diff --git "a/checkpoint-28368/trainer_state.json" "b/checkpoint-28368/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-28368/trainer_state.json" @@ -0,0 +1,198609 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 3.0, + "eval_steps": 500, + "global_step": 28368, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00010575296108291032, + "grad_norm": 65.5, + "learning_rate": 3.523608174770966e-09, + "loss": 3.555, + "step": 1 + }, + { + "epoch": 0.00021150592216582064, + "grad_norm": 68.0, + "learning_rate": 7.047216349541932e-09, + "loss": 3.5704, + "step": 2 + }, + { + "epoch": 0.00031725888324873094, + "grad_norm": 75.0, + "learning_rate": 1.0570824524312897e-08, + "loss": 3.8265, + "step": 3 + }, + { + "epoch": 0.00042301184433164127, + "grad_norm": 70.0, + "learning_rate": 1.4094432699083864e-08, + "loss": 3.9112, + "step": 4 + }, + { + "epoch": 0.0005287648054145516, + "grad_norm": 64.5, + "learning_rate": 1.761804087385483e-08, + "loss": 3.8201, + "step": 5 + }, + { + "epoch": 0.0006345177664974619, + "grad_norm": 78.0, + "learning_rate": 2.1141649048625795e-08, + "loss": 3.7823, + "step": 6 + }, + { + "epoch": 0.0007402707275803723, + "grad_norm": 62.25, + "learning_rate": 2.466525722339676e-08, + "loss": 3.9749, + "step": 7 + }, + { + "epoch": 0.0008460236886632825, + "grad_norm": 67.5, + "learning_rate": 2.8188865398167727e-08, + "loss": 3.9768, + "step": 8 + }, + { + "epoch": 0.0009517766497461929, + "grad_norm": 69.0, + "learning_rate": 3.171247357293869e-08, + "loss": 3.8636, + "step": 9 + }, + { + "epoch": 0.0010575296108291032, + "grad_norm": 56.0, + "learning_rate": 3.523608174770966e-08, + "loss": 3.6337, + "step": 10 + }, + { + "epoch": 0.0011632825719120135, + "grad_norm": 66.0, + "learning_rate": 3.875968992248062e-08, + "loss": 3.8189, + "step": 11 + }, + { + "epoch": 0.0012690355329949238, + "grad_norm": 55.25, + "learning_rate": 4.228329809725159e-08, + "loss": 3.9148, + "step": 12 + }, + { + "epoch": 0.0013747884940778343, + "grad_norm": 65.5, + "learning_rate": 4.580690627202256e-08, + "loss": 3.9351, + "step": 13 + }, + { + "epoch": 0.0014805414551607445, + "grad_norm": 53.75, + "learning_rate": 4.933051444679352e-08, + "loss": 3.5825, + "step": 14 + }, + { + "epoch": 0.0015862944162436548, + "grad_norm": 61.75, + "learning_rate": 5.285412262156449e-08, + "loss": 3.5867, + "step": 15 + }, + { + "epoch": 0.001692047377326565, + "grad_norm": 68.5, + "learning_rate": 5.6377730796335455e-08, + "loss": 3.8553, + "step": 16 + }, + { + "epoch": 0.0017978003384094754, + "grad_norm": 52.75, + "learning_rate": 5.990133897110641e-08, + "loss": 3.4511, + "step": 17 + }, + { + "epoch": 0.0019035532994923859, + "grad_norm": 53.5, + "learning_rate": 6.342494714587738e-08, + "loss": 3.6486, + "step": 18 + }, + { + "epoch": 0.002009306260575296, + "grad_norm": 65.0, + "learning_rate": 6.694855532064835e-08, + "loss": 3.7229, + "step": 19 + }, + { + "epoch": 0.0021150592216582064, + "grad_norm": 51.25, + "learning_rate": 7.047216349541932e-08, + "loss": 3.7859, + "step": 20 + }, + { + "epoch": 0.002220812182741117, + "grad_norm": 70.5, + "learning_rate": 7.399577167019028e-08, + "loss": 3.6738, + "step": 21 + }, + { + "epoch": 0.002326565143824027, + "grad_norm": 50.0, + "learning_rate": 7.751937984496124e-08, + "loss": 3.6319, + "step": 22 + }, + { + "epoch": 0.0024323181049069375, + "grad_norm": 55.75, + "learning_rate": 8.10429880197322e-08, + "loss": 3.8704, + "step": 23 + }, + { + "epoch": 0.0025380710659898475, + "grad_norm": 68.5, + "learning_rate": 8.456659619450318e-08, + "loss": 3.8875, + "step": 24 + }, + { + "epoch": 0.002643824027072758, + "grad_norm": 62.5, + "learning_rate": 8.809020436927414e-08, + "loss": 3.5903, + "step": 25 + }, + { + "epoch": 0.0027495769881556685, + "grad_norm": 65.5, + "learning_rate": 9.161381254404511e-08, + "loss": 3.7099, + "step": 26 + }, + { + "epoch": 0.0028553299492385786, + "grad_norm": 61.5, + "learning_rate": 9.513742071881608e-08, + "loss": 3.6988, + "step": 27 + }, + { + "epoch": 0.002961082910321489, + "grad_norm": 57.0, + "learning_rate": 9.866102889358704e-08, + "loss": 3.7658, + "step": 28 + }, + { + "epoch": 0.003066835871404399, + "grad_norm": 57.5, + "learning_rate": 1.0218463706835801e-07, + "loss": 3.5928, + "step": 29 + }, + { + "epoch": 0.0031725888324873096, + "grad_norm": 61.5, + "learning_rate": 1.0570824524312897e-07, + "loss": 3.7315, + "step": 30 + }, + { + "epoch": 0.00327834179357022, + "grad_norm": 65.5, + "learning_rate": 1.0923185341789993e-07, + "loss": 3.7655, + "step": 31 + }, + { + "epoch": 0.00338409475465313, + "grad_norm": 41.5, + "learning_rate": 1.1275546159267091e-07, + "loss": 3.2478, + "step": 32 + }, + { + "epoch": 0.0034898477157360407, + "grad_norm": 57.75, + "learning_rate": 1.1627906976744187e-07, + "loss": 3.5252, + "step": 33 + }, + { + "epoch": 0.0035956006768189507, + "grad_norm": 39.75, + "learning_rate": 1.1980267794221282e-07, + "loss": 3.4351, + "step": 34 + }, + { + "epoch": 0.003701353637901861, + "grad_norm": 59.25, + "learning_rate": 1.233262861169838e-07, + "loss": 3.626, + "step": 35 + }, + { + "epoch": 0.0038071065989847717, + "grad_norm": 77.5, + "learning_rate": 1.2684989429175477e-07, + "loss": 3.6728, + "step": 36 + }, + { + "epoch": 0.003912859560067682, + "grad_norm": 69.0, + "learning_rate": 1.3037350246652573e-07, + "loss": 4.2837, + "step": 37 + }, + { + "epoch": 0.004018612521150592, + "grad_norm": 58.25, + "learning_rate": 1.338971106412967e-07, + "loss": 3.8975, + "step": 38 + }, + { + "epoch": 0.004124365482233503, + "grad_norm": 48.75, + "learning_rate": 1.3742071881606765e-07, + "loss": 3.5995, + "step": 39 + }, + { + "epoch": 0.004230118443316413, + "grad_norm": 63.5, + "learning_rate": 1.4094432699083864e-07, + "loss": 3.6022, + "step": 40 + }, + { + "epoch": 0.004335871404399323, + "grad_norm": 44.5, + "learning_rate": 1.444679351656096e-07, + "loss": 3.2825, + "step": 41 + }, + { + "epoch": 0.004441624365482234, + "grad_norm": 63.25, + "learning_rate": 1.4799154334038056e-07, + "loss": 3.8275, + "step": 42 + }, + { + "epoch": 0.004547377326565144, + "grad_norm": 58.25, + "learning_rate": 1.5151515151515152e-07, + "loss": 3.8747, + "step": 43 + }, + { + "epoch": 0.004653130287648054, + "grad_norm": 61.25, + "learning_rate": 1.5503875968992249e-07, + "loss": 3.4283, + "step": 44 + }, + { + "epoch": 0.004758883248730965, + "grad_norm": 63.75, + "learning_rate": 1.5856236786469345e-07, + "loss": 3.6574, + "step": 45 + }, + { + "epoch": 0.004864636209813875, + "grad_norm": 69.5, + "learning_rate": 1.620859760394644e-07, + "loss": 4.0838, + "step": 46 + }, + { + "epoch": 0.004970389170896785, + "grad_norm": 60.5, + "learning_rate": 1.656095842142354e-07, + "loss": 3.7493, + "step": 47 + }, + { + "epoch": 0.005076142131979695, + "grad_norm": 66.0, + "learning_rate": 1.6913319238900636e-07, + "loss": 3.7588, + "step": 48 + }, + { + "epoch": 0.005181895093062606, + "grad_norm": 62.0, + "learning_rate": 1.7265680056377732e-07, + "loss": 3.7568, + "step": 49 + }, + { + "epoch": 0.005287648054145516, + "grad_norm": 64.5, + "learning_rate": 1.7618040873854828e-07, + "loss": 3.8459, + "step": 50 + }, + { + "epoch": 0.005393401015228426, + "grad_norm": 70.0, + "learning_rate": 1.7970401691331924e-07, + "loss": 4.1238, + "step": 51 + }, + { + "epoch": 0.005499153976311337, + "grad_norm": 57.0, + "learning_rate": 1.8322762508809023e-07, + "loss": 3.8423, + "step": 52 + }, + { + "epoch": 0.005604906937394247, + "grad_norm": 63.75, + "learning_rate": 1.867512332628612e-07, + "loss": 3.6152, + "step": 53 + }, + { + "epoch": 0.005710659898477157, + "grad_norm": 63.75, + "learning_rate": 1.9027484143763215e-07, + "loss": 4.0896, + "step": 54 + }, + { + "epoch": 0.005816412859560068, + "grad_norm": 63.25, + "learning_rate": 1.9379844961240311e-07, + "loss": 3.4999, + "step": 55 + }, + { + "epoch": 0.005922165820642978, + "grad_norm": 77.5, + "learning_rate": 1.9732205778717408e-07, + "loss": 3.871, + "step": 56 + }, + { + "epoch": 0.006027918781725888, + "grad_norm": 60.75, + "learning_rate": 2.0084566596194504e-07, + "loss": 3.7725, + "step": 57 + }, + { + "epoch": 0.006133671742808798, + "grad_norm": 62.0, + "learning_rate": 2.0436927413671602e-07, + "loss": 3.9003, + "step": 58 + }, + { + "epoch": 0.006239424703891709, + "grad_norm": 61.25, + "learning_rate": 2.0789288231148699e-07, + "loss": 4.0019, + "step": 59 + }, + { + "epoch": 0.006345177664974619, + "grad_norm": 60.5, + "learning_rate": 2.1141649048625795e-07, + "loss": 3.7613, + "step": 60 + }, + { + "epoch": 0.006450930626057529, + "grad_norm": 56.5, + "learning_rate": 2.149400986610289e-07, + "loss": 3.5079, + "step": 61 + }, + { + "epoch": 0.00655668358714044, + "grad_norm": 71.5, + "learning_rate": 2.1846370683579987e-07, + "loss": 3.6341, + "step": 62 + }, + { + "epoch": 0.00666243654822335, + "grad_norm": 65.5, + "learning_rate": 2.2198731501057086e-07, + "loss": 3.727, + "step": 63 + }, + { + "epoch": 0.00676818950930626, + "grad_norm": 84.5, + "learning_rate": 2.2551092318534182e-07, + "loss": 4.0097, + "step": 64 + }, + { + "epoch": 0.006873942470389171, + "grad_norm": 58.75, + "learning_rate": 2.2903453136011278e-07, + "loss": 3.6626, + "step": 65 + }, + { + "epoch": 0.006979695431472081, + "grad_norm": 64.5, + "learning_rate": 2.3255813953488374e-07, + "loss": 3.8272, + "step": 66 + }, + { + "epoch": 0.007085448392554991, + "grad_norm": 63.75, + "learning_rate": 2.360817477096547e-07, + "loss": 3.8128, + "step": 67 + }, + { + "epoch": 0.0071912013536379014, + "grad_norm": 63.25, + "learning_rate": 2.3960535588442564e-07, + "loss": 3.7855, + "step": 68 + }, + { + "epoch": 0.007296954314720812, + "grad_norm": 60.25, + "learning_rate": 2.431289640591966e-07, + "loss": 3.7052, + "step": 69 + }, + { + "epoch": 0.007402707275803722, + "grad_norm": 59.25, + "learning_rate": 2.466525722339676e-07, + "loss": 3.4868, + "step": 70 + }, + { + "epoch": 0.0075084602368866325, + "grad_norm": 60.5, + "learning_rate": 2.5017618040873855e-07, + "loss": 3.5455, + "step": 71 + }, + { + "epoch": 0.007614213197969543, + "grad_norm": 69.5, + "learning_rate": 2.5369978858350954e-07, + "loss": 3.8048, + "step": 72 + }, + { + "epoch": 0.0077199661590524535, + "grad_norm": 44.0, + "learning_rate": 2.5722339675828047e-07, + "loss": 3.6932, + "step": 73 + }, + { + "epoch": 0.007825719120135364, + "grad_norm": 64.0, + "learning_rate": 2.6074700493305146e-07, + "loss": 3.772, + "step": 74 + }, + { + "epoch": 0.007931472081218274, + "grad_norm": 53.75, + "learning_rate": 2.6427061310782245e-07, + "loss": 3.8735, + "step": 75 + }, + { + "epoch": 0.008037225042301184, + "grad_norm": 65.5, + "learning_rate": 2.677942212825934e-07, + "loss": 3.8704, + "step": 76 + }, + { + "epoch": 0.008142978003384095, + "grad_norm": 50.25, + "learning_rate": 2.7131782945736437e-07, + "loss": 3.828, + "step": 77 + }, + { + "epoch": 0.008248730964467006, + "grad_norm": 68.0, + "learning_rate": 2.748414376321353e-07, + "loss": 3.8291, + "step": 78 + }, + { + "epoch": 0.008354483925549916, + "grad_norm": 48.0, + "learning_rate": 2.783650458069063e-07, + "loss": 3.4683, + "step": 79 + }, + { + "epoch": 0.008460236886632826, + "grad_norm": 67.0, + "learning_rate": 2.818886539816773e-07, + "loss": 3.8545, + "step": 80 + }, + { + "epoch": 0.008565989847715736, + "grad_norm": 50.25, + "learning_rate": 2.854122621564482e-07, + "loss": 3.6997, + "step": 81 + }, + { + "epoch": 0.008671742808798646, + "grad_norm": 56.25, + "learning_rate": 2.889358703312192e-07, + "loss": 3.7988, + "step": 82 + }, + { + "epoch": 0.008777495769881556, + "grad_norm": 66.5, + "learning_rate": 2.9245947850599014e-07, + "loss": 3.9501, + "step": 83 + }, + { + "epoch": 0.008883248730964468, + "grad_norm": 84.5, + "learning_rate": 2.959830866807611e-07, + "loss": 3.7707, + "step": 84 + }, + { + "epoch": 0.008989001692047378, + "grad_norm": 68.5, + "learning_rate": 2.995066948555321e-07, + "loss": 3.804, + "step": 85 + }, + { + "epoch": 0.009094754653130288, + "grad_norm": 59.75, + "learning_rate": 3.0303030303030305e-07, + "loss": 3.6813, + "step": 86 + }, + { + "epoch": 0.009200507614213198, + "grad_norm": 59.25, + "learning_rate": 3.0655391120507404e-07, + "loss": 3.7826, + "step": 87 + }, + { + "epoch": 0.009306260575296108, + "grad_norm": 58.25, + "learning_rate": 3.1007751937984497e-07, + "loss": 3.6606, + "step": 88 + }, + { + "epoch": 0.009412013536379018, + "grad_norm": 62.75, + "learning_rate": 3.1360112755461596e-07, + "loss": 3.6379, + "step": 89 + }, + { + "epoch": 0.00951776649746193, + "grad_norm": 55.0, + "learning_rate": 3.171247357293869e-07, + "loss": 3.7395, + "step": 90 + }, + { + "epoch": 0.00962351945854484, + "grad_norm": 52.75, + "learning_rate": 3.206483439041579e-07, + "loss": 3.8119, + "step": 91 + }, + { + "epoch": 0.00972927241962775, + "grad_norm": 53.5, + "learning_rate": 3.241719520789288e-07, + "loss": 4.0847, + "step": 92 + }, + { + "epoch": 0.00983502538071066, + "grad_norm": 53.0, + "learning_rate": 3.2769556025369986e-07, + "loss": 3.4029, + "step": 93 + }, + { + "epoch": 0.00994077834179357, + "grad_norm": 43.75, + "learning_rate": 3.312191684284708e-07, + "loss": 3.3181, + "step": 94 + }, + { + "epoch": 0.01004653130287648, + "grad_norm": 67.5, + "learning_rate": 3.347427766032418e-07, + "loss": 3.8324, + "step": 95 + }, + { + "epoch": 0.01015228426395939, + "grad_norm": 67.0, + "learning_rate": 3.382663847780127e-07, + "loss": 3.4051, + "step": 96 + }, + { + "epoch": 0.010258037225042302, + "grad_norm": 58.5, + "learning_rate": 3.417899929527837e-07, + "loss": 3.7327, + "step": 97 + }, + { + "epoch": 0.010363790186125212, + "grad_norm": 54.75, + "learning_rate": 3.4531360112755464e-07, + "loss": 3.4315, + "step": 98 + }, + { + "epoch": 0.010469543147208122, + "grad_norm": 63.0, + "learning_rate": 3.488372093023256e-07, + "loss": 4.0246, + "step": 99 + }, + { + "epoch": 0.010575296108291032, + "grad_norm": 49.75, + "learning_rate": 3.5236081747709656e-07, + "loss": 3.5897, + "step": 100 + }, + { + "epoch": 0.010681049069373942, + "grad_norm": 51.25, + "learning_rate": 3.5588442565186755e-07, + "loss": 3.7466, + "step": 101 + }, + { + "epoch": 0.010786802030456852, + "grad_norm": 56.75, + "learning_rate": 3.594080338266385e-07, + "loss": 3.5741, + "step": 102 + }, + { + "epoch": 0.010892554991539764, + "grad_norm": 52.0, + "learning_rate": 3.629316420014094e-07, + "loss": 3.6855, + "step": 103 + }, + { + "epoch": 0.010998307952622674, + "grad_norm": 57.5, + "learning_rate": 3.6645525017618046e-07, + "loss": 3.678, + "step": 104 + }, + { + "epoch": 0.011104060913705584, + "grad_norm": 52.75, + "learning_rate": 3.6997885835095145e-07, + "loss": 3.4721, + "step": 105 + }, + { + "epoch": 0.011209813874788494, + "grad_norm": 53.0, + "learning_rate": 3.735024665257224e-07, + "loss": 3.8013, + "step": 106 + }, + { + "epoch": 0.011315566835871404, + "grad_norm": 47.75, + "learning_rate": 3.7702607470049337e-07, + "loss": 3.5052, + "step": 107 + }, + { + "epoch": 0.011421319796954314, + "grad_norm": 65.0, + "learning_rate": 3.805496828752643e-07, + "loss": 3.8599, + "step": 108 + }, + { + "epoch": 0.011527072758037224, + "grad_norm": 42.0, + "learning_rate": 3.840732910500353e-07, + "loss": 3.3915, + "step": 109 + }, + { + "epoch": 0.011632825719120136, + "grad_norm": 49.75, + "learning_rate": 3.8759689922480623e-07, + "loss": 3.9633, + "step": 110 + }, + { + "epoch": 0.011738578680203046, + "grad_norm": 60.5, + "learning_rate": 3.9112050739957716e-07, + "loss": 3.8584, + "step": 111 + }, + { + "epoch": 0.011844331641285956, + "grad_norm": 53.0, + "learning_rate": 3.9464411557434815e-07, + "loss": 3.6439, + "step": 112 + }, + { + "epoch": 0.011950084602368866, + "grad_norm": 61.25, + "learning_rate": 3.981677237491191e-07, + "loss": 3.7225, + "step": 113 + }, + { + "epoch": 0.012055837563451776, + "grad_norm": 51.5, + "learning_rate": 4.0169133192389007e-07, + "loss": 3.8408, + "step": 114 + }, + { + "epoch": 0.012161590524534686, + "grad_norm": 51.0, + "learning_rate": 4.052149400986611e-07, + "loss": 3.5629, + "step": 115 + }, + { + "epoch": 0.012267343485617596, + "grad_norm": 57.25, + "learning_rate": 4.0873854827343205e-07, + "loss": 3.7337, + "step": 116 + }, + { + "epoch": 0.012373096446700508, + "grad_norm": 55.0, + "learning_rate": 4.1226215644820304e-07, + "loss": 3.5655, + "step": 117 + }, + { + "epoch": 0.012478849407783418, + "grad_norm": 60.75, + "learning_rate": 4.1578576462297397e-07, + "loss": 3.6658, + "step": 118 + }, + { + "epoch": 0.012584602368866328, + "grad_norm": 61.25, + "learning_rate": 4.193093727977449e-07, + "loss": 3.6663, + "step": 119 + }, + { + "epoch": 0.012690355329949238, + "grad_norm": 53.25, + "learning_rate": 4.228329809725159e-07, + "loss": 3.8232, + "step": 120 + }, + { + "epoch": 0.012796108291032149, + "grad_norm": 46.0, + "learning_rate": 4.2635658914728683e-07, + "loss": 3.6954, + "step": 121 + }, + { + "epoch": 0.012901861252115059, + "grad_norm": 52.25, + "learning_rate": 4.298801973220578e-07, + "loss": 3.8238, + "step": 122 + }, + { + "epoch": 0.01300761421319797, + "grad_norm": 47.0, + "learning_rate": 4.3340380549682875e-07, + "loss": 3.6867, + "step": 123 + }, + { + "epoch": 0.01311336717428088, + "grad_norm": 54.75, + "learning_rate": 4.3692741367159974e-07, + "loss": 3.5859, + "step": 124 + }, + { + "epoch": 0.01321912013536379, + "grad_norm": 40.75, + "learning_rate": 4.404510218463707e-07, + "loss": 3.5069, + "step": 125 + }, + { + "epoch": 0.0133248730964467, + "grad_norm": 56.5, + "learning_rate": 4.439746300211417e-07, + "loss": 3.79, + "step": 126 + }, + { + "epoch": 0.01343062605752961, + "grad_norm": 47.25, + "learning_rate": 4.4749823819591265e-07, + "loss": 3.6145, + "step": 127 + }, + { + "epoch": 0.01353637901861252, + "grad_norm": 51.0, + "learning_rate": 4.5102184637068364e-07, + "loss": 3.7078, + "step": 128 + }, + { + "epoch": 0.01364213197969543, + "grad_norm": 67.5, + "learning_rate": 4.5454545454545457e-07, + "loss": 3.3779, + "step": 129 + }, + { + "epoch": 0.013747884940778343, + "grad_norm": 39.25, + "learning_rate": 4.5806906272022556e-07, + "loss": 3.5768, + "step": 130 + }, + { + "epoch": 0.013853637901861253, + "grad_norm": 44.5, + "learning_rate": 4.615926708949965e-07, + "loss": 3.5933, + "step": 131 + }, + { + "epoch": 0.013959390862944163, + "grad_norm": 45.0, + "learning_rate": 4.651162790697675e-07, + "loss": 3.3684, + "step": 132 + }, + { + "epoch": 0.014065143824027073, + "grad_norm": 48.75, + "learning_rate": 4.686398872445384e-07, + "loss": 3.7802, + "step": 133 + }, + { + "epoch": 0.014170896785109983, + "grad_norm": 52.5, + "learning_rate": 4.721634954193094e-07, + "loss": 3.6978, + "step": 134 + }, + { + "epoch": 0.014276649746192893, + "grad_norm": 48.75, + "learning_rate": 4.7568710359408034e-07, + "loss": 3.5434, + "step": 135 + }, + { + "epoch": 0.014382402707275803, + "grad_norm": 43.75, + "learning_rate": 4.792107117688513e-07, + "loss": 3.6639, + "step": 136 + }, + { + "epoch": 0.014488155668358715, + "grad_norm": 51.75, + "learning_rate": 4.827343199436224e-07, + "loss": 3.3968, + "step": 137 + }, + { + "epoch": 0.014593908629441625, + "grad_norm": 50.0, + "learning_rate": 4.862579281183933e-07, + "loss": 3.5999, + "step": 138 + }, + { + "epoch": 0.014699661590524535, + "grad_norm": 62.75, + "learning_rate": 4.897815362931642e-07, + "loss": 3.8126, + "step": 139 + }, + { + "epoch": 0.014805414551607445, + "grad_norm": 47.25, + "learning_rate": 4.933051444679352e-07, + "loss": 3.5118, + "step": 140 + }, + { + "epoch": 0.014911167512690355, + "grad_norm": 45.0, + "learning_rate": 4.968287526427062e-07, + "loss": 3.6211, + "step": 141 + }, + { + "epoch": 0.015016920473773265, + "grad_norm": 45.5, + "learning_rate": 5.003523608174771e-07, + "loss": 3.4845, + "step": 142 + }, + { + "epoch": 0.015122673434856177, + "grad_norm": 45.25, + "learning_rate": 5.038759689922481e-07, + "loss": 3.5677, + "step": 143 + }, + { + "epoch": 0.015228426395939087, + "grad_norm": 54.75, + "learning_rate": 5.073995771670191e-07, + "loss": 3.6965, + "step": 144 + }, + { + "epoch": 0.015334179357021997, + "grad_norm": 36.5, + "learning_rate": 5.109231853417901e-07, + "loss": 3.6467, + "step": 145 + }, + { + "epoch": 0.015439932318104907, + "grad_norm": 50.0, + "learning_rate": 5.144467935165609e-07, + "loss": 3.7696, + "step": 146 + }, + { + "epoch": 0.015545685279187817, + "grad_norm": 39.0, + "learning_rate": 5.179704016913319e-07, + "loss": 3.7168, + "step": 147 + }, + { + "epoch": 0.015651438240270727, + "grad_norm": 50.5, + "learning_rate": 5.214940098661029e-07, + "loss": 3.4837, + "step": 148 + }, + { + "epoch": 0.01575719120135364, + "grad_norm": 47.0, + "learning_rate": 5.250176180408739e-07, + "loss": 3.6969, + "step": 149 + }, + { + "epoch": 0.015862944162436547, + "grad_norm": 40.75, + "learning_rate": 5.285412262156449e-07, + "loss": 3.3707, + "step": 150 + }, + { + "epoch": 0.01596869712351946, + "grad_norm": 46.5, + "learning_rate": 5.320648343904159e-07, + "loss": 3.4378, + "step": 151 + }, + { + "epoch": 0.016074450084602367, + "grad_norm": 52.0, + "learning_rate": 5.355884425651868e-07, + "loss": 3.7879, + "step": 152 + }, + { + "epoch": 0.01618020304568528, + "grad_norm": 46.0, + "learning_rate": 5.391120507399578e-07, + "loss": 3.5452, + "step": 153 + }, + { + "epoch": 0.01628595600676819, + "grad_norm": 44.0, + "learning_rate": 5.426356589147287e-07, + "loss": 3.623, + "step": 154 + }, + { + "epoch": 0.0163917089678511, + "grad_norm": 50.5, + "learning_rate": 5.461592670894997e-07, + "loss": 3.5399, + "step": 155 + }, + { + "epoch": 0.01649746192893401, + "grad_norm": 45.5, + "learning_rate": 5.496828752642706e-07, + "loss": 3.5832, + "step": 156 + }, + { + "epoch": 0.01660321489001692, + "grad_norm": 44.25, + "learning_rate": 5.532064834390416e-07, + "loss": 3.9145, + "step": 157 + }, + { + "epoch": 0.01670896785109983, + "grad_norm": 46.5, + "learning_rate": 5.567300916138126e-07, + "loss": 3.6728, + "step": 158 + }, + { + "epoch": 0.01681472081218274, + "grad_norm": 47.75, + "learning_rate": 5.602536997885836e-07, + "loss": 3.4508, + "step": 159 + }, + { + "epoch": 0.01692047377326565, + "grad_norm": 45.5, + "learning_rate": 5.637773079633546e-07, + "loss": 3.537, + "step": 160 + }, + { + "epoch": 0.017026226734348563, + "grad_norm": 43.5, + "learning_rate": 5.673009161381255e-07, + "loss": 3.7183, + "step": 161 + }, + { + "epoch": 0.01713197969543147, + "grad_norm": 44.5, + "learning_rate": 5.708245243128964e-07, + "loss": 3.6791, + "step": 162 + }, + { + "epoch": 0.017237732656514383, + "grad_norm": 47.75, + "learning_rate": 5.743481324876674e-07, + "loss": 3.6137, + "step": 163 + }, + { + "epoch": 0.01734348561759729, + "grad_norm": 39.75, + "learning_rate": 5.778717406624384e-07, + "loss": 3.4017, + "step": 164 + }, + { + "epoch": 0.017449238578680203, + "grad_norm": 39.25, + "learning_rate": 5.813953488372094e-07, + "loss": 3.4877, + "step": 165 + }, + { + "epoch": 0.01755499153976311, + "grad_norm": 38.5, + "learning_rate": 5.849189570119803e-07, + "loss": 3.546, + "step": 166 + }, + { + "epoch": 0.017660744500846023, + "grad_norm": 30.75, + "learning_rate": 5.884425651867513e-07, + "loss": 3.0729, + "step": 167 + }, + { + "epoch": 0.017766497461928935, + "grad_norm": 45.0, + "learning_rate": 5.919661733615223e-07, + "loss": 3.5765, + "step": 168 + }, + { + "epoch": 0.017872250423011844, + "grad_norm": 37.5, + "learning_rate": 5.954897815362931e-07, + "loss": 3.4605, + "step": 169 + }, + { + "epoch": 0.017978003384094755, + "grad_norm": 45.0, + "learning_rate": 5.990133897110642e-07, + "loss": 3.4181, + "step": 170 + }, + { + "epoch": 0.018083756345177664, + "grad_norm": 43.5, + "learning_rate": 6.025369978858352e-07, + "loss": 3.5062, + "step": 171 + }, + { + "epoch": 0.018189509306260575, + "grad_norm": 40.5, + "learning_rate": 6.060606060606061e-07, + "loss": 3.7169, + "step": 172 + }, + { + "epoch": 0.018295262267343487, + "grad_norm": 39.25, + "learning_rate": 6.095842142353771e-07, + "loss": 3.2096, + "step": 173 + }, + { + "epoch": 0.018401015228426396, + "grad_norm": 28.625, + "learning_rate": 6.131078224101481e-07, + "loss": 3.4358, + "step": 174 + }, + { + "epoch": 0.018506768189509307, + "grad_norm": 39.25, + "learning_rate": 6.16631430584919e-07, + "loss": 3.4729, + "step": 175 + }, + { + "epoch": 0.018612521150592216, + "grad_norm": 36.0, + "learning_rate": 6.201550387596899e-07, + "loss": 3.7674, + "step": 176 + }, + { + "epoch": 0.018718274111675128, + "grad_norm": 41.25, + "learning_rate": 6.236786469344609e-07, + "loss": 4.0072, + "step": 177 + }, + { + "epoch": 0.018824027072758036, + "grad_norm": 31.75, + "learning_rate": 6.272022551092319e-07, + "loss": 3.2824, + "step": 178 + }, + { + "epoch": 0.018929780033840948, + "grad_norm": 39.5, + "learning_rate": 6.307258632840028e-07, + "loss": 3.5507, + "step": 179 + }, + { + "epoch": 0.01903553299492386, + "grad_norm": 33.5, + "learning_rate": 6.342494714587738e-07, + "loss": 3.1986, + "step": 180 + }, + { + "epoch": 0.019141285956006768, + "grad_norm": 37.25, + "learning_rate": 6.377730796335448e-07, + "loss": 3.5168, + "step": 181 + }, + { + "epoch": 0.01924703891708968, + "grad_norm": 49.5, + "learning_rate": 6.412966878083158e-07, + "loss": 3.4363, + "step": 182 + }, + { + "epoch": 0.019352791878172588, + "grad_norm": 35.75, + "learning_rate": 6.448202959830866e-07, + "loss": 3.5291, + "step": 183 + }, + { + "epoch": 0.0194585448392555, + "grad_norm": 40.0, + "learning_rate": 6.483439041578576e-07, + "loss": 3.6676, + "step": 184 + }, + { + "epoch": 0.019564297800338408, + "grad_norm": 29.875, + "learning_rate": 6.518675123326286e-07, + "loss": 3.1858, + "step": 185 + }, + { + "epoch": 0.01967005076142132, + "grad_norm": 31.875, + "learning_rate": 6.553911205073997e-07, + "loss": 3.4472, + "step": 186 + }, + { + "epoch": 0.01977580372250423, + "grad_norm": 38.5, + "learning_rate": 6.589147286821707e-07, + "loss": 3.2957, + "step": 187 + }, + { + "epoch": 0.01988155668358714, + "grad_norm": 32.0, + "learning_rate": 6.624383368569416e-07, + "loss": 3.3289, + "step": 188 + }, + { + "epoch": 0.01998730964467005, + "grad_norm": 30.5, + "learning_rate": 6.659619450317126e-07, + "loss": 3.267, + "step": 189 + }, + { + "epoch": 0.02009306260575296, + "grad_norm": 32.25, + "learning_rate": 6.694855532064836e-07, + "loss": 3.3508, + "step": 190 + }, + { + "epoch": 0.020198815566835872, + "grad_norm": 34.25, + "learning_rate": 6.730091613812544e-07, + "loss": 3.7722, + "step": 191 + }, + { + "epoch": 0.02030456852791878, + "grad_norm": 36.0, + "learning_rate": 6.765327695560254e-07, + "loss": 3.5876, + "step": 192 + }, + { + "epoch": 0.020410321489001692, + "grad_norm": 36.25, + "learning_rate": 6.800563777307964e-07, + "loss": 3.5856, + "step": 193 + }, + { + "epoch": 0.020516074450084604, + "grad_norm": 33.75, + "learning_rate": 6.835799859055674e-07, + "loss": 3.5235, + "step": 194 + }, + { + "epoch": 0.020621827411167512, + "grad_norm": 29.375, + "learning_rate": 6.871035940803383e-07, + "loss": 3.385, + "step": 195 + }, + { + "epoch": 0.020727580372250424, + "grad_norm": 29.5, + "learning_rate": 6.906272022551093e-07, + "loss": 3.2106, + "step": 196 + }, + { + "epoch": 0.020833333333333332, + "grad_norm": 27.5, + "learning_rate": 6.941508104298803e-07, + "loss": 3.4844, + "step": 197 + }, + { + "epoch": 0.020939086294416244, + "grad_norm": 23.875, + "learning_rate": 6.976744186046513e-07, + "loss": 3.2768, + "step": 198 + }, + { + "epoch": 0.021044839255499152, + "grad_norm": 36.0, + "learning_rate": 7.011980267794221e-07, + "loss": 3.2607, + "step": 199 + }, + { + "epoch": 0.021150592216582064, + "grad_norm": 23.75, + "learning_rate": 7.047216349541931e-07, + "loss": 3.2054, + "step": 200 + }, + { + "epoch": 0.021256345177664976, + "grad_norm": 32.25, + "learning_rate": 7.082452431289641e-07, + "loss": 3.3788, + "step": 201 + }, + { + "epoch": 0.021362098138747884, + "grad_norm": 25.375, + "learning_rate": 7.117688513037351e-07, + "loss": 3.3577, + "step": 202 + }, + { + "epoch": 0.021467851099830796, + "grad_norm": 26.5, + "learning_rate": 7.15292459478506e-07, + "loss": 3.3225, + "step": 203 + }, + { + "epoch": 0.021573604060913704, + "grad_norm": 28.25, + "learning_rate": 7.18816067653277e-07, + "loss": 3.4624, + "step": 204 + }, + { + "epoch": 0.021679357021996616, + "grad_norm": 29.0, + "learning_rate": 7.22339675828048e-07, + "loss": 3.2953, + "step": 205 + }, + { + "epoch": 0.021785109983079528, + "grad_norm": 23.125, + "learning_rate": 7.258632840028188e-07, + "loss": 2.975, + "step": 206 + }, + { + "epoch": 0.021890862944162436, + "grad_norm": 30.0, + "learning_rate": 7.293868921775898e-07, + "loss": 3.2574, + "step": 207 + }, + { + "epoch": 0.021996615905245348, + "grad_norm": 32.25, + "learning_rate": 7.329105003523609e-07, + "loss": 3.3113, + "step": 208 + }, + { + "epoch": 0.022102368866328256, + "grad_norm": 32.25, + "learning_rate": 7.364341085271319e-07, + "loss": 3.5148, + "step": 209 + }, + { + "epoch": 0.022208121827411168, + "grad_norm": 30.75, + "learning_rate": 7.399577167019029e-07, + "loss": 3.596, + "step": 210 + }, + { + "epoch": 0.022313874788494076, + "grad_norm": 25.0, + "learning_rate": 7.434813248766738e-07, + "loss": 3.279, + "step": 211 + }, + { + "epoch": 0.02241962774957699, + "grad_norm": 28.0, + "learning_rate": 7.470049330514448e-07, + "loss": 3.379, + "step": 212 + }, + { + "epoch": 0.0225253807106599, + "grad_norm": 24.625, + "learning_rate": 7.505285412262158e-07, + "loss": 3.1378, + "step": 213 + }, + { + "epoch": 0.02263113367174281, + "grad_norm": 26.25, + "learning_rate": 7.540521494009867e-07, + "loss": 3.365, + "step": 214 + }, + { + "epoch": 0.02273688663282572, + "grad_norm": 25.625, + "learning_rate": 7.575757575757576e-07, + "loss": 3.1455, + "step": 215 + }, + { + "epoch": 0.02284263959390863, + "grad_norm": 21.25, + "learning_rate": 7.610993657505286e-07, + "loss": 3.3527, + "step": 216 + }, + { + "epoch": 0.02294839255499154, + "grad_norm": 26.875, + "learning_rate": 7.646229739252996e-07, + "loss": 3.4171, + "step": 217 + }, + { + "epoch": 0.02305414551607445, + "grad_norm": 25.375, + "learning_rate": 7.681465821000706e-07, + "loss": 3.2335, + "step": 218 + }, + { + "epoch": 0.02315989847715736, + "grad_norm": 27.875, + "learning_rate": 7.716701902748415e-07, + "loss": 3.3241, + "step": 219 + }, + { + "epoch": 0.023265651438240272, + "grad_norm": 26.75, + "learning_rate": 7.751937984496125e-07, + "loss": 3.4242, + "step": 220 + }, + { + "epoch": 0.02337140439932318, + "grad_norm": 24.0, + "learning_rate": 7.787174066243834e-07, + "loss": 3.4493, + "step": 221 + }, + { + "epoch": 0.023477157360406092, + "grad_norm": 25.0, + "learning_rate": 7.822410147991543e-07, + "loss": 3.489, + "step": 222 + }, + { + "epoch": 0.023582910321489, + "grad_norm": 28.0, + "learning_rate": 7.857646229739253e-07, + "loss": 3.4168, + "step": 223 + }, + { + "epoch": 0.023688663282571912, + "grad_norm": 21.375, + "learning_rate": 7.892882311486963e-07, + "loss": 3.2, + "step": 224 + }, + { + "epoch": 0.02379441624365482, + "grad_norm": 22.875, + "learning_rate": 7.928118393234673e-07, + "loss": 3.1934, + "step": 225 + }, + { + "epoch": 0.023900169204737733, + "grad_norm": 23.375, + "learning_rate": 7.963354474982382e-07, + "loss": 3.2064, + "step": 226 + }, + { + "epoch": 0.024005922165820644, + "grad_norm": 29.0, + "learning_rate": 7.998590556730092e-07, + "loss": 3.2974, + "step": 227 + }, + { + "epoch": 0.024111675126903553, + "grad_norm": 24.0, + "learning_rate": 8.033826638477801e-07, + "loss": 3.2801, + "step": 228 + }, + { + "epoch": 0.024217428087986465, + "grad_norm": 23.25, + "learning_rate": 8.069062720225511e-07, + "loss": 3.2264, + "step": 229 + }, + { + "epoch": 0.024323181049069373, + "grad_norm": 21.25, + "learning_rate": 8.104298801973222e-07, + "loss": 3.4614, + "step": 230 + }, + { + "epoch": 0.024428934010152285, + "grad_norm": 20.0, + "learning_rate": 8.139534883720931e-07, + "loss": 3.2032, + "step": 231 + }, + { + "epoch": 0.024534686971235193, + "grad_norm": 20.0, + "learning_rate": 8.174770965468641e-07, + "loss": 3.2047, + "step": 232 + }, + { + "epoch": 0.024640439932318105, + "grad_norm": 21.5, + "learning_rate": 8.210007047216351e-07, + "loss": 3.4226, + "step": 233 + }, + { + "epoch": 0.024746192893401017, + "grad_norm": 19.5, + "learning_rate": 8.245243128964061e-07, + "loss": 3.3676, + "step": 234 + }, + { + "epoch": 0.024851945854483925, + "grad_norm": 20.875, + "learning_rate": 8.28047921071177e-07, + "loss": 3.2258, + "step": 235 + }, + { + "epoch": 0.024957698815566837, + "grad_norm": 17.625, + "learning_rate": 8.315715292459479e-07, + "loss": 3.0411, + "step": 236 + }, + { + "epoch": 0.025063451776649745, + "grad_norm": 18.0, + "learning_rate": 8.350951374207189e-07, + "loss": 3.0421, + "step": 237 + }, + { + "epoch": 0.025169204737732657, + "grad_norm": 21.625, + "learning_rate": 8.386187455954898e-07, + "loss": 3.3592, + "step": 238 + }, + { + "epoch": 0.025274957698815565, + "grad_norm": 23.0, + "learning_rate": 8.421423537702608e-07, + "loss": 3.3078, + "step": 239 + }, + { + "epoch": 0.025380710659898477, + "grad_norm": 20.625, + "learning_rate": 8.456659619450318e-07, + "loss": 3.2938, + "step": 240 + }, + { + "epoch": 0.02548646362098139, + "grad_norm": 21.625, + "learning_rate": 8.491895701198028e-07, + "loss": 2.8207, + "step": 241 + }, + { + "epoch": 0.025592216582064297, + "grad_norm": 16.875, + "learning_rate": 8.527131782945737e-07, + "loss": 3.1796, + "step": 242 + }, + { + "epoch": 0.02569796954314721, + "grad_norm": 24.25, + "learning_rate": 8.562367864693446e-07, + "loss": 3.2474, + "step": 243 + }, + { + "epoch": 0.025803722504230117, + "grad_norm": 21.5, + "learning_rate": 8.597603946441156e-07, + "loss": 3.2877, + "step": 244 + }, + { + "epoch": 0.02590947546531303, + "grad_norm": 20.75, + "learning_rate": 8.632840028188866e-07, + "loss": 3.3085, + "step": 245 + }, + { + "epoch": 0.02601522842639594, + "grad_norm": 17.375, + "learning_rate": 8.668076109936575e-07, + "loss": 3.2206, + "step": 246 + }, + { + "epoch": 0.02612098138747885, + "grad_norm": 21.75, + "learning_rate": 8.703312191684285e-07, + "loss": 3.1233, + "step": 247 + }, + { + "epoch": 0.02622673434856176, + "grad_norm": 19.5, + "learning_rate": 8.738548273431995e-07, + "loss": 3.2112, + "step": 248 + }, + { + "epoch": 0.02633248730964467, + "grad_norm": 19.25, + "learning_rate": 8.773784355179705e-07, + "loss": 3.1399, + "step": 249 + }, + { + "epoch": 0.02643824027072758, + "grad_norm": 19.25, + "learning_rate": 8.809020436927413e-07, + "loss": 3.2693, + "step": 250 + }, + { + "epoch": 0.02654399323181049, + "grad_norm": 18.0, + "learning_rate": 8.844256518675123e-07, + "loss": 3.3879, + "step": 251 + }, + { + "epoch": 0.0266497461928934, + "grad_norm": 18.25, + "learning_rate": 8.879492600422834e-07, + "loss": 3.2233, + "step": 252 + }, + { + "epoch": 0.026755499153976313, + "grad_norm": 17.25, + "learning_rate": 8.914728682170544e-07, + "loss": 3.1425, + "step": 253 + }, + { + "epoch": 0.02686125211505922, + "grad_norm": 16.375, + "learning_rate": 8.949964763918253e-07, + "loss": 3.0997, + "step": 254 + }, + { + "epoch": 0.026967005076142133, + "grad_norm": 17.625, + "learning_rate": 8.985200845665963e-07, + "loss": 3.0831, + "step": 255 + }, + { + "epoch": 0.02707275803722504, + "grad_norm": 17.125, + "learning_rate": 9.020436927413673e-07, + "loss": 3.0776, + "step": 256 + }, + { + "epoch": 0.027178510998307953, + "grad_norm": 17.0, + "learning_rate": 9.055673009161383e-07, + "loss": 3.1954, + "step": 257 + }, + { + "epoch": 0.02728426395939086, + "grad_norm": 19.375, + "learning_rate": 9.090909090909091e-07, + "loss": 3.2033, + "step": 258 + }, + { + "epoch": 0.027390016920473773, + "grad_norm": 20.0, + "learning_rate": 9.126145172656801e-07, + "loss": 3.2158, + "step": 259 + }, + { + "epoch": 0.027495769881556685, + "grad_norm": 18.375, + "learning_rate": 9.161381254404511e-07, + "loss": 3.0714, + "step": 260 + }, + { + "epoch": 0.027601522842639593, + "grad_norm": 19.625, + "learning_rate": 9.196617336152221e-07, + "loss": 3.1148, + "step": 261 + }, + { + "epoch": 0.027707275803722505, + "grad_norm": 18.125, + "learning_rate": 9.23185341789993e-07, + "loss": 3.4516, + "step": 262 + }, + { + "epoch": 0.027813028764805414, + "grad_norm": 15.5625, + "learning_rate": 9.26708949964764e-07, + "loss": 3.1625, + "step": 263 + }, + { + "epoch": 0.027918781725888325, + "grad_norm": 15.375, + "learning_rate": 9.30232558139535e-07, + "loss": 3.1497, + "step": 264 + }, + { + "epoch": 0.028024534686971234, + "grad_norm": 19.75, + "learning_rate": 9.33756166314306e-07, + "loss": 3.0481, + "step": 265 + }, + { + "epoch": 0.028130287648054145, + "grad_norm": 20.875, + "learning_rate": 9.372797744890768e-07, + "loss": 3.5285, + "step": 266 + }, + { + "epoch": 0.028236040609137057, + "grad_norm": 15.0, + "learning_rate": 9.408033826638478e-07, + "loss": 2.9041, + "step": 267 + }, + { + "epoch": 0.028341793570219966, + "grad_norm": 15.6875, + "learning_rate": 9.443269908386188e-07, + "loss": 3.2835, + "step": 268 + }, + { + "epoch": 0.028447546531302877, + "grad_norm": 14.8125, + "learning_rate": 9.478505990133897e-07, + "loss": 2.895, + "step": 269 + }, + { + "epoch": 0.028553299492385786, + "grad_norm": 14.625, + "learning_rate": 9.513742071881607e-07, + "loss": 3.2074, + "step": 270 + }, + { + "epoch": 0.028659052453468697, + "grad_norm": 15.5625, + "learning_rate": 9.548978153629316e-07, + "loss": 3.2881, + "step": 271 + }, + { + "epoch": 0.028764805414551606, + "grad_norm": 18.5, + "learning_rate": 9.584214235377026e-07, + "loss": 3.1233, + "step": 272 + }, + { + "epoch": 0.028870558375634518, + "grad_norm": 13.75, + "learning_rate": 9.619450317124738e-07, + "loss": 3.2328, + "step": 273 + }, + { + "epoch": 0.02897631133671743, + "grad_norm": 17.75, + "learning_rate": 9.654686398872447e-07, + "loss": 3.1506, + "step": 274 + }, + { + "epoch": 0.029082064297800338, + "grad_norm": 14.375, + "learning_rate": 9.689922480620157e-07, + "loss": 2.9507, + "step": 275 + }, + { + "epoch": 0.02918781725888325, + "grad_norm": 15.6875, + "learning_rate": 9.725158562367865e-07, + "loss": 3.2562, + "step": 276 + }, + { + "epoch": 0.029293570219966158, + "grad_norm": 15.9375, + "learning_rate": 9.760394644115575e-07, + "loss": 3.2507, + "step": 277 + }, + { + "epoch": 0.02939932318104907, + "grad_norm": 15.6875, + "learning_rate": 9.795630725863285e-07, + "loss": 3.173, + "step": 278 + }, + { + "epoch": 0.02950507614213198, + "grad_norm": 15.8125, + "learning_rate": 9.830866807610995e-07, + "loss": 3.2369, + "step": 279 + }, + { + "epoch": 0.02961082910321489, + "grad_norm": 14.6875, + "learning_rate": 9.866102889358705e-07, + "loss": 3.1647, + "step": 280 + }, + { + "epoch": 0.0297165820642978, + "grad_norm": 12.9375, + "learning_rate": 9.901338971106414e-07, + "loss": 3.067, + "step": 281 + }, + { + "epoch": 0.02982233502538071, + "grad_norm": 13.1875, + "learning_rate": 9.936575052854124e-07, + "loss": 3.2583, + "step": 282 + }, + { + "epoch": 0.02992808798646362, + "grad_norm": 15.4375, + "learning_rate": 9.971811134601834e-07, + "loss": 3.231, + "step": 283 + }, + { + "epoch": 0.03003384094754653, + "grad_norm": 14.0, + "learning_rate": 1.0007047216349542e-06, + "loss": 3.1909, + "step": 284 + }, + { + "epoch": 0.030139593908629442, + "grad_norm": 13.25, + "learning_rate": 1.0042283298097252e-06, + "loss": 2.9142, + "step": 285 + }, + { + "epoch": 0.030245346869712354, + "grad_norm": 14.0625, + "learning_rate": 1.0077519379844962e-06, + "loss": 3.1298, + "step": 286 + }, + { + "epoch": 0.030351099830795262, + "grad_norm": 12.6875, + "learning_rate": 1.0112755461592672e-06, + "loss": 3.0403, + "step": 287 + }, + { + "epoch": 0.030456852791878174, + "grad_norm": 16.5, + "learning_rate": 1.0147991543340381e-06, + "loss": 3.0758, + "step": 288 + }, + { + "epoch": 0.030562605752961082, + "grad_norm": 12.625, + "learning_rate": 1.0183227625088091e-06, + "loss": 3.1574, + "step": 289 + }, + { + "epoch": 0.030668358714043994, + "grad_norm": 11.5625, + "learning_rate": 1.0218463706835801e-06, + "loss": 3.0579, + "step": 290 + }, + { + "epoch": 0.030774111675126902, + "grad_norm": 14.0, + "learning_rate": 1.025369978858351e-06, + "loss": 2.9732, + "step": 291 + }, + { + "epoch": 0.030879864636209814, + "grad_norm": 13.0, + "learning_rate": 1.0288935870331219e-06, + "loss": 2.9344, + "step": 292 + }, + { + "epoch": 0.030985617597292726, + "grad_norm": 13.8125, + "learning_rate": 1.0324171952078929e-06, + "loss": 2.949, + "step": 293 + }, + { + "epoch": 0.031091370558375634, + "grad_norm": 11.375, + "learning_rate": 1.0359408033826639e-06, + "loss": 2.8956, + "step": 294 + }, + { + "epoch": 0.031197123519458546, + "grad_norm": 12.75, + "learning_rate": 1.039464411557435e-06, + "loss": 3.1051, + "step": 295 + }, + { + "epoch": 0.031302876480541454, + "grad_norm": 13.75, + "learning_rate": 1.0429880197322058e-06, + "loss": 3.1083, + "step": 296 + }, + { + "epoch": 0.03140862944162436, + "grad_norm": 12.0625, + "learning_rate": 1.0465116279069768e-06, + "loss": 3.1032, + "step": 297 + }, + { + "epoch": 0.03151438240270728, + "grad_norm": 14.5625, + "learning_rate": 1.0500352360817478e-06, + "loss": 2.7733, + "step": 298 + }, + { + "epoch": 0.031620135363790186, + "grad_norm": 13.125, + "learning_rate": 1.0535588442565188e-06, + "loss": 2.9945, + "step": 299 + }, + { + "epoch": 0.031725888324873094, + "grad_norm": 13.6875, + "learning_rate": 1.0570824524312898e-06, + "loss": 3.0439, + "step": 300 + }, + { + "epoch": 0.03183164128595601, + "grad_norm": 14.5, + "learning_rate": 1.0606060606060608e-06, + "loss": 3.3183, + "step": 301 + }, + { + "epoch": 0.03193739424703892, + "grad_norm": 12.3125, + "learning_rate": 1.0641296687808318e-06, + "loss": 2.9687, + "step": 302 + }, + { + "epoch": 0.032043147208121826, + "grad_norm": 12.75, + "learning_rate": 1.0676532769556025e-06, + "loss": 2.7299, + "step": 303 + }, + { + "epoch": 0.032148900169204735, + "grad_norm": 14.875, + "learning_rate": 1.0711768851303735e-06, + "loss": 3.1886, + "step": 304 + }, + { + "epoch": 0.03225465313028765, + "grad_norm": 13.4375, + "learning_rate": 1.0747004933051445e-06, + "loss": 3.065, + "step": 305 + }, + { + "epoch": 0.03236040609137056, + "grad_norm": 11.8125, + "learning_rate": 1.0782241014799155e-06, + "loss": 3.1231, + "step": 306 + }, + { + "epoch": 0.03246615905245347, + "grad_norm": 12.375, + "learning_rate": 1.0817477096546865e-06, + "loss": 2.9281, + "step": 307 + }, + { + "epoch": 0.03257191201353638, + "grad_norm": 12.5625, + "learning_rate": 1.0852713178294575e-06, + "loss": 3.2123, + "step": 308 + }, + { + "epoch": 0.03267766497461929, + "grad_norm": 11.625, + "learning_rate": 1.0887949260042285e-06, + "loss": 3.0347, + "step": 309 + }, + { + "epoch": 0.0327834179357022, + "grad_norm": 11.75, + "learning_rate": 1.0923185341789995e-06, + "loss": 2.9814, + "step": 310 + }, + { + "epoch": 0.03288917089678511, + "grad_norm": 12.25, + "learning_rate": 1.0958421423537702e-06, + "loss": 2.9134, + "step": 311 + }, + { + "epoch": 0.03299492385786802, + "grad_norm": 13.1875, + "learning_rate": 1.0993657505285412e-06, + "loss": 3.1155, + "step": 312 + }, + { + "epoch": 0.03310067681895093, + "grad_norm": 12.375, + "learning_rate": 1.1028893587033122e-06, + "loss": 3.0666, + "step": 313 + }, + { + "epoch": 0.03320642978003384, + "grad_norm": 12.6875, + "learning_rate": 1.1064129668780832e-06, + "loss": 2.9597, + "step": 314 + }, + { + "epoch": 0.033312182741116754, + "grad_norm": 13.0625, + "learning_rate": 1.1099365750528542e-06, + "loss": 3.1623, + "step": 315 + }, + { + "epoch": 0.03341793570219966, + "grad_norm": 12.5625, + "learning_rate": 1.1134601832276252e-06, + "loss": 3.0193, + "step": 316 + }, + { + "epoch": 0.03352368866328257, + "grad_norm": 11.9375, + "learning_rate": 1.1169837914023962e-06, + "loss": 2.9404, + "step": 317 + }, + { + "epoch": 0.03362944162436548, + "grad_norm": 12.5625, + "learning_rate": 1.1205073995771671e-06, + "loss": 2.8875, + "step": 318 + }, + { + "epoch": 0.033735194585448394, + "grad_norm": 12.0625, + "learning_rate": 1.1240310077519381e-06, + "loss": 2.9812, + "step": 319 + }, + { + "epoch": 0.0338409475465313, + "grad_norm": 12.625, + "learning_rate": 1.1275546159267091e-06, + "loss": 3.1964, + "step": 320 + }, + { + "epoch": 0.03394670050761421, + "grad_norm": 12.25, + "learning_rate": 1.1310782241014801e-06, + "loss": 2.9208, + "step": 321 + }, + { + "epoch": 0.034052453468697126, + "grad_norm": 10.9375, + "learning_rate": 1.134601832276251e-06, + "loss": 3.0401, + "step": 322 + }, + { + "epoch": 0.034158206429780034, + "grad_norm": 10.0625, + "learning_rate": 1.1381254404510219e-06, + "loss": 3.045, + "step": 323 + }, + { + "epoch": 0.03426395939086294, + "grad_norm": 12.25, + "learning_rate": 1.1416490486257929e-06, + "loss": 3.043, + "step": 324 + }, + { + "epoch": 0.03436971235194585, + "grad_norm": 11.1875, + "learning_rate": 1.1451726568005638e-06, + "loss": 2.8909, + "step": 325 + }, + { + "epoch": 0.034475465313028766, + "grad_norm": 13.8125, + "learning_rate": 1.1486962649753348e-06, + "loss": 3.099, + "step": 326 + }, + { + "epoch": 0.034581218274111675, + "grad_norm": 11.3125, + "learning_rate": 1.1522198731501058e-06, + "loss": 3.3562, + "step": 327 + }, + { + "epoch": 0.03468697123519458, + "grad_norm": 10.625, + "learning_rate": 1.1557434813248768e-06, + "loss": 3.2262, + "step": 328 + }, + { + "epoch": 0.0347927241962775, + "grad_norm": 11.4375, + "learning_rate": 1.1592670894996478e-06, + "loss": 3.0989, + "step": 329 + }, + { + "epoch": 0.03489847715736041, + "grad_norm": 10.8125, + "learning_rate": 1.1627906976744188e-06, + "loss": 2.9162, + "step": 330 + }, + { + "epoch": 0.035004230118443315, + "grad_norm": 12.0, + "learning_rate": 1.1663143058491896e-06, + "loss": 2.7171, + "step": 331 + }, + { + "epoch": 0.03510998307952622, + "grad_norm": 12.625, + "learning_rate": 1.1698379140239606e-06, + "loss": 3.0176, + "step": 332 + }, + { + "epoch": 0.03521573604060914, + "grad_norm": 11.0625, + "learning_rate": 1.1733615221987315e-06, + "loss": 3.1157, + "step": 333 + }, + { + "epoch": 0.03532148900169205, + "grad_norm": 11.375, + "learning_rate": 1.1768851303735025e-06, + "loss": 2.8643, + "step": 334 + }, + { + "epoch": 0.035427241962774955, + "grad_norm": 15.5, + "learning_rate": 1.1804087385482735e-06, + "loss": 2.7006, + "step": 335 + }, + { + "epoch": 0.03553299492385787, + "grad_norm": 10.4375, + "learning_rate": 1.1839323467230445e-06, + "loss": 3.0363, + "step": 336 + }, + { + "epoch": 0.03563874788494078, + "grad_norm": 13.0625, + "learning_rate": 1.1874559548978155e-06, + "loss": 3.2555, + "step": 337 + }, + { + "epoch": 0.03574450084602369, + "grad_norm": 11.375, + "learning_rate": 1.1909795630725863e-06, + "loss": 2.952, + "step": 338 + }, + { + "epoch": 0.0358502538071066, + "grad_norm": 10.3125, + "learning_rate": 1.1945031712473575e-06, + "loss": 2.8793, + "step": 339 + }, + { + "epoch": 0.03595600676818951, + "grad_norm": 11.25, + "learning_rate": 1.1980267794221285e-06, + "loss": 3.0475, + "step": 340 + }, + { + "epoch": 0.03606175972927242, + "grad_norm": 11.375, + "learning_rate": 1.2015503875968994e-06, + "loss": 3.006, + "step": 341 + }, + { + "epoch": 0.03616751269035533, + "grad_norm": 10.5625, + "learning_rate": 1.2050739957716704e-06, + "loss": 2.9646, + "step": 342 + }, + { + "epoch": 0.03627326565143824, + "grad_norm": 9.875, + "learning_rate": 1.2085976039464412e-06, + "loss": 2.9881, + "step": 343 + }, + { + "epoch": 0.03637901861252115, + "grad_norm": 10.3125, + "learning_rate": 1.2121212121212122e-06, + "loss": 2.9851, + "step": 344 + }, + { + "epoch": 0.03648477157360406, + "grad_norm": 10.3125, + "learning_rate": 1.2156448202959832e-06, + "loss": 2.9429, + "step": 345 + }, + { + "epoch": 0.036590524534686975, + "grad_norm": 10.625, + "learning_rate": 1.2191684284707542e-06, + "loss": 3.0807, + "step": 346 + }, + { + "epoch": 0.03669627749576988, + "grad_norm": 10.0625, + "learning_rate": 1.2226920366455252e-06, + "loss": 3.0994, + "step": 347 + }, + { + "epoch": 0.03680203045685279, + "grad_norm": 11.125, + "learning_rate": 1.2262156448202961e-06, + "loss": 3.1917, + "step": 348 + }, + { + "epoch": 0.0369077834179357, + "grad_norm": 12.9375, + "learning_rate": 1.2297392529950671e-06, + "loss": 3.4028, + "step": 349 + }, + { + "epoch": 0.037013536379018615, + "grad_norm": 10.125, + "learning_rate": 1.233262861169838e-06, + "loss": 3.0315, + "step": 350 + }, + { + "epoch": 0.03711928934010152, + "grad_norm": 9.875, + "learning_rate": 1.236786469344609e-06, + "loss": 2.9717, + "step": 351 + }, + { + "epoch": 0.03722504230118443, + "grad_norm": 11.5, + "learning_rate": 1.2403100775193799e-06, + "loss": 3.4263, + "step": 352 + }, + { + "epoch": 0.03733079526226735, + "grad_norm": 12.0625, + "learning_rate": 1.2438336856941509e-06, + "loss": 3.0615, + "step": 353 + }, + { + "epoch": 0.037436548223350255, + "grad_norm": 10.3125, + "learning_rate": 1.2473572938689219e-06, + "loss": 3.0467, + "step": 354 + }, + { + "epoch": 0.03754230118443316, + "grad_norm": 9.9375, + "learning_rate": 1.2508809020436928e-06, + "loss": 3.0969, + "step": 355 + }, + { + "epoch": 0.03764805414551607, + "grad_norm": 13.3125, + "learning_rate": 1.2544045102184638e-06, + "loss": 3.0876, + "step": 356 + }, + { + "epoch": 0.03775380710659899, + "grad_norm": 9.875, + "learning_rate": 1.2579281183932348e-06, + "loss": 3.0386, + "step": 357 + }, + { + "epoch": 0.037859560067681895, + "grad_norm": 13.0625, + "learning_rate": 1.2614517265680056e-06, + "loss": 3.0185, + "step": 358 + }, + { + "epoch": 0.037965313028764804, + "grad_norm": 10.3125, + "learning_rate": 1.2649753347427768e-06, + "loss": 2.9272, + "step": 359 + }, + { + "epoch": 0.03807106598984772, + "grad_norm": 11.9375, + "learning_rate": 1.2684989429175476e-06, + "loss": 3.1216, + "step": 360 + }, + { + "epoch": 0.03817681895093063, + "grad_norm": 10.875, + "learning_rate": 1.2720225510923188e-06, + "loss": 2.8979, + "step": 361 + }, + { + "epoch": 0.038282571912013535, + "grad_norm": 10.0, + "learning_rate": 1.2755461592670896e-06, + "loss": 2.9956, + "step": 362 + }, + { + "epoch": 0.038388324873096444, + "grad_norm": 9.875, + "learning_rate": 1.2790697674418605e-06, + "loss": 2.8573, + "step": 363 + }, + { + "epoch": 0.03849407783417936, + "grad_norm": 10.3125, + "learning_rate": 1.2825933756166315e-06, + "loss": 3.0014, + "step": 364 + }, + { + "epoch": 0.03859983079526227, + "grad_norm": 10.9375, + "learning_rate": 1.2861169837914025e-06, + "loss": 3.1294, + "step": 365 + }, + { + "epoch": 0.038705583756345176, + "grad_norm": 10.0625, + "learning_rate": 1.2896405919661733e-06, + "loss": 3.1351, + "step": 366 + }, + { + "epoch": 0.03881133671742809, + "grad_norm": 9.25, + "learning_rate": 1.2931642001409445e-06, + "loss": 2.9083, + "step": 367 + }, + { + "epoch": 0.038917089678511, + "grad_norm": 9.4375, + "learning_rate": 1.2966878083157153e-06, + "loss": 2.8369, + "step": 368 + }, + { + "epoch": 0.03902284263959391, + "grad_norm": 10.8125, + "learning_rate": 1.3002114164904865e-06, + "loss": 3.0386, + "step": 369 + }, + { + "epoch": 0.039128595600676816, + "grad_norm": 9.625, + "learning_rate": 1.3037350246652572e-06, + "loss": 2.949, + "step": 370 + }, + { + "epoch": 0.03923434856175973, + "grad_norm": 9.5625, + "learning_rate": 1.3072586328400282e-06, + "loss": 2.8351, + "step": 371 + }, + { + "epoch": 0.03934010152284264, + "grad_norm": 10.625, + "learning_rate": 1.3107822410147994e-06, + "loss": 3.2478, + "step": 372 + }, + { + "epoch": 0.03944585448392555, + "grad_norm": 9.9375, + "learning_rate": 1.3143058491895702e-06, + "loss": 2.9608, + "step": 373 + }, + { + "epoch": 0.03955160744500846, + "grad_norm": 10.125, + "learning_rate": 1.3178294573643414e-06, + "loss": 2.9897, + "step": 374 + }, + { + "epoch": 0.03965736040609137, + "grad_norm": 10.3125, + "learning_rate": 1.3213530655391122e-06, + "loss": 2.6826, + "step": 375 + }, + { + "epoch": 0.03976311336717428, + "grad_norm": 9.9375, + "learning_rate": 1.3248766737138832e-06, + "loss": 2.7083, + "step": 376 + }, + { + "epoch": 0.03986886632825719, + "grad_norm": 11.25, + "learning_rate": 1.3284002818886542e-06, + "loss": 2.6146, + "step": 377 + }, + { + "epoch": 0.0399746192893401, + "grad_norm": 11.25, + "learning_rate": 1.3319238900634251e-06, + "loss": 3.0488, + "step": 378 + }, + { + "epoch": 0.04008037225042301, + "grad_norm": 9.0, + "learning_rate": 1.335447498238196e-06, + "loss": 2.8475, + "step": 379 + }, + { + "epoch": 0.04018612521150592, + "grad_norm": 9.75, + "learning_rate": 1.3389711064129671e-06, + "loss": 3.1055, + "step": 380 + }, + { + "epoch": 0.040291878172588835, + "grad_norm": 9.125, + "learning_rate": 1.342494714587738e-06, + "loss": 2.6959, + "step": 381 + }, + { + "epoch": 0.040397631133671744, + "grad_norm": 9.375, + "learning_rate": 1.3460183227625089e-06, + "loss": 2.9737, + "step": 382 + }, + { + "epoch": 0.04050338409475465, + "grad_norm": 10.5, + "learning_rate": 1.3495419309372799e-06, + "loss": 2.9434, + "step": 383 + }, + { + "epoch": 0.04060913705583756, + "grad_norm": 9.0625, + "learning_rate": 1.3530655391120509e-06, + "loss": 3.027, + "step": 384 + }, + { + "epoch": 0.040714890016920476, + "grad_norm": 9.4375, + "learning_rate": 1.3565891472868216e-06, + "loss": 2.7472, + "step": 385 + }, + { + "epoch": 0.040820642978003384, + "grad_norm": 9.3125, + "learning_rate": 1.3601127554615928e-06, + "loss": 2.9017, + "step": 386 + }, + { + "epoch": 0.04092639593908629, + "grad_norm": 9.25, + "learning_rate": 1.3636363636363636e-06, + "loss": 2.8187, + "step": 387 + }, + { + "epoch": 0.04103214890016921, + "grad_norm": 9.625, + "learning_rate": 1.3671599718111348e-06, + "loss": 3.0454, + "step": 388 + }, + { + "epoch": 0.041137901861252116, + "grad_norm": 9.625, + "learning_rate": 1.3706835799859056e-06, + "loss": 3.0132, + "step": 389 + }, + { + "epoch": 0.041243654822335024, + "grad_norm": 10.25, + "learning_rate": 1.3742071881606766e-06, + "loss": 2.9829, + "step": 390 + }, + { + "epoch": 0.04134940778341793, + "grad_norm": 9.875, + "learning_rate": 1.3777307963354476e-06, + "loss": 3.0246, + "step": 391 + }, + { + "epoch": 0.04145516074450085, + "grad_norm": 12.625, + "learning_rate": 1.3812544045102186e-06, + "loss": 3.0387, + "step": 392 + }, + { + "epoch": 0.041560913705583756, + "grad_norm": 9.5625, + "learning_rate": 1.3847780126849898e-06, + "loss": 2.9387, + "step": 393 + }, + { + "epoch": 0.041666666666666664, + "grad_norm": 12.25, + "learning_rate": 1.3883016208597605e-06, + "loss": 2.6643, + "step": 394 + }, + { + "epoch": 0.04177241962774958, + "grad_norm": 10.0625, + "learning_rate": 1.3918252290345315e-06, + "loss": 3.0207, + "step": 395 + }, + { + "epoch": 0.04187817258883249, + "grad_norm": 10.1875, + "learning_rate": 1.3953488372093025e-06, + "loss": 3.0532, + "step": 396 + }, + { + "epoch": 0.041983925549915396, + "grad_norm": 9.5625, + "learning_rate": 1.3988724453840735e-06, + "loss": 2.9767, + "step": 397 + }, + { + "epoch": 0.042089678510998305, + "grad_norm": 10.625, + "learning_rate": 1.4023960535588443e-06, + "loss": 2.8685, + "step": 398 + }, + { + "epoch": 0.04219543147208122, + "grad_norm": 9.875, + "learning_rate": 1.4059196617336155e-06, + "loss": 2.907, + "step": 399 + }, + { + "epoch": 0.04230118443316413, + "grad_norm": 12.125, + "learning_rate": 1.4094432699083862e-06, + "loss": 2.8543, + "step": 400 + }, + { + "epoch": 0.042406937394247037, + "grad_norm": 9.5625, + "learning_rate": 1.4129668780831574e-06, + "loss": 3.1113, + "step": 401 + }, + { + "epoch": 0.04251269035532995, + "grad_norm": 9.6875, + "learning_rate": 1.4164904862579282e-06, + "loss": 3.0078, + "step": 402 + }, + { + "epoch": 0.04261844331641286, + "grad_norm": 11.75, + "learning_rate": 1.4200140944326992e-06, + "loss": 2.8093, + "step": 403 + }, + { + "epoch": 0.04272419627749577, + "grad_norm": 9.0625, + "learning_rate": 1.4235377026074702e-06, + "loss": 2.747, + "step": 404 + }, + { + "epoch": 0.04282994923857868, + "grad_norm": 9.8125, + "learning_rate": 1.4270613107822412e-06, + "loss": 3.1362, + "step": 405 + }, + { + "epoch": 0.04293570219966159, + "grad_norm": 8.9375, + "learning_rate": 1.430584918957012e-06, + "loss": 2.9416, + "step": 406 + }, + { + "epoch": 0.0430414551607445, + "grad_norm": 9.6875, + "learning_rate": 1.4341085271317832e-06, + "loss": 3.1504, + "step": 407 + }, + { + "epoch": 0.04314720812182741, + "grad_norm": 9.375, + "learning_rate": 1.437632135306554e-06, + "loss": 2.8461, + "step": 408 + }, + { + "epoch": 0.043252961082910324, + "grad_norm": 10.1875, + "learning_rate": 1.4411557434813251e-06, + "loss": 2.8614, + "step": 409 + }, + { + "epoch": 0.04335871404399323, + "grad_norm": 9.0, + "learning_rate": 1.444679351656096e-06, + "loss": 3.0282, + "step": 410 + }, + { + "epoch": 0.04346446700507614, + "grad_norm": 9.8125, + "learning_rate": 1.448202959830867e-06, + "loss": 2.9139, + "step": 411 + }, + { + "epoch": 0.043570219966159056, + "grad_norm": 9.875, + "learning_rate": 1.4517265680056377e-06, + "loss": 3.036, + "step": 412 + }, + { + "epoch": 0.043675972927241964, + "grad_norm": 11.0625, + "learning_rate": 1.4552501761804089e-06, + "loss": 2.9626, + "step": 413 + }, + { + "epoch": 0.04378172588832487, + "grad_norm": 10.375, + "learning_rate": 1.4587737843551796e-06, + "loss": 2.9921, + "step": 414 + }, + { + "epoch": 0.04388747884940778, + "grad_norm": 8.75, + "learning_rate": 1.4622973925299508e-06, + "loss": 2.8192, + "step": 415 + }, + { + "epoch": 0.043993231810490696, + "grad_norm": 10.0625, + "learning_rate": 1.4658210007047218e-06, + "loss": 3.0319, + "step": 416 + }, + { + "epoch": 0.044098984771573604, + "grad_norm": 9.875, + "learning_rate": 1.4693446088794926e-06, + "loss": 2.8327, + "step": 417 + }, + { + "epoch": 0.04420473773265651, + "grad_norm": 10.0625, + "learning_rate": 1.4728682170542638e-06, + "loss": 3.0083, + "step": 418 + }, + { + "epoch": 0.04431049069373943, + "grad_norm": 9.125, + "learning_rate": 1.4763918252290346e-06, + "loss": 2.674, + "step": 419 + }, + { + "epoch": 0.044416243654822336, + "grad_norm": 11.1875, + "learning_rate": 1.4799154334038058e-06, + "loss": 2.7831, + "step": 420 + }, + { + "epoch": 0.044521996615905245, + "grad_norm": 8.9375, + "learning_rate": 1.4834390415785766e-06, + "loss": 2.9749, + "step": 421 + }, + { + "epoch": 0.04462774957698815, + "grad_norm": 8.6875, + "learning_rate": 1.4869626497533476e-06, + "loss": 2.9581, + "step": 422 + }, + { + "epoch": 0.04473350253807107, + "grad_norm": 9.3125, + "learning_rate": 1.4904862579281185e-06, + "loss": 2.9759, + "step": 423 + }, + { + "epoch": 0.04483925549915398, + "grad_norm": 9.25, + "learning_rate": 1.4940098661028895e-06, + "loss": 3.0514, + "step": 424 + }, + { + "epoch": 0.044945008460236885, + "grad_norm": 8.9375, + "learning_rate": 1.4975334742776603e-06, + "loss": 2.8017, + "step": 425 + }, + { + "epoch": 0.0450507614213198, + "grad_norm": 9.125, + "learning_rate": 1.5010570824524315e-06, + "loss": 2.9373, + "step": 426 + }, + { + "epoch": 0.04515651438240271, + "grad_norm": 10.5, + "learning_rate": 1.5045806906272023e-06, + "loss": 2.9503, + "step": 427 + }, + { + "epoch": 0.04526226734348562, + "grad_norm": 9.1875, + "learning_rate": 1.5081042988019735e-06, + "loss": 3.0655, + "step": 428 + }, + { + "epoch": 0.045368020304568525, + "grad_norm": 9.4375, + "learning_rate": 1.5116279069767443e-06, + "loss": 2.9002, + "step": 429 + }, + { + "epoch": 0.04547377326565144, + "grad_norm": 10.25, + "learning_rate": 1.5151515151515152e-06, + "loss": 2.8539, + "step": 430 + }, + { + "epoch": 0.04557952622673435, + "grad_norm": 10.125, + "learning_rate": 1.5186751233262862e-06, + "loss": 3.0796, + "step": 431 + }, + { + "epoch": 0.04568527918781726, + "grad_norm": 9.875, + "learning_rate": 1.5221987315010572e-06, + "loss": 2.9602, + "step": 432 + }, + { + "epoch": 0.04579103214890017, + "grad_norm": 9.3125, + "learning_rate": 1.525722339675828e-06, + "loss": 2.8304, + "step": 433 + }, + { + "epoch": 0.04589678510998308, + "grad_norm": 9.1875, + "learning_rate": 1.5292459478505992e-06, + "loss": 2.8444, + "step": 434 + }, + { + "epoch": 0.04600253807106599, + "grad_norm": 11.4375, + "learning_rate": 1.53276955602537e-06, + "loss": 3.1211, + "step": 435 + }, + { + "epoch": 0.0461082910321489, + "grad_norm": 10.875, + "learning_rate": 1.5362931642001412e-06, + "loss": 2.9202, + "step": 436 + }, + { + "epoch": 0.04621404399323181, + "grad_norm": 9.125, + "learning_rate": 1.5398167723749122e-06, + "loss": 2.8633, + "step": 437 + }, + { + "epoch": 0.04631979695431472, + "grad_norm": 9.1875, + "learning_rate": 1.543340380549683e-06, + "loss": 2.9306, + "step": 438 + }, + { + "epoch": 0.04642554991539763, + "grad_norm": 10.125, + "learning_rate": 1.5468639887244541e-06, + "loss": 2.9843, + "step": 439 + }, + { + "epoch": 0.046531302876480544, + "grad_norm": 9.5, + "learning_rate": 1.550387596899225e-06, + "loss": 2.8668, + "step": 440 + }, + { + "epoch": 0.04663705583756345, + "grad_norm": 9.375, + "learning_rate": 1.5539112050739961e-06, + "loss": 2.8407, + "step": 441 + }, + { + "epoch": 0.04674280879864636, + "grad_norm": 9.5, + "learning_rate": 1.5574348132487669e-06, + "loss": 3.0332, + "step": 442 + }, + { + "epoch": 0.04684856175972927, + "grad_norm": 10.4375, + "learning_rate": 1.5609584214235379e-06, + "loss": 2.9532, + "step": 443 + }, + { + "epoch": 0.046954314720812185, + "grad_norm": 8.6875, + "learning_rate": 1.5644820295983086e-06, + "loss": 2.8713, + "step": 444 + }, + { + "epoch": 0.04706006768189509, + "grad_norm": 9.1875, + "learning_rate": 1.5680056377730798e-06, + "loss": 2.8109, + "step": 445 + }, + { + "epoch": 0.047165820642978, + "grad_norm": 8.625, + "learning_rate": 1.5715292459478506e-06, + "loss": 2.8698, + "step": 446 + }, + { + "epoch": 0.04727157360406092, + "grad_norm": 11.1875, + "learning_rate": 1.5750528541226218e-06, + "loss": 2.7498, + "step": 447 + }, + { + "epoch": 0.047377326565143825, + "grad_norm": 10.0, + "learning_rate": 1.5785764622973926e-06, + "loss": 2.9365, + "step": 448 + }, + { + "epoch": 0.04748307952622673, + "grad_norm": 10.9375, + "learning_rate": 1.5821000704721636e-06, + "loss": 3.0387, + "step": 449 + }, + { + "epoch": 0.04758883248730964, + "grad_norm": 9.125, + "learning_rate": 1.5856236786469346e-06, + "loss": 2.8643, + "step": 450 + }, + { + "epoch": 0.04769458544839256, + "grad_norm": 10.625, + "learning_rate": 1.5891472868217056e-06, + "loss": 2.7773, + "step": 451 + }, + { + "epoch": 0.047800338409475465, + "grad_norm": 8.75, + "learning_rate": 1.5926708949964763e-06, + "loss": 3.0679, + "step": 452 + }, + { + "epoch": 0.047906091370558374, + "grad_norm": 8.75, + "learning_rate": 1.5961945031712475e-06, + "loss": 2.9428, + "step": 453 + }, + { + "epoch": 0.04801184433164129, + "grad_norm": 9.5, + "learning_rate": 1.5997181113460183e-06, + "loss": 2.91, + "step": 454 + }, + { + "epoch": 0.0481175972927242, + "grad_norm": 11.25, + "learning_rate": 1.6032417195207895e-06, + "loss": 2.6975, + "step": 455 + }, + { + "epoch": 0.048223350253807105, + "grad_norm": 10.0625, + "learning_rate": 1.6067653276955603e-06, + "loss": 2.8516, + "step": 456 + }, + { + "epoch": 0.048329103214890014, + "grad_norm": 9.125, + "learning_rate": 1.6102889358703313e-06, + "loss": 2.9495, + "step": 457 + }, + { + "epoch": 0.04843485617597293, + "grad_norm": 9.8125, + "learning_rate": 1.6138125440451023e-06, + "loss": 2.9531, + "step": 458 + }, + { + "epoch": 0.04854060913705584, + "grad_norm": 8.75, + "learning_rate": 1.6173361522198733e-06, + "loss": 3.0741, + "step": 459 + }, + { + "epoch": 0.048646362098138746, + "grad_norm": 8.3125, + "learning_rate": 1.6208597603946445e-06, + "loss": 2.8202, + "step": 460 + }, + { + "epoch": 0.04875211505922166, + "grad_norm": 9.1875, + "learning_rate": 1.6243833685694152e-06, + "loss": 2.7913, + "step": 461 + }, + { + "epoch": 0.04885786802030457, + "grad_norm": 8.4375, + "learning_rate": 1.6279069767441862e-06, + "loss": 2.7094, + "step": 462 + }, + { + "epoch": 0.04896362098138748, + "grad_norm": 9.1875, + "learning_rate": 1.6314305849189572e-06, + "loss": 2.929, + "step": 463 + }, + { + "epoch": 0.049069373942470386, + "grad_norm": 9.1875, + "learning_rate": 1.6349541930937282e-06, + "loss": 2.8219, + "step": 464 + }, + { + "epoch": 0.0491751269035533, + "grad_norm": 9.5625, + "learning_rate": 1.638477801268499e-06, + "loss": 3.0709, + "step": 465 + }, + { + "epoch": 0.04928087986463621, + "grad_norm": 10.9375, + "learning_rate": 1.6420014094432702e-06, + "loss": 2.8241, + "step": 466 + }, + { + "epoch": 0.04938663282571912, + "grad_norm": 9.0, + "learning_rate": 1.645525017618041e-06, + "loss": 3.1018, + "step": 467 + }, + { + "epoch": 0.04949238578680203, + "grad_norm": 8.8125, + "learning_rate": 1.6490486257928121e-06, + "loss": 2.9837, + "step": 468 + }, + { + "epoch": 0.04959813874788494, + "grad_norm": 8.3125, + "learning_rate": 1.652572233967583e-06, + "loss": 2.7916, + "step": 469 + }, + { + "epoch": 0.04970389170896785, + "grad_norm": 9.0625, + "learning_rate": 1.656095842142354e-06, + "loss": 3.0319, + "step": 470 + }, + { + "epoch": 0.04980964467005076, + "grad_norm": 8.3125, + "learning_rate": 1.6596194503171247e-06, + "loss": 2.8162, + "step": 471 + }, + { + "epoch": 0.04991539763113367, + "grad_norm": 9.0, + "learning_rate": 1.6631430584918959e-06, + "loss": 2.9298, + "step": 472 + }, + { + "epoch": 0.05002115059221658, + "grad_norm": 9.375, + "learning_rate": 1.6666666666666667e-06, + "loss": 2.904, + "step": 473 + }, + { + "epoch": 0.05012690355329949, + "grad_norm": 8.9375, + "learning_rate": 1.6701902748414379e-06, + "loss": 3.1226, + "step": 474 + }, + { + "epoch": 0.050232656514382405, + "grad_norm": 10.0625, + "learning_rate": 1.6737138830162086e-06, + "loss": 3.3281, + "step": 475 + }, + { + "epoch": 0.050338409475465314, + "grad_norm": 10.0625, + "learning_rate": 1.6772374911909796e-06, + "loss": 2.8227, + "step": 476 + }, + { + "epoch": 0.05044416243654822, + "grad_norm": 8.5, + "learning_rate": 1.6807610993657506e-06, + "loss": 2.9102, + "step": 477 + }, + { + "epoch": 0.05054991539763113, + "grad_norm": 11.375, + "learning_rate": 1.6842847075405216e-06, + "loss": 2.6628, + "step": 478 + }, + { + "epoch": 0.050655668358714045, + "grad_norm": 8.25, + "learning_rate": 1.6878083157152924e-06, + "loss": 2.8268, + "step": 479 + }, + { + "epoch": 0.050761421319796954, + "grad_norm": 8.875, + "learning_rate": 1.6913319238900636e-06, + "loss": 3.1022, + "step": 480 + }, + { + "epoch": 0.05086717428087986, + "grad_norm": 11.25, + "learning_rate": 1.6948555320648346e-06, + "loss": 2.695, + "step": 481 + }, + { + "epoch": 0.05097292724196278, + "grad_norm": 8.625, + "learning_rate": 1.6983791402396056e-06, + "loss": 2.8664, + "step": 482 + }, + { + "epoch": 0.051078680203045686, + "grad_norm": 9.0, + "learning_rate": 1.7019027484143765e-06, + "loss": 2.795, + "step": 483 + }, + { + "epoch": 0.051184433164128594, + "grad_norm": 8.875, + "learning_rate": 1.7054263565891473e-06, + "loss": 2.8407, + "step": 484 + }, + { + "epoch": 0.05129018612521151, + "grad_norm": 9.25, + "learning_rate": 1.7089499647639185e-06, + "loss": 2.952, + "step": 485 + }, + { + "epoch": 0.05139593908629442, + "grad_norm": 10.0, + "learning_rate": 1.7124735729386893e-06, + "loss": 3.3295, + "step": 486 + }, + { + "epoch": 0.051501692047377326, + "grad_norm": 9.0625, + "learning_rate": 1.7159971811134605e-06, + "loss": 2.7126, + "step": 487 + }, + { + "epoch": 0.051607445008460234, + "grad_norm": 9.4375, + "learning_rate": 1.7195207892882313e-06, + "loss": 2.8709, + "step": 488 + }, + { + "epoch": 0.05171319796954315, + "grad_norm": 9.8125, + "learning_rate": 1.7230443974630023e-06, + "loss": 2.7861, + "step": 489 + }, + { + "epoch": 0.05181895093062606, + "grad_norm": 11.375, + "learning_rate": 1.7265680056377732e-06, + "loss": 2.8504, + "step": 490 + }, + { + "epoch": 0.051924703891708966, + "grad_norm": 8.125, + "learning_rate": 1.7300916138125442e-06, + "loss": 2.9187, + "step": 491 + }, + { + "epoch": 0.05203045685279188, + "grad_norm": 8.3125, + "learning_rate": 1.733615221987315e-06, + "loss": 2.9907, + "step": 492 + }, + { + "epoch": 0.05213620981387479, + "grad_norm": 9.1875, + "learning_rate": 1.7371388301620862e-06, + "loss": 2.9847, + "step": 493 + }, + { + "epoch": 0.0522419627749577, + "grad_norm": 8.75, + "learning_rate": 1.740662438336857e-06, + "loss": 2.8677, + "step": 494 + }, + { + "epoch": 0.052347715736040606, + "grad_norm": 10.5625, + "learning_rate": 1.7441860465116282e-06, + "loss": 2.7738, + "step": 495 + }, + { + "epoch": 0.05245346869712352, + "grad_norm": 9.5625, + "learning_rate": 1.747709654686399e-06, + "loss": 2.8749, + "step": 496 + }, + { + "epoch": 0.05255922165820643, + "grad_norm": 9.1875, + "learning_rate": 1.75123326286117e-06, + "loss": 3.0514, + "step": 497 + }, + { + "epoch": 0.05266497461928934, + "grad_norm": 8.6875, + "learning_rate": 1.754756871035941e-06, + "loss": 2.7204, + "step": 498 + }, + { + "epoch": 0.052770727580372254, + "grad_norm": 8.1875, + "learning_rate": 1.758280479210712e-06, + "loss": 2.8242, + "step": 499 + }, + { + "epoch": 0.05287648054145516, + "grad_norm": 8.4375, + "learning_rate": 1.7618040873854827e-06, + "loss": 2.8311, + "step": 500 + }, + { + "epoch": 0.05298223350253807, + "grad_norm": 8.8125, + "learning_rate": 1.765327695560254e-06, + "loss": 2.9026, + "step": 501 + }, + { + "epoch": 0.05308798646362098, + "grad_norm": 8.5, + "learning_rate": 1.7688513037350247e-06, + "loss": 2.7285, + "step": 502 + }, + { + "epoch": 0.053193739424703894, + "grad_norm": 8.9375, + "learning_rate": 1.7723749119097959e-06, + "loss": 2.8597, + "step": 503 + }, + { + "epoch": 0.0532994923857868, + "grad_norm": 8.5, + "learning_rate": 1.7758985200845669e-06, + "loss": 2.9608, + "step": 504 + }, + { + "epoch": 0.05340524534686971, + "grad_norm": 8.9375, + "learning_rate": 1.7794221282593376e-06, + "loss": 2.8394, + "step": 505 + }, + { + "epoch": 0.053510998307952626, + "grad_norm": 8.625, + "learning_rate": 1.7829457364341088e-06, + "loss": 3.0286, + "step": 506 + }, + { + "epoch": 0.053616751269035534, + "grad_norm": 8.6875, + "learning_rate": 1.7864693446088796e-06, + "loss": 2.8343, + "step": 507 + }, + { + "epoch": 0.05372250423011844, + "grad_norm": 8.4375, + "learning_rate": 1.7899929527836506e-06, + "loss": 2.88, + "step": 508 + }, + { + "epoch": 0.05382825719120135, + "grad_norm": 8.9375, + "learning_rate": 1.7935165609584216e-06, + "loss": 2.7837, + "step": 509 + }, + { + "epoch": 0.053934010152284266, + "grad_norm": 9.625, + "learning_rate": 1.7970401691331926e-06, + "loss": 2.9412, + "step": 510 + }, + { + "epoch": 0.054039763113367174, + "grad_norm": 8.75, + "learning_rate": 1.8005637773079634e-06, + "loss": 2.7244, + "step": 511 + }, + { + "epoch": 0.05414551607445008, + "grad_norm": 9.25, + "learning_rate": 1.8040873854827346e-06, + "loss": 2.7892, + "step": 512 + }, + { + "epoch": 0.054251269035533, + "grad_norm": 8.9375, + "learning_rate": 1.8076109936575053e-06, + "loss": 3.0005, + "step": 513 + }, + { + "epoch": 0.054357021996615906, + "grad_norm": 9.4375, + "learning_rate": 1.8111346018322765e-06, + "loss": 3.0647, + "step": 514 + }, + { + "epoch": 0.054462774957698815, + "grad_norm": 9.5, + "learning_rate": 1.8146582100070473e-06, + "loss": 2.8632, + "step": 515 + }, + { + "epoch": 0.05456852791878172, + "grad_norm": 10.125, + "learning_rate": 1.8181818181818183e-06, + "loss": 3.0604, + "step": 516 + }, + { + "epoch": 0.05467428087986464, + "grad_norm": 9.75, + "learning_rate": 1.8217054263565893e-06, + "loss": 3.1164, + "step": 517 + }, + { + "epoch": 0.054780033840947547, + "grad_norm": 7.71875, + "learning_rate": 1.8252290345313603e-06, + "loss": 2.9984, + "step": 518 + }, + { + "epoch": 0.054885786802030455, + "grad_norm": 8.375, + "learning_rate": 1.828752642706131e-06, + "loss": 2.9318, + "step": 519 + }, + { + "epoch": 0.05499153976311337, + "grad_norm": 8.4375, + "learning_rate": 1.8322762508809022e-06, + "loss": 2.8741, + "step": 520 + }, + { + "epoch": 0.05509729272419628, + "grad_norm": 8.75, + "learning_rate": 1.835799859055673e-06, + "loss": 2.6578, + "step": 521 + }, + { + "epoch": 0.05520304568527919, + "grad_norm": 8.0625, + "learning_rate": 1.8393234672304442e-06, + "loss": 2.957, + "step": 522 + }, + { + "epoch": 0.055308798646362095, + "grad_norm": 8.375, + "learning_rate": 1.842847075405215e-06, + "loss": 3.0142, + "step": 523 + }, + { + "epoch": 0.05541455160744501, + "grad_norm": 8.25, + "learning_rate": 1.846370683579986e-06, + "loss": 2.6888, + "step": 524 + }, + { + "epoch": 0.05552030456852792, + "grad_norm": 8.9375, + "learning_rate": 1.8498942917547572e-06, + "loss": 2.7897, + "step": 525 + }, + { + "epoch": 0.05562605752961083, + "grad_norm": 7.9375, + "learning_rate": 1.853417899929528e-06, + "loss": 2.8645, + "step": 526 + }, + { + "epoch": 0.05573181049069374, + "grad_norm": 9.25, + "learning_rate": 1.8569415081042992e-06, + "loss": 2.79, + "step": 527 + }, + { + "epoch": 0.05583756345177665, + "grad_norm": 8.75, + "learning_rate": 1.86046511627907e-06, + "loss": 2.8971, + "step": 528 + }, + { + "epoch": 0.05594331641285956, + "grad_norm": 8.5625, + "learning_rate": 1.863988724453841e-06, + "loss": 2.6958, + "step": 529 + }, + { + "epoch": 0.05604906937394247, + "grad_norm": 8.4375, + "learning_rate": 1.867512332628612e-06, + "loss": 2.8655, + "step": 530 + }, + { + "epoch": 0.05615482233502538, + "grad_norm": 8.875, + "learning_rate": 1.871035940803383e-06, + "loss": 2.6478, + "step": 531 + }, + { + "epoch": 0.05626057529610829, + "grad_norm": 8.9375, + "learning_rate": 1.8745595489781537e-06, + "loss": 2.9156, + "step": 532 + }, + { + "epoch": 0.0563663282571912, + "grad_norm": 8.6875, + "learning_rate": 1.8780831571529249e-06, + "loss": 2.9467, + "step": 533 + }, + { + "epoch": 0.056472081218274114, + "grad_norm": 8.875, + "learning_rate": 1.8816067653276956e-06, + "loss": 2.9773, + "step": 534 + }, + { + "epoch": 0.05657783417935702, + "grad_norm": 9.125, + "learning_rate": 1.8851303735024668e-06, + "loss": 3.0491, + "step": 535 + }, + { + "epoch": 0.05668358714043993, + "grad_norm": 9.125, + "learning_rate": 1.8886539816772376e-06, + "loss": 2.7805, + "step": 536 + }, + { + "epoch": 0.05678934010152284, + "grad_norm": 8.5625, + "learning_rate": 1.8921775898520086e-06, + "loss": 2.8048, + "step": 537 + }, + { + "epoch": 0.056895093062605755, + "grad_norm": 8.25, + "learning_rate": 1.8957011980267794e-06, + "loss": 2.8941, + "step": 538 + }, + { + "epoch": 0.05700084602368866, + "grad_norm": 8.375, + "learning_rate": 1.8992248062015506e-06, + "loss": 2.9099, + "step": 539 + }, + { + "epoch": 0.05710659898477157, + "grad_norm": 8.3125, + "learning_rate": 1.9027484143763214e-06, + "loss": 2.8036, + "step": 540 + }, + { + "epoch": 0.05721235194585449, + "grad_norm": 9.5625, + "learning_rate": 1.9062720225510926e-06, + "loss": 2.84, + "step": 541 + }, + { + "epoch": 0.057318104906937395, + "grad_norm": 8.4375, + "learning_rate": 1.909795630725863e-06, + "loss": 2.7414, + "step": 542 + }, + { + "epoch": 0.0574238578680203, + "grad_norm": 8.625, + "learning_rate": 1.9133192389006345e-06, + "loss": 2.7904, + "step": 543 + }, + { + "epoch": 0.05752961082910321, + "grad_norm": 9.375, + "learning_rate": 1.916842847075405e-06, + "loss": 3.2143, + "step": 544 + }, + { + "epoch": 0.05763536379018613, + "grad_norm": 7.78125, + "learning_rate": 1.9203664552501765e-06, + "loss": 2.758, + "step": 545 + }, + { + "epoch": 0.057741116751269035, + "grad_norm": 7.875, + "learning_rate": 1.9238900634249475e-06, + "loss": 2.9333, + "step": 546 + }, + { + "epoch": 0.057846869712351943, + "grad_norm": 8.1875, + "learning_rate": 1.927413671599718e-06, + "loss": 2.7104, + "step": 547 + }, + { + "epoch": 0.05795262267343486, + "grad_norm": 8.4375, + "learning_rate": 1.9309372797744895e-06, + "loss": 2.8744, + "step": 548 + }, + { + "epoch": 0.05805837563451777, + "grad_norm": 8.125, + "learning_rate": 1.93446088794926e-06, + "loss": 2.9468, + "step": 549 + }, + { + "epoch": 0.058164128595600675, + "grad_norm": 9.6875, + "learning_rate": 1.9379844961240315e-06, + "loss": 2.8179, + "step": 550 + }, + { + "epoch": 0.058269881556683584, + "grad_norm": 8.0625, + "learning_rate": 1.941508104298802e-06, + "loss": 2.8298, + "step": 551 + }, + { + "epoch": 0.0583756345177665, + "grad_norm": 8.4375, + "learning_rate": 1.945031712473573e-06, + "loss": 2.6977, + "step": 552 + }, + { + "epoch": 0.05848138747884941, + "grad_norm": 9.5625, + "learning_rate": 1.948555320648344e-06, + "loss": 2.8417, + "step": 553 + }, + { + "epoch": 0.058587140439932316, + "grad_norm": 8.5625, + "learning_rate": 1.952078928823115e-06, + "loss": 2.9777, + "step": 554 + }, + { + "epoch": 0.05869289340101523, + "grad_norm": 8.3125, + "learning_rate": 1.955602536997886e-06, + "loss": 2.8297, + "step": 555 + }, + { + "epoch": 0.05879864636209814, + "grad_norm": 8.5, + "learning_rate": 1.959126145172657e-06, + "loss": 2.9486, + "step": 556 + }, + { + "epoch": 0.05890439932318105, + "grad_norm": 8.375, + "learning_rate": 1.962649753347428e-06, + "loss": 2.8211, + "step": 557 + }, + { + "epoch": 0.05901015228426396, + "grad_norm": 7.96875, + "learning_rate": 1.966173361522199e-06, + "loss": 2.681, + "step": 558 + }, + { + "epoch": 0.05911590524534687, + "grad_norm": 8.1875, + "learning_rate": 1.96969696969697e-06, + "loss": 2.7128, + "step": 559 + }, + { + "epoch": 0.05922165820642978, + "grad_norm": 9.9375, + "learning_rate": 1.973220577871741e-06, + "loss": 3.1161, + "step": 560 + }, + { + "epoch": 0.05932741116751269, + "grad_norm": 8.25, + "learning_rate": 1.976744186046512e-06, + "loss": 2.8341, + "step": 561 + }, + { + "epoch": 0.0594331641285956, + "grad_norm": 8.5, + "learning_rate": 1.980267794221283e-06, + "loss": 2.7904, + "step": 562 + }, + { + "epoch": 0.05953891708967851, + "grad_norm": 8.0, + "learning_rate": 1.9837914023960535e-06, + "loss": 2.902, + "step": 563 + }, + { + "epoch": 0.05964467005076142, + "grad_norm": 8.375, + "learning_rate": 1.987315010570825e-06, + "loss": 2.7812, + "step": 564 + }, + { + "epoch": 0.059750423011844335, + "grad_norm": 9.375, + "learning_rate": 1.9908386187455954e-06, + "loss": 2.8535, + "step": 565 + }, + { + "epoch": 0.05985617597292724, + "grad_norm": 13.1875, + "learning_rate": 1.994362226920367e-06, + "loss": 2.9303, + "step": 566 + }, + { + "epoch": 0.05996192893401015, + "grad_norm": 8.3125, + "learning_rate": 1.9978858350951374e-06, + "loss": 2.9548, + "step": 567 + }, + { + "epoch": 0.06006768189509306, + "grad_norm": 8.6875, + "learning_rate": 2.0014094432699084e-06, + "loss": 2.6878, + "step": 568 + }, + { + "epoch": 0.060173434856175975, + "grad_norm": 8.375, + "learning_rate": 2.00493305144468e-06, + "loss": 2.8796, + "step": 569 + }, + { + "epoch": 0.060279187817258884, + "grad_norm": 8.5625, + "learning_rate": 2.0084566596194504e-06, + "loss": 3.1336, + "step": 570 + }, + { + "epoch": 0.06038494077834179, + "grad_norm": 8.625, + "learning_rate": 2.0119802677942214e-06, + "loss": 2.8421, + "step": 571 + }, + { + "epoch": 0.06049069373942471, + "grad_norm": 8.5, + "learning_rate": 2.0155038759689923e-06, + "loss": 2.7373, + "step": 572 + }, + { + "epoch": 0.060596446700507615, + "grad_norm": 8.25, + "learning_rate": 2.0190274841437633e-06, + "loss": 2.7465, + "step": 573 + }, + { + "epoch": 0.060702199661590524, + "grad_norm": 8.625, + "learning_rate": 2.0225510923185343e-06, + "loss": 2.9216, + "step": 574 + }, + { + "epoch": 0.06080795262267343, + "grad_norm": 8.25, + "learning_rate": 2.0260747004933053e-06, + "loss": 2.8131, + "step": 575 + }, + { + "epoch": 0.06091370558375635, + "grad_norm": 9.375, + "learning_rate": 2.0295983086680763e-06, + "loss": 2.7528, + "step": 576 + }, + { + "epoch": 0.061019458544839256, + "grad_norm": 8.75, + "learning_rate": 2.0331219168428473e-06, + "loss": 2.8186, + "step": 577 + }, + { + "epoch": 0.061125211505922164, + "grad_norm": 8.5625, + "learning_rate": 2.0366455250176183e-06, + "loss": 2.8535, + "step": 578 + }, + { + "epoch": 0.06123096446700508, + "grad_norm": 8.375, + "learning_rate": 2.0401691331923893e-06, + "loss": 2.9135, + "step": 579 + }, + { + "epoch": 0.06133671742808799, + "grad_norm": 8.625, + "learning_rate": 2.0436927413671602e-06, + "loss": 2.9106, + "step": 580 + }, + { + "epoch": 0.061442470389170896, + "grad_norm": 8.5625, + "learning_rate": 2.0472163495419312e-06, + "loss": 2.8578, + "step": 581 + }, + { + "epoch": 0.061548223350253804, + "grad_norm": 8.0625, + "learning_rate": 2.050739957716702e-06, + "loss": 2.7345, + "step": 582 + }, + { + "epoch": 0.06165397631133672, + "grad_norm": 8.125, + "learning_rate": 2.054263565891473e-06, + "loss": 2.8763, + "step": 583 + }, + { + "epoch": 0.06175972927241963, + "grad_norm": 8.625, + "learning_rate": 2.0577871740662438e-06, + "loss": 2.8948, + "step": 584 + }, + { + "epoch": 0.061865482233502536, + "grad_norm": 8.4375, + "learning_rate": 2.061310782241015e-06, + "loss": 2.9429, + "step": 585 + }, + { + "epoch": 0.06197123519458545, + "grad_norm": 8.9375, + "learning_rate": 2.0648343904157857e-06, + "loss": 3.1011, + "step": 586 + }, + { + "epoch": 0.06207698815566836, + "grad_norm": 8.5625, + "learning_rate": 2.0683579985905567e-06, + "loss": 2.855, + "step": 587 + }, + { + "epoch": 0.06218274111675127, + "grad_norm": 8.625, + "learning_rate": 2.0718816067653277e-06, + "loss": 2.9079, + "step": 588 + }, + { + "epoch": 0.062288494077834176, + "grad_norm": 8.0625, + "learning_rate": 2.0754052149400987e-06, + "loss": 2.8778, + "step": 589 + }, + { + "epoch": 0.06239424703891709, + "grad_norm": 8.1875, + "learning_rate": 2.07892882311487e-06, + "loss": 2.9059, + "step": 590 + }, + { + "epoch": 0.0625, + "grad_norm": 8.3125, + "learning_rate": 2.0824524312896407e-06, + "loss": 2.8873, + "step": 591 + }, + { + "epoch": 0.06260575296108291, + "grad_norm": 8.25, + "learning_rate": 2.0859760394644117e-06, + "loss": 2.8984, + "step": 592 + }, + { + "epoch": 0.06271150592216582, + "grad_norm": 8.875, + "learning_rate": 2.0894996476391827e-06, + "loss": 2.9381, + "step": 593 + }, + { + "epoch": 0.06281725888324872, + "grad_norm": 8.25, + "learning_rate": 2.0930232558139536e-06, + "loss": 2.8617, + "step": 594 + }, + { + "epoch": 0.06292301184433165, + "grad_norm": 8.5, + "learning_rate": 2.0965468639887246e-06, + "loss": 2.3662, + "step": 595 + }, + { + "epoch": 0.06302876480541456, + "grad_norm": 8.25, + "learning_rate": 2.1000704721634956e-06, + "loss": 2.7942, + "step": 596 + }, + { + "epoch": 0.06313451776649746, + "grad_norm": 8.0, + "learning_rate": 2.1035940803382666e-06, + "loss": 2.6511, + "step": 597 + }, + { + "epoch": 0.06324027072758037, + "grad_norm": 7.90625, + "learning_rate": 2.1071176885130376e-06, + "loss": 2.6217, + "step": 598 + }, + { + "epoch": 0.06334602368866328, + "grad_norm": 8.8125, + "learning_rate": 2.1106412966878086e-06, + "loss": 2.775, + "step": 599 + }, + { + "epoch": 0.06345177664974619, + "grad_norm": 8.25, + "learning_rate": 2.1141649048625796e-06, + "loss": 3.0492, + "step": 600 + }, + { + "epoch": 0.0635575296108291, + "grad_norm": 7.8125, + "learning_rate": 2.11768851303735e-06, + "loss": 2.7044, + "step": 601 + }, + { + "epoch": 0.06366328257191202, + "grad_norm": 8.4375, + "learning_rate": 2.1212121212121216e-06, + "loss": 2.6018, + "step": 602 + }, + { + "epoch": 0.06376903553299493, + "grad_norm": 8.375, + "learning_rate": 2.124735729386892e-06, + "loss": 2.8451, + "step": 603 + }, + { + "epoch": 0.06387478849407784, + "grad_norm": 9.5625, + "learning_rate": 2.1282593375616635e-06, + "loss": 3.2038, + "step": 604 + }, + { + "epoch": 0.06398054145516074, + "grad_norm": 8.875, + "learning_rate": 2.131782945736434e-06, + "loss": 2.9647, + "step": 605 + }, + { + "epoch": 0.06408629441624365, + "grad_norm": 7.9375, + "learning_rate": 2.135306553911205e-06, + "loss": 2.6134, + "step": 606 + }, + { + "epoch": 0.06419204737732656, + "grad_norm": 7.8125, + "learning_rate": 2.138830162085976e-06, + "loss": 2.9617, + "step": 607 + }, + { + "epoch": 0.06429780033840947, + "grad_norm": 8.375, + "learning_rate": 2.142353770260747e-06, + "loss": 2.7116, + "step": 608 + }, + { + "epoch": 0.06440355329949239, + "grad_norm": 7.9375, + "learning_rate": 2.145877378435518e-06, + "loss": 2.7686, + "step": 609 + }, + { + "epoch": 0.0645093062605753, + "grad_norm": 8.4375, + "learning_rate": 2.149400986610289e-06, + "loss": 2.8761, + "step": 610 + }, + { + "epoch": 0.06461505922165821, + "grad_norm": 8.625, + "learning_rate": 2.15292459478506e-06, + "loss": 2.8118, + "step": 611 + }, + { + "epoch": 0.06472081218274112, + "grad_norm": 8.5625, + "learning_rate": 2.156448202959831e-06, + "loss": 2.9027, + "step": 612 + }, + { + "epoch": 0.06482656514382402, + "grad_norm": 8.5625, + "learning_rate": 2.159971811134602e-06, + "loss": 2.7531, + "step": 613 + }, + { + "epoch": 0.06493231810490693, + "grad_norm": 8.5625, + "learning_rate": 2.163495419309373e-06, + "loss": 3.0356, + "step": 614 + }, + { + "epoch": 0.06503807106598984, + "grad_norm": 8.4375, + "learning_rate": 2.167019027484144e-06, + "loss": 3.0117, + "step": 615 + }, + { + "epoch": 0.06514382402707276, + "grad_norm": 8.125, + "learning_rate": 2.170542635658915e-06, + "loss": 2.8803, + "step": 616 + }, + { + "epoch": 0.06524957698815567, + "grad_norm": 8.3125, + "learning_rate": 2.174066243833686e-06, + "loss": 2.7699, + "step": 617 + }, + { + "epoch": 0.06535532994923858, + "grad_norm": 9.875, + "learning_rate": 2.177589852008457e-06, + "loss": 2.8248, + "step": 618 + }, + { + "epoch": 0.06546108291032149, + "grad_norm": 8.3125, + "learning_rate": 2.181113460183228e-06, + "loss": 2.7645, + "step": 619 + }, + { + "epoch": 0.0655668358714044, + "grad_norm": 8.5, + "learning_rate": 2.184637068357999e-06, + "loss": 2.8403, + "step": 620 + }, + { + "epoch": 0.0656725888324873, + "grad_norm": 7.8125, + "learning_rate": 2.18816067653277e-06, + "loss": 2.7322, + "step": 621 + }, + { + "epoch": 0.06577834179357021, + "grad_norm": 7.96875, + "learning_rate": 2.1916842847075405e-06, + "loss": 2.8199, + "step": 622 + }, + { + "epoch": 0.06588409475465314, + "grad_norm": 7.84375, + "learning_rate": 2.195207892882312e-06, + "loss": 2.5899, + "step": 623 + }, + { + "epoch": 0.06598984771573604, + "grad_norm": 8.4375, + "learning_rate": 2.1987315010570824e-06, + "loss": 3.0329, + "step": 624 + }, + { + "epoch": 0.06609560067681895, + "grad_norm": 8.875, + "learning_rate": 2.202255109231854e-06, + "loss": 3.0405, + "step": 625 + }, + { + "epoch": 0.06620135363790186, + "grad_norm": 8.0, + "learning_rate": 2.2057787174066244e-06, + "loss": 2.8867, + "step": 626 + }, + { + "epoch": 0.06630710659898477, + "grad_norm": 8.75, + "learning_rate": 2.2093023255813954e-06, + "loss": 2.7017, + "step": 627 + }, + { + "epoch": 0.06641285956006768, + "grad_norm": 7.53125, + "learning_rate": 2.2128259337561664e-06, + "loss": 2.6407, + "step": 628 + }, + { + "epoch": 0.06651861252115059, + "grad_norm": 9.0625, + "learning_rate": 2.2163495419309374e-06, + "loss": 3.0091, + "step": 629 + }, + { + "epoch": 0.06662436548223351, + "grad_norm": 8.375, + "learning_rate": 2.2198731501057084e-06, + "loss": 2.9184, + "step": 630 + }, + { + "epoch": 0.06673011844331642, + "grad_norm": 9.9375, + "learning_rate": 2.2233967582804794e-06, + "loss": 3.0185, + "step": 631 + }, + { + "epoch": 0.06683587140439932, + "grad_norm": 8.5, + "learning_rate": 2.2269203664552503e-06, + "loss": 2.7773, + "step": 632 + }, + { + "epoch": 0.06694162436548223, + "grad_norm": 7.6875, + "learning_rate": 2.2304439746300213e-06, + "loss": 2.8724, + "step": 633 + }, + { + "epoch": 0.06704737732656514, + "grad_norm": 8.375, + "learning_rate": 2.2339675828047923e-06, + "loss": 2.736, + "step": 634 + }, + { + "epoch": 0.06715313028764805, + "grad_norm": 7.9375, + "learning_rate": 2.2374911909795633e-06, + "loss": 2.757, + "step": 635 + }, + { + "epoch": 0.06725888324873096, + "grad_norm": 8.0, + "learning_rate": 2.2410147991543343e-06, + "loss": 2.9643, + "step": 636 + }, + { + "epoch": 0.06736463620981388, + "grad_norm": 8.0625, + "learning_rate": 2.2445384073291053e-06, + "loss": 2.8939, + "step": 637 + }, + { + "epoch": 0.06747038917089679, + "grad_norm": 9.1875, + "learning_rate": 2.2480620155038763e-06, + "loss": 2.8158, + "step": 638 + }, + { + "epoch": 0.0675761421319797, + "grad_norm": 9.3125, + "learning_rate": 2.2515856236786473e-06, + "loss": 2.7125, + "step": 639 + }, + { + "epoch": 0.0676818950930626, + "grad_norm": 9.5625, + "learning_rate": 2.2551092318534182e-06, + "loss": 2.7622, + "step": 640 + }, + { + "epoch": 0.06778764805414551, + "grad_norm": 7.8125, + "learning_rate": 2.258632840028189e-06, + "loss": 2.7422, + "step": 641 + }, + { + "epoch": 0.06789340101522842, + "grad_norm": 9.1875, + "learning_rate": 2.2621564482029602e-06, + "loss": 2.8343, + "step": 642 + }, + { + "epoch": 0.06799915397631133, + "grad_norm": 8.25, + "learning_rate": 2.2656800563777308e-06, + "loss": 2.8356, + "step": 643 + }, + { + "epoch": 0.06810490693739425, + "grad_norm": 9.1875, + "learning_rate": 2.269203664552502e-06, + "loss": 3.1141, + "step": 644 + }, + { + "epoch": 0.06821065989847716, + "grad_norm": 8.6875, + "learning_rate": 2.2727272727272728e-06, + "loss": 3.0477, + "step": 645 + }, + { + "epoch": 0.06831641285956007, + "grad_norm": 8.3125, + "learning_rate": 2.2762508809020437e-06, + "loss": 2.7913, + "step": 646 + }, + { + "epoch": 0.06842216582064298, + "grad_norm": 8.0625, + "learning_rate": 2.2797744890768147e-06, + "loss": 2.7181, + "step": 647 + }, + { + "epoch": 0.06852791878172589, + "grad_norm": 8.5, + "learning_rate": 2.2832980972515857e-06, + "loss": 2.8892, + "step": 648 + }, + { + "epoch": 0.0686336717428088, + "grad_norm": 7.8125, + "learning_rate": 2.2868217054263567e-06, + "loss": 2.6898, + "step": 649 + }, + { + "epoch": 0.0687394247038917, + "grad_norm": 7.75, + "learning_rate": 2.2903453136011277e-06, + "loss": 2.8404, + "step": 650 + }, + { + "epoch": 0.06884517766497462, + "grad_norm": 8.625, + "learning_rate": 2.2938689217758987e-06, + "loss": 2.6429, + "step": 651 + }, + { + "epoch": 0.06895093062605753, + "grad_norm": 9.5, + "learning_rate": 2.2973925299506697e-06, + "loss": 2.7195, + "step": 652 + }, + { + "epoch": 0.06905668358714044, + "grad_norm": 8.875, + "learning_rate": 2.3009161381254407e-06, + "loss": 2.7213, + "step": 653 + }, + { + "epoch": 0.06916243654822335, + "grad_norm": 9.1875, + "learning_rate": 2.3044397463002116e-06, + "loss": 2.7529, + "step": 654 + }, + { + "epoch": 0.06926818950930626, + "grad_norm": 9.3125, + "learning_rate": 2.3079633544749826e-06, + "loss": 2.7189, + "step": 655 + }, + { + "epoch": 0.06937394247038917, + "grad_norm": 8.625, + "learning_rate": 2.3114869626497536e-06, + "loss": 3.1161, + "step": 656 + }, + { + "epoch": 0.06947969543147207, + "grad_norm": 7.78125, + "learning_rate": 2.3150105708245246e-06, + "loss": 2.7381, + "step": 657 + }, + { + "epoch": 0.069585448392555, + "grad_norm": 8.5, + "learning_rate": 2.3185341789992956e-06, + "loss": 2.8175, + "step": 658 + }, + { + "epoch": 0.0696912013536379, + "grad_norm": 8.0625, + "learning_rate": 2.3220577871740666e-06, + "loss": 2.7607, + "step": 659 + }, + { + "epoch": 0.06979695431472081, + "grad_norm": 8.125, + "learning_rate": 2.3255813953488376e-06, + "loss": 2.7253, + "step": 660 + }, + { + "epoch": 0.06990270727580372, + "grad_norm": 8.8125, + "learning_rate": 2.3291050035236086e-06, + "loss": 2.9595, + "step": 661 + }, + { + "epoch": 0.07000846023688663, + "grad_norm": 7.9375, + "learning_rate": 2.332628611698379e-06, + "loss": 2.8311, + "step": 662 + }, + { + "epoch": 0.07011421319796954, + "grad_norm": 8.875, + "learning_rate": 2.3361522198731505e-06, + "loss": 2.9719, + "step": 663 + }, + { + "epoch": 0.07021996615905245, + "grad_norm": 8.0, + "learning_rate": 2.339675828047921e-06, + "loss": 2.7527, + "step": 664 + }, + { + "epoch": 0.07032571912013537, + "grad_norm": 8.875, + "learning_rate": 2.343199436222692e-06, + "loss": 2.9043, + "step": 665 + }, + { + "epoch": 0.07043147208121828, + "grad_norm": 8.0625, + "learning_rate": 2.346723044397463e-06, + "loss": 2.8043, + "step": 666 + }, + { + "epoch": 0.07053722504230119, + "grad_norm": 8.75, + "learning_rate": 2.350246652572234e-06, + "loss": 2.4113, + "step": 667 + }, + { + "epoch": 0.0706429780033841, + "grad_norm": 9.0, + "learning_rate": 2.353770260747005e-06, + "loss": 2.9967, + "step": 668 + }, + { + "epoch": 0.070748730964467, + "grad_norm": 9.4375, + "learning_rate": 2.357293868921776e-06, + "loss": 2.6786, + "step": 669 + }, + { + "epoch": 0.07085448392554991, + "grad_norm": 7.9375, + "learning_rate": 2.360817477096547e-06, + "loss": 2.6808, + "step": 670 + }, + { + "epoch": 0.07096023688663283, + "grad_norm": 8.25, + "learning_rate": 2.364341085271318e-06, + "loss": 2.9015, + "step": 671 + }, + { + "epoch": 0.07106598984771574, + "grad_norm": 8.5625, + "learning_rate": 2.367864693446089e-06, + "loss": 2.9672, + "step": 672 + }, + { + "epoch": 0.07117174280879865, + "grad_norm": 8.625, + "learning_rate": 2.37138830162086e-06, + "loss": 2.9305, + "step": 673 + }, + { + "epoch": 0.07127749576988156, + "grad_norm": 8.3125, + "learning_rate": 2.374911909795631e-06, + "loss": 2.999, + "step": 674 + }, + { + "epoch": 0.07138324873096447, + "grad_norm": 7.71875, + "learning_rate": 2.378435517970402e-06, + "loss": 2.7163, + "step": 675 + }, + { + "epoch": 0.07148900169204737, + "grad_norm": 8.375, + "learning_rate": 2.3819591261451725e-06, + "loss": 2.6723, + "step": 676 + }, + { + "epoch": 0.07159475465313028, + "grad_norm": 7.46875, + "learning_rate": 2.385482734319944e-06, + "loss": 2.8568, + "step": 677 + }, + { + "epoch": 0.0717005076142132, + "grad_norm": 7.78125, + "learning_rate": 2.389006342494715e-06, + "loss": 2.5685, + "step": 678 + }, + { + "epoch": 0.07180626057529611, + "grad_norm": 8.3125, + "learning_rate": 2.392529950669486e-06, + "loss": 2.7444, + "step": 679 + }, + { + "epoch": 0.07191201353637902, + "grad_norm": 8.0625, + "learning_rate": 2.396053558844257e-06, + "loss": 2.7316, + "step": 680 + }, + { + "epoch": 0.07201776649746193, + "grad_norm": 8.375, + "learning_rate": 2.3995771670190275e-06, + "loss": 2.9459, + "step": 681 + }, + { + "epoch": 0.07212351945854484, + "grad_norm": 9.25, + "learning_rate": 2.403100775193799e-06, + "loss": 2.6203, + "step": 682 + }, + { + "epoch": 0.07222927241962775, + "grad_norm": 8.3125, + "learning_rate": 2.4066243833685695e-06, + "loss": 2.8151, + "step": 683 + }, + { + "epoch": 0.07233502538071065, + "grad_norm": 8.4375, + "learning_rate": 2.410147991543341e-06, + "loss": 2.7918, + "step": 684 + }, + { + "epoch": 0.07244077834179358, + "grad_norm": 8.25, + "learning_rate": 2.4136715997181114e-06, + "loss": 2.7216, + "step": 685 + }, + { + "epoch": 0.07254653130287649, + "grad_norm": 9.4375, + "learning_rate": 2.4171952078928824e-06, + "loss": 2.9787, + "step": 686 + }, + { + "epoch": 0.0726522842639594, + "grad_norm": 8.0625, + "learning_rate": 2.4207188160676534e-06, + "loss": 2.8487, + "step": 687 + }, + { + "epoch": 0.0727580372250423, + "grad_norm": 7.9375, + "learning_rate": 2.4242424242424244e-06, + "loss": 2.9151, + "step": 688 + }, + { + "epoch": 0.07286379018612521, + "grad_norm": 7.9375, + "learning_rate": 2.4277660324171954e-06, + "loss": 2.955, + "step": 689 + }, + { + "epoch": 0.07296954314720812, + "grad_norm": 8.125, + "learning_rate": 2.4312896405919664e-06, + "loss": 2.8102, + "step": 690 + }, + { + "epoch": 0.07307529610829103, + "grad_norm": 7.71875, + "learning_rate": 2.4348132487667374e-06, + "loss": 2.8267, + "step": 691 + }, + { + "epoch": 0.07318104906937395, + "grad_norm": 7.84375, + "learning_rate": 2.4383368569415083e-06, + "loss": 2.7964, + "step": 692 + }, + { + "epoch": 0.07328680203045686, + "grad_norm": 9.4375, + "learning_rate": 2.4418604651162793e-06, + "loss": 3.1879, + "step": 693 + }, + { + "epoch": 0.07339255499153977, + "grad_norm": 8.4375, + "learning_rate": 2.4453840732910503e-06, + "loss": 2.6342, + "step": 694 + }, + { + "epoch": 0.07349830795262267, + "grad_norm": 8.0, + "learning_rate": 2.448907681465821e-06, + "loss": 2.735, + "step": 695 + }, + { + "epoch": 0.07360406091370558, + "grad_norm": 7.34375, + "learning_rate": 2.4524312896405923e-06, + "loss": 2.6143, + "step": 696 + }, + { + "epoch": 0.07370981387478849, + "grad_norm": 7.9375, + "learning_rate": 2.455954897815363e-06, + "loss": 2.7019, + "step": 697 + }, + { + "epoch": 0.0738155668358714, + "grad_norm": 7.71875, + "learning_rate": 2.4594785059901343e-06, + "loss": 2.7462, + "step": 698 + }, + { + "epoch": 0.07392131979695432, + "grad_norm": 8.8125, + "learning_rate": 2.4630021141649053e-06, + "loss": 2.7576, + "step": 699 + }, + { + "epoch": 0.07402707275803723, + "grad_norm": 7.59375, + "learning_rate": 2.466525722339676e-06, + "loss": 2.7452, + "step": 700 + }, + { + "epoch": 0.07413282571912014, + "grad_norm": 7.90625, + "learning_rate": 2.4700493305144472e-06, + "loss": 2.6232, + "step": 701 + }, + { + "epoch": 0.07423857868020305, + "grad_norm": 8.0625, + "learning_rate": 2.473572938689218e-06, + "loss": 2.7871, + "step": 702 + }, + { + "epoch": 0.07434433164128595, + "grad_norm": 10.625, + "learning_rate": 2.477096546863989e-06, + "loss": 2.7983, + "step": 703 + }, + { + "epoch": 0.07445008460236886, + "grad_norm": 7.21875, + "learning_rate": 2.4806201550387598e-06, + "loss": 2.7883, + "step": 704 + }, + { + "epoch": 0.07455583756345177, + "grad_norm": 8.9375, + "learning_rate": 2.4841437632135308e-06, + "loss": 2.9282, + "step": 705 + }, + { + "epoch": 0.0746615905245347, + "grad_norm": 8.375, + "learning_rate": 2.4876673713883017e-06, + "loss": 2.6707, + "step": 706 + }, + { + "epoch": 0.0747673434856176, + "grad_norm": 9.5, + "learning_rate": 2.4911909795630727e-06, + "loss": 2.7633, + "step": 707 + }, + { + "epoch": 0.07487309644670051, + "grad_norm": 9.125, + "learning_rate": 2.4947145877378437e-06, + "loss": 2.7473, + "step": 708 + }, + { + "epoch": 0.07497884940778342, + "grad_norm": 9.25, + "learning_rate": 2.4982381959126147e-06, + "loss": 2.9258, + "step": 709 + }, + { + "epoch": 0.07508460236886633, + "grad_norm": 8.625, + "learning_rate": 2.5017618040873857e-06, + "loss": 2.861, + "step": 710 + }, + { + "epoch": 0.07519035532994923, + "grad_norm": 8.3125, + "learning_rate": 2.5052854122621567e-06, + "loss": 2.9084, + "step": 711 + }, + { + "epoch": 0.07529610829103214, + "grad_norm": 7.78125, + "learning_rate": 2.5088090204369277e-06, + "loss": 2.8743, + "step": 712 + }, + { + "epoch": 0.07540186125211507, + "grad_norm": 8.375, + "learning_rate": 2.5123326286116982e-06, + "loss": 2.7711, + "step": 713 + }, + { + "epoch": 0.07550761421319797, + "grad_norm": 9.8125, + "learning_rate": 2.5158562367864696e-06, + "loss": 2.7809, + "step": 714 + }, + { + "epoch": 0.07561336717428088, + "grad_norm": 9.0, + "learning_rate": 2.5193798449612406e-06, + "loss": 2.7893, + "step": 715 + }, + { + "epoch": 0.07571912013536379, + "grad_norm": 8.0625, + "learning_rate": 2.522903453136011e-06, + "loss": 2.8789, + "step": 716 + }, + { + "epoch": 0.0758248730964467, + "grad_norm": 8.0, + "learning_rate": 2.526427061310782e-06, + "loss": 2.8319, + "step": 717 + }, + { + "epoch": 0.07593062605752961, + "grad_norm": 7.6875, + "learning_rate": 2.5299506694855536e-06, + "loss": 2.8605, + "step": 718 + }, + { + "epoch": 0.07603637901861252, + "grad_norm": 7.84375, + "learning_rate": 2.5334742776603246e-06, + "loss": 2.927, + "step": 719 + }, + { + "epoch": 0.07614213197969544, + "grad_norm": 9.1875, + "learning_rate": 2.536997885835095e-06, + "loss": 2.691, + "step": 720 + }, + { + "epoch": 0.07624788494077835, + "grad_norm": 9.1875, + "learning_rate": 2.5405214940098666e-06, + "loss": 2.9975, + "step": 721 + }, + { + "epoch": 0.07635363790186125, + "grad_norm": 8.8125, + "learning_rate": 2.5440451021846376e-06, + "loss": 2.8906, + "step": 722 + }, + { + "epoch": 0.07645939086294416, + "grad_norm": 7.6875, + "learning_rate": 2.547568710359408e-06, + "loss": 2.8544, + "step": 723 + }, + { + "epoch": 0.07656514382402707, + "grad_norm": 9.4375, + "learning_rate": 2.551092318534179e-06, + "loss": 2.6734, + "step": 724 + }, + { + "epoch": 0.07667089678510998, + "grad_norm": 9.125, + "learning_rate": 2.5546159267089505e-06, + "loss": 2.9802, + "step": 725 + }, + { + "epoch": 0.07677664974619289, + "grad_norm": 9.0625, + "learning_rate": 2.558139534883721e-06, + "loss": 2.6894, + "step": 726 + }, + { + "epoch": 0.07688240270727581, + "grad_norm": 8.25, + "learning_rate": 2.561663143058492e-06, + "loss": 2.8441, + "step": 727 + }, + { + "epoch": 0.07698815566835872, + "grad_norm": 7.90625, + "learning_rate": 2.565186751233263e-06, + "loss": 2.7109, + "step": 728 + }, + { + "epoch": 0.07709390862944163, + "grad_norm": 7.34375, + "learning_rate": 2.5687103594080345e-06, + "loss": 2.6921, + "step": 729 + }, + { + "epoch": 0.07719966159052453, + "grad_norm": 9.875, + "learning_rate": 2.572233967582805e-06, + "loss": 2.734, + "step": 730 + }, + { + "epoch": 0.07730541455160744, + "grad_norm": 7.84375, + "learning_rate": 2.575757575757576e-06, + "loss": 2.631, + "step": 731 + }, + { + "epoch": 0.07741116751269035, + "grad_norm": 8.6875, + "learning_rate": 2.5792811839323466e-06, + "loss": 2.8275, + "step": 732 + }, + { + "epoch": 0.07751692047377326, + "grad_norm": 8.75, + "learning_rate": 2.582804792107118e-06, + "loss": 3.2445, + "step": 733 + }, + { + "epoch": 0.07762267343485618, + "grad_norm": 8.0625, + "learning_rate": 2.586328400281889e-06, + "loss": 2.7603, + "step": 734 + }, + { + "epoch": 0.07772842639593909, + "grad_norm": 7.75, + "learning_rate": 2.5898520084566595e-06, + "loss": 2.7474, + "step": 735 + }, + { + "epoch": 0.077834179357022, + "grad_norm": 8.5, + "learning_rate": 2.5933756166314305e-06, + "loss": 2.9152, + "step": 736 + }, + { + "epoch": 0.07793993231810491, + "grad_norm": 8.1875, + "learning_rate": 2.596899224806202e-06, + "loss": 2.7586, + "step": 737 + }, + { + "epoch": 0.07804568527918782, + "grad_norm": 9.625, + "learning_rate": 2.600422832980973e-06, + "loss": 2.8464, + "step": 738 + }, + { + "epoch": 0.07815143824027072, + "grad_norm": 7.75, + "learning_rate": 2.6039464411557435e-06, + "loss": 2.7477, + "step": 739 + }, + { + "epoch": 0.07825719120135363, + "grad_norm": 8.3125, + "learning_rate": 2.6074700493305145e-06, + "loss": 2.9824, + "step": 740 + }, + { + "epoch": 0.07836294416243655, + "grad_norm": 8.125, + "learning_rate": 2.610993657505286e-06, + "loss": 2.7645, + "step": 741 + }, + { + "epoch": 0.07846869712351946, + "grad_norm": 8.1875, + "learning_rate": 2.6145172656800565e-06, + "loss": 2.77, + "step": 742 + }, + { + "epoch": 0.07857445008460237, + "grad_norm": 8.5625, + "learning_rate": 2.6180408738548275e-06, + "loss": 2.7721, + "step": 743 + }, + { + "epoch": 0.07868020304568528, + "grad_norm": 8.25, + "learning_rate": 2.621564482029599e-06, + "loss": 2.6418, + "step": 744 + }, + { + "epoch": 0.07878595600676819, + "grad_norm": 8.6875, + "learning_rate": 2.6250880902043694e-06, + "loss": 2.6398, + "step": 745 + }, + { + "epoch": 0.0788917089678511, + "grad_norm": 7.5, + "learning_rate": 2.6286116983791404e-06, + "loss": 2.5761, + "step": 746 + }, + { + "epoch": 0.078997461928934, + "grad_norm": 8.9375, + "learning_rate": 2.6321353065539114e-06, + "loss": 2.7132, + "step": 747 + }, + { + "epoch": 0.07910321489001693, + "grad_norm": 8.4375, + "learning_rate": 2.635658914728683e-06, + "loss": 2.6916, + "step": 748 + }, + { + "epoch": 0.07920896785109983, + "grad_norm": 7.8125, + "learning_rate": 2.6391825229034534e-06, + "loss": 2.8894, + "step": 749 + }, + { + "epoch": 0.07931472081218274, + "grad_norm": 8.6875, + "learning_rate": 2.6427061310782244e-06, + "loss": 2.9742, + "step": 750 + }, + { + "epoch": 0.07942047377326565, + "grad_norm": 10.5625, + "learning_rate": 2.646229739252995e-06, + "loss": 2.4034, + "step": 751 + }, + { + "epoch": 0.07952622673434856, + "grad_norm": 8.125, + "learning_rate": 2.6497533474277663e-06, + "loss": 2.7003, + "step": 752 + }, + { + "epoch": 0.07963197969543147, + "grad_norm": 7.84375, + "learning_rate": 2.6532769556025373e-06, + "loss": 2.6671, + "step": 753 + }, + { + "epoch": 0.07973773265651438, + "grad_norm": 8.0625, + "learning_rate": 2.6568005637773083e-06, + "loss": 3.024, + "step": 754 + }, + { + "epoch": 0.0798434856175973, + "grad_norm": 8.25, + "learning_rate": 2.660324171952079e-06, + "loss": 2.691, + "step": 755 + }, + { + "epoch": 0.0799492385786802, + "grad_norm": 8.375, + "learning_rate": 2.6638477801268503e-06, + "loss": 2.8093, + "step": 756 + }, + { + "epoch": 0.08005499153976312, + "grad_norm": 8.5625, + "learning_rate": 2.6673713883016213e-06, + "loss": 3.0313, + "step": 757 + }, + { + "epoch": 0.08016074450084602, + "grad_norm": 8.1875, + "learning_rate": 2.670894996476392e-06, + "loss": 2.9489, + "step": 758 + }, + { + "epoch": 0.08026649746192893, + "grad_norm": 7.5625, + "learning_rate": 2.674418604651163e-06, + "loss": 2.7807, + "step": 759 + }, + { + "epoch": 0.08037225042301184, + "grad_norm": 7.9375, + "learning_rate": 2.6779422128259342e-06, + "loss": 2.7754, + "step": 760 + }, + { + "epoch": 0.08047800338409475, + "grad_norm": 8.125, + "learning_rate": 2.681465821000705e-06, + "loss": 2.7404, + "step": 761 + }, + { + "epoch": 0.08058375634517767, + "grad_norm": 8.125, + "learning_rate": 2.684989429175476e-06, + "loss": 2.7128, + "step": 762 + }, + { + "epoch": 0.08068950930626058, + "grad_norm": 8.25, + "learning_rate": 2.6885130373502468e-06, + "loss": 2.8131, + "step": 763 + }, + { + "epoch": 0.08079526226734349, + "grad_norm": 8.0625, + "learning_rate": 2.6920366455250178e-06, + "loss": 2.8838, + "step": 764 + }, + { + "epoch": 0.0809010152284264, + "grad_norm": 8.0, + "learning_rate": 2.6955602536997888e-06, + "loss": 2.7924, + "step": 765 + }, + { + "epoch": 0.0810067681895093, + "grad_norm": 7.5, + "learning_rate": 2.6990838618745597e-06, + "loss": 2.872, + "step": 766 + }, + { + "epoch": 0.08111252115059221, + "grad_norm": 8.0, + "learning_rate": 2.702607470049331e-06, + "loss": 2.7604, + "step": 767 + }, + { + "epoch": 0.08121827411167512, + "grad_norm": 8.0, + "learning_rate": 2.7061310782241017e-06, + "loss": 2.7085, + "step": 768 + }, + { + "epoch": 0.08132402707275804, + "grad_norm": 8.6875, + "learning_rate": 2.7096546863988727e-06, + "loss": 2.9519, + "step": 769 + }, + { + "epoch": 0.08142978003384095, + "grad_norm": 8.5625, + "learning_rate": 2.7131782945736433e-06, + "loss": 2.7978, + "step": 770 + }, + { + "epoch": 0.08153553299492386, + "grad_norm": 9.25, + "learning_rate": 2.7167019027484147e-06, + "loss": 2.8292, + "step": 771 + }, + { + "epoch": 0.08164128595600677, + "grad_norm": 7.96875, + "learning_rate": 2.7202255109231857e-06, + "loss": 2.7296, + "step": 772 + }, + { + "epoch": 0.08174703891708968, + "grad_norm": 7.375, + "learning_rate": 2.7237491190979567e-06, + "loss": 2.7882, + "step": 773 + }, + { + "epoch": 0.08185279187817258, + "grad_norm": 7.90625, + "learning_rate": 2.7272727272727272e-06, + "loss": 2.7848, + "step": 774 + }, + { + "epoch": 0.08195854483925549, + "grad_norm": 7.84375, + "learning_rate": 2.7307963354474986e-06, + "loss": 2.6468, + "step": 775 + }, + { + "epoch": 0.08206429780033841, + "grad_norm": 8.6875, + "learning_rate": 2.7343199436222696e-06, + "loss": 2.8896, + "step": 776 + }, + { + "epoch": 0.08217005076142132, + "grad_norm": 8.5625, + "learning_rate": 2.73784355179704e-06, + "loss": 2.9049, + "step": 777 + }, + { + "epoch": 0.08227580372250423, + "grad_norm": 8.875, + "learning_rate": 2.741367159971811e-06, + "loss": 2.934, + "step": 778 + }, + { + "epoch": 0.08238155668358714, + "grad_norm": 24.125, + "learning_rate": 2.7448907681465826e-06, + "loss": 2.9239, + "step": 779 + }, + { + "epoch": 0.08248730964467005, + "grad_norm": 8.125, + "learning_rate": 2.748414376321353e-06, + "loss": 2.8982, + "step": 780 + }, + { + "epoch": 0.08259306260575296, + "grad_norm": 7.4375, + "learning_rate": 2.751937984496124e-06, + "loss": 2.6752, + "step": 781 + }, + { + "epoch": 0.08269881556683586, + "grad_norm": 8.0, + "learning_rate": 2.755461592670895e-06, + "loss": 2.9464, + "step": 782 + }, + { + "epoch": 0.08280456852791879, + "grad_norm": 7.875, + "learning_rate": 2.7589852008456665e-06, + "loss": 2.8701, + "step": 783 + }, + { + "epoch": 0.0829103214890017, + "grad_norm": 8.4375, + "learning_rate": 2.762508809020437e-06, + "loss": 2.9149, + "step": 784 + }, + { + "epoch": 0.0830160744500846, + "grad_norm": 7.9375, + "learning_rate": 2.766032417195208e-06, + "loss": 2.8296, + "step": 785 + }, + { + "epoch": 0.08312182741116751, + "grad_norm": 8.0625, + "learning_rate": 2.7695560253699795e-06, + "loss": 2.8281, + "step": 786 + }, + { + "epoch": 0.08322758037225042, + "grad_norm": 7.90625, + "learning_rate": 2.77307963354475e-06, + "loss": 2.7596, + "step": 787 + }, + { + "epoch": 0.08333333333333333, + "grad_norm": 7.84375, + "learning_rate": 2.776603241719521e-06, + "loss": 2.6853, + "step": 788 + }, + { + "epoch": 0.08343908629441624, + "grad_norm": 8.125, + "learning_rate": 2.7801268498942916e-06, + "loss": 2.7592, + "step": 789 + }, + { + "epoch": 0.08354483925549916, + "grad_norm": 8.1875, + "learning_rate": 2.783650458069063e-06, + "loss": 3.0103, + "step": 790 + }, + { + "epoch": 0.08365059221658207, + "grad_norm": 8.25, + "learning_rate": 2.787174066243834e-06, + "loss": 2.768, + "step": 791 + }, + { + "epoch": 0.08375634517766498, + "grad_norm": 8.1875, + "learning_rate": 2.790697674418605e-06, + "loss": 2.7518, + "step": 792 + }, + { + "epoch": 0.08386209813874788, + "grad_norm": 8.3125, + "learning_rate": 2.7942212825933756e-06, + "loss": 2.7637, + "step": 793 + }, + { + "epoch": 0.08396785109983079, + "grad_norm": 8.5625, + "learning_rate": 2.797744890768147e-06, + "loss": 2.9934, + "step": 794 + }, + { + "epoch": 0.0840736040609137, + "grad_norm": 8.75, + "learning_rate": 2.801268498942918e-06, + "loss": 2.6614, + "step": 795 + }, + { + "epoch": 0.08417935702199661, + "grad_norm": 8.75, + "learning_rate": 2.8047921071176885e-06, + "loss": 2.7671, + "step": 796 + }, + { + "epoch": 0.08428510998307953, + "grad_norm": 8.3125, + "learning_rate": 2.8083157152924595e-06, + "loss": 2.7505, + "step": 797 + }, + { + "epoch": 0.08439086294416244, + "grad_norm": 8.375, + "learning_rate": 2.811839323467231e-06, + "loss": 2.8381, + "step": 798 + }, + { + "epoch": 0.08449661590524535, + "grad_norm": 10.6875, + "learning_rate": 2.8153629316420015e-06, + "loss": 2.6302, + "step": 799 + }, + { + "epoch": 0.08460236886632826, + "grad_norm": 9.0625, + "learning_rate": 2.8188865398167725e-06, + "loss": 2.7141, + "step": 800 + }, + { + "epoch": 0.08470812182741116, + "grad_norm": 8.9375, + "learning_rate": 2.8224101479915435e-06, + "loss": 2.7801, + "step": 801 + }, + { + "epoch": 0.08481387478849407, + "grad_norm": 8.0625, + "learning_rate": 2.825933756166315e-06, + "loss": 2.7378, + "step": 802 + }, + { + "epoch": 0.08491962774957698, + "grad_norm": 8.0625, + "learning_rate": 2.8294573643410855e-06, + "loss": 2.8204, + "step": 803 + }, + { + "epoch": 0.0850253807106599, + "grad_norm": 8.5, + "learning_rate": 2.8329809725158564e-06, + "loss": 3.0395, + "step": 804 + }, + { + "epoch": 0.08513113367174281, + "grad_norm": 7.625, + "learning_rate": 2.836504580690627e-06, + "loss": 2.6985, + "step": 805 + }, + { + "epoch": 0.08523688663282572, + "grad_norm": 8.1875, + "learning_rate": 2.8400281888653984e-06, + "loss": 3.0016, + "step": 806 + }, + { + "epoch": 0.08534263959390863, + "grad_norm": 7.75, + "learning_rate": 2.8435517970401694e-06, + "loss": 2.5917, + "step": 807 + }, + { + "epoch": 0.08544839255499154, + "grad_norm": 7.96875, + "learning_rate": 2.8470754052149404e-06, + "loss": 2.6237, + "step": 808 + }, + { + "epoch": 0.08555414551607445, + "grad_norm": 8.4375, + "learning_rate": 2.8505990133897114e-06, + "loss": 2.751, + "step": 809 + }, + { + "epoch": 0.08565989847715735, + "grad_norm": 8.4375, + "learning_rate": 2.8541226215644824e-06, + "loss": 2.8495, + "step": 810 + }, + { + "epoch": 0.08576565143824028, + "grad_norm": 7.84375, + "learning_rate": 2.8576462297392534e-06, + "loss": 2.5675, + "step": 811 + }, + { + "epoch": 0.08587140439932318, + "grad_norm": 9.5, + "learning_rate": 2.861169837914024e-06, + "loss": 2.8038, + "step": 812 + }, + { + "epoch": 0.08597715736040609, + "grad_norm": 7.5625, + "learning_rate": 2.8646934460887953e-06, + "loss": 2.7137, + "step": 813 + }, + { + "epoch": 0.086082910321489, + "grad_norm": 8.25, + "learning_rate": 2.8682170542635663e-06, + "loss": 2.8668, + "step": 814 + }, + { + "epoch": 0.08618866328257191, + "grad_norm": 7.96875, + "learning_rate": 2.871740662438337e-06, + "loss": 2.7287, + "step": 815 + }, + { + "epoch": 0.08629441624365482, + "grad_norm": 8.5625, + "learning_rate": 2.875264270613108e-06, + "loss": 2.7365, + "step": 816 + }, + { + "epoch": 0.08640016920473774, + "grad_norm": 7.8125, + "learning_rate": 2.8787878787878793e-06, + "loss": 2.7071, + "step": 817 + }, + { + "epoch": 0.08650592216582065, + "grad_norm": 7.78125, + "learning_rate": 2.8823114869626503e-06, + "loss": 2.6485, + "step": 818 + }, + { + "epoch": 0.08661167512690356, + "grad_norm": 8.375, + "learning_rate": 2.885835095137421e-06, + "loss": 2.9005, + "step": 819 + }, + { + "epoch": 0.08671742808798646, + "grad_norm": 9.0625, + "learning_rate": 2.889358703312192e-06, + "loss": 2.7671, + "step": 820 + }, + { + "epoch": 0.08682318104906937, + "grad_norm": 8.0, + "learning_rate": 2.8928823114869632e-06, + "loss": 2.969, + "step": 821 + }, + { + "epoch": 0.08692893401015228, + "grad_norm": 8.875, + "learning_rate": 2.896405919661734e-06, + "loss": 2.4395, + "step": 822 + }, + { + "epoch": 0.08703468697123519, + "grad_norm": 7.65625, + "learning_rate": 2.8999295278365048e-06, + "loss": 2.7549, + "step": 823 + }, + { + "epoch": 0.08714043993231811, + "grad_norm": 7.90625, + "learning_rate": 2.9034531360112753e-06, + "loss": 2.7009, + "step": 824 + }, + { + "epoch": 0.08724619289340102, + "grad_norm": 8.125, + "learning_rate": 2.9069767441860468e-06, + "loss": 2.7214, + "step": 825 + }, + { + "epoch": 0.08735194585448393, + "grad_norm": 7.65625, + "learning_rate": 2.9105003523608177e-06, + "loss": 2.8015, + "step": 826 + }, + { + "epoch": 0.08745769881556684, + "grad_norm": 8.0625, + "learning_rate": 2.9140239605355887e-06, + "loss": 2.6368, + "step": 827 + }, + { + "epoch": 0.08756345177664974, + "grad_norm": 9.3125, + "learning_rate": 2.9175475687103593e-06, + "loss": 2.7248, + "step": 828 + }, + { + "epoch": 0.08766920473773265, + "grad_norm": 8.75, + "learning_rate": 2.9210711768851307e-06, + "loss": 2.729, + "step": 829 + }, + { + "epoch": 0.08777495769881556, + "grad_norm": 8.25, + "learning_rate": 2.9245947850599017e-06, + "loss": 2.6927, + "step": 830 + }, + { + "epoch": 0.08788071065989848, + "grad_norm": 8.5, + "learning_rate": 2.9281183932346723e-06, + "loss": 2.8389, + "step": 831 + }, + { + "epoch": 0.08798646362098139, + "grad_norm": 8.375, + "learning_rate": 2.9316420014094437e-06, + "loss": 2.8589, + "step": 832 + }, + { + "epoch": 0.0880922165820643, + "grad_norm": 8.0625, + "learning_rate": 2.9351656095842147e-06, + "loss": 2.682, + "step": 833 + }, + { + "epoch": 0.08819796954314721, + "grad_norm": 7.8125, + "learning_rate": 2.9386892177589852e-06, + "loss": 2.9769, + "step": 834 + }, + { + "epoch": 0.08830372250423012, + "grad_norm": 8.375, + "learning_rate": 2.9422128259337562e-06, + "loss": 2.864, + "step": 835 + }, + { + "epoch": 0.08840947546531303, + "grad_norm": 8.9375, + "learning_rate": 2.9457364341085276e-06, + "loss": 2.6968, + "step": 836 + }, + { + "epoch": 0.08851522842639593, + "grad_norm": 8.5625, + "learning_rate": 2.9492600422832986e-06, + "loss": 2.8777, + "step": 837 + }, + { + "epoch": 0.08862098138747886, + "grad_norm": 9.125, + "learning_rate": 2.952783650458069e-06, + "loss": 2.6914, + "step": 838 + }, + { + "epoch": 0.08872673434856176, + "grad_norm": 8.0625, + "learning_rate": 2.95630725863284e-06, + "loss": 2.9202, + "step": 839 + }, + { + "epoch": 0.08883248730964467, + "grad_norm": 7.78125, + "learning_rate": 2.9598308668076116e-06, + "loss": 2.8764, + "step": 840 + }, + { + "epoch": 0.08893824027072758, + "grad_norm": 8.0, + "learning_rate": 2.963354474982382e-06, + "loss": 2.8037, + "step": 841 + }, + { + "epoch": 0.08904399323181049, + "grad_norm": 8.625, + "learning_rate": 2.966878083157153e-06, + "loss": 2.5802, + "step": 842 + }, + { + "epoch": 0.0891497461928934, + "grad_norm": 8.5625, + "learning_rate": 2.970401691331924e-06, + "loss": 2.5913, + "step": 843 + }, + { + "epoch": 0.0892554991539763, + "grad_norm": 8.3125, + "learning_rate": 2.973925299506695e-06, + "loss": 2.6131, + "step": 844 + }, + { + "epoch": 0.08936125211505923, + "grad_norm": 8.0, + "learning_rate": 2.977448907681466e-06, + "loss": 2.6048, + "step": 845 + }, + { + "epoch": 0.08946700507614214, + "grad_norm": 8.125, + "learning_rate": 2.980972515856237e-06, + "loss": 2.8855, + "step": 846 + }, + { + "epoch": 0.08957275803722504, + "grad_norm": 8.5, + "learning_rate": 2.9844961240310076e-06, + "loss": 2.7397, + "step": 847 + }, + { + "epoch": 0.08967851099830795, + "grad_norm": 7.4375, + "learning_rate": 2.988019732205779e-06, + "loss": 2.7213, + "step": 848 + }, + { + "epoch": 0.08978426395939086, + "grad_norm": 8.0625, + "learning_rate": 2.99154334038055e-06, + "loss": 2.7795, + "step": 849 + }, + { + "epoch": 0.08989001692047377, + "grad_norm": 8.125, + "learning_rate": 2.9950669485553206e-06, + "loss": 2.8538, + "step": 850 + }, + { + "epoch": 0.08999576988155668, + "grad_norm": 7.90625, + "learning_rate": 2.9985905567300916e-06, + "loss": 2.8934, + "step": 851 + }, + { + "epoch": 0.0901015228426396, + "grad_norm": 8.75, + "learning_rate": 3.002114164904863e-06, + "loss": 2.6529, + "step": 852 + }, + { + "epoch": 0.09020727580372251, + "grad_norm": 8.5625, + "learning_rate": 3.0056377730796336e-06, + "loss": 2.6064, + "step": 853 + }, + { + "epoch": 0.09031302876480542, + "grad_norm": 7.9375, + "learning_rate": 3.0091613812544046e-06, + "loss": 2.7934, + "step": 854 + }, + { + "epoch": 0.09041878172588833, + "grad_norm": 7.59375, + "learning_rate": 3.012684989429176e-06, + "loss": 2.7383, + "step": 855 + }, + { + "epoch": 0.09052453468697123, + "grad_norm": 7.5, + "learning_rate": 3.016208597603947e-06, + "loss": 2.7565, + "step": 856 + }, + { + "epoch": 0.09063028764805414, + "grad_norm": 8.0625, + "learning_rate": 3.0197322057787175e-06, + "loss": 2.6586, + "step": 857 + }, + { + "epoch": 0.09073604060913705, + "grad_norm": 8.9375, + "learning_rate": 3.0232558139534885e-06, + "loss": 2.7825, + "step": 858 + }, + { + "epoch": 0.09084179357021997, + "grad_norm": 8.0625, + "learning_rate": 3.02677942212826e-06, + "loss": 2.9203, + "step": 859 + }, + { + "epoch": 0.09094754653130288, + "grad_norm": 8.4375, + "learning_rate": 3.0303030303030305e-06, + "loss": 3.1157, + "step": 860 + }, + { + "epoch": 0.09105329949238579, + "grad_norm": 8.0, + "learning_rate": 3.0338266384778015e-06, + "loss": 2.8874, + "step": 861 + }, + { + "epoch": 0.0911590524534687, + "grad_norm": 8.125, + "learning_rate": 3.0373502466525725e-06, + "loss": 2.9741, + "step": 862 + }, + { + "epoch": 0.0912648054145516, + "grad_norm": 7.75, + "learning_rate": 3.0408738548273435e-06, + "loss": 2.7562, + "step": 863 + }, + { + "epoch": 0.09137055837563451, + "grad_norm": 8.0625, + "learning_rate": 3.0443974630021144e-06, + "loss": 2.9398, + "step": 864 + }, + { + "epoch": 0.09147631133671742, + "grad_norm": 8.1875, + "learning_rate": 3.0479210711768854e-06, + "loss": 2.87, + "step": 865 + }, + { + "epoch": 0.09158206429780034, + "grad_norm": 7.75, + "learning_rate": 3.051444679351656e-06, + "loss": 2.7413, + "step": 866 + }, + { + "epoch": 0.09168781725888325, + "grad_norm": 8.3125, + "learning_rate": 3.0549682875264274e-06, + "loss": 2.6844, + "step": 867 + }, + { + "epoch": 0.09179357021996616, + "grad_norm": 7.875, + "learning_rate": 3.0584918957011984e-06, + "loss": 2.7846, + "step": 868 + }, + { + "epoch": 0.09189932318104907, + "grad_norm": 8.5625, + "learning_rate": 3.062015503875969e-06, + "loss": 2.8051, + "step": 869 + }, + { + "epoch": 0.09200507614213198, + "grad_norm": 8.4375, + "learning_rate": 3.06553911205074e-06, + "loss": 2.5735, + "step": 870 + }, + { + "epoch": 0.09211082910321489, + "grad_norm": 7.46875, + "learning_rate": 3.0690627202255114e-06, + "loss": 2.7783, + "step": 871 + }, + { + "epoch": 0.0922165820642978, + "grad_norm": 8.1875, + "learning_rate": 3.0725863284002823e-06, + "loss": 2.6385, + "step": 872 + }, + { + "epoch": 0.09232233502538072, + "grad_norm": 7.96875, + "learning_rate": 3.076109936575053e-06, + "loss": 2.7731, + "step": 873 + }, + { + "epoch": 0.09242808798646363, + "grad_norm": 8.8125, + "learning_rate": 3.0796335447498243e-06, + "loss": 2.9694, + "step": 874 + }, + { + "epoch": 0.09253384094754653, + "grad_norm": 7.96875, + "learning_rate": 3.0831571529245953e-06, + "loss": 2.7977, + "step": 875 + }, + { + "epoch": 0.09263959390862944, + "grad_norm": 8.625, + "learning_rate": 3.086680761099366e-06, + "loss": 2.7899, + "step": 876 + }, + { + "epoch": 0.09274534686971235, + "grad_norm": 8.1875, + "learning_rate": 3.090204369274137e-06, + "loss": 2.5416, + "step": 877 + }, + { + "epoch": 0.09285109983079526, + "grad_norm": 7.75, + "learning_rate": 3.0937279774489083e-06, + "loss": 2.7629, + "step": 878 + }, + { + "epoch": 0.09295685279187817, + "grad_norm": 8.1875, + "learning_rate": 3.097251585623679e-06, + "loss": 2.8894, + "step": 879 + }, + { + "epoch": 0.09306260575296109, + "grad_norm": 7.46875, + "learning_rate": 3.10077519379845e-06, + "loss": 2.5222, + "step": 880 + }, + { + "epoch": 0.093168358714044, + "grad_norm": 7.78125, + "learning_rate": 3.104298801973221e-06, + "loss": 2.7861, + "step": 881 + }, + { + "epoch": 0.0932741116751269, + "grad_norm": 8.0, + "learning_rate": 3.1078224101479922e-06, + "loss": 2.8062, + "step": 882 + }, + { + "epoch": 0.09337986463620981, + "grad_norm": 7.78125, + "learning_rate": 3.1113460183227628e-06, + "loss": 2.7126, + "step": 883 + }, + { + "epoch": 0.09348561759729272, + "grad_norm": 8.5, + "learning_rate": 3.1148696264975338e-06, + "loss": 2.9444, + "step": 884 + }, + { + "epoch": 0.09359137055837563, + "grad_norm": 7.96875, + "learning_rate": 3.1183932346723043e-06, + "loss": 2.825, + "step": 885 + }, + { + "epoch": 0.09369712351945854, + "grad_norm": 9.5, + "learning_rate": 3.1219168428470757e-06, + "loss": 2.7658, + "step": 886 + }, + { + "epoch": 0.09380287648054146, + "grad_norm": 7.84375, + "learning_rate": 3.1254404510218467e-06, + "loss": 2.7339, + "step": 887 + }, + { + "epoch": 0.09390862944162437, + "grad_norm": 8.9375, + "learning_rate": 3.1289640591966173e-06, + "loss": 2.8197, + "step": 888 + }, + { + "epoch": 0.09401438240270728, + "grad_norm": 7.84375, + "learning_rate": 3.1324876673713883e-06, + "loss": 2.7514, + "step": 889 + }, + { + "epoch": 0.09412013536379019, + "grad_norm": 7.8125, + "learning_rate": 3.1360112755461597e-06, + "loss": 2.7466, + "step": 890 + }, + { + "epoch": 0.0942258883248731, + "grad_norm": 8.375, + "learning_rate": 3.1395348837209307e-06, + "loss": 2.6309, + "step": 891 + }, + { + "epoch": 0.094331641285956, + "grad_norm": 8.8125, + "learning_rate": 3.1430584918957013e-06, + "loss": 2.7274, + "step": 892 + }, + { + "epoch": 0.09443739424703891, + "grad_norm": 7.6875, + "learning_rate": 3.1465821000704722e-06, + "loss": 2.6553, + "step": 893 + }, + { + "epoch": 0.09454314720812183, + "grad_norm": 8.0, + "learning_rate": 3.1501057082452436e-06, + "loss": 2.718, + "step": 894 + }, + { + "epoch": 0.09464890016920474, + "grad_norm": 8.5, + "learning_rate": 3.1536293164200142e-06, + "loss": 2.9519, + "step": 895 + }, + { + "epoch": 0.09475465313028765, + "grad_norm": 8.125, + "learning_rate": 3.157152924594785e-06, + "loss": 2.747, + "step": 896 + }, + { + "epoch": 0.09486040609137056, + "grad_norm": 8.5625, + "learning_rate": 3.1606765327695566e-06, + "loss": 2.8658, + "step": 897 + }, + { + "epoch": 0.09496615905245347, + "grad_norm": 8.1875, + "learning_rate": 3.164200140944327e-06, + "loss": 2.8329, + "step": 898 + }, + { + "epoch": 0.09507191201353637, + "grad_norm": 7.875, + "learning_rate": 3.167723749119098e-06, + "loss": 2.7106, + "step": 899 + }, + { + "epoch": 0.09517766497461928, + "grad_norm": 9.0, + "learning_rate": 3.171247357293869e-06, + "loss": 3.0433, + "step": 900 + }, + { + "epoch": 0.0952834179357022, + "grad_norm": 7.75, + "learning_rate": 3.1747709654686406e-06, + "loss": 2.7284, + "step": 901 + }, + { + "epoch": 0.09538917089678511, + "grad_norm": 8.9375, + "learning_rate": 3.178294573643411e-06, + "loss": 2.6401, + "step": 902 + }, + { + "epoch": 0.09549492385786802, + "grad_norm": 8.125, + "learning_rate": 3.181818181818182e-06, + "loss": 2.8881, + "step": 903 + }, + { + "epoch": 0.09560067681895093, + "grad_norm": 8.5625, + "learning_rate": 3.1853417899929527e-06, + "loss": 2.6037, + "step": 904 + }, + { + "epoch": 0.09570642978003384, + "grad_norm": 8.375, + "learning_rate": 3.188865398167724e-06, + "loss": 2.8307, + "step": 905 + }, + { + "epoch": 0.09581218274111675, + "grad_norm": 8.625, + "learning_rate": 3.192389006342495e-06, + "loss": 2.6773, + "step": 906 + }, + { + "epoch": 0.09591793570219966, + "grad_norm": 7.5625, + "learning_rate": 3.195912614517266e-06, + "loss": 2.7252, + "step": 907 + }, + { + "epoch": 0.09602368866328258, + "grad_norm": 7.65625, + "learning_rate": 3.1994362226920366e-06, + "loss": 2.7652, + "step": 908 + }, + { + "epoch": 0.09612944162436549, + "grad_norm": 8.0, + "learning_rate": 3.202959830866808e-06, + "loss": 2.5122, + "step": 909 + }, + { + "epoch": 0.0962351945854484, + "grad_norm": 7.59375, + "learning_rate": 3.206483439041579e-06, + "loss": 2.7362, + "step": 910 + }, + { + "epoch": 0.0963409475465313, + "grad_norm": 7.90625, + "learning_rate": 3.2100070472163496e-06, + "loss": 2.836, + "step": 911 + }, + { + "epoch": 0.09644670050761421, + "grad_norm": 9.125, + "learning_rate": 3.2135306553911206e-06, + "loss": 2.5082, + "step": 912 + }, + { + "epoch": 0.09655245346869712, + "grad_norm": 8.0625, + "learning_rate": 3.217054263565892e-06, + "loss": 2.7792, + "step": 913 + }, + { + "epoch": 0.09665820642978003, + "grad_norm": 9.1875, + "learning_rate": 3.2205778717406626e-06, + "loss": 2.8709, + "step": 914 + }, + { + "epoch": 0.09676395939086295, + "grad_norm": 8.125, + "learning_rate": 3.2241014799154335e-06, + "loss": 2.808, + "step": 915 + }, + { + "epoch": 0.09686971235194586, + "grad_norm": 8.75, + "learning_rate": 3.2276250880902045e-06, + "loss": 3.0286, + "step": 916 + }, + { + "epoch": 0.09697546531302877, + "grad_norm": 7.84375, + "learning_rate": 3.231148696264976e-06, + "loss": 2.6905, + "step": 917 + }, + { + "epoch": 0.09708121827411167, + "grad_norm": 8.0625, + "learning_rate": 3.2346723044397465e-06, + "loss": 2.7504, + "step": 918 + }, + { + "epoch": 0.09718697123519458, + "grad_norm": 8.125, + "learning_rate": 3.2381959126145175e-06, + "loss": 2.7057, + "step": 919 + }, + { + "epoch": 0.09729272419627749, + "grad_norm": 7.75, + "learning_rate": 3.241719520789289e-06, + "loss": 2.6545, + "step": 920 + }, + { + "epoch": 0.0973984771573604, + "grad_norm": 7.84375, + "learning_rate": 3.2452431289640595e-06, + "loss": 2.7152, + "step": 921 + }, + { + "epoch": 0.09750423011844332, + "grad_norm": 7.59375, + "learning_rate": 3.2487667371388305e-06, + "loss": 2.7619, + "step": 922 + }, + { + "epoch": 0.09760998307952623, + "grad_norm": 8.375, + "learning_rate": 3.252290345313601e-06, + "loss": 2.9597, + "step": 923 + }, + { + "epoch": 0.09771573604060914, + "grad_norm": 8.1875, + "learning_rate": 3.2558139534883724e-06, + "loss": 2.6868, + "step": 924 + }, + { + "epoch": 0.09782148900169205, + "grad_norm": 8.3125, + "learning_rate": 3.2593375616631434e-06, + "loss": 2.7163, + "step": 925 + }, + { + "epoch": 0.09792724196277496, + "grad_norm": 8.25, + "learning_rate": 3.2628611698379144e-06, + "loss": 2.7665, + "step": 926 + }, + { + "epoch": 0.09803299492385786, + "grad_norm": 7.9375, + "learning_rate": 3.266384778012685e-06, + "loss": 2.7085, + "step": 927 + }, + { + "epoch": 0.09813874788494077, + "grad_norm": 8.5625, + "learning_rate": 3.2699083861874564e-06, + "loss": 2.6782, + "step": 928 + }, + { + "epoch": 0.0982445008460237, + "grad_norm": 8.3125, + "learning_rate": 3.2734319943622274e-06, + "loss": 2.7597, + "step": 929 + }, + { + "epoch": 0.0983502538071066, + "grad_norm": 7.65625, + "learning_rate": 3.276955602536998e-06, + "loss": 2.7049, + "step": 930 + }, + { + "epoch": 0.09845600676818951, + "grad_norm": 7.75, + "learning_rate": 3.280479210711769e-06, + "loss": 2.4215, + "step": 931 + }, + { + "epoch": 0.09856175972927242, + "grad_norm": 7.90625, + "learning_rate": 3.2840028188865403e-06, + "loss": 2.709, + "step": 932 + }, + { + "epoch": 0.09866751269035533, + "grad_norm": 9.5, + "learning_rate": 3.287526427061311e-06, + "loss": 2.7455, + "step": 933 + }, + { + "epoch": 0.09877326565143824, + "grad_norm": 7.09375, + "learning_rate": 3.291050035236082e-06, + "loss": 2.6546, + "step": 934 + }, + { + "epoch": 0.09887901861252114, + "grad_norm": 7.6875, + "learning_rate": 3.294573643410853e-06, + "loss": 2.7872, + "step": 935 + }, + { + "epoch": 0.09898477157360407, + "grad_norm": 8.125, + "learning_rate": 3.2980972515856243e-06, + "loss": 2.7731, + "step": 936 + }, + { + "epoch": 0.09909052453468697, + "grad_norm": 8.6875, + "learning_rate": 3.301620859760395e-06, + "loss": 2.8685, + "step": 937 + }, + { + "epoch": 0.09919627749576988, + "grad_norm": 7.5, + "learning_rate": 3.305144467935166e-06, + "loss": 2.8278, + "step": 938 + }, + { + "epoch": 0.09930203045685279, + "grad_norm": 7.78125, + "learning_rate": 3.3086680761099373e-06, + "loss": 2.749, + "step": 939 + }, + { + "epoch": 0.0994077834179357, + "grad_norm": 8.375, + "learning_rate": 3.312191684284708e-06, + "loss": 2.8703, + "step": 940 + }, + { + "epoch": 0.09951353637901861, + "grad_norm": 7.78125, + "learning_rate": 3.315715292459479e-06, + "loss": 2.7769, + "step": 941 + }, + { + "epoch": 0.09961928934010152, + "grad_norm": 7.875, + "learning_rate": 3.3192389006342494e-06, + "loss": 2.6711, + "step": 942 + }, + { + "epoch": 0.09972504230118444, + "grad_norm": 8.375, + "learning_rate": 3.3227625088090208e-06, + "loss": 2.8075, + "step": 943 + }, + { + "epoch": 0.09983079526226735, + "grad_norm": 7.6875, + "learning_rate": 3.3262861169837918e-06, + "loss": 2.6862, + "step": 944 + }, + { + "epoch": 0.09993654822335026, + "grad_norm": 8.375, + "learning_rate": 3.3298097251585628e-06, + "loss": 2.8816, + "step": 945 + }, + { + "epoch": 0.10004230118443316, + "grad_norm": 7.5, + "learning_rate": 3.3333333333333333e-06, + "loss": 2.7382, + "step": 946 + }, + { + "epoch": 0.10014805414551607, + "grad_norm": 8.5625, + "learning_rate": 3.3368569415081047e-06, + "loss": 2.7106, + "step": 947 + }, + { + "epoch": 0.10025380710659898, + "grad_norm": 7.78125, + "learning_rate": 3.3403805496828757e-06, + "loss": 2.8475, + "step": 948 + }, + { + "epoch": 0.10035956006768189, + "grad_norm": 8.25, + "learning_rate": 3.3439041578576463e-06, + "loss": 2.7824, + "step": 949 + }, + { + "epoch": 0.10046531302876481, + "grad_norm": 8.0625, + "learning_rate": 3.3474277660324173e-06, + "loss": 2.8653, + "step": 950 + }, + { + "epoch": 0.10057106598984772, + "grad_norm": 8.375, + "learning_rate": 3.3509513742071887e-06, + "loss": 2.7733, + "step": 951 + }, + { + "epoch": 0.10067681895093063, + "grad_norm": 8.875, + "learning_rate": 3.3544749823819593e-06, + "loss": 2.8515, + "step": 952 + }, + { + "epoch": 0.10078257191201354, + "grad_norm": 8.5, + "learning_rate": 3.3579985905567302e-06, + "loss": 2.6907, + "step": 953 + }, + { + "epoch": 0.10088832487309644, + "grad_norm": 8.3125, + "learning_rate": 3.3615221987315012e-06, + "loss": 2.871, + "step": 954 + }, + { + "epoch": 0.10099407783417935, + "grad_norm": 8.0625, + "learning_rate": 3.3650458069062726e-06, + "loss": 2.9063, + "step": 955 + }, + { + "epoch": 0.10109983079526226, + "grad_norm": 8.25, + "learning_rate": 3.368569415081043e-06, + "loss": 2.8625, + "step": 956 + }, + { + "epoch": 0.10120558375634518, + "grad_norm": 8.75, + "learning_rate": 3.372093023255814e-06, + "loss": 3.0793, + "step": 957 + }, + { + "epoch": 0.10131133671742809, + "grad_norm": 7.8125, + "learning_rate": 3.3756166314305848e-06, + "loss": 2.7514, + "step": 958 + }, + { + "epoch": 0.101417089678511, + "grad_norm": 7.78125, + "learning_rate": 3.379140239605356e-06, + "loss": 2.851, + "step": 959 + }, + { + "epoch": 0.10152284263959391, + "grad_norm": 8.375, + "learning_rate": 3.382663847780127e-06, + "loss": 2.771, + "step": 960 + }, + { + "epoch": 0.10162859560067682, + "grad_norm": 7.6875, + "learning_rate": 3.386187455954898e-06, + "loss": 2.6502, + "step": 961 + }, + { + "epoch": 0.10173434856175972, + "grad_norm": 7.65625, + "learning_rate": 3.389711064129669e-06, + "loss": 2.6949, + "step": 962 + }, + { + "epoch": 0.10184010152284265, + "grad_norm": 7.625, + "learning_rate": 3.39323467230444e-06, + "loss": 2.8741, + "step": 963 + }, + { + "epoch": 0.10194585448392555, + "grad_norm": 7.875, + "learning_rate": 3.396758280479211e-06, + "loss": 2.744, + "step": 964 + }, + { + "epoch": 0.10205160744500846, + "grad_norm": 7.59375, + "learning_rate": 3.4002818886539817e-06, + "loss": 2.7998, + "step": 965 + }, + { + "epoch": 0.10215736040609137, + "grad_norm": 7.6875, + "learning_rate": 3.403805496828753e-06, + "loss": 2.7342, + "step": 966 + }, + { + "epoch": 0.10226311336717428, + "grad_norm": 7.71875, + "learning_rate": 3.407329105003524e-06, + "loss": 2.8187, + "step": 967 + }, + { + "epoch": 0.10236886632825719, + "grad_norm": 9.5, + "learning_rate": 3.4108527131782946e-06, + "loss": 2.8142, + "step": 968 + }, + { + "epoch": 0.1024746192893401, + "grad_norm": 7.6875, + "learning_rate": 3.4143763213530656e-06, + "loss": 2.6595, + "step": 969 + }, + { + "epoch": 0.10258037225042302, + "grad_norm": 7.4375, + "learning_rate": 3.417899929527837e-06, + "loss": 2.7047, + "step": 970 + }, + { + "epoch": 0.10268612521150593, + "grad_norm": 8.5625, + "learning_rate": 3.421423537702608e-06, + "loss": 2.6621, + "step": 971 + }, + { + "epoch": 0.10279187817258884, + "grad_norm": 8.625, + "learning_rate": 3.4249471458773786e-06, + "loss": 2.5041, + "step": 972 + }, + { + "epoch": 0.10289763113367174, + "grad_norm": 8.375, + "learning_rate": 3.4284707540521496e-06, + "loss": 2.9098, + "step": 973 + }, + { + "epoch": 0.10300338409475465, + "grad_norm": 8.25, + "learning_rate": 3.431994362226921e-06, + "loss": 2.7249, + "step": 974 + }, + { + "epoch": 0.10310913705583756, + "grad_norm": 7.8125, + "learning_rate": 3.4355179704016915e-06, + "loss": 2.8528, + "step": 975 + }, + { + "epoch": 0.10321489001692047, + "grad_norm": 7.78125, + "learning_rate": 3.4390415785764625e-06, + "loss": 2.7797, + "step": 976 + }, + { + "epoch": 0.10332064297800339, + "grad_norm": 8.125, + "learning_rate": 3.442565186751233e-06, + "loss": 2.841, + "step": 977 + }, + { + "epoch": 0.1034263959390863, + "grad_norm": 7.9375, + "learning_rate": 3.4460887949260045e-06, + "loss": 2.7198, + "step": 978 + }, + { + "epoch": 0.10353214890016921, + "grad_norm": 7.75, + "learning_rate": 3.4496124031007755e-06, + "loss": 2.8368, + "step": 979 + }, + { + "epoch": 0.10363790186125212, + "grad_norm": 7.15625, + "learning_rate": 3.4531360112755465e-06, + "loss": 2.6188, + "step": 980 + }, + { + "epoch": 0.10374365482233502, + "grad_norm": 8.3125, + "learning_rate": 3.456659619450317e-06, + "loss": 2.6667, + "step": 981 + }, + { + "epoch": 0.10384940778341793, + "grad_norm": 8.1875, + "learning_rate": 3.4601832276250885e-06, + "loss": 2.7503, + "step": 982 + }, + { + "epoch": 0.10395516074450084, + "grad_norm": 8.25, + "learning_rate": 3.4637068357998595e-06, + "loss": 2.7677, + "step": 983 + }, + { + "epoch": 0.10406091370558376, + "grad_norm": 7.8125, + "learning_rate": 3.46723044397463e-06, + "loss": 2.9198, + "step": 984 + }, + { + "epoch": 0.10416666666666667, + "grad_norm": 9.25, + "learning_rate": 3.4707540521494014e-06, + "loss": 2.8639, + "step": 985 + }, + { + "epoch": 0.10427241962774958, + "grad_norm": 8.0625, + "learning_rate": 3.4742776603241724e-06, + "loss": 2.7354, + "step": 986 + }, + { + "epoch": 0.10437817258883249, + "grad_norm": 8.0, + "learning_rate": 3.477801268498943e-06, + "loss": 2.8481, + "step": 987 + }, + { + "epoch": 0.1044839255499154, + "grad_norm": 8.0625, + "learning_rate": 3.481324876673714e-06, + "loss": 2.9481, + "step": 988 + }, + { + "epoch": 0.1045896785109983, + "grad_norm": 7.84375, + "learning_rate": 3.4848484848484854e-06, + "loss": 2.625, + "step": 989 + }, + { + "epoch": 0.10469543147208121, + "grad_norm": 8.1875, + "learning_rate": 3.4883720930232564e-06, + "loss": 2.7982, + "step": 990 + }, + { + "epoch": 0.10480118443316414, + "grad_norm": 7.78125, + "learning_rate": 3.491895701198027e-06, + "loss": 2.8424, + "step": 991 + }, + { + "epoch": 0.10490693739424704, + "grad_norm": 8.0625, + "learning_rate": 3.495419309372798e-06, + "loss": 2.6371, + "step": 992 + }, + { + "epoch": 0.10501269035532995, + "grad_norm": 7.875, + "learning_rate": 3.4989429175475693e-06, + "loss": 2.7213, + "step": 993 + }, + { + "epoch": 0.10511844331641286, + "grad_norm": 8.1875, + "learning_rate": 3.50246652572234e-06, + "loss": 2.5454, + "step": 994 + }, + { + "epoch": 0.10522419627749577, + "grad_norm": 9.4375, + "learning_rate": 3.505990133897111e-06, + "loss": 2.6628, + "step": 995 + }, + { + "epoch": 0.10532994923857868, + "grad_norm": 7.78125, + "learning_rate": 3.509513742071882e-06, + "loss": 2.6424, + "step": 996 + }, + { + "epoch": 0.10543570219966159, + "grad_norm": 8.125, + "learning_rate": 3.513037350246653e-06, + "loss": 2.6895, + "step": 997 + }, + { + "epoch": 0.10554145516074451, + "grad_norm": 8.1875, + "learning_rate": 3.516560958421424e-06, + "loss": 2.7121, + "step": 998 + }, + { + "epoch": 0.10564720812182742, + "grad_norm": 9.625, + "learning_rate": 3.520084566596195e-06, + "loss": 3.0068, + "step": 999 + }, + { + "epoch": 0.10575296108291032, + "grad_norm": 7.6875, + "learning_rate": 3.5236081747709654e-06, + "loss": 2.8092, + "step": 1000 + }, + { + "epoch": 0.10585871404399323, + "grad_norm": 8.1875, + "learning_rate": 3.527131782945737e-06, + "loss": 2.7499, + "step": 1001 + }, + { + "epoch": 0.10596446700507614, + "grad_norm": 8.0, + "learning_rate": 3.530655391120508e-06, + "loss": 2.7285, + "step": 1002 + }, + { + "epoch": 0.10607021996615905, + "grad_norm": 8.1875, + "learning_rate": 3.5341789992952784e-06, + "loss": 2.6407, + "step": 1003 + }, + { + "epoch": 0.10617597292724196, + "grad_norm": 8.125, + "learning_rate": 3.5377026074700493e-06, + "loss": 2.8461, + "step": 1004 + }, + { + "epoch": 0.10628172588832488, + "grad_norm": 7.84375, + "learning_rate": 3.5412262156448208e-06, + "loss": 2.8794, + "step": 1005 + }, + { + "epoch": 0.10638747884940779, + "grad_norm": 8.0, + "learning_rate": 3.5447498238195917e-06, + "loss": 2.8471, + "step": 1006 + }, + { + "epoch": 0.1064932318104907, + "grad_norm": 8.0, + "learning_rate": 3.5482734319943623e-06, + "loss": 2.6236, + "step": 1007 + }, + { + "epoch": 0.1065989847715736, + "grad_norm": 8.0, + "learning_rate": 3.5517970401691337e-06, + "loss": 2.8274, + "step": 1008 + }, + { + "epoch": 0.10670473773265651, + "grad_norm": 8.3125, + "learning_rate": 3.5553206483439047e-06, + "loss": 2.7063, + "step": 1009 + }, + { + "epoch": 0.10681049069373942, + "grad_norm": 8.0625, + "learning_rate": 3.5588442565186753e-06, + "loss": 2.8838, + "step": 1010 + }, + { + "epoch": 0.10691624365482233, + "grad_norm": 7.59375, + "learning_rate": 3.5623678646934463e-06, + "loss": 2.616, + "step": 1011 + }, + { + "epoch": 0.10702199661590525, + "grad_norm": 7.6875, + "learning_rate": 3.5658914728682177e-06, + "loss": 2.8164, + "step": 1012 + }, + { + "epoch": 0.10712774957698816, + "grad_norm": 7.40625, + "learning_rate": 3.5694150810429882e-06, + "loss": 2.7465, + "step": 1013 + }, + { + "epoch": 0.10723350253807107, + "grad_norm": 7.65625, + "learning_rate": 3.5729386892177592e-06, + "loss": 2.6622, + "step": 1014 + }, + { + "epoch": 0.10733925549915398, + "grad_norm": 8.9375, + "learning_rate": 3.5764622973925302e-06, + "loss": 2.6004, + "step": 1015 + }, + { + "epoch": 0.10744500846023688, + "grad_norm": 7.4375, + "learning_rate": 3.579985905567301e-06, + "loss": 2.5872, + "step": 1016 + }, + { + "epoch": 0.1075507614213198, + "grad_norm": 9.1875, + "learning_rate": 3.583509513742072e-06, + "loss": 2.8429, + "step": 1017 + }, + { + "epoch": 0.1076565143824027, + "grad_norm": 7.59375, + "learning_rate": 3.587033121916843e-06, + "loss": 2.8753, + "step": 1018 + }, + { + "epoch": 0.10776226734348562, + "grad_norm": 8.0625, + "learning_rate": 3.5905567300916137e-06, + "loss": 2.9216, + "step": 1019 + }, + { + "epoch": 0.10786802030456853, + "grad_norm": 7.71875, + "learning_rate": 3.594080338266385e-06, + "loss": 2.6604, + "step": 1020 + }, + { + "epoch": 0.10797377326565144, + "grad_norm": 8.25, + "learning_rate": 3.597603946441156e-06, + "loss": 2.8527, + "step": 1021 + }, + { + "epoch": 0.10807952622673435, + "grad_norm": 7.78125, + "learning_rate": 3.6011275546159267e-06, + "loss": 2.9064, + "step": 1022 + }, + { + "epoch": 0.10818527918781726, + "grad_norm": 8.3125, + "learning_rate": 3.6046511627906977e-06, + "loss": 2.8432, + "step": 1023 + }, + { + "epoch": 0.10829103214890017, + "grad_norm": 8.875, + "learning_rate": 3.608174770965469e-06, + "loss": 2.9381, + "step": 1024 + }, + { + "epoch": 0.10839678510998307, + "grad_norm": 8.875, + "learning_rate": 3.61169837914024e-06, + "loss": 2.6708, + "step": 1025 + }, + { + "epoch": 0.108502538071066, + "grad_norm": 8.125, + "learning_rate": 3.6152219873150107e-06, + "loss": 2.8281, + "step": 1026 + }, + { + "epoch": 0.1086082910321489, + "grad_norm": 8.6875, + "learning_rate": 3.618745595489782e-06, + "loss": 2.8912, + "step": 1027 + }, + { + "epoch": 0.10871404399323181, + "grad_norm": 11.8125, + "learning_rate": 3.622269203664553e-06, + "loss": 2.8168, + "step": 1028 + }, + { + "epoch": 0.10881979695431472, + "grad_norm": 8.875, + "learning_rate": 3.6257928118393236e-06, + "loss": 2.4871, + "step": 1029 + }, + { + "epoch": 0.10892554991539763, + "grad_norm": 7.90625, + "learning_rate": 3.6293164200140946e-06, + "loss": 2.8002, + "step": 1030 + }, + { + "epoch": 0.10903130287648054, + "grad_norm": 8.6875, + "learning_rate": 3.632840028188866e-06, + "loss": 2.5199, + "step": 1031 + }, + { + "epoch": 0.10913705583756345, + "grad_norm": 7.59375, + "learning_rate": 3.6363636363636366e-06, + "loss": 2.7712, + "step": 1032 + }, + { + "epoch": 0.10924280879864637, + "grad_norm": 9.0625, + "learning_rate": 3.6398872445384076e-06, + "loss": 2.6995, + "step": 1033 + }, + { + "epoch": 0.10934856175972928, + "grad_norm": 7.59375, + "learning_rate": 3.6434108527131786e-06, + "loss": 2.7077, + "step": 1034 + }, + { + "epoch": 0.10945431472081218, + "grad_norm": 9.375, + "learning_rate": 3.64693446088795e-06, + "loss": 2.7848, + "step": 1035 + }, + { + "epoch": 0.10956006768189509, + "grad_norm": 7.84375, + "learning_rate": 3.6504580690627205e-06, + "loss": 2.8569, + "step": 1036 + }, + { + "epoch": 0.109665820642978, + "grad_norm": 7.9375, + "learning_rate": 3.6539816772374915e-06, + "loss": 2.7076, + "step": 1037 + }, + { + "epoch": 0.10977157360406091, + "grad_norm": 7.75, + "learning_rate": 3.657505285412262e-06, + "loss": 2.5797, + "step": 1038 + }, + { + "epoch": 0.10987732656514382, + "grad_norm": 8.0625, + "learning_rate": 3.6610288935870335e-06, + "loss": 2.7841, + "step": 1039 + }, + { + "epoch": 0.10998307952622674, + "grad_norm": 7.90625, + "learning_rate": 3.6645525017618045e-06, + "loss": 2.921, + "step": 1040 + }, + { + "epoch": 0.11008883248730965, + "grad_norm": 7.65625, + "learning_rate": 3.668076109936575e-06, + "loss": 2.9157, + "step": 1041 + }, + { + "epoch": 0.11019458544839256, + "grad_norm": 7.65625, + "learning_rate": 3.671599718111346e-06, + "loss": 2.8395, + "step": 1042 + }, + { + "epoch": 0.11030033840947547, + "grad_norm": 7.875, + "learning_rate": 3.6751233262861175e-06, + "loss": 2.7907, + "step": 1043 + }, + { + "epoch": 0.11040609137055837, + "grad_norm": 7.78125, + "learning_rate": 3.6786469344608884e-06, + "loss": 2.8491, + "step": 1044 + }, + { + "epoch": 0.11051184433164128, + "grad_norm": 7.6875, + "learning_rate": 3.682170542635659e-06, + "loss": 2.8247, + "step": 1045 + }, + { + "epoch": 0.11061759729272419, + "grad_norm": 7.65625, + "learning_rate": 3.68569415081043e-06, + "loss": 2.606, + "step": 1046 + }, + { + "epoch": 0.11072335025380711, + "grad_norm": 7.5625, + "learning_rate": 3.6892177589852014e-06, + "loss": 2.6741, + "step": 1047 + }, + { + "epoch": 0.11082910321489002, + "grad_norm": 7.6875, + "learning_rate": 3.692741367159972e-06, + "loss": 2.7343, + "step": 1048 + }, + { + "epoch": 0.11093485617597293, + "grad_norm": 8.0625, + "learning_rate": 3.696264975334743e-06, + "loss": 2.9497, + "step": 1049 + }, + { + "epoch": 0.11104060913705584, + "grad_norm": 7.625, + "learning_rate": 3.6997885835095144e-06, + "loss": 2.6245, + "step": 1050 + }, + { + "epoch": 0.11114636209813875, + "grad_norm": 7.78125, + "learning_rate": 3.703312191684285e-06, + "loss": 2.7709, + "step": 1051 + }, + { + "epoch": 0.11125211505922165, + "grad_norm": 8.1875, + "learning_rate": 3.706835799859056e-06, + "loss": 2.6182, + "step": 1052 + }, + { + "epoch": 0.11135786802030456, + "grad_norm": 7.84375, + "learning_rate": 3.710359408033827e-06, + "loss": 2.5421, + "step": 1053 + }, + { + "epoch": 0.11146362098138748, + "grad_norm": 8.125, + "learning_rate": 3.7138830162085983e-06, + "loss": 2.6673, + "step": 1054 + }, + { + "epoch": 0.11156937394247039, + "grad_norm": 7.6875, + "learning_rate": 3.717406624383369e-06, + "loss": 2.8234, + "step": 1055 + }, + { + "epoch": 0.1116751269035533, + "grad_norm": 10.0625, + "learning_rate": 3.72093023255814e-06, + "loss": 2.7392, + "step": 1056 + }, + { + "epoch": 0.11178087986463621, + "grad_norm": 8.4375, + "learning_rate": 3.7244538407329104e-06, + "loss": 2.8976, + "step": 1057 + }, + { + "epoch": 0.11188663282571912, + "grad_norm": 6.84375, + "learning_rate": 3.727977448907682e-06, + "loss": 2.7049, + "step": 1058 + }, + { + "epoch": 0.11199238578680203, + "grad_norm": 7.40625, + "learning_rate": 3.731501057082453e-06, + "loss": 2.7156, + "step": 1059 + }, + { + "epoch": 0.11209813874788493, + "grad_norm": 8.9375, + "learning_rate": 3.735024665257224e-06, + "loss": 2.9071, + "step": 1060 + }, + { + "epoch": 0.11220389170896786, + "grad_norm": 8.125, + "learning_rate": 3.7385482734319944e-06, + "loss": 2.9711, + "step": 1061 + }, + { + "epoch": 0.11230964467005077, + "grad_norm": 7.96875, + "learning_rate": 3.742071881606766e-06, + "loss": 2.903, + "step": 1062 + }, + { + "epoch": 0.11241539763113367, + "grad_norm": 7.84375, + "learning_rate": 3.7455954897815368e-06, + "loss": 2.7527, + "step": 1063 + }, + { + "epoch": 0.11252115059221658, + "grad_norm": 8.5, + "learning_rate": 3.7491190979563073e-06, + "loss": 2.5183, + "step": 1064 + }, + { + "epoch": 0.11262690355329949, + "grad_norm": 8.0625, + "learning_rate": 3.7526427061310783e-06, + "loss": 2.8787, + "step": 1065 + }, + { + "epoch": 0.1127326565143824, + "grad_norm": 7.5, + "learning_rate": 3.7561663143058497e-06, + "loss": 2.668, + "step": 1066 + }, + { + "epoch": 0.1128384094754653, + "grad_norm": 8.5, + "learning_rate": 3.7596899224806203e-06, + "loss": 2.8611, + "step": 1067 + }, + { + "epoch": 0.11294416243654823, + "grad_norm": 8.375, + "learning_rate": 3.7632135306553913e-06, + "loss": 2.491, + "step": 1068 + }, + { + "epoch": 0.11304991539763114, + "grad_norm": 8.4375, + "learning_rate": 3.7667371388301623e-06, + "loss": 2.6421, + "step": 1069 + }, + { + "epoch": 0.11315566835871405, + "grad_norm": 8.0625, + "learning_rate": 3.7702607470049337e-06, + "loss": 2.833, + "step": 1070 + }, + { + "epoch": 0.11326142131979695, + "grad_norm": 8.75, + "learning_rate": 3.7737843551797043e-06, + "loss": 2.6749, + "step": 1071 + }, + { + "epoch": 0.11336717428087986, + "grad_norm": 7.84375, + "learning_rate": 3.7773079633544753e-06, + "loss": 2.7013, + "step": 1072 + }, + { + "epoch": 0.11347292724196277, + "grad_norm": 8.125, + "learning_rate": 3.7808315715292467e-06, + "loss": 2.7661, + "step": 1073 + }, + { + "epoch": 0.11357868020304568, + "grad_norm": 7.6875, + "learning_rate": 3.7843551797040172e-06, + "loss": 2.8041, + "step": 1074 + }, + { + "epoch": 0.1136844331641286, + "grad_norm": 7.71875, + "learning_rate": 3.7878787878787882e-06, + "loss": 2.6713, + "step": 1075 + }, + { + "epoch": 0.11379018612521151, + "grad_norm": 8.25, + "learning_rate": 3.7914023960535588e-06, + "loss": 2.8764, + "step": 1076 + }, + { + "epoch": 0.11389593908629442, + "grad_norm": 8.25, + "learning_rate": 3.79492600422833e-06, + "loss": 3.0791, + "step": 1077 + }, + { + "epoch": 0.11400169204737733, + "grad_norm": 7.84375, + "learning_rate": 3.798449612403101e-06, + "loss": 2.6677, + "step": 1078 + }, + { + "epoch": 0.11410744500846023, + "grad_norm": 7.78125, + "learning_rate": 3.801973220577872e-06, + "loss": 2.9061, + "step": 1079 + }, + { + "epoch": 0.11421319796954314, + "grad_norm": 8.3125, + "learning_rate": 3.8054968287526427e-06, + "loss": 2.7203, + "step": 1080 + }, + { + "epoch": 0.11431895093062605, + "grad_norm": 7.75, + "learning_rate": 3.809020436927414e-06, + "loss": 2.6077, + "step": 1081 + }, + { + "epoch": 0.11442470389170897, + "grad_norm": 8.25, + "learning_rate": 3.812544045102185e-06, + "loss": 2.582, + "step": 1082 + }, + { + "epoch": 0.11453045685279188, + "grad_norm": 7.78125, + "learning_rate": 3.816067653276956e-06, + "loss": 2.7511, + "step": 1083 + }, + { + "epoch": 0.11463620981387479, + "grad_norm": 8.0625, + "learning_rate": 3.819591261451726e-06, + "loss": 2.9279, + "step": 1084 + }, + { + "epoch": 0.1147419627749577, + "grad_norm": 7.53125, + "learning_rate": 3.823114869626498e-06, + "loss": 2.567, + "step": 1085 + }, + { + "epoch": 0.1148477157360406, + "grad_norm": 8.375, + "learning_rate": 3.826638477801269e-06, + "loss": 2.956, + "step": 1086 + }, + { + "epoch": 0.11495346869712351, + "grad_norm": 7.96875, + "learning_rate": 3.83016208597604e-06, + "loss": 2.6674, + "step": 1087 + }, + { + "epoch": 0.11505922165820642, + "grad_norm": 8.5625, + "learning_rate": 3.83368569415081e-06, + "loss": 2.7492, + "step": 1088 + }, + { + "epoch": 0.11516497461928935, + "grad_norm": 7.5625, + "learning_rate": 3.837209302325582e-06, + "loss": 2.8757, + "step": 1089 + }, + { + "epoch": 0.11527072758037225, + "grad_norm": 8.0, + "learning_rate": 3.840732910500353e-06, + "loss": 2.5894, + "step": 1090 + }, + { + "epoch": 0.11537648054145516, + "grad_norm": 8.375, + "learning_rate": 3.844256518675123e-06, + "loss": 2.9392, + "step": 1091 + }, + { + "epoch": 0.11548223350253807, + "grad_norm": 8.0625, + "learning_rate": 3.847780126849895e-06, + "loss": 2.7175, + "step": 1092 + }, + { + "epoch": 0.11558798646362098, + "grad_norm": 7.75, + "learning_rate": 3.851303735024666e-06, + "loss": 2.6261, + "step": 1093 + }, + { + "epoch": 0.11569373942470389, + "grad_norm": 14.875, + "learning_rate": 3.854827343199436e-06, + "loss": 2.5795, + "step": 1094 + }, + { + "epoch": 0.1157994923857868, + "grad_norm": 8.1875, + "learning_rate": 3.858350951374207e-06, + "loss": 2.6461, + "step": 1095 + }, + { + "epoch": 0.11590524534686972, + "grad_norm": 8.5625, + "learning_rate": 3.861874559548979e-06, + "loss": 2.7479, + "step": 1096 + }, + { + "epoch": 0.11601099830795263, + "grad_norm": 8.1875, + "learning_rate": 3.86539816772375e-06, + "loss": 2.884, + "step": 1097 + }, + { + "epoch": 0.11611675126903553, + "grad_norm": 8.8125, + "learning_rate": 3.86892177589852e-06, + "loss": 2.8949, + "step": 1098 + }, + { + "epoch": 0.11622250423011844, + "grad_norm": 8.5625, + "learning_rate": 3.872445384073291e-06, + "loss": 2.6613, + "step": 1099 + }, + { + "epoch": 0.11632825719120135, + "grad_norm": 8.125, + "learning_rate": 3.875968992248063e-06, + "loss": 2.765, + "step": 1100 + }, + { + "epoch": 0.11643401015228426, + "grad_norm": 7.71875, + "learning_rate": 3.879492600422833e-06, + "loss": 2.8384, + "step": 1101 + }, + { + "epoch": 0.11653976311336717, + "grad_norm": 8.0625, + "learning_rate": 3.883016208597604e-06, + "loss": 2.8369, + "step": 1102 + }, + { + "epoch": 0.11664551607445009, + "grad_norm": 8.5625, + "learning_rate": 3.886539816772375e-06, + "loss": 2.8124, + "step": 1103 + }, + { + "epoch": 0.116751269035533, + "grad_norm": 7.8125, + "learning_rate": 3.890063424947146e-06, + "loss": 2.7814, + "step": 1104 + }, + { + "epoch": 0.1168570219966159, + "grad_norm": 8.3125, + "learning_rate": 3.893587033121917e-06, + "loss": 2.7572, + "step": 1105 + }, + { + "epoch": 0.11696277495769881, + "grad_norm": 7.90625, + "learning_rate": 3.897110641296688e-06, + "loss": 2.8949, + "step": 1106 + }, + { + "epoch": 0.11706852791878172, + "grad_norm": 7.46875, + "learning_rate": 3.900634249471459e-06, + "loss": 2.7298, + "step": 1107 + }, + { + "epoch": 0.11717428087986463, + "grad_norm": 9.1875, + "learning_rate": 3.90415785764623e-06, + "loss": 2.8488, + "step": 1108 + }, + { + "epoch": 0.11728003384094755, + "grad_norm": 7.65625, + "learning_rate": 3.907681465821001e-06, + "loss": 2.8319, + "step": 1109 + }, + { + "epoch": 0.11738578680203046, + "grad_norm": 8.5, + "learning_rate": 3.911205073995772e-06, + "loss": 2.9979, + "step": 1110 + }, + { + "epoch": 0.11749153976311337, + "grad_norm": 8.0, + "learning_rate": 3.914728682170543e-06, + "loss": 2.7812, + "step": 1111 + }, + { + "epoch": 0.11759729272419628, + "grad_norm": 7.96875, + "learning_rate": 3.918252290345314e-06, + "loss": 2.4872, + "step": 1112 + }, + { + "epoch": 0.11770304568527919, + "grad_norm": 8.0, + "learning_rate": 3.921775898520085e-06, + "loss": 2.613, + "step": 1113 + }, + { + "epoch": 0.1178087986463621, + "grad_norm": 8.1875, + "learning_rate": 3.925299506694856e-06, + "loss": 2.5855, + "step": 1114 + }, + { + "epoch": 0.117914551607445, + "grad_norm": 9.1875, + "learning_rate": 3.928823114869627e-06, + "loss": 2.7945, + "step": 1115 + }, + { + "epoch": 0.11802030456852793, + "grad_norm": 7.90625, + "learning_rate": 3.932346723044398e-06, + "loss": 2.7411, + "step": 1116 + }, + { + "epoch": 0.11812605752961083, + "grad_norm": 10.125, + "learning_rate": 3.935870331219169e-06, + "loss": 2.7898, + "step": 1117 + }, + { + "epoch": 0.11823181049069374, + "grad_norm": 7.5625, + "learning_rate": 3.93939393939394e-06, + "loss": 2.7559, + "step": 1118 + }, + { + "epoch": 0.11833756345177665, + "grad_norm": 8.6875, + "learning_rate": 3.942917547568711e-06, + "loss": 2.4153, + "step": 1119 + }, + { + "epoch": 0.11844331641285956, + "grad_norm": 7.5625, + "learning_rate": 3.946441155743482e-06, + "loss": 2.7178, + "step": 1120 + }, + { + "epoch": 0.11854906937394247, + "grad_norm": 7.625, + "learning_rate": 3.949964763918253e-06, + "loss": 2.6588, + "step": 1121 + }, + { + "epoch": 0.11865482233502538, + "grad_norm": 8.625, + "learning_rate": 3.953488372093024e-06, + "loss": 2.5786, + "step": 1122 + }, + { + "epoch": 0.1187605752961083, + "grad_norm": 8.5625, + "learning_rate": 3.957011980267795e-06, + "loss": 2.688, + "step": 1123 + }, + { + "epoch": 0.1188663282571912, + "grad_norm": 8.25, + "learning_rate": 3.960535588442566e-06, + "loss": 2.9948, + "step": 1124 + }, + { + "epoch": 0.11897208121827411, + "grad_norm": 7.84375, + "learning_rate": 3.964059196617337e-06, + "loss": 2.6393, + "step": 1125 + }, + { + "epoch": 0.11907783417935702, + "grad_norm": 7.84375, + "learning_rate": 3.967582804792107e-06, + "loss": 2.7523, + "step": 1126 + }, + { + "epoch": 0.11918358714043993, + "grad_norm": 7.8125, + "learning_rate": 3.971106412966879e-06, + "loss": 2.6895, + "step": 1127 + }, + { + "epoch": 0.11928934010152284, + "grad_norm": 8.125, + "learning_rate": 3.97463002114165e-06, + "loss": 2.6714, + "step": 1128 + }, + { + "epoch": 0.11939509306260575, + "grad_norm": 8.125, + "learning_rate": 3.97815362931642e-06, + "loss": 2.5874, + "step": 1129 + }, + { + "epoch": 0.11950084602368867, + "grad_norm": 7.84375, + "learning_rate": 3.981677237491191e-06, + "loss": 2.7003, + "step": 1130 + }, + { + "epoch": 0.11960659898477158, + "grad_norm": 7.46875, + "learning_rate": 3.985200845665963e-06, + "loss": 2.517, + "step": 1131 + }, + { + "epoch": 0.11971235194585449, + "grad_norm": 7.59375, + "learning_rate": 3.988724453840734e-06, + "loss": 2.6385, + "step": 1132 + }, + { + "epoch": 0.1198181049069374, + "grad_norm": 8.5, + "learning_rate": 3.992248062015504e-06, + "loss": 2.73, + "step": 1133 + }, + { + "epoch": 0.1199238578680203, + "grad_norm": 8.25, + "learning_rate": 3.995771670190275e-06, + "loss": 2.6259, + "step": 1134 + }, + { + "epoch": 0.12002961082910321, + "grad_norm": 8.0625, + "learning_rate": 3.999295278365047e-06, + "loss": 2.8053, + "step": 1135 + }, + { + "epoch": 0.12013536379018612, + "grad_norm": 7.5, + "learning_rate": 4.002818886539817e-06, + "loss": 2.6044, + "step": 1136 + }, + { + "epoch": 0.12024111675126904, + "grad_norm": 8.75, + "learning_rate": 4.006342494714588e-06, + "loss": 2.8909, + "step": 1137 + }, + { + "epoch": 0.12034686971235195, + "grad_norm": 7.875, + "learning_rate": 4.00986610288936e-06, + "loss": 2.4164, + "step": 1138 + }, + { + "epoch": 0.12045262267343486, + "grad_norm": 7.6875, + "learning_rate": 4.01338971106413e-06, + "loss": 2.7767, + "step": 1139 + }, + { + "epoch": 0.12055837563451777, + "grad_norm": 8.0625, + "learning_rate": 4.016913319238901e-06, + "loss": 2.8255, + "step": 1140 + }, + { + "epoch": 0.12066412859560068, + "grad_norm": 7.375, + "learning_rate": 4.020436927413672e-06, + "loss": 2.6695, + "step": 1141 + }, + { + "epoch": 0.12076988155668358, + "grad_norm": 7.65625, + "learning_rate": 4.023960535588443e-06, + "loss": 2.8111, + "step": 1142 + }, + { + "epoch": 0.12087563451776649, + "grad_norm": 8.0625, + "learning_rate": 4.027484143763214e-06, + "loss": 2.7809, + "step": 1143 + }, + { + "epoch": 0.12098138747884941, + "grad_norm": 8.4375, + "learning_rate": 4.031007751937985e-06, + "loss": 2.9548, + "step": 1144 + }, + { + "epoch": 0.12108714043993232, + "grad_norm": 7.96875, + "learning_rate": 4.034531360112756e-06, + "loss": 2.6476, + "step": 1145 + }, + { + "epoch": 0.12119289340101523, + "grad_norm": 8.1875, + "learning_rate": 4.038054968287527e-06, + "loss": 2.8782, + "step": 1146 + }, + { + "epoch": 0.12129864636209814, + "grad_norm": 8.6875, + "learning_rate": 4.041578576462298e-06, + "loss": 2.7903, + "step": 1147 + }, + { + "epoch": 0.12140439932318105, + "grad_norm": 8.0, + "learning_rate": 4.045102184637069e-06, + "loss": 2.6024, + "step": 1148 + }, + { + "epoch": 0.12151015228426396, + "grad_norm": 7.9375, + "learning_rate": 4.04862579281184e-06, + "loss": 2.8512, + "step": 1149 + }, + { + "epoch": 0.12161590524534686, + "grad_norm": 7.71875, + "learning_rate": 4.052149400986611e-06, + "loss": 2.7486, + "step": 1150 + }, + { + "epoch": 0.12172165820642979, + "grad_norm": 8.0625, + "learning_rate": 4.055673009161382e-06, + "loss": 2.6022, + "step": 1151 + }, + { + "epoch": 0.1218274111675127, + "grad_norm": 7.875, + "learning_rate": 4.059196617336153e-06, + "loss": 2.6918, + "step": 1152 + }, + { + "epoch": 0.1219331641285956, + "grad_norm": 8.0625, + "learning_rate": 4.0627202255109236e-06, + "loss": 2.6992, + "step": 1153 + }, + { + "epoch": 0.12203891708967851, + "grad_norm": 7.625, + "learning_rate": 4.0662438336856946e-06, + "loss": 2.6445, + "step": 1154 + }, + { + "epoch": 0.12214467005076142, + "grad_norm": 7.625, + "learning_rate": 4.0697674418604655e-06, + "loss": 2.5229, + "step": 1155 + }, + { + "epoch": 0.12225042301184433, + "grad_norm": 8.5, + "learning_rate": 4.0732910500352365e-06, + "loss": 2.4618, + "step": 1156 + }, + { + "epoch": 0.12235617597292724, + "grad_norm": 8.0625, + "learning_rate": 4.0768146582100075e-06, + "loss": 2.6554, + "step": 1157 + }, + { + "epoch": 0.12246192893401016, + "grad_norm": 8.25, + "learning_rate": 4.0803382663847785e-06, + "loss": 2.7929, + "step": 1158 + }, + { + "epoch": 0.12256768189509307, + "grad_norm": 7.90625, + "learning_rate": 4.0838618745595495e-06, + "loss": 2.5576, + "step": 1159 + }, + { + "epoch": 0.12267343485617598, + "grad_norm": 7.5, + "learning_rate": 4.0873854827343205e-06, + "loss": 2.7291, + "step": 1160 + }, + { + "epoch": 0.12277918781725888, + "grad_norm": 7.5625, + "learning_rate": 4.0909090909090915e-06, + "loss": 2.5773, + "step": 1161 + }, + { + "epoch": 0.12288494077834179, + "grad_norm": 7.96875, + "learning_rate": 4.0944326990838625e-06, + "loss": 2.6753, + "step": 1162 + }, + { + "epoch": 0.1229906937394247, + "grad_norm": 8.4375, + "learning_rate": 4.0979563072586335e-06, + "loss": 2.946, + "step": 1163 + }, + { + "epoch": 0.12309644670050761, + "grad_norm": 8.0, + "learning_rate": 4.101479915433404e-06, + "loss": 2.8228, + "step": 1164 + }, + { + "epoch": 0.12320219966159053, + "grad_norm": 7.8125, + "learning_rate": 4.1050035236081754e-06, + "loss": 2.7558, + "step": 1165 + }, + { + "epoch": 0.12330795262267344, + "grad_norm": 7.96875, + "learning_rate": 4.108527131782946e-06, + "loss": 2.6983, + "step": 1166 + }, + { + "epoch": 0.12341370558375635, + "grad_norm": 7.875, + "learning_rate": 4.1120507399577166e-06, + "loss": 2.7032, + "step": 1167 + }, + { + "epoch": 0.12351945854483926, + "grad_norm": 8.75, + "learning_rate": 4.1155743481324875e-06, + "loss": 2.8349, + "step": 1168 + }, + { + "epoch": 0.12362521150592216, + "grad_norm": 7.84375, + "learning_rate": 4.119097956307259e-06, + "loss": 2.6607, + "step": 1169 + }, + { + "epoch": 0.12373096446700507, + "grad_norm": 7.65625, + "learning_rate": 4.12262156448203e-06, + "loss": 2.7335, + "step": 1170 + }, + { + "epoch": 0.12383671742808798, + "grad_norm": 7.875, + "learning_rate": 4.1261451726568005e-06, + "loss": 2.6447, + "step": 1171 + }, + { + "epoch": 0.1239424703891709, + "grad_norm": 7.90625, + "learning_rate": 4.1296687808315715e-06, + "loss": 2.6237, + "step": 1172 + }, + { + "epoch": 0.12404822335025381, + "grad_norm": 8.5625, + "learning_rate": 4.133192389006343e-06, + "loss": 2.7873, + "step": 1173 + }, + { + "epoch": 0.12415397631133672, + "grad_norm": 7.9375, + "learning_rate": 4.1367159971811135e-06, + "loss": 2.9064, + "step": 1174 + }, + { + "epoch": 0.12425972927241963, + "grad_norm": 7.5, + "learning_rate": 4.1402396053558845e-06, + "loss": 2.7037, + "step": 1175 + }, + { + "epoch": 0.12436548223350254, + "grad_norm": 7.28125, + "learning_rate": 4.1437632135306554e-06, + "loss": 2.5781, + "step": 1176 + }, + { + "epoch": 0.12447123519458544, + "grad_norm": 7.65625, + "learning_rate": 4.1472868217054264e-06, + "loss": 2.719, + "step": 1177 + }, + { + "epoch": 0.12457698815566835, + "grad_norm": 8.1875, + "learning_rate": 4.150810429880197e-06, + "loss": 3.0749, + "step": 1178 + }, + { + "epoch": 0.12468274111675128, + "grad_norm": 7.75, + "learning_rate": 4.154334038054968e-06, + "loss": 2.6748, + "step": 1179 + }, + { + "epoch": 0.12478849407783418, + "grad_norm": 8.375, + "learning_rate": 4.15785764622974e-06, + "loss": 2.7501, + "step": 1180 + }, + { + "epoch": 0.12489424703891709, + "grad_norm": 7.71875, + "learning_rate": 4.16138125440451e-06, + "loss": 2.8778, + "step": 1181 + }, + { + "epoch": 0.125, + "grad_norm": 8.1875, + "learning_rate": 4.164904862579281e-06, + "loss": 2.7608, + "step": 1182 + }, + { + "epoch": 0.12510575296108292, + "grad_norm": 7.8125, + "learning_rate": 4.168428470754052e-06, + "loss": 2.9555, + "step": 1183 + }, + { + "epoch": 0.12521150592216582, + "grad_norm": 8.375, + "learning_rate": 4.171952078928823e-06, + "loss": 2.8133, + "step": 1184 + }, + { + "epoch": 0.12531725888324874, + "grad_norm": 8.75, + "learning_rate": 4.175475687103594e-06, + "loss": 3.0349, + "step": 1185 + }, + { + "epoch": 0.12542301184433163, + "grad_norm": 7.59375, + "learning_rate": 4.178999295278365e-06, + "loss": 2.7233, + "step": 1186 + }, + { + "epoch": 0.12552876480541456, + "grad_norm": 7.53125, + "learning_rate": 4.182522903453136e-06, + "loss": 2.6763, + "step": 1187 + }, + { + "epoch": 0.12563451776649745, + "grad_norm": 8.8125, + "learning_rate": 4.186046511627907e-06, + "loss": 2.5793, + "step": 1188 + }, + { + "epoch": 0.12574027072758037, + "grad_norm": 9.5625, + "learning_rate": 4.189570119802678e-06, + "loss": 2.6594, + "step": 1189 + }, + { + "epoch": 0.1258460236886633, + "grad_norm": 7.875, + "learning_rate": 4.193093727977449e-06, + "loss": 2.7904, + "step": 1190 + }, + { + "epoch": 0.1259517766497462, + "grad_norm": 8.125, + "learning_rate": 4.19661733615222e-06, + "loss": 2.6075, + "step": 1191 + }, + { + "epoch": 0.1260575296108291, + "grad_norm": 7.84375, + "learning_rate": 4.200140944326991e-06, + "loss": 2.633, + "step": 1192 + }, + { + "epoch": 0.126163282571912, + "grad_norm": 7.78125, + "learning_rate": 4.203664552501762e-06, + "loss": 2.6451, + "step": 1193 + }, + { + "epoch": 0.12626903553299493, + "grad_norm": 7.5, + "learning_rate": 4.207188160676533e-06, + "loss": 2.7155, + "step": 1194 + }, + { + "epoch": 0.12637478849407782, + "grad_norm": 8.8125, + "learning_rate": 4.210711768851304e-06, + "loss": 2.6852, + "step": 1195 + }, + { + "epoch": 0.12648054145516074, + "grad_norm": 8.125, + "learning_rate": 4.214235377026075e-06, + "loss": 2.5479, + "step": 1196 + }, + { + "epoch": 0.12658629441624367, + "grad_norm": 8.0625, + "learning_rate": 4.217758985200846e-06, + "loss": 2.7297, + "step": 1197 + }, + { + "epoch": 0.12669204737732656, + "grad_norm": 8.5625, + "learning_rate": 4.221282593375617e-06, + "loss": 2.949, + "step": 1198 + }, + { + "epoch": 0.12679780033840948, + "grad_norm": 7.375, + "learning_rate": 4.224806201550387e-06, + "loss": 2.6527, + "step": 1199 + }, + { + "epoch": 0.12690355329949238, + "grad_norm": 8.0625, + "learning_rate": 4.228329809725159e-06, + "loss": 2.758, + "step": 1200 + }, + { + "epoch": 0.1270093062605753, + "grad_norm": 7.5625, + "learning_rate": 4.23185341789993e-06, + "loss": 2.6419, + "step": 1201 + }, + { + "epoch": 0.1271150592216582, + "grad_norm": 7.46875, + "learning_rate": 4.2353770260747e-06, + "loss": 2.6681, + "step": 1202 + }, + { + "epoch": 0.12722081218274112, + "grad_norm": 8.125, + "learning_rate": 4.238900634249472e-06, + "loss": 2.6364, + "step": 1203 + }, + { + "epoch": 0.12732656514382404, + "grad_norm": 8.4375, + "learning_rate": 4.242424242424243e-06, + "loss": 3.0272, + "step": 1204 + }, + { + "epoch": 0.12743231810490693, + "grad_norm": 8.25, + "learning_rate": 4.245947850599014e-06, + "loss": 2.8453, + "step": 1205 + }, + { + "epoch": 0.12753807106598986, + "grad_norm": 7.78125, + "learning_rate": 4.249471458773784e-06, + "loss": 2.5608, + "step": 1206 + }, + { + "epoch": 0.12764382402707275, + "grad_norm": 8.75, + "learning_rate": 4.252995066948556e-06, + "loss": 2.7727, + "step": 1207 + }, + { + "epoch": 0.12774957698815567, + "grad_norm": 7.34375, + "learning_rate": 4.256518675123327e-06, + "loss": 2.5744, + "step": 1208 + }, + { + "epoch": 0.12785532994923857, + "grad_norm": 8.75, + "learning_rate": 4.260042283298097e-06, + "loss": 2.641, + "step": 1209 + }, + { + "epoch": 0.1279610829103215, + "grad_norm": 8.0625, + "learning_rate": 4.263565891472868e-06, + "loss": 2.7561, + "step": 1210 + }, + { + "epoch": 0.1280668358714044, + "grad_norm": 8.0, + "learning_rate": 4.26708949964764e-06, + "loss": 2.8598, + "step": 1211 + }, + { + "epoch": 0.1281725888324873, + "grad_norm": 7.75, + "learning_rate": 4.27061310782241e-06, + "loss": 2.8148, + "step": 1212 + }, + { + "epoch": 0.12827834179357023, + "grad_norm": 8.3125, + "learning_rate": 4.274136715997181e-06, + "loss": 2.761, + "step": 1213 + }, + { + "epoch": 0.12838409475465312, + "grad_norm": 9.3125, + "learning_rate": 4.277660324171952e-06, + "loss": 2.7351, + "step": 1214 + }, + { + "epoch": 0.12848984771573604, + "grad_norm": 7.84375, + "learning_rate": 4.281183932346724e-06, + "loss": 2.6323, + "step": 1215 + }, + { + "epoch": 0.12859560067681894, + "grad_norm": 7.96875, + "learning_rate": 4.284707540521494e-06, + "loss": 3.0348, + "step": 1216 + }, + { + "epoch": 0.12870135363790186, + "grad_norm": 8.25, + "learning_rate": 4.288231148696265e-06, + "loss": 2.6736, + "step": 1217 + }, + { + "epoch": 0.12880710659898478, + "grad_norm": 7.96875, + "learning_rate": 4.291754756871036e-06, + "loss": 2.6407, + "step": 1218 + }, + { + "epoch": 0.12891285956006768, + "grad_norm": 8.1875, + "learning_rate": 4.295278365045807e-06, + "loss": 2.7407, + "step": 1219 + }, + { + "epoch": 0.1290186125211506, + "grad_norm": 7.4375, + "learning_rate": 4.298801973220578e-06, + "loss": 2.694, + "step": 1220 + }, + { + "epoch": 0.1291243654822335, + "grad_norm": 8.0, + "learning_rate": 4.302325581395349e-06, + "loss": 2.657, + "step": 1221 + }, + { + "epoch": 0.12923011844331642, + "grad_norm": 8.3125, + "learning_rate": 4.30584918957012e-06, + "loss": 2.6047, + "step": 1222 + }, + { + "epoch": 0.1293358714043993, + "grad_norm": 7.96875, + "learning_rate": 4.309372797744891e-06, + "loss": 2.6312, + "step": 1223 + }, + { + "epoch": 0.12944162436548223, + "grad_norm": 7.625, + "learning_rate": 4.312896405919662e-06, + "loss": 2.8136, + "step": 1224 + }, + { + "epoch": 0.12954737732656516, + "grad_norm": 7.71875, + "learning_rate": 4.316420014094433e-06, + "loss": 2.9357, + "step": 1225 + }, + { + "epoch": 0.12965313028764805, + "grad_norm": 7.5625, + "learning_rate": 4.319943622269204e-06, + "loss": 2.7084, + "step": 1226 + }, + { + "epoch": 0.12975888324873097, + "grad_norm": 8.5, + "learning_rate": 4.323467230443975e-06, + "loss": 2.7023, + "step": 1227 + }, + { + "epoch": 0.12986463620981387, + "grad_norm": 7.96875, + "learning_rate": 4.326990838618746e-06, + "loss": 2.7384, + "step": 1228 + }, + { + "epoch": 0.1299703891708968, + "grad_norm": 8.25, + "learning_rate": 4.330514446793517e-06, + "loss": 2.8484, + "step": 1229 + }, + { + "epoch": 0.13007614213197968, + "grad_norm": 7.71875, + "learning_rate": 4.334038054968288e-06, + "loss": 2.6442, + "step": 1230 + }, + { + "epoch": 0.1301818950930626, + "grad_norm": 7.78125, + "learning_rate": 4.337561663143059e-06, + "loss": 2.4833, + "step": 1231 + }, + { + "epoch": 0.13028764805414553, + "grad_norm": 7.90625, + "learning_rate": 4.34108527131783e-06, + "loss": 2.8376, + "step": 1232 + }, + { + "epoch": 0.13039340101522842, + "grad_norm": 8.375, + "learning_rate": 4.344608879492601e-06, + "loss": 2.8934, + "step": 1233 + }, + { + "epoch": 0.13049915397631134, + "grad_norm": 8.3125, + "learning_rate": 4.348132487667372e-06, + "loss": 2.608, + "step": 1234 + }, + { + "epoch": 0.13060490693739424, + "grad_norm": 7.9375, + "learning_rate": 4.351656095842143e-06, + "loss": 2.727, + "step": 1235 + }, + { + "epoch": 0.13071065989847716, + "grad_norm": 8.125, + "learning_rate": 4.355179704016914e-06, + "loss": 2.8348, + "step": 1236 + }, + { + "epoch": 0.13081641285956006, + "grad_norm": 8.375, + "learning_rate": 4.358703312191684e-06, + "loss": 2.5305, + "step": 1237 + }, + { + "epoch": 0.13092216582064298, + "grad_norm": 8.1875, + "learning_rate": 4.362226920366456e-06, + "loss": 2.5551, + "step": 1238 + }, + { + "epoch": 0.1310279187817259, + "grad_norm": 8.1875, + "learning_rate": 4.365750528541227e-06, + "loss": 2.6681, + "step": 1239 + }, + { + "epoch": 0.1311336717428088, + "grad_norm": 7.21875, + "learning_rate": 4.369274136715998e-06, + "loss": 2.8741, + "step": 1240 + }, + { + "epoch": 0.13123942470389172, + "grad_norm": 8.0, + "learning_rate": 4.372797744890768e-06, + "loss": 2.8763, + "step": 1241 + }, + { + "epoch": 0.1313451776649746, + "grad_norm": 7.40625, + "learning_rate": 4.37632135306554e-06, + "loss": 2.8576, + "step": 1242 + }, + { + "epoch": 0.13145093062605753, + "grad_norm": 9.25, + "learning_rate": 4.379844961240311e-06, + "loss": 2.93, + "step": 1243 + }, + { + "epoch": 0.13155668358714043, + "grad_norm": 8.0625, + "learning_rate": 4.383368569415081e-06, + "loss": 2.5836, + "step": 1244 + }, + { + "epoch": 0.13166243654822335, + "grad_norm": 7.84375, + "learning_rate": 4.386892177589853e-06, + "loss": 2.8961, + "step": 1245 + }, + { + "epoch": 0.13176818950930627, + "grad_norm": 7.84375, + "learning_rate": 4.390415785764624e-06, + "loss": 2.8864, + "step": 1246 + }, + { + "epoch": 0.13187394247038917, + "grad_norm": 7.75, + "learning_rate": 4.393939393939394e-06, + "loss": 2.5684, + "step": 1247 + }, + { + "epoch": 0.1319796954314721, + "grad_norm": 9.0625, + "learning_rate": 4.397463002114165e-06, + "loss": 2.6399, + "step": 1248 + }, + { + "epoch": 0.13208544839255498, + "grad_norm": 7.96875, + "learning_rate": 4.400986610288937e-06, + "loss": 2.5755, + "step": 1249 + }, + { + "epoch": 0.1321912013536379, + "grad_norm": 8.625, + "learning_rate": 4.404510218463708e-06, + "loss": 2.9408, + "step": 1250 + }, + { + "epoch": 0.1322969543147208, + "grad_norm": 8.125, + "learning_rate": 4.408033826638478e-06, + "loss": 2.7385, + "step": 1251 + }, + { + "epoch": 0.13240270727580372, + "grad_norm": 7.6875, + "learning_rate": 4.411557434813249e-06, + "loss": 2.7033, + "step": 1252 + }, + { + "epoch": 0.13250846023688664, + "grad_norm": 8.3125, + "learning_rate": 4.415081042988021e-06, + "loss": 2.5923, + "step": 1253 + }, + { + "epoch": 0.13261421319796954, + "grad_norm": 8.125, + "learning_rate": 4.418604651162791e-06, + "loss": 2.526, + "step": 1254 + }, + { + "epoch": 0.13271996615905246, + "grad_norm": 7.90625, + "learning_rate": 4.422128259337562e-06, + "loss": 2.6422, + "step": 1255 + }, + { + "epoch": 0.13282571912013535, + "grad_norm": 8.0, + "learning_rate": 4.425651867512333e-06, + "loss": 2.7433, + "step": 1256 + }, + { + "epoch": 0.13293147208121828, + "grad_norm": 9.625, + "learning_rate": 4.429175475687104e-06, + "loss": 2.6049, + "step": 1257 + }, + { + "epoch": 0.13303722504230117, + "grad_norm": 7.3125, + "learning_rate": 4.432699083861875e-06, + "loss": 2.729, + "step": 1258 + }, + { + "epoch": 0.1331429780033841, + "grad_norm": 8.1875, + "learning_rate": 4.436222692036646e-06, + "loss": 2.7385, + "step": 1259 + }, + { + "epoch": 0.13324873096446702, + "grad_norm": 8.6875, + "learning_rate": 4.439746300211417e-06, + "loss": 2.8919, + "step": 1260 + }, + { + "epoch": 0.1333544839255499, + "grad_norm": 8.6875, + "learning_rate": 4.443269908386188e-06, + "loss": 2.782, + "step": 1261 + }, + { + "epoch": 0.13346023688663283, + "grad_norm": 7.34375, + "learning_rate": 4.446793516560959e-06, + "loss": 2.7439, + "step": 1262 + }, + { + "epoch": 0.13356598984771573, + "grad_norm": 8.0, + "learning_rate": 4.45031712473573e-06, + "loss": 2.5544, + "step": 1263 + }, + { + "epoch": 0.13367174280879865, + "grad_norm": 7.5625, + "learning_rate": 4.453840732910501e-06, + "loss": 2.6856, + "step": 1264 + }, + { + "epoch": 0.13377749576988154, + "grad_norm": 7.4375, + "learning_rate": 4.457364341085272e-06, + "loss": 2.6479, + "step": 1265 + }, + { + "epoch": 0.13388324873096447, + "grad_norm": 7.8125, + "learning_rate": 4.460887949260043e-06, + "loss": 2.549, + "step": 1266 + }, + { + "epoch": 0.1339890016920474, + "grad_norm": 8.1875, + "learning_rate": 4.464411557434814e-06, + "loss": 2.6799, + "step": 1267 + }, + { + "epoch": 0.13409475465313028, + "grad_norm": 7.625, + "learning_rate": 4.467935165609585e-06, + "loss": 2.649, + "step": 1268 + }, + { + "epoch": 0.1342005076142132, + "grad_norm": 7.46875, + "learning_rate": 4.471458773784356e-06, + "loss": 2.6709, + "step": 1269 + }, + { + "epoch": 0.1343062605752961, + "grad_norm": 7.4375, + "learning_rate": 4.474982381959127e-06, + "loss": 2.7394, + "step": 1270 + }, + { + "epoch": 0.13441201353637902, + "grad_norm": 7.90625, + "learning_rate": 4.478505990133898e-06, + "loss": 2.8942, + "step": 1271 + }, + { + "epoch": 0.13451776649746192, + "grad_norm": 8.625, + "learning_rate": 4.482029598308669e-06, + "loss": 2.7273, + "step": 1272 + }, + { + "epoch": 0.13462351945854484, + "grad_norm": 7.8125, + "learning_rate": 4.4855532064834396e-06, + "loss": 2.7531, + "step": 1273 + }, + { + "epoch": 0.13472927241962776, + "grad_norm": 8.0, + "learning_rate": 4.4890768146582106e-06, + "loss": 2.851, + "step": 1274 + }, + { + "epoch": 0.13483502538071065, + "grad_norm": 8.4375, + "learning_rate": 4.4926004228329815e-06, + "loss": 2.5123, + "step": 1275 + }, + { + "epoch": 0.13494077834179358, + "grad_norm": 7.71875, + "learning_rate": 4.4961240310077525e-06, + "loss": 2.8861, + "step": 1276 + }, + { + "epoch": 0.13504653130287647, + "grad_norm": 7.5625, + "learning_rate": 4.4996476391825235e-06, + "loss": 2.7067, + "step": 1277 + }, + { + "epoch": 0.1351522842639594, + "grad_norm": 7.90625, + "learning_rate": 4.5031712473572945e-06, + "loss": 2.9755, + "step": 1278 + }, + { + "epoch": 0.1352580372250423, + "grad_norm": 8.0625, + "learning_rate": 4.506694855532065e-06, + "loss": 2.5313, + "step": 1279 + }, + { + "epoch": 0.1353637901861252, + "grad_norm": 7.46875, + "learning_rate": 4.5102184637068365e-06, + "loss": 2.7308, + "step": 1280 + }, + { + "epoch": 0.13546954314720813, + "grad_norm": 7.8125, + "learning_rate": 4.5137420718816075e-06, + "loss": 2.7879, + "step": 1281 + }, + { + "epoch": 0.13557529610829103, + "grad_norm": 8.5625, + "learning_rate": 4.517265680056378e-06, + "loss": 2.4263, + "step": 1282 + }, + { + "epoch": 0.13568104906937395, + "grad_norm": 7.875, + "learning_rate": 4.520789288231149e-06, + "loss": 2.5521, + "step": 1283 + }, + { + "epoch": 0.13578680203045684, + "grad_norm": 7.96875, + "learning_rate": 4.5243128964059204e-06, + "loss": 2.7625, + "step": 1284 + }, + { + "epoch": 0.13589255499153977, + "grad_norm": 7.15625, + "learning_rate": 4.5278365045806914e-06, + "loss": 2.5772, + "step": 1285 + }, + { + "epoch": 0.13599830795262266, + "grad_norm": 8.75, + "learning_rate": 4.5313601127554616e-06, + "loss": 2.5985, + "step": 1286 + }, + { + "epoch": 0.13610406091370558, + "grad_norm": 7.6875, + "learning_rate": 4.5348837209302326e-06, + "loss": 2.7156, + "step": 1287 + }, + { + "epoch": 0.1362098138747885, + "grad_norm": 8.0, + "learning_rate": 4.538407329105004e-06, + "loss": 2.7148, + "step": 1288 + }, + { + "epoch": 0.1363155668358714, + "grad_norm": 7.8125, + "learning_rate": 4.5419309372797745e-06, + "loss": 2.6398, + "step": 1289 + }, + { + "epoch": 0.13642131979695432, + "grad_norm": 7.9375, + "learning_rate": 4.5454545454545455e-06, + "loss": 2.6423, + "step": 1290 + }, + { + "epoch": 0.13652707275803722, + "grad_norm": 7.28125, + "learning_rate": 4.548978153629317e-06, + "loss": 2.6743, + "step": 1291 + }, + { + "epoch": 0.13663282571912014, + "grad_norm": 8.0, + "learning_rate": 4.5525017618040875e-06, + "loss": 2.7262, + "step": 1292 + }, + { + "epoch": 0.13673857868020303, + "grad_norm": 7.46875, + "learning_rate": 4.5560253699788585e-06, + "loss": 2.6299, + "step": 1293 + }, + { + "epoch": 0.13684433164128595, + "grad_norm": 7.78125, + "learning_rate": 4.5595489781536295e-06, + "loss": 2.667, + "step": 1294 + }, + { + "epoch": 0.13695008460236888, + "grad_norm": 7.96875, + "learning_rate": 4.563072586328401e-06, + "loss": 2.963, + "step": 1295 + }, + { + "epoch": 0.13705583756345177, + "grad_norm": 8.0625, + "learning_rate": 4.5665961945031714e-06, + "loss": 2.6013, + "step": 1296 + }, + { + "epoch": 0.1371615905245347, + "grad_norm": 7.4375, + "learning_rate": 4.5701198026779424e-06, + "loss": 2.6017, + "step": 1297 + }, + { + "epoch": 0.1372673434856176, + "grad_norm": 7.90625, + "learning_rate": 4.573643410852713e-06, + "loss": 2.7391, + "step": 1298 + }, + { + "epoch": 0.1373730964467005, + "grad_norm": 8.75, + "learning_rate": 4.577167019027484e-06, + "loss": 2.9898, + "step": 1299 + }, + { + "epoch": 0.1374788494077834, + "grad_norm": 7.71875, + "learning_rate": 4.580690627202255e-06, + "loss": 2.7114, + "step": 1300 + }, + { + "epoch": 0.13758460236886633, + "grad_norm": 8.1875, + "learning_rate": 4.584214235377026e-06, + "loss": 2.5817, + "step": 1301 + }, + { + "epoch": 0.13769035532994925, + "grad_norm": 8.25, + "learning_rate": 4.587737843551797e-06, + "loss": 2.55, + "step": 1302 + }, + { + "epoch": 0.13779610829103214, + "grad_norm": 7.75, + "learning_rate": 4.591261451726568e-06, + "loss": 2.7329, + "step": 1303 + }, + { + "epoch": 0.13790186125211507, + "grad_norm": 7.8125, + "learning_rate": 4.594785059901339e-06, + "loss": 2.7533, + "step": 1304 + }, + { + "epoch": 0.13800761421319796, + "grad_norm": 8.5, + "learning_rate": 4.59830866807611e-06, + "loss": 2.7127, + "step": 1305 + }, + { + "epoch": 0.13811336717428088, + "grad_norm": 8.875, + "learning_rate": 4.601832276250881e-06, + "loss": 2.9427, + "step": 1306 + }, + { + "epoch": 0.13821912013536378, + "grad_norm": 7.40625, + "learning_rate": 4.605355884425652e-06, + "loss": 2.5343, + "step": 1307 + }, + { + "epoch": 0.1383248730964467, + "grad_norm": 8.3125, + "learning_rate": 4.608879492600423e-06, + "loss": 2.5954, + "step": 1308 + }, + { + "epoch": 0.13843062605752962, + "grad_norm": 7.65625, + "learning_rate": 4.612403100775194e-06, + "loss": 2.8486, + "step": 1309 + }, + { + "epoch": 0.13853637901861252, + "grad_norm": 8.0625, + "learning_rate": 4.615926708949965e-06, + "loss": 2.7423, + "step": 1310 + }, + { + "epoch": 0.13864213197969544, + "grad_norm": 7.625, + "learning_rate": 4.619450317124736e-06, + "loss": 2.7549, + "step": 1311 + }, + { + "epoch": 0.13874788494077833, + "grad_norm": 8.0, + "learning_rate": 4.622973925299507e-06, + "loss": 2.7928, + "step": 1312 + }, + { + "epoch": 0.13885363790186125, + "grad_norm": 7.84375, + "learning_rate": 4.626497533474278e-06, + "loss": 2.5699, + "step": 1313 + }, + { + "epoch": 0.13895939086294415, + "grad_norm": 7.9375, + "learning_rate": 4.630021141649049e-06, + "loss": 2.5818, + "step": 1314 + }, + { + "epoch": 0.13906514382402707, + "grad_norm": 11.25, + "learning_rate": 4.63354474982382e-06, + "loss": 2.7136, + "step": 1315 + }, + { + "epoch": 0.13917089678511, + "grad_norm": 7.875, + "learning_rate": 4.637068357998591e-06, + "loss": 2.6404, + "step": 1316 + }, + { + "epoch": 0.1392766497461929, + "grad_norm": 8.25, + "learning_rate": 4.640591966173361e-06, + "loss": 2.8471, + "step": 1317 + }, + { + "epoch": 0.1393824027072758, + "grad_norm": 8.5625, + "learning_rate": 4.644115574348133e-06, + "loss": 3.0231, + "step": 1318 + }, + { + "epoch": 0.1394881556683587, + "grad_norm": 8.1875, + "learning_rate": 4.647639182522904e-06, + "loss": 2.7721, + "step": 1319 + }, + { + "epoch": 0.13959390862944163, + "grad_norm": 11.75, + "learning_rate": 4.651162790697675e-06, + "loss": 2.6952, + "step": 1320 + }, + { + "epoch": 0.13969966159052452, + "grad_norm": 15.25, + "learning_rate": 4.654686398872445e-06, + "loss": 2.7662, + "step": 1321 + }, + { + "epoch": 0.13980541455160744, + "grad_norm": 9.1875, + "learning_rate": 4.658210007047217e-06, + "loss": 2.6883, + "step": 1322 + }, + { + "epoch": 0.13991116751269037, + "grad_norm": 7.78125, + "learning_rate": 4.661733615221988e-06, + "loss": 2.7262, + "step": 1323 + }, + { + "epoch": 0.14001692047377326, + "grad_norm": 7.875, + "learning_rate": 4.665257223396758e-06, + "loss": 2.6963, + "step": 1324 + }, + { + "epoch": 0.14012267343485618, + "grad_norm": 8.25, + "learning_rate": 4.668780831571529e-06, + "loss": 2.7961, + "step": 1325 + }, + { + "epoch": 0.14022842639593908, + "grad_norm": 8.5, + "learning_rate": 4.672304439746301e-06, + "loss": 2.8243, + "step": 1326 + }, + { + "epoch": 0.140334179357022, + "grad_norm": 7.625, + "learning_rate": 4.675828047921071e-06, + "loss": 2.653, + "step": 1327 + }, + { + "epoch": 0.1404399323181049, + "grad_norm": 8.4375, + "learning_rate": 4.679351656095842e-06, + "loss": 2.6839, + "step": 1328 + }, + { + "epoch": 0.14054568527918782, + "grad_norm": 7.96875, + "learning_rate": 4.682875264270613e-06, + "loss": 2.8615, + "step": 1329 + }, + { + "epoch": 0.14065143824027074, + "grad_norm": 9.875, + "learning_rate": 4.686398872445384e-06, + "loss": 2.5778, + "step": 1330 + }, + { + "epoch": 0.14075719120135363, + "grad_norm": 7.71875, + "learning_rate": 4.689922480620155e-06, + "loss": 2.7909, + "step": 1331 + }, + { + "epoch": 0.14086294416243655, + "grad_norm": 7.59375, + "learning_rate": 4.693446088794926e-06, + "loss": 2.5294, + "step": 1332 + }, + { + "epoch": 0.14096869712351945, + "grad_norm": 7.78125, + "learning_rate": 4.696969696969698e-06, + "loss": 2.6288, + "step": 1333 + }, + { + "epoch": 0.14107445008460237, + "grad_norm": 10.5, + "learning_rate": 4.700493305144468e-06, + "loss": 2.709, + "step": 1334 + }, + { + "epoch": 0.1411802030456853, + "grad_norm": 7.8125, + "learning_rate": 4.704016913319239e-06, + "loss": 2.7881, + "step": 1335 + }, + { + "epoch": 0.1412859560067682, + "grad_norm": 7.9375, + "learning_rate": 4.70754052149401e-06, + "loss": 2.7182, + "step": 1336 + }, + { + "epoch": 0.1413917089678511, + "grad_norm": 8.0, + "learning_rate": 4.711064129668781e-06, + "loss": 2.8109, + "step": 1337 + }, + { + "epoch": 0.141497461928934, + "grad_norm": 8.25, + "learning_rate": 4.714587737843552e-06, + "loss": 2.7543, + "step": 1338 + }, + { + "epoch": 0.14160321489001693, + "grad_norm": 8.375, + "learning_rate": 4.718111346018323e-06, + "loss": 2.7642, + "step": 1339 + }, + { + "epoch": 0.14170896785109982, + "grad_norm": 7.59375, + "learning_rate": 4.721634954193094e-06, + "loss": 2.7754, + "step": 1340 + }, + { + "epoch": 0.14181472081218274, + "grad_norm": 12.4375, + "learning_rate": 4.725158562367865e-06, + "loss": 2.7934, + "step": 1341 + }, + { + "epoch": 0.14192047377326567, + "grad_norm": 8.625, + "learning_rate": 4.728682170542636e-06, + "loss": 2.6764, + "step": 1342 + }, + { + "epoch": 0.14202622673434856, + "grad_norm": 8.75, + "learning_rate": 4.732205778717407e-06, + "loss": 2.5425, + "step": 1343 + }, + { + "epoch": 0.14213197969543148, + "grad_norm": 8.375, + "learning_rate": 4.735729386892178e-06, + "loss": 2.7148, + "step": 1344 + }, + { + "epoch": 0.14223773265651438, + "grad_norm": 8.125, + "learning_rate": 4.739252995066949e-06, + "loss": 2.565, + "step": 1345 + }, + { + "epoch": 0.1423434856175973, + "grad_norm": 12.75, + "learning_rate": 4.74277660324172e-06, + "loss": 2.7555, + "step": 1346 + }, + { + "epoch": 0.1424492385786802, + "grad_norm": 7.75, + "learning_rate": 4.746300211416491e-06, + "loss": 2.7761, + "step": 1347 + }, + { + "epoch": 0.14255499153976312, + "grad_norm": 7.53125, + "learning_rate": 4.749823819591262e-06, + "loss": 2.6045, + "step": 1348 + }, + { + "epoch": 0.14266074450084604, + "grad_norm": 7.40625, + "learning_rate": 4.753347427766033e-06, + "loss": 2.5231, + "step": 1349 + }, + { + "epoch": 0.14276649746192893, + "grad_norm": 8.5, + "learning_rate": 4.756871035940804e-06, + "loss": 2.7608, + "step": 1350 + }, + { + "epoch": 0.14287225042301185, + "grad_norm": 8.0, + "learning_rate": 4.760394644115575e-06, + "loss": 2.9908, + "step": 1351 + }, + { + "epoch": 0.14297800338409475, + "grad_norm": 8.5625, + "learning_rate": 4.763918252290345e-06, + "loss": 2.7377, + "step": 1352 + }, + { + "epoch": 0.14308375634517767, + "grad_norm": 8.3125, + "learning_rate": 4.767441860465117e-06, + "loss": 2.6636, + "step": 1353 + }, + { + "epoch": 0.14318950930626057, + "grad_norm": 8.0625, + "learning_rate": 4.770965468639888e-06, + "loss": 2.5658, + "step": 1354 + }, + { + "epoch": 0.1432952622673435, + "grad_norm": 8.5625, + "learning_rate": 4.774489076814658e-06, + "loss": 2.5432, + "step": 1355 + }, + { + "epoch": 0.1434010152284264, + "grad_norm": 9.0625, + "learning_rate": 4.77801268498943e-06, + "loss": 2.7954, + "step": 1356 + }, + { + "epoch": 0.1435067681895093, + "grad_norm": 8.0, + "learning_rate": 4.781536293164201e-06, + "loss": 2.6632, + "step": 1357 + }, + { + "epoch": 0.14361252115059223, + "grad_norm": 7.3125, + "learning_rate": 4.785059901338972e-06, + "loss": 2.5478, + "step": 1358 + }, + { + "epoch": 0.14371827411167512, + "grad_norm": 8.625, + "learning_rate": 4.788583509513742e-06, + "loss": 3.0223, + "step": 1359 + }, + { + "epoch": 0.14382402707275804, + "grad_norm": 8.3125, + "learning_rate": 4.792107117688514e-06, + "loss": 2.7543, + "step": 1360 + }, + { + "epoch": 0.14392978003384094, + "grad_norm": 7.625, + "learning_rate": 4.795630725863285e-06, + "loss": 2.7688, + "step": 1361 + }, + { + "epoch": 0.14403553299492386, + "grad_norm": 9.0625, + "learning_rate": 4.799154334038055e-06, + "loss": 2.8062, + "step": 1362 + }, + { + "epoch": 0.14414128595600678, + "grad_norm": 8.625, + "learning_rate": 4.802677942212826e-06, + "loss": 2.6389, + "step": 1363 + }, + { + "epoch": 0.14424703891708968, + "grad_norm": 8.375, + "learning_rate": 4.806201550387598e-06, + "loss": 2.8056, + "step": 1364 + }, + { + "epoch": 0.1443527918781726, + "grad_norm": 7.9375, + "learning_rate": 4.809725158562368e-06, + "loss": 2.6259, + "step": 1365 + }, + { + "epoch": 0.1444585448392555, + "grad_norm": 7.65625, + "learning_rate": 4.813248766737139e-06, + "loss": 2.5622, + "step": 1366 + }, + { + "epoch": 0.14456429780033841, + "grad_norm": 13.125, + "learning_rate": 4.81677237491191e-06, + "loss": 2.5381, + "step": 1367 + }, + { + "epoch": 0.1446700507614213, + "grad_norm": 8.5625, + "learning_rate": 4.820295983086682e-06, + "loss": 2.7305, + "step": 1368 + }, + { + "epoch": 0.14477580372250423, + "grad_norm": 8.25, + "learning_rate": 4.823819591261452e-06, + "loss": 2.6369, + "step": 1369 + }, + { + "epoch": 0.14488155668358715, + "grad_norm": 8.375, + "learning_rate": 4.827343199436223e-06, + "loss": 2.8085, + "step": 1370 + }, + { + "epoch": 0.14498730964467005, + "grad_norm": 8.8125, + "learning_rate": 4.830866807610994e-06, + "loss": 2.832, + "step": 1371 + }, + { + "epoch": 0.14509306260575297, + "grad_norm": 8.375, + "learning_rate": 4.834390415785765e-06, + "loss": 2.742, + "step": 1372 + }, + { + "epoch": 0.14519881556683586, + "grad_norm": 7.59375, + "learning_rate": 4.837914023960536e-06, + "loss": 2.7113, + "step": 1373 + }, + { + "epoch": 0.1453045685279188, + "grad_norm": 7.9375, + "learning_rate": 4.841437632135307e-06, + "loss": 2.7032, + "step": 1374 + }, + { + "epoch": 0.14541032148900168, + "grad_norm": 9.5, + "learning_rate": 4.844961240310078e-06, + "loss": 2.7028, + "step": 1375 + }, + { + "epoch": 0.1455160744500846, + "grad_norm": 9.4375, + "learning_rate": 4.848484848484849e-06, + "loss": 2.9236, + "step": 1376 + }, + { + "epoch": 0.14562182741116753, + "grad_norm": 8.0625, + "learning_rate": 4.85200845665962e-06, + "loss": 2.6987, + "step": 1377 + }, + { + "epoch": 0.14572758037225042, + "grad_norm": 8.3125, + "learning_rate": 4.855532064834391e-06, + "loss": 2.4757, + "step": 1378 + }, + { + "epoch": 0.14583333333333334, + "grad_norm": 8.5625, + "learning_rate": 4.859055673009162e-06, + "loss": 2.4638, + "step": 1379 + }, + { + "epoch": 0.14593908629441624, + "grad_norm": 8.4375, + "learning_rate": 4.862579281183933e-06, + "loss": 2.7321, + "step": 1380 + }, + { + "epoch": 0.14604483925549916, + "grad_norm": 7.625, + "learning_rate": 4.866102889358704e-06, + "loss": 2.7726, + "step": 1381 + }, + { + "epoch": 0.14615059221658205, + "grad_norm": 9.1875, + "learning_rate": 4.869626497533475e-06, + "loss": 2.6823, + "step": 1382 + }, + { + "epoch": 0.14625634517766498, + "grad_norm": 12.375, + "learning_rate": 4.873150105708246e-06, + "loss": 2.4739, + "step": 1383 + }, + { + "epoch": 0.1463620981387479, + "grad_norm": 7.8125, + "learning_rate": 4.876673713883017e-06, + "loss": 2.6064, + "step": 1384 + }, + { + "epoch": 0.1464678510998308, + "grad_norm": 8.125, + "learning_rate": 4.880197322057788e-06, + "loss": 2.8447, + "step": 1385 + }, + { + "epoch": 0.14657360406091371, + "grad_norm": 8.625, + "learning_rate": 4.883720930232559e-06, + "loss": 2.7365, + "step": 1386 + }, + { + "epoch": 0.1466793570219966, + "grad_norm": 9.6875, + "learning_rate": 4.88724453840733e-06, + "loss": 2.564, + "step": 1387 + }, + { + "epoch": 0.14678510998307953, + "grad_norm": 7.15625, + "learning_rate": 4.890768146582101e-06, + "loss": 2.5112, + "step": 1388 + }, + { + "epoch": 0.14689086294416243, + "grad_norm": 8.125, + "learning_rate": 4.894291754756872e-06, + "loss": 2.5092, + "step": 1389 + }, + { + "epoch": 0.14699661590524535, + "grad_norm": 9.8125, + "learning_rate": 4.897815362931642e-06, + "loss": 2.6767, + "step": 1390 + }, + { + "epoch": 0.14710236886632827, + "grad_norm": 8.25, + "learning_rate": 4.901338971106414e-06, + "loss": 2.7298, + "step": 1391 + }, + { + "epoch": 0.14720812182741116, + "grad_norm": 7.9375, + "learning_rate": 4.904862579281185e-06, + "loss": 2.6644, + "step": 1392 + }, + { + "epoch": 0.1473138747884941, + "grad_norm": 7.65625, + "learning_rate": 4.9083861874559556e-06, + "loss": 2.5967, + "step": 1393 + }, + { + "epoch": 0.14741962774957698, + "grad_norm": 7.625, + "learning_rate": 4.911909795630726e-06, + "loss": 2.5991, + "step": 1394 + }, + { + "epoch": 0.1475253807106599, + "grad_norm": 10.0625, + "learning_rate": 4.9154334038054975e-06, + "loss": 2.5522, + "step": 1395 + }, + { + "epoch": 0.1476311336717428, + "grad_norm": 8.125, + "learning_rate": 4.9189570119802685e-06, + "loss": 2.5648, + "step": 1396 + }, + { + "epoch": 0.14773688663282572, + "grad_norm": 7.53125, + "learning_rate": 4.922480620155039e-06, + "loss": 2.5002, + "step": 1397 + }, + { + "epoch": 0.14784263959390864, + "grad_norm": 8.875, + "learning_rate": 4.9260042283298105e-06, + "loss": 2.6319, + "step": 1398 + }, + { + "epoch": 0.14794839255499154, + "grad_norm": 8.375, + "learning_rate": 4.9295278365045815e-06, + "loss": 2.8477, + "step": 1399 + }, + { + "epoch": 0.14805414551607446, + "grad_norm": 8.1875, + "learning_rate": 4.933051444679352e-06, + "loss": 2.7283, + "step": 1400 + }, + { + "epoch": 0.14815989847715735, + "grad_norm": 8.75, + "learning_rate": 4.936575052854123e-06, + "loss": 2.6507, + "step": 1401 + }, + { + "epoch": 0.14826565143824028, + "grad_norm": 9.6875, + "learning_rate": 4.9400986610288945e-06, + "loss": 2.728, + "step": 1402 + }, + { + "epoch": 0.14837140439932317, + "grad_norm": 8.3125, + "learning_rate": 4.9436222692036655e-06, + "loss": 2.36, + "step": 1403 + }, + { + "epoch": 0.1484771573604061, + "grad_norm": 8.625, + "learning_rate": 4.947145877378436e-06, + "loss": 2.6348, + "step": 1404 + }, + { + "epoch": 0.14858291032148901, + "grad_norm": 8.4375, + "learning_rate": 4.950669485553207e-06, + "loss": 2.6236, + "step": 1405 + }, + { + "epoch": 0.1486886632825719, + "grad_norm": 9.0625, + "learning_rate": 4.954193093727978e-06, + "loss": 2.8273, + "step": 1406 + }, + { + "epoch": 0.14879441624365483, + "grad_norm": 9.4375, + "learning_rate": 4.9577167019027486e-06, + "loss": 2.9888, + "step": 1407 + }, + { + "epoch": 0.14890016920473773, + "grad_norm": 12.5625, + "learning_rate": 4.9612403100775195e-06, + "loss": 2.8279, + "step": 1408 + }, + { + "epoch": 0.14900592216582065, + "grad_norm": 7.65625, + "learning_rate": 4.9647639182522905e-06, + "loss": 2.7718, + "step": 1409 + }, + { + "epoch": 0.14911167512690354, + "grad_norm": 10.8125, + "learning_rate": 4.9682875264270615e-06, + "loss": 2.8803, + "step": 1410 + }, + { + "epoch": 0.14921742808798646, + "grad_norm": 7.84375, + "learning_rate": 4.9718111346018325e-06, + "loss": 2.8829, + "step": 1411 + }, + { + "epoch": 0.1493231810490694, + "grad_norm": 8.9375, + "learning_rate": 4.9753347427766035e-06, + "loss": 2.6811, + "step": 1412 + }, + { + "epoch": 0.14942893401015228, + "grad_norm": 8.1875, + "learning_rate": 4.9788583509513745e-06, + "loss": 2.8291, + "step": 1413 + }, + { + "epoch": 0.1495346869712352, + "grad_norm": 7.03125, + "learning_rate": 4.9823819591261455e-06, + "loss": 2.7168, + "step": 1414 + }, + { + "epoch": 0.1496404399323181, + "grad_norm": 7.59375, + "learning_rate": 4.9859055673009165e-06, + "loss": 2.6757, + "step": 1415 + }, + { + "epoch": 0.14974619289340102, + "grad_norm": 7.875, + "learning_rate": 4.9894291754756874e-06, + "loss": 2.5623, + "step": 1416 + }, + { + "epoch": 0.14985194585448391, + "grad_norm": 7.59375, + "learning_rate": 4.9929527836504584e-06, + "loss": 2.7724, + "step": 1417 + }, + { + "epoch": 0.14995769881556684, + "grad_norm": 7.96875, + "learning_rate": 4.996476391825229e-06, + "loss": 2.6863, + "step": 1418 + }, + { + "epoch": 0.15006345177664976, + "grad_norm": 8.3125, + "learning_rate": 5e-06, + "loss": 2.8476, + "step": 1419 + }, + { + "epoch": 0.15016920473773265, + "grad_norm": 7.9375, + "learning_rate": 4.999999994134262e-06, + "loss": 2.7508, + "step": 1420 + }, + { + "epoch": 0.15027495769881558, + "grad_norm": 10.5, + "learning_rate": 4.9999999765370455e-06, + "loss": 2.6764, + "step": 1421 + }, + { + "epoch": 0.15038071065989847, + "grad_norm": 7.75, + "learning_rate": 4.999999947208352e-06, + "loss": 2.5746, + "step": 1422 + }, + { + "epoch": 0.1504864636209814, + "grad_norm": 7.75, + "learning_rate": 4.999999906148181e-06, + "loss": 2.6933, + "step": 1423 + }, + { + "epoch": 0.1505922165820643, + "grad_norm": 8.4375, + "learning_rate": 4.999999853356533e-06, + "loss": 2.6968, + "step": 1424 + }, + { + "epoch": 0.1506979695431472, + "grad_norm": 7.96875, + "learning_rate": 4.999999788833408e-06, + "loss": 2.6568, + "step": 1425 + }, + { + "epoch": 0.15080372250423013, + "grad_norm": 9.75, + "learning_rate": 4.999999712578806e-06, + "loss": 2.7167, + "step": 1426 + }, + { + "epoch": 0.15090947546531303, + "grad_norm": 8.125, + "learning_rate": 4.999999624592728e-06, + "loss": 2.524, + "step": 1427 + }, + { + "epoch": 0.15101522842639595, + "grad_norm": 8.375, + "learning_rate": 4.9999995248751756e-06, + "loss": 2.7195, + "step": 1428 + }, + { + "epoch": 0.15112098138747884, + "grad_norm": 8.9375, + "learning_rate": 4.999999413426146e-06, + "loss": 2.8105, + "step": 1429 + }, + { + "epoch": 0.15122673434856176, + "grad_norm": 10.125, + "learning_rate": 4.999999290245643e-06, + "loss": 2.7081, + "step": 1430 + }, + { + "epoch": 0.15133248730964466, + "grad_norm": 8.375, + "learning_rate": 4.999999155333665e-06, + "loss": 2.5111, + "step": 1431 + }, + { + "epoch": 0.15143824027072758, + "grad_norm": 8.0625, + "learning_rate": 4.999999008690214e-06, + "loss": 2.8602, + "step": 1432 + }, + { + "epoch": 0.1515439932318105, + "grad_norm": 7.59375, + "learning_rate": 4.99999885031529e-06, + "loss": 2.5883, + "step": 1433 + }, + { + "epoch": 0.1516497461928934, + "grad_norm": 7.84375, + "learning_rate": 4.9999986802088934e-06, + "loss": 2.8359, + "step": 1434 + }, + { + "epoch": 0.15175549915397632, + "grad_norm": 8.5625, + "learning_rate": 4.999998498371026e-06, + "loss": 2.6529, + "step": 1435 + }, + { + "epoch": 0.15186125211505921, + "grad_norm": 8.125, + "learning_rate": 4.999998304801687e-06, + "loss": 2.5083, + "step": 1436 + }, + { + "epoch": 0.15196700507614214, + "grad_norm": 13.1875, + "learning_rate": 4.99999809950088e-06, + "loss": 2.7885, + "step": 1437 + }, + { + "epoch": 0.15207275803722503, + "grad_norm": 8.3125, + "learning_rate": 4.999997882468603e-06, + "loss": 2.5248, + "step": 1438 + }, + { + "epoch": 0.15217851099830795, + "grad_norm": 7.5625, + "learning_rate": 4.9999976537048596e-06, + "loss": 2.6041, + "step": 1439 + }, + { + "epoch": 0.15228426395939088, + "grad_norm": 10.5, + "learning_rate": 4.999997413209649e-06, + "loss": 2.6352, + "step": 1440 + }, + { + "epoch": 0.15239001692047377, + "grad_norm": 7.78125, + "learning_rate": 4.9999971609829735e-06, + "loss": 2.7355, + "step": 1441 + }, + { + "epoch": 0.1524957698815567, + "grad_norm": 8.5625, + "learning_rate": 4.999996897024833e-06, + "loss": 2.9625, + "step": 1442 + }, + { + "epoch": 0.1526015228426396, + "grad_norm": 8.625, + "learning_rate": 4.99999662133523e-06, + "loss": 2.8565, + "step": 1443 + }, + { + "epoch": 0.1527072758037225, + "grad_norm": 7.875, + "learning_rate": 4.999996333914165e-06, + "loss": 2.6452, + "step": 1444 + }, + { + "epoch": 0.1528130287648054, + "grad_norm": 7.90625, + "learning_rate": 4.99999603476164e-06, + "loss": 2.6305, + "step": 1445 + }, + { + "epoch": 0.15291878172588833, + "grad_norm": 8.875, + "learning_rate": 4.999995723877657e-06, + "loss": 2.6412, + "step": 1446 + }, + { + "epoch": 0.15302453468697125, + "grad_norm": 7.40625, + "learning_rate": 4.9999954012622144e-06, + "loss": 2.621, + "step": 1447 + }, + { + "epoch": 0.15313028764805414, + "grad_norm": 8.0625, + "learning_rate": 4.9999950669153174e-06, + "loss": 2.9465, + "step": 1448 + }, + { + "epoch": 0.15323604060913706, + "grad_norm": 7.90625, + "learning_rate": 4.999994720836966e-06, + "loss": 2.7272, + "step": 1449 + }, + { + "epoch": 0.15334179357021996, + "grad_norm": 7.34375, + "learning_rate": 4.9999943630271605e-06, + "loss": 2.6622, + "step": 1450 + }, + { + "epoch": 0.15344754653130288, + "grad_norm": 8.125, + "learning_rate": 4.999993993485906e-06, + "loss": 2.8811, + "step": 1451 + }, + { + "epoch": 0.15355329949238578, + "grad_norm": 7.84375, + "learning_rate": 4.9999936122132e-06, + "loss": 2.6973, + "step": 1452 + }, + { + "epoch": 0.1536590524534687, + "grad_norm": 7.46875, + "learning_rate": 4.999993219209047e-06, + "loss": 2.5626, + "step": 1453 + }, + { + "epoch": 0.15376480541455162, + "grad_norm": 7.5625, + "learning_rate": 4.99999281447345e-06, + "loss": 2.6739, + "step": 1454 + }, + { + "epoch": 0.15387055837563451, + "grad_norm": 7.46875, + "learning_rate": 4.999992398006407e-06, + "loss": 2.5709, + "step": 1455 + }, + { + "epoch": 0.15397631133671744, + "grad_norm": 7.59375, + "learning_rate": 4.999991969807923e-06, + "loss": 2.6133, + "step": 1456 + }, + { + "epoch": 0.15408206429780033, + "grad_norm": 8.4375, + "learning_rate": 4.999991529878e-06, + "loss": 2.721, + "step": 1457 + }, + { + "epoch": 0.15418781725888325, + "grad_norm": 8.375, + "learning_rate": 4.9999910782166384e-06, + "loss": 2.5463, + "step": 1458 + }, + { + "epoch": 0.15429357021996615, + "grad_norm": 7.75, + "learning_rate": 4.999990614823841e-06, + "loss": 2.58, + "step": 1459 + }, + { + "epoch": 0.15439932318104907, + "grad_norm": 7.875, + "learning_rate": 4.99999013969961e-06, + "loss": 2.6539, + "step": 1460 + }, + { + "epoch": 0.154505076142132, + "grad_norm": 8.0, + "learning_rate": 4.999989652843949e-06, + "loss": 2.7806, + "step": 1461 + }, + { + "epoch": 0.1546108291032149, + "grad_norm": 7.65625, + "learning_rate": 4.999989154256857e-06, + "loss": 2.4831, + "step": 1462 + }, + { + "epoch": 0.1547165820642978, + "grad_norm": 8.6875, + "learning_rate": 4.99998864393834e-06, + "loss": 2.8254, + "step": 1463 + }, + { + "epoch": 0.1548223350253807, + "grad_norm": 7.75, + "learning_rate": 4.999988121888397e-06, + "loss": 2.7524, + "step": 1464 + }, + { + "epoch": 0.15492808798646363, + "grad_norm": 7.1875, + "learning_rate": 4.9999875881070335e-06, + "loss": 2.6158, + "step": 1465 + }, + { + "epoch": 0.15503384094754652, + "grad_norm": 7.625, + "learning_rate": 4.999987042594251e-06, + "loss": 2.6912, + "step": 1466 + }, + { + "epoch": 0.15513959390862944, + "grad_norm": 9.5625, + "learning_rate": 4.999986485350051e-06, + "loss": 2.7687, + "step": 1467 + }, + { + "epoch": 0.15524534686971236, + "grad_norm": 7.78125, + "learning_rate": 4.999985916374437e-06, + "loss": 2.5256, + "step": 1468 + }, + { + "epoch": 0.15535109983079526, + "grad_norm": 7.71875, + "learning_rate": 4.999985335667412e-06, + "loss": 2.7488, + "step": 1469 + }, + { + "epoch": 0.15545685279187818, + "grad_norm": 8.0, + "learning_rate": 4.999984743228978e-06, + "loss": 2.5637, + "step": 1470 + }, + { + "epoch": 0.15556260575296108, + "grad_norm": 8.0, + "learning_rate": 4.999984139059139e-06, + "loss": 2.6603, + "step": 1471 + }, + { + "epoch": 0.155668358714044, + "grad_norm": 8.125, + "learning_rate": 4.9999835231578954e-06, + "loss": 2.6939, + "step": 1472 + }, + { + "epoch": 0.1557741116751269, + "grad_norm": 7.5625, + "learning_rate": 4.999982895525252e-06, + "loss": 2.6627, + "step": 1473 + }, + { + "epoch": 0.15587986463620981, + "grad_norm": 8.9375, + "learning_rate": 4.999982256161212e-06, + "loss": 2.911, + "step": 1474 + }, + { + "epoch": 0.15598561759729274, + "grad_norm": 8.75, + "learning_rate": 4.999981605065777e-06, + "loss": 2.5405, + "step": 1475 + }, + { + "epoch": 0.15609137055837563, + "grad_norm": 11.375, + "learning_rate": 4.9999809422389515e-06, + "loss": 2.6548, + "step": 1476 + }, + { + "epoch": 0.15619712351945855, + "grad_norm": 8.0625, + "learning_rate": 4.999980267680736e-06, + "loss": 2.7041, + "step": 1477 + }, + { + "epoch": 0.15630287648054145, + "grad_norm": 8.625, + "learning_rate": 4.999979581391138e-06, + "loss": 2.6708, + "step": 1478 + }, + { + "epoch": 0.15640862944162437, + "grad_norm": 7.75, + "learning_rate": 4.999978883370157e-06, + "loss": 2.7874, + "step": 1479 + }, + { + "epoch": 0.15651438240270726, + "grad_norm": 8.75, + "learning_rate": 4.999978173617798e-06, + "loss": 2.5681, + "step": 1480 + }, + { + "epoch": 0.15662013536379019, + "grad_norm": 9.0625, + "learning_rate": 4.999977452134064e-06, + "loss": 2.6744, + "step": 1481 + }, + { + "epoch": 0.1567258883248731, + "grad_norm": 7.6875, + "learning_rate": 4.999976718918957e-06, + "loss": 2.5081, + "step": 1482 + }, + { + "epoch": 0.156831641285956, + "grad_norm": 8.1875, + "learning_rate": 4.999975973972483e-06, + "loss": 2.8534, + "step": 1483 + }, + { + "epoch": 0.15693739424703892, + "grad_norm": 8.75, + "learning_rate": 4.9999752172946436e-06, + "loss": 2.735, + "step": 1484 + }, + { + "epoch": 0.15704314720812182, + "grad_norm": 8.0625, + "learning_rate": 4.999974448885443e-06, + "loss": 2.7182, + "step": 1485 + }, + { + "epoch": 0.15714890016920474, + "grad_norm": 7.46875, + "learning_rate": 4.999973668744885e-06, + "loss": 2.5844, + "step": 1486 + }, + { + "epoch": 0.15725465313028764, + "grad_norm": 9.6875, + "learning_rate": 4.999972876872974e-06, + "loss": 2.8411, + "step": 1487 + }, + { + "epoch": 0.15736040609137056, + "grad_norm": 8.375, + "learning_rate": 4.999972073269711e-06, + "loss": 2.7027, + "step": 1488 + }, + { + "epoch": 0.15746615905245348, + "grad_norm": 7.9375, + "learning_rate": 4.999971257935102e-06, + "loss": 2.5852, + "step": 1489 + }, + { + "epoch": 0.15757191201353637, + "grad_norm": 7.75, + "learning_rate": 4.999970430869151e-06, + "loss": 2.4769, + "step": 1490 + }, + { + "epoch": 0.1576776649746193, + "grad_norm": 7.71875, + "learning_rate": 4.999969592071861e-06, + "loss": 2.5642, + "step": 1491 + }, + { + "epoch": 0.1577834179357022, + "grad_norm": 8.0625, + "learning_rate": 4.9999687415432365e-06, + "loss": 2.6102, + "step": 1492 + }, + { + "epoch": 0.1578891708967851, + "grad_norm": 8.375, + "learning_rate": 4.9999678792832805e-06, + "loss": 2.6825, + "step": 1493 + }, + { + "epoch": 0.157994923857868, + "grad_norm": 7.71875, + "learning_rate": 4.999967005291998e-06, + "loss": 2.8423, + "step": 1494 + }, + { + "epoch": 0.15810067681895093, + "grad_norm": 7.84375, + "learning_rate": 4.9999661195693925e-06, + "loss": 2.6997, + "step": 1495 + }, + { + "epoch": 0.15820642978003385, + "grad_norm": 7.65625, + "learning_rate": 4.999965222115468e-06, + "loss": 2.8613, + "step": 1496 + }, + { + "epoch": 0.15831218274111675, + "grad_norm": 8.4375, + "learning_rate": 4.99996431293023e-06, + "loss": 2.6945, + "step": 1497 + }, + { + "epoch": 0.15841793570219967, + "grad_norm": 8.5625, + "learning_rate": 4.999963392013681e-06, + "loss": 2.6467, + "step": 1498 + }, + { + "epoch": 0.15852368866328256, + "grad_norm": 8.375, + "learning_rate": 4.9999624593658275e-06, + "loss": 2.6817, + "step": 1499 + }, + { + "epoch": 0.15862944162436549, + "grad_norm": 10.75, + "learning_rate": 4.999961514986672e-06, + "loss": 2.8445, + "step": 1500 + }, + { + "epoch": 0.15873519458544838, + "grad_norm": 8.4375, + "learning_rate": 4.99996055887622e-06, + "loss": 2.7665, + "step": 1501 + }, + { + "epoch": 0.1588409475465313, + "grad_norm": 8.5, + "learning_rate": 4.999959591034475e-06, + "loss": 2.7048, + "step": 1502 + }, + { + "epoch": 0.15894670050761422, + "grad_norm": 8.1875, + "learning_rate": 4.999958611461443e-06, + "loss": 2.4344, + "step": 1503 + }, + { + "epoch": 0.15905245346869712, + "grad_norm": 11.5, + "learning_rate": 4.999957620157127e-06, + "loss": 2.6835, + "step": 1504 + }, + { + "epoch": 0.15915820642978004, + "grad_norm": 10.1875, + "learning_rate": 4.9999566171215315e-06, + "loss": 2.7109, + "step": 1505 + }, + { + "epoch": 0.15926395939086294, + "grad_norm": 7.8125, + "learning_rate": 4.999955602354662e-06, + "loss": 2.4957, + "step": 1506 + }, + { + "epoch": 0.15936971235194586, + "grad_norm": 8.9375, + "learning_rate": 4.999954575856525e-06, + "loss": 2.6775, + "step": 1507 + }, + { + "epoch": 0.15947546531302875, + "grad_norm": 12.0, + "learning_rate": 4.999953537627123e-06, + "loss": 2.5477, + "step": 1508 + }, + { + "epoch": 0.15958121827411167, + "grad_norm": 16.875, + "learning_rate": 4.9999524876664605e-06, + "loss": 2.8145, + "step": 1509 + }, + { + "epoch": 0.1596869712351946, + "grad_norm": 7.53125, + "learning_rate": 4.999951425974544e-06, + "loss": 2.5811, + "step": 1510 + }, + { + "epoch": 0.1597927241962775, + "grad_norm": 7.71875, + "learning_rate": 4.9999503525513785e-06, + "loss": 2.5546, + "step": 1511 + }, + { + "epoch": 0.1598984771573604, + "grad_norm": 7.96875, + "learning_rate": 4.999949267396968e-06, + "loss": 2.5886, + "step": 1512 + }, + { + "epoch": 0.1600042301184433, + "grad_norm": 8.5625, + "learning_rate": 4.9999481705113184e-06, + "loss": 2.7139, + "step": 1513 + }, + { + "epoch": 0.16010998307952623, + "grad_norm": 7.21875, + "learning_rate": 4.999947061894434e-06, + "loss": 2.5581, + "step": 1514 + }, + { + "epoch": 0.16021573604060912, + "grad_norm": 7.8125, + "learning_rate": 4.99994594154632e-06, + "loss": 2.7342, + "step": 1515 + }, + { + "epoch": 0.16032148900169205, + "grad_norm": 12.125, + "learning_rate": 4.999944809466983e-06, + "loss": 2.8523, + "step": 1516 + }, + { + "epoch": 0.16042724196277497, + "grad_norm": 8.0, + "learning_rate": 4.999943665656427e-06, + "loss": 2.7413, + "step": 1517 + }, + { + "epoch": 0.16053299492385786, + "grad_norm": 17.125, + "learning_rate": 4.999942510114658e-06, + "loss": 2.6443, + "step": 1518 + }, + { + "epoch": 0.16063874788494079, + "grad_norm": 8.3125, + "learning_rate": 4.999941342841682e-06, + "loss": 3.0184, + "step": 1519 + }, + { + "epoch": 0.16074450084602368, + "grad_norm": 8.25, + "learning_rate": 4.9999401638375024e-06, + "loss": 2.8065, + "step": 1520 + }, + { + "epoch": 0.1608502538071066, + "grad_norm": 7.53125, + "learning_rate": 4.999938973102127e-06, + "loss": 2.5967, + "step": 1521 + }, + { + "epoch": 0.1609560067681895, + "grad_norm": 7.25, + "learning_rate": 4.99993777063556e-06, + "loss": 2.6966, + "step": 1522 + }, + { + "epoch": 0.16106175972927242, + "grad_norm": 14.1875, + "learning_rate": 4.999936556437808e-06, + "loss": 2.6269, + "step": 1523 + }, + { + "epoch": 0.16116751269035534, + "grad_norm": 8.125, + "learning_rate": 4.999935330508875e-06, + "loss": 2.48, + "step": 1524 + }, + { + "epoch": 0.16127326565143824, + "grad_norm": 8.375, + "learning_rate": 4.999934092848769e-06, + "loss": 2.9399, + "step": 1525 + }, + { + "epoch": 0.16137901861252116, + "grad_norm": 9.1875, + "learning_rate": 4.999932843457494e-06, + "loss": 2.5868, + "step": 1526 + }, + { + "epoch": 0.16148477157360405, + "grad_norm": 9.25, + "learning_rate": 4.999931582335057e-06, + "loss": 2.7436, + "step": 1527 + }, + { + "epoch": 0.16159052453468697, + "grad_norm": 8.1875, + "learning_rate": 4.999930309481464e-06, + "loss": 2.7797, + "step": 1528 + }, + { + "epoch": 0.16169627749576987, + "grad_norm": 8.25, + "learning_rate": 4.99992902489672e-06, + "loss": 2.7464, + "step": 1529 + }, + { + "epoch": 0.1618020304568528, + "grad_norm": 8.125, + "learning_rate": 4.999927728580831e-06, + "loss": 2.7446, + "step": 1530 + }, + { + "epoch": 0.1619077834179357, + "grad_norm": 7.625, + "learning_rate": 4.999926420533805e-06, + "loss": 2.6963, + "step": 1531 + }, + { + "epoch": 0.1620135363790186, + "grad_norm": 8.4375, + "learning_rate": 4.999925100755645e-06, + "loss": 2.8916, + "step": 1532 + }, + { + "epoch": 0.16211928934010153, + "grad_norm": 8.0, + "learning_rate": 4.999923769246361e-06, + "loss": 2.9015, + "step": 1533 + }, + { + "epoch": 0.16222504230118442, + "grad_norm": 8.5, + "learning_rate": 4.9999224260059555e-06, + "loss": 2.7983, + "step": 1534 + }, + { + "epoch": 0.16233079526226735, + "grad_norm": 8.625, + "learning_rate": 4.999921071034437e-06, + "loss": 2.644, + "step": 1535 + }, + { + "epoch": 0.16243654822335024, + "grad_norm": 7.875, + "learning_rate": 4.999919704331811e-06, + "loss": 2.4066, + "step": 1536 + }, + { + "epoch": 0.16254230118443316, + "grad_norm": 8.4375, + "learning_rate": 4.999918325898084e-06, + "loss": 2.7636, + "step": 1537 + }, + { + "epoch": 0.16264805414551609, + "grad_norm": 8.0625, + "learning_rate": 4.999916935733264e-06, + "loss": 2.8773, + "step": 1538 + }, + { + "epoch": 0.16275380710659898, + "grad_norm": 8.75, + "learning_rate": 4.999915533837355e-06, + "loss": 2.6949, + "step": 1539 + }, + { + "epoch": 0.1628595600676819, + "grad_norm": 8.8125, + "learning_rate": 4.999914120210366e-06, + "loss": 2.6093, + "step": 1540 + }, + { + "epoch": 0.1629653130287648, + "grad_norm": 9.75, + "learning_rate": 4.999912694852302e-06, + "loss": 2.5288, + "step": 1541 + }, + { + "epoch": 0.16307106598984772, + "grad_norm": 7.375, + "learning_rate": 4.999911257763171e-06, + "loss": 2.5331, + "step": 1542 + }, + { + "epoch": 0.1631768189509306, + "grad_norm": 7.46875, + "learning_rate": 4.999909808942977e-06, + "loss": 2.5996, + "step": 1543 + }, + { + "epoch": 0.16328257191201354, + "grad_norm": 7.28125, + "learning_rate": 4.99990834839173e-06, + "loss": 2.636, + "step": 1544 + }, + { + "epoch": 0.16338832487309646, + "grad_norm": 8.1875, + "learning_rate": 4.999906876109435e-06, + "loss": 2.763, + "step": 1545 + }, + { + "epoch": 0.16349407783417935, + "grad_norm": 7.53125, + "learning_rate": 4.9999053920960995e-06, + "loss": 2.506, + "step": 1546 + }, + { + "epoch": 0.16359983079526227, + "grad_norm": 7.5, + "learning_rate": 4.99990389635173e-06, + "loss": 2.5631, + "step": 1547 + }, + { + "epoch": 0.16370558375634517, + "grad_norm": 8.0625, + "learning_rate": 4.999902388876335e-06, + "loss": 2.7089, + "step": 1548 + }, + { + "epoch": 0.1638113367174281, + "grad_norm": 8.625, + "learning_rate": 4.99990086966992e-06, + "loss": 2.7466, + "step": 1549 + }, + { + "epoch": 0.16391708967851099, + "grad_norm": 8.1875, + "learning_rate": 4.999899338732492e-06, + "loss": 2.5365, + "step": 1550 + }, + { + "epoch": 0.1640228426395939, + "grad_norm": 9.125, + "learning_rate": 4.99989779606406e-06, + "loss": 2.4627, + "step": 1551 + }, + { + "epoch": 0.16412859560067683, + "grad_norm": 8.4375, + "learning_rate": 4.99989624166463e-06, + "loss": 2.579, + "step": 1552 + }, + { + "epoch": 0.16423434856175972, + "grad_norm": 8.375, + "learning_rate": 4.999894675534208e-06, + "loss": 2.7176, + "step": 1553 + }, + { + "epoch": 0.16434010152284265, + "grad_norm": 8.5, + "learning_rate": 4.999893097672804e-06, + "loss": 3.0293, + "step": 1554 + }, + { + "epoch": 0.16444585448392554, + "grad_norm": 8.9375, + "learning_rate": 4.9998915080804235e-06, + "loss": 2.6593, + "step": 1555 + }, + { + "epoch": 0.16455160744500846, + "grad_norm": 9.625, + "learning_rate": 4.999889906757075e-06, + "loss": 2.5995, + "step": 1556 + }, + { + "epoch": 0.16465736040609136, + "grad_norm": 7.78125, + "learning_rate": 4.999888293702765e-06, + "loss": 2.8143, + "step": 1557 + }, + { + "epoch": 0.16476311336717428, + "grad_norm": 8.1875, + "learning_rate": 4.999886668917502e-06, + "loss": 2.6467, + "step": 1558 + }, + { + "epoch": 0.1648688663282572, + "grad_norm": 8.125, + "learning_rate": 4.999885032401294e-06, + "loss": 2.8067, + "step": 1559 + }, + { + "epoch": 0.1649746192893401, + "grad_norm": 7.71875, + "learning_rate": 4.999883384154147e-06, + "loss": 2.7995, + "step": 1560 + }, + { + "epoch": 0.16508037225042302, + "grad_norm": 8.0, + "learning_rate": 4.99988172417607e-06, + "loss": 2.6127, + "step": 1561 + }, + { + "epoch": 0.1651861252115059, + "grad_norm": 7.46875, + "learning_rate": 4.999880052467071e-06, + "loss": 2.444, + "step": 1562 + }, + { + "epoch": 0.16529187817258884, + "grad_norm": 8.375, + "learning_rate": 4.999878369027157e-06, + "loss": 2.8738, + "step": 1563 + }, + { + "epoch": 0.16539763113367173, + "grad_norm": 8.4375, + "learning_rate": 4.999876673856336e-06, + "loss": 2.9185, + "step": 1564 + }, + { + "epoch": 0.16550338409475465, + "grad_norm": 7.90625, + "learning_rate": 4.999874966954616e-06, + "loss": 2.5639, + "step": 1565 + }, + { + "epoch": 0.16560913705583757, + "grad_norm": 8.1875, + "learning_rate": 4.999873248322004e-06, + "loss": 2.3616, + "step": 1566 + }, + { + "epoch": 0.16571489001692047, + "grad_norm": 7.875, + "learning_rate": 4.999871517958512e-06, + "loss": 2.6849, + "step": 1567 + }, + { + "epoch": 0.1658206429780034, + "grad_norm": 8.4375, + "learning_rate": 4.999869775864143e-06, + "loss": 2.4691, + "step": 1568 + }, + { + "epoch": 0.16592639593908629, + "grad_norm": 10.5, + "learning_rate": 4.99986802203891e-06, + "loss": 2.6754, + "step": 1569 + }, + { + "epoch": 0.1660321489001692, + "grad_norm": 8.875, + "learning_rate": 4.999866256482817e-06, + "loss": 2.6315, + "step": 1570 + }, + { + "epoch": 0.1661379018612521, + "grad_norm": 8.5625, + "learning_rate": 4.999864479195874e-06, + "loss": 2.8528, + "step": 1571 + }, + { + "epoch": 0.16624365482233502, + "grad_norm": 10.0625, + "learning_rate": 4.99986269017809e-06, + "loss": 2.822, + "step": 1572 + }, + { + "epoch": 0.16634940778341795, + "grad_norm": 8.5, + "learning_rate": 4.999860889429473e-06, + "loss": 2.7644, + "step": 1573 + }, + { + "epoch": 0.16645516074450084, + "grad_norm": 7.59375, + "learning_rate": 4.999859076950031e-06, + "loss": 2.6249, + "step": 1574 + }, + { + "epoch": 0.16656091370558376, + "grad_norm": 7.84375, + "learning_rate": 4.999857252739773e-06, + "loss": 2.7502, + "step": 1575 + }, + { + "epoch": 0.16666666666666666, + "grad_norm": 8.0, + "learning_rate": 4.9998554167987066e-06, + "loss": 2.9168, + "step": 1576 + }, + { + "epoch": 0.16677241962774958, + "grad_norm": 9.3125, + "learning_rate": 4.999853569126841e-06, + "loss": 3.0142, + "step": 1577 + }, + { + "epoch": 0.16687817258883247, + "grad_norm": 8.625, + "learning_rate": 4.999851709724186e-06, + "loss": 2.6578, + "step": 1578 + }, + { + "epoch": 0.1669839255499154, + "grad_norm": 8.0, + "learning_rate": 4.9998498385907484e-06, + "loss": 2.5975, + "step": 1579 + }, + { + "epoch": 0.16708967851099832, + "grad_norm": 7.5625, + "learning_rate": 4.999847955726539e-06, + "loss": 2.7658, + "step": 1580 + }, + { + "epoch": 0.1671954314720812, + "grad_norm": 8.125, + "learning_rate": 4.999846061131565e-06, + "loss": 2.3176, + "step": 1581 + }, + { + "epoch": 0.16730118443316414, + "grad_norm": 7.78125, + "learning_rate": 4.999844154805835e-06, + "loss": 2.6715, + "step": 1582 + }, + { + "epoch": 0.16740693739424703, + "grad_norm": 7.84375, + "learning_rate": 4.99984223674936e-06, + "loss": 2.6697, + "step": 1583 + }, + { + "epoch": 0.16751269035532995, + "grad_norm": 8.5625, + "learning_rate": 4.999840306962147e-06, + "loss": 2.3422, + "step": 1584 + }, + { + "epoch": 0.16761844331641285, + "grad_norm": 7.75, + "learning_rate": 4.999838365444206e-06, + "loss": 2.6983, + "step": 1585 + }, + { + "epoch": 0.16772419627749577, + "grad_norm": 8.125, + "learning_rate": 4.999836412195546e-06, + "loss": 2.5833, + "step": 1586 + }, + { + "epoch": 0.1678299492385787, + "grad_norm": 7.59375, + "learning_rate": 4.9998344472161765e-06, + "loss": 2.5525, + "step": 1587 + }, + { + "epoch": 0.16793570219966159, + "grad_norm": 7.96875, + "learning_rate": 4.999832470506107e-06, + "loss": 2.7419, + "step": 1588 + }, + { + "epoch": 0.1680414551607445, + "grad_norm": 7.46875, + "learning_rate": 4.9998304820653446e-06, + "loss": 2.5492, + "step": 1589 + }, + { + "epoch": 0.1681472081218274, + "grad_norm": 8.0625, + "learning_rate": 4.9998284818939e-06, + "loss": 2.7188, + "step": 1590 + }, + { + "epoch": 0.16825296108291032, + "grad_norm": 7.625, + "learning_rate": 4.999826469991784e-06, + "loss": 2.5712, + "step": 1591 + }, + { + "epoch": 0.16835871404399322, + "grad_norm": 7.96875, + "learning_rate": 4.999824446359004e-06, + "loss": 2.6546, + "step": 1592 + }, + { + "epoch": 0.16846446700507614, + "grad_norm": 9.125, + "learning_rate": 4.999822410995571e-06, + "loss": 2.4818, + "step": 1593 + }, + { + "epoch": 0.16857021996615906, + "grad_norm": 8.4375, + "learning_rate": 4.999820363901493e-06, + "loss": 2.699, + "step": 1594 + }, + { + "epoch": 0.16867597292724196, + "grad_norm": 8.0625, + "learning_rate": 4.999818305076781e-06, + "loss": 2.8097, + "step": 1595 + }, + { + "epoch": 0.16878172588832488, + "grad_norm": 8.5625, + "learning_rate": 4.999816234521444e-06, + "loss": 2.6147, + "step": 1596 + }, + { + "epoch": 0.16888747884940777, + "grad_norm": 8.5, + "learning_rate": 4.999814152235492e-06, + "loss": 2.7679, + "step": 1597 + }, + { + "epoch": 0.1689932318104907, + "grad_norm": 7.90625, + "learning_rate": 4.999812058218934e-06, + "loss": 2.6793, + "step": 1598 + }, + { + "epoch": 0.1690989847715736, + "grad_norm": 9.25, + "learning_rate": 4.99980995247178e-06, + "loss": 2.6692, + "step": 1599 + }, + { + "epoch": 0.1692047377326565, + "grad_norm": 7.96875, + "learning_rate": 4.999807834994041e-06, + "loss": 2.6722, + "step": 1600 + }, + { + "epoch": 0.16931049069373943, + "grad_norm": 8.375, + "learning_rate": 4.999805705785725e-06, + "loss": 2.5867, + "step": 1601 + }, + { + "epoch": 0.16941624365482233, + "grad_norm": 8.375, + "learning_rate": 4.9998035648468455e-06, + "loss": 2.6295, + "step": 1602 + }, + { + "epoch": 0.16952199661590525, + "grad_norm": 8.4375, + "learning_rate": 4.999801412177409e-06, + "loss": 2.6622, + "step": 1603 + }, + { + "epoch": 0.16962774957698815, + "grad_norm": 7.03125, + "learning_rate": 4.999799247777427e-06, + "loss": 2.5771, + "step": 1604 + }, + { + "epoch": 0.16973350253807107, + "grad_norm": 7.875, + "learning_rate": 4.999797071646909e-06, + "loss": 2.6334, + "step": 1605 + }, + { + "epoch": 0.16983925549915396, + "grad_norm": 9.1875, + "learning_rate": 4.999794883785866e-06, + "loss": 2.6669, + "step": 1606 + }, + { + "epoch": 0.16994500846023688, + "grad_norm": 8.5, + "learning_rate": 4.999792684194309e-06, + "loss": 2.7212, + "step": 1607 + }, + { + "epoch": 0.1700507614213198, + "grad_norm": 9.75, + "learning_rate": 4.999790472872246e-06, + "loss": 2.824, + "step": 1608 + }, + { + "epoch": 0.1701565143824027, + "grad_norm": 8.0625, + "learning_rate": 4.999788249819689e-06, + "loss": 2.5828, + "step": 1609 + }, + { + "epoch": 0.17026226734348562, + "grad_norm": 7.71875, + "learning_rate": 4.999786015036648e-06, + "loss": 2.6128, + "step": 1610 + }, + { + "epoch": 0.17036802030456852, + "grad_norm": 8.0, + "learning_rate": 4.999783768523135e-06, + "loss": 2.8638, + "step": 1611 + }, + { + "epoch": 0.17047377326565144, + "grad_norm": 7.84375, + "learning_rate": 4.999781510279157e-06, + "loss": 2.6961, + "step": 1612 + }, + { + "epoch": 0.17057952622673433, + "grad_norm": 9.25, + "learning_rate": 4.999779240304728e-06, + "loss": 2.6708, + "step": 1613 + }, + { + "epoch": 0.17068527918781726, + "grad_norm": 8.0, + "learning_rate": 4.9997769585998565e-06, + "loss": 2.5804, + "step": 1614 + }, + { + "epoch": 0.17079103214890018, + "grad_norm": 7.15625, + "learning_rate": 4.9997746651645555e-06, + "loss": 2.6759, + "step": 1615 + }, + { + "epoch": 0.17089678510998307, + "grad_norm": 7.9375, + "learning_rate": 4.999772359998833e-06, + "loss": 2.5964, + "step": 1616 + }, + { + "epoch": 0.171002538071066, + "grad_norm": 8.125, + "learning_rate": 4.999770043102702e-06, + "loss": 2.8566, + "step": 1617 + }, + { + "epoch": 0.1711082910321489, + "grad_norm": 8.125, + "learning_rate": 4.999767714476173e-06, + "loss": 2.5802, + "step": 1618 + }, + { + "epoch": 0.1712140439932318, + "grad_norm": 7.6875, + "learning_rate": 4.999765374119255e-06, + "loss": 2.6432, + "step": 1619 + }, + { + "epoch": 0.1713197969543147, + "grad_norm": 9.1875, + "learning_rate": 4.999763022031962e-06, + "loss": 2.6454, + "step": 1620 + }, + { + "epoch": 0.17142554991539763, + "grad_norm": 10.9375, + "learning_rate": 4.999760658214303e-06, + "loss": 2.6601, + "step": 1621 + }, + { + "epoch": 0.17153130287648055, + "grad_norm": 8.5, + "learning_rate": 4.99975828266629e-06, + "loss": 2.6006, + "step": 1622 + }, + { + "epoch": 0.17163705583756345, + "grad_norm": 9.375, + "learning_rate": 4.999755895387932e-06, + "loss": 2.8594, + "step": 1623 + }, + { + "epoch": 0.17174280879864637, + "grad_norm": 8.6875, + "learning_rate": 4.999753496379244e-06, + "loss": 2.8462, + "step": 1624 + }, + { + "epoch": 0.17184856175972926, + "grad_norm": 9.8125, + "learning_rate": 4.9997510856402335e-06, + "loss": 2.7933, + "step": 1625 + }, + { + "epoch": 0.17195431472081218, + "grad_norm": 7.8125, + "learning_rate": 4.999748663170915e-06, + "loss": 2.5908, + "step": 1626 + }, + { + "epoch": 0.1720600676818951, + "grad_norm": 8.9375, + "learning_rate": 4.999746228971297e-06, + "loss": 2.6275, + "step": 1627 + }, + { + "epoch": 0.172165820642978, + "grad_norm": 8.5625, + "learning_rate": 4.9997437830413935e-06, + "loss": 2.5869, + "step": 1628 + }, + { + "epoch": 0.17227157360406092, + "grad_norm": 9.0, + "learning_rate": 4.9997413253812144e-06, + "loss": 2.725, + "step": 1629 + }, + { + "epoch": 0.17237732656514382, + "grad_norm": 8.6875, + "learning_rate": 4.999738855990771e-06, + "loss": 2.7387, + "step": 1630 + }, + { + "epoch": 0.17248307952622674, + "grad_norm": 8.75, + "learning_rate": 4.999736374870077e-06, + "loss": 2.6658, + "step": 1631 + }, + { + "epoch": 0.17258883248730963, + "grad_norm": 9.25, + "learning_rate": 4.999733882019141e-06, + "loss": 2.53, + "step": 1632 + }, + { + "epoch": 0.17269458544839256, + "grad_norm": 10.1875, + "learning_rate": 4.999731377437977e-06, + "loss": 2.3471, + "step": 1633 + }, + { + "epoch": 0.17280033840947548, + "grad_norm": 7.4375, + "learning_rate": 4.999728861126596e-06, + "loss": 2.612, + "step": 1634 + }, + { + "epoch": 0.17290609137055837, + "grad_norm": 8.5, + "learning_rate": 4.99972633308501e-06, + "loss": 2.559, + "step": 1635 + }, + { + "epoch": 0.1730118443316413, + "grad_norm": 7.65625, + "learning_rate": 4.99972379331323e-06, + "loss": 2.6755, + "step": 1636 + }, + { + "epoch": 0.1731175972927242, + "grad_norm": 8.0, + "learning_rate": 4.999721241811269e-06, + "loss": 2.7616, + "step": 1637 + }, + { + "epoch": 0.1732233502538071, + "grad_norm": 8.5625, + "learning_rate": 4.999718678579138e-06, + "loss": 2.6353, + "step": 1638 + }, + { + "epoch": 0.17332910321489, + "grad_norm": 10.4375, + "learning_rate": 4.99971610361685e-06, + "loss": 2.4341, + "step": 1639 + }, + { + "epoch": 0.17343485617597293, + "grad_norm": 8.125, + "learning_rate": 4.999713516924417e-06, + "loss": 2.6153, + "step": 1640 + }, + { + "epoch": 0.17354060913705585, + "grad_norm": 9.3125, + "learning_rate": 4.999710918501851e-06, + "loss": 2.7743, + "step": 1641 + }, + { + "epoch": 0.17364636209813875, + "grad_norm": 7.78125, + "learning_rate": 4.999708308349164e-06, + "loss": 2.7892, + "step": 1642 + }, + { + "epoch": 0.17375211505922167, + "grad_norm": 8.9375, + "learning_rate": 4.999705686466367e-06, + "loss": 2.6434, + "step": 1643 + }, + { + "epoch": 0.17385786802030456, + "grad_norm": 9.5, + "learning_rate": 4.999703052853475e-06, + "loss": 2.6044, + "step": 1644 + }, + { + "epoch": 0.17396362098138748, + "grad_norm": 8.4375, + "learning_rate": 4.999700407510498e-06, + "loss": 2.4498, + "step": 1645 + }, + { + "epoch": 0.17406937394247038, + "grad_norm": 12.125, + "learning_rate": 4.99969775043745e-06, + "loss": 2.7015, + "step": 1646 + }, + { + "epoch": 0.1741751269035533, + "grad_norm": 7.3125, + "learning_rate": 4.999695081634342e-06, + "loss": 2.6189, + "step": 1647 + }, + { + "epoch": 0.17428087986463622, + "grad_norm": 8.0625, + "learning_rate": 4.999692401101188e-06, + "loss": 2.7283, + "step": 1648 + }, + { + "epoch": 0.17438663282571912, + "grad_norm": 7.1875, + "learning_rate": 4.9996897088379995e-06, + "loss": 2.6846, + "step": 1649 + }, + { + "epoch": 0.17449238578680204, + "grad_norm": 7.9375, + "learning_rate": 4.99968700484479e-06, + "loss": 2.8329, + "step": 1650 + }, + { + "epoch": 0.17459813874788493, + "grad_norm": 8.8125, + "learning_rate": 4.999684289121571e-06, + "loss": 2.6934, + "step": 1651 + }, + { + "epoch": 0.17470389170896786, + "grad_norm": 8.25, + "learning_rate": 4.9996815616683565e-06, + "loss": 2.7133, + "step": 1652 + }, + { + "epoch": 0.17480964467005075, + "grad_norm": 8.5625, + "learning_rate": 4.999678822485159e-06, + "loss": 2.8549, + "step": 1653 + }, + { + "epoch": 0.17491539763113367, + "grad_norm": 8.25, + "learning_rate": 4.99967607157199e-06, + "loss": 2.7533, + "step": 1654 + }, + { + "epoch": 0.1750211505922166, + "grad_norm": 7.625, + "learning_rate": 4.9996733089288656e-06, + "loss": 2.6493, + "step": 1655 + }, + { + "epoch": 0.1751269035532995, + "grad_norm": 7.875, + "learning_rate": 4.999670534555795e-06, + "loss": 2.7397, + "step": 1656 + }, + { + "epoch": 0.1752326565143824, + "grad_norm": 10.125, + "learning_rate": 4.999667748452793e-06, + "loss": 2.7305, + "step": 1657 + }, + { + "epoch": 0.1753384094754653, + "grad_norm": 8.75, + "learning_rate": 4.999664950619873e-06, + "loss": 2.5299, + "step": 1658 + }, + { + "epoch": 0.17544416243654823, + "grad_norm": 8.5, + "learning_rate": 4.999662141057048e-06, + "loss": 2.5838, + "step": 1659 + }, + { + "epoch": 0.17554991539763112, + "grad_norm": 7.21875, + "learning_rate": 4.9996593197643304e-06, + "loss": 2.5698, + "step": 1660 + }, + { + "epoch": 0.17565566835871405, + "grad_norm": 8.1875, + "learning_rate": 4.999656486741734e-06, + "loss": 2.476, + "step": 1661 + }, + { + "epoch": 0.17576142131979697, + "grad_norm": 7.125, + "learning_rate": 4.999653641989271e-06, + "loss": 2.731, + "step": 1662 + }, + { + "epoch": 0.17586717428087986, + "grad_norm": 8.125, + "learning_rate": 4.999650785506958e-06, + "loss": 2.5861, + "step": 1663 + }, + { + "epoch": 0.17597292724196278, + "grad_norm": 8.4375, + "learning_rate": 4.999647917294805e-06, + "loss": 2.7823, + "step": 1664 + }, + { + "epoch": 0.17607868020304568, + "grad_norm": 9.0625, + "learning_rate": 4.999645037352826e-06, + "loss": 2.7968, + "step": 1665 + }, + { + "epoch": 0.1761844331641286, + "grad_norm": 7.65625, + "learning_rate": 4.999642145681036e-06, + "loss": 2.4996, + "step": 1666 + }, + { + "epoch": 0.1762901861252115, + "grad_norm": 8.125, + "learning_rate": 4.999639242279448e-06, + "loss": 2.7682, + "step": 1667 + }, + { + "epoch": 0.17639593908629442, + "grad_norm": 9.875, + "learning_rate": 4.9996363271480755e-06, + "loss": 2.6357, + "step": 1668 + }, + { + "epoch": 0.17650169204737734, + "grad_norm": 8.125, + "learning_rate": 4.99963340028693e-06, + "loss": 2.7505, + "step": 1669 + }, + { + "epoch": 0.17660744500846023, + "grad_norm": 8.4375, + "learning_rate": 4.99963046169603e-06, + "loss": 2.5237, + "step": 1670 + }, + { + "epoch": 0.17671319796954316, + "grad_norm": 7.96875, + "learning_rate": 4.999627511375385e-06, + "loss": 2.7432, + "step": 1671 + }, + { + "epoch": 0.17681895093062605, + "grad_norm": 8.125, + "learning_rate": 4.99962454932501e-06, + "loss": 2.7517, + "step": 1672 + }, + { + "epoch": 0.17692470389170897, + "grad_norm": 7.75, + "learning_rate": 4.99962157554492e-06, + "loss": 2.622, + "step": 1673 + }, + { + "epoch": 0.17703045685279187, + "grad_norm": 7.78125, + "learning_rate": 4.999618590035129e-06, + "loss": 2.7548, + "step": 1674 + }, + { + "epoch": 0.1771362098138748, + "grad_norm": 8.3125, + "learning_rate": 4.999615592795649e-06, + "loss": 2.7838, + "step": 1675 + }, + { + "epoch": 0.1772419627749577, + "grad_norm": 8.0625, + "learning_rate": 4.999612583826497e-06, + "loss": 2.7731, + "step": 1676 + }, + { + "epoch": 0.1773477157360406, + "grad_norm": 8.4375, + "learning_rate": 4.999609563127684e-06, + "loss": 2.604, + "step": 1677 + }, + { + "epoch": 0.17745346869712353, + "grad_norm": 7.96875, + "learning_rate": 4.999606530699225e-06, + "loss": 2.6009, + "step": 1678 + }, + { + "epoch": 0.17755922165820642, + "grad_norm": 8.375, + "learning_rate": 4.999603486541136e-06, + "loss": 2.4824, + "step": 1679 + }, + { + "epoch": 0.17766497461928935, + "grad_norm": 8.875, + "learning_rate": 4.9996004306534305e-06, + "loss": 2.8697, + "step": 1680 + }, + { + "epoch": 0.17777072758037224, + "grad_norm": 17.0, + "learning_rate": 4.999597363036122e-06, + "loss": 2.8188, + "step": 1681 + }, + { + "epoch": 0.17787648054145516, + "grad_norm": 10.5, + "learning_rate": 4.999594283689225e-06, + "loss": 2.5656, + "step": 1682 + }, + { + "epoch": 0.17798223350253808, + "grad_norm": 8.625, + "learning_rate": 4.999591192612754e-06, + "loss": 2.7476, + "step": 1683 + }, + { + "epoch": 0.17808798646362098, + "grad_norm": 7.625, + "learning_rate": 4.999588089806725e-06, + "loss": 2.6532, + "step": 1684 + }, + { + "epoch": 0.1781937394247039, + "grad_norm": 7.65625, + "learning_rate": 4.99958497527115e-06, + "loss": 2.6345, + "step": 1685 + }, + { + "epoch": 0.1782994923857868, + "grad_norm": 8.125, + "learning_rate": 4.999581849006046e-06, + "loss": 2.8961, + "step": 1686 + }, + { + "epoch": 0.17840524534686972, + "grad_norm": 10.25, + "learning_rate": 4.999578711011427e-06, + "loss": 2.6759, + "step": 1687 + }, + { + "epoch": 0.1785109983079526, + "grad_norm": 8.0625, + "learning_rate": 4.999575561287307e-06, + "loss": 2.6225, + "step": 1688 + }, + { + "epoch": 0.17861675126903553, + "grad_norm": 7.90625, + "learning_rate": 4.9995723998337005e-06, + "loss": 2.7891, + "step": 1689 + }, + { + "epoch": 0.17872250423011846, + "grad_norm": 8.0625, + "learning_rate": 4.999569226650625e-06, + "loss": 2.7405, + "step": 1690 + }, + { + "epoch": 0.17882825719120135, + "grad_norm": 8.875, + "learning_rate": 4.999566041738091e-06, + "loss": 2.7515, + "step": 1691 + }, + { + "epoch": 0.17893401015228427, + "grad_norm": 7.75, + "learning_rate": 4.999562845096117e-06, + "loss": 2.569, + "step": 1692 + }, + { + "epoch": 0.17903976311336717, + "grad_norm": 7.6875, + "learning_rate": 4.999559636724716e-06, + "loss": 2.7857, + "step": 1693 + }, + { + "epoch": 0.1791455160744501, + "grad_norm": 7.90625, + "learning_rate": 4.999556416623905e-06, + "loss": 2.7567, + "step": 1694 + }, + { + "epoch": 0.17925126903553298, + "grad_norm": 8.3125, + "learning_rate": 4.999553184793699e-06, + "loss": 2.7773, + "step": 1695 + }, + { + "epoch": 0.1793570219966159, + "grad_norm": 10.375, + "learning_rate": 4.99954994123411e-06, + "loss": 2.6245, + "step": 1696 + }, + { + "epoch": 0.17946277495769883, + "grad_norm": 7.9375, + "learning_rate": 4.999546685945157e-06, + "loss": 2.5987, + "step": 1697 + }, + { + "epoch": 0.17956852791878172, + "grad_norm": 7.875, + "learning_rate": 4.999543418926852e-06, + "loss": 2.5991, + "step": 1698 + }, + { + "epoch": 0.17967428087986465, + "grad_norm": 7.375, + "learning_rate": 4.999540140179213e-06, + "loss": 2.6435, + "step": 1699 + }, + { + "epoch": 0.17978003384094754, + "grad_norm": 8.25, + "learning_rate": 4.999536849702254e-06, + "loss": 2.6867, + "step": 1700 + }, + { + "epoch": 0.17988578680203046, + "grad_norm": 7.75, + "learning_rate": 4.999533547495992e-06, + "loss": 2.66, + "step": 1701 + }, + { + "epoch": 0.17999153976311336, + "grad_norm": 7.71875, + "learning_rate": 4.99953023356044e-06, + "loss": 2.6149, + "step": 1702 + }, + { + "epoch": 0.18009729272419628, + "grad_norm": 8.5625, + "learning_rate": 4.999526907895616e-06, + "loss": 2.6132, + "step": 1703 + }, + { + "epoch": 0.1802030456852792, + "grad_norm": 7.8125, + "learning_rate": 4.999523570501533e-06, + "loss": 2.4801, + "step": 1704 + }, + { + "epoch": 0.1803087986463621, + "grad_norm": 7.84375, + "learning_rate": 4.999520221378209e-06, + "loss": 2.2845, + "step": 1705 + }, + { + "epoch": 0.18041455160744502, + "grad_norm": 8.75, + "learning_rate": 4.999516860525659e-06, + "loss": 2.7032, + "step": 1706 + }, + { + "epoch": 0.1805203045685279, + "grad_norm": 7.375, + "learning_rate": 4.999513487943899e-06, + "loss": 2.6925, + "step": 1707 + }, + { + "epoch": 0.18062605752961083, + "grad_norm": 7.84375, + "learning_rate": 4.999510103632944e-06, + "loss": 2.6406, + "step": 1708 + }, + { + "epoch": 0.18073181049069373, + "grad_norm": 7.8125, + "learning_rate": 4.9995067075928104e-06, + "loss": 2.5154, + "step": 1709 + }, + { + "epoch": 0.18083756345177665, + "grad_norm": 7.9375, + "learning_rate": 4.999503299823515e-06, + "loss": 2.7014, + "step": 1710 + }, + { + "epoch": 0.18094331641285957, + "grad_norm": 8.3125, + "learning_rate": 4.999499880325071e-06, + "loss": 2.7338, + "step": 1711 + }, + { + "epoch": 0.18104906937394247, + "grad_norm": 8.0625, + "learning_rate": 4.999496449097497e-06, + "loss": 2.6327, + "step": 1712 + }, + { + "epoch": 0.1811548223350254, + "grad_norm": 8.3125, + "learning_rate": 4.9994930061408095e-06, + "loss": 2.6006, + "step": 1713 + }, + { + "epoch": 0.18126057529610828, + "grad_norm": 8.0, + "learning_rate": 4.999489551455023e-06, + "loss": 2.8353, + "step": 1714 + }, + { + "epoch": 0.1813663282571912, + "grad_norm": 8.3125, + "learning_rate": 4.999486085040154e-06, + "loss": 2.7449, + "step": 1715 + }, + { + "epoch": 0.1814720812182741, + "grad_norm": 8.125, + "learning_rate": 4.999482606896218e-06, + "loss": 2.7882, + "step": 1716 + }, + { + "epoch": 0.18157783417935702, + "grad_norm": 7.4375, + "learning_rate": 4.999479117023234e-06, + "loss": 2.5897, + "step": 1717 + }, + { + "epoch": 0.18168358714043994, + "grad_norm": 10.3125, + "learning_rate": 4.999475615421216e-06, + "loss": 2.9612, + "step": 1718 + }, + { + "epoch": 0.18178934010152284, + "grad_norm": 8.125, + "learning_rate": 4.999472102090182e-06, + "loss": 2.4059, + "step": 1719 + }, + { + "epoch": 0.18189509306260576, + "grad_norm": 7.6875, + "learning_rate": 4.999468577030147e-06, + "loss": 2.5103, + "step": 1720 + }, + { + "epoch": 0.18200084602368866, + "grad_norm": 8.0, + "learning_rate": 4.999465040241128e-06, + "loss": 2.6771, + "step": 1721 + }, + { + "epoch": 0.18210659898477158, + "grad_norm": 8.25, + "learning_rate": 4.999461491723142e-06, + "loss": 2.7886, + "step": 1722 + }, + { + "epoch": 0.18221235194585447, + "grad_norm": 7.96875, + "learning_rate": 4.999457931476206e-06, + "loss": 2.6818, + "step": 1723 + }, + { + "epoch": 0.1823181049069374, + "grad_norm": 7.75, + "learning_rate": 4.999454359500336e-06, + "loss": 2.8019, + "step": 1724 + }, + { + "epoch": 0.18242385786802032, + "grad_norm": 7.34375, + "learning_rate": 4.999450775795549e-06, + "loss": 2.6631, + "step": 1725 + }, + { + "epoch": 0.1825296108291032, + "grad_norm": 7.90625, + "learning_rate": 4.999447180361861e-06, + "loss": 2.5371, + "step": 1726 + }, + { + "epoch": 0.18263536379018613, + "grad_norm": 7.4375, + "learning_rate": 4.99944357319929e-06, + "loss": 2.8328, + "step": 1727 + }, + { + "epoch": 0.18274111675126903, + "grad_norm": 33.0, + "learning_rate": 4.999439954307853e-06, + "loss": 2.5675, + "step": 1728 + }, + { + "epoch": 0.18284686971235195, + "grad_norm": 8.5, + "learning_rate": 4.999436323687566e-06, + "loss": 2.689, + "step": 1729 + }, + { + "epoch": 0.18295262267343484, + "grad_norm": 8.75, + "learning_rate": 4.999432681338447e-06, + "loss": 2.7756, + "step": 1730 + }, + { + "epoch": 0.18305837563451777, + "grad_norm": 9.125, + "learning_rate": 4.999429027260512e-06, + "loss": 2.312, + "step": 1731 + }, + { + "epoch": 0.1831641285956007, + "grad_norm": 8.25, + "learning_rate": 4.999425361453779e-06, + "loss": 2.7636, + "step": 1732 + }, + { + "epoch": 0.18326988155668358, + "grad_norm": 8.5, + "learning_rate": 4.999421683918265e-06, + "loss": 2.4409, + "step": 1733 + }, + { + "epoch": 0.1833756345177665, + "grad_norm": 7.6875, + "learning_rate": 4.999417994653988e-06, + "loss": 2.6843, + "step": 1734 + }, + { + "epoch": 0.1834813874788494, + "grad_norm": 7.6875, + "learning_rate": 4.999414293660964e-06, + "loss": 2.6819, + "step": 1735 + }, + { + "epoch": 0.18358714043993232, + "grad_norm": 8.0625, + "learning_rate": 4.999410580939211e-06, + "loss": 2.7638, + "step": 1736 + }, + { + "epoch": 0.18369289340101522, + "grad_norm": 7.53125, + "learning_rate": 4.999406856488747e-06, + "loss": 2.624, + "step": 1737 + }, + { + "epoch": 0.18379864636209814, + "grad_norm": 8.0, + "learning_rate": 4.999403120309587e-06, + "loss": 2.8222, + "step": 1738 + }, + { + "epoch": 0.18390439932318106, + "grad_norm": 8.125, + "learning_rate": 4.999399372401752e-06, + "loss": 2.7919, + "step": 1739 + }, + { + "epoch": 0.18401015228426396, + "grad_norm": 9.125, + "learning_rate": 4.9993956127652575e-06, + "loss": 2.6974, + "step": 1740 + }, + { + "epoch": 0.18411590524534688, + "grad_norm": 7.71875, + "learning_rate": 4.999391841400122e-06, + "loss": 2.9125, + "step": 1741 + }, + { + "epoch": 0.18422165820642977, + "grad_norm": 8.5, + "learning_rate": 4.999388058306363e-06, + "loss": 2.4621, + "step": 1742 + }, + { + "epoch": 0.1843274111675127, + "grad_norm": 7.90625, + "learning_rate": 4.9993842634839965e-06, + "loss": 2.7907, + "step": 1743 + }, + { + "epoch": 0.1844331641285956, + "grad_norm": 7.53125, + "learning_rate": 4.9993804569330426e-06, + "loss": 2.5584, + "step": 1744 + }, + { + "epoch": 0.1845389170896785, + "grad_norm": 8.1875, + "learning_rate": 4.999376638653519e-06, + "loss": 2.8539, + "step": 1745 + }, + { + "epoch": 0.18464467005076143, + "grad_norm": 8.0, + "learning_rate": 4.999372808645443e-06, + "loss": 2.6452, + "step": 1746 + }, + { + "epoch": 0.18475042301184433, + "grad_norm": 8.0, + "learning_rate": 4.999368966908833e-06, + "loss": 2.6252, + "step": 1747 + }, + { + "epoch": 0.18485617597292725, + "grad_norm": 8.1875, + "learning_rate": 4.999365113443705e-06, + "loss": 2.4987, + "step": 1748 + }, + { + "epoch": 0.18496192893401014, + "grad_norm": 7.9375, + "learning_rate": 4.9993612482500805e-06, + "loss": 2.7765, + "step": 1749 + }, + { + "epoch": 0.18506768189509307, + "grad_norm": 7.9375, + "learning_rate": 4.999357371327975e-06, + "loss": 2.8416, + "step": 1750 + }, + { + "epoch": 0.18517343485617596, + "grad_norm": 8.0625, + "learning_rate": 4.999353482677407e-06, + "loss": 2.7046, + "step": 1751 + }, + { + "epoch": 0.18527918781725888, + "grad_norm": 8.6875, + "learning_rate": 4.999349582298396e-06, + "loss": 2.7847, + "step": 1752 + }, + { + "epoch": 0.1853849407783418, + "grad_norm": 8.625, + "learning_rate": 4.999345670190961e-06, + "loss": 2.9171, + "step": 1753 + }, + { + "epoch": 0.1854906937394247, + "grad_norm": 8.4375, + "learning_rate": 4.999341746355116e-06, + "loss": 2.9064, + "step": 1754 + }, + { + "epoch": 0.18559644670050762, + "grad_norm": 9.8125, + "learning_rate": 4.999337810790884e-06, + "loss": 2.606, + "step": 1755 + }, + { + "epoch": 0.18570219966159052, + "grad_norm": 8.0625, + "learning_rate": 4.999333863498283e-06, + "loss": 2.7359, + "step": 1756 + }, + { + "epoch": 0.18580795262267344, + "grad_norm": 8.0, + "learning_rate": 4.999329904477328e-06, + "loss": 2.3635, + "step": 1757 + }, + { + "epoch": 0.18591370558375633, + "grad_norm": 7.78125, + "learning_rate": 4.999325933728042e-06, + "loss": 2.701, + "step": 1758 + }, + { + "epoch": 0.18601945854483926, + "grad_norm": 10.0625, + "learning_rate": 4.999321951250441e-06, + "loss": 2.8197, + "step": 1759 + }, + { + "epoch": 0.18612521150592218, + "grad_norm": 8.25, + "learning_rate": 4.999317957044543e-06, + "loss": 2.7741, + "step": 1760 + }, + { + "epoch": 0.18623096446700507, + "grad_norm": 8.75, + "learning_rate": 4.999313951110369e-06, + "loss": 2.4684, + "step": 1761 + }, + { + "epoch": 0.186336717428088, + "grad_norm": 7.75, + "learning_rate": 4.9993099334479365e-06, + "loss": 2.5523, + "step": 1762 + }, + { + "epoch": 0.1864424703891709, + "grad_norm": 8.0, + "learning_rate": 4.999305904057265e-06, + "loss": 2.6232, + "step": 1763 + }, + { + "epoch": 0.1865482233502538, + "grad_norm": 7.78125, + "learning_rate": 4.9993018629383735e-06, + "loss": 2.6086, + "step": 1764 + }, + { + "epoch": 0.1866539763113367, + "grad_norm": 6.90625, + "learning_rate": 4.9992978100912805e-06, + "loss": 2.6542, + "step": 1765 + }, + { + "epoch": 0.18675972927241963, + "grad_norm": 7.1875, + "learning_rate": 4.999293745516004e-06, + "loss": 2.7508, + "step": 1766 + }, + { + "epoch": 0.18686548223350255, + "grad_norm": 8.4375, + "learning_rate": 4.999289669212565e-06, + "loss": 2.8558, + "step": 1767 + }, + { + "epoch": 0.18697123519458544, + "grad_norm": 11.3125, + "learning_rate": 4.999285581180981e-06, + "loss": 2.6561, + "step": 1768 + }, + { + "epoch": 0.18707698815566837, + "grad_norm": 8.125, + "learning_rate": 4.999281481421273e-06, + "loss": 2.8333, + "step": 1769 + }, + { + "epoch": 0.18718274111675126, + "grad_norm": 7.875, + "learning_rate": 4.999277369933458e-06, + "loss": 2.6584, + "step": 1770 + }, + { + "epoch": 0.18728849407783418, + "grad_norm": 7.84375, + "learning_rate": 4.9992732467175575e-06, + "loss": 2.8041, + "step": 1771 + }, + { + "epoch": 0.18739424703891708, + "grad_norm": 7.65625, + "learning_rate": 4.999269111773589e-06, + "loss": 2.8057, + "step": 1772 + }, + { + "epoch": 0.1875, + "grad_norm": 11.8125, + "learning_rate": 4.999264965101573e-06, + "loss": 2.7309, + "step": 1773 + }, + { + "epoch": 0.18760575296108292, + "grad_norm": 11.0, + "learning_rate": 4.999260806701529e-06, + "loss": 2.5186, + "step": 1774 + }, + { + "epoch": 0.18771150592216582, + "grad_norm": 7.75, + "learning_rate": 4.9992566365734764e-06, + "loss": 2.6332, + "step": 1775 + }, + { + "epoch": 0.18781725888324874, + "grad_norm": 7.9375, + "learning_rate": 4.999252454717435e-06, + "loss": 2.6528, + "step": 1776 + }, + { + "epoch": 0.18792301184433163, + "grad_norm": 8.5, + "learning_rate": 4.999248261133423e-06, + "loss": 2.5133, + "step": 1777 + }, + { + "epoch": 0.18802876480541456, + "grad_norm": 9.3125, + "learning_rate": 4.999244055821461e-06, + "loss": 2.6799, + "step": 1778 + }, + { + "epoch": 0.18813451776649745, + "grad_norm": 8.4375, + "learning_rate": 4.9992398387815685e-06, + "loss": 2.5435, + "step": 1779 + }, + { + "epoch": 0.18824027072758037, + "grad_norm": 7.65625, + "learning_rate": 4.9992356100137665e-06, + "loss": 2.6371, + "step": 1780 + }, + { + "epoch": 0.1883460236886633, + "grad_norm": 8.5, + "learning_rate": 4.999231369518073e-06, + "loss": 2.7912, + "step": 1781 + }, + { + "epoch": 0.1884517766497462, + "grad_norm": 7.78125, + "learning_rate": 4.9992271172945095e-06, + "loss": 2.879, + "step": 1782 + }, + { + "epoch": 0.1885575296108291, + "grad_norm": 8.1875, + "learning_rate": 4.999222853343095e-06, + "loss": 2.5883, + "step": 1783 + }, + { + "epoch": 0.188663282571912, + "grad_norm": 8.6875, + "learning_rate": 4.9992185776638505e-06, + "loss": 2.6438, + "step": 1784 + }, + { + "epoch": 0.18876903553299493, + "grad_norm": 8.5625, + "learning_rate": 4.9992142902567944e-06, + "loss": 2.6166, + "step": 1785 + }, + { + "epoch": 0.18887478849407782, + "grad_norm": 8.5, + "learning_rate": 4.999209991121948e-06, + "loss": 2.7591, + "step": 1786 + }, + { + "epoch": 0.18898054145516074, + "grad_norm": 8.5625, + "learning_rate": 4.9992056802593325e-06, + "loss": 2.7166, + "step": 1787 + }, + { + "epoch": 0.18908629441624367, + "grad_norm": 8.3125, + "learning_rate": 4.9992013576689655e-06, + "loss": 2.6637, + "step": 1788 + }, + { + "epoch": 0.18919204737732656, + "grad_norm": 7.78125, + "learning_rate": 4.999197023350869e-06, + "loss": 2.6967, + "step": 1789 + }, + { + "epoch": 0.18929780033840948, + "grad_norm": 9.75, + "learning_rate": 4.999192677305063e-06, + "loss": 2.7215, + "step": 1790 + }, + { + "epoch": 0.18940355329949238, + "grad_norm": 7.84375, + "learning_rate": 4.999188319531568e-06, + "loss": 2.416, + "step": 1791 + }, + { + "epoch": 0.1895093062605753, + "grad_norm": 8.875, + "learning_rate": 4.999183950030404e-06, + "loss": 2.5707, + "step": 1792 + }, + { + "epoch": 0.1896150592216582, + "grad_norm": 8.0625, + "learning_rate": 4.999179568801592e-06, + "loss": 2.7895, + "step": 1793 + }, + { + "epoch": 0.18972081218274112, + "grad_norm": 7.9375, + "learning_rate": 4.999175175845153e-06, + "loss": 2.6876, + "step": 1794 + }, + { + "epoch": 0.18982656514382404, + "grad_norm": 7.28125, + "learning_rate": 4.999170771161107e-06, + "loss": 2.4447, + "step": 1795 + }, + { + "epoch": 0.18993231810490693, + "grad_norm": 8.4375, + "learning_rate": 4.999166354749474e-06, + "loss": 2.7774, + "step": 1796 + }, + { + "epoch": 0.19003807106598986, + "grad_norm": 8.4375, + "learning_rate": 4.999161926610275e-06, + "loss": 2.3334, + "step": 1797 + }, + { + "epoch": 0.19014382402707275, + "grad_norm": 8.1875, + "learning_rate": 4.999157486743532e-06, + "loss": 2.7592, + "step": 1798 + }, + { + "epoch": 0.19024957698815567, + "grad_norm": 7.5, + "learning_rate": 4.999153035149265e-06, + "loss": 2.6297, + "step": 1799 + }, + { + "epoch": 0.19035532994923857, + "grad_norm": 8.125, + "learning_rate": 4.9991485718274955e-06, + "loss": 2.4502, + "step": 1800 + }, + { + "epoch": 0.1904610829103215, + "grad_norm": 8.6875, + "learning_rate": 4.999144096778243e-06, + "loss": 2.8703, + "step": 1801 + }, + { + "epoch": 0.1905668358714044, + "grad_norm": 7.875, + "learning_rate": 4.99913961000153e-06, + "loss": 2.8429, + "step": 1802 + }, + { + "epoch": 0.1906725888324873, + "grad_norm": 8.3125, + "learning_rate": 4.999135111497376e-06, + "loss": 2.6059, + "step": 1803 + }, + { + "epoch": 0.19077834179357023, + "grad_norm": 12.0, + "learning_rate": 4.999130601265804e-06, + "loss": 2.3834, + "step": 1804 + }, + { + "epoch": 0.19088409475465312, + "grad_norm": 11.125, + "learning_rate": 4.999126079306833e-06, + "loss": 2.835, + "step": 1805 + }, + { + "epoch": 0.19098984771573604, + "grad_norm": 7.9375, + "learning_rate": 4.999121545620487e-06, + "loss": 2.819, + "step": 1806 + }, + { + "epoch": 0.19109560067681894, + "grad_norm": 10.0625, + "learning_rate": 4.999117000206784e-06, + "loss": 2.6123, + "step": 1807 + }, + { + "epoch": 0.19120135363790186, + "grad_norm": 9.4375, + "learning_rate": 4.999112443065748e-06, + "loss": 2.7972, + "step": 1808 + }, + { + "epoch": 0.19130710659898478, + "grad_norm": 8.6875, + "learning_rate": 4.9991078741974e-06, + "loss": 2.5042, + "step": 1809 + }, + { + "epoch": 0.19141285956006768, + "grad_norm": 8.0625, + "learning_rate": 4.999103293601759e-06, + "loss": 2.6617, + "step": 1810 + }, + { + "epoch": 0.1915186125211506, + "grad_norm": 7.53125, + "learning_rate": 4.99909870127885e-06, + "loss": 2.7297, + "step": 1811 + }, + { + "epoch": 0.1916243654822335, + "grad_norm": 7.71875, + "learning_rate": 4.999094097228693e-06, + "loss": 2.6884, + "step": 1812 + }, + { + "epoch": 0.19173011844331642, + "grad_norm": 8.0, + "learning_rate": 4.999089481451308e-06, + "loss": 2.5748, + "step": 1813 + }, + { + "epoch": 0.1918358714043993, + "grad_norm": 7.21875, + "learning_rate": 4.9990848539467195e-06, + "loss": 2.54, + "step": 1814 + }, + { + "epoch": 0.19194162436548223, + "grad_norm": 9.9375, + "learning_rate": 4.999080214714947e-06, + "loss": 2.5239, + "step": 1815 + }, + { + "epoch": 0.19204737732656516, + "grad_norm": 8.3125, + "learning_rate": 4.999075563756014e-06, + "loss": 2.6052, + "step": 1816 + }, + { + "epoch": 0.19215313028764805, + "grad_norm": 7.90625, + "learning_rate": 4.9990709010699416e-06, + "loss": 2.677, + "step": 1817 + }, + { + "epoch": 0.19225888324873097, + "grad_norm": 7.84375, + "learning_rate": 4.999066226656751e-06, + "loss": 2.5403, + "step": 1818 + }, + { + "epoch": 0.19236463620981387, + "grad_norm": 7.96875, + "learning_rate": 4.999061540516464e-06, + "loss": 2.7334, + "step": 1819 + }, + { + "epoch": 0.1924703891708968, + "grad_norm": 11.0625, + "learning_rate": 4.9990568426491034e-06, + "loss": 2.5455, + "step": 1820 + }, + { + "epoch": 0.19257614213197968, + "grad_norm": 8.0, + "learning_rate": 4.999052133054692e-06, + "loss": 2.7584, + "step": 1821 + }, + { + "epoch": 0.1926818950930626, + "grad_norm": 8.1875, + "learning_rate": 4.99904741173325e-06, + "loss": 2.6309, + "step": 1822 + }, + { + "epoch": 0.19278764805414553, + "grad_norm": 8.0, + "learning_rate": 4.9990426786848015e-06, + "loss": 2.6873, + "step": 1823 + }, + { + "epoch": 0.19289340101522842, + "grad_norm": 9.0, + "learning_rate": 4.999037933909367e-06, + "loss": 2.7073, + "step": 1824 + }, + { + "epoch": 0.19299915397631134, + "grad_norm": 7.71875, + "learning_rate": 4.999033177406971e-06, + "loss": 2.7722, + "step": 1825 + }, + { + "epoch": 0.19310490693739424, + "grad_norm": 7.15625, + "learning_rate": 4.999028409177633e-06, + "loss": 2.7005, + "step": 1826 + }, + { + "epoch": 0.19321065989847716, + "grad_norm": 8.1875, + "learning_rate": 4.999023629221377e-06, + "loss": 2.6018, + "step": 1827 + }, + { + "epoch": 0.19331641285956006, + "grad_norm": 7.65625, + "learning_rate": 4.9990188375382254e-06, + "loss": 2.6333, + "step": 1828 + }, + { + "epoch": 0.19342216582064298, + "grad_norm": 8.0, + "learning_rate": 4.999014034128201e-06, + "loss": 2.7947, + "step": 1829 + }, + { + "epoch": 0.1935279187817259, + "grad_norm": 7.46875, + "learning_rate": 4.999009218991325e-06, + "loss": 2.6191, + "step": 1830 + }, + { + "epoch": 0.1936336717428088, + "grad_norm": 7.875, + "learning_rate": 4.999004392127621e-06, + "loss": 2.6723, + "step": 1831 + }, + { + "epoch": 0.19373942470389172, + "grad_norm": 7.90625, + "learning_rate": 4.9989995535371115e-06, + "loss": 2.7302, + "step": 1832 + }, + { + "epoch": 0.1938451776649746, + "grad_norm": 7.6875, + "learning_rate": 4.9989947032198185e-06, + "loss": 2.3673, + "step": 1833 + }, + { + "epoch": 0.19395093062605753, + "grad_norm": 8.9375, + "learning_rate": 4.998989841175767e-06, + "loss": 2.4476, + "step": 1834 + }, + { + "epoch": 0.19405668358714043, + "grad_norm": 7.5625, + "learning_rate": 4.998984967404976e-06, + "loss": 2.8259, + "step": 1835 + }, + { + "epoch": 0.19416243654822335, + "grad_norm": 11.9375, + "learning_rate": 4.998980081907472e-06, + "loss": 2.7559, + "step": 1836 + }, + { + "epoch": 0.19426818950930627, + "grad_norm": 7.96875, + "learning_rate": 4.998975184683277e-06, + "loss": 2.6583, + "step": 1837 + }, + { + "epoch": 0.19437394247038917, + "grad_norm": 8.9375, + "learning_rate": 4.998970275732412e-06, + "loss": 2.6577, + "step": 1838 + }, + { + "epoch": 0.1944796954314721, + "grad_norm": 7.6875, + "learning_rate": 4.998965355054903e-06, + "loss": 2.7734, + "step": 1839 + }, + { + "epoch": 0.19458544839255498, + "grad_norm": 7.84375, + "learning_rate": 4.998960422650771e-06, + "loss": 2.4832, + "step": 1840 + }, + { + "epoch": 0.1946912013536379, + "grad_norm": 7.5, + "learning_rate": 4.99895547852004e-06, + "loss": 2.3933, + "step": 1841 + }, + { + "epoch": 0.1947969543147208, + "grad_norm": 7.84375, + "learning_rate": 4.998950522662732e-06, + "loss": 2.6429, + "step": 1842 + }, + { + "epoch": 0.19490270727580372, + "grad_norm": 7.8125, + "learning_rate": 4.998945555078872e-06, + "loss": 2.5441, + "step": 1843 + }, + { + "epoch": 0.19500846023688664, + "grad_norm": 8.375, + "learning_rate": 4.9989405757684825e-06, + "loss": 2.6185, + "step": 1844 + }, + { + "epoch": 0.19511421319796954, + "grad_norm": 10.375, + "learning_rate": 4.9989355847315875e-06, + "loss": 2.4671, + "step": 1845 + }, + { + "epoch": 0.19521996615905246, + "grad_norm": 8.375, + "learning_rate": 4.998930581968209e-06, + "loss": 2.5016, + "step": 1846 + }, + { + "epoch": 0.19532571912013535, + "grad_norm": 8.625, + "learning_rate": 4.998925567478372e-06, + "loss": 2.7731, + "step": 1847 + }, + { + "epoch": 0.19543147208121828, + "grad_norm": 8.75, + "learning_rate": 4.998920541262099e-06, + "loss": 2.7119, + "step": 1848 + }, + { + "epoch": 0.19553722504230117, + "grad_norm": 7.5625, + "learning_rate": 4.998915503319413e-06, + "loss": 2.6307, + "step": 1849 + }, + { + "epoch": 0.1956429780033841, + "grad_norm": 7.75, + "learning_rate": 4.9989104536503394e-06, + "loss": 2.6257, + "step": 1850 + }, + { + "epoch": 0.19574873096446702, + "grad_norm": 8.1875, + "learning_rate": 4.998905392254902e-06, + "loss": 2.6248, + "step": 1851 + }, + { + "epoch": 0.1958544839255499, + "grad_norm": 8.0625, + "learning_rate": 4.998900319133122e-06, + "loss": 2.5657, + "step": 1852 + }, + { + "epoch": 0.19596023688663283, + "grad_norm": 9.125, + "learning_rate": 4.998895234285026e-06, + "loss": 2.7993, + "step": 1853 + }, + { + "epoch": 0.19606598984771573, + "grad_norm": 8.9375, + "learning_rate": 4.998890137710636e-06, + "loss": 2.7416, + "step": 1854 + }, + { + "epoch": 0.19617174280879865, + "grad_norm": 8.75, + "learning_rate": 4.998885029409977e-06, + "loss": 2.492, + "step": 1855 + }, + { + "epoch": 0.19627749576988154, + "grad_norm": 8.0, + "learning_rate": 4.998879909383072e-06, + "loss": 2.5825, + "step": 1856 + }, + { + "epoch": 0.19638324873096447, + "grad_norm": 7.40625, + "learning_rate": 4.998874777629946e-06, + "loss": 2.6402, + "step": 1857 + }, + { + "epoch": 0.1964890016920474, + "grad_norm": 8.0625, + "learning_rate": 4.998869634150623e-06, + "loss": 2.5891, + "step": 1858 + }, + { + "epoch": 0.19659475465313028, + "grad_norm": 8.25, + "learning_rate": 4.998864478945126e-06, + "loss": 2.9032, + "step": 1859 + }, + { + "epoch": 0.1967005076142132, + "grad_norm": 9.6875, + "learning_rate": 4.99885931201348e-06, + "loss": 2.4709, + "step": 1860 + }, + { + "epoch": 0.1968062605752961, + "grad_norm": 8.0, + "learning_rate": 4.99885413335571e-06, + "loss": 2.5834, + "step": 1861 + }, + { + "epoch": 0.19691201353637902, + "grad_norm": 7.5, + "learning_rate": 4.998848942971839e-06, + "loss": 2.6274, + "step": 1862 + }, + { + "epoch": 0.19701776649746192, + "grad_norm": 7.59375, + "learning_rate": 4.998843740861893e-06, + "loss": 2.71, + "step": 1863 + }, + { + "epoch": 0.19712351945854484, + "grad_norm": 7.96875, + "learning_rate": 4.998838527025894e-06, + "loss": 2.6091, + "step": 1864 + }, + { + "epoch": 0.19722927241962776, + "grad_norm": 7.09375, + "learning_rate": 4.998833301463868e-06, + "loss": 2.5498, + "step": 1865 + }, + { + "epoch": 0.19733502538071065, + "grad_norm": 8.1875, + "learning_rate": 4.9988280641758394e-06, + "loss": 2.7214, + "step": 1866 + }, + { + "epoch": 0.19744077834179358, + "grad_norm": 8.3125, + "learning_rate": 4.998822815161833e-06, + "loss": 2.6211, + "step": 1867 + }, + { + "epoch": 0.19754653130287647, + "grad_norm": 8.0625, + "learning_rate": 4.9988175544218735e-06, + "loss": 2.5163, + "step": 1868 + }, + { + "epoch": 0.1976522842639594, + "grad_norm": 7.96875, + "learning_rate": 4.998812281955984e-06, + "loss": 2.755, + "step": 1869 + }, + { + "epoch": 0.1977580372250423, + "grad_norm": 8.375, + "learning_rate": 4.998806997764191e-06, + "loss": 2.8245, + "step": 1870 + }, + { + "epoch": 0.1978637901861252, + "grad_norm": 9.25, + "learning_rate": 4.998801701846518e-06, + "loss": 2.4187, + "step": 1871 + }, + { + "epoch": 0.19796954314720813, + "grad_norm": 10.4375, + "learning_rate": 4.998796394202992e-06, + "loss": 2.8604, + "step": 1872 + }, + { + "epoch": 0.19807529610829103, + "grad_norm": 8.25, + "learning_rate": 4.998791074833636e-06, + "loss": 2.7996, + "step": 1873 + }, + { + "epoch": 0.19818104906937395, + "grad_norm": 7.8125, + "learning_rate": 4.9987857437384755e-06, + "loss": 2.7081, + "step": 1874 + }, + { + "epoch": 0.19828680203045684, + "grad_norm": 9.1875, + "learning_rate": 4.998780400917535e-06, + "loss": 2.9542, + "step": 1875 + }, + { + "epoch": 0.19839255499153977, + "grad_norm": 7.9375, + "learning_rate": 4.998775046370841e-06, + "loss": 2.6399, + "step": 1876 + }, + { + "epoch": 0.19849830795262266, + "grad_norm": 8.1875, + "learning_rate": 4.998769680098416e-06, + "loss": 2.5624, + "step": 1877 + }, + { + "epoch": 0.19860406091370558, + "grad_norm": 7.375, + "learning_rate": 4.9987643021002885e-06, + "loss": 2.5503, + "step": 1878 + }, + { + "epoch": 0.1987098138747885, + "grad_norm": 8.3125, + "learning_rate": 4.998758912376481e-06, + "loss": 2.6709, + "step": 1879 + }, + { + "epoch": 0.1988155668358714, + "grad_norm": 8.125, + "learning_rate": 4.99875351092702e-06, + "loss": 2.5398, + "step": 1880 + }, + { + "epoch": 0.19892131979695432, + "grad_norm": 9.625, + "learning_rate": 4.998748097751932e-06, + "loss": 2.4579, + "step": 1881 + }, + { + "epoch": 0.19902707275803722, + "grad_norm": 8.1875, + "learning_rate": 4.99874267285124e-06, + "loss": 2.9502, + "step": 1882 + }, + { + "epoch": 0.19913282571912014, + "grad_norm": 7.6875, + "learning_rate": 4.998737236224971e-06, + "loss": 2.5415, + "step": 1883 + }, + { + "epoch": 0.19923857868020303, + "grad_norm": 8.1875, + "learning_rate": 4.99873178787315e-06, + "loss": 2.7016, + "step": 1884 + }, + { + "epoch": 0.19934433164128595, + "grad_norm": 7.84375, + "learning_rate": 4.998726327795802e-06, + "loss": 2.5383, + "step": 1885 + }, + { + "epoch": 0.19945008460236888, + "grad_norm": 8.8125, + "learning_rate": 4.998720855992954e-06, + "loss": 2.5773, + "step": 1886 + }, + { + "epoch": 0.19955583756345177, + "grad_norm": 7.84375, + "learning_rate": 4.998715372464631e-06, + "loss": 2.8139, + "step": 1887 + }, + { + "epoch": 0.1996615905245347, + "grad_norm": 8.625, + "learning_rate": 4.998709877210858e-06, + "loss": 2.5698, + "step": 1888 + }, + { + "epoch": 0.1997673434856176, + "grad_norm": 7.5625, + "learning_rate": 4.998704370231662e-06, + "loss": 2.4931, + "step": 1889 + }, + { + "epoch": 0.1998730964467005, + "grad_norm": 11.1875, + "learning_rate": 4.998698851527068e-06, + "loss": 2.5137, + "step": 1890 + }, + { + "epoch": 0.1999788494077834, + "grad_norm": 7.71875, + "learning_rate": 4.998693321097104e-06, + "loss": 2.7377, + "step": 1891 + }, + { + "epoch": 0.20008460236886633, + "grad_norm": 8.1875, + "learning_rate": 4.998687778941792e-06, + "loss": 2.9007, + "step": 1892 + }, + { + "epoch": 0.20019035532994925, + "grad_norm": 7.375, + "learning_rate": 4.998682225061161e-06, + "loss": 2.6274, + "step": 1893 + }, + { + "epoch": 0.20029610829103214, + "grad_norm": 7.75, + "learning_rate": 4.998676659455236e-06, + "loss": 2.6775, + "step": 1894 + }, + { + "epoch": 0.20040186125211507, + "grad_norm": 7.75, + "learning_rate": 4.998671082124044e-06, + "loss": 2.6473, + "step": 1895 + }, + { + "epoch": 0.20050761421319796, + "grad_norm": 7.625, + "learning_rate": 4.9986654930676095e-06, + "loss": 2.6634, + "step": 1896 + }, + { + "epoch": 0.20061336717428088, + "grad_norm": 7.625, + "learning_rate": 4.99865989228596e-06, + "loss": 2.5627, + "step": 1897 + }, + { + "epoch": 0.20071912013536378, + "grad_norm": 8.4375, + "learning_rate": 4.998654279779122e-06, + "loss": 2.7274, + "step": 1898 + }, + { + "epoch": 0.2008248730964467, + "grad_norm": 7.75, + "learning_rate": 4.998648655547121e-06, + "loss": 2.7737, + "step": 1899 + }, + { + "epoch": 0.20093062605752962, + "grad_norm": 7.5, + "learning_rate": 4.998643019589984e-06, + "loss": 2.64, + "step": 1900 + }, + { + "epoch": 0.20103637901861252, + "grad_norm": 7.65625, + "learning_rate": 4.998637371907737e-06, + "loss": 2.6035, + "step": 1901 + }, + { + "epoch": 0.20114213197969544, + "grad_norm": 7.84375, + "learning_rate": 4.998631712500407e-06, + "loss": 2.7503, + "step": 1902 + }, + { + "epoch": 0.20124788494077833, + "grad_norm": 8.125, + "learning_rate": 4.998626041368021e-06, + "loss": 2.54, + "step": 1903 + }, + { + "epoch": 0.20135363790186125, + "grad_norm": 8.25, + "learning_rate": 4.9986203585106036e-06, + "loss": 2.6494, + "step": 1904 + }, + { + "epoch": 0.20145939086294415, + "grad_norm": 7.96875, + "learning_rate": 4.998614663928183e-06, + "loss": 2.5725, + "step": 1905 + }, + { + "epoch": 0.20156514382402707, + "grad_norm": 7.5625, + "learning_rate": 4.998608957620786e-06, + "loss": 2.4867, + "step": 1906 + }, + { + "epoch": 0.20167089678511, + "grad_norm": 7.71875, + "learning_rate": 4.998603239588439e-06, + "loss": 2.6064, + "step": 1907 + }, + { + "epoch": 0.2017766497461929, + "grad_norm": 7.625, + "learning_rate": 4.998597509831169e-06, + "loss": 2.5894, + "step": 1908 + }, + { + "epoch": 0.2018824027072758, + "grad_norm": 8.4375, + "learning_rate": 4.998591768349003e-06, + "loss": 2.8164, + "step": 1909 + }, + { + "epoch": 0.2019881556683587, + "grad_norm": 8.5, + "learning_rate": 4.9985860151419675e-06, + "loss": 2.6633, + "step": 1910 + }, + { + "epoch": 0.20209390862944163, + "grad_norm": 7.625, + "learning_rate": 4.998580250210089e-06, + "loss": 2.6374, + "step": 1911 + }, + { + "epoch": 0.20219966159052452, + "grad_norm": 8.25, + "learning_rate": 4.998574473553396e-06, + "loss": 2.7625, + "step": 1912 + }, + { + "epoch": 0.20230541455160744, + "grad_norm": 8.1875, + "learning_rate": 4.998568685171915e-06, + "loss": 2.6198, + "step": 1913 + }, + { + "epoch": 0.20241116751269037, + "grad_norm": 8.1875, + "learning_rate": 4.998562885065673e-06, + "loss": 2.5411, + "step": 1914 + }, + { + "epoch": 0.20251692047377326, + "grad_norm": 8.5, + "learning_rate": 4.998557073234696e-06, + "loss": 2.4162, + "step": 1915 + }, + { + "epoch": 0.20262267343485618, + "grad_norm": 7.4375, + "learning_rate": 4.998551249679014e-06, + "loss": 2.6221, + "step": 1916 + }, + { + "epoch": 0.20272842639593908, + "grad_norm": 7.84375, + "learning_rate": 4.998545414398652e-06, + "loss": 2.3284, + "step": 1917 + }, + { + "epoch": 0.202834179357022, + "grad_norm": 7.78125, + "learning_rate": 4.99853956739364e-06, + "loss": 2.6607, + "step": 1918 + }, + { + "epoch": 0.2029399323181049, + "grad_norm": 7.9375, + "learning_rate": 4.998533708664001e-06, + "loss": 2.4797, + "step": 1919 + }, + { + "epoch": 0.20304568527918782, + "grad_norm": 7.625, + "learning_rate": 4.998527838209767e-06, + "loss": 2.7329, + "step": 1920 + }, + { + "epoch": 0.20315143824027074, + "grad_norm": 7.6875, + "learning_rate": 4.998521956030963e-06, + "loss": 2.5638, + "step": 1921 + }, + { + "epoch": 0.20325719120135363, + "grad_norm": 7.90625, + "learning_rate": 4.998516062127618e-06, + "loss": 2.6398, + "step": 1922 + }, + { + "epoch": 0.20336294416243655, + "grad_norm": 8.375, + "learning_rate": 4.998510156499758e-06, + "loss": 2.4982, + "step": 1923 + }, + { + "epoch": 0.20346869712351945, + "grad_norm": 8.3125, + "learning_rate": 4.998504239147413e-06, + "loss": 2.6401, + "step": 1924 + }, + { + "epoch": 0.20357445008460237, + "grad_norm": 9.625, + "learning_rate": 4.998498310070608e-06, + "loss": 2.7284, + "step": 1925 + }, + { + "epoch": 0.2036802030456853, + "grad_norm": 7.6875, + "learning_rate": 4.998492369269373e-06, + "loss": 2.7327, + "step": 1926 + }, + { + "epoch": 0.2037859560067682, + "grad_norm": 8.9375, + "learning_rate": 4.998486416743735e-06, + "loss": 2.7774, + "step": 1927 + }, + { + "epoch": 0.2038917089678511, + "grad_norm": 7.9375, + "learning_rate": 4.998480452493722e-06, + "loss": 2.4966, + "step": 1928 + }, + { + "epoch": 0.203997461928934, + "grad_norm": 7.84375, + "learning_rate": 4.998474476519362e-06, + "loss": 2.6598, + "step": 1929 + }, + { + "epoch": 0.20410321489001693, + "grad_norm": 7.625, + "learning_rate": 4.998468488820685e-06, + "loss": 2.5205, + "step": 1930 + }, + { + "epoch": 0.20420896785109982, + "grad_norm": 7.625, + "learning_rate": 4.998462489397715e-06, + "loss": 2.6648, + "step": 1931 + }, + { + "epoch": 0.20431472081218274, + "grad_norm": 7.46875, + "learning_rate": 4.998456478250483e-06, + "loss": 2.5886, + "step": 1932 + }, + { + "epoch": 0.20442047377326567, + "grad_norm": 9.625, + "learning_rate": 4.998450455379017e-06, + "loss": 2.6859, + "step": 1933 + }, + { + "epoch": 0.20452622673434856, + "grad_norm": 8.25, + "learning_rate": 4.998444420783344e-06, + "loss": 2.6202, + "step": 1934 + }, + { + "epoch": 0.20463197969543148, + "grad_norm": 7.96875, + "learning_rate": 4.998438374463494e-06, + "loss": 2.7047, + "step": 1935 + }, + { + "epoch": 0.20473773265651438, + "grad_norm": 17.5, + "learning_rate": 4.998432316419493e-06, + "loss": 2.6815, + "step": 1936 + }, + { + "epoch": 0.2048434856175973, + "grad_norm": 8.5625, + "learning_rate": 4.998426246651373e-06, + "loss": 2.7152, + "step": 1937 + }, + { + "epoch": 0.2049492385786802, + "grad_norm": 8.6875, + "learning_rate": 4.99842016515916e-06, + "loss": 2.7486, + "step": 1938 + }, + { + "epoch": 0.20505499153976312, + "grad_norm": 9.125, + "learning_rate": 4.998414071942882e-06, + "loss": 2.899, + "step": 1939 + }, + { + "epoch": 0.20516074450084604, + "grad_norm": 8.4375, + "learning_rate": 4.998407967002569e-06, + "loss": 2.7897, + "step": 1940 + }, + { + "epoch": 0.20526649746192893, + "grad_norm": 8.0625, + "learning_rate": 4.998401850338249e-06, + "loss": 2.8069, + "step": 1941 + }, + { + "epoch": 0.20537225042301185, + "grad_norm": 7.375, + "learning_rate": 4.998395721949952e-06, + "loss": 2.6214, + "step": 1942 + }, + { + "epoch": 0.20547800338409475, + "grad_norm": 8.4375, + "learning_rate": 4.998389581837705e-06, + "loss": 2.775, + "step": 1943 + }, + { + "epoch": 0.20558375634517767, + "grad_norm": 7.96875, + "learning_rate": 4.998383430001537e-06, + "loss": 2.8424, + "step": 1944 + }, + { + "epoch": 0.20568950930626057, + "grad_norm": 7.4375, + "learning_rate": 4.9983772664414784e-06, + "loss": 2.744, + "step": 1945 + }, + { + "epoch": 0.2057952622673435, + "grad_norm": 9.5, + "learning_rate": 4.998371091157556e-06, + "loss": 2.7128, + "step": 1946 + }, + { + "epoch": 0.2059010152284264, + "grad_norm": 7.96875, + "learning_rate": 4.9983649041498e-06, + "loss": 2.5037, + "step": 1947 + }, + { + "epoch": 0.2060067681895093, + "grad_norm": 9.9375, + "learning_rate": 4.99835870541824e-06, + "loss": 2.9134, + "step": 1948 + }, + { + "epoch": 0.20611252115059223, + "grad_norm": 8.375, + "learning_rate": 4.998352494962904e-06, + "loss": 2.8394, + "step": 1949 + }, + { + "epoch": 0.20621827411167512, + "grad_norm": 8.1875, + "learning_rate": 4.998346272783822e-06, + "loss": 2.6833, + "step": 1950 + }, + { + "epoch": 0.20632402707275804, + "grad_norm": 8.25, + "learning_rate": 4.9983400388810225e-06, + "loss": 2.5918, + "step": 1951 + }, + { + "epoch": 0.20642978003384094, + "grad_norm": 7.46875, + "learning_rate": 4.998333793254534e-06, + "loss": 2.5106, + "step": 1952 + }, + { + "epoch": 0.20653553299492386, + "grad_norm": 8.0625, + "learning_rate": 4.998327535904388e-06, + "loss": 2.7351, + "step": 1953 + }, + { + "epoch": 0.20664128595600678, + "grad_norm": 7.75, + "learning_rate": 4.998321266830612e-06, + "loss": 2.5843, + "step": 1954 + }, + { + "epoch": 0.20674703891708968, + "grad_norm": 7.6875, + "learning_rate": 4.998314986033237e-06, + "loss": 2.671, + "step": 1955 + }, + { + "epoch": 0.2068527918781726, + "grad_norm": 8.125, + "learning_rate": 4.998308693512291e-06, + "loss": 2.7486, + "step": 1956 + }, + { + "epoch": 0.2069585448392555, + "grad_norm": 7.6875, + "learning_rate": 4.998302389267803e-06, + "loss": 2.5398, + "step": 1957 + }, + { + "epoch": 0.20706429780033841, + "grad_norm": 9.0625, + "learning_rate": 4.998296073299805e-06, + "loss": 2.4896, + "step": 1958 + }, + { + "epoch": 0.2071700507614213, + "grad_norm": 8.1875, + "learning_rate": 4.998289745608324e-06, + "loss": 3.0601, + "step": 1959 + }, + { + "epoch": 0.20727580372250423, + "grad_norm": 10.5625, + "learning_rate": 4.998283406193393e-06, + "loss": 2.6766, + "step": 1960 + }, + { + "epoch": 0.20738155668358715, + "grad_norm": 8.25, + "learning_rate": 4.998277055055039e-06, + "loss": 2.819, + "step": 1961 + }, + { + "epoch": 0.20748730964467005, + "grad_norm": 7.75, + "learning_rate": 4.998270692193291e-06, + "loss": 2.6535, + "step": 1962 + }, + { + "epoch": 0.20759306260575297, + "grad_norm": 9.5, + "learning_rate": 4.9982643176081825e-06, + "loss": 2.7226, + "step": 1963 + }, + { + "epoch": 0.20769881556683586, + "grad_norm": 8.5, + "learning_rate": 4.9982579312997404e-06, + "loss": 2.7826, + "step": 1964 + }, + { + "epoch": 0.2078045685279188, + "grad_norm": 8.4375, + "learning_rate": 4.998251533267996e-06, + "loss": 2.489, + "step": 1965 + }, + { + "epoch": 0.20791032148900168, + "grad_norm": 8.375, + "learning_rate": 4.998245123512978e-06, + "loss": 2.7839, + "step": 1966 + }, + { + "epoch": 0.2080160744500846, + "grad_norm": 8.5625, + "learning_rate": 4.998238702034718e-06, + "loss": 2.3132, + "step": 1967 + }, + { + "epoch": 0.20812182741116753, + "grad_norm": 7.9375, + "learning_rate": 4.998232268833245e-06, + "loss": 2.5712, + "step": 1968 + }, + { + "epoch": 0.20822758037225042, + "grad_norm": 7.90625, + "learning_rate": 4.99822582390859e-06, + "loss": 2.6472, + "step": 1969 + }, + { + "epoch": 0.20833333333333334, + "grad_norm": 7.78125, + "learning_rate": 4.998219367260783e-06, + "loss": 2.6962, + "step": 1970 + }, + { + "epoch": 0.20843908629441624, + "grad_norm": 7.96875, + "learning_rate": 4.9982128988898545e-06, + "loss": 2.7924, + "step": 1971 + }, + { + "epoch": 0.20854483925549916, + "grad_norm": 7.40625, + "learning_rate": 4.998206418795835e-06, + "loss": 2.5747, + "step": 1972 + }, + { + "epoch": 0.20865059221658205, + "grad_norm": 7.75, + "learning_rate": 4.998199926978753e-06, + "loss": 2.5606, + "step": 1973 + }, + { + "epoch": 0.20875634517766498, + "grad_norm": 7.65625, + "learning_rate": 4.998193423438641e-06, + "loss": 2.6722, + "step": 1974 + }, + { + "epoch": 0.2088620981387479, + "grad_norm": 7.5625, + "learning_rate": 4.99818690817553e-06, + "loss": 2.5923, + "step": 1975 + }, + { + "epoch": 0.2089678510998308, + "grad_norm": 8.1875, + "learning_rate": 4.998180381189449e-06, + "loss": 2.5069, + "step": 1976 + }, + { + "epoch": 0.20907360406091371, + "grad_norm": 9.0625, + "learning_rate": 4.998173842480429e-06, + "loss": 2.6673, + "step": 1977 + }, + { + "epoch": 0.2091793570219966, + "grad_norm": 8.625, + "learning_rate": 4.9981672920485005e-06, + "loss": 2.7386, + "step": 1978 + }, + { + "epoch": 0.20928510998307953, + "grad_norm": 7.65625, + "learning_rate": 4.998160729893695e-06, + "loss": 2.6414, + "step": 1979 + }, + { + "epoch": 0.20939086294416243, + "grad_norm": 7.4375, + "learning_rate": 4.998154156016043e-06, + "loss": 2.6325, + "step": 1980 + }, + { + "epoch": 0.20949661590524535, + "grad_norm": 8.25, + "learning_rate": 4.9981475704155756e-06, + "loss": 2.5977, + "step": 1981 + }, + { + "epoch": 0.20960236886632827, + "grad_norm": 8.125, + "learning_rate": 4.998140973092323e-06, + "loss": 2.7308, + "step": 1982 + }, + { + "epoch": 0.20970812182741116, + "grad_norm": 7.53125, + "learning_rate": 4.998134364046317e-06, + "loss": 2.6283, + "step": 1983 + }, + { + "epoch": 0.2098138747884941, + "grad_norm": 8.5625, + "learning_rate": 4.998127743277587e-06, + "loss": 2.6697, + "step": 1984 + }, + { + "epoch": 0.20991962774957698, + "grad_norm": 9.75, + "learning_rate": 4.998121110786166e-06, + "loss": 2.9138, + "step": 1985 + }, + { + "epoch": 0.2100253807106599, + "grad_norm": 7.5, + "learning_rate": 4.998114466572084e-06, + "loss": 2.5715, + "step": 1986 + }, + { + "epoch": 0.2101311336717428, + "grad_norm": 21.625, + "learning_rate": 4.998107810635373e-06, + "loss": 2.7307, + "step": 1987 + }, + { + "epoch": 0.21023688663282572, + "grad_norm": 7.75, + "learning_rate": 4.998101142976063e-06, + "loss": 2.6988, + "step": 1988 + }, + { + "epoch": 0.21034263959390864, + "grad_norm": 8.625, + "learning_rate": 4.998094463594186e-06, + "loss": 2.7093, + "step": 1989 + }, + { + "epoch": 0.21044839255499154, + "grad_norm": 8.5, + "learning_rate": 4.9980877724897745e-06, + "loss": 2.6947, + "step": 1990 + }, + { + "epoch": 0.21055414551607446, + "grad_norm": 8.0625, + "learning_rate": 4.998081069662858e-06, + "loss": 2.5245, + "step": 1991 + }, + { + "epoch": 0.21065989847715735, + "grad_norm": 7.28125, + "learning_rate": 4.998074355113469e-06, + "loss": 2.466, + "step": 1992 + }, + { + "epoch": 0.21076565143824028, + "grad_norm": 7.53125, + "learning_rate": 4.9980676288416386e-06, + "loss": 2.5869, + "step": 1993 + }, + { + "epoch": 0.21087140439932317, + "grad_norm": 7.78125, + "learning_rate": 4.998060890847398e-06, + "loss": 2.5779, + "step": 1994 + }, + { + "epoch": 0.2109771573604061, + "grad_norm": 9.5625, + "learning_rate": 4.998054141130779e-06, + "loss": 2.4412, + "step": 1995 + }, + { + "epoch": 0.21108291032148901, + "grad_norm": 8.375, + "learning_rate": 4.998047379691815e-06, + "loss": 2.7339, + "step": 1996 + }, + { + "epoch": 0.2111886632825719, + "grad_norm": 9.375, + "learning_rate": 4.998040606530536e-06, + "loss": 2.6618, + "step": 1997 + }, + { + "epoch": 0.21129441624365483, + "grad_norm": 7.625, + "learning_rate": 4.998033821646974e-06, + "loss": 2.5915, + "step": 1998 + }, + { + "epoch": 0.21140016920473773, + "grad_norm": 8.25, + "learning_rate": 4.99802702504116e-06, + "loss": 2.6289, + "step": 1999 + }, + { + "epoch": 0.21150592216582065, + "grad_norm": 8.5, + "learning_rate": 4.998020216713128e-06, + "loss": 2.6913, + "step": 2000 + }, + { + "epoch": 0.21161167512690354, + "grad_norm": 9.4375, + "learning_rate": 4.998013396662908e-06, + "loss": 2.6928, + "step": 2001 + }, + { + "epoch": 0.21171742808798646, + "grad_norm": 8.6875, + "learning_rate": 4.998006564890533e-06, + "loss": 2.9065, + "step": 2002 + }, + { + "epoch": 0.2118231810490694, + "grad_norm": 8.75, + "learning_rate": 4.997999721396035e-06, + "loss": 2.7737, + "step": 2003 + }, + { + "epoch": 0.21192893401015228, + "grad_norm": 7.15625, + "learning_rate": 4.997992866179446e-06, + "loss": 2.5369, + "step": 2004 + }, + { + "epoch": 0.2120346869712352, + "grad_norm": 8.5625, + "learning_rate": 4.997985999240798e-06, + "loss": 2.6232, + "step": 2005 + }, + { + "epoch": 0.2121404399323181, + "grad_norm": 7.90625, + "learning_rate": 4.9979791205801235e-06, + "loss": 2.7994, + "step": 2006 + }, + { + "epoch": 0.21224619289340102, + "grad_norm": 7.875, + "learning_rate": 4.997972230197454e-06, + "loss": 2.6578, + "step": 2007 + }, + { + "epoch": 0.21235194585448391, + "grad_norm": 8.0625, + "learning_rate": 4.997965328092823e-06, + "loss": 2.8326, + "step": 2008 + }, + { + "epoch": 0.21245769881556684, + "grad_norm": 7.6875, + "learning_rate": 4.997958414266262e-06, + "loss": 2.5792, + "step": 2009 + }, + { + "epoch": 0.21256345177664976, + "grad_norm": 7.71875, + "learning_rate": 4.997951488717804e-06, + "loss": 2.6578, + "step": 2010 + }, + { + "epoch": 0.21266920473773265, + "grad_norm": 7.03125, + "learning_rate": 4.997944551447481e-06, + "loss": 2.6027, + "step": 2011 + }, + { + "epoch": 0.21277495769881558, + "grad_norm": 7.46875, + "learning_rate": 4.997937602455327e-06, + "loss": 2.7817, + "step": 2012 + }, + { + "epoch": 0.21288071065989847, + "grad_norm": 7.5625, + "learning_rate": 4.997930641741372e-06, + "loss": 2.6143, + "step": 2013 + }, + { + "epoch": 0.2129864636209814, + "grad_norm": 7.78125, + "learning_rate": 4.997923669305651e-06, + "loss": 2.61, + "step": 2014 + }, + { + "epoch": 0.2130922165820643, + "grad_norm": 7.59375, + "learning_rate": 4.9979166851481955e-06, + "loss": 2.619, + "step": 2015 + }, + { + "epoch": 0.2131979695431472, + "grad_norm": 7.90625, + "learning_rate": 4.997909689269039e-06, + "loss": 2.7264, + "step": 2016 + }, + { + "epoch": 0.21330372250423013, + "grad_norm": 7.5625, + "learning_rate": 4.997902681668213e-06, + "loss": 2.7771, + "step": 2017 + }, + { + "epoch": 0.21340947546531303, + "grad_norm": 7.375, + "learning_rate": 4.997895662345753e-06, + "loss": 2.55, + "step": 2018 + }, + { + "epoch": 0.21351522842639595, + "grad_norm": 7.71875, + "learning_rate": 4.9978886313016886e-06, + "loss": 2.7081, + "step": 2019 + }, + { + "epoch": 0.21362098138747884, + "grad_norm": 8.0625, + "learning_rate": 4.9978815885360556e-06, + "loss": 2.6677, + "step": 2020 + }, + { + "epoch": 0.21372673434856176, + "grad_norm": 8.875, + "learning_rate": 4.997874534048885e-06, + "loss": 2.5192, + "step": 2021 + }, + { + "epoch": 0.21383248730964466, + "grad_norm": 8.375, + "learning_rate": 4.997867467840211e-06, + "loss": 2.7813, + "step": 2022 + }, + { + "epoch": 0.21393824027072758, + "grad_norm": 8.0625, + "learning_rate": 4.997860389910066e-06, + "loss": 2.5616, + "step": 2023 + }, + { + "epoch": 0.2140439932318105, + "grad_norm": 12.0, + "learning_rate": 4.997853300258485e-06, + "loss": 2.5579, + "step": 2024 + }, + { + "epoch": 0.2141497461928934, + "grad_norm": 7.84375, + "learning_rate": 4.9978461988855e-06, + "loss": 2.7338, + "step": 2025 + }, + { + "epoch": 0.21425549915397632, + "grad_norm": 7.71875, + "learning_rate": 4.997839085791144e-06, + "loss": 2.8965, + "step": 2026 + }, + { + "epoch": 0.21436125211505921, + "grad_norm": 11.5625, + "learning_rate": 4.9978319609754505e-06, + "loss": 2.3056, + "step": 2027 + }, + { + "epoch": 0.21446700507614214, + "grad_norm": 7.75, + "learning_rate": 4.997824824438453e-06, + "loss": 2.5669, + "step": 2028 + }, + { + "epoch": 0.21457275803722503, + "grad_norm": 7.6875, + "learning_rate": 4.997817676180186e-06, + "loss": 2.5505, + "step": 2029 + }, + { + "epoch": 0.21467851099830795, + "grad_norm": 7.625, + "learning_rate": 4.997810516200682e-06, + "loss": 2.7201, + "step": 2030 + }, + { + "epoch": 0.21478426395939088, + "grad_norm": 7.84375, + "learning_rate": 4.997803344499974e-06, + "loss": 2.7146, + "step": 2031 + }, + { + "epoch": 0.21489001692047377, + "grad_norm": 12.375, + "learning_rate": 4.997796161078097e-06, + "loss": 2.8346, + "step": 2032 + }, + { + "epoch": 0.2149957698815567, + "grad_norm": 7.3125, + "learning_rate": 4.997788965935085e-06, + "loss": 2.436, + "step": 2033 + }, + { + "epoch": 0.2151015228426396, + "grad_norm": 8.3125, + "learning_rate": 4.99778175907097e-06, + "loss": 2.7191, + "step": 2034 + }, + { + "epoch": 0.2152072758037225, + "grad_norm": 7.59375, + "learning_rate": 4.997774540485787e-06, + "loss": 2.4845, + "step": 2035 + }, + { + "epoch": 0.2153130287648054, + "grad_norm": 7.59375, + "learning_rate": 4.997767310179569e-06, + "loss": 2.5335, + "step": 2036 + }, + { + "epoch": 0.21541878172588833, + "grad_norm": 7.6875, + "learning_rate": 4.997760068152352e-06, + "loss": 2.8148, + "step": 2037 + }, + { + "epoch": 0.21552453468697125, + "grad_norm": 7.625, + "learning_rate": 4.997752814404168e-06, + "loss": 2.7426, + "step": 2038 + }, + { + "epoch": 0.21563028764805414, + "grad_norm": 7.4375, + "learning_rate": 4.99774554893505e-06, + "loss": 2.7131, + "step": 2039 + }, + { + "epoch": 0.21573604060913706, + "grad_norm": 8.625, + "learning_rate": 4.997738271745035e-06, + "loss": 2.7822, + "step": 2040 + }, + { + "epoch": 0.21584179357021996, + "grad_norm": 7.71875, + "learning_rate": 4.997730982834157e-06, + "loss": 2.7015, + "step": 2041 + }, + { + "epoch": 0.21594754653130288, + "grad_norm": 11.0625, + "learning_rate": 4.997723682202447e-06, + "loss": 2.8267, + "step": 2042 + }, + { + "epoch": 0.21605329949238578, + "grad_norm": 7.96875, + "learning_rate": 4.997716369849942e-06, + "loss": 2.6449, + "step": 2043 + }, + { + "epoch": 0.2161590524534687, + "grad_norm": 7.71875, + "learning_rate": 4.997709045776676e-06, + "loss": 2.8438, + "step": 2044 + }, + { + "epoch": 0.21626480541455162, + "grad_norm": 7.5625, + "learning_rate": 4.997701709982683e-06, + "loss": 2.5132, + "step": 2045 + }, + { + "epoch": 0.21637055837563451, + "grad_norm": 7.46875, + "learning_rate": 4.997694362467997e-06, + "loss": 2.5002, + "step": 2046 + }, + { + "epoch": 0.21647631133671744, + "grad_norm": 8.25, + "learning_rate": 4.997687003232653e-06, + "loss": 2.6744, + "step": 2047 + }, + { + "epoch": 0.21658206429780033, + "grad_norm": 7.96875, + "learning_rate": 4.997679632276685e-06, + "loss": 2.5781, + "step": 2048 + }, + { + "epoch": 0.21668781725888325, + "grad_norm": 7.84375, + "learning_rate": 4.997672249600129e-06, + "loss": 2.678, + "step": 2049 + }, + { + "epoch": 0.21679357021996615, + "grad_norm": 8.4375, + "learning_rate": 4.997664855203017e-06, + "loss": 2.6078, + "step": 2050 + }, + { + "epoch": 0.21689932318104907, + "grad_norm": 8.0, + "learning_rate": 4.9976574490853866e-06, + "loss": 2.7953, + "step": 2051 + }, + { + "epoch": 0.217005076142132, + "grad_norm": 8.1875, + "learning_rate": 4.997650031247272e-06, + "loss": 2.4923, + "step": 2052 + }, + { + "epoch": 0.2171108291032149, + "grad_norm": 10.0625, + "learning_rate": 4.997642601688706e-06, + "loss": 2.7178, + "step": 2053 + }, + { + "epoch": 0.2172165820642978, + "grad_norm": 7.625, + "learning_rate": 4.997635160409725e-06, + "loss": 2.5723, + "step": 2054 + }, + { + "epoch": 0.2173223350253807, + "grad_norm": 9.8125, + "learning_rate": 4.997627707410364e-06, + "loss": 2.3657, + "step": 2055 + }, + { + "epoch": 0.21742808798646363, + "grad_norm": 7.59375, + "learning_rate": 4.997620242690657e-06, + "loss": 2.6353, + "step": 2056 + }, + { + "epoch": 0.21753384094754652, + "grad_norm": 8.1875, + "learning_rate": 4.997612766250641e-06, + "loss": 2.7017, + "step": 2057 + }, + { + "epoch": 0.21763959390862944, + "grad_norm": 7.625, + "learning_rate": 4.997605278090349e-06, + "loss": 2.799, + "step": 2058 + }, + { + "epoch": 0.21774534686971236, + "grad_norm": 8.5, + "learning_rate": 4.997597778209816e-06, + "loss": 2.5519, + "step": 2059 + }, + { + "epoch": 0.21785109983079526, + "grad_norm": 7.5625, + "learning_rate": 4.99759026660908e-06, + "loss": 2.6205, + "step": 2060 + }, + { + "epoch": 0.21795685279187818, + "grad_norm": 7.71875, + "learning_rate": 4.997582743288173e-06, + "loss": 2.5904, + "step": 2061 + }, + { + "epoch": 0.21806260575296108, + "grad_norm": 8.0, + "learning_rate": 4.997575208247132e-06, + "loss": 2.6048, + "step": 2062 + }, + { + "epoch": 0.218168358714044, + "grad_norm": 8.375, + "learning_rate": 4.9975676614859925e-06, + "loss": 2.8727, + "step": 2063 + }, + { + "epoch": 0.2182741116751269, + "grad_norm": 7.78125, + "learning_rate": 4.997560103004789e-06, + "loss": 2.6234, + "step": 2064 + }, + { + "epoch": 0.21837986463620981, + "grad_norm": 7.6875, + "learning_rate": 4.997552532803557e-06, + "loss": 2.7648, + "step": 2065 + }, + { + "epoch": 0.21848561759729274, + "grad_norm": 8.5625, + "learning_rate": 4.997544950882333e-06, + "loss": 2.71, + "step": 2066 + }, + { + "epoch": 0.21859137055837563, + "grad_norm": 8.875, + "learning_rate": 4.997537357241152e-06, + "loss": 2.7717, + "step": 2067 + }, + { + "epoch": 0.21869712351945855, + "grad_norm": 7.65625, + "learning_rate": 4.9975297518800495e-06, + "loss": 2.6599, + "step": 2068 + }, + { + "epoch": 0.21880287648054145, + "grad_norm": 8.125, + "learning_rate": 4.997522134799061e-06, + "loss": 2.8491, + "step": 2069 + }, + { + "epoch": 0.21890862944162437, + "grad_norm": 7.78125, + "learning_rate": 4.997514505998224e-06, + "loss": 2.6446, + "step": 2070 + }, + { + "epoch": 0.21901438240270726, + "grad_norm": 9.625, + "learning_rate": 4.9975068654775705e-06, + "loss": 2.589, + "step": 2071 + }, + { + "epoch": 0.21912013536379019, + "grad_norm": 9.0, + "learning_rate": 4.99749921323714e-06, + "loss": 2.6224, + "step": 2072 + }, + { + "epoch": 0.2192258883248731, + "grad_norm": 8.0, + "learning_rate": 4.997491549276967e-06, + "loss": 2.6624, + "step": 2073 + }, + { + "epoch": 0.219331641285956, + "grad_norm": 8.125, + "learning_rate": 4.997483873597087e-06, + "loss": 2.6023, + "step": 2074 + }, + { + "epoch": 0.21943739424703892, + "grad_norm": 7.46875, + "learning_rate": 4.997476186197539e-06, + "loss": 2.4544, + "step": 2075 + }, + { + "epoch": 0.21954314720812182, + "grad_norm": 9.0625, + "learning_rate": 4.997468487078354e-06, + "loss": 2.4098, + "step": 2076 + }, + { + "epoch": 0.21964890016920474, + "grad_norm": 7.5625, + "learning_rate": 4.997460776239572e-06, + "loss": 2.6389, + "step": 2077 + }, + { + "epoch": 0.21975465313028764, + "grad_norm": 8.0625, + "learning_rate": 4.997453053681229e-06, + "loss": 2.4858, + "step": 2078 + }, + { + "epoch": 0.21986040609137056, + "grad_norm": 9.5, + "learning_rate": 4.9974453194033575e-06, + "loss": 2.6351, + "step": 2079 + }, + { + "epoch": 0.21996615905245348, + "grad_norm": 8.0, + "learning_rate": 4.997437573405999e-06, + "loss": 2.4468, + "step": 2080 + }, + { + "epoch": 0.22007191201353637, + "grad_norm": 10.0, + "learning_rate": 4.997429815689187e-06, + "loss": 2.7354, + "step": 2081 + }, + { + "epoch": 0.2201776649746193, + "grad_norm": 7.875, + "learning_rate": 4.9974220462529574e-06, + "loss": 2.6089, + "step": 2082 + }, + { + "epoch": 0.2202834179357022, + "grad_norm": 8.625, + "learning_rate": 4.997414265097348e-06, + "loss": 2.5093, + "step": 2083 + }, + { + "epoch": 0.2203891708967851, + "grad_norm": 7.71875, + "learning_rate": 4.997406472222395e-06, + "loss": 2.7161, + "step": 2084 + }, + { + "epoch": 0.220494923857868, + "grad_norm": 7.875, + "learning_rate": 4.9973986676281335e-06, + "loss": 2.6019, + "step": 2085 + }, + { + "epoch": 0.22060067681895093, + "grad_norm": 8.375, + "learning_rate": 4.997390851314603e-06, + "loss": 2.9422, + "step": 2086 + }, + { + "epoch": 0.22070642978003385, + "grad_norm": 8.3125, + "learning_rate": 4.997383023281838e-06, + "loss": 2.6997, + "step": 2087 + }, + { + "epoch": 0.22081218274111675, + "grad_norm": 8.6875, + "learning_rate": 4.997375183529877e-06, + "loss": 2.7078, + "step": 2088 + }, + { + "epoch": 0.22091793570219967, + "grad_norm": 8.125, + "learning_rate": 4.997367332058754e-06, + "loss": 2.7138, + "step": 2089 + }, + { + "epoch": 0.22102368866328256, + "grad_norm": 8.625, + "learning_rate": 4.997359468868508e-06, + "loss": 2.848, + "step": 2090 + }, + { + "epoch": 0.22112944162436549, + "grad_norm": 10.125, + "learning_rate": 4.997351593959176e-06, + "loss": 2.3117, + "step": 2091 + }, + { + "epoch": 0.22123519458544838, + "grad_norm": 7.78125, + "learning_rate": 4.997343707330794e-06, + "loss": 2.7956, + "step": 2092 + }, + { + "epoch": 0.2213409475465313, + "grad_norm": 7.6875, + "learning_rate": 4.997335808983399e-06, + "loss": 2.5175, + "step": 2093 + }, + { + "epoch": 0.22144670050761422, + "grad_norm": 7.46875, + "learning_rate": 4.997327898917029e-06, + "loss": 2.7368, + "step": 2094 + }, + { + "epoch": 0.22155245346869712, + "grad_norm": 7.9375, + "learning_rate": 4.997319977131721e-06, + "loss": 2.7319, + "step": 2095 + }, + { + "epoch": 0.22165820642978004, + "grad_norm": 10.6875, + "learning_rate": 4.99731204362751e-06, + "loss": 2.8392, + "step": 2096 + }, + { + "epoch": 0.22176395939086294, + "grad_norm": 9.125, + "learning_rate": 4.997304098404436e-06, + "loss": 2.6514, + "step": 2097 + }, + { + "epoch": 0.22186971235194586, + "grad_norm": 7.875, + "learning_rate": 4.997296141462536e-06, + "loss": 2.6037, + "step": 2098 + }, + { + "epoch": 0.22197546531302875, + "grad_norm": 8.375, + "learning_rate": 4.997288172801845e-06, + "loss": 2.7498, + "step": 2099 + }, + { + "epoch": 0.22208121827411167, + "grad_norm": 7.84375, + "learning_rate": 4.997280192422403e-06, + "loss": 2.4078, + "step": 2100 + }, + { + "epoch": 0.2221869712351946, + "grad_norm": 7.625, + "learning_rate": 4.997272200324246e-06, + "loss": 2.6359, + "step": 2101 + }, + { + "epoch": 0.2222927241962775, + "grad_norm": 7.90625, + "learning_rate": 4.997264196507412e-06, + "loss": 2.8291, + "step": 2102 + }, + { + "epoch": 0.2223984771573604, + "grad_norm": 9.125, + "learning_rate": 4.997256180971939e-06, + "loss": 2.5576, + "step": 2103 + }, + { + "epoch": 0.2225042301184433, + "grad_norm": 7.6875, + "learning_rate": 4.997248153717864e-06, + "loss": 2.6446, + "step": 2104 + }, + { + "epoch": 0.22260998307952623, + "grad_norm": 7.65625, + "learning_rate": 4.9972401147452245e-06, + "loss": 2.713, + "step": 2105 + }, + { + "epoch": 0.22271573604060912, + "grad_norm": 7.5625, + "learning_rate": 4.997232064054058e-06, + "loss": 2.7052, + "step": 2106 + }, + { + "epoch": 0.22282148900169205, + "grad_norm": 7.9375, + "learning_rate": 4.997224001644403e-06, + "loss": 2.6041, + "step": 2107 + }, + { + "epoch": 0.22292724196277497, + "grad_norm": 7.6875, + "learning_rate": 4.997215927516297e-06, + "loss": 2.6225, + "step": 2108 + }, + { + "epoch": 0.22303299492385786, + "grad_norm": 8.625, + "learning_rate": 4.997207841669779e-06, + "loss": 2.8486, + "step": 2109 + }, + { + "epoch": 0.22313874788494079, + "grad_norm": 7.59375, + "learning_rate": 4.997199744104885e-06, + "loss": 2.5407, + "step": 2110 + }, + { + "epoch": 0.22324450084602368, + "grad_norm": 7.53125, + "learning_rate": 4.9971916348216534e-06, + "loss": 2.6719, + "step": 2111 + }, + { + "epoch": 0.2233502538071066, + "grad_norm": 8.0625, + "learning_rate": 4.997183513820123e-06, + "loss": 2.7953, + "step": 2112 + }, + { + "epoch": 0.2234560067681895, + "grad_norm": 9.1875, + "learning_rate": 4.9971753811003325e-06, + "loss": 2.4966, + "step": 2113 + }, + { + "epoch": 0.22356175972927242, + "grad_norm": 7.90625, + "learning_rate": 4.997167236662318e-06, + "loss": 2.7144, + "step": 2114 + }, + { + "epoch": 0.22366751269035534, + "grad_norm": 8.0625, + "learning_rate": 4.997159080506121e-06, + "loss": 2.3585, + "step": 2115 + }, + { + "epoch": 0.22377326565143824, + "grad_norm": 13.6875, + "learning_rate": 4.9971509126317765e-06, + "loss": 2.5506, + "step": 2116 + }, + { + "epoch": 0.22387901861252116, + "grad_norm": 7.53125, + "learning_rate": 4.997142733039323e-06, + "loss": 2.6007, + "step": 2117 + }, + { + "epoch": 0.22398477157360405, + "grad_norm": 7.5, + "learning_rate": 4.997134541728802e-06, + "loss": 2.5256, + "step": 2118 + }, + { + "epoch": 0.22409052453468697, + "grad_norm": 7.3125, + "learning_rate": 4.9971263387002485e-06, + "loss": 2.5288, + "step": 2119 + }, + { + "epoch": 0.22419627749576987, + "grad_norm": 7.96875, + "learning_rate": 4.997118123953703e-06, + "loss": 2.7286, + "step": 2120 + }, + { + "epoch": 0.2243020304568528, + "grad_norm": 7.3125, + "learning_rate": 4.997109897489203e-06, + "loss": 2.4545, + "step": 2121 + }, + { + "epoch": 0.2244077834179357, + "grad_norm": 7.65625, + "learning_rate": 4.997101659306788e-06, + "loss": 2.5492, + "step": 2122 + }, + { + "epoch": 0.2245135363790186, + "grad_norm": 8.8125, + "learning_rate": 4.997093409406496e-06, + "loss": 2.5512, + "step": 2123 + }, + { + "epoch": 0.22461928934010153, + "grad_norm": 7.625, + "learning_rate": 4.997085147788366e-06, + "loss": 2.6433, + "step": 2124 + }, + { + "epoch": 0.22472504230118442, + "grad_norm": 9.375, + "learning_rate": 4.9970768744524365e-06, + "loss": 2.6787, + "step": 2125 + }, + { + "epoch": 0.22483079526226735, + "grad_norm": 7.84375, + "learning_rate": 4.997068589398746e-06, + "loss": 2.7206, + "step": 2126 + }, + { + "epoch": 0.22493654822335024, + "grad_norm": 7.90625, + "learning_rate": 4.997060292627335e-06, + "loss": 2.5202, + "step": 2127 + }, + { + "epoch": 0.22504230118443316, + "grad_norm": 8.0, + "learning_rate": 4.99705198413824e-06, + "loss": 2.6877, + "step": 2128 + }, + { + "epoch": 0.22514805414551609, + "grad_norm": 9.5, + "learning_rate": 4.997043663931502e-06, + "loss": 2.6527, + "step": 2129 + }, + { + "epoch": 0.22525380710659898, + "grad_norm": 7.75, + "learning_rate": 4.9970353320071596e-06, + "loss": 2.7884, + "step": 2130 + }, + { + "epoch": 0.2253595600676819, + "grad_norm": 8.3125, + "learning_rate": 4.9970269883652514e-06, + "loss": 2.5843, + "step": 2131 + }, + { + "epoch": 0.2254653130287648, + "grad_norm": 7.90625, + "learning_rate": 4.9970186330058165e-06, + "loss": 2.829, + "step": 2132 + }, + { + "epoch": 0.22557106598984772, + "grad_norm": 8.125, + "learning_rate": 4.997010265928895e-06, + "loss": 2.7948, + "step": 2133 + }, + { + "epoch": 0.2256768189509306, + "grad_norm": 7.40625, + "learning_rate": 4.997001887134525e-06, + "loss": 2.6239, + "step": 2134 + }, + { + "epoch": 0.22578257191201354, + "grad_norm": 7.40625, + "learning_rate": 4.996993496622747e-06, + "loss": 2.666, + "step": 2135 + }, + { + "epoch": 0.22588832487309646, + "grad_norm": 9.625, + "learning_rate": 4.996985094393599e-06, + "loss": 2.5738, + "step": 2136 + }, + { + "epoch": 0.22599407783417935, + "grad_norm": 7.875, + "learning_rate": 4.996976680447122e-06, + "loss": 2.7546, + "step": 2137 + }, + { + "epoch": 0.22609983079526227, + "grad_norm": 7.96875, + "learning_rate": 4.996968254783354e-06, + "loss": 2.4551, + "step": 2138 + }, + { + "epoch": 0.22620558375634517, + "grad_norm": 7.65625, + "learning_rate": 4.9969598174023356e-06, + "loss": 2.6143, + "step": 2139 + }, + { + "epoch": 0.2263113367174281, + "grad_norm": 8.875, + "learning_rate": 4.996951368304106e-06, + "loss": 2.5284, + "step": 2140 + }, + { + "epoch": 0.22641708967851099, + "grad_norm": 8.25, + "learning_rate": 4.996942907488705e-06, + "loss": 2.9146, + "step": 2141 + }, + { + "epoch": 0.2265228426395939, + "grad_norm": 8.75, + "learning_rate": 4.996934434956172e-06, + "loss": 2.6923, + "step": 2142 + }, + { + "epoch": 0.22662859560067683, + "grad_norm": 7.65625, + "learning_rate": 4.9969259507065465e-06, + "loss": 2.6131, + "step": 2143 + }, + { + "epoch": 0.22673434856175972, + "grad_norm": 8.1875, + "learning_rate": 4.996917454739869e-06, + "loss": 2.4152, + "step": 2144 + }, + { + "epoch": 0.22684010152284265, + "grad_norm": 8.625, + "learning_rate": 4.996908947056179e-06, + "loss": 2.5877, + "step": 2145 + }, + { + "epoch": 0.22694585448392554, + "grad_norm": 7.59375, + "learning_rate": 4.996900427655517e-06, + "loss": 2.5876, + "step": 2146 + }, + { + "epoch": 0.22705160744500846, + "grad_norm": 7.75, + "learning_rate": 4.9968918965379225e-06, + "loss": 2.7012, + "step": 2147 + }, + { + "epoch": 0.22715736040609136, + "grad_norm": 7.90625, + "learning_rate": 4.996883353703434e-06, + "loss": 2.6333, + "step": 2148 + }, + { + "epoch": 0.22726311336717428, + "grad_norm": 8.625, + "learning_rate": 4.996874799152095e-06, + "loss": 2.5396, + "step": 2149 + }, + { + "epoch": 0.2273688663282572, + "grad_norm": 7.3125, + "learning_rate": 4.996866232883942e-06, + "loss": 2.5526, + "step": 2150 + }, + { + "epoch": 0.2274746192893401, + "grad_norm": 8.5, + "learning_rate": 4.996857654899019e-06, + "loss": 2.4475, + "step": 2151 + }, + { + "epoch": 0.22758037225042302, + "grad_norm": 7.78125, + "learning_rate": 4.996849065197363e-06, + "loss": 2.8291, + "step": 2152 + }, + { + "epoch": 0.2276861252115059, + "grad_norm": 8.125, + "learning_rate": 4.9968404637790154e-06, + "loss": 2.5906, + "step": 2153 + }, + { + "epoch": 0.22779187817258884, + "grad_norm": 7.71875, + "learning_rate": 4.996831850644017e-06, + "loss": 2.5527, + "step": 2154 + }, + { + "epoch": 0.22789763113367173, + "grad_norm": 8.875, + "learning_rate": 4.996823225792408e-06, + "loss": 2.5978, + "step": 2155 + }, + { + "epoch": 0.22800338409475465, + "grad_norm": 7.5625, + "learning_rate": 4.996814589224228e-06, + "loss": 2.4846, + "step": 2156 + }, + { + "epoch": 0.22810913705583757, + "grad_norm": 8.0625, + "learning_rate": 4.996805940939519e-06, + "loss": 2.6699, + "step": 2157 + }, + { + "epoch": 0.22821489001692047, + "grad_norm": 9.0625, + "learning_rate": 4.996797280938321e-06, + "loss": 2.6993, + "step": 2158 + }, + { + "epoch": 0.2283206429780034, + "grad_norm": 8.125, + "learning_rate": 4.996788609220674e-06, + "loss": 2.7596, + "step": 2159 + }, + { + "epoch": 0.22842639593908629, + "grad_norm": 9.875, + "learning_rate": 4.996779925786619e-06, + "loss": 2.6534, + "step": 2160 + }, + { + "epoch": 0.2285321489001692, + "grad_norm": 8.75, + "learning_rate": 4.996771230636198e-06, + "loss": 2.6208, + "step": 2161 + }, + { + "epoch": 0.2286379018612521, + "grad_norm": 8.0, + "learning_rate": 4.9967625237694505e-06, + "loss": 2.6634, + "step": 2162 + }, + { + "epoch": 0.22874365482233502, + "grad_norm": 8.0625, + "learning_rate": 4.996753805186417e-06, + "loss": 2.637, + "step": 2163 + }, + { + "epoch": 0.22884940778341795, + "grad_norm": 7.90625, + "learning_rate": 4.99674507488714e-06, + "loss": 2.5981, + "step": 2164 + }, + { + "epoch": 0.22895516074450084, + "grad_norm": 7.9375, + "learning_rate": 4.996736332871658e-06, + "loss": 2.5964, + "step": 2165 + }, + { + "epoch": 0.22906091370558376, + "grad_norm": 7.34375, + "learning_rate": 4.996727579140015e-06, + "loss": 2.5875, + "step": 2166 + }, + { + "epoch": 0.22916666666666666, + "grad_norm": 7.5, + "learning_rate": 4.99671881369225e-06, + "loss": 2.5568, + "step": 2167 + }, + { + "epoch": 0.22927241962774958, + "grad_norm": 7.8125, + "learning_rate": 4.996710036528404e-06, + "loss": 2.7954, + "step": 2168 + }, + { + "epoch": 0.22937817258883247, + "grad_norm": 7.84375, + "learning_rate": 4.99670124764852e-06, + "loss": 2.6369, + "step": 2169 + }, + { + "epoch": 0.2294839255499154, + "grad_norm": 9.5, + "learning_rate": 4.996692447052638e-06, + "loss": 2.6811, + "step": 2170 + }, + { + "epoch": 0.22958967851099832, + "grad_norm": 7.90625, + "learning_rate": 4.996683634740799e-06, + "loss": 2.5304, + "step": 2171 + }, + { + "epoch": 0.2296954314720812, + "grad_norm": 8.125, + "learning_rate": 4.996674810713045e-06, + "loss": 2.5757, + "step": 2172 + }, + { + "epoch": 0.22980118443316414, + "grad_norm": 7.90625, + "learning_rate": 4.996665974969417e-06, + "loss": 2.7545, + "step": 2173 + }, + { + "epoch": 0.22990693739424703, + "grad_norm": 7.5, + "learning_rate": 4.996657127509957e-06, + "loss": 2.3169, + "step": 2174 + }, + { + "epoch": 0.23001269035532995, + "grad_norm": 9.5625, + "learning_rate": 4.996648268334706e-06, + "loss": 2.438, + "step": 2175 + }, + { + "epoch": 0.23011844331641285, + "grad_norm": 8.5, + "learning_rate": 4.996639397443706e-06, + "loss": 2.697, + "step": 2176 + }, + { + "epoch": 0.23022419627749577, + "grad_norm": 7.6875, + "learning_rate": 4.996630514836999e-06, + "loss": 2.7488, + "step": 2177 + }, + { + "epoch": 0.2303299492385787, + "grad_norm": 8.1875, + "learning_rate": 4.996621620514625e-06, + "loss": 2.5705, + "step": 2178 + }, + { + "epoch": 0.23043570219966159, + "grad_norm": 8.375, + "learning_rate": 4.996612714476627e-06, + "loss": 2.646, + "step": 2179 + }, + { + "epoch": 0.2305414551607445, + "grad_norm": 7.875, + "learning_rate": 4.996603796723047e-06, + "loss": 2.6197, + "step": 2180 + }, + { + "epoch": 0.2306472081218274, + "grad_norm": 8.0625, + "learning_rate": 4.996594867253926e-06, + "loss": 2.5274, + "step": 2181 + }, + { + "epoch": 0.23075296108291032, + "grad_norm": 8.1875, + "learning_rate": 4.9965859260693065e-06, + "loss": 2.7053, + "step": 2182 + }, + { + "epoch": 0.23085871404399322, + "grad_norm": 8.0625, + "learning_rate": 4.99657697316923e-06, + "loss": 2.6247, + "step": 2183 + }, + { + "epoch": 0.23096446700507614, + "grad_norm": 8.375, + "learning_rate": 4.99656800855374e-06, + "loss": 2.8091, + "step": 2184 + }, + { + "epoch": 0.23107021996615906, + "grad_norm": 7.90625, + "learning_rate": 4.996559032222877e-06, + "loss": 2.8756, + "step": 2185 + }, + { + "epoch": 0.23117597292724196, + "grad_norm": 7.71875, + "learning_rate": 4.996550044176682e-06, + "loss": 2.4979, + "step": 2186 + }, + { + "epoch": 0.23128172588832488, + "grad_norm": 8.125, + "learning_rate": 4.996541044415201e-06, + "loss": 2.5699, + "step": 2187 + }, + { + "epoch": 0.23138747884940777, + "grad_norm": 8.25, + "learning_rate": 4.996532032938472e-06, + "loss": 2.32, + "step": 2188 + }, + { + "epoch": 0.2314932318104907, + "grad_norm": 7.6875, + "learning_rate": 4.99652300974654e-06, + "loss": 2.6703, + "step": 2189 + }, + { + "epoch": 0.2315989847715736, + "grad_norm": 7.6875, + "learning_rate": 4.996513974839447e-06, + "loss": 2.7156, + "step": 2190 + }, + { + "epoch": 0.2317047377326565, + "grad_norm": 7.625, + "learning_rate": 4.996504928217235e-06, + "loss": 2.7664, + "step": 2191 + }, + { + "epoch": 0.23181049069373943, + "grad_norm": 8.6875, + "learning_rate": 4.996495869879946e-06, + "loss": 2.5851, + "step": 2192 + }, + { + "epoch": 0.23191624365482233, + "grad_norm": 7.84375, + "learning_rate": 4.996486799827622e-06, + "loss": 2.7535, + "step": 2193 + }, + { + "epoch": 0.23202199661590525, + "grad_norm": 7.1875, + "learning_rate": 4.996477718060307e-06, + "loss": 2.655, + "step": 2194 + }, + { + "epoch": 0.23212774957698815, + "grad_norm": 8.1875, + "learning_rate": 4.996468624578044e-06, + "loss": 2.6049, + "step": 2195 + }, + { + "epoch": 0.23223350253807107, + "grad_norm": 7.53125, + "learning_rate": 4.996459519380874e-06, + "loss": 2.6411, + "step": 2196 + }, + { + "epoch": 0.23233925549915396, + "grad_norm": 8.625, + "learning_rate": 4.996450402468841e-06, + "loss": 2.5603, + "step": 2197 + }, + { + "epoch": 0.23244500846023688, + "grad_norm": 7.34375, + "learning_rate": 4.996441273841987e-06, + "loss": 2.5929, + "step": 2198 + }, + { + "epoch": 0.2325507614213198, + "grad_norm": 8.3125, + "learning_rate": 4.996432133500355e-06, + "loss": 2.6476, + "step": 2199 + }, + { + "epoch": 0.2326565143824027, + "grad_norm": 7.65625, + "learning_rate": 4.996422981443988e-06, + "loss": 2.6737, + "step": 2200 + }, + { + "epoch": 0.23276226734348562, + "grad_norm": 7.46875, + "learning_rate": 4.996413817672929e-06, + "loss": 2.7049, + "step": 2201 + }, + { + "epoch": 0.23286802030456852, + "grad_norm": 8.0, + "learning_rate": 4.996404642187221e-06, + "loss": 2.7603, + "step": 2202 + }, + { + "epoch": 0.23297377326565144, + "grad_norm": 7.90625, + "learning_rate": 4.996395454986908e-06, + "loss": 2.6548, + "step": 2203 + }, + { + "epoch": 0.23307952622673433, + "grad_norm": 7.5625, + "learning_rate": 4.99638625607203e-06, + "loss": 2.7115, + "step": 2204 + }, + { + "epoch": 0.23318527918781726, + "grad_norm": 8.8125, + "learning_rate": 4.996377045442634e-06, + "loss": 2.7392, + "step": 2205 + }, + { + "epoch": 0.23329103214890018, + "grad_norm": 7.875, + "learning_rate": 4.9963678230987615e-06, + "loss": 2.4993, + "step": 2206 + }, + { + "epoch": 0.23339678510998307, + "grad_norm": 8.375, + "learning_rate": 4.996358589040454e-06, + "loss": 2.772, + "step": 2207 + }, + { + "epoch": 0.233502538071066, + "grad_norm": 8.3125, + "learning_rate": 4.996349343267758e-06, + "loss": 2.4937, + "step": 2208 + }, + { + "epoch": 0.2336082910321489, + "grad_norm": 7.4375, + "learning_rate": 4.996340085780716e-06, + "loss": 2.5282, + "step": 2209 + }, + { + "epoch": 0.2337140439932318, + "grad_norm": 7.625, + "learning_rate": 4.9963308165793704e-06, + "loss": 2.5912, + "step": 2210 + }, + { + "epoch": 0.2338197969543147, + "grad_norm": 7.78125, + "learning_rate": 4.9963215356637645e-06, + "loss": 2.5906, + "step": 2211 + }, + { + "epoch": 0.23392554991539763, + "grad_norm": 8.3125, + "learning_rate": 4.9963122430339435e-06, + "loss": 2.7387, + "step": 2212 + }, + { + "epoch": 0.23403130287648055, + "grad_norm": 7.5625, + "learning_rate": 4.99630293868995e-06, + "loss": 2.6773, + "step": 2213 + }, + { + "epoch": 0.23413705583756345, + "grad_norm": 7.625, + "learning_rate": 4.996293622631827e-06, + "loss": 2.6682, + "step": 2214 + }, + { + "epoch": 0.23424280879864637, + "grad_norm": 8.875, + "learning_rate": 4.99628429485962e-06, + "loss": 2.4217, + "step": 2215 + }, + { + "epoch": 0.23434856175972926, + "grad_norm": 7.375, + "learning_rate": 4.996274955373371e-06, + "loss": 2.6343, + "step": 2216 + }, + { + "epoch": 0.23445431472081218, + "grad_norm": 8.8125, + "learning_rate": 4.996265604173125e-06, + "loss": 2.5985, + "step": 2217 + }, + { + "epoch": 0.2345600676818951, + "grad_norm": 8.0, + "learning_rate": 4.996256241258925e-06, + "loss": 2.717, + "step": 2218 + }, + { + "epoch": 0.234665820642978, + "grad_norm": 7.875, + "learning_rate": 4.996246866630816e-06, + "loss": 2.7333, + "step": 2219 + }, + { + "epoch": 0.23477157360406092, + "grad_norm": 7.4375, + "learning_rate": 4.996237480288842e-06, + "loss": 2.6606, + "step": 2220 + }, + { + "epoch": 0.23487732656514382, + "grad_norm": 7.65625, + "learning_rate": 4.996228082233045e-06, + "loss": 2.8039, + "step": 2221 + }, + { + "epoch": 0.23498307952622674, + "grad_norm": 7.5, + "learning_rate": 4.996218672463471e-06, + "loss": 2.6897, + "step": 2222 + }, + { + "epoch": 0.23508883248730963, + "grad_norm": 7.65625, + "learning_rate": 4.996209250980164e-06, + "loss": 2.532, + "step": 2223 + }, + { + "epoch": 0.23519458544839256, + "grad_norm": 7.71875, + "learning_rate": 4.996199817783167e-06, + "loss": 2.7125, + "step": 2224 + }, + { + "epoch": 0.23530033840947548, + "grad_norm": 7.5625, + "learning_rate": 4.996190372872526e-06, + "loss": 2.674, + "step": 2225 + }, + { + "epoch": 0.23540609137055837, + "grad_norm": 7.65625, + "learning_rate": 4.996180916248284e-06, + "loss": 2.5275, + "step": 2226 + }, + { + "epoch": 0.2355118443316413, + "grad_norm": 7.75, + "learning_rate": 4.996171447910486e-06, + "loss": 2.5681, + "step": 2227 + }, + { + "epoch": 0.2356175972927242, + "grad_norm": 7.65625, + "learning_rate": 4.996161967859176e-06, + "loss": 2.678, + "step": 2228 + }, + { + "epoch": 0.2357233502538071, + "grad_norm": 7.5625, + "learning_rate": 4.996152476094401e-06, + "loss": 2.6003, + "step": 2229 + }, + { + "epoch": 0.23582910321489, + "grad_norm": 7.96875, + "learning_rate": 4.9961429726162e-06, + "loss": 2.6593, + "step": 2230 + }, + { + "epoch": 0.23593485617597293, + "grad_norm": 7.4375, + "learning_rate": 4.996133457424624e-06, + "loss": 2.7618, + "step": 2231 + }, + { + "epoch": 0.23604060913705585, + "grad_norm": 8.25, + "learning_rate": 4.996123930519712e-06, + "loss": 2.5355, + "step": 2232 + }, + { + "epoch": 0.23614636209813875, + "grad_norm": 7.75, + "learning_rate": 4.996114391901513e-06, + "loss": 2.7754, + "step": 2233 + }, + { + "epoch": 0.23625211505922167, + "grad_norm": 7.28125, + "learning_rate": 4.996104841570069e-06, + "loss": 2.5779, + "step": 2234 + }, + { + "epoch": 0.23635786802030456, + "grad_norm": 8.375, + "learning_rate": 4.996095279525426e-06, + "loss": 2.5859, + "step": 2235 + }, + { + "epoch": 0.23646362098138748, + "grad_norm": 7.34375, + "learning_rate": 4.996085705767629e-06, + "loss": 2.7843, + "step": 2236 + }, + { + "epoch": 0.23656937394247038, + "grad_norm": 7.90625, + "learning_rate": 4.9960761202967225e-06, + "loss": 2.6083, + "step": 2237 + }, + { + "epoch": 0.2366751269035533, + "grad_norm": 8.0625, + "learning_rate": 4.996066523112751e-06, + "loss": 2.8844, + "step": 2238 + }, + { + "epoch": 0.23678087986463622, + "grad_norm": 7.6875, + "learning_rate": 4.996056914215762e-06, + "loss": 2.747, + "step": 2239 + }, + { + "epoch": 0.23688663282571912, + "grad_norm": 9.0, + "learning_rate": 4.9960472936057976e-06, + "loss": 2.5768, + "step": 2240 + }, + { + "epoch": 0.23699238578680204, + "grad_norm": 8.0625, + "learning_rate": 4.996037661282904e-06, + "loss": 2.7355, + "step": 2241 + }, + { + "epoch": 0.23709813874788493, + "grad_norm": 7.75, + "learning_rate": 4.9960280172471265e-06, + "loss": 2.5481, + "step": 2242 + }, + { + "epoch": 0.23720389170896786, + "grad_norm": 7.65625, + "learning_rate": 4.99601836149851e-06, + "loss": 2.5601, + "step": 2243 + }, + { + "epoch": 0.23730964467005075, + "grad_norm": 7.78125, + "learning_rate": 4.996008694037101e-06, + "loss": 2.6071, + "step": 2244 + }, + { + "epoch": 0.23741539763113367, + "grad_norm": 8.125, + "learning_rate": 4.995999014862944e-06, + "loss": 2.7551, + "step": 2245 + }, + { + "epoch": 0.2375211505922166, + "grad_norm": 7.59375, + "learning_rate": 4.995989323976084e-06, + "loss": 2.7665, + "step": 2246 + }, + { + "epoch": 0.2376269035532995, + "grad_norm": 7.8125, + "learning_rate": 4.995979621376567e-06, + "loss": 2.5224, + "step": 2247 + }, + { + "epoch": 0.2377326565143824, + "grad_norm": 7.78125, + "learning_rate": 4.995969907064438e-06, + "loss": 2.6375, + "step": 2248 + }, + { + "epoch": 0.2378384094754653, + "grad_norm": 7.40625, + "learning_rate": 4.995960181039744e-06, + "loss": 2.6572, + "step": 2249 + }, + { + "epoch": 0.23794416243654823, + "grad_norm": 8.875, + "learning_rate": 4.99595044330253e-06, + "loss": 2.7551, + "step": 2250 + }, + { + "epoch": 0.23804991539763112, + "grad_norm": 7.9375, + "learning_rate": 4.99594069385284e-06, + "loss": 2.6311, + "step": 2251 + }, + { + "epoch": 0.23815566835871405, + "grad_norm": 8.75, + "learning_rate": 4.995930932690722e-06, + "loss": 2.3336, + "step": 2252 + }, + { + "epoch": 0.23826142131979697, + "grad_norm": 8.5625, + "learning_rate": 4.9959211598162205e-06, + "loss": 2.4563, + "step": 2253 + }, + { + "epoch": 0.23836717428087986, + "grad_norm": 8.0625, + "learning_rate": 4.995911375229382e-06, + "loss": 2.6485, + "step": 2254 + }, + { + "epoch": 0.23847292724196278, + "grad_norm": 7.875, + "learning_rate": 4.9959015789302526e-06, + "loss": 2.7707, + "step": 2255 + }, + { + "epoch": 0.23857868020304568, + "grad_norm": 7.5625, + "learning_rate": 4.995891770918877e-06, + "loss": 2.5195, + "step": 2256 + }, + { + "epoch": 0.2386844331641286, + "grad_norm": 7.625, + "learning_rate": 4.995881951195303e-06, + "loss": 2.7497, + "step": 2257 + }, + { + "epoch": 0.2387901861252115, + "grad_norm": 8.0, + "learning_rate": 4.995872119759575e-06, + "loss": 2.7277, + "step": 2258 + }, + { + "epoch": 0.23889593908629442, + "grad_norm": 8.5, + "learning_rate": 4.995862276611741e-06, + "loss": 2.5897, + "step": 2259 + }, + { + "epoch": 0.23900169204737734, + "grad_norm": 7.46875, + "learning_rate": 4.995852421751844e-06, + "loss": 2.4189, + "step": 2260 + }, + { + "epoch": 0.23910744500846023, + "grad_norm": 8.875, + "learning_rate": 4.9958425551799335e-06, + "loss": 2.5189, + "step": 2261 + }, + { + "epoch": 0.23921319796954316, + "grad_norm": 7.84375, + "learning_rate": 4.995832676896055e-06, + "loss": 2.4464, + "step": 2262 + }, + { + "epoch": 0.23931895093062605, + "grad_norm": 7.9375, + "learning_rate": 4.995822786900255e-06, + "loss": 2.6118, + "step": 2263 + }, + { + "epoch": 0.23942470389170897, + "grad_norm": 7.8125, + "learning_rate": 4.9958128851925794e-06, + "loss": 2.7472, + "step": 2264 + }, + { + "epoch": 0.23953045685279187, + "grad_norm": 7.8125, + "learning_rate": 4.995802971773074e-06, + "loss": 2.683, + "step": 2265 + }, + { + "epoch": 0.2396362098138748, + "grad_norm": 8.3125, + "learning_rate": 4.995793046641786e-06, + "loss": 2.7595, + "step": 2266 + }, + { + "epoch": 0.2397419627749577, + "grad_norm": 8.1875, + "learning_rate": 4.9957831097987615e-06, + "loss": 2.6066, + "step": 2267 + }, + { + "epoch": 0.2398477157360406, + "grad_norm": 8.5, + "learning_rate": 4.995773161244049e-06, + "loss": 2.6394, + "step": 2268 + }, + { + "epoch": 0.23995346869712353, + "grad_norm": 8.625, + "learning_rate": 4.995763200977692e-06, + "loss": 2.4528, + "step": 2269 + }, + { + "epoch": 0.24005922165820642, + "grad_norm": 7.78125, + "learning_rate": 4.99575322899974e-06, + "loss": 2.2778, + "step": 2270 + }, + { + "epoch": 0.24016497461928935, + "grad_norm": 7.25, + "learning_rate": 4.99574324531024e-06, + "loss": 2.5045, + "step": 2271 + }, + { + "epoch": 0.24027072758037224, + "grad_norm": 7.40625, + "learning_rate": 4.995733249909236e-06, + "loss": 2.6143, + "step": 2272 + }, + { + "epoch": 0.24037648054145516, + "grad_norm": 8.3125, + "learning_rate": 4.9957232427967765e-06, + "loss": 2.818, + "step": 2273 + }, + { + "epoch": 0.24048223350253808, + "grad_norm": 7.75, + "learning_rate": 4.995713223972909e-06, + "loss": 2.5238, + "step": 2274 + }, + { + "epoch": 0.24058798646362098, + "grad_norm": 8.0625, + "learning_rate": 4.995703193437681e-06, + "loss": 2.5358, + "step": 2275 + }, + { + "epoch": 0.2406937394247039, + "grad_norm": 8.0, + "learning_rate": 4.995693151191138e-06, + "loss": 2.5628, + "step": 2276 + }, + { + "epoch": 0.2407994923857868, + "grad_norm": 8.3125, + "learning_rate": 4.995683097233327e-06, + "loss": 2.6861, + "step": 2277 + }, + { + "epoch": 0.24090524534686972, + "grad_norm": 8.5, + "learning_rate": 4.995673031564296e-06, + "loss": 2.7526, + "step": 2278 + }, + { + "epoch": 0.2410109983079526, + "grad_norm": 7.75, + "learning_rate": 4.995662954184093e-06, + "loss": 2.5391, + "step": 2279 + }, + { + "epoch": 0.24111675126903553, + "grad_norm": 7.5625, + "learning_rate": 4.9956528650927635e-06, + "loss": 2.6143, + "step": 2280 + }, + { + "epoch": 0.24122250423011846, + "grad_norm": 7.53125, + "learning_rate": 4.995642764290357e-06, + "loss": 2.6674, + "step": 2281 + }, + { + "epoch": 0.24132825719120135, + "grad_norm": 7.96875, + "learning_rate": 4.995632651776919e-06, + "loss": 2.496, + "step": 2282 + }, + { + "epoch": 0.24143401015228427, + "grad_norm": 8.0625, + "learning_rate": 4.995622527552498e-06, + "loss": 2.6419, + "step": 2283 + }, + { + "epoch": 0.24153976311336717, + "grad_norm": 7.4375, + "learning_rate": 4.99561239161714e-06, + "loss": 2.6564, + "step": 2284 + }, + { + "epoch": 0.2416455160744501, + "grad_norm": 7.625, + "learning_rate": 4.995602243970895e-06, + "loss": 2.2574, + "step": 2285 + }, + { + "epoch": 0.24175126903553298, + "grad_norm": 8.25, + "learning_rate": 4.995592084613809e-06, + "loss": 2.9663, + "step": 2286 + }, + { + "epoch": 0.2418570219966159, + "grad_norm": 7.65625, + "learning_rate": 4.99558191354593e-06, + "loss": 2.5372, + "step": 2287 + }, + { + "epoch": 0.24196277495769883, + "grad_norm": 10.375, + "learning_rate": 4.995571730767306e-06, + "loss": 2.7796, + "step": 2288 + }, + { + "epoch": 0.24206852791878172, + "grad_norm": 10.125, + "learning_rate": 4.995561536277984e-06, + "loss": 2.6682, + "step": 2289 + }, + { + "epoch": 0.24217428087986465, + "grad_norm": 7.9375, + "learning_rate": 4.995551330078012e-06, + "loss": 2.7634, + "step": 2290 + }, + { + "epoch": 0.24228003384094754, + "grad_norm": 7.96875, + "learning_rate": 4.995541112167439e-06, + "loss": 2.6426, + "step": 2291 + }, + { + "epoch": 0.24238578680203046, + "grad_norm": 7.78125, + "learning_rate": 4.995530882546312e-06, + "loss": 2.5776, + "step": 2292 + }, + { + "epoch": 0.24249153976311336, + "grad_norm": 7.90625, + "learning_rate": 4.995520641214679e-06, + "loss": 2.7807, + "step": 2293 + }, + { + "epoch": 0.24259729272419628, + "grad_norm": 7.78125, + "learning_rate": 4.995510388172589e-06, + "loss": 2.6026, + "step": 2294 + }, + { + "epoch": 0.2427030456852792, + "grad_norm": 8.375, + "learning_rate": 4.995500123420088e-06, + "loss": 2.54, + "step": 2295 + }, + { + "epoch": 0.2428087986463621, + "grad_norm": 7.65625, + "learning_rate": 4.9954898469572265e-06, + "loss": 2.5085, + "step": 2296 + }, + { + "epoch": 0.24291455160744502, + "grad_norm": 8.1875, + "learning_rate": 4.995479558784052e-06, + "loss": 2.6788, + "step": 2297 + }, + { + "epoch": 0.2430203045685279, + "grad_norm": 7.71875, + "learning_rate": 4.995469258900612e-06, + "loss": 2.3979, + "step": 2298 + }, + { + "epoch": 0.24312605752961083, + "grad_norm": 9.0625, + "learning_rate": 4.9954589473069555e-06, + "loss": 2.9606, + "step": 2299 + }, + { + "epoch": 0.24323181049069373, + "grad_norm": 7.90625, + "learning_rate": 4.9954486240031305e-06, + "loss": 2.6073, + "step": 2300 + }, + { + "epoch": 0.24333756345177665, + "grad_norm": 7.84375, + "learning_rate": 4.995438288989186e-06, + "loss": 2.661, + "step": 2301 + }, + { + "epoch": 0.24344331641285957, + "grad_norm": 7.90625, + "learning_rate": 4.995427942265171e-06, + "loss": 2.5209, + "step": 2302 + }, + { + "epoch": 0.24354906937394247, + "grad_norm": 7.75, + "learning_rate": 4.995417583831133e-06, + "loss": 2.6008, + "step": 2303 + }, + { + "epoch": 0.2436548223350254, + "grad_norm": 8.9375, + "learning_rate": 4.995407213687119e-06, + "loss": 2.5736, + "step": 2304 + }, + { + "epoch": 0.24376057529610828, + "grad_norm": 7.6875, + "learning_rate": 4.995396831833181e-06, + "loss": 2.5861, + "step": 2305 + }, + { + "epoch": 0.2438663282571912, + "grad_norm": 8.625, + "learning_rate": 4.9953864382693664e-06, + "loss": 2.7752, + "step": 2306 + }, + { + "epoch": 0.2439720812182741, + "grad_norm": 8.0625, + "learning_rate": 4.9953760329957244e-06, + "loss": 2.6728, + "step": 2307 + }, + { + "epoch": 0.24407783417935702, + "grad_norm": 7.96875, + "learning_rate": 4.995365616012302e-06, + "loss": 2.6708, + "step": 2308 + }, + { + "epoch": 0.24418358714043994, + "grad_norm": 8.0625, + "learning_rate": 4.99535518731915e-06, + "loss": 2.7714, + "step": 2309 + }, + { + "epoch": 0.24428934010152284, + "grad_norm": 8.9375, + "learning_rate": 4.995344746916317e-06, + "loss": 2.6293, + "step": 2310 + }, + { + "epoch": 0.24439509306260576, + "grad_norm": 8.1875, + "learning_rate": 4.99533429480385e-06, + "loss": 2.6625, + "step": 2311 + }, + { + "epoch": 0.24450084602368866, + "grad_norm": 8.75, + "learning_rate": 4.995323830981801e-06, + "loss": 2.4628, + "step": 2312 + }, + { + "epoch": 0.24460659898477158, + "grad_norm": 8.1875, + "learning_rate": 4.995313355450217e-06, + "loss": 2.348, + "step": 2313 + }, + { + "epoch": 0.24471235194585447, + "grad_norm": 7.96875, + "learning_rate": 4.9953028682091495e-06, + "loss": 2.8817, + "step": 2314 + }, + { + "epoch": 0.2448181049069374, + "grad_norm": 7.8125, + "learning_rate": 4.995292369258645e-06, + "loss": 2.665, + "step": 2315 + }, + { + "epoch": 0.24492385786802032, + "grad_norm": 7.9375, + "learning_rate": 4.995281858598754e-06, + "loss": 2.5928, + "step": 2316 + }, + { + "epoch": 0.2450296108291032, + "grad_norm": 7.4375, + "learning_rate": 4.995271336229527e-06, + "loss": 2.7522, + "step": 2317 + }, + { + "epoch": 0.24513536379018613, + "grad_norm": 8.0625, + "learning_rate": 4.995260802151011e-06, + "loss": 2.8034, + "step": 2318 + }, + { + "epoch": 0.24524111675126903, + "grad_norm": 7.96875, + "learning_rate": 4.995250256363256e-06, + "loss": 2.6799, + "step": 2319 + }, + { + "epoch": 0.24534686971235195, + "grad_norm": 8.375, + "learning_rate": 4.995239698866313e-06, + "loss": 2.7045, + "step": 2320 + }, + { + "epoch": 0.24545262267343484, + "grad_norm": 8.1875, + "learning_rate": 4.995229129660231e-06, + "loss": 2.5488, + "step": 2321 + }, + { + "epoch": 0.24555837563451777, + "grad_norm": 7.5625, + "learning_rate": 4.995218548745059e-06, + "loss": 2.6892, + "step": 2322 + }, + { + "epoch": 0.2456641285956007, + "grad_norm": 7.78125, + "learning_rate": 4.995207956120846e-06, + "loss": 2.5879, + "step": 2323 + }, + { + "epoch": 0.24576988155668358, + "grad_norm": 7.71875, + "learning_rate": 4.995197351787643e-06, + "loss": 2.5581, + "step": 2324 + }, + { + "epoch": 0.2458756345177665, + "grad_norm": 7.9375, + "learning_rate": 4.995186735745499e-06, + "loss": 2.6868, + "step": 2325 + }, + { + "epoch": 0.2459813874788494, + "grad_norm": 8.4375, + "learning_rate": 4.995176107994465e-06, + "loss": 2.911, + "step": 2326 + }, + { + "epoch": 0.24608714043993232, + "grad_norm": 9.125, + "learning_rate": 4.995165468534589e-06, + "loss": 2.8772, + "step": 2327 + }, + { + "epoch": 0.24619289340101522, + "grad_norm": 8.4375, + "learning_rate": 4.995154817365922e-06, + "loss": 2.6608, + "step": 2328 + }, + { + "epoch": 0.24629864636209814, + "grad_norm": 7.9375, + "learning_rate": 4.995144154488516e-06, + "loss": 2.5492, + "step": 2329 + }, + { + "epoch": 0.24640439932318106, + "grad_norm": 7.3125, + "learning_rate": 4.995133479902417e-06, + "loss": 2.6338, + "step": 2330 + }, + { + "epoch": 0.24651015228426396, + "grad_norm": 7.78125, + "learning_rate": 4.995122793607677e-06, + "loss": 2.6026, + "step": 2331 + }, + { + "epoch": 0.24661590524534688, + "grad_norm": 8.625, + "learning_rate": 4.995112095604346e-06, + "loss": 2.8578, + "step": 2332 + }, + { + "epoch": 0.24672165820642977, + "grad_norm": 8.0625, + "learning_rate": 4.9951013858924755e-06, + "loss": 2.6096, + "step": 2333 + }, + { + "epoch": 0.2468274111675127, + "grad_norm": 7.875, + "learning_rate": 4.995090664472114e-06, + "loss": 2.5634, + "step": 2334 + }, + { + "epoch": 0.2469331641285956, + "grad_norm": 7.1875, + "learning_rate": 4.995079931343313e-06, + "loss": 2.5111, + "step": 2335 + }, + { + "epoch": 0.2470389170896785, + "grad_norm": 8.1875, + "learning_rate": 4.995069186506122e-06, + "loss": 2.7512, + "step": 2336 + }, + { + "epoch": 0.24714467005076143, + "grad_norm": 7.96875, + "learning_rate": 4.995058429960592e-06, + "loss": 2.7061, + "step": 2337 + }, + { + "epoch": 0.24725042301184433, + "grad_norm": 7.9375, + "learning_rate": 4.995047661706773e-06, + "loss": 2.7554, + "step": 2338 + }, + { + "epoch": 0.24735617597292725, + "grad_norm": 7.5, + "learning_rate": 4.9950368817447155e-06, + "loss": 2.6335, + "step": 2339 + }, + { + "epoch": 0.24746192893401014, + "grad_norm": 8.5, + "learning_rate": 4.99502609007447e-06, + "loss": 2.6845, + "step": 2340 + }, + { + "epoch": 0.24756768189509307, + "grad_norm": 7.84375, + "learning_rate": 4.9950152866960885e-06, + "loss": 2.6305, + "step": 2341 + }, + { + "epoch": 0.24767343485617596, + "grad_norm": 14.125, + "learning_rate": 4.995004471609621e-06, + "loss": 2.4841, + "step": 2342 + }, + { + "epoch": 0.24777918781725888, + "grad_norm": 7.625, + "learning_rate": 4.9949936448151174e-06, + "loss": 2.7993, + "step": 2343 + }, + { + "epoch": 0.2478849407783418, + "grad_norm": 8.0, + "learning_rate": 4.994982806312628e-06, + "loss": 2.3304, + "step": 2344 + }, + { + "epoch": 0.2479906937394247, + "grad_norm": 7.96875, + "learning_rate": 4.9949719561022055e-06, + "loss": 2.8885, + "step": 2345 + }, + { + "epoch": 0.24809644670050762, + "grad_norm": 7.78125, + "learning_rate": 4.9949610941839e-06, + "loss": 2.6412, + "step": 2346 + }, + { + "epoch": 0.24820219966159052, + "grad_norm": 7.34375, + "learning_rate": 4.9949502205577634e-06, + "loss": 2.4509, + "step": 2347 + }, + { + "epoch": 0.24830795262267344, + "grad_norm": 8.25, + "learning_rate": 4.994939335223845e-06, + "loss": 2.6496, + "step": 2348 + }, + { + "epoch": 0.24841370558375633, + "grad_norm": 7.59375, + "learning_rate": 4.994928438182196e-06, + "loss": 2.4976, + "step": 2349 + }, + { + "epoch": 0.24851945854483926, + "grad_norm": 7.75, + "learning_rate": 4.994917529432869e-06, + "loss": 2.6133, + "step": 2350 + }, + { + "epoch": 0.24862521150592218, + "grad_norm": 8.5625, + "learning_rate": 4.994906608975915e-06, + "loss": 2.6974, + "step": 2351 + }, + { + "epoch": 0.24873096446700507, + "grad_norm": 7.65625, + "learning_rate": 4.994895676811384e-06, + "loss": 2.6163, + "step": 2352 + }, + { + "epoch": 0.248836717428088, + "grad_norm": 8.0625, + "learning_rate": 4.994884732939328e-06, + "loss": 2.7199, + "step": 2353 + }, + { + "epoch": 0.2489424703891709, + "grad_norm": 7.5625, + "learning_rate": 4.994873777359798e-06, + "loss": 2.5761, + "step": 2354 + }, + { + "epoch": 0.2490482233502538, + "grad_norm": 7.5, + "learning_rate": 4.9948628100728455e-06, + "loss": 2.7264, + "step": 2355 + }, + { + "epoch": 0.2491539763113367, + "grad_norm": 8.125, + "learning_rate": 4.994851831078523e-06, + "loss": 2.8256, + "step": 2356 + }, + { + "epoch": 0.24925972927241963, + "grad_norm": 8.375, + "learning_rate": 4.994840840376881e-06, + "loss": 2.5146, + "step": 2357 + }, + { + "epoch": 0.24936548223350255, + "grad_norm": 7.5, + "learning_rate": 4.994829837967971e-06, + "loss": 2.4979, + "step": 2358 + }, + { + "epoch": 0.24947123519458544, + "grad_norm": 7.375, + "learning_rate": 4.994818823851846e-06, + "loss": 2.5774, + "step": 2359 + }, + { + "epoch": 0.24957698815566837, + "grad_norm": 8.25, + "learning_rate": 4.994807798028555e-06, + "loss": 2.6665, + "step": 2360 + }, + { + "epoch": 0.24968274111675126, + "grad_norm": 8.8125, + "learning_rate": 4.994796760498152e-06, + "loss": 2.7612, + "step": 2361 + }, + { + "epoch": 0.24978849407783418, + "grad_norm": 8.375, + "learning_rate": 4.994785711260688e-06, + "loss": 2.5293, + "step": 2362 + }, + { + "epoch": 0.24989424703891708, + "grad_norm": 7.8125, + "learning_rate": 4.994774650316216e-06, + "loss": 2.7399, + "step": 2363 + }, + { + "epoch": 0.25, + "grad_norm": 18.25, + "learning_rate": 4.9947635776647865e-06, + "loss": 2.8234, + "step": 2364 + }, + { + "epoch": 0.2501057529610829, + "grad_norm": 8.6875, + "learning_rate": 4.994752493306451e-06, + "loss": 2.6033, + "step": 2365 + }, + { + "epoch": 0.25021150592216584, + "grad_norm": 7.6875, + "learning_rate": 4.994741397241263e-06, + "loss": 2.6632, + "step": 2366 + }, + { + "epoch": 0.25031725888324874, + "grad_norm": 8.6875, + "learning_rate": 4.994730289469273e-06, + "loss": 2.6943, + "step": 2367 + }, + { + "epoch": 0.25042301184433163, + "grad_norm": 9.25, + "learning_rate": 4.994719169990535e-06, + "loss": 2.7312, + "step": 2368 + }, + { + "epoch": 0.2505287648054145, + "grad_norm": 7.53125, + "learning_rate": 4.994708038805101e-06, + "loss": 2.7025, + "step": 2369 + }, + { + "epoch": 0.2506345177664975, + "grad_norm": 7.59375, + "learning_rate": 4.994696895913021e-06, + "loss": 2.5826, + "step": 2370 + }, + { + "epoch": 0.25074027072758037, + "grad_norm": 8.1875, + "learning_rate": 4.994685741314349e-06, + "loss": 2.813, + "step": 2371 + }, + { + "epoch": 0.25084602368866327, + "grad_norm": 7.59375, + "learning_rate": 4.994674575009137e-06, + "loss": 2.5811, + "step": 2372 + }, + { + "epoch": 0.2509517766497462, + "grad_norm": 7.90625, + "learning_rate": 4.994663396997439e-06, + "loss": 2.9065, + "step": 2373 + }, + { + "epoch": 0.2510575296108291, + "grad_norm": 8.0, + "learning_rate": 4.994652207279305e-06, + "loss": 2.8518, + "step": 2374 + }, + { + "epoch": 0.251163282571912, + "grad_norm": 8.1875, + "learning_rate": 4.9946410058547875e-06, + "loss": 2.7473, + "step": 2375 + }, + { + "epoch": 0.2512690355329949, + "grad_norm": 8.3125, + "learning_rate": 4.994629792723941e-06, + "loss": 2.458, + "step": 2376 + }, + { + "epoch": 0.25137478849407785, + "grad_norm": 8.3125, + "learning_rate": 4.994618567886818e-06, + "loss": 2.9182, + "step": 2377 + }, + { + "epoch": 0.25148054145516074, + "grad_norm": 7.96875, + "learning_rate": 4.9946073313434685e-06, + "loss": 2.6328, + "step": 2378 + }, + { + "epoch": 0.25158629441624364, + "grad_norm": 9.0, + "learning_rate": 4.994596083093948e-06, + "loss": 2.8544, + "step": 2379 + }, + { + "epoch": 0.2516920473773266, + "grad_norm": 8.375, + "learning_rate": 4.994584823138309e-06, + "loss": 2.4829, + "step": 2380 + }, + { + "epoch": 0.2517978003384095, + "grad_norm": 7.5, + "learning_rate": 4.9945735514766024e-06, + "loss": 2.4154, + "step": 2381 + }, + { + "epoch": 0.2519035532994924, + "grad_norm": 7.34375, + "learning_rate": 4.994562268108883e-06, + "loss": 2.5404, + "step": 2382 + }, + { + "epoch": 0.25200930626057527, + "grad_norm": 7.375, + "learning_rate": 4.994550973035203e-06, + "loss": 2.385, + "step": 2383 + }, + { + "epoch": 0.2521150592216582, + "grad_norm": 7.6875, + "learning_rate": 4.994539666255616e-06, + "loss": 2.8155, + "step": 2384 + }, + { + "epoch": 0.2522208121827411, + "grad_norm": 7.75, + "learning_rate": 4.994528347770173e-06, + "loss": 2.3406, + "step": 2385 + }, + { + "epoch": 0.252326565143824, + "grad_norm": 7.4375, + "learning_rate": 4.9945170175789305e-06, + "loss": 2.7023, + "step": 2386 + }, + { + "epoch": 0.25243231810490696, + "grad_norm": 7.65625, + "learning_rate": 4.994505675681939e-06, + "loss": 2.569, + "step": 2387 + }, + { + "epoch": 0.25253807106598986, + "grad_norm": 9.125, + "learning_rate": 4.994494322079253e-06, + "loss": 2.9844, + "step": 2388 + }, + { + "epoch": 0.25264382402707275, + "grad_norm": 7.9375, + "learning_rate": 4.994482956770925e-06, + "loss": 2.6915, + "step": 2389 + }, + { + "epoch": 0.25274957698815564, + "grad_norm": 7.6875, + "learning_rate": 4.994471579757008e-06, + "loss": 2.7064, + "step": 2390 + }, + { + "epoch": 0.2528553299492386, + "grad_norm": 7.1875, + "learning_rate": 4.994460191037558e-06, + "loss": 2.3953, + "step": 2391 + }, + { + "epoch": 0.2529610829103215, + "grad_norm": 7.90625, + "learning_rate": 4.994448790612625e-06, + "loss": 2.6615, + "step": 2392 + }, + { + "epoch": 0.2530668358714044, + "grad_norm": 7.625, + "learning_rate": 4.994437378482264e-06, + "loss": 2.5987, + "step": 2393 + }, + { + "epoch": 0.25317258883248733, + "grad_norm": 8.0, + "learning_rate": 4.994425954646529e-06, + "loss": 2.672, + "step": 2394 + }, + { + "epoch": 0.2532783417935702, + "grad_norm": 7.9375, + "learning_rate": 4.994414519105474e-06, + "loss": 2.8208, + "step": 2395 + }, + { + "epoch": 0.2533840947546531, + "grad_norm": 7.9375, + "learning_rate": 4.99440307185915e-06, + "loss": 2.7344, + "step": 2396 + }, + { + "epoch": 0.253489847715736, + "grad_norm": 8.0, + "learning_rate": 4.994391612907614e-06, + "loss": 2.4487, + "step": 2397 + }, + { + "epoch": 0.25359560067681897, + "grad_norm": 7.375, + "learning_rate": 4.9943801422509184e-06, + "loss": 2.557, + "step": 2398 + }, + { + "epoch": 0.25370135363790186, + "grad_norm": 7.8125, + "learning_rate": 4.994368659889116e-06, + "loss": 2.6856, + "step": 2399 + }, + { + "epoch": 0.25380710659898476, + "grad_norm": 8.25, + "learning_rate": 4.994357165822263e-06, + "loss": 2.5659, + "step": 2400 + }, + { + "epoch": 0.2539128595600677, + "grad_norm": 8.0, + "learning_rate": 4.99434566005041e-06, + "loss": 2.6398, + "step": 2401 + }, + { + "epoch": 0.2540186125211506, + "grad_norm": 8.375, + "learning_rate": 4.994334142573615e-06, + "loss": 2.588, + "step": 2402 + }, + { + "epoch": 0.2541243654822335, + "grad_norm": 8.4375, + "learning_rate": 4.9943226133919285e-06, + "loss": 2.7405, + "step": 2403 + }, + { + "epoch": 0.2542301184433164, + "grad_norm": 7.8125, + "learning_rate": 4.994311072505407e-06, + "loss": 2.4118, + "step": 2404 + }, + { + "epoch": 0.25433587140439934, + "grad_norm": 9.3125, + "learning_rate": 4.994299519914103e-06, + "loss": 2.6957, + "step": 2405 + }, + { + "epoch": 0.25444162436548223, + "grad_norm": 8.0, + "learning_rate": 4.9942879556180725e-06, + "loss": 2.7145, + "step": 2406 + }, + { + "epoch": 0.2545473773265651, + "grad_norm": 8.25, + "learning_rate": 4.994276379617367e-06, + "loss": 2.6988, + "step": 2407 + }, + { + "epoch": 0.2546531302876481, + "grad_norm": 8.4375, + "learning_rate": 4.994264791912045e-06, + "loss": 2.6554, + "step": 2408 + }, + { + "epoch": 0.25475888324873097, + "grad_norm": 8.4375, + "learning_rate": 4.994253192502157e-06, + "loss": 2.6713, + "step": 2409 + }, + { + "epoch": 0.25486463620981387, + "grad_norm": 8.25, + "learning_rate": 4.9942415813877585e-06, + "loss": 2.6018, + "step": 2410 + }, + { + "epoch": 0.25497038917089676, + "grad_norm": 7.96875, + "learning_rate": 4.9942299585689055e-06, + "loss": 2.5419, + "step": 2411 + }, + { + "epoch": 0.2550761421319797, + "grad_norm": 8.75, + "learning_rate": 4.994218324045651e-06, + "loss": 2.3494, + "step": 2412 + }, + { + "epoch": 0.2551818950930626, + "grad_norm": 8.125, + "learning_rate": 4.99420667781805e-06, + "loss": 2.7155, + "step": 2413 + }, + { + "epoch": 0.2552876480541455, + "grad_norm": 7.75, + "learning_rate": 4.994195019886157e-06, + "loss": 2.6121, + "step": 2414 + }, + { + "epoch": 0.25539340101522845, + "grad_norm": 8.5625, + "learning_rate": 4.994183350250027e-06, + "loss": 2.6936, + "step": 2415 + }, + { + "epoch": 0.25549915397631134, + "grad_norm": 7.125, + "learning_rate": 4.994171668909714e-06, + "loss": 2.5394, + "step": 2416 + }, + { + "epoch": 0.25560490693739424, + "grad_norm": 8.3125, + "learning_rate": 4.994159975865274e-06, + "loss": 2.8451, + "step": 2417 + }, + { + "epoch": 0.25571065989847713, + "grad_norm": 7.84375, + "learning_rate": 4.9941482711167614e-06, + "loss": 2.6876, + "step": 2418 + }, + { + "epoch": 0.2558164128595601, + "grad_norm": 8.0, + "learning_rate": 4.994136554664231e-06, + "loss": 2.6599, + "step": 2419 + }, + { + "epoch": 0.255922165820643, + "grad_norm": 7.59375, + "learning_rate": 4.994124826507738e-06, + "loss": 2.6833, + "step": 2420 + }, + { + "epoch": 0.25602791878172587, + "grad_norm": 8.3125, + "learning_rate": 4.994113086647337e-06, + "loss": 2.8262, + "step": 2421 + }, + { + "epoch": 0.2561336717428088, + "grad_norm": 7.59375, + "learning_rate": 4.994101335083083e-06, + "loss": 2.7979, + "step": 2422 + }, + { + "epoch": 0.2562394247038917, + "grad_norm": 7.90625, + "learning_rate": 4.994089571815032e-06, + "loss": 2.5458, + "step": 2423 + }, + { + "epoch": 0.2563451776649746, + "grad_norm": 8.125, + "learning_rate": 4.994077796843239e-06, + "loss": 2.7301, + "step": 2424 + }, + { + "epoch": 0.2564509306260575, + "grad_norm": 7.90625, + "learning_rate": 4.994066010167758e-06, + "loss": 2.666, + "step": 2425 + }, + { + "epoch": 0.25655668358714045, + "grad_norm": 8.25, + "learning_rate": 4.994054211788646e-06, + "loss": 2.7838, + "step": 2426 + }, + { + "epoch": 0.25666243654822335, + "grad_norm": 7.8125, + "learning_rate": 4.994042401705958e-06, + "loss": 2.8112, + "step": 2427 + }, + { + "epoch": 0.25676818950930624, + "grad_norm": 7.53125, + "learning_rate": 4.994030579919747e-06, + "loss": 2.5911, + "step": 2428 + }, + { + "epoch": 0.2568739424703892, + "grad_norm": 8.0, + "learning_rate": 4.9940187464300725e-06, + "loss": 2.6882, + "step": 2429 + }, + { + "epoch": 0.2569796954314721, + "grad_norm": 7.875, + "learning_rate": 4.9940069012369874e-06, + "loss": 2.7162, + "step": 2430 + }, + { + "epoch": 0.257085448392555, + "grad_norm": 7.65625, + "learning_rate": 4.993995044340548e-06, + "loss": 2.6134, + "step": 2431 + }, + { + "epoch": 0.2571912013536379, + "grad_norm": 7.625, + "learning_rate": 4.9939831757408095e-06, + "loss": 2.6914, + "step": 2432 + }, + { + "epoch": 0.2572969543147208, + "grad_norm": 7.625, + "learning_rate": 4.9939712954378285e-06, + "loss": 2.6254, + "step": 2433 + }, + { + "epoch": 0.2574027072758037, + "grad_norm": 7.6875, + "learning_rate": 4.99395940343166e-06, + "loss": 2.6736, + "step": 2434 + }, + { + "epoch": 0.2575084602368866, + "grad_norm": 8.375, + "learning_rate": 4.99394749972236e-06, + "loss": 2.9461, + "step": 2435 + }, + { + "epoch": 0.25761421319796957, + "grad_norm": 8.0, + "learning_rate": 4.9939355843099845e-06, + "loss": 2.6327, + "step": 2436 + }, + { + "epoch": 0.25771996615905246, + "grad_norm": 14.125, + "learning_rate": 4.993923657194589e-06, + "loss": 2.9098, + "step": 2437 + }, + { + "epoch": 0.25782571912013535, + "grad_norm": 9.6875, + "learning_rate": 4.99391171837623e-06, + "loss": 2.6493, + "step": 2438 + }, + { + "epoch": 0.25793147208121825, + "grad_norm": 8.125, + "learning_rate": 4.993899767854963e-06, + "loss": 2.5464, + "step": 2439 + }, + { + "epoch": 0.2580372250423012, + "grad_norm": 8.0625, + "learning_rate": 4.993887805630845e-06, + "loss": 2.7896, + "step": 2440 + }, + { + "epoch": 0.2581429780033841, + "grad_norm": 8.875, + "learning_rate": 4.993875831703931e-06, + "loss": 2.5859, + "step": 2441 + }, + { + "epoch": 0.258248730964467, + "grad_norm": 7.9375, + "learning_rate": 4.993863846074277e-06, + "loss": 2.6562, + "step": 2442 + }, + { + "epoch": 0.25835448392554994, + "grad_norm": 7.5, + "learning_rate": 4.993851848741941e-06, + "loss": 2.5169, + "step": 2443 + }, + { + "epoch": 0.25846023688663283, + "grad_norm": 7.59375, + "learning_rate": 4.9938398397069775e-06, + "loss": 2.4453, + "step": 2444 + }, + { + "epoch": 0.2585659898477157, + "grad_norm": 8.0625, + "learning_rate": 4.993827818969445e-06, + "loss": 2.3491, + "step": 2445 + }, + { + "epoch": 0.2586717428087986, + "grad_norm": 7.96875, + "learning_rate": 4.993815786529397e-06, + "loss": 2.8163, + "step": 2446 + }, + { + "epoch": 0.25877749576988157, + "grad_norm": 8.5, + "learning_rate": 4.993803742386892e-06, + "loss": 2.9246, + "step": 2447 + }, + { + "epoch": 0.25888324873096447, + "grad_norm": 8.375, + "learning_rate": 4.993791686541986e-06, + "loss": 2.813, + "step": 2448 + }, + { + "epoch": 0.25898900169204736, + "grad_norm": 8.25, + "learning_rate": 4.993779618994735e-06, + "loss": 2.9667, + "step": 2449 + }, + { + "epoch": 0.2590947546531303, + "grad_norm": 7.6875, + "learning_rate": 4.993767539745197e-06, + "loss": 2.5755, + "step": 2450 + }, + { + "epoch": 0.2592005076142132, + "grad_norm": 8.1875, + "learning_rate": 4.993755448793427e-06, + "loss": 2.423, + "step": 2451 + }, + { + "epoch": 0.2593062605752961, + "grad_norm": 7.625, + "learning_rate": 4.993743346139483e-06, + "loss": 2.7165, + "step": 2452 + }, + { + "epoch": 0.259412013536379, + "grad_norm": 7.78125, + "learning_rate": 4.993731231783421e-06, + "loss": 2.5798, + "step": 2453 + }, + { + "epoch": 0.25951776649746194, + "grad_norm": 8.3125, + "learning_rate": 4.9937191057252985e-06, + "loss": 2.804, + "step": 2454 + }, + { + "epoch": 0.25962351945854484, + "grad_norm": 9.125, + "learning_rate": 4.9937069679651726e-06, + "loss": 2.4904, + "step": 2455 + }, + { + "epoch": 0.25972927241962773, + "grad_norm": 7.21875, + "learning_rate": 4.9936948185031e-06, + "loss": 2.4786, + "step": 2456 + }, + { + "epoch": 0.2598350253807107, + "grad_norm": 7.1875, + "learning_rate": 4.993682657339136e-06, + "loss": 2.6142, + "step": 2457 + }, + { + "epoch": 0.2599407783417936, + "grad_norm": 8.0, + "learning_rate": 4.993670484473341e-06, + "loss": 2.6681, + "step": 2458 + }, + { + "epoch": 0.26004653130287647, + "grad_norm": 8.1875, + "learning_rate": 4.993658299905769e-06, + "loss": 2.9148, + "step": 2459 + }, + { + "epoch": 0.26015228426395937, + "grad_norm": 8.625, + "learning_rate": 4.993646103636479e-06, + "loss": 2.3704, + "step": 2460 + }, + { + "epoch": 0.2602580372250423, + "grad_norm": 8.0625, + "learning_rate": 4.993633895665527e-06, + "loss": 2.8996, + "step": 2461 + }, + { + "epoch": 0.2603637901861252, + "grad_norm": 7.875, + "learning_rate": 4.993621675992972e-06, + "loss": 2.6994, + "step": 2462 + }, + { + "epoch": 0.2604695431472081, + "grad_norm": 7.9375, + "learning_rate": 4.99360944461887e-06, + "loss": 2.7288, + "step": 2463 + }, + { + "epoch": 0.26057529610829105, + "grad_norm": 7.5625, + "learning_rate": 4.993597201543279e-06, + "loss": 2.7367, + "step": 2464 + }, + { + "epoch": 0.26068104906937395, + "grad_norm": 7.875, + "learning_rate": 4.993584946766256e-06, + "loss": 2.4873, + "step": 2465 + }, + { + "epoch": 0.26078680203045684, + "grad_norm": 8.0, + "learning_rate": 4.993572680287858e-06, + "loss": 2.8621, + "step": 2466 + }, + { + "epoch": 0.26089255499153974, + "grad_norm": 7.84375, + "learning_rate": 4.993560402108144e-06, + "loss": 2.6941, + "step": 2467 + }, + { + "epoch": 0.2609983079526227, + "grad_norm": 8.0, + "learning_rate": 4.99354811222717e-06, + "loss": 2.6531, + "step": 2468 + }, + { + "epoch": 0.2611040609137056, + "grad_norm": 7.65625, + "learning_rate": 4.993535810644995e-06, + "loss": 2.5585, + "step": 2469 + }, + { + "epoch": 0.2612098138747885, + "grad_norm": 7.5625, + "learning_rate": 4.993523497361677e-06, + "loss": 2.5933, + "step": 2470 + }, + { + "epoch": 0.2613155668358714, + "grad_norm": 8.4375, + "learning_rate": 4.993511172377272e-06, + "loss": 2.3466, + "step": 2471 + }, + { + "epoch": 0.2614213197969543, + "grad_norm": 8.625, + "learning_rate": 4.99349883569184e-06, + "loss": 2.8974, + "step": 2472 + }, + { + "epoch": 0.2615270727580372, + "grad_norm": 8.375, + "learning_rate": 4.993486487305437e-06, + "loss": 2.642, + "step": 2473 + }, + { + "epoch": 0.2616328257191201, + "grad_norm": 8.25, + "learning_rate": 4.9934741272181216e-06, + "loss": 2.4507, + "step": 2474 + }, + { + "epoch": 0.26173857868020306, + "grad_norm": 8.4375, + "learning_rate": 4.993461755429952e-06, + "loss": 2.7344, + "step": 2475 + }, + { + "epoch": 0.26184433164128595, + "grad_norm": 7.78125, + "learning_rate": 4.993449371940987e-06, + "loss": 2.7194, + "step": 2476 + }, + { + "epoch": 0.26195008460236885, + "grad_norm": 8.25, + "learning_rate": 4.9934369767512824e-06, + "loss": 2.6, + "step": 2477 + }, + { + "epoch": 0.2620558375634518, + "grad_norm": 7.5625, + "learning_rate": 4.993424569860899e-06, + "loss": 2.5914, + "step": 2478 + }, + { + "epoch": 0.2621615905245347, + "grad_norm": 7.15625, + "learning_rate": 4.9934121512698935e-06, + "loss": 2.4914, + "step": 2479 + }, + { + "epoch": 0.2622673434856176, + "grad_norm": 8.4375, + "learning_rate": 4.993399720978325e-06, + "loss": 2.7138, + "step": 2480 + }, + { + "epoch": 0.2623730964467005, + "grad_norm": 9.625, + "learning_rate": 4.9933872789862506e-06, + "loss": 2.4358, + "step": 2481 + }, + { + "epoch": 0.26247884940778343, + "grad_norm": 8.625, + "learning_rate": 4.99337482529373e-06, + "loss": 2.6089, + "step": 2482 + }, + { + "epoch": 0.2625846023688663, + "grad_norm": 7.21875, + "learning_rate": 4.993362359900821e-06, + "loss": 2.4458, + "step": 2483 + }, + { + "epoch": 0.2626903553299492, + "grad_norm": 8.1875, + "learning_rate": 4.993349882807582e-06, + "loss": 2.2095, + "step": 2484 + }, + { + "epoch": 0.26279610829103217, + "grad_norm": 8.8125, + "learning_rate": 4.993337394014072e-06, + "loss": 2.7794, + "step": 2485 + }, + { + "epoch": 0.26290186125211507, + "grad_norm": 8.375, + "learning_rate": 4.993324893520349e-06, + "loss": 2.7633, + "step": 2486 + }, + { + "epoch": 0.26300761421319796, + "grad_norm": 8.25, + "learning_rate": 4.993312381326472e-06, + "loss": 2.8204, + "step": 2487 + }, + { + "epoch": 0.26311336717428085, + "grad_norm": 7.84375, + "learning_rate": 4.9932998574325e-06, + "loss": 2.5616, + "step": 2488 + }, + { + "epoch": 0.2632191201353638, + "grad_norm": 7.3125, + "learning_rate": 4.993287321838492e-06, + "loss": 2.5122, + "step": 2489 + }, + { + "epoch": 0.2633248730964467, + "grad_norm": 7.78125, + "learning_rate": 4.993274774544504e-06, + "loss": 2.6568, + "step": 2490 + }, + { + "epoch": 0.2634306260575296, + "grad_norm": 7.53125, + "learning_rate": 4.9932622155506e-06, + "loss": 2.6886, + "step": 2491 + }, + { + "epoch": 0.26353637901861254, + "grad_norm": 7.90625, + "learning_rate": 4.993249644856835e-06, + "loss": 2.4555, + "step": 2492 + }, + { + "epoch": 0.26364213197969544, + "grad_norm": 7.65625, + "learning_rate": 4.993237062463269e-06, + "loss": 2.6433, + "step": 2493 + }, + { + "epoch": 0.26374788494077833, + "grad_norm": 7.84375, + "learning_rate": 4.993224468369961e-06, + "loss": 2.6636, + "step": 2494 + }, + { + "epoch": 0.2638536379018612, + "grad_norm": 7.90625, + "learning_rate": 4.993211862576971e-06, + "loss": 2.562, + "step": 2495 + }, + { + "epoch": 0.2639593908629442, + "grad_norm": 7.1875, + "learning_rate": 4.993199245084357e-06, + "loss": 2.5632, + "step": 2496 + }, + { + "epoch": 0.26406514382402707, + "grad_norm": 8.25, + "learning_rate": 4.993186615892179e-06, + "loss": 2.862, + "step": 2497 + }, + { + "epoch": 0.26417089678510997, + "grad_norm": 8.5, + "learning_rate": 4.9931739750004956e-06, + "loss": 2.6875, + "step": 2498 + }, + { + "epoch": 0.2642766497461929, + "grad_norm": 9.75, + "learning_rate": 4.993161322409366e-06, + "loss": 2.5328, + "step": 2499 + }, + { + "epoch": 0.2643824027072758, + "grad_norm": 8.375, + "learning_rate": 4.99314865811885e-06, + "loss": 2.361, + "step": 2500 + }, + { + "epoch": 0.2644881556683587, + "grad_norm": 7.6875, + "learning_rate": 4.993135982129007e-06, + "loss": 2.5712, + "step": 2501 + }, + { + "epoch": 0.2645939086294416, + "grad_norm": 8.0625, + "learning_rate": 4.993123294439898e-06, + "loss": 2.7706, + "step": 2502 + }, + { + "epoch": 0.26469966159052455, + "grad_norm": 7.6875, + "learning_rate": 4.993110595051579e-06, + "loss": 2.5486, + "step": 2503 + }, + { + "epoch": 0.26480541455160744, + "grad_norm": 7.71875, + "learning_rate": 4.993097883964113e-06, + "loss": 2.5295, + "step": 2504 + }, + { + "epoch": 0.26491116751269034, + "grad_norm": 7.8125, + "learning_rate": 4.993085161177558e-06, + "loss": 2.636, + "step": 2505 + }, + { + "epoch": 0.2650169204737733, + "grad_norm": 8.5, + "learning_rate": 4.993072426691974e-06, + "loss": 2.6994, + "step": 2506 + }, + { + "epoch": 0.2651226734348562, + "grad_norm": 8.0, + "learning_rate": 4.99305968050742e-06, + "loss": 2.5198, + "step": 2507 + }, + { + "epoch": 0.2652284263959391, + "grad_norm": 7.59375, + "learning_rate": 4.993046922623957e-06, + "loss": 2.4879, + "step": 2508 + }, + { + "epoch": 0.26533417935702197, + "grad_norm": 7.4375, + "learning_rate": 4.993034153041644e-06, + "loss": 2.548, + "step": 2509 + }, + { + "epoch": 0.2654399323181049, + "grad_norm": 8.0, + "learning_rate": 4.993021371760542e-06, + "loss": 2.592, + "step": 2510 + }, + { + "epoch": 0.2655456852791878, + "grad_norm": 8.625, + "learning_rate": 4.99300857878071e-06, + "loss": 2.3118, + "step": 2511 + }, + { + "epoch": 0.2656514382402707, + "grad_norm": 9.1875, + "learning_rate": 4.9929957741022085e-06, + "loss": 2.6705, + "step": 2512 + }, + { + "epoch": 0.26575719120135366, + "grad_norm": 7.53125, + "learning_rate": 4.992982957725098e-06, + "loss": 2.526, + "step": 2513 + }, + { + "epoch": 0.26586294416243655, + "grad_norm": 7.9375, + "learning_rate": 4.992970129649437e-06, + "loss": 2.8333, + "step": 2514 + }, + { + "epoch": 0.26596869712351945, + "grad_norm": 7.6875, + "learning_rate": 4.992957289875287e-06, + "loss": 2.7098, + "step": 2515 + }, + { + "epoch": 0.26607445008460234, + "grad_norm": 7.34375, + "learning_rate": 4.992944438402708e-06, + "loss": 2.695, + "step": 2516 + }, + { + "epoch": 0.2661802030456853, + "grad_norm": 7.53125, + "learning_rate": 4.992931575231761e-06, + "loss": 2.559, + "step": 2517 + }, + { + "epoch": 0.2662859560067682, + "grad_norm": 7.5, + "learning_rate": 4.992918700362505e-06, + "loss": 2.5723, + "step": 2518 + }, + { + "epoch": 0.2663917089678511, + "grad_norm": 8.125, + "learning_rate": 4.992905813795001e-06, + "loss": 2.6063, + "step": 2519 + }, + { + "epoch": 0.26649746192893403, + "grad_norm": 7.5625, + "learning_rate": 4.992892915529311e-06, + "loss": 2.45, + "step": 2520 + }, + { + "epoch": 0.2666032148900169, + "grad_norm": 7.84375, + "learning_rate": 4.992880005565492e-06, + "loss": 2.7918, + "step": 2521 + }, + { + "epoch": 0.2667089678510998, + "grad_norm": 7.65625, + "learning_rate": 4.992867083903609e-06, + "loss": 2.5951, + "step": 2522 + }, + { + "epoch": 0.2668147208121827, + "grad_norm": 7.78125, + "learning_rate": 4.992854150543718e-06, + "loss": 2.6371, + "step": 2523 + }, + { + "epoch": 0.26692047377326567, + "grad_norm": 7.9375, + "learning_rate": 4.992841205485884e-06, + "loss": 2.7071, + "step": 2524 + }, + { + "epoch": 0.26702622673434856, + "grad_norm": 8.875, + "learning_rate": 4.992828248730164e-06, + "loss": 2.9575, + "step": 2525 + }, + { + "epoch": 0.26713197969543145, + "grad_norm": 8.5, + "learning_rate": 4.992815280276622e-06, + "loss": 2.6167, + "step": 2526 + }, + { + "epoch": 0.2672377326565144, + "grad_norm": 8.25, + "learning_rate": 4.992802300125317e-06, + "loss": 2.7709, + "step": 2527 + }, + { + "epoch": 0.2673434856175973, + "grad_norm": 7.34375, + "learning_rate": 4.992789308276309e-06, + "loss": 2.512, + "step": 2528 + }, + { + "epoch": 0.2674492385786802, + "grad_norm": 8.8125, + "learning_rate": 4.992776304729662e-06, + "loss": 2.4377, + "step": 2529 + }, + { + "epoch": 0.2675549915397631, + "grad_norm": 8.3125, + "learning_rate": 4.992763289485435e-06, + "loss": 2.7528, + "step": 2530 + }, + { + "epoch": 0.26766074450084604, + "grad_norm": 7.53125, + "learning_rate": 4.992750262543689e-06, + "loss": 2.5512, + "step": 2531 + }, + { + "epoch": 0.26776649746192893, + "grad_norm": 8.875, + "learning_rate": 4.992737223904486e-06, + "loss": 2.6519, + "step": 2532 + }, + { + "epoch": 0.2678722504230118, + "grad_norm": 8.125, + "learning_rate": 4.9927241735678864e-06, + "loss": 2.3626, + "step": 2533 + }, + { + "epoch": 0.2679780033840948, + "grad_norm": 7.75, + "learning_rate": 4.992711111533951e-06, + "loss": 2.5034, + "step": 2534 + }, + { + "epoch": 0.26808375634517767, + "grad_norm": 8.0, + "learning_rate": 4.9926980378027425e-06, + "loss": 2.4478, + "step": 2535 + }, + { + "epoch": 0.26818950930626057, + "grad_norm": 7.53125, + "learning_rate": 4.992684952374322e-06, + "loss": 2.7031, + "step": 2536 + }, + { + "epoch": 0.26829526226734346, + "grad_norm": 8.0, + "learning_rate": 4.99267185524875e-06, + "loss": 2.5885, + "step": 2537 + }, + { + "epoch": 0.2684010152284264, + "grad_norm": 7.53125, + "learning_rate": 4.992658746426089e-06, + "loss": 2.7062, + "step": 2538 + }, + { + "epoch": 0.2685067681895093, + "grad_norm": 8.0625, + "learning_rate": 4.992645625906399e-06, + "loss": 2.6713, + "step": 2539 + }, + { + "epoch": 0.2686125211505922, + "grad_norm": 8.25, + "learning_rate": 4.992632493689743e-06, + "loss": 2.7777, + "step": 2540 + }, + { + "epoch": 0.26871827411167515, + "grad_norm": 7.625, + "learning_rate": 4.992619349776182e-06, + "loss": 2.5716, + "step": 2541 + }, + { + "epoch": 0.26882402707275804, + "grad_norm": 8.1875, + "learning_rate": 4.992606194165779e-06, + "loss": 2.8643, + "step": 2542 + }, + { + "epoch": 0.26892978003384094, + "grad_norm": 13.5, + "learning_rate": 4.9925930268585934e-06, + "loss": 2.7307, + "step": 2543 + }, + { + "epoch": 0.26903553299492383, + "grad_norm": 9.0625, + "learning_rate": 4.992579847854688e-06, + "loss": 2.4279, + "step": 2544 + }, + { + "epoch": 0.2691412859560068, + "grad_norm": 7.78125, + "learning_rate": 4.992566657154125e-06, + "loss": 2.6498, + "step": 2545 + }, + { + "epoch": 0.2692470389170897, + "grad_norm": 7.875, + "learning_rate": 4.9925534547569664e-06, + "loss": 2.7319, + "step": 2546 + }, + { + "epoch": 0.26935279187817257, + "grad_norm": 7.9375, + "learning_rate": 4.992540240663274e-06, + "loss": 2.7311, + "step": 2547 + }, + { + "epoch": 0.2694585448392555, + "grad_norm": 7.5625, + "learning_rate": 4.99252701487311e-06, + "loss": 2.5451, + "step": 2548 + }, + { + "epoch": 0.2695642978003384, + "grad_norm": 7.59375, + "learning_rate": 4.992513777386535e-06, + "loss": 2.25, + "step": 2549 + }, + { + "epoch": 0.2696700507614213, + "grad_norm": 7.65625, + "learning_rate": 4.992500528203613e-06, + "loss": 2.6632, + "step": 2550 + }, + { + "epoch": 0.2697758037225042, + "grad_norm": 7.65625, + "learning_rate": 4.992487267324405e-06, + "loss": 2.5395, + "step": 2551 + }, + { + "epoch": 0.26988155668358715, + "grad_norm": 8.5625, + "learning_rate": 4.992473994748974e-06, + "loss": 2.5606, + "step": 2552 + }, + { + "epoch": 0.26998730964467005, + "grad_norm": 8.3125, + "learning_rate": 4.992460710477381e-06, + "loss": 2.5565, + "step": 2553 + }, + { + "epoch": 0.27009306260575294, + "grad_norm": 8.75, + "learning_rate": 4.99244741450969e-06, + "loss": 2.6096, + "step": 2554 + }, + { + "epoch": 0.2701988155668359, + "grad_norm": 7.84375, + "learning_rate": 4.992434106845964e-06, + "loss": 2.6879, + "step": 2555 + }, + { + "epoch": 0.2703045685279188, + "grad_norm": 7.5625, + "learning_rate": 4.992420787486262e-06, + "loss": 2.5584, + "step": 2556 + }, + { + "epoch": 0.2704103214890017, + "grad_norm": 7.78125, + "learning_rate": 4.99240745643065e-06, + "loss": 2.5683, + "step": 2557 + }, + { + "epoch": 0.2705160744500846, + "grad_norm": 8.25, + "learning_rate": 4.992394113679188e-06, + "loss": 2.7551, + "step": 2558 + }, + { + "epoch": 0.2706218274111675, + "grad_norm": 7.4375, + "learning_rate": 4.9923807592319405e-06, + "loss": 2.6067, + "step": 2559 + }, + { + "epoch": 0.2707275803722504, + "grad_norm": 8.125, + "learning_rate": 4.992367393088969e-06, + "loss": 2.5769, + "step": 2560 + }, + { + "epoch": 0.2708333333333333, + "grad_norm": 8.4375, + "learning_rate": 4.992354015250338e-06, + "loss": 2.6958, + "step": 2561 + }, + { + "epoch": 0.27093908629441626, + "grad_norm": 8.25, + "learning_rate": 4.992340625716107e-06, + "loss": 2.8936, + "step": 2562 + }, + { + "epoch": 0.27104483925549916, + "grad_norm": 7.875, + "learning_rate": 4.992327224486342e-06, + "loss": 2.6909, + "step": 2563 + }, + { + "epoch": 0.27115059221658205, + "grad_norm": 7.90625, + "learning_rate": 4.992313811561105e-06, + "loss": 2.6428, + "step": 2564 + }, + { + "epoch": 0.27125634517766495, + "grad_norm": 9.9375, + "learning_rate": 4.992300386940457e-06, + "loss": 2.6333, + "step": 2565 + }, + { + "epoch": 0.2713620981387479, + "grad_norm": 7.71875, + "learning_rate": 4.992286950624464e-06, + "loss": 2.8908, + "step": 2566 + }, + { + "epoch": 0.2714678510998308, + "grad_norm": 7.46875, + "learning_rate": 4.992273502613188e-06, + "loss": 2.7571, + "step": 2567 + }, + { + "epoch": 0.2715736040609137, + "grad_norm": 8.25, + "learning_rate": 4.9922600429066905e-06, + "loss": 2.846, + "step": 2568 + }, + { + "epoch": 0.27167935702199664, + "grad_norm": 7.59375, + "learning_rate": 4.9922465715050365e-06, + "loss": 2.6383, + "step": 2569 + }, + { + "epoch": 0.27178510998307953, + "grad_norm": 7.90625, + "learning_rate": 4.9922330884082885e-06, + "loss": 2.6951, + "step": 2570 + }, + { + "epoch": 0.2718908629441624, + "grad_norm": 7.875, + "learning_rate": 4.99221959361651e-06, + "loss": 2.6763, + "step": 2571 + }, + { + "epoch": 0.2719966159052453, + "grad_norm": 8.3125, + "learning_rate": 4.992206087129764e-06, + "loss": 2.248, + "step": 2572 + }, + { + "epoch": 0.27210236886632827, + "grad_norm": 7.875, + "learning_rate": 4.992192568948114e-06, + "loss": 2.5581, + "step": 2573 + }, + { + "epoch": 0.27220812182741116, + "grad_norm": 7.625, + "learning_rate": 4.992179039071624e-06, + "loss": 2.5666, + "step": 2574 + }, + { + "epoch": 0.27231387478849406, + "grad_norm": 8.625, + "learning_rate": 4.992165497500357e-06, + "loss": 2.6503, + "step": 2575 + }, + { + "epoch": 0.272419627749577, + "grad_norm": 8.3125, + "learning_rate": 4.992151944234376e-06, + "loss": 2.6704, + "step": 2576 + }, + { + "epoch": 0.2725253807106599, + "grad_norm": 8.125, + "learning_rate": 4.992138379273745e-06, + "loss": 2.8204, + "step": 2577 + }, + { + "epoch": 0.2726311336717428, + "grad_norm": 8.0625, + "learning_rate": 4.992124802618529e-06, + "loss": 2.6327, + "step": 2578 + }, + { + "epoch": 0.2727368866328257, + "grad_norm": 8.1875, + "learning_rate": 4.99211121426879e-06, + "loss": 2.7478, + "step": 2579 + }, + { + "epoch": 0.27284263959390864, + "grad_norm": 7.40625, + "learning_rate": 4.9920976142245925e-06, + "loss": 2.5718, + "step": 2580 + }, + { + "epoch": 0.27294839255499154, + "grad_norm": 7.375, + "learning_rate": 4.992084002486e-06, + "loss": 2.4724, + "step": 2581 + }, + { + "epoch": 0.27305414551607443, + "grad_norm": 8.4375, + "learning_rate": 4.992070379053077e-06, + "loss": 2.7598, + "step": 2582 + }, + { + "epoch": 0.2731598984771574, + "grad_norm": 8.6875, + "learning_rate": 4.992056743925886e-06, + "loss": 2.7619, + "step": 2583 + }, + { + "epoch": 0.2732656514382403, + "grad_norm": 7.53125, + "learning_rate": 4.992043097104492e-06, + "loss": 2.7624, + "step": 2584 + }, + { + "epoch": 0.27337140439932317, + "grad_norm": 7.53125, + "learning_rate": 4.99202943858896e-06, + "loss": 2.7446, + "step": 2585 + }, + { + "epoch": 0.27347715736040606, + "grad_norm": 8.8125, + "learning_rate": 4.992015768379352e-06, + "loss": 2.9602, + "step": 2586 + }, + { + "epoch": 0.273582910321489, + "grad_norm": 8.375, + "learning_rate": 4.9920020864757335e-06, + "loss": 2.6724, + "step": 2587 + }, + { + "epoch": 0.2736886632825719, + "grad_norm": 7.96875, + "learning_rate": 4.991988392878168e-06, + "loss": 2.8562, + "step": 2588 + }, + { + "epoch": 0.2737944162436548, + "grad_norm": 8.75, + "learning_rate": 4.991974687586721e-06, + "loss": 2.8142, + "step": 2589 + }, + { + "epoch": 0.27390016920473775, + "grad_norm": 7.875, + "learning_rate": 4.991960970601455e-06, + "loss": 2.6228, + "step": 2590 + }, + { + "epoch": 0.27400592216582065, + "grad_norm": 12.5625, + "learning_rate": 4.991947241922436e-06, + "loss": 2.5024, + "step": 2591 + }, + { + "epoch": 0.27411167512690354, + "grad_norm": 7.8125, + "learning_rate": 4.991933501549728e-06, + "loss": 2.818, + "step": 2592 + }, + { + "epoch": 0.27421742808798644, + "grad_norm": 7.8125, + "learning_rate": 4.991919749483395e-06, + "loss": 2.68, + "step": 2593 + }, + { + "epoch": 0.2743231810490694, + "grad_norm": 8.25, + "learning_rate": 4.991905985723501e-06, + "loss": 2.8945, + "step": 2594 + }, + { + "epoch": 0.2744289340101523, + "grad_norm": 8.0, + "learning_rate": 4.991892210270113e-06, + "loss": 2.7717, + "step": 2595 + }, + { + "epoch": 0.2745346869712352, + "grad_norm": 7.59375, + "learning_rate": 4.991878423123292e-06, + "loss": 2.7653, + "step": 2596 + }, + { + "epoch": 0.2746404399323181, + "grad_norm": 7.78125, + "learning_rate": 4.9918646242831055e-06, + "loss": 2.6584, + "step": 2597 + }, + { + "epoch": 0.274746192893401, + "grad_norm": 7.59375, + "learning_rate": 4.991850813749619e-06, + "loss": 2.7102, + "step": 2598 + }, + { + "epoch": 0.2748519458544839, + "grad_norm": 7.71875, + "learning_rate": 4.991836991522894e-06, + "loss": 2.49, + "step": 2599 + }, + { + "epoch": 0.2749576988155668, + "grad_norm": 11.8125, + "learning_rate": 4.991823157602997e-06, + "loss": 2.8318, + "step": 2600 + }, + { + "epoch": 0.27506345177664976, + "grad_norm": 7.65625, + "learning_rate": 4.991809311989994e-06, + "loss": 2.5842, + "step": 2601 + }, + { + "epoch": 0.27516920473773265, + "grad_norm": 7.28125, + "learning_rate": 4.991795454683949e-06, + "loss": 2.6623, + "step": 2602 + }, + { + "epoch": 0.27527495769881555, + "grad_norm": 7.6875, + "learning_rate": 4.991781585684926e-06, + "loss": 2.5927, + "step": 2603 + }, + { + "epoch": 0.2753807106598985, + "grad_norm": 7.96875, + "learning_rate": 4.991767704992993e-06, + "loss": 2.8185, + "step": 2604 + }, + { + "epoch": 0.2754864636209814, + "grad_norm": 8.25, + "learning_rate": 4.991753812608211e-06, + "loss": 2.7605, + "step": 2605 + }, + { + "epoch": 0.2755922165820643, + "grad_norm": 8.1875, + "learning_rate": 4.991739908530649e-06, + "loss": 2.6164, + "step": 2606 + }, + { + "epoch": 0.2756979695431472, + "grad_norm": 7.4375, + "learning_rate": 4.99172599276037e-06, + "loss": 2.6578, + "step": 2607 + }, + { + "epoch": 0.27580372250423013, + "grad_norm": 8.3125, + "learning_rate": 4.991712065297441e-06, + "loss": 2.7885, + "step": 2608 + }, + { + "epoch": 0.275909475465313, + "grad_norm": 7.78125, + "learning_rate": 4.991698126141925e-06, + "loss": 2.6127, + "step": 2609 + }, + { + "epoch": 0.2760152284263959, + "grad_norm": 8.3125, + "learning_rate": 4.991684175293889e-06, + "loss": 2.7878, + "step": 2610 + }, + { + "epoch": 0.27612098138747887, + "grad_norm": 8.625, + "learning_rate": 4.991670212753399e-06, + "loss": 2.6879, + "step": 2611 + }, + { + "epoch": 0.27622673434856176, + "grad_norm": 9.5, + "learning_rate": 4.991656238520519e-06, + "loss": 2.5551, + "step": 2612 + }, + { + "epoch": 0.27633248730964466, + "grad_norm": 7.6875, + "learning_rate": 4.991642252595315e-06, + "loss": 2.4855, + "step": 2613 + }, + { + "epoch": 0.27643824027072755, + "grad_norm": 8.875, + "learning_rate": 4.991628254977855e-06, + "loss": 2.7258, + "step": 2614 + }, + { + "epoch": 0.2765439932318105, + "grad_norm": 7.5, + "learning_rate": 4.9916142456682e-06, + "loss": 2.5822, + "step": 2615 + }, + { + "epoch": 0.2766497461928934, + "grad_norm": 7.71875, + "learning_rate": 4.99160022466642e-06, + "loss": 2.5529, + "step": 2616 + }, + { + "epoch": 0.2767554991539763, + "grad_norm": 8.125, + "learning_rate": 4.991586191972579e-06, + "loss": 2.7046, + "step": 2617 + }, + { + "epoch": 0.27686125211505924, + "grad_norm": 8.375, + "learning_rate": 4.991572147586743e-06, + "loss": 2.7379, + "step": 2618 + }, + { + "epoch": 0.27696700507614214, + "grad_norm": 7.53125, + "learning_rate": 4.991558091508978e-06, + "loss": 2.6985, + "step": 2619 + }, + { + "epoch": 0.27707275803722503, + "grad_norm": 8.0625, + "learning_rate": 4.991544023739349e-06, + "loss": 2.4856, + "step": 2620 + }, + { + "epoch": 0.2771785109983079, + "grad_norm": 11.0, + "learning_rate": 4.991529944277924e-06, + "loss": 2.5114, + "step": 2621 + }, + { + "epoch": 0.2772842639593909, + "grad_norm": 7.625, + "learning_rate": 4.991515853124767e-06, + "loss": 2.5138, + "step": 2622 + }, + { + "epoch": 0.27739001692047377, + "grad_norm": 8.625, + "learning_rate": 4.9915017502799454e-06, + "loss": 2.6114, + "step": 2623 + }, + { + "epoch": 0.27749576988155666, + "grad_norm": 7.84375, + "learning_rate": 4.991487635743526e-06, + "loss": 2.6674, + "step": 2624 + }, + { + "epoch": 0.2776015228426396, + "grad_norm": 7.0625, + "learning_rate": 4.991473509515573e-06, + "loss": 2.5872, + "step": 2625 + }, + { + "epoch": 0.2777072758037225, + "grad_norm": 7.625, + "learning_rate": 4.991459371596154e-06, + "loss": 2.5418, + "step": 2626 + }, + { + "epoch": 0.2778130287648054, + "grad_norm": 8.25, + "learning_rate": 4.9914452219853354e-06, + "loss": 2.6154, + "step": 2627 + }, + { + "epoch": 0.2779187817258883, + "grad_norm": 7.65625, + "learning_rate": 4.9914310606831835e-06, + "loss": 2.7327, + "step": 2628 + }, + { + "epoch": 0.27802453468697125, + "grad_norm": 7.78125, + "learning_rate": 4.991416887689764e-06, + "loss": 2.6936, + "step": 2629 + }, + { + "epoch": 0.27813028764805414, + "grad_norm": 8.4375, + "learning_rate": 4.991402703005145e-06, + "loss": 2.7133, + "step": 2630 + }, + { + "epoch": 0.27823604060913704, + "grad_norm": 8.5625, + "learning_rate": 4.991388506629391e-06, + "loss": 2.8193, + "step": 2631 + }, + { + "epoch": 0.27834179357022, + "grad_norm": 7.84375, + "learning_rate": 4.99137429856257e-06, + "loss": 2.6243, + "step": 2632 + }, + { + "epoch": 0.2784475465313029, + "grad_norm": 7.5, + "learning_rate": 4.991360078804749e-06, + "loss": 2.8443, + "step": 2633 + }, + { + "epoch": 0.2785532994923858, + "grad_norm": 7.4375, + "learning_rate": 4.991345847355993e-06, + "loss": 2.7052, + "step": 2634 + }, + { + "epoch": 0.27865905245346867, + "grad_norm": 7.3125, + "learning_rate": 4.991331604216371e-06, + "loss": 2.5494, + "step": 2635 + }, + { + "epoch": 0.2787648054145516, + "grad_norm": 9.125, + "learning_rate": 4.991317349385948e-06, + "loss": 2.394, + "step": 2636 + }, + { + "epoch": 0.2788705583756345, + "grad_norm": 7.4375, + "learning_rate": 4.991303082864791e-06, + "loss": 2.5652, + "step": 2637 + }, + { + "epoch": 0.2789763113367174, + "grad_norm": 8.3125, + "learning_rate": 4.991288804652968e-06, + "loss": 2.5538, + "step": 2638 + }, + { + "epoch": 0.27908206429780036, + "grad_norm": 8.9375, + "learning_rate": 4.991274514750546e-06, + "loss": 2.7258, + "step": 2639 + }, + { + "epoch": 0.27918781725888325, + "grad_norm": 7.46875, + "learning_rate": 4.991260213157592e-06, + "loss": 2.4869, + "step": 2640 + }, + { + "epoch": 0.27929357021996615, + "grad_norm": 8.6875, + "learning_rate": 4.991245899874172e-06, + "loss": 2.4627, + "step": 2641 + }, + { + "epoch": 0.27939932318104904, + "grad_norm": 7.875, + "learning_rate": 4.991231574900354e-06, + "loss": 2.8955, + "step": 2642 + }, + { + "epoch": 0.279505076142132, + "grad_norm": 7.75, + "learning_rate": 4.991217238236204e-06, + "loss": 2.6508, + "step": 2643 + }, + { + "epoch": 0.2796108291032149, + "grad_norm": 7.625, + "learning_rate": 4.991202889881792e-06, + "loss": 2.7257, + "step": 2644 + }, + { + "epoch": 0.2797165820642978, + "grad_norm": 7.625, + "learning_rate": 4.991188529837183e-06, + "loss": 2.5611, + "step": 2645 + }, + { + "epoch": 0.27982233502538073, + "grad_norm": 7.84375, + "learning_rate": 4.9911741581024455e-06, + "loss": 2.7054, + "step": 2646 + }, + { + "epoch": 0.2799280879864636, + "grad_norm": 7.78125, + "learning_rate": 4.991159774677646e-06, + "loss": 2.5398, + "step": 2647 + }, + { + "epoch": 0.2800338409475465, + "grad_norm": 7.21875, + "learning_rate": 4.9911453795628525e-06, + "loss": 2.5701, + "step": 2648 + }, + { + "epoch": 0.2801395939086294, + "grad_norm": 7.53125, + "learning_rate": 4.991130972758134e-06, + "loss": 2.5589, + "step": 2649 + }, + { + "epoch": 0.28024534686971236, + "grad_norm": 7.625, + "learning_rate": 4.991116554263555e-06, + "loss": 2.4937, + "step": 2650 + }, + { + "epoch": 0.28035109983079526, + "grad_norm": 12.125, + "learning_rate": 4.991102124079186e-06, + "loss": 2.4377, + "step": 2651 + }, + { + "epoch": 0.28045685279187815, + "grad_norm": 7.53125, + "learning_rate": 4.9910876822050925e-06, + "loss": 2.541, + "step": 2652 + }, + { + "epoch": 0.2805626057529611, + "grad_norm": 9.5625, + "learning_rate": 4.991073228641344e-06, + "loss": 2.8482, + "step": 2653 + }, + { + "epoch": 0.280668358714044, + "grad_norm": 10.3125, + "learning_rate": 4.991058763388007e-06, + "loss": 2.5599, + "step": 2654 + }, + { + "epoch": 0.2807741116751269, + "grad_norm": 7.5625, + "learning_rate": 4.991044286445151e-06, + "loss": 2.6418, + "step": 2655 + }, + { + "epoch": 0.2808798646362098, + "grad_norm": 8.9375, + "learning_rate": 4.991029797812842e-06, + "loss": 2.7933, + "step": 2656 + }, + { + "epoch": 0.28098561759729274, + "grad_norm": 7.65625, + "learning_rate": 4.9910152974911496e-06, + "loss": 2.7134, + "step": 2657 + }, + { + "epoch": 0.28109137055837563, + "grad_norm": 10.5625, + "learning_rate": 4.9910007854801406e-06, + "loss": 2.6159, + "step": 2658 + }, + { + "epoch": 0.2811971235194585, + "grad_norm": 7.84375, + "learning_rate": 4.990986261779885e-06, + "loss": 2.6095, + "step": 2659 + }, + { + "epoch": 0.2813028764805415, + "grad_norm": 7.375, + "learning_rate": 4.990971726390448e-06, + "loss": 2.4881, + "step": 2660 + }, + { + "epoch": 0.28140862944162437, + "grad_norm": 8.0, + "learning_rate": 4.9909571793119e-06, + "loss": 2.6211, + "step": 2661 + }, + { + "epoch": 0.28151438240270726, + "grad_norm": 10.4375, + "learning_rate": 4.990942620544309e-06, + "loss": 2.7068, + "step": 2662 + }, + { + "epoch": 0.2816201353637902, + "grad_norm": 8.625, + "learning_rate": 4.990928050087743e-06, + "loss": 2.7068, + "step": 2663 + }, + { + "epoch": 0.2817258883248731, + "grad_norm": 12.0, + "learning_rate": 4.9909134679422706e-06, + "loss": 2.9258, + "step": 2664 + }, + { + "epoch": 0.281831641285956, + "grad_norm": 8.4375, + "learning_rate": 4.99089887410796e-06, + "loss": 2.6183, + "step": 2665 + }, + { + "epoch": 0.2819373942470389, + "grad_norm": 7.40625, + "learning_rate": 4.990884268584879e-06, + "loss": 2.6493, + "step": 2666 + }, + { + "epoch": 0.28204314720812185, + "grad_norm": 7.59375, + "learning_rate": 4.990869651373098e-06, + "loss": 2.567, + "step": 2667 + }, + { + "epoch": 0.28214890016920474, + "grad_norm": 9.3125, + "learning_rate": 4.990855022472684e-06, + "loss": 2.6607, + "step": 2668 + }, + { + "epoch": 0.28225465313028764, + "grad_norm": 7.09375, + "learning_rate": 4.990840381883706e-06, + "loss": 2.4727, + "step": 2669 + }, + { + "epoch": 0.2823604060913706, + "grad_norm": 8.8125, + "learning_rate": 4.990825729606233e-06, + "loss": 2.6676, + "step": 2670 + }, + { + "epoch": 0.2824661590524535, + "grad_norm": 8.4375, + "learning_rate": 4.990811065640333e-06, + "loss": 2.751, + "step": 2671 + }, + { + "epoch": 0.2825719120135364, + "grad_norm": 8.3125, + "learning_rate": 4.990796389986076e-06, + "loss": 2.787, + "step": 2672 + }, + { + "epoch": 0.28267766497461927, + "grad_norm": 8.0, + "learning_rate": 4.990781702643529e-06, + "loss": 2.669, + "step": 2673 + }, + { + "epoch": 0.2827834179357022, + "grad_norm": 8.0625, + "learning_rate": 4.9907670036127635e-06, + "loss": 2.7032, + "step": 2674 + }, + { + "epoch": 0.2828891708967851, + "grad_norm": 8.1875, + "learning_rate": 4.990752292893847e-06, + "loss": 2.5569, + "step": 2675 + }, + { + "epoch": 0.282994923857868, + "grad_norm": 8.8125, + "learning_rate": 4.9907375704868475e-06, + "loss": 2.6454, + "step": 2676 + }, + { + "epoch": 0.28310067681895096, + "grad_norm": 8.5625, + "learning_rate": 4.990722836391837e-06, + "loss": 2.7652, + "step": 2677 + }, + { + "epoch": 0.28320642978003385, + "grad_norm": 7.59375, + "learning_rate": 4.990708090608883e-06, + "loss": 2.419, + "step": 2678 + }, + { + "epoch": 0.28331218274111675, + "grad_norm": 7.75, + "learning_rate": 4.990693333138053e-06, + "loss": 2.4393, + "step": 2679 + }, + { + "epoch": 0.28341793570219964, + "grad_norm": 8.0625, + "learning_rate": 4.990678563979418e-06, + "loss": 2.3433, + "step": 2680 + }, + { + "epoch": 0.2835236886632826, + "grad_norm": 8.6875, + "learning_rate": 4.990663783133048e-06, + "loss": 2.6457, + "step": 2681 + }, + { + "epoch": 0.2836294416243655, + "grad_norm": 8.1875, + "learning_rate": 4.990648990599011e-06, + "loss": 2.5977, + "step": 2682 + }, + { + "epoch": 0.2837351945854484, + "grad_norm": 7.90625, + "learning_rate": 4.990634186377377e-06, + "loss": 2.4751, + "step": 2683 + }, + { + "epoch": 0.28384094754653133, + "grad_norm": 7.53125, + "learning_rate": 4.990619370468216e-06, + "loss": 2.6751, + "step": 2684 + }, + { + "epoch": 0.2839467005076142, + "grad_norm": 7.46875, + "learning_rate": 4.990604542871596e-06, + "loss": 2.6515, + "step": 2685 + }, + { + "epoch": 0.2840524534686971, + "grad_norm": 8.125, + "learning_rate": 4.990589703587588e-06, + "loss": 2.3275, + "step": 2686 + }, + { + "epoch": 0.28415820642978, + "grad_norm": 10.125, + "learning_rate": 4.990574852616261e-06, + "loss": 2.6911, + "step": 2687 + }, + { + "epoch": 0.28426395939086296, + "grad_norm": 10.0625, + "learning_rate": 4.990559989957685e-06, + "loss": 2.8652, + "step": 2688 + }, + { + "epoch": 0.28436971235194586, + "grad_norm": 8.5625, + "learning_rate": 4.99054511561193e-06, + "loss": 2.7996, + "step": 2689 + }, + { + "epoch": 0.28447546531302875, + "grad_norm": 8.6875, + "learning_rate": 4.990530229579065e-06, + "loss": 2.6995, + "step": 2690 + }, + { + "epoch": 0.2845812182741117, + "grad_norm": 8.5, + "learning_rate": 4.99051533185916e-06, + "loss": 2.8211, + "step": 2691 + }, + { + "epoch": 0.2846869712351946, + "grad_norm": 8.625, + "learning_rate": 4.990500422452284e-06, + "loss": 2.4143, + "step": 2692 + }, + { + "epoch": 0.2847927241962775, + "grad_norm": 7.3125, + "learning_rate": 4.99048550135851e-06, + "loss": 2.6098, + "step": 2693 + }, + { + "epoch": 0.2848984771573604, + "grad_norm": 7.5625, + "learning_rate": 4.990470568577905e-06, + "loss": 2.6685, + "step": 2694 + }, + { + "epoch": 0.28500423011844334, + "grad_norm": 7.71875, + "learning_rate": 4.990455624110541e-06, + "loss": 2.4881, + "step": 2695 + }, + { + "epoch": 0.28510998307952623, + "grad_norm": 7.46875, + "learning_rate": 4.990440667956487e-06, + "loss": 2.6667, + "step": 2696 + }, + { + "epoch": 0.2852157360406091, + "grad_norm": 7.59375, + "learning_rate": 4.990425700115813e-06, + "loss": 2.5716, + "step": 2697 + }, + { + "epoch": 0.2853214890016921, + "grad_norm": 7.875, + "learning_rate": 4.990410720588591e-06, + "loss": 2.68, + "step": 2698 + }, + { + "epoch": 0.28542724196277497, + "grad_norm": 7.53125, + "learning_rate": 4.990395729374889e-06, + "loss": 2.5115, + "step": 2699 + }, + { + "epoch": 0.28553299492385786, + "grad_norm": 7.90625, + "learning_rate": 4.990380726474778e-06, + "loss": 2.683, + "step": 2700 + }, + { + "epoch": 0.28563874788494076, + "grad_norm": 8.1875, + "learning_rate": 4.9903657118883295e-06, + "loss": 2.4312, + "step": 2701 + }, + { + "epoch": 0.2857445008460237, + "grad_norm": 8.125, + "learning_rate": 4.990350685615614e-06, + "loss": 2.7856, + "step": 2702 + }, + { + "epoch": 0.2858502538071066, + "grad_norm": 8.6875, + "learning_rate": 4.9903356476567e-06, + "loss": 2.8684, + "step": 2703 + }, + { + "epoch": 0.2859560067681895, + "grad_norm": 7.6875, + "learning_rate": 4.99032059801166e-06, + "loss": 2.4886, + "step": 2704 + }, + { + "epoch": 0.28606175972927245, + "grad_norm": 7.96875, + "learning_rate": 4.990305536680563e-06, + "loss": 2.4933, + "step": 2705 + }, + { + "epoch": 0.28616751269035534, + "grad_norm": 7.84375, + "learning_rate": 4.990290463663481e-06, + "loss": 2.6296, + "step": 2706 + }, + { + "epoch": 0.28627326565143824, + "grad_norm": 7.84375, + "learning_rate": 4.990275378960485e-06, + "loss": 2.6406, + "step": 2707 + }, + { + "epoch": 0.28637901861252113, + "grad_norm": 8.375, + "learning_rate": 4.990260282571645e-06, + "loss": 2.5792, + "step": 2708 + }, + { + "epoch": 0.2864847715736041, + "grad_norm": 8.3125, + "learning_rate": 4.990245174497031e-06, + "loss": 2.5422, + "step": 2709 + }, + { + "epoch": 0.286590524534687, + "grad_norm": 7.53125, + "learning_rate": 4.9902300547367154e-06, + "loss": 2.5361, + "step": 2710 + }, + { + "epoch": 0.28669627749576987, + "grad_norm": 7.90625, + "learning_rate": 4.99021492329077e-06, + "loss": 2.5954, + "step": 2711 + }, + { + "epoch": 0.2868020304568528, + "grad_norm": 8.0625, + "learning_rate": 4.990199780159263e-06, + "loss": 2.709, + "step": 2712 + }, + { + "epoch": 0.2869077834179357, + "grad_norm": 9.0, + "learning_rate": 4.990184625342267e-06, + "loss": 2.89, + "step": 2713 + }, + { + "epoch": 0.2870135363790186, + "grad_norm": 8.625, + "learning_rate": 4.990169458839853e-06, + "loss": 2.9148, + "step": 2714 + }, + { + "epoch": 0.2871192893401015, + "grad_norm": 7.8125, + "learning_rate": 4.990154280652092e-06, + "loss": 2.7387, + "step": 2715 + }, + { + "epoch": 0.28722504230118445, + "grad_norm": 9.75, + "learning_rate": 4.990139090779056e-06, + "loss": 2.6082, + "step": 2716 + }, + { + "epoch": 0.28733079526226735, + "grad_norm": 10.5, + "learning_rate": 4.990123889220815e-06, + "loss": 2.6477, + "step": 2717 + }, + { + "epoch": 0.28743654822335024, + "grad_norm": 8.6875, + "learning_rate": 4.990108675977441e-06, + "loss": 2.9515, + "step": 2718 + }, + { + "epoch": 0.2875423011844332, + "grad_norm": 7.5, + "learning_rate": 4.990093451049006e-06, + "loss": 2.6633, + "step": 2719 + }, + { + "epoch": 0.2876480541455161, + "grad_norm": 8.25, + "learning_rate": 4.990078214435581e-06, + "loss": 2.6974, + "step": 2720 + }, + { + "epoch": 0.287753807106599, + "grad_norm": 7.96875, + "learning_rate": 4.990062966137237e-06, + "loss": 2.6176, + "step": 2721 + }, + { + "epoch": 0.2878595600676819, + "grad_norm": 8.0625, + "learning_rate": 4.990047706154045e-06, + "loss": 2.7455, + "step": 2722 + }, + { + "epoch": 0.2879653130287648, + "grad_norm": 7.59375, + "learning_rate": 4.990032434486079e-06, + "loss": 2.6906, + "step": 2723 + }, + { + "epoch": 0.2880710659898477, + "grad_norm": 8.9375, + "learning_rate": 4.990017151133408e-06, + "loss": 2.5684, + "step": 2724 + }, + { + "epoch": 0.2881768189509306, + "grad_norm": 7.625, + "learning_rate": 4.9900018560961065e-06, + "loss": 2.6471, + "step": 2725 + }, + { + "epoch": 0.28828257191201356, + "grad_norm": 8.0625, + "learning_rate": 4.989986549374244e-06, + "loss": 2.2254, + "step": 2726 + }, + { + "epoch": 0.28838832487309646, + "grad_norm": 9.125, + "learning_rate": 4.9899712309678915e-06, + "loss": 2.8446, + "step": 2727 + }, + { + "epoch": 0.28849407783417935, + "grad_norm": 7.9375, + "learning_rate": 4.989955900877124e-06, + "loss": 2.4404, + "step": 2728 + }, + { + "epoch": 0.28859983079526225, + "grad_norm": 8.875, + "learning_rate": 4.989940559102011e-06, + "loss": 2.8273, + "step": 2729 + }, + { + "epoch": 0.2887055837563452, + "grad_norm": 7.71875, + "learning_rate": 4.989925205642625e-06, + "loss": 2.5043, + "step": 2730 + }, + { + "epoch": 0.2888113367174281, + "grad_norm": 10.1875, + "learning_rate": 4.98990984049904e-06, + "loss": 2.7121, + "step": 2731 + }, + { + "epoch": 0.288917089678511, + "grad_norm": 8.0, + "learning_rate": 4.9898944636713255e-06, + "loss": 2.6317, + "step": 2732 + }, + { + "epoch": 0.28902284263959394, + "grad_norm": 7.34375, + "learning_rate": 4.989879075159554e-06, + "loss": 2.5221, + "step": 2733 + }, + { + "epoch": 0.28912859560067683, + "grad_norm": 7.3125, + "learning_rate": 4.989863674963799e-06, + "loss": 2.602, + "step": 2734 + }, + { + "epoch": 0.2892343485617597, + "grad_norm": 7.53125, + "learning_rate": 4.989848263084132e-06, + "loss": 2.6824, + "step": 2735 + }, + { + "epoch": 0.2893401015228426, + "grad_norm": 8.0625, + "learning_rate": 4.989832839520625e-06, + "loss": 2.8703, + "step": 2736 + }, + { + "epoch": 0.28944585448392557, + "grad_norm": 7.21875, + "learning_rate": 4.989817404273351e-06, + "loss": 2.4994, + "step": 2737 + }, + { + "epoch": 0.28955160744500846, + "grad_norm": 8.125, + "learning_rate": 4.989801957342383e-06, + "loss": 2.4715, + "step": 2738 + }, + { + "epoch": 0.28965736040609136, + "grad_norm": 7.90625, + "learning_rate": 4.989786498727792e-06, + "loss": 2.6549, + "step": 2739 + }, + { + "epoch": 0.2897631133671743, + "grad_norm": 7.21875, + "learning_rate": 4.989771028429651e-06, + "loss": 2.5318, + "step": 2740 + }, + { + "epoch": 0.2898688663282572, + "grad_norm": 8.625, + "learning_rate": 4.989755546448033e-06, + "loss": 2.7238, + "step": 2741 + }, + { + "epoch": 0.2899746192893401, + "grad_norm": 8.5625, + "learning_rate": 4.989740052783011e-06, + "loss": 2.7022, + "step": 2742 + }, + { + "epoch": 0.290080372250423, + "grad_norm": 7.78125, + "learning_rate": 4.989724547434657e-06, + "loss": 2.6965, + "step": 2743 + }, + { + "epoch": 0.29018612521150594, + "grad_norm": 8.9375, + "learning_rate": 4.9897090304030435e-06, + "loss": 2.8937, + "step": 2744 + }, + { + "epoch": 0.29029187817258884, + "grad_norm": 7.875, + "learning_rate": 4.989693501688244e-06, + "loss": 2.7733, + "step": 2745 + }, + { + "epoch": 0.29039763113367173, + "grad_norm": 8.375, + "learning_rate": 4.989677961290331e-06, + "loss": 2.3967, + "step": 2746 + }, + { + "epoch": 0.2905033840947547, + "grad_norm": 7.59375, + "learning_rate": 4.989662409209378e-06, + "loss": 2.3178, + "step": 2747 + }, + { + "epoch": 0.2906091370558376, + "grad_norm": 8.0625, + "learning_rate": 4.989646845445458e-06, + "loss": 2.5432, + "step": 2748 + }, + { + "epoch": 0.29071489001692047, + "grad_norm": 7.90625, + "learning_rate": 4.989631269998643e-06, + "loss": 2.7265, + "step": 2749 + }, + { + "epoch": 0.29082064297800336, + "grad_norm": 7.53125, + "learning_rate": 4.989615682869007e-06, + "loss": 2.61, + "step": 2750 + }, + { + "epoch": 0.2909263959390863, + "grad_norm": 7.84375, + "learning_rate": 4.989600084056622e-06, + "loss": 2.5493, + "step": 2751 + }, + { + "epoch": 0.2910321489001692, + "grad_norm": 8.125, + "learning_rate": 4.989584473561563e-06, + "loss": 2.8518, + "step": 2752 + }, + { + "epoch": 0.2911379018612521, + "grad_norm": 7.78125, + "learning_rate": 4.989568851383901e-06, + "loss": 2.7151, + "step": 2753 + }, + { + "epoch": 0.29124365482233505, + "grad_norm": 8.3125, + "learning_rate": 4.989553217523712e-06, + "loss": 2.6651, + "step": 2754 + }, + { + "epoch": 0.29134940778341795, + "grad_norm": 8.375, + "learning_rate": 4.989537571981067e-06, + "loss": 2.7894, + "step": 2755 + }, + { + "epoch": 0.29145516074450084, + "grad_norm": 7.78125, + "learning_rate": 4.98952191475604e-06, + "loss": 2.6863, + "step": 2756 + }, + { + "epoch": 0.29156091370558374, + "grad_norm": 8.3125, + "learning_rate": 4.989506245848706e-06, + "loss": 2.5587, + "step": 2757 + }, + { + "epoch": 0.2916666666666667, + "grad_norm": 7.65625, + "learning_rate": 4.989490565259136e-06, + "loss": 2.5266, + "step": 2758 + }, + { + "epoch": 0.2917724196277496, + "grad_norm": 7.65625, + "learning_rate": 4.989474872987406e-06, + "loss": 2.4897, + "step": 2759 + }, + { + "epoch": 0.2918781725888325, + "grad_norm": 8.0625, + "learning_rate": 4.989459169033589e-06, + "loss": 2.5301, + "step": 2760 + }, + { + "epoch": 0.2919839255499154, + "grad_norm": 8.625, + "learning_rate": 4.989443453397757e-06, + "loss": 2.7136, + "step": 2761 + }, + { + "epoch": 0.2920896785109983, + "grad_norm": 7.40625, + "learning_rate": 4.989427726079986e-06, + "loss": 2.5751, + "step": 2762 + }, + { + "epoch": 0.2921954314720812, + "grad_norm": 8.375, + "learning_rate": 4.989411987080348e-06, + "loss": 2.7122, + "step": 2763 + }, + { + "epoch": 0.2923011844331641, + "grad_norm": 7.375, + "learning_rate": 4.989396236398919e-06, + "loss": 2.6862, + "step": 2764 + }, + { + "epoch": 0.29240693739424706, + "grad_norm": 7.75, + "learning_rate": 4.98938047403577e-06, + "loss": 2.6215, + "step": 2765 + }, + { + "epoch": 0.29251269035532995, + "grad_norm": 9.6875, + "learning_rate": 4.989364699990978e-06, + "loss": 2.7205, + "step": 2766 + }, + { + "epoch": 0.29261844331641285, + "grad_norm": 8.0, + "learning_rate": 4.989348914264615e-06, + "loss": 2.7929, + "step": 2767 + }, + { + "epoch": 0.2927241962774958, + "grad_norm": 7.65625, + "learning_rate": 4.989333116856755e-06, + "loss": 2.6708, + "step": 2768 + }, + { + "epoch": 0.2928299492385787, + "grad_norm": 7.90625, + "learning_rate": 4.989317307767474e-06, + "loss": 2.6403, + "step": 2769 + }, + { + "epoch": 0.2929357021996616, + "grad_norm": 8.625, + "learning_rate": 4.989301486996844e-06, + "loss": 2.6331, + "step": 2770 + }, + { + "epoch": 0.2930414551607445, + "grad_norm": 7.96875, + "learning_rate": 4.98928565454494e-06, + "loss": 2.8621, + "step": 2771 + }, + { + "epoch": 0.29314720812182743, + "grad_norm": 8.125, + "learning_rate": 4.989269810411838e-06, + "loss": 2.6512, + "step": 2772 + }, + { + "epoch": 0.2932529610829103, + "grad_norm": 8.4375, + "learning_rate": 4.989253954597609e-06, + "loss": 2.8392, + "step": 2773 + }, + { + "epoch": 0.2933587140439932, + "grad_norm": 7.625, + "learning_rate": 4.98923808710233e-06, + "loss": 2.6477, + "step": 2774 + }, + { + "epoch": 0.29346446700507617, + "grad_norm": 7.9375, + "learning_rate": 4.989222207926076e-06, + "loss": 2.6758, + "step": 2775 + }, + { + "epoch": 0.29357021996615906, + "grad_norm": 8.3125, + "learning_rate": 4.989206317068918e-06, + "loss": 2.8553, + "step": 2776 + }, + { + "epoch": 0.29367597292724196, + "grad_norm": 8.125, + "learning_rate": 4.989190414530934e-06, + "loss": 2.6492, + "step": 2777 + }, + { + "epoch": 0.29378172588832485, + "grad_norm": 8.375, + "learning_rate": 4.9891745003121976e-06, + "loss": 2.741, + "step": 2778 + }, + { + "epoch": 0.2938874788494078, + "grad_norm": 8.5, + "learning_rate": 4.989158574412782e-06, + "loss": 2.5874, + "step": 2779 + }, + { + "epoch": 0.2939932318104907, + "grad_norm": 8.125, + "learning_rate": 4.989142636832765e-06, + "loss": 2.6381, + "step": 2780 + }, + { + "epoch": 0.2940989847715736, + "grad_norm": 7.78125, + "learning_rate": 4.989126687572219e-06, + "loss": 2.6161, + "step": 2781 + }, + { + "epoch": 0.29420473773265654, + "grad_norm": 8.125, + "learning_rate": 4.989110726631219e-06, + "loss": 2.6055, + "step": 2782 + }, + { + "epoch": 0.29431049069373943, + "grad_norm": 7.4375, + "learning_rate": 4.989094754009841e-06, + "loss": 2.5402, + "step": 2783 + }, + { + "epoch": 0.29441624365482233, + "grad_norm": 7.75, + "learning_rate": 4.989078769708159e-06, + "loss": 2.8603, + "step": 2784 + }, + { + "epoch": 0.2945219966159052, + "grad_norm": 7.4375, + "learning_rate": 4.989062773726248e-06, + "loss": 2.5826, + "step": 2785 + }, + { + "epoch": 0.2946277495769882, + "grad_norm": 7.75, + "learning_rate": 4.9890467660641835e-06, + "loss": 2.4815, + "step": 2786 + }, + { + "epoch": 0.29473350253807107, + "grad_norm": 7.71875, + "learning_rate": 4.989030746722041e-06, + "loss": 2.4302, + "step": 2787 + }, + { + "epoch": 0.29483925549915396, + "grad_norm": 7.8125, + "learning_rate": 4.989014715699896e-06, + "loss": 2.5025, + "step": 2788 + }, + { + "epoch": 0.2949450084602369, + "grad_norm": 7.4375, + "learning_rate": 4.988998672997821e-06, + "loss": 2.7718, + "step": 2789 + }, + { + "epoch": 0.2950507614213198, + "grad_norm": 7.90625, + "learning_rate": 4.988982618615895e-06, + "loss": 2.6602, + "step": 2790 + }, + { + "epoch": 0.2951565143824027, + "grad_norm": 8.5625, + "learning_rate": 4.98896655255419e-06, + "loss": 2.5062, + "step": 2791 + }, + { + "epoch": 0.2952622673434856, + "grad_norm": 7.6875, + "learning_rate": 4.988950474812784e-06, + "loss": 2.5917, + "step": 2792 + }, + { + "epoch": 0.29536802030456855, + "grad_norm": 7.96875, + "learning_rate": 4.9889343853917505e-06, + "loss": 2.7424, + "step": 2793 + }, + { + "epoch": 0.29547377326565144, + "grad_norm": 7.6875, + "learning_rate": 4.988918284291166e-06, + "loss": 2.7171, + "step": 2794 + }, + { + "epoch": 0.29557952622673433, + "grad_norm": 7.40625, + "learning_rate": 4.988902171511107e-06, + "loss": 2.5925, + "step": 2795 + }, + { + "epoch": 0.2956852791878173, + "grad_norm": 8.5625, + "learning_rate": 4.988886047051647e-06, + "loss": 2.743, + "step": 2796 + }, + { + "epoch": 0.2957910321489002, + "grad_norm": 7.78125, + "learning_rate": 4.988869910912863e-06, + "loss": 2.5561, + "step": 2797 + }, + { + "epoch": 0.2958967851099831, + "grad_norm": 7.34375, + "learning_rate": 4.9888537630948295e-06, + "loss": 2.4645, + "step": 2798 + }, + { + "epoch": 0.29600253807106597, + "grad_norm": 7.625, + "learning_rate": 4.988837603597625e-06, + "loss": 2.5548, + "step": 2799 + }, + { + "epoch": 0.2961082910321489, + "grad_norm": 7.9375, + "learning_rate": 4.988821432421322e-06, + "loss": 2.4106, + "step": 2800 + }, + { + "epoch": 0.2962140439932318, + "grad_norm": 7.875, + "learning_rate": 4.988805249565999e-06, + "loss": 2.5737, + "step": 2801 + }, + { + "epoch": 0.2963197969543147, + "grad_norm": 8.1875, + "learning_rate": 4.98878905503173e-06, + "loss": 2.5662, + "step": 2802 + }, + { + "epoch": 0.29642554991539766, + "grad_norm": 8.25, + "learning_rate": 4.988772848818592e-06, + "loss": 2.3748, + "step": 2803 + }, + { + "epoch": 0.29653130287648055, + "grad_norm": 7.4375, + "learning_rate": 4.988756630926661e-06, + "loss": 2.5758, + "step": 2804 + }, + { + "epoch": 0.29663705583756345, + "grad_norm": 8.25, + "learning_rate": 4.9887404013560136e-06, + "loss": 2.7934, + "step": 2805 + }, + { + "epoch": 0.29674280879864634, + "grad_norm": 8.0625, + "learning_rate": 4.988724160106725e-06, + "loss": 2.5298, + "step": 2806 + }, + { + "epoch": 0.2968485617597293, + "grad_norm": 7.28125, + "learning_rate": 4.9887079071788715e-06, + "loss": 2.545, + "step": 2807 + }, + { + "epoch": 0.2969543147208122, + "grad_norm": 10.0, + "learning_rate": 4.988691642572529e-06, + "loss": 2.7168, + "step": 2808 + }, + { + "epoch": 0.2970600676818951, + "grad_norm": 7.96875, + "learning_rate": 4.988675366287776e-06, + "loss": 2.5832, + "step": 2809 + }, + { + "epoch": 0.29716582064297803, + "grad_norm": 7.5, + "learning_rate": 4.988659078324686e-06, + "loss": 2.6228, + "step": 2810 + }, + { + "epoch": 0.2972715736040609, + "grad_norm": 8.0625, + "learning_rate": 4.988642778683338e-06, + "loss": 2.6563, + "step": 2811 + }, + { + "epoch": 0.2973773265651438, + "grad_norm": 8.3125, + "learning_rate": 4.988626467363806e-06, + "loss": 2.6263, + "step": 2812 + }, + { + "epoch": 0.2974830795262267, + "grad_norm": 9.0, + "learning_rate": 4.9886101443661685e-06, + "loss": 2.7251, + "step": 2813 + }, + { + "epoch": 0.29758883248730966, + "grad_norm": 7.65625, + "learning_rate": 4.988593809690501e-06, + "loss": 2.6191, + "step": 2814 + }, + { + "epoch": 0.29769458544839256, + "grad_norm": 8.4375, + "learning_rate": 4.988577463336881e-06, + "loss": 2.853, + "step": 2815 + }, + { + "epoch": 0.29780033840947545, + "grad_norm": 7.90625, + "learning_rate": 4.988561105305385e-06, + "loss": 2.7996, + "step": 2816 + }, + { + "epoch": 0.2979060913705584, + "grad_norm": 7.8125, + "learning_rate": 4.988544735596088e-06, + "loss": 2.5039, + "step": 2817 + }, + { + "epoch": 0.2980118443316413, + "grad_norm": 7.5625, + "learning_rate": 4.9885283542090705e-06, + "loss": 2.6856, + "step": 2818 + }, + { + "epoch": 0.2981175972927242, + "grad_norm": 8.9375, + "learning_rate": 4.988511961144406e-06, + "loss": 2.4736, + "step": 2819 + }, + { + "epoch": 0.2982233502538071, + "grad_norm": 7.3125, + "learning_rate": 4.988495556402172e-06, + "loss": 2.5963, + "step": 2820 + }, + { + "epoch": 0.29832910321489003, + "grad_norm": 7.9375, + "learning_rate": 4.988479139982448e-06, + "loss": 2.6103, + "step": 2821 + }, + { + "epoch": 0.29843485617597293, + "grad_norm": 7.5625, + "learning_rate": 4.9884627118853075e-06, + "loss": 2.5305, + "step": 2822 + }, + { + "epoch": 0.2985406091370558, + "grad_norm": 7.5625, + "learning_rate": 4.98844627211083e-06, + "loss": 2.6902, + "step": 2823 + }, + { + "epoch": 0.2986463620981388, + "grad_norm": 7.75, + "learning_rate": 4.988429820659091e-06, + "loss": 2.7974, + "step": 2824 + }, + { + "epoch": 0.29875211505922167, + "grad_norm": 7.625, + "learning_rate": 4.988413357530169e-06, + "loss": 2.5553, + "step": 2825 + }, + { + "epoch": 0.29885786802030456, + "grad_norm": 7.875, + "learning_rate": 4.988396882724142e-06, + "loss": 2.5616, + "step": 2826 + }, + { + "epoch": 0.29896362098138746, + "grad_norm": 11.125, + "learning_rate": 4.988380396241085e-06, + "loss": 2.5966, + "step": 2827 + }, + { + "epoch": 0.2990693739424704, + "grad_norm": 7.96875, + "learning_rate": 4.988363898081077e-06, + "loss": 2.3119, + "step": 2828 + }, + { + "epoch": 0.2991751269035533, + "grad_norm": 8.1875, + "learning_rate": 4.988347388244194e-06, + "loss": 2.5973, + "step": 2829 + }, + { + "epoch": 0.2992808798646362, + "grad_norm": 7.65625, + "learning_rate": 4.988330866730515e-06, + "loss": 2.3819, + "step": 2830 + }, + { + "epoch": 0.29938663282571915, + "grad_norm": 7.84375, + "learning_rate": 4.988314333540118e-06, + "loss": 2.6486, + "step": 2831 + }, + { + "epoch": 0.29949238578680204, + "grad_norm": 7.46875, + "learning_rate": 4.988297788673078e-06, + "loss": 2.5248, + "step": 2832 + }, + { + "epoch": 0.29959813874788493, + "grad_norm": 7.71875, + "learning_rate": 4.988281232129475e-06, + "loss": 2.6715, + "step": 2833 + }, + { + "epoch": 0.29970389170896783, + "grad_norm": 9.6875, + "learning_rate": 4.988264663909386e-06, + "loss": 2.3906, + "step": 2834 + }, + { + "epoch": 0.2998096446700508, + "grad_norm": 8.1875, + "learning_rate": 4.988248084012888e-06, + "loss": 2.6565, + "step": 2835 + }, + { + "epoch": 0.2999153976311337, + "grad_norm": 8.3125, + "learning_rate": 4.9882314924400584e-06, + "loss": 2.5808, + "step": 2836 + }, + { + "epoch": 0.30002115059221657, + "grad_norm": 9.0625, + "learning_rate": 4.988214889190978e-06, + "loss": 3.2284, + "step": 2837 + }, + { + "epoch": 0.3001269035532995, + "grad_norm": 8.5, + "learning_rate": 4.988198274265722e-06, + "loss": 2.65, + "step": 2838 + }, + { + "epoch": 0.3002326565143824, + "grad_norm": 8.625, + "learning_rate": 4.988181647664369e-06, + "loss": 2.5381, + "step": 2839 + }, + { + "epoch": 0.3003384094754653, + "grad_norm": 7.53125, + "learning_rate": 4.988165009386997e-06, + "loss": 2.6177, + "step": 2840 + }, + { + "epoch": 0.3004441624365482, + "grad_norm": 8.0625, + "learning_rate": 4.988148359433685e-06, + "loss": 2.7428, + "step": 2841 + }, + { + "epoch": 0.30054991539763115, + "grad_norm": 7.75, + "learning_rate": 4.98813169780451e-06, + "loss": 2.4471, + "step": 2842 + }, + { + "epoch": 0.30065566835871405, + "grad_norm": 8.25, + "learning_rate": 4.98811502449955e-06, + "loss": 2.5902, + "step": 2843 + }, + { + "epoch": 0.30076142131979694, + "grad_norm": 7.84375, + "learning_rate": 4.9880983395188845e-06, + "loss": 2.6461, + "step": 2844 + }, + { + "epoch": 0.3008671742808799, + "grad_norm": 7.25, + "learning_rate": 4.98808164286259e-06, + "loss": 2.5556, + "step": 2845 + }, + { + "epoch": 0.3009729272419628, + "grad_norm": 7.65625, + "learning_rate": 4.988064934530747e-06, + "loss": 2.5109, + "step": 2846 + }, + { + "epoch": 0.3010786802030457, + "grad_norm": 7.625, + "learning_rate": 4.9880482145234335e-06, + "loss": 2.653, + "step": 2847 + }, + { + "epoch": 0.3011844331641286, + "grad_norm": 8.0625, + "learning_rate": 4.988031482840726e-06, + "loss": 2.7763, + "step": 2848 + }, + { + "epoch": 0.3012901861252115, + "grad_norm": 7.46875, + "learning_rate": 4.988014739482705e-06, + "loss": 2.7006, + "step": 2849 + }, + { + "epoch": 0.3013959390862944, + "grad_norm": 12.75, + "learning_rate": 4.987997984449447e-06, + "loss": 2.7091, + "step": 2850 + }, + { + "epoch": 0.3015016920473773, + "grad_norm": 7.8125, + "learning_rate": 4.987981217741034e-06, + "loss": 2.6797, + "step": 2851 + }, + { + "epoch": 0.30160744500846026, + "grad_norm": 7.9375, + "learning_rate": 4.987964439357542e-06, + "loss": 2.6877, + "step": 2852 + }, + { + "epoch": 0.30171319796954316, + "grad_norm": 7.65625, + "learning_rate": 4.98794764929905e-06, + "loss": 2.4311, + "step": 2853 + }, + { + "epoch": 0.30181895093062605, + "grad_norm": 7.9375, + "learning_rate": 4.9879308475656375e-06, + "loss": 2.6641, + "step": 2854 + }, + { + "epoch": 0.30192470389170895, + "grad_norm": 7.6875, + "learning_rate": 4.987914034157383e-06, + "loss": 2.6236, + "step": 2855 + }, + { + "epoch": 0.3020304568527919, + "grad_norm": 7.8125, + "learning_rate": 4.987897209074366e-06, + "loss": 2.5199, + "step": 2856 + }, + { + "epoch": 0.3021362098138748, + "grad_norm": 7.6875, + "learning_rate": 4.987880372316664e-06, + "loss": 2.6121, + "step": 2857 + }, + { + "epoch": 0.3022419627749577, + "grad_norm": 8.8125, + "learning_rate": 4.9878635238843575e-06, + "loss": 2.713, + "step": 2858 + }, + { + "epoch": 0.30234771573604063, + "grad_norm": 8.4375, + "learning_rate": 4.987846663777525e-06, + "loss": 2.3181, + "step": 2859 + }, + { + "epoch": 0.30245346869712353, + "grad_norm": 7.40625, + "learning_rate": 4.987829791996245e-06, + "loss": 2.6684, + "step": 2860 + }, + { + "epoch": 0.3025592216582064, + "grad_norm": 7.5625, + "learning_rate": 4.987812908540598e-06, + "loss": 2.6176, + "step": 2861 + }, + { + "epoch": 0.3026649746192893, + "grad_norm": 8.25, + "learning_rate": 4.987796013410662e-06, + "loss": 2.6842, + "step": 2862 + }, + { + "epoch": 0.30277072758037227, + "grad_norm": 7.875, + "learning_rate": 4.987779106606516e-06, + "loss": 2.7922, + "step": 2863 + }, + { + "epoch": 0.30287648054145516, + "grad_norm": 8.5, + "learning_rate": 4.987762188128242e-06, + "loss": 2.8345, + "step": 2864 + }, + { + "epoch": 0.30298223350253806, + "grad_norm": 8.6875, + "learning_rate": 4.9877452579759165e-06, + "loss": 2.5577, + "step": 2865 + }, + { + "epoch": 0.303087986463621, + "grad_norm": 8.375, + "learning_rate": 4.98772831614962e-06, + "loss": 2.6533, + "step": 2866 + }, + { + "epoch": 0.3031937394247039, + "grad_norm": 8.125, + "learning_rate": 4.987711362649432e-06, + "loss": 2.8148, + "step": 2867 + }, + { + "epoch": 0.3032994923857868, + "grad_norm": 12.0, + "learning_rate": 4.987694397475432e-06, + "loss": 2.718, + "step": 2868 + }, + { + "epoch": 0.3034052453468697, + "grad_norm": 8.3125, + "learning_rate": 4.9876774206277e-06, + "loss": 2.8318, + "step": 2869 + }, + { + "epoch": 0.30351099830795264, + "grad_norm": 8.0625, + "learning_rate": 4.987660432106315e-06, + "loss": 2.5536, + "step": 2870 + }, + { + "epoch": 0.30361675126903553, + "grad_norm": 7.5625, + "learning_rate": 4.987643431911357e-06, + "loss": 2.4738, + "step": 2871 + }, + { + "epoch": 0.30372250423011843, + "grad_norm": 7.3125, + "learning_rate": 4.987626420042906e-06, + "loss": 2.5433, + "step": 2872 + }, + { + "epoch": 0.3038282571912014, + "grad_norm": 7.9375, + "learning_rate": 4.98760939650104e-06, + "loss": 2.686, + "step": 2873 + }, + { + "epoch": 0.3039340101522843, + "grad_norm": 7.65625, + "learning_rate": 4.987592361285842e-06, + "loss": 2.3929, + "step": 2874 + }, + { + "epoch": 0.30403976311336717, + "grad_norm": 8.0625, + "learning_rate": 4.987575314397391e-06, + "loss": 2.3835, + "step": 2875 + }, + { + "epoch": 0.30414551607445006, + "grad_norm": 7.15625, + "learning_rate": 4.987558255835766e-06, + "loss": 2.4774, + "step": 2876 + }, + { + "epoch": 0.304251269035533, + "grad_norm": 7.96875, + "learning_rate": 4.987541185601047e-06, + "loss": 2.773, + "step": 2877 + }, + { + "epoch": 0.3043570219966159, + "grad_norm": 7.78125, + "learning_rate": 4.987524103693315e-06, + "loss": 2.6394, + "step": 2878 + }, + { + "epoch": 0.3044627749576988, + "grad_norm": 7.3125, + "learning_rate": 4.987507010112649e-06, + "loss": 2.7101, + "step": 2879 + }, + { + "epoch": 0.30456852791878175, + "grad_norm": 8.1875, + "learning_rate": 4.987489904859131e-06, + "loss": 2.5963, + "step": 2880 + }, + { + "epoch": 0.30467428087986465, + "grad_norm": 8.625, + "learning_rate": 4.98747278793284e-06, + "loss": 2.9058, + "step": 2881 + }, + { + "epoch": 0.30478003384094754, + "grad_norm": 8.0, + "learning_rate": 4.987455659333856e-06, + "loss": 2.6428, + "step": 2882 + }, + { + "epoch": 0.30488578680203043, + "grad_norm": 7.46875, + "learning_rate": 4.98743851906226e-06, + "loss": 2.4885, + "step": 2883 + }, + { + "epoch": 0.3049915397631134, + "grad_norm": 8.0, + "learning_rate": 4.987421367118132e-06, + "loss": 2.4421, + "step": 2884 + }, + { + "epoch": 0.3050972927241963, + "grad_norm": 7.96875, + "learning_rate": 4.987404203501554e-06, + "loss": 2.9148, + "step": 2885 + }, + { + "epoch": 0.3052030456852792, + "grad_norm": 7.4375, + "learning_rate": 4.987387028212604e-06, + "loss": 2.7375, + "step": 2886 + }, + { + "epoch": 0.3053087986463621, + "grad_norm": 11.0625, + "learning_rate": 4.9873698412513656e-06, + "loss": 2.6201, + "step": 2887 + }, + { + "epoch": 0.305414551607445, + "grad_norm": 8.3125, + "learning_rate": 4.987352642617917e-06, + "loss": 2.8232, + "step": 2888 + }, + { + "epoch": 0.3055203045685279, + "grad_norm": 8.8125, + "learning_rate": 4.9873354323123395e-06, + "loss": 2.6844, + "step": 2889 + }, + { + "epoch": 0.3056260575296108, + "grad_norm": 7.5625, + "learning_rate": 4.987318210334715e-06, + "loss": 2.6437, + "step": 2890 + }, + { + "epoch": 0.30573181049069376, + "grad_norm": 7.625, + "learning_rate": 4.987300976685122e-06, + "loss": 2.6914, + "step": 2891 + }, + { + "epoch": 0.30583756345177665, + "grad_norm": 7.90625, + "learning_rate": 4.987283731363643e-06, + "loss": 2.712, + "step": 2892 + }, + { + "epoch": 0.30594331641285955, + "grad_norm": 9.0, + "learning_rate": 4.98726647437036e-06, + "loss": 2.8344, + "step": 2893 + }, + { + "epoch": 0.3060490693739425, + "grad_norm": 7.375, + "learning_rate": 4.987249205705352e-06, + "loss": 2.6505, + "step": 2894 + }, + { + "epoch": 0.3061548223350254, + "grad_norm": 7.9375, + "learning_rate": 4.987231925368701e-06, + "loss": 2.6437, + "step": 2895 + }, + { + "epoch": 0.3062605752961083, + "grad_norm": 8.0, + "learning_rate": 4.987214633360487e-06, + "loss": 2.6997, + "step": 2896 + }, + { + "epoch": 0.3063663282571912, + "grad_norm": 7.6875, + "learning_rate": 4.987197329680793e-06, + "loss": 2.6439, + "step": 2897 + }, + { + "epoch": 0.30647208121827413, + "grad_norm": 7.84375, + "learning_rate": 4.987180014329698e-06, + "loss": 2.6492, + "step": 2898 + }, + { + "epoch": 0.306577834179357, + "grad_norm": 8.125, + "learning_rate": 4.987162687307285e-06, + "loss": 2.42, + "step": 2899 + }, + { + "epoch": 0.3066835871404399, + "grad_norm": 7.65625, + "learning_rate": 4.987145348613635e-06, + "loss": 2.4938, + "step": 2900 + }, + { + "epoch": 0.30678934010152287, + "grad_norm": 7.75, + "learning_rate": 4.987127998248829e-06, + "loss": 2.4965, + "step": 2901 + }, + { + "epoch": 0.30689509306260576, + "grad_norm": 8.375, + "learning_rate": 4.9871106362129484e-06, + "loss": 2.5102, + "step": 2902 + }, + { + "epoch": 0.30700084602368866, + "grad_norm": 7.96875, + "learning_rate": 4.987093262506075e-06, + "loss": 2.5825, + "step": 2903 + }, + { + "epoch": 0.30710659898477155, + "grad_norm": 7.96875, + "learning_rate": 4.98707587712829e-06, + "loss": 2.6596, + "step": 2904 + }, + { + "epoch": 0.3072123519458545, + "grad_norm": 8.0, + "learning_rate": 4.987058480079675e-06, + "loss": 2.49, + "step": 2905 + }, + { + "epoch": 0.3073181049069374, + "grad_norm": 7.96875, + "learning_rate": 4.987041071360311e-06, + "loss": 2.5842, + "step": 2906 + }, + { + "epoch": 0.3074238578680203, + "grad_norm": 7.5, + "learning_rate": 4.98702365097028e-06, + "loss": 2.6263, + "step": 2907 + }, + { + "epoch": 0.30752961082910324, + "grad_norm": 7.5, + "learning_rate": 4.987006218909666e-06, + "loss": 2.6441, + "step": 2908 + }, + { + "epoch": 0.30763536379018613, + "grad_norm": 7.625, + "learning_rate": 4.986988775178548e-06, + "loss": 2.4931, + "step": 2909 + }, + { + "epoch": 0.30774111675126903, + "grad_norm": 7.6875, + "learning_rate": 4.98697131977701e-06, + "loss": 2.6968, + "step": 2910 + }, + { + "epoch": 0.3078468697123519, + "grad_norm": 7.8125, + "learning_rate": 4.986953852705132e-06, + "loss": 2.7347, + "step": 2911 + }, + { + "epoch": 0.3079526226734349, + "grad_norm": 7.9375, + "learning_rate": 4.986936373962996e-06, + "loss": 2.714, + "step": 2912 + }, + { + "epoch": 0.30805837563451777, + "grad_norm": 8.25, + "learning_rate": 4.9869188835506845e-06, + "loss": 2.7135, + "step": 2913 + }, + { + "epoch": 0.30816412859560066, + "grad_norm": 8.1875, + "learning_rate": 4.986901381468281e-06, + "loss": 2.6031, + "step": 2914 + }, + { + "epoch": 0.3082698815566836, + "grad_norm": 8.25, + "learning_rate": 4.986883867715866e-06, + "loss": 2.2819, + "step": 2915 + }, + { + "epoch": 0.3083756345177665, + "grad_norm": 8.125, + "learning_rate": 4.986866342293521e-06, + "loss": 2.4892, + "step": 2916 + }, + { + "epoch": 0.3084813874788494, + "grad_norm": 7.875, + "learning_rate": 4.986848805201331e-06, + "loss": 2.6256, + "step": 2917 + }, + { + "epoch": 0.3085871404399323, + "grad_norm": 7.90625, + "learning_rate": 4.9868312564393765e-06, + "loss": 2.5764, + "step": 2918 + }, + { + "epoch": 0.30869289340101524, + "grad_norm": 8.1875, + "learning_rate": 4.986813696007739e-06, + "loss": 2.5933, + "step": 2919 + }, + { + "epoch": 0.30879864636209814, + "grad_norm": 8.25, + "learning_rate": 4.986796123906502e-06, + "loss": 2.5418, + "step": 2920 + }, + { + "epoch": 0.30890439932318103, + "grad_norm": 7.75, + "learning_rate": 4.986778540135749e-06, + "loss": 2.6464, + "step": 2921 + }, + { + "epoch": 0.309010152284264, + "grad_norm": 9.875, + "learning_rate": 4.98676094469556e-06, + "loss": 2.6031, + "step": 2922 + }, + { + "epoch": 0.3091159052453469, + "grad_norm": 7.84375, + "learning_rate": 4.98674333758602e-06, + "loss": 2.7052, + "step": 2923 + }, + { + "epoch": 0.3092216582064298, + "grad_norm": 7.71875, + "learning_rate": 4.98672571880721e-06, + "loss": 2.5927, + "step": 2924 + }, + { + "epoch": 0.30932741116751267, + "grad_norm": 8.1875, + "learning_rate": 4.986708088359214e-06, + "loss": 2.3573, + "step": 2925 + }, + { + "epoch": 0.3094331641285956, + "grad_norm": 7.25, + "learning_rate": 4.986690446242113e-06, + "loss": 2.53, + "step": 2926 + }, + { + "epoch": 0.3095389170896785, + "grad_norm": 8.875, + "learning_rate": 4.986672792455991e-06, + "loss": 2.7392, + "step": 2927 + }, + { + "epoch": 0.3096446700507614, + "grad_norm": 8.1875, + "learning_rate": 4.98665512700093e-06, + "loss": 2.7405, + "step": 2928 + }, + { + "epoch": 0.30975042301184436, + "grad_norm": 11.1875, + "learning_rate": 4.986637449877014e-06, + "loss": 2.6065, + "step": 2929 + }, + { + "epoch": 0.30985617597292725, + "grad_norm": 7.8125, + "learning_rate": 4.986619761084326e-06, + "loss": 2.5052, + "step": 2930 + }, + { + "epoch": 0.30996192893401014, + "grad_norm": 8.1875, + "learning_rate": 4.986602060622948e-06, + "loss": 2.7137, + "step": 2931 + }, + { + "epoch": 0.31006768189509304, + "grad_norm": 8.1875, + "learning_rate": 4.986584348492964e-06, + "loss": 2.7912, + "step": 2932 + }, + { + "epoch": 0.310173434856176, + "grad_norm": 9.0625, + "learning_rate": 4.986566624694456e-06, + "loss": 2.7277, + "step": 2933 + }, + { + "epoch": 0.3102791878172589, + "grad_norm": 8.1875, + "learning_rate": 4.986548889227508e-06, + "loss": 2.9011, + "step": 2934 + }, + { + "epoch": 0.3103849407783418, + "grad_norm": 8.125, + "learning_rate": 4.986531142092203e-06, + "loss": 2.6791, + "step": 2935 + }, + { + "epoch": 0.31049069373942473, + "grad_norm": 8.0, + "learning_rate": 4.986513383288625e-06, + "loss": 2.5836, + "step": 2936 + }, + { + "epoch": 0.3105964467005076, + "grad_norm": 8.5, + "learning_rate": 4.986495612816855e-06, + "loss": 2.5072, + "step": 2937 + }, + { + "epoch": 0.3107021996615905, + "grad_norm": 8.0625, + "learning_rate": 4.98647783067698e-06, + "loss": 2.646, + "step": 2938 + }, + { + "epoch": 0.3108079526226734, + "grad_norm": 8.0, + "learning_rate": 4.9864600368690794e-06, + "loss": 2.5106, + "step": 2939 + }, + { + "epoch": 0.31091370558375636, + "grad_norm": 7.875, + "learning_rate": 4.986442231393241e-06, + "loss": 2.8034, + "step": 2940 + }, + { + "epoch": 0.31101945854483926, + "grad_norm": 7.59375, + "learning_rate": 4.9864244142495444e-06, + "loss": 2.5845, + "step": 2941 + }, + { + "epoch": 0.31112521150592215, + "grad_norm": 8.125, + "learning_rate": 4.986406585438075e-06, + "loss": 2.533, + "step": 2942 + }, + { + "epoch": 0.3112309644670051, + "grad_norm": 7.09375, + "learning_rate": 4.986388744958918e-06, + "loss": 2.5582, + "step": 2943 + }, + { + "epoch": 0.311336717428088, + "grad_norm": 7.3125, + "learning_rate": 4.986370892812154e-06, + "loss": 2.5513, + "step": 2944 + }, + { + "epoch": 0.3114424703891709, + "grad_norm": 8.0, + "learning_rate": 4.986353028997869e-06, + "loss": 2.693, + "step": 2945 + }, + { + "epoch": 0.3115482233502538, + "grad_norm": 7.6875, + "learning_rate": 4.986335153516145e-06, + "loss": 2.493, + "step": 2946 + }, + { + "epoch": 0.31165397631133673, + "grad_norm": 7.75, + "learning_rate": 4.986317266367068e-06, + "loss": 2.7879, + "step": 2947 + }, + { + "epoch": 0.31175972927241963, + "grad_norm": 8.1875, + "learning_rate": 4.986299367550721e-06, + "loss": 2.3209, + "step": 2948 + }, + { + "epoch": 0.3118654822335025, + "grad_norm": 8.3125, + "learning_rate": 4.986281457067188e-06, + "loss": 2.4346, + "step": 2949 + }, + { + "epoch": 0.3119712351945855, + "grad_norm": 7.90625, + "learning_rate": 4.986263534916552e-06, + "loss": 2.3932, + "step": 2950 + }, + { + "epoch": 0.31207698815566837, + "grad_norm": 7.375, + "learning_rate": 4.986245601098899e-06, + "loss": 2.6393, + "step": 2951 + }, + { + "epoch": 0.31218274111675126, + "grad_norm": 7.5, + "learning_rate": 4.986227655614312e-06, + "loss": 2.4981, + "step": 2952 + }, + { + "epoch": 0.31228849407783416, + "grad_norm": 7.625, + "learning_rate": 4.9862096984628764e-06, + "loss": 2.5561, + "step": 2953 + }, + { + "epoch": 0.3123942470389171, + "grad_norm": 8.5, + "learning_rate": 4.986191729644674e-06, + "loss": 2.5157, + "step": 2954 + }, + { + "epoch": 0.3125, + "grad_norm": 7.59375, + "learning_rate": 4.986173749159792e-06, + "loss": 2.6812, + "step": 2955 + }, + { + "epoch": 0.3126057529610829, + "grad_norm": 8.125, + "learning_rate": 4.986155757008312e-06, + "loss": 2.5939, + "step": 2956 + }, + { + "epoch": 0.31271150592216584, + "grad_norm": 7.59375, + "learning_rate": 4.98613775319032e-06, + "loss": 2.6942, + "step": 2957 + }, + { + "epoch": 0.31281725888324874, + "grad_norm": 8.9375, + "learning_rate": 4.9861197377059015e-06, + "loss": 2.8416, + "step": 2958 + }, + { + "epoch": 0.31292301184433163, + "grad_norm": 7.8125, + "learning_rate": 4.986101710555139e-06, + "loss": 2.692, + "step": 2959 + }, + { + "epoch": 0.3130287648054145, + "grad_norm": 7.84375, + "learning_rate": 4.986083671738117e-06, + "loss": 2.4606, + "step": 2960 + }, + { + "epoch": 0.3131345177664975, + "grad_norm": 7.625, + "learning_rate": 4.986065621254923e-06, + "loss": 2.6316, + "step": 2961 + }, + { + "epoch": 0.31324027072758037, + "grad_norm": 7.78125, + "learning_rate": 4.986047559105639e-06, + "loss": 2.6971, + "step": 2962 + }, + { + "epoch": 0.31334602368866327, + "grad_norm": 7.875, + "learning_rate": 4.9860294852903505e-06, + "loss": 2.7475, + "step": 2963 + }, + { + "epoch": 0.3134517766497462, + "grad_norm": 7.6875, + "learning_rate": 4.986011399809142e-06, + "loss": 2.7325, + "step": 2964 + }, + { + "epoch": 0.3135575296108291, + "grad_norm": 7.5625, + "learning_rate": 4.9859933026621e-06, + "loss": 2.6502, + "step": 2965 + }, + { + "epoch": 0.313663282571912, + "grad_norm": 8.125, + "learning_rate": 4.9859751938493076e-06, + "loss": 2.8497, + "step": 2966 + }, + { + "epoch": 0.3137690355329949, + "grad_norm": 8.125, + "learning_rate": 4.9859570733708495e-06, + "loss": 2.6183, + "step": 2967 + }, + { + "epoch": 0.31387478849407785, + "grad_norm": 8.3125, + "learning_rate": 4.985938941226812e-06, + "loss": 2.5191, + "step": 2968 + }, + { + "epoch": 0.31398054145516074, + "grad_norm": 8.0, + "learning_rate": 4.98592079741728e-06, + "loss": 2.6625, + "step": 2969 + }, + { + "epoch": 0.31408629441624364, + "grad_norm": 7.96875, + "learning_rate": 4.985902641942338e-06, + "loss": 2.5889, + "step": 2970 + }, + { + "epoch": 0.3141920473773266, + "grad_norm": 8.5625, + "learning_rate": 4.985884474802073e-06, + "loss": 2.7863, + "step": 2971 + }, + { + "epoch": 0.3142978003384095, + "grad_norm": 7.625, + "learning_rate": 4.985866295996567e-06, + "loss": 2.6038, + "step": 2972 + }, + { + "epoch": 0.3144035532994924, + "grad_norm": 8.25, + "learning_rate": 4.985848105525908e-06, + "loss": 2.4594, + "step": 2973 + }, + { + "epoch": 0.31450930626057527, + "grad_norm": 7.9375, + "learning_rate": 4.98582990339018e-06, + "loss": 2.5021, + "step": 2974 + }, + { + "epoch": 0.3146150592216582, + "grad_norm": 7.8125, + "learning_rate": 4.98581168958947e-06, + "loss": 2.5839, + "step": 2975 + }, + { + "epoch": 0.3147208121827411, + "grad_norm": 7.5625, + "learning_rate": 4.985793464123861e-06, + "loss": 2.4923, + "step": 2976 + }, + { + "epoch": 0.314826565143824, + "grad_norm": 8.25, + "learning_rate": 4.985775226993441e-06, + "loss": 2.697, + "step": 2977 + }, + { + "epoch": 0.31493231810490696, + "grad_norm": 8.75, + "learning_rate": 4.985756978198295e-06, + "loss": 2.8227, + "step": 2978 + }, + { + "epoch": 0.31503807106598986, + "grad_norm": 8.4375, + "learning_rate": 4.985738717738507e-06, + "loss": 2.4795, + "step": 2979 + }, + { + "epoch": 0.31514382402707275, + "grad_norm": 8.4375, + "learning_rate": 4.985720445614164e-06, + "loss": 2.2002, + "step": 2980 + }, + { + "epoch": 0.31524957698815564, + "grad_norm": 8.25, + "learning_rate": 4.985702161825352e-06, + "loss": 2.5768, + "step": 2981 + }, + { + "epoch": 0.3153553299492386, + "grad_norm": 7.71875, + "learning_rate": 4.985683866372156e-06, + "loss": 2.49, + "step": 2982 + }, + { + "epoch": 0.3154610829103215, + "grad_norm": 7.375, + "learning_rate": 4.985665559254662e-06, + "loss": 2.6047, + "step": 2983 + }, + { + "epoch": 0.3155668358714044, + "grad_norm": 7.9375, + "learning_rate": 4.985647240472957e-06, + "loss": 2.7663, + "step": 2984 + }, + { + "epoch": 0.31567258883248733, + "grad_norm": 8.3125, + "learning_rate": 4.985628910027126e-06, + "loss": 2.5001, + "step": 2985 + }, + { + "epoch": 0.3157783417935702, + "grad_norm": 7.59375, + "learning_rate": 4.985610567917254e-06, + "loss": 2.6956, + "step": 2986 + }, + { + "epoch": 0.3158840947546531, + "grad_norm": 7.6875, + "learning_rate": 4.985592214143429e-06, + "loss": 2.3756, + "step": 2987 + }, + { + "epoch": 0.315989847715736, + "grad_norm": 8.375, + "learning_rate": 4.985573848705736e-06, + "loss": 2.4644, + "step": 2988 + }, + { + "epoch": 0.31609560067681897, + "grad_norm": 7.5625, + "learning_rate": 4.985555471604262e-06, + "loss": 2.6528, + "step": 2989 + }, + { + "epoch": 0.31620135363790186, + "grad_norm": 7.65625, + "learning_rate": 4.9855370828390924e-06, + "loss": 2.6412, + "step": 2990 + }, + { + "epoch": 0.31630710659898476, + "grad_norm": 8.1875, + "learning_rate": 4.985518682410314e-06, + "loss": 2.263, + "step": 2991 + }, + { + "epoch": 0.3164128595600677, + "grad_norm": 8.375, + "learning_rate": 4.985500270318013e-06, + "loss": 2.6204, + "step": 2992 + }, + { + "epoch": 0.3165186125211506, + "grad_norm": 7.90625, + "learning_rate": 4.985481846562275e-06, + "loss": 2.6586, + "step": 2993 + }, + { + "epoch": 0.3166243654822335, + "grad_norm": 8.0625, + "learning_rate": 4.985463411143188e-06, + "loss": 2.6672, + "step": 2994 + }, + { + "epoch": 0.3167301184433164, + "grad_norm": 8.4375, + "learning_rate": 4.985444964060838e-06, + "loss": 2.6412, + "step": 2995 + }, + { + "epoch": 0.31683587140439934, + "grad_norm": 8.625, + "learning_rate": 4.98542650531531e-06, + "loss": 2.5799, + "step": 2996 + }, + { + "epoch": 0.31694162436548223, + "grad_norm": 8.0625, + "learning_rate": 4.985408034906694e-06, + "loss": 2.8249, + "step": 2997 + }, + { + "epoch": 0.3170473773265651, + "grad_norm": 7.75, + "learning_rate": 4.985389552835073e-06, + "loss": 2.4449, + "step": 2998 + }, + { + "epoch": 0.3171531302876481, + "grad_norm": 8.125, + "learning_rate": 4.985371059100536e-06, + "loss": 2.7694, + "step": 2999 + }, + { + "epoch": 0.31725888324873097, + "grad_norm": 8.1875, + "learning_rate": 4.985352553703168e-06, + "loss": 2.6443, + "step": 3000 + }, + { + "epoch": 0.31736463620981387, + "grad_norm": 8.875, + "learning_rate": 4.985334036643058e-06, + "loss": 2.6677, + "step": 3001 + }, + { + "epoch": 0.31747038917089676, + "grad_norm": 7.59375, + "learning_rate": 4.985315507920291e-06, + "loss": 2.7182, + "step": 3002 + }, + { + "epoch": 0.3175761421319797, + "grad_norm": 8.0625, + "learning_rate": 4.9852969675349556e-06, + "loss": 2.211, + "step": 3003 + }, + { + "epoch": 0.3176818950930626, + "grad_norm": 7.84375, + "learning_rate": 4.985278415487138e-06, + "loss": 2.5979, + "step": 3004 + }, + { + "epoch": 0.3177876480541455, + "grad_norm": 7.71875, + "learning_rate": 4.9852598517769245e-06, + "loss": 2.3659, + "step": 3005 + }, + { + "epoch": 0.31789340101522845, + "grad_norm": 7.84375, + "learning_rate": 4.9852412764044035e-06, + "loss": 2.7286, + "step": 3006 + }, + { + "epoch": 0.31799915397631134, + "grad_norm": 7.75, + "learning_rate": 4.985222689369661e-06, + "loss": 2.7527, + "step": 3007 + }, + { + "epoch": 0.31810490693739424, + "grad_norm": 8.125, + "learning_rate": 4.985204090672786e-06, + "loss": 2.7461, + "step": 3008 + }, + { + "epoch": 0.31821065989847713, + "grad_norm": 8.0625, + "learning_rate": 4.985185480313863e-06, + "loss": 2.5035, + "step": 3009 + }, + { + "epoch": 0.3183164128595601, + "grad_norm": 8.375, + "learning_rate": 4.985166858292983e-06, + "loss": 2.7261, + "step": 3010 + }, + { + "epoch": 0.318422165820643, + "grad_norm": 7.625, + "learning_rate": 4.98514822461023e-06, + "loss": 2.6631, + "step": 3011 + }, + { + "epoch": 0.31852791878172587, + "grad_norm": 7.375, + "learning_rate": 4.985129579265693e-06, + "loss": 2.6169, + "step": 3012 + }, + { + "epoch": 0.3186336717428088, + "grad_norm": 9.875, + "learning_rate": 4.985110922259459e-06, + "loss": 2.9255, + "step": 3013 + }, + { + "epoch": 0.3187394247038917, + "grad_norm": 7.84375, + "learning_rate": 4.985092253591616e-06, + "loss": 2.4789, + "step": 3014 + }, + { + "epoch": 0.3188451776649746, + "grad_norm": 8.5, + "learning_rate": 4.985073573262251e-06, + "loss": 2.4751, + "step": 3015 + }, + { + "epoch": 0.3189509306260575, + "grad_norm": 8.25, + "learning_rate": 4.985054881271453e-06, + "loss": 2.564, + "step": 3016 + }, + { + "epoch": 0.31905668358714045, + "grad_norm": 8.4375, + "learning_rate": 4.985036177619309e-06, + "loss": 2.6647, + "step": 3017 + }, + { + "epoch": 0.31916243654822335, + "grad_norm": 7.53125, + "learning_rate": 4.985017462305906e-06, + "loss": 2.6038, + "step": 3018 + }, + { + "epoch": 0.31926818950930624, + "grad_norm": 8.4375, + "learning_rate": 4.984998735331332e-06, + "loss": 3.012, + "step": 3019 + }, + { + "epoch": 0.3193739424703892, + "grad_norm": 8.1875, + "learning_rate": 4.984979996695676e-06, + "loss": 2.718, + "step": 3020 + }, + { + "epoch": 0.3194796954314721, + "grad_norm": 7.5625, + "learning_rate": 4.9849612463990245e-06, + "loss": 2.6879, + "step": 3021 + }, + { + "epoch": 0.319585448392555, + "grad_norm": 8.0, + "learning_rate": 4.984942484441466e-06, + "loss": 2.5498, + "step": 3022 + }, + { + "epoch": 0.3196912013536379, + "grad_norm": 9.4375, + "learning_rate": 4.98492371082309e-06, + "loss": 2.5811, + "step": 3023 + }, + { + "epoch": 0.3197969543147208, + "grad_norm": 7.875, + "learning_rate": 4.984904925543982e-06, + "loss": 2.629, + "step": 3024 + }, + { + "epoch": 0.3199027072758037, + "grad_norm": 7.5, + "learning_rate": 4.984886128604232e-06, + "loss": 2.5391, + "step": 3025 + }, + { + "epoch": 0.3200084602368866, + "grad_norm": 8.4375, + "learning_rate": 4.984867320003928e-06, + "loss": 2.6093, + "step": 3026 + }, + { + "epoch": 0.32011421319796957, + "grad_norm": 7.5625, + "learning_rate": 4.984848499743157e-06, + "loss": 2.7227, + "step": 3027 + }, + { + "epoch": 0.32021996615905246, + "grad_norm": 7.1875, + "learning_rate": 4.984829667822009e-06, + "loss": 2.4842, + "step": 3028 + }, + { + "epoch": 0.32032571912013535, + "grad_norm": 7.625, + "learning_rate": 4.984810824240572e-06, + "loss": 2.489, + "step": 3029 + }, + { + "epoch": 0.32043147208121825, + "grad_norm": 8.0, + "learning_rate": 4.984791968998933e-06, + "loss": 2.5727, + "step": 3030 + }, + { + "epoch": 0.3205372250423012, + "grad_norm": 8.375, + "learning_rate": 4.984773102097182e-06, + "loss": 2.9827, + "step": 3031 + }, + { + "epoch": 0.3206429780033841, + "grad_norm": 8.3125, + "learning_rate": 4.984754223535408e-06, + "loss": 2.654, + "step": 3032 + }, + { + "epoch": 0.320748730964467, + "grad_norm": 7.6875, + "learning_rate": 4.984735333313697e-06, + "loss": 2.7104, + "step": 3033 + }, + { + "epoch": 0.32085448392554994, + "grad_norm": 9.5, + "learning_rate": 4.984716431432141e-06, + "loss": 2.7213, + "step": 3034 + }, + { + "epoch": 0.32096023688663283, + "grad_norm": 7.5, + "learning_rate": 4.9846975178908255e-06, + "loss": 2.8692, + "step": 3035 + }, + { + "epoch": 0.3210659898477157, + "grad_norm": 8.8125, + "learning_rate": 4.984678592689842e-06, + "loss": 2.5646, + "step": 3036 + }, + { + "epoch": 0.3211717428087986, + "grad_norm": 7.8125, + "learning_rate": 4.984659655829277e-06, + "loss": 2.346, + "step": 3037 + }, + { + "epoch": 0.32127749576988157, + "grad_norm": 9.1875, + "learning_rate": 4.984640707309221e-06, + "loss": 2.6819, + "step": 3038 + }, + { + "epoch": 0.32138324873096447, + "grad_norm": 7.625, + "learning_rate": 4.984621747129762e-06, + "loss": 2.4333, + "step": 3039 + }, + { + "epoch": 0.32148900169204736, + "grad_norm": 7.09375, + "learning_rate": 4.98460277529099e-06, + "loss": 2.4951, + "step": 3040 + }, + { + "epoch": 0.3215947546531303, + "grad_norm": 7.3125, + "learning_rate": 4.984583791792993e-06, + "loss": 2.3685, + "step": 3041 + }, + { + "epoch": 0.3217005076142132, + "grad_norm": 7.5625, + "learning_rate": 4.98456479663586e-06, + "loss": 2.7578, + "step": 3042 + }, + { + "epoch": 0.3218062605752961, + "grad_norm": 7.25, + "learning_rate": 4.984545789819682e-06, + "loss": 2.569, + "step": 3043 + }, + { + "epoch": 0.321912013536379, + "grad_norm": 8.125, + "learning_rate": 4.984526771344544e-06, + "loss": 2.6965, + "step": 3044 + }, + { + "epoch": 0.32201776649746194, + "grad_norm": 8.3125, + "learning_rate": 4.9845077412105405e-06, + "loss": 2.6076, + "step": 3045 + }, + { + "epoch": 0.32212351945854484, + "grad_norm": 7.875, + "learning_rate": 4.9844886994177576e-06, + "loss": 2.6898, + "step": 3046 + }, + { + "epoch": 0.32222927241962773, + "grad_norm": 9.0625, + "learning_rate": 4.984469645966284e-06, + "loss": 2.4043, + "step": 3047 + }, + { + "epoch": 0.3223350253807107, + "grad_norm": 7.5, + "learning_rate": 4.984450580856212e-06, + "loss": 2.5702, + "step": 3048 + }, + { + "epoch": 0.3224407783417936, + "grad_norm": 7.6875, + "learning_rate": 4.98443150408763e-06, + "loss": 2.7059, + "step": 3049 + }, + { + "epoch": 0.32254653130287647, + "grad_norm": 7.8125, + "learning_rate": 4.984412415660626e-06, + "loss": 2.4714, + "step": 3050 + }, + { + "epoch": 0.32265228426395937, + "grad_norm": 7.40625, + "learning_rate": 4.984393315575291e-06, + "loss": 2.5132, + "step": 3051 + }, + { + "epoch": 0.3227580372250423, + "grad_norm": 8.4375, + "learning_rate": 4.984374203831714e-06, + "loss": 2.7877, + "step": 3052 + }, + { + "epoch": 0.3228637901861252, + "grad_norm": 7.3125, + "learning_rate": 4.984355080429985e-06, + "loss": 2.5977, + "step": 3053 + }, + { + "epoch": 0.3229695431472081, + "grad_norm": 12.3125, + "learning_rate": 4.984335945370193e-06, + "loss": 2.5085, + "step": 3054 + }, + { + "epoch": 0.32307529610829105, + "grad_norm": 7.875, + "learning_rate": 4.98431679865243e-06, + "loss": 2.4986, + "step": 3055 + }, + { + "epoch": 0.32318104906937395, + "grad_norm": 7.8125, + "learning_rate": 4.984297640276783e-06, + "loss": 2.601, + "step": 3056 + }, + { + "epoch": 0.32328680203045684, + "grad_norm": 7.6875, + "learning_rate": 4.984278470243345e-06, + "loss": 2.7312, + "step": 3057 + }, + { + "epoch": 0.32339255499153974, + "grad_norm": 7.625, + "learning_rate": 4.984259288552202e-06, + "loss": 2.72, + "step": 3058 + }, + { + "epoch": 0.3234983079526227, + "grad_norm": 7.21875, + "learning_rate": 4.984240095203447e-06, + "loss": 2.5826, + "step": 3059 + }, + { + "epoch": 0.3236040609137056, + "grad_norm": 7.6875, + "learning_rate": 4.98422089019717e-06, + "loss": 2.6163, + "step": 3060 + }, + { + "epoch": 0.3237098138747885, + "grad_norm": 7.78125, + "learning_rate": 4.984201673533459e-06, + "loss": 2.5727, + "step": 3061 + }, + { + "epoch": 0.3238155668358714, + "grad_norm": 7.6875, + "learning_rate": 4.9841824452124065e-06, + "loss": 2.8738, + "step": 3062 + }, + { + "epoch": 0.3239213197969543, + "grad_norm": 7.5, + "learning_rate": 4.984163205234101e-06, + "loss": 2.5273, + "step": 3063 + }, + { + "epoch": 0.3240270727580372, + "grad_norm": 8.0, + "learning_rate": 4.9841439535986345e-06, + "loss": 2.8124, + "step": 3064 + }, + { + "epoch": 0.3241328257191201, + "grad_norm": 7.75, + "learning_rate": 4.9841246903060955e-06, + "loss": 2.642, + "step": 3065 + }, + { + "epoch": 0.32423857868020306, + "grad_norm": 8.0625, + "learning_rate": 4.984105415356576e-06, + "loss": 2.8805, + "step": 3066 + }, + { + "epoch": 0.32434433164128595, + "grad_norm": 8.5, + "learning_rate": 4.984086128750166e-06, + "loss": 2.7036, + "step": 3067 + }, + { + "epoch": 0.32445008460236885, + "grad_norm": 7.96875, + "learning_rate": 4.9840668304869554e-06, + "loss": 2.6657, + "step": 3068 + }, + { + "epoch": 0.3245558375634518, + "grad_norm": 14.6875, + "learning_rate": 4.984047520567035e-06, + "loss": 2.3492, + "step": 3069 + }, + { + "epoch": 0.3246615905245347, + "grad_norm": 8.5625, + "learning_rate": 4.984028198990495e-06, + "loss": 2.5903, + "step": 3070 + }, + { + "epoch": 0.3247673434856176, + "grad_norm": 7.96875, + "learning_rate": 4.984008865757427e-06, + "loss": 2.4928, + "step": 3071 + }, + { + "epoch": 0.3248730964467005, + "grad_norm": 7.5625, + "learning_rate": 4.983989520867923e-06, + "loss": 2.7497, + "step": 3072 + }, + { + "epoch": 0.32497884940778343, + "grad_norm": 7.875, + "learning_rate": 4.98397016432207e-06, + "loss": 2.6031, + "step": 3073 + }, + { + "epoch": 0.3250846023688663, + "grad_norm": 7.4375, + "learning_rate": 4.983950796119961e-06, + "loss": 2.5995, + "step": 3074 + }, + { + "epoch": 0.3251903553299492, + "grad_norm": 8.375, + "learning_rate": 4.983931416261687e-06, + "loss": 2.5231, + "step": 3075 + }, + { + "epoch": 0.32529610829103217, + "grad_norm": 7.25, + "learning_rate": 4.983912024747339e-06, + "loss": 2.4606, + "step": 3076 + }, + { + "epoch": 0.32540186125211507, + "grad_norm": 8.0625, + "learning_rate": 4.983892621577008e-06, + "loss": 2.6726, + "step": 3077 + }, + { + "epoch": 0.32550761421319796, + "grad_norm": 7.96875, + "learning_rate": 4.983873206750785e-06, + "loss": 2.8648, + "step": 3078 + }, + { + "epoch": 0.32561336717428085, + "grad_norm": 7.65625, + "learning_rate": 4.9838537802687605e-06, + "loss": 2.5695, + "step": 3079 + }, + { + "epoch": 0.3257191201353638, + "grad_norm": 7.375, + "learning_rate": 4.983834342131026e-06, + "loss": 2.6862, + "step": 3080 + }, + { + "epoch": 0.3258248730964467, + "grad_norm": 7.5625, + "learning_rate": 4.983814892337673e-06, + "loss": 2.544, + "step": 3081 + }, + { + "epoch": 0.3259306260575296, + "grad_norm": 7.65625, + "learning_rate": 4.983795430888793e-06, + "loss": 2.5578, + "step": 3082 + }, + { + "epoch": 0.32603637901861254, + "grad_norm": 8.1875, + "learning_rate": 4.983775957784476e-06, + "loss": 2.655, + "step": 3083 + }, + { + "epoch": 0.32614213197969544, + "grad_norm": 8.5625, + "learning_rate": 4.983756473024814e-06, + "loss": 2.5093, + "step": 3084 + }, + { + "epoch": 0.32624788494077833, + "grad_norm": 7.84375, + "learning_rate": 4.9837369766099e-06, + "loss": 2.8218, + "step": 3085 + }, + { + "epoch": 0.3263536379018612, + "grad_norm": 7.46875, + "learning_rate": 4.983717468539822e-06, + "loss": 2.6876, + "step": 3086 + }, + { + "epoch": 0.3264593908629442, + "grad_norm": 7.875, + "learning_rate": 4.983697948814676e-06, + "loss": 2.7159, + "step": 3087 + }, + { + "epoch": 0.32656514382402707, + "grad_norm": 8.4375, + "learning_rate": 4.983678417434551e-06, + "loss": 2.4572, + "step": 3088 + }, + { + "epoch": 0.32667089678510997, + "grad_norm": 8.25, + "learning_rate": 4.983658874399539e-06, + "loss": 2.5392, + "step": 3089 + }, + { + "epoch": 0.3267766497461929, + "grad_norm": 7.4375, + "learning_rate": 4.983639319709731e-06, + "loss": 2.6312, + "step": 3090 + }, + { + "epoch": 0.3268824027072758, + "grad_norm": 8.8125, + "learning_rate": 4.983619753365219e-06, + "loss": 2.7939, + "step": 3091 + }, + { + "epoch": 0.3269881556683587, + "grad_norm": 8.1875, + "learning_rate": 4.983600175366096e-06, + "loss": 2.8416, + "step": 3092 + }, + { + "epoch": 0.3270939086294416, + "grad_norm": 7.75, + "learning_rate": 4.983580585712453e-06, + "loss": 2.5916, + "step": 3093 + }, + { + "epoch": 0.32719966159052455, + "grad_norm": 8.375, + "learning_rate": 4.983560984404382e-06, + "loss": 2.6868, + "step": 3094 + }, + { + "epoch": 0.32730541455160744, + "grad_norm": 8.6875, + "learning_rate": 4.983541371441975e-06, + "loss": 2.9008, + "step": 3095 + }, + { + "epoch": 0.32741116751269034, + "grad_norm": 7.96875, + "learning_rate": 4.983521746825324e-06, + "loss": 2.4233, + "step": 3096 + }, + { + "epoch": 0.3275169204737733, + "grad_norm": 8.0625, + "learning_rate": 4.983502110554522e-06, + "loss": 2.648, + "step": 3097 + }, + { + "epoch": 0.3276226734348562, + "grad_norm": 7.84375, + "learning_rate": 4.98348246262966e-06, + "loss": 2.6805, + "step": 3098 + }, + { + "epoch": 0.3277284263959391, + "grad_norm": 7.53125, + "learning_rate": 4.983462803050829e-06, + "loss": 2.4649, + "step": 3099 + }, + { + "epoch": 0.32783417935702197, + "grad_norm": 7.5625, + "learning_rate": 4.983443131818124e-06, + "loss": 2.5782, + "step": 3100 + }, + { + "epoch": 0.3279399323181049, + "grad_norm": 8.0625, + "learning_rate": 4.983423448931637e-06, + "loss": 2.7564, + "step": 3101 + }, + { + "epoch": 0.3280456852791878, + "grad_norm": 7.5625, + "learning_rate": 4.983403754391458e-06, + "loss": 2.6395, + "step": 3102 + }, + { + "epoch": 0.3281514382402707, + "grad_norm": 8.4375, + "learning_rate": 4.9833840481976816e-06, + "loss": 2.6395, + "step": 3103 + }, + { + "epoch": 0.32825719120135366, + "grad_norm": 7.46875, + "learning_rate": 4.9833643303503985e-06, + "loss": 2.6188, + "step": 3104 + }, + { + "epoch": 0.32836294416243655, + "grad_norm": 7.84375, + "learning_rate": 4.983344600849703e-06, + "loss": 2.5689, + "step": 3105 + }, + { + "epoch": 0.32846869712351945, + "grad_norm": 8.5, + "learning_rate": 4.9833248596956875e-06, + "loss": 2.6606, + "step": 3106 + }, + { + "epoch": 0.32857445008460234, + "grad_norm": 8.125, + "learning_rate": 4.9833051068884434e-06, + "loss": 2.5396, + "step": 3107 + }, + { + "epoch": 0.3286802030456853, + "grad_norm": 7.875, + "learning_rate": 4.983285342428064e-06, + "loss": 2.5881, + "step": 3108 + }, + { + "epoch": 0.3287859560067682, + "grad_norm": 8.1875, + "learning_rate": 4.983265566314642e-06, + "loss": 2.7548, + "step": 3109 + }, + { + "epoch": 0.3288917089678511, + "grad_norm": 7.8125, + "learning_rate": 4.9832457785482715e-06, + "loss": 2.5392, + "step": 3110 + }, + { + "epoch": 0.32899746192893403, + "grad_norm": 7.84375, + "learning_rate": 4.983225979129042e-06, + "loss": 2.6145, + "step": 3111 + }, + { + "epoch": 0.3291032148900169, + "grad_norm": 7.5, + "learning_rate": 4.9832061680570506e-06, + "loss": 2.5418, + "step": 3112 + }, + { + "epoch": 0.3292089678510998, + "grad_norm": 7.375, + "learning_rate": 4.983186345332388e-06, + "loss": 2.6392, + "step": 3113 + }, + { + "epoch": 0.3293147208121827, + "grad_norm": 7.40625, + "learning_rate": 4.983166510955147e-06, + "loss": 2.5607, + "step": 3114 + }, + { + "epoch": 0.32942047377326567, + "grad_norm": 7.34375, + "learning_rate": 4.983146664925421e-06, + "loss": 2.4411, + "step": 3115 + }, + { + "epoch": 0.32952622673434856, + "grad_norm": 8.25, + "learning_rate": 4.9831268072433036e-06, + "loss": 2.587, + "step": 3116 + }, + { + "epoch": 0.32963197969543145, + "grad_norm": 8.0, + "learning_rate": 4.983106937908887e-06, + "loss": 2.7822, + "step": 3117 + }, + { + "epoch": 0.3297377326565144, + "grad_norm": 8.0, + "learning_rate": 4.983087056922266e-06, + "loss": 2.7678, + "step": 3118 + }, + { + "epoch": 0.3298434856175973, + "grad_norm": 7.15625, + "learning_rate": 4.983067164283533e-06, + "loss": 2.5368, + "step": 3119 + }, + { + "epoch": 0.3299492385786802, + "grad_norm": 7.9375, + "learning_rate": 4.98304725999278e-06, + "loss": 2.642, + "step": 3120 + }, + { + "epoch": 0.3300549915397631, + "grad_norm": 7.46875, + "learning_rate": 4.983027344050103e-06, + "loss": 2.5826, + "step": 3121 + }, + { + "epoch": 0.33016074450084604, + "grad_norm": 8.5625, + "learning_rate": 4.983007416455594e-06, + "loss": 2.9239, + "step": 3122 + }, + { + "epoch": 0.33026649746192893, + "grad_norm": 8.875, + "learning_rate": 4.9829874772093465e-06, + "loss": 2.592, + "step": 3123 + }, + { + "epoch": 0.3303722504230118, + "grad_norm": 7.53125, + "learning_rate": 4.982967526311454e-06, + "loss": 2.4313, + "step": 3124 + }, + { + "epoch": 0.3304780033840948, + "grad_norm": 8.0, + "learning_rate": 4.982947563762011e-06, + "loss": 2.6658, + "step": 3125 + }, + { + "epoch": 0.33058375634517767, + "grad_norm": 8.6875, + "learning_rate": 4.98292758956111e-06, + "loss": 2.7539, + "step": 3126 + }, + { + "epoch": 0.33068950930626057, + "grad_norm": 7.6875, + "learning_rate": 4.982907603708846e-06, + "loss": 2.5901, + "step": 3127 + }, + { + "epoch": 0.33079526226734346, + "grad_norm": 8.3125, + "learning_rate": 4.982887606205312e-06, + "loss": 2.7246, + "step": 3128 + }, + { + "epoch": 0.3309010152284264, + "grad_norm": 8.0, + "learning_rate": 4.982867597050601e-06, + "loss": 2.5335, + "step": 3129 + }, + { + "epoch": 0.3310067681895093, + "grad_norm": 7.75, + "learning_rate": 4.982847576244809e-06, + "loss": 2.6042, + "step": 3130 + }, + { + "epoch": 0.3311125211505922, + "grad_norm": 8.25, + "learning_rate": 4.982827543788028e-06, + "loss": 2.6436, + "step": 3131 + }, + { + "epoch": 0.33121827411167515, + "grad_norm": 8.3125, + "learning_rate": 4.982807499680353e-06, + "loss": 2.6166, + "step": 3132 + }, + { + "epoch": 0.33132402707275804, + "grad_norm": 7.625, + "learning_rate": 4.982787443921878e-06, + "loss": 2.4373, + "step": 3133 + }, + { + "epoch": 0.33142978003384094, + "grad_norm": 7.375, + "learning_rate": 4.982767376512697e-06, + "loss": 2.5947, + "step": 3134 + }, + { + "epoch": 0.33153553299492383, + "grad_norm": 7.8125, + "learning_rate": 4.982747297452904e-06, + "loss": 2.6674, + "step": 3135 + }, + { + "epoch": 0.3316412859560068, + "grad_norm": 8.125, + "learning_rate": 4.982727206742593e-06, + "loss": 2.7547, + "step": 3136 + }, + { + "epoch": 0.3317470389170897, + "grad_norm": 7.96875, + "learning_rate": 4.982707104381859e-06, + "loss": 2.4929, + "step": 3137 + }, + { + "epoch": 0.33185279187817257, + "grad_norm": 7.8125, + "learning_rate": 4.982686990370796e-06, + "loss": 2.4606, + "step": 3138 + }, + { + "epoch": 0.3319585448392555, + "grad_norm": 7.625, + "learning_rate": 4.9826668647094975e-06, + "loss": 2.6773, + "step": 3139 + }, + { + "epoch": 0.3320642978003384, + "grad_norm": 7.5, + "learning_rate": 4.98264672739806e-06, + "loss": 2.4858, + "step": 3140 + }, + { + "epoch": 0.3321700507614213, + "grad_norm": 8.125, + "learning_rate": 4.982626578436576e-06, + "loss": 2.4874, + "step": 3141 + }, + { + "epoch": 0.3322758037225042, + "grad_norm": 7.625, + "learning_rate": 4.982606417825141e-06, + "loss": 2.6637, + "step": 3142 + }, + { + "epoch": 0.33238155668358715, + "grad_norm": 7.59375, + "learning_rate": 4.98258624556385e-06, + "loss": 2.7264, + "step": 3143 + }, + { + "epoch": 0.33248730964467005, + "grad_norm": 7.34375, + "learning_rate": 4.982566061652796e-06, + "loss": 2.4316, + "step": 3144 + }, + { + "epoch": 0.33259306260575294, + "grad_norm": 8.4375, + "learning_rate": 4.982545866092075e-06, + "loss": 2.736, + "step": 3145 + }, + { + "epoch": 0.3326988155668359, + "grad_norm": 8.125, + "learning_rate": 4.982525658881783e-06, + "loss": 2.7561, + "step": 3146 + }, + { + "epoch": 0.3328045685279188, + "grad_norm": 8.375, + "learning_rate": 4.982505440022012e-06, + "loss": 2.568, + "step": 3147 + }, + { + "epoch": 0.3329103214890017, + "grad_norm": 7.40625, + "learning_rate": 4.982485209512859e-06, + "loss": 2.4925, + "step": 3148 + }, + { + "epoch": 0.3330160744500846, + "grad_norm": 7.90625, + "learning_rate": 4.982464967354418e-06, + "loss": 2.5932, + "step": 3149 + }, + { + "epoch": 0.3331218274111675, + "grad_norm": 8.0625, + "learning_rate": 4.982444713546783e-06, + "loss": 2.7326, + "step": 3150 + }, + { + "epoch": 0.3332275803722504, + "grad_norm": 8.125, + "learning_rate": 4.982424448090052e-06, + "loss": 2.8213, + "step": 3151 + }, + { + "epoch": 0.3333333333333333, + "grad_norm": 7.6875, + "learning_rate": 4.982404170984317e-06, + "loss": 2.6014, + "step": 3152 + }, + { + "epoch": 0.33343908629441626, + "grad_norm": 9.0625, + "learning_rate": 4.982383882229676e-06, + "loss": 2.7305, + "step": 3153 + }, + { + "epoch": 0.33354483925549916, + "grad_norm": 10.9375, + "learning_rate": 4.982363581826222e-06, + "loss": 2.7051, + "step": 3154 + }, + { + "epoch": 0.33365059221658205, + "grad_norm": 8.9375, + "learning_rate": 4.982343269774051e-06, + "loss": 2.6552, + "step": 3155 + }, + { + "epoch": 0.33375634517766495, + "grad_norm": 8.5, + "learning_rate": 4.982322946073258e-06, + "loss": 2.8067, + "step": 3156 + }, + { + "epoch": 0.3338620981387479, + "grad_norm": 7.3125, + "learning_rate": 4.982302610723938e-06, + "loss": 2.5281, + "step": 3157 + }, + { + "epoch": 0.3339678510998308, + "grad_norm": 8.125, + "learning_rate": 4.982282263726188e-06, + "loss": 2.7202, + "step": 3158 + }, + { + "epoch": 0.3340736040609137, + "grad_norm": 8.375, + "learning_rate": 4.982261905080102e-06, + "loss": 2.3968, + "step": 3159 + }, + { + "epoch": 0.33417935702199664, + "grad_norm": 7.78125, + "learning_rate": 4.982241534785777e-06, + "loss": 2.7467, + "step": 3160 + }, + { + "epoch": 0.33428510998307953, + "grad_norm": 8.0, + "learning_rate": 4.982221152843308e-06, + "loss": 2.7547, + "step": 3161 + }, + { + "epoch": 0.3343908629441624, + "grad_norm": 8.4375, + "learning_rate": 4.982200759252789e-06, + "loss": 2.6913, + "step": 3162 + }, + { + "epoch": 0.3344966159052453, + "grad_norm": 7.59375, + "learning_rate": 4.982180354014317e-06, + "loss": 2.5873, + "step": 3163 + }, + { + "epoch": 0.33460236886632827, + "grad_norm": 7.78125, + "learning_rate": 4.9821599371279885e-06, + "loss": 2.8872, + "step": 3164 + }, + { + "epoch": 0.33470812182741116, + "grad_norm": 7.78125, + "learning_rate": 4.982139508593899e-06, + "loss": 2.7754, + "step": 3165 + }, + { + "epoch": 0.33481387478849406, + "grad_norm": 8.0625, + "learning_rate": 4.982119068412142e-06, + "loss": 2.5619, + "step": 3166 + }, + { + "epoch": 0.334919627749577, + "grad_norm": 7.5625, + "learning_rate": 4.982098616582817e-06, + "loss": 2.7946, + "step": 3167 + }, + { + "epoch": 0.3350253807106599, + "grad_norm": 7.625, + "learning_rate": 4.982078153106018e-06, + "loss": 2.6169, + "step": 3168 + }, + { + "epoch": 0.3351311336717428, + "grad_norm": 7.5, + "learning_rate": 4.982057677981842e-06, + "loss": 2.5145, + "step": 3169 + }, + { + "epoch": 0.3352368866328257, + "grad_norm": 8.375, + "learning_rate": 4.982037191210384e-06, + "loss": 2.5139, + "step": 3170 + }, + { + "epoch": 0.33534263959390864, + "grad_norm": 9.25, + "learning_rate": 4.98201669279174e-06, + "loss": 2.5183, + "step": 3171 + }, + { + "epoch": 0.33544839255499154, + "grad_norm": 7.625, + "learning_rate": 4.981996182726007e-06, + "loss": 2.5436, + "step": 3172 + }, + { + "epoch": 0.33555414551607443, + "grad_norm": 8.0625, + "learning_rate": 4.981975661013281e-06, + "loss": 2.4588, + "step": 3173 + }, + { + "epoch": 0.3356598984771574, + "grad_norm": 7.65625, + "learning_rate": 4.981955127653659e-06, + "loss": 2.5129, + "step": 3174 + }, + { + "epoch": 0.3357656514382403, + "grad_norm": 8.25, + "learning_rate": 4.981934582647236e-06, + "loss": 2.499, + "step": 3175 + }, + { + "epoch": 0.33587140439932317, + "grad_norm": 7.75, + "learning_rate": 4.981914025994109e-06, + "loss": 2.6659, + "step": 3176 + }, + { + "epoch": 0.33597715736040606, + "grad_norm": 7.8125, + "learning_rate": 4.981893457694375e-06, + "loss": 2.6486, + "step": 3177 + }, + { + "epoch": 0.336082910321489, + "grad_norm": 7.71875, + "learning_rate": 4.9818728777481294e-06, + "loss": 2.514, + "step": 3178 + }, + { + "epoch": 0.3361886632825719, + "grad_norm": 9.0625, + "learning_rate": 4.98185228615547e-06, + "loss": 2.7855, + "step": 3179 + }, + { + "epoch": 0.3362944162436548, + "grad_norm": 7.71875, + "learning_rate": 4.981831682916493e-06, + "loss": 2.4288, + "step": 3180 + }, + { + "epoch": 0.33640016920473775, + "grad_norm": 7.8125, + "learning_rate": 4.981811068031295e-06, + "loss": 2.2106, + "step": 3181 + }, + { + "epoch": 0.33650592216582065, + "grad_norm": 7.6875, + "learning_rate": 4.981790441499972e-06, + "loss": 2.4737, + "step": 3182 + }, + { + "epoch": 0.33661167512690354, + "grad_norm": 7.96875, + "learning_rate": 4.981769803322622e-06, + "loss": 2.4387, + "step": 3183 + }, + { + "epoch": 0.33671742808798644, + "grad_norm": 7.78125, + "learning_rate": 4.981749153499341e-06, + "loss": 2.5555, + "step": 3184 + }, + { + "epoch": 0.3368231810490694, + "grad_norm": 9.9375, + "learning_rate": 4.981728492030227e-06, + "loss": 2.6484, + "step": 3185 + }, + { + "epoch": 0.3369289340101523, + "grad_norm": 7.875, + "learning_rate": 4.981707818915375e-06, + "loss": 2.754, + "step": 3186 + }, + { + "epoch": 0.3370346869712352, + "grad_norm": 8.5, + "learning_rate": 4.981687134154884e-06, + "loss": 2.5569, + "step": 3187 + }, + { + "epoch": 0.3371404399323181, + "grad_norm": 7.40625, + "learning_rate": 4.98166643774885e-06, + "loss": 2.3805, + "step": 3188 + }, + { + "epoch": 0.337246192893401, + "grad_norm": 8.1875, + "learning_rate": 4.98164572969737e-06, + "loss": 2.6247, + "step": 3189 + }, + { + "epoch": 0.3373519458544839, + "grad_norm": 7.8125, + "learning_rate": 4.981625010000542e-06, + "loss": 2.5806, + "step": 3190 + }, + { + "epoch": 0.3374576988155668, + "grad_norm": 8.0625, + "learning_rate": 4.981604278658463e-06, + "loss": 2.5549, + "step": 3191 + }, + { + "epoch": 0.33756345177664976, + "grad_norm": 7.96875, + "learning_rate": 4.9815835356712295e-06, + "loss": 2.5569, + "step": 3192 + }, + { + "epoch": 0.33766920473773265, + "grad_norm": 7.65625, + "learning_rate": 4.98156278103894e-06, + "loss": 2.6449, + "step": 3193 + }, + { + "epoch": 0.33777495769881555, + "grad_norm": 7.96875, + "learning_rate": 4.981542014761691e-06, + "loss": 2.6019, + "step": 3194 + }, + { + "epoch": 0.3378807106598985, + "grad_norm": 7.84375, + "learning_rate": 4.98152123683958e-06, + "loss": 2.5926, + "step": 3195 + }, + { + "epoch": 0.3379864636209814, + "grad_norm": 8.3125, + "learning_rate": 4.981500447272705e-06, + "loss": 2.6669, + "step": 3196 + }, + { + "epoch": 0.3380922165820643, + "grad_norm": 8.0625, + "learning_rate": 4.981479646061164e-06, + "loss": 2.8375, + "step": 3197 + }, + { + "epoch": 0.3381979695431472, + "grad_norm": 8.375, + "learning_rate": 4.981458833205052e-06, + "loss": 2.8255, + "step": 3198 + }, + { + "epoch": 0.33830372250423013, + "grad_norm": 8.75, + "learning_rate": 4.98143800870447e-06, + "loss": 2.68, + "step": 3199 + }, + { + "epoch": 0.338409475465313, + "grad_norm": 8.0625, + "learning_rate": 4.981417172559514e-06, + "loss": 2.6642, + "step": 3200 + }, + { + "epoch": 0.3385152284263959, + "grad_norm": 7.65625, + "learning_rate": 4.9813963247702825e-06, + "loss": 2.5127, + "step": 3201 + }, + { + "epoch": 0.33862098138747887, + "grad_norm": 7.40625, + "learning_rate": 4.981375465336872e-06, + "loss": 2.4608, + "step": 3202 + }, + { + "epoch": 0.33872673434856176, + "grad_norm": 8.1875, + "learning_rate": 4.981354594259382e-06, + "loss": 2.6085, + "step": 3203 + }, + { + "epoch": 0.33883248730964466, + "grad_norm": 7.40625, + "learning_rate": 4.981333711537909e-06, + "loss": 2.4631, + "step": 3204 + }, + { + "epoch": 0.33893824027072755, + "grad_norm": 7.8125, + "learning_rate": 4.981312817172554e-06, + "loss": 2.5903, + "step": 3205 + }, + { + "epoch": 0.3390439932318105, + "grad_norm": 7.9375, + "learning_rate": 4.98129191116341e-06, + "loss": 2.6401, + "step": 3206 + }, + { + "epoch": 0.3391497461928934, + "grad_norm": 7.65625, + "learning_rate": 4.981270993510579e-06, + "loss": 2.4577, + "step": 3207 + }, + { + "epoch": 0.3392554991539763, + "grad_norm": 7.875, + "learning_rate": 4.981250064214158e-06, + "loss": 2.4407, + "step": 3208 + }, + { + "epoch": 0.33936125211505924, + "grad_norm": 7.65625, + "learning_rate": 4.981229123274245e-06, + "loss": 2.448, + "step": 3209 + }, + { + "epoch": 0.33946700507614214, + "grad_norm": 8.375, + "learning_rate": 4.981208170690939e-06, + "loss": 2.7389, + "step": 3210 + }, + { + "epoch": 0.33957275803722503, + "grad_norm": 8.0625, + "learning_rate": 4.981187206464338e-06, + "loss": 2.5076, + "step": 3211 + }, + { + "epoch": 0.3396785109983079, + "grad_norm": 7.84375, + "learning_rate": 4.981166230594539e-06, + "loss": 2.6029, + "step": 3212 + }, + { + "epoch": 0.3397842639593909, + "grad_norm": 8.5, + "learning_rate": 4.981145243081642e-06, + "loss": 2.6863, + "step": 3213 + }, + { + "epoch": 0.33989001692047377, + "grad_norm": 7.78125, + "learning_rate": 4.981124243925746e-06, + "loss": 2.7079, + "step": 3214 + }, + { + "epoch": 0.33999576988155666, + "grad_norm": 8.5, + "learning_rate": 4.981103233126948e-06, + "loss": 2.9865, + "step": 3215 + }, + { + "epoch": 0.3401015228426396, + "grad_norm": 7.71875, + "learning_rate": 4.981082210685347e-06, + "loss": 2.6686, + "step": 3216 + }, + { + "epoch": 0.3402072758037225, + "grad_norm": 10.4375, + "learning_rate": 4.981061176601043e-06, + "loss": 2.7146, + "step": 3217 + }, + { + "epoch": 0.3403130287648054, + "grad_norm": 7.90625, + "learning_rate": 4.981040130874133e-06, + "loss": 2.6879, + "step": 3218 + }, + { + "epoch": 0.3404187817258883, + "grad_norm": 8.5, + "learning_rate": 4.981019073504716e-06, + "loss": 2.4008, + "step": 3219 + }, + { + "epoch": 0.34052453468697125, + "grad_norm": 7.8125, + "learning_rate": 4.980998004492892e-06, + "loss": 2.7824, + "step": 3220 + }, + { + "epoch": 0.34063028764805414, + "grad_norm": 7.96875, + "learning_rate": 4.980976923838759e-06, + "loss": 2.7631, + "step": 3221 + }, + { + "epoch": 0.34073604060913704, + "grad_norm": 7.21875, + "learning_rate": 4.980955831542415e-06, + "loss": 2.6635, + "step": 3222 + }, + { + "epoch": 0.34084179357022, + "grad_norm": 7.6875, + "learning_rate": 4.98093472760396e-06, + "loss": 2.7586, + "step": 3223 + }, + { + "epoch": 0.3409475465313029, + "grad_norm": 7.59375, + "learning_rate": 4.980913612023494e-06, + "loss": 2.5181, + "step": 3224 + }, + { + "epoch": 0.3410532994923858, + "grad_norm": 8.0625, + "learning_rate": 4.980892484801115e-06, + "loss": 2.5125, + "step": 3225 + }, + { + "epoch": 0.34115905245346867, + "grad_norm": 7.71875, + "learning_rate": 4.9808713459369215e-06, + "loss": 2.5838, + "step": 3226 + }, + { + "epoch": 0.3412648054145516, + "grad_norm": 8.125, + "learning_rate": 4.980850195431014e-06, + "loss": 2.7047, + "step": 3227 + }, + { + "epoch": 0.3413705583756345, + "grad_norm": 7.46875, + "learning_rate": 4.9808290332834905e-06, + "loss": 2.5913, + "step": 3228 + }, + { + "epoch": 0.3414763113367174, + "grad_norm": 7.1875, + "learning_rate": 4.980807859494452e-06, + "loss": 2.6208, + "step": 3229 + }, + { + "epoch": 0.34158206429780036, + "grad_norm": 7.53125, + "learning_rate": 4.980786674063996e-06, + "loss": 2.5149, + "step": 3230 + }, + { + "epoch": 0.34168781725888325, + "grad_norm": 7.625, + "learning_rate": 4.980765476992223e-06, + "loss": 2.6674, + "step": 3231 + }, + { + "epoch": 0.34179357021996615, + "grad_norm": 8.6875, + "learning_rate": 4.980744268279232e-06, + "loss": 2.9115, + "step": 3232 + }, + { + "epoch": 0.34189932318104904, + "grad_norm": 7.5625, + "learning_rate": 4.9807230479251225e-06, + "loss": 2.6502, + "step": 3233 + }, + { + "epoch": 0.342005076142132, + "grad_norm": 7.59375, + "learning_rate": 4.9807018159299955e-06, + "loss": 2.4048, + "step": 3234 + }, + { + "epoch": 0.3421108291032149, + "grad_norm": 8.5, + "learning_rate": 4.980680572293948e-06, + "loss": 2.8167, + "step": 3235 + }, + { + "epoch": 0.3422165820642978, + "grad_norm": 7.59375, + "learning_rate": 4.980659317017082e-06, + "loss": 2.5539, + "step": 3236 + }, + { + "epoch": 0.34232233502538073, + "grad_norm": 8.0625, + "learning_rate": 4.980638050099497e-06, + "loss": 2.6952, + "step": 3237 + }, + { + "epoch": 0.3424280879864636, + "grad_norm": 7.9375, + "learning_rate": 4.980616771541291e-06, + "loss": 2.7739, + "step": 3238 + }, + { + "epoch": 0.3425338409475465, + "grad_norm": 7.6875, + "learning_rate": 4.980595481342565e-06, + "loss": 2.6456, + "step": 3239 + }, + { + "epoch": 0.3426395939086294, + "grad_norm": 7.71875, + "learning_rate": 4.9805741795034195e-06, + "loss": 2.6683, + "step": 3240 + }, + { + "epoch": 0.34274534686971236, + "grad_norm": 8.0625, + "learning_rate": 4.980552866023954e-06, + "loss": 2.7952, + "step": 3241 + }, + { + "epoch": 0.34285109983079526, + "grad_norm": 8.0, + "learning_rate": 4.9805315409042675e-06, + "loss": 2.6529, + "step": 3242 + }, + { + "epoch": 0.34295685279187815, + "grad_norm": 7.125, + "learning_rate": 4.980510204144462e-06, + "loss": 2.6864, + "step": 3243 + }, + { + "epoch": 0.3430626057529611, + "grad_norm": 8.8125, + "learning_rate": 4.980488855744636e-06, + "loss": 2.5478, + "step": 3244 + }, + { + "epoch": 0.343168358714044, + "grad_norm": 7.6875, + "learning_rate": 4.98046749570489e-06, + "loss": 2.6534, + "step": 3245 + }, + { + "epoch": 0.3432741116751269, + "grad_norm": 9.75, + "learning_rate": 4.980446124025325e-06, + "loss": 2.6248, + "step": 3246 + }, + { + "epoch": 0.3433798646362098, + "grad_norm": 7.78125, + "learning_rate": 4.980424740706041e-06, + "loss": 2.5749, + "step": 3247 + }, + { + "epoch": 0.34348561759729274, + "grad_norm": 7.5625, + "learning_rate": 4.980403345747139e-06, + "loss": 2.6214, + "step": 3248 + }, + { + "epoch": 0.34359137055837563, + "grad_norm": 8.8125, + "learning_rate": 4.980381939148717e-06, + "loss": 2.4261, + "step": 3249 + }, + { + "epoch": 0.3436971235194585, + "grad_norm": 7.375, + "learning_rate": 4.980360520910878e-06, + "loss": 2.5398, + "step": 3250 + }, + { + "epoch": 0.3438028764805415, + "grad_norm": 7.71875, + "learning_rate": 4.98033909103372e-06, + "loss": 2.6288, + "step": 3251 + }, + { + "epoch": 0.34390862944162437, + "grad_norm": 8.1875, + "learning_rate": 4.980317649517347e-06, + "loss": 2.6507, + "step": 3252 + }, + { + "epoch": 0.34401438240270726, + "grad_norm": 7.5, + "learning_rate": 4.9802961963618565e-06, + "loss": 2.5042, + "step": 3253 + }, + { + "epoch": 0.3441201353637902, + "grad_norm": 8.1875, + "learning_rate": 4.9802747315673515e-06, + "loss": 2.7607, + "step": 3254 + }, + { + "epoch": 0.3442258883248731, + "grad_norm": 7.875, + "learning_rate": 4.98025325513393e-06, + "loss": 2.8571, + "step": 3255 + }, + { + "epoch": 0.344331641285956, + "grad_norm": 7.375, + "learning_rate": 4.980231767061695e-06, + "loss": 2.4747, + "step": 3256 + }, + { + "epoch": 0.3444373942470389, + "grad_norm": 7.75, + "learning_rate": 4.9802102673507465e-06, + "loss": 2.7118, + "step": 3257 + }, + { + "epoch": 0.34454314720812185, + "grad_norm": 8.5, + "learning_rate": 4.980188756001186e-06, + "loss": 2.6536, + "step": 3258 + }, + { + "epoch": 0.34464890016920474, + "grad_norm": 7.65625, + "learning_rate": 4.980167233013115e-06, + "loss": 2.7811, + "step": 3259 + }, + { + "epoch": 0.34475465313028764, + "grad_norm": 8.375, + "learning_rate": 4.980145698386632e-06, + "loss": 2.6097, + "step": 3260 + }, + { + "epoch": 0.3448604060913706, + "grad_norm": 8.0, + "learning_rate": 4.98012415212184e-06, + "loss": 2.5116, + "step": 3261 + }, + { + "epoch": 0.3449661590524535, + "grad_norm": 7.71875, + "learning_rate": 4.980102594218839e-06, + "loss": 2.5918, + "step": 3262 + }, + { + "epoch": 0.3450719120135364, + "grad_norm": 7.90625, + "learning_rate": 4.980081024677732e-06, + "loss": 2.3606, + "step": 3263 + }, + { + "epoch": 0.34517766497461927, + "grad_norm": 8.0625, + "learning_rate": 4.980059443498619e-06, + "loss": 2.559, + "step": 3264 + }, + { + "epoch": 0.3452834179357022, + "grad_norm": 7.96875, + "learning_rate": 4.980037850681601e-06, + "loss": 2.5042, + "step": 3265 + }, + { + "epoch": 0.3453891708967851, + "grad_norm": 7.9375, + "learning_rate": 4.98001624622678e-06, + "loss": 2.6283, + "step": 3266 + }, + { + "epoch": 0.345494923857868, + "grad_norm": 7.78125, + "learning_rate": 4.979994630134256e-06, + "loss": 2.5191, + "step": 3267 + }, + { + "epoch": 0.34560067681895096, + "grad_norm": 8.4375, + "learning_rate": 4.979973002404133e-06, + "loss": 2.6029, + "step": 3268 + }, + { + "epoch": 0.34570642978003385, + "grad_norm": 7.84375, + "learning_rate": 4.97995136303651e-06, + "loss": 2.7062, + "step": 3269 + }, + { + "epoch": 0.34581218274111675, + "grad_norm": 7.375, + "learning_rate": 4.979929712031489e-06, + "loss": 2.5725, + "step": 3270 + }, + { + "epoch": 0.34591793570219964, + "grad_norm": 7.46875, + "learning_rate": 4.979908049389173e-06, + "loss": 2.5913, + "step": 3271 + }, + { + "epoch": 0.3460236886632826, + "grad_norm": 8.5625, + "learning_rate": 4.979886375109663e-06, + "loss": 2.8369, + "step": 3272 + }, + { + "epoch": 0.3461294416243655, + "grad_norm": 8.125, + "learning_rate": 4.9798646891930605e-06, + "loss": 2.7009, + "step": 3273 + }, + { + "epoch": 0.3462351945854484, + "grad_norm": 7.1875, + "learning_rate": 4.979842991639467e-06, + "loss": 2.6221, + "step": 3274 + }, + { + "epoch": 0.34634094754653133, + "grad_norm": 7.65625, + "learning_rate": 4.9798212824489834e-06, + "loss": 2.6071, + "step": 3275 + }, + { + "epoch": 0.3464467005076142, + "grad_norm": 7.65625, + "learning_rate": 4.9797995616217134e-06, + "loss": 2.7096, + "step": 3276 + }, + { + "epoch": 0.3465524534686971, + "grad_norm": 8.0625, + "learning_rate": 4.979777829157759e-06, + "loss": 2.75, + "step": 3277 + }, + { + "epoch": 0.34665820642978, + "grad_norm": 11.25, + "learning_rate": 4.979756085057221e-06, + "loss": 2.6317, + "step": 3278 + }, + { + "epoch": 0.34676395939086296, + "grad_norm": 13.5, + "learning_rate": 4.979734329320202e-06, + "loss": 2.5944, + "step": 3279 + }, + { + "epoch": 0.34686971235194586, + "grad_norm": 7.875, + "learning_rate": 4.979712561946803e-06, + "loss": 2.5645, + "step": 3280 + }, + { + "epoch": 0.34697546531302875, + "grad_norm": 7.625, + "learning_rate": 4.979690782937129e-06, + "loss": 2.6455, + "step": 3281 + }, + { + "epoch": 0.3470812182741117, + "grad_norm": 7.65625, + "learning_rate": 4.979668992291279e-06, + "loss": 2.6624, + "step": 3282 + }, + { + "epoch": 0.3471869712351946, + "grad_norm": 9.1875, + "learning_rate": 4.979647190009357e-06, + "loss": 2.439, + "step": 3283 + }, + { + "epoch": 0.3472927241962775, + "grad_norm": 8.3125, + "learning_rate": 4.979625376091465e-06, + "loss": 2.3595, + "step": 3284 + }, + { + "epoch": 0.3473984771573604, + "grad_norm": 7.96875, + "learning_rate": 4.979603550537704e-06, + "loss": 2.6478, + "step": 3285 + }, + { + "epoch": 0.34750423011844334, + "grad_norm": 8.0, + "learning_rate": 4.9795817133481794e-06, + "loss": 2.6154, + "step": 3286 + }, + { + "epoch": 0.34760998307952623, + "grad_norm": 7.65625, + "learning_rate": 4.979559864522991e-06, + "loss": 2.4696, + "step": 3287 + }, + { + "epoch": 0.3477157360406091, + "grad_norm": 7.21875, + "learning_rate": 4.979538004062242e-06, + "loss": 2.5265, + "step": 3288 + }, + { + "epoch": 0.3478214890016921, + "grad_norm": 7.28125, + "learning_rate": 4.979516131966036e-06, + "loss": 2.5749, + "step": 3289 + }, + { + "epoch": 0.34792724196277497, + "grad_norm": 8.125, + "learning_rate": 4.979494248234475e-06, + "loss": 2.5894, + "step": 3290 + }, + { + "epoch": 0.34803299492385786, + "grad_norm": 7.875, + "learning_rate": 4.97947235286766e-06, + "loss": 2.4739, + "step": 3291 + }, + { + "epoch": 0.34813874788494076, + "grad_norm": 8.1875, + "learning_rate": 4.9794504458656965e-06, + "loss": 2.5594, + "step": 3292 + }, + { + "epoch": 0.3482445008460237, + "grad_norm": 8.1875, + "learning_rate": 4.979428527228687e-06, + "loss": 2.8674, + "step": 3293 + }, + { + "epoch": 0.3483502538071066, + "grad_norm": 8.625, + "learning_rate": 4.979406596956732e-06, + "loss": 3.1294, + "step": 3294 + }, + { + "epoch": 0.3484560067681895, + "grad_norm": 8.125, + "learning_rate": 4.979384655049937e-06, + "loss": 2.5855, + "step": 3295 + }, + { + "epoch": 0.34856175972927245, + "grad_norm": 8.25, + "learning_rate": 4.979362701508403e-06, + "loss": 2.5149, + "step": 3296 + }, + { + "epoch": 0.34866751269035534, + "grad_norm": 8.125, + "learning_rate": 4.9793407363322346e-06, + "loss": 2.7549, + "step": 3297 + }, + { + "epoch": 0.34877326565143824, + "grad_norm": 7.8125, + "learning_rate": 4.979318759521533e-06, + "loss": 2.5582, + "step": 3298 + }, + { + "epoch": 0.34887901861252113, + "grad_norm": 8.625, + "learning_rate": 4.9792967710764025e-06, + "loss": 2.6289, + "step": 3299 + }, + { + "epoch": 0.3489847715736041, + "grad_norm": 8.5625, + "learning_rate": 4.979274770996947e-06, + "loss": 2.6918, + "step": 3300 + }, + { + "epoch": 0.349090524534687, + "grad_norm": 8.25, + "learning_rate": 4.979252759283269e-06, + "loss": 2.7651, + "step": 3301 + }, + { + "epoch": 0.34919627749576987, + "grad_norm": 7.84375, + "learning_rate": 4.9792307359354716e-06, + "loss": 2.5514, + "step": 3302 + }, + { + "epoch": 0.3493020304568528, + "grad_norm": 7.5, + "learning_rate": 4.979208700953659e-06, + "loss": 2.6494, + "step": 3303 + }, + { + "epoch": 0.3494077834179357, + "grad_norm": 7.78125, + "learning_rate": 4.979186654337933e-06, + "loss": 2.7144, + "step": 3304 + }, + { + "epoch": 0.3495135363790186, + "grad_norm": 8.125, + "learning_rate": 4.9791645960883985e-06, + "loss": 2.74, + "step": 3305 + }, + { + "epoch": 0.3496192893401015, + "grad_norm": 7.75, + "learning_rate": 4.979142526205159e-06, + "loss": 2.4685, + "step": 3306 + }, + { + "epoch": 0.34972504230118445, + "grad_norm": 7.9375, + "learning_rate": 4.9791204446883154e-06, + "loss": 2.634, + "step": 3307 + }, + { + "epoch": 0.34983079526226735, + "grad_norm": 8.1875, + "learning_rate": 4.979098351537975e-06, + "loss": 2.7518, + "step": 3308 + }, + { + "epoch": 0.34993654822335024, + "grad_norm": 8.875, + "learning_rate": 4.979076246754241e-06, + "loss": 2.4183, + "step": 3309 + }, + { + "epoch": 0.3500423011844332, + "grad_norm": 7.6875, + "learning_rate": 4.9790541303372146e-06, + "loss": 2.5591, + "step": 3310 + }, + { + "epoch": 0.3501480541455161, + "grad_norm": 7.625, + "learning_rate": 4.979032002287001e-06, + "loss": 2.5074, + "step": 3311 + }, + { + "epoch": 0.350253807106599, + "grad_norm": 7.59375, + "learning_rate": 4.979009862603705e-06, + "loss": 2.4098, + "step": 3312 + }, + { + "epoch": 0.3503595600676819, + "grad_norm": 7.53125, + "learning_rate": 4.978987711287429e-06, + "loss": 2.678, + "step": 3313 + }, + { + "epoch": 0.3504653130287648, + "grad_norm": 8.4375, + "learning_rate": 4.978965548338278e-06, + "loss": 2.7877, + "step": 3314 + }, + { + "epoch": 0.3505710659898477, + "grad_norm": 9.5, + "learning_rate": 4.978943373756355e-06, + "loss": 2.6398, + "step": 3315 + }, + { + "epoch": 0.3506768189509306, + "grad_norm": 8.4375, + "learning_rate": 4.978921187541765e-06, + "loss": 2.6569, + "step": 3316 + }, + { + "epoch": 0.35078257191201356, + "grad_norm": 7.59375, + "learning_rate": 4.978898989694611e-06, + "loss": 2.4961, + "step": 3317 + }, + { + "epoch": 0.35088832487309646, + "grad_norm": 7.96875, + "learning_rate": 4.978876780214998e-06, + "loss": 2.6791, + "step": 3318 + }, + { + "epoch": 0.35099407783417935, + "grad_norm": 7.4375, + "learning_rate": 4.978854559103031e-06, + "loss": 2.5774, + "step": 3319 + }, + { + "epoch": 0.35109983079526225, + "grad_norm": 7.75, + "learning_rate": 4.978832326358813e-06, + "loss": 2.5403, + "step": 3320 + }, + { + "epoch": 0.3512055837563452, + "grad_norm": 7.9375, + "learning_rate": 4.978810081982448e-06, + "loss": 2.5433, + "step": 3321 + }, + { + "epoch": 0.3513113367174281, + "grad_norm": 9.0625, + "learning_rate": 4.978787825974042e-06, + "loss": 2.5895, + "step": 3322 + }, + { + "epoch": 0.351417089678511, + "grad_norm": 8.0, + "learning_rate": 4.978765558333698e-06, + "loss": 2.7133, + "step": 3323 + }, + { + "epoch": 0.35152284263959394, + "grad_norm": 7.90625, + "learning_rate": 4.978743279061521e-06, + "loss": 2.8692, + "step": 3324 + }, + { + "epoch": 0.35162859560067683, + "grad_norm": 8.6875, + "learning_rate": 4.978720988157616e-06, + "loss": 2.7783, + "step": 3325 + }, + { + "epoch": 0.3517343485617597, + "grad_norm": 7.75, + "learning_rate": 4.978698685622087e-06, + "loss": 2.667, + "step": 3326 + }, + { + "epoch": 0.3518401015228426, + "grad_norm": 7.1875, + "learning_rate": 4.978676371455039e-06, + "loss": 2.4969, + "step": 3327 + }, + { + "epoch": 0.35194585448392557, + "grad_norm": 7.9375, + "learning_rate": 4.978654045656575e-06, + "loss": 2.5032, + "step": 3328 + }, + { + "epoch": 0.35205160744500846, + "grad_norm": 7.875, + "learning_rate": 4.9786317082268036e-06, + "loss": 2.5974, + "step": 3329 + }, + { + "epoch": 0.35215736040609136, + "grad_norm": 8.0625, + "learning_rate": 4.978609359165826e-06, + "loss": 2.8897, + "step": 3330 + }, + { + "epoch": 0.3522631133671743, + "grad_norm": 7.375, + "learning_rate": 4.978586998473749e-06, + "loss": 2.5649, + "step": 3331 + }, + { + "epoch": 0.3523688663282572, + "grad_norm": 7.90625, + "learning_rate": 4.978564626150677e-06, + "loss": 2.6629, + "step": 3332 + }, + { + "epoch": 0.3524746192893401, + "grad_norm": 7.9375, + "learning_rate": 4.9785422421967145e-06, + "loss": 2.5669, + "step": 3333 + }, + { + "epoch": 0.352580372250423, + "grad_norm": 7.5625, + "learning_rate": 4.978519846611967e-06, + "loss": 2.4708, + "step": 3334 + }, + { + "epoch": 0.35268612521150594, + "grad_norm": 9.5, + "learning_rate": 4.978497439396539e-06, + "loss": 2.4566, + "step": 3335 + }, + { + "epoch": 0.35279187817258884, + "grad_norm": 7.8125, + "learning_rate": 4.978475020550537e-06, + "loss": 2.5374, + "step": 3336 + }, + { + "epoch": 0.35289763113367173, + "grad_norm": 7.65625, + "learning_rate": 4.978452590074065e-06, + "loss": 2.3589, + "step": 3337 + }, + { + "epoch": 0.3530033840947547, + "grad_norm": 8.125, + "learning_rate": 4.9784301479672284e-06, + "loss": 2.4173, + "step": 3338 + }, + { + "epoch": 0.3531091370558376, + "grad_norm": 8.25, + "learning_rate": 4.978407694230133e-06, + "loss": 2.6358, + "step": 3339 + }, + { + "epoch": 0.35321489001692047, + "grad_norm": 7.96875, + "learning_rate": 4.978385228862884e-06, + "loss": 2.7235, + "step": 3340 + }, + { + "epoch": 0.35332064297800336, + "grad_norm": 9.0, + "learning_rate": 4.978362751865587e-06, + "loss": 2.5798, + "step": 3341 + }, + { + "epoch": 0.3534263959390863, + "grad_norm": 8.375, + "learning_rate": 4.978340263238347e-06, + "loss": 2.7766, + "step": 3342 + }, + { + "epoch": 0.3535321489001692, + "grad_norm": 7.75, + "learning_rate": 4.978317762981269e-06, + "loss": 2.8646, + "step": 3343 + }, + { + "epoch": 0.3536379018612521, + "grad_norm": 8.375, + "learning_rate": 4.97829525109446e-06, + "loss": 2.6715, + "step": 3344 + }, + { + "epoch": 0.35374365482233505, + "grad_norm": 7.78125, + "learning_rate": 4.9782727275780256e-06, + "loss": 2.5651, + "step": 3345 + }, + { + "epoch": 0.35384940778341795, + "grad_norm": 8.1875, + "learning_rate": 4.978250192432071e-06, + "loss": 2.6018, + "step": 3346 + }, + { + "epoch": 0.35395516074450084, + "grad_norm": 7.75, + "learning_rate": 4.9782276456567e-06, + "loss": 2.7098, + "step": 3347 + }, + { + "epoch": 0.35406091370558374, + "grad_norm": 8.1875, + "learning_rate": 4.9782050872520215e-06, + "loss": 2.6012, + "step": 3348 + }, + { + "epoch": 0.3541666666666667, + "grad_norm": 8.5, + "learning_rate": 4.978182517218141e-06, + "loss": 2.6969, + "step": 3349 + }, + { + "epoch": 0.3542724196277496, + "grad_norm": 8.25, + "learning_rate": 4.9781599355551616e-06, + "loss": 2.7892, + "step": 3350 + }, + { + "epoch": 0.3543781725888325, + "grad_norm": 8.125, + "learning_rate": 4.978137342263192e-06, + "loss": 2.5534, + "step": 3351 + }, + { + "epoch": 0.3544839255499154, + "grad_norm": 8.5625, + "learning_rate": 4.978114737342338e-06, + "loss": 2.694, + "step": 3352 + }, + { + "epoch": 0.3545896785109983, + "grad_norm": 8.0, + "learning_rate": 4.9780921207927034e-06, + "loss": 2.5465, + "step": 3353 + }, + { + "epoch": 0.3546954314720812, + "grad_norm": 7.75, + "learning_rate": 4.978069492614398e-06, + "loss": 2.617, + "step": 3354 + }, + { + "epoch": 0.3548011844331641, + "grad_norm": 7.65625, + "learning_rate": 4.9780468528075245e-06, + "loss": 2.5069, + "step": 3355 + }, + { + "epoch": 0.35490693739424706, + "grad_norm": 7.5625, + "learning_rate": 4.978024201372191e-06, + "loss": 2.5637, + "step": 3356 + }, + { + "epoch": 0.35501269035532995, + "grad_norm": 7.71875, + "learning_rate": 4.978001538308504e-06, + "loss": 2.3845, + "step": 3357 + }, + { + "epoch": 0.35511844331641285, + "grad_norm": 8.4375, + "learning_rate": 4.9779788636165685e-06, + "loss": 2.7617, + "step": 3358 + }, + { + "epoch": 0.3552241962774958, + "grad_norm": 7.59375, + "learning_rate": 4.977956177296491e-06, + "loss": 2.6208, + "step": 3359 + }, + { + "epoch": 0.3553299492385787, + "grad_norm": 7.34375, + "learning_rate": 4.977933479348381e-06, + "loss": 2.538, + "step": 3360 + }, + { + "epoch": 0.3554357021996616, + "grad_norm": 7.96875, + "learning_rate": 4.97791076977234e-06, + "loss": 2.8367, + "step": 3361 + }, + { + "epoch": 0.3555414551607445, + "grad_norm": 7.28125, + "learning_rate": 4.977888048568479e-06, + "loss": 2.6495, + "step": 3362 + }, + { + "epoch": 0.35564720812182743, + "grad_norm": 7.8125, + "learning_rate": 4.977865315736901e-06, + "loss": 2.6435, + "step": 3363 + }, + { + "epoch": 0.3557529610829103, + "grad_norm": 7.78125, + "learning_rate": 4.977842571277715e-06, + "loss": 2.5211, + "step": 3364 + }, + { + "epoch": 0.3558587140439932, + "grad_norm": 7.28125, + "learning_rate": 4.977819815191028e-06, + "loss": 2.4885, + "step": 3365 + }, + { + "epoch": 0.35596446700507617, + "grad_norm": 8.125, + "learning_rate": 4.9777970474769465e-06, + "loss": 2.7381, + "step": 3366 + }, + { + "epoch": 0.35607021996615906, + "grad_norm": 8.6875, + "learning_rate": 4.9777742681355755e-06, + "loss": 2.594, + "step": 3367 + }, + { + "epoch": 0.35617597292724196, + "grad_norm": 8.25, + "learning_rate": 4.977751477167024e-06, + "loss": 2.7324, + "step": 3368 + }, + { + "epoch": 0.35628172588832485, + "grad_norm": 8.8125, + "learning_rate": 4.9777286745713985e-06, + "loss": 2.7729, + "step": 3369 + }, + { + "epoch": 0.3563874788494078, + "grad_norm": 8.0, + "learning_rate": 4.977705860348805e-06, + "loss": 2.8089, + "step": 3370 + }, + { + "epoch": 0.3564932318104907, + "grad_norm": 7.90625, + "learning_rate": 4.977683034499351e-06, + "loss": 2.658, + "step": 3371 + }, + { + "epoch": 0.3565989847715736, + "grad_norm": 8.125, + "learning_rate": 4.9776601970231445e-06, + "loss": 2.6963, + "step": 3372 + }, + { + "epoch": 0.35670473773265654, + "grad_norm": 7.1875, + "learning_rate": 4.9776373479202925e-06, + "loss": 2.4332, + "step": 3373 + }, + { + "epoch": 0.35681049069373943, + "grad_norm": 7.90625, + "learning_rate": 4.9776144871909e-06, + "loss": 2.446, + "step": 3374 + }, + { + "epoch": 0.35691624365482233, + "grad_norm": 7.5625, + "learning_rate": 4.977591614835078e-06, + "loss": 2.6644, + "step": 3375 + }, + { + "epoch": 0.3570219966159052, + "grad_norm": 7.6875, + "learning_rate": 4.977568730852931e-06, + "loss": 2.3475, + "step": 3376 + }, + { + "epoch": 0.3571277495769882, + "grad_norm": 7.71875, + "learning_rate": 4.977545835244567e-06, + "loss": 2.5413, + "step": 3377 + }, + { + "epoch": 0.35723350253807107, + "grad_norm": 7.625, + "learning_rate": 4.977522928010094e-06, + "loss": 2.7249, + "step": 3378 + }, + { + "epoch": 0.35733925549915396, + "grad_norm": 7.625, + "learning_rate": 4.977500009149619e-06, + "loss": 2.7447, + "step": 3379 + }, + { + "epoch": 0.3574450084602369, + "grad_norm": 7.46875, + "learning_rate": 4.97747707866325e-06, + "loss": 2.6626, + "step": 3380 + }, + { + "epoch": 0.3575507614213198, + "grad_norm": 8.0625, + "learning_rate": 4.977454136551094e-06, + "loss": 2.5189, + "step": 3381 + }, + { + "epoch": 0.3576565143824027, + "grad_norm": 7.84375, + "learning_rate": 4.977431182813259e-06, + "loss": 2.7937, + "step": 3382 + }, + { + "epoch": 0.3577622673434856, + "grad_norm": 8.5, + "learning_rate": 4.977408217449853e-06, + "loss": 2.3366, + "step": 3383 + }, + { + "epoch": 0.35786802030456855, + "grad_norm": 7.65625, + "learning_rate": 4.977385240460983e-06, + "loss": 2.4755, + "step": 3384 + }, + { + "epoch": 0.35797377326565144, + "grad_norm": 8.3125, + "learning_rate": 4.977362251846757e-06, + "loss": 2.704, + "step": 3385 + }, + { + "epoch": 0.35807952622673433, + "grad_norm": 8.4375, + "learning_rate": 4.977339251607284e-06, + "loss": 2.5081, + "step": 3386 + }, + { + "epoch": 0.3581852791878173, + "grad_norm": 7.78125, + "learning_rate": 4.977316239742671e-06, + "loss": 2.7048, + "step": 3387 + }, + { + "epoch": 0.3582910321489002, + "grad_norm": 7.46875, + "learning_rate": 4.977293216253025e-06, + "loss": 2.5624, + "step": 3388 + }, + { + "epoch": 0.3583967851099831, + "grad_norm": 7.375, + "learning_rate": 4.977270181138457e-06, + "loss": 2.5574, + "step": 3389 + }, + { + "epoch": 0.35850253807106597, + "grad_norm": 8.8125, + "learning_rate": 4.977247134399072e-06, + "loss": 2.6206, + "step": 3390 + }, + { + "epoch": 0.3586082910321489, + "grad_norm": 7.8125, + "learning_rate": 4.977224076034979e-06, + "loss": 2.5547, + "step": 3391 + }, + { + "epoch": 0.3587140439932318, + "grad_norm": 8.3125, + "learning_rate": 4.977201006046287e-06, + "loss": 2.5091, + "step": 3392 + }, + { + "epoch": 0.3588197969543147, + "grad_norm": 7.9375, + "learning_rate": 4.977177924433104e-06, + "loss": 2.7, + "step": 3393 + }, + { + "epoch": 0.35892554991539766, + "grad_norm": 8.0625, + "learning_rate": 4.977154831195539e-06, + "loss": 2.6879, + "step": 3394 + }, + { + "epoch": 0.35903130287648055, + "grad_norm": 7.75, + "learning_rate": 4.977131726333698e-06, + "loss": 2.3697, + "step": 3395 + }, + { + "epoch": 0.35913705583756345, + "grad_norm": 7.65625, + "learning_rate": 4.9771086098476915e-06, + "loss": 2.5798, + "step": 3396 + }, + { + "epoch": 0.35924280879864634, + "grad_norm": 7.375, + "learning_rate": 4.977085481737627e-06, + "loss": 2.5132, + "step": 3397 + }, + { + "epoch": 0.3593485617597293, + "grad_norm": 8.1875, + "learning_rate": 4.977062342003614e-06, + "loss": 2.6302, + "step": 3398 + }, + { + "epoch": 0.3594543147208122, + "grad_norm": 7.21875, + "learning_rate": 4.97703919064576e-06, + "loss": 2.5923, + "step": 3399 + }, + { + "epoch": 0.3595600676818951, + "grad_norm": 7.9375, + "learning_rate": 4.977016027664174e-06, + "loss": 2.5346, + "step": 3400 + }, + { + "epoch": 0.35966582064297803, + "grad_norm": 7.5, + "learning_rate": 4.976992853058966e-06, + "loss": 2.6236, + "step": 3401 + }, + { + "epoch": 0.3597715736040609, + "grad_norm": 7.78125, + "learning_rate": 4.976969666830243e-06, + "loss": 2.5323, + "step": 3402 + }, + { + "epoch": 0.3598773265651438, + "grad_norm": 7.46875, + "learning_rate": 4.976946468978114e-06, + "loss": 2.6218, + "step": 3403 + }, + { + "epoch": 0.3599830795262267, + "grad_norm": 7.625, + "learning_rate": 4.976923259502688e-06, + "loss": 2.3861, + "step": 3404 + }, + { + "epoch": 0.36008883248730966, + "grad_norm": 8.125, + "learning_rate": 4.9769000384040755e-06, + "loss": 2.6984, + "step": 3405 + }, + { + "epoch": 0.36019458544839256, + "grad_norm": 7.625, + "learning_rate": 4.9768768056823835e-06, + "loss": 2.6651, + "step": 3406 + }, + { + "epoch": 0.36030033840947545, + "grad_norm": 8.875, + "learning_rate": 4.9768535613377215e-06, + "loss": 2.7731, + "step": 3407 + }, + { + "epoch": 0.3604060913705584, + "grad_norm": 9.4375, + "learning_rate": 4.976830305370198e-06, + "loss": 2.7224, + "step": 3408 + }, + { + "epoch": 0.3605118443316413, + "grad_norm": 7.71875, + "learning_rate": 4.976807037779924e-06, + "loss": 2.5915, + "step": 3409 + }, + { + "epoch": 0.3606175972927242, + "grad_norm": 7.40625, + "learning_rate": 4.976783758567007e-06, + "loss": 2.6374, + "step": 3410 + }, + { + "epoch": 0.3607233502538071, + "grad_norm": 7.15625, + "learning_rate": 4.976760467731557e-06, + "loss": 2.512, + "step": 3411 + }, + { + "epoch": 0.36082910321489003, + "grad_norm": 8.0625, + "learning_rate": 4.976737165273683e-06, + "loss": 2.7593, + "step": 3412 + }, + { + "epoch": 0.36093485617597293, + "grad_norm": 8.0, + "learning_rate": 4.976713851193494e-06, + "loss": 2.5755, + "step": 3413 + }, + { + "epoch": 0.3610406091370558, + "grad_norm": 9.25, + "learning_rate": 4.976690525491101e-06, + "loss": 2.4085, + "step": 3414 + }, + { + "epoch": 0.3611463620981388, + "grad_norm": 8.375, + "learning_rate": 4.9766671881666115e-06, + "loss": 2.592, + "step": 3415 + }, + { + "epoch": 0.36125211505922167, + "grad_norm": 7.65625, + "learning_rate": 4.976643839220136e-06, + "loss": 2.6479, + "step": 3416 + }, + { + "epoch": 0.36135786802030456, + "grad_norm": 7.875, + "learning_rate": 4.976620478651783e-06, + "loss": 2.6765, + "step": 3417 + }, + { + "epoch": 0.36146362098138746, + "grad_norm": 7.78125, + "learning_rate": 4.9765971064616645e-06, + "loss": 2.5734, + "step": 3418 + }, + { + "epoch": 0.3615693739424704, + "grad_norm": 7.53125, + "learning_rate": 4.976573722649888e-06, + "loss": 2.5006, + "step": 3419 + }, + { + "epoch": 0.3616751269035533, + "grad_norm": 8.6875, + "learning_rate": 4.976550327216564e-06, + "loss": 2.5247, + "step": 3420 + }, + { + "epoch": 0.3617808798646362, + "grad_norm": 7.75, + "learning_rate": 4.976526920161802e-06, + "loss": 2.6309, + "step": 3421 + }, + { + "epoch": 0.36188663282571915, + "grad_norm": 7.71875, + "learning_rate": 4.976503501485712e-06, + "loss": 2.4025, + "step": 3422 + }, + { + "epoch": 0.36199238578680204, + "grad_norm": 7.78125, + "learning_rate": 4.976480071188405e-06, + "loss": 2.7175, + "step": 3423 + }, + { + "epoch": 0.36209813874788493, + "grad_norm": 7.84375, + "learning_rate": 4.976456629269988e-06, + "loss": 2.5664, + "step": 3424 + }, + { + "epoch": 0.36220389170896783, + "grad_norm": 7.625, + "learning_rate": 4.976433175730574e-06, + "loss": 2.7332, + "step": 3425 + }, + { + "epoch": 0.3623096446700508, + "grad_norm": 8.25, + "learning_rate": 4.976409710570271e-06, + "loss": 2.6086, + "step": 3426 + }, + { + "epoch": 0.3624153976311337, + "grad_norm": 8.5, + "learning_rate": 4.9763862337891914e-06, + "loss": 2.5394, + "step": 3427 + }, + { + "epoch": 0.36252115059221657, + "grad_norm": 8.125, + "learning_rate": 4.976362745387443e-06, + "loss": 2.5946, + "step": 3428 + }, + { + "epoch": 0.3626269035532995, + "grad_norm": 7.875, + "learning_rate": 4.976339245365137e-06, + "loss": 2.5588, + "step": 3429 + }, + { + "epoch": 0.3627326565143824, + "grad_norm": 7.40625, + "learning_rate": 4.976315733722385e-06, + "loss": 2.5323, + "step": 3430 + }, + { + "epoch": 0.3628384094754653, + "grad_norm": 7.625, + "learning_rate": 4.976292210459294e-06, + "loss": 2.7151, + "step": 3431 + }, + { + "epoch": 0.3629441624365482, + "grad_norm": 7.40625, + "learning_rate": 4.976268675575978e-06, + "loss": 2.4232, + "step": 3432 + }, + { + "epoch": 0.36304991539763115, + "grad_norm": 8.1875, + "learning_rate": 4.976245129072545e-06, + "loss": 2.6805, + "step": 3433 + }, + { + "epoch": 0.36315566835871405, + "grad_norm": 7.875, + "learning_rate": 4.976221570949107e-06, + "loss": 2.6685, + "step": 3434 + }, + { + "epoch": 0.36326142131979694, + "grad_norm": 8.8125, + "learning_rate": 4.976198001205774e-06, + "loss": 2.6038, + "step": 3435 + }, + { + "epoch": 0.3633671742808799, + "grad_norm": 8.1875, + "learning_rate": 4.976174419842656e-06, + "loss": 2.6937, + "step": 3436 + }, + { + "epoch": 0.3634729272419628, + "grad_norm": 8.0625, + "learning_rate": 4.976150826859865e-06, + "loss": 2.6174, + "step": 3437 + }, + { + "epoch": 0.3635786802030457, + "grad_norm": 8.125, + "learning_rate": 4.976127222257509e-06, + "loss": 2.7579, + "step": 3438 + }, + { + "epoch": 0.3636844331641286, + "grad_norm": 7.5, + "learning_rate": 4.976103606035703e-06, + "loss": 2.3785, + "step": 3439 + }, + { + "epoch": 0.3637901861252115, + "grad_norm": 7.6875, + "learning_rate": 4.976079978194554e-06, + "loss": 2.6583, + "step": 3440 + }, + { + "epoch": 0.3638959390862944, + "grad_norm": 7.4375, + "learning_rate": 4.976056338734175e-06, + "loss": 2.5159, + "step": 3441 + }, + { + "epoch": 0.3640016920473773, + "grad_norm": 7.9375, + "learning_rate": 4.976032687654675e-06, + "loss": 2.7994, + "step": 3442 + }, + { + "epoch": 0.36410744500846026, + "grad_norm": 7.90625, + "learning_rate": 4.976009024956168e-06, + "loss": 2.7212, + "step": 3443 + }, + { + "epoch": 0.36421319796954316, + "grad_norm": 7.78125, + "learning_rate": 4.975985350638763e-06, + "loss": 2.5806, + "step": 3444 + }, + { + "epoch": 0.36431895093062605, + "grad_norm": 8.0, + "learning_rate": 4.97596166470257e-06, + "loss": 2.204, + "step": 3445 + }, + { + "epoch": 0.36442470389170895, + "grad_norm": 7.84375, + "learning_rate": 4.9759379671477025e-06, + "loss": 2.6323, + "step": 3446 + }, + { + "epoch": 0.3645304568527919, + "grad_norm": 7.9375, + "learning_rate": 4.9759142579742704e-06, + "loss": 2.4338, + "step": 3447 + }, + { + "epoch": 0.3646362098138748, + "grad_norm": 7.875, + "learning_rate": 4.9758905371823865e-06, + "loss": 2.5391, + "step": 3448 + }, + { + "epoch": 0.3647419627749577, + "grad_norm": 8.1875, + "learning_rate": 4.975866804772159e-06, + "loss": 2.745, + "step": 3449 + }, + { + "epoch": 0.36484771573604063, + "grad_norm": 7.53125, + "learning_rate": 4.975843060743702e-06, + "loss": 2.5064, + "step": 3450 + }, + { + "epoch": 0.36495346869712353, + "grad_norm": 7.53125, + "learning_rate": 4.9758193050971275e-06, + "loss": 2.54, + "step": 3451 + }, + { + "epoch": 0.3650592216582064, + "grad_norm": 7.46875, + "learning_rate": 4.975795537832544e-06, + "loss": 2.6138, + "step": 3452 + }, + { + "epoch": 0.3651649746192893, + "grad_norm": 8.1875, + "learning_rate": 4.9757717589500645e-06, + "loss": 2.7236, + "step": 3453 + }, + { + "epoch": 0.36527072758037227, + "grad_norm": 8.1875, + "learning_rate": 4.975747968449801e-06, + "loss": 2.643, + "step": 3454 + }, + { + "epoch": 0.36537648054145516, + "grad_norm": 8.0, + "learning_rate": 4.975724166331865e-06, + "loss": 2.805, + "step": 3455 + }, + { + "epoch": 0.36548223350253806, + "grad_norm": 7.625, + "learning_rate": 4.975700352596369e-06, + "loss": 2.5681, + "step": 3456 + }, + { + "epoch": 0.365587986463621, + "grad_norm": 7.875, + "learning_rate": 4.975676527243423e-06, + "loss": 2.6591, + "step": 3457 + }, + { + "epoch": 0.3656937394247039, + "grad_norm": 9.4375, + "learning_rate": 4.975652690273138e-06, + "loss": 2.5911, + "step": 3458 + }, + { + "epoch": 0.3657994923857868, + "grad_norm": 8.4375, + "learning_rate": 4.975628841685629e-06, + "loss": 2.7494, + "step": 3459 + }, + { + "epoch": 0.3659052453468697, + "grad_norm": 7.59375, + "learning_rate": 4.975604981481007e-06, + "loss": 2.5973, + "step": 3460 + }, + { + "epoch": 0.36601099830795264, + "grad_norm": 7.40625, + "learning_rate": 4.975581109659382e-06, + "loss": 2.6127, + "step": 3461 + }, + { + "epoch": 0.36611675126903553, + "grad_norm": 8.0, + "learning_rate": 4.975557226220868e-06, + "loss": 2.7141, + "step": 3462 + }, + { + "epoch": 0.36622250423011843, + "grad_norm": 7.84375, + "learning_rate": 4.975533331165576e-06, + "loss": 2.7346, + "step": 3463 + }, + { + "epoch": 0.3663282571912014, + "grad_norm": 7.90625, + "learning_rate": 4.975509424493619e-06, + "loss": 2.6858, + "step": 3464 + }, + { + "epoch": 0.3664340101522843, + "grad_norm": 8.5, + "learning_rate": 4.975485506205108e-06, + "loss": 2.7337, + "step": 3465 + }, + { + "epoch": 0.36653976311336717, + "grad_norm": 8.0, + "learning_rate": 4.975461576300157e-06, + "loss": 2.6624, + "step": 3466 + }, + { + "epoch": 0.36664551607445006, + "grad_norm": 7.5625, + "learning_rate": 4.975437634778876e-06, + "loss": 2.6091, + "step": 3467 + }, + { + "epoch": 0.366751269035533, + "grad_norm": 8.9375, + "learning_rate": 4.97541368164138e-06, + "loss": 2.6247, + "step": 3468 + }, + { + "epoch": 0.3668570219966159, + "grad_norm": 9.0, + "learning_rate": 4.975389716887779e-06, + "loss": 2.6813, + "step": 3469 + }, + { + "epoch": 0.3669627749576988, + "grad_norm": 8.1875, + "learning_rate": 4.9753657405181864e-06, + "loss": 2.5119, + "step": 3470 + }, + { + "epoch": 0.36706852791878175, + "grad_norm": 7.9375, + "learning_rate": 4.975341752532715e-06, + "loss": 2.6414, + "step": 3471 + }, + { + "epoch": 0.36717428087986465, + "grad_norm": 8.0625, + "learning_rate": 4.9753177529314775e-06, + "loss": 2.6031, + "step": 3472 + }, + { + "epoch": 0.36728003384094754, + "grad_norm": 8.1875, + "learning_rate": 4.975293741714586e-06, + "loss": 2.4316, + "step": 3473 + }, + { + "epoch": 0.36738578680203043, + "grad_norm": 7.875, + "learning_rate": 4.975269718882153e-06, + "loss": 2.6564, + "step": 3474 + }, + { + "epoch": 0.3674915397631134, + "grad_norm": 8.3125, + "learning_rate": 4.975245684434292e-06, + "loss": 2.8826, + "step": 3475 + }, + { + "epoch": 0.3675972927241963, + "grad_norm": 7.46875, + "learning_rate": 4.975221638371115e-06, + "loss": 2.6445, + "step": 3476 + }, + { + "epoch": 0.3677030456852792, + "grad_norm": 8.125, + "learning_rate": 4.975197580692735e-06, + "loss": 2.4455, + "step": 3477 + }, + { + "epoch": 0.3678087986463621, + "grad_norm": 8.3125, + "learning_rate": 4.975173511399266e-06, + "loss": 2.6046, + "step": 3478 + }, + { + "epoch": 0.367914551607445, + "grad_norm": 8.5625, + "learning_rate": 4.975149430490819e-06, + "loss": 2.4662, + "step": 3479 + }, + { + "epoch": 0.3680203045685279, + "grad_norm": 7.9375, + "learning_rate": 4.975125337967509e-06, + "loss": 2.6273, + "step": 3480 + }, + { + "epoch": 0.3681260575296108, + "grad_norm": 8.4375, + "learning_rate": 4.975101233829447e-06, + "loss": 2.5233, + "step": 3481 + }, + { + "epoch": 0.36823181049069376, + "grad_norm": 7.40625, + "learning_rate": 4.975077118076747e-06, + "loss": 2.7915, + "step": 3482 + }, + { + "epoch": 0.36833756345177665, + "grad_norm": 7.875, + "learning_rate": 4.975052990709524e-06, + "loss": 2.6999, + "step": 3483 + }, + { + "epoch": 0.36844331641285955, + "grad_norm": 8.0, + "learning_rate": 4.975028851727889e-06, + "loss": 2.5267, + "step": 3484 + }, + { + "epoch": 0.3685490693739425, + "grad_norm": 7.53125, + "learning_rate": 4.975004701131955e-06, + "loss": 2.4857, + "step": 3485 + }, + { + "epoch": 0.3686548223350254, + "grad_norm": 11.0, + "learning_rate": 4.974980538921836e-06, + "loss": 2.493, + "step": 3486 + }, + { + "epoch": 0.3687605752961083, + "grad_norm": 8.1875, + "learning_rate": 4.974956365097647e-06, + "loss": 2.5536, + "step": 3487 + }, + { + "epoch": 0.3688663282571912, + "grad_norm": 7.625, + "learning_rate": 4.974932179659499e-06, + "loss": 2.4263, + "step": 3488 + }, + { + "epoch": 0.36897208121827413, + "grad_norm": 7.3125, + "learning_rate": 4.974907982607506e-06, + "loss": 2.4781, + "step": 3489 + }, + { + "epoch": 0.369077834179357, + "grad_norm": 10.375, + "learning_rate": 4.974883773941784e-06, + "loss": 2.6529, + "step": 3490 + }, + { + "epoch": 0.3691835871404399, + "grad_norm": 7.4375, + "learning_rate": 4.974859553662443e-06, + "loss": 2.5655, + "step": 3491 + }, + { + "epoch": 0.36928934010152287, + "grad_norm": 8.25, + "learning_rate": 4.974835321769599e-06, + "loss": 2.7241, + "step": 3492 + }, + { + "epoch": 0.36939509306260576, + "grad_norm": 7.53125, + "learning_rate": 4.974811078263364e-06, + "loss": 2.4418, + "step": 3493 + }, + { + "epoch": 0.36950084602368866, + "grad_norm": 7.8125, + "learning_rate": 4.974786823143854e-06, + "loss": 2.6021, + "step": 3494 + }, + { + "epoch": 0.36960659898477155, + "grad_norm": 7.78125, + "learning_rate": 4.9747625564111815e-06, + "loss": 2.7248, + "step": 3495 + }, + { + "epoch": 0.3697123519458545, + "grad_norm": 8.3125, + "learning_rate": 4.97473827806546e-06, + "loss": 2.6521, + "step": 3496 + }, + { + "epoch": 0.3698181049069374, + "grad_norm": 12.75, + "learning_rate": 4.974713988106803e-06, + "loss": 2.6779, + "step": 3497 + }, + { + "epoch": 0.3699238578680203, + "grad_norm": 7.53125, + "learning_rate": 4.9746896865353266e-06, + "loss": 2.7425, + "step": 3498 + }, + { + "epoch": 0.37002961082910324, + "grad_norm": 7.71875, + "learning_rate": 4.974665373351143e-06, + "loss": 2.508, + "step": 3499 + }, + { + "epoch": 0.37013536379018613, + "grad_norm": 8.1875, + "learning_rate": 4.974641048554367e-06, + "loss": 2.6976, + "step": 3500 + }, + { + "epoch": 0.37024111675126903, + "grad_norm": 7.90625, + "learning_rate": 4.974616712145114e-06, + "loss": 2.7685, + "step": 3501 + }, + { + "epoch": 0.3703468697123519, + "grad_norm": 7.375, + "learning_rate": 4.974592364123495e-06, + "loss": 2.4408, + "step": 3502 + }, + { + "epoch": 0.3704526226734349, + "grad_norm": 8.375, + "learning_rate": 4.9745680044896276e-06, + "loss": 2.6189, + "step": 3503 + }, + { + "epoch": 0.37055837563451777, + "grad_norm": 7.90625, + "learning_rate": 4.974543633243623e-06, + "loss": 2.5171, + "step": 3504 + }, + { + "epoch": 0.37066412859560066, + "grad_norm": 7.65625, + "learning_rate": 4.974519250385598e-06, + "loss": 2.5047, + "step": 3505 + }, + { + "epoch": 0.3707698815566836, + "grad_norm": 7.59375, + "learning_rate": 4.974494855915667e-06, + "loss": 2.476, + "step": 3506 + }, + { + "epoch": 0.3708756345177665, + "grad_norm": 8.0, + "learning_rate": 4.974470449833942e-06, + "loss": 2.5689, + "step": 3507 + }, + { + "epoch": 0.3709813874788494, + "grad_norm": 8.1875, + "learning_rate": 4.974446032140541e-06, + "loss": 2.8212, + "step": 3508 + }, + { + "epoch": 0.3710871404399323, + "grad_norm": 7.59375, + "learning_rate": 4.9744216028355755e-06, + "loss": 2.3962, + "step": 3509 + }, + { + "epoch": 0.37119289340101524, + "grad_norm": 8.0625, + "learning_rate": 4.974397161919162e-06, + "loss": 2.5713, + "step": 3510 + }, + { + "epoch": 0.37129864636209814, + "grad_norm": 8.125, + "learning_rate": 4.974372709391414e-06, + "loss": 2.3692, + "step": 3511 + }, + { + "epoch": 0.37140439932318103, + "grad_norm": 8.125, + "learning_rate": 4.9743482452524475e-06, + "loss": 2.6783, + "step": 3512 + }, + { + "epoch": 0.371510152284264, + "grad_norm": 8.25, + "learning_rate": 4.974323769502376e-06, + "loss": 2.7025, + "step": 3513 + }, + { + "epoch": 0.3716159052453469, + "grad_norm": 8.6875, + "learning_rate": 4.974299282141316e-06, + "loss": 2.5146, + "step": 3514 + }, + { + "epoch": 0.3717216582064298, + "grad_norm": 7.96875, + "learning_rate": 4.974274783169381e-06, + "loss": 2.5046, + "step": 3515 + }, + { + "epoch": 0.37182741116751267, + "grad_norm": 7.875, + "learning_rate": 4.9742502725866855e-06, + "loss": 2.7605, + "step": 3516 + }, + { + "epoch": 0.3719331641285956, + "grad_norm": 8.0625, + "learning_rate": 4.974225750393347e-06, + "loss": 2.7551, + "step": 3517 + }, + { + "epoch": 0.3720389170896785, + "grad_norm": 9.8125, + "learning_rate": 4.974201216589478e-06, + "loss": 2.8707, + "step": 3518 + }, + { + "epoch": 0.3721446700507614, + "grad_norm": 7.78125, + "learning_rate": 4.974176671175195e-06, + "loss": 2.3567, + "step": 3519 + }, + { + "epoch": 0.37225042301184436, + "grad_norm": 7.46875, + "learning_rate": 4.974152114150612e-06, + "loss": 2.372, + "step": 3520 + }, + { + "epoch": 0.37235617597292725, + "grad_norm": 10.125, + "learning_rate": 4.974127545515846e-06, + "loss": 2.67, + "step": 3521 + }, + { + "epoch": 0.37246192893401014, + "grad_norm": 8.625, + "learning_rate": 4.974102965271011e-06, + "loss": 2.3719, + "step": 3522 + }, + { + "epoch": 0.37256768189509304, + "grad_norm": 8.1875, + "learning_rate": 4.974078373416222e-06, + "loss": 2.7844, + "step": 3523 + }, + { + "epoch": 0.372673434856176, + "grad_norm": 7.84375, + "learning_rate": 4.9740537699515965e-06, + "loss": 2.4915, + "step": 3524 + }, + { + "epoch": 0.3727791878172589, + "grad_norm": 8.0625, + "learning_rate": 4.9740291548772464e-06, + "loss": 2.723, + "step": 3525 + }, + { + "epoch": 0.3728849407783418, + "grad_norm": 7.40625, + "learning_rate": 4.974004528193292e-06, + "loss": 2.5667, + "step": 3526 + }, + { + "epoch": 0.37299069373942473, + "grad_norm": 8.1875, + "learning_rate": 4.973979889899844e-06, + "loss": 2.5514, + "step": 3527 + }, + { + "epoch": 0.3730964467005076, + "grad_norm": 7.75, + "learning_rate": 4.9739552399970206e-06, + "loss": 2.6201, + "step": 3528 + }, + { + "epoch": 0.3732021996615905, + "grad_norm": 7.84375, + "learning_rate": 4.973930578484938e-06, + "loss": 2.6268, + "step": 3529 + }, + { + "epoch": 0.3733079526226734, + "grad_norm": 7.75, + "learning_rate": 4.9739059053637105e-06, + "loss": 2.5341, + "step": 3530 + }, + { + "epoch": 0.37341370558375636, + "grad_norm": 8.625, + "learning_rate": 4.9738812206334544e-06, + "loss": 2.5348, + "step": 3531 + }, + { + "epoch": 0.37351945854483926, + "grad_norm": 7.9375, + "learning_rate": 4.973856524294285e-06, + "loss": 2.4707, + "step": 3532 + }, + { + "epoch": 0.37362521150592215, + "grad_norm": 7.34375, + "learning_rate": 4.973831816346319e-06, + "loss": 2.6202, + "step": 3533 + }, + { + "epoch": 0.3737309644670051, + "grad_norm": 7.71875, + "learning_rate": 4.973807096789672e-06, + "loss": 2.5791, + "step": 3534 + }, + { + "epoch": 0.373836717428088, + "grad_norm": 7.84375, + "learning_rate": 4.97378236562446e-06, + "loss": 2.6527, + "step": 3535 + }, + { + "epoch": 0.3739424703891709, + "grad_norm": 7.71875, + "learning_rate": 4.9737576228508e-06, + "loss": 2.5091, + "step": 3536 + }, + { + "epoch": 0.3740482233502538, + "grad_norm": 7.5, + "learning_rate": 4.973732868468805e-06, + "loss": 2.6599, + "step": 3537 + }, + { + "epoch": 0.37415397631133673, + "grad_norm": 7.9375, + "learning_rate": 4.973708102478596e-06, + "loss": 2.6843, + "step": 3538 + }, + { + "epoch": 0.37425972927241963, + "grad_norm": 7.96875, + "learning_rate": 4.973683324880285e-06, + "loss": 2.8682, + "step": 3539 + }, + { + "epoch": 0.3743654822335025, + "grad_norm": 8.0625, + "learning_rate": 4.9736585356739904e-06, + "loss": 2.5805, + "step": 3540 + }, + { + "epoch": 0.3744712351945855, + "grad_norm": 7.75, + "learning_rate": 4.973633734859828e-06, + "loss": 2.6738, + "step": 3541 + }, + { + "epoch": 0.37457698815566837, + "grad_norm": 7.90625, + "learning_rate": 4.9736089224379144e-06, + "loss": 2.3978, + "step": 3542 + }, + { + "epoch": 0.37468274111675126, + "grad_norm": 8.0625, + "learning_rate": 4.9735840984083646e-06, + "loss": 2.6568, + "step": 3543 + }, + { + "epoch": 0.37478849407783416, + "grad_norm": 7.8125, + "learning_rate": 4.973559262771298e-06, + "loss": 2.5731, + "step": 3544 + }, + { + "epoch": 0.3748942470389171, + "grad_norm": 10.3125, + "learning_rate": 4.973534415526828e-06, + "loss": 2.6437, + "step": 3545 + }, + { + "epoch": 0.375, + "grad_norm": 7.96875, + "learning_rate": 4.973509556675074e-06, + "loss": 2.3338, + "step": 3546 + }, + { + "epoch": 0.3751057529610829, + "grad_norm": 7.6875, + "learning_rate": 4.97348468621615e-06, + "loss": 2.4783, + "step": 3547 + }, + { + "epoch": 0.37521150592216584, + "grad_norm": 8.3125, + "learning_rate": 4.973459804150175e-06, + "loss": 2.655, + "step": 3548 + }, + { + "epoch": 0.37531725888324874, + "grad_norm": 7.9375, + "learning_rate": 4.9734349104772636e-06, + "loss": 2.5199, + "step": 3549 + }, + { + "epoch": 0.37542301184433163, + "grad_norm": 7.875, + "learning_rate": 4.973410005197534e-06, + "loss": 2.4519, + "step": 3550 + }, + { + "epoch": 0.3755287648054145, + "grad_norm": 8.125, + "learning_rate": 4.973385088311103e-06, + "loss": 2.7327, + "step": 3551 + }, + { + "epoch": 0.3756345177664975, + "grad_norm": 7.6875, + "learning_rate": 4.973360159818088e-06, + "loss": 2.6056, + "step": 3552 + }, + { + "epoch": 0.37574027072758037, + "grad_norm": 7.65625, + "learning_rate": 4.973335219718605e-06, + "loss": 2.5139, + "step": 3553 + }, + { + "epoch": 0.37584602368866327, + "grad_norm": 8.125, + "learning_rate": 4.9733102680127714e-06, + "loss": 2.6584, + "step": 3554 + }, + { + "epoch": 0.3759517766497462, + "grad_norm": 7.5625, + "learning_rate": 4.973285304700705e-06, + "loss": 2.4807, + "step": 3555 + }, + { + "epoch": 0.3760575296108291, + "grad_norm": 7.71875, + "learning_rate": 4.973260329782521e-06, + "loss": 2.7355, + "step": 3556 + }, + { + "epoch": 0.376163282571912, + "grad_norm": 7.59375, + "learning_rate": 4.973235343258338e-06, + "loss": 2.6219, + "step": 3557 + }, + { + "epoch": 0.3762690355329949, + "grad_norm": 8.0, + "learning_rate": 4.973210345128273e-06, + "loss": 2.4821, + "step": 3558 + }, + { + "epoch": 0.37637478849407785, + "grad_norm": 7.71875, + "learning_rate": 4.973185335392444e-06, + "loss": 2.8437, + "step": 3559 + }, + { + "epoch": 0.37648054145516074, + "grad_norm": 7.625, + "learning_rate": 4.973160314050967e-06, + "loss": 2.7118, + "step": 3560 + }, + { + "epoch": 0.37658629441624364, + "grad_norm": 8.25, + "learning_rate": 4.97313528110396e-06, + "loss": 2.5271, + "step": 3561 + }, + { + "epoch": 0.3766920473773266, + "grad_norm": 7.8125, + "learning_rate": 4.9731102365515414e-06, + "loss": 2.6681, + "step": 3562 + }, + { + "epoch": 0.3767978003384095, + "grad_norm": 7.53125, + "learning_rate": 4.973085180393827e-06, + "loss": 2.6897, + "step": 3563 + }, + { + "epoch": 0.3769035532994924, + "grad_norm": 7.34375, + "learning_rate": 4.973060112630936e-06, + "loss": 2.4743, + "step": 3564 + }, + { + "epoch": 0.37700930626057527, + "grad_norm": 7.875, + "learning_rate": 4.973035033262986e-06, + "loss": 2.6962, + "step": 3565 + }, + { + "epoch": 0.3771150592216582, + "grad_norm": 7.625, + "learning_rate": 4.973009942290093e-06, + "loss": 2.5726, + "step": 3566 + }, + { + "epoch": 0.3772208121827411, + "grad_norm": 8.4375, + "learning_rate": 4.972984839712376e-06, + "loss": 2.672, + "step": 3567 + }, + { + "epoch": 0.377326565143824, + "grad_norm": 8.1875, + "learning_rate": 4.972959725529952e-06, + "loss": 2.838, + "step": 3568 + }, + { + "epoch": 0.37743231810490696, + "grad_norm": 8.0625, + "learning_rate": 4.97293459974294e-06, + "loss": 2.557, + "step": 3569 + }, + { + "epoch": 0.37753807106598986, + "grad_norm": 8.5, + "learning_rate": 4.972909462351457e-06, + "loss": 2.418, + "step": 3570 + }, + { + "epoch": 0.37764382402707275, + "grad_norm": 8.0625, + "learning_rate": 4.972884313355623e-06, + "loss": 2.6769, + "step": 3571 + }, + { + "epoch": 0.37774957698815564, + "grad_norm": 8.8125, + "learning_rate": 4.972859152755553e-06, + "loss": 2.4369, + "step": 3572 + }, + { + "epoch": 0.3778553299492386, + "grad_norm": 8.0625, + "learning_rate": 4.972833980551365e-06, + "loss": 2.6398, + "step": 3573 + }, + { + "epoch": 0.3779610829103215, + "grad_norm": 8.0, + "learning_rate": 4.97280879674318e-06, + "loss": 2.641, + "step": 3574 + }, + { + "epoch": 0.3780668358714044, + "grad_norm": 7.84375, + "learning_rate": 4.972783601331115e-06, + "loss": 2.642, + "step": 3575 + }, + { + "epoch": 0.37817258883248733, + "grad_norm": 9.0, + "learning_rate": 4.9727583943152865e-06, + "loss": 2.5908, + "step": 3576 + }, + { + "epoch": 0.3782783417935702, + "grad_norm": 8.6875, + "learning_rate": 4.972733175695815e-06, + "loss": 2.4014, + "step": 3577 + }, + { + "epoch": 0.3783840947546531, + "grad_norm": 8.0625, + "learning_rate": 4.972707945472817e-06, + "loss": 2.6914, + "step": 3578 + }, + { + "epoch": 0.378489847715736, + "grad_norm": 7.8125, + "learning_rate": 4.972682703646414e-06, + "loss": 2.5846, + "step": 3579 + }, + { + "epoch": 0.37859560067681897, + "grad_norm": 7.78125, + "learning_rate": 4.972657450216721e-06, + "loss": 2.6069, + "step": 3580 + }, + { + "epoch": 0.37870135363790186, + "grad_norm": 8.0, + "learning_rate": 4.9726321851838585e-06, + "loss": 2.5683, + "step": 3581 + }, + { + "epoch": 0.37880710659898476, + "grad_norm": 8.0625, + "learning_rate": 4.972606908547944e-06, + "loss": 2.6982, + "step": 3582 + }, + { + "epoch": 0.3789128595600677, + "grad_norm": 8.25, + "learning_rate": 4.972581620309096e-06, + "loss": 2.464, + "step": 3583 + }, + { + "epoch": 0.3790186125211506, + "grad_norm": 8.375, + "learning_rate": 4.972556320467435e-06, + "loss": 2.7786, + "step": 3584 + }, + { + "epoch": 0.3791243654822335, + "grad_norm": 8.375, + "learning_rate": 4.972531009023077e-06, + "loss": 2.6967, + "step": 3585 + }, + { + "epoch": 0.3792301184433164, + "grad_norm": 7.375, + "learning_rate": 4.972505685976143e-06, + "loss": 2.4808, + "step": 3586 + }, + { + "epoch": 0.37933587140439934, + "grad_norm": 7.65625, + "learning_rate": 4.972480351326752e-06, + "loss": 2.6585, + "step": 3587 + }, + { + "epoch": 0.37944162436548223, + "grad_norm": 7.8125, + "learning_rate": 4.9724550050750196e-06, + "loss": 2.6515, + "step": 3588 + }, + { + "epoch": 0.3795473773265651, + "grad_norm": 8.0, + "learning_rate": 4.9724296472210685e-06, + "loss": 2.5491, + "step": 3589 + }, + { + "epoch": 0.3796531302876481, + "grad_norm": 7.5625, + "learning_rate": 4.972404277765016e-06, + "loss": 2.3778, + "step": 3590 + }, + { + "epoch": 0.37975888324873097, + "grad_norm": 6.96875, + "learning_rate": 4.972378896706982e-06, + "loss": 2.5896, + "step": 3591 + }, + { + "epoch": 0.37986463620981387, + "grad_norm": 7.75, + "learning_rate": 4.972353504047085e-06, + "loss": 2.6368, + "step": 3592 + }, + { + "epoch": 0.37997038917089676, + "grad_norm": 8.1875, + "learning_rate": 4.972328099785443e-06, + "loss": 2.5043, + "step": 3593 + }, + { + "epoch": 0.3800761421319797, + "grad_norm": 7.84375, + "learning_rate": 4.972302683922177e-06, + "loss": 2.6061, + "step": 3594 + }, + { + "epoch": 0.3801818950930626, + "grad_norm": 7.8125, + "learning_rate": 4.972277256457407e-06, + "loss": 2.6281, + "step": 3595 + }, + { + "epoch": 0.3802876480541455, + "grad_norm": 9.125, + "learning_rate": 4.972251817391249e-06, + "loss": 2.4226, + "step": 3596 + }, + { + "epoch": 0.38039340101522845, + "grad_norm": 8.125, + "learning_rate": 4.972226366723825e-06, + "loss": 2.6215, + "step": 3597 + }, + { + "epoch": 0.38049915397631134, + "grad_norm": 8.0625, + "learning_rate": 4.9722009044552534e-06, + "loss": 2.7418, + "step": 3598 + }, + { + "epoch": 0.38060490693739424, + "grad_norm": 7.5, + "learning_rate": 4.972175430585654e-06, + "loss": 2.4898, + "step": 3599 + }, + { + "epoch": 0.38071065989847713, + "grad_norm": 7.53125, + "learning_rate": 4.972149945115147e-06, + "loss": 2.5475, + "step": 3600 + }, + { + "epoch": 0.3808164128595601, + "grad_norm": 8.375, + "learning_rate": 4.972124448043852e-06, + "loss": 2.599, + "step": 3601 + }, + { + "epoch": 0.380922165820643, + "grad_norm": 7.65625, + "learning_rate": 4.972098939371886e-06, + "loss": 2.5519, + "step": 3602 + }, + { + "epoch": 0.38102791878172587, + "grad_norm": 9.25, + "learning_rate": 4.972073419099373e-06, + "loss": 3.0932, + "step": 3603 + }, + { + "epoch": 0.3811336717428088, + "grad_norm": 7.875, + "learning_rate": 4.972047887226429e-06, + "loss": 2.7348, + "step": 3604 + }, + { + "epoch": 0.3812394247038917, + "grad_norm": 9.25, + "learning_rate": 4.972022343753175e-06, + "loss": 2.6198, + "step": 3605 + }, + { + "epoch": 0.3813451776649746, + "grad_norm": 7.6875, + "learning_rate": 4.971996788679733e-06, + "loss": 2.5527, + "step": 3606 + }, + { + "epoch": 0.3814509306260575, + "grad_norm": 7.4375, + "learning_rate": 4.97197122200622e-06, + "loss": 2.5015, + "step": 3607 + }, + { + "epoch": 0.38155668358714045, + "grad_norm": 7.78125, + "learning_rate": 4.971945643732757e-06, + "loss": 2.6985, + "step": 3608 + }, + { + "epoch": 0.38166243654822335, + "grad_norm": 9.0, + "learning_rate": 4.9719200538594636e-06, + "loss": 2.6119, + "step": 3609 + }, + { + "epoch": 0.38176818950930624, + "grad_norm": 7.875, + "learning_rate": 4.971894452386461e-06, + "loss": 2.8844, + "step": 3610 + }, + { + "epoch": 0.3818739424703892, + "grad_norm": 7.65625, + "learning_rate": 4.971868839313868e-06, + "loss": 2.5064, + "step": 3611 + }, + { + "epoch": 0.3819796954314721, + "grad_norm": 8.125, + "learning_rate": 4.971843214641806e-06, + "loss": 2.6155, + "step": 3612 + }, + { + "epoch": 0.382085448392555, + "grad_norm": 8.0, + "learning_rate": 4.971817578370394e-06, + "loss": 2.7188, + "step": 3613 + }, + { + "epoch": 0.3821912013536379, + "grad_norm": 7.15625, + "learning_rate": 4.971791930499755e-06, + "loss": 2.5202, + "step": 3614 + }, + { + "epoch": 0.3822969543147208, + "grad_norm": 7.5, + "learning_rate": 4.971766271030007e-06, + "loss": 2.6925, + "step": 3615 + }, + { + "epoch": 0.3824027072758037, + "grad_norm": 8.5, + "learning_rate": 4.9717405999612695e-06, + "loss": 2.7214, + "step": 3616 + }, + { + "epoch": 0.3825084602368866, + "grad_norm": 7.84375, + "learning_rate": 4.971714917293665e-06, + "loss": 2.4906, + "step": 3617 + }, + { + "epoch": 0.38261421319796957, + "grad_norm": 8.0625, + "learning_rate": 4.971689223027313e-06, + "loss": 2.4346, + "step": 3618 + }, + { + "epoch": 0.38271996615905246, + "grad_norm": 9.5, + "learning_rate": 4.971663517162334e-06, + "loss": 2.5749, + "step": 3619 + }, + { + "epoch": 0.38282571912013535, + "grad_norm": 7.71875, + "learning_rate": 4.97163779969885e-06, + "loss": 2.3888, + "step": 3620 + }, + { + "epoch": 0.38293147208121825, + "grad_norm": 8.1875, + "learning_rate": 4.97161207063698e-06, + "loss": 2.6928, + "step": 3621 + }, + { + "epoch": 0.3830372250423012, + "grad_norm": 7.375, + "learning_rate": 4.971586329976845e-06, + "loss": 2.6503, + "step": 3622 + }, + { + "epoch": 0.3831429780033841, + "grad_norm": 8.5, + "learning_rate": 4.971560577718567e-06, + "loss": 2.6634, + "step": 3623 + }, + { + "epoch": 0.383248730964467, + "grad_norm": 7.625, + "learning_rate": 4.971534813862266e-06, + "loss": 2.5978, + "step": 3624 + }, + { + "epoch": 0.38335448392554994, + "grad_norm": 7.96875, + "learning_rate": 4.971509038408063e-06, + "loss": 2.6396, + "step": 3625 + }, + { + "epoch": 0.38346023688663283, + "grad_norm": 7.65625, + "learning_rate": 4.971483251356078e-06, + "loss": 2.7059, + "step": 3626 + }, + { + "epoch": 0.3835659898477157, + "grad_norm": 7.84375, + "learning_rate": 4.971457452706434e-06, + "loss": 2.3714, + "step": 3627 + }, + { + "epoch": 0.3836717428087986, + "grad_norm": 10.6875, + "learning_rate": 4.97143164245925e-06, + "loss": 2.521, + "step": 3628 + }, + { + "epoch": 0.38377749576988157, + "grad_norm": 8.1875, + "learning_rate": 4.9714058206146485e-06, + "loss": 2.4969, + "step": 3629 + }, + { + "epoch": 0.38388324873096447, + "grad_norm": 9.5625, + "learning_rate": 4.9713799871727506e-06, + "loss": 2.5548, + "step": 3630 + }, + { + "epoch": 0.38398900169204736, + "grad_norm": 8.0, + "learning_rate": 4.971354142133677e-06, + "loss": 2.2668, + "step": 3631 + }, + { + "epoch": 0.3840947546531303, + "grad_norm": 7.5625, + "learning_rate": 4.971328285497549e-06, + "loss": 2.5179, + "step": 3632 + }, + { + "epoch": 0.3842005076142132, + "grad_norm": 7.71875, + "learning_rate": 4.9713024172644885e-06, + "loss": 2.4218, + "step": 3633 + }, + { + "epoch": 0.3843062605752961, + "grad_norm": 8.125, + "learning_rate": 4.971276537434616e-06, + "loss": 2.5892, + "step": 3634 + }, + { + "epoch": 0.384412013536379, + "grad_norm": 7.6875, + "learning_rate": 4.971250646008054e-06, + "loss": 2.7886, + "step": 3635 + }, + { + "epoch": 0.38451776649746194, + "grad_norm": 8.75, + "learning_rate": 4.971224742984922e-06, + "loss": 2.5566, + "step": 3636 + }, + { + "epoch": 0.38462351945854484, + "grad_norm": 7.5625, + "learning_rate": 4.971198828365344e-06, + "loss": 2.5586, + "step": 3637 + }, + { + "epoch": 0.38472927241962773, + "grad_norm": 8.6875, + "learning_rate": 4.971172902149441e-06, + "loss": 2.6916, + "step": 3638 + }, + { + "epoch": 0.3848350253807107, + "grad_norm": 7.40625, + "learning_rate": 4.9711469643373345e-06, + "loss": 2.5771, + "step": 3639 + }, + { + "epoch": 0.3849407783417936, + "grad_norm": 9.0625, + "learning_rate": 4.971121014929145e-06, + "loss": 2.5535, + "step": 3640 + }, + { + "epoch": 0.38504653130287647, + "grad_norm": 7.8125, + "learning_rate": 4.971095053924996e-06, + "loss": 2.6823, + "step": 3641 + }, + { + "epoch": 0.38515228426395937, + "grad_norm": 7.96875, + "learning_rate": 4.971069081325007e-06, + "loss": 2.4887, + "step": 3642 + }, + { + "epoch": 0.3852580372250423, + "grad_norm": 7.3125, + "learning_rate": 4.971043097129303e-06, + "loss": 2.5428, + "step": 3643 + }, + { + "epoch": 0.3853637901861252, + "grad_norm": 7.625, + "learning_rate": 4.971017101338004e-06, + "loss": 2.5581, + "step": 3644 + }, + { + "epoch": 0.3854695431472081, + "grad_norm": 7.84375, + "learning_rate": 4.970991093951232e-06, + "loss": 2.7083, + "step": 3645 + }, + { + "epoch": 0.38557529610829105, + "grad_norm": 7.84375, + "learning_rate": 4.9709650749691095e-06, + "loss": 2.624, + "step": 3646 + }, + { + "epoch": 0.38568104906937395, + "grad_norm": 7.75, + "learning_rate": 4.9709390443917585e-06, + "loss": 2.4955, + "step": 3647 + }, + { + "epoch": 0.38578680203045684, + "grad_norm": 7.4375, + "learning_rate": 4.9709130022193016e-06, + "loss": 2.3558, + "step": 3648 + }, + { + "epoch": 0.38589255499153974, + "grad_norm": 7.59375, + "learning_rate": 4.9708869484518605e-06, + "loss": 2.6447, + "step": 3649 + }, + { + "epoch": 0.3859983079526227, + "grad_norm": 8.3125, + "learning_rate": 4.970860883089557e-06, + "loss": 2.7099, + "step": 3650 + }, + { + "epoch": 0.3861040609137056, + "grad_norm": 7.625, + "learning_rate": 4.970834806132514e-06, + "loss": 2.6612, + "step": 3651 + }, + { + "epoch": 0.3862098138747885, + "grad_norm": 7.75, + "learning_rate": 4.970808717580854e-06, + "loss": 2.558, + "step": 3652 + }, + { + "epoch": 0.3863155668358714, + "grad_norm": 7.40625, + "learning_rate": 4.9707826174347e-06, + "loss": 2.4398, + "step": 3653 + }, + { + "epoch": 0.3864213197969543, + "grad_norm": 9.125, + "learning_rate": 4.970756505694172e-06, + "loss": 2.8253, + "step": 3654 + }, + { + "epoch": 0.3865270727580372, + "grad_norm": 7.4375, + "learning_rate": 4.970730382359395e-06, + "loss": 2.5487, + "step": 3655 + }, + { + "epoch": 0.3866328257191201, + "grad_norm": 7.375, + "learning_rate": 4.970704247430491e-06, + "loss": 2.641, + "step": 3656 + }, + { + "epoch": 0.38673857868020306, + "grad_norm": 7.59375, + "learning_rate": 4.9706781009075825e-06, + "loss": 2.6719, + "step": 3657 + }, + { + "epoch": 0.38684433164128595, + "grad_norm": 8.0, + "learning_rate": 4.970651942790792e-06, + "loss": 2.5377, + "step": 3658 + }, + { + "epoch": 0.38695008460236885, + "grad_norm": 7.9375, + "learning_rate": 4.970625773080242e-06, + "loss": 2.7018, + "step": 3659 + }, + { + "epoch": 0.3870558375634518, + "grad_norm": 8.375, + "learning_rate": 4.970599591776056e-06, + "loss": 2.6729, + "step": 3660 + }, + { + "epoch": 0.3871615905245347, + "grad_norm": 8.25, + "learning_rate": 4.970573398878357e-06, + "loss": 2.7696, + "step": 3661 + }, + { + "epoch": 0.3872673434856176, + "grad_norm": 8.375, + "learning_rate": 4.970547194387267e-06, + "loss": 2.4471, + "step": 3662 + }, + { + "epoch": 0.3873730964467005, + "grad_norm": 7.5625, + "learning_rate": 4.97052097830291e-06, + "loss": 2.5634, + "step": 3663 + }, + { + "epoch": 0.38747884940778343, + "grad_norm": 8.0625, + "learning_rate": 4.9704947506254075e-06, + "loss": 2.4402, + "step": 3664 + }, + { + "epoch": 0.3875846023688663, + "grad_norm": 7.78125, + "learning_rate": 4.970468511354884e-06, + "loss": 2.6591, + "step": 3665 + }, + { + "epoch": 0.3876903553299492, + "grad_norm": 7.96875, + "learning_rate": 4.970442260491462e-06, + "loss": 2.6674, + "step": 3666 + }, + { + "epoch": 0.38779610829103217, + "grad_norm": 7.5, + "learning_rate": 4.970415998035265e-06, + "loss": 2.5424, + "step": 3667 + }, + { + "epoch": 0.38790186125211507, + "grad_norm": 7.3125, + "learning_rate": 4.9703897239864164e-06, + "loss": 2.4894, + "step": 3668 + }, + { + "epoch": 0.38800761421319796, + "grad_norm": 8.1875, + "learning_rate": 4.9703634383450385e-06, + "loss": 2.5476, + "step": 3669 + }, + { + "epoch": 0.38811336717428085, + "grad_norm": 7.375, + "learning_rate": 4.970337141111256e-06, + "loss": 2.4484, + "step": 3670 + }, + { + "epoch": 0.3882191201353638, + "grad_norm": 7.46875, + "learning_rate": 4.970310832285191e-06, + "loss": 2.5846, + "step": 3671 + }, + { + "epoch": 0.3883248730964467, + "grad_norm": 7.75, + "learning_rate": 4.970284511866969e-06, + "loss": 2.6116, + "step": 3672 + }, + { + "epoch": 0.3884306260575296, + "grad_norm": 7.875, + "learning_rate": 4.970258179856711e-06, + "loss": 2.793, + "step": 3673 + }, + { + "epoch": 0.38853637901861254, + "grad_norm": 7.75, + "learning_rate": 4.970231836254541e-06, + "loss": 2.7864, + "step": 3674 + }, + { + "epoch": 0.38864213197969544, + "grad_norm": 7.625, + "learning_rate": 4.970205481060583e-06, + "loss": 2.4566, + "step": 3675 + }, + { + "epoch": 0.38874788494077833, + "grad_norm": 8.625, + "learning_rate": 4.9701791142749635e-06, + "loss": 2.9032, + "step": 3676 + }, + { + "epoch": 0.3888536379018612, + "grad_norm": 7.5625, + "learning_rate": 4.970152735897802e-06, + "loss": 2.7482, + "step": 3677 + }, + { + "epoch": 0.3889593908629442, + "grad_norm": 8.1875, + "learning_rate": 4.970126345929223e-06, + "loss": 2.2572, + "step": 3678 + }, + { + "epoch": 0.38906514382402707, + "grad_norm": 7.6875, + "learning_rate": 4.970099944369353e-06, + "loss": 2.3555, + "step": 3679 + }, + { + "epoch": 0.38917089678510997, + "grad_norm": 8.1875, + "learning_rate": 4.970073531218313e-06, + "loss": 2.3157, + "step": 3680 + }, + { + "epoch": 0.3892766497461929, + "grad_norm": 7.625, + "learning_rate": 4.970047106476229e-06, + "loss": 2.5922, + "step": 3681 + }, + { + "epoch": 0.3893824027072758, + "grad_norm": 8.125, + "learning_rate": 4.9700206701432245e-06, + "loss": 2.6924, + "step": 3682 + }, + { + "epoch": 0.3894881556683587, + "grad_norm": 7.8125, + "learning_rate": 4.969994222219422e-06, + "loss": 2.5815, + "step": 3683 + }, + { + "epoch": 0.3895939086294416, + "grad_norm": 7.6875, + "learning_rate": 4.969967762704948e-06, + "loss": 2.6638, + "step": 3684 + }, + { + "epoch": 0.38969966159052455, + "grad_norm": 7.59375, + "learning_rate": 4.969941291599924e-06, + "loss": 2.5343, + "step": 3685 + }, + { + "epoch": 0.38980541455160744, + "grad_norm": 7.9375, + "learning_rate": 4.969914808904477e-06, + "loss": 2.6395, + "step": 3686 + }, + { + "epoch": 0.38991116751269034, + "grad_norm": 13.875, + "learning_rate": 4.969888314618729e-06, + "loss": 2.3276, + "step": 3687 + }, + { + "epoch": 0.3900169204737733, + "grad_norm": 7.75, + "learning_rate": 4.969861808742806e-06, + "loss": 2.5592, + "step": 3688 + }, + { + "epoch": 0.3901226734348562, + "grad_norm": 7.96875, + "learning_rate": 4.969835291276832e-06, + "loss": 2.6199, + "step": 3689 + }, + { + "epoch": 0.3902284263959391, + "grad_norm": 8.8125, + "learning_rate": 4.96980876222093e-06, + "loss": 2.7776, + "step": 3690 + }, + { + "epoch": 0.39033417935702197, + "grad_norm": 8.0625, + "learning_rate": 4.969782221575227e-06, + "loss": 2.6728, + "step": 3691 + }, + { + "epoch": 0.3904399323181049, + "grad_norm": 7.65625, + "learning_rate": 4.969755669339845e-06, + "loss": 2.656, + "step": 3692 + }, + { + "epoch": 0.3905456852791878, + "grad_norm": 8.6875, + "learning_rate": 4.96972910551491e-06, + "loss": 2.753, + "step": 3693 + }, + { + "epoch": 0.3906514382402707, + "grad_norm": 7.3125, + "learning_rate": 4.969702530100547e-06, + "loss": 2.3635, + "step": 3694 + }, + { + "epoch": 0.39075719120135366, + "grad_norm": 7.5625, + "learning_rate": 4.969675943096879e-06, + "loss": 2.5118, + "step": 3695 + }, + { + "epoch": 0.39086294416243655, + "grad_norm": 7.8125, + "learning_rate": 4.969649344504033e-06, + "loss": 2.4271, + "step": 3696 + }, + { + "epoch": 0.39096869712351945, + "grad_norm": 8.1875, + "learning_rate": 4.969622734322131e-06, + "loss": 2.5532, + "step": 3697 + }, + { + "epoch": 0.39107445008460234, + "grad_norm": 8.375, + "learning_rate": 4.969596112551302e-06, + "loss": 2.5821, + "step": 3698 + }, + { + "epoch": 0.3911802030456853, + "grad_norm": 7.625, + "learning_rate": 4.969569479191666e-06, + "loss": 2.564, + "step": 3699 + }, + { + "epoch": 0.3912859560067682, + "grad_norm": 9.1875, + "learning_rate": 4.9695428342433515e-06, + "loss": 2.6383, + "step": 3700 + }, + { + "epoch": 0.3913917089678511, + "grad_norm": 7.5, + "learning_rate": 4.969516177706482e-06, + "loss": 2.4487, + "step": 3701 + }, + { + "epoch": 0.39149746192893403, + "grad_norm": 7.96875, + "learning_rate": 4.9694895095811844e-06, + "loss": 2.5716, + "step": 3702 + }, + { + "epoch": 0.3916032148900169, + "grad_norm": 7.75, + "learning_rate": 4.969462829867582e-06, + "loss": 2.5742, + "step": 3703 + }, + { + "epoch": 0.3917089678510998, + "grad_norm": 8.5, + "learning_rate": 4.9694361385658e-06, + "loss": 2.742, + "step": 3704 + }, + { + "epoch": 0.3918147208121827, + "grad_norm": 8.25, + "learning_rate": 4.969409435675965e-06, + "loss": 2.7599, + "step": 3705 + }, + { + "epoch": 0.39192047377326567, + "grad_norm": 7.8125, + "learning_rate": 4.9693827211982e-06, + "loss": 2.5577, + "step": 3706 + }, + { + "epoch": 0.39202622673434856, + "grad_norm": 8.1875, + "learning_rate": 4.9693559951326334e-06, + "loss": 2.5533, + "step": 3707 + }, + { + "epoch": 0.39213197969543145, + "grad_norm": 7.3125, + "learning_rate": 4.969329257479388e-06, + "loss": 2.5781, + "step": 3708 + }, + { + "epoch": 0.3922377326565144, + "grad_norm": 7.625, + "learning_rate": 4.969302508238591e-06, + "loss": 2.5848, + "step": 3709 + }, + { + "epoch": 0.3923434856175973, + "grad_norm": 7.75, + "learning_rate": 4.969275747410367e-06, + "loss": 2.6005, + "step": 3710 + }, + { + "epoch": 0.3924492385786802, + "grad_norm": 7.75, + "learning_rate": 4.969248974994842e-06, + "loss": 2.6146, + "step": 3711 + }, + { + "epoch": 0.3925549915397631, + "grad_norm": 8.4375, + "learning_rate": 4.9692221909921416e-06, + "loss": 2.6442, + "step": 3712 + }, + { + "epoch": 0.39266074450084604, + "grad_norm": 8.0, + "learning_rate": 4.96919539540239e-06, + "loss": 2.5374, + "step": 3713 + }, + { + "epoch": 0.39276649746192893, + "grad_norm": 7.59375, + "learning_rate": 4.969168588225715e-06, + "loss": 2.5749, + "step": 3714 + }, + { + "epoch": 0.3928722504230118, + "grad_norm": 7.90625, + "learning_rate": 4.969141769462243e-06, + "loss": 2.3755, + "step": 3715 + }, + { + "epoch": 0.3929780033840948, + "grad_norm": 8.5625, + "learning_rate": 4.969114939112096e-06, + "loss": 2.7018, + "step": 3716 + }, + { + "epoch": 0.39308375634517767, + "grad_norm": 7.75, + "learning_rate": 4.969088097175404e-06, + "loss": 2.6808, + "step": 3717 + }, + { + "epoch": 0.39318950930626057, + "grad_norm": 7.875, + "learning_rate": 4.969061243652291e-06, + "loss": 2.6546, + "step": 3718 + }, + { + "epoch": 0.39329526226734346, + "grad_norm": 8.125, + "learning_rate": 4.969034378542884e-06, + "loss": 2.7029, + "step": 3719 + }, + { + "epoch": 0.3934010152284264, + "grad_norm": 7.90625, + "learning_rate": 4.9690075018473084e-06, + "loss": 2.6768, + "step": 3720 + }, + { + "epoch": 0.3935067681895093, + "grad_norm": 7.09375, + "learning_rate": 4.96898061356569e-06, + "loss": 2.534, + "step": 3721 + }, + { + "epoch": 0.3936125211505922, + "grad_norm": 7.625, + "learning_rate": 4.968953713698156e-06, + "loss": 2.6258, + "step": 3722 + }, + { + "epoch": 0.39371827411167515, + "grad_norm": 7.65625, + "learning_rate": 4.968926802244831e-06, + "loss": 2.6983, + "step": 3723 + }, + { + "epoch": 0.39382402707275804, + "grad_norm": 8.6875, + "learning_rate": 4.968899879205843e-06, + "loss": 2.8747, + "step": 3724 + }, + { + "epoch": 0.39392978003384094, + "grad_norm": 9.4375, + "learning_rate": 4.968872944581316e-06, + "loss": 2.6009, + "step": 3725 + }, + { + "epoch": 0.39403553299492383, + "grad_norm": 7.84375, + "learning_rate": 4.96884599837138e-06, + "loss": 2.7197, + "step": 3726 + }, + { + "epoch": 0.3941412859560068, + "grad_norm": 7.46875, + "learning_rate": 4.968819040576159e-06, + "loss": 2.6418, + "step": 3727 + }, + { + "epoch": 0.3942470389170897, + "grad_norm": 8.1875, + "learning_rate": 4.968792071195779e-06, + "loss": 2.5544, + "step": 3728 + }, + { + "epoch": 0.39435279187817257, + "grad_norm": 8.375, + "learning_rate": 4.968765090230368e-06, + "loss": 2.5223, + "step": 3729 + }, + { + "epoch": 0.3944585448392555, + "grad_norm": 7.34375, + "learning_rate": 4.968738097680053e-06, + "loss": 2.6729, + "step": 3730 + }, + { + "epoch": 0.3945642978003384, + "grad_norm": 7.71875, + "learning_rate": 4.968711093544959e-06, + "loss": 2.6748, + "step": 3731 + }, + { + "epoch": 0.3946700507614213, + "grad_norm": 7.6875, + "learning_rate": 4.9686840778252135e-06, + "loss": 2.5381, + "step": 3732 + }, + { + "epoch": 0.3947758037225042, + "grad_norm": 7.875, + "learning_rate": 4.968657050520944e-06, + "loss": 2.6038, + "step": 3733 + }, + { + "epoch": 0.39488155668358715, + "grad_norm": 8.0, + "learning_rate": 4.968630011632276e-06, + "loss": 2.627, + "step": 3734 + }, + { + "epoch": 0.39498730964467005, + "grad_norm": 9.0625, + "learning_rate": 4.9686029611593364e-06, + "loss": 2.5817, + "step": 3735 + }, + { + "epoch": 0.39509306260575294, + "grad_norm": 8.0, + "learning_rate": 4.968575899102254e-06, + "loss": 2.7251, + "step": 3736 + }, + { + "epoch": 0.3951988155668359, + "grad_norm": 7.90625, + "learning_rate": 4.968548825461154e-06, + "loss": 2.5712, + "step": 3737 + }, + { + "epoch": 0.3953045685279188, + "grad_norm": 7.96875, + "learning_rate": 4.968521740236163e-06, + "loss": 2.6057, + "step": 3738 + }, + { + "epoch": 0.3954103214890017, + "grad_norm": 8.1875, + "learning_rate": 4.96849464342741e-06, + "loss": 2.2718, + "step": 3739 + }, + { + "epoch": 0.3955160744500846, + "grad_norm": 8.1875, + "learning_rate": 4.968467535035021e-06, + "loss": 2.2673, + "step": 3740 + }, + { + "epoch": 0.3956218274111675, + "grad_norm": 7.09375, + "learning_rate": 4.968440415059124e-06, + "loss": 2.4668, + "step": 3741 + }, + { + "epoch": 0.3957275803722504, + "grad_norm": 7.375, + "learning_rate": 4.968413283499846e-06, + "loss": 2.6246, + "step": 3742 + }, + { + "epoch": 0.3958333333333333, + "grad_norm": 7.78125, + "learning_rate": 4.968386140357312e-06, + "loss": 2.5791, + "step": 3743 + }, + { + "epoch": 0.39593908629441626, + "grad_norm": 8.25, + "learning_rate": 4.968358985631653e-06, + "loss": 2.5949, + "step": 3744 + }, + { + "epoch": 0.39604483925549916, + "grad_norm": 7.5, + "learning_rate": 4.968331819322995e-06, + "loss": 2.154, + "step": 3745 + }, + { + "epoch": 0.39615059221658205, + "grad_norm": 8.1875, + "learning_rate": 4.968304641431464e-06, + "loss": 2.5858, + "step": 3746 + }, + { + "epoch": 0.39625634517766495, + "grad_norm": 7.75, + "learning_rate": 4.9682774519571896e-06, + "loss": 2.7446, + "step": 3747 + }, + { + "epoch": 0.3963620981387479, + "grad_norm": 7.6875, + "learning_rate": 4.968250250900299e-06, + "loss": 2.6068, + "step": 3748 + }, + { + "epoch": 0.3964678510998308, + "grad_norm": 7.5, + "learning_rate": 4.968223038260918e-06, + "loss": 2.6205, + "step": 3749 + }, + { + "epoch": 0.3965736040609137, + "grad_norm": 8.5, + "learning_rate": 4.968195814039177e-06, + "loss": 2.5831, + "step": 3750 + }, + { + "epoch": 0.39667935702199664, + "grad_norm": 7.90625, + "learning_rate": 4.968168578235203e-06, + "loss": 2.7323, + "step": 3751 + }, + { + "epoch": 0.39678510998307953, + "grad_norm": 7.53125, + "learning_rate": 4.968141330849122e-06, + "loss": 2.3925, + "step": 3752 + }, + { + "epoch": 0.3968908629441624, + "grad_norm": 8.75, + "learning_rate": 4.968114071881064e-06, + "loss": 2.9327, + "step": 3753 + }, + { + "epoch": 0.3969966159052453, + "grad_norm": 7.5, + "learning_rate": 4.968086801331156e-06, + "loss": 2.4364, + "step": 3754 + }, + { + "epoch": 0.39710236886632827, + "grad_norm": 8.1875, + "learning_rate": 4.9680595191995255e-06, + "loss": 2.48, + "step": 3755 + }, + { + "epoch": 0.39720812182741116, + "grad_norm": 9.625, + "learning_rate": 4.968032225486301e-06, + "loss": 2.8705, + "step": 3756 + }, + { + "epoch": 0.39731387478849406, + "grad_norm": 7.625, + "learning_rate": 4.968004920191611e-06, + "loss": 2.51, + "step": 3757 + }, + { + "epoch": 0.397419627749577, + "grad_norm": 7.59375, + "learning_rate": 4.967977603315583e-06, + "loss": 2.6521, + "step": 3758 + }, + { + "epoch": 0.3975253807106599, + "grad_norm": 8.1875, + "learning_rate": 4.967950274858346e-06, + "loss": 2.5353, + "step": 3759 + }, + { + "epoch": 0.3976311336717428, + "grad_norm": 7.125, + "learning_rate": 4.967922934820028e-06, + "loss": 2.4911, + "step": 3760 + }, + { + "epoch": 0.3977368866328257, + "grad_norm": 8.25, + "learning_rate": 4.967895583200757e-06, + "loss": 2.4245, + "step": 3761 + }, + { + "epoch": 0.39784263959390864, + "grad_norm": 7.875, + "learning_rate": 4.9678682200006605e-06, + "loss": 2.497, + "step": 3762 + }, + { + "epoch": 0.39794839255499154, + "grad_norm": 10.25, + "learning_rate": 4.967840845219868e-06, + "loss": 2.4246, + "step": 3763 + }, + { + "epoch": 0.39805414551607443, + "grad_norm": 8.25, + "learning_rate": 4.967813458858508e-06, + "loss": 2.3329, + "step": 3764 + }, + { + "epoch": 0.3981598984771574, + "grad_norm": 8.0625, + "learning_rate": 4.967786060916709e-06, + "loss": 2.7827, + "step": 3765 + }, + { + "epoch": 0.3982656514382403, + "grad_norm": 7.9375, + "learning_rate": 4.967758651394599e-06, + "loss": 2.693, + "step": 3766 + }, + { + "epoch": 0.39837140439932317, + "grad_norm": 9.125, + "learning_rate": 4.967731230292307e-06, + "loss": 2.6102, + "step": 3767 + }, + { + "epoch": 0.39847715736040606, + "grad_norm": 8.0625, + "learning_rate": 4.967703797609961e-06, + "loss": 2.6069, + "step": 3768 + }, + { + "epoch": 0.398582910321489, + "grad_norm": 7.78125, + "learning_rate": 4.967676353347691e-06, + "loss": 2.4039, + "step": 3769 + }, + { + "epoch": 0.3986886632825719, + "grad_norm": 7.78125, + "learning_rate": 4.967648897505625e-06, + "loss": 2.6817, + "step": 3770 + }, + { + "epoch": 0.3987944162436548, + "grad_norm": 7.65625, + "learning_rate": 4.967621430083892e-06, + "loss": 2.5689, + "step": 3771 + }, + { + "epoch": 0.39890016920473775, + "grad_norm": 8.0, + "learning_rate": 4.967593951082621e-06, + "loss": 2.8882, + "step": 3772 + }, + { + "epoch": 0.39900592216582065, + "grad_norm": 8.25, + "learning_rate": 4.96756646050194e-06, + "loss": 2.5816, + "step": 3773 + }, + { + "epoch": 0.39911167512690354, + "grad_norm": 8.5625, + "learning_rate": 4.96753895834198e-06, + "loss": 2.5379, + "step": 3774 + }, + { + "epoch": 0.39921742808798644, + "grad_norm": 7.71875, + "learning_rate": 4.967511444602868e-06, + "loss": 2.5511, + "step": 3775 + }, + { + "epoch": 0.3993231810490694, + "grad_norm": 7.90625, + "learning_rate": 4.9674839192847345e-06, + "loss": 2.5578, + "step": 3776 + }, + { + "epoch": 0.3994289340101523, + "grad_norm": 8.8125, + "learning_rate": 4.967456382387707e-06, + "loss": 2.4128, + "step": 3777 + }, + { + "epoch": 0.3995346869712352, + "grad_norm": 7.84375, + "learning_rate": 4.967428833911917e-06, + "loss": 2.5956, + "step": 3778 + }, + { + "epoch": 0.3996404399323181, + "grad_norm": 7.75, + "learning_rate": 4.967401273857492e-06, + "loss": 2.6154, + "step": 3779 + }, + { + "epoch": 0.399746192893401, + "grad_norm": 8.625, + "learning_rate": 4.967373702224563e-06, + "loss": 2.6645, + "step": 3780 + }, + { + "epoch": 0.3998519458544839, + "grad_norm": 7.375, + "learning_rate": 4.967346119013257e-06, + "loss": 2.5687, + "step": 3781 + }, + { + "epoch": 0.3999576988155668, + "grad_norm": 8.1875, + "learning_rate": 4.967318524223705e-06, + "loss": 2.5758, + "step": 3782 + }, + { + "epoch": 0.40006345177664976, + "grad_norm": 7.59375, + "learning_rate": 4.967290917856037e-06, + "loss": 2.4124, + "step": 3783 + }, + { + "epoch": 0.40016920473773265, + "grad_norm": 7.21875, + "learning_rate": 4.9672632999103815e-06, + "loss": 2.5336, + "step": 3784 + }, + { + "epoch": 0.40027495769881555, + "grad_norm": 7.90625, + "learning_rate": 4.967235670386868e-06, + "loss": 2.7985, + "step": 3785 + }, + { + "epoch": 0.4003807106598985, + "grad_norm": 7.34375, + "learning_rate": 4.967208029285627e-06, + "loss": 2.444, + "step": 3786 + }, + { + "epoch": 0.4004864636209814, + "grad_norm": 10.625, + "learning_rate": 4.967180376606789e-06, + "loss": 2.5385, + "step": 3787 + }, + { + "epoch": 0.4005922165820643, + "grad_norm": 7.8125, + "learning_rate": 4.96715271235048e-06, + "loss": 2.5692, + "step": 3788 + }, + { + "epoch": 0.4006979695431472, + "grad_norm": 7.78125, + "learning_rate": 4.967125036516834e-06, + "loss": 2.5142, + "step": 3789 + }, + { + "epoch": 0.40080372250423013, + "grad_norm": 8.25, + "learning_rate": 4.967097349105979e-06, + "loss": 2.5994, + "step": 3790 + }, + { + "epoch": 0.400909475465313, + "grad_norm": 7.5, + "learning_rate": 4.967069650118045e-06, + "loss": 2.5429, + "step": 3791 + }, + { + "epoch": 0.4010152284263959, + "grad_norm": 7.75, + "learning_rate": 4.9670419395531615e-06, + "loss": 2.6138, + "step": 3792 + }, + { + "epoch": 0.40112098138747887, + "grad_norm": 8.3125, + "learning_rate": 4.96701421741146e-06, + "loss": 2.7651, + "step": 3793 + }, + { + "epoch": 0.40122673434856176, + "grad_norm": 8.4375, + "learning_rate": 4.966986483693069e-06, + "loss": 2.5448, + "step": 3794 + }, + { + "epoch": 0.40133248730964466, + "grad_norm": 7.75, + "learning_rate": 4.96695873839812e-06, + "loss": 2.8168, + "step": 3795 + }, + { + "epoch": 0.40143824027072755, + "grad_norm": 8.125, + "learning_rate": 4.966930981526743e-06, + "loss": 2.6995, + "step": 3796 + }, + { + "epoch": 0.4015439932318105, + "grad_norm": 8.0, + "learning_rate": 4.9669032130790675e-06, + "loss": 2.7869, + "step": 3797 + }, + { + "epoch": 0.4016497461928934, + "grad_norm": 7.875, + "learning_rate": 4.966875433055224e-06, + "loss": 2.7218, + "step": 3798 + }, + { + "epoch": 0.4017554991539763, + "grad_norm": 7.75, + "learning_rate": 4.966847641455342e-06, + "loss": 2.7686, + "step": 3799 + }, + { + "epoch": 0.40186125211505924, + "grad_norm": 7.75, + "learning_rate": 4.9668198382795545e-06, + "loss": 2.6213, + "step": 3800 + }, + { + "epoch": 0.40196700507614214, + "grad_norm": 8.1875, + "learning_rate": 4.9667920235279895e-06, + "loss": 2.5918, + "step": 3801 + }, + { + "epoch": 0.40207275803722503, + "grad_norm": 8.0625, + "learning_rate": 4.966764197200779e-06, + "loss": 2.5381, + "step": 3802 + }, + { + "epoch": 0.4021785109983079, + "grad_norm": 8.25, + "learning_rate": 4.966736359298053e-06, + "loss": 2.5721, + "step": 3803 + }, + { + "epoch": 0.4022842639593909, + "grad_norm": 7.75, + "learning_rate": 4.966708509819942e-06, + "loss": 2.4661, + "step": 3804 + }, + { + "epoch": 0.40239001692047377, + "grad_norm": 7.40625, + "learning_rate": 4.966680648766576e-06, + "loss": 2.6567, + "step": 3805 + }, + { + "epoch": 0.40249576988155666, + "grad_norm": 7.84375, + "learning_rate": 4.966652776138088e-06, + "loss": 2.5782, + "step": 3806 + }, + { + "epoch": 0.4026015228426396, + "grad_norm": 11.0, + "learning_rate": 4.966624891934606e-06, + "loss": 2.7352, + "step": 3807 + }, + { + "epoch": 0.4027072758037225, + "grad_norm": 8.25, + "learning_rate": 4.966596996156263e-06, + "loss": 2.6162, + "step": 3808 + }, + { + "epoch": 0.4028130287648054, + "grad_norm": 7.8125, + "learning_rate": 4.9665690888031885e-06, + "loss": 2.3169, + "step": 3809 + }, + { + "epoch": 0.4029187817258883, + "grad_norm": 7.875, + "learning_rate": 4.966541169875515e-06, + "loss": 2.5995, + "step": 3810 + }, + { + "epoch": 0.40302453468697125, + "grad_norm": 8.5, + "learning_rate": 4.966513239373371e-06, + "loss": 2.4543, + "step": 3811 + }, + { + "epoch": 0.40313028764805414, + "grad_norm": 8.125, + "learning_rate": 4.96648529729689e-06, + "loss": 2.688, + "step": 3812 + }, + { + "epoch": 0.40323604060913704, + "grad_norm": 7.25, + "learning_rate": 4.966457343646203e-06, + "loss": 2.5566, + "step": 3813 + }, + { + "epoch": 0.40334179357022, + "grad_norm": 9.3125, + "learning_rate": 4.9664293784214405e-06, + "loss": 2.5229, + "step": 3814 + }, + { + "epoch": 0.4034475465313029, + "grad_norm": 8.25, + "learning_rate": 4.966401401622733e-06, + "loss": 2.7488, + "step": 3815 + }, + { + "epoch": 0.4035532994923858, + "grad_norm": 7.59375, + "learning_rate": 4.9663734132502114e-06, + "loss": 2.5467, + "step": 3816 + }, + { + "epoch": 0.40365905245346867, + "grad_norm": 7.625, + "learning_rate": 4.966345413304009e-06, + "loss": 2.4515, + "step": 3817 + }, + { + "epoch": 0.4037648054145516, + "grad_norm": 7.78125, + "learning_rate": 4.966317401784256e-06, + "loss": 2.7017, + "step": 3818 + }, + { + "epoch": 0.4038705583756345, + "grad_norm": 8.5625, + "learning_rate": 4.9662893786910844e-06, + "loss": 2.6709, + "step": 3819 + }, + { + "epoch": 0.4039763113367174, + "grad_norm": 8.0625, + "learning_rate": 4.966261344024626e-06, + "loss": 2.7231, + "step": 3820 + }, + { + "epoch": 0.40408206429780036, + "grad_norm": 7.625, + "learning_rate": 4.9662332977850115e-06, + "loss": 2.554, + "step": 3821 + }, + { + "epoch": 0.40418781725888325, + "grad_norm": 7.84375, + "learning_rate": 4.966205239972373e-06, + "loss": 2.6394, + "step": 3822 + }, + { + "epoch": 0.40429357021996615, + "grad_norm": 8.5, + "learning_rate": 4.966177170586841e-06, + "loss": 2.9546, + "step": 3823 + }, + { + "epoch": 0.40439932318104904, + "grad_norm": 7.6875, + "learning_rate": 4.9661490896285495e-06, + "loss": 2.4904, + "step": 3824 + }, + { + "epoch": 0.404505076142132, + "grad_norm": 7.71875, + "learning_rate": 4.966120997097628e-06, + "loss": 2.4847, + "step": 3825 + }, + { + "epoch": 0.4046108291032149, + "grad_norm": 7.25, + "learning_rate": 4.966092892994209e-06, + "loss": 2.691, + "step": 3826 + }, + { + "epoch": 0.4047165820642978, + "grad_norm": 8.375, + "learning_rate": 4.966064777318425e-06, + "loss": 2.7881, + "step": 3827 + }, + { + "epoch": 0.40482233502538073, + "grad_norm": 8.125, + "learning_rate": 4.966036650070408e-06, + "loss": 2.6378, + "step": 3828 + }, + { + "epoch": 0.4049280879864636, + "grad_norm": 8.0, + "learning_rate": 4.9660085112502885e-06, + "loss": 2.7255, + "step": 3829 + }, + { + "epoch": 0.4050338409475465, + "grad_norm": 7.53125, + "learning_rate": 4.965980360858201e-06, + "loss": 2.6472, + "step": 3830 + }, + { + "epoch": 0.4051395939086294, + "grad_norm": 8.1875, + "learning_rate": 4.965952198894275e-06, + "loss": 2.7968, + "step": 3831 + }, + { + "epoch": 0.40524534686971236, + "grad_norm": 7.8125, + "learning_rate": 4.965924025358645e-06, + "loss": 2.5617, + "step": 3832 + }, + { + "epoch": 0.40535109983079526, + "grad_norm": 7.625, + "learning_rate": 4.965895840251441e-06, + "loss": 2.5918, + "step": 3833 + }, + { + "epoch": 0.40545685279187815, + "grad_norm": 7.65625, + "learning_rate": 4.965867643572797e-06, + "loss": 2.2177, + "step": 3834 + }, + { + "epoch": 0.4055626057529611, + "grad_norm": 7.875, + "learning_rate": 4.965839435322845e-06, + "loss": 2.7099, + "step": 3835 + }, + { + "epoch": 0.405668358714044, + "grad_norm": 9.0625, + "learning_rate": 4.965811215501717e-06, + "loss": 2.7194, + "step": 3836 + }, + { + "epoch": 0.4057741116751269, + "grad_norm": 7.53125, + "learning_rate": 4.965782984109544e-06, + "loss": 2.6349, + "step": 3837 + }, + { + "epoch": 0.4058798646362098, + "grad_norm": 7.8125, + "learning_rate": 4.965754741146461e-06, + "loss": 2.5984, + "step": 3838 + }, + { + "epoch": 0.40598561759729274, + "grad_norm": 7.9375, + "learning_rate": 4.9657264866126e-06, + "loss": 2.4918, + "step": 3839 + }, + { + "epoch": 0.40609137055837563, + "grad_norm": 8.25, + "learning_rate": 4.965698220508092e-06, + "loss": 2.5511, + "step": 3840 + }, + { + "epoch": 0.4061971235194585, + "grad_norm": 7.46875, + "learning_rate": 4.9656699428330715e-06, + "loss": 2.641, + "step": 3841 + }, + { + "epoch": 0.4063028764805415, + "grad_norm": 9.0625, + "learning_rate": 4.9656416535876695e-06, + "loss": 2.8265, + "step": 3842 + }, + { + "epoch": 0.40640862944162437, + "grad_norm": 8.375, + "learning_rate": 4.965613352772021e-06, + "loss": 2.8918, + "step": 3843 + }, + { + "epoch": 0.40651438240270726, + "grad_norm": 7.84375, + "learning_rate": 4.965585040386256e-06, + "loss": 2.5404, + "step": 3844 + }, + { + "epoch": 0.4066201353637902, + "grad_norm": 7.75, + "learning_rate": 4.965556716430509e-06, + "loss": 2.3565, + "step": 3845 + }, + { + "epoch": 0.4067258883248731, + "grad_norm": 7.59375, + "learning_rate": 4.9655283809049134e-06, + "loss": 2.5757, + "step": 3846 + }, + { + "epoch": 0.406831641285956, + "grad_norm": 7.5625, + "learning_rate": 4.965500033809601e-06, + "loss": 2.607, + "step": 3847 + }, + { + "epoch": 0.4069373942470389, + "grad_norm": 10.375, + "learning_rate": 4.965471675144705e-06, + "loss": 2.5487, + "step": 3848 + }, + { + "epoch": 0.40704314720812185, + "grad_norm": 8.0, + "learning_rate": 4.96544330491036e-06, + "loss": 2.6411, + "step": 3849 + }, + { + "epoch": 0.40714890016920474, + "grad_norm": 7.34375, + "learning_rate": 4.965414923106697e-06, + "loss": 2.5913, + "step": 3850 + }, + { + "epoch": 0.40725465313028764, + "grad_norm": 7.5, + "learning_rate": 4.9653865297338495e-06, + "loss": 2.6164, + "step": 3851 + }, + { + "epoch": 0.4073604060913706, + "grad_norm": 7.53125, + "learning_rate": 4.965358124791952e-06, + "loss": 2.6311, + "step": 3852 + }, + { + "epoch": 0.4074661590524535, + "grad_norm": 7.59375, + "learning_rate": 4.965329708281138e-06, + "loss": 2.6959, + "step": 3853 + }, + { + "epoch": 0.4075719120135364, + "grad_norm": 7.9375, + "learning_rate": 4.965301280201539e-06, + "loss": 2.6107, + "step": 3854 + }, + { + "epoch": 0.40767766497461927, + "grad_norm": 7.15625, + "learning_rate": 4.965272840553289e-06, + "loss": 2.5911, + "step": 3855 + }, + { + "epoch": 0.4077834179357022, + "grad_norm": 8.625, + "learning_rate": 4.965244389336522e-06, + "loss": 2.6555, + "step": 3856 + }, + { + "epoch": 0.4078891708967851, + "grad_norm": 8.125, + "learning_rate": 4.965215926551373e-06, + "loss": 2.5804, + "step": 3857 + }, + { + "epoch": 0.407994923857868, + "grad_norm": 7.8125, + "learning_rate": 4.965187452197971e-06, + "loss": 2.569, + "step": 3858 + }, + { + "epoch": 0.40810067681895096, + "grad_norm": 7.75, + "learning_rate": 4.9651589662764545e-06, + "loss": 2.7628, + "step": 3859 + }, + { + "epoch": 0.40820642978003385, + "grad_norm": 8.0625, + "learning_rate": 4.965130468786955e-06, + "loss": 2.4841, + "step": 3860 + }, + { + "epoch": 0.40831218274111675, + "grad_norm": 7.5, + "learning_rate": 4.9651019597296055e-06, + "loss": 2.5529, + "step": 3861 + }, + { + "epoch": 0.40841793570219964, + "grad_norm": 7.59375, + "learning_rate": 4.965073439104542e-06, + "loss": 2.3971, + "step": 3862 + }, + { + "epoch": 0.4085236886632826, + "grad_norm": 7.75, + "learning_rate": 4.965044906911897e-06, + "loss": 2.6582, + "step": 3863 + }, + { + "epoch": 0.4086294416243655, + "grad_norm": 8.125, + "learning_rate": 4.965016363151804e-06, + "loss": 2.7721, + "step": 3864 + }, + { + "epoch": 0.4087351945854484, + "grad_norm": 7.53125, + "learning_rate": 4.964987807824396e-06, + "loss": 2.4996, + "step": 3865 + }, + { + "epoch": 0.40884094754653133, + "grad_norm": 7.96875, + "learning_rate": 4.9649592409298095e-06, + "loss": 2.6154, + "step": 3866 + }, + { + "epoch": 0.4089467005076142, + "grad_norm": 7.3125, + "learning_rate": 4.964930662468177e-06, + "loss": 2.3538, + "step": 3867 + }, + { + "epoch": 0.4090524534686971, + "grad_norm": 7.40625, + "learning_rate": 4.964902072439634e-06, + "loss": 2.7179, + "step": 3868 + }, + { + "epoch": 0.40915820642978, + "grad_norm": 7.84375, + "learning_rate": 4.964873470844313e-06, + "loss": 2.5405, + "step": 3869 + }, + { + "epoch": 0.40926395939086296, + "grad_norm": 8.25, + "learning_rate": 4.964844857682349e-06, + "loss": 2.7328, + "step": 3870 + }, + { + "epoch": 0.40936971235194586, + "grad_norm": 8.3125, + "learning_rate": 4.964816232953876e-06, + "loss": 2.6663, + "step": 3871 + }, + { + "epoch": 0.40947546531302875, + "grad_norm": 7.65625, + "learning_rate": 4.9647875966590295e-06, + "loss": 2.5797, + "step": 3872 + }, + { + "epoch": 0.4095812182741117, + "grad_norm": 7.40625, + "learning_rate": 4.964758948797942e-06, + "loss": 2.645, + "step": 3873 + }, + { + "epoch": 0.4096869712351946, + "grad_norm": 7.6875, + "learning_rate": 4.964730289370748e-06, + "loss": 2.7065, + "step": 3874 + }, + { + "epoch": 0.4097927241962775, + "grad_norm": 7.59375, + "learning_rate": 4.964701618377584e-06, + "loss": 2.4807, + "step": 3875 + }, + { + "epoch": 0.4098984771573604, + "grad_norm": 7.75, + "learning_rate": 4.964672935818583e-06, + "loss": 2.6484, + "step": 3876 + }, + { + "epoch": 0.41000423011844334, + "grad_norm": 7.65625, + "learning_rate": 4.96464424169388e-06, + "loss": 2.618, + "step": 3877 + }, + { + "epoch": 0.41010998307952623, + "grad_norm": 7.21875, + "learning_rate": 4.9646155360036095e-06, + "loss": 2.5163, + "step": 3878 + }, + { + "epoch": 0.4102157360406091, + "grad_norm": 8.0, + "learning_rate": 4.9645868187479065e-06, + "loss": 2.7526, + "step": 3879 + }, + { + "epoch": 0.4103214890016921, + "grad_norm": 7.65625, + "learning_rate": 4.964558089926905e-06, + "loss": 2.679, + "step": 3880 + }, + { + "epoch": 0.41042724196277497, + "grad_norm": 7.8125, + "learning_rate": 4.964529349540741e-06, + "loss": 2.56, + "step": 3881 + }, + { + "epoch": 0.41053299492385786, + "grad_norm": 7.53125, + "learning_rate": 4.9645005975895485e-06, + "loss": 2.6633, + "step": 3882 + }, + { + "epoch": 0.41063874788494076, + "grad_norm": 8.0625, + "learning_rate": 4.964471834073462e-06, + "loss": 2.5875, + "step": 3883 + }, + { + "epoch": 0.4107445008460237, + "grad_norm": 7.75, + "learning_rate": 4.964443058992619e-06, + "loss": 2.7107, + "step": 3884 + }, + { + "epoch": 0.4108502538071066, + "grad_norm": 7.5, + "learning_rate": 4.96441427234715e-06, + "loss": 2.491, + "step": 3885 + }, + { + "epoch": 0.4109560067681895, + "grad_norm": 8.5, + "learning_rate": 4.964385474137195e-06, + "loss": 2.6572, + "step": 3886 + }, + { + "epoch": 0.41106175972927245, + "grad_norm": 7.5, + "learning_rate": 4.964356664362886e-06, + "loss": 2.5132, + "step": 3887 + }, + { + "epoch": 0.41116751269035534, + "grad_norm": 8.75, + "learning_rate": 4.964327843024358e-06, + "loss": 2.7006, + "step": 3888 + }, + { + "epoch": 0.41127326565143824, + "grad_norm": 7.6875, + "learning_rate": 4.964299010121749e-06, + "loss": 2.4928, + "step": 3889 + }, + { + "epoch": 0.41137901861252113, + "grad_norm": 8.25, + "learning_rate": 4.964270165655192e-06, + "loss": 2.6781, + "step": 3890 + }, + { + "epoch": 0.4114847715736041, + "grad_norm": 7.6875, + "learning_rate": 4.964241309624822e-06, + "loss": 2.7055, + "step": 3891 + }, + { + "epoch": 0.411590524534687, + "grad_norm": 7.46875, + "learning_rate": 4.964212442030777e-06, + "loss": 2.6702, + "step": 3892 + }, + { + "epoch": 0.41169627749576987, + "grad_norm": 7.65625, + "learning_rate": 4.9641835628731895e-06, + "loss": 2.5806, + "step": 3893 + }, + { + "epoch": 0.4118020304568528, + "grad_norm": 8.625, + "learning_rate": 4.964154672152197e-06, + "loss": 2.5912, + "step": 3894 + }, + { + "epoch": 0.4119077834179357, + "grad_norm": 7.90625, + "learning_rate": 4.9641257698679346e-06, + "loss": 2.7661, + "step": 3895 + }, + { + "epoch": 0.4120135363790186, + "grad_norm": 7.8125, + "learning_rate": 4.964096856020537e-06, + "loss": 2.6514, + "step": 3896 + }, + { + "epoch": 0.4121192893401015, + "grad_norm": 13.125, + "learning_rate": 4.9640679306101415e-06, + "loss": 2.4834, + "step": 3897 + }, + { + "epoch": 0.41222504230118445, + "grad_norm": 7.71875, + "learning_rate": 4.964038993636883e-06, + "loss": 2.6265, + "step": 3898 + }, + { + "epoch": 0.41233079526226735, + "grad_norm": 8.125, + "learning_rate": 4.964010045100896e-06, + "loss": 2.6029, + "step": 3899 + }, + { + "epoch": 0.41243654822335024, + "grad_norm": 7.375, + "learning_rate": 4.963981085002319e-06, + "loss": 2.5699, + "step": 3900 + }, + { + "epoch": 0.4125423011844332, + "grad_norm": 8.0, + "learning_rate": 4.963952113341286e-06, + "loss": 2.5431, + "step": 3901 + }, + { + "epoch": 0.4126480541455161, + "grad_norm": 8.3125, + "learning_rate": 4.963923130117934e-06, + "loss": 2.908, + "step": 3902 + }, + { + "epoch": 0.412753807106599, + "grad_norm": 7.5, + "learning_rate": 4.963894135332397e-06, + "loss": 2.5677, + "step": 3903 + }, + { + "epoch": 0.4128595600676819, + "grad_norm": 7.6875, + "learning_rate": 4.963865128984814e-06, + "loss": 2.622, + "step": 3904 + }, + { + "epoch": 0.4129653130287648, + "grad_norm": 8.25, + "learning_rate": 4.963836111075318e-06, + "loss": 2.6586, + "step": 3905 + }, + { + "epoch": 0.4130710659898477, + "grad_norm": 7.375, + "learning_rate": 4.963807081604048e-06, + "loss": 2.5832, + "step": 3906 + }, + { + "epoch": 0.4131768189509306, + "grad_norm": 8.1875, + "learning_rate": 4.96377804057114e-06, + "loss": 2.719, + "step": 3907 + }, + { + "epoch": 0.41328257191201356, + "grad_norm": 7.84375, + "learning_rate": 4.963748987976728e-06, + "loss": 2.6209, + "step": 3908 + }, + { + "epoch": 0.41338832487309646, + "grad_norm": 7.5, + "learning_rate": 4.963719923820949e-06, + "loss": 2.5845, + "step": 3909 + }, + { + "epoch": 0.41349407783417935, + "grad_norm": 7.53125, + "learning_rate": 4.963690848103941e-06, + "loss": 2.6661, + "step": 3910 + }, + { + "epoch": 0.41359983079526225, + "grad_norm": 7.375, + "learning_rate": 4.96366176082584e-06, + "loss": 2.3983, + "step": 3911 + }, + { + "epoch": 0.4137055837563452, + "grad_norm": 8.5, + "learning_rate": 4.963632661986781e-06, + "loss": 2.6467, + "step": 3912 + }, + { + "epoch": 0.4138113367174281, + "grad_norm": 9.0, + "learning_rate": 4.963603551586902e-06, + "loss": 2.5687, + "step": 3913 + }, + { + "epoch": 0.413917089678511, + "grad_norm": 8.3125, + "learning_rate": 4.963574429626339e-06, + "loss": 2.5743, + "step": 3914 + }, + { + "epoch": 0.41402284263959394, + "grad_norm": 8.25, + "learning_rate": 4.963545296105229e-06, + "loss": 2.4708, + "step": 3915 + }, + { + "epoch": 0.41412859560067683, + "grad_norm": 7.59375, + "learning_rate": 4.9635161510237074e-06, + "loss": 2.5949, + "step": 3916 + }, + { + "epoch": 0.4142343485617597, + "grad_norm": 7.84375, + "learning_rate": 4.963486994381913e-06, + "loss": 2.8107, + "step": 3917 + }, + { + "epoch": 0.4143401015228426, + "grad_norm": 8.1875, + "learning_rate": 4.963457826179982e-06, + "loss": 2.4356, + "step": 3918 + }, + { + "epoch": 0.41444585448392557, + "grad_norm": 8.5625, + "learning_rate": 4.9634286464180494e-06, + "loss": 2.6658, + "step": 3919 + }, + { + "epoch": 0.41455160744500846, + "grad_norm": 8.4375, + "learning_rate": 4.963399455096255e-06, + "loss": 2.5677, + "step": 3920 + }, + { + "epoch": 0.41465736040609136, + "grad_norm": 7.625, + "learning_rate": 4.9633702522147346e-06, + "loss": 2.4473, + "step": 3921 + }, + { + "epoch": 0.4147631133671743, + "grad_norm": 10.375, + "learning_rate": 4.963341037773624e-06, + "loss": 2.6435, + "step": 3922 + }, + { + "epoch": 0.4148688663282572, + "grad_norm": 7.75, + "learning_rate": 4.9633118117730626e-06, + "loss": 2.5796, + "step": 3923 + }, + { + "epoch": 0.4149746192893401, + "grad_norm": 7.96875, + "learning_rate": 4.9632825742131856e-06, + "loss": 2.621, + "step": 3924 + }, + { + "epoch": 0.415080372250423, + "grad_norm": 7.6875, + "learning_rate": 4.96325332509413e-06, + "loss": 2.6277, + "step": 3925 + }, + { + "epoch": 0.41518612521150594, + "grad_norm": 7.625, + "learning_rate": 4.963224064416036e-06, + "loss": 2.671, + "step": 3926 + }, + { + "epoch": 0.41529187817258884, + "grad_norm": 8.375, + "learning_rate": 4.963194792179038e-06, + "loss": 2.6131, + "step": 3927 + }, + { + "epoch": 0.41539763113367173, + "grad_norm": 7.78125, + "learning_rate": 4.963165508383274e-06, + "loss": 2.6357, + "step": 3928 + }, + { + "epoch": 0.4155033840947547, + "grad_norm": 7.65625, + "learning_rate": 4.963136213028882e-06, + "loss": 2.6443, + "step": 3929 + }, + { + "epoch": 0.4156091370558376, + "grad_norm": 8.5625, + "learning_rate": 4.963106906115999e-06, + "loss": 2.5916, + "step": 3930 + }, + { + "epoch": 0.41571489001692047, + "grad_norm": 8.125, + "learning_rate": 4.963077587644762e-06, + "loss": 2.6411, + "step": 3931 + }, + { + "epoch": 0.41582064297800336, + "grad_norm": 8.0, + "learning_rate": 4.963048257615311e-06, + "loss": 2.6891, + "step": 3932 + }, + { + "epoch": 0.4159263959390863, + "grad_norm": 7.625, + "learning_rate": 4.9630189160277795e-06, + "loss": 2.6333, + "step": 3933 + }, + { + "epoch": 0.4160321489001692, + "grad_norm": 8.125, + "learning_rate": 4.962989562882309e-06, + "loss": 2.6428, + "step": 3934 + }, + { + "epoch": 0.4161379018612521, + "grad_norm": 7.5, + "learning_rate": 4.962960198179036e-06, + "loss": 2.615, + "step": 3935 + }, + { + "epoch": 0.41624365482233505, + "grad_norm": 7.25, + "learning_rate": 4.962930821918098e-06, + "loss": 2.5134, + "step": 3936 + }, + { + "epoch": 0.41634940778341795, + "grad_norm": 8.6875, + "learning_rate": 4.9629014340996325e-06, + "loss": 2.5227, + "step": 3937 + }, + { + "epoch": 0.41645516074450084, + "grad_norm": 7.875, + "learning_rate": 4.962872034723778e-06, + "loss": 2.6972, + "step": 3938 + }, + { + "epoch": 0.41656091370558374, + "grad_norm": 7.71875, + "learning_rate": 4.9628426237906725e-06, + "loss": 2.6148, + "step": 3939 + }, + { + "epoch": 0.4166666666666667, + "grad_norm": 7.75, + "learning_rate": 4.962813201300454e-06, + "loss": 2.5765, + "step": 3940 + }, + { + "epoch": 0.4167724196277496, + "grad_norm": 7.875, + "learning_rate": 4.96278376725326e-06, + "loss": 2.6093, + "step": 3941 + }, + { + "epoch": 0.4168781725888325, + "grad_norm": 7.5, + "learning_rate": 4.9627543216492295e-06, + "loss": 2.7174, + "step": 3942 + }, + { + "epoch": 0.4169839255499154, + "grad_norm": 8.0, + "learning_rate": 4.9627248644884995e-06, + "loss": 2.3872, + "step": 3943 + }, + { + "epoch": 0.4170896785109983, + "grad_norm": 7.46875, + "learning_rate": 4.962695395771209e-06, + "loss": 2.6057, + "step": 3944 + }, + { + "epoch": 0.4171954314720812, + "grad_norm": 8.125, + "learning_rate": 4.962665915497496e-06, + "loss": 2.6214, + "step": 3945 + }, + { + "epoch": 0.4173011844331641, + "grad_norm": 7.84375, + "learning_rate": 4.9626364236675e-06, + "loss": 2.51, + "step": 3946 + }, + { + "epoch": 0.41740693739424706, + "grad_norm": 8.1875, + "learning_rate": 4.962606920281358e-06, + "loss": 2.8812, + "step": 3947 + }, + { + "epoch": 0.41751269035532995, + "grad_norm": 8.3125, + "learning_rate": 4.962577405339209e-06, + "loss": 2.7481, + "step": 3948 + }, + { + "epoch": 0.41761844331641285, + "grad_norm": 7.25, + "learning_rate": 4.9625478788411916e-06, + "loss": 2.6229, + "step": 3949 + }, + { + "epoch": 0.4177241962774958, + "grad_norm": 7.875, + "learning_rate": 4.962518340787444e-06, + "loss": 2.4878, + "step": 3950 + }, + { + "epoch": 0.4178299492385787, + "grad_norm": 8.5, + "learning_rate": 4.962488791178105e-06, + "loss": 2.7081, + "step": 3951 + }, + { + "epoch": 0.4179357021996616, + "grad_norm": 7.4375, + "learning_rate": 4.962459230013312e-06, + "loss": 2.7016, + "step": 3952 + }, + { + "epoch": 0.4180414551607445, + "grad_norm": 8.0625, + "learning_rate": 4.962429657293207e-06, + "loss": 2.2641, + "step": 3953 + }, + { + "epoch": 0.41814720812182743, + "grad_norm": 7.96875, + "learning_rate": 4.962400073017926e-06, + "loss": 2.6655, + "step": 3954 + }, + { + "epoch": 0.4182529610829103, + "grad_norm": 7.34375, + "learning_rate": 4.962370477187608e-06, + "loss": 2.5784, + "step": 3955 + }, + { + "epoch": 0.4183587140439932, + "grad_norm": 7.6875, + "learning_rate": 4.9623408698023924e-06, + "loss": 2.5549, + "step": 3956 + }, + { + "epoch": 0.41846446700507617, + "grad_norm": 7.9375, + "learning_rate": 4.962311250862418e-06, + "loss": 2.6343, + "step": 3957 + }, + { + "epoch": 0.41857021996615906, + "grad_norm": 8.25, + "learning_rate": 4.962281620367825e-06, + "loss": 2.8774, + "step": 3958 + }, + { + "epoch": 0.41867597292724196, + "grad_norm": 8.25, + "learning_rate": 4.96225197831875e-06, + "loss": 2.5853, + "step": 3959 + }, + { + "epoch": 0.41878172588832485, + "grad_norm": 7.8125, + "learning_rate": 4.962222324715334e-06, + "loss": 2.5398, + "step": 3960 + }, + { + "epoch": 0.4188874788494078, + "grad_norm": 7.625, + "learning_rate": 4.962192659557717e-06, + "loss": 2.5945, + "step": 3961 + }, + { + "epoch": 0.4189932318104907, + "grad_norm": 8.125, + "learning_rate": 4.9621629828460354e-06, + "loss": 2.7218, + "step": 3962 + }, + { + "epoch": 0.4190989847715736, + "grad_norm": 8.0625, + "learning_rate": 4.96213329458043e-06, + "loss": 2.7935, + "step": 3963 + }, + { + "epoch": 0.41920473773265654, + "grad_norm": 7.46875, + "learning_rate": 4.962103594761041e-06, + "loss": 2.6277, + "step": 3964 + }, + { + "epoch": 0.41931049069373943, + "grad_norm": 7.75, + "learning_rate": 4.962073883388005e-06, + "loss": 2.6183, + "step": 3965 + }, + { + "epoch": 0.41941624365482233, + "grad_norm": 7.53125, + "learning_rate": 4.9620441604614654e-06, + "loss": 2.6988, + "step": 3966 + }, + { + "epoch": 0.4195219966159052, + "grad_norm": 7.875, + "learning_rate": 4.962014425981558e-06, + "loss": 2.7574, + "step": 3967 + }, + { + "epoch": 0.4196277495769882, + "grad_norm": 7.34375, + "learning_rate": 4.961984679948425e-06, + "loss": 2.5758, + "step": 3968 + }, + { + "epoch": 0.41973350253807107, + "grad_norm": 7.96875, + "learning_rate": 4.961954922362203e-06, + "loss": 2.7066, + "step": 3969 + }, + { + "epoch": 0.41983925549915396, + "grad_norm": 7.4375, + "learning_rate": 4.961925153223035e-06, + "loss": 2.5888, + "step": 3970 + }, + { + "epoch": 0.4199450084602369, + "grad_norm": 7.25, + "learning_rate": 4.961895372531059e-06, + "loss": 2.6173, + "step": 3971 + }, + { + "epoch": 0.4200507614213198, + "grad_norm": 8.125, + "learning_rate": 4.961865580286415e-06, + "loss": 2.5457, + "step": 3972 + }, + { + "epoch": 0.4201565143824027, + "grad_norm": 7.1875, + "learning_rate": 4.961835776489242e-06, + "loss": 2.5068, + "step": 3973 + }, + { + "epoch": 0.4202622673434856, + "grad_norm": 7.5625, + "learning_rate": 4.96180596113968e-06, + "loss": 2.6704, + "step": 3974 + }, + { + "epoch": 0.42036802030456855, + "grad_norm": 8.125, + "learning_rate": 4.96177613423787e-06, + "loss": 2.7378, + "step": 3975 + }, + { + "epoch": 0.42047377326565144, + "grad_norm": 7.625, + "learning_rate": 4.961746295783952e-06, + "loss": 2.7024, + "step": 3976 + }, + { + "epoch": 0.42057952622673433, + "grad_norm": 7.6875, + "learning_rate": 4.9617164457780655e-06, + "loss": 2.3955, + "step": 3977 + }, + { + "epoch": 0.4206852791878173, + "grad_norm": 7.5, + "learning_rate": 4.9616865842203504e-06, + "loss": 2.6461, + "step": 3978 + }, + { + "epoch": 0.4207910321489002, + "grad_norm": 8.0, + "learning_rate": 4.961656711110947e-06, + "loss": 2.7064, + "step": 3979 + }, + { + "epoch": 0.4208967851099831, + "grad_norm": 8.0625, + "learning_rate": 4.9616268264499945e-06, + "loss": 2.7258, + "step": 3980 + }, + { + "epoch": 0.42100253807106597, + "grad_norm": 7.59375, + "learning_rate": 4.961596930237636e-06, + "loss": 2.7064, + "step": 3981 + }, + { + "epoch": 0.4211082910321489, + "grad_norm": 7.53125, + "learning_rate": 4.961567022474007e-06, + "loss": 2.511, + "step": 3982 + }, + { + "epoch": 0.4212140439932318, + "grad_norm": 7.8125, + "learning_rate": 4.961537103159253e-06, + "loss": 2.6974, + "step": 3983 + }, + { + "epoch": 0.4213197969543147, + "grad_norm": 7.90625, + "learning_rate": 4.961507172293511e-06, + "loss": 2.4829, + "step": 3984 + }, + { + "epoch": 0.42142554991539766, + "grad_norm": 8.5, + "learning_rate": 4.961477229876924e-06, + "loss": 2.4651, + "step": 3985 + }, + { + "epoch": 0.42153130287648055, + "grad_norm": 8.25, + "learning_rate": 4.96144727590963e-06, + "loss": 2.7486, + "step": 3986 + }, + { + "epoch": 0.42163705583756345, + "grad_norm": 7.8125, + "learning_rate": 4.9614173103917705e-06, + "loss": 2.7045, + "step": 3987 + }, + { + "epoch": 0.42174280879864634, + "grad_norm": 7.5, + "learning_rate": 4.961387333323486e-06, + "loss": 2.5977, + "step": 3988 + }, + { + "epoch": 0.4218485617597293, + "grad_norm": 7.84375, + "learning_rate": 4.9613573447049185e-06, + "loss": 2.8232, + "step": 3989 + }, + { + "epoch": 0.4219543147208122, + "grad_norm": 8.5625, + "learning_rate": 4.961327344536207e-06, + "loss": 2.918, + "step": 3990 + }, + { + "epoch": 0.4220600676818951, + "grad_norm": 8.25, + "learning_rate": 4.961297332817493e-06, + "loss": 2.6977, + "step": 3991 + }, + { + "epoch": 0.42216582064297803, + "grad_norm": 7.6875, + "learning_rate": 4.961267309548917e-06, + "loss": 2.6123, + "step": 3992 + }, + { + "epoch": 0.4222715736040609, + "grad_norm": 8.625, + "learning_rate": 4.9612372747306205e-06, + "loss": 2.7706, + "step": 3993 + }, + { + "epoch": 0.4223773265651438, + "grad_norm": 7.6875, + "learning_rate": 4.961207228362744e-06, + "loss": 2.7436, + "step": 3994 + }, + { + "epoch": 0.4224830795262267, + "grad_norm": 7.75, + "learning_rate": 4.961177170445428e-06, + "loss": 2.5551, + "step": 3995 + }, + { + "epoch": 0.42258883248730966, + "grad_norm": 7.9375, + "learning_rate": 4.961147100978815e-06, + "loss": 2.6524, + "step": 3996 + }, + { + "epoch": 0.42269458544839256, + "grad_norm": 7.625, + "learning_rate": 4.961117019963045e-06, + "loss": 2.6581, + "step": 3997 + }, + { + "epoch": 0.42280033840947545, + "grad_norm": 7.96875, + "learning_rate": 4.961086927398258e-06, + "loss": 2.9132, + "step": 3998 + }, + { + "epoch": 0.4229060913705584, + "grad_norm": 8.5, + "learning_rate": 4.961056823284599e-06, + "loss": 2.7187, + "step": 3999 + }, + { + "epoch": 0.4230118443316413, + "grad_norm": 7.75, + "learning_rate": 4.961026707622206e-06, + "loss": 2.482, + "step": 4000 + }, + { + "epoch": 0.4231175972927242, + "grad_norm": 8.0625, + "learning_rate": 4.96099658041122e-06, + "loss": 2.7266, + "step": 4001 + }, + { + "epoch": 0.4232233502538071, + "grad_norm": 7.71875, + "learning_rate": 4.960966441651785e-06, + "loss": 2.638, + "step": 4002 + }, + { + "epoch": 0.42332910321489003, + "grad_norm": 7.8125, + "learning_rate": 4.960936291344041e-06, + "loss": 2.7195, + "step": 4003 + }, + { + "epoch": 0.42343485617597293, + "grad_norm": 14.125, + "learning_rate": 4.960906129488128e-06, + "loss": 2.45, + "step": 4004 + }, + { + "epoch": 0.4235406091370558, + "grad_norm": 8.8125, + "learning_rate": 4.96087595608419e-06, + "loss": 2.7289, + "step": 4005 + }, + { + "epoch": 0.4236463620981388, + "grad_norm": 8.6875, + "learning_rate": 4.960845771132368e-06, + "loss": 2.8612, + "step": 4006 + }, + { + "epoch": 0.42375211505922167, + "grad_norm": 7.75, + "learning_rate": 4.960815574632802e-06, + "loss": 2.5925, + "step": 4007 + }, + { + "epoch": 0.42385786802030456, + "grad_norm": 8.75, + "learning_rate": 4.9607853665856365e-06, + "loss": 2.6576, + "step": 4008 + }, + { + "epoch": 0.42396362098138746, + "grad_norm": 9.0625, + "learning_rate": 4.960755146991011e-06, + "loss": 2.6656, + "step": 4009 + }, + { + "epoch": 0.4240693739424704, + "grad_norm": 7.71875, + "learning_rate": 4.960724915849068e-06, + "loss": 2.7154, + "step": 4010 + }, + { + "epoch": 0.4241751269035533, + "grad_norm": 7.25, + "learning_rate": 4.960694673159949e-06, + "loss": 2.4958, + "step": 4011 + }, + { + "epoch": 0.4242808798646362, + "grad_norm": 7.28125, + "learning_rate": 4.960664418923797e-06, + "loss": 2.5152, + "step": 4012 + }, + { + "epoch": 0.42438663282571915, + "grad_norm": 8.0, + "learning_rate": 4.960634153140752e-06, + "loss": 2.5518, + "step": 4013 + }, + { + "epoch": 0.42449238578680204, + "grad_norm": 7.34375, + "learning_rate": 4.960603875810959e-06, + "loss": 2.6027, + "step": 4014 + }, + { + "epoch": 0.42459813874788493, + "grad_norm": 7.59375, + "learning_rate": 4.960573586934557e-06, + "loss": 2.5513, + "step": 4015 + }, + { + "epoch": 0.42470389170896783, + "grad_norm": 7.8125, + "learning_rate": 4.960543286511691e-06, + "loss": 2.7342, + "step": 4016 + }, + { + "epoch": 0.4248096446700508, + "grad_norm": 16.375, + "learning_rate": 4.960512974542499e-06, + "loss": 2.6328, + "step": 4017 + }, + { + "epoch": 0.4249153976311337, + "grad_norm": 8.3125, + "learning_rate": 4.960482651027129e-06, + "loss": 2.7338, + "step": 4018 + }, + { + "epoch": 0.42502115059221657, + "grad_norm": 7.78125, + "learning_rate": 4.9604523159657185e-06, + "loss": 2.5628, + "step": 4019 + }, + { + "epoch": 0.4251269035532995, + "grad_norm": 11.4375, + "learning_rate": 4.960421969358412e-06, + "loss": 2.8323, + "step": 4020 + }, + { + "epoch": 0.4252326565143824, + "grad_norm": 8.1875, + "learning_rate": 4.960391611205351e-06, + "loss": 2.723, + "step": 4021 + }, + { + "epoch": 0.4253384094754653, + "grad_norm": 8.0625, + "learning_rate": 4.960361241506679e-06, + "loss": 2.6729, + "step": 4022 + }, + { + "epoch": 0.4254441624365482, + "grad_norm": 8.5, + "learning_rate": 4.960330860262537e-06, + "loss": 2.5199, + "step": 4023 + }, + { + "epoch": 0.42554991539763115, + "grad_norm": 7.9375, + "learning_rate": 4.96030046747307e-06, + "loss": 2.5701, + "step": 4024 + }, + { + "epoch": 0.42565566835871405, + "grad_norm": 7.84375, + "learning_rate": 4.960270063138418e-06, + "loss": 2.3367, + "step": 4025 + }, + { + "epoch": 0.42576142131979694, + "grad_norm": 8.1875, + "learning_rate": 4.960239647258725e-06, + "loss": 2.7207, + "step": 4026 + }, + { + "epoch": 0.4258671742808799, + "grad_norm": 7.3125, + "learning_rate": 4.960209219834134e-06, + "loss": 2.6778, + "step": 4027 + }, + { + "epoch": 0.4259729272419628, + "grad_norm": 8.9375, + "learning_rate": 4.960178780864787e-06, + "loss": 2.5414, + "step": 4028 + }, + { + "epoch": 0.4260786802030457, + "grad_norm": 7.90625, + "learning_rate": 4.960148330350828e-06, + "loss": 2.6114, + "step": 4029 + }, + { + "epoch": 0.4261844331641286, + "grad_norm": 7.75, + "learning_rate": 4.960117868292398e-06, + "loss": 2.6331, + "step": 4030 + }, + { + "epoch": 0.4262901861252115, + "grad_norm": 7.59375, + "learning_rate": 4.9600873946896415e-06, + "loss": 2.5102, + "step": 4031 + }, + { + "epoch": 0.4263959390862944, + "grad_norm": 8.125, + "learning_rate": 4.960056909542701e-06, + "loss": 2.7289, + "step": 4032 + }, + { + "epoch": 0.4265016920473773, + "grad_norm": 8.1875, + "learning_rate": 4.960026412851719e-06, + "loss": 2.6502, + "step": 4033 + }, + { + "epoch": 0.42660744500846026, + "grad_norm": 7.96875, + "learning_rate": 4.95999590461684e-06, + "loss": 2.4806, + "step": 4034 + }, + { + "epoch": 0.42671319796954316, + "grad_norm": 7.84375, + "learning_rate": 4.959965384838206e-06, + "loss": 2.7537, + "step": 4035 + }, + { + "epoch": 0.42681895093062605, + "grad_norm": 7.9375, + "learning_rate": 4.95993485351596e-06, + "loss": 2.4471, + "step": 4036 + }, + { + "epoch": 0.42692470389170895, + "grad_norm": 8.8125, + "learning_rate": 4.9599043106502455e-06, + "loss": 2.7005, + "step": 4037 + }, + { + "epoch": 0.4270304568527919, + "grad_norm": 8.4375, + "learning_rate": 4.959873756241208e-06, + "loss": 2.6168, + "step": 4038 + }, + { + "epoch": 0.4271362098138748, + "grad_norm": 7.75, + "learning_rate": 4.959843190288987e-06, + "loss": 2.6955, + "step": 4039 + }, + { + "epoch": 0.4272419627749577, + "grad_norm": 7.71875, + "learning_rate": 4.959812612793729e-06, + "loss": 2.615, + "step": 4040 + }, + { + "epoch": 0.42734771573604063, + "grad_norm": 7.9375, + "learning_rate": 4.959782023755576e-06, + "loss": 2.8946, + "step": 4041 + }, + { + "epoch": 0.42745346869712353, + "grad_norm": 7.46875, + "learning_rate": 4.959751423174673e-06, + "loss": 2.5308, + "step": 4042 + }, + { + "epoch": 0.4275592216582064, + "grad_norm": 7.53125, + "learning_rate": 4.959720811051161e-06, + "loss": 2.5748, + "step": 4043 + }, + { + "epoch": 0.4276649746192893, + "grad_norm": 11.75, + "learning_rate": 4.959690187385187e-06, + "loss": 2.3616, + "step": 4044 + }, + { + "epoch": 0.42777072758037227, + "grad_norm": 8.125, + "learning_rate": 4.959659552176891e-06, + "loss": 2.6326, + "step": 4045 + }, + { + "epoch": 0.42787648054145516, + "grad_norm": 8.1875, + "learning_rate": 4.95962890542642e-06, + "loss": 2.7242, + "step": 4046 + }, + { + "epoch": 0.42798223350253806, + "grad_norm": 8.3125, + "learning_rate": 4.959598247133916e-06, + "loss": 2.6522, + "step": 4047 + }, + { + "epoch": 0.428087986463621, + "grad_norm": 9.0625, + "learning_rate": 4.959567577299523e-06, + "loss": 2.2408, + "step": 4048 + }, + { + "epoch": 0.4281937394247039, + "grad_norm": 7.78125, + "learning_rate": 4.9595368959233856e-06, + "loss": 2.434, + "step": 4049 + }, + { + "epoch": 0.4282994923857868, + "grad_norm": 7.40625, + "learning_rate": 4.959506203005648e-06, + "loss": 2.5412, + "step": 4050 + }, + { + "epoch": 0.4284052453468697, + "grad_norm": 7.625, + "learning_rate": 4.9594754985464535e-06, + "loss": 2.466, + "step": 4051 + }, + { + "epoch": 0.42851099830795264, + "grad_norm": 7.3125, + "learning_rate": 4.959444782545946e-06, + "loss": 2.6863, + "step": 4052 + }, + { + "epoch": 0.42861675126903553, + "grad_norm": 8.0, + "learning_rate": 4.95941405500427e-06, + "loss": 2.6552, + "step": 4053 + }, + { + "epoch": 0.42872250423011843, + "grad_norm": 7.71875, + "learning_rate": 4.95938331592157e-06, + "loss": 2.6723, + "step": 4054 + }, + { + "epoch": 0.4288282571912014, + "grad_norm": 7.78125, + "learning_rate": 4.9593525652979905e-06, + "loss": 2.685, + "step": 4055 + }, + { + "epoch": 0.4289340101522843, + "grad_norm": 8.625, + "learning_rate": 4.959321803133674e-06, + "loss": 2.7296, + "step": 4056 + }, + { + "epoch": 0.42903976311336717, + "grad_norm": 8.125, + "learning_rate": 4.959291029428767e-06, + "loss": 2.4903, + "step": 4057 + }, + { + "epoch": 0.42914551607445006, + "grad_norm": 7.6875, + "learning_rate": 4.9592602441834134e-06, + "loss": 2.675, + "step": 4058 + }, + { + "epoch": 0.429251269035533, + "grad_norm": 7.875, + "learning_rate": 4.959229447397756e-06, + "loss": 2.2996, + "step": 4059 + }, + { + "epoch": 0.4293570219966159, + "grad_norm": 7.25, + "learning_rate": 4.959198639071942e-06, + "loss": 2.4617, + "step": 4060 + }, + { + "epoch": 0.4294627749576988, + "grad_norm": 8.125, + "learning_rate": 4.959167819206113e-06, + "loss": 2.7613, + "step": 4061 + }, + { + "epoch": 0.42956852791878175, + "grad_norm": 7.6875, + "learning_rate": 4.959136987800416e-06, + "loss": 2.4735, + "step": 4062 + }, + { + "epoch": 0.42967428087986465, + "grad_norm": 7.96875, + "learning_rate": 4.959106144854995e-06, + "loss": 2.4752, + "step": 4063 + }, + { + "epoch": 0.42978003384094754, + "grad_norm": 8.1875, + "learning_rate": 4.959075290369995e-06, + "loss": 2.7539, + "step": 4064 + }, + { + "epoch": 0.42988578680203043, + "grad_norm": 8.0, + "learning_rate": 4.9590444243455595e-06, + "loss": 2.5316, + "step": 4065 + }, + { + "epoch": 0.4299915397631134, + "grad_norm": 8.0, + "learning_rate": 4.959013546781834e-06, + "loss": 2.5384, + "step": 4066 + }, + { + "epoch": 0.4300972927241963, + "grad_norm": 7.6875, + "learning_rate": 4.958982657678964e-06, + "loss": 2.7254, + "step": 4067 + }, + { + "epoch": 0.4302030456852792, + "grad_norm": 7.5, + "learning_rate": 4.958951757037095e-06, + "loss": 2.5262, + "step": 4068 + }, + { + "epoch": 0.4303087986463621, + "grad_norm": 7.5, + "learning_rate": 4.9589208448563705e-06, + "loss": 2.5701, + "step": 4069 + }, + { + "epoch": 0.430414551607445, + "grad_norm": 7.71875, + "learning_rate": 4.958889921136936e-06, + "loss": 2.6911, + "step": 4070 + }, + { + "epoch": 0.4305203045685279, + "grad_norm": 7.78125, + "learning_rate": 4.958858985878936e-06, + "loss": 2.561, + "step": 4071 + }, + { + "epoch": 0.4306260575296108, + "grad_norm": 8.375, + "learning_rate": 4.9588280390825176e-06, + "loss": 2.4335, + "step": 4072 + }, + { + "epoch": 0.43073181049069376, + "grad_norm": 7.6875, + "learning_rate": 4.9587970807478245e-06, + "loss": 2.688, + "step": 4073 + }, + { + "epoch": 0.43083756345177665, + "grad_norm": 15.0, + "learning_rate": 4.958766110875003e-06, + "loss": 2.801, + "step": 4074 + }, + { + "epoch": 0.43094331641285955, + "grad_norm": 7.875, + "learning_rate": 4.958735129464196e-06, + "loss": 2.5089, + "step": 4075 + }, + { + "epoch": 0.4310490693739425, + "grad_norm": 7.15625, + "learning_rate": 4.958704136515552e-06, + "loss": 2.6402, + "step": 4076 + }, + { + "epoch": 0.4311548223350254, + "grad_norm": 7.75, + "learning_rate": 4.958673132029215e-06, + "loss": 2.7236, + "step": 4077 + }, + { + "epoch": 0.4312605752961083, + "grad_norm": 7.53125, + "learning_rate": 4.95864211600533e-06, + "loss": 2.6807, + "step": 4078 + }, + { + "epoch": 0.4313663282571912, + "grad_norm": 8.4375, + "learning_rate": 4.958611088444043e-06, + "loss": 2.7486, + "step": 4079 + }, + { + "epoch": 0.43147208121827413, + "grad_norm": 8.4375, + "learning_rate": 4.9585800493455e-06, + "loss": 2.7043, + "step": 4080 + }, + { + "epoch": 0.431577834179357, + "grad_norm": 7.78125, + "learning_rate": 4.958548998709847e-06, + "loss": 2.6131, + "step": 4081 + }, + { + "epoch": 0.4316835871404399, + "grad_norm": 7.3125, + "learning_rate": 4.958517936537228e-06, + "loss": 2.4552, + "step": 4082 + }, + { + "epoch": 0.43178934010152287, + "grad_norm": 8.3125, + "learning_rate": 4.95848686282779e-06, + "loss": 2.5166, + "step": 4083 + }, + { + "epoch": 0.43189509306260576, + "grad_norm": 7.625, + "learning_rate": 4.958455777581679e-06, + "loss": 2.4899, + "step": 4084 + }, + { + "epoch": 0.43200084602368866, + "grad_norm": 7.84375, + "learning_rate": 4.95842468079904e-06, + "loss": 2.6593, + "step": 4085 + }, + { + "epoch": 0.43210659898477155, + "grad_norm": 8.25, + "learning_rate": 4.95839357248002e-06, + "loss": 2.5792, + "step": 4086 + }, + { + "epoch": 0.4322123519458545, + "grad_norm": 8.125, + "learning_rate": 4.958362452624764e-06, + "loss": 2.4911, + "step": 4087 + }, + { + "epoch": 0.4323181049069374, + "grad_norm": 8.375, + "learning_rate": 4.9583313212334194e-06, + "loss": 2.2654, + "step": 4088 + }, + { + "epoch": 0.4324238578680203, + "grad_norm": 7.625, + "learning_rate": 4.9583001783061305e-06, + "loss": 2.5113, + "step": 4089 + }, + { + "epoch": 0.43252961082910324, + "grad_norm": 7.40625, + "learning_rate": 4.958269023843044e-06, + "loss": 2.5616, + "step": 4090 + }, + { + "epoch": 0.43263536379018613, + "grad_norm": 8.25, + "learning_rate": 4.958237857844307e-06, + "loss": 2.6644, + "step": 4091 + }, + { + "epoch": 0.43274111675126903, + "grad_norm": 8.375, + "learning_rate": 4.958206680310065e-06, + "loss": 2.5933, + "step": 4092 + }, + { + "epoch": 0.4328468697123519, + "grad_norm": 7.25, + "learning_rate": 4.9581754912404646e-06, + "loss": 2.5562, + "step": 4093 + }, + { + "epoch": 0.4329526226734349, + "grad_norm": 7.5, + "learning_rate": 4.958144290635651e-06, + "loss": 2.6804, + "step": 4094 + }, + { + "epoch": 0.43305837563451777, + "grad_norm": 7.4375, + "learning_rate": 4.958113078495773e-06, + "loss": 2.5591, + "step": 4095 + }, + { + "epoch": 0.43316412859560066, + "grad_norm": 8.5625, + "learning_rate": 4.9580818548209744e-06, + "loss": 2.8427, + "step": 4096 + }, + { + "epoch": 0.4332698815566836, + "grad_norm": 8.3125, + "learning_rate": 4.958050619611403e-06, + "loss": 2.5405, + "step": 4097 + }, + { + "epoch": 0.4333756345177665, + "grad_norm": 8.8125, + "learning_rate": 4.958019372867207e-06, + "loss": 2.4296, + "step": 4098 + }, + { + "epoch": 0.4334813874788494, + "grad_norm": 8.0625, + "learning_rate": 4.95798811458853e-06, + "loss": 2.7566, + "step": 4099 + }, + { + "epoch": 0.4335871404399323, + "grad_norm": 7.875, + "learning_rate": 4.95795684477552e-06, + "loss": 2.7384, + "step": 4100 + }, + { + "epoch": 0.43369289340101524, + "grad_norm": 7.3125, + "learning_rate": 4.957925563428324e-06, + "loss": 2.4782, + "step": 4101 + }, + { + "epoch": 0.43379864636209814, + "grad_norm": 10.0, + "learning_rate": 4.957894270547089e-06, + "loss": 2.4727, + "step": 4102 + }, + { + "epoch": 0.43390439932318103, + "grad_norm": 7.34375, + "learning_rate": 4.957862966131961e-06, + "loss": 2.5388, + "step": 4103 + }, + { + "epoch": 0.434010152284264, + "grad_norm": 7.25, + "learning_rate": 4.957831650183087e-06, + "loss": 2.6106, + "step": 4104 + }, + { + "epoch": 0.4341159052453469, + "grad_norm": 7.96875, + "learning_rate": 4.957800322700615e-06, + "loss": 2.5749, + "step": 4105 + }, + { + "epoch": 0.4342216582064298, + "grad_norm": 8.375, + "learning_rate": 4.957768983684691e-06, + "loss": 2.6106, + "step": 4106 + }, + { + "epoch": 0.43432741116751267, + "grad_norm": 7.71875, + "learning_rate": 4.957737633135462e-06, + "loss": 2.5858, + "step": 4107 + }, + { + "epoch": 0.4344331641285956, + "grad_norm": 7.8125, + "learning_rate": 4.957706271053076e-06, + "loss": 2.647, + "step": 4108 + }, + { + "epoch": 0.4345389170896785, + "grad_norm": 8.625, + "learning_rate": 4.95767489743768e-06, + "loss": 2.4623, + "step": 4109 + }, + { + "epoch": 0.4346446700507614, + "grad_norm": 7.78125, + "learning_rate": 4.95764351228942e-06, + "loss": 2.6344, + "step": 4110 + }, + { + "epoch": 0.43475042301184436, + "grad_norm": 7.4375, + "learning_rate": 4.957612115608445e-06, + "loss": 2.5281, + "step": 4111 + }, + { + "epoch": 0.43485617597292725, + "grad_norm": 8.25, + "learning_rate": 4.9575807073949005e-06, + "loss": 2.8331, + "step": 4112 + }, + { + "epoch": 0.43496192893401014, + "grad_norm": 7.9375, + "learning_rate": 4.957549287648935e-06, + "loss": 2.738, + "step": 4113 + }, + { + "epoch": 0.43506768189509304, + "grad_norm": 7.375, + "learning_rate": 4.957517856370696e-06, + "loss": 2.6808, + "step": 4114 + }, + { + "epoch": 0.435173434856176, + "grad_norm": 8.25, + "learning_rate": 4.957486413560331e-06, + "loss": 2.6663, + "step": 4115 + }, + { + "epoch": 0.4352791878172589, + "grad_norm": 9.125, + "learning_rate": 4.9574549592179875e-06, + "loss": 2.68, + "step": 4116 + }, + { + "epoch": 0.4353849407783418, + "grad_norm": 8.0, + "learning_rate": 4.957423493343813e-06, + "loss": 2.6268, + "step": 4117 + }, + { + "epoch": 0.43549069373942473, + "grad_norm": 7.4375, + "learning_rate": 4.957392015937955e-06, + "loss": 2.7285, + "step": 4118 + }, + { + "epoch": 0.4355964467005076, + "grad_norm": 7.6875, + "learning_rate": 4.9573605270005606e-06, + "loss": 2.691, + "step": 4119 + }, + { + "epoch": 0.4357021996615905, + "grad_norm": 7.71875, + "learning_rate": 4.957329026531778e-06, + "loss": 2.6914, + "step": 4120 + }, + { + "epoch": 0.4358079526226734, + "grad_norm": 7.78125, + "learning_rate": 4.957297514531757e-06, + "loss": 2.6171, + "step": 4121 + }, + { + "epoch": 0.43591370558375636, + "grad_norm": 8.0625, + "learning_rate": 4.957265991000643e-06, + "loss": 2.4077, + "step": 4122 + }, + { + "epoch": 0.43601945854483926, + "grad_norm": 7.96875, + "learning_rate": 4.957234455938584e-06, + "loss": 2.8315, + "step": 4123 + }, + { + "epoch": 0.43612521150592215, + "grad_norm": 8.125, + "learning_rate": 4.95720290934573e-06, + "loss": 2.4973, + "step": 4124 + }, + { + "epoch": 0.4362309644670051, + "grad_norm": 8.0625, + "learning_rate": 4.957171351222226e-06, + "loss": 2.5434, + "step": 4125 + }, + { + "epoch": 0.436336717428088, + "grad_norm": 8.0, + "learning_rate": 4.957139781568223e-06, + "loss": 2.7687, + "step": 4126 + }, + { + "epoch": 0.4364424703891709, + "grad_norm": 7.46875, + "learning_rate": 4.957108200383868e-06, + "loss": 2.6166, + "step": 4127 + }, + { + "epoch": 0.4365482233502538, + "grad_norm": 7.84375, + "learning_rate": 4.957076607669308e-06, + "loss": 2.5949, + "step": 4128 + }, + { + "epoch": 0.43665397631133673, + "grad_norm": 7.625, + "learning_rate": 4.957045003424693e-06, + "loss": 2.5963, + "step": 4129 + }, + { + "epoch": 0.43675972927241963, + "grad_norm": 7.78125, + "learning_rate": 4.957013387650171e-06, + "loss": 2.7959, + "step": 4130 + }, + { + "epoch": 0.4368654822335025, + "grad_norm": 8.1875, + "learning_rate": 4.95698176034589e-06, + "loss": 2.7215, + "step": 4131 + }, + { + "epoch": 0.4369712351945855, + "grad_norm": 7.5, + "learning_rate": 4.956950121511998e-06, + "loss": 2.6107, + "step": 4132 + }, + { + "epoch": 0.43707698815566837, + "grad_norm": 7.71875, + "learning_rate": 4.9569184711486455e-06, + "loss": 2.4358, + "step": 4133 + }, + { + "epoch": 0.43718274111675126, + "grad_norm": 7.65625, + "learning_rate": 4.956886809255977e-06, + "loss": 2.605, + "step": 4134 + }, + { + "epoch": 0.43728849407783416, + "grad_norm": 8.125, + "learning_rate": 4.956855135834146e-06, + "loss": 2.654, + "step": 4135 + }, + { + "epoch": 0.4373942470389171, + "grad_norm": 7.75, + "learning_rate": 4.956823450883297e-06, + "loss": 2.7777, + "step": 4136 + }, + { + "epoch": 0.4375, + "grad_norm": 7.78125, + "learning_rate": 4.956791754403581e-06, + "loss": 2.4923, + "step": 4137 + }, + { + "epoch": 0.4376057529610829, + "grad_norm": 8.1875, + "learning_rate": 4.956760046395146e-06, + "loss": 2.5061, + "step": 4138 + }, + { + "epoch": 0.43771150592216584, + "grad_norm": 7.65625, + "learning_rate": 4.95672832685814e-06, + "loss": 2.4371, + "step": 4139 + }, + { + "epoch": 0.43781725888324874, + "grad_norm": 7.75, + "learning_rate": 4.9566965957927135e-06, + "loss": 2.5616, + "step": 4140 + }, + { + "epoch": 0.43792301184433163, + "grad_norm": 7.6875, + "learning_rate": 4.956664853199014e-06, + "loss": 2.5678, + "step": 4141 + }, + { + "epoch": 0.4380287648054145, + "grad_norm": 8.0, + "learning_rate": 4.956633099077193e-06, + "loss": 2.6081, + "step": 4142 + }, + { + "epoch": 0.4381345177664975, + "grad_norm": 7.28125, + "learning_rate": 4.956601333427396e-06, + "loss": 2.486, + "step": 4143 + }, + { + "epoch": 0.43824027072758037, + "grad_norm": 8.0625, + "learning_rate": 4.956569556249774e-06, + "loss": 2.6065, + "step": 4144 + }, + { + "epoch": 0.43834602368866327, + "grad_norm": 8.25, + "learning_rate": 4.9565377675444755e-06, + "loss": 2.5077, + "step": 4145 + }, + { + "epoch": 0.4384517766497462, + "grad_norm": 7.78125, + "learning_rate": 4.9565059673116506e-06, + "loss": 2.7127, + "step": 4146 + }, + { + "epoch": 0.4385575296108291, + "grad_norm": 8.25, + "learning_rate": 4.9564741555514475e-06, + "loss": 2.6184, + "step": 4147 + }, + { + "epoch": 0.438663282571912, + "grad_norm": 7.71875, + "learning_rate": 4.956442332264016e-06, + "loss": 2.5863, + "step": 4148 + }, + { + "epoch": 0.4387690355329949, + "grad_norm": 8.5625, + "learning_rate": 4.956410497449505e-06, + "loss": 2.6247, + "step": 4149 + }, + { + "epoch": 0.43887478849407785, + "grad_norm": 7.46875, + "learning_rate": 4.956378651108064e-06, + "loss": 2.4994, + "step": 4150 + }, + { + "epoch": 0.43898054145516074, + "grad_norm": 8.25, + "learning_rate": 4.956346793239843e-06, + "loss": 2.6644, + "step": 4151 + }, + { + "epoch": 0.43908629441624364, + "grad_norm": 7.65625, + "learning_rate": 4.956314923844991e-06, + "loss": 2.5976, + "step": 4152 + }, + { + "epoch": 0.4391920473773266, + "grad_norm": 9.5, + "learning_rate": 4.956283042923658e-06, + "loss": 2.7835, + "step": 4153 + }, + { + "epoch": 0.4392978003384095, + "grad_norm": 8.125, + "learning_rate": 4.956251150475992e-06, + "loss": 2.6725, + "step": 4154 + }, + { + "epoch": 0.4394035532994924, + "grad_norm": 7.65625, + "learning_rate": 4.9562192465021455e-06, + "loss": 2.7423, + "step": 4155 + }, + { + "epoch": 0.43950930626057527, + "grad_norm": 7.90625, + "learning_rate": 4.956187331002267e-06, + "loss": 2.3152, + "step": 4156 + }, + { + "epoch": 0.4396150592216582, + "grad_norm": 7.90625, + "learning_rate": 4.956155403976504e-06, + "loss": 2.6294, + "step": 4157 + }, + { + "epoch": 0.4397208121827411, + "grad_norm": 8.5625, + "learning_rate": 4.956123465425009e-06, + "loss": 2.5538, + "step": 4158 + }, + { + "epoch": 0.439826565143824, + "grad_norm": 7.625, + "learning_rate": 4.956091515347932e-06, + "loss": 2.5856, + "step": 4159 + }, + { + "epoch": 0.43993231810490696, + "grad_norm": 7.75, + "learning_rate": 4.95605955374542e-06, + "loss": 2.7488, + "step": 4160 + }, + { + "epoch": 0.44003807106598986, + "grad_norm": 8.0, + "learning_rate": 4.956027580617627e-06, + "loss": 2.4759, + "step": 4161 + }, + { + "epoch": 0.44014382402707275, + "grad_norm": 8.0625, + "learning_rate": 4.9559955959647e-06, + "loss": 2.7286, + "step": 4162 + }, + { + "epoch": 0.44024957698815564, + "grad_norm": 7.65625, + "learning_rate": 4.9559635997867905e-06, + "loss": 2.6885, + "step": 4163 + }, + { + "epoch": 0.4403553299492386, + "grad_norm": 7.4375, + "learning_rate": 4.955931592084048e-06, + "loss": 2.5399, + "step": 4164 + }, + { + "epoch": 0.4404610829103215, + "grad_norm": 8.875, + "learning_rate": 4.955899572856624e-06, + "loss": 2.6343, + "step": 4165 + }, + { + "epoch": 0.4405668358714044, + "grad_norm": 8.75, + "learning_rate": 4.955867542104667e-06, + "loss": 2.636, + "step": 4166 + }, + { + "epoch": 0.44067258883248733, + "grad_norm": 8.25, + "learning_rate": 4.955835499828327e-06, + "loss": 2.6075, + "step": 4167 + }, + { + "epoch": 0.4407783417935702, + "grad_norm": 7.65625, + "learning_rate": 4.955803446027757e-06, + "loss": 2.6059, + "step": 4168 + }, + { + "epoch": 0.4408840947546531, + "grad_norm": 7.40625, + "learning_rate": 4.955771380703105e-06, + "loss": 2.5469, + "step": 4169 + }, + { + "epoch": 0.440989847715736, + "grad_norm": 8.125, + "learning_rate": 4.955739303854522e-06, + "loss": 2.8538, + "step": 4170 + }, + { + "epoch": 0.44109560067681897, + "grad_norm": 7.71875, + "learning_rate": 4.95570721548216e-06, + "loss": 2.5172, + "step": 4171 + }, + { + "epoch": 0.44120135363790186, + "grad_norm": 8.75, + "learning_rate": 4.9556751155861674e-06, + "loss": 2.2569, + "step": 4172 + }, + { + "epoch": 0.44130710659898476, + "grad_norm": 7.71875, + "learning_rate": 4.955643004166696e-06, + "loss": 2.4705, + "step": 4173 + }, + { + "epoch": 0.4414128595600677, + "grad_norm": 7.75, + "learning_rate": 4.955610881223897e-06, + "loss": 2.6655, + "step": 4174 + }, + { + "epoch": 0.4415186125211506, + "grad_norm": 7.5625, + "learning_rate": 4.955578746757919e-06, + "loss": 2.6764, + "step": 4175 + }, + { + "epoch": 0.4416243654822335, + "grad_norm": 7.84375, + "learning_rate": 4.955546600768916e-06, + "loss": 2.6182, + "step": 4176 + }, + { + "epoch": 0.4417301184433164, + "grad_norm": 7.40625, + "learning_rate": 4.955514443257036e-06, + "loss": 2.5207, + "step": 4177 + }, + { + "epoch": 0.44183587140439934, + "grad_norm": 8.625, + "learning_rate": 4.955482274222431e-06, + "loss": 2.6327, + "step": 4178 + }, + { + "epoch": 0.44194162436548223, + "grad_norm": 8.0625, + "learning_rate": 4.955450093665252e-06, + "loss": 2.5703, + "step": 4179 + }, + { + "epoch": 0.4420473773265651, + "grad_norm": 8.6875, + "learning_rate": 4.95541790158565e-06, + "loss": 2.6406, + "step": 4180 + }, + { + "epoch": 0.4421531302876481, + "grad_norm": 8.0625, + "learning_rate": 4.955385697983776e-06, + "loss": 2.5353, + "step": 4181 + }, + { + "epoch": 0.44225888324873097, + "grad_norm": 8.75, + "learning_rate": 4.955353482859781e-06, + "loss": 2.6406, + "step": 4182 + }, + { + "epoch": 0.44236463620981387, + "grad_norm": 7.84375, + "learning_rate": 4.955321256213817e-06, + "loss": 2.7417, + "step": 4183 + }, + { + "epoch": 0.44247038917089676, + "grad_norm": 8.25, + "learning_rate": 4.955289018046034e-06, + "loss": 2.5179, + "step": 4184 + }, + { + "epoch": 0.4425761421319797, + "grad_norm": 7.4375, + "learning_rate": 4.9552567683565845e-06, + "loss": 2.5777, + "step": 4185 + }, + { + "epoch": 0.4426818950930626, + "grad_norm": 8.125, + "learning_rate": 4.955224507145618e-06, + "loss": 2.5638, + "step": 4186 + }, + { + "epoch": 0.4427876480541455, + "grad_norm": 8.0625, + "learning_rate": 4.955192234413288e-06, + "loss": 2.6849, + "step": 4187 + }, + { + "epoch": 0.44289340101522845, + "grad_norm": 7.6875, + "learning_rate": 4.955159950159744e-06, + "loss": 2.4597, + "step": 4188 + }, + { + "epoch": 0.44299915397631134, + "grad_norm": 8.125, + "learning_rate": 4.955127654385139e-06, + "loss": 2.5672, + "step": 4189 + }, + { + "epoch": 0.44310490693739424, + "grad_norm": 7.5, + "learning_rate": 4.955095347089625e-06, + "loss": 2.5176, + "step": 4190 + }, + { + "epoch": 0.44321065989847713, + "grad_norm": 8.125, + "learning_rate": 4.9550630282733514e-06, + "loss": 2.6988, + "step": 4191 + }, + { + "epoch": 0.4433164128595601, + "grad_norm": 7.59375, + "learning_rate": 4.955030697936472e-06, + "loss": 2.5185, + "step": 4192 + }, + { + "epoch": 0.443422165820643, + "grad_norm": 7.96875, + "learning_rate": 4.954998356079136e-06, + "loss": 2.5837, + "step": 4193 + }, + { + "epoch": 0.44352791878172587, + "grad_norm": 8.0625, + "learning_rate": 4.9549660027014976e-06, + "loss": 2.6181, + "step": 4194 + }, + { + "epoch": 0.4436336717428088, + "grad_norm": 8.1875, + "learning_rate": 4.954933637803709e-06, + "loss": 2.6122, + "step": 4195 + }, + { + "epoch": 0.4437394247038917, + "grad_norm": 7.40625, + "learning_rate": 4.954901261385919e-06, + "loss": 2.5697, + "step": 4196 + }, + { + "epoch": 0.4438451776649746, + "grad_norm": 7.71875, + "learning_rate": 4.954868873448283e-06, + "loss": 2.5541, + "step": 4197 + }, + { + "epoch": 0.4439509306260575, + "grad_norm": 7.96875, + "learning_rate": 4.95483647399095e-06, + "loss": 2.5115, + "step": 4198 + }, + { + "epoch": 0.44405668358714045, + "grad_norm": 7.6875, + "learning_rate": 4.954804063014074e-06, + "loss": 2.6273, + "step": 4199 + }, + { + "epoch": 0.44416243654822335, + "grad_norm": 8.1875, + "learning_rate": 4.954771640517807e-06, + "loss": 2.8232, + "step": 4200 + }, + { + "epoch": 0.44426818950930624, + "grad_norm": 7.625, + "learning_rate": 4.9547392065023e-06, + "loss": 2.681, + "step": 4201 + }, + { + "epoch": 0.4443739424703892, + "grad_norm": 7.78125, + "learning_rate": 4.954706760967706e-06, + "loss": 2.6302, + "step": 4202 + }, + { + "epoch": 0.4444796954314721, + "grad_norm": 8.625, + "learning_rate": 4.954674303914178e-06, + "loss": 2.7325, + "step": 4203 + }, + { + "epoch": 0.444585448392555, + "grad_norm": 8.1875, + "learning_rate": 4.954641835341866e-06, + "loss": 2.4083, + "step": 4204 + }, + { + "epoch": 0.4446912013536379, + "grad_norm": 7.78125, + "learning_rate": 4.954609355250924e-06, + "loss": 2.4662, + "step": 4205 + }, + { + "epoch": 0.4447969543147208, + "grad_norm": 7.0625, + "learning_rate": 4.954576863641505e-06, + "loss": 2.4765, + "step": 4206 + }, + { + "epoch": 0.4449027072758037, + "grad_norm": 8.3125, + "learning_rate": 4.95454436051376e-06, + "loss": 2.367, + "step": 4207 + }, + { + "epoch": 0.4450084602368866, + "grad_norm": 7.625, + "learning_rate": 4.954511845867842e-06, + "loss": 2.6242, + "step": 4208 + }, + { + "epoch": 0.44511421319796957, + "grad_norm": 8.0625, + "learning_rate": 4.9544793197039045e-06, + "loss": 2.5392, + "step": 4209 + }, + { + "epoch": 0.44521996615905246, + "grad_norm": 8.75, + "learning_rate": 4.954446782022099e-06, + "loss": 2.2693, + "step": 4210 + }, + { + "epoch": 0.44532571912013535, + "grad_norm": 7.75, + "learning_rate": 4.9544142328225794e-06, + "loss": 2.7571, + "step": 4211 + }, + { + "epoch": 0.44543147208121825, + "grad_norm": 7.75, + "learning_rate": 4.954381672105496e-06, + "loss": 2.6296, + "step": 4212 + }, + { + "epoch": 0.4455372250423012, + "grad_norm": 7.625, + "learning_rate": 4.9543490998710045e-06, + "loss": 2.5473, + "step": 4213 + }, + { + "epoch": 0.4456429780033841, + "grad_norm": 7.59375, + "learning_rate": 4.954316516119257e-06, + "loss": 2.4613, + "step": 4214 + }, + { + "epoch": 0.445748730964467, + "grad_norm": 7.90625, + "learning_rate": 4.954283920850404e-06, + "loss": 2.6354, + "step": 4215 + }, + { + "epoch": 0.44585448392554994, + "grad_norm": 7.4375, + "learning_rate": 4.954251314064602e-06, + "loss": 2.6164, + "step": 4216 + }, + { + "epoch": 0.44596023688663283, + "grad_norm": 7.5, + "learning_rate": 4.9542186957620015e-06, + "loss": 2.7023, + "step": 4217 + }, + { + "epoch": 0.4460659898477157, + "grad_norm": 8.125, + "learning_rate": 4.954186065942757e-06, + "loss": 2.5204, + "step": 4218 + }, + { + "epoch": 0.4461717428087986, + "grad_norm": 8.6875, + "learning_rate": 4.95415342460702e-06, + "loss": 2.6825, + "step": 4219 + }, + { + "epoch": 0.44627749576988157, + "grad_norm": 7.90625, + "learning_rate": 4.954120771754946e-06, + "loss": 2.7959, + "step": 4220 + }, + { + "epoch": 0.44638324873096447, + "grad_norm": 7.75, + "learning_rate": 4.954088107386686e-06, + "loss": 2.5883, + "step": 4221 + }, + { + "epoch": 0.44648900169204736, + "grad_norm": 7.28125, + "learning_rate": 4.954055431502393e-06, + "loss": 2.5699, + "step": 4222 + }, + { + "epoch": 0.4465947546531303, + "grad_norm": 8.125, + "learning_rate": 4.954022744102224e-06, + "loss": 2.5568, + "step": 4223 + }, + { + "epoch": 0.4467005076142132, + "grad_norm": 7.8125, + "learning_rate": 4.953990045186329e-06, + "loss": 2.4683, + "step": 4224 + }, + { + "epoch": 0.4468062605752961, + "grad_norm": 10.6875, + "learning_rate": 4.953957334754862e-06, + "loss": 2.9396, + "step": 4225 + }, + { + "epoch": 0.446912013536379, + "grad_norm": 8.3125, + "learning_rate": 4.953924612807977e-06, + "loss": 2.7132, + "step": 4226 + }, + { + "epoch": 0.44701776649746194, + "grad_norm": 8.1875, + "learning_rate": 4.953891879345828e-06, + "loss": 2.6586, + "step": 4227 + }, + { + "epoch": 0.44712351945854484, + "grad_norm": 7.53125, + "learning_rate": 4.953859134368567e-06, + "loss": 2.7658, + "step": 4228 + }, + { + "epoch": 0.44722927241962773, + "grad_norm": 8.0625, + "learning_rate": 4.95382637787635e-06, + "loss": 2.5795, + "step": 4229 + }, + { + "epoch": 0.4473350253807107, + "grad_norm": 7.90625, + "learning_rate": 4.953793609869328e-06, + "loss": 2.4152, + "step": 4230 + }, + { + "epoch": 0.4474407783417936, + "grad_norm": 7.90625, + "learning_rate": 4.953760830347657e-06, + "loss": 2.7416, + "step": 4231 + }, + { + "epoch": 0.44754653130287647, + "grad_norm": 7.78125, + "learning_rate": 4.95372803931149e-06, + "loss": 2.7261, + "step": 4232 + }, + { + "epoch": 0.44765228426395937, + "grad_norm": 7.96875, + "learning_rate": 4.95369523676098e-06, + "loss": 2.2687, + "step": 4233 + }, + { + "epoch": 0.4477580372250423, + "grad_norm": 7.6875, + "learning_rate": 4.953662422696282e-06, + "loss": 2.7312, + "step": 4234 + }, + { + "epoch": 0.4478637901861252, + "grad_norm": 8.25, + "learning_rate": 4.95362959711755e-06, + "loss": 2.431, + "step": 4235 + }, + { + "epoch": 0.4479695431472081, + "grad_norm": 10.1875, + "learning_rate": 4.953596760024938e-06, + "loss": 2.6411, + "step": 4236 + }, + { + "epoch": 0.44807529610829105, + "grad_norm": 7.96875, + "learning_rate": 4.9535639114186e-06, + "loss": 2.3864, + "step": 4237 + }, + { + "epoch": 0.44818104906937395, + "grad_norm": 7.5625, + "learning_rate": 4.95353105129869e-06, + "loss": 2.5062, + "step": 4238 + }, + { + "epoch": 0.44828680203045684, + "grad_norm": 7.46875, + "learning_rate": 4.953498179665362e-06, + "loss": 2.6385, + "step": 4239 + }, + { + "epoch": 0.44839255499153974, + "grad_norm": 7.71875, + "learning_rate": 4.95346529651877e-06, + "loss": 2.5165, + "step": 4240 + }, + { + "epoch": 0.4484983079526227, + "grad_norm": 8.0, + "learning_rate": 4.953432401859069e-06, + "loss": 2.8786, + "step": 4241 + }, + { + "epoch": 0.4486040609137056, + "grad_norm": 8.8125, + "learning_rate": 4.953399495686414e-06, + "loss": 3.0241, + "step": 4242 + }, + { + "epoch": 0.4487098138747885, + "grad_norm": 7.53125, + "learning_rate": 4.953366578000958e-06, + "loss": 2.6163, + "step": 4243 + }, + { + "epoch": 0.4488155668358714, + "grad_norm": 7.90625, + "learning_rate": 4.9533336488028565e-06, + "loss": 2.818, + "step": 4244 + }, + { + "epoch": 0.4489213197969543, + "grad_norm": 8.3125, + "learning_rate": 4.953300708092263e-06, + "loss": 2.5643, + "step": 4245 + }, + { + "epoch": 0.4490270727580372, + "grad_norm": 8.875, + "learning_rate": 4.953267755869333e-06, + "loss": 2.728, + "step": 4246 + }, + { + "epoch": 0.4491328257191201, + "grad_norm": 11.25, + "learning_rate": 4.95323479213422e-06, + "loss": 2.7123, + "step": 4247 + }, + { + "epoch": 0.44923857868020306, + "grad_norm": 7.625, + "learning_rate": 4.953201816887081e-06, + "loss": 2.5561, + "step": 4248 + }, + { + "epoch": 0.44934433164128595, + "grad_norm": 8.0625, + "learning_rate": 4.953168830128068e-06, + "loss": 2.819, + "step": 4249 + }, + { + "epoch": 0.44945008460236885, + "grad_norm": 8.0, + "learning_rate": 4.953135831857337e-06, + "loss": 2.5984, + "step": 4250 + }, + { + "epoch": 0.4495558375634518, + "grad_norm": 7.5625, + "learning_rate": 4.953102822075043e-06, + "loss": 2.4011, + "step": 4251 + }, + { + "epoch": 0.4496615905245347, + "grad_norm": 7.5, + "learning_rate": 4.953069800781342e-06, + "loss": 2.5353, + "step": 4252 + }, + { + "epoch": 0.4497673434856176, + "grad_norm": 8.125, + "learning_rate": 4.953036767976386e-06, + "loss": 2.4832, + "step": 4253 + }, + { + "epoch": 0.4498730964467005, + "grad_norm": 7.96875, + "learning_rate": 4.953003723660332e-06, + "loss": 2.6778, + "step": 4254 + }, + { + "epoch": 0.44997884940778343, + "grad_norm": 7.9375, + "learning_rate": 4.952970667833335e-06, + "loss": 2.3829, + "step": 4255 + }, + { + "epoch": 0.4500846023688663, + "grad_norm": 7.78125, + "learning_rate": 4.9529376004955495e-06, + "loss": 2.549, + "step": 4256 + }, + { + "epoch": 0.4501903553299492, + "grad_norm": 7.90625, + "learning_rate": 4.952904521647132e-06, + "loss": 2.6586, + "step": 4257 + }, + { + "epoch": 0.45029610829103217, + "grad_norm": 10.0, + "learning_rate": 4.952871431288237e-06, + "loss": 2.4148, + "step": 4258 + }, + { + "epoch": 0.45040186125211507, + "grad_norm": 7.25, + "learning_rate": 4.952838329419018e-06, + "loss": 2.5133, + "step": 4259 + }, + { + "epoch": 0.45050761421319796, + "grad_norm": 8.5625, + "learning_rate": 4.952805216039633e-06, + "loss": 2.5865, + "step": 4260 + }, + { + "epoch": 0.45061336717428085, + "grad_norm": 8.1875, + "learning_rate": 4.952772091150236e-06, + "loss": 2.6046, + "step": 4261 + }, + { + "epoch": 0.4507191201353638, + "grad_norm": 7.8125, + "learning_rate": 4.9527389547509825e-06, + "loss": 2.6115, + "step": 4262 + }, + { + "epoch": 0.4508248730964467, + "grad_norm": 7.28125, + "learning_rate": 4.952705806842029e-06, + "loss": 2.3798, + "step": 4263 + }, + { + "epoch": 0.4509306260575296, + "grad_norm": 7.5625, + "learning_rate": 4.952672647423529e-06, + "loss": 2.5144, + "step": 4264 + }, + { + "epoch": 0.45103637901861254, + "grad_norm": 7.96875, + "learning_rate": 4.95263947649564e-06, + "loss": 2.6315, + "step": 4265 + }, + { + "epoch": 0.45114213197969544, + "grad_norm": 7.5, + "learning_rate": 4.952606294058517e-06, + "loss": 2.507, + "step": 4266 + }, + { + "epoch": 0.45124788494077833, + "grad_norm": 8.375, + "learning_rate": 4.952573100112316e-06, + "loss": 2.599, + "step": 4267 + }, + { + "epoch": 0.4513536379018612, + "grad_norm": 9.125, + "learning_rate": 4.952539894657192e-06, + "loss": 2.4362, + "step": 4268 + }, + { + "epoch": 0.4514593908629442, + "grad_norm": 8.0625, + "learning_rate": 4.952506677693302e-06, + "loss": 2.8204, + "step": 4269 + }, + { + "epoch": 0.45156514382402707, + "grad_norm": 7.75, + "learning_rate": 4.952473449220801e-06, + "loss": 2.629, + "step": 4270 + }, + { + "epoch": 0.45167089678510997, + "grad_norm": 8.375, + "learning_rate": 4.952440209239845e-06, + "loss": 2.2582, + "step": 4271 + }, + { + "epoch": 0.4517766497461929, + "grad_norm": 7.71875, + "learning_rate": 4.9524069577505906e-06, + "loss": 2.509, + "step": 4272 + }, + { + "epoch": 0.4518824027072758, + "grad_norm": 7.375, + "learning_rate": 4.952373694753193e-06, + "loss": 2.5229, + "step": 4273 + }, + { + "epoch": 0.4519881556683587, + "grad_norm": 7.8125, + "learning_rate": 4.9523404202478085e-06, + "loss": 2.4218, + "step": 4274 + }, + { + "epoch": 0.4520939086294416, + "grad_norm": 7.75, + "learning_rate": 4.952307134234593e-06, + "loss": 2.5342, + "step": 4275 + }, + { + "epoch": 0.45219966159052455, + "grad_norm": 7.84375, + "learning_rate": 4.9522738367137036e-06, + "loss": 2.7181, + "step": 4276 + }, + { + "epoch": 0.45230541455160744, + "grad_norm": 8.375, + "learning_rate": 4.952240527685296e-06, + "loss": 2.6894, + "step": 4277 + }, + { + "epoch": 0.45241116751269034, + "grad_norm": 7.84375, + "learning_rate": 4.952207207149526e-06, + "loss": 2.6466, + "step": 4278 + }, + { + "epoch": 0.4525169204737733, + "grad_norm": 7.40625, + "learning_rate": 4.952173875106552e-06, + "loss": 2.5959, + "step": 4279 + }, + { + "epoch": 0.4526226734348562, + "grad_norm": 7.84375, + "learning_rate": 4.952140531556527e-06, + "loss": 2.5179, + "step": 4280 + }, + { + "epoch": 0.4527284263959391, + "grad_norm": 9.375, + "learning_rate": 4.952107176499611e-06, + "loss": 2.4561, + "step": 4281 + }, + { + "epoch": 0.45283417935702197, + "grad_norm": 8.3125, + "learning_rate": 4.952073809935958e-06, + "loss": 2.5815, + "step": 4282 + }, + { + "epoch": 0.4529399323181049, + "grad_norm": 8.25, + "learning_rate": 4.952040431865725e-06, + "loss": 2.2628, + "step": 4283 + }, + { + "epoch": 0.4530456852791878, + "grad_norm": 7.875, + "learning_rate": 4.95200704228907e-06, + "loss": 2.7169, + "step": 4284 + }, + { + "epoch": 0.4531514382402707, + "grad_norm": 7.5625, + "learning_rate": 4.951973641206148e-06, + "loss": 2.6712, + "step": 4285 + }, + { + "epoch": 0.45325719120135366, + "grad_norm": 8.25, + "learning_rate": 4.951940228617117e-06, + "loss": 2.5788, + "step": 4286 + }, + { + "epoch": 0.45336294416243655, + "grad_norm": 7.1875, + "learning_rate": 4.951906804522134e-06, + "loss": 2.4922, + "step": 4287 + }, + { + "epoch": 0.45346869712351945, + "grad_norm": 9.75, + "learning_rate": 4.951873368921354e-06, + "loss": 2.706, + "step": 4288 + }, + { + "epoch": 0.45357445008460234, + "grad_norm": 7.34375, + "learning_rate": 4.951839921814936e-06, + "loss": 2.432, + "step": 4289 + }, + { + "epoch": 0.4536802030456853, + "grad_norm": 7.59375, + "learning_rate": 4.951806463203035e-06, + "loss": 2.1891, + "step": 4290 + }, + { + "epoch": 0.4537859560067682, + "grad_norm": 7.71875, + "learning_rate": 4.951772993085809e-06, + "loss": 2.8116, + "step": 4291 + }, + { + "epoch": 0.4538917089678511, + "grad_norm": 8.25, + "learning_rate": 4.951739511463416e-06, + "loss": 2.805, + "step": 4292 + }, + { + "epoch": 0.45399746192893403, + "grad_norm": 8.125, + "learning_rate": 4.951706018336012e-06, + "loss": 2.7021, + "step": 4293 + }, + { + "epoch": 0.4541032148900169, + "grad_norm": 7.59375, + "learning_rate": 4.951672513703755e-06, + "loss": 2.5478, + "step": 4294 + }, + { + "epoch": 0.4542089678510998, + "grad_norm": 8.1875, + "learning_rate": 4.951638997566801e-06, + "loss": 2.6036, + "step": 4295 + }, + { + "epoch": 0.4543147208121827, + "grad_norm": 7.9375, + "learning_rate": 4.951605469925307e-06, + "loss": 2.606, + "step": 4296 + }, + { + "epoch": 0.45442047377326567, + "grad_norm": 7.625, + "learning_rate": 4.951571930779432e-06, + "loss": 2.489, + "step": 4297 + }, + { + "epoch": 0.45452622673434856, + "grad_norm": 8.0, + "learning_rate": 4.9515383801293324e-06, + "loss": 2.6529, + "step": 4298 + }, + { + "epoch": 0.45463197969543145, + "grad_norm": 8.0625, + "learning_rate": 4.951504817975166e-06, + "loss": 2.6531, + "step": 4299 + }, + { + "epoch": 0.4547377326565144, + "grad_norm": 7.53125, + "learning_rate": 4.9514712443170896e-06, + "loss": 2.4914, + "step": 4300 + }, + { + "epoch": 0.4548434856175973, + "grad_norm": 7.75, + "learning_rate": 4.951437659155263e-06, + "loss": 2.6146, + "step": 4301 + }, + { + "epoch": 0.4549492385786802, + "grad_norm": 8.125, + "learning_rate": 4.95140406248984e-06, + "loss": 2.7289, + "step": 4302 + }, + { + "epoch": 0.4550549915397631, + "grad_norm": 8.5625, + "learning_rate": 4.951370454320981e-06, + "loss": 2.7869, + "step": 4303 + }, + { + "epoch": 0.45516074450084604, + "grad_norm": 7.625, + "learning_rate": 4.951336834648843e-06, + "loss": 2.59, + "step": 4304 + }, + { + "epoch": 0.45526649746192893, + "grad_norm": 7.90625, + "learning_rate": 4.951303203473584e-06, + "loss": 2.6682, + "step": 4305 + }, + { + "epoch": 0.4553722504230118, + "grad_norm": 7.25, + "learning_rate": 4.951269560795362e-06, + "loss": 2.6694, + "step": 4306 + }, + { + "epoch": 0.4554780033840948, + "grad_norm": 8.375, + "learning_rate": 4.951235906614334e-06, + "loss": 2.6588, + "step": 4307 + }, + { + "epoch": 0.45558375634517767, + "grad_norm": 7.96875, + "learning_rate": 4.951202240930658e-06, + "loss": 2.6782, + "step": 4308 + }, + { + "epoch": 0.45568950930626057, + "grad_norm": 7.90625, + "learning_rate": 4.9511685637444925e-06, + "loss": 2.2773, + "step": 4309 + }, + { + "epoch": 0.45579526226734346, + "grad_norm": 8.6875, + "learning_rate": 4.951134875055996e-06, + "loss": 2.8261, + "step": 4310 + }, + { + "epoch": 0.4559010152284264, + "grad_norm": 7.40625, + "learning_rate": 4.951101174865326e-06, + "loss": 2.4408, + "step": 4311 + }, + { + "epoch": 0.4560067681895093, + "grad_norm": 7.75, + "learning_rate": 4.951067463172641e-06, + "loss": 2.6357, + "step": 4312 + }, + { + "epoch": 0.4561125211505922, + "grad_norm": 7.78125, + "learning_rate": 4.951033739978098e-06, + "loss": 2.5946, + "step": 4313 + }, + { + "epoch": 0.45621827411167515, + "grad_norm": 7.78125, + "learning_rate": 4.951000005281857e-06, + "loss": 2.6195, + "step": 4314 + }, + { + "epoch": 0.45632402707275804, + "grad_norm": 8.1875, + "learning_rate": 4.950966259084074e-06, + "loss": 2.5979, + "step": 4315 + }, + { + "epoch": 0.45642978003384094, + "grad_norm": 7.9375, + "learning_rate": 4.95093250138491e-06, + "loss": 2.7241, + "step": 4316 + }, + { + "epoch": 0.45653553299492383, + "grad_norm": 8.3125, + "learning_rate": 4.9508987321845214e-06, + "loss": 2.4492, + "step": 4317 + }, + { + "epoch": 0.4566412859560068, + "grad_norm": 8.1875, + "learning_rate": 4.950864951483068e-06, + "loss": 2.6139, + "step": 4318 + }, + { + "epoch": 0.4567470389170897, + "grad_norm": 7.4375, + "learning_rate": 4.950831159280708e-06, + "loss": 2.6471, + "step": 4319 + }, + { + "epoch": 0.45685279187817257, + "grad_norm": 7.46875, + "learning_rate": 4.9507973555775985e-06, + "loss": 2.698, + "step": 4320 + }, + { + "epoch": 0.4569585448392555, + "grad_norm": 8.0625, + "learning_rate": 4.950763540373899e-06, + "loss": 2.6658, + "step": 4321 + }, + { + "epoch": 0.4570642978003384, + "grad_norm": 7.9375, + "learning_rate": 4.9507297136697705e-06, + "loss": 2.8791, + "step": 4322 + }, + { + "epoch": 0.4571700507614213, + "grad_norm": 7.3125, + "learning_rate": 4.950695875465369e-06, + "loss": 2.5501, + "step": 4323 + }, + { + "epoch": 0.4572758037225042, + "grad_norm": 8.25, + "learning_rate": 4.950662025760854e-06, + "loss": 2.5418, + "step": 4324 + }, + { + "epoch": 0.45738155668358715, + "grad_norm": 7.59375, + "learning_rate": 4.950628164556384e-06, + "loss": 2.349, + "step": 4325 + }, + { + "epoch": 0.45748730964467005, + "grad_norm": 8.6875, + "learning_rate": 4.950594291852119e-06, + "loss": 2.5562, + "step": 4326 + }, + { + "epoch": 0.45759306260575294, + "grad_norm": 8.125, + "learning_rate": 4.950560407648216e-06, + "loss": 2.5755, + "step": 4327 + }, + { + "epoch": 0.4576988155668359, + "grad_norm": 7.9375, + "learning_rate": 4.950526511944837e-06, + "loss": 2.6661, + "step": 4328 + }, + { + "epoch": 0.4578045685279188, + "grad_norm": 8.4375, + "learning_rate": 4.950492604742138e-06, + "loss": 2.4812, + "step": 4329 + }, + { + "epoch": 0.4579103214890017, + "grad_norm": 7.75, + "learning_rate": 4.95045868604028e-06, + "loss": 2.5956, + "step": 4330 + }, + { + "epoch": 0.4580160744500846, + "grad_norm": 8.125, + "learning_rate": 4.950424755839422e-06, + "loss": 2.5336, + "step": 4331 + }, + { + "epoch": 0.4581218274111675, + "grad_norm": 7.78125, + "learning_rate": 4.9503908141397215e-06, + "loss": 2.5785, + "step": 4332 + }, + { + "epoch": 0.4582275803722504, + "grad_norm": 7.46875, + "learning_rate": 4.95035686094134e-06, + "loss": 2.633, + "step": 4333 + }, + { + "epoch": 0.4583333333333333, + "grad_norm": 8.3125, + "learning_rate": 4.950322896244436e-06, + "loss": 2.4712, + "step": 4334 + }, + { + "epoch": 0.45843908629441626, + "grad_norm": 7.65625, + "learning_rate": 4.950288920049169e-06, + "loss": 2.6041, + "step": 4335 + }, + { + "epoch": 0.45854483925549916, + "grad_norm": 7.3125, + "learning_rate": 4.950254932355697e-06, + "loss": 2.3551, + "step": 4336 + }, + { + "epoch": 0.45865059221658205, + "grad_norm": 7.59375, + "learning_rate": 4.950220933164181e-06, + "loss": 2.7245, + "step": 4337 + }, + { + "epoch": 0.45875634517766495, + "grad_norm": 7.875, + "learning_rate": 4.950186922474781e-06, + "loss": 2.6563, + "step": 4338 + }, + { + "epoch": 0.4588620981387479, + "grad_norm": 9.25, + "learning_rate": 4.950152900287656e-06, + "loss": 2.8952, + "step": 4339 + }, + { + "epoch": 0.4589678510998308, + "grad_norm": 8.0, + "learning_rate": 4.950118866602965e-06, + "loss": 2.4692, + "step": 4340 + }, + { + "epoch": 0.4590736040609137, + "grad_norm": 8.125, + "learning_rate": 4.950084821420869e-06, + "loss": 2.5891, + "step": 4341 + }, + { + "epoch": 0.45917935702199664, + "grad_norm": 7.28125, + "learning_rate": 4.9500507647415265e-06, + "loss": 2.5664, + "step": 4342 + }, + { + "epoch": 0.45928510998307953, + "grad_norm": 8.0625, + "learning_rate": 4.950016696565098e-06, + "loss": 2.5906, + "step": 4343 + }, + { + "epoch": 0.4593908629441624, + "grad_norm": 8.75, + "learning_rate": 4.949982616891742e-06, + "loss": 2.468, + "step": 4344 + }, + { + "epoch": 0.4594966159052453, + "grad_norm": 8.0625, + "learning_rate": 4.949948525721621e-06, + "loss": 2.6734, + "step": 4345 + }, + { + "epoch": 0.45960236886632827, + "grad_norm": 7.65625, + "learning_rate": 4.949914423054892e-06, + "loss": 2.669, + "step": 4346 + }, + { + "epoch": 0.45970812182741116, + "grad_norm": 7.90625, + "learning_rate": 4.949880308891718e-06, + "loss": 2.7574, + "step": 4347 + }, + { + "epoch": 0.45981387478849406, + "grad_norm": 7.9375, + "learning_rate": 4.949846183232257e-06, + "loss": 2.6044, + "step": 4348 + }, + { + "epoch": 0.459919627749577, + "grad_norm": 7.625, + "learning_rate": 4.94981204607667e-06, + "loss": 2.4819, + "step": 4349 + }, + { + "epoch": 0.4600253807106599, + "grad_norm": 8.5, + "learning_rate": 4.9497778974251176e-06, + "loss": 2.8575, + "step": 4350 + }, + { + "epoch": 0.4601311336717428, + "grad_norm": 8.0625, + "learning_rate": 4.949743737277759e-06, + "loss": 2.6676, + "step": 4351 + }, + { + "epoch": 0.4602368866328257, + "grad_norm": 7.8125, + "learning_rate": 4.949709565634755e-06, + "loss": 2.5525, + "step": 4352 + }, + { + "epoch": 0.46034263959390864, + "grad_norm": 7.5625, + "learning_rate": 4.949675382496265e-06, + "loss": 2.6756, + "step": 4353 + }, + { + "epoch": 0.46044839255499154, + "grad_norm": 7.40625, + "learning_rate": 4.949641187862452e-06, + "loss": 2.6185, + "step": 4354 + }, + { + "epoch": 0.46055414551607443, + "grad_norm": 8.3125, + "learning_rate": 4.949606981733474e-06, + "loss": 2.6979, + "step": 4355 + }, + { + "epoch": 0.4606598984771574, + "grad_norm": 7.6875, + "learning_rate": 4.949572764109492e-06, + "loss": 2.647, + "step": 4356 + }, + { + "epoch": 0.4607656514382403, + "grad_norm": 7.9375, + "learning_rate": 4.949538534990667e-06, + "loss": 2.5883, + "step": 4357 + }, + { + "epoch": 0.46087140439932317, + "grad_norm": 8.375, + "learning_rate": 4.949504294377159e-06, + "loss": 2.6959, + "step": 4358 + }, + { + "epoch": 0.46097715736040606, + "grad_norm": 7.46875, + "learning_rate": 4.94947004226913e-06, + "loss": 2.5065, + "step": 4359 + }, + { + "epoch": 0.461082910321489, + "grad_norm": 7.875, + "learning_rate": 4.949435778666739e-06, + "loss": 2.6789, + "step": 4360 + }, + { + "epoch": 0.4611886632825719, + "grad_norm": 8.625, + "learning_rate": 4.9494015035701485e-06, + "loss": 2.5318, + "step": 4361 + }, + { + "epoch": 0.4612944162436548, + "grad_norm": 7.65625, + "learning_rate": 4.949367216979518e-06, + "loss": 2.2827, + "step": 4362 + }, + { + "epoch": 0.46140016920473775, + "grad_norm": 7.71875, + "learning_rate": 4.949332918895009e-06, + "loss": 2.6414, + "step": 4363 + }, + { + "epoch": 0.46150592216582065, + "grad_norm": 7.78125, + "learning_rate": 4.949298609316782e-06, + "loss": 2.6011, + "step": 4364 + }, + { + "epoch": 0.46161167512690354, + "grad_norm": 7.71875, + "learning_rate": 4.9492642882449994e-06, + "loss": 2.8706, + "step": 4365 + }, + { + "epoch": 0.46171742808798644, + "grad_norm": 7.78125, + "learning_rate": 4.9492299556798205e-06, + "loss": 2.5198, + "step": 4366 + }, + { + "epoch": 0.4618231810490694, + "grad_norm": 8.6875, + "learning_rate": 4.949195611621408e-06, + "loss": 2.7219, + "step": 4367 + }, + { + "epoch": 0.4619289340101523, + "grad_norm": 8.1875, + "learning_rate": 4.949161256069921e-06, + "loss": 2.7294, + "step": 4368 + }, + { + "epoch": 0.4620346869712352, + "grad_norm": 7.625, + "learning_rate": 4.949126889025522e-06, + "loss": 2.5226, + "step": 4369 + }, + { + "epoch": 0.4621404399323181, + "grad_norm": 8.3125, + "learning_rate": 4.949092510488373e-06, + "loss": 2.4455, + "step": 4370 + }, + { + "epoch": 0.462246192893401, + "grad_norm": 7.4375, + "learning_rate": 4.949058120458634e-06, + "loss": 2.6057, + "step": 4371 + }, + { + "epoch": 0.4623519458544839, + "grad_norm": 8.125, + "learning_rate": 4.949023718936466e-06, + "loss": 2.4781, + "step": 4372 + }, + { + "epoch": 0.4624576988155668, + "grad_norm": 9.9375, + "learning_rate": 4.948989305922033e-06, + "loss": 2.3789, + "step": 4373 + }, + { + "epoch": 0.46256345177664976, + "grad_norm": 8.125, + "learning_rate": 4.948954881415494e-06, + "loss": 2.5861, + "step": 4374 + }, + { + "epoch": 0.46266920473773265, + "grad_norm": 8.3125, + "learning_rate": 4.948920445417012e-06, + "loss": 2.6093, + "step": 4375 + }, + { + "epoch": 0.46277495769881555, + "grad_norm": 7.84375, + "learning_rate": 4.9488859979267465e-06, + "loss": 2.5153, + "step": 4376 + }, + { + "epoch": 0.4628807106598985, + "grad_norm": 8.25, + "learning_rate": 4.948851538944861e-06, + "loss": 2.6833, + "step": 4377 + }, + { + "epoch": 0.4629864636209814, + "grad_norm": 8.375, + "learning_rate": 4.9488170684715174e-06, + "loss": 2.6981, + "step": 4378 + }, + { + "epoch": 0.4630922165820643, + "grad_norm": 7.71875, + "learning_rate": 4.9487825865068765e-06, + "loss": 2.5916, + "step": 4379 + }, + { + "epoch": 0.4631979695431472, + "grad_norm": 8.125, + "learning_rate": 4.9487480930511e-06, + "loss": 2.5395, + "step": 4380 + }, + { + "epoch": 0.46330372250423013, + "grad_norm": 15.625, + "learning_rate": 4.948713588104351e-06, + "loss": 2.5831, + "step": 4381 + }, + { + "epoch": 0.463409475465313, + "grad_norm": 8.0625, + "learning_rate": 4.94867907166679e-06, + "loss": 2.5972, + "step": 4382 + }, + { + "epoch": 0.4635152284263959, + "grad_norm": 8.1875, + "learning_rate": 4.94864454373858e-06, + "loss": 2.6647, + "step": 4383 + }, + { + "epoch": 0.46362098138747887, + "grad_norm": 8.875, + "learning_rate": 4.948610004319883e-06, + "loss": 2.7137, + "step": 4384 + }, + { + "epoch": 0.46372673434856176, + "grad_norm": 7.84375, + "learning_rate": 4.94857545341086e-06, + "loss": 2.6393, + "step": 4385 + }, + { + "epoch": 0.46383248730964466, + "grad_norm": 7.8125, + "learning_rate": 4.948540891011674e-06, + "loss": 2.4978, + "step": 4386 + }, + { + "epoch": 0.46393824027072755, + "grad_norm": 8.0625, + "learning_rate": 4.948506317122487e-06, + "loss": 2.6754, + "step": 4387 + }, + { + "epoch": 0.4640439932318105, + "grad_norm": 8.75, + "learning_rate": 4.948471731743461e-06, + "loss": 2.9051, + "step": 4388 + }, + { + "epoch": 0.4641497461928934, + "grad_norm": 7.96875, + "learning_rate": 4.94843713487476e-06, + "loss": 2.7282, + "step": 4389 + }, + { + "epoch": 0.4642554991539763, + "grad_norm": 8.1875, + "learning_rate": 4.948402526516543e-06, + "loss": 2.6413, + "step": 4390 + }, + { + "epoch": 0.46436125211505924, + "grad_norm": 7.5625, + "learning_rate": 4.948367906668975e-06, + "loss": 2.6621, + "step": 4391 + }, + { + "epoch": 0.46446700507614214, + "grad_norm": 8.375, + "learning_rate": 4.9483332753322176e-06, + "loss": 2.6025, + "step": 4392 + }, + { + "epoch": 0.46457275803722503, + "grad_norm": 7.84375, + "learning_rate": 4.948298632506434e-06, + "loss": 2.6406, + "step": 4393 + }, + { + "epoch": 0.4646785109983079, + "grad_norm": 7.75, + "learning_rate": 4.948263978191785e-06, + "loss": 2.6584, + "step": 4394 + }, + { + "epoch": 0.4647842639593909, + "grad_norm": 7.6875, + "learning_rate": 4.9482293123884365e-06, + "loss": 2.6173, + "step": 4395 + }, + { + "epoch": 0.46489001692047377, + "grad_norm": 8.0625, + "learning_rate": 4.948194635096548e-06, + "loss": 2.717, + "step": 4396 + }, + { + "epoch": 0.46499576988155666, + "grad_norm": 8.75, + "learning_rate": 4.9481599463162835e-06, + "loss": 2.6908, + "step": 4397 + }, + { + "epoch": 0.4651015228426396, + "grad_norm": 7.875, + "learning_rate": 4.948125246047805e-06, + "loss": 2.62, + "step": 4398 + }, + { + "epoch": 0.4652072758037225, + "grad_norm": 7.65625, + "learning_rate": 4.948090534291277e-06, + "loss": 2.6496, + "step": 4399 + }, + { + "epoch": 0.4653130287648054, + "grad_norm": 8.1875, + "learning_rate": 4.94805581104686e-06, + "loss": 2.7623, + "step": 4400 + }, + { + "epoch": 0.4654187817258883, + "grad_norm": 8.0, + "learning_rate": 4.94802107631472e-06, + "loss": 2.5705, + "step": 4401 + }, + { + "epoch": 0.46552453468697125, + "grad_norm": 9.125, + "learning_rate": 4.947986330095017e-06, + "loss": 2.6473, + "step": 4402 + }, + { + "epoch": 0.46563028764805414, + "grad_norm": 7.25, + "learning_rate": 4.947951572387916e-06, + "loss": 2.427, + "step": 4403 + }, + { + "epoch": 0.46573604060913704, + "grad_norm": 7.75, + "learning_rate": 4.947916803193579e-06, + "loss": 2.6847, + "step": 4404 + }, + { + "epoch": 0.46584179357022, + "grad_norm": 8.1875, + "learning_rate": 4.94788202251217e-06, + "loss": 2.4156, + "step": 4405 + }, + { + "epoch": 0.4659475465313029, + "grad_norm": 8.3125, + "learning_rate": 4.947847230343851e-06, + "loss": 2.6023, + "step": 4406 + }, + { + "epoch": 0.4660532994923858, + "grad_norm": 8.75, + "learning_rate": 4.947812426688787e-06, + "loss": 2.6214, + "step": 4407 + }, + { + "epoch": 0.46615905245346867, + "grad_norm": 7.8125, + "learning_rate": 4.94777761154714e-06, + "loss": 2.4756, + "step": 4408 + }, + { + "epoch": 0.4662648054145516, + "grad_norm": 7.25, + "learning_rate": 4.947742784919074e-06, + "loss": 2.2961, + "step": 4409 + }, + { + "epoch": 0.4663705583756345, + "grad_norm": 7.75, + "learning_rate": 4.947707946804752e-06, + "loss": 2.7582, + "step": 4410 + }, + { + "epoch": 0.4664763113367174, + "grad_norm": 8.0625, + "learning_rate": 4.947673097204338e-06, + "loss": 2.4435, + "step": 4411 + }, + { + "epoch": 0.46658206429780036, + "grad_norm": 7.0625, + "learning_rate": 4.947638236117995e-06, + "loss": 2.3149, + "step": 4412 + }, + { + "epoch": 0.46668781725888325, + "grad_norm": 8.375, + "learning_rate": 4.947603363545887e-06, + "loss": 2.647, + "step": 4413 + }, + { + "epoch": 0.46679357021996615, + "grad_norm": 7.6875, + "learning_rate": 4.947568479488178e-06, + "loss": 2.6029, + "step": 4414 + }, + { + "epoch": 0.46689932318104904, + "grad_norm": 9.125, + "learning_rate": 4.94753358394503e-06, + "loss": 2.6546, + "step": 4415 + }, + { + "epoch": 0.467005076142132, + "grad_norm": 7.96875, + "learning_rate": 4.947498676916608e-06, + "loss": 2.5659, + "step": 4416 + }, + { + "epoch": 0.4671108291032149, + "grad_norm": 7.21875, + "learning_rate": 4.947463758403076e-06, + "loss": 2.5943, + "step": 4417 + }, + { + "epoch": 0.4672165820642978, + "grad_norm": 7.78125, + "learning_rate": 4.947428828404597e-06, + "loss": 2.6695, + "step": 4418 + }, + { + "epoch": 0.46732233502538073, + "grad_norm": 8.0625, + "learning_rate": 4.947393886921337e-06, + "loss": 2.5697, + "step": 4419 + }, + { + "epoch": 0.4674280879864636, + "grad_norm": 8.375, + "learning_rate": 4.947358933953457e-06, + "loss": 2.8206, + "step": 4420 + }, + { + "epoch": 0.4675338409475465, + "grad_norm": 8.0, + "learning_rate": 4.947323969501123e-06, + "loss": 2.7385, + "step": 4421 + }, + { + "epoch": 0.4676395939086294, + "grad_norm": 8.25, + "learning_rate": 4.947288993564499e-06, + "loss": 2.6594, + "step": 4422 + }, + { + "epoch": 0.46774534686971236, + "grad_norm": 9.375, + "learning_rate": 4.947254006143748e-06, + "loss": 2.4975, + "step": 4423 + }, + { + "epoch": 0.46785109983079526, + "grad_norm": 7.96875, + "learning_rate": 4.947219007239035e-06, + "loss": 2.7508, + "step": 4424 + }, + { + "epoch": 0.46795685279187815, + "grad_norm": 8.0625, + "learning_rate": 4.947183996850523e-06, + "loss": 2.476, + "step": 4425 + }, + { + "epoch": 0.4680626057529611, + "grad_norm": 7.625, + "learning_rate": 4.947148974978379e-06, + "loss": 2.6402, + "step": 4426 + }, + { + "epoch": 0.468168358714044, + "grad_norm": 7.84375, + "learning_rate": 4.947113941622765e-06, + "loss": 2.4765, + "step": 4427 + }, + { + "epoch": 0.4682741116751269, + "grad_norm": 7.34375, + "learning_rate": 4.947078896783846e-06, + "loss": 2.6372, + "step": 4428 + }, + { + "epoch": 0.4683798646362098, + "grad_norm": 7.96875, + "learning_rate": 4.947043840461787e-06, + "loss": 2.6384, + "step": 4429 + }, + { + "epoch": 0.46848561759729274, + "grad_norm": 8.125, + "learning_rate": 4.9470087726567515e-06, + "loss": 2.6377, + "step": 4430 + }, + { + "epoch": 0.46859137055837563, + "grad_norm": 7.78125, + "learning_rate": 4.946973693368905e-06, + "loss": 2.4304, + "step": 4431 + }, + { + "epoch": 0.4686971235194585, + "grad_norm": 7.90625, + "learning_rate": 4.946938602598412e-06, + "loss": 2.3123, + "step": 4432 + }, + { + "epoch": 0.4688028764805415, + "grad_norm": 7.75, + "learning_rate": 4.9469035003454366e-06, + "loss": 2.7273, + "step": 4433 + }, + { + "epoch": 0.46890862944162437, + "grad_norm": 8.3125, + "learning_rate": 4.946868386610144e-06, + "loss": 2.499, + "step": 4434 + }, + { + "epoch": 0.46901438240270726, + "grad_norm": 8.1875, + "learning_rate": 4.946833261392698e-06, + "loss": 2.6704, + "step": 4435 + }, + { + "epoch": 0.4691201353637902, + "grad_norm": 7.1875, + "learning_rate": 4.946798124693265e-06, + "loss": 2.5696, + "step": 4436 + }, + { + "epoch": 0.4692258883248731, + "grad_norm": 9.8125, + "learning_rate": 4.9467629765120096e-06, + "loss": 2.3482, + "step": 4437 + }, + { + "epoch": 0.469331641285956, + "grad_norm": 7.6875, + "learning_rate": 4.946727816849096e-06, + "loss": 2.633, + "step": 4438 + }, + { + "epoch": 0.4694373942470389, + "grad_norm": 8.1875, + "learning_rate": 4.946692645704688e-06, + "loss": 2.6744, + "step": 4439 + }, + { + "epoch": 0.46954314720812185, + "grad_norm": 8.3125, + "learning_rate": 4.946657463078954e-06, + "loss": 2.6514, + "step": 4440 + }, + { + "epoch": 0.46964890016920474, + "grad_norm": 7.1875, + "learning_rate": 4.946622268972056e-06, + "loss": 2.6107, + "step": 4441 + }, + { + "epoch": 0.46975465313028764, + "grad_norm": 7.6875, + "learning_rate": 4.94658706338416e-06, + "loss": 2.598, + "step": 4442 + }, + { + "epoch": 0.4698604060913706, + "grad_norm": 7.71875, + "learning_rate": 4.946551846315433e-06, + "loss": 2.7085, + "step": 4443 + }, + { + "epoch": 0.4699661590524535, + "grad_norm": 7.84375, + "learning_rate": 4.9465166177660376e-06, + "loss": 2.6226, + "step": 4444 + }, + { + "epoch": 0.4700719120135364, + "grad_norm": 8.1875, + "learning_rate": 4.946481377736141e-06, + "loss": 2.4746, + "step": 4445 + }, + { + "epoch": 0.47017766497461927, + "grad_norm": 7.09375, + "learning_rate": 4.9464461262259075e-06, + "loss": 2.4494, + "step": 4446 + }, + { + "epoch": 0.4702834179357022, + "grad_norm": 8.1875, + "learning_rate": 4.946410863235502e-06, + "loss": 2.5575, + "step": 4447 + }, + { + "epoch": 0.4703891708967851, + "grad_norm": 7.46875, + "learning_rate": 4.946375588765092e-06, + "loss": 2.4508, + "step": 4448 + }, + { + "epoch": 0.470494923857868, + "grad_norm": 7.6875, + "learning_rate": 4.946340302814842e-06, + "loss": 2.837, + "step": 4449 + }, + { + "epoch": 0.47060067681895096, + "grad_norm": 8.1875, + "learning_rate": 4.946305005384917e-06, + "loss": 2.672, + "step": 4450 + }, + { + "epoch": 0.47070642978003385, + "grad_norm": 7.59375, + "learning_rate": 4.946269696475483e-06, + "loss": 2.5447, + "step": 4451 + }, + { + "epoch": 0.47081218274111675, + "grad_norm": 7.84375, + "learning_rate": 4.946234376086707e-06, + "loss": 2.7622, + "step": 4452 + }, + { + "epoch": 0.47091793570219964, + "grad_norm": 7.875, + "learning_rate": 4.946199044218753e-06, + "loss": 2.42, + "step": 4453 + }, + { + "epoch": 0.4710236886632826, + "grad_norm": 7.75, + "learning_rate": 4.946163700871787e-06, + "loss": 2.4439, + "step": 4454 + }, + { + "epoch": 0.4711294416243655, + "grad_norm": 7.59375, + "learning_rate": 4.9461283460459745e-06, + "loss": 2.8076, + "step": 4455 + }, + { + "epoch": 0.4712351945854484, + "grad_norm": 7.5625, + "learning_rate": 4.9460929797414825e-06, + "loss": 2.2777, + "step": 4456 + }, + { + "epoch": 0.47134094754653133, + "grad_norm": 7.84375, + "learning_rate": 4.946057601958477e-06, + "loss": 2.6639, + "step": 4457 + }, + { + "epoch": 0.4714467005076142, + "grad_norm": 7.90625, + "learning_rate": 4.946022212697124e-06, + "loss": 2.4901, + "step": 4458 + }, + { + "epoch": 0.4715524534686971, + "grad_norm": 7.375, + "learning_rate": 4.945986811957588e-06, + "loss": 2.6513, + "step": 4459 + }, + { + "epoch": 0.47165820642978, + "grad_norm": 8.4375, + "learning_rate": 4.945951399740037e-06, + "loss": 2.7297, + "step": 4460 + }, + { + "epoch": 0.47176395939086296, + "grad_norm": 8.375, + "learning_rate": 4.945915976044636e-06, + "loss": 2.233, + "step": 4461 + }, + { + "epoch": 0.47186971235194586, + "grad_norm": 8.0625, + "learning_rate": 4.9458805408715525e-06, + "loss": 2.46, + "step": 4462 + }, + { + "epoch": 0.47197546531302875, + "grad_norm": 7.46875, + "learning_rate": 4.9458450942209515e-06, + "loss": 2.5536, + "step": 4463 + }, + { + "epoch": 0.4720812182741117, + "grad_norm": 8.0625, + "learning_rate": 4.945809636092999e-06, + "loss": 2.5188, + "step": 4464 + }, + { + "epoch": 0.4721869712351946, + "grad_norm": 7.84375, + "learning_rate": 4.945774166487864e-06, + "loss": 2.6226, + "step": 4465 + }, + { + "epoch": 0.4722927241962775, + "grad_norm": 7.6875, + "learning_rate": 4.945738685405709e-06, + "loss": 2.2713, + "step": 4466 + }, + { + "epoch": 0.4723984771573604, + "grad_norm": 8.625, + "learning_rate": 4.945703192846704e-06, + "loss": 2.8993, + "step": 4467 + }, + { + "epoch": 0.47250423011844334, + "grad_norm": 7.9375, + "learning_rate": 4.945667688811013e-06, + "loss": 2.6566, + "step": 4468 + }, + { + "epoch": 0.47260998307952623, + "grad_norm": 7.71875, + "learning_rate": 4.945632173298805e-06, + "loss": 2.638, + "step": 4469 + }, + { + "epoch": 0.4727157360406091, + "grad_norm": 8.0, + "learning_rate": 4.945596646310245e-06, + "loss": 2.4586, + "step": 4470 + }, + { + "epoch": 0.4728214890016921, + "grad_norm": 7.5, + "learning_rate": 4.9455611078455e-06, + "loss": 2.6689, + "step": 4471 + }, + { + "epoch": 0.47292724196277497, + "grad_norm": 8.125, + "learning_rate": 4.945525557904737e-06, + "loss": 2.7095, + "step": 4472 + }, + { + "epoch": 0.47303299492385786, + "grad_norm": 8.5625, + "learning_rate": 4.945489996488123e-06, + "loss": 2.352, + "step": 4473 + }, + { + "epoch": 0.47313874788494076, + "grad_norm": 8.625, + "learning_rate": 4.945454423595825e-06, + "loss": 2.5928, + "step": 4474 + }, + { + "epoch": 0.4732445008460237, + "grad_norm": 7.78125, + "learning_rate": 4.945418839228008e-06, + "loss": 2.4716, + "step": 4475 + }, + { + "epoch": 0.4733502538071066, + "grad_norm": 7.59375, + "learning_rate": 4.945383243384841e-06, + "loss": 2.6093, + "step": 4476 + }, + { + "epoch": 0.4734560067681895, + "grad_norm": 8.0, + "learning_rate": 4.945347636066492e-06, + "loss": 2.7022, + "step": 4477 + }, + { + "epoch": 0.47356175972927245, + "grad_norm": 8.5625, + "learning_rate": 4.9453120172731255e-06, + "loss": 2.7303, + "step": 4478 + }, + { + "epoch": 0.47366751269035534, + "grad_norm": 8.25, + "learning_rate": 4.94527638700491e-06, + "loss": 2.6405, + "step": 4479 + }, + { + "epoch": 0.47377326565143824, + "grad_norm": 7.96875, + "learning_rate": 4.945240745262013e-06, + "loss": 2.5877, + "step": 4480 + }, + { + "epoch": 0.47387901861252113, + "grad_norm": 7.6875, + "learning_rate": 4.9452050920446005e-06, + "loss": 2.669, + "step": 4481 + }, + { + "epoch": 0.4739847715736041, + "grad_norm": 9.6875, + "learning_rate": 4.945169427352841e-06, + "loss": 2.2675, + "step": 4482 + }, + { + "epoch": 0.474090524534687, + "grad_norm": 7.59375, + "learning_rate": 4.945133751186902e-06, + "loss": 2.6948, + "step": 4483 + }, + { + "epoch": 0.47419627749576987, + "grad_norm": 7.75, + "learning_rate": 4.945098063546948e-06, + "loss": 2.5573, + "step": 4484 + }, + { + "epoch": 0.4743020304568528, + "grad_norm": 8.875, + "learning_rate": 4.945062364433151e-06, + "loss": 2.5763, + "step": 4485 + }, + { + "epoch": 0.4744077834179357, + "grad_norm": 7.65625, + "learning_rate": 4.945026653845676e-06, + "loss": 2.4245, + "step": 4486 + }, + { + "epoch": 0.4745135363790186, + "grad_norm": 7.625, + "learning_rate": 4.94499093178469e-06, + "loss": 2.4667, + "step": 4487 + }, + { + "epoch": 0.4746192893401015, + "grad_norm": 7.8125, + "learning_rate": 4.944955198250362e-06, + "loss": 2.5464, + "step": 4488 + }, + { + "epoch": 0.47472504230118445, + "grad_norm": 8.9375, + "learning_rate": 4.9449194532428605e-06, + "loss": 2.6475, + "step": 4489 + }, + { + "epoch": 0.47483079526226735, + "grad_norm": 7.59375, + "learning_rate": 4.94488369676235e-06, + "loss": 2.6385, + "step": 4490 + }, + { + "epoch": 0.47493654822335024, + "grad_norm": 7.6875, + "learning_rate": 4.944847928809001e-06, + "loss": 2.5584, + "step": 4491 + }, + { + "epoch": 0.4750423011844332, + "grad_norm": 7.375, + "learning_rate": 4.9448121493829804e-06, + "loss": 2.3056, + "step": 4492 + }, + { + "epoch": 0.4751480541455161, + "grad_norm": 7.28125, + "learning_rate": 4.944776358484457e-06, + "loss": 2.4569, + "step": 4493 + }, + { + "epoch": 0.475253807106599, + "grad_norm": 8.6875, + "learning_rate": 4.944740556113598e-06, + "loss": 2.7529, + "step": 4494 + }, + { + "epoch": 0.4753595600676819, + "grad_norm": 8.1875, + "learning_rate": 4.94470474227057e-06, + "loss": 2.5348, + "step": 4495 + }, + { + "epoch": 0.4754653130287648, + "grad_norm": 7.9375, + "learning_rate": 4.9446689169555435e-06, + "loss": 2.6086, + "step": 4496 + }, + { + "epoch": 0.4755710659898477, + "grad_norm": 7.5, + "learning_rate": 4.9446330801686855e-06, + "loss": 2.4908, + "step": 4497 + }, + { + "epoch": 0.4756768189509306, + "grad_norm": 7.96875, + "learning_rate": 4.944597231910164e-06, + "loss": 2.7081, + "step": 4498 + }, + { + "epoch": 0.47578257191201356, + "grad_norm": 7.84375, + "learning_rate": 4.944561372180148e-06, + "loss": 2.6146, + "step": 4499 + }, + { + "epoch": 0.47588832487309646, + "grad_norm": 8.0, + "learning_rate": 4.9445255009788055e-06, + "loss": 2.5123, + "step": 4500 + }, + { + "epoch": 0.47599407783417935, + "grad_norm": 8.0, + "learning_rate": 4.944489618306304e-06, + "loss": 2.6583, + "step": 4501 + }, + { + "epoch": 0.47609983079526225, + "grad_norm": 8.0, + "learning_rate": 4.9444537241628125e-06, + "loss": 2.779, + "step": 4502 + }, + { + "epoch": 0.4762055837563452, + "grad_norm": 7.5625, + "learning_rate": 4.9444178185485e-06, + "loss": 2.4073, + "step": 4503 + }, + { + "epoch": 0.4763113367174281, + "grad_norm": 8.1875, + "learning_rate": 4.9443819014635345e-06, + "loss": 2.494, + "step": 4504 + }, + { + "epoch": 0.476417089678511, + "grad_norm": 8.125, + "learning_rate": 4.944345972908083e-06, + "loss": 2.7546, + "step": 4505 + }, + { + "epoch": 0.47652284263959394, + "grad_norm": 8.0625, + "learning_rate": 4.944310032882317e-06, + "loss": 2.4805, + "step": 4506 + }, + { + "epoch": 0.47662859560067683, + "grad_norm": 7.75, + "learning_rate": 4.9442740813864034e-06, + "loss": 2.5967, + "step": 4507 + }, + { + "epoch": 0.4767343485617597, + "grad_norm": 8.125, + "learning_rate": 4.944238118420511e-06, + "loss": 2.3843, + "step": 4508 + }, + { + "epoch": 0.4768401015228426, + "grad_norm": 8.0, + "learning_rate": 4.944202143984809e-06, + "loss": 2.4893, + "step": 4509 + }, + { + "epoch": 0.47694585448392557, + "grad_norm": 7.875, + "learning_rate": 4.9441661580794655e-06, + "loss": 2.8383, + "step": 4510 + }, + { + "epoch": 0.47705160744500846, + "grad_norm": 8.125, + "learning_rate": 4.94413016070465e-06, + "loss": 2.6612, + "step": 4511 + }, + { + "epoch": 0.47715736040609136, + "grad_norm": 7.4375, + "learning_rate": 4.9440941518605316e-06, + "loss": 2.4953, + "step": 4512 + }, + { + "epoch": 0.4772631133671743, + "grad_norm": 7.75, + "learning_rate": 4.944058131547278e-06, + "loss": 2.6682, + "step": 4513 + }, + { + "epoch": 0.4773688663282572, + "grad_norm": 8.0, + "learning_rate": 4.94402209976506e-06, + "loss": 2.6851, + "step": 4514 + }, + { + "epoch": 0.4774746192893401, + "grad_norm": 7.84375, + "learning_rate": 4.943986056514046e-06, + "loss": 2.5894, + "step": 4515 + }, + { + "epoch": 0.477580372250423, + "grad_norm": 8.9375, + "learning_rate": 4.943950001794405e-06, + "loss": 2.6412, + "step": 4516 + }, + { + "epoch": 0.47768612521150594, + "grad_norm": 8.1875, + "learning_rate": 4.943913935606306e-06, + "loss": 2.71, + "step": 4517 + }, + { + "epoch": 0.47779187817258884, + "grad_norm": 8.25, + "learning_rate": 4.943877857949918e-06, + "loss": 2.4834, + "step": 4518 + }, + { + "epoch": 0.47789763113367173, + "grad_norm": 8.1875, + "learning_rate": 4.9438417688254115e-06, + "loss": 2.4994, + "step": 4519 + }, + { + "epoch": 0.4780033840947547, + "grad_norm": 8.125, + "learning_rate": 4.943805668232954e-06, + "loss": 2.8127, + "step": 4520 + }, + { + "epoch": 0.4781091370558376, + "grad_norm": 9.0625, + "learning_rate": 4.943769556172717e-06, + "loss": 2.2758, + "step": 4521 + }, + { + "epoch": 0.47821489001692047, + "grad_norm": 7.53125, + "learning_rate": 4.943733432644868e-06, + "loss": 2.5825, + "step": 4522 + }, + { + "epoch": 0.47832064297800336, + "grad_norm": 7.5625, + "learning_rate": 4.943697297649579e-06, + "loss": 2.6356, + "step": 4523 + }, + { + "epoch": 0.4784263959390863, + "grad_norm": 8.0, + "learning_rate": 4.943661151187017e-06, + "loss": 2.5294, + "step": 4524 + }, + { + "epoch": 0.4785321489001692, + "grad_norm": 8.75, + "learning_rate": 4.943624993257352e-06, + "loss": 2.7217, + "step": 4525 + }, + { + "epoch": 0.4786379018612521, + "grad_norm": 7.5, + "learning_rate": 4.943588823860755e-06, + "loss": 2.5621, + "step": 4526 + }, + { + "epoch": 0.47874365482233505, + "grad_norm": 7.6875, + "learning_rate": 4.943552642997396e-06, + "loss": 2.5525, + "step": 4527 + }, + { + "epoch": 0.47884940778341795, + "grad_norm": 7.15625, + "learning_rate": 4.943516450667443e-06, + "loss": 2.4353, + "step": 4528 + }, + { + "epoch": 0.47895516074450084, + "grad_norm": 9.125, + "learning_rate": 4.9434802468710666e-06, + "loss": 2.5604, + "step": 4529 + }, + { + "epoch": 0.47906091370558374, + "grad_norm": 8.375, + "learning_rate": 4.943444031608436e-06, + "loss": 2.7091, + "step": 4530 + }, + { + "epoch": 0.4791666666666667, + "grad_norm": 8.25, + "learning_rate": 4.943407804879723e-06, + "loss": 2.5855, + "step": 4531 + }, + { + "epoch": 0.4792724196277496, + "grad_norm": 8.125, + "learning_rate": 4.943371566685096e-06, + "loss": 2.5941, + "step": 4532 + }, + { + "epoch": 0.4793781725888325, + "grad_norm": 8.0, + "learning_rate": 4.943335317024726e-06, + "loss": 2.6365, + "step": 4533 + }, + { + "epoch": 0.4794839255499154, + "grad_norm": 7.78125, + "learning_rate": 4.943299055898782e-06, + "loss": 2.8061, + "step": 4534 + }, + { + "epoch": 0.4795896785109983, + "grad_norm": 7.59375, + "learning_rate": 4.943262783307434e-06, + "loss": 2.6171, + "step": 4535 + }, + { + "epoch": 0.4796954314720812, + "grad_norm": 7.53125, + "learning_rate": 4.943226499250855e-06, + "loss": 2.6282, + "step": 4536 + }, + { + "epoch": 0.4798011844331641, + "grad_norm": 8.75, + "learning_rate": 4.943190203729212e-06, + "loss": 2.7357, + "step": 4537 + }, + { + "epoch": 0.47990693739424706, + "grad_norm": 8.0, + "learning_rate": 4.943153896742678e-06, + "loss": 2.4791, + "step": 4538 + }, + { + "epoch": 0.48001269035532995, + "grad_norm": 7.9375, + "learning_rate": 4.94311757829142e-06, + "loss": 2.7197, + "step": 4539 + }, + { + "epoch": 0.48011844331641285, + "grad_norm": 7.5625, + "learning_rate": 4.9430812483756116e-06, + "loss": 2.5989, + "step": 4540 + }, + { + "epoch": 0.4802241962774958, + "grad_norm": 7.625, + "learning_rate": 4.943044906995422e-06, + "loss": 2.5928, + "step": 4541 + }, + { + "epoch": 0.4803299492385787, + "grad_norm": 8.5625, + "learning_rate": 4.943008554151021e-06, + "loss": 2.7083, + "step": 4542 + }, + { + "epoch": 0.4804357021996616, + "grad_norm": 8.0625, + "learning_rate": 4.942972189842581e-06, + "loss": 2.6967, + "step": 4543 + }, + { + "epoch": 0.4805414551607445, + "grad_norm": 8.375, + "learning_rate": 4.942935814070272e-06, + "loss": 2.7655, + "step": 4544 + }, + { + "epoch": 0.48064720812182743, + "grad_norm": 7.59375, + "learning_rate": 4.942899426834263e-06, + "loss": 2.7343, + "step": 4545 + }, + { + "epoch": 0.4807529610829103, + "grad_norm": 7.59375, + "learning_rate": 4.942863028134726e-06, + "loss": 2.6177, + "step": 4546 + }, + { + "epoch": 0.4808587140439932, + "grad_norm": 8.3125, + "learning_rate": 4.942826617971834e-06, + "loss": 2.6203, + "step": 4547 + }, + { + "epoch": 0.48096446700507617, + "grad_norm": 7.625, + "learning_rate": 4.9427901963457545e-06, + "loss": 2.4309, + "step": 4548 + }, + { + "epoch": 0.48107021996615906, + "grad_norm": 7.4375, + "learning_rate": 4.942753763256659e-06, + "loss": 2.557, + "step": 4549 + }, + { + "epoch": 0.48117597292724196, + "grad_norm": 7.53125, + "learning_rate": 4.94271731870472e-06, + "loss": 2.668, + "step": 4550 + }, + { + "epoch": 0.48128172588832485, + "grad_norm": 7.875, + "learning_rate": 4.942680862690108e-06, + "loss": 2.7717, + "step": 4551 + }, + { + "epoch": 0.4813874788494078, + "grad_norm": 7.8125, + "learning_rate": 4.9426443952129925e-06, + "loss": 2.7169, + "step": 4552 + }, + { + "epoch": 0.4814932318104907, + "grad_norm": 7.40625, + "learning_rate": 4.942607916273547e-06, + "loss": 2.5009, + "step": 4553 + }, + { + "epoch": 0.4815989847715736, + "grad_norm": 7.4375, + "learning_rate": 4.942571425871941e-06, + "loss": 2.541, + "step": 4554 + }, + { + "epoch": 0.48170473773265654, + "grad_norm": 8.875, + "learning_rate": 4.942534924008346e-06, + "loss": 2.7909, + "step": 4555 + }, + { + "epoch": 0.48181049069373943, + "grad_norm": 7.5625, + "learning_rate": 4.942498410682934e-06, + "loss": 2.4291, + "step": 4556 + }, + { + "epoch": 0.48191624365482233, + "grad_norm": 9.5625, + "learning_rate": 4.9424618858958765e-06, + "loss": 2.6876, + "step": 4557 + }, + { + "epoch": 0.4820219966159052, + "grad_norm": 8.1875, + "learning_rate": 4.942425349647343e-06, + "loss": 2.848, + "step": 4558 + }, + { + "epoch": 0.4821277495769882, + "grad_norm": 7.53125, + "learning_rate": 4.942388801937507e-06, + "loss": 2.2984, + "step": 4559 + }, + { + "epoch": 0.48223350253807107, + "grad_norm": 7.90625, + "learning_rate": 4.942352242766539e-06, + "loss": 2.5833, + "step": 4560 + }, + { + "epoch": 0.48233925549915396, + "grad_norm": 8.0, + "learning_rate": 4.942315672134611e-06, + "loss": 2.5149, + "step": 4561 + }, + { + "epoch": 0.4824450084602369, + "grad_norm": 8.0625, + "learning_rate": 4.9422790900418945e-06, + "loss": 2.6641, + "step": 4562 + }, + { + "epoch": 0.4825507614213198, + "grad_norm": 7.6875, + "learning_rate": 4.942242496488561e-06, + "loss": 2.4712, + "step": 4563 + }, + { + "epoch": 0.4826565143824027, + "grad_norm": 7.625, + "learning_rate": 4.9422058914747825e-06, + "loss": 2.6281, + "step": 4564 + }, + { + "epoch": 0.4827622673434856, + "grad_norm": 7.71875, + "learning_rate": 4.94216927500073e-06, + "loss": 2.5735, + "step": 4565 + }, + { + "epoch": 0.48286802030456855, + "grad_norm": 8.375, + "learning_rate": 4.942132647066577e-06, + "loss": 2.6241, + "step": 4566 + }, + { + "epoch": 0.48297377326565144, + "grad_norm": 7.8125, + "learning_rate": 4.942096007672493e-06, + "loss": 2.6924, + "step": 4567 + }, + { + "epoch": 0.48307952622673433, + "grad_norm": 8.5, + "learning_rate": 4.9420593568186514e-06, + "loss": 2.496, + "step": 4568 + }, + { + "epoch": 0.4831852791878173, + "grad_norm": 7.5, + "learning_rate": 4.9420226945052245e-06, + "loss": 2.7503, + "step": 4569 + }, + { + "epoch": 0.4832910321489002, + "grad_norm": 7.59375, + "learning_rate": 4.941986020732384e-06, + "loss": 2.6842, + "step": 4570 + }, + { + "epoch": 0.4833967851099831, + "grad_norm": 7.65625, + "learning_rate": 4.941949335500301e-06, + "loss": 2.473, + "step": 4571 + }, + { + "epoch": 0.48350253807106597, + "grad_norm": 7.65625, + "learning_rate": 4.94191263880915e-06, + "loss": 2.5135, + "step": 4572 + }, + { + "epoch": 0.4836082910321489, + "grad_norm": 7.78125, + "learning_rate": 4.9418759306591e-06, + "loss": 2.5773, + "step": 4573 + }, + { + "epoch": 0.4837140439932318, + "grad_norm": 7.3125, + "learning_rate": 4.9418392110503255e-06, + "loss": 2.4675, + "step": 4574 + }, + { + "epoch": 0.4838197969543147, + "grad_norm": 8.125, + "learning_rate": 4.941802479982999e-06, + "loss": 2.6782, + "step": 4575 + }, + { + "epoch": 0.48392554991539766, + "grad_norm": 7.96875, + "learning_rate": 4.9417657374572915e-06, + "loss": 2.3753, + "step": 4576 + }, + { + "epoch": 0.48403130287648055, + "grad_norm": 7.78125, + "learning_rate": 4.9417289834733765e-06, + "loss": 2.7548, + "step": 4577 + }, + { + "epoch": 0.48413705583756345, + "grad_norm": 7.875, + "learning_rate": 4.941692218031425e-06, + "loss": 2.6697, + "step": 4578 + }, + { + "epoch": 0.48424280879864634, + "grad_norm": 8.0625, + "learning_rate": 4.941655441131611e-06, + "loss": 2.2264, + "step": 4579 + }, + { + "epoch": 0.4843485617597293, + "grad_norm": 7.5625, + "learning_rate": 4.941618652774107e-06, + "loss": 2.6949, + "step": 4580 + }, + { + "epoch": 0.4844543147208122, + "grad_norm": 7.6875, + "learning_rate": 4.941581852959085e-06, + "loss": 2.6491, + "step": 4581 + }, + { + "epoch": 0.4845600676818951, + "grad_norm": 8.0, + "learning_rate": 4.941545041686719e-06, + "loss": 2.6914, + "step": 4582 + }, + { + "epoch": 0.48466582064297803, + "grad_norm": 8.0625, + "learning_rate": 4.9415082189571786e-06, + "loss": 2.6921, + "step": 4583 + }, + { + "epoch": 0.4847715736040609, + "grad_norm": 8.0, + "learning_rate": 4.9414713847706395e-06, + "loss": 2.5915, + "step": 4584 + }, + { + "epoch": 0.4848773265651438, + "grad_norm": 7.875, + "learning_rate": 4.941434539127274e-06, + "loss": 2.5693, + "step": 4585 + }, + { + "epoch": 0.4849830795262267, + "grad_norm": 8.1875, + "learning_rate": 4.941397682027255e-06, + "loss": 2.6621, + "step": 4586 + }, + { + "epoch": 0.48508883248730966, + "grad_norm": 7.90625, + "learning_rate": 4.9413608134707545e-06, + "loss": 2.2496, + "step": 4587 + }, + { + "epoch": 0.48519458544839256, + "grad_norm": 7.84375, + "learning_rate": 4.941323933457947e-06, + "loss": 2.5691, + "step": 4588 + }, + { + "epoch": 0.48530033840947545, + "grad_norm": 8.0, + "learning_rate": 4.9412870419890045e-06, + "loss": 2.6747, + "step": 4589 + }, + { + "epoch": 0.4854060913705584, + "grad_norm": 8.0625, + "learning_rate": 4.9412501390641e-06, + "loss": 2.4196, + "step": 4590 + }, + { + "epoch": 0.4855118443316413, + "grad_norm": 7.4375, + "learning_rate": 4.941213224683407e-06, + "loss": 2.6995, + "step": 4591 + }, + { + "epoch": 0.4856175972927242, + "grad_norm": 7.875, + "learning_rate": 4.9411762988470994e-06, + "loss": 2.7614, + "step": 4592 + }, + { + "epoch": 0.4857233502538071, + "grad_norm": 7.40625, + "learning_rate": 4.9411393615553495e-06, + "loss": 2.6602, + "step": 4593 + }, + { + "epoch": 0.48582910321489003, + "grad_norm": 7.875, + "learning_rate": 4.941102412808331e-06, + "loss": 2.4587, + "step": 4594 + }, + { + "epoch": 0.48593485617597293, + "grad_norm": 8.25, + "learning_rate": 4.941065452606216e-06, + "loss": 2.6874, + "step": 4595 + }, + { + "epoch": 0.4860406091370558, + "grad_norm": 7.5, + "learning_rate": 4.941028480949181e-06, + "loss": 2.5921, + "step": 4596 + }, + { + "epoch": 0.4861463620981388, + "grad_norm": 8.0, + "learning_rate": 4.940991497837396e-06, + "loss": 2.6153, + "step": 4597 + }, + { + "epoch": 0.48625211505922167, + "grad_norm": 7.59375, + "learning_rate": 4.940954503271038e-06, + "loss": 2.5632, + "step": 4598 + }, + { + "epoch": 0.48635786802030456, + "grad_norm": 7.78125, + "learning_rate": 4.940917497250279e-06, + "loss": 2.5622, + "step": 4599 + }, + { + "epoch": 0.48646362098138746, + "grad_norm": 7.9375, + "learning_rate": 4.9408804797752905e-06, + "loss": 2.6199, + "step": 4600 + }, + { + "epoch": 0.4865693739424704, + "grad_norm": 8.1875, + "learning_rate": 4.940843450846249e-06, + "loss": 2.561, + "step": 4601 + }, + { + "epoch": 0.4866751269035533, + "grad_norm": 8.1875, + "learning_rate": 4.940806410463328e-06, + "loss": 2.6118, + "step": 4602 + }, + { + "epoch": 0.4867808798646362, + "grad_norm": 8.1875, + "learning_rate": 4.9407693586267e-06, + "loss": 2.8864, + "step": 4603 + }, + { + "epoch": 0.48688663282571915, + "grad_norm": 8.9375, + "learning_rate": 4.94073229533654e-06, + "loss": 2.7386, + "step": 4604 + }, + { + "epoch": 0.48699238578680204, + "grad_norm": 7.84375, + "learning_rate": 4.940695220593022e-06, + "loss": 2.7166, + "step": 4605 + }, + { + "epoch": 0.48709813874788493, + "grad_norm": 7.8125, + "learning_rate": 4.940658134396319e-06, + "loss": 2.3742, + "step": 4606 + }, + { + "epoch": 0.48720389170896783, + "grad_norm": 8.25, + "learning_rate": 4.940621036746606e-06, + "loss": 2.5947, + "step": 4607 + }, + { + "epoch": 0.4873096446700508, + "grad_norm": 8.5, + "learning_rate": 4.940583927644056e-06, + "loss": 2.4567, + "step": 4608 + }, + { + "epoch": 0.4874153976311337, + "grad_norm": 7.84375, + "learning_rate": 4.940546807088845e-06, + "loss": 2.5475, + "step": 4609 + }, + { + "epoch": 0.48752115059221657, + "grad_norm": 7.5625, + "learning_rate": 4.940509675081145e-06, + "loss": 2.6472, + "step": 4610 + }, + { + "epoch": 0.4876269035532995, + "grad_norm": 8.0625, + "learning_rate": 4.940472531621131e-06, + "loss": 2.6686, + "step": 4611 + }, + { + "epoch": 0.4877326565143824, + "grad_norm": 8.0, + "learning_rate": 4.940435376708978e-06, + "loss": 2.4844, + "step": 4612 + }, + { + "epoch": 0.4878384094754653, + "grad_norm": 7.4375, + "learning_rate": 4.94039821034486e-06, + "loss": 2.6113, + "step": 4613 + }, + { + "epoch": 0.4879441624365482, + "grad_norm": 8.3125, + "learning_rate": 4.940361032528951e-06, + "loss": 2.693, + "step": 4614 + }, + { + "epoch": 0.48804991539763115, + "grad_norm": 8.0625, + "learning_rate": 4.940323843261426e-06, + "loss": 2.3997, + "step": 4615 + }, + { + "epoch": 0.48815566835871405, + "grad_norm": 7.9375, + "learning_rate": 4.940286642542459e-06, + "loss": 2.5477, + "step": 4616 + }, + { + "epoch": 0.48826142131979694, + "grad_norm": 7.9375, + "learning_rate": 4.940249430372225e-06, + "loss": 2.678, + "step": 4617 + }, + { + "epoch": 0.4883671742808799, + "grad_norm": 7.40625, + "learning_rate": 4.940212206750898e-06, + "loss": 2.4974, + "step": 4618 + }, + { + "epoch": 0.4884729272419628, + "grad_norm": 7.78125, + "learning_rate": 4.940174971678654e-06, + "loss": 2.6972, + "step": 4619 + }, + { + "epoch": 0.4885786802030457, + "grad_norm": 7.875, + "learning_rate": 4.940137725155666e-06, + "loss": 2.64, + "step": 4620 + }, + { + "epoch": 0.4886844331641286, + "grad_norm": 8.25, + "learning_rate": 4.940100467182111e-06, + "loss": 2.4491, + "step": 4621 + }, + { + "epoch": 0.4887901861252115, + "grad_norm": 7.875, + "learning_rate": 4.940063197758161e-06, + "loss": 2.8068, + "step": 4622 + }, + { + "epoch": 0.4888959390862944, + "grad_norm": 8.3125, + "learning_rate": 4.940025916883993e-06, + "loss": 2.6347, + "step": 4623 + }, + { + "epoch": 0.4890016920473773, + "grad_norm": 7.9375, + "learning_rate": 4.939988624559781e-06, + "loss": 2.7187, + "step": 4624 + }, + { + "epoch": 0.48910744500846026, + "grad_norm": 8.25, + "learning_rate": 4.9399513207857e-06, + "loss": 2.8623, + "step": 4625 + }, + { + "epoch": 0.48921319796954316, + "grad_norm": 8.5625, + "learning_rate": 4.9399140055619265e-06, + "loss": 2.5768, + "step": 4626 + }, + { + "epoch": 0.48931895093062605, + "grad_norm": 9.25, + "learning_rate": 4.9398766788886335e-06, + "loss": 2.5931, + "step": 4627 + }, + { + "epoch": 0.48942470389170895, + "grad_norm": 8.625, + "learning_rate": 4.939839340765997e-06, + "loss": 2.4702, + "step": 4628 + }, + { + "epoch": 0.4895304568527919, + "grad_norm": 7.625, + "learning_rate": 4.939801991194193e-06, + "loss": 2.6116, + "step": 4629 + }, + { + "epoch": 0.4896362098138748, + "grad_norm": 7.875, + "learning_rate": 4.939764630173396e-06, + "loss": 2.5808, + "step": 4630 + }, + { + "epoch": 0.4897419627749577, + "grad_norm": 8.25, + "learning_rate": 4.9397272577037815e-06, + "loss": 2.6105, + "step": 4631 + }, + { + "epoch": 0.48984771573604063, + "grad_norm": 7.625, + "learning_rate": 4.939689873785525e-06, + "loss": 2.477, + "step": 4632 + }, + { + "epoch": 0.48995346869712353, + "grad_norm": 8.1875, + "learning_rate": 4.9396524784188e-06, + "loss": 2.6726, + "step": 4633 + }, + { + "epoch": 0.4900592216582064, + "grad_norm": 7.59375, + "learning_rate": 4.939615071603786e-06, + "loss": 2.7051, + "step": 4634 + }, + { + "epoch": 0.4901649746192893, + "grad_norm": 7.5, + "learning_rate": 4.939577653340655e-06, + "loss": 2.7316, + "step": 4635 + }, + { + "epoch": 0.49027072758037227, + "grad_norm": 8.0, + "learning_rate": 4.939540223629584e-06, + "loss": 2.7921, + "step": 4636 + }, + { + "epoch": 0.49037648054145516, + "grad_norm": 7.65625, + "learning_rate": 4.939502782470749e-06, + "loss": 2.5724, + "step": 4637 + }, + { + "epoch": 0.49048223350253806, + "grad_norm": 8.0625, + "learning_rate": 4.939465329864324e-06, + "loss": 2.3949, + "step": 4638 + }, + { + "epoch": 0.490587986463621, + "grad_norm": 8.3125, + "learning_rate": 4.939427865810486e-06, + "loss": 2.508, + "step": 4639 + }, + { + "epoch": 0.4906937394247039, + "grad_norm": 7.46875, + "learning_rate": 4.939390390309412e-06, + "loss": 2.6696, + "step": 4640 + }, + { + "epoch": 0.4907994923857868, + "grad_norm": 7.5, + "learning_rate": 4.939352903361276e-06, + "loss": 2.5819, + "step": 4641 + }, + { + "epoch": 0.4909052453468697, + "grad_norm": 8.3125, + "learning_rate": 4.939315404966254e-06, + "loss": 2.7079, + "step": 4642 + }, + { + "epoch": 0.49101099830795264, + "grad_norm": 7.40625, + "learning_rate": 4.939277895124523e-06, + "loss": 2.5288, + "step": 4643 + }, + { + "epoch": 0.49111675126903553, + "grad_norm": 7.34375, + "learning_rate": 4.9392403738362586e-06, + "loss": 2.601, + "step": 4644 + }, + { + "epoch": 0.49122250423011843, + "grad_norm": 8.1875, + "learning_rate": 4.939202841101636e-06, + "loss": 2.6346, + "step": 4645 + }, + { + "epoch": 0.4913282571912014, + "grad_norm": 8.0, + "learning_rate": 4.939165296920833e-06, + "loss": 2.2166, + "step": 4646 + }, + { + "epoch": 0.4914340101522843, + "grad_norm": 8.0, + "learning_rate": 4.939127741294024e-06, + "loss": 2.4999, + "step": 4647 + }, + { + "epoch": 0.49153976311336717, + "grad_norm": 7.65625, + "learning_rate": 4.939090174221387e-06, + "loss": 2.566, + "step": 4648 + }, + { + "epoch": 0.49164551607445006, + "grad_norm": 7.46875, + "learning_rate": 4.939052595703097e-06, + "loss": 2.6961, + "step": 4649 + }, + { + "epoch": 0.491751269035533, + "grad_norm": 7.6875, + "learning_rate": 4.93901500573933e-06, + "loss": 2.5721, + "step": 4650 + }, + { + "epoch": 0.4918570219966159, + "grad_norm": 7.84375, + "learning_rate": 4.9389774043302635e-06, + "loss": 2.5374, + "step": 4651 + }, + { + "epoch": 0.4919627749576988, + "grad_norm": 8.375, + "learning_rate": 4.938939791476074e-06, + "loss": 2.5536, + "step": 4652 + }, + { + "epoch": 0.49206852791878175, + "grad_norm": 7.5, + "learning_rate": 4.938902167176937e-06, + "loss": 2.5561, + "step": 4653 + }, + { + "epoch": 0.49217428087986465, + "grad_norm": 7.40625, + "learning_rate": 4.93886453143303e-06, + "loss": 2.6039, + "step": 4654 + }, + { + "epoch": 0.49228003384094754, + "grad_norm": 7.59375, + "learning_rate": 4.938826884244529e-06, + "loss": 2.4623, + "step": 4655 + }, + { + "epoch": 0.49238578680203043, + "grad_norm": 7.65625, + "learning_rate": 4.938789225611611e-06, + "loss": 2.5893, + "step": 4656 + }, + { + "epoch": 0.4924915397631134, + "grad_norm": 8.4375, + "learning_rate": 4.938751555534452e-06, + "loss": 2.939, + "step": 4657 + }, + { + "epoch": 0.4925972927241963, + "grad_norm": 8.0, + "learning_rate": 4.93871387401323e-06, + "loss": 2.8049, + "step": 4658 + }, + { + "epoch": 0.4927030456852792, + "grad_norm": 8.25, + "learning_rate": 4.938676181048121e-06, + "loss": 2.5704, + "step": 4659 + }, + { + "epoch": 0.4928087986463621, + "grad_norm": 7.78125, + "learning_rate": 4.938638476639302e-06, + "loss": 2.5412, + "step": 4660 + }, + { + "epoch": 0.492914551607445, + "grad_norm": 7.90625, + "learning_rate": 4.93860076078695e-06, + "loss": 2.4359, + "step": 4661 + }, + { + "epoch": 0.4930203045685279, + "grad_norm": 8.125, + "learning_rate": 4.938563033491242e-06, + "loss": 2.7197, + "step": 4662 + }, + { + "epoch": 0.4931260575296108, + "grad_norm": 7.75, + "learning_rate": 4.938525294752356e-06, + "loss": 2.6811, + "step": 4663 + }, + { + "epoch": 0.49323181049069376, + "grad_norm": 7.4375, + "learning_rate": 4.938487544570467e-06, + "loss": 2.6247, + "step": 4664 + }, + { + "epoch": 0.49333756345177665, + "grad_norm": 8.125, + "learning_rate": 4.938449782945753e-06, + "loss": 2.6045, + "step": 4665 + }, + { + "epoch": 0.49344331641285955, + "grad_norm": 7.9375, + "learning_rate": 4.9384120098783925e-06, + "loss": 2.6659, + "step": 4666 + }, + { + "epoch": 0.4935490693739425, + "grad_norm": 8.0625, + "learning_rate": 4.938374225368561e-06, + "loss": 2.4905, + "step": 4667 + }, + { + "epoch": 0.4936548223350254, + "grad_norm": 7.65625, + "learning_rate": 4.938336429416437e-06, + "loss": 2.5509, + "step": 4668 + }, + { + "epoch": 0.4937605752961083, + "grad_norm": 7.84375, + "learning_rate": 4.938298622022197e-06, + "loss": 2.5513, + "step": 4669 + }, + { + "epoch": 0.4938663282571912, + "grad_norm": 7.9375, + "learning_rate": 4.9382608031860195e-06, + "loss": 2.6743, + "step": 4670 + }, + { + "epoch": 0.49397208121827413, + "grad_norm": 8.1875, + "learning_rate": 4.93822297290808e-06, + "loss": 2.4878, + "step": 4671 + }, + { + "epoch": 0.494077834179357, + "grad_norm": 7.84375, + "learning_rate": 4.938185131188558e-06, + "loss": 2.5591, + "step": 4672 + }, + { + "epoch": 0.4941835871404399, + "grad_norm": 8.6875, + "learning_rate": 4.93814727802763e-06, + "loss": 2.8898, + "step": 4673 + }, + { + "epoch": 0.49428934010152287, + "grad_norm": 8.25, + "learning_rate": 4.938109413425474e-06, + "loss": 2.4062, + "step": 4674 + }, + { + "epoch": 0.49439509306260576, + "grad_norm": 7.5, + "learning_rate": 4.938071537382268e-06, + "loss": 2.5454, + "step": 4675 + }, + { + "epoch": 0.49450084602368866, + "grad_norm": 8.125, + "learning_rate": 4.93803364989819e-06, + "loss": 2.74, + "step": 4676 + }, + { + "epoch": 0.49460659898477155, + "grad_norm": 7.96875, + "learning_rate": 4.937995750973417e-06, + "loss": 2.6792, + "step": 4677 + }, + { + "epoch": 0.4947123519458545, + "grad_norm": 7.8125, + "learning_rate": 4.937957840608126e-06, + "loss": 2.6566, + "step": 4678 + }, + { + "epoch": 0.4948181049069374, + "grad_norm": 8.6875, + "learning_rate": 4.937919918802496e-06, + "loss": 2.6794, + "step": 4679 + }, + { + "epoch": 0.4949238578680203, + "grad_norm": 7.78125, + "learning_rate": 4.9378819855567064e-06, + "loss": 2.4559, + "step": 4680 + }, + { + "epoch": 0.49502961082910324, + "grad_norm": 8.375, + "learning_rate": 4.937844040870932e-06, + "loss": 2.8108, + "step": 4681 + }, + { + "epoch": 0.49513536379018613, + "grad_norm": 7.90625, + "learning_rate": 4.9378060847453534e-06, + "loss": 2.5434, + "step": 4682 + }, + { + "epoch": 0.49524111675126903, + "grad_norm": 7.65625, + "learning_rate": 4.9377681171801485e-06, + "loss": 2.6941, + "step": 4683 + }, + { + "epoch": 0.4953468697123519, + "grad_norm": 7.5, + "learning_rate": 4.937730138175494e-06, + "loss": 2.6379, + "step": 4684 + }, + { + "epoch": 0.4954526226734349, + "grad_norm": 8.0625, + "learning_rate": 4.9376921477315686e-06, + "loss": 2.8126, + "step": 4685 + }, + { + "epoch": 0.49555837563451777, + "grad_norm": 7.9375, + "learning_rate": 4.937654145848552e-06, + "loss": 2.7497, + "step": 4686 + }, + { + "epoch": 0.49566412859560066, + "grad_norm": 7.65625, + "learning_rate": 4.93761613252662e-06, + "loss": 2.6263, + "step": 4687 + }, + { + "epoch": 0.4957698815566836, + "grad_norm": 7.9375, + "learning_rate": 4.9375781077659535e-06, + "loss": 2.7081, + "step": 4688 + }, + { + "epoch": 0.4958756345177665, + "grad_norm": 8.25, + "learning_rate": 4.93754007156673e-06, + "loss": 2.3826, + "step": 4689 + }, + { + "epoch": 0.4959813874788494, + "grad_norm": 8.25, + "learning_rate": 4.937502023929127e-06, + "loss": 2.5203, + "step": 4690 + }, + { + "epoch": 0.4960871404399323, + "grad_norm": 7.1875, + "learning_rate": 4.9374639648533254e-06, + "loss": 2.472, + "step": 4691 + }, + { + "epoch": 0.49619289340101524, + "grad_norm": 8.0625, + "learning_rate": 4.937425894339501e-06, + "loss": 2.5641, + "step": 4692 + }, + { + "epoch": 0.49629864636209814, + "grad_norm": 8.4375, + "learning_rate": 4.937387812387834e-06, + "loss": 2.6142, + "step": 4693 + }, + { + "epoch": 0.49640439932318103, + "grad_norm": 8.4375, + "learning_rate": 4.937349718998503e-06, + "loss": 2.7028, + "step": 4694 + }, + { + "epoch": 0.496510152284264, + "grad_norm": 8.1875, + "learning_rate": 4.937311614171687e-06, + "loss": 2.5497, + "step": 4695 + }, + { + "epoch": 0.4966159052453469, + "grad_norm": 8.0, + "learning_rate": 4.937273497907564e-06, + "loss": 2.4353, + "step": 4696 + }, + { + "epoch": 0.4967216582064298, + "grad_norm": 8.0625, + "learning_rate": 4.937235370206313e-06, + "loss": 2.5737, + "step": 4697 + }, + { + "epoch": 0.49682741116751267, + "grad_norm": 7.6875, + "learning_rate": 4.937197231068114e-06, + "loss": 2.5166, + "step": 4698 + }, + { + "epoch": 0.4969331641285956, + "grad_norm": 9.875, + "learning_rate": 4.937159080493145e-06, + "loss": 2.8086, + "step": 4699 + }, + { + "epoch": 0.4970389170896785, + "grad_norm": 7.90625, + "learning_rate": 4.9371209184815835e-06, + "loss": 2.5284, + "step": 4700 + }, + { + "epoch": 0.4971446700507614, + "grad_norm": 8.5, + "learning_rate": 4.937082745033612e-06, + "loss": 2.6194, + "step": 4701 + }, + { + "epoch": 0.49725042301184436, + "grad_norm": 7.8125, + "learning_rate": 4.9370445601494075e-06, + "loss": 2.6844, + "step": 4702 + }, + { + "epoch": 0.49735617597292725, + "grad_norm": 9.75, + "learning_rate": 4.93700636382915e-06, + "loss": 2.5134, + "step": 4703 + }, + { + "epoch": 0.49746192893401014, + "grad_norm": 7.71875, + "learning_rate": 4.936968156073017e-06, + "loss": 2.4047, + "step": 4704 + }, + { + "epoch": 0.49756768189509304, + "grad_norm": 7.375, + "learning_rate": 4.9369299368811915e-06, + "loss": 2.4554, + "step": 4705 + }, + { + "epoch": 0.497673434856176, + "grad_norm": 8.125, + "learning_rate": 4.936891706253849e-06, + "loss": 2.5091, + "step": 4706 + }, + { + "epoch": 0.4977791878172589, + "grad_norm": 7.96875, + "learning_rate": 4.93685346419117e-06, + "loss": 2.4354, + "step": 4707 + }, + { + "epoch": 0.4978849407783418, + "grad_norm": 7.875, + "learning_rate": 4.936815210693335e-06, + "loss": 2.5964, + "step": 4708 + }, + { + "epoch": 0.49799069373942473, + "grad_norm": 7.84375, + "learning_rate": 4.936776945760523e-06, + "loss": 2.6107, + "step": 4709 + }, + { + "epoch": 0.4980964467005076, + "grad_norm": 7.84375, + "learning_rate": 4.936738669392913e-06, + "loss": 2.5408, + "step": 4710 + }, + { + "epoch": 0.4982021996615905, + "grad_norm": 7.75, + "learning_rate": 4.9367003815906845e-06, + "loss": 2.4983, + "step": 4711 + }, + { + "epoch": 0.4983079526226734, + "grad_norm": 8.3125, + "learning_rate": 4.936662082354018e-06, + "loss": 2.498, + "step": 4712 + }, + { + "epoch": 0.49841370558375636, + "grad_norm": 8.4375, + "learning_rate": 4.936623771683093e-06, + "loss": 2.6295, + "step": 4713 + }, + { + "epoch": 0.49851945854483926, + "grad_norm": 7.3125, + "learning_rate": 4.93658544957809e-06, + "loss": 2.7066, + "step": 4714 + }, + { + "epoch": 0.49862521150592215, + "grad_norm": 8.5625, + "learning_rate": 4.9365471160391875e-06, + "loss": 2.6168, + "step": 4715 + }, + { + "epoch": 0.4987309644670051, + "grad_norm": 8.625, + "learning_rate": 4.936508771066566e-06, + "loss": 2.6592, + "step": 4716 + }, + { + "epoch": 0.498836717428088, + "grad_norm": 7.75, + "learning_rate": 4.936470414660405e-06, + "loss": 2.2299, + "step": 4717 + }, + { + "epoch": 0.4989424703891709, + "grad_norm": 8.625, + "learning_rate": 4.9364320468208845e-06, + "loss": 2.6991, + "step": 4718 + }, + { + "epoch": 0.4990482233502538, + "grad_norm": 8.625, + "learning_rate": 4.936393667548185e-06, + "loss": 2.6343, + "step": 4719 + }, + { + "epoch": 0.49915397631133673, + "grad_norm": 7.53125, + "learning_rate": 4.936355276842487e-06, + "loss": 2.7639, + "step": 4720 + }, + { + "epoch": 0.49925972927241963, + "grad_norm": 8.0625, + "learning_rate": 4.9363168747039704e-06, + "loss": 2.7172, + "step": 4721 + }, + { + "epoch": 0.4993654822335025, + "grad_norm": 7.75, + "learning_rate": 4.936278461132814e-06, + "loss": 2.6603, + "step": 4722 + }, + { + "epoch": 0.4994712351945855, + "grad_norm": 7.84375, + "learning_rate": 4.9362400361292e-06, + "loss": 2.7521, + "step": 4723 + }, + { + "epoch": 0.49957698815566837, + "grad_norm": 8.125, + "learning_rate": 4.936201599693307e-06, + "loss": 2.5781, + "step": 4724 + }, + { + "epoch": 0.49968274111675126, + "grad_norm": 7.84375, + "learning_rate": 4.936163151825318e-06, + "loss": 2.6049, + "step": 4725 + }, + { + "epoch": 0.49978849407783416, + "grad_norm": 7.78125, + "learning_rate": 4.9361246925254106e-06, + "loss": 2.5661, + "step": 4726 + }, + { + "epoch": 0.4998942470389171, + "grad_norm": 11.1875, + "learning_rate": 4.936086221793766e-06, + "loss": 2.6176, + "step": 4727 + }, + { + "epoch": 0.5, + "grad_norm": 8.5, + "learning_rate": 4.936047739630566e-06, + "loss": 2.4125, + "step": 4728 + }, + { + "epoch": 0.500105752961083, + "grad_norm": 7.75, + "learning_rate": 4.93600924603599e-06, + "loss": 2.5055, + "step": 4729 + }, + { + "epoch": 0.5002115059221658, + "grad_norm": 8.0, + "learning_rate": 4.935970741010218e-06, + "loss": 2.5361, + "step": 4730 + }, + { + "epoch": 0.5003172588832487, + "grad_norm": 7.8125, + "learning_rate": 4.935932224553432e-06, + "loss": 2.5251, + "step": 4731 + }, + { + "epoch": 0.5004230118443317, + "grad_norm": 7.4375, + "learning_rate": 4.935893696665813e-06, + "loss": 2.5342, + "step": 4732 + }, + { + "epoch": 0.5005287648054145, + "grad_norm": 7.4375, + "learning_rate": 4.935855157347541e-06, + "loss": 2.509, + "step": 4733 + }, + { + "epoch": 0.5006345177664975, + "grad_norm": 7.53125, + "learning_rate": 4.935816606598797e-06, + "loss": 2.5818, + "step": 4734 + }, + { + "epoch": 0.5007402707275804, + "grad_norm": 7.5, + "learning_rate": 4.935778044419761e-06, + "loss": 2.439, + "step": 4735 + }, + { + "epoch": 0.5008460236886633, + "grad_norm": 7.59375, + "learning_rate": 4.935739470810615e-06, + "loss": 2.517, + "step": 4736 + }, + { + "epoch": 0.5009517766497462, + "grad_norm": 8.0625, + "learning_rate": 4.935700885771541e-06, + "loss": 2.6271, + "step": 4737 + }, + { + "epoch": 0.501057529610829, + "grad_norm": 8.5625, + "learning_rate": 4.935662289302718e-06, + "loss": 2.9399, + "step": 4738 + }, + { + "epoch": 0.501163282571912, + "grad_norm": 7.5625, + "learning_rate": 4.935623681404328e-06, + "loss": 2.6806, + "step": 4739 + }, + { + "epoch": 0.501269035532995, + "grad_norm": 7.8125, + "learning_rate": 4.935585062076552e-06, + "loss": 2.6186, + "step": 4740 + }, + { + "epoch": 0.5013747884940778, + "grad_norm": 7.53125, + "learning_rate": 4.935546431319572e-06, + "loss": 2.51, + "step": 4741 + }, + { + "epoch": 0.5014805414551607, + "grad_norm": 7.15625, + "learning_rate": 4.935507789133569e-06, + "loss": 2.4849, + "step": 4742 + }, + { + "epoch": 0.5015862944162437, + "grad_norm": 8.0625, + "learning_rate": 4.935469135518724e-06, + "loss": 2.4981, + "step": 4743 + }, + { + "epoch": 0.5016920473773265, + "grad_norm": 8.3125, + "learning_rate": 4.935430470475218e-06, + "loss": 2.5763, + "step": 4744 + }, + { + "epoch": 0.5017978003384095, + "grad_norm": 7.40625, + "learning_rate": 4.935391794003232e-06, + "loss": 2.534, + "step": 4745 + }, + { + "epoch": 0.5019035532994924, + "grad_norm": 8.375, + "learning_rate": 4.935353106102949e-06, + "loss": 2.8295, + "step": 4746 + }, + { + "epoch": 0.5020093062605753, + "grad_norm": 7.59375, + "learning_rate": 4.935314406774551e-06, + "loss": 2.4121, + "step": 4747 + }, + { + "epoch": 0.5021150592216582, + "grad_norm": 8.25, + "learning_rate": 4.935275696018217e-06, + "loss": 2.3549, + "step": 4748 + }, + { + "epoch": 0.5022208121827412, + "grad_norm": 7.5, + "learning_rate": 4.935236973834131e-06, + "loss": 2.5257, + "step": 4749 + }, + { + "epoch": 0.502326565143824, + "grad_norm": 8.375, + "learning_rate": 4.935198240222473e-06, + "loss": 2.6825, + "step": 4750 + }, + { + "epoch": 0.502432318104907, + "grad_norm": 8.25, + "learning_rate": 4.935159495183427e-06, + "loss": 2.6936, + "step": 4751 + }, + { + "epoch": 0.5025380710659898, + "grad_norm": 7.84375, + "learning_rate": 4.935120738717172e-06, + "loss": 2.6095, + "step": 4752 + }, + { + "epoch": 0.5026438240270727, + "grad_norm": 7.6875, + "learning_rate": 4.935081970823892e-06, + "loss": 2.6198, + "step": 4753 + }, + { + "epoch": 0.5027495769881557, + "grad_norm": 7.3125, + "learning_rate": 4.9350431915037685e-06, + "loss": 2.6206, + "step": 4754 + }, + { + "epoch": 0.5028553299492385, + "grad_norm": 8.25, + "learning_rate": 4.935004400756982e-06, + "loss": 2.6917, + "step": 4755 + }, + { + "epoch": 0.5029610829103215, + "grad_norm": 7.96875, + "learning_rate": 4.934965598583718e-06, + "loss": 2.5335, + "step": 4756 + }, + { + "epoch": 0.5030668358714044, + "grad_norm": 8.8125, + "learning_rate": 4.934926784984155e-06, + "loss": 2.3132, + "step": 4757 + }, + { + "epoch": 0.5031725888324873, + "grad_norm": 8.0, + "learning_rate": 4.934887959958475e-06, + "loss": 2.3977, + "step": 4758 + }, + { + "epoch": 0.5032783417935702, + "grad_norm": 7.75, + "learning_rate": 4.934849123506864e-06, + "loss": 2.5932, + "step": 4759 + }, + { + "epoch": 0.5033840947546532, + "grad_norm": 8.125, + "learning_rate": 4.9348102756295e-06, + "loss": 2.6227, + "step": 4760 + }, + { + "epoch": 0.503489847715736, + "grad_norm": 7.90625, + "learning_rate": 4.934771416326569e-06, + "loss": 2.5566, + "step": 4761 + }, + { + "epoch": 0.503595600676819, + "grad_norm": 7.4375, + "learning_rate": 4.934732545598249e-06, + "loss": 2.5319, + "step": 4762 + }, + { + "epoch": 0.5037013536379019, + "grad_norm": 7.75, + "learning_rate": 4.934693663444727e-06, + "loss": 2.5949, + "step": 4763 + }, + { + "epoch": 0.5038071065989848, + "grad_norm": 7.625, + "learning_rate": 4.934654769866184e-06, + "loss": 2.6845, + "step": 4764 + }, + { + "epoch": 0.5039128595600677, + "grad_norm": 7.78125, + "learning_rate": 4.9346158648628006e-06, + "loss": 2.5999, + "step": 4765 + }, + { + "epoch": 0.5040186125211505, + "grad_norm": 8.125, + "learning_rate": 4.934576948434761e-06, + "loss": 2.5828, + "step": 4766 + }, + { + "epoch": 0.5041243654822335, + "grad_norm": 7.53125, + "learning_rate": 4.934538020582247e-06, + "loss": 2.6241, + "step": 4767 + }, + { + "epoch": 0.5042301184433164, + "grad_norm": 7.6875, + "learning_rate": 4.934499081305442e-06, + "loss": 2.3294, + "step": 4768 + }, + { + "epoch": 0.5043358714043993, + "grad_norm": 8.5625, + "learning_rate": 4.934460130604529e-06, + "loss": 2.7415, + "step": 4769 + }, + { + "epoch": 0.5044416243654822, + "grad_norm": 7.4375, + "learning_rate": 4.9344211684796894e-06, + "loss": 2.6587, + "step": 4770 + }, + { + "epoch": 0.5045473773265652, + "grad_norm": 7.875, + "learning_rate": 4.934382194931108e-06, + "loss": 2.7571, + "step": 4771 + }, + { + "epoch": 0.504653130287648, + "grad_norm": 8.0625, + "learning_rate": 4.9343432099589665e-06, + "loss": 2.5934, + "step": 4772 + }, + { + "epoch": 0.504758883248731, + "grad_norm": 8.4375, + "learning_rate": 4.934304213563447e-06, + "loss": 2.4058, + "step": 4773 + }, + { + "epoch": 0.5048646362098139, + "grad_norm": 8.25, + "learning_rate": 4.934265205744734e-06, + "loss": 2.6576, + "step": 4774 + }, + { + "epoch": 0.5049703891708968, + "grad_norm": 7.6875, + "learning_rate": 4.93422618650301e-06, + "loss": 2.6122, + "step": 4775 + }, + { + "epoch": 0.5050761421319797, + "grad_norm": 8.0, + "learning_rate": 4.934187155838458e-06, + "loss": 2.6028, + "step": 4776 + }, + { + "epoch": 0.5051818950930627, + "grad_norm": 7.5, + "learning_rate": 4.93414811375126e-06, + "loss": 2.5782, + "step": 4777 + }, + { + "epoch": 0.5052876480541455, + "grad_norm": 7.71875, + "learning_rate": 4.934109060241602e-06, + "loss": 2.5524, + "step": 4778 + }, + { + "epoch": 0.5053934010152284, + "grad_norm": 8.0, + "learning_rate": 4.934069995309665e-06, + "loss": 2.754, + "step": 4779 + }, + { + "epoch": 0.5054991539763113, + "grad_norm": 7.65625, + "learning_rate": 4.934030918955634e-06, + "loss": 2.5633, + "step": 4780 + }, + { + "epoch": 0.5056049069373942, + "grad_norm": 7.78125, + "learning_rate": 4.93399183117969e-06, + "loss": 2.5019, + "step": 4781 + }, + { + "epoch": 0.5057106598984772, + "grad_norm": 7.75, + "learning_rate": 4.933952731982019e-06, + "loss": 2.5197, + "step": 4782 + }, + { + "epoch": 0.50581641285956, + "grad_norm": 7.84375, + "learning_rate": 4.9339136213628025e-06, + "loss": 2.676, + "step": 4783 + }, + { + "epoch": 0.505922165820643, + "grad_norm": 7.5, + "learning_rate": 4.933874499322225e-06, + "loss": 2.4672, + "step": 4784 + }, + { + "epoch": 0.5060279187817259, + "grad_norm": 7.875, + "learning_rate": 4.93383536586047e-06, + "loss": 2.6524, + "step": 4785 + }, + { + "epoch": 0.5061336717428088, + "grad_norm": 7.9375, + "learning_rate": 4.93379622097772e-06, + "loss": 2.3057, + "step": 4786 + }, + { + "epoch": 0.5062394247038917, + "grad_norm": 7.6875, + "learning_rate": 4.933757064674161e-06, + "loss": 2.4902, + "step": 4787 + }, + { + "epoch": 0.5063451776649747, + "grad_norm": 11.1875, + "learning_rate": 4.933717896949976e-06, + "loss": 2.5782, + "step": 4788 + }, + { + "epoch": 0.5064509306260575, + "grad_norm": 8.4375, + "learning_rate": 4.933678717805347e-06, + "loss": 2.7718, + "step": 4789 + }, + { + "epoch": 0.5065566835871405, + "grad_norm": 7.375, + "learning_rate": 4.9336395272404595e-06, + "loss": 2.3852, + "step": 4790 + }, + { + "epoch": 0.5066624365482234, + "grad_norm": 8.0625, + "learning_rate": 4.933600325255497e-06, + "loss": 2.8004, + "step": 4791 + }, + { + "epoch": 0.5067681895093062, + "grad_norm": 7.90625, + "learning_rate": 4.9335611118506436e-06, + "loss": 2.3775, + "step": 4792 + }, + { + "epoch": 0.5068739424703892, + "grad_norm": 8.1875, + "learning_rate": 4.933521887026083e-06, + "loss": 2.4828, + "step": 4793 + }, + { + "epoch": 0.506979695431472, + "grad_norm": 9.125, + "learning_rate": 4.933482650782e-06, + "loss": 2.4991, + "step": 4794 + }, + { + "epoch": 0.507085448392555, + "grad_norm": 8.5, + "learning_rate": 4.933443403118577e-06, + "loss": 2.4144, + "step": 4795 + }, + { + "epoch": 0.5071912013536379, + "grad_norm": 8.625, + "learning_rate": 4.9334041440360004e-06, + "loss": 2.754, + "step": 4796 + }, + { + "epoch": 0.5072969543147208, + "grad_norm": 7.8125, + "learning_rate": 4.933364873534453e-06, + "loss": 2.6621, + "step": 4797 + }, + { + "epoch": 0.5074027072758037, + "grad_norm": 7.65625, + "learning_rate": 4.93332559161412e-06, + "loss": 2.7081, + "step": 4798 + }, + { + "epoch": 0.5075084602368867, + "grad_norm": 7.5625, + "learning_rate": 4.933286298275185e-06, + "loss": 2.6647, + "step": 4799 + }, + { + "epoch": 0.5076142131979695, + "grad_norm": 7.78125, + "learning_rate": 4.933246993517832e-06, + "loss": 2.6876, + "step": 4800 + }, + { + "epoch": 0.5077199661590525, + "grad_norm": 7.96875, + "learning_rate": 4.9332076773422465e-06, + "loss": 2.6509, + "step": 4801 + }, + { + "epoch": 0.5078257191201354, + "grad_norm": 7.59375, + "learning_rate": 4.933168349748613e-06, + "loss": 2.6069, + "step": 4802 + }, + { + "epoch": 0.5079314720812182, + "grad_norm": 7.84375, + "learning_rate": 4.9331290107371146e-06, + "loss": 2.4901, + "step": 4803 + }, + { + "epoch": 0.5080372250423012, + "grad_norm": 9.6875, + "learning_rate": 4.933089660307937e-06, + "loss": 2.5795, + "step": 4804 + }, + { + "epoch": 0.5081429780033841, + "grad_norm": 7.78125, + "learning_rate": 4.933050298461265e-06, + "loss": 2.2445, + "step": 4805 + }, + { + "epoch": 0.508248730964467, + "grad_norm": 7.71875, + "learning_rate": 4.933010925197283e-06, + "loss": 2.5913, + "step": 4806 + }, + { + "epoch": 0.5083544839255499, + "grad_norm": 8.0625, + "learning_rate": 4.932971540516176e-06, + "loss": 2.5448, + "step": 4807 + }, + { + "epoch": 0.5084602368866328, + "grad_norm": 7.375, + "learning_rate": 4.932932144418128e-06, + "loss": 2.5766, + "step": 4808 + }, + { + "epoch": 0.5085659898477157, + "grad_norm": 7.53125, + "learning_rate": 4.932892736903326e-06, + "loss": 2.4758, + "step": 4809 + }, + { + "epoch": 0.5086717428087987, + "grad_norm": 8.0, + "learning_rate": 4.9328533179719514e-06, + "loss": 2.8212, + "step": 4810 + }, + { + "epoch": 0.5087774957698815, + "grad_norm": 8.375, + "learning_rate": 4.932813887624192e-06, + "loss": 2.8787, + "step": 4811 + }, + { + "epoch": 0.5088832487309645, + "grad_norm": 7.71875, + "learning_rate": 4.932774445860232e-06, + "loss": 2.6781, + "step": 4812 + }, + { + "epoch": 0.5089890016920474, + "grad_norm": 8.1875, + "learning_rate": 4.932734992680257e-06, + "loss": 2.7013, + "step": 4813 + }, + { + "epoch": 0.5090947546531303, + "grad_norm": 7.4375, + "learning_rate": 4.932695528084451e-06, + "loss": 2.4157, + "step": 4814 + }, + { + "epoch": 0.5092005076142132, + "grad_norm": 8.3125, + "learning_rate": 4.932656052073e-06, + "loss": 2.5098, + "step": 4815 + }, + { + "epoch": 0.5093062605752962, + "grad_norm": 7.5625, + "learning_rate": 4.932616564646089e-06, + "loss": 2.5504, + "step": 4816 + }, + { + "epoch": 0.509412013536379, + "grad_norm": 8.3125, + "learning_rate": 4.932577065803904e-06, + "loss": 2.5499, + "step": 4817 + }, + { + "epoch": 0.5095177664974619, + "grad_norm": 8.5, + "learning_rate": 4.932537555546629e-06, + "loss": 2.669, + "step": 4818 + }, + { + "epoch": 0.5096235194585449, + "grad_norm": 7.53125, + "learning_rate": 4.93249803387445e-06, + "loss": 2.5366, + "step": 4819 + }, + { + "epoch": 0.5097292724196277, + "grad_norm": 7.59375, + "learning_rate": 4.932458500787552e-06, + "loss": 2.495, + "step": 4820 + }, + { + "epoch": 0.5098350253807107, + "grad_norm": 8.5, + "learning_rate": 4.932418956286123e-06, + "loss": 2.6629, + "step": 4821 + }, + { + "epoch": 0.5099407783417935, + "grad_norm": 7.71875, + "learning_rate": 4.932379400370345e-06, + "loss": 2.6078, + "step": 4822 + }, + { + "epoch": 0.5100465313028765, + "grad_norm": 8.9375, + "learning_rate": 4.932339833040406e-06, + "loss": 2.6205, + "step": 4823 + }, + { + "epoch": 0.5101522842639594, + "grad_norm": 7.5, + "learning_rate": 4.932300254296492e-06, + "loss": 2.6576, + "step": 4824 + }, + { + "epoch": 0.5102580372250423, + "grad_norm": 7.71875, + "learning_rate": 4.932260664138786e-06, + "loss": 2.7071, + "step": 4825 + }, + { + "epoch": 0.5103637901861252, + "grad_norm": 7.78125, + "learning_rate": 4.932221062567476e-06, + "loss": 2.5624, + "step": 4826 + }, + { + "epoch": 0.5104695431472082, + "grad_norm": 7.375, + "learning_rate": 4.932181449582747e-06, + "loss": 2.5523, + "step": 4827 + }, + { + "epoch": 0.510575296108291, + "grad_norm": 7.5, + "learning_rate": 4.9321418251847855e-06, + "loss": 2.4678, + "step": 4828 + }, + { + "epoch": 0.510681049069374, + "grad_norm": 7.875, + "learning_rate": 4.932102189373778e-06, + "loss": 2.7478, + "step": 4829 + }, + { + "epoch": 0.5107868020304569, + "grad_norm": 7.40625, + "learning_rate": 4.932062542149908e-06, + "loss": 2.5004, + "step": 4830 + }, + { + "epoch": 0.5108925549915397, + "grad_norm": 7.375, + "learning_rate": 4.932022883513364e-06, + "loss": 2.5765, + "step": 4831 + }, + { + "epoch": 0.5109983079526227, + "grad_norm": 8.1875, + "learning_rate": 4.931983213464331e-06, + "loss": 2.643, + "step": 4832 + }, + { + "epoch": 0.5111040609137056, + "grad_norm": 7.71875, + "learning_rate": 4.931943532002996e-06, + "loss": 2.5901, + "step": 4833 + }, + { + "epoch": 0.5112098138747885, + "grad_norm": 7.625, + "learning_rate": 4.931903839129545e-06, + "loss": 2.3799, + "step": 4834 + }, + { + "epoch": 0.5113155668358714, + "grad_norm": 7.5, + "learning_rate": 4.931864134844162e-06, + "loss": 2.5404, + "step": 4835 + }, + { + "epoch": 0.5114213197969543, + "grad_norm": 7.9375, + "learning_rate": 4.931824419147037e-06, + "loss": 2.4995, + "step": 4836 + }, + { + "epoch": 0.5115270727580372, + "grad_norm": 7.96875, + "learning_rate": 4.9317846920383535e-06, + "loss": 2.6283, + "step": 4837 + }, + { + "epoch": 0.5116328257191202, + "grad_norm": 7.9375, + "learning_rate": 4.9317449535183e-06, + "loss": 2.5109, + "step": 4838 + }, + { + "epoch": 0.511738578680203, + "grad_norm": 8.3125, + "learning_rate": 4.931705203587062e-06, + "loss": 2.5407, + "step": 4839 + }, + { + "epoch": 0.511844331641286, + "grad_norm": 8.0625, + "learning_rate": 4.931665442244825e-06, + "loss": 2.3774, + "step": 4840 + }, + { + "epoch": 0.5119500846023689, + "grad_norm": 8.4375, + "learning_rate": 4.9316256694917765e-06, + "loss": 2.8568, + "step": 4841 + }, + { + "epoch": 0.5120558375634517, + "grad_norm": 7.6875, + "learning_rate": 4.931585885328104e-06, + "loss": 2.5125, + "step": 4842 + }, + { + "epoch": 0.5121615905245347, + "grad_norm": 8.5, + "learning_rate": 4.931546089753994e-06, + "loss": 2.3647, + "step": 4843 + }, + { + "epoch": 0.5122673434856176, + "grad_norm": 7.65625, + "learning_rate": 4.931506282769632e-06, + "loss": 2.6309, + "step": 4844 + }, + { + "epoch": 0.5123730964467005, + "grad_norm": 7.59375, + "learning_rate": 4.931466464375205e-06, + "loss": 2.429, + "step": 4845 + }, + { + "epoch": 0.5124788494077834, + "grad_norm": 7.9375, + "learning_rate": 4.931426634570901e-06, + "loss": 2.5643, + "step": 4846 + }, + { + "epoch": 0.5125846023688664, + "grad_norm": 7.59375, + "learning_rate": 4.9313867933569056e-06, + "loss": 2.6425, + "step": 4847 + }, + { + "epoch": 0.5126903553299492, + "grad_norm": 8.0, + "learning_rate": 4.931346940733406e-06, + "loss": 2.6325, + "step": 4848 + }, + { + "epoch": 0.5127961082910322, + "grad_norm": 7.375, + "learning_rate": 4.931307076700591e-06, + "loss": 2.646, + "step": 4849 + }, + { + "epoch": 0.512901861252115, + "grad_norm": 7.90625, + "learning_rate": 4.9312672012586445e-06, + "loss": 2.6662, + "step": 4850 + }, + { + "epoch": 0.513007614213198, + "grad_norm": 8.5625, + "learning_rate": 4.9312273144077575e-06, + "loss": 2.6633, + "step": 4851 + }, + { + "epoch": 0.5131133671742809, + "grad_norm": 8.25, + "learning_rate": 4.9311874161481134e-06, + "loss": 2.4947, + "step": 4852 + }, + { + "epoch": 0.5132191201353637, + "grad_norm": 7.625, + "learning_rate": 4.931147506479902e-06, + "loss": 2.5265, + "step": 4853 + }, + { + "epoch": 0.5133248730964467, + "grad_norm": 7.875, + "learning_rate": 4.931107585403309e-06, + "loss": 2.7284, + "step": 4854 + }, + { + "epoch": 0.5134306260575296, + "grad_norm": 7.6875, + "learning_rate": 4.9310676529185235e-06, + "loss": 2.6455, + "step": 4855 + }, + { + "epoch": 0.5135363790186125, + "grad_norm": 7.34375, + "learning_rate": 4.931027709025731e-06, + "loss": 2.4655, + "step": 4856 + }, + { + "epoch": 0.5136421319796954, + "grad_norm": 8.3125, + "learning_rate": 4.930987753725119e-06, + "loss": 2.889, + "step": 4857 + }, + { + "epoch": 0.5137478849407784, + "grad_norm": 7.84375, + "learning_rate": 4.930947787016877e-06, + "loss": 2.6582, + "step": 4858 + }, + { + "epoch": 0.5138536379018612, + "grad_norm": 7.90625, + "learning_rate": 4.9309078089011906e-06, + "loss": 2.6522, + "step": 4859 + }, + { + "epoch": 0.5139593908629442, + "grad_norm": 7.9375, + "learning_rate": 4.9308678193782476e-06, + "loss": 2.7947, + "step": 4860 + }, + { + "epoch": 0.5140651438240271, + "grad_norm": 7.625, + "learning_rate": 4.9308278184482365e-06, + "loss": 2.5784, + "step": 4861 + }, + { + "epoch": 0.51417089678511, + "grad_norm": 7.625, + "learning_rate": 4.930787806111345e-06, + "loss": 2.5554, + "step": 4862 + }, + { + "epoch": 0.5142766497461929, + "grad_norm": 8.125, + "learning_rate": 4.9307477823677605e-06, + "loss": 2.4345, + "step": 4863 + }, + { + "epoch": 0.5143824027072758, + "grad_norm": 7.625, + "learning_rate": 4.93070774721767e-06, + "loss": 2.4738, + "step": 4864 + }, + { + "epoch": 0.5144881556683587, + "grad_norm": 8.25, + "learning_rate": 4.930667700661263e-06, + "loss": 2.7304, + "step": 4865 + }, + { + "epoch": 0.5145939086294417, + "grad_norm": 7.5625, + "learning_rate": 4.930627642698727e-06, + "loss": 2.7089, + "step": 4866 + }, + { + "epoch": 0.5146996615905245, + "grad_norm": 8.0, + "learning_rate": 4.930587573330248e-06, + "loss": 2.5791, + "step": 4867 + }, + { + "epoch": 0.5148054145516074, + "grad_norm": 7.65625, + "learning_rate": 4.930547492556017e-06, + "loss": 2.2796, + "step": 4868 + }, + { + "epoch": 0.5149111675126904, + "grad_norm": 7.5625, + "learning_rate": 4.93050740037622e-06, + "loss": 2.4022, + "step": 4869 + }, + { + "epoch": 0.5150169204737732, + "grad_norm": 7.5, + "learning_rate": 4.930467296791046e-06, + "loss": 2.5638, + "step": 4870 + }, + { + "epoch": 0.5151226734348562, + "grad_norm": 7.5625, + "learning_rate": 4.930427181800683e-06, + "loss": 2.3018, + "step": 4871 + }, + { + "epoch": 0.5152284263959391, + "grad_norm": 7.875, + "learning_rate": 4.930387055405319e-06, + "loss": 2.7564, + "step": 4872 + }, + { + "epoch": 0.515334179357022, + "grad_norm": 7.9375, + "learning_rate": 4.930346917605143e-06, + "loss": 2.5838, + "step": 4873 + }, + { + "epoch": 0.5154399323181049, + "grad_norm": 8.0, + "learning_rate": 4.9303067684003425e-06, + "loss": 2.5724, + "step": 4874 + }, + { + "epoch": 0.5155456852791879, + "grad_norm": 7.6875, + "learning_rate": 4.9302666077911065e-06, + "loss": 2.7617, + "step": 4875 + }, + { + "epoch": 0.5156514382402707, + "grad_norm": 8.75, + "learning_rate": 4.930226435777623e-06, + "loss": 2.4516, + "step": 4876 + }, + { + "epoch": 0.5157571912013537, + "grad_norm": 8.0, + "learning_rate": 4.930186252360081e-06, + "loss": 2.7777, + "step": 4877 + }, + { + "epoch": 0.5158629441624365, + "grad_norm": 7.4375, + "learning_rate": 4.930146057538669e-06, + "loss": 2.5402, + "step": 4878 + }, + { + "epoch": 0.5159686971235194, + "grad_norm": 8.25, + "learning_rate": 4.930105851313575e-06, + "loss": 2.6938, + "step": 4879 + }, + { + "epoch": 0.5160744500846024, + "grad_norm": 8.0, + "learning_rate": 4.9300656336849885e-06, + "loss": 2.6229, + "step": 4880 + }, + { + "epoch": 0.5161802030456852, + "grad_norm": 7.59375, + "learning_rate": 4.930025404653098e-06, + "loss": 2.5731, + "step": 4881 + }, + { + "epoch": 0.5162859560067682, + "grad_norm": 7.59375, + "learning_rate": 4.929985164218092e-06, + "loss": 2.6875, + "step": 4882 + }, + { + "epoch": 0.5163917089678511, + "grad_norm": 7.71875, + "learning_rate": 4.929944912380159e-06, + "loss": 2.5551, + "step": 4883 + }, + { + "epoch": 0.516497461928934, + "grad_norm": 7.8125, + "learning_rate": 4.929904649139489e-06, + "loss": 2.5685, + "step": 4884 + }, + { + "epoch": 0.5166032148900169, + "grad_norm": 8.4375, + "learning_rate": 4.9298643744962696e-06, + "loss": 2.5867, + "step": 4885 + }, + { + "epoch": 0.5167089678510999, + "grad_norm": 7.375, + "learning_rate": 4.92982408845069e-06, + "loss": 2.4475, + "step": 4886 + }, + { + "epoch": 0.5168147208121827, + "grad_norm": 7.6875, + "learning_rate": 4.929783791002941e-06, + "loss": 2.4468, + "step": 4887 + }, + { + "epoch": 0.5169204737732657, + "grad_norm": 7.71875, + "learning_rate": 4.92974348215321e-06, + "loss": 2.6177, + "step": 4888 + }, + { + "epoch": 0.5170262267343486, + "grad_norm": 7.78125, + "learning_rate": 4.929703161901686e-06, + "loss": 2.739, + "step": 4889 + }, + { + "epoch": 0.5171319796954315, + "grad_norm": 8.0, + "learning_rate": 4.92966283024856e-06, + "loss": 2.6889, + "step": 4890 + }, + { + "epoch": 0.5172377326565144, + "grad_norm": 7.78125, + "learning_rate": 4.929622487194019e-06, + "loss": 2.5334, + "step": 4891 + }, + { + "epoch": 0.5173434856175972, + "grad_norm": 8.3125, + "learning_rate": 4.929582132738253e-06, + "loss": 2.4959, + "step": 4892 + }, + { + "epoch": 0.5174492385786802, + "grad_norm": 7.9375, + "learning_rate": 4.929541766881453e-06, + "loss": 2.8237, + "step": 4893 + }, + { + "epoch": 0.5175549915397631, + "grad_norm": 7.875, + "learning_rate": 4.9295013896238066e-06, + "loss": 2.6365, + "step": 4894 + }, + { + "epoch": 0.517660744500846, + "grad_norm": 8.0625, + "learning_rate": 4.929461000965503e-06, + "loss": 2.3778, + "step": 4895 + }, + { + "epoch": 0.5177664974619289, + "grad_norm": 8.4375, + "learning_rate": 4.9294206009067335e-06, + "loss": 2.711, + "step": 4896 + }, + { + "epoch": 0.5178722504230119, + "grad_norm": 7.6875, + "learning_rate": 4.929380189447686e-06, + "loss": 2.528, + "step": 4897 + }, + { + "epoch": 0.5179780033840947, + "grad_norm": 7.875, + "learning_rate": 4.929339766588552e-06, + "loss": 2.4797, + "step": 4898 + }, + { + "epoch": 0.5180837563451777, + "grad_norm": 7.4375, + "learning_rate": 4.929299332329519e-06, + "loss": 2.2302, + "step": 4899 + }, + { + "epoch": 0.5181895093062606, + "grad_norm": 7.8125, + "learning_rate": 4.929258886670778e-06, + "loss": 2.5959, + "step": 4900 + }, + { + "epoch": 0.5182952622673435, + "grad_norm": 8.4375, + "learning_rate": 4.929218429612519e-06, + "loss": 2.7472, + "step": 4901 + }, + { + "epoch": 0.5184010152284264, + "grad_norm": 7.34375, + "learning_rate": 4.929177961154931e-06, + "loss": 2.3541, + "step": 4902 + }, + { + "epoch": 0.5185067681895094, + "grad_norm": 7.96875, + "learning_rate": 4.929137481298205e-06, + "loss": 2.6572, + "step": 4903 + }, + { + "epoch": 0.5186125211505922, + "grad_norm": 7.875, + "learning_rate": 4.92909699004253e-06, + "loss": 2.5811, + "step": 4904 + }, + { + "epoch": 0.5187182741116751, + "grad_norm": 8.0, + "learning_rate": 4.9290564873880954e-06, + "loss": 2.7137, + "step": 4905 + }, + { + "epoch": 0.518824027072758, + "grad_norm": 7.375, + "learning_rate": 4.929015973335093e-06, + "loss": 2.5261, + "step": 4906 + }, + { + "epoch": 0.5189297800338409, + "grad_norm": 8.3125, + "learning_rate": 4.928975447883711e-06, + "loss": 2.5505, + "step": 4907 + }, + { + "epoch": 0.5190355329949239, + "grad_norm": 8.125, + "learning_rate": 4.928934911034142e-06, + "loss": 2.5436, + "step": 4908 + }, + { + "epoch": 0.5191412859560067, + "grad_norm": 7.875, + "learning_rate": 4.928894362786575e-06, + "loss": 2.335, + "step": 4909 + }, + { + "epoch": 0.5192470389170897, + "grad_norm": 8.3125, + "learning_rate": 4.928853803141199e-06, + "loss": 2.5967, + "step": 4910 + }, + { + "epoch": 0.5193527918781726, + "grad_norm": 8.3125, + "learning_rate": 4.9288132320982066e-06, + "loss": 2.8917, + "step": 4911 + }, + { + "epoch": 0.5194585448392555, + "grad_norm": 8.25, + "learning_rate": 4.928772649657786e-06, + "loss": 2.7527, + "step": 4912 + }, + { + "epoch": 0.5195642978003384, + "grad_norm": 7.65625, + "learning_rate": 4.928732055820129e-06, + "loss": 2.5392, + "step": 4913 + }, + { + "epoch": 0.5196700507614214, + "grad_norm": 8.125, + "learning_rate": 4.928691450585426e-06, + "loss": 2.6219, + "step": 4914 + }, + { + "epoch": 0.5197758037225042, + "grad_norm": 7.84375, + "learning_rate": 4.928650833953867e-06, + "loss": 2.715, + "step": 4915 + }, + { + "epoch": 0.5198815566835872, + "grad_norm": 8.5, + "learning_rate": 4.928610205925644e-06, + "loss": 2.8227, + "step": 4916 + }, + { + "epoch": 0.5199873096446701, + "grad_norm": 7.8125, + "learning_rate": 4.928569566500945e-06, + "loss": 2.7154, + "step": 4917 + }, + { + "epoch": 0.5200930626057529, + "grad_norm": 8.3125, + "learning_rate": 4.928528915679963e-06, + "loss": 2.549, + "step": 4918 + }, + { + "epoch": 0.5201988155668359, + "grad_norm": 8.25, + "learning_rate": 4.928488253462888e-06, + "loss": 2.7126, + "step": 4919 + }, + { + "epoch": 0.5203045685279187, + "grad_norm": 7.65625, + "learning_rate": 4.928447579849911e-06, + "loss": 2.6559, + "step": 4920 + }, + { + "epoch": 0.5204103214890017, + "grad_norm": 11.0, + "learning_rate": 4.928406894841222e-06, + "loss": 2.5595, + "step": 4921 + }, + { + "epoch": 0.5205160744500846, + "grad_norm": 7.65625, + "learning_rate": 4.928366198437014e-06, + "loss": 2.5634, + "step": 4922 + }, + { + "epoch": 0.5206218274111675, + "grad_norm": 7.53125, + "learning_rate": 4.928325490637476e-06, + "loss": 2.4424, + "step": 4923 + }, + { + "epoch": 0.5207275803722504, + "grad_norm": 7.90625, + "learning_rate": 4.928284771442799e-06, + "loss": 2.6203, + "step": 4924 + }, + { + "epoch": 0.5208333333333334, + "grad_norm": 9.0625, + "learning_rate": 4.928244040853175e-06, + "loss": 2.9337, + "step": 4925 + }, + { + "epoch": 0.5209390862944162, + "grad_norm": 7.8125, + "learning_rate": 4.928203298868795e-06, + "loss": 2.509, + "step": 4926 + }, + { + "epoch": 0.5210448392554992, + "grad_norm": 8.375, + "learning_rate": 4.92816254548985e-06, + "loss": 2.7747, + "step": 4927 + }, + { + "epoch": 0.5211505922165821, + "grad_norm": 8.125, + "learning_rate": 4.9281217807165315e-06, + "loss": 2.3855, + "step": 4928 + }, + { + "epoch": 0.521256345177665, + "grad_norm": 7.875, + "learning_rate": 4.92808100454903e-06, + "loss": 2.6519, + "step": 4929 + }, + { + "epoch": 0.5213620981387479, + "grad_norm": 8.0, + "learning_rate": 4.928040216987537e-06, + "loss": 2.755, + "step": 4930 + }, + { + "epoch": 0.5214678510998308, + "grad_norm": 8.25, + "learning_rate": 4.927999418032245e-06, + "loss": 2.6205, + "step": 4931 + }, + { + "epoch": 0.5215736040609137, + "grad_norm": 8.125, + "learning_rate": 4.927958607683345e-06, + "loss": 2.5763, + "step": 4932 + }, + { + "epoch": 0.5216793570219966, + "grad_norm": 8.5625, + "learning_rate": 4.9279177859410285e-06, + "loss": 2.4864, + "step": 4933 + }, + { + "epoch": 0.5217851099830795, + "grad_norm": 7.75, + "learning_rate": 4.927876952805486e-06, + "loss": 2.2043, + "step": 4934 + }, + { + "epoch": 0.5218908629441624, + "grad_norm": 7.96875, + "learning_rate": 4.927836108276911e-06, + "loss": 2.5915, + "step": 4935 + }, + { + "epoch": 0.5219966159052454, + "grad_norm": 8.0625, + "learning_rate": 4.927795252355493e-06, + "loss": 2.4931, + "step": 4936 + }, + { + "epoch": 0.5221023688663282, + "grad_norm": 7.46875, + "learning_rate": 4.927754385041425e-06, + "loss": 2.5635, + "step": 4937 + }, + { + "epoch": 0.5222081218274112, + "grad_norm": 7.9375, + "learning_rate": 4.927713506334899e-06, + "loss": 2.569, + "step": 4938 + }, + { + "epoch": 0.5223138747884941, + "grad_norm": 8.8125, + "learning_rate": 4.927672616236105e-06, + "loss": 2.2591, + "step": 4939 + }, + { + "epoch": 0.522419627749577, + "grad_norm": 8.125, + "learning_rate": 4.927631714745238e-06, + "loss": 2.552, + "step": 4940 + }, + { + "epoch": 0.5225253807106599, + "grad_norm": 8.75, + "learning_rate": 4.927590801862488e-06, + "loss": 2.6888, + "step": 4941 + }, + { + "epoch": 0.5226311336717429, + "grad_norm": 7.9375, + "learning_rate": 4.927549877588047e-06, + "loss": 2.553, + "step": 4942 + }, + { + "epoch": 0.5227368866328257, + "grad_norm": 7.6875, + "learning_rate": 4.927508941922106e-06, + "loss": 2.6554, + "step": 4943 + }, + { + "epoch": 0.5228426395939086, + "grad_norm": 8.75, + "learning_rate": 4.927467994864861e-06, + "loss": 2.4366, + "step": 4944 + }, + { + "epoch": 0.5229483925549916, + "grad_norm": 7.9375, + "learning_rate": 4.9274270364164995e-06, + "loss": 2.796, + "step": 4945 + }, + { + "epoch": 0.5230541455160744, + "grad_norm": 7.90625, + "learning_rate": 4.927386066577216e-06, + "loss": 2.6078, + "step": 4946 + }, + { + "epoch": 0.5231598984771574, + "grad_norm": 7.90625, + "learning_rate": 4.927345085347202e-06, + "loss": 2.7566, + "step": 4947 + }, + { + "epoch": 0.5232656514382402, + "grad_norm": 9.0625, + "learning_rate": 4.927304092726652e-06, + "loss": 2.4988, + "step": 4948 + }, + { + "epoch": 0.5233714043993232, + "grad_norm": 7.78125, + "learning_rate": 4.927263088715755e-06, + "loss": 2.6074, + "step": 4949 + }, + { + "epoch": 0.5234771573604061, + "grad_norm": 7.96875, + "learning_rate": 4.927222073314706e-06, + "loss": 2.5961, + "step": 4950 + }, + { + "epoch": 0.523582910321489, + "grad_norm": 7.8125, + "learning_rate": 4.927181046523696e-06, + "loss": 2.5068, + "step": 4951 + }, + { + "epoch": 0.5236886632825719, + "grad_norm": 7.875, + "learning_rate": 4.927140008342918e-06, + "loss": 2.3819, + "step": 4952 + }, + { + "epoch": 0.5237944162436549, + "grad_norm": 8.625, + "learning_rate": 4.927098958772565e-06, + "loss": 2.7728, + "step": 4953 + }, + { + "epoch": 0.5239001692047377, + "grad_norm": 8.5625, + "learning_rate": 4.927057897812829e-06, + "loss": 2.5567, + "step": 4954 + }, + { + "epoch": 0.5240059221658206, + "grad_norm": 7.5, + "learning_rate": 4.927016825463903e-06, + "loss": 2.6017, + "step": 4955 + }, + { + "epoch": 0.5241116751269036, + "grad_norm": 8.0, + "learning_rate": 4.926975741725979e-06, + "loss": 2.5583, + "step": 4956 + }, + { + "epoch": 0.5242174280879864, + "grad_norm": 7.8125, + "learning_rate": 4.926934646599251e-06, + "loss": 2.8302, + "step": 4957 + }, + { + "epoch": 0.5243231810490694, + "grad_norm": 8.1875, + "learning_rate": 4.926893540083911e-06, + "loss": 2.8968, + "step": 4958 + }, + { + "epoch": 0.5244289340101523, + "grad_norm": 7.84375, + "learning_rate": 4.926852422180153e-06, + "loss": 2.6283, + "step": 4959 + }, + { + "epoch": 0.5245346869712352, + "grad_norm": 8.0625, + "learning_rate": 4.926811292888168e-06, + "loss": 2.6788, + "step": 4960 + }, + { + "epoch": 0.5246404399323181, + "grad_norm": 7.65625, + "learning_rate": 4.926770152208151e-06, + "loss": 2.5866, + "step": 4961 + }, + { + "epoch": 0.524746192893401, + "grad_norm": 8.0, + "learning_rate": 4.926729000140295e-06, + "loss": 2.8689, + "step": 4962 + }, + { + "epoch": 0.5248519458544839, + "grad_norm": 7.9375, + "learning_rate": 4.926687836684791e-06, + "loss": 2.6598, + "step": 4963 + }, + { + "epoch": 0.5249576988155669, + "grad_norm": 7.84375, + "learning_rate": 4.926646661841833e-06, + "loss": 2.58, + "step": 4964 + }, + { + "epoch": 0.5250634517766497, + "grad_norm": 8.5, + "learning_rate": 4.9266054756116154e-06, + "loss": 2.6281, + "step": 4965 + }, + { + "epoch": 0.5251692047377327, + "grad_norm": 8.25, + "learning_rate": 4.9265642779943314e-06, + "loss": 2.3694, + "step": 4966 + }, + { + "epoch": 0.5252749576988156, + "grad_norm": 8.3125, + "learning_rate": 4.926523068990172e-06, + "loss": 2.8324, + "step": 4967 + }, + { + "epoch": 0.5253807106598984, + "grad_norm": 7.8125, + "learning_rate": 4.926481848599334e-06, + "loss": 2.5664, + "step": 4968 + }, + { + "epoch": 0.5254864636209814, + "grad_norm": 7.78125, + "learning_rate": 4.926440616822008e-06, + "loss": 2.5365, + "step": 4969 + }, + { + "epoch": 0.5255922165820643, + "grad_norm": 7.625, + "learning_rate": 4.9263993736583895e-06, + "loss": 2.5293, + "step": 4970 + }, + { + "epoch": 0.5256979695431472, + "grad_norm": 7.9375, + "learning_rate": 4.9263581191086694e-06, + "loss": 2.459, + "step": 4971 + }, + { + "epoch": 0.5258037225042301, + "grad_norm": 8.0625, + "learning_rate": 4.926316853173044e-06, + "loss": 2.6607, + "step": 4972 + }, + { + "epoch": 0.5259094754653131, + "grad_norm": 7.46875, + "learning_rate": 4.926275575851707e-06, + "loss": 2.544, + "step": 4973 + }, + { + "epoch": 0.5260152284263959, + "grad_norm": 8.8125, + "learning_rate": 4.92623428714485e-06, + "loss": 2.825, + "step": 4974 + }, + { + "epoch": 0.5261209813874789, + "grad_norm": 9.8125, + "learning_rate": 4.926192987052667e-06, + "loss": 2.4914, + "step": 4975 + }, + { + "epoch": 0.5262267343485617, + "grad_norm": 7.375, + "learning_rate": 4.9261516755753536e-06, + "loss": 2.5245, + "step": 4976 + }, + { + "epoch": 0.5263324873096447, + "grad_norm": 7.5625, + "learning_rate": 4.926110352713103e-06, + "loss": 2.5333, + "step": 4977 + }, + { + "epoch": 0.5264382402707276, + "grad_norm": 7.96875, + "learning_rate": 4.926069018466108e-06, + "loss": 2.6308, + "step": 4978 + }, + { + "epoch": 0.5265439932318104, + "grad_norm": 7.6875, + "learning_rate": 4.926027672834563e-06, + "loss": 2.5556, + "step": 4979 + }, + { + "epoch": 0.5266497461928934, + "grad_norm": 7.8125, + "learning_rate": 4.925986315818664e-06, + "loss": 2.4871, + "step": 4980 + }, + { + "epoch": 0.5267554991539763, + "grad_norm": 7.34375, + "learning_rate": 4.925944947418601e-06, + "loss": 2.4593, + "step": 4981 + }, + { + "epoch": 0.5268612521150592, + "grad_norm": 7.78125, + "learning_rate": 4.925903567634572e-06, + "loss": 2.5397, + "step": 4982 + }, + { + "epoch": 0.5269670050761421, + "grad_norm": 7.5625, + "learning_rate": 4.92586217646677e-06, + "loss": 2.4584, + "step": 4983 + }, + { + "epoch": 0.5270727580372251, + "grad_norm": 7.875, + "learning_rate": 4.925820773915389e-06, + "loss": 2.613, + "step": 4984 + }, + { + "epoch": 0.5271785109983079, + "grad_norm": 7.65625, + "learning_rate": 4.9257793599806224e-06, + "loss": 2.6336, + "step": 4985 + }, + { + "epoch": 0.5272842639593909, + "grad_norm": 8.1875, + "learning_rate": 4.925737934662666e-06, + "loss": 2.5806, + "step": 4986 + }, + { + "epoch": 0.5273900169204738, + "grad_norm": 7.90625, + "learning_rate": 4.925696497961713e-06, + "loss": 2.708, + "step": 4987 + }, + { + "epoch": 0.5274957698815567, + "grad_norm": 8.625, + "learning_rate": 4.925655049877959e-06, + "loss": 2.5286, + "step": 4988 + }, + { + "epoch": 0.5276015228426396, + "grad_norm": 8.25, + "learning_rate": 4.925613590411597e-06, + "loss": 2.674, + "step": 4989 + }, + { + "epoch": 0.5277072758037225, + "grad_norm": 7.6875, + "learning_rate": 4.925572119562824e-06, + "loss": 2.5766, + "step": 4990 + }, + { + "epoch": 0.5278130287648054, + "grad_norm": 7.75, + "learning_rate": 4.925530637331832e-06, + "loss": 2.776, + "step": 4991 + }, + { + "epoch": 0.5279187817258884, + "grad_norm": 8.375, + "learning_rate": 4.9254891437188165e-06, + "loss": 2.7035, + "step": 4992 + }, + { + "epoch": 0.5280245346869712, + "grad_norm": 7.8125, + "learning_rate": 4.925447638723973e-06, + "loss": 2.5231, + "step": 4993 + }, + { + "epoch": 0.5281302876480541, + "grad_norm": 7.5625, + "learning_rate": 4.9254061223474955e-06, + "loss": 2.585, + "step": 4994 + }, + { + "epoch": 0.5282360406091371, + "grad_norm": 8.3125, + "learning_rate": 4.9253645945895785e-06, + "loss": 2.7494, + "step": 4995 + }, + { + "epoch": 0.5283417935702199, + "grad_norm": 7.9375, + "learning_rate": 4.925323055450419e-06, + "loss": 2.4986, + "step": 4996 + }, + { + "epoch": 0.5284475465313029, + "grad_norm": 7.875, + "learning_rate": 4.925281504930209e-06, + "loss": 2.2934, + "step": 4997 + }, + { + "epoch": 0.5285532994923858, + "grad_norm": 7.9375, + "learning_rate": 4.925239943029145e-06, + "loss": 2.7717, + "step": 4998 + }, + { + "epoch": 0.5286590524534687, + "grad_norm": 7.90625, + "learning_rate": 4.925198369747421e-06, + "loss": 2.7678, + "step": 4999 + }, + { + "epoch": 0.5287648054145516, + "grad_norm": 7.5, + "learning_rate": 4.9251567850852345e-06, + "loss": 2.5514, + "step": 5000 + }, + { + "epoch": 0.5288705583756346, + "grad_norm": 8.6875, + "learning_rate": 4.925115189042778e-06, + "loss": 2.4766, + "step": 5001 + }, + { + "epoch": 0.5289763113367174, + "grad_norm": 7.9375, + "learning_rate": 4.925073581620249e-06, + "loss": 2.5154, + "step": 5002 + }, + { + "epoch": 0.5290820642978004, + "grad_norm": 7.5625, + "learning_rate": 4.92503196281784e-06, + "loss": 2.5318, + "step": 5003 + }, + { + "epoch": 0.5291878172588832, + "grad_norm": 7.71875, + "learning_rate": 4.9249903326357486e-06, + "loss": 2.603, + "step": 5004 + }, + { + "epoch": 0.5292935702199661, + "grad_norm": 7.71875, + "learning_rate": 4.924948691074169e-06, + "loss": 2.6346, + "step": 5005 + }, + { + "epoch": 0.5293993231810491, + "grad_norm": 9.5, + "learning_rate": 4.924907038133297e-06, + "loss": 2.3116, + "step": 5006 + }, + { + "epoch": 0.5295050761421319, + "grad_norm": 7.625, + "learning_rate": 4.924865373813329e-06, + "loss": 2.5728, + "step": 5007 + }, + { + "epoch": 0.5296108291032149, + "grad_norm": 7.65625, + "learning_rate": 4.924823698114458e-06, + "loss": 2.3851, + "step": 5008 + }, + { + "epoch": 0.5297165820642978, + "grad_norm": 7.71875, + "learning_rate": 4.924782011036882e-06, + "loss": 2.6291, + "step": 5009 + }, + { + "epoch": 0.5298223350253807, + "grad_norm": 8.375, + "learning_rate": 4.924740312580795e-06, + "loss": 2.6512, + "step": 5010 + }, + { + "epoch": 0.5299280879864636, + "grad_norm": 7.4375, + "learning_rate": 4.924698602746395e-06, + "loss": 2.4316, + "step": 5011 + }, + { + "epoch": 0.5300338409475466, + "grad_norm": 7.8125, + "learning_rate": 4.924656881533875e-06, + "loss": 2.5519, + "step": 5012 + }, + { + "epoch": 0.5301395939086294, + "grad_norm": 8.1875, + "learning_rate": 4.924615148943433e-06, + "loss": 2.7818, + "step": 5013 + }, + { + "epoch": 0.5302453468697124, + "grad_norm": 8.5, + "learning_rate": 4.924573404975262e-06, + "loss": 2.2973, + "step": 5014 + }, + { + "epoch": 0.5303510998307953, + "grad_norm": 8.625, + "learning_rate": 4.924531649629561e-06, + "loss": 2.6027, + "step": 5015 + }, + { + "epoch": 0.5304568527918782, + "grad_norm": 7.71875, + "learning_rate": 4.924489882906523e-06, + "loss": 2.5133, + "step": 5016 + }, + { + "epoch": 0.5305626057529611, + "grad_norm": 7.875, + "learning_rate": 4.924448104806346e-06, + "loss": 2.6, + "step": 5017 + }, + { + "epoch": 0.5306683587140439, + "grad_norm": 7.96875, + "learning_rate": 4.924406315329226e-06, + "loss": 2.5928, + "step": 5018 + }, + { + "epoch": 0.5307741116751269, + "grad_norm": 7.5625, + "learning_rate": 4.92436451447536e-06, + "loss": 2.5026, + "step": 5019 + }, + { + "epoch": 0.5308798646362098, + "grad_norm": 7.65625, + "learning_rate": 4.924322702244942e-06, + "loss": 2.3937, + "step": 5020 + }, + { + "epoch": 0.5309856175972927, + "grad_norm": 8.375, + "learning_rate": 4.924280878638168e-06, + "loss": 2.5979, + "step": 5021 + }, + { + "epoch": 0.5310913705583756, + "grad_norm": 7.4375, + "learning_rate": 4.9242390436552355e-06, + "loss": 2.8108, + "step": 5022 + }, + { + "epoch": 0.5311971235194586, + "grad_norm": 8.6875, + "learning_rate": 4.924197197296342e-06, + "loss": 2.504, + "step": 5023 + }, + { + "epoch": 0.5313028764805414, + "grad_norm": 7.0625, + "learning_rate": 4.924155339561681e-06, + "loss": 2.4923, + "step": 5024 + }, + { + "epoch": 0.5314086294416244, + "grad_norm": 8.25, + "learning_rate": 4.9241134704514516e-06, + "loss": 2.6668, + "step": 5025 + }, + { + "epoch": 0.5315143824027073, + "grad_norm": 8.1875, + "learning_rate": 4.924071589965849e-06, + "loss": 2.7498, + "step": 5026 + }, + { + "epoch": 0.5316201353637902, + "grad_norm": 7.625, + "learning_rate": 4.924029698105069e-06, + "loss": 2.58, + "step": 5027 + }, + { + "epoch": 0.5317258883248731, + "grad_norm": 7.84375, + "learning_rate": 4.92398779486931e-06, + "loss": 2.6295, + "step": 5028 + }, + { + "epoch": 0.5318316412859561, + "grad_norm": 7.28125, + "learning_rate": 4.923945880258767e-06, + "loss": 2.48, + "step": 5029 + }, + { + "epoch": 0.5319373942470389, + "grad_norm": 7.96875, + "learning_rate": 4.9239039542736375e-06, + "loss": 2.7975, + "step": 5030 + }, + { + "epoch": 0.5320431472081218, + "grad_norm": 9.0, + "learning_rate": 4.923862016914119e-06, + "loss": 2.7797, + "step": 5031 + }, + { + "epoch": 0.5321489001692047, + "grad_norm": 7.65625, + "learning_rate": 4.923820068180406e-06, + "loss": 2.6973, + "step": 5032 + }, + { + "epoch": 0.5322546531302876, + "grad_norm": 7.53125, + "learning_rate": 4.9237781080726974e-06, + "loss": 2.6235, + "step": 5033 + }, + { + "epoch": 0.5323604060913706, + "grad_norm": 7.9375, + "learning_rate": 4.923736136591189e-06, + "loss": 2.6384, + "step": 5034 + }, + { + "epoch": 0.5324661590524534, + "grad_norm": 7.8125, + "learning_rate": 4.923694153736079e-06, + "loss": 2.4988, + "step": 5035 + }, + { + "epoch": 0.5325719120135364, + "grad_norm": 7.4375, + "learning_rate": 4.923652159507562e-06, + "loss": 2.5243, + "step": 5036 + }, + { + "epoch": 0.5326776649746193, + "grad_norm": 8.0625, + "learning_rate": 4.923610153905838e-06, + "loss": 2.6584, + "step": 5037 + }, + { + "epoch": 0.5327834179357022, + "grad_norm": 8.0625, + "learning_rate": 4.923568136931103e-06, + "loss": 2.6833, + "step": 5038 + }, + { + "epoch": 0.5328891708967851, + "grad_norm": 7.46875, + "learning_rate": 4.923526108583553e-06, + "loss": 2.6257, + "step": 5039 + }, + { + "epoch": 0.5329949238578681, + "grad_norm": 7.65625, + "learning_rate": 4.923484068863386e-06, + "loss": 2.5568, + "step": 5040 + }, + { + "epoch": 0.5331006768189509, + "grad_norm": 8.4375, + "learning_rate": 4.923442017770801e-06, + "loss": 2.8996, + "step": 5041 + }, + { + "epoch": 0.5332064297800339, + "grad_norm": 7.625, + "learning_rate": 4.9233999553059924e-06, + "loss": 2.6063, + "step": 5042 + }, + { + "epoch": 0.5333121827411168, + "grad_norm": 9.25, + "learning_rate": 4.92335788146916e-06, + "loss": 2.6117, + "step": 5043 + }, + { + "epoch": 0.5334179357021996, + "grad_norm": 7.59375, + "learning_rate": 4.923315796260499e-06, + "loss": 2.524, + "step": 5044 + }, + { + "epoch": 0.5335236886632826, + "grad_norm": 8.375, + "learning_rate": 4.9232736996802085e-06, + "loss": 2.7297, + "step": 5045 + }, + { + "epoch": 0.5336294416243654, + "grad_norm": 7.875, + "learning_rate": 4.923231591728485e-06, + "loss": 2.6214, + "step": 5046 + }, + { + "epoch": 0.5337351945854484, + "grad_norm": 7.875, + "learning_rate": 4.9231894724055285e-06, + "loss": 2.6786, + "step": 5047 + }, + { + "epoch": 0.5338409475465313, + "grad_norm": 8.0, + "learning_rate": 4.923147341711533e-06, + "loss": 2.4399, + "step": 5048 + }, + { + "epoch": 0.5339467005076142, + "grad_norm": 7.53125, + "learning_rate": 4.9231051996467e-06, + "loss": 2.5868, + "step": 5049 + }, + { + "epoch": 0.5340524534686971, + "grad_norm": 7.1875, + "learning_rate": 4.923063046211223e-06, + "loss": 2.5395, + "step": 5050 + }, + { + "epoch": 0.5341582064297801, + "grad_norm": 7.375, + "learning_rate": 4.9230208814053035e-06, + "loss": 2.6468, + "step": 5051 + }, + { + "epoch": 0.5342639593908629, + "grad_norm": 8.875, + "learning_rate": 4.922978705229138e-06, + "loss": 2.745, + "step": 5052 + }, + { + "epoch": 0.5343697123519459, + "grad_norm": 7.96875, + "learning_rate": 4.922936517682924e-06, + "loss": 2.6196, + "step": 5053 + }, + { + "epoch": 0.5344754653130288, + "grad_norm": 8.125, + "learning_rate": 4.92289431876686e-06, + "loss": 2.6479, + "step": 5054 + }, + { + "epoch": 0.5345812182741116, + "grad_norm": 7.78125, + "learning_rate": 4.922852108481145e-06, + "loss": 2.5548, + "step": 5055 + }, + { + "epoch": 0.5346869712351946, + "grad_norm": 9.8125, + "learning_rate": 4.922809886825974e-06, + "loss": 2.6506, + "step": 5056 + }, + { + "epoch": 0.5347927241962775, + "grad_norm": 7.5, + "learning_rate": 4.922767653801548e-06, + "loss": 2.401, + "step": 5057 + }, + { + "epoch": 0.5348984771573604, + "grad_norm": 8.1875, + "learning_rate": 4.922725409408064e-06, + "loss": 2.7556, + "step": 5058 + }, + { + "epoch": 0.5350042301184433, + "grad_norm": 7.8125, + "learning_rate": 4.922683153645721e-06, + "loss": 2.8617, + "step": 5059 + }, + { + "epoch": 0.5351099830795262, + "grad_norm": 7.84375, + "learning_rate": 4.922640886514716e-06, + "loss": 2.5822, + "step": 5060 + }, + { + "epoch": 0.5352157360406091, + "grad_norm": 7.625, + "learning_rate": 4.922598608015248e-06, + "loss": 2.6251, + "step": 5061 + }, + { + "epoch": 0.5353214890016921, + "grad_norm": 8.0625, + "learning_rate": 4.922556318147517e-06, + "loss": 2.5348, + "step": 5062 + }, + { + "epoch": 0.5354272419627749, + "grad_norm": 7.6875, + "learning_rate": 4.922514016911719e-06, + "loss": 2.559, + "step": 5063 + }, + { + "epoch": 0.5355329949238579, + "grad_norm": 7.6875, + "learning_rate": 4.922471704308054e-06, + "loss": 2.5252, + "step": 5064 + }, + { + "epoch": 0.5356387478849408, + "grad_norm": 8.3125, + "learning_rate": 4.92242938033672e-06, + "loss": 2.663, + "step": 5065 + }, + { + "epoch": 0.5357445008460237, + "grad_norm": 7.625, + "learning_rate": 4.9223870449979154e-06, + "loss": 2.5207, + "step": 5066 + }, + { + "epoch": 0.5358502538071066, + "grad_norm": 8.375, + "learning_rate": 4.92234469829184e-06, + "loss": 2.5961, + "step": 5067 + }, + { + "epoch": 0.5359560067681896, + "grad_norm": 8.125, + "learning_rate": 4.922302340218689e-06, + "loss": 2.77, + "step": 5068 + }, + { + "epoch": 0.5360617597292724, + "grad_norm": 7.875, + "learning_rate": 4.922259970778667e-06, + "loss": 2.7439, + "step": 5069 + }, + { + "epoch": 0.5361675126903553, + "grad_norm": 7.28125, + "learning_rate": 4.922217589971968e-06, + "loss": 2.5394, + "step": 5070 + }, + { + "epoch": 0.5362732656514383, + "grad_norm": 8.5625, + "learning_rate": 4.922175197798793e-06, + "loss": 2.3907, + "step": 5071 + }, + { + "epoch": 0.5363790186125211, + "grad_norm": 8.25, + "learning_rate": 4.92213279425934e-06, + "loss": 2.4488, + "step": 5072 + }, + { + "epoch": 0.5364847715736041, + "grad_norm": 8.0, + "learning_rate": 4.922090379353809e-06, + "loss": 2.6592, + "step": 5073 + }, + { + "epoch": 0.5365905245346869, + "grad_norm": 7.46875, + "learning_rate": 4.922047953082397e-06, + "loss": 2.6529, + "step": 5074 + }, + { + "epoch": 0.5366962774957699, + "grad_norm": 8.0625, + "learning_rate": 4.922005515445306e-06, + "loss": 2.5925, + "step": 5075 + }, + { + "epoch": 0.5368020304568528, + "grad_norm": 7.875, + "learning_rate": 4.921963066442733e-06, + "loss": 2.5584, + "step": 5076 + }, + { + "epoch": 0.5369077834179357, + "grad_norm": 7.5, + "learning_rate": 4.921920606074878e-06, + "loss": 2.3632, + "step": 5077 + }, + { + "epoch": 0.5370135363790186, + "grad_norm": 7.90625, + "learning_rate": 4.921878134341941e-06, + "loss": 2.6554, + "step": 5078 + }, + { + "epoch": 0.5371192893401016, + "grad_norm": 8.8125, + "learning_rate": 4.921835651244119e-06, + "loss": 2.6179, + "step": 5079 + }, + { + "epoch": 0.5372250423011844, + "grad_norm": 8.625, + "learning_rate": 4.921793156781613e-06, + "loss": 2.7932, + "step": 5080 + }, + { + "epoch": 0.5373307952622673, + "grad_norm": 7.6875, + "learning_rate": 4.9217506509546235e-06, + "loss": 2.5648, + "step": 5081 + }, + { + "epoch": 0.5374365482233503, + "grad_norm": 7.84375, + "learning_rate": 4.921708133763347e-06, + "loss": 2.5964, + "step": 5082 + }, + { + "epoch": 0.5375423011844331, + "grad_norm": 10.0625, + "learning_rate": 4.921665605207987e-06, + "loss": 2.5237, + "step": 5083 + }, + { + "epoch": 0.5376480541455161, + "grad_norm": 7.375, + "learning_rate": 4.921623065288739e-06, + "loss": 2.4983, + "step": 5084 + }, + { + "epoch": 0.537753807106599, + "grad_norm": 8.375, + "learning_rate": 4.9215805140058046e-06, + "loss": 2.7925, + "step": 5085 + }, + { + "epoch": 0.5378595600676819, + "grad_norm": 7.5, + "learning_rate": 4.921537951359383e-06, + "loss": 2.5712, + "step": 5086 + }, + { + "epoch": 0.5379653130287648, + "grad_norm": 7.96875, + "learning_rate": 4.9214953773496746e-06, + "loss": 2.5226, + "step": 5087 + }, + { + "epoch": 0.5380710659898477, + "grad_norm": 10.375, + "learning_rate": 4.921452791976879e-06, + "loss": 2.6206, + "step": 5088 + }, + { + "epoch": 0.5381768189509306, + "grad_norm": 7.6875, + "learning_rate": 4.921410195241195e-06, + "loss": 2.6795, + "step": 5089 + }, + { + "epoch": 0.5382825719120136, + "grad_norm": 7.46875, + "learning_rate": 4.921367587142824e-06, + "loss": 2.4311, + "step": 5090 + }, + { + "epoch": 0.5383883248730964, + "grad_norm": 7.625, + "learning_rate": 4.921324967681964e-06, + "loss": 2.5326, + "step": 5091 + }, + { + "epoch": 0.5384940778341794, + "grad_norm": 9.0, + "learning_rate": 4.921282336858817e-06, + "loss": 2.5109, + "step": 5092 + }, + { + "epoch": 0.5385998307952623, + "grad_norm": 12.8125, + "learning_rate": 4.921239694673582e-06, + "loss": 2.5739, + "step": 5093 + }, + { + "epoch": 0.5387055837563451, + "grad_norm": 7.65625, + "learning_rate": 4.9211970411264605e-06, + "loss": 2.477, + "step": 5094 + }, + { + "epoch": 0.5388113367174281, + "grad_norm": 8.25, + "learning_rate": 4.92115437621765e-06, + "loss": 2.5362, + "step": 5095 + }, + { + "epoch": 0.538917089678511, + "grad_norm": 7.9375, + "learning_rate": 4.921111699947353e-06, + "loss": 2.6631, + "step": 5096 + }, + { + "epoch": 0.5390228426395939, + "grad_norm": 8.0625, + "learning_rate": 4.92106901231577e-06, + "loss": 2.6954, + "step": 5097 + }, + { + "epoch": 0.5391285956006768, + "grad_norm": 7.84375, + "learning_rate": 4.921026313323098e-06, + "loss": 2.3296, + "step": 5098 + }, + { + "epoch": 0.5392343485617598, + "grad_norm": 7.28125, + "learning_rate": 4.9209836029695414e-06, + "loss": 2.421, + "step": 5099 + }, + { + "epoch": 0.5393401015228426, + "grad_norm": 7.53125, + "learning_rate": 4.920940881255298e-06, + "loss": 2.8048, + "step": 5100 + }, + { + "epoch": 0.5394458544839256, + "grad_norm": 7.625, + "learning_rate": 4.92089814818057e-06, + "loss": 2.4352, + "step": 5101 + }, + { + "epoch": 0.5395516074450084, + "grad_norm": 7.34375, + "learning_rate": 4.920855403745557e-06, + "loss": 2.447, + "step": 5102 + }, + { + "epoch": 0.5396573604060914, + "grad_norm": 7.625, + "learning_rate": 4.920812647950459e-06, + "loss": 2.6538, + "step": 5103 + }, + { + "epoch": 0.5397631133671743, + "grad_norm": 8.3125, + "learning_rate": 4.920769880795478e-06, + "loss": 2.7051, + "step": 5104 + }, + { + "epoch": 0.5398688663282571, + "grad_norm": 8.4375, + "learning_rate": 4.920727102280813e-06, + "loss": 2.7058, + "step": 5105 + }, + { + "epoch": 0.5399746192893401, + "grad_norm": 8.9375, + "learning_rate": 4.920684312406667e-06, + "loss": 2.4324, + "step": 5106 + }, + { + "epoch": 0.540080372250423, + "grad_norm": 8.0625, + "learning_rate": 4.920641511173239e-06, + "loss": 2.7229, + "step": 5107 + }, + { + "epoch": 0.5401861252115059, + "grad_norm": 8.375, + "learning_rate": 4.920598698580731e-06, + "loss": 2.652, + "step": 5108 + }, + { + "epoch": 0.5402918781725888, + "grad_norm": 8.3125, + "learning_rate": 4.920555874629343e-06, + "loss": 2.4913, + "step": 5109 + }, + { + "epoch": 0.5403976311336718, + "grad_norm": 7.75, + "learning_rate": 4.920513039319276e-06, + "loss": 2.6411, + "step": 5110 + }, + { + "epoch": 0.5405033840947546, + "grad_norm": 9.1875, + "learning_rate": 4.9204701926507305e-06, + "loss": 2.4788, + "step": 5111 + }, + { + "epoch": 0.5406091370558376, + "grad_norm": 8.4375, + "learning_rate": 4.92042733462391e-06, + "loss": 2.5176, + "step": 5112 + }, + { + "epoch": 0.5407148900169205, + "grad_norm": 7.46875, + "learning_rate": 4.920384465239012e-06, + "loss": 2.5508, + "step": 5113 + }, + { + "epoch": 0.5408206429780034, + "grad_norm": 7.375, + "learning_rate": 4.920341584496241e-06, + "loss": 2.778, + "step": 5114 + }, + { + "epoch": 0.5409263959390863, + "grad_norm": 9.875, + "learning_rate": 4.9202986923957965e-06, + "loss": 2.7104, + "step": 5115 + }, + { + "epoch": 0.5410321489001692, + "grad_norm": 7.6875, + "learning_rate": 4.92025578893788e-06, + "loss": 2.4808, + "step": 5116 + }, + { + "epoch": 0.5411379018612521, + "grad_norm": 7.5625, + "learning_rate": 4.9202128741226925e-06, + "loss": 2.4137, + "step": 5117 + }, + { + "epoch": 0.541243654822335, + "grad_norm": 7.78125, + "learning_rate": 4.9201699479504365e-06, + "loss": 2.5574, + "step": 5118 + }, + { + "epoch": 0.5413494077834179, + "grad_norm": 7.5, + "learning_rate": 4.9201270104213126e-06, + "loss": 2.5842, + "step": 5119 + }, + { + "epoch": 0.5414551607445008, + "grad_norm": 7.6875, + "learning_rate": 4.920084061535522e-06, + "loss": 2.6285, + "step": 5120 + }, + { + "epoch": 0.5415609137055838, + "grad_norm": 7.90625, + "learning_rate": 4.920041101293268e-06, + "loss": 2.6681, + "step": 5121 + }, + { + "epoch": 0.5416666666666666, + "grad_norm": 7.71875, + "learning_rate": 4.91999812969475e-06, + "loss": 2.6721, + "step": 5122 + }, + { + "epoch": 0.5417724196277496, + "grad_norm": 7.71875, + "learning_rate": 4.91995514674017e-06, + "loss": 2.7805, + "step": 5123 + }, + { + "epoch": 0.5418781725888325, + "grad_norm": 7.8125, + "learning_rate": 4.9199121524297305e-06, + "loss": 2.4732, + "step": 5124 + }, + { + "epoch": 0.5419839255499154, + "grad_norm": 7.78125, + "learning_rate": 4.919869146763633e-06, + "loss": 2.8802, + "step": 5125 + }, + { + "epoch": 0.5420896785109983, + "grad_norm": 7.71875, + "learning_rate": 4.91982612974208e-06, + "loss": 2.6242, + "step": 5126 + }, + { + "epoch": 0.5421954314720813, + "grad_norm": 7.6875, + "learning_rate": 4.919783101365272e-06, + "loss": 2.6122, + "step": 5127 + }, + { + "epoch": 0.5423011844331641, + "grad_norm": 7.90625, + "learning_rate": 4.919740061633411e-06, + "loss": 2.7632, + "step": 5128 + }, + { + "epoch": 0.5424069373942471, + "grad_norm": 7.75, + "learning_rate": 4.919697010546701e-06, + "loss": 2.7041, + "step": 5129 + }, + { + "epoch": 0.5425126903553299, + "grad_norm": 8.375, + "learning_rate": 4.919653948105342e-06, + "loss": 2.6956, + "step": 5130 + }, + { + "epoch": 0.5426184433164128, + "grad_norm": 8.4375, + "learning_rate": 4.919610874309536e-06, + "loss": 2.7133, + "step": 5131 + }, + { + "epoch": 0.5427241962774958, + "grad_norm": 7.84375, + "learning_rate": 4.919567789159486e-06, + "loss": 2.6403, + "step": 5132 + }, + { + "epoch": 0.5428299492385786, + "grad_norm": 7.53125, + "learning_rate": 4.919524692655394e-06, + "loss": 2.47, + "step": 5133 + }, + { + "epoch": 0.5429357021996616, + "grad_norm": 8.125, + "learning_rate": 4.919481584797462e-06, + "loss": 2.6202, + "step": 5134 + }, + { + "epoch": 0.5430414551607445, + "grad_norm": 8.375, + "learning_rate": 4.919438465585893e-06, + "loss": 2.4559, + "step": 5135 + }, + { + "epoch": 0.5431472081218274, + "grad_norm": 8.0, + "learning_rate": 4.919395335020888e-06, + "loss": 2.3763, + "step": 5136 + }, + { + "epoch": 0.5432529610829103, + "grad_norm": 9.875, + "learning_rate": 4.9193521931026515e-06, + "loss": 2.4328, + "step": 5137 + }, + { + "epoch": 0.5433587140439933, + "grad_norm": 8.375, + "learning_rate": 4.919309039831384e-06, + "loss": 2.5464, + "step": 5138 + }, + { + "epoch": 0.5434644670050761, + "grad_norm": 8.3125, + "learning_rate": 4.919265875207287e-06, + "loss": 2.3824, + "step": 5139 + }, + { + "epoch": 0.5435702199661591, + "grad_norm": 7.625, + "learning_rate": 4.919222699230567e-06, + "loss": 2.5934, + "step": 5140 + }, + { + "epoch": 0.543675972927242, + "grad_norm": 8.25, + "learning_rate": 4.919179511901424e-06, + "loss": 2.6174, + "step": 5141 + }, + { + "epoch": 0.5437817258883249, + "grad_norm": 7.875, + "learning_rate": 4.91913631322006e-06, + "loss": 2.6408, + "step": 5142 + }, + { + "epoch": 0.5438874788494078, + "grad_norm": 8.1875, + "learning_rate": 4.919093103186679e-06, + "loss": 2.5539, + "step": 5143 + }, + { + "epoch": 0.5439932318104906, + "grad_norm": 7.78125, + "learning_rate": 4.919049881801483e-06, + "loss": 2.6398, + "step": 5144 + }, + { + "epoch": 0.5440989847715736, + "grad_norm": 8.375, + "learning_rate": 4.919006649064676e-06, + "loss": 2.5223, + "step": 5145 + }, + { + "epoch": 0.5442047377326565, + "grad_norm": 7.65625, + "learning_rate": 4.91896340497646e-06, + "loss": 2.6046, + "step": 5146 + }, + { + "epoch": 0.5443104906937394, + "grad_norm": 7.90625, + "learning_rate": 4.918920149537038e-06, + "loss": 2.5493, + "step": 5147 + }, + { + "epoch": 0.5444162436548223, + "grad_norm": 7.65625, + "learning_rate": 4.918876882746613e-06, + "loss": 2.5632, + "step": 5148 + }, + { + "epoch": 0.5445219966159053, + "grad_norm": 10.4375, + "learning_rate": 4.918833604605389e-06, + "loss": 2.5955, + "step": 5149 + }, + { + "epoch": 0.5446277495769881, + "grad_norm": 7.84375, + "learning_rate": 4.918790315113566e-06, + "loss": 2.5132, + "step": 5150 + }, + { + "epoch": 0.5447335025380711, + "grad_norm": 7.8125, + "learning_rate": 4.918747014271351e-06, + "loss": 2.5247, + "step": 5151 + }, + { + "epoch": 0.544839255499154, + "grad_norm": 8.1875, + "learning_rate": 4.918703702078945e-06, + "loss": 2.5156, + "step": 5152 + }, + { + "epoch": 0.5449450084602369, + "grad_norm": 8.25, + "learning_rate": 4.918660378536551e-06, + "loss": 2.5394, + "step": 5153 + }, + { + "epoch": 0.5450507614213198, + "grad_norm": 8.1875, + "learning_rate": 4.918617043644375e-06, + "loss": 2.4901, + "step": 5154 + }, + { + "epoch": 0.5451565143824028, + "grad_norm": 8.1875, + "learning_rate": 4.918573697402617e-06, + "loss": 2.5985, + "step": 5155 + }, + { + "epoch": 0.5452622673434856, + "grad_norm": 7.75, + "learning_rate": 4.918530339811483e-06, + "loss": 2.7433, + "step": 5156 + }, + { + "epoch": 0.5453680203045685, + "grad_norm": 7.75, + "learning_rate": 4.918486970871174e-06, + "loss": 2.4906, + "step": 5157 + }, + { + "epoch": 0.5454737732656514, + "grad_norm": 8.9375, + "learning_rate": 4.918443590581895e-06, + "loss": 2.7887, + "step": 5158 + }, + { + "epoch": 0.5455795262267343, + "grad_norm": 8.5625, + "learning_rate": 4.918400198943849e-06, + "loss": 2.4424, + "step": 5159 + }, + { + "epoch": 0.5456852791878173, + "grad_norm": 8.75, + "learning_rate": 4.918356795957241e-06, + "loss": 2.8281, + "step": 5160 + }, + { + "epoch": 0.5457910321489001, + "grad_norm": 8.375, + "learning_rate": 4.918313381622272e-06, + "loss": 2.5136, + "step": 5161 + }, + { + "epoch": 0.5458967851099831, + "grad_norm": 7.5, + "learning_rate": 4.918269955939149e-06, + "loss": 2.5887, + "step": 5162 + }, + { + "epoch": 0.546002538071066, + "grad_norm": 8.1875, + "learning_rate": 4.918226518908073e-06, + "loss": 2.6992, + "step": 5163 + }, + { + "epoch": 0.5461082910321489, + "grad_norm": 8.4375, + "learning_rate": 4.91818307052925e-06, + "loss": 2.8157, + "step": 5164 + }, + { + "epoch": 0.5462140439932318, + "grad_norm": 7.5625, + "learning_rate": 4.918139610802882e-06, + "loss": 2.6778, + "step": 5165 + }, + { + "epoch": 0.5463197969543148, + "grad_norm": 7.9375, + "learning_rate": 4.9180961397291734e-06, + "loss": 2.5804, + "step": 5166 + }, + { + "epoch": 0.5464255499153976, + "grad_norm": 7.75, + "learning_rate": 4.91805265730833e-06, + "loss": 2.5894, + "step": 5167 + }, + { + "epoch": 0.5465313028764806, + "grad_norm": 7.96875, + "learning_rate": 4.918009163540553e-06, + "loss": 2.7422, + "step": 5168 + }, + { + "epoch": 0.5466370558375635, + "grad_norm": 8.1875, + "learning_rate": 4.917965658426049e-06, + "loss": 2.3217, + "step": 5169 + }, + { + "epoch": 0.5467428087986463, + "grad_norm": 7.90625, + "learning_rate": 4.917922141965022e-06, + "loss": 2.5824, + "step": 5170 + }, + { + "epoch": 0.5468485617597293, + "grad_norm": 8.4375, + "learning_rate": 4.917878614157673e-06, + "loss": 2.5357, + "step": 5171 + }, + { + "epoch": 0.5469543147208121, + "grad_norm": 7.8125, + "learning_rate": 4.91783507500421e-06, + "loss": 2.5214, + "step": 5172 + }, + { + "epoch": 0.5470600676818951, + "grad_norm": 7.40625, + "learning_rate": 4.917791524504835e-06, + "loss": 2.5351, + "step": 5173 + }, + { + "epoch": 0.547165820642978, + "grad_norm": 8.125, + "learning_rate": 4.917747962659754e-06, + "loss": 2.5834, + "step": 5174 + }, + { + "epoch": 0.5472715736040609, + "grad_norm": 7.9375, + "learning_rate": 4.91770438946917e-06, + "loss": 2.5235, + "step": 5175 + }, + { + "epoch": 0.5473773265651438, + "grad_norm": 9.0625, + "learning_rate": 4.917660804933289e-06, + "loss": 2.2762, + "step": 5176 + }, + { + "epoch": 0.5474830795262268, + "grad_norm": 11.375, + "learning_rate": 4.917617209052313e-06, + "loss": 2.6889, + "step": 5177 + }, + { + "epoch": 0.5475888324873096, + "grad_norm": 7.65625, + "learning_rate": 4.91757360182645e-06, + "loss": 2.5057, + "step": 5178 + }, + { + "epoch": 0.5476945854483926, + "grad_norm": 8.125, + "learning_rate": 4.917529983255902e-06, + "loss": 2.8326, + "step": 5179 + }, + { + "epoch": 0.5478003384094755, + "grad_norm": 8.25, + "learning_rate": 4.917486353340875e-06, + "loss": 2.5278, + "step": 5180 + }, + { + "epoch": 0.5479060913705583, + "grad_norm": 8.25, + "learning_rate": 4.917442712081572e-06, + "loss": 2.6759, + "step": 5181 + }, + { + "epoch": 0.5480118443316413, + "grad_norm": 7.75, + "learning_rate": 4.9173990594782e-06, + "loss": 2.6566, + "step": 5182 + }, + { + "epoch": 0.5481175972927242, + "grad_norm": 8.3125, + "learning_rate": 4.917355395530962e-06, + "loss": 2.9414, + "step": 5183 + }, + { + "epoch": 0.5482233502538071, + "grad_norm": 7.75, + "learning_rate": 4.917311720240066e-06, + "loss": 2.5694, + "step": 5184 + }, + { + "epoch": 0.54832910321489, + "grad_norm": 7.78125, + "learning_rate": 4.917268033605712e-06, + "loss": 2.7994, + "step": 5185 + }, + { + "epoch": 0.5484348561759729, + "grad_norm": 7.71875, + "learning_rate": 4.9172243356281105e-06, + "loss": 2.5255, + "step": 5186 + }, + { + "epoch": 0.5485406091370558, + "grad_norm": 7.71875, + "learning_rate": 4.917180626307462e-06, + "loss": 2.6979, + "step": 5187 + }, + { + "epoch": 0.5486463620981388, + "grad_norm": 8.25, + "learning_rate": 4.917136905643973e-06, + "loss": 2.6099, + "step": 5188 + }, + { + "epoch": 0.5487521150592216, + "grad_norm": 7.5625, + "learning_rate": 4.91709317363785e-06, + "loss": 2.6757, + "step": 5189 + }, + { + "epoch": 0.5488578680203046, + "grad_norm": 7.65625, + "learning_rate": 4.917049430289298e-06, + "loss": 2.6866, + "step": 5190 + }, + { + "epoch": 0.5489636209813875, + "grad_norm": 7.875, + "learning_rate": 4.91700567559852e-06, + "loss": 2.7708, + "step": 5191 + }, + { + "epoch": 0.5490693739424704, + "grad_norm": 7.4375, + "learning_rate": 4.916961909565724e-06, + "loss": 2.7411, + "step": 5192 + }, + { + "epoch": 0.5491751269035533, + "grad_norm": 7.5625, + "learning_rate": 4.916918132191114e-06, + "loss": 2.5447, + "step": 5193 + }, + { + "epoch": 0.5492808798646363, + "grad_norm": 7.6875, + "learning_rate": 4.916874343474896e-06, + "loss": 2.8166, + "step": 5194 + }, + { + "epoch": 0.5493866328257191, + "grad_norm": 7.6875, + "learning_rate": 4.916830543417274e-06, + "loss": 2.4544, + "step": 5195 + }, + { + "epoch": 0.549492385786802, + "grad_norm": 7.71875, + "learning_rate": 4.916786732018456e-06, + "loss": 2.5637, + "step": 5196 + }, + { + "epoch": 0.549598138747885, + "grad_norm": 8.0625, + "learning_rate": 4.9167429092786454e-06, + "loss": 2.5456, + "step": 5197 + }, + { + "epoch": 0.5497038917089678, + "grad_norm": 7.78125, + "learning_rate": 4.916699075198049e-06, + "loss": 2.5253, + "step": 5198 + }, + { + "epoch": 0.5498096446700508, + "grad_norm": 7.59375, + "learning_rate": 4.916655229776872e-06, + "loss": 2.5198, + "step": 5199 + }, + { + "epoch": 0.5499153976311336, + "grad_norm": 7.625, + "learning_rate": 4.916611373015321e-06, + "loss": 2.4157, + "step": 5200 + }, + { + "epoch": 0.5500211505922166, + "grad_norm": 8.0, + "learning_rate": 4.916567504913601e-06, + "loss": 2.8177, + "step": 5201 + }, + { + "epoch": 0.5501269035532995, + "grad_norm": 8.25, + "learning_rate": 4.9165236254719175e-06, + "loss": 2.5743, + "step": 5202 + }, + { + "epoch": 0.5502326565143824, + "grad_norm": 7.78125, + "learning_rate": 4.916479734690478e-06, + "loss": 2.7505, + "step": 5203 + }, + { + "epoch": 0.5503384094754653, + "grad_norm": 7.96875, + "learning_rate": 4.916435832569486e-06, + "loss": 2.5153, + "step": 5204 + }, + { + "epoch": 0.5504441624365483, + "grad_norm": 8.0, + "learning_rate": 4.9163919191091504e-06, + "loss": 2.5997, + "step": 5205 + }, + { + "epoch": 0.5505499153976311, + "grad_norm": 8.0625, + "learning_rate": 4.9163479943096745e-06, + "loss": 2.7615, + "step": 5206 + }, + { + "epoch": 0.550655668358714, + "grad_norm": 8.0625, + "learning_rate": 4.916304058171267e-06, + "loss": 2.6369, + "step": 5207 + }, + { + "epoch": 0.550761421319797, + "grad_norm": 7.65625, + "learning_rate": 4.916260110694132e-06, + "loss": 2.6436, + "step": 5208 + }, + { + "epoch": 0.5508671742808798, + "grad_norm": 7.75, + "learning_rate": 4.916216151878477e-06, + "loss": 2.4659, + "step": 5209 + }, + { + "epoch": 0.5509729272419628, + "grad_norm": 7.84375, + "learning_rate": 4.9161721817245065e-06, + "loss": 2.6686, + "step": 5210 + }, + { + "epoch": 0.5510786802030457, + "grad_norm": 8.0, + "learning_rate": 4.916128200232429e-06, + "loss": 2.4457, + "step": 5211 + }, + { + "epoch": 0.5511844331641286, + "grad_norm": 7.9375, + "learning_rate": 4.916084207402451e-06, + "loss": 2.7335, + "step": 5212 + }, + { + "epoch": 0.5512901861252115, + "grad_norm": 7.28125, + "learning_rate": 4.916040203234777e-06, + "loss": 2.6255, + "step": 5213 + }, + { + "epoch": 0.5513959390862944, + "grad_norm": 8.375, + "learning_rate": 4.9159961877296146e-06, + "loss": 2.6526, + "step": 5214 + }, + { + "epoch": 0.5515016920473773, + "grad_norm": 7.625, + "learning_rate": 4.91595216088717e-06, + "loss": 2.4713, + "step": 5215 + }, + { + "epoch": 0.5516074450084603, + "grad_norm": 7.46875, + "learning_rate": 4.91590812270765e-06, + "loss": 2.5674, + "step": 5216 + }, + { + "epoch": 0.5517131979695431, + "grad_norm": 8.75, + "learning_rate": 4.915864073191261e-06, + "loss": 2.5477, + "step": 5217 + }, + { + "epoch": 0.551818950930626, + "grad_norm": 8.3125, + "learning_rate": 4.915820012338211e-06, + "loss": 2.7116, + "step": 5218 + }, + { + "epoch": 0.551924703891709, + "grad_norm": 8.4375, + "learning_rate": 4.915775940148705e-06, + "loss": 2.7046, + "step": 5219 + }, + { + "epoch": 0.5520304568527918, + "grad_norm": 7.53125, + "learning_rate": 4.91573185662295e-06, + "loss": 2.6225, + "step": 5220 + }, + { + "epoch": 0.5521362098138748, + "grad_norm": 7.875, + "learning_rate": 4.915687761761154e-06, + "loss": 2.5265, + "step": 5221 + }, + { + "epoch": 0.5522419627749577, + "grad_norm": 7.84375, + "learning_rate": 4.915643655563523e-06, + "loss": 2.4814, + "step": 5222 + }, + { + "epoch": 0.5523477157360406, + "grad_norm": 7.96875, + "learning_rate": 4.915599538030264e-06, + "loss": 2.5795, + "step": 5223 + }, + { + "epoch": 0.5524534686971235, + "grad_norm": 8.1875, + "learning_rate": 4.915555409161586e-06, + "loss": 2.492, + "step": 5224 + }, + { + "epoch": 0.5525592216582065, + "grad_norm": 7.84375, + "learning_rate": 4.915511268957693e-06, + "loss": 2.6652, + "step": 5225 + }, + { + "epoch": 0.5526649746192893, + "grad_norm": 9.5, + "learning_rate": 4.915467117418794e-06, + "loss": 2.724, + "step": 5226 + }, + { + "epoch": 0.5527707275803723, + "grad_norm": 8.875, + "learning_rate": 4.915422954545096e-06, + "loss": 2.8142, + "step": 5227 + }, + { + "epoch": 0.5528764805414551, + "grad_norm": 8.75, + "learning_rate": 4.915378780336805e-06, + "loss": 2.7564, + "step": 5228 + }, + { + "epoch": 0.5529822335025381, + "grad_norm": 7.625, + "learning_rate": 4.91533459479413e-06, + "loss": 2.6421, + "step": 5229 + }, + { + "epoch": 0.553087986463621, + "grad_norm": 7.59375, + "learning_rate": 4.915290397917277e-06, + "loss": 2.6185, + "step": 5230 + }, + { + "epoch": 0.5531937394247038, + "grad_norm": 7.75, + "learning_rate": 4.915246189706454e-06, + "loss": 2.7605, + "step": 5231 + }, + { + "epoch": 0.5532994923857868, + "grad_norm": 7.4375, + "learning_rate": 4.915201970161869e-06, + "loss": 2.7009, + "step": 5232 + }, + { + "epoch": 0.5534052453468697, + "grad_norm": 9.4375, + "learning_rate": 4.915157739283728e-06, + "loss": 2.6974, + "step": 5233 + }, + { + "epoch": 0.5535109983079526, + "grad_norm": 8.0, + "learning_rate": 4.91511349707224e-06, + "loss": 2.468, + "step": 5234 + }, + { + "epoch": 0.5536167512690355, + "grad_norm": 7.59375, + "learning_rate": 4.915069243527613e-06, + "loss": 2.613, + "step": 5235 + }, + { + "epoch": 0.5537225042301185, + "grad_norm": 7.8125, + "learning_rate": 4.915024978650052e-06, + "loss": 2.7784, + "step": 5236 + }, + { + "epoch": 0.5538282571912013, + "grad_norm": 7.9375, + "learning_rate": 4.9149807024397685e-06, + "loss": 2.9141, + "step": 5237 + }, + { + "epoch": 0.5539340101522843, + "grad_norm": 7.75, + "learning_rate": 4.914936414896967e-06, + "loss": 2.5337, + "step": 5238 + }, + { + "epoch": 0.5540397631133672, + "grad_norm": 7.8125, + "learning_rate": 4.914892116021857e-06, + "loss": 2.5541, + "step": 5239 + }, + { + "epoch": 0.5541455160744501, + "grad_norm": 8.125, + "learning_rate": 4.914847805814645e-06, + "loss": 2.674, + "step": 5240 + }, + { + "epoch": 0.554251269035533, + "grad_norm": 8.6875, + "learning_rate": 4.914803484275542e-06, + "loss": 2.8883, + "step": 5241 + }, + { + "epoch": 0.5543570219966159, + "grad_norm": 8.375, + "learning_rate": 4.914759151404752e-06, + "loss": 2.6121, + "step": 5242 + }, + { + "epoch": 0.5544627749576988, + "grad_norm": 7.6875, + "learning_rate": 4.9147148072024855e-06, + "loss": 2.3151, + "step": 5243 + }, + { + "epoch": 0.5545685279187818, + "grad_norm": 7.5625, + "learning_rate": 4.91467045166895e-06, + "loss": 2.5579, + "step": 5244 + }, + { + "epoch": 0.5546742808798646, + "grad_norm": 7.75, + "learning_rate": 4.914626084804354e-06, + "loss": 2.6105, + "step": 5245 + }, + { + "epoch": 0.5547800338409475, + "grad_norm": 8.5625, + "learning_rate": 4.914581706608904e-06, + "loss": 2.6188, + "step": 5246 + }, + { + "epoch": 0.5548857868020305, + "grad_norm": 7.96875, + "learning_rate": 4.914537317082811e-06, + "loss": 2.5709, + "step": 5247 + }, + { + "epoch": 0.5549915397631133, + "grad_norm": 7.6875, + "learning_rate": 4.914492916226281e-06, + "loss": 2.6733, + "step": 5248 + }, + { + "epoch": 0.5550972927241963, + "grad_norm": 7.65625, + "learning_rate": 4.914448504039524e-06, + "loss": 2.608, + "step": 5249 + }, + { + "epoch": 0.5552030456852792, + "grad_norm": 7.96875, + "learning_rate": 4.9144040805227475e-06, + "loss": 2.6036, + "step": 5250 + }, + { + "epoch": 0.5553087986463621, + "grad_norm": 8.0625, + "learning_rate": 4.914359645676159e-06, + "loss": 2.6174, + "step": 5251 + }, + { + "epoch": 0.555414551607445, + "grad_norm": 7.78125, + "learning_rate": 4.914315199499969e-06, + "loss": 2.6813, + "step": 5252 + }, + { + "epoch": 0.555520304568528, + "grad_norm": 7.5625, + "learning_rate": 4.9142707419943855e-06, + "loss": 2.5198, + "step": 5253 + }, + { + "epoch": 0.5556260575296108, + "grad_norm": 7.9375, + "learning_rate": 4.914226273159617e-06, + "loss": 2.6423, + "step": 5254 + }, + { + "epoch": 0.5557318104906938, + "grad_norm": 8.0625, + "learning_rate": 4.914181792995871e-06, + "loss": 2.6402, + "step": 5255 + }, + { + "epoch": 0.5558375634517766, + "grad_norm": 7.90625, + "learning_rate": 4.9141373015033576e-06, + "loss": 2.6987, + "step": 5256 + }, + { + "epoch": 0.5559433164128595, + "grad_norm": 7.96875, + "learning_rate": 4.9140927986822855e-06, + "loss": 2.4651, + "step": 5257 + }, + { + "epoch": 0.5560490693739425, + "grad_norm": 8.125, + "learning_rate": 4.914048284532863e-06, + "loss": 2.5675, + "step": 5258 + }, + { + "epoch": 0.5561548223350253, + "grad_norm": 7.8125, + "learning_rate": 4.914003759055299e-06, + "loss": 2.6215, + "step": 5259 + }, + { + "epoch": 0.5562605752961083, + "grad_norm": 7.46875, + "learning_rate": 4.913959222249803e-06, + "loss": 2.5013, + "step": 5260 + }, + { + "epoch": 0.5563663282571912, + "grad_norm": 8.6875, + "learning_rate": 4.913914674116583e-06, + "loss": 2.6312, + "step": 5261 + }, + { + "epoch": 0.5564720812182741, + "grad_norm": 7.75, + "learning_rate": 4.9138701146558494e-06, + "loss": 2.6449, + "step": 5262 + }, + { + "epoch": 0.556577834179357, + "grad_norm": 8.4375, + "learning_rate": 4.913825543867811e-06, + "loss": 2.717, + "step": 5263 + }, + { + "epoch": 0.55668358714044, + "grad_norm": 7.78125, + "learning_rate": 4.913780961752675e-06, + "loss": 2.4976, + "step": 5264 + }, + { + "epoch": 0.5567893401015228, + "grad_norm": 7.375, + "learning_rate": 4.913736368310654e-06, + "loss": 2.6745, + "step": 5265 + }, + { + "epoch": 0.5568950930626058, + "grad_norm": 7.625, + "learning_rate": 4.9136917635419545e-06, + "loss": 2.516, + "step": 5266 + }, + { + "epoch": 0.5570008460236887, + "grad_norm": 8.1875, + "learning_rate": 4.913647147446787e-06, + "loss": 2.4915, + "step": 5267 + }, + { + "epoch": 0.5571065989847716, + "grad_norm": 8.5, + "learning_rate": 4.91360252002536e-06, + "loss": 3.0283, + "step": 5268 + }, + { + "epoch": 0.5572123519458545, + "grad_norm": 8.0625, + "learning_rate": 4.913557881277884e-06, + "loss": 2.7941, + "step": 5269 + }, + { + "epoch": 0.5573181049069373, + "grad_norm": 7.6875, + "learning_rate": 4.913513231204568e-06, + "loss": 2.6277, + "step": 5270 + }, + { + "epoch": 0.5574238578680203, + "grad_norm": 7.9375, + "learning_rate": 4.913468569805621e-06, + "loss": 2.4437, + "step": 5271 + }, + { + "epoch": 0.5575296108291032, + "grad_norm": 8.125, + "learning_rate": 4.913423897081253e-06, + "loss": 2.4406, + "step": 5272 + }, + { + "epoch": 0.5576353637901861, + "grad_norm": 7.9375, + "learning_rate": 4.913379213031674e-06, + "loss": 2.7245, + "step": 5273 + }, + { + "epoch": 0.557741116751269, + "grad_norm": 8.0, + "learning_rate": 4.913334517657093e-06, + "loss": 2.2505, + "step": 5274 + }, + { + "epoch": 0.557846869712352, + "grad_norm": 7.78125, + "learning_rate": 4.913289810957721e-06, + "loss": 2.5171, + "step": 5275 + }, + { + "epoch": 0.5579526226734348, + "grad_norm": 7.6875, + "learning_rate": 4.913245092933767e-06, + "loss": 2.6453, + "step": 5276 + }, + { + "epoch": 0.5580583756345178, + "grad_norm": 8.5, + "learning_rate": 4.91320036358544e-06, + "loss": 2.4804, + "step": 5277 + }, + { + "epoch": 0.5581641285956007, + "grad_norm": 7.71875, + "learning_rate": 4.913155622912951e-06, + "loss": 2.5625, + "step": 5278 + }, + { + "epoch": 0.5582698815566836, + "grad_norm": 7.78125, + "learning_rate": 4.91311087091651e-06, + "loss": 2.6, + "step": 5279 + }, + { + "epoch": 0.5583756345177665, + "grad_norm": 7.5625, + "learning_rate": 4.913066107596326e-06, + "loss": 2.5237, + "step": 5280 + }, + { + "epoch": 0.5584813874788495, + "grad_norm": 8.3125, + "learning_rate": 4.91302133295261e-06, + "loss": 2.7434, + "step": 5281 + }, + { + "epoch": 0.5585871404399323, + "grad_norm": 7.84375, + "learning_rate": 4.912976546985571e-06, + "loss": 2.694, + "step": 5282 + }, + { + "epoch": 0.5586928934010152, + "grad_norm": 8.375, + "learning_rate": 4.912931749695421e-06, + "loss": 2.9164, + "step": 5283 + }, + { + "epoch": 0.5587986463620981, + "grad_norm": 7.4375, + "learning_rate": 4.912886941082369e-06, + "loss": 2.6206, + "step": 5284 + }, + { + "epoch": 0.558904399323181, + "grad_norm": 8.0, + "learning_rate": 4.912842121146625e-06, + "loss": 2.6467, + "step": 5285 + }, + { + "epoch": 0.559010152284264, + "grad_norm": 8.0, + "learning_rate": 4.912797289888399e-06, + "loss": 2.283, + "step": 5286 + }, + { + "epoch": 0.5591159052453468, + "grad_norm": 9.25, + "learning_rate": 4.912752447307904e-06, + "loss": 2.5271, + "step": 5287 + }, + { + "epoch": 0.5592216582064298, + "grad_norm": 8.875, + "learning_rate": 4.912707593405347e-06, + "loss": 2.5999, + "step": 5288 + }, + { + "epoch": 0.5593274111675127, + "grad_norm": 8.0625, + "learning_rate": 4.9126627281809405e-06, + "loss": 2.4527, + "step": 5289 + }, + { + "epoch": 0.5594331641285956, + "grad_norm": 8.375, + "learning_rate": 4.912617851634894e-06, + "loss": 2.3668, + "step": 5290 + }, + { + "epoch": 0.5595389170896785, + "grad_norm": 8.625, + "learning_rate": 4.91257296376742e-06, + "loss": 2.7163, + "step": 5291 + }, + { + "epoch": 0.5596446700507615, + "grad_norm": 8.375, + "learning_rate": 4.912528064578727e-06, + "loss": 2.6008, + "step": 5292 + }, + { + "epoch": 0.5597504230118443, + "grad_norm": 7.6875, + "learning_rate": 4.912483154069027e-06, + "loss": 2.6077, + "step": 5293 + }, + { + "epoch": 0.5598561759729273, + "grad_norm": 8.0625, + "learning_rate": 4.912438232238529e-06, + "loss": 2.6082, + "step": 5294 + }, + { + "epoch": 0.5599619289340102, + "grad_norm": 7.8125, + "learning_rate": 4.912393299087445e-06, + "loss": 2.5692, + "step": 5295 + }, + { + "epoch": 0.560067681895093, + "grad_norm": 7.96875, + "learning_rate": 4.912348354615988e-06, + "loss": 2.12, + "step": 5296 + }, + { + "epoch": 0.560173434856176, + "grad_norm": 7.96875, + "learning_rate": 4.912303398824365e-06, + "loss": 2.6756, + "step": 5297 + }, + { + "epoch": 0.5602791878172588, + "grad_norm": 7.9375, + "learning_rate": 4.912258431712789e-06, + "loss": 2.5982, + "step": 5298 + }, + { + "epoch": 0.5603849407783418, + "grad_norm": 8.625, + "learning_rate": 4.912213453281471e-06, + "loss": 2.5522, + "step": 5299 + }, + { + "epoch": 0.5604906937394247, + "grad_norm": 7.4375, + "learning_rate": 4.912168463530622e-06, + "loss": 2.586, + "step": 5300 + }, + { + "epoch": 0.5605964467005076, + "grad_norm": 7.71875, + "learning_rate": 4.912123462460453e-06, + "loss": 2.6361, + "step": 5301 + }, + { + "epoch": 0.5607021996615905, + "grad_norm": 7.96875, + "learning_rate": 4.912078450071175e-06, + "loss": 2.5041, + "step": 5302 + }, + { + "epoch": 0.5608079526226735, + "grad_norm": 7.53125, + "learning_rate": 4.912033426362999e-06, + "loss": 2.3324, + "step": 5303 + }, + { + "epoch": 0.5609137055837563, + "grad_norm": 8.0625, + "learning_rate": 4.911988391336136e-06, + "loss": 2.5563, + "step": 5304 + }, + { + "epoch": 0.5610194585448393, + "grad_norm": 7.78125, + "learning_rate": 4.911943344990799e-06, + "loss": 2.4396, + "step": 5305 + }, + { + "epoch": 0.5611252115059222, + "grad_norm": 7.53125, + "learning_rate": 4.911898287327198e-06, + "loss": 2.5481, + "step": 5306 + }, + { + "epoch": 0.561230964467005, + "grad_norm": 7.875, + "learning_rate": 4.911853218345545e-06, + "loss": 2.6827, + "step": 5307 + }, + { + "epoch": 0.561336717428088, + "grad_norm": 8.375, + "learning_rate": 4.911808138046051e-06, + "loss": 2.4102, + "step": 5308 + }, + { + "epoch": 0.561442470389171, + "grad_norm": 8.1875, + "learning_rate": 4.911763046428928e-06, + "loss": 2.5073, + "step": 5309 + }, + { + "epoch": 0.5615482233502538, + "grad_norm": 7.75, + "learning_rate": 4.911717943494387e-06, + "loss": 2.6599, + "step": 5310 + }, + { + "epoch": 0.5616539763113367, + "grad_norm": 8.375, + "learning_rate": 4.9116728292426406e-06, + "loss": 2.5193, + "step": 5311 + }, + { + "epoch": 0.5617597292724196, + "grad_norm": 8.6875, + "learning_rate": 4.9116277036739e-06, + "loss": 2.4966, + "step": 5312 + }, + { + "epoch": 0.5618654822335025, + "grad_norm": 7.875, + "learning_rate": 4.911582566788376e-06, + "loss": 2.5753, + "step": 5313 + }, + { + "epoch": 0.5619712351945855, + "grad_norm": 7.9375, + "learning_rate": 4.911537418586283e-06, + "loss": 2.5267, + "step": 5314 + }, + { + "epoch": 0.5620769881556683, + "grad_norm": 7.875, + "learning_rate": 4.91149225906783e-06, + "loss": 2.6854, + "step": 5315 + }, + { + "epoch": 0.5621827411167513, + "grad_norm": 8.1875, + "learning_rate": 4.9114470882332295e-06, + "loss": 2.5528, + "step": 5316 + }, + { + "epoch": 0.5622884940778342, + "grad_norm": 8.3125, + "learning_rate": 4.911401906082696e-06, + "loss": 2.6907, + "step": 5317 + }, + { + "epoch": 0.562394247038917, + "grad_norm": 7.90625, + "learning_rate": 4.911356712616438e-06, + "loss": 2.4936, + "step": 5318 + }, + { + "epoch": 0.5625, + "grad_norm": 7.65625, + "learning_rate": 4.91131150783467e-06, + "loss": 2.3613, + "step": 5319 + }, + { + "epoch": 0.562605752961083, + "grad_norm": 8.125, + "learning_rate": 4.911266291737602e-06, + "loss": 2.4856, + "step": 5320 + }, + { + "epoch": 0.5627115059221658, + "grad_norm": 8.375, + "learning_rate": 4.9112210643254485e-06, + "loss": 2.6799, + "step": 5321 + }, + { + "epoch": 0.5628172588832487, + "grad_norm": 7.8125, + "learning_rate": 4.911175825598421e-06, + "loss": 2.5954, + "step": 5322 + }, + { + "epoch": 0.5629230118443317, + "grad_norm": 9.0625, + "learning_rate": 4.911130575556731e-06, + "loss": 2.5655, + "step": 5323 + }, + { + "epoch": 0.5630287648054145, + "grad_norm": 8.0625, + "learning_rate": 4.911085314200591e-06, + "loss": 2.6181, + "step": 5324 + }, + { + "epoch": 0.5631345177664975, + "grad_norm": 7.96875, + "learning_rate": 4.911040041530214e-06, + "loss": 2.7948, + "step": 5325 + }, + { + "epoch": 0.5632402707275804, + "grad_norm": 10.375, + "learning_rate": 4.910994757545812e-06, + "loss": 2.7372, + "step": 5326 + }, + { + "epoch": 0.5633460236886633, + "grad_norm": 8.125, + "learning_rate": 4.910949462247599e-06, + "loss": 2.5361, + "step": 5327 + }, + { + "epoch": 0.5634517766497462, + "grad_norm": 8.875, + "learning_rate": 4.9109041556357836e-06, + "loss": 2.611, + "step": 5328 + }, + { + "epoch": 0.563557529610829, + "grad_norm": 7.9375, + "learning_rate": 4.910858837710582e-06, + "loss": 2.6528, + "step": 5329 + }, + { + "epoch": 0.563663282571912, + "grad_norm": 7.71875, + "learning_rate": 4.910813508472207e-06, + "loss": 2.5189, + "step": 5330 + }, + { + "epoch": 0.563769035532995, + "grad_norm": 7.53125, + "learning_rate": 4.9107681679208695e-06, + "loss": 2.6044, + "step": 5331 + }, + { + "epoch": 0.5638747884940778, + "grad_norm": 7.75, + "learning_rate": 4.910722816056783e-06, + "loss": 2.4644, + "step": 5332 + }, + { + "epoch": 0.5639805414551607, + "grad_norm": 8.25, + "learning_rate": 4.910677452880161e-06, + "loss": 2.7622, + "step": 5333 + }, + { + "epoch": 0.5640862944162437, + "grad_norm": 7.65625, + "learning_rate": 4.910632078391215e-06, + "loss": 2.7145, + "step": 5334 + }, + { + "epoch": 0.5641920473773265, + "grad_norm": 7.90625, + "learning_rate": 4.910586692590158e-06, + "loss": 2.4751, + "step": 5335 + }, + { + "epoch": 0.5642978003384095, + "grad_norm": 8.25, + "learning_rate": 4.910541295477205e-06, + "loss": 2.7743, + "step": 5336 + }, + { + "epoch": 0.5644035532994924, + "grad_norm": 8.125, + "learning_rate": 4.910495887052566e-06, + "loss": 2.3902, + "step": 5337 + }, + { + "epoch": 0.5645093062605753, + "grad_norm": 8.4375, + "learning_rate": 4.910450467316456e-06, + "loss": 2.7854, + "step": 5338 + }, + { + "epoch": 0.5646150592216582, + "grad_norm": 7.75, + "learning_rate": 4.9104050362690885e-06, + "loss": 2.676, + "step": 5339 + }, + { + "epoch": 0.5647208121827412, + "grad_norm": 7.28125, + "learning_rate": 4.9103595939106755e-06, + "loss": 2.3948, + "step": 5340 + }, + { + "epoch": 0.564826565143824, + "grad_norm": 8.25, + "learning_rate": 4.910314140241431e-06, + "loss": 2.7376, + "step": 5341 + }, + { + "epoch": 0.564932318104907, + "grad_norm": 7.78125, + "learning_rate": 4.910268675261568e-06, + "loss": 2.5691, + "step": 5342 + }, + { + "epoch": 0.5650380710659898, + "grad_norm": 8.375, + "learning_rate": 4.910223198971299e-06, + "loss": 2.675, + "step": 5343 + }, + { + "epoch": 0.5651438240270727, + "grad_norm": 7.8125, + "learning_rate": 4.91017771137084e-06, + "loss": 2.6769, + "step": 5344 + }, + { + "epoch": 0.5652495769881557, + "grad_norm": 8.4375, + "learning_rate": 4.9101322124604015e-06, + "loss": 2.8824, + "step": 5345 + }, + { + "epoch": 0.5653553299492385, + "grad_norm": 9.0625, + "learning_rate": 4.910086702240198e-06, + "loss": 2.6727, + "step": 5346 + }, + { + "epoch": 0.5654610829103215, + "grad_norm": 8.75, + "learning_rate": 4.910041180710444e-06, + "loss": 2.8362, + "step": 5347 + }, + { + "epoch": 0.5655668358714044, + "grad_norm": 8.1875, + "learning_rate": 4.9099956478713525e-06, + "loss": 2.7998, + "step": 5348 + }, + { + "epoch": 0.5656725888324873, + "grad_norm": 8.0625, + "learning_rate": 4.909950103723137e-06, + "loss": 2.6443, + "step": 5349 + }, + { + "epoch": 0.5657783417935702, + "grad_norm": 8.0, + "learning_rate": 4.90990454826601e-06, + "loss": 2.2814, + "step": 5350 + }, + { + "epoch": 0.5658840947546532, + "grad_norm": 7.5, + "learning_rate": 4.909858981500189e-06, + "loss": 2.563, + "step": 5351 + }, + { + "epoch": 0.565989847715736, + "grad_norm": 8.0625, + "learning_rate": 4.9098134034258835e-06, + "loss": 2.6819, + "step": 5352 + }, + { + "epoch": 0.566095600676819, + "grad_norm": 8.375, + "learning_rate": 4.909767814043309e-06, + "loss": 2.4047, + "step": 5353 + }, + { + "epoch": 0.5662013536379019, + "grad_norm": 9.0, + "learning_rate": 4.909722213352681e-06, + "loss": 2.6351, + "step": 5354 + }, + { + "epoch": 0.5663071065989848, + "grad_norm": 8.5625, + "learning_rate": 4.909676601354212e-06, + "loss": 2.6253, + "step": 5355 + }, + { + "epoch": 0.5664128595600677, + "grad_norm": 7.78125, + "learning_rate": 4.909630978048116e-06, + "loss": 2.5738, + "step": 5356 + }, + { + "epoch": 0.5665186125211505, + "grad_norm": 7.3125, + "learning_rate": 4.909585343434607e-06, + "loss": 2.5841, + "step": 5357 + }, + { + "epoch": 0.5666243654822335, + "grad_norm": 7.875, + "learning_rate": 4.9095396975139e-06, + "loss": 2.6911, + "step": 5358 + }, + { + "epoch": 0.5667301184433164, + "grad_norm": 8.0625, + "learning_rate": 4.909494040286208e-06, + "loss": 2.692, + "step": 5359 + }, + { + "epoch": 0.5668358714043993, + "grad_norm": 7.78125, + "learning_rate": 4.909448371751747e-06, + "loss": 2.6369, + "step": 5360 + }, + { + "epoch": 0.5669416243654822, + "grad_norm": 7.65625, + "learning_rate": 4.909402691910729e-06, + "loss": 2.5815, + "step": 5361 + }, + { + "epoch": 0.5670473773265652, + "grad_norm": 8.25, + "learning_rate": 4.9093570007633705e-06, + "loss": 2.6451, + "step": 5362 + }, + { + "epoch": 0.567153130287648, + "grad_norm": 8.1875, + "learning_rate": 4.909311298309884e-06, + "loss": 2.9376, + "step": 5363 + }, + { + "epoch": 0.567258883248731, + "grad_norm": 7.875, + "learning_rate": 4.909265584550485e-06, + "loss": 2.5894, + "step": 5364 + }, + { + "epoch": 0.5673646362098139, + "grad_norm": 9.0, + "learning_rate": 4.909219859485389e-06, + "loss": 2.7019, + "step": 5365 + }, + { + "epoch": 0.5674703891708968, + "grad_norm": 8.0625, + "learning_rate": 4.909174123114809e-06, + "loss": 2.6097, + "step": 5366 + }, + { + "epoch": 0.5675761421319797, + "grad_norm": 8.125, + "learning_rate": 4.909128375438961e-06, + "loss": 2.47, + "step": 5367 + }, + { + "epoch": 0.5676818950930627, + "grad_norm": 7.5625, + "learning_rate": 4.909082616458058e-06, + "loss": 2.7388, + "step": 5368 + }, + { + "epoch": 0.5677876480541455, + "grad_norm": 7.71875, + "learning_rate": 4.909036846172315e-06, + "loss": 2.5783, + "step": 5369 + }, + { + "epoch": 0.5678934010152284, + "grad_norm": 7.875, + "learning_rate": 4.908991064581949e-06, + "loss": 2.5542, + "step": 5370 + }, + { + "epoch": 0.5679991539763113, + "grad_norm": 7.90625, + "learning_rate": 4.908945271687172e-06, + "loss": 2.4984, + "step": 5371 + }, + { + "epoch": 0.5681049069373942, + "grad_norm": 9.5625, + "learning_rate": 4.9088994674882e-06, + "loss": 2.6696, + "step": 5372 + }, + { + "epoch": 0.5682106598984772, + "grad_norm": 8.1875, + "learning_rate": 4.908853651985248e-06, + "loss": 2.6618, + "step": 5373 + }, + { + "epoch": 0.56831641285956, + "grad_norm": 7.5625, + "learning_rate": 4.908807825178531e-06, + "loss": 2.5467, + "step": 5374 + }, + { + "epoch": 0.568422165820643, + "grad_norm": 8.125, + "learning_rate": 4.908761987068265e-06, + "loss": 2.5306, + "step": 5375 + }, + { + "epoch": 0.5685279187817259, + "grad_norm": 7.71875, + "learning_rate": 4.908716137654663e-06, + "loss": 2.6104, + "step": 5376 + }, + { + "epoch": 0.5686336717428088, + "grad_norm": 8.0, + "learning_rate": 4.908670276937941e-06, + "loss": 2.8187, + "step": 5377 + }, + { + "epoch": 0.5687394247038917, + "grad_norm": 8.25, + "learning_rate": 4.908624404918317e-06, + "loss": 2.7053, + "step": 5378 + }, + { + "epoch": 0.5688451776649747, + "grad_norm": 8.5625, + "learning_rate": 4.9085785215960006e-06, + "loss": 2.663, + "step": 5379 + }, + { + "epoch": 0.5689509306260575, + "grad_norm": 8.0625, + "learning_rate": 4.908532626971212e-06, + "loss": 2.8986, + "step": 5380 + }, + { + "epoch": 0.5690566835871405, + "grad_norm": 7.84375, + "learning_rate": 4.908486721044164e-06, + "loss": 2.6833, + "step": 5381 + }, + { + "epoch": 0.5691624365482234, + "grad_norm": 8.25, + "learning_rate": 4.908440803815074e-06, + "loss": 2.4383, + "step": 5382 + }, + { + "epoch": 0.5692681895093062, + "grad_norm": 7.28125, + "learning_rate": 4.9083948752841554e-06, + "loss": 2.4932, + "step": 5383 + }, + { + "epoch": 0.5693739424703892, + "grad_norm": 8.0625, + "learning_rate": 4.908348935451625e-06, + "loss": 2.5278, + "step": 5384 + }, + { + "epoch": 0.569479695431472, + "grad_norm": 8.0625, + "learning_rate": 4.908302984317697e-06, + "loss": 2.5855, + "step": 5385 + }, + { + "epoch": 0.569585448392555, + "grad_norm": 9.0, + "learning_rate": 4.90825702188259e-06, + "loss": 2.6541, + "step": 5386 + }, + { + "epoch": 0.5696912013536379, + "grad_norm": 8.4375, + "learning_rate": 4.9082110481465165e-06, + "loss": 2.5909, + "step": 5387 + }, + { + "epoch": 0.5697969543147208, + "grad_norm": 7.84375, + "learning_rate": 4.908165063109694e-06, + "loss": 2.4525, + "step": 5388 + }, + { + "epoch": 0.5699027072758037, + "grad_norm": 7.9375, + "learning_rate": 4.908119066772337e-06, + "loss": 2.2958, + "step": 5389 + }, + { + "epoch": 0.5700084602368867, + "grad_norm": 7.96875, + "learning_rate": 4.9080730591346624e-06, + "loss": 2.5649, + "step": 5390 + }, + { + "epoch": 0.5701142131979695, + "grad_norm": 7.65625, + "learning_rate": 4.908027040196887e-06, + "loss": 2.6177, + "step": 5391 + }, + { + "epoch": 0.5702199661590525, + "grad_norm": 8.125, + "learning_rate": 4.907981009959224e-06, + "loss": 2.4504, + "step": 5392 + }, + { + "epoch": 0.5703257191201354, + "grad_norm": 8.125, + "learning_rate": 4.907934968421892e-06, + "loss": 2.8051, + "step": 5393 + }, + { + "epoch": 0.5704314720812182, + "grad_norm": 8.375, + "learning_rate": 4.907888915585105e-06, + "loss": 2.5389, + "step": 5394 + }, + { + "epoch": 0.5705372250423012, + "grad_norm": 8.375, + "learning_rate": 4.90784285144908e-06, + "loss": 2.6249, + "step": 5395 + }, + { + "epoch": 0.5706429780033841, + "grad_norm": 8.5625, + "learning_rate": 4.9077967760140345e-06, + "loss": 2.5763, + "step": 5396 + }, + { + "epoch": 0.570748730964467, + "grad_norm": 8.0625, + "learning_rate": 4.907750689280183e-06, + "loss": 2.5786, + "step": 5397 + }, + { + "epoch": 0.5708544839255499, + "grad_norm": 7.84375, + "learning_rate": 4.907704591247742e-06, + "loss": 2.5658, + "step": 5398 + }, + { + "epoch": 0.5709602368866328, + "grad_norm": 7.90625, + "learning_rate": 4.9076584819169285e-06, + "loss": 2.5586, + "step": 5399 + }, + { + "epoch": 0.5710659898477157, + "grad_norm": 7.5625, + "learning_rate": 4.907612361287958e-06, + "loss": 2.6252, + "step": 5400 + }, + { + "epoch": 0.5711717428087987, + "grad_norm": 8.0, + "learning_rate": 4.907566229361047e-06, + "loss": 2.6636, + "step": 5401 + }, + { + "epoch": 0.5712774957698815, + "grad_norm": 8.0625, + "learning_rate": 4.907520086136413e-06, + "loss": 2.5998, + "step": 5402 + }, + { + "epoch": 0.5713832487309645, + "grad_norm": 7.625, + "learning_rate": 4.907473931614271e-06, + "loss": 2.317, + "step": 5403 + }, + { + "epoch": 0.5714890016920474, + "grad_norm": 7.90625, + "learning_rate": 4.90742776579484e-06, + "loss": 2.7806, + "step": 5404 + }, + { + "epoch": 0.5715947546531303, + "grad_norm": 8.5625, + "learning_rate": 4.907381588678334e-06, + "loss": 2.6532, + "step": 5405 + }, + { + "epoch": 0.5717005076142132, + "grad_norm": 7.5, + "learning_rate": 4.9073354002649715e-06, + "loss": 2.5568, + "step": 5406 + }, + { + "epoch": 0.5718062605752962, + "grad_norm": 7.5625, + "learning_rate": 4.907289200554968e-06, + "loss": 2.5911, + "step": 5407 + }, + { + "epoch": 0.571912013536379, + "grad_norm": 7.65625, + "learning_rate": 4.9072429895485404e-06, + "loss": 2.6887, + "step": 5408 + }, + { + "epoch": 0.5720177664974619, + "grad_norm": 7.84375, + "learning_rate": 4.907196767245907e-06, + "loss": 2.6208, + "step": 5409 + }, + { + "epoch": 0.5721235194585449, + "grad_norm": 7.78125, + "learning_rate": 4.907150533647283e-06, + "loss": 2.5748, + "step": 5410 + }, + { + "epoch": 0.5722292724196277, + "grad_norm": 7.84375, + "learning_rate": 4.907104288752886e-06, + "loss": 2.4375, + "step": 5411 + }, + { + "epoch": 0.5723350253807107, + "grad_norm": 7.59375, + "learning_rate": 4.907058032562933e-06, + "loss": 2.6177, + "step": 5412 + }, + { + "epoch": 0.5724407783417935, + "grad_norm": 7.59375, + "learning_rate": 4.907011765077642e-06, + "loss": 2.4796, + "step": 5413 + }, + { + "epoch": 0.5725465313028765, + "grad_norm": 7.875, + "learning_rate": 4.906965486297229e-06, + "loss": 2.6002, + "step": 5414 + }, + { + "epoch": 0.5726522842639594, + "grad_norm": 7.78125, + "learning_rate": 4.906919196221911e-06, + "loss": 2.4091, + "step": 5415 + }, + { + "epoch": 0.5727580372250423, + "grad_norm": 8.0, + "learning_rate": 4.906872894851905e-06, + "loss": 2.5121, + "step": 5416 + }, + { + "epoch": 0.5728637901861252, + "grad_norm": 7.59375, + "learning_rate": 4.906826582187429e-06, + "loss": 2.5926, + "step": 5417 + }, + { + "epoch": 0.5729695431472082, + "grad_norm": 8.25, + "learning_rate": 4.906780258228702e-06, + "loss": 2.6654, + "step": 5418 + }, + { + "epoch": 0.573075296108291, + "grad_norm": 8.3125, + "learning_rate": 4.906733922975938e-06, + "loss": 2.5555, + "step": 5419 + }, + { + "epoch": 0.573181049069374, + "grad_norm": 8.0625, + "learning_rate": 4.9066875764293566e-06, + "loss": 2.6026, + "step": 5420 + }, + { + "epoch": 0.5732868020304569, + "grad_norm": 9.125, + "learning_rate": 4.906641218589174e-06, + "loss": 2.5624, + "step": 5421 + }, + { + "epoch": 0.5733925549915397, + "grad_norm": 7.5625, + "learning_rate": 4.906594849455609e-06, + "loss": 2.6052, + "step": 5422 + }, + { + "epoch": 0.5734983079526227, + "grad_norm": 8.1875, + "learning_rate": 4.906548469028879e-06, + "loss": 2.8387, + "step": 5423 + }, + { + "epoch": 0.5736040609137056, + "grad_norm": 8.125, + "learning_rate": 4.906502077309199e-06, + "loss": 2.619, + "step": 5424 + }, + { + "epoch": 0.5737098138747885, + "grad_norm": 7.6875, + "learning_rate": 4.906455674296791e-06, + "loss": 2.5709, + "step": 5425 + }, + { + "epoch": 0.5738155668358714, + "grad_norm": 7.84375, + "learning_rate": 4.906409259991871e-06, + "loss": 2.231, + "step": 5426 + }, + { + "epoch": 0.5739213197969543, + "grad_norm": 7.9375, + "learning_rate": 4.906362834394656e-06, + "loss": 2.7093, + "step": 5427 + }, + { + "epoch": 0.5740270727580372, + "grad_norm": 7.4375, + "learning_rate": 4.9063163975053644e-06, + "loss": 2.3997, + "step": 5428 + }, + { + "epoch": 0.5741328257191202, + "grad_norm": 7.4375, + "learning_rate": 4.906269949324215e-06, + "loss": 2.5396, + "step": 5429 + }, + { + "epoch": 0.574238578680203, + "grad_norm": 7.75, + "learning_rate": 4.906223489851424e-06, + "loss": 2.5423, + "step": 5430 + }, + { + "epoch": 0.574344331641286, + "grad_norm": 7.90625, + "learning_rate": 4.906177019087209e-06, + "loss": 2.601, + "step": 5431 + }, + { + "epoch": 0.5744500846023689, + "grad_norm": 10.75, + "learning_rate": 4.906130537031791e-06, + "loss": 2.6018, + "step": 5432 + }, + { + "epoch": 0.5745558375634517, + "grad_norm": 7.875, + "learning_rate": 4.906084043685386e-06, + "loss": 2.6738, + "step": 5433 + }, + { + "epoch": 0.5746615905245347, + "grad_norm": 7.78125, + "learning_rate": 4.906037539048213e-06, + "loss": 2.485, + "step": 5434 + }, + { + "epoch": 0.5747673434856176, + "grad_norm": 8.4375, + "learning_rate": 4.905991023120489e-06, + "loss": 2.872, + "step": 5435 + }, + { + "epoch": 0.5748730964467005, + "grad_norm": 8.25, + "learning_rate": 4.905944495902434e-06, + "loss": 2.6679, + "step": 5436 + }, + { + "epoch": 0.5749788494077834, + "grad_norm": 7.5625, + "learning_rate": 4.905897957394265e-06, + "loss": 2.7058, + "step": 5437 + }, + { + "epoch": 0.5750846023688664, + "grad_norm": 7.6875, + "learning_rate": 4.9058514075962014e-06, + "loss": 2.5439, + "step": 5438 + }, + { + "epoch": 0.5751903553299492, + "grad_norm": 7.40625, + "learning_rate": 4.905804846508461e-06, + "loss": 2.4574, + "step": 5439 + }, + { + "epoch": 0.5752961082910322, + "grad_norm": 8.375, + "learning_rate": 4.905758274131262e-06, + "loss": 2.7626, + "step": 5440 + }, + { + "epoch": 0.575401861252115, + "grad_norm": 8.0625, + "learning_rate": 4.905711690464823e-06, + "loss": 2.5501, + "step": 5441 + }, + { + "epoch": 0.575507614213198, + "grad_norm": 8.0625, + "learning_rate": 4.905665095509363e-06, + "loss": 2.5001, + "step": 5442 + }, + { + "epoch": 0.5756133671742809, + "grad_norm": 7.4375, + "learning_rate": 4.9056184892651015e-06, + "loss": 2.512, + "step": 5443 + }, + { + "epoch": 0.5757191201353637, + "grad_norm": 7.53125, + "learning_rate": 4.905571871732255e-06, + "loss": 2.5508, + "step": 5444 + }, + { + "epoch": 0.5758248730964467, + "grad_norm": 7.75, + "learning_rate": 4.905525242911045e-06, + "loss": 2.587, + "step": 5445 + }, + { + "epoch": 0.5759306260575296, + "grad_norm": 7.875, + "learning_rate": 4.905478602801688e-06, + "loss": 2.7373, + "step": 5446 + }, + { + "epoch": 0.5760363790186125, + "grad_norm": 8.1875, + "learning_rate": 4.905431951404404e-06, + "loss": 2.5738, + "step": 5447 + }, + { + "epoch": 0.5761421319796954, + "grad_norm": 8.0, + "learning_rate": 4.905385288719412e-06, + "loss": 2.7946, + "step": 5448 + }, + { + "epoch": 0.5762478849407784, + "grad_norm": 7.71875, + "learning_rate": 4.90533861474693e-06, + "loss": 2.54, + "step": 5449 + }, + { + "epoch": 0.5763536379018612, + "grad_norm": 7.75, + "learning_rate": 4.905291929487178e-06, + "loss": 2.6655, + "step": 5450 + }, + { + "epoch": 0.5764593908629442, + "grad_norm": 7.65625, + "learning_rate": 4.905245232940374e-06, + "loss": 2.6155, + "step": 5451 + }, + { + "epoch": 0.5765651438240271, + "grad_norm": 8.0625, + "learning_rate": 4.905198525106738e-06, + "loss": 2.7817, + "step": 5452 + }, + { + "epoch": 0.57667089678511, + "grad_norm": 7.8125, + "learning_rate": 4.90515180598649e-06, + "loss": 2.5496, + "step": 5453 + }, + { + "epoch": 0.5767766497461929, + "grad_norm": 7.84375, + "learning_rate": 4.905105075579847e-06, + "loss": 2.5053, + "step": 5454 + }, + { + "epoch": 0.5768824027072758, + "grad_norm": 7.40625, + "learning_rate": 4.905058333887031e-06, + "loss": 2.5689, + "step": 5455 + }, + { + "epoch": 0.5769881556683587, + "grad_norm": 8.125, + "learning_rate": 4.905011580908259e-06, + "loss": 2.737, + "step": 5456 + }, + { + "epoch": 0.5770939086294417, + "grad_norm": 7.46875, + "learning_rate": 4.904964816643751e-06, + "loss": 2.5275, + "step": 5457 + }, + { + "epoch": 0.5771996615905245, + "grad_norm": 7.1875, + "learning_rate": 4.904918041093727e-06, + "loss": 2.5748, + "step": 5458 + }, + { + "epoch": 0.5773054145516074, + "grad_norm": 7.84375, + "learning_rate": 4.9048712542584056e-06, + "loss": 2.5454, + "step": 5459 + }, + { + "epoch": 0.5774111675126904, + "grad_norm": 8.875, + "learning_rate": 4.904824456138008e-06, + "loss": 2.5794, + "step": 5460 + }, + { + "epoch": 0.5775169204737732, + "grad_norm": 8.25, + "learning_rate": 4.904777646732753e-06, + "loss": 2.6125, + "step": 5461 + }, + { + "epoch": 0.5776226734348562, + "grad_norm": 7.53125, + "learning_rate": 4.904730826042859e-06, + "loss": 2.5926, + "step": 5462 + }, + { + "epoch": 0.5777284263959391, + "grad_norm": 7.96875, + "learning_rate": 4.904683994068546e-06, + "loss": 2.6184, + "step": 5463 + }, + { + "epoch": 0.577834179357022, + "grad_norm": 7.71875, + "learning_rate": 4.904637150810036e-06, + "loss": 2.4695, + "step": 5464 + }, + { + "epoch": 0.5779399323181049, + "grad_norm": 7.53125, + "learning_rate": 4.904590296267547e-06, + "loss": 2.498, + "step": 5465 + }, + { + "epoch": 0.5780456852791879, + "grad_norm": 7.75, + "learning_rate": 4.9045434304412995e-06, + "loss": 2.6526, + "step": 5466 + }, + { + "epoch": 0.5781514382402707, + "grad_norm": 7.9375, + "learning_rate": 4.9044965533315125e-06, + "loss": 2.72, + "step": 5467 + }, + { + "epoch": 0.5782571912013537, + "grad_norm": 7.40625, + "learning_rate": 4.904449664938406e-06, + "loss": 2.4698, + "step": 5468 + }, + { + "epoch": 0.5783629441624365, + "grad_norm": 7.8125, + "learning_rate": 4.904402765262201e-06, + "loss": 2.5067, + "step": 5469 + }, + { + "epoch": 0.5784686971235194, + "grad_norm": 8.0, + "learning_rate": 4.9043558543031185e-06, + "loss": 2.8236, + "step": 5470 + }, + { + "epoch": 0.5785744500846024, + "grad_norm": 7.96875, + "learning_rate": 4.9043089320613764e-06, + "loss": 2.7058, + "step": 5471 + }, + { + "epoch": 0.5786802030456852, + "grad_norm": 7.875, + "learning_rate": 4.904261998537196e-06, + "loss": 2.3457, + "step": 5472 + }, + { + "epoch": 0.5787859560067682, + "grad_norm": 7.9375, + "learning_rate": 4.904215053730797e-06, + "loss": 2.7518, + "step": 5473 + }, + { + "epoch": 0.5788917089678511, + "grad_norm": 7.65625, + "learning_rate": 4.9041680976424e-06, + "loss": 2.5657, + "step": 5474 + }, + { + "epoch": 0.578997461928934, + "grad_norm": 7.6875, + "learning_rate": 4.904121130272226e-06, + "loss": 2.6589, + "step": 5475 + }, + { + "epoch": 0.5791032148900169, + "grad_norm": 7.53125, + "learning_rate": 4.904074151620494e-06, + "loss": 2.6643, + "step": 5476 + }, + { + "epoch": 0.5792089678510999, + "grad_norm": 7.78125, + "learning_rate": 4.904027161687425e-06, + "loss": 2.6065, + "step": 5477 + }, + { + "epoch": 0.5793147208121827, + "grad_norm": 8.3125, + "learning_rate": 4.903980160473241e-06, + "loss": 2.5915, + "step": 5478 + }, + { + "epoch": 0.5794204737732657, + "grad_norm": 7.5625, + "learning_rate": 4.90393314797816e-06, + "loss": 2.5923, + "step": 5479 + }, + { + "epoch": 0.5795262267343486, + "grad_norm": 7.53125, + "learning_rate": 4.903886124202405e-06, + "loss": 2.6172, + "step": 5480 + }, + { + "epoch": 0.5796319796954315, + "grad_norm": 9.0625, + "learning_rate": 4.903839089146195e-06, + "loss": 2.4536, + "step": 5481 + }, + { + "epoch": 0.5797377326565144, + "grad_norm": 7.71875, + "learning_rate": 4.903792042809752e-06, + "loss": 2.6425, + "step": 5482 + }, + { + "epoch": 0.5798434856175972, + "grad_norm": 9.0, + "learning_rate": 4.903744985193295e-06, + "loss": 2.4934, + "step": 5483 + }, + { + "epoch": 0.5799492385786802, + "grad_norm": 8.6875, + "learning_rate": 4.903697916297047e-06, + "loss": 2.7046, + "step": 5484 + }, + { + "epoch": 0.5800549915397631, + "grad_norm": 7.5625, + "learning_rate": 4.903650836121227e-06, + "loss": 2.6426, + "step": 5485 + }, + { + "epoch": 0.580160744500846, + "grad_norm": 8.1875, + "learning_rate": 4.903603744666057e-06, + "loss": 2.6335, + "step": 5486 + }, + { + "epoch": 0.5802664974619289, + "grad_norm": 7.90625, + "learning_rate": 4.903556641931757e-06, + "loss": 2.6048, + "step": 5487 + }, + { + "epoch": 0.5803722504230119, + "grad_norm": 7.78125, + "learning_rate": 4.90350952791855e-06, + "loss": 2.6872, + "step": 5488 + }, + { + "epoch": 0.5804780033840947, + "grad_norm": 8.25, + "learning_rate": 4.903462402626654e-06, + "loss": 2.5981, + "step": 5489 + }, + { + "epoch": 0.5805837563451777, + "grad_norm": 7.65625, + "learning_rate": 4.9034152660562926e-06, + "loss": 2.7585, + "step": 5490 + }, + { + "epoch": 0.5806895093062606, + "grad_norm": 8.0, + "learning_rate": 4.9033681182076875e-06, + "loss": 2.6205, + "step": 5491 + }, + { + "epoch": 0.5807952622673435, + "grad_norm": 7.90625, + "learning_rate": 4.903320959081057e-06, + "loss": 2.7335, + "step": 5492 + }, + { + "epoch": 0.5809010152284264, + "grad_norm": 9.5, + "learning_rate": 4.9032737886766255e-06, + "loss": 2.6537, + "step": 5493 + }, + { + "epoch": 0.5810067681895094, + "grad_norm": 7.5, + "learning_rate": 4.903226606994612e-06, + "loss": 2.545, + "step": 5494 + }, + { + "epoch": 0.5811125211505922, + "grad_norm": 7.59375, + "learning_rate": 4.90317941403524e-06, + "loss": 2.5873, + "step": 5495 + }, + { + "epoch": 0.5812182741116751, + "grad_norm": 8.4375, + "learning_rate": 4.903132209798729e-06, + "loss": 2.6624, + "step": 5496 + }, + { + "epoch": 0.581324027072758, + "grad_norm": 8.0625, + "learning_rate": 4.903084994285302e-06, + "loss": 2.7985, + "step": 5497 + }, + { + "epoch": 0.5814297800338409, + "grad_norm": 8.125, + "learning_rate": 4.90303776749518e-06, + "loss": 2.7462, + "step": 5498 + }, + { + "epoch": 0.5815355329949239, + "grad_norm": 7.53125, + "learning_rate": 4.902990529428584e-06, + "loss": 2.6055, + "step": 5499 + }, + { + "epoch": 0.5816412859560067, + "grad_norm": 8.125, + "learning_rate": 4.902943280085737e-06, + "loss": 2.4696, + "step": 5500 + }, + { + "epoch": 0.5817470389170897, + "grad_norm": 7.5625, + "learning_rate": 4.902896019466859e-06, + "loss": 2.5689, + "step": 5501 + }, + { + "epoch": 0.5818527918781726, + "grad_norm": 7.8125, + "learning_rate": 4.902848747572173e-06, + "loss": 2.6481, + "step": 5502 + }, + { + "epoch": 0.5819585448392555, + "grad_norm": 7.75, + "learning_rate": 4.9028014644019e-06, + "loss": 2.6568, + "step": 5503 + }, + { + "epoch": 0.5820642978003384, + "grad_norm": 7.9375, + "learning_rate": 4.902754169956265e-06, + "loss": 2.5689, + "step": 5504 + }, + { + "epoch": 0.5821700507614214, + "grad_norm": 8.0625, + "learning_rate": 4.9027068642354845e-06, + "loss": 2.3905, + "step": 5505 + }, + { + "epoch": 0.5822758037225042, + "grad_norm": 8.5625, + "learning_rate": 4.9026595472397844e-06, + "loss": 2.7316, + "step": 5506 + }, + { + "epoch": 0.5823815566835872, + "grad_norm": 7.75, + "learning_rate": 4.9026122189693855e-06, + "loss": 2.4473, + "step": 5507 + }, + { + "epoch": 0.5824873096446701, + "grad_norm": 8.0, + "learning_rate": 4.9025648794245105e-06, + "loss": 2.4439, + "step": 5508 + }, + { + "epoch": 0.5825930626057529, + "grad_norm": 7.96875, + "learning_rate": 4.902517528605381e-06, + "loss": 2.7244, + "step": 5509 + }, + { + "epoch": 0.5826988155668359, + "grad_norm": 7.90625, + "learning_rate": 4.90247016651222e-06, + "loss": 2.6023, + "step": 5510 + }, + { + "epoch": 0.5828045685279187, + "grad_norm": 8.125, + "learning_rate": 4.902422793145248e-06, + "loss": 2.8311, + "step": 5511 + }, + { + "epoch": 0.5829103214890017, + "grad_norm": 8.125, + "learning_rate": 4.902375408504689e-06, + "loss": 2.8809, + "step": 5512 + }, + { + "epoch": 0.5830160744500846, + "grad_norm": 7.46875, + "learning_rate": 4.902328012590764e-06, + "loss": 2.649, + "step": 5513 + }, + { + "epoch": 0.5831218274111675, + "grad_norm": 8.25, + "learning_rate": 4.902280605403696e-06, + "loss": 2.7033, + "step": 5514 + }, + { + "epoch": 0.5832275803722504, + "grad_norm": 8.0, + "learning_rate": 4.902233186943709e-06, + "loss": 2.6966, + "step": 5515 + }, + { + "epoch": 0.5833333333333334, + "grad_norm": 7.96875, + "learning_rate": 4.902185757211023e-06, + "loss": 2.7785, + "step": 5516 + }, + { + "epoch": 0.5834390862944162, + "grad_norm": 7.5, + "learning_rate": 4.902138316205863e-06, + "loss": 2.4652, + "step": 5517 + }, + { + "epoch": 0.5835448392554992, + "grad_norm": 8.0, + "learning_rate": 4.902090863928449e-06, + "loss": 2.484, + "step": 5518 + }, + { + "epoch": 0.5836505922165821, + "grad_norm": 7.6875, + "learning_rate": 4.9020434003790054e-06, + "loss": 2.4782, + "step": 5519 + }, + { + "epoch": 0.583756345177665, + "grad_norm": 7.71875, + "learning_rate": 4.9019959255577544e-06, + "loss": 2.6674, + "step": 5520 + }, + { + "epoch": 0.5838620981387479, + "grad_norm": 7.84375, + "learning_rate": 4.901948439464919e-06, + "loss": 2.5558, + "step": 5521 + }, + { + "epoch": 0.5839678510998308, + "grad_norm": 8.75, + "learning_rate": 4.901900942100722e-06, + "loss": 2.5942, + "step": 5522 + }, + { + "epoch": 0.5840736040609137, + "grad_norm": 8.125, + "learning_rate": 4.9018534334653866e-06, + "loss": 2.6655, + "step": 5523 + }, + { + "epoch": 0.5841793570219966, + "grad_norm": 8.25, + "learning_rate": 4.901805913559135e-06, + "loss": 2.8925, + "step": 5524 + }, + { + "epoch": 0.5842851099830795, + "grad_norm": 7.5625, + "learning_rate": 4.90175838238219e-06, + "loss": 2.6639, + "step": 5525 + }, + { + "epoch": 0.5843908629441624, + "grad_norm": 8.6875, + "learning_rate": 4.9017108399347754e-06, + "loss": 2.4702, + "step": 5526 + }, + { + "epoch": 0.5844966159052454, + "grad_norm": 8.0, + "learning_rate": 4.901663286217116e-06, + "loss": 2.669, + "step": 5527 + }, + { + "epoch": 0.5846023688663282, + "grad_norm": 7.875, + "learning_rate": 4.90161572122943e-06, + "loss": 2.5377, + "step": 5528 + }, + { + "epoch": 0.5847081218274112, + "grad_norm": 7.875, + "learning_rate": 4.901568144971945e-06, + "loss": 2.6796, + "step": 5529 + }, + { + "epoch": 0.5848138747884941, + "grad_norm": 7.65625, + "learning_rate": 4.901520557444884e-06, + "loss": 2.396, + "step": 5530 + }, + { + "epoch": 0.584919627749577, + "grad_norm": 8.0625, + "learning_rate": 4.901472958648468e-06, + "loss": 2.2738, + "step": 5531 + }, + { + "epoch": 0.5850253807106599, + "grad_norm": 8.0625, + "learning_rate": 4.9014253485829215e-06, + "loss": 2.4652, + "step": 5532 + }, + { + "epoch": 0.5851311336717429, + "grad_norm": 7.75, + "learning_rate": 4.901377727248468e-06, + "loss": 2.6498, + "step": 5533 + }, + { + "epoch": 0.5852368866328257, + "grad_norm": 8.125, + "learning_rate": 4.901330094645332e-06, + "loss": 2.7625, + "step": 5534 + }, + { + "epoch": 0.5853426395939086, + "grad_norm": 7.84375, + "learning_rate": 4.901282450773735e-06, + "loss": 2.3827, + "step": 5535 + }, + { + "epoch": 0.5854483925549916, + "grad_norm": 7.59375, + "learning_rate": 4.9012347956339015e-06, + "loss": 2.6156, + "step": 5536 + }, + { + "epoch": 0.5855541455160744, + "grad_norm": 7.75, + "learning_rate": 4.901187129226056e-06, + "loss": 2.7086, + "step": 5537 + }, + { + "epoch": 0.5856598984771574, + "grad_norm": 7.71875, + "learning_rate": 4.901139451550421e-06, + "loss": 2.4846, + "step": 5538 + }, + { + "epoch": 0.5857656514382402, + "grad_norm": 8.0625, + "learning_rate": 4.901091762607219e-06, + "loss": 2.5651, + "step": 5539 + }, + { + "epoch": 0.5858714043993232, + "grad_norm": 7.9375, + "learning_rate": 4.9010440623966775e-06, + "loss": 2.591, + "step": 5540 + }, + { + "epoch": 0.5859771573604061, + "grad_norm": 8.4375, + "learning_rate": 4.900996350919017e-06, + "loss": 2.7721, + "step": 5541 + }, + { + "epoch": 0.586082910321489, + "grad_norm": 7.96875, + "learning_rate": 4.900948628174463e-06, + "loss": 2.5596, + "step": 5542 + }, + { + "epoch": 0.5861886632825719, + "grad_norm": 7.4375, + "learning_rate": 4.900900894163239e-06, + "loss": 2.5519, + "step": 5543 + }, + { + "epoch": 0.5862944162436549, + "grad_norm": 7.53125, + "learning_rate": 4.90085314888557e-06, + "loss": 2.7469, + "step": 5544 + }, + { + "epoch": 0.5864001692047377, + "grad_norm": 8.6875, + "learning_rate": 4.9008053923416785e-06, + "loss": 2.299, + "step": 5545 + }, + { + "epoch": 0.5865059221658206, + "grad_norm": 7.75, + "learning_rate": 4.900757624531789e-06, + "loss": 2.638, + "step": 5546 + }, + { + "epoch": 0.5866116751269036, + "grad_norm": 7.625, + "learning_rate": 4.900709845456125e-06, + "loss": 2.4113, + "step": 5547 + }, + { + "epoch": 0.5867174280879864, + "grad_norm": 8.375, + "learning_rate": 4.900662055114912e-06, + "loss": 2.44, + "step": 5548 + }, + { + "epoch": 0.5868231810490694, + "grad_norm": 7.375, + "learning_rate": 4.900614253508375e-06, + "loss": 2.4404, + "step": 5549 + }, + { + "epoch": 0.5869289340101523, + "grad_norm": 7.5625, + "learning_rate": 4.900566440636736e-06, + "loss": 2.504, + "step": 5550 + }, + { + "epoch": 0.5870346869712352, + "grad_norm": 7.46875, + "learning_rate": 4.900518616500221e-06, + "loss": 2.5578, + "step": 5551 + }, + { + "epoch": 0.5871404399323181, + "grad_norm": 8.375, + "learning_rate": 4.9004707810990534e-06, + "loss": 2.5144, + "step": 5552 + }, + { + "epoch": 0.587246192893401, + "grad_norm": 7.65625, + "learning_rate": 4.900422934433459e-06, + "loss": 2.4358, + "step": 5553 + }, + { + "epoch": 0.5873519458544839, + "grad_norm": 7.90625, + "learning_rate": 4.9003750765036605e-06, + "loss": 2.702, + "step": 5554 + }, + { + "epoch": 0.5874576988155669, + "grad_norm": 8.125, + "learning_rate": 4.900327207309884e-06, + "loss": 2.6853, + "step": 5555 + }, + { + "epoch": 0.5875634517766497, + "grad_norm": 8.0625, + "learning_rate": 4.900279326852353e-06, + "loss": 2.3991, + "step": 5556 + }, + { + "epoch": 0.5876692047377327, + "grad_norm": 7.71875, + "learning_rate": 4.900231435131294e-06, + "loss": 2.5923, + "step": 5557 + }, + { + "epoch": 0.5877749576988156, + "grad_norm": 9.0, + "learning_rate": 4.900183532146929e-06, + "loss": 2.6281, + "step": 5558 + }, + { + "epoch": 0.5878807106598984, + "grad_norm": 8.5, + "learning_rate": 4.900135617899486e-06, + "loss": 2.8438, + "step": 5559 + }, + { + "epoch": 0.5879864636209814, + "grad_norm": 8.0625, + "learning_rate": 4.900087692389187e-06, + "loss": 2.6706, + "step": 5560 + }, + { + "epoch": 0.5880922165820643, + "grad_norm": 7.53125, + "learning_rate": 4.900039755616259e-06, + "loss": 2.5374, + "step": 5561 + }, + { + "epoch": 0.5881979695431472, + "grad_norm": 8.0, + "learning_rate": 4.899991807580925e-06, + "loss": 2.579, + "step": 5562 + }, + { + "epoch": 0.5883037225042301, + "grad_norm": 9.375, + "learning_rate": 4.899943848283411e-06, + "loss": 2.3746, + "step": 5563 + }, + { + "epoch": 0.5884094754653131, + "grad_norm": 8.0, + "learning_rate": 4.8998958777239435e-06, + "loss": 2.4872, + "step": 5564 + }, + { + "epoch": 0.5885152284263959, + "grad_norm": 8.125, + "learning_rate": 4.899847895902745e-06, + "loss": 2.3895, + "step": 5565 + }, + { + "epoch": 0.5886209813874789, + "grad_norm": 8.0, + "learning_rate": 4.899799902820041e-06, + "loss": 2.7341, + "step": 5566 + }, + { + "epoch": 0.5887267343485617, + "grad_norm": 8.0, + "learning_rate": 4.8997518984760595e-06, + "loss": 2.5124, + "step": 5567 + }, + { + "epoch": 0.5888324873096447, + "grad_norm": 7.90625, + "learning_rate": 4.899703882871023e-06, + "loss": 2.7733, + "step": 5568 + }, + { + "epoch": 0.5889382402707276, + "grad_norm": 7.84375, + "learning_rate": 4.899655856005157e-06, + "loss": 2.5543, + "step": 5569 + }, + { + "epoch": 0.5890439932318104, + "grad_norm": 7.96875, + "learning_rate": 4.899607817878689e-06, + "loss": 2.5583, + "step": 5570 + }, + { + "epoch": 0.5891497461928934, + "grad_norm": 7.78125, + "learning_rate": 4.899559768491842e-06, + "loss": 2.6528, + "step": 5571 + }, + { + "epoch": 0.5892554991539763, + "grad_norm": 7.96875, + "learning_rate": 4.899511707844842e-06, + "loss": 2.5099, + "step": 5572 + }, + { + "epoch": 0.5893612521150592, + "grad_norm": 13.9375, + "learning_rate": 4.899463635937914e-06, + "loss": 2.6249, + "step": 5573 + }, + { + "epoch": 0.5894670050761421, + "grad_norm": 8.0625, + "learning_rate": 4.899415552771286e-06, + "loss": 2.4996, + "step": 5574 + }, + { + "epoch": 0.5895727580372251, + "grad_norm": 7.59375, + "learning_rate": 4.899367458345182e-06, + "loss": 2.6513, + "step": 5575 + }, + { + "epoch": 0.5896785109983079, + "grad_norm": 7.03125, + "learning_rate": 4.899319352659828e-06, + "loss": 2.4352, + "step": 5576 + }, + { + "epoch": 0.5897842639593909, + "grad_norm": 9.1875, + "learning_rate": 4.8992712357154495e-06, + "loss": 2.5385, + "step": 5577 + }, + { + "epoch": 0.5898900169204738, + "grad_norm": 8.3125, + "learning_rate": 4.899223107512272e-06, + "loss": 2.5511, + "step": 5578 + }, + { + "epoch": 0.5899957698815567, + "grad_norm": 7.5625, + "learning_rate": 4.8991749680505216e-06, + "loss": 2.6287, + "step": 5579 + }, + { + "epoch": 0.5901015228426396, + "grad_norm": 8.0625, + "learning_rate": 4.899126817330425e-06, + "loss": 2.5634, + "step": 5580 + }, + { + "epoch": 0.5902072758037225, + "grad_norm": 7.21875, + "learning_rate": 4.899078655352206e-06, + "loss": 2.431, + "step": 5581 + }, + { + "epoch": 0.5903130287648054, + "grad_norm": 7.84375, + "learning_rate": 4.899030482116094e-06, + "loss": 2.6758, + "step": 5582 + }, + { + "epoch": 0.5904187817258884, + "grad_norm": 7.46875, + "learning_rate": 4.898982297622312e-06, + "loss": 2.6442, + "step": 5583 + }, + { + "epoch": 0.5905245346869712, + "grad_norm": 7.84375, + "learning_rate": 4.8989341018710876e-06, + "loss": 2.5905, + "step": 5584 + }, + { + "epoch": 0.5906302876480541, + "grad_norm": 8.1875, + "learning_rate": 4.898885894862646e-06, + "loss": 2.741, + "step": 5585 + }, + { + "epoch": 0.5907360406091371, + "grad_norm": 8.625, + "learning_rate": 4.8988376765972144e-06, + "loss": 2.7399, + "step": 5586 + }, + { + "epoch": 0.5908417935702199, + "grad_norm": 7.71875, + "learning_rate": 4.898789447075018e-06, + "loss": 2.7074, + "step": 5587 + }, + { + "epoch": 0.5909475465313029, + "grad_norm": 7.1875, + "learning_rate": 4.898741206296285e-06, + "loss": 2.6919, + "step": 5588 + }, + { + "epoch": 0.5910532994923858, + "grad_norm": 8.0625, + "learning_rate": 4.89869295426124e-06, + "loss": 2.8565, + "step": 5589 + }, + { + "epoch": 0.5911590524534687, + "grad_norm": 8.25, + "learning_rate": 4.89864469097011e-06, + "loss": 2.7029, + "step": 5590 + }, + { + "epoch": 0.5912648054145516, + "grad_norm": 8.625, + "learning_rate": 4.8985964164231215e-06, + "loss": 2.4333, + "step": 5591 + }, + { + "epoch": 0.5913705583756346, + "grad_norm": 7.59375, + "learning_rate": 4.898548130620501e-06, + "loss": 2.6938, + "step": 5592 + }, + { + "epoch": 0.5914763113367174, + "grad_norm": 7.46875, + "learning_rate": 4.898499833562476e-06, + "loss": 2.6024, + "step": 5593 + }, + { + "epoch": 0.5915820642978004, + "grad_norm": 8.0625, + "learning_rate": 4.898451525249271e-06, + "loss": 2.6129, + "step": 5594 + }, + { + "epoch": 0.5916878172588832, + "grad_norm": 7.78125, + "learning_rate": 4.898403205681114e-06, + "loss": 2.6836, + "step": 5595 + }, + { + "epoch": 0.5917935702199661, + "grad_norm": 8.1875, + "learning_rate": 4.898354874858232e-06, + "loss": 2.789, + "step": 5596 + }, + { + "epoch": 0.5918993231810491, + "grad_norm": 7.625, + "learning_rate": 4.898306532780851e-06, + "loss": 2.477, + "step": 5597 + }, + { + "epoch": 0.5920050761421319, + "grad_norm": 7.96875, + "learning_rate": 4.898258179449199e-06, + "loss": 2.5475, + "step": 5598 + }, + { + "epoch": 0.5921108291032149, + "grad_norm": 7.4375, + "learning_rate": 4.898209814863501e-06, + "loss": 2.673, + "step": 5599 + }, + { + "epoch": 0.5922165820642978, + "grad_norm": 8.125, + "learning_rate": 4.8981614390239875e-06, + "loss": 2.7362, + "step": 5600 + }, + { + "epoch": 0.5923223350253807, + "grad_norm": 7.8125, + "learning_rate": 4.898113051930881e-06, + "loss": 2.654, + "step": 5601 + }, + { + "epoch": 0.5924280879864636, + "grad_norm": 9.1875, + "learning_rate": 4.898064653584411e-06, + "loss": 2.6591, + "step": 5602 + }, + { + "epoch": 0.5925338409475466, + "grad_norm": 8.625, + "learning_rate": 4.8980162439848055e-06, + "loss": 2.569, + "step": 5603 + }, + { + "epoch": 0.5926395939086294, + "grad_norm": 7.625, + "learning_rate": 4.89796782313229e-06, + "loss": 2.6928, + "step": 5604 + }, + { + "epoch": 0.5927453468697124, + "grad_norm": 8.0625, + "learning_rate": 4.897919391027092e-06, + "loss": 2.3864, + "step": 5605 + }, + { + "epoch": 0.5928510998307953, + "grad_norm": 8.1875, + "learning_rate": 4.89787094766944e-06, + "loss": 2.7098, + "step": 5606 + }, + { + "epoch": 0.5929568527918782, + "grad_norm": 7.71875, + "learning_rate": 4.897822493059558e-06, + "loss": 2.5724, + "step": 5607 + }, + { + "epoch": 0.5930626057529611, + "grad_norm": 7.625, + "learning_rate": 4.897774027197678e-06, + "loss": 2.6673, + "step": 5608 + }, + { + "epoch": 0.5931683587140439, + "grad_norm": 7.59375, + "learning_rate": 4.897725550084025e-06, + "loss": 2.4774, + "step": 5609 + }, + { + "epoch": 0.5932741116751269, + "grad_norm": 7.59375, + "learning_rate": 4.8976770617188254e-06, + "loss": 2.4922, + "step": 5610 + }, + { + "epoch": 0.5933798646362098, + "grad_norm": 8.5, + "learning_rate": 4.897628562102309e-06, + "loss": 2.583, + "step": 5611 + }, + { + "epoch": 0.5934856175972927, + "grad_norm": 7.5625, + "learning_rate": 4.897580051234702e-06, + "loss": 2.506, + "step": 5612 + }, + { + "epoch": 0.5935913705583756, + "grad_norm": 7.53125, + "learning_rate": 4.897531529116232e-06, + "loss": 2.6672, + "step": 5613 + }, + { + "epoch": 0.5936971235194586, + "grad_norm": 8.5625, + "learning_rate": 4.897482995747127e-06, + "loss": 2.8255, + "step": 5614 + }, + { + "epoch": 0.5938028764805414, + "grad_norm": 7.9375, + "learning_rate": 4.897434451127615e-06, + "loss": 2.4529, + "step": 5615 + }, + { + "epoch": 0.5939086294416244, + "grad_norm": 7.75, + "learning_rate": 4.897385895257924e-06, + "loss": 2.6517, + "step": 5616 + }, + { + "epoch": 0.5940143824027073, + "grad_norm": 8.5, + "learning_rate": 4.897337328138282e-06, + "loss": 2.7618, + "step": 5617 + }, + { + "epoch": 0.5941201353637902, + "grad_norm": 7.5625, + "learning_rate": 4.897288749768916e-06, + "loss": 2.6908, + "step": 5618 + }, + { + "epoch": 0.5942258883248731, + "grad_norm": 7.875, + "learning_rate": 4.897240160150054e-06, + "loss": 2.5393, + "step": 5619 + }, + { + "epoch": 0.5943316412859561, + "grad_norm": 7.34375, + "learning_rate": 4.897191559281924e-06, + "loss": 2.5112, + "step": 5620 + }, + { + "epoch": 0.5944373942470389, + "grad_norm": 7.59375, + "learning_rate": 4.897142947164755e-06, + "loss": 2.5302, + "step": 5621 + }, + { + "epoch": 0.5945431472081218, + "grad_norm": 8.3125, + "learning_rate": 4.897094323798775e-06, + "loss": 2.3763, + "step": 5622 + }, + { + "epoch": 0.5946489001692047, + "grad_norm": 7.6875, + "learning_rate": 4.897045689184211e-06, + "loss": 2.6004, + "step": 5623 + }, + { + "epoch": 0.5947546531302876, + "grad_norm": 8.25, + "learning_rate": 4.896997043321292e-06, + "loss": 2.7215, + "step": 5624 + }, + { + "epoch": 0.5948604060913706, + "grad_norm": 7.6875, + "learning_rate": 4.896948386210246e-06, + "loss": 2.5828, + "step": 5625 + }, + { + "epoch": 0.5949661590524534, + "grad_norm": 7.4375, + "learning_rate": 4.896899717851302e-06, + "loss": 2.546, + "step": 5626 + }, + { + "epoch": 0.5950719120135364, + "grad_norm": 7.78125, + "learning_rate": 4.896851038244688e-06, + "loss": 2.6933, + "step": 5627 + }, + { + "epoch": 0.5951776649746193, + "grad_norm": 7.4375, + "learning_rate": 4.896802347390632e-06, + "loss": 2.5998, + "step": 5628 + }, + { + "epoch": 0.5952834179357022, + "grad_norm": 7.6875, + "learning_rate": 4.896753645289362e-06, + "loss": 2.3236, + "step": 5629 + }, + { + "epoch": 0.5953891708967851, + "grad_norm": 7.90625, + "learning_rate": 4.896704931941109e-06, + "loss": 2.5853, + "step": 5630 + }, + { + "epoch": 0.5954949238578681, + "grad_norm": 8.875, + "learning_rate": 4.8966562073461e-06, + "loss": 2.3525, + "step": 5631 + }, + { + "epoch": 0.5956006768189509, + "grad_norm": 7.8125, + "learning_rate": 4.896607471504563e-06, + "loss": 2.4205, + "step": 5632 + }, + { + "epoch": 0.5957064297800339, + "grad_norm": 7.5625, + "learning_rate": 4.896558724416727e-06, + "loss": 2.6536, + "step": 5633 + }, + { + "epoch": 0.5958121827411168, + "grad_norm": 7.46875, + "learning_rate": 4.896509966082821e-06, + "loss": 2.7662, + "step": 5634 + }, + { + "epoch": 0.5959179357021996, + "grad_norm": 7.8125, + "learning_rate": 4.896461196503074e-06, + "loss": 2.5026, + "step": 5635 + }, + { + "epoch": 0.5960236886632826, + "grad_norm": 8.3125, + "learning_rate": 4.896412415677715e-06, + "loss": 2.6474, + "step": 5636 + }, + { + "epoch": 0.5961294416243654, + "grad_norm": 8.25, + "learning_rate": 4.896363623606973e-06, + "loss": 2.6826, + "step": 5637 + }, + { + "epoch": 0.5962351945854484, + "grad_norm": 7.875, + "learning_rate": 4.896314820291076e-06, + "loss": 2.6035, + "step": 5638 + }, + { + "epoch": 0.5963409475465313, + "grad_norm": 8.8125, + "learning_rate": 4.896266005730254e-06, + "loss": 2.708, + "step": 5639 + }, + { + "epoch": 0.5964467005076142, + "grad_norm": 7.90625, + "learning_rate": 4.896217179924735e-06, + "loss": 2.499, + "step": 5640 + }, + { + "epoch": 0.5965524534686971, + "grad_norm": 7.875, + "learning_rate": 4.89616834287475e-06, + "loss": 2.6613, + "step": 5641 + }, + { + "epoch": 0.5966582064297801, + "grad_norm": 7.5625, + "learning_rate": 4.896119494580526e-06, + "loss": 2.4672, + "step": 5642 + }, + { + "epoch": 0.5967639593908629, + "grad_norm": 7.8125, + "learning_rate": 4.896070635042293e-06, + "loss": 2.4905, + "step": 5643 + }, + { + "epoch": 0.5968697123519459, + "grad_norm": 8.125, + "learning_rate": 4.8960217642602805e-06, + "loss": 2.3574, + "step": 5644 + }, + { + "epoch": 0.5969754653130288, + "grad_norm": 7.25, + "learning_rate": 4.895972882234719e-06, + "loss": 2.4395, + "step": 5645 + }, + { + "epoch": 0.5970812182741116, + "grad_norm": 8.0625, + "learning_rate": 4.8959239889658354e-06, + "loss": 2.5376, + "step": 5646 + }, + { + "epoch": 0.5971869712351946, + "grad_norm": 10.25, + "learning_rate": 4.895875084453861e-06, + "loss": 2.5773, + "step": 5647 + }, + { + "epoch": 0.5972927241962775, + "grad_norm": 7.4375, + "learning_rate": 4.895826168699025e-06, + "loss": 2.4027, + "step": 5648 + }, + { + "epoch": 0.5973984771573604, + "grad_norm": 8.125, + "learning_rate": 4.895777241701556e-06, + "loss": 2.6875, + "step": 5649 + }, + { + "epoch": 0.5975042301184433, + "grad_norm": 7.78125, + "learning_rate": 4.895728303461684e-06, + "loss": 2.4578, + "step": 5650 + }, + { + "epoch": 0.5976099830795262, + "grad_norm": 7.90625, + "learning_rate": 4.89567935397964e-06, + "loss": 2.5209, + "step": 5651 + }, + { + "epoch": 0.5977157360406091, + "grad_norm": 7.78125, + "learning_rate": 4.895630393255651e-06, + "loss": 2.5887, + "step": 5652 + }, + { + "epoch": 0.5978214890016921, + "grad_norm": 7.125, + "learning_rate": 4.89558142128995e-06, + "loss": 2.5455, + "step": 5653 + }, + { + "epoch": 0.5979272419627749, + "grad_norm": 7.28125, + "learning_rate": 4.895532438082764e-06, + "loss": 2.4659, + "step": 5654 + }, + { + "epoch": 0.5980329949238579, + "grad_norm": 7.84375, + "learning_rate": 4.8954834436343245e-06, + "loss": 2.5283, + "step": 5655 + }, + { + "epoch": 0.5981387478849408, + "grad_norm": 8.375, + "learning_rate": 4.895434437944862e-06, + "loss": 2.4148, + "step": 5656 + }, + { + "epoch": 0.5982445008460237, + "grad_norm": 7.90625, + "learning_rate": 4.895385421014604e-06, + "loss": 2.3549, + "step": 5657 + }, + { + "epoch": 0.5983502538071066, + "grad_norm": 7.6875, + "learning_rate": 4.8953363928437814e-06, + "loss": 2.5475, + "step": 5658 + }, + { + "epoch": 0.5984560067681896, + "grad_norm": 8.1875, + "learning_rate": 4.895287353432626e-06, + "loss": 2.616, + "step": 5659 + }, + { + "epoch": 0.5985617597292724, + "grad_norm": 8.125, + "learning_rate": 4.895238302781365e-06, + "loss": 2.4415, + "step": 5660 + }, + { + "epoch": 0.5986675126903553, + "grad_norm": 7.75, + "learning_rate": 4.8951892408902326e-06, + "loss": 2.5454, + "step": 5661 + }, + { + "epoch": 0.5987732656514383, + "grad_norm": 8.5, + "learning_rate": 4.895140167759455e-06, + "loss": 2.608, + "step": 5662 + }, + { + "epoch": 0.5988790186125211, + "grad_norm": 7.8125, + "learning_rate": 4.895091083389265e-06, + "loss": 2.7266, + "step": 5663 + }, + { + "epoch": 0.5989847715736041, + "grad_norm": 8.125, + "learning_rate": 4.895041987779892e-06, + "loss": 2.6839, + "step": 5664 + }, + { + "epoch": 0.5990905245346869, + "grad_norm": 7.875, + "learning_rate": 4.894992880931566e-06, + "loss": 2.6265, + "step": 5665 + }, + { + "epoch": 0.5991962774957699, + "grad_norm": 7.90625, + "learning_rate": 4.894943762844519e-06, + "loss": 2.6099, + "step": 5666 + }, + { + "epoch": 0.5993020304568528, + "grad_norm": 7.71875, + "learning_rate": 4.8948946335189795e-06, + "loss": 2.514, + "step": 5667 + }, + { + "epoch": 0.5994077834179357, + "grad_norm": 7.6875, + "learning_rate": 4.894845492955178e-06, + "loss": 2.3214, + "step": 5668 + }, + { + "epoch": 0.5995135363790186, + "grad_norm": 7.875, + "learning_rate": 4.894796341153348e-06, + "loss": 2.5458, + "step": 5669 + }, + { + "epoch": 0.5996192893401016, + "grad_norm": 7.625, + "learning_rate": 4.894747178113717e-06, + "loss": 2.5814, + "step": 5670 + }, + { + "epoch": 0.5997250423011844, + "grad_norm": 7.46875, + "learning_rate": 4.8946980038365174e-06, + "loss": 2.6516, + "step": 5671 + }, + { + "epoch": 0.5998307952622673, + "grad_norm": 8.6875, + "learning_rate": 4.89464881832198e-06, + "loss": 2.8352, + "step": 5672 + }, + { + "epoch": 0.5999365482233503, + "grad_norm": 7.71875, + "learning_rate": 4.894599621570334e-06, + "loss": 2.4324, + "step": 5673 + }, + { + "epoch": 0.6000423011844331, + "grad_norm": 8.4375, + "learning_rate": 4.894550413581813e-06, + "loss": 2.625, + "step": 5674 + }, + { + "epoch": 0.6001480541455161, + "grad_norm": 7.625, + "learning_rate": 4.894501194356644e-06, + "loss": 2.6723, + "step": 5675 + }, + { + "epoch": 0.600253807106599, + "grad_norm": 7.5, + "learning_rate": 4.894451963895061e-06, + "loss": 2.6114, + "step": 5676 + }, + { + "epoch": 0.6003595600676819, + "grad_norm": 7.96875, + "learning_rate": 4.894402722197295e-06, + "loss": 2.5139, + "step": 5677 + }, + { + "epoch": 0.6004653130287648, + "grad_norm": 8.0625, + "learning_rate": 4.894353469263575e-06, + "loss": 2.7695, + "step": 5678 + }, + { + "epoch": 0.6005710659898477, + "grad_norm": 8.0625, + "learning_rate": 4.894304205094135e-06, + "loss": 2.7263, + "step": 5679 + }, + { + "epoch": 0.6006768189509306, + "grad_norm": 8.3125, + "learning_rate": 4.894254929689204e-06, + "loss": 2.565, + "step": 5680 + }, + { + "epoch": 0.6007825719120136, + "grad_norm": 7.8125, + "learning_rate": 4.8942056430490135e-06, + "loss": 2.6907, + "step": 5681 + }, + { + "epoch": 0.6008883248730964, + "grad_norm": 8.75, + "learning_rate": 4.894156345173795e-06, + "loss": 2.6554, + "step": 5682 + }, + { + "epoch": 0.6009940778341794, + "grad_norm": 7.53125, + "learning_rate": 4.89410703606378e-06, + "loss": 2.5589, + "step": 5683 + }, + { + "epoch": 0.6010998307952623, + "grad_norm": 8.375, + "learning_rate": 4.894057715719201e-06, + "loss": 2.0957, + "step": 5684 + }, + { + "epoch": 0.6012055837563451, + "grad_norm": 8.0625, + "learning_rate": 4.8940083841402864e-06, + "loss": 2.6472, + "step": 5685 + }, + { + "epoch": 0.6013113367174281, + "grad_norm": 8.125, + "learning_rate": 4.893959041327271e-06, + "loss": 2.6262, + "step": 5686 + }, + { + "epoch": 0.601417089678511, + "grad_norm": 7.59375, + "learning_rate": 4.893909687280384e-06, + "loss": 2.7304, + "step": 5687 + }, + { + "epoch": 0.6015228426395939, + "grad_norm": 8.125, + "learning_rate": 4.893860321999858e-06, + "loss": 2.4549, + "step": 5688 + }, + { + "epoch": 0.6016285956006768, + "grad_norm": 7.59375, + "learning_rate": 4.893810945485925e-06, + "loss": 2.5628, + "step": 5689 + }, + { + "epoch": 0.6017343485617598, + "grad_norm": 7.71875, + "learning_rate": 4.893761557738817e-06, + "loss": 2.5756, + "step": 5690 + }, + { + "epoch": 0.6018401015228426, + "grad_norm": 8.125, + "learning_rate": 4.893712158758763e-06, + "loss": 2.7104, + "step": 5691 + }, + { + "epoch": 0.6019458544839256, + "grad_norm": 8.0625, + "learning_rate": 4.8936627485459985e-06, + "loss": 2.6542, + "step": 5692 + }, + { + "epoch": 0.6020516074450084, + "grad_norm": 7.8125, + "learning_rate": 4.8936133271007535e-06, + "loss": 2.6958, + "step": 5693 + }, + { + "epoch": 0.6021573604060914, + "grad_norm": 9.0, + "learning_rate": 4.8935638944232595e-06, + "loss": 2.7151, + "step": 5694 + }, + { + "epoch": 0.6022631133671743, + "grad_norm": 7.96875, + "learning_rate": 4.8935144505137496e-06, + "loss": 2.5632, + "step": 5695 + }, + { + "epoch": 0.6023688663282571, + "grad_norm": 7.9375, + "learning_rate": 4.893464995372455e-06, + "loss": 2.6063, + "step": 5696 + }, + { + "epoch": 0.6024746192893401, + "grad_norm": 7.96875, + "learning_rate": 4.893415528999607e-06, + "loss": 2.5879, + "step": 5697 + }, + { + "epoch": 0.602580372250423, + "grad_norm": 7.53125, + "learning_rate": 4.893366051395441e-06, + "loss": 2.6211, + "step": 5698 + }, + { + "epoch": 0.6026861252115059, + "grad_norm": 7.96875, + "learning_rate": 4.893316562560186e-06, + "loss": 2.7402, + "step": 5699 + }, + { + "epoch": 0.6027918781725888, + "grad_norm": 7.75, + "learning_rate": 4.893267062494075e-06, + "loss": 2.6159, + "step": 5700 + }, + { + "epoch": 0.6028976311336718, + "grad_norm": 8.0, + "learning_rate": 4.893217551197341e-06, + "loss": 2.6081, + "step": 5701 + }, + { + "epoch": 0.6030033840947546, + "grad_norm": 7.96875, + "learning_rate": 4.893168028670215e-06, + "loss": 2.6745, + "step": 5702 + }, + { + "epoch": 0.6031091370558376, + "grad_norm": 7.5, + "learning_rate": 4.893118494912931e-06, + "loss": 2.6283, + "step": 5703 + }, + { + "epoch": 0.6032148900169205, + "grad_norm": 8.8125, + "learning_rate": 4.89306894992572e-06, + "loss": 2.6734, + "step": 5704 + }, + { + "epoch": 0.6033206429780034, + "grad_norm": 8.0625, + "learning_rate": 4.893019393708816e-06, + "loss": 2.4064, + "step": 5705 + }, + { + "epoch": 0.6034263959390863, + "grad_norm": 7.8125, + "learning_rate": 4.89296982626245e-06, + "loss": 2.7026, + "step": 5706 + }, + { + "epoch": 0.6035321489001692, + "grad_norm": 8.125, + "learning_rate": 4.892920247586855e-06, + "loss": 2.654, + "step": 5707 + }, + { + "epoch": 0.6036379018612521, + "grad_norm": 7.96875, + "learning_rate": 4.892870657682265e-06, + "loss": 2.6713, + "step": 5708 + }, + { + "epoch": 0.603743654822335, + "grad_norm": 7.6875, + "learning_rate": 4.8928210565489105e-06, + "loss": 2.7106, + "step": 5709 + }, + { + "epoch": 0.6038494077834179, + "grad_norm": 8.0625, + "learning_rate": 4.892771444187027e-06, + "loss": 2.4924, + "step": 5710 + }, + { + "epoch": 0.6039551607445008, + "grad_norm": 8.0625, + "learning_rate": 4.892721820596844e-06, + "loss": 2.6989, + "step": 5711 + }, + { + "epoch": 0.6040609137055838, + "grad_norm": 8.4375, + "learning_rate": 4.892672185778596e-06, + "loss": 2.5107, + "step": 5712 + }, + { + "epoch": 0.6041666666666666, + "grad_norm": 7.53125, + "learning_rate": 4.8926225397325176e-06, + "loss": 2.4568, + "step": 5713 + }, + { + "epoch": 0.6042724196277496, + "grad_norm": 8.3125, + "learning_rate": 4.8925728824588395e-06, + "loss": 2.9691, + "step": 5714 + }, + { + "epoch": 0.6043781725888325, + "grad_norm": 7.75, + "learning_rate": 4.892523213957795e-06, + "loss": 2.623, + "step": 5715 + }, + { + "epoch": 0.6044839255499154, + "grad_norm": 8.0625, + "learning_rate": 4.892473534229618e-06, + "loss": 2.6482, + "step": 5716 + }, + { + "epoch": 0.6045896785109983, + "grad_norm": 8.25, + "learning_rate": 4.8924238432745405e-06, + "loss": 2.7554, + "step": 5717 + }, + { + "epoch": 0.6046954314720813, + "grad_norm": 7.71875, + "learning_rate": 4.892374141092798e-06, + "loss": 2.7086, + "step": 5718 + }, + { + "epoch": 0.6048011844331641, + "grad_norm": 7.90625, + "learning_rate": 4.892324427684621e-06, + "loss": 2.6984, + "step": 5719 + }, + { + "epoch": 0.6049069373942471, + "grad_norm": 8.0, + "learning_rate": 4.8922747030502434e-06, + "loss": 2.7856, + "step": 5720 + }, + { + "epoch": 0.6050126903553299, + "grad_norm": 7.8125, + "learning_rate": 4.892224967189899e-06, + "loss": 2.2882, + "step": 5721 + }, + { + "epoch": 0.6051184433164128, + "grad_norm": 7.71875, + "learning_rate": 4.892175220103822e-06, + "loss": 2.7415, + "step": 5722 + }, + { + "epoch": 0.6052241962774958, + "grad_norm": 7.9375, + "learning_rate": 4.892125461792244e-06, + "loss": 2.6506, + "step": 5723 + }, + { + "epoch": 0.6053299492385786, + "grad_norm": 7.375, + "learning_rate": 4.892075692255401e-06, + "loss": 2.4228, + "step": 5724 + }, + { + "epoch": 0.6054357021996616, + "grad_norm": 8.4375, + "learning_rate": 4.892025911493524e-06, + "loss": 2.6982, + "step": 5725 + }, + { + "epoch": 0.6055414551607445, + "grad_norm": 7.59375, + "learning_rate": 4.891976119506849e-06, + "loss": 2.5845, + "step": 5726 + }, + { + "epoch": 0.6056472081218274, + "grad_norm": 7.5, + "learning_rate": 4.891926316295607e-06, + "loss": 2.7111, + "step": 5727 + }, + { + "epoch": 0.6057529610829103, + "grad_norm": 7.6875, + "learning_rate": 4.891876501860033e-06, + "loss": 2.6581, + "step": 5728 + }, + { + "epoch": 0.6058587140439933, + "grad_norm": 7.6875, + "learning_rate": 4.8918266762003615e-06, + "loss": 2.5473, + "step": 5729 + }, + { + "epoch": 0.6059644670050761, + "grad_norm": 7.15625, + "learning_rate": 4.891776839316827e-06, + "loss": 2.4469, + "step": 5730 + }, + { + "epoch": 0.6060702199661591, + "grad_norm": 8.375, + "learning_rate": 4.89172699120966e-06, + "loss": 2.2543, + "step": 5731 + }, + { + "epoch": 0.606175972927242, + "grad_norm": 7.5625, + "learning_rate": 4.891677131879097e-06, + "loss": 2.5844, + "step": 5732 + }, + { + "epoch": 0.6062817258883249, + "grad_norm": 7.84375, + "learning_rate": 4.89162726132537e-06, + "loss": 2.539, + "step": 5733 + }, + { + "epoch": 0.6063874788494078, + "grad_norm": 7.71875, + "learning_rate": 4.8915773795487165e-06, + "loss": 2.6072, + "step": 5734 + }, + { + "epoch": 0.6064932318104906, + "grad_norm": 7.75, + "learning_rate": 4.891527486549367e-06, + "loss": 2.5315, + "step": 5735 + }, + { + "epoch": 0.6065989847715736, + "grad_norm": 8.0625, + "learning_rate": 4.8914775823275585e-06, + "loss": 2.5546, + "step": 5736 + }, + { + "epoch": 0.6067047377326565, + "grad_norm": 7.75, + "learning_rate": 4.891427666883523e-06, + "loss": 2.7227, + "step": 5737 + }, + { + "epoch": 0.6068104906937394, + "grad_norm": 8.25, + "learning_rate": 4.891377740217496e-06, + "loss": 2.5333, + "step": 5738 + }, + { + "epoch": 0.6069162436548223, + "grad_norm": 8.0, + "learning_rate": 4.891327802329711e-06, + "loss": 2.5974, + "step": 5739 + }, + { + "epoch": 0.6070219966159053, + "grad_norm": 8.375, + "learning_rate": 4.891277853220402e-06, + "loss": 2.7119, + "step": 5740 + }, + { + "epoch": 0.6071277495769881, + "grad_norm": 7.625, + "learning_rate": 4.8912278928898045e-06, + "loss": 2.6428, + "step": 5741 + }, + { + "epoch": 0.6072335025380711, + "grad_norm": 7.25, + "learning_rate": 4.891177921338153e-06, + "loss": 2.4198, + "step": 5742 + }, + { + "epoch": 0.607339255499154, + "grad_norm": 7.5625, + "learning_rate": 4.891127938565681e-06, + "loss": 2.5224, + "step": 5743 + }, + { + "epoch": 0.6074450084602369, + "grad_norm": 8.1875, + "learning_rate": 4.891077944572623e-06, + "loss": 2.7077, + "step": 5744 + }, + { + "epoch": 0.6075507614213198, + "grad_norm": 8.0, + "learning_rate": 4.891027939359215e-06, + "loss": 2.5291, + "step": 5745 + }, + { + "epoch": 0.6076565143824028, + "grad_norm": 8.5, + "learning_rate": 4.8909779229256906e-06, + "loss": 2.5689, + "step": 5746 + }, + { + "epoch": 0.6077622673434856, + "grad_norm": 8.0625, + "learning_rate": 4.890927895272284e-06, + "loss": 2.6917, + "step": 5747 + }, + { + "epoch": 0.6078680203045685, + "grad_norm": 8.625, + "learning_rate": 4.890877856399231e-06, + "loss": 2.6732, + "step": 5748 + }, + { + "epoch": 0.6079737732656514, + "grad_norm": 7.59375, + "learning_rate": 4.890827806306767e-06, + "loss": 2.6376, + "step": 5749 + }, + { + "epoch": 0.6080795262267343, + "grad_norm": 7.71875, + "learning_rate": 4.890777744995124e-06, + "loss": 2.4605, + "step": 5750 + }, + { + "epoch": 0.6081852791878173, + "grad_norm": 7.84375, + "learning_rate": 4.89072767246454e-06, + "loss": 2.6186, + "step": 5751 + }, + { + "epoch": 0.6082910321489001, + "grad_norm": 8.1875, + "learning_rate": 4.890677588715249e-06, + "loss": 2.3595, + "step": 5752 + }, + { + "epoch": 0.6083967851099831, + "grad_norm": 7.5, + "learning_rate": 4.890627493747486e-06, + "loss": 2.5562, + "step": 5753 + }, + { + "epoch": 0.608502538071066, + "grad_norm": 8.0, + "learning_rate": 4.890577387561486e-06, + "loss": 2.5495, + "step": 5754 + }, + { + "epoch": 0.6086082910321489, + "grad_norm": 8.625, + "learning_rate": 4.890527270157484e-06, + "loss": 2.7301, + "step": 5755 + }, + { + "epoch": 0.6087140439932318, + "grad_norm": 8.0625, + "learning_rate": 4.890477141535714e-06, + "loss": 2.3684, + "step": 5756 + }, + { + "epoch": 0.6088197969543148, + "grad_norm": 7.9375, + "learning_rate": 4.8904270016964125e-06, + "loss": 2.7678, + "step": 5757 + }, + { + "epoch": 0.6089255499153976, + "grad_norm": 7.84375, + "learning_rate": 4.890376850639816e-06, + "loss": 2.6055, + "step": 5758 + }, + { + "epoch": 0.6090313028764806, + "grad_norm": 8.6875, + "learning_rate": 4.890326688366159e-06, + "loss": 2.5865, + "step": 5759 + }, + { + "epoch": 0.6091370558375635, + "grad_norm": 7.46875, + "learning_rate": 4.890276514875675e-06, + "loss": 2.6496, + "step": 5760 + }, + { + "epoch": 0.6092428087986463, + "grad_norm": 8.125, + "learning_rate": 4.890226330168601e-06, + "loss": 2.7289, + "step": 5761 + }, + { + "epoch": 0.6093485617597293, + "grad_norm": 8.25, + "learning_rate": 4.890176134245172e-06, + "loss": 2.4114, + "step": 5762 + }, + { + "epoch": 0.6094543147208121, + "grad_norm": 8.1875, + "learning_rate": 4.8901259271056256e-06, + "loss": 2.6833, + "step": 5763 + }, + { + "epoch": 0.6095600676818951, + "grad_norm": 8.25, + "learning_rate": 4.890075708750195e-06, + "loss": 2.6819, + "step": 5764 + }, + { + "epoch": 0.609665820642978, + "grad_norm": 19.75, + "learning_rate": 4.8900254791791165e-06, + "loss": 2.6986, + "step": 5765 + }, + { + "epoch": 0.6097715736040609, + "grad_norm": 8.1875, + "learning_rate": 4.889975238392626e-06, + "loss": 2.7414, + "step": 5766 + }, + { + "epoch": 0.6098773265651438, + "grad_norm": 8.5, + "learning_rate": 4.889924986390959e-06, + "loss": 2.3188, + "step": 5767 + }, + { + "epoch": 0.6099830795262268, + "grad_norm": 8.0625, + "learning_rate": 4.889874723174352e-06, + "loss": 2.6076, + "step": 5768 + }, + { + "epoch": 0.6100888324873096, + "grad_norm": 7.75, + "learning_rate": 4.889824448743039e-06, + "loss": 2.4396, + "step": 5769 + }, + { + "epoch": 0.6101945854483926, + "grad_norm": 7.84375, + "learning_rate": 4.889774163097259e-06, + "loss": 2.6755, + "step": 5770 + }, + { + "epoch": 0.6103003384094755, + "grad_norm": 8.25, + "learning_rate": 4.8897238662372445e-06, + "loss": 2.3931, + "step": 5771 + }, + { + "epoch": 0.6104060913705583, + "grad_norm": 8.1875, + "learning_rate": 4.889673558163235e-06, + "loss": 2.4774, + "step": 5772 + }, + { + "epoch": 0.6105118443316413, + "grad_norm": 8.0625, + "learning_rate": 4.889623238875464e-06, + "loss": 2.6147, + "step": 5773 + }, + { + "epoch": 0.6106175972927242, + "grad_norm": 7.59375, + "learning_rate": 4.8895729083741685e-06, + "loss": 2.5366, + "step": 5774 + }, + { + "epoch": 0.6107233502538071, + "grad_norm": 8.75, + "learning_rate": 4.889522566659584e-06, + "loss": 2.7723, + "step": 5775 + }, + { + "epoch": 0.61082910321489, + "grad_norm": 7.78125, + "learning_rate": 4.8894722137319485e-06, + "loss": 2.6816, + "step": 5776 + }, + { + "epoch": 0.6109348561759729, + "grad_norm": 7.875, + "learning_rate": 4.889421849591497e-06, + "loss": 2.6751, + "step": 5777 + }, + { + "epoch": 0.6110406091370558, + "grad_norm": 8.0625, + "learning_rate": 4.889371474238465e-06, + "loss": 2.514, + "step": 5778 + }, + { + "epoch": 0.6111463620981388, + "grad_norm": 7.375, + "learning_rate": 4.88932108767309e-06, + "loss": 2.4149, + "step": 5779 + }, + { + "epoch": 0.6112521150592216, + "grad_norm": 7.65625, + "learning_rate": 4.88927068989561e-06, + "loss": 2.4343, + "step": 5780 + }, + { + "epoch": 0.6113578680203046, + "grad_norm": 7.5625, + "learning_rate": 4.889220280906258e-06, + "loss": 2.5504, + "step": 5781 + }, + { + "epoch": 0.6114636209813875, + "grad_norm": 7.375, + "learning_rate": 4.889169860705274e-06, + "loss": 2.4288, + "step": 5782 + }, + { + "epoch": 0.6115693739424704, + "grad_norm": 8.75, + "learning_rate": 4.889119429292891e-06, + "loss": 2.9139, + "step": 5783 + }, + { + "epoch": 0.6116751269035533, + "grad_norm": 8.125, + "learning_rate": 4.8890689866693494e-06, + "loss": 2.4349, + "step": 5784 + }, + { + "epoch": 0.6117808798646363, + "grad_norm": 8.125, + "learning_rate": 4.889018532834883e-06, + "loss": 2.7632, + "step": 5785 + }, + { + "epoch": 0.6118866328257191, + "grad_norm": 8.625, + "learning_rate": 4.88896806778973e-06, + "loss": 2.5972, + "step": 5786 + }, + { + "epoch": 0.611992385786802, + "grad_norm": 7.96875, + "learning_rate": 4.888917591534128e-06, + "loss": 2.6842, + "step": 5787 + }, + { + "epoch": 0.612098138747885, + "grad_norm": 8.5, + "learning_rate": 4.8888671040683115e-06, + "loss": 2.3808, + "step": 5788 + }, + { + "epoch": 0.6122038917089678, + "grad_norm": 8.5625, + "learning_rate": 4.888816605392519e-06, + "loss": 2.7889, + "step": 5789 + }, + { + "epoch": 0.6123096446700508, + "grad_norm": 8.5, + "learning_rate": 4.888766095506987e-06, + "loss": 2.8536, + "step": 5790 + }, + { + "epoch": 0.6124153976311336, + "grad_norm": 7.9375, + "learning_rate": 4.888715574411953e-06, + "loss": 2.358, + "step": 5791 + }, + { + "epoch": 0.6125211505922166, + "grad_norm": 7.4375, + "learning_rate": 4.888665042107654e-06, + "loss": 2.5021, + "step": 5792 + }, + { + "epoch": 0.6126269035532995, + "grad_norm": 7.84375, + "learning_rate": 4.8886144985943265e-06, + "loss": 2.5419, + "step": 5793 + }, + { + "epoch": 0.6127326565143824, + "grad_norm": 7.59375, + "learning_rate": 4.888563943872208e-06, + "loss": 2.686, + "step": 5794 + }, + { + "epoch": 0.6128384094754653, + "grad_norm": 7.8125, + "learning_rate": 4.888513377941536e-06, + "loss": 2.4641, + "step": 5795 + }, + { + "epoch": 0.6129441624365483, + "grad_norm": 7.78125, + "learning_rate": 4.888462800802547e-06, + "loss": 2.556, + "step": 5796 + }, + { + "epoch": 0.6130499153976311, + "grad_norm": 7.71875, + "learning_rate": 4.88841221245548e-06, + "loss": 2.7441, + "step": 5797 + }, + { + "epoch": 0.613155668358714, + "grad_norm": 8.1875, + "learning_rate": 4.88836161290057e-06, + "loss": 2.5256, + "step": 5798 + }, + { + "epoch": 0.613261421319797, + "grad_norm": 8.5, + "learning_rate": 4.888311002138056e-06, + "loss": 2.8456, + "step": 5799 + }, + { + "epoch": 0.6133671742808798, + "grad_norm": 7.6875, + "learning_rate": 4.888260380168177e-06, + "loss": 2.685, + "step": 5800 + }, + { + "epoch": 0.6134729272419628, + "grad_norm": 7.6875, + "learning_rate": 4.888209746991167e-06, + "loss": 2.7116, + "step": 5801 + }, + { + "epoch": 0.6135786802030457, + "grad_norm": 8.5625, + "learning_rate": 4.888159102607266e-06, + "loss": 2.3108, + "step": 5802 + }, + { + "epoch": 0.6136844331641286, + "grad_norm": 8.25, + "learning_rate": 4.888108447016711e-06, + "loss": 2.6694, + "step": 5803 + }, + { + "epoch": 0.6137901861252115, + "grad_norm": 7.65625, + "learning_rate": 4.88805778021974e-06, + "loss": 2.4541, + "step": 5804 + }, + { + "epoch": 0.6138959390862944, + "grad_norm": 7.65625, + "learning_rate": 4.88800710221659e-06, + "loss": 2.4327, + "step": 5805 + }, + { + "epoch": 0.6140016920473773, + "grad_norm": 7.15625, + "learning_rate": 4.8879564130074994e-06, + "loss": 2.4439, + "step": 5806 + }, + { + "epoch": 0.6141074450084603, + "grad_norm": 8.4375, + "learning_rate": 4.887905712592707e-06, + "loss": 2.5803, + "step": 5807 + }, + { + "epoch": 0.6142131979695431, + "grad_norm": 7.96875, + "learning_rate": 4.8878550009724485e-06, + "loss": 2.6662, + "step": 5808 + }, + { + "epoch": 0.614318950930626, + "grad_norm": 7.71875, + "learning_rate": 4.887804278146963e-06, + "loss": 2.7295, + "step": 5809 + }, + { + "epoch": 0.614424703891709, + "grad_norm": 8.25, + "learning_rate": 4.887753544116489e-06, + "loss": 2.728, + "step": 5810 + }, + { + "epoch": 0.6145304568527918, + "grad_norm": 7.84375, + "learning_rate": 4.887702798881265e-06, + "loss": 2.683, + "step": 5811 + }, + { + "epoch": 0.6146362098138748, + "grad_norm": 7.875, + "learning_rate": 4.887652042441527e-06, + "loss": 2.739, + "step": 5812 + }, + { + "epoch": 0.6147419627749577, + "grad_norm": 7.53125, + "learning_rate": 4.887601274797515e-06, + "loss": 2.6077, + "step": 5813 + }, + { + "epoch": 0.6148477157360406, + "grad_norm": 7.9375, + "learning_rate": 4.8875504959494664e-06, + "loss": 2.5979, + "step": 5814 + }, + { + "epoch": 0.6149534686971235, + "grad_norm": 7.5625, + "learning_rate": 4.887499705897621e-06, + "loss": 2.7086, + "step": 5815 + }, + { + "epoch": 0.6150592216582065, + "grad_norm": 8.5625, + "learning_rate": 4.887448904642215e-06, + "loss": 2.7532, + "step": 5816 + }, + { + "epoch": 0.6151649746192893, + "grad_norm": 7.75, + "learning_rate": 4.887398092183488e-06, + "loss": 2.5992, + "step": 5817 + }, + { + "epoch": 0.6152707275803723, + "grad_norm": 7.625, + "learning_rate": 4.887347268521678e-06, + "loss": 2.3699, + "step": 5818 + }, + { + "epoch": 0.6153764805414551, + "grad_norm": 7.78125, + "learning_rate": 4.887296433657023e-06, + "loss": 2.5476, + "step": 5819 + }, + { + "epoch": 0.6154822335025381, + "grad_norm": 8.1875, + "learning_rate": 4.8872455875897635e-06, + "loss": 2.6944, + "step": 5820 + }, + { + "epoch": 0.615587986463621, + "grad_norm": 7.84375, + "learning_rate": 4.887194730320136e-06, + "loss": 2.6387, + "step": 5821 + }, + { + "epoch": 0.6156937394247038, + "grad_norm": 8.25, + "learning_rate": 4.8871438618483805e-06, + "loss": 2.7891, + "step": 5822 + }, + { + "epoch": 0.6157994923857868, + "grad_norm": 7.625, + "learning_rate": 4.887092982174734e-06, + "loss": 2.5735, + "step": 5823 + }, + { + "epoch": 0.6159052453468697, + "grad_norm": 8.1875, + "learning_rate": 4.8870420912994374e-06, + "loss": 2.7284, + "step": 5824 + }, + { + "epoch": 0.6160109983079526, + "grad_norm": 7.8125, + "learning_rate": 4.886991189222728e-06, + "loss": 2.582, + "step": 5825 + }, + { + "epoch": 0.6161167512690355, + "grad_norm": 7.53125, + "learning_rate": 4.886940275944846e-06, + "loss": 2.5381, + "step": 5826 + }, + { + "epoch": 0.6162225042301185, + "grad_norm": 8.0625, + "learning_rate": 4.88688935146603e-06, + "loss": 2.4112, + "step": 5827 + }, + { + "epoch": 0.6163282571912013, + "grad_norm": 7.8125, + "learning_rate": 4.886838415786516e-06, + "loss": 2.5453, + "step": 5828 + }, + { + "epoch": 0.6164340101522843, + "grad_norm": 8.0, + "learning_rate": 4.886787468906547e-06, + "loss": 2.6391, + "step": 5829 + }, + { + "epoch": 0.6165397631133672, + "grad_norm": 8.3125, + "learning_rate": 4.886736510826361e-06, + "loss": 2.6279, + "step": 5830 + }, + { + "epoch": 0.6166455160744501, + "grad_norm": 7.6875, + "learning_rate": 4.886685541546197e-06, + "loss": 2.517, + "step": 5831 + }, + { + "epoch": 0.616751269035533, + "grad_norm": 7.46875, + "learning_rate": 4.8866345610662924e-06, + "loss": 2.5838, + "step": 5832 + }, + { + "epoch": 0.6168570219966159, + "grad_norm": 8.0, + "learning_rate": 4.886583569386889e-06, + "loss": 2.616, + "step": 5833 + }, + { + "epoch": 0.6169627749576988, + "grad_norm": 7.71875, + "learning_rate": 4.886532566508224e-06, + "loss": 2.297, + "step": 5834 + }, + { + "epoch": 0.6170685279187818, + "grad_norm": 7.65625, + "learning_rate": 4.886481552430539e-06, + "loss": 2.6032, + "step": 5835 + }, + { + "epoch": 0.6171742808798646, + "grad_norm": 7.75, + "learning_rate": 4.886430527154072e-06, + "loss": 2.6669, + "step": 5836 + }, + { + "epoch": 0.6172800338409475, + "grad_norm": 7.40625, + "learning_rate": 4.8863794906790615e-06, + "loss": 2.5526, + "step": 5837 + }, + { + "epoch": 0.6173857868020305, + "grad_norm": 8.125, + "learning_rate": 4.886328443005749e-06, + "loss": 2.5491, + "step": 5838 + }, + { + "epoch": 0.6174915397631133, + "grad_norm": 8.125, + "learning_rate": 4.886277384134373e-06, + "loss": 2.6312, + "step": 5839 + }, + { + "epoch": 0.6175972927241963, + "grad_norm": 8.75, + "learning_rate": 4.886226314065172e-06, + "loss": 2.6233, + "step": 5840 + }, + { + "epoch": 0.6177030456852792, + "grad_norm": 7.9375, + "learning_rate": 4.886175232798388e-06, + "loss": 2.3362, + "step": 5841 + }, + { + "epoch": 0.6178087986463621, + "grad_norm": 8.3125, + "learning_rate": 4.88612414033426e-06, + "loss": 2.6977, + "step": 5842 + }, + { + "epoch": 0.617914551607445, + "grad_norm": 8.0625, + "learning_rate": 4.886073036673026e-06, + "loss": 2.6513, + "step": 5843 + }, + { + "epoch": 0.618020304568528, + "grad_norm": 7.53125, + "learning_rate": 4.8860219218149275e-06, + "loss": 2.4945, + "step": 5844 + }, + { + "epoch": 0.6181260575296108, + "grad_norm": 7.875, + "learning_rate": 4.885970795760204e-06, + "loss": 2.6812, + "step": 5845 + }, + { + "epoch": 0.6182318104906938, + "grad_norm": 8.8125, + "learning_rate": 4.885919658509096e-06, + "loss": 2.7516, + "step": 5846 + }, + { + "epoch": 0.6183375634517766, + "grad_norm": 7.75, + "learning_rate": 4.885868510061842e-06, + "loss": 2.6029, + "step": 5847 + }, + { + "epoch": 0.6184433164128595, + "grad_norm": 8.3125, + "learning_rate": 4.885817350418684e-06, + "loss": 2.4745, + "step": 5848 + }, + { + "epoch": 0.6185490693739425, + "grad_norm": 8.25, + "learning_rate": 4.885766179579859e-06, + "loss": 2.1951, + "step": 5849 + }, + { + "epoch": 0.6186548223350253, + "grad_norm": 7.59375, + "learning_rate": 4.885714997545611e-06, + "loss": 2.5792, + "step": 5850 + }, + { + "epoch": 0.6187605752961083, + "grad_norm": 7.5, + "learning_rate": 4.885663804316177e-06, + "loss": 2.4836, + "step": 5851 + }, + { + "epoch": 0.6188663282571912, + "grad_norm": 7.8125, + "learning_rate": 4.885612599891799e-06, + "loss": 2.5577, + "step": 5852 + }, + { + "epoch": 0.6189720812182741, + "grad_norm": 7.125, + "learning_rate": 4.885561384272717e-06, + "loss": 2.4831, + "step": 5853 + }, + { + "epoch": 0.619077834179357, + "grad_norm": 7.65625, + "learning_rate": 4.885510157459171e-06, + "loss": 2.696, + "step": 5854 + }, + { + "epoch": 0.61918358714044, + "grad_norm": 7.78125, + "learning_rate": 4.8854589194514014e-06, + "loss": 2.5798, + "step": 5855 + }, + { + "epoch": 0.6192893401015228, + "grad_norm": 8.5625, + "learning_rate": 4.885407670249648e-06, + "loss": 2.776, + "step": 5856 + }, + { + "epoch": 0.6193950930626058, + "grad_norm": 9.1875, + "learning_rate": 4.885356409854153e-06, + "loss": 2.6382, + "step": 5857 + }, + { + "epoch": 0.6195008460236887, + "grad_norm": 8.0625, + "learning_rate": 4.8853051382651555e-06, + "loss": 2.2381, + "step": 5858 + }, + { + "epoch": 0.6196065989847716, + "grad_norm": 8.1875, + "learning_rate": 4.8852538554828965e-06, + "loss": 2.5335, + "step": 5859 + }, + { + "epoch": 0.6197123519458545, + "grad_norm": 9.125, + "learning_rate": 4.885202561507616e-06, + "loss": 2.856, + "step": 5860 + }, + { + "epoch": 0.6198181049069373, + "grad_norm": 8.0625, + "learning_rate": 4.8851512563395565e-06, + "loss": 2.6031, + "step": 5861 + }, + { + "epoch": 0.6199238578680203, + "grad_norm": 10.5625, + "learning_rate": 4.885099939978957e-06, + "loss": 2.5745, + "step": 5862 + }, + { + "epoch": 0.6200296108291032, + "grad_norm": 7.71875, + "learning_rate": 4.88504861242606e-06, + "loss": 2.7041, + "step": 5863 + }, + { + "epoch": 0.6201353637901861, + "grad_norm": 8.5, + "learning_rate": 4.884997273681105e-06, + "loss": 2.5325, + "step": 5864 + }, + { + "epoch": 0.620241116751269, + "grad_norm": 7.875, + "learning_rate": 4.884945923744332e-06, + "loss": 2.6138, + "step": 5865 + }, + { + "epoch": 0.620346869712352, + "grad_norm": 10.25, + "learning_rate": 4.8848945626159835e-06, + "loss": 2.7009, + "step": 5866 + }, + { + "epoch": 0.6204526226734348, + "grad_norm": 8.125, + "learning_rate": 4.884843190296301e-06, + "loss": 2.5787, + "step": 5867 + }, + { + "epoch": 0.6205583756345178, + "grad_norm": 8.375, + "learning_rate": 4.884791806785525e-06, + "loss": 2.5813, + "step": 5868 + }, + { + "epoch": 0.6206641285956007, + "grad_norm": 7.875, + "learning_rate": 4.884740412083896e-06, + "loss": 2.7219, + "step": 5869 + }, + { + "epoch": 0.6207698815566836, + "grad_norm": 8.375, + "learning_rate": 4.884689006191655e-06, + "loss": 2.6768, + "step": 5870 + }, + { + "epoch": 0.6208756345177665, + "grad_norm": 7.875, + "learning_rate": 4.884637589109045e-06, + "loss": 2.4261, + "step": 5871 + }, + { + "epoch": 0.6209813874788495, + "grad_norm": 8.5, + "learning_rate": 4.884586160836304e-06, + "loss": 2.6596, + "step": 5872 + }, + { + "epoch": 0.6210871404399323, + "grad_norm": 7.65625, + "learning_rate": 4.8845347213736775e-06, + "loss": 2.5548, + "step": 5873 + }, + { + "epoch": 0.6211928934010152, + "grad_norm": 7.53125, + "learning_rate": 4.884483270721404e-06, + "loss": 2.5044, + "step": 5874 + }, + { + "epoch": 0.6212986463620981, + "grad_norm": 7.71875, + "learning_rate": 4.884431808879726e-06, + "loss": 2.7193, + "step": 5875 + }, + { + "epoch": 0.621404399323181, + "grad_norm": 7.78125, + "learning_rate": 4.884380335848884e-06, + "loss": 2.4954, + "step": 5876 + }, + { + "epoch": 0.621510152284264, + "grad_norm": 8.9375, + "learning_rate": 4.884328851629121e-06, + "loss": 2.3967, + "step": 5877 + }, + { + "epoch": 0.6216159052453468, + "grad_norm": 8.25, + "learning_rate": 4.884277356220677e-06, + "loss": 2.779, + "step": 5878 + }, + { + "epoch": 0.6217216582064298, + "grad_norm": 7.5625, + "learning_rate": 4.884225849623796e-06, + "loss": 2.4315, + "step": 5879 + }, + { + "epoch": 0.6218274111675127, + "grad_norm": 8.0, + "learning_rate": 4.884174331838717e-06, + "loss": 2.7008, + "step": 5880 + }, + { + "epoch": 0.6219331641285956, + "grad_norm": 8.3125, + "learning_rate": 4.884122802865683e-06, + "loss": 2.5197, + "step": 5881 + }, + { + "epoch": 0.6220389170896785, + "grad_norm": 7.75, + "learning_rate": 4.884071262704936e-06, + "loss": 2.525, + "step": 5882 + }, + { + "epoch": 0.6221446700507615, + "grad_norm": 7.78125, + "learning_rate": 4.884019711356718e-06, + "loss": 2.6357, + "step": 5883 + }, + { + "epoch": 0.6222504230118443, + "grad_norm": 8.125, + "learning_rate": 4.88396814882127e-06, + "loss": 2.8994, + "step": 5884 + }, + { + "epoch": 0.6223561759729273, + "grad_norm": 7.8125, + "learning_rate": 4.883916575098835e-06, + "loss": 2.471, + "step": 5885 + }, + { + "epoch": 0.6224619289340102, + "grad_norm": 8.0625, + "learning_rate": 4.883864990189654e-06, + "loss": 2.7335, + "step": 5886 + }, + { + "epoch": 0.622567681895093, + "grad_norm": 8.5, + "learning_rate": 4.883813394093969e-06, + "loss": 2.7695, + "step": 5887 + }, + { + "epoch": 0.622673434856176, + "grad_norm": 7.8125, + "learning_rate": 4.883761786812023e-06, + "loss": 2.7963, + "step": 5888 + }, + { + "epoch": 0.6227791878172588, + "grad_norm": 7.625, + "learning_rate": 4.883710168344059e-06, + "loss": 2.5571, + "step": 5889 + }, + { + "epoch": 0.6228849407783418, + "grad_norm": 7.9375, + "learning_rate": 4.883658538690317e-06, + "loss": 2.5754, + "step": 5890 + }, + { + "epoch": 0.6229906937394247, + "grad_norm": 7.625, + "learning_rate": 4.883606897851041e-06, + "loss": 2.6163, + "step": 5891 + }, + { + "epoch": 0.6230964467005076, + "grad_norm": 8.3125, + "learning_rate": 4.883555245826472e-06, + "loss": 2.5454, + "step": 5892 + }, + { + "epoch": 0.6232021996615905, + "grad_norm": 7.84375, + "learning_rate": 4.883503582616854e-06, + "loss": 2.5751, + "step": 5893 + }, + { + "epoch": 0.6233079526226735, + "grad_norm": 7.71875, + "learning_rate": 4.883451908222428e-06, + "loss": 2.5905, + "step": 5894 + }, + { + "epoch": 0.6234137055837563, + "grad_norm": 9.3125, + "learning_rate": 4.883400222643437e-06, + "loss": 2.8222, + "step": 5895 + }, + { + "epoch": 0.6235194585448393, + "grad_norm": 8.9375, + "learning_rate": 4.883348525880123e-06, + "loss": 2.6849, + "step": 5896 + }, + { + "epoch": 0.6236252115059222, + "grad_norm": 8.6875, + "learning_rate": 4.88329681793273e-06, + "loss": 2.7098, + "step": 5897 + }, + { + "epoch": 0.623730964467005, + "grad_norm": 8.1875, + "learning_rate": 4.883245098801499e-06, + "loss": 2.606, + "step": 5898 + }, + { + "epoch": 0.623836717428088, + "grad_norm": 8.375, + "learning_rate": 4.8831933684866736e-06, + "loss": 2.7007, + "step": 5899 + }, + { + "epoch": 0.623942470389171, + "grad_norm": 7.8125, + "learning_rate": 4.8831416269884975e-06, + "loss": 2.5369, + "step": 5900 + }, + { + "epoch": 0.6240482233502538, + "grad_norm": 8.5, + "learning_rate": 4.8830898743072114e-06, + "loss": 2.6105, + "step": 5901 + }, + { + "epoch": 0.6241539763113367, + "grad_norm": 8.1875, + "learning_rate": 4.883038110443059e-06, + "loss": 2.4737, + "step": 5902 + }, + { + "epoch": 0.6242597292724196, + "grad_norm": 8.75, + "learning_rate": 4.882986335396283e-06, + "loss": 2.6743, + "step": 5903 + }, + { + "epoch": 0.6243654822335025, + "grad_norm": 7.84375, + "learning_rate": 4.882934549167129e-06, + "loss": 2.3654, + "step": 5904 + }, + { + "epoch": 0.6244712351945855, + "grad_norm": 7.5, + "learning_rate": 4.8828827517558355e-06, + "loss": 2.4972, + "step": 5905 + }, + { + "epoch": 0.6245769881556683, + "grad_norm": 11.75, + "learning_rate": 4.882830943162648e-06, + "loss": 2.5609, + "step": 5906 + }, + { + "epoch": 0.6246827411167513, + "grad_norm": 8.5, + "learning_rate": 4.8827791233878105e-06, + "loss": 2.9111, + "step": 5907 + }, + { + "epoch": 0.6247884940778342, + "grad_norm": 7.5, + "learning_rate": 4.882727292431565e-06, + "loss": 2.7131, + "step": 5908 + }, + { + "epoch": 0.624894247038917, + "grad_norm": 7.84375, + "learning_rate": 4.882675450294155e-06, + "loss": 2.588, + "step": 5909 + }, + { + "epoch": 0.625, + "grad_norm": 7.65625, + "learning_rate": 4.882623596975822e-06, + "loss": 2.6349, + "step": 5910 + }, + { + "epoch": 0.625105752961083, + "grad_norm": 8.25, + "learning_rate": 4.882571732476812e-06, + "loss": 2.3887, + "step": 5911 + }, + { + "epoch": 0.6252115059221658, + "grad_norm": 7.59375, + "learning_rate": 4.882519856797367e-06, + "loss": 2.6818, + "step": 5912 + }, + { + "epoch": 0.6253172588832487, + "grad_norm": 7.96875, + "learning_rate": 4.882467969937731e-06, + "loss": 2.5353, + "step": 5913 + }, + { + "epoch": 0.6254230118443317, + "grad_norm": 7.8125, + "learning_rate": 4.8824160718981476e-06, + "loss": 2.7687, + "step": 5914 + }, + { + "epoch": 0.6255287648054145, + "grad_norm": 7.71875, + "learning_rate": 4.882364162678859e-06, + "loss": 2.7046, + "step": 5915 + }, + { + "epoch": 0.6256345177664975, + "grad_norm": 7.71875, + "learning_rate": 4.88231224228011e-06, + "loss": 2.5695, + "step": 5916 + }, + { + "epoch": 0.6257402707275804, + "grad_norm": 8.0625, + "learning_rate": 4.882260310702145e-06, + "loss": 2.6067, + "step": 5917 + }, + { + "epoch": 0.6258460236886633, + "grad_norm": 7.4375, + "learning_rate": 4.882208367945206e-06, + "loss": 2.5765, + "step": 5918 + }, + { + "epoch": 0.6259517766497462, + "grad_norm": 8.125, + "learning_rate": 4.882156414009537e-06, + "loss": 2.7675, + "step": 5919 + }, + { + "epoch": 0.626057529610829, + "grad_norm": 8.375, + "learning_rate": 4.882104448895382e-06, + "loss": 2.6794, + "step": 5920 + }, + { + "epoch": 0.626163282571912, + "grad_norm": 7.625, + "learning_rate": 4.882052472602986e-06, + "loss": 2.5883, + "step": 5921 + }, + { + "epoch": 0.626269035532995, + "grad_norm": 8.625, + "learning_rate": 4.882000485132591e-06, + "loss": 2.7408, + "step": 5922 + }, + { + "epoch": 0.6263747884940778, + "grad_norm": 8.5625, + "learning_rate": 4.881948486484442e-06, + "loss": 2.7968, + "step": 5923 + }, + { + "epoch": 0.6264805414551607, + "grad_norm": 8.3125, + "learning_rate": 4.881896476658784e-06, + "loss": 2.5878, + "step": 5924 + }, + { + "epoch": 0.6265862944162437, + "grad_norm": 8.0625, + "learning_rate": 4.88184445565586e-06, + "loss": 2.4803, + "step": 5925 + }, + { + "epoch": 0.6266920473773265, + "grad_norm": 7.90625, + "learning_rate": 4.881792423475913e-06, + "loss": 2.4134, + "step": 5926 + }, + { + "epoch": 0.6267978003384095, + "grad_norm": 12.0, + "learning_rate": 4.881740380119189e-06, + "loss": 2.2429, + "step": 5927 + }, + { + "epoch": 0.6269035532994924, + "grad_norm": 8.0, + "learning_rate": 4.881688325585931e-06, + "loss": 2.5033, + "step": 5928 + }, + { + "epoch": 0.6270093062605753, + "grad_norm": 7.75, + "learning_rate": 4.881636259876385e-06, + "loss": 2.5301, + "step": 5929 + }, + { + "epoch": 0.6271150592216582, + "grad_norm": 7.90625, + "learning_rate": 4.881584182990793e-06, + "loss": 2.6663, + "step": 5930 + }, + { + "epoch": 0.6272208121827412, + "grad_norm": 8.3125, + "learning_rate": 4.8815320949294e-06, + "loss": 2.6032, + "step": 5931 + }, + { + "epoch": 0.627326565143824, + "grad_norm": 7.78125, + "learning_rate": 4.881479995692451e-06, + "loss": 2.5759, + "step": 5932 + }, + { + "epoch": 0.627432318104907, + "grad_norm": 7.59375, + "learning_rate": 4.881427885280191e-06, + "loss": 2.5371, + "step": 5933 + }, + { + "epoch": 0.6275380710659898, + "grad_norm": 8.0625, + "learning_rate": 4.8813757636928635e-06, + "loss": 2.7139, + "step": 5934 + }, + { + "epoch": 0.6276438240270727, + "grad_norm": 7.65625, + "learning_rate": 4.8813236309307144e-06, + "loss": 2.3383, + "step": 5935 + }, + { + "epoch": 0.6277495769881557, + "grad_norm": 8.1875, + "learning_rate": 4.881271486993987e-06, + "loss": 2.4516, + "step": 5936 + }, + { + "epoch": 0.6278553299492385, + "grad_norm": 7.28125, + "learning_rate": 4.881219331882926e-06, + "loss": 2.5361, + "step": 5937 + }, + { + "epoch": 0.6279610829103215, + "grad_norm": 8.3125, + "learning_rate": 4.881167165597777e-06, + "loss": 2.4045, + "step": 5938 + }, + { + "epoch": 0.6280668358714044, + "grad_norm": 8.375, + "learning_rate": 4.881114988138784e-06, + "loss": 2.566, + "step": 5939 + }, + { + "epoch": 0.6281725888324873, + "grad_norm": 7.75, + "learning_rate": 4.881062799506192e-06, + "loss": 2.5132, + "step": 5940 + }, + { + "epoch": 0.6282783417935702, + "grad_norm": 8.5625, + "learning_rate": 4.881010599700247e-06, + "loss": 2.5386, + "step": 5941 + }, + { + "epoch": 0.6283840947546532, + "grad_norm": 8.125, + "learning_rate": 4.8809583887211916e-06, + "loss": 2.6527, + "step": 5942 + }, + { + "epoch": 0.628489847715736, + "grad_norm": 7.71875, + "learning_rate": 4.880906166569274e-06, + "loss": 2.7673, + "step": 5943 + }, + { + "epoch": 0.628595600676819, + "grad_norm": 8.25, + "learning_rate": 4.8808539332447356e-06, + "loss": 2.5963, + "step": 5944 + }, + { + "epoch": 0.6287013536379019, + "grad_norm": 8.6875, + "learning_rate": 4.880801688747825e-06, + "loss": 2.591, + "step": 5945 + }, + { + "epoch": 0.6288071065989848, + "grad_norm": 7.96875, + "learning_rate": 4.880749433078785e-06, + "loss": 2.5185, + "step": 5946 + }, + { + "epoch": 0.6289128595600677, + "grad_norm": 7.46875, + "learning_rate": 4.880697166237862e-06, + "loss": 2.5242, + "step": 5947 + }, + { + "epoch": 0.6290186125211505, + "grad_norm": 8.125, + "learning_rate": 4.880644888225301e-06, + "loss": 2.5071, + "step": 5948 + }, + { + "epoch": 0.6291243654822335, + "grad_norm": 12.5625, + "learning_rate": 4.880592599041347e-06, + "loss": 2.5148, + "step": 5949 + }, + { + "epoch": 0.6292301184433164, + "grad_norm": 7.96875, + "learning_rate": 4.880540298686245e-06, + "loss": 2.6454, + "step": 5950 + }, + { + "epoch": 0.6293358714043993, + "grad_norm": 7.625, + "learning_rate": 4.8804879871602415e-06, + "loss": 2.5828, + "step": 5951 + }, + { + "epoch": 0.6294416243654822, + "grad_norm": 8.125, + "learning_rate": 4.880435664463582e-06, + "loss": 2.6268, + "step": 5952 + }, + { + "epoch": 0.6295473773265652, + "grad_norm": 7.5625, + "learning_rate": 4.880383330596511e-06, + "loss": 2.5614, + "step": 5953 + }, + { + "epoch": 0.629653130287648, + "grad_norm": 7.53125, + "learning_rate": 4.8803309855592736e-06, + "loss": 2.6993, + "step": 5954 + }, + { + "epoch": 0.629758883248731, + "grad_norm": 7.65625, + "learning_rate": 4.880278629352118e-06, + "loss": 2.6528, + "step": 5955 + }, + { + "epoch": 0.6298646362098139, + "grad_norm": 8.125, + "learning_rate": 4.880226261975288e-06, + "loss": 2.779, + "step": 5956 + }, + { + "epoch": 0.6299703891708968, + "grad_norm": 7.90625, + "learning_rate": 4.8801738834290295e-06, + "loss": 2.6199, + "step": 5957 + }, + { + "epoch": 0.6300761421319797, + "grad_norm": 7.875, + "learning_rate": 4.880121493713588e-06, + "loss": 2.661, + "step": 5958 + }, + { + "epoch": 0.6301818950930627, + "grad_norm": 8.5625, + "learning_rate": 4.880069092829209e-06, + "loss": 2.5604, + "step": 5959 + }, + { + "epoch": 0.6302876480541455, + "grad_norm": 8.1875, + "learning_rate": 4.880016680776141e-06, + "loss": 2.6282, + "step": 5960 + }, + { + "epoch": 0.6303934010152284, + "grad_norm": 7.84375, + "learning_rate": 4.879964257554627e-06, + "loss": 2.6796, + "step": 5961 + }, + { + "epoch": 0.6304991539763113, + "grad_norm": 7.875, + "learning_rate": 4.879911823164915e-06, + "loss": 2.6045, + "step": 5962 + }, + { + "epoch": 0.6306049069373942, + "grad_norm": 7.9375, + "learning_rate": 4.879859377607249e-06, + "loss": 2.501, + "step": 5963 + }, + { + "epoch": 0.6307106598984772, + "grad_norm": 7.4375, + "learning_rate": 4.8798069208818765e-06, + "loss": 2.5147, + "step": 5964 + }, + { + "epoch": 0.63081641285956, + "grad_norm": 8.0625, + "learning_rate": 4.879754452989044e-06, + "loss": 2.3414, + "step": 5965 + }, + { + "epoch": 0.630922165820643, + "grad_norm": 7.875, + "learning_rate": 4.879701973928996e-06, + "loss": 2.5533, + "step": 5966 + }, + { + "epoch": 0.6310279187817259, + "grad_norm": 7.875, + "learning_rate": 4.879649483701982e-06, + "loss": 2.5514, + "step": 5967 + }, + { + "epoch": 0.6311336717428088, + "grad_norm": 7.59375, + "learning_rate": 4.879596982308245e-06, + "loss": 2.5626, + "step": 5968 + }, + { + "epoch": 0.6312394247038917, + "grad_norm": 7.78125, + "learning_rate": 4.879544469748032e-06, + "loss": 2.5205, + "step": 5969 + }, + { + "epoch": 0.6313451776649747, + "grad_norm": 7.75, + "learning_rate": 4.87949194602159e-06, + "loss": 2.6443, + "step": 5970 + }, + { + "epoch": 0.6314509306260575, + "grad_norm": 7.3125, + "learning_rate": 4.879439411129167e-06, + "loss": 2.5514, + "step": 5971 + }, + { + "epoch": 0.6315566835871405, + "grad_norm": 8.125, + "learning_rate": 4.879386865071006e-06, + "loss": 2.7786, + "step": 5972 + }, + { + "epoch": 0.6316624365482234, + "grad_norm": 7.875, + "learning_rate": 4.879334307847357e-06, + "loss": 2.6718, + "step": 5973 + }, + { + "epoch": 0.6317681895093062, + "grad_norm": 7.40625, + "learning_rate": 4.879281739458465e-06, + "loss": 2.5101, + "step": 5974 + }, + { + "epoch": 0.6318739424703892, + "grad_norm": 7.78125, + "learning_rate": 4.879229159904576e-06, + "loss": 2.4178, + "step": 5975 + }, + { + "epoch": 0.631979695431472, + "grad_norm": 8.375, + "learning_rate": 4.879176569185938e-06, + "loss": 2.5744, + "step": 5976 + }, + { + "epoch": 0.632085448392555, + "grad_norm": 7.4375, + "learning_rate": 4.879123967302798e-06, + "loss": 2.607, + "step": 5977 + }, + { + "epoch": 0.6321912013536379, + "grad_norm": 7.75, + "learning_rate": 4.879071354255402e-06, + "loss": 2.7124, + "step": 5978 + }, + { + "epoch": 0.6322969543147208, + "grad_norm": 7.96875, + "learning_rate": 4.879018730043998e-06, + "loss": 2.6995, + "step": 5979 + }, + { + "epoch": 0.6324027072758037, + "grad_norm": 7.84375, + "learning_rate": 4.87896609466883e-06, + "loss": 2.5807, + "step": 5980 + }, + { + "epoch": 0.6325084602368867, + "grad_norm": 8.0, + "learning_rate": 4.878913448130149e-06, + "loss": 2.6295, + "step": 5981 + }, + { + "epoch": 0.6326142131979695, + "grad_norm": 8.25, + "learning_rate": 4.878860790428199e-06, + "loss": 2.623, + "step": 5982 + }, + { + "epoch": 0.6327199661590525, + "grad_norm": 7.90625, + "learning_rate": 4.8788081215632285e-06, + "loss": 2.6193, + "step": 5983 + }, + { + "epoch": 0.6328257191201354, + "grad_norm": 7.6875, + "learning_rate": 4.878755441535485e-06, + "loss": 2.5838, + "step": 5984 + }, + { + "epoch": 0.6329314720812182, + "grad_norm": 7.34375, + "learning_rate": 4.878702750345215e-06, + "loss": 2.3568, + "step": 5985 + }, + { + "epoch": 0.6330372250423012, + "grad_norm": 8.0625, + "learning_rate": 4.878650047992664e-06, + "loss": 2.5857, + "step": 5986 + }, + { + "epoch": 0.6331429780033841, + "grad_norm": 7.8125, + "learning_rate": 4.878597334478083e-06, + "loss": 2.6037, + "step": 5987 + }, + { + "epoch": 0.633248730964467, + "grad_norm": 7.75, + "learning_rate": 4.878544609801718e-06, + "loss": 2.6205, + "step": 5988 + }, + { + "epoch": 0.6333544839255499, + "grad_norm": 7.5625, + "learning_rate": 4.878491873963815e-06, + "loss": 2.2232, + "step": 5989 + }, + { + "epoch": 0.6334602368866328, + "grad_norm": 8.125, + "learning_rate": 4.878439126964623e-06, + "loss": 2.9181, + "step": 5990 + }, + { + "epoch": 0.6335659898477157, + "grad_norm": 8.3125, + "learning_rate": 4.878386368804388e-06, + "loss": 2.7595, + "step": 5991 + }, + { + "epoch": 0.6336717428087987, + "grad_norm": 7.71875, + "learning_rate": 4.8783335994833585e-06, + "loss": 2.6389, + "step": 5992 + }, + { + "epoch": 0.6337774957698815, + "grad_norm": 8.125, + "learning_rate": 4.8782808190017825e-06, + "loss": 2.7523, + "step": 5993 + }, + { + "epoch": 0.6338832487309645, + "grad_norm": 7.46875, + "learning_rate": 4.878228027359907e-06, + "loss": 2.5269, + "step": 5994 + }, + { + "epoch": 0.6339890016920474, + "grad_norm": 7.46875, + "learning_rate": 4.878175224557981e-06, + "loss": 2.4448, + "step": 5995 + }, + { + "epoch": 0.6340947546531303, + "grad_norm": 8.125, + "learning_rate": 4.878122410596251e-06, + "loss": 2.7386, + "step": 5996 + }, + { + "epoch": 0.6342005076142132, + "grad_norm": 7.96875, + "learning_rate": 4.8780695854749644e-06, + "loss": 2.6093, + "step": 5997 + }, + { + "epoch": 0.6343062605752962, + "grad_norm": 8.0, + "learning_rate": 4.8780167491943706e-06, + "loss": 2.7991, + "step": 5998 + }, + { + "epoch": 0.634412013536379, + "grad_norm": 8.4375, + "learning_rate": 4.8779639017547165e-06, + "loss": 2.5726, + "step": 5999 + }, + { + "epoch": 0.6345177664974619, + "grad_norm": 8.25, + "learning_rate": 4.8779110431562495e-06, + "loss": 2.3789, + "step": 6000 + }, + { + "epoch": 0.6346235194585449, + "grad_norm": 7.25, + "learning_rate": 4.8778581733992195e-06, + "loss": 2.5923, + "step": 6001 + }, + { + "epoch": 0.6347292724196277, + "grad_norm": 7.34375, + "learning_rate": 4.877805292483874e-06, + "loss": 2.4735, + "step": 6002 + }, + { + "epoch": 0.6348350253807107, + "grad_norm": 7.8125, + "learning_rate": 4.87775240041046e-06, + "loss": 2.5164, + "step": 6003 + }, + { + "epoch": 0.6349407783417935, + "grad_norm": 7.375, + "learning_rate": 4.877699497179226e-06, + "loss": 2.3879, + "step": 6004 + }, + { + "epoch": 0.6350465313028765, + "grad_norm": 7.96875, + "learning_rate": 4.877646582790422e-06, + "loss": 2.6165, + "step": 6005 + }, + { + "epoch": 0.6351522842639594, + "grad_norm": 7.71875, + "learning_rate": 4.8775936572442935e-06, + "loss": 2.6081, + "step": 6006 + }, + { + "epoch": 0.6352580372250423, + "grad_norm": 8.0, + "learning_rate": 4.877540720541092e-06, + "loss": 2.5822, + "step": 6007 + }, + { + "epoch": 0.6353637901861252, + "grad_norm": 7.84375, + "learning_rate": 4.877487772681063e-06, + "loss": 2.4918, + "step": 6008 + }, + { + "epoch": 0.6354695431472082, + "grad_norm": 7.96875, + "learning_rate": 4.877434813664457e-06, + "loss": 2.5243, + "step": 6009 + }, + { + "epoch": 0.635575296108291, + "grad_norm": 9.375, + "learning_rate": 4.877381843491521e-06, + "loss": 2.5992, + "step": 6010 + }, + { + "epoch": 0.635681049069374, + "grad_norm": 7.78125, + "learning_rate": 4.877328862162505e-06, + "loss": 2.4344, + "step": 6011 + }, + { + "epoch": 0.6357868020304569, + "grad_norm": 7.96875, + "learning_rate": 4.877275869677657e-06, + "loss": 2.6643, + "step": 6012 + }, + { + "epoch": 0.6358925549915397, + "grad_norm": 7.6875, + "learning_rate": 4.877222866037226e-06, + "loss": 2.601, + "step": 6013 + }, + { + "epoch": 0.6359983079526227, + "grad_norm": 8.0625, + "learning_rate": 4.877169851241459e-06, + "loss": 2.5743, + "step": 6014 + }, + { + "epoch": 0.6361040609137056, + "grad_norm": 7.65625, + "learning_rate": 4.877116825290607e-06, + "loss": 2.6698, + "step": 6015 + }, + { + "epoch": 0.6362098138747885, + "grad_norm": 8.25, + "learning_rate": 4.877063788184917e-06, + "loss": 2.7059, + "step": 6016 + }, + { + "epoch": 0.6363155668358714, + "grad_norm": 8.3125, + "learning_rate": 4.87701073992464e-06, + "loss": 2.8481, + "step": 6017 + }, + { + "epoch": 0.6364213197969543, + "grad_norm": 9.125, + "learning_rate": 4.876957680510024e-06, + "loss": 2.5696, + "step": 6018 + }, + { + "epoch": 0.6365270727580372, + "grad_norm": 7.5625, + "learning_rate": 4.876904609941316e-06, + "loss": 2.6649, + "step": 6019 + }, + { + "epoch": 0.6366328257191202, + "grad_norm": 7.90625, + "learning_rate": 4.8768515282187685e-06, + "loss": 2.4442, + "step": 6020 + }, + { + "epoch": 0.636738578680203, + "grad_norm": 8.1875, + "learning_rate": 4.876798435342628e-06, + "loss": 2.6246, + "step": 6021 + }, + { + "epoch": 0.636844331641286, + "grad_norm": 7.53125, + "learning_rate": 4.876745331313145e-06, + "loss": 2.5044, + "step": 6022 + }, + { + "epoch": 0.6369500846023689, + "grad_norm": 7.65625, + "learning_rate": 4.876692216130568e-06, + "loss": 2.5388, + "step": 6023 + }, + { + "epoch": 0.6370558375634517, + "grad_norm": 7.9375, + "learning_rate": 4.876639089795147e-06, + "loss": 2.5802, + "step": 6024 + }, + { + "epoch": 0.6371615905245347, + "grad_norm": 8.5625, + "learning_rate": 4.87658595230713e-06, + "loss": 2.7091, + "step": 6025 + }, + { + "epoch": 0.6372673434856176, + "grad_norm": 8.375, + "learning_rate": 4.876532803666767e-06, + "loss": 2.7803, + "step": 6026 + }, + { + "epoch": 0.6373730964467005, + "grad_norm": 7.6875, + "learning_rate": 4.876479643874308e-06, + "loss": 2.638, + "step": 6027 + }, + { + "epoch": 0.6374788494077834, + "grad_norm": 7.53125, + "learning_rate": 4.8764264729300014e-06, + "loss": 2.6248, + "step": 6028 + }, + { + "epoch": 0.6375846023688664, + "grad_norm": 7.75, + "learning_rate": 4.876373290834098e-06, + "loss": 2.4814, + "step": 6029 + }, + { + "epoch": 0.6376903553299492, + "grad_norm": 7.40625, + "learning_rate": 4.876320097586846e-06, + "loss": 2.4088, + "step": 6030 + }, + { + "epoch": 0.6377961082910322, + "grad_norm": 8.8125, + "learning_rate": 4.876266893188496e-06, + "loss": 2.8618, + "step": 6031 + }, + { + "epoch": 0.637901861252115, + "grad_norm": 7.84375, + "learning_rate": 4.876213677639298e-06, + "loss": 2.7024, + "step": 6032 + }, + { + "epoch": 0.638007614213198, + "grad_norm": 8.3125, + "learning_rate": 4.8761604509395e-06, + "loss": 2.6697, + "step": 6033 + }, + { + "epoch": 0.6381133671742809, + "grad_norm": 7.25, + "learning_rate": 4.876107213089353e-06, + "loss": 2.4868, + "step": 6034 + }, + { + "epoch": 0.6382191201353637, + "grad_norm": 7.75, + "learning_rate": 4.876053964089107e-06, + "loss": 2.7537, + "step": 6035 + }, + { + "epoch": 0.6383248730964467, + "grad_norm": 8.0625, + "learning_rate": 4.8760007039390115e-06, + "loss": 2.7171, + "step": 6036 + }, + { + "epoch": 0.6384306260575296, + "grad_norm": 8.75, + "learning_rate": 4.875947432639316e-06, + "loss": 3.1318, + "step": 6037 + }, + { + "epoch": 0.6385363790186125, + "grad_norm": 7.59375, + "learning_rate": 4.875894150190272e-06, + "loss": 2.5095, + "step": 6038 + }, + { + "epoch": 0.6386421319796954, + "grad_norm": 7.40625, + "learning_rate": 4.875840856592128e-06, + "loss": 2.5126, + "step": 6039 + }, + { + "epoch": 0.6387478849407784, + "grad_norm": 7.84375, + "learning_rate": 4.875787551845133e-06, + "loss": 2.5212, + "step": 6040 + }, + { + "epoch": 0.6388536379018612, + "grad_norm": 8.0625, + "learning_rate": 4.87573423594954e-06, + "loss": 2.4448, + "step": 6041 + }, + { + "epoch": 0.6389593908629442, + "grad_norm": 8.8125, + "learning_rate": 4.875680908905599e-06, + "loss": 2.8822, + "step": 6042 + }, + { + "epoch": 0.6390651438240271, + "grad_norm": 8.4375, + "learning_rate": 4.875627570713558e-06, + "loss": 2.5926, + "step": 6043 + }, + { + "epoch": 0.63917089678511, + "grad_norm": 8.0, + "learning_rate": 4.875574221373668e-06, + "loss": 2.6586, + "step": 6044 + }, + { + "epoch": 0.6392766497461929, + "grad_norm": 8.0625, + "learning_rate": 4.87552086088618e-06, + "loss": 2.7483, + "step": 6045 + }, + { + "epoch": 0.6393824027072758, + "grad_norm": 8.3125, + "learning_rate": 4.875467489251345e-06, + "loss": 2.6046, + "step": 6046 + }, + { + "epoch": 0.6394881556683587, + "grad_norm": 7.875, + "learning_rate": 4.875414106469411e-06, + "loss": 2.5584, + "step": 6047 + }, + { + "epoch": 0.6395939086294417, + "grad_norm": 8.625, + "learning_rate": 4.875360712540631e-06, + "loss": 2.4998, + "step": 6048 + }, + { + "epoch": 0.6396996615905245, + "grad_norm": 7.65625, + "learning_rate": 4.875307307465255e-06, + "loss": 2.6583, + "step": 6049 + }, + { + "epoch": 0.6398054145516074, + "grad_norm": 10.25, + "learning_rate": 4.875253891243533e-06, + "loss": 2.5703, + "step": 6050 + }, + { + "epoch": 0.6399111675126904, + "grad_norm": 8.75, + "learning_rate": 4.875200463875715e-06, + "loss": 2.6611, + "step": 6051 + }, + { + "epoch": 0.6400169204737732, + "grad_norm": 8.3125, + "learning_rate": 4.875147025362054e-06, + "loss": 2.5451, + "step": 6052 + }, + { + "epoch": 0.6401226734348562, + "grad_norm": 7.75, + "learning_rate": 4.8750935757027996e-06, + "loss": 2.589, + "step": 6053 + }, + { + "epoch": 0.6402284263959391, + "grad_norm": 8.4375, + "learning_rate": 4.875040114898201e-06, + "loss": 2.4565, + "step": 6054 + }, + { + "epoch": 0.640334179357022, + "grad_norm": 7.6875, + "learning_rate": 4.874986642948511e-06, + "loss": 2.6601, + "step": 6055 + }, + { + "epoch": 0.6404399323181049, + "grad_norm": 8.9375, + "learning_rate": 4.8749331598539805e-06, + "loss": 2.7901, + "step": 6056 + }, + { + "epoch": 0.6405456852791879, + "grad_norm": 7.9375, + "learning_rate": 4.87487966561486e-06, + "loss": 2.6855, + "step": 6057 + }, + { + "epoch": 0.6406514382402707, + "grad_norm": 7.75, + "learning_rate": 4.8748261602314004e-06, + "loss": 2.4606, + "step": 6058 + }, + { + "epoch": 0.6407571912013537, + "grad_norm": 7.96875, + "learning_rate": 4.8747726437038525e-06, + "loss": 2.5923, + "step": 6059 + }, + { + "epoch": 0.6408629441624365, + "grad_norm": 7.71875, + "learning_rate": 4.874719116032468e-06, + "loss": 2.2544, + "step": 6060 + }, + { + "epoch": 0.6409686971235194, + "grad_norm": 7.3125, + "learning_rate": 4.874665577217499e-06, + "loss": 2.699, + "step": 6061 + }, + { + "epoch": 0.6410744500846024, + "grad_norm": 8.0625, + "learning_rate": 4.874612027259194e-06, + "loss": 2.5979, + "step": 6062 + }, + { + "epoch": 0.6411802030456852, + "grad_norm": 7.875, + "learning_rate": 4.874558466157806e-06, + "loss": 2.6412, + "step": 6063 + }, + { + "epoch": 0.6412859560067682, + "grad_norm": 7.59375, + "learning_rate": 4.874504893913587e-06, + "loss": 2.5058, + "step": 6064 + }, + { + "epoch": 0.6413917089678511, + "grad_norm": 7.78125, + "learning_rate": 4.874451310526788e-06, + "loss": 2.5879, + "step": 6065 + }, + { + "epoch": 0.641497461928934, + "grad_norm": 7.90625, + "learning_rate": 4.87439771599766e-06, + "loss": 2.5206, + "step": 6066 + }, + { + "epoch": 0.6416032148900169, + "grad_norm": 8.1875, + "learning_rate": 4.874344110326454e-06, + "loss": 2.6344, + "step": 6067 + }, + { + "epoch": 0.6417089678510999, + "grad_norm": 8.1875, + "learning_rate": 4.8742904935134225e-06, + "loss": 2.4089, + "step": 6068 + }, + { + "epoch": 0.6418147208121827, + "grad_norm": 7.59375, + "learning_rate": 4.8742368655588165e-06, + "loss": 2.584, + "step": 6069 + }, + { + "epoch": 0.6419204737732657, + "grad_norm": 7.625, + "learning_rate": 4.874183226462889e-06, + "loss": 2.5669, + "step": 6070 + }, + { + "epoch": 0.6420262267343486, + "grad_norm": 7.09375, + "learning_rate": 4.8741295762258895e-06, + "loss": 2.4903, + "step": 6071 + }, + { + "epoch": 0.6421319796954315, + "grad_norm": 8.0625, + "learning_rate": 4.874075914848071e-06, + "loss": 2.4408, + "step": 6072 + }, + { + "epoch": 0.6422377326565144, + "grad_norm": 8.5, + "learning_rate": 4.874022242329686e-06, + "loss": 2.6997, + "step": 6073 + }, + { + "epoch": 0.6423434856175972, + "grad_norm": 8.4375, + "learning_rate": 4.873968558670984e-06, + "loss": 2.733, + "step": 6074 + }, + { + "epoch": 0.6424492385786802, + "grad_norm": 9.0625, + "learning_rate": 4.87391486387222e-06, + "loss": 2.8079, + "step": 6075 + }, + { + "epoch": 0.6425549915397631, + "grad_norm": 7.5625, + "learning_rate": 4.873861157933643e-06, + "loss": 2.5614, + "step": 6076 + }, + { + "epoch": 0.642660744500846, + "grad_norm": 8.8125, + "learning_rate": 4.873807440855508e-06, + "loss": 3.1285, + "step": 6077 + }, + { + "epoch": 0.6427664974619289, + "grad_norm": 7.46875, + "learning_rate": 4.8737537126380645e-06, + "loss": 2.7076, + "step": 6078 + }, + { + "epoch": 0.6428722504230119, + "grad_norm": 7.59375, + "learning_rate": 4.873699973281566e-06, + "loss": 2.4816, + "step": 6079 + }, + { + "epoch": 0.6429780033840947, + "grad_norm": 7.40625, + "learning_rate": 4.873646222786265e-06, + "loss": 2.554, + "step": 6080 + }, + { + "epoch": 0.6430837563451777, + "grad_norm": 9.1875, + "learning_rate": 4.8735924611524124e-06, + "loss": 3.0028, + "step": 6081 + }, + { + "epoch": 0.6431895093062606, + "grad_norm": 7.90625, + "learning_rate": 4.873538688380261e-06, + "loss": 2.6341, + "step": 6082 + }, + { + "epoch": 0.6432952622673435, + "grad_norm": 7.8125, + "learning_rate": 4.873484904470063e-06, + "loss": 2.7073, + "step": 6083 + }, + { + "epoch": 0.6434010152284264, + "grad_norm": 8.5, + "learning_rate": 4.873431109422071e-06, + "loss": 2.6355, + "step": 6084 + }, + { + "epoch": 0.6435067681895094, + "grad_norm": 7.96875, + "learning_rate": 4.873377303236538e-06, + "loss": 2.8222, + "step": 6085 + }, + { + "epoch": 0.6436125211505922, + "grad_norm": 8.25, + "learning_rate": 4.873323485913717e-06, + "loss": 2.666, + "step": 6086 + }, + { + "epoch": 0.6437182741116751, + "grad_norm": 8.1875, + "learning_rate": 4.873269657453858e-06, + "loss": 2.5358, + "step": 6087 + }, + { + "epoch": 0.643824027072758, + "grad_norm": 8.0625, + "learning_rate": 4.8732158178572154e-06, + "loss": 2.6228, + "step": 6088 + }, + { + "epoch": 0.6439297800338409, + "grad_norm": 7.8125, + "learning_rate": 4.873161967124042e-06, + "loss": 2.6895, + "step": 6089 + }, + { + "epoch": 0.6440355329949239, + "grad_norm": 7.65625, + "learning_rate": 4.87310810525459e-06, + "loss": 2.5631, + "step": 6090 + }, + { + "epoch": 0.6441412859560067, + "grad_norm": 8.0625, + "learning_rate": 4.873054232249112e-06, + "loss": 2.6039, + "step": 6091 + }, + { + "epoch": 0.6442470389170897, + "grad_norm": 8.0625, + "learning_rate": 4.873000348107862e-06, + "loss": 2.566, + "step": 6092 + }, + { + "epoch": 0.6443527918781726, + "grad_norm": 8.25, + "learning_rate": 4.872946452831091e-06, + "loss": 2.595, + "step": 6093 + }, + { + "epoch": 0.6444585448392555, + "grad_norm": 7.71875, + "learning_rate": 4.872892546419052e-06, + "loss": 2.5428, + "step": 6094 + }, + { + "epoch": 0.6445642978003384, + "grad_norm": 8.0625, + "learning_rate": 4.872838628871999e-06, + "loss": 2.6292, + "step": 6095 + }, + { + "epoch": 0.6446700507614214, + "grad_norm": 8.9375, + "learning_rate": 4.872784700190186e-06, + "loss": 2.4939, + "step": 6096 + }, + { + "epoch": 0.6447758037225042, + "grad_norm": 7.9375, + "learning_rate": 4.872730760373864e-06, + "loss": 2.7179, + "step": 6097 + }, + { + "epoch": 0.6448815566835872, + "grad_norm": 8.5625, + "learning_rate": 4.872676809423288e-06, + "loss": 2.8658, + "step": 6098 + }, + { + "epoch": 0.6449873096446701, + "grad_norm": 7.9375, + "learning_rate": 4.872622847338709e-06, + "loss": 2.5345, + "step": 6099 + }, + { + "epoch": 0.6450930626057529, + "grad_norm": 7.65625, + "learning_rate": 4.872568874120381e-06, + "loss": 2.469, + "step": 6100 + }, + { + "epoch": 0.6451988155668359, + "grad_norm": 8.0, + "learning_rate": 4.872514889768559e-06, + "loss": 2.6965, + "step": 6101 + }, + { + "epoch": 0.6453045685279187, + "grad_norm": 7.8125, + "learning_rate": 4.872460894283494e-06, + "loss": 2.6166, + "step": 6102 + }, + { + "epoch": 0.6454103214890017, + "grad_norm": 7.96875, + "learning_rate": 4.87240688766544e-06, + "loss": 2.6954, + "step": 6103 + }, + { + "epoch": 0.6455160744500846, + "grad_norm": 7.8125, + "learning_rate": 4.872352869914652e-06, + "loss": 2.5787, + "step": 6104 + }, + { + "epoch": 0.6456218274111675, + "grad_norm": 8.0625, + "learning_rate": 4.872298841031381e-06, + "loss": 2.6645, + "step": 6105 + }, + { + "epoch": 0.6457275803722504, + "grad_norm": 8.5, + "learning_rate": 4.872244801015883e-06, + "loss": 2.3973, + "step": 6106 + }, + { + "epoch": 0.6458333333333334, + "grad_norm": 7.875, + "learning_rate": 4.8721907498684094e-06, + "loss": 2.7977, + "step": 6107 + }, + { + "epoch": 0.6459390862944162, + "grad_norm": 7.78125, + "learning_rate": 4.872136687589215e-06, + "loss": 2.6051, + "step": 6108 + }, + { + "epoch": 0.6460448392554992, + "grad_norm": 7.71875, + "learning_rate": 4.872082614178554e-06, + "loss": 2.4415, + "step": 6109 + }, + { + "epoch": 0.6461505922165821, + "grad_norm": 8.25, + "learning_rate": 4.872028529636679e-06, + "loss": 2.7338, + "step": 6110 + }, + { + "epoch": 0.646256345177665, + "grad_norm": 7.5625, + "learning_rate": 4.871974433963844e-06, + "loss": 2.5604, + "step": 6111 + }, + { + "epoch": 0.6463620981387479, + "grad_norm": 8.0625, + "learning_rate": 4.871920327160303e-06, + "loss": 2.6652, + "step": 6112 + }, + { + "epoch": 0.6464678510998308, + "grad_norm": 8.4375, + "learning_rate": 4.87186620922631e-06, + "loss": 2.6224, + "step": 6113 + }, + { + "epoch": 0.6465736040609137, + "grad_norm": 7.71875, + "learning_rate": 4.871812080162119e-06, + "loss": 2.4052, + "step": 6114 + }, + { + "epoch": 0.6466793570219966, + "grad_norm": 8.3125, + "learning_rate": 4.871757939967984e-06, + "loss": 2.6968, + "step": 6115 + }, + { + "epoch": 0.6467851099830795, + "grad_norm": 7.8125, + "learning_rate": 4.8717037886441585e-06, + "loss": 2.7283, + "step": 6116 + }, + { + "epoch": 0.6468908629441624, + "grad_norm": 8.5, + "learning_rate": 4.871649626190899e-06, + "loss": 2.6921, + "step": 6117 + }, + { + "epoch": 0.6469966159052454, + "grad_norm": 8.0, + "learning_rate": 4.871595452608455e-06, + "loss": 2.6877, + "step": 6118 + }, + { + "epoch": 0.6471023688663282, + "grad_norm": 8.625, + "learning_rate": 4.8715412678970856e-06, + "loss": 2.5322, + "step": 6119 + }, + { + "epoch": 0.6472081218274112, + "grad_norm": 8.0, + "learning_rate": 4.871487072057042e-06, + "loss": 2.732, + "step": 6120 + }, + { + "epoch": 0.6473138747884941, + "grad_norm": 8.5625, + "learning_rate": 4.87143286508858e-06, + "loss": 2.7116, + "step": 6121 + }, + { + "epoch": 0.647419627749577, + "grad_norm": 8.6875, + "learning_rate": 4.871378646991953e-06, + "loss": 2.5612, + "step": 6122 + }, + { + "epoch": 0.6475253807106599, + "grad_norm": 7.875, + "learning_rate": 4.871324417767416e-06, + "loss": 2.538, + "step": 6123 + }, + { + "epoch": 0.6476311336717429, + "grad_norm": 8.6875, + "learning_rate": 4.871270177415224e-06, + "loss": 2.8294, + "step": 6124 + }, + { + "epoch": 0.6477368866328257, + "grad_norm": 8.3125, + "learning_rate": 4.871215925935629e-06, + "loss": 2.5563, + "step": 6125 + }, + { + "epoch": 0.6478426395939086, + "grad_norm": 7.5, + "learning_rate": 4.87116166332889e-06, + "loss": 2.6262, + "step": 6126 + }, + { + "epoch": 0.6479483925549916, + "grad_norm": 9.5, + "learning_rate": 4.871107389595258e-06, + "loss": 2.5421, + "step": 6127 + }, + { + "epoch": 0.6480541455160744, + "grad_norm": 8.9375, + "learning_rate": 4.871053104734988e-06, + "loss": 2.6097, + "step": 6128 + }, + { + "epoch": 0.6481598984771574, + "grad_norm": 7.75, + "learning_rate": 4.870998808748336e-06, + "loss": 2.5336, + "step": 6129 + }, + { + "epoch": 0.6482656514382402, + "grad_norm": 7.5, + "learning_rate": 4.8709445016355565e-06, + "loss": 2.5515, + "step": 6130 + }, + { + "epoch": 0.6483714043993232, + "grad_norm": 9.0, + "learning_rate": 4.870890183396904e-06, + "loss": 2.6036, + "step": 6131 + }, + { + "epoch": 0.6484771573604061, + "grad_norm": 8.625, + "learning_rate": 4.870835854032634e-06, + "loss": 2.8194, + "step": 6132 + }, + { + "epoch": 0.648582910321489, + "grad_norm": 7.28125, + "learning_rate": 4.870781513543e-06, + "loss": 2.6026, + "step": 6133 + }, + { + "epoch": 0.6486886632825719, + "grad_norm": 8.125, + "learning_rate": 4.870727161928259e-06, + "loss": 2.3757, + "step": 6134 + }, + { + "epoch": 0.6487944162436549, + "grad_norm": 8.0, + "learning_rate": 4.870672799188665e-06, + "loss": 2.5461, + "step": 6135 + }, + { + "epoch": 0.6489001692047377, + "grad_norm": 7.8125, + "learning_rate": 4.870618425324472e-06, + "loss": 2.4808, + "step": 6136 + }, + { + "epoch": 0.6490059221658206, + "grad_norm": 7.5, + "learning_rate": 4.870564040335937e-06, + "loss": 2.4925, + "step": 6137 + }, + { + "epoch": 0.6491116751269036, + "grad_norm": 7.5625, + "learning_rate": 4.870509644223314e-06, + "loss": 2.5174, + "step": 6138 + }, + { + "epoch": 0.6492174280879864, + "grad_norm": 7.71875, + "learning_rate": 4.8704552369868595e-06, + "loss": 2.506, + "step": 6139 + }, + { + "epoch": 0.6493231810490694, + "grad_norm": 7.53125, + "learning_rate": 4.8704008186268275e-06, + "loss": 2.8501, + "step": 6140 + }, + { + "epoch": 0.6494289340101523, + "grad_norm": 8.25, + "learning_rate": 4.870346389143474e-06, + "loss": 2.7052, + "step": 6141 + }, + { + "epoch": 0.6495346869712352, + "grad_norm": 7.65625, + "learning_rate": 4.870291948537055e-06, + "loss": 2.4447, + "step": 6142 + }, + { + "epoch": 0.6496404399323181, + "grad_norm": 7.84375, + "learning_rate": 4.870237496807825e-06, + "loss": 2.5789, + "step": 6143 + }, + { + "epoch": 0.649746192893401, + "grad_norm": 7.8125, + "learning_rate": 4.8701830339560395e-06, + "loss": 2.5716, + "step": 6144 + }, + { + "epoch": 0.6498519458544839, + "grad_norm": 7.4375, + "learning_rate": 4.870128559981954e-06, + "loss": 2.664, + "step": 6145 + }, + { + "epoch": 0.6499576988155669, + "grad_norm": 8.3125, + "learning_rate": 4.870074074885825e-06, + "loss": 2.5745, + "step": 6146 + }, + { + "epoch": 0.6500634517766497, + "grad_norm": 8.375, + "learning_rate": 4.870019578667908e-06, + "loss": 2.747, + "step": 6147 + }, + { + "epoch": 0.6501692047377327, + "grad_norm": 7.78125, + "learning_rate": 4.869965071328458e-06, + "loss": 2.756, + "step": 6148 + }, + { + "epoch": 0.6502749576988156, + "grad_norm": 8.25, + "learning_rate": 4.869910552867732e-06, + "loss": 2.5596, + "step": 6149 + }, + { + "epoch": 0.6503807106598984, + "grad_norm": 8.0625, + "learning_rate": 4.869856023285984e-06, + "loss": 2.5787, + "step": 6150 + }, + { + "epoch": 0.6504864636209814, + "grad_norm": 8.0, + "learning_rate": 4.8698014825834715e-06, + "loss": 2.3789, + "step": 6151 + }, + { + "epoch": 0.6505922165820643, + "grad_norm": 7.75, + "learning_rate": 4.869746930760449e-06, + "loss": 2.7352, + "step": 6152 + }, + { + "epoch": 0.6506979695431472, + "grad_norm": 7.75, + "learning_rate": 4.8696923678171734e-06, + "loss": 2.6092, + "step": 6153 + }, + { + "epoch": 0.6508037225042301, + "grad_norm": 7.53125, + "learning_rate": 4.869637793753902e-06, + "loss": 2.6145, + "step": 6154 + }, + { + "epoch": 0.6509094754653131, + "grad_norm": 7.4375, + "learning_rate": 4.869583208570889e-06, + "loss": 2.5886, + "step": 6155 + }, + { + "epoch": 0.6510152284263959, + "grad_norm": 7.65625, + "learning_rate": 4.869528612268391e-06, + "loss": 2.5227, + "step": 6156 + }, + { + "epoch": 0.6511209813874789, + "grad_norm": 8.0, + "learning_rate": 4.8694740048466635e-06, + "loss": 2.6012, + "step": 6157 + }, + { + "epoch": 0.6512267343485617, + "grad_norm": 17.375, + "learning_rate": 4.869419386305965e-06, + "loss": 2.4614, + "step": 6158 + }, + { + "epoch": 0.6513324873096447, + "grad_norm": 8.375, + "learning_rate": 4.8693647566465495e-06, + "loss": 2.79, + "step": 6159 + }, + { + "epoch": 0.6514382402707276, + "grad_norm": 8.3125, + "learning_rate": 4.869310115868675e-06, + "loss": 2.4528, + "step": 6160 + }, + { + "epoch": 0.6515439932318104, + "grad_norm": 7.84375, + "learning_rate": 4.869255463972596e-06, + "loss": 2.4028, + "step": 6161 + }, + { + "epoch": 0.6516497461928934, + "grad_norm": 8.0625, + "learning_rate": 4.869200800958571e-06, + "loss": 2.4469, + "step": 6162 + }, + { + "epoch": 0.6517554991539763, + "grad_norm": 7.53125, + "learning_rate": 4.869146126826856e-06, + "loss": 2.5975, + "step": 6163 + }, + { + "epoch": 0.6518612521150592, + "grad_norm": 7.65625, + "learning_rate": 4.869091441577706e-06, + "loss": 2.5192, + "step": 6164 + }, + { + "epoch": 0.6519670050761421, + "grad_norm": 8.5625, + "learning_rate": 4.86903674521138e-06, + "loss": 2.5259, + "step": 6165 + }, + { + "epoch": 0.6520727580372251, + "grad_norm": 9.125, + "learning_rate": 4.868982037728132e-06, + "loss": 2.9259, + "step": 6166 + }, + { + "epoch": 0.6521785109983079, + "grad_norm": 7.84375, + "learning_rate": 4.868927319128222e-06, + "loss": 2.7741, + "step": 6167 + }, + { + "epoch": 0.6522842639593909, + "grad_norm": 7.6875, + "learning_rate": 4.868872589411903e-06, + "loss": 2.5837, + "step": 6168 + }, + { + "epoch": 0.6523900169204738, + "grad_norm": 20.25, + "learning_rate": 4.868817848579434e-06, + "loss": 2.5941, + "step": 6169 + }, + { + "epoch": 0.6524957698815567, + "grad_norm": 7.96875, + "learning_rate": 4.868763096631072e-06, + "loss": 2.763, + "step": 6170 + }, + { + "epoch": 0.6526015228426396, + "grad_norm": 7.875, + "learning_rate": 4.868708333567075e-06, + "loss": 2.476, + "step": 6171 + }, + { + "epoch": 0.6527072758037225, + "grad_norm": 7.4375, + "learning_rate": 4.868653559387697e-06, + "loss": 2.4899, + "step": 6172 + }, + { + "epoch": 0.6528130287648054, + "grad_norm": 8.625, + "learning_rate": 4.868598774093198e-06, + "loss": 2.5677, + "step": 6173 + }, + { + "epoch": 0.6529187817258884, + "grad_norm": 8.5625, + "learning_rate": 4.868543977683832e-06, + "loss": 2.5164, + "step": 6174 + }, + { + "epoch": 0.6530245346869712, + "grad_norm": 7.5, + "learning_rate": 4.868489170159859e-06, + "loss": 2.5943, + "step": 6175 + }, + { + "epoch": 0.6531302876480541, + "grad_norm": 7.96875, + "learning_rate": 4.868434351521535e-06, + "loss": 2.5003, + "step": 6176 + }, + { + "epoch": 0.6532360406091371, + "grad_norm": 8.1875, + "learning_rate": 4.8683795217691165e-06, + "loss": 2.6222, + "step": 6177 + }, + { + "epoch": 0.6533417935702199, + "grad_norm": 7.9375, + "learning_rate": 4.868324680902862e-06, + "loss": 2.5151, + "step": 6178 + }, + { + "epoch": 0.6534475465313029, + "grad_norm": 7.6875, + "learning_rate": 4.868269828923028e-06, + "loss": 2.6911, + "step": 6179 + }, + { + "epoch": 0.6535532994923858, + "grad_norm": 7.375, + "learning_rate": 4.868214965829873e-06, + "loss": 2.5734, + "step": 6180 + }, + { + "epoch": 0.6536590524534687, + "grad_norm": 9.1875, + "learning_rate": 4.868160091623653e-06, + "loss": 2.6129, + "step": 6181 + }, + { + "epoch": 0.6537648054145516, + "grad_norm": 8.25, + "learning_rate": 4.8681052063046266e-06, + "loss": 2.6101, + "step": 6182 + }, + { + "epoch": 0.6538705583756346, + "grad_norm": 7.96875, + "learning_rate": 4.8680503098730505e-06, + "loss": 2.5174, + "step": 6183 + }, + { + "epoch": 0.6539763113367174, + "grad_norm": 8.0625, + "learning_rate": 4.867995402329183e-06, + "loss": 2.6378, + "step": 6184 + }, + { + "epoch": 0.6540820642978004, + "grad_norm": 8.125, + "learning_rate": 4.867940483673282e-06, + "loss": 2.6294, + "step": 6185 + }, + { + "epoch": 0.6541878172588832, + "grad_norm": 8.25, + "learning_rate": 4.8678855539056044e-06, + "loss": 2.6218, + "step": 6186 + }, + { + "epoch": 0.6542935702199661, + "grad_norm": 7.5, + "learning_rate": 4.867830613026409e-06, + "loss": 2.5188, + "step": 6187 + }, + { + "epoch": 0.6543993231810491, + "grad_norm": 7.625, + "learning_rate": 4.867775661035951e-06, + "loss": 2.4606, + "step": 6188 + }, + { + "epoch": 0.6545050761421319, + "grad_norm": 7.40625, + "learning_rate": 4.8677206979344925e-06, + "loss": 2.6929, + "step": 6189 + }, + { + "epoch": 0.6546108291032149, + "grad_norm": 7.78125, + "learning_rate": 4.867665723722288e-06, + "loss": 2.5019, + "step": 6190 + }, + { + "epoch": 0.6547165820642978, + "grad_norm": 7.5625, + "learning_rate": 4.867610738399597e-06, + "loss": 2.5175, + "step": 6191 + }, + { + "epoch": 0.6548223350253807, + "grad_norm": 8.25, + "learning_rate": 4.867555741966676e-06, + "loss": 2.6567, + "step": 6192 + }, + { + "epoch": 0.6549280879864636, + "grad_norm": 7.71875, + "learning_rate": 4.867500734423786e-06, + "loss": 2.496, + "step": 6193 + }, + { + "epoch": 0.6550338409475466, + "grad_norm": 7.9375, + "learning_rate": 4.867445715771182e-06, + "loss": 2.5278, + "step": 6194 + }, + { + "epoch": 0.6551395939086294, + "grad_norm": 8.25, + "learning_rate": 4.867390686009124e-06, + "loss": 2.6676, + "step": 6195 + }, + { + "epoch": 0.6552453468697124, + "grad_norm": 7.625, + "learning_rate": 4.8673356451378694e-06, + "loss": 2.5875, + "step": 6196 + }, + { + "epoch": 0.6553510998307953, + "grad_norm": 7.6875, + "learning_rate": 4.867280593157677e-06, + "loss": 2.1102, + "step": 6197 + }, + { + "epoch": 0.6554568527918782, + "grad_norm": 8.0, + "learning_rate": 4.867225530068805e-06, + "loss": 2.5503, + "step": 6198 + }, + { + "epoch": 0.6555626057529611, + "grad_norm": 8.375, + "learning_rate": 4.867170455871513e-06, + "loss": 2.6385, + "step": 6199 + }, + { + "epoch": 0.6556683587140439, + "grad_norm": 7.75, + "learning_rate": 4.867115370566057e-06, + "loss": 2.5847, + "step": 6200 + }, + { + "epoch": 0.6557741116751269, + "grad_norm": 8.375, + "learning_rate": 4.867060274152696e-06, + "loss": 2.8953, + "step": 6201 + }, + { + "epoch": 0.6558798646362098, + "grad_norm": 8.375, + "learning_rate": 4.86700516663169e-06, + "loss": 2.8693, + "step": 6202 + }, + { + "epoch": 0.6559856175972927, + "grad_norm": 8.8125, + "learning_rate": 4.866950048003297e-06, + "loss": 2.5954, + "step": 6203 + }, + { + "epoch": 0.6560913705583756, + "grad_norm": 8.0, + "learning_rate": 4.8668949182677746e-06, + "loss": 2.4544, + "step": 6204 + }, + { + "epoch": 0.6561971235194586, + "grad_norm": 7.28125, + "learning_rate": 4.866839777425384e-06, + "loss": 2.4727, + "step": 6205 + }, + { + "epoch": 0.6563028764805414, + "grad_norm": 7.375, + "learning_rate": 4.866784625476381e-06, + "loss": 2.4374, + "step": 6206 + }, + { + "epoch": 0.6564086294416244, + "grad_norm": 7.8125, + "learning_rate": 4.866729462421026e-06, + "loss": 2.7078, + "step": 6207 + }, + { + "epoch": 0.6565143824027073, + "grad_norm": 7.9375, + "learning_rate": 4.866674288259578e-06, + "loss": 2.552, + "step": 6208 + }, + { + "epoch": 0.6566201353637902, + "grad_norm": 8.0, + "learning_rate": 4.866619102992295e-06, + "loss": 2.5812, + "step": 6209 + }, + { + "epoch": 0.6567258883248731, + "grad_norm": 8.625, + "learning_rate": 4.866563906619436e-06, + "loss": 2.6309, + "step": 6210 + }, + { + "epoch": 0.6568316412859561, + "grad_norm": 7.90625, + "learning_rate": 4.866508699141262e-06, + "loss": 2.3925, + "step": 6211 + }, + { + "epoch": 0.6569373942470389, + "grad_norm": 8.0, + "learning_rate": 4.866453480558028e-06, + "loss": 2.4923, + "step": 6212 + }, + { + "epoch": 0.6570431472081218, + "grad_norm": 8.125, + "learning_rate": 4.8663982508699985e-06, + "loss": 2.7162, + "step": 6213 + }, + { + "epoch": 0.6571489001692047, + "grad_norm": 8.25, + "learning_rate": 4.866343010077428e-06, + "loss": 2.6675, + "step": 6214 + }, + { + "epoch": 0.6572546531302876, + "grad_norm": 7.84375, + "learning_rate": 4.866287758180579e-06, + "loss": 2.7, + "step": 6215 + }, + { + "epoch": 0.6573604060913706, + "grad_norm": 7.84375, + "learning_rate": 4.866232495179708e-06, + "loss": 2.6134, + "step": 6216 + }, + { + "epoch": 0.6574661590524534, + "grad_norm": 8.4375, + "learning_rate": 4.866177221075076e-06, + "loss": 2.7422, + "step": 6217 + }, + { + "epoch": 0.6575719120135364, + "grad_norm": 8.75, + "learning_rate": 4.866121935866943e-06, + "loss": 2.6146, + "step": 6218 + }, + { + "epoch": 0.6576776649746193, + "grad_norm": 7.75, + "learning_rate": 4.866066639555566e-06, + "loss": 2.5003, + "step": 6219 + }, + { + "epoch": 0.6577834179357022, + "grad_norm": 8.0625, + "learning_rate": 4.866011332141207e-06, + "loss": 2.5106, + "step": 6220 + }, + { + "epoch": 0.6578891708967851, + "grad_norm": 8.25, + "learning_rate": 4.8659560136241244e-06, + "loss": 2.543, + "step": 6221 + }, + { + "epoch": 0.6579949238578681, + "grad_norm": 8.5625, + "learning_rate": 4.865900684004577e-06, + "loss": 2.6128, + "step": 6222 + }, + { + "epoch": 0.6581006768189509, + "grad_norm": 7.96875, + "learning_rate": 4.865845343282826e-06, + "loss": 2.4729, + "step": 6223 + }, + { + "epoch": 0.6582064297800339, + "grad_norm": 7.875, + "learning_rate": 4.86578999145913e-06, + "loss": 2.5157, + "step": 6224 + }, + { + "epoch": 0.6583121827411168, + "grad_norm": 8.0, + "learning_rate": 4.86573462853375e-06, + "loss": 2.3882, + "step": 6225 + }, + { + "epoch": 0.6584179357021996, + "grad_norm": 7.40625, + "learning_rate": 4.865679254506944e-06, + "loss": 2.5862, + "step": 6226 + }, + { + "epoch": 0.6585236886632826, + "grad_norm": 8.5625, + "learning_rate": 4.865623869378973e-06, + "loss": 2.4148, + "step": 6227 + }, + { + "epoch": 0.6586294416243654, + "grad_norm": 8.4375, + "learning_rate": 4.8655684731500976e-06, + "loss": 2.4802, + "step": 6228 + }, + { + "epoch": 0.6587351945854484, + "grad_norm": 7.90625, + "learning_rate": 4.8655130658205756e-06, + "loss": 2.4245, + "step": 6229 + }, + { + "epoch": 0.6588409475465313, + "grad_norm": 7.8125, + "learning_rate": 4.865457647390669e-06, + "loss": 2.393, + "step": 6230 + }, + { + "epoch": 0.6589467005076142, + "grad_norm": 8.25, + "learning_rate": 4.865402217860636e-06, + "loss": 2.5911, + "step": 6231 + }, + { + "epoch": 0.6590524534686971, + "grad_norm": 7.65625, + "learning_rate": 4.86534677723074e-06, + "loss": 2.5676, + "step": 6232 + }, + { + "epoch": 0.6591582064297801, + "grad_norm": 7.8125, + "learning_rate": 4.865291325501237e-06, + "loss": 2.447, + "step": 6233 + }, + { + "epoch": 0.6592639593908629, + "grad_norm": 7.71875, + "learning_rate": 4.8652358626723895e-06, + "loss": 2.4775, + "step": 6234 + }, + { + "epoch": 0.6593697123519459, + "grad_norm": 7.90625, + "learning_rate": 4.865180388744458e-06, + "loss": 2.7255, + "step": 6235 + }, + { + "epoch": 0.6594754653130288, + "grad_norm": 7.4375, + "learning_rate": 4.865124903717702e-06, + "loss": 2.5377, + "step": 6236 + }, + { + "epoch": 0.6595812182741116, + "grad_norm": 7.75, + "learning_rate": 4.865069407592382e-06, + "loss": 2.629, + "step": 6237 + }, + { + "epoch": 0.6596869712351946, + "grad_norm": 7.78125, + "learning_rate": 4.865013900368759e-06, + "loss": 2.6214, + "step": 6238 + }, + { + "epoch": 0.6597927241962775, + "grad_norm": 8.0, + "learning_rate": 4.8649583820470925e-06, + "loss": 2.8084, + "step": 6239 + }, + { + "epoch": 0.6598984771573604, + "grad_norm": 8.0625, + "learning_rate": 4.864902852627644e-06, + "loss": 2.541, + "step": 6240 + }, + { + "epoch": 0.6600042301184433, + "grad_norm": 7.59375, + "learning_rate": 4.864847312110673e-06, + "loss": 2.5204, + "step": 6241 + }, + { + "epoch": 0.6601099830795262, + "grad_norm": 8.375, + "learning_rate": 4.864791760496441e-06, + "loss": 2.7327, + "step": 6242 + }, + { + "epoch": 0.6602157360406091, + "grad_norm": 7.8125, + "learning_rate": 4.864736197785209e-06, + "loss": 2.4279, + "step": 6243 + }, + { + "epoch": 0.6603214890016921, + "grad_norm": 7.78125, + "learning_rate": 4.864680623977237e-06, + "loss": 2.485, + "step": 6244 + }, + { + "epoch": 0.6604272419627749, + "grad_norm": 7.78125, + "learning_rate": 4.864625039072785e-06, + "loss": 2.3379, + "step": 6245 + }, + { + "epoch": 0.6605329949238579, + "grad_norm": 8.5, + "learning_rate": 4.864569443072116e-06, + "loss": 2.4977, + "step": 6246 + }, + { + "epoch": 0.6606387478849408, + "grad_norm": 8.125, + "learning_rate": 4.864513835975489e-06, + "loss": 2.7549, + "step": 6247 + }, + { + "epoch": 0.6607445008460237, + "grad_norm": 8.125, + "learning_rate": 4.864458217783166e-06, + "loss": 2.6523, + "step": 6248 + }, + { + "epoch": 0.6608502538071066, + "grad_norm": 7.53125, + "learning_rate": 4.864402588495407e-06, + "loss": 2.5197, + "step": 6249 + }, + { + "epoch": 0.6609560067681896, + "grad_norm": 7.9375, + "learning_rate": 4.864346948112474e-06, + "loss": 2.7665, + "step": 6250 + }, + { + "epoch": 0.6610617597292724, + "grad_norm": 7.5625, + "learning_rate": 4.8642912966346275e-06, + "loss": 2.4169, + "step": 6251 + }, + { + "epoch": 0.6611675126903553, + "grad_norm": 7.875, + "learning_rate": 4.86423563406213e-06, + "loss": 2.5907, + "step": 6252 + }, + { + "epoch": 0.6612732656514383, + "grad_norm": 7.59375, + "learning_rate": 4.864179960395241e-06, + "loss": 2.6705, + "step": 6253 + }, + { + "epoch": 0.6613790186125211, + "grad_norm": 7.5, + "learning_rate": 4.8641242756342225e-06, + "loss": 2.5536, + "step": 6254 + }, + { + "epoch": 0.6614847715736041, + "grad_norm": 7.21875, + "learning_rate": 4.864068579779335e-06, + "loss": 2.4975, + "step": 6255 + }, + { + "epoch": 0.6615905245346869, + "grad_norm": 8.875, + "learning_rate": 4.864012872830841e-06, + "loss": 2.594, + "step": 6256 + }, + { + "epoch": 0.6616962774957699, + "grad_norm": 7.40625, + "learning_rate": 4.863957154789001e-06, + "loss": 2.3998, + "step": 6257 + }, + { + "epoch": 0.6618020304568528, + "grad_norm": 7.5, + "learning_rate": 4.863901425654078e-06, + "loss": 2.6269, + "step": 6258 + }, + { + "epoch": 0.6619077834179357, + "grad_norm": 8.25, + "learning_rate": 4.863845685426331e-06, + "loss": 2.7718, + "step": 6259 + }, + { + "epoch": 0.6620135363790186, + "grad_norm": 7.6875, + "learning_rate": 4.863789934106023e-06, + "loss": 2.6016, + "step": 6260 + }, + { + "epoch": 0.6621192893401016, + "grad_norm": 8.3125, + "learning_rate": 4.863734171693417e-06, + "loss": 2.7248, + "step": 6261 + }, + { + "epoch": 0.6622250423011844, + "grad_norm": 7.59375, + "learning_rate": 4.863678398188772e-06, + "loss": 2.6587, + "step": 6262 + }, + { + "epoch": 0.6623307952622673, + "grad_norm": 7.625, + "learning_rate": 4.863622613592351e-06, + "loss": 2.635, + "step": 6263 + }, + { + "epoch": 0.6624365482233503, + "grad_norm": 8.25, + "learning_rate": 4.8635668179044174e-06, + "loss": 2.6027, + "step": 6264 + }, + { + "epoch": 0.6625423011844331, + "grad_norm": 7.71875, + "learning_rate": 4.863511011125229e-06, + "loss": 2.3892, + "step": 6265 + }, + { + "epoch": 0.6626480541455161, + "grad_norm": 8.25, + "learning_rate": 4.863455193255052e-06, + "loss": 2.4011, + "step": 6266 + }, + { + "epoch": 0.662753807106599, + "grad_norm": 7.90625, + "learning_rate": 4.8633993642941454e-06, + "loss": 2.7402, + "step": 6267 + }, + { + "epoch": 0.6628595600676819, + "grad_norm": 8.1875, + "learning_rate": 4.863343524242773e-06, + "loss": 2.491, + "step": 6268 + }, + { + "epoch": 0.6629653130287648, + "grad_norm": 7.5, + "learning_rate": 4.863287673101194e-06, + "loss": 2.4455, + "step": 6269 + }, + { + "epoch": 0.6630710659898477, + "grad_norm": 8.125, + "learning_rate": 4.863231810869675e-06, + "loss": 2.6665, + "step": 6270 + }, + { + "epoch": 0.6631768189509306, + "grad_norm": 7.90625, + "learning_rate": 4.863175937548475e-06, + "loss": 2.4827, + "step": 6271 + }, + { + "epoch": 0.6632825719120136, + "grad_norm": 7.65625, + "learning_rate": 4.863120053137856e-06, + "loss": 2.5036, + "step": 6272 + }, + { + "epoch": 0.6633883248730964, + "grad_norm": 7.71875, + "learning_rate": 4.863064157638081e-06, + "loss": 2.6853, + "step": 6273 + }, + { + "epoch": 0.6634940778341794, + "grad_norm": 7.5625, + "learning_rate": 4.863008251049414e-06, + "loss": 2.6407, + "step": 6274 + }, + { + "epoch": 0.6635998307952623, + "grad_norm": 8.5, + "learning_rate": 4.862952333372114e-06, + "loss": 2.5934, + "step": 6275 + }, + { + "epoch": 0.6637055837563451, + "grad_norm": 7.78125, + "learning_rate": 4.862896404606446e-06, + "loss": 2.6342, + "step": 6276 + }, + { + "epoch": 0.6638113367174281, + "grad_norm": 8.4375, + "learning_rate": 4.8628404647526714e-06, + "loss": 2.5171, + "step": 6277 + }, + { + "epoch": 0.663917089678511, + "grad_norm": 7.9375, + "learning_rate": 4.862784513811052e-06, + "loss": 2.6904, + "step": 6278 + }, + { + "epoch": 0.6640228426395939, + "grad_norm": 8.6875, + "learning_rate": 4.862728551781852e-06, + "loss": 2.3311, + "step": 6279 + }, + { + "epoch": 0.6641285956006768, + "grad_norm": 7.75, + "learning_rate": 4.862672578665334e-06, + "loss": 2.34, + "step": 6280 + }, + { + "epoch": 0.6642343485617598, + "grad_norm": 8.0, + "learning_rate": 4.862616594461759e-06, + "loss": 2.5939, + "step": 6281 + }, + { + "epoch": 0.6643401015228426, + "grad_norm": 7.78125, + "learning_rate": 4.86256059917139e-06, + "loss": 2.6423, + "step": 6282 + }, + { + "epoch": 0.6644458544839256, + "grad_norm": 7.96875, + "learning_rate": 4.862504592794492e-06, + "loss": 2.55, + "step": 6283 + }, + { + "epoch": 0.6645516074450084, + "grad_norm": 8.0625, + "learning_rate": 4.8624485753313245e-06, + "loss": 2.5795, + "step": 6284 + }, + { + "epoch": 0.6646573604060914, + "grad_norm": 7.6875, + "learning_rate": 4.862392546782153e-06, + "loss": 2.5726, + "step": 6285 + }, + { + "epoch": 0.6647631133671743, + "grad_norm": 7.59375, + "learning_rate": 4.862336507147239e-06, + "loss": 2.3974, + "step": 6286 + }, + { + "epoch": 0.6648688663282571, + "grad_norm": 8.125, + "learning_rate": 4.8622804564268464e-06, + "loss": 2.6123, + "step": 6287 + }, + { + "epoch": 0.6649746192893401, + "grad_norm": 7.78125, + "learning_rate": 4.862224394621238e-06, + "loss": 2.6643, + "step": 6288 + }, + { + "epoch": 0.665080372250423, + "grad_norm": 7.71875, + "learning_rate": 4.862168321730676e-06, + "loss": 2.5916, + "step": 6289 + }, + { + "epoch": 0.6651861252115059, + "grad_norm": 8.0, + "learning_rate": 4.8621122377554245e-06, + "loss": 2.5933, + "step": 6290 + }, + { + "epoch": 0.6652918781725888, + "grad_norm": 7.96875, + "learning_rate": 4.862056142695746e-06, + "loss": 2.5983, + "step": 6291 + }, + { + "epoch": 0.6653976311336718, + "grad_norm": 8.0, + "learning_rate": 4.862000036551904e-06, + "loss": 2.6786, + "step": 6292 + }, + { + "epoch": 0.6655033840947546, + "grad_norm": 8.0625, + "learning_rate": 4.861943919324163e-06, + "loss": 2.706, + "step": 6293 + }, + { + "epoch": 0.6656091370558376, + "grad_norm": 7.875, + "learning_rate": 4.861887791012783e-06, + "loss": 2.7153, + "step": 6294 + }, + { + "epoch": 0.6657148900169205, + "grad_norm": 8.4375, + "learning_rate": 4.861831651618031e-06, + "loss": 2.6596, + "step": 6295 + }, + { + "epoch": 0.6658206429780034, + "grad_norm": 8.0, + "learning_rate": 4.861775501140169e-06, + "loss": 2.5892, + "step": 6296 + }, + { + "epoch": 0.6659263959390863, + "grad_norm": 7.59375, + "learning_rate": 4.861719339579461e-06, + "loss": 2.6178, + "step": 6297 + }, + { + "epoch": 0.6660321489001692, + "grad_norm": 7.84375, + "learning_rate": 4.861663166936169e-06, + "loss": 2.5696, + "step": 6298 + }, + { + "epoch": 0.6661379018612521, + "grad_norm": 7.96875, + "learning_rate": 4.861606983210559e-06, + "loss": 2.6354, + "step": 6299 + }, + { + "epoch": 0.666243654822335, + "grad_norm": 7.65625, + "learning_rate": 4.861550788402891e-06, + "loss": 2.4911, + "step": 6300 + }, + { + "epoch": 0.6663494077834179, + "grad_norm": 7.90625, + "learning_rate": 4.861494582513433e-06, + "loss": 2.4515, + "step": 6301 + }, + { + "epoch": 0.6664551607445008, + "grad_norm": 9.1875, + "learning_rate": 4.861438365542446e-06, + "loss": 2.4006, + "step": 6302 + }, + { + "epoch": 0.6665609137055838, + "grad_norm": 8.375, + "learning_rate": 4.8613821374901945e-06, + "loss": 2.5386, + "step": 6303 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 7.875, + "learning_rate": 4.861325898356942e-06, + "loss": 2.6273, + "step": 6304 + }, + { + "epoch": 0.6667724196277496, + "grad_norm": 7.4375, + "learning_rate": 4.861269648142954e-06, + "loss": 2.4763, + "step": 6305 + }, + { + "epoch": 0.6668781725888325, + "grad_norm": 7.875, + "learning_rate": 4.861213386848493e-06, + "loss": 2.5371, + "step": 6306 + }, + { + "epoch": 0.6669839255499154, + "grad_norm": 7.71875, + "learning_rate": 4.861157114473822e-06, + "loss": 2.662, + "step": 6307 + }, + { + "epoch": 0.6670896785109983, + "grad_norm": 7.84375, + "learning_rate": 4.861100831019207e-06, + "loss": 2.6545, + "step": 6308 + }, + { + "epoch": 0.6671954314720813, + "grad_norm": 7.65625, + "learning_rate": 4.861044536484912e-06, + "loss": 2.4912, + "step": 6309 + }, + { + "epoch": 0.6673011844331641, + "grad_norm": 7.9375, + "learning_rate": 4.860988230871201e-06, + "loss": 2.6268, + "step": 6310 + }, + { + "epoch": 0.6674069373942471, + "grad_norm": 8.4375, + "learning_rate": 4.860931914178337e-06, + "loss": 2.6855, + "step": 6311 + }, + { + "epoch": 0.6675126903553299, + "grad_norm": 7.5625, + "learning_rate": 4.860875586406585e-06, + "loss": 2.4486, + "step": 6312 + }, + { + "epoch": 0.6676184433164128, + "grad_norm": 7.5, + "learning_rate": 4.86081924755621e-06, + "loss": 2.3035, + "step": 6313 + }, + { + "epoch": 0.6677241962774958, + "grad_norm": 7.625, + "learning_rate": 4.860762897627475e-06, + "loss": 2.5656, + "step": 6314 + }, + { + "epoch": 0.6678299492385786, + "grad_norm": 8.125, + "learning_rate": 4.860706536620646e-06, + "loss": 2.6691, + "step": 6315 + }, + { + "epoch": 0.6679357021996616, + "grad_norm": 7.59375, + "learning_rate": 4.860650164535986e-06, + "loss": 2.3009, + "step": 6316 + }, + { + "epoch": 0.6680414551607445, + "grad_norm": 8.1875, + "learning_rate": 4.860593781373762e-06, + "loss": 2.712, + "step": 6317 + }, + { + "epoch": 0.6681472081218274, + "grad_norm": 7.4375, + "learning_rate": 4.860537387134235e-06, + "loss": 2.5136, + "step": 6318 + }, + { + "epoch": 0.6682529610829103, + "grad_norm": 7.84375, + "learning_rate": 4.860480981817672e-06, + "loss": 2.4536, + "step": 6319 + }, + { + "epoch": 0.6683587140439933, + "grad_norm": 7.75, + "learning_rate": 4.860424565424337e-06, + "loss": 2.6488, + "step": 6320 + }, + { + "epoch": 0.6684644670050761, + "grad_norm": 7.90625, + "learning_rate": 4.860368137954495e-06, + "loss": 2.5455, + "step": 6321 + }, + { + "epoch": 0.6685702199661591, + "grad_norm": 7.8125, + "learning_rate": 4.8603116994084105e-06, + "loss": 2.5141, + "step": 6322 + }, + { + "epoch": 0.668675972927242, + "grad_norm": 8.0, + "learning_rate": 4.8602552497863494e-06, + "loss": 2.6517, + "step": 6323 + }, + { + "epoch": 0.6687817258883249, + "grad_norm": 8.0, + "learning_rate": 4.860198789088575e-06, + "loss": 2.7295, + "step": 6324 + }, + { + "epoch": 0.6688874788494078, + "grad_norm": 7.96875, + "learning_rate": 4.860142317315352e-06, + "loss": 2.6597, + "step": 6325 + }, + { + "epoch": 0.6689932318104906, + "grad_norm": 7.96875, + "learning_rate": 4.8600858344669484e-06, + "loss": 2.6883, + "step": 6326 + }, + { + "epoch": 0.6690989847715736, + "grad_norm": 8.0, + "learning_rate": 4.860029340543626e-06, + "loss": 2.383, + "step": 6327 + }, + { + "epoch": 0.6692047377326565, + "grad_norm": 8.3125, + "learning_rate": 4.859972835545652e-06, + "loss": 2.6206, + "step": 6328 + }, + { + "epoch": 0.6693104906937394, + "grad_norm": 8.1875, + "learning_rate": 4.85991631947329e-06, + "loss": 2.5243, + "step": 6329 + }, + { + "epoch": 0.6694162436548223, + "grad_norm": 9.5, + "learning_rate": 4.859859792326806e-06, + "loss": 2.4736, + "step": 6330 + }, + { + "epoch": 0.6695219966159053, + "grad_norm": 7.875, + "learning_rate": 4.859803254106466e-06, + "loss": 2.6184, + "step": 6331 + }, + { + "epoch": 0.6696277495769881, + "grad_norm": 8.625, + "learning_rate": 4.859746704812533e-06, + "loss": 2.6324, + "step": 6332 + }, + { + "epoch": 0.6697335025380711, + "grad_norm": 7.875, + "learning_rate": 4.8596901444452734e-06, + "loss": 2.7322, + "step": 6333 + }, + { + "epoch": 0.669839255499154, + "grad_norm": 7.9375, + "learning_rate": 4.859633573004955e-06, + "loss": 2.6871, + "step": 6334 + }, + { + "epoch": 0.6699450084602369, + "grad_norm": 7.9375, + "learning_rate": 4.85957699049184e-06, + "loss": 2.7524, + "step": 6335 + }, + { + "epoch": 0.6700507614213198, + "grad_norm": 8.0, + "learning_rate": 4.859520396906195e-06, + "loss": 2.3402, + "step": 6336 + }, + { + "epoch": 0.6701565143824028, + "grad_norm": 8.5, + "learning_rate": 4.859463792248287e-06, + "loss": 2.6507, + "step": 6337 + }, + { + "epoch": 0.6702622673434856, + "grad_norm": 7.375, + "learning_rate": 4.85940717651838e-06, + "loss": 2.5132, + "step": 6338 + }, + { + "epoch": 0.6703680203045685, + "grad_norm": 7.59375, + "learning_rate": 4.8593505497167395e-06, + "loss": 2.5858, + "step": 6339 + }, + { + "epoch": 0.6704737732656514, + "grad_norm": 8.5, + "learning_rate": 4.859293911843632e-06, + "loss": 2.6621, + "step": 6340 + }, + { + "epoch": 0.6705795262267343, + "grad_norm": 8.625, + "learning_rate": 4.8592372628993235e-06, + "loss": 2.6744, + "step": 6341 + }, + { + "epoch": 0.6706852791878173, + "grad_norm": 7.96875, + "learning_rate": 4.859180602884079e-06, + "loss": 2.5638, + "step": 6342 + }, + { + "epoch": 0.6707910321489001, + "grad_norm": 8.0625, + "learning_rate": 4.859123931798165e-06, + "loss": 2.6501, + "step": 6343 + }, + { + "epoch": 0.6708967851099831, + "grad_norm": 8.125, + "learning_rate": 4.8590672496418476e-06, + "loss": 2.5937, + "step": 6344 + }, + { + "epoch": 0.671002538071066, + "grad_norm": 8.0, + "learning_rate": 4.859010556415392e-06, + "loss": 2.6147, + "step": 6345 + }, + { + "epoch": 0.6711082910321489, + "grad_norm": 8.375, + "learning_rate": 4.858953852119064e-06, + "loss": 2.6379, + "step": 6346 + }, + { + "epoch": 0.6712140439932318, + "grad_norm": 7.46875, + "learning_rate": 4.858897136753131e-06, + "loss": 2.6436, + "step": 6347 + }, + { + "epoch": 0.6713197969543148, + "grad_norm": 7.59375, + "learning_rate": 4.858840410317859e-06, + "loss": 2.5578, + "step": 6348 + }, + { + "epoch": 0.6714255499153976, + "grad_norm": 7.875, + "learning_rate": 4.858783672813513e-06, + "loss": 2.7042, + "step": 6349 + }, + { + "epoch": 0.6715313028764806, + "grad_norm": 7.34375, + "learning_rate": 4.8587269242403604e-06, + "loss": 2.593, + "step": 6350 + }, + { + "epoch": 0.6716370558375635, + "grad_norm": 8.0625, + "learning_rate": 4.858670164598666e-06, + "loss": 2.5698, + "step": 6351 + }, + { + "epoch": 0.6717428087986463, + "grad_norm": 7.71875, + "learning_rate": 4.858613393888699e-06, + "loss": 2.4715, + "step": 6352 + }, + { + "epoch": 0.6718485617597293, + "grad_norm": 8.125, + "learning_rate": 4.8585566121107226e-06, + "loss": 2.5316, + "step": 6353 + }, + { + "epoch": 0.6719543147208121, + "grad_norm": 7.84375, + "learning_rate": 4.858499819265005e-06, + "loss": 2.5874, + "step": 6354 + }, + { + "epoch": 0.6720600676818951, + "grad_norm": 8.5, + "learning_rate": 4.858443015351812e-06, + "loss": 2.6531, + "step": 6355 + }, + { + "epoch": 0.672165820642978, + "grad_norm": 7.5625, + "learning_rate": 4.858386200371411e-06, + "loss": 2.6748, + "step": 6356 + }, + { + "epoch": 0.6722715736040609, + "grad_norm": 7.59375, + "learning_rate": 4.858329374324068e-06, + "loss": 2.438, + "step": 6357 + }, + { + "epoch": 0.6723773265651438, + "grad_norm": 8.125, + "learning_rate": 4.85827253721005e-06, + "loss": 2.6374, + "step": 6358 + }, + { + "epoch": 0.6724830795262268, + "grad_norm": 8.3125, + "learning_rate": 4.858215689029622e-06, + "loss": 2.8387, + "step": 6359 + }, + { + "epoch": 0.6725888324873096, + "grad_norm": 8.0625, + "learning_rate": 4.858158829783054e-06, + "loss": 2.5895, + "step": 6360 + }, + { + "epoch": 0.6726945854483926, + "grad_norm": 7.59375, + "learning_rate": 4.858101959470611e-06, + "loss": 2.5108, + "step": 6361 + }, + { + "epoch": 0.6728003384094755, + "grad_norm": 7.71875, + "learning_rate": 4.858045078092559e-06, + "loss": 2.4681, + "step": 6362 + }, + { + "epoch": 0.6729060913705583, + "grad_norm": 7.90625, + "learning_rate": 4.857988185649166e-06, + "loss": 2.6761, + "step": 6363 + }, + { + "epoch": 0.6730118443316413, + "grad_norm": 8.0625, + "learning_rate": 4.8579312821407e-06, + "loss": 2.5993, + "step": 6364 + }, + { + "epoch": 0.6731175972927242, + "grad_norm": 7.6875, + "learning_rate": 4.857874367567425e-06, + "loss": 2.5836, + "step": 6365 + }, + { + "epoch": 0.6732233502538071, + "grad_norm": 8.25, + "learning_rate": 4.857817441929611e-06, + "loss": 2.4232, + "step": 6366 + }, + { + "epoch": 0.67332910321489, + "grad_norm": 8.0625, + "learning_rate": 4.8577605052275246e-06, + "loss": 2.6848, + "step": 6367 + }, + { + "epoch": 0.6734348561759729, + "grad_norm": 7.75, + "learning_rate": 4.8577035574614315e-06, + "loss": 2.558, + "step": 6368 + }, + { + "epoch": 0.6735406091370558, + "grad_norm": 8.0625, + "learning_rate": 4.8576465986316e-06, + "loss": 2.5217, + "step": 6369 + }, + { + "epoch": 0.6736463620981388, + "grad_norm": 7.625, + "learning_rate": 4.857589628738298e-06, + "loss": 2.5955, + "step": 6370 + }, + { + "epoch": 0.6737521150592216, + "grad_norm": 7.71875, + "learning_rate": 4.857532647781791e-06, + "loss": 2.4631, + "step": 6371 + }, + { + "epoch": 0.6738578680203046, + "grad_norm": 8.0625, + "learning_rate": 4.857475655762348e-06, + "loss": 2.5972, + "step": 6372 + }, + { + "epoch": 0.6739636209813875, + "grad_norm": 8.375, + "learning_rate": 4.857418652680236e-06, + "loss": 2.7386, + "step": 6373 + }, + { + "epoch": 0.6740693739424704, + "grad_norm": 8.0625, + "learning_rate": 4.857361638535722e-06, + "loss": 2.5732, + "step": 6374 + }, + { + "epoch": 0.6741751269035533, + "grad_norm": 7.78125, + "learning_rate": 4.857304613329075e-06, + "loss": 2.5826, + "step": 6375 + }, + { + "epoch": 0.6742808798646363, + "grad_norm": 8.5625, + "learning_rate": 4.857247577060561e-06, + "loss": 2.9471, + "step": 6376 + }, + { + "epoch": 0.6743866328257191, + "grad_norm": 8.1875, + "learning_rate": 4.857190529730448e-06, + "loss": 2.6494, + "step": 6377 + }, + { + "epoch": 0.674492385786802, + "grad_norm": 7.71875, + "learning_rate": 4.857133471339004e-06, + "loss": 2.3709, + "step": 6378 + }, + { + "epoch": 0.674598138747885, + "grad_norm": 7.71875, + "learning_rate": 4.857076401886496e-06, + "loss": 2.3947, + "step": 6379 + }, + { + "epoch": 0.6747038917089678, + "grad_norm": 7.625, + "learning_rate": 4.857019321373193e-06, + "loss": 2.5482, + "step": 6380 + }, + { + "epoch": 0.6748096446700508, + "grad_norm": 7.59375, + "learning_rate": 4.856962229799362e-06, + "loss": 2.5618, + "step": 6381 + }, + { + "epoch": 0.6749153976311336, + "grad_norm": 8.3125, + "learning_rate": 4.856905127165272e-06, + "loss": 2.6534, + "step": 6382 + }, + { + "epoch": 0.6750211505922166, + "grad_norm": 8.25, + "learning_rate": 4.85684801347119e-06, + "loss": 2.8052, + "step": 6383 + }, + { + "epoch": 0.6751269035532995, + "grad_norm": 7.875, + "learning_rate": 4.856790888717384e-06, + "loss": 2.5704, + "step": 6384 + }, + { + "epoch": 0.6752326565143824, + "grad_norm": 8.0, + "learning_rate": 4.856733752904122e-06, + "loss": 2.4973, + "step": 6385 + }, + { + "epoch": 0.6753384094754653, + "grad_norm": 9.25, + "learning_rate": 4.856676606031673e-06, + "loss": 2.9146, + "step": 6386 + }, + { + "epoch": 0.6754441624365483, + "grad_norm": 7.8125, + "learning_rate": 4.8566194481003036e-06, + "loss": 2.6113, + "step": 6387 + }, + { + "epoch": 0.6755499153976311, + "grad_norm": 7.4375, + "learning_rate": 4.856562279110284e-06, + "loss": 2.4173, + "step": 6388 + }, + { + "epoch": 0.675655668358714, + "grad_norm": 7.6875, + "learning_rate": 4.856505099061881e-06, + "loss": 2.6221, + "step": 6389 + }, + { + "epoch": 0.675761421319797, + "grad_norm": 8.3125, + "learning_rate": 4.8564479079553626e-06, + "loss": 2.5886, + "step": 6390 + }, + { + "epoch": 0.6758671742808798, + "grad_norm": 7.9375, + "learning_rate": 4.856390705790999e-06, + "loss": 2.7293, + "step": 6391 + }, + { + "epoch": 0.6759729272419628, + "grad_norm": 7.6875, + "learning_rate": 4.856333492569057e-06, + "loss": 2.5762, + "step": 6392 + }, + { + "epoch": 0.6760786802030457, + "grad_norm": 8.5, + "learning_rate": 4.856276268289806e-06, + "loss": 2.3564, + "step": 6393 + }, + { + "epoch": 0.6761844331641286, + "grad_norm": 8.125, + "learning_rate": 4.856219032953514e-06, + "loss": 2.5491, + "step": 6394 + }, + { + "epoch": 0.6762901861252115, + "grad_norm": 8.125, + "learning_rate": 4.856161786560449e-06, + "loss": 2.5488, + "step": 6395 + }, + { + "epoch": 0.6763959390862944, + "grad_norm": 8.3125, + "learning_rate": 4.8561045291108815e-06, + "loss": 2.7275, + "step": 6396 + }, + { + "epoch": 0.6765016920473773, + "grad_norm": 9.4375, + "learning_rate": 4.8560472606050785e-06, + "loss": 2.6565, + "step": 6397 + }, + { + "epoch": 0.6766074450084603, + "grad_norm": 8.625, + "learning_rate": 4.855989981043309e-06, + "loss": 2.8066, + "step": 6398 + }, + { + "epoch": 0.6767131979695431, + "grad_norm": 7.6875, + "learning_rate": 4.8559326904258434e-06, + "loss": 2.4354, + "step": 6399 + }, + { + "epoch": 0.676818950930626, + "grad_norm": 7.875, + "learning_rate": 4.8558753887529475e-06, + "loss": 2.5783, + "step": 6400 + }, + { + "epoch": 0.676924703891709, + "grad_norm": 7.96875, + "learning_rate": 4.855818076024893e-06, + "loss": 2.6234, + "step": 6401 + }, + { + "epoch": 0.6770304568527918, + "grad_norm": 8.25, + "learning_rate": 4.8557607522419474e-06, + "loss": 2.7431, + "step": 6402 + }, + { + "epoch": 0.6771362098138748, + "grad_norm": 8.4375, + "learning_rate": 4.85570341740438e-06, + "loss": 2.505, + "step": 6403 + }, + { + "epoch": 0.6772419627749577, + "grad_norm": 8.0, + "learning_rate": 4.8556460715124605e-06, + "loss": 2.6482, + "step": 6404 + }, + { + "epoch": 0.6773477157360406, + "grad_norm": 8.4375, + "learning_rate": 4.855588714566457e-06, + "loss": 2.7643, + "step": 6405 + }, + { + "epoch": 0.6774534686971235, + "grad_norm": 8.5625, + "learning_rate": 4.855531346566639e-06, + "loss": 2.3979, + "step": 6406 + }, + { + "epoch": 0.6775592216582065, + "grad_norm": 8.25, + "learning_rate": 4.855473967513275e-06, + "loss": 2.5256, + "step": 6407 + }, + { + "epoch": 0.6776649746192893, + "grad_norm": 9.9375, + "learning_rate": 4.855416577406636e-06, + "loss": 2.4875, + "step": 6408 + }, + { + "epoch": 0.6777707275803723, + "grad_norm": 8.3125, + "learning_rate": 4.855359176246991e-06, + "loss": 2.6637, + "step": 6409 + }, + { + "epoch": 0.6778764805414551, + "grad_norm": 7.8125, + "learning_rate": 4.855301764034607e-06, + "loss": 2.483, + "step": 6410 + }, + { + "epoch": 0.6779822335025381, + "grad_norm": 8.0, + "learning_rate": 4.8552443407697565e-06, + "loss": 2.5613, + "step": 6411 + }, + { + "epoch": 0.678087986463621, + "grad_norm": 7.875, + "learning_rate": 4.855186906452707e-06, + "loss": 2.5883, + "step": 6412 + }, + { + "epoch": 0.6781937394247038, + "grad_norm": 7.78125, + "learning_rate": 4.855129461083729e-06, + "loss": 2.5543, + "step": 6413 + }, + { + "epoch": 0.6782994923857868, + "grad_norm": 8.5, + "learning_rate": 4.85507200466309e-06, + "loss": 2.4573, + "step": 6414 + }, + { + "epoch": 0.6784052453468697, + "grad_norm": 7.6875, + "learning_rate": 4.855014537191063e-06, + "loss": 2.4594, + "step": 6415 + }, + { + "epoch": 0.6785109983079526, + "grad_norm": 7.84375, + "learning_rate": 4.854957058667916e-06, + "loss": 2.6876, + "step": 6416 + }, + { + "epoch": 0.6786167512690355, + "grad_norm": 7.75, + "learning_rate": 4.8548995690939185e-06, + "loss": 2.646, + "step": 6417 + }, + { + "epoch": 0.6787225042301185, + "grad_norm": 7.65625, + "learning_rate": 4.8548420684693395e-06, + "loss": 2.5762, + "step": 6418 + }, + { + "epoch": 0.6788282571912013, + "grad_norm": 9.1875, + "learning_rate": 4.854784556794451e-06, + "loss": 2.379, + "step": 6419 + }, + { + "epoch": 0.6789340101522843, + "grad_norm": 7.9375, + "learning_rate": 4.854727034069521e-06, + "loss": 2.3967, + "step": 6420 + }, + { + "epoch": 0.6790397631133672, + "grad_norm": 7.75, + "learning_rate": 4.854669500294819e-06, + "loss": 2.507, + "step": 6421 + }, + { + "epoch": 0.6791455160744501, + "grad_norm": 7.90625, + "learning_rate": 4.854611955470618e-06, + "loss": 2.4316, + "step": 6422 + }, + { + "epoch": 0.679251269035533, + "grad_norm": 8.0, + "learning_rate": 4.854554399597185e-06, + "loss": 2.6677, + "step": 6423 + }, + { + "epoch": 0.6793570219966159, + "grad_norm": 8.5625, + "learning_rate": 4.8544968326747916e-06, + "loss": 2.8491, + "step": 6424 + }, + { + "epoch": 0.6794627749576988, + "grad_norm": 7.59375, + "learning_rate": 4.854439254703707e-06, + "loss": 2.6666, + "step": 6425 + }, + { + "epoch": 0.6795685279187818, + "grad_norm": 7.8125, + "learning_rate": 4.854381665684202e-06, + "loss": 2.5494, + "step": 6426 + }, + { + "epoch": 0.6796742808798646, + "grad_norm": 7.8125, + "learning_rate": 4.8543240656165474e-06, + "loss": 2.3804, + "step": 6427 + }, + { + "epoch": 0.6797800338409475, + "grad_norm": 8.0625, + "learning_rate": 4.854266454501012e-06, + "loss": 2.4987, + "step": 6428 + }, + { + "epoch": 0.6798857868020305, + "grad_norm": 7.90625, + "learning_rate": 4.854208832337868e-06, + "loss": 2.6277, + "step": 6429 + }, + { + "epoch": 0.6799915397631133, + "grad_norm": 8.0625, + "learning_rate": 4.854151199127384e-06, + "loss": 2.6705, + "step": 6430 + }, + { + "epoch": 0.6800972927241963, + "grad_norm": 7.46875, + "learning_rate": 4.854093554869832e-06, + "loss": 2.5198, + "step": 6431 + }, + { + "epoch": 0.6802030456852792, + "grad_norm": 7.90625, + "learning_rate": 4.854035899565481e-06, + "loss": 2.7093, + "step": 6432 + }, + { + "epoch": 0.6803087986463621, + "grad_norm": 7.53125, + "learning_rate": 4.853978233214602e-06, + "loss": 2.5577, + "step": 6433 + }, + { + "epoch": 0.680414551607445, + "grad_norm": 8.6875, + "learning_rate": 4.853920555817468e-06, + "loss": 2.5314, + "step": 6434 + }, + { + "epoch": 0.680520304568528, + "grad_norm": 7.875, + "learning_rate": 4.853862867374345e-06, + "loss": 2.6122, + "step": 6435 + }, + { + "epoch": 0.6806260575296108, + "grad_norm": 7.8125, + "learning_rate": 4.853805167885507e-06, + "loss": 2.6529, + "step": 6436 + }, + { + "epoch": 0.6807318104906938, + "grad_norm": 7.53125, + "learning_rate": 4.853747457351224e-06, + "loss": 2.414, + "step": 6437 + }, + { + "epoch": 0.6808375634517766, + "grad_norm": 8.25, + "learning_rate": 4.853689735771767e-06, + "loss": 2.6287, + "step": 6438 + }, + { + "epoch": 0.6809433164128595, + "grad_norm": 7.78125, + "learning_rate": 4.853632003147407e-06, + "loss": 2.5004, + "step": 6439 + }, + { + "epoch": 0.6810490693739425, + "grad_norm": 7.8125, + "learning_rate": 4.853574259478415e-06, + "loss": 2.5789, + "step": 6440 + }, + { + "epoch": 0.6811548223350253, + "grad_norm": 7.90625, + "learning_rate": 4.85351650476506e-06, + "loss": 2.6143, + "step": 6441 + }, + { + "epoch": 0.6812605752961083, + "grad_norm": 8.0625, + "learning_rate": 4.853458739007616e-06, + "loss": 2.6043, + "step": 6442 + }, + { + "epoch": 0.6813663282571912, + "grad_norm": 7.96875, + "learning_rate": 4.853400962206352e-06, + "loss": 2.7665, + "step": 6443 + }, + { + "epoch": 0.6814720812182741, + "grad_norm": 7.8125, + "learning_rate": 4.853343174361541e-06, + "loss": 2.6133, + "step": 6444 + }, + { + "epoch": 0.681577834179357, + "grad_norm": 7.875, + "learning_rate": 4.853285375473451e-06, + "loss": 2.622, + "step": 6445 + }, + { + "epoch": 0.68168358714044, + "grad_norm": 7.875, + "learning_rate": 4.853227565542356e-06, + "loss": 2.5535, + "step": 6446 + }, + { + "epoch": 0.6817893401015228, + "grad_norm": 8.0625, + "learning_rate": 4.8531697445685276e-06, + "loss": 2.4928, + "step": 6447 + }, + { + "epoch": 0.6818950930626058, + "grad_norm": 8.3125, + "learning_rate": 4.853111912552234e-06, + "loss": 2.6688, + "step": 6448 + }, + { + "epoch": 0.6820008460236887, + "grad_norm": 8.0625, + "learning_rate": 4.85305406949375e-06, + "loss": 2.5674, + "step": 6449 + }, + { + "epoch": 0.6821065989847716, + "grad_norm": 7.8125, + "learning_rate": 4.8529962153933455e-06, + "loss": 2.5633, + "step": 6450 + }, + { + "epoch": 0.6822123519458545, + "grad_norm": 8.4375, + "learning_rate": 4.852938350251292e-06, + "loss": 2.6658, + "step": 6451 + }, + { + "epoch": 0.6823181049069373, + "grad_norm": 7.8125, + "learning_rate": 4.852880474067862e-06, + "loss": 2.5986, + "step": 6452 + }, + { + "epoch": 0.6824238578680203, + "grad_norm": 8.5625, + "learning_rate": 4.8528225868433246e-06, + "loss": 2.5984, + "step": 6453 + }, + { + "epoch": 0.6825296108291032, + "grad_norm": 8.25, + "learning_rate": 4.852764688577954e-06, + "loss": 2.5275, + "step": 6454 + }, + { + "epoch": 0.6826353637901861, + "grad_norm": 7.84375, + "learning_rate": 4.85270677927202e-06, + "loss": 2.6988, + "step": 6455 + }, + { + "epoch": 0.682741116751269, + "grad_norm": 8.4375, + "learning_rate": 4.852648858925796e-06, + "loss": 2.7895, + "step": 6456 + }, + { + "epoch": 0.682846869712352, + "grad_norm": 8.0, + "learning_rate": 4.852590927539553e-06, + "loss": 2.5648, + "step": 6457 + }, + { + "epoch": 0.6829526226734348, + "grad_norm": 7.75, + "learning_rate": 4.852532985113564e-06, + "loss": 2.4538, + "step": 6458 + }, + { + "epoch": 0.6830583756345178, + "grad_norm": 7.4375, + "learning_rate": 4.852475031648098e-06, + "loss": 2.4586, + "step": 6459 + }, + { + "epoch": 0.6831641285956007, + "grad_norm": 8.1875, + "learning_rate": 4.85241706714343e-06, + "loss": 2.4905, + "step": 6460 + }, + { + "epoch": 0.6832698815566836, + "grad_norm": 7.75, + "learning_rate": 4.85235909159983e-06, + "loss": 2.6367, + "step": 6461 + }, + { + "epoch": 0.6833756345177665, + "grad_norm": 8.5625, + "learning_rate": 4.852301105017572e-06, + "loss": 2.8434, + "step": 6462 + }, + { + "epoch": 0.6834813874788495, + "grad_norm": 8.3125, + "learning_rate": 4.852243107396926e-06, + "loss": 2.4718, + "step": 6463 + }, + { + "epoch": 0.6835871404399323, + "grad_norm": 7.875, + "learning_rate": 4.852185098738166e-06, + "loss": 2.5262, + "step": 6464 + }, + { + "epoch": 0.6836928934010152, + "grad_norm": 8.25, + "learning_rate": 4.852127079041562e-06, + "loss": 2.6246, + "step": 6465 + }, + { + "epoch": 0.6837986463620981, + "grad_norm": 7.53125, + "learning_rate": 4.852069048307388e-06, + "loss": 2.4748, + "step": 6466 + }, + { + "epoch": 0.683904399323181, + "grad_norm": 7.4375, + "learning_rate": 4.8520110065359175e-06, + "loss": 2.717, + "step": 6467 + }, + { + "epoch": 0.684010152284264, + "grad_norm": 8.5625, + "learning_rate": 4.85195295372742e-06, + "loss": 2.7095, + "step": 6468 + }, + { + "epoch": 0.6841159052453468, + "grad_norm": 8.9375, + "learning_rate": 4.851894889882169e-06, + "loss": 2.6507, + "step": 6469 + }, + { + "epoch": 0.6842216582064298, + "grad_norm": 8.3125, + "learning_rate": 4.851836815000438e-06, + "loss": 2.6594, + "step": 6470 + }, + { + "epoch": 0.6843274111675127, + "grad_norm": 8.3125, + "learning_rate": 4.851778729082498e-06, + "loss": 2.8149, + "step": 6471 + }, + { + "epoch": 0.6844331641285956, + "grad_norm": 8.3125, + "learning_rate": 4.851720632128623e-06, + "loss": 2.5681, + "step": 6472 + }, + { + "epoch": 0.6845389170896785, + "grad_norm": 7.5625, + "learning_rate": 4.851662524139083e-06, + "loss": 2.5282, + "step": 6473 + }, + { + "epoch": 0.6846446700507615, + "grad_norm": 7.84375, + "learning_rate": 4.851604405114155e-06, + "loss": 2.625, + "step": 6474 + }, + { + "epoch": 0.6847504230118443, + "grad_norm": 7.6875, + "learning_rate": 4.851546275054108e-06, + "loss": 2.6303, + "step": 6475 + }, + { + "epoch": 0.6848561759729273, + "grad_norm": 7.875, + "learning_rate": 4.851488133959217e-06, + "loss": 2.4648, + "step": 6476 + }, + { + "epoch": 0.6849619289340102, + "grad_norm": 7.4375, + "learning_rate": 4.851429981829753e-06, + "loss": 2.6066, + "step": 6477 + }, + { + "epoch": 0.685067681895093, + "grad_norm": 8.6875, + "learning_rate": 4.85137181866599e-06, + "loss": 2.4006, + "step": 6478 + }, + { + "epoch": 0.685173434856176, + "grad_norm": 7.46875, + "learning_rate": 4.8513136444682e-06, + "loss": 2.7278, + "step": 6479 + }, + { + "epoch": 0.6852791878172588, + "grad_norm": 7.5, + "learning_rate": 4.8512554592366575e-06, + "loss": 2.5154, + "step": 6480 + }, + { + "epoch": 0.6853849407783418, + "grad_norm": 8.4375, + "learning_rate": 4.8511972629716356e-06, + "loss": 2.7244, + "step": 6481 + }, + { + "epoch": 0.6854906937394247, + "grad_norm": 8.125, + "learning_rate": 4.851139055673405e-06, + "loss": 2.5864, + "step": 6482 + }, + { + "epoch": 0.6855964467005076, + "grad_norm": 7.6875, + "learning_rate": 4.851080837342241e-06, + "loss": 2.6737, + "step": 6483 + }, + { + "epoch": 0.6857021996615905, + "grad_norm": 8.0625, + "learning_rate": 4.851022607978416e-06, + "loss": 2.5811, + "step": 6484 + }, + { + "epoch": 0.6858079526226735, + "grad_norm": 8.4375, + "learning_rate": 4.8509643675822045e-06, + "loss": 2.5546, + "step": 6485 + }, + { + "epoch": 0.6859137055837563, + "grad_norm": 7.53125, + "learning_rate": 4.850906116153878e-06, + "loss": 2.5455, + "step": 6486 + }, + { + "epoch": 0.6860194585448393, + "grad_norm": 7.59375, + "learning_rate": 4.850847853693711e-06, + "loss": 2.5105, + "step": 6487 + }, + { + "epoch": 0.6861252115059222, + "grad_norm": 8.375, + "learning_rate": 4.850789580201976e-06, + "loss": 2.5571, + "step": 6488 + }, + { + "epoch": 0.686230964467005, + "grad_norm": 8.4375, + "learning_rate": 4.8507312956789466e-06, + "loss": 2.5946, + "step": 6489 + }, + { + "epoch": 0.686336717428088, + "grad_norm": 8.0, + "learning_rate": 4.850673000124897e-06, + "loss": 2.7493, + "step": 6490 + }, + { + "epoch": 0.686442470389171, + "grad_norm": 7.96875, + "learning_rate": 4.8506146935401e-06, + "loss": 2.6637, + "step": 6491 + }, + { + "epoch": 0.6865482233502538, + "grad_norm": 8.5625, + "learning_rate": 4.85055637592483e-06, + "loss": 2.6306, + "step": 6492 + }, + { + "epoch": 0.6866539763113367, + "grad_norm": 8.1875, + "learning_rate": 4.8504980472793606e-06, + "loss": 2.4899, + "step": 6493 + }, + { + "epoch": 0.6867597292724196, + "grad_norm": 8.0625, + "learning_rate": 4.8504397076039645e-06, + "loss": 2.6224, + "step": 6494 + }, + { + "epoch": 0.6868654822335025, + "grad_norm": 7.96875, + "learning_rate": 4.850381356898917e-06, + "loss": 2.6801, + "step": 6495 + }, + { + "epoch": 0.6869712351945855, + "grad_norm": 8.6875, + "learning_rate": 4.85032299516449e-06, + "loss": 2.7989, + "step": 6496 + }, + { + "epoch": 0.6870769881556683, + "grad_norm": 8.25, + "learning_rate": 4.85026462240096e-06, + "loss": 2.7652, + "step": 6497 + }, + { + "epoch": 0.6871827411167513, + "grad_norm": 7.6875, + "learning_rate": 4.850206238608598e-06, + "loss": 2.7167, + "step": 6498 + }, + { + "epoch": 0.6872884940778342, + "grad_norm": 7.4375, + "learning_rate": 4.850147843787679e-06, + "loss": 2.6981, + "step": 6499 + }, + { + "epoch": 0.687394247038917, + "grad_norm": 7.875, + "learning_rate": 4.850089437938479e-06, + "loss": 2.6097, + "step": 6500 + }, + { + "epoch": 0.6875, + "grad_norm": 7.71875, + "learning_rate": 4.850031021061269e-06, + "loss": 2.5811, + "step": 6501 + }, + { + "epoch": 0.687605752961083, + "grad_norm": 7.96875, + "learning_rate": 4.849972593156325e-06, + "loss": 2.6632, + "step": 6502 + }, + { + "epoch": 0.6877115059221658, + "grad_norm": 7.78125, + "learning_rate": 4.849914154223921e-06, + "loss": 2.7024, + "step": 6503 + }, + { + "epoch": 0.6878172588832487, + "grad_norm": 7.40625, + "learning_rate": 4.84985570426433e-06, + "loss": 2.3774, + "step": 6504 + }, + { + "epoch": 0.6879230118443317, + "grad_norm": 7.8125, + "learning_rate": 4.8497972432778285e-06, + "loss": 2.713, + "step": 6505 + }, + { + "epoch": 0.6880287648054145, + "grad_norm": 7.46875, + "learning_rate": 4.849738771264689e-06, + "loss": 2.5857, + "step": 6506 + }, + { + "epoch": 0.6881345177664975, + "grad_norm": 8.5625, + "learning_rate": 4.849680288225186e-06, + "loss": 2.6104, + "step": 6507 + }, + { + "epoch": 0.6882402707275804, + "grad_norm": 8.4375, + "learning_rate": 4.849621794159596e-06, + "loss": 2.6794, + "step": 6508 + }, + { + "epoch": 0.6883460236886633, + "grad_norm": 8.125, + "learning_rate": 4.84956328906819e-06, + "loss": 2.8178, + "step": 6509 + }, + { + "epoch": 0.6884517766497462, + "grad_norm": 7.4375, + "learning_rate": 4.849504772951246e-06, + "loss": 2.7345, + "step": 6510 + }, + { + "epoch": 0.688557529610829, + "grad_norm": 7.65625, + "learning_rate": 4.8494462458090355e-06, + "loss": 2.5214, + "step": 6511 + }, + { + "epoch": 0.688663282571912, + "grad_norm": 8.0625, + "learning_rate": 4.849387707641835e-06, + "loss": 2.73, + "step": 6512 + }, + { + "epoch": 0.688769035532995, + "grad_norm": 7.625, + "learning_rate": 4.849329158449919e-06, + "loss": 2.4891, + "step": 6513 + }, + { + "epoch": 0.6888747884940778, + "grad_norm": 7.78125, + "learning_rate": 4.849270598233563e-06, + "loss": 2.4966, + "step": 6514 + }, + { + "epoch": 0.6889805414551607, + "grad_norm": 7.75, + "learning_rate": 4.84921202699304e-06, + "loss": 2.4331, + "step": 6515 + }, + { + "epoch": 0.6890862944162437, + "grad_norm": 7.78125, + "learning_rate": 4.8491534447286256e-06, + "loss": 2.6996, + "step": 6516 + }, + { + "epoch": 0.6891920473773265, + "grad_norm": 7.8125, + "learning_rate": 4.849094851440595e-06, + "loss": 2.4261, + "step": 6517 + }, + { + "epoch": 0.6892978003384095, + "grad_norm": 7.78125, + "learning_rate": 4.849036247129224e-06, + "loss": 2.6531, + "step": 6518 + }, + { + "epoch": 0.6894035532994924, + "grad_norm": 8.0625, + "learning_rate": 4.848977631794786e-06, + "loss": 2.5431, + "step": 6519 + }, + { + "epoch": 0.6895093062605753, + "grad_norm": 8.375, + "learning_rate": 4.8489190054375565e-06, + "loss": 2.6833, + "step": 6520 + }, + { + "epoch": 0.6896150592216582, + "grad_norm": 7.5, + "learning_rate": 4.8488603680578105e-06, + "loss": 2.7495, + "step": 6521 + }, + { + "epoch": 0.6897208121827412, + "grad_norm": 8.0, + "learning_rate": 4.848801719655824e-06, + "loss": 2.5226, + "step": 6522 + }, + { + "epoch": 0.689826565143824, + "grad_norm": 8.0625, + "learning_rate": 4.848743060231871e-06, + "loss": 2.5466, + "step": 6523 + }, + { + "epoch": 0.689932318104907, + "grad_norm": 7.71875, + "learning_rate": 4.848684389786228e-06, + "loss": 2.4233, + "step": 6524 + }, + { + "epoch": 0.6900380710659898, + "grad_norm": 7.8125, + "learning_rate": 4.84862570831917e-06, + "loss": 2.5691, + "step": 6525 + }, + { + "epoch": 0.6901438240270727, + "grad_norm": 8.5, + "learning_rate": 4.8485670158309714e-06, + "loss": 2.7065, + "step": 6526 + }, + { + "epoch": 0.6902495769881557, + "grad_norm": 7.40625, + "learning_rate": 4.848508312321909e-06, + "loss": 2.5175, + "step": 6527 + }, + { + "epoch": 0.6903553299492385, + "grad_norm": 7.71875, + "learning_rate": 4.8484495977922564e-06, + "loss": 2.5563, + "step": 6528 + }, + { + "epoch": 0.6904610829103215, + "grad_norm": 8.0, + "learning_rate": 4.848390872242291e-06, + "loss": 2.5436, + "step": 6529 + }, + { + "epoch": 0.6905668358714044, + "grad_norm": 7.84375, + "learning_rate": 4.8483321356722875e-06, + "loss": 2.4786, + "step": 6530 + }, + { + "epoch": 0.6906725888324873, + "grad_norm": 7.84375, + "learning_rate": 4.8482733880825224e-06, + "loss": 2.4904, + "step": 6531 + }, + { + "epoch": 0.6907783417935702, + "grad_norm": 7.5, + "learning_rate": 4.84821462947327e-06, + "loss": 2.6726, + "step": 6532 + }, + { + "epoch": 0.6908840947546532, + "grad_norm": 7.90625, + "learning_rate": 4.8481558598448075e-06, + "loss": 2.6861, + "step": 6533 + }, + { + "epoch": 0.690989847715736, + "grad_norm": 7.75, + "learning_rate": 4.848097079197409e-06, + "loss": 2.5209, + "step": 6534 + }, + { + "epoch": 0.691095600676819, + "grad_norm": 7.5625, + "learning_rate": 4.848038287531353e-06, + "loss": 2.6134, + "step": 6535 + }, + { + "epoch": 0.6912013536379019, + "grad_norm": 7.75, + "learning_rate": 4.847979484846912e-06, + "loss": 2.491, + "step": 6536 + }, + { + "epoch": 0.6913071065989848, + "grad_norm": 8.125, + "learning_rate": 4.847920671144364e-06, + "loss": 2.7013, + "step": 6537 + }, + { + "epoch": 0.6914128595600677, + "grad_norm": 7.84375, + "learning_rate": 4.847861846423984e-06, + "loss": 2.5636, + "step": 6538 + }, + { + "epoch": 0.6915186125211505, + "grad_norm": 9.0, + "learning_rate": 4.847803010686049e-06, + "loss": 2.8167, + "step": 6539 + }, + { + "epoch": 0.6916243654822335, + "grad_norm": 7.3125, + "learning_rate": 4.8477441639308346e-06, + "loss": 2.7357, + "step": 6540 + }, + { + "epoch": 0.6917301184433164, + "grad_norm": 7.875, + "learning_rate": 4.847685306158617e-06, + "loss": 2.7429, + "step": 6541 + }, + { + "epoch": 0.6918358714043993, + "grad_norm": 10.5625, + "learning_rate": 4.847626437369673e-06, + "loss": 2.5829, + "step": 6542 + }, + { + "epoch": 0.6919416243654822, + "grad_norm": 8.1875, + "learning_rate": 4.847567557564278e-06, + "loss": 2.7456, + "step": 6543 + }, + { + "epoch": 0.6920473773265652, + "grad_norm": 7.9375, + "learning_rate": 4.847508666742709e-06, + "loss": 2.6557, + "step": 6544 + }, + { + "epoch": 0.692153130287648, + "grad_norm": 8.375, + "learning_rate": 4.847449764905241e-06, + "loss": 2.6087, + "step": 6545 + }, + { + "epoch": 0.692258883248731, + "grad_norm": 8.1875, + "learning_rate": 4.847390852052152e-06, + "loss": 2.9228, + "step": 6546 + }, + { + "epoch": 0.6923646362098139, + "grad_norm": 8.3125, + "learning_rate": 4.847331928183717e-06, + "loss": 2.6681, + "step": 6547 + }, + { + "epoch": 0.6924703891708968, + "grad_norm": 7.375, + "learning_rate": 4.847272993300214e-06, + "loss": 2.449, + "step": 6548 + }, + { + "epoch": 0.6925761421319797, + "grad_norm": 7.5625, + "learning_rate": 4.847214047401918e-06, + "loss": 2.5898, + "step": 6549 + }, + { + "epoch": 0.6926818950930627, + "grad_norm": 7.59375, + "learning_rate": 4.847155090489108e-06, + "loss": 2.455, + "step": 6550 + }, + { + "epoch": 0.6927876480541455, + "grad_norm": 8.1875, + "learning_rate": 4.847096122562058e-06, + "loss": 2.8271, + "step": 6551 + }, + { + "epoch": 0.6928934010152284, + "grad_norm": 8.8125, + "learning_rate": 4.8470371436210455e-06, + "loss": 2.197, + "step": 6552 + }, + { + "epoch": 0.6929991539763113, + "grad_norm": 7.78125, + "learning_rate": 4.846978153666348e-06, + "loss": 2.6516, + "step": 6553 + }, + { + "epoch": 0.6931049069373942, + "grad_norm": 7.84375, + "learning_rate": 4.8469191526982424e-06, + "loss": 2.7367, + "step": 6554 + }, + { + "epoch": 0.6932106598984772, + "grad_norm": 7.65625, + "learning_rate": 4.846860140717005e-06, + "loss": 2.7, + "step": 6555 + }, + { + "epoch": 0.69331641285956, + "grad_norm": 7.6875, + "learning_rate": 4.846801117722913e-06, + "loss": 2.5555, + "step": 6556 + }, + { + "epoch": 0.693422165820643, + "grad_norm": 7.875, + "learning_rate": 4.846742083716242e-06, + "loss": 2.4712, + "step": 6557 + }, + { + "epoch": 0.6935279187817259, + "grad_norm": 8.6875, + "learning_rate": 4.846683038697271e-06, + "loss": 2.5905, + "step": 6558 + }, + { + "epoch": 0.6936336717428088, + "grad_norm": 9.125, + "learning_rate": 4.8466239826662756e-06, + "loss": 2.6356, + "step": 6559 + }, + { + "epoch": 0.6937394247038917, + "grad_norm": 7.84375, + "learning_rate": 4.846564915623534e-06, + "loss": 2.3558, + "step": 6560 + }, + { + "epoch": 0.6938451776649747, + "grad_norm": 8.0625, + "learning_rate": 4.846505837569323e-06, + "loss": 2.4608, + "step": 6561 + }, + { + "epoch": 0.6939509306260575, + "grad_norm": 7.9375, + "learning_rate": 4.84644674850392e-06, + "loss": 2.3489, + "step": 6562 + }, + { + "epoch": 0.6940566835871405, + "grad_norm": 8.125, + "learning_rate": 4.846387648427602e-06, + "loss": 2.5579, + "step": 6563 + }, + { + "epoch": 0.6941624365482234, + "grad_norm": 8.6875, + "learning_rate": 4.846328537340646e-06, + "loss": 2.5976, + "step": 6564 + }, + { + "epoch": 0.6942681895093062, + "grad_norm": 8.125, + "learning_rate": 4.846269415243331e-06, + "loss": 2.4654, + "step": 6565 + }, + { + "epoch": 0.6943739424703892, + "grad_norm": 7.90625, + "learning_rate": 4.846210282135932e-06, + "loss": 2.4723, + "step": 6566 + }, + { + "epoch": 0.694479695431472, + "grad_norm": 8.0625, + "learning_rate": 4.846151138018728e-06, + "loss": 2.4637, + "step": 6567 + }, + { + "epoch": 0.694585448392555, + "grad_norm": 8.375, + "learning_rate": 4.846091982891996e-06, + "loss": 2.6597, + "step": 6568 + }, + { + "epoch": 0.6946912013536379, + "grad_norm": 8.125, + "learning_rate": 4.8460328167560145e-06, + "loss": 2.5557, + "step": 6569 + }, + { + "epoch": 0.6947969543147208, + "grad_norm": 7.5625, + "learning_rate": 4.84597363961106e-06, + "loss": 2.3404, + "step": 6570 + }, + { + "epoch": 0.6949027072758037, + "grad_norm": 8.3125, + "learning_rate": 4.845914451457411e-06, + "loss": 2.6234, + "step": 6571 + }, + { + "epoch": 0.6950084602368867, + "grad_norm": 8.875, + "learning_rate": 4.845855252295345e-06, + "loss": 2.2668, + "step": 6572 + }, + { + "epoch": 0.6951142131979695, + "grad_norm": 8.4375, + "learning_rate": 4.845796042125139e-06, + "loss": 2.7415, + "step": 6573 + }, + { + "epoch": 0.6952199661590525, + "grad_norm": 8.6875, + "learning_rate": 4.845736820947072e-06, + "loss": 2.6435, + "step": 6574 + }, + { + "epoch": 0.6953257191201354, + "grad_norm": 7.375, + "learning_rate": 4.845677588761421e-06, + "loss": 2.6392, + "step": 6575 + }, + { + "epoch": 0.6954314720812182, + "grad_norm": 7.75, + "learning_rate": 4.845618345568466e-06, + "loss": 2.3836, + "step": 6576 + }, + { + "epoch": 0.6955372250423012, + "grad_norm": 7.4375, + "learning_rate": 4.845559091368481e-06, + "loss": 2.4689, + "step": 6577 + }, + { + "epoch": 0.6956429780033841, + "grad_norm": 7.90625, + "learning_rate": 4.845499826161749e-06, + "loss": 2.6677, + "step": 6578 + }, + { + "epoch": 0.695748730964467, + "grad_norm": 8.125, + "learning_rate": 4.845440549948544e-06, + "loss": 2.5029, + "step": 6579 + }, + { + "epoch": 0.6958544839255499, + "grad_norm": 8.0625, + "learning_rate": 4.8453812627291465e-06, + "loss": 2.6059, + "step": 6580 + }, + { + "epoch": 0.6959602368866328, + "grad_norm": 7.90625, + "learning_rate": 4.8453219645038335e-06, + "loss": 2.5174, + "step": 6581 + }, + { + "epoch": 0.6960659898477157, + "grad_norm": 8.25, + "learning_rate": 4.845262655272884e-06, + "loss": 2.7767, + "step": 6582 + }, + { + "epoch": 0.6961717428087987, + "grad_norm": 7.59375, + "learning_rate": 4.8452033350365754e-06, + "loss": 2.497, + "step": 6583 + }, + { + "epoch": 0.6962774957698815, + "grad_norm": 7.875, + "learning_rate": 4.845144003795188e-06, + "loss": 2.5731, + "step": 6584 + }, + { + "epoch": 0.6963832487309645, + "grad_norm": 8.25, + "learning_rate": 4.845084661548997e-06, + "loss": 2.5038, + "step": 6585 + }, + { + "epoch": 0.6964890016920474, + "grad_norm": 8.4375, + "learning_rate": 4.8450253082982844e-06, + "loss": 2.4999, + "step": 6586 + }, + { + "epoch": 0.6965947546531303, + "grad_norm": 7.96875, + "learning_rate": 4.844965944043327e-06, + "loss": 2.6364, + "step": 6587 + }, + { + "epoch": 0.6967005076142132, + "grad_norm": 7.9375, + "learning_rate": 4.844906568784403e-06, + "loss": 2.7098, + "step": 6588 + }, + { + "epoch": 0.6968062605752962, + "grad_norm": 7.96875, + "learning_rate": 4.844847182521791e-06, + "loss": 2.5545, + "step": 6589 + }, + { + "epoch": 0.696912013536379, + "grad_norm": 7.65625, + "learning_rate": 4.844787785255772e-06, + "loss": 2.598, + "step": 6590 + }, + { + "epoch": 0.6970177664974619, + "grad_norm": 7.75, + "learning_rate": 4.844728376986621e-06, + "loss": 2.5532, + "step": 6591 + }, + { + "epoch": 0.6971235194585449, + "grad_norm": 7.875, + "learning_rate": 4.844668957714619e-06, + "loss": 2.7553, + "step": 6592 + }, + { + "epoch": 0.6972292724196277, + "grad_norm": 9.0625, + "learning_rate": 4.844609527440045e-06, + "loss": 2.9045, + "step": 6593 + }, + { + "epoch": 0.6973350253807107, + "grad_norm": 8.25, + "learning_rate": 4.844550086163178e-06, + "loss": 2.4927, + "step": 6594 + }, + { + "epoch": 0.6974407783417935, + "grad_norm": 7.96875, + "learning_rate": 4.844490633884296e-06, + "loss": 2.6006, + "step": 6595 + }, + { + "epoch": 0.6975465313028765, + "grad_norm": 8.3125, + "learning_rate": 4.844431170603677e-06, + "loss": 2.4831, + "step": 6596 + }, + { + "epoch": 0.6976522842639594, + "grad_norm": 8.3125, + "learning_rate": 4.844371696321603e-06, + "loss": 2.6642, + "step": 6597 + }, + { + "epoch": 0.6977580372250423, + "grad_norm": 8.75, + "learning_rate": 4.844312211038351e-06, + "loss": 2.7085, + "step": 6598 + }, + { + "epoch": 0.6978637901861252, + "grad_norm": 8.1875, + "learning_rate": 4.8442527147541995e-06, + "loss": 2.6608, + "step": 6599 + }, + { + "epoch": 0.6979695431472082, + "grad_norm": 7.5625, + "learning_rate": 4.84419320746943e-06, + "loss": 2.498, + "step": 6600 + }, + { + "epoch": 0.698075296108291, + "grad_norm": 7.78125, + "learning_rate": 4.844133689184321e-06, + "loss": 2.3662, + "step": 6601 + }, + { + "epoch": 0.698181049069374, + "grad_norm": 7.59375, + "learning_rate": 4.84407415989915e-06, + "loss": 2.4495, + "step": 6602 + }, + { + "epoch": 0.6982868020304569, + "grad_norm": 7.625, + "learning_rate": 4.844014619614198e-06, + "loss": 2.5788, + "step": 6603 + }, + { + "epoch": 0.6983925549915397, + "grad_norm": 8.25, + "learning_rate": 4.843955068329745e-06, + "loss": 2.6878, + "step": 6604 + }, + { + "epoch": 0.6984983079526227, + "grad_norm": 7.8125, + "learning_rate": 4.843895506046068e-06, + "loss": 2.5133, + "step": 6605 + }, + { + "epoch": 0.6986040609137056, + "grad_norm": 8.0, + "learning_rate": 4.843835932763449e-06, + "loss": 2.7984, + "step": 6606 + }, + { + "epoch": 0.6987098138747885, + "grad_norm": 8.3125, + "learning_rate": 4.843776348482165e-06, + "loss": 2.0919, + "step": 6607 + }, + { + "epoch": 0.6988155668358714, + "grad_norm": 7.9375, + "learning_rate": 4.843716753202499e-06, + "loss": 2.552, + "step": 6608 + }, + { + "epoch": 0.6989213197969543, + "grad_norm": 7.59375, + "learning_rate": 4.843657146924729e-06, + "loss": 2.434, + "step": 6609 + }, + { + "epoch": 0.6990270727580372, + "grad_norm": 8.625, + "learning_rate": 4.843597529649133e-06, + "loss": 2.5499, + "step": 6610 + }, + { + "epoch": 0.6991328257191202, + "grad_norm": 8.0625, + "learning_rate": 4.843537901375993e-06, + "loss": 2.4449, + "step": 6611 + }, + { + "epoch": 0.699238578680203, + "grad_norm": 7.46875, + "learning_rate": 4.843478262105589e-06, + "loss": 2.542, + "step": 6612 + }, + { + "epoch": 0.699344331641286, + "grad_norm": 7.6875, + "learning_rate": 4.843418611838198e-06, + "loss": 2.5703, + "step": 6613 + }, + { + "epoch": 0.6994500846023689, + "grad_norm": 7.53125, + "learning_rate": 4.843358950574104e-06, + "loss": 2.5403, + "step": 6614 + }, + { + "epoch": 0.6995558375634517, + "grad_norm": 7.875, + "learning_rate": 4.8432992783135834e-06, + "loss": 2.6465, + "step": 6615 + }, + { + "epoch": 0.6996615905245347, + "grad_norm": 7.875, + "learning_rate": 4.843239595056918e-06, + "loss": 2.6212, + "step": 6616 + }, + { + "epoch": 0.6997673434856176, + "grad_norm": 7.625, + "learning_rate": 4.843179900804388e-06, + "loss": 2.3969, + "step": 6617 + }, + { + "epoch": 0.6998730964467005, + "grad_norm": 8.5, + "learning_rate": 4.843120195556273e-06, + "loss": 2.7106, + "step": 6618 + }, + { + "epoch": 0.6999788494077834, + "grad_norm": 7.625, + "learning_rate": 4.843060479312853e-06, + "loss": 2.4758, + "step": 6619 + }, + { + "epoch": 0.7000846023688664, + "grad_norm": 7.375, + "learning_rate": 4.843000752074408e-06, + "loss": 2.4931, + "step": 6620 + }, + { + "epoch": 0.7001903553299492, + "grad_norm": 7.9375, + "learning_rate": 4.842941013841219e-06, + "loss": 2.4929, + "step": 6621 + }, + { + "epoch": 0.7002961082910322, + "grad_norm": 7.59375, + "learning_rate": 4.842881264613567e-06, + "loss": 2.5177, + "step": 6622 + }, + { + "epoch": 0.700401861252115, + "grad_norm": 8.0625, + "learning_rate": 4.84282150439173e-06, + "loss": 2.7177, + "step": 6623 + }, + { + "epoch": 0.700507614213198, + "grad_norm": 8.9375, + "learning_rate": 4.842761733175991e-06, + "loss": 2.7048, + "step": 6624 + }, + { + "epoch": 0.7006133671742809, + "grad_norm": 7.28125, + "learning_rate": 4.842701950966629e-06, + "loss": 2.4693, + "step": 6625 + }, + { + "epoch": 0.7007191201353637, + "grad_norm": 7.78125, + "learning_rate": 4.842642157763924e-06, + "loss": 2.5765, + "step": 6626 + }, + { + "epoch": 0.7008248730964467, + "grad_norm": 7.9375, + "learning_rate": 4.842582353568158e-06, + "loss": 2.6927, + "step": 6627 + }, + { + "epoch": 0.7009306260575296, + "grad_norm": 7.84375, + "learning_rate": 4.842522538379611e-06, + "loss": 2.5967, + "step": 6628 + }, + { + "epoch": 0.7010363790186125, + "grad_norm": 8.875, + "learning_rate": 4.842462712198565e-06, + "loss": 2.7264, + "step": 6629 + }, + { + "epoch": 0.7011421319796954, + "grad_norm": 8.75, + "learning_rate": 4.842402875025298e-06, + "loss": 2.5698, + "step": 6630 + }, + { + "epoch": 0.7012478849407784, + "grad_norm": 8.0, + "learning_rate": 4.8423430268600936e-06, + "loss": 2.6021, + "step": 6631 + }, + { + "epoch": 0.7013536379018612, + "grad_norm": 7.75, + "learning_rate": 4.84228316770323e-06, + "loss": 2.5992, + "step": 6632 + }, + { + "epoch": 0.7014593908629442, + "grad_norm": 7.6875, + "learning_rate": 4.84222329755499e-06, + "loss": 2.6199, + "step": 6633 + }, + { + "epoch": 0.7015651438240271, + "grad_norm": 8.5, + "learning_rate": 4.842163416415655e-06, + "loss": 2.4227, + "step": 6634 + }, + { + "epoch": 0.70167089678511, + "grad_norm": 7.4375, + "learning_rate": 4.842103524285504e-06, + "loss": 2.5101, + "step": 6635 + }, + { + "epoch": 0.7017766497461929, + "grad_norm": 7.65625, + "learning_rate": 4.842043621164818e-06, + "loss": 2.5665, + "step": 6636 + }, + { + "epoch": 0.7018824027072758, + "grad_norm": 8.25, + "learning_rate": 4.841983707053881e-06, + "loss": 2.6143, + "step": 6637 + }, + { + "epoch": 0.7019881556683587, + "grad_norm": 7.90625, + "learning_rate": 4.841923781952972e-06, + "loss": 2.4989, + "step": 6638 + }, + { + "epoch": 0.7020939086294417, + "grad_norm": 8.875, + "learning_rate": 4.841863845862372e-06, + "loss": 2.671, + "step": 6639 + }, + { + "epoch": 0.7021996615905245, + "grad_norm": 7.875, + "learning_rate": 4.841803898782362e-06, + "loss": 2.3383, + "step": 6640 + }, + { + "epoch": 0.7023054145516074, + "grad_norm": 7.71875, + "learning_rate": 4.841743940713226e-06, + "loss": 2.5756, + "step": 6641 + }, + { + "epoch": 0.7024111675126904, + "grad_norm": 7.5, + "learning_rate": 4.841683971655242e-06, + "loss": 2.5945, + "step": 6642 + }, + { + "epoch": 0.7025169204737732, + "grad_norm": 8.1875, + "learning_rate": 4.841623991608693e-06, + "loss": 2.6941, + "step": 6643 + }, + { + "epoch": 0.7026226734348562, + "grad_norm": 7.96875, + "learning_rate": 4.841564000573861e-06, + "loss": 2.7567, + "step": 6644 + }, + { + "epoch": 0.7027284263959391, + "grad_norm": 8.125, + "learning_rate": 4.841503998551026e-06, + "loss": 2.7504, + "step": 6645 + }, + { + "epoch": 0.702834179357022, + "grad_norm": 7.9375, + "learning_rate": 4.84144398554047e-06, + "loss": 2.5947, + "step": 6646 + }, + { + "epoch": 0.7029399323181049, + "grad_norm": 7.78125, + "learning_rate": 4.841383961542476e-06, + "loss": 2.5167, + "step": 6647 + }, + { + "epoch": 0.7030456852791879, + "grad_norm": 8.0625, + "learning_rate": 4.841323926557325e-06, + "loss": 2.6227, + "step": 6648 + }, + { + "epoch": 0.7031514382402707, + "grad_norm": 7.65625, + "learning_rate": 4.841263880585297e-06, + "loss": 2.6596, + "step": 6649 + }, + { + "epoch": 0.7032571912013537, + "grad_norm": 8.625, + "learning_rate": 4.841203823626676e-06, + "loss": 2.7211, + "step": 6650 + }, + { + "epoch": 0.7033629441624365, + "grad_norm": 7.96875, + "learning_rate": 4.8411437556817435e-06, + "loss": 2.421, + "step": 6651 + }, + { + "epoch": 0.7034686971235194, + "grad_norm": 8.125, + "learning_rate": 4.841083676750779e-06, + "loss": 2.6764, + "step": 6652 + }, + { + "epoch": 0.7035744500846024, + "grad_norm": 8.75, + "learning_rate": 4.841023586834068e-06, + "loss": 2.5503, + "step": 6653 + }, + { + "epoch": 0.7036802030456852, + "grad_norm": 8.375, + "learning_rate": 4.840963485931891e-06, + "loss": 2.5585, + "step": 6654 + }, + { + "epoch": 0.7037859560067682, + "grad_norm": 7.5, + "learning_rate": 4.840903374044529e-06, + "loss": 2.5954, + "step": 6655 + }, + { + "epoch": 0.7038917089678511, + "grad_norm": 7.5625, + "learning_rate": 4.840843251172265e-06, + "loss": 2.6234, + "step": 6656 + }, + { + "epoch": 0.703997461928934, + "grad_norm": 7.78125, + "learning_rate": 4.84078311731538e-06, + "loss": 2.5538, + "step": 6657 + }, + { + "epoch": 0.7041032148900169, + "grad_norm": 7.625, + "learning_rate": 4.840722972474158e-06, + "loss": 2.4919, + "step": 6658 + }, + { + "epoch": 0.7042089678510999, + "grad_norm": 7.5625, + "learning_rate": 4.840662816648881e-06, + "loss": 2.5005, + "step": 6659 + }, + { + "epoch": 0.7043147208121827, + "grad_norm": 8.25, + "learning_rate": 4.84060264983983e-06, + "loss": 2.5919, + "step": 6660 + }, + { + "epoch": 0.7044204737732657, + "grad_norm": 7.25, + "learning_rate": 4.8405424720472886e-06, + "loss": 2.6705, + "step": 6661 + }, + { + "epoch": 0.7045262267343486, + "grad_norm": 7.75, + "learning_rate": 4.840482283271538e-06, + "loss": 2.6888, + "step": 6662 + }, + { + "epoch": 0.7046319796954315, + "grad_norm": 8.875, + "learning_rate": 4.8404220835128615e-06, + "loss": 2.6163, + "step": 6663 + }, + { + "epoch": 0.7047377326565144, + "grad_norm": 8.5625, + "learning_rate": 4.840361872771541e-06, + "loss": 2.4523, + "step": 6664 + }, + { + "epoch": 0.7048434856175972, + "grad_norm": 7.625, + "learning_rate": 4.84030165104786e-06, + "loss": 2.5137, + "step": 6665 + }, + { + "epoch": 0.7049492385786802, + "grad_norm": 8.5625, + "learning_rate": 4.8402414183421005e-06, + "loss": 2.5613, + "step": 6666 + }, + { + "epoch": 0.7050549915397631, + "grad_norm": 7.53125, + "learning_rate": 4.840181174654544e-06, + "loss": 2.5807, + "step": 6667 + }, + { + "epoch": 0.705160744500846, + "grad_norm": 7.6875, + "learning_rate": 4.840120919985476e-06, + "loss": 2.6477, + "step": 6668 + }, + { + "epoch": 0.7052664974619289, + "grad_norm": 7.625, + "learning_rate": 4.840060654335177e-06, + "loss": 2.6622, + "step": 6669 + }, + { + "epoch": 0.7053722504230119, + "grad_norm": 7.875, + "learning_rate": 4.84000037770393e-06, + "loss": 2.7309, + "step": 6670 + }, + { + "epoch": 0.7054780033840947, + "grad_norm": 7.96875, + "learning_rate": 4.839940090092019e-06, + "loss": 2.3288, + "step": 6671 + }, + { + "epoch": 0.7055837563451777, + "grad_norm": 8.125, + "learning_rate": 4.839879791499725e-06, + "loss": 2.6938, + "step": 6672 + }, + { + "epoch": 0.7056895093062606, + "grad_norm": 7.40625, + "learning_rate": 4.839819481927333e-06, + "loss": 2.5076, + "step": 6673 + }, + { + "epoch": 0.7057952622673435, + "grad_norm": 8.0, + "learning_rate": 4.839759161375125e-06, + "loss": 2.617, + "step": 6674 + }, + { + "epoch": 0.7059010152284264, + "grad_norm": 8.0625, + "learning_rate": 4.839698829843384e-06, + "loss": 2.6685, + "step": 6675 + }, + { + "epoch": 0.7060067681895094, + "grad_norm": 8.125, + "learning_rate": 4.839638487332394e-06, + "loss": 2.7487, + "step": 6676 + }, + { + "epoch": 0.7061125211505922, + "grad_norm": 7.71875, + "learning_rate": 4.8395781338424365e-06, + "loss": 2.3694, + "step": 6677 + }, + { + "epoch": 0.7062182741116751, + "grad_norm": 8.3125, + "learning_rate": 4.839517769373796e-06, + "loss": 2.525, + "step": 6678 + }, + { + "epoch": 0.706324027072758, + "grad_norm": 7.9375, + "learning_rate": 4.839457393926756e-06, + "loss": 2.641, + "step": 6679 + }, + { + "epoch": 0.7064297800338409, + "grad_norm": 8.0625, + "learning_rate": 4.839397007501598e-06, + "loss": 2.5508, + "step": 6680 + }, + { + "epoch": 0.7065355329949239, + "grad_norm": 7.84375, + "learning_rate": 4.839336610098607e-06, + "loss": 2.6108, + "step": 6681 + }, + { + "epoch": 0.7066412859560067, + "grad_norm": 7.84375, + "learning_rate": 4.839276201718067e-06, + "loss": 2.4864, + "step": 6682 + }, + { + "epoch": 0.7067470389170897, + "grad_norm": 7.4375, + "learning_rate": 4.83921578236026e-06, + "loss": 2.5407, + "step": 6683 + }, + { + "epoch": 0.7068527918781726, + "grad_norm": 7.96875, + "learning_rate": 4.8391553520254694e-06, + "loss": 2.6626, + "step": 6684 + }, + { + "epoch": 0.7069585448392555, + "grad_norm": 8.0625, + "learning_rate": 4.8390949107139805e-06, + "loss": 2.7063, + "step": 6685 + }, + { + "epoch": 0.7070642978003384, + "grad_norm": 7.3125, + "learning_rate": 4.839034458426075e-06, + "loss": 2.2588, + "step": 6686 + }, + { + "epoch": 0.7071700507614214, + "grad_norm": 8.0625, + "learning_rate": 4.838973995162038e-06, + "loss": 2.5563, + "step": 6687 + }, + { + "epoch": 0.7072758037225042, + "grad_norm": 8.0, + "learning_rate": 4.838913520922152e-06, + "loss": 2.0364, + "step": 6688 + }, + { + "epoch": 0.7073815566835872, + "grad_norm": 8.5625, + "learning_rate": 4.838853035706702e-06, + "loss": 2.7389, + "step": 6689 + }, + { + "epoch": 0.7074873096446701, + "grad_norm": 7.625, + "learning_rate": 4.838792539515971e-06, + "loss": 2.5214, + "step": 6690 + }, + { + "epoch": 0.7075930626057529, + "grad_norm": 8.0625, + "learning_rate": 4.838732032350243e-06, + "loss": 2.6757, + "step": 6691 + }, + { + "epoch": 0.7076988155668359, + "grad_norm": 8.375, + "learning_rate": 4.838671514209803e-06, + "loss": 2.4548, + "step": 6692 + }, + { + "epoch": 0.7078045685279187, + "grad_norm": 7.625, + "learning_rate": 4.838610985094933e-06, + "loss": 2.6323, + "step": 6693 + }, + { + "epoch": 0.7079103214890017, + "grad_norm": 8.4375, + "learning_rate": 4.8385504450059185e-06, + "loss": 2.6454, + "step": 6694 + }, + { + "epoch": 0.7080160744500846, + "grad_norm": 7.84375, + "learning_rate": 4.838489893943043e-06, + "loss": 2.6842, + "step": 6695 + }, + { + "epoch": 0.7081218274111675, + "grad_norm": 7.34375, + "learning_rate": 4.838429331906591e-06, + "loss": 2.5876, + "step": 6696 + }, + { + "epoch": 0.7082275803722504, + "grad_norm": 8.1875, + "learning_rate": 4.838368758896847e-06, + "loss": 2.7284, + "step": 6697 + }, + { + "epoch": 0.7083333333333334, + "grad_norm": 7.96875, + "learning_rate": 4.838308174914094e-06, + "loss": 2.454, + "step": 6698 + }, + { + "epoch": 0.7084390862944162, + "grad_norm": 7.8125, + "learning_rate": 4.838247579958617e-06, + "loss": 2.6603, + "step": 6699 + }, + { + "epoch": 0.7085448392554992, + "grad_norm": 7.625, + "learning_rate": 4.838186974030702e-06, + "loss": 2.4095, + "step": 6700 + }, + { + "epoch": 0.7086505922165821, + "grad_norm": 7.5625, + "learning_rate": 4.83812635713063e-06, + "loss": 2.3641, + "step": 6701 + }, + { + "epoch": 0.708756345177665, + "grad_norm": 7.71875, + "learning_rate": 4.838065729258688e-06, + "loss": 2.5119, + "step": 6702 + }, + { + "epoch": 0.7088620981387479, + "grad_norm": 9.9375, + "learning_rate": 4.83800509041516e-06, + "loss": 2.6992, + "step": 6703 + }, + { + "epoch": 0.7089678510998308, + "grad_norm": 7.875, + "learning_rate": 4.83794444060033e-06, + "loss": 2.582, + "step": 6704 + }, + { + "epoch": 0.7090736040609137, + "grad_norm": 7.84375, + "learning_rate": 4.837883779814482e-06, + "loss": 2.659, + "step": 6705 + }, + { + "epoch": 0.7091793570219966, + "grad_norm": 9.1875, + "learning_rate": 4.837823108057903e-06, + "loss": 3.0241, + "step": 6706 + }, + { + "epoch": 0.7092851099830795, + "grad_norm": 8.0625, + "learning_rate": 4.837762425330876e-06, + "loss": 2.7514, + "step": 6707 + }, + { + "epoch": 0.7093908629441624, + "grad_norm": 7.75, + "learning_rate": 4.8377017316336856e-06, + "loss": 2.6258, + "step": 6708 + }, + { + "epoch": 0.7094966159052454, + "grad_norm": 7.9375, + "learning_rate": 4.8376410269666175e-06, + "loss": 2.4978, + "step": 6709 + }, + { + "epoch": 0.7096023688663282, + "grad_norm": 7.96875, + "learning_rate": 4.837580311329956e-06, + "loss": 2.5911, + "step": 6710 + }, + { + "epoch": 0.7097081218274112, + "grad_norm": 8.3125, + "learning_rate": 4.837519584723985e-06, + "loss": 2.6884, + "step": 6711 + }, + { + "epoch": 0.7098138747884941, + "grad_norm": 7.4375, + "learning_rate": 4.837458847148992e-06, + "loss": 2.456, + "step": 6712 + }, + { + "epoch": 0.709919627749577, + "grad_norm": 7.96875, + "learning_rate": 4.83739809860526e-06, + "loss": 2.5372, + "step": 6713 + }, + { + "epoch": 0.7100253807106599, + "grad_norm": 7.625, + "learning_rate": 4.837337339093076e-06, + "loss": 2.7482, + "step": 6714 + }, + { + "epoch": 0.7101311336717429, + "grad_norm": 8.5625, + "learning_rate": 4.837276568612722e-06, + "loss": 2.758, + "step": 6715 + }, + { + "epoch": 0.7102368866328257, + "grad_norm": 8.375, + "learning_rate": 4.837215787164486e-06, + "loss": 2.7758, + "step": 6716 + }, + { + "epoch": 0.7103426395939086, + "grad_norm": 8.125, + "learning_rate": 4.837154994748652e-06, + "loss": 2.7631, + "step": 6717 + }, + { + "epoch": 0.7104483925549916, + "grad_norm": 8.3125, + "learning_rate": 4.837094191365505e-06, + "loss": 2.7161, + "step": 6718 + }, + { + "epoch": 0.7105541455160744, + "grad_norm": 8.5625, + "learning_rate": 4.837033377015332e-06, + "loss": 2.7488, + "step": 6719 + }, + { + "epoch": 0.7106598984771574, + "grad_norm": 8.375, + "learning_rate": 4.836972551698416e-06, + "loss": 2.635, + "step": 6720 + }, + { + "epoch": 0.7107656514382402, + "grad_norm": 7.6875, + "learning_rate": 4.836911715415043e-06, + "loss": 2.4835, + "step": 6721 + }, + { + "epoch": 0.7108714043993232, + "grad_norm": 7.4375, + "learning_rate": 4.836850868165501e-06, + "loss": 2.4488, + "step": 6722 + }, + { + "epoch": 0.7109771573604061, + "grad_norm": 7.90625, + "learning_rate": 4.8367900099500725e-06, + "loss": 2.7558, + "step": 6723 + }, + { + "epoch": 0.711082910321489, + "grad_norm": 7.9375, + "learning_rate": 4.836729140769044e-06, + "loss": 2.7682, + "step": 6724 + }, + { + "epoch": 0.7111886632825719, + "grad_norm": 7.59375, + "learning_rate": 4.8366682606227025e-06, + "loss": 2.5532, + "step": 6725 + }, + { + "epoch": 0.7112944162436549, + "grad_norm": 7.4375, + "learning_rate": 4.836607369511331e-06, + "loss": 2.4181, + "step": 6726 + }, + { + "epoch": 0.7114001692047377, + "grad_norm": 8.4375, + "learning_rate": 4.836546467435218e-06, + "loss": 2.5189, + "step": 6727 + }, + { + "epoch": 0.7115059221658206, + "grad_norm": 10.1875, + "learning_rate": 4.836485554394648e-06, + "loss": 2.8565, + "step": 6728 + }, + { + "epoch": 0.7116116751269036, + "grad_norm": 7.6875, + "learning_rate": 4.836424630389906e-06, + "loss": 2.5251, + "step": 6729 + }, + { + "epoch": 0.7117174280879864, + "grad_norm": 8.875, + "learning_rate": 4.836363695421279e-06, + "loss": 2.555, + "step": 6730 + }, + { + "epoch": 0.7118231810490694, + "grad_norm": 8.25, + "learning_rate": 4.8363027494890526e-06, + "loss": 2.7305, + "step": 6731 + }, + { + "epoch": 0.7119289340101523, + "grad_norm": 8.375, + "learning_rate": 4.836241792593514e-06, + "loss": 2.6223, + "step": 6732 + }, + { + "epoch": 0.7120346869712352, + "grad_norm": 7.21875, + "learning_rate": 4.836180824734947e-06, + "loss": 2.4632, + "step": 6733 + }, + { + "epoch": 0.7121404399323181, + "grad_norm": 8.4375, + "learning_rate": 4.8361198459136394e-06, + "loss": 2.9636, + "step": 6734 + }, + { + "epoch": 0.712246192893401, + "grad_norm": 8.1875, + "learning_rate": 4.836058856129876e-06, + "loss": 2.7905, + "step": 6735 + }, + { + "epoch": 0.7123519458544839, + "grad_norm": 7.5, + "learning_rate": 4.835997855383945e-06, + "loss": 2.53, + "step": 6736 + }, + { + "epoch": 0.7124576988155669, + "grad_norm": 8.4375, + "learning_rate": 4.835936843676131e-06, + "loss": 2.6483, + "step": 6737 + }, + { + "epoch": 0.7125634517766497, + "grad_norm": 7.84375, + "learning_rate": 4.83587582100672e-06, + "loss": 2.6563, + "step": 6738 + }, + { + "epoch": 0.7126692047377327, + "grad_norm": 7.75, + "learning_rate": 4.835814787376e-06, + "loss": 2.6382, + "step": 6739 + }, + { + "epoch": 0.7127749576988156, + "grad_norm": 7.75, + "learning_rate": 4.8357537427842564e-06, + "loss": 2.5435, + "step": 6740 + }, + { + "epoch": 0.7128807106598984, + "grad_norm": 7.8125, + "learning_rate": 4.835692687231776e-06, + "loss": 2.5771, + "step": 6741 + }, + { + "epoch": 0.7129864636209814, + "grad_norm": 8.1875, + "learning_rate": 4.8356316207188445e-06, + "loss": 2.5194, + "step": 6742 + }, + { + "epoch": 0.7130922165820643, + "grad_norm": 8.375, + "learning_rate": 4.835570543245749e-06, + "loss": 2.4971, + "step": 6743 + }, + { + "epoch": 0.7131979695431472, + "grad_norm": 8.625, + "learning_rate": 4.835509454812777e-06, + "loss": 2.6262, + "step": 6744 + }, + { + "epoch": 0.7133037225042301, + "grad_norm": 7.59375, + "learning_rate": 4.8354483554202135e-06, + "loss": 2.3694, + "step": 6745 + }, + { + "epoch": 0.7134094754653131, + "grad_norm": 8.25, + "learning_rate": 4.835387245068347e-06, + "loss": 2.3277, + "step": 6746 + }, + { + "epoch": 0.7135152284263959, + "grad_norm": 7.71875, + "learning_rate": 4.835326123757462e-06, + "loss": 2.518, + "step": 6747 + }, + { + "epoch": 0.7136209813874789, + "grad_norm": 8.0, + "learning_rate": 4.835264991487847e-06, + "loss": 2.4813, + "step": 6748 + }, + { + "epoch": 0.7137267343485617, + "grad_norm": 7.78125, + "learning_rate": 4.835203848259789e-06, + "loss": 2.7195, + "step": 6749 + }, + { + "epoch": 0.7138324873096447, + "grad_norm": 8.0625, + "learning_rate": 4.835142694073574e-06, + "loss": 2.3397, + "step": 6750 + }, + { + "epoch": 0.7139382402707276, + "grad_norm": 8.1875, + "learning_rate": 4.8350815289294895e-06, + "loss": 2.7006, + "step": 6751 + }, + { + "epoch": 0.7140439932318104, + "grad_norm": 8.0625, + "learning_rate": 4.835020352827822e-06, + "loss": 2.5458, + "step": 6752 + }, + { + "epoch": 0.7141497461928934, + "grad_norm": 8.875, + "learning_rate": 4.83495916576886e-06, + "loss": 2.4375, + "step": 6753 + }, + { + "epoch": 0.7142554991539763, + "grad_norm": 8.0625, + "learning_rate": 4.834897967752889e-06, + "loss": 2.4908, + "step": 6754 + }, + { + "epoch": 0.7143612521150592, + "grad_norm": 8.25, + "learning_rate": 4.8348367587801965e-06, + "loss": 2.6332, + "step": 6755 + }, + { + "epoch": 0.7144670050761421, + "grad_norm": 7.21875, + "learning_rate": 4.83477553885107e-06, + "loss": 2.4251, + "step": 6756 + }, + { + "epoch": 0.7145727580372251, + "grad_norm": 7.875, + "learning_rate": 4.834714307965797e-06, + "loss": 2.7781, + "step": 6757 + }, + { + "epoch": 0.7146785109983079, + "grad_norm": 7.40625, + "learning_rate": 4.834653066124665e-06, + "loss": 2.4856, + "step": 6758 + }, + { + "epoch": 0.7147842639593909, + "grad_norm": 7.4375, + "learning_rate": 4.83459181332796e-06, + "loss": 2.54, + "step": 6759 + }, + { + "epoch": 0.7148900169204738, + "grad_norm": 8.0625, + "learning_rate": 4.834530549575972e-06, + "loss": 2.6817, + "step": 6760 + }, + { + "epoch": 0.7149957698815567, + "grad_norm": 7.5, + "learning_rate": 4.834469274868986e-06, + "loss": 2.3194, + "step": 6761 + }, + { + "epoch": 0.7151015228426396, + "grad_norm": 7.71875, + "learning_rate": 4.83440798920729e-06, + "loss": 2.6051, + "step": 6762 + }, + { + "epoch": 0.7152072758037225, + "grad_norm": 7.8125, + "learning_rate": 4.834346692591172e-06, + "loss": 2.6169, + "step": 6763 + }, + { + "epoch": 0.7153130287648054, + "grad_norm": 8.5, + "learning_rate": 4.83428538502092e-06, + "loss": 2.4775, + "step": 6764 + }, + { + "epoch": 0.7154187817258884, + "grad_norm": 7.65625, + "learning_rate": 4.834224066496822e-06, + "loss": 2.7175, + "step": 6765 + }, + { + "epoch": 0.7155245346869712, + "grad_norm": 7.84375, + "learning_rate": 4.834162737019165e-06, + "loss": 2.6707, + "step": 6766 + }, + { + "epoch": 0.7156302876480541, + "grad_norm": 8.5625, + "learning_rate": 4.834101396588237e-06, + "loss": 2.5746, + "step": 6767 + }, + { + "epoch": 0.7157360406091371, + "grad_norm": 8.1875, + "learning_rate": 4.834040045204325e-06, + "loss": 2.6889, + "step": 6768 + }, + { + "epoch": 0.7158417935702199, + "grad_norm": 7.84375, + "learning_rate": 4.8339786828677185e-06, + "loss": 2.7226, + "step": 6769 + }, + { + "epoch": 0.7159475465313029, + "grad_norm": 7.9375, + "learning_rate": 4.833917309578704e-06, + "loss": 2.6136, + "step": 6770 + }, + { + "epoch": 0.7160532994923858, + "grad_norm": 7.59375, + "learning_rate": 4.83385592533757e-06, + "loss": 2.5807, + "step": 6771 + }, + { + "epoch": 0.7161590524534687, + "grad_norm": 8.0625, + "learning_rate": 4.833794530144605e-06, + "loss": 2.7313, + "step": 6772 + }, + { + "epoch": 0.7162648054145516, + "grad_norm": 7.4375, + "learning_rate": 4.833733124000096e-06, + "loss": 2.7468, + "step": 6773 + }, + { + "epoch": 0.7163705583756346, + "grad_norm": 8.5, + "learning_rate": 4.833671706904333e-06, + "loss": 2.6842, + "step": 6774 + }, + { + "epoch": 0.7164763113367174, + "grad_norm": 8.0625, + "learning_rate": 4.833610278857603e-06, + "loss": 2.4555, + "step": 6775 + }, + { + "epoch": 0.7165820642978004, + "grad_norm": 7.5625, + "learning_rate": 4.8335488398601935e-06, + "loss": 2.5027, + "step": 6776 + }, + { + "epoch": 0.7166878172588832, + "grad_norm": 7.9375, + "learning_rate": 4.8334873899123945e-06, + "loss": 2.5489, + "step": 6777 + }, + { + "epoch": 0.7167935702199661, + "grad_norm": 7.59375, + "learning_rate": 4.833425929014493e-06, + "loss": 2.674, + "step": 6778 + }, + { + "epoch": 0.7168993231810491, + "grad_norm": 8.0625, + "learning_rate": 4.8333644571667786e-06, + "loss": 2.767, + "step": 6779 + }, + { + "epoch": 0.7170050761421319, + "grad_norm": 8.25, + "learning_rate": 4.833302974369539e-06, + "loss": 2.521, + "step": 6780 + }, + { + "epoch": 0.7171108291032149, + "grad_norm": 8.0, + "learning_rate": 4.8332414806230625e-06, + "loss": 2.4382, + "step": 6781 + }, + { + "epoch": 0.7172165820642978, + "grad_norm": 7.8125, + "learning_rate": 4.833179975927638e-06, + "loss": 2.5214, + "step": 6782 + }, + { + "epoch": 0.7173223350253807, + "grad_norm": 7.84375, + "learning_rate": 4.8331184602835544e-06, + "loss": 2.4972, + "step": 6783 + }, + { + "epoch": 0.7174280879864636, + "grad_norm": 8.0, + "learning_rate": 4.833056933691099e-06, + "loss": 2.527, + "step": 6784 + }, + { + "epoch": 0.7175338409475466, + "grad_norm": 7.4375, + "learning_rate": 4.8329953961505635e-06, + "loss": 2.4487, + "step": 6785 + }, + { + "epoch": 0.7176395939086294, + "grad_norm": 9.0625, + "learning_rate": 4.832933847662234e-06, + "loss": 2.6157, + "step": 6786 + }, + { + "epoch": 0.7177453468697124, + "grad_norm": 8.4375, + "learning_rate": 4.832872288226399e-06, + "loss": 2.622, + "step": 6787 + }, + { + "epoch": 0.7178510998307953, + "grad_norm": 8.1875, + "learning_rate": 4.8328107178433496e-06, + "loss": 2.4853, + "step": 6788 + }, + { + "epoch": 0.7179568527918782, + "grad_norm": 7.625, + "learning_rate": 4.832749136513373e-06, + "loss": 2.5962, + "step": 6789 + }, + { + "epoch": 0.7180626057529611, + "grad_norm": 8.1875, + "learning_rate": 4.832687544236759e-06, + "loss": 2.8985, + "step": 6790 + }, + { + "epoch": 0.7181683587140439, + "grad_norm": 7.6875, + "learning_rate": 4.8326259410137966e-06, + "loss": 2.522, + "step": 6791 + }, + { + "epoch": 0.7182741116751269, + "grad_norm": 7.46875, + "learning_rate": 4.8325643268447744e-06, + "loss": 2.568, + "step": 6792 + }, + { + "epoch": 0.7183798646362098, + "grad_norm": 7.28125, + "learning_rate": 4.832502701729982e-06, + "loss": 2.5322, + "step": 6793 + }, + { + "epoch": 0.7184856175972927, + "grad_norm": 7.5625, + "learning_rate": 4.832441065669708e-06, + "loss": 2.5767, + "step": 6794 + }, + { + "epoch": 0.7185913705583756, + "grad_norm": 8.1875, + "learning_rate": 4.832379418664243e-06, + "loss": 2.5521, + "step": 6795 + }, + { + "epoch": 0.7186971235194586, + "grad_norm": 7.53125, + "learning_rate": 4.832317760713875e-06, + "loss": 2.4652, + "step": 6796 + }, + { + "epoch": 0.7188028764805414, + "grad_norm": 7.6875, + "learning_rate": 4.832256091818892e-06, + "loss": 2.3981, + "step": 6797 + }, + { + "epoch": 0.7189086294416244, + "grad_norm": 7.9375, + "learning_rate": 4.832194411979587e-06, + "loss": 2.5687, + "step": 6798 + }, + { + "epoch": 0.7190143824027073, + "grad_norm": 8.125, + "learning_rate": 4.832132721196247e-06, + "loss": 2.7057, + "step": 6799 + }, + { + "epoch": 0.7191201353637902, + "grad_norm": 7.9375, + "learning_rate": 4.832071019469161e-06, + "loss": 2.6425, + "step": 6800 + }, + { + "epoch": 0.7192258883248731, + "grad_norm": 7.25, + "learning_rate": 4.832009306798621e-06, + "loss": 2.4462, + "step": 6801 + }, + { + "epoch": 0.7193316412859561, + "grad_norm": 8.4375, + "learning_rate": 4.831947583184914e-06, + "loss": 2.6133, + "step": 6802 + }, + { + "epoch": 0.7194373942470389, + "grad_norm": 7.375, + "learning_rate": 4.8318858486283305e-06, + "loss": 2.5821, + "step": 6803 + }, + { + "epoch": 0.7195431472081218, + "grad_norm": 7.78125, + "learning_rate": 4.8318241031291614e-06, + "loss": 2.5469, + "step": 6804 + }, + { + "epoch": 0.7196489001692047, + "grad_norm": 8.125, + "learning_rate": 4.831762346687695e-06, + "loss": 2.5968, + "step": 6805 + }, + { + "epoch": 0.7197546531302876, + "grad_norm": 8.125, + "learning_rate": 4.831700579304221e-06, + "loss": 2.5854, + "step": 6806 + }, + { + "epoch": 0.7198604060913706, + "grad_norm": 8.5, + "learning_rate": 4.83163880097903e-06, + "loss": 2.7883, + "step": 6807 + }, + { + "epoch": 0.7199661590524534, + "grad_norm": 8.375, + "learning_rate": 4.8315770117124125e-06, + "loss": 2.6548, + "step": 6808 + }, + { + "epoch": 0.7200719120135364, + "grad_norm": 7.84375, + "learning_rate": 4.831515211504657e-06, + "loss": 2.609, + "step": 6809 + }, + { + "epoch": 0.7201776649746193, + "grad_norm": 8.0, + "learning_rate": 4.831453400356054e-06, + "loss": 2.5899, + "step": 6810 + }, + { + "epoch": 0.7202834179357022, + "grad_norm": 7.59375, + "learning_rate": 4.831391578266895e-06, + "loss": 2.6145, + "step": 6811 + }, + { + "epoch": 0.7203891708967851, + "grad_norm": 8.6875, + "learning_rate": 4.831329745237468e-06, + "loss": 2.4612, + "step": 6812 + }, + { + "epoch": 0.7204949238578681, + "grad_norm": 8.0, + "learning_rate": 4.831267901268063e-06, + "loss": 2.5745, + "step": 6813 + }, + { + "epoch": 0.7206006768189509, + "grad_norm": 9.375, + "learning_rate": 4.831206046358973e-06, + "loss": 2.5127, + "step": 6814 + }, + { + "epoch": 0.7207064297800339, + "grad_norm": 7.96875, + "learning_rate": 4.8311441805104855e-06, + "loss": 2.6402, + "step": 6815 + }, + { + "epoch": 0.7208121827411168, + "grad_norm": 8.1875, + "learning_rate": 4.831082303722892e-06, + "loss": 2.6572, + "step": 6816 + }, + { + "epoch": 0.7209179357021996, + "grad_norm": 7.84375, + "learning_rate": 4.831020415996483e-06, + "loss": 2.8058, + "step": 6817 + }, + { + "epoch": 0.7210236886632826, + "grad_norm": 7.8125, + "learning_rate": 4.830958517331547e-06, + "loss": 2.5295, + "step": 6818 + }, + { + "epoch": 0.7211294416243654, + "grad_norm": 7.71875, + "learning_rate": 4.830896607728377e-06, + "loss": 2.431, + "step": 6819 + }, + { + "epoch": 0.7212351945854484, + "grad_norm": 8.4375, + "learning_rate": 4.830834687187263e-06, + "loss": 2.6449, + "step": 6820 + }, + { + "epoch": 0.7213409475465313, + "grad_norm": 7.84375, + "learning_rate": 4.830772755708495e-06, + "loss": 2.6381, + "step": 6821 + }, + { + "epoch": 0.7214467005076142, + "grad_norm": 21.25, + "learning_rate": 4.830710813292363e-06, + "loss": 2.5981, + "step": 6822 + }, + { + "epoch": 0.7215524534686971, + "grad_norm": 8.1875, + "learning_rate": 4.830648859939159e-06, + "loss": 2.3007, + "step": 6823 + }, + { + "epoch": 0.7216582064297801, + "grad_norm": 7.71875, + "learning_rate": 4.8305868956491715e-06, + "loss": 2.5287, + "step": 6824 + }, + { + "epoch": 0.7217639593908629, + "grad_norm": 8.4375, + "learning_rate": 4.830524920422694e-06, + "loss": 2.7222, + "step": 6825 + }, + { + "epoch": 0.7218697123519459, + "grad_norm": 9.375, + "learning_rate": 4.830462934260017e-06, + "loss": 2.6008, + "step": 6826 + }, + { + "epoch": 0.7219754653130288, + "grad_norm": 8.0625, + "learning_rate": 4.83040093716143e-06, + "loss": 2.6902, + "step": 6827 + }, + { + "epoch": 0.7220812182741116, + "grad_norm": 7.90625, + "learning_rate": 4.8303389291272254e-06, + "loss": 2.5793, + "step": 6828 + }, + { + "epoch": 0.7221869712351946, + "grad_norm": 7.6875, + "learning_rate": 4.830276910157692e-06, + "loss": 2.6608, + "step": 6829 + }, + { + "epoch": 0.7222927241962775, + "grad_norm": 8.125, + "learning_rate": 4.830214880253122e-06, + "loss": 2.6461, + "step": 6830 + }, + { + "epoch": 0.7223984771573604, + "grad_norm": 7.9375, + "learning_rate": 4.830152839413807e-06, + "loss": 2.5619, + "step": 6831 + }, + { + "epoch": 0.7225042301184433, + "grad_norm": 8.125, + "learning_rate": 4.830090787640038e-06, + "loss": 2.7234, + "step": 6832 + }, + { + "epoch": 0.7226099830795262, + "grad_norm": 7.71875, + "learning_rate": 4.830028724932106e-06, + "loss": 2.4281, + "step": 6833 + }, + { + "epoch": 0.7227157360406091, + "grad_norm": 7.5625, + "learning_rate": 4.829966651290302e-06, + "loss": 2.4871, + "step": 6834 + }, + { + "epoch": 0.7228214890016921, + "grad_norm": 8.375, + "learning_rate": 4.829904566714917e-06, + "loss": 2.371, + "step": 6835 + }, + { + "epoch": 0.7229272419627749, + "grad_norm": 7.78125, + "learning_rate": 4.829842471206243e-06, + "loss": 2.6974, + "step": 6836 + }, + { + "epoch": 0.7230329949238579, + "grad_norm": 7.8125, + "learning_rate": 4.829780364764571e-06, + "loss": 2.1953, + "step": 6837 + }, + { + "epoch": 0.7231387478849408, + "grad_norm": 8.4375, + "learning_rate": 4.829718247390193e-06, + "loss": 2.6672, + "step": 6838 + }, + { + "epoch": 0.7232445008460237, + "grad_norm": 7.96875, + "learning_rate": 4.8296561190834e-06, + "loss": 2.6743, + "step": 6839 + }, + { + "epoch": 0.7233502538071066, + "grad_norm": 8.0, + "learning_rate": 4.829593979844483e-06, + "loss": 2.5431, + "step": 6840 + }, + { + "epoch": 0.7234560067681896, + "grad_norm": 7.84375, + "learning_rate": 4.8295318296737345e-06, + "loss": 2.3714, + "step": 6841 + }, + { + "epoch": 0.7235617597292724, + "grad_norm": 7.15625, + "learning_rate": 4.829469668571447e-06, + "loss": 2.5307, + "step": 6842 + }, + { + "epoch": 0.7236675126903553, + "grad_norm": 7.96875, + "learning_rate": 4.82940749653791e-06, + "loss": 2.5547, + "step": 6843 + }, + { + "epoch": 0.7237732656514383, + "grad_norm": 8.75, + "learning_rate": 4.829345313573417e-06, + "loss": 2.4738, + "step": 6844 + }, + { + "epoch": 0.7238790186125211, + "grad_norm": 7.875, + "learning_rate": 4.829283119678258e-06, + "loss": 2.6643, + "step": 6845 + }, + { + "epoch": 0.7239847715736041, + "grad_norm": 7.375, + "learning_rate": 4.829220914852727e-06, + "loss": 2.4752, + "step": 6846 + }, + { + "epoch": 0.7240905245346869, + "grad_norm": 7.5625, + "learning_rate": 4.829158699097114e-06, + "loss": 2.5768, + "step": 6847 + }, + { + "epoch": 0.7241962774957699, + "grad_norm": 7.59375, + "learning_rate": 4.829096472411712e-06, + "loss": 2.5992, + "step": 6848 + }, + { + "epoch": 0.7243020304568528, + "grad_norm": 8.0625, + "learning_rate": 4.829034234796813e-06, + "loss": 2.5858, + "step": 6849 + }, + { + "epoch": 0.7244077834179357, + "grad_norm": 7.5, + "learning_rate": 4.828971986252709e-06, + "loss": 2.424, + "step": 6850 + }, + { + "epoch": 0.7245135363790186, + "grad_norm": 8.1875, + "learning_rate": 4.8289097267796925e-06, + "loss": 2.7106, + "step": 6851 + }, + { + "epoch": 0.7246192893401016, + "grad_norm": 7.90625, + "learning_rate": 4.828847456378054e-06, + "loss": 2.2489, + "step": 6852 + }, + { + "epoch": 0.7247250423011844, + "grad_norm": 8.625, + "learning_rate": 4.828785175048088e-06, + "loss": 2.5278, + "step": 6853 + }, + { + "epoch": 0.7248307952622673, + "grad_norm": 8.6875, + "learning_rate": 4.828722882790084e-06, + "loss": 2.8532, + "step": 6854 + }, + { + "epoch": 0.7249365482233503, + "grad_norm": 7.90625, + "learning_rate": 4.828660579604338e-06, + "loss": 2.5068, + "step": 6855 + }, + { + "epoch": 0.7250423011844331, + "grad_norm": 8.5625, + "learning_rate": 4.828598265491139e-06, + "loss": 2.6052, + "step": 6856 + }, + { + "epoch": 0.7251480541455161, + "grad_norm": 8.25, + "learning_rate": 4.8285359404507816e-06, + "loss": 2.5354, + "step": 6857 + }, + { + "epoch": 0.725253807106599, + "grad_norm": 10.0, + "learning_rate": 4.828473604483556e-06, + "loss": 2.5968, + "step": 6858 + }, + { + "epoch": 0.7253595600676819, + "grad_norm": 8.125, + "learning_rate": 4.828411257589757e-06, + "loss": 2.8349, + "step": 6859 + }, + { + "epoch": 0.7254653130287648, + "grad_norm": 7.9375, + "learning_rate": 4.828348899769676e-06, + "loss": 2.6465, + "step": 6860 + }, + { + "epoch": 0.7255710659898477, + "grad_norm": 8.125, + "learning_rate": 4.828286531023606e-06, + "loss": 2.6797, + "step": 6861 + }, + { + "epoch": 0.7256768189509306, + "grad_norm": 8.0, + "learning_rate": 4.8282241513518395e-06, + "loss": 2.6711, + "step": 6862 + }, + { + "epoch": 0.7257825719120136, + "grad_norm": 7.96875, + "learning_rate": 4.828161760754669e-06, + "loss": 2.7229, + "step": 6863 + }, + { + "epoch": 0.7258883248730964, + "grad_norm": 8.375, + "learning_rate": 4.828099359232388e-06, + "loss": 2.3878, + "step": 6864 + }, + { + "epoch": 0.7259940778341794, + "grad_norm": 9.0625, + "learning_rate": 4.828036946785289e-06, + "loss": 2.497, + "step": 6865 + }, + { + "epoch": 0.7260998307952623, + "grad_norm": 8.3125, + "learning_rate": 4.827974523413664e-06, + "loss": 2.6107, + "step": 6866 + }, + { + "epoch": 0.7262055837563451, + "grad_norm": 7.78125, + "learning_rate": 4.827912089117808e-06, + "loss": 2.6825, + "step": 6867 + }, + { + "epoch": 0.7263113367174281, + "grad_norm": 7.9375, + "learning_rate": 4.827849643898011e-06, + "loss": 2.4054, + "step": 6868 + }, + { + "epoch": 0.726417089678511, + "grad_norm": 8.1875, + "learning_rate": 4.827787187754568e-06, + "loss": 2.6161, + "step": 6869 + }, + { + "epoch": 0.7265228426395939, + "grad_norm": 8.125, + "learning_rate": 4.827724720687773e-06, + "loss": 2.7132, + "step": 6870 + }, + { + "epoch": 0.7266285956006768, + "grad_norm": 7.78125, + "learning_rate": 4.827662242697917e-06, + "loss": 2.5158, + "step": 6871 + }, + { + "epoch": 0.7267343485617598, + "grad_norm": 8.0625, + "learning_rate": 4.827599753785294e-06, + "loss": 2.5275, + "step": 6872 + }, + { + "epoch": 0.7268401015228426, + "grad_norm": 8.4375, + "learning_rate": 4.827537253950198e-06, + "loss": 2.5967, + "step": 6873 + }, + { + "epoch": 0.7269458544839256, + "grad_norm": 8.5, + "learning_rate": 4.8274747431929206e-06, + "loss": 2.7073, + "step": 6874 + }, + { + "epoch": 0.7270516074450084, + "grad_norm": 7.96875, + "learning_rate": 4.827412221513757e-06, + "loss": 2.572, + "step": 6875 + }, + { + "epoch": 0.7271573604060914, + "grad_norm": 8.5625, + "learning_rate": 4.827349688912999e-06, + "loss": 2.3279, + "step": 6876 + }, + { + "epoch": 0.7272631133671743, + "grad_norm": 7.53125, + "learning_rate": 4.827287145390941e-06, + "loss": 2.5553, + "step": 6877 + }, + { + "epoch": 0.7273688663282571, + "grad_norm": 9.875, + "learning_rate": 4.827224590947876e-06, + "loss": 2.6254, + "step": 6878 + }, + { + "epoch": 0.7274746192893401, + "grad_norm": 8.4375, + "learning_rate": 4.827162025584099e-06, + "loss": 2.4959, + "step": 6879 + }, + { + "epoch": 0.727580372250423, + "grad_norm": 7.875, + "learning_rate": 4.827099449299901e-06, + "loss": 2.5187, + "step": 6880 + }, + { + "epoch": 0.7276861252115059, + "grad_norm": 8.0625, + "learning_rate": 4.827036862095577e-06, + "loss": 2.6914, + "step": 6881 + }, + { + "epoch": 0.7277918781725888, + "grad_norm": 7.375, + "learning_rate": 4.8269742639714215e-06, + "loss": 2.6254, + "step": 6882 + }, + { + "epoch": 0.7278976311336718, + "grad_norm": 7.90625, + "learning_rate": 4.826911654927728e-06, + "loss": 2.6573, + "step": 6883 + }, + { + "epoch": 0.7280033840947546, + "grad_norm": 8.4375, + "learning_rate": 4.826849034964789e-06, + "loss": 2.6557, + "step": 6884 + }, + { + "epoch": 0.7281091370558376, + "grad_norm": 8.0, + "learning_rate": 4.826786404082899e-06, + "loss": 2.6581, + "step": 6885 + }, + { + "epoch": 0.7282148900169205, + "grad_norm": 8.5625, + "learning_rate": 4.826723762282352e-06, + "loss": 2.5888, + "step": 6886 + }, + { + "epoch": 0.7283206429780034, + "grad_norm": 8.0625, + "learning_rate": 4.826661109563442e-06, + "loss": 2.6597, + "step": 6887 + }, + { + "epoch": 0.7284263959390863, + "grad_norm": 8.4375, + "learning_rate": 4.826598445926463e-06, + "loss": 2.5721, + "step": 6888 + }, + { + "epoch": 0.7285321489001692, + "grad_norm": 7.78125, + "learning_rate": 4.82653577137171e-06, + "loss": 2.5732, + "step": 6889 + }, + { + "epoch": 0.7286379018612521, + "grad_norm": 8.0625, + "learning_rate": 4.826473085899474e-06, + "loss": 2.5157, + "step": 6890 + }, + { + "epoch": 0.728743654822335, + "grad_norm": 8.125, + "learning_rate": 4.826410389510053e-06, + "loss": 2.5939, + "step": 6891 + }, + { + "epoch": 0.7288494077834179, + "grad_norm": 8.375, + "learning_rate": 4.82634768220374e-06, + "loss": 2.5773, + "step": 6892 + }, + { + "epoch": 0.7289551607445008, + "grad_norm": 8.0625, + "learning_rate": 4.826284963980827e-06, + "loss": 2.9101, + "step": 6893 + }, + { + "epoch": 0.7290609137055838, + "grad_norm": 7.75, + "learning_rate": 4.826222234841612e-06, + "loss": 2.6594, + "step": 6894 + }, + { + "epoch": 0.7291666666666666, + "grad_norm": 8.1875, + "learning_rate": 4.8261594947863865e-06, + "loss": 2.0898, + "step": 6895 + }, + { + "epoch": 0.7292724196277496, + "grad_norm": 8.75, + "learning_rate": 4.826096743815445e-06, + "loss": 2.9881, + "step": 6896 + }, + { + "epoch": 0.7293781725888325, + "grad_norm": 7.90625, + "learning_rate": 4.826033981929084e-06, + "loss": 2.7737, + "step": 6897 + }, + { + "epoch": 0.7294839255499154, + "grad_norm": 7.84375, + "learning_rate": 4.825971209127597e-06, + "loss": 2.6167, + "step": 6898 + }, + { + "epoch": 0.7295896785109983, + "grad_norm": 7.90625, + "learning_rate": 4.825908425411278e-06, + "loss": 2.6882, + "step": 6899 + }, + { + "epoch": 0.7296954314720813, + "grad_norm": 7.96875, + "learning_rate": 4.825845630780422e-06, + "loss": 2.6461, + "step": 6900 + }, + { + "epoch": 0.7298011844331641, + "grad_norm": 7.34375, + "learning_rate": 4.825782825235323e-06, + "loss": 2.2619, + "step": 6901 + }, + { + "epoch": 0.7299069373942471, + "grad_norm": 9.125, + "learning_rate": 4.825720008776277e-06, + "loss": 2.6641, + "step": 6902 + }, + { + "epoch": 0.7300126903553299, + "grad_norm": 8.0625, + "learning_rate": 4.825657181403578e-06, + "loss": 2.8898, + "step": 6903 + }, + { + "epoch": 0.7301184433164128, + "grad_norm": 8.625, + "learning_rate": 4.825594343117521e-06, + "loss": 2.7094, + "step": 6904 + }, + { + "epoch": 0.7302241962774958, + "grad_norm": 8.375, + "learning_rate": 4.825531493918402e-06, + "loss": 2.3867, + "step": 6905 + }, + { + "epoch": 0.7303299492385786, + "grad_norm": 8.375, + "learning_rate": 4.825468633806514e-06, + "loss": 2.6356, + "step": 6906 + }, + { + "epoch": 0.7304357021996616, + "grad_norm": 7.8125, + "learning_rate": 4.825405762782153e-06, + "loss": 2.5643, + "step": 6907 + }, + { + "epoch": 0.7305414551607445, + "grad_norm": 8.0, + "learning_rate": 4.825342880845612e-06, + "loss": 2.5457, + "step": 6908 + }, + { + "epoch": 0.7306472081218274, + "grad_norm": 7.84375, + "learning_rate": 4.82527998799719e-06, + "loss": 2.6061, + "step": 6909 + }, + { + "epoch": 0.7307529610829103, + "grad_norm": 8.0625, + "learning_rate": 4.825217084237179e-06, + "loss": 2.8067, + "step": 6910 + }, + { + "epoch": 0.7308587140439933, + "grad_norm": 8.0625, + "learning_rate": 4.8251541695658765e-06, + "loss": 2.4425, + "step": 6911 + }, + { + "epoch": 0.7309644670050761, + "grad_norm": 8.0, + "learning_rate": 4.825091243983575e-06, + "loss": 2.5716, + "step": 6912 + }, + { + "epoch": 0.7310702199661591, + "grad_norm": 8.1875, + "learning_rate": 4.8250283074905725e-06, + "loss": 2.6535, + "step": 6913 + }, + { + "epoch": 0.731175972927242, + "grad_norm": 8.125, + "learning_rate": 4.824965360087162e-06, + "loss": 2.1879, + "step": 6914 + }, + { + "epoch": 0.7312817258883249, + "grad_norm": 8.125, + "learning_rate": 4.82490240177364e-06, + "loss": 2.5733, + "step": 6915 + }, + { + "epoch": 0.7313874788494078, + "grad_norm": 8.3125, + "learning_rate": 4.824839432550302e-06, + "loss": 2.8922, + "step": 6916 + }, + { + "epoch": 0.7314932318104906, + "grad_norm": 7.90625, + "learning_rate": 4.824776452417444e-06, + "loss": 2.7427, + "step": 6917 + }, + { + "epoch": 0.7315989847715736, + "grad_norm": 8.125, + "learning_rate": 4.82471346137536e-06, + "loss": 2.5103, + "step": 6918 + }, + { + "epoch": 0.7317047377326565, + "grad_norm": 7.53125, + "learning_rate": 4.824650459424347e-06, + "loss": 2.3216, + "step": 6919 + }, + { + "epoch": 0.7318104906937394, + "grad_norm": 8.125, + "learning_rate": 4.8245874465647e-06, + "loss": 2.6162, + "step": 6920 + }, + { + "epoch": 0.7319162436548223, + "grad_norm": 7.5625, + "learning_rate": 4.824524422796715e-06, + "loss": 2.485, + "step": 6921 + }, + { + "epoch": 0.7320219966159053, + "grad_norm": 7.59375, + "learning_rate": 4.824461388120688e-06, + "loss": 2.649, + "step": 6922 + }, + { + "epoch": 0.7321277495769881, + "grad_norm": 8.4375, + "learning_rate": 4.824398342536914e-06, + "loss": 2.5224, + "step": 6923 + }, + { + "epoch": 0.7322335025380711, + "grad_norm": 7.5625, + "learning_rate": 4.824335286045689e-06, + "loss": 2.4539, + "step": 6924 + }, + { + "epoch": 0.732339255499154, + "grad_norm": 7.90625, + "learning_rate": 4.824272218647309e-06, + "loss": 2.5919, + "step": 6925 + }, + { + "epoch": 0.7324450084602369, + "grad_norm": 8.0625, + "learning_rate": 4.82420914034207e-06, + "loss": 2.6229, + "step": 6926 + }, + { + "epoch": 0.7325507614213198, + "grad_norm": 8.0, + "learning_rate": 4.824146051130269e-06, + "loss": 2.7409, + "step": 6927 + }, + { + "epoch": 0.7326565143824028, + "grad_norm": 7.5, + "learning_rate": 4.8240829510122004e-06, + "loss": 2.5223, + "step": 6928 + }, + { + "epoch": 0.7327622673434856, + "grad_norm": 9.4375, + "learning_rate": 4.824019839988161e-06, + "loss": 2.5812, + "step": 6929 + }, + { + "epoch": 0.7328680203045685, + "grad_norm": 7.8125, + "learning_rate": 4.823956718058448e-06, + "loss": 2.5944, + "step": 6930 + }, + { + "epoch": 0.7329737732656514, + "grad_norm": 8.1875, + "learning_rate": 4.8238935852233546e-06, + "loss": 2.8904, + "step": 6931 + }, + { + "epoch": 0.7330795262267343, + "grad_norm": 8.125, + "learning_rate": 4.823830441483181e-06, + "loss": 2.919, + "step": 6932 + }, + { + "epoch": 0.7331852791878173, + "grad_norm": 8.0, + "learning_rate": 4.82376728683822e-06, + "loss": 2.4913, + "step": 6933 + }, + { + "epoch": 0.7332910321489001, + "grad_norm": 7.6875, + "learning_rate": 4.8237041212887706e-06, + "loss": 2.5382, + "step": 6934 + }, + { + "epoch": 0.7333967851099831, + "grad_norm": 7.5, + "learning_rate": 4.823640944835127e-06, + "loss": 2.5554, + "step": 6935 + }, + { + "epoch": 0.733502538071066, + "grad_norm": 7.375, + "learning_rate": 4.823577757477588e-06, + "loss": 2.5471, + "step": 6936 + }, + { + "epoch": 0.7336082910321489, + "grad_norm": 8.0625, + "learning_rate": 4.823514559216449e-06, + "loss": 2.5289, + "step": 6937 + }, + { + "epoch": 0.7337140439932318, + "grad_norm": 7.75, + "learning_rate": 4.823451350052006e-06, + "loss": 2.5995, + "step": 6938 + }, + { + "epoch": 0.7338197969543148, + "grad_norm": 7.71875, + "learning_rate": 4.823388129984555e-06, + "loss": 2.8852, + "step": 6939 + }, + { + "epoch": 0.7339255499153976, + "grad_norm": 7.53125, + "learning_rate": 4.823324899014395e-06, + "loss": 2.5813, + "step": 6940 + }, + { + "epoch": 0.7340313028764806, + "grad_norm": 7.59375, + "learning_rate": 4.8232616571418215e-06, + "loss": 2.6352, + "step": 6941 + }, + { + "epoch": 0.7341370558375635, + "grad_norm": 7.96875, + "learning_rate": 4.823198404367131e-06, + "loss": 2.4692, + "step": 6942 + }, + { + "epoch": 0.7342428087986463, + "grad_norm": 7.28125, + "learning_rate": 4.82313514069062e-06, + "loss": 2.5256, + "step": 6943 + }, + { + "epoch": 0.7343485617597293, + "grad_norm": 7.65625, + "learning_rate": 4.823071866112587e-06, + "loss": 2.5625, + "step": 6944 + }, + { + "epoch": 0.7344543147208121, + "grad_norm": 7.71875, + "learning_rate": 4.823008580633326e-06, + "loss": 2.8205, + "step": 6945 + }, + { + "epoch": 0.7345600676818951, + "grad_norm": 7.75, + "learning_rate": 4.822945284253138e-06, + "loss": 2.5872, + "step": 6946 + }, + { + "epoch": 0.734665820642978, + "grad_norm": 8.0625, + "learning_rate": 4.822881976972317e-06, + "loss": 2.5826, + "step": 6947 + }, + { + "epoch": 0.7347715736040609, + "grad_norm": 8.375, + "learning_rate": 4.8228186587911605e-06, + "loss": 2.7398, + "step": 6948 + }, + { + "epoch": 0.7348773265651438, + "grad_norm": 7.90625, + "learning_rate": 4.822755329709966e-06, + "loss": 2.7709, + "step": 6949 + }, + { + "epoch": 0.7349830795262268, + "grad_norm": 7.625, + "learning_rate": 4.822691989729031e-06, + "loss": 2.6085, + "step": 6950 + }, + { + "epoch": 0.7350888324873096, + "grad_norm": 9.25, + "learning_rate": 4.822628638848652e-06, + "loss": 2.6693, + "step": 6951 + }, + { + "epoch": 0.7351945854483926, + "grad_norm": 8.3125, + "learning_rate": 4.8225652770691276e-06, + "loss": 2.5387, + "step": 6952 + }, + { + "epoch": 0.7353003384094755, + "grad_norm": 8.0, + "learning_rate": 4.822501904390753e-06, + "loss": 2.5966, + "step": 6953 + }, + { + "epoch": 0.7354060913705583, + "grad_norm": 7.90625, + "learning_rate": 4.822438520813828e-06, + "loss": 2.5901, + "step": 6954 + }, + { + "epoch": 0.7355118443316413, + "grad_norm": 7.9375, + "learning_rate": 4.822375126338648e-06, + "loss": 2.668, + "step": 6955 + }, + { + "epoch": 0.7356175972927242, + "grad_norm": 7.4375, + "learning_rate": 4.822311720965512e-06, + "loss": 2.5365, + "step": 6956 + }, + { + "epoch": 0.7357233502538071, + "grad_norm": 8.25, + "learning_rate": 4.822248304694717e-06, + "loss": 2.8267, + "step": 6957 + }, + { + "epoch": 0.73582910321489, + "grad_norm": 8.375, + "learning_rate": 4.82218487752656e-06, + "loss": 2.4037, + "step": 6958 + }, + { + "epoch": 0.7359348561759729, + "grad_norm": 7.75, + "learning_rate": 4.822121439461339e-06, + "loss": 2.4532, + "step": 6959 + }, + { + "epoch": 0.7360406091370558, + "grad_norm": 9.3125, + "learning_rate": 4.8220579904993524e-06, + "loss": 2.8837, + "step": 6960 + }, + { + "epoch": 0.7361463620981388, + "grad_norm": 7.53125, + "learning_rate": 4.821994530640897e-06, + "loss": 2.4288, + "step": 6961 + }, + { + "epoch": 0.7362521150592216, + "grad_norm": 8.5, + "learning_rate": 4.821931059886271e-06, + "loss": 2.6302, + "step": 6962 + }, + { + "epoch": 0.7363578680203046, + "grad_norm": 8.4375, + "learning_rate": 4.821867578235773e-06, + "loss": 2.5411, + "step": 6963 + }, + { + "epoch": 0.7364636209813875, + "grad_norm": 7.90625, + "learning_rate": 4.821804085689699e-06, + "loss": 2.4932, + "step": 6964 + }, + { + "epoch": 0.7365693739424704, + "grad_norm": 7.5625, + "learning_rate": 4.8217405822483475e-06, + "loss": 2.6291, + "step": 6965 + }, + { + "epoch": 0.7366751269035533, + "grad_norm": 7.65625, + "learning_rate": 4.821677067912018e-06, + "loss": 2.6428, + "step": 6966 + }, + { + "epoch": 0.7367808798646363, + "grad_norm": 8.25, + "learning_rate": 4.821613542681008e-06, + "loss": 2.5844, + "step": 6967 + }, + { + "epoch": 0.7368866328257191, + "grad_norm": 7.78125, + "learning_rate": 4.8215500065556145e-06, + "loss": 2.3634, + "step": 6968 + }, + { + "epoch": 0.736992385786802, + "grad_norm": 7.90625, + "learning_rate": 4.821486459536136e-06, + "loss": 2.2453, + "step": 6969 + }, + { + "epoch": 0.737098138747885, + "grad_norm": 8.25, + "learning_rate": 4.821422901622871e-06, + "loss": 2.6778, + "step": 6970 + }, + { + "epoch": 0.7372038917089678, + "grad_norm": 9.0625, + "learning_rate": 4.8213593328161175e-06, + "loss": 2.3693, + "step": 6971 + }, + { + "epoch": 0.7373096446700508, + "grad_norm": 7.875, + "learning_rate": 4.821295753116175e-06, + "loss": 2.4169, + "step": 6972 + }, + { + "epoch": 0.7374153976311336, + "grad_norm": 8.6875, + "learning_rate": 4.82123216252334e-06, + "loss": 2.4475, + "step": 6973 + }, + { + "epoch": 0.7375211505922166, + "grad_norm": 8.0, + "learning_rate": 4.821168561037913e-06, + "loss": 2.6355, + "step": 6974 + }, + { + "epoch": 0.7376269035532995, + "grad_norm": 8.5, + "learning_rate": 4.82110494866019e-06, + "loss": 2.4685, + "step": 6975 + }, + { + "epoch": 0.7377326565143824, + "grad_norm": 11.25, + "learning_rate": 4.821041325390471e-06, + "loss": 2.5149, + "step": 6976 + }, + { + "epoch": 0.7378384094754653, + "grad_norm": 8.1875, + "learning_rate": 4.820977691229055e-06, + "loss": 2.5735, + "step": 6977 + }, + { + "epoch": 0.7379441624365483, + "grad_norm": 7.875, + "learning_rate": 4.820914046176239e-06, + "loss": 2.536, + "step": 6978 + }, + { + "epoch": 0.7380499153976311, + "grad_norm": 8.25, + "learning_rate": 4.820850390232324e-06, + "loss": 2.6443, + "step": 6979 + }, + { + "epoch": 0.738155668358714, + "grad_norm": 7.96875, + "learning_rate": 4.8207867233976065e-06, + "loss": 2.435, + "step": 6980 + }, + { + "epoch": 0.738261421319797, + "grad_norm": 8.1875, + "learning_rate": 4.8207230456723865e-06, + "loss": 2.5379, + "step": 6981 + }, + { + "epoch": 0.7383671742808798, + "grad_norm": 7.40625, + "learning_rate": 4.820659357056961e-06, + "loss": 2.4174, + "step": 6982 + }, + { + "epoch": 0.7384729272419628, + "grad_norm": 7.34375, + "learning_rate": 4.820595657551632e-06, + "loss": 2.4877, + "step": 6983 + }, + { + "epoch": 0.7385786802030457, + "grad_norm": 8.125, + "learning_rate": 4.820531947156696e-06, + "loss": 2.7002, + "step": 6984 + }, + { + "epoch": 0.7386844331641286, + "grad_norm": 7.8125, + "learning_rate": 4.8204682258724525e-06, + "loss": 2.4762, + "step": 6985 + }, + { + "epoch": 0.7387901861252115, + "grad_norm": 8.75, + "learning_rate": 4.8204044936992e-06, + "loss": 2.6366, + "step": 6986 + }, + { + "epoch": 0.7388959390862944, + "grad_norm": 8.0, + "learning_rate": 4.820340750637239e-06, + "loss": 2.6633, + "step": 6987 + }, + { + "epoch": 0.7390016920473773, + "grad_norm": 7.6875, + "learning_rate": 4.820276996686868e-06, + "loss": 2.5867, + "step": 6988 + }, + { + "epoch": 0.7391074450084603, + "grad_norm": 7.75, + "learning_rate": 4.820213231848386e-06, + "loss": 2.5313, + "step": 6989 + }, + { + "epoch": 0.7392131979695431, + "grad_norm": 7.96875, + "learning_rate": 4.820149456122092e-06, + "loss": 2.7402, + "step": 6990 + }, + { + "epoch": 0.739318950930626, + "grad_norm": 7.65625, + "learning_rate": 4.820085669508285e-06, + "loss": 2.6498, + "step": 6991 + }, + { + "epoch": 0.739424703891709, + "grad_norm": 8.3125, + "learning_rate": 4.820021872007265e-06, + "loss": 2.7657, + "step": 6992 + }, + { + "epoch": 0.7395304568527918, + "grad_norm": 8.0, + "learning_rate": 4.819958063619332e-06, + "loss": 2.644, + "step": 6993 + }, + { + "epoch": 0.7396362098138748, + "grad_norm": 8.5625, + "learning_rate": 4.819894244344784e-06, + "loss": 2.5805, + "step": 6994 + }, + { + "epoch": 0.7397419627749577, + "grad_norm": 7.8125, + "learning_rate": 4.819830414183922e-06, + "loss": 2.4478, + "step": 6995 + }, + { + "epoch": 0.7398477157360406, + "grad_norm": 8.0625, + "learning_rate": 4.819766573137044e-06, + "loss": 2.7327, + "step": 6996 + }, + { + "epoch": 0.7399534686971235, + "grad_norm": 7.8125, + "learning_rate": 4.81970272120445e-06, + "loss": 2.6043, + "step": 6997 + }, + { + "epoch": 0.7400592216582065, + "grad_norm": 8.25, + "learning_rate": 4.81963885838644e-06, + "loss": 2.6592, + "step": 6998 + }, + { + "epoch": 0.7401649746192893, + "grad_norm": 7.96875, + "learning_rate": 4.819574984683314e-06, + "loss": 2.399, + "step": 6999 + }, + { + "epoch": 0.7402707275803723, + "grad_norm": 7.96875, + "learning_rate": 4.81951110009537e-06, + "loss": 2.7577, + "step": 7000 + }, + { + "epoch": 0.7403764805414551, + "grad_norm": 7.90625, + "learning_rate": 4.81944720462291e-06, + "loss": 2.5335, + "step": 7001 + }, + { + "epoch": 0.7404822335025381, + "grad_norm": 8.25, + "learning_rate": 4.819383298266233e-06, + "loss": 2.6595, + "step": 7002 + }, + { + "epoch": 0.740587986463621, + "grad_norm": 7.6875, + "learning_rate": 4.8193193810256385e-06, + "loss": 2.5479, + "step": 7003 + }, + { + "epoch": 0.7406937394247038, + "grad_norm": 8.25, + "learning_rate": 4.819255452901427e-06, + "loss": 2.5105, + "step": 7004 + }, + { + "epoch": 0.7407994923857868, + "grad_norm": 7.875, + "learning_rate": 4.819191513893898e-06, + "loss": 2.8934, + "step": 7005 + }, + { + "epoch": 0.7409052453468697, + "grad_norm": 8.0, + "learning_rate": 4.819127564003352e-06, + "loss": 2.5716, + "step": 7006 + }, + { + "epoch": 0.7410109983079526, + "grad_norm": 8.125, + "learning_rate": 4.819063603230088e-06, + "loss": 2.7581, + "step": 7007 + }, + { + "epoch": 0.7411167512690355, + "grad_norm": 8.5, + "learning_rate": 4.818999631574408e-06, + "loss": 2.7647, + "step": 7008 + }, + { + "epoch": 0.7412225042301185, + "grad_norm": 7.625, + "learning_rate": 4.81893564903661e-06, + "loss": 2.5878, + "step": 7009 + }, + { + "epoch": 0.7413282571912013, + "grad_norm": 7.78125, + "learning_rate": 4.8188716556169965e-06, + "loss": 2.5432, + "step": 7010 + }, + { + "epoch": 0.7414340101522843, + "grad_norm": 7.65625, + "learning_rate": 4.8188076513158665e-06, + "loss": 2.4776, + "step": 7011 + }, + { + "epoch": 0.7415397631133672, + "grad_norm": 7.84375, + "learning_rate": 4.818743636133521e-06, + "loss": 2.6572, + "step": 7012 + }, + { + "epoch": 0.7416455160744501, + "grad_norm": 9.625, + "learning_rate": 4.8186796100702585e-06, + "loss": 2.4737, + "step": 7013 + }, + { + "epoch": 0.741751269035533, + "grad_norm": 7.875, + "learning_rate": 4.818615573126382e-06, + "loss": 2.4784, + "step": 7014 + }, + { + "epoch": 0.7418570219966159, + "grad_norm": 7.46875, + "learning_rate": 4.81855152530219e-06, + "loss": 2.5721, + "step": 7015 + }, + { + "epoch": 0.7419627749576988, + "grad_norm": 7.8125, + "learning_rate": 4.8184874665979855e-06, + "loss": 2.5116, + "step": 7016 + }, + { + "epoch": 0.7420685279187818, + "grad_norm": 8.375, + "learning_rate": 4.818423397014066e-06, + "loss": 2.6142, + "step": 7017 + }, + { + "epoch": 0.7421742808798646, + "grad_norm": 7.875, + "learning_rate": 4.818359316550735e-06, + "loss": 2.5588, + "step": 7018 + }, + { + "epoch": 0.7422800338409475, + "grad_norm": 8.5, + "learning_rate": 4.8182952252082905e-06, + "loss": 2.5009, + "step": 7019 + }, + { + "epoch": 0.7423857868020305, + "grad_norm": 8.6875, + "learning_rate": 4.8182311229870355e-06, + "loss": 2.8868, + "step": 7020 + }, + { + "epoch": 0.7424915397631133, + "grad_norm": 8.1875, + "learning_rate": 4.81816700988727e-06, + "loss": 2.6532, + "step": 7021 + }, + { + "epoch": 0.7425972927241963, + "grad_norm": 8.5625, + "learning_rate": 4.818102885909294e-06, + "loss": 2.6316, + "step": 7022 + }, + { + "epoch": 0.7427030456852792, + "grad_norm": 8.25, + "learning_rate": 4.81803875105341e-06, + "loss": 2.5659, + "step": 7023 + }, + { + "epoch": 0.7428087986463621, + "grad_norm": 7.3125, + "learning_rate": 4.817974605319918e-06, + "loss": 2.5356, + "step": 7024 + }, + { + "epoch": 0.742914551607445, + "grad_norm": 7.625, + "learning_rate": 4.81791044870912e-06, + "loss": 2.482, + "step": 7025 + }, + { + "epoch": 0.743020304568528, + "grad_norm": 7.84375, + "learning_rate": 4.817846281221315e-06, + "loss": 2.4771, + "step": 7026 + }, + { + "epoch": 0.7431260575296108, + "grad_norm": 8.1875, + "learning_rate": 4.817782102856806e-06, + "loss": 2.7167, + "step": 7027 + }, + { + "epoch": 0.7432318104906938, + "grad_norm": 7.71875, + "learning_rate": 4.8177179136158935e-06, + "loss": 2.6252, + "step": 7028 + }, + { + "epoch": 0.7433375634517766, + "grad_norm": 7.96875, + "learning_rate": 4.8176537134988784e-06, + "loss": 2.5231, + "step": 7029 + }, + { + "epoch": 0.7434433164128595, + "grad_norm": 7.75, + "learning_rate": 4.817589502506063e-06, + "loss": 2.5378, + "step": 7030 + }, + { + "epoch": 0.7435490693739425, + "grad_norm": 9.75, + "learning_rate": 4.8175252806377475e-06, + "loss": 2.829, + "step": 7031 + }, + { + "epoch": 0.7436548223350253, + "grad_norm": 8.375, + "learning_rate": 4.817461047894234e-06, + "loss": 2.4237, + "step": 7032 + }, + { + "epoch": 0.7437605752961083, + "grad_norm": 8.1875, + "learning_rate": 4.817396804275823e-06, + "loss": 2.4832, + "step": 7033 + }, + { + "epoch": 0.7438663282571912, + "grad_norm": 7.8125, + "learning_rate": 4.8173325497828175e-06, + "loss": 2.5601, + "step": 7034 + }, + { + "epoch": 0.7439720812182741, + "grad_norm": 7.875, + "learning_rate": 4.817268284415516e-06, + "loss": 2.6156, + "step": 7035 + }, + { + "epoch": 0.744077834179357, + "grad_norm": 8.1875, + "learning_rate": 4.817204008174224e-06, + "loss": 2.7187, + "step": 7036 + }, + { + "epoch": 0.74418358714044, + "grad_norm": 7.625, + "learning_rate": 4.817139721059241e-06, + "loss": 2.6327, + "step": 7037 + }, + { + "epoch": 0.7442893401015228, + "grad_norm": 8.75, + "learning_rate": 4.817075423070869e-06, + "loss": 2.5222, + "step": 7038 + }, + { + "epoch": 0.7443950930626058, + "grad_norm": 8.1875, + "learning_rate": 4.81701111420941e-06, + "loss": 2.5189, + "step": 7039 + }, + { + "epoch": 0.7445008460236887, + "grad_norm": 7.3125, + "learning_rate": 4.816946794475165e-06, + "loss": 2.5914, + "step": 7040 + }, + { + "epoch": 0.7446065989847716, + "grad_norm": 8.1875, + "learning_rate": 4.816882463868437e-06, + "loss": 2.6837, + "step": 7041 + }, + { + "epoch": 0.7447123519458545, + "grad_norm": 7.71875, + "learning_rate": 4.816818122389526e-06, + "loss": 2.6253, + "step": 7042 + }, + { + "epoch": 0.7448181049069373, + "grad_norm": 8.5, + "learning_rate": 4.816753770038736e-06, + "loss": 2.4017, + "step": 7043 + }, + { + "epoch": 0.7449238578680203, + "grad_norm": 8.0625, + "learning_rate": 4.816689406816367e-06, + "loss": 2.7164, + "step": 7044 + }, + { + "epoch": 0.7450296108291032, + "grad_norm": 7.9375, + "learning_rate": 4.816625032722723e-06, + "loss": 2.6185, + "step": 7045 + }, + { + "epoch": 0.7451353637901861, + "grad_norm": 8.5, + "learning_rate": 4.816560647758105e-06, + "loss": 2.7822, + "step": 7046 + }, + { + "epoch": 0.745241116751269, + "grad_norm": 8.125, + "learning_rate": 4.816496251922816e-06, + "loss": 2.7158, + "step": 7047 + }, + { + "epoch": 0.745346869712352, + "grad_norm": 7.71875, + "learning_rate": 4.816431845217157e-06, + "loss": 2.4986, + "step": 7048 + }, + { + "epoch": 0.7454526226734348, + "grad_norm": 7.375, + "learning_rate": 4.81636742764143e-06, + "loss": 2.539, + "step": 7049 + }, + { + "epoch": 0.7455583756345178, + "grad_norm": 7.875, + "learning_rate": 4.8163029991959385e-06, + "loss": 2.6015, + "step": 7050 + }, + { + "epoch": 0.7456641285956007, + "grad_norm": 8.375, + "learning_rate": 4.816238559880985e-06, + "loss": 2.6624, + "step": 7051 + }, + { + "epoch": 0.7457698815566836, + "grad_norm": 8.0625, + "learning_rate": 4.816174109696871e-06, + "loss": 2.3394, + "step": 7052 + }, + { + "epoch": 0.7458756345177665, + "grad_norm": 8.4375, + "learning_rate": 4.816109648643898e-06, + "loss": 2.3264, + "step": 7053 + }, + { + "epoch": 0.7459813874788495, + "grad_norm": 8.75, + "learning_rate": 4.8160451767223706e-06, + "loss": 2.4048, + "step": 7054 + }, + { + "epoch": 0.7460871404399323, + "grad_norm": 7.90625, + "learning_rate": 4.81598069393259e-06, + "loss": 2.689, + "step": 7055 + }, + { + "epoch": 0.7461928934010152, + "grad_norm": 8.25, + "learning_rate": 4.815916200274861e-06, + "loss": 2.5947, + "step": 7056 + }, + { + "epoch": 0.7462986463620981, + "grad_norm": 8.875, + "learning_rate": 4.815851695749482e-06, + "loss": 2.8172, + "step": 7057 + }, + { + "epoch": 0.746404399323181, + "grad_norm": 8.4375, + "learning_rate": 4.8157871803567595e-06, + "loss": 2.3894, + "step": 7058 + }, + { + "epoch": 0.746510152284264, + "grad_norm": 8.6875, + "learning_rate": 4.815722654096995e-06, + "loss": 2.6713, + "step": 7059 + }, + { + "epoch": 0.7466159052453468, + "grad_norm": 8.4375, + "learning_rate": 4.81565811697049e-06, + "loss": 2.5682, + "step": 7060 + }, + { + "epoch": 0.7467216582064298, + "grad_norm": 8.0, + "learning_rate": 4.815593568977549e-06, + "loss": 2.6916, + "step": 7061 + }, + { + "epoch": 0.7468274111675127, + "grad_norm": 8.875, + "learning_rate": 4.815529010118475e-06, + "loss": 2.7772, + "step": 7062 + }, + { + "epoch": 0.7469331641285956, + "grad_norm": 7.78125, + "learning_rate": 4.81546444039357e-06, + "loss": 2.7273, + "step": 7063 + }, + { + "epoch": 0.7470389170896785, + "grad_norm": 8.5625, + "learning_rate": 4.815399859803138e-06, + "loss": 2.3948, + "step": 7064 + }, + { + "epoch": 0.7471446700507615, + "grad_norm": 8.3125, + "learning_rate": 4.815335268347481e-06, + "loss": 2.6321, + "step": 7065 + }, + { + "epoch": 0.7472504230118443, + "grad_norm": 8.25, + "learning_rate": 4.815270666026902e-06, + "loss": 2.5641, + "step": 7066 + }, + { + "epoch": 0.7473561759729273, + "grad_norm": 8.0625, + "learning_rate": 4.8152060528417055e-06, + "loss": 2.6198, + "step": 7067 + }, + { + "epoch": 0.7474619289340102, + "grad_norm": 8.375, + "learning_rate": 4.815141428792193e-06, + "loss": 2.7006, + "step": 7068 + }, + { + "epoch": 0.747567681895093, + "grad_norm": 7.8125, + "learning_rate": 4.8150767938786695e-06, + "loss": 2.4467, + "step": 7069 + }, + { + "epoch": 0.747673434856176, + "grad_norm": 8.3125, + "learning_rate": 4.8150121481014365e-06, + "loss": 2.5432, + "step": 7070 + }, + { + "epoch": 0.7477791878172588, + "grad_norm": 8.0625, + "learning_rate": 4.814947491460799e-06, + "loss": 2.3319, + "step": 7071 + }, + { + "epoch": 0.7478849407783418, + "grad_norm": 8.25, + "learning_rate": 4.81488282395706e-06, + "loss": 2.8285, + "step": 7072 + }, + { + "epoch": 0.7479906937394247, + "grad_norm": 7.34375, + "learning_rate": 4.814818145590522e-06, + "loss": 2.4154, + "step": 7073 + }, + { + "epoch": 0.7480964467005076, + "grad_norm": 8.0, + "learning_rate": 4.81475345636149e-06, + "loss": 2.5225, + "step": 7074 + }, + { + "epoch": 0.7482021996615905, + "grad_norm": 8.125, + "learning_rate": 4.8146887562702655e-06, + "loss": 2.461, + "step": 7075 + }, + { + "epoch": 0.7483079526226735, + "grad_norm": 7.96875, + "learning_rate": 4.814624045317154e-06, + "loss": 2.7715, + "step": 7076 + }, + { + "epoch": 0.7484137055837563, + "grad_norm": 8.0625, + "learning_rate": 4.814559323502459e-06, + "loss": 2.6515, + "step": 7077 + }, + { + "epoch": 0.7485194585448393, + "grad_norm": 7.71875, + "learning_rate": 4.8144945908264825e-06, + "loss": 2.5255, + "step": 7078 + }, + { + "epoch": 0.7486252115059222, + "grad_norm": 8.5, + "learning_rate": 4.81442984728953e-06, + "loss": 2.7037, + "step": 7079 + }, + { + "epoch": 0.748730964467005, + "grad_norm": 7.875, + "learning_rate": 4.814365092891905e-06, + "loss": 2.4534, + "step": 7080 + }, + { + "epoch": 0.748836717428088, + "grad_norm": 8.625, + "learning_rate": 4.814300327633911e-06, + "loss": 2.4539, + "step": 7081 + }, + { + "epoch": 0.748942470389171, + "grad_norm": 8.25, + "learning_rate": 4.814235551515853e-06, + "loss": 2.7171, + "step": 7082 + }, + { + "epoch": 0.7490482233502538, + "grad_norm": 7.71875, + "learning_rate": 4.8141707645380335e-06, + "loss": 2.7421, + "step": 7083 + }, + { + "epoch": 0.7491539763113367, + "grad_norm": 8.375, + "learning_rate": 4.814105966700756e-06, + "loss": 2.7445, + "step": 7084 + }, + { + "epoch": 0.7492597292724196, + "grad_norm": 8.5, + "learning_rate": 4.814041158004327e-06, + "loss": 2.5547, + "step": 7085 + }, + { + "epoch": 0.7493654822335025, + "grad_norm": 7.84375, + "learning_rate": 4.813976338449048e-06, + "loss": 2.5086, + "step": 7086 + }, + { + "epoch": 0.7494712351945855, + "grad_norm": 8.0, + "learning_rate": 4.813911508035225e-06, + "loss": 2.5674, + "step": 7087 + }, + { + "epoch": 0.7495769881556683, + "grad_norm": 7.90625, + "learning_rate": 4.813846666763162e-06, + "loss": 2.6274, + "step": 7088 + }, + { + "epoch": 0.7496827411167513, + "grad_norm": 7.8125, + "learning_rate": 4.813781814633163e-06, + "loss": 2.5779, + "step": 7089 + }, + { + "epoch": 0.7497884940778342, + "grad_norm": 8.9375, + "learning_rate": 4.813716951645532e-06, + "loss": 2.5897, + "step": 7090 + }, + { + "epoch": 0.749894247038917, + "grad_norm": 8.6875, + "learning_rate": 4.813652077800574e-06, + "loss": 2.8487, + "step": 7091 + }, + { + "epoch": 0.75, + "grad_norm": 8.25, + "learning_rate": 4.8135871930985925e-06, + "loss": 2.6801, + "step": 7092 + }, + { + "epoch": 0.750105752961083, + "grad_norm": 7.53125, + "learning_rate": 4.8135222975398925e-06, + "loss": 2.4196, + "step": 7093 + }, + { + "epoch": 0.7502115059221658, + "grad_norm": 7.5625, + "learning_rate": 4.813457391124778e-06, + "loss": 2.6867, + "step": 7094 + }, + { + "epoch": 0.7503172588832487, + "grad_norm": 8.3125, + "learning_rate": 4.813392473853556e-06, + "loss": 2.6203, + "step": 7095 + }, + { + "epoch": 0.7504230118443317, + "grad_norm": 8.25, + "learning_rate": 4.813327545726527e-06, + "loss": 2.7049, + "step": 7096 + }, + { + "epoch": 0.7505287648054145, + "grad_norm": 8.25, + "learning_rate": 4.813262606743999e-06, + "loss": 2.5505, + "step": 7097 + }, + { + "epoch": 0.7506345177664975, + "grad_norm": 7.8125, + "learning_rate": 4.813197656906276e-06, + "loss": 2.4304, + "step": 7098 + }, + { + "epoch": 0.7507402707275804, + "grad_norm": 8.1875, + "learning_rate": 4.813132696213663e-06, + "loss": 2.6436, + "step": 7099 + }, + { + "epoch": 0.7508460236886633, + "grad_norm": 8.25, + "learning_rate": 4.813067724666462e-06, + "loss": 2.7362, + "step": 7100 + }, + { + "epoch": 0.7509517766497462, + "grad_norm": 8.4375, + "learning_rate": 4.813002742264983e-06, + "loss": 2.8317, + "step": 7101 + }, + { + "epoch": 0.751057529610829, + "grad_norm": 8.375, + "learning_rate": 4.812937749009526e-06, + "loss": 2.425, + "step": 7102 + }, + { + "epoch": 0.751163282571912, + "grad_norm": 8.25, + "learning_rate": 4.812872744900399e-06, + "loss": 2.6252, + "step": 7103 + }, + { + "epoch": 0.751269035532995, + "grad_norm": 7.875, + "learning_rate": 4.812807729937906e-06, + "loss": 2.5566, + "step": 7104 + }, + { + "epoch": 0.7513747884940778, + "grad_norm": 9.0, + "learning_rate": 4.812742704122353e-06, + "loss": 2.4047, + "step": 7105 + }, + { + "epoch": 0.7514805414551607, + "grad_norm": 8.5, + "learning_rate": 4.812677667454043e-06, + "loss": 2.6601, + "step": 7106 + }, + { + "epoch": 0.7515862944162437, + "grad_norm": 9.0625, + "learning_rate": 4.812612619933283e-06, + "loss": 2.5318, + "step": 7107 + }, + { + "epoch": 0.7516920473773265, + "grad_norm": 7.375, + "learning_rate": 4.812547561560378e-06, + "loss": 2.594, + "step": 7108 + }, + { + "epoch": 0.7517978003384095, + "grad_norm": 8.25, + "learning_rate": 4.812482492335633e-06, + "loss": 2.5942, + "step": 7109 + }, + { + "epoch": 0.7519035532994924, + "grad_norm": 7.8125, + "learning_rate": 4.812417412259354e-06, + "loss": 2.4874, + "step": 7110 + }, + { + "epoch": 0.7520093062605753, + "grad_norm": 7.59375, + "learning_rate": 4.8123523213318455e-06, + "loss": 2.559, + "step": 7111 + }, + { + "epoch": 0.7521150592216582, + "grad_norm": 7.84375, + "learning_rate": 4.812287219553413e-06, + "loss": 2.6473, + "step": 7112 + }, + { + "epoch": 0.7522208121827412, + "grad_norm": 9.25, + "learning_rate": 4.812222106924363e-06, + "loss": 2.5596, + "step": 7113 + }, + { + "epoch": 0.752326565143824, + "grad_norm": 8.1875, + "learning_rate": 4.812156983444999e-06, + "loss": 2.5346, + "step": 7114 + }, + { + "epoch": 0.752432318104907, + "grad_norm": 8.1875, + "learning_rate": 4.812091849115629e-06, + "loss": 2.5639, + "step": 7115 + }, + { + "epoch": 0.7525380710659898, + "grad_norm": 8.0, + "learning_rate": 4.8120267039365565e-06, + "loss": 2.4465, + "step": 7116 + }, + { + "epoch": 0.7526438240270727, + "grad_norm": 8.1875, + "learning_rate": 4.81196154790809e-06, + "loss": 2.4121, + "step": 7117 + }, + { + "epoch": 0.7527495769881557, + "grad_norm": 7.875, + "learning_rate": 4.811896381030532e-06, + "loss": 2.6413, + "step": 7118 + }, + { + "epoch": 0.7528553299492385, + "grad_norm": 7.40625, + "learning_rate": 4.81183120330419e-06, + "loss": 2.4764, + "step": 7119 + }, + { + "epoch": 0.7529610829103215, + "grad_norm": 8.3125, + "learning_rate": 4.81176601472937e-06, + "loss": 2.6279, + "step": 7120 + }, + { + "epoch": 0.7530668358714044, + "grad_norm": 8.0, + "learning_rate": 4.811700815306377e-06, + "loss": 2.54, + "step": 7121 + }, + { + "epoch": 0.7531725888324873, + "grad_norm": 8.5, + "learning_rate": 4.811635605035519e-06, + "loss": 2.7439, + "step": 7122 + }, + { + "epoch": 0.7532783417935702, + "grad_norm": 7.8125, + "learning_rate": 4.811570383917099e-06, + "loss": 2.4579, + "step": 7123 + }, + { + "epoch": 0.7533840947546532, + "grad_norm": 7.875, + "learning_rate": 4.811505151951426e-06, + "loss": 2.6878, + "step": 7124 + }, + { + "epoch": 0.753489847715736, + "grad_norm": 8.25, + "learning_rate": 4.811439909138803e-06, + "loss": 2.7574, + "step": 7125 + }, + { + "epoch": 0.753595600676819, + "grad_norm": 7.84375, + "learning_rate": 4.8113746554795395e-06, + "loss": 2.4143, + "step": 7126 + }, + { + "epoch": 0.7537013536379019, + "grad_norm": 8.1875, + "learning_rate": 4.811309390973939e-06, + "loss": 2.6757, + "step": 7127 + }, + { + "epoch": 0.7538071065989848, + "grad_norm": 7.9375, + "learning_rate": 4.811244115622309e-06, + "loss": 2.6613, + "step": 7128 + }, + { + "epoch": 0.7539128595600677, + "grad_norm": 7.84375, + "learning_rate": 4.8111788294249555e-06, + "loss": 2.7474, + "step": 7129 + }, + { + "epoch": 0.7540186125211505, + "grad_norm": 7.84375, + "learning_rate": 4.8111135323821854e-06, + "loss": 2.5249, + "step": 7130 + }, + { + "epoch": 0.7541243654822335, + "grad_norm": 8.0625, + "learning_rate": 4.8110482244943045e-06, + "loss": 2.6091, + "step": 7131 + }, + { + "epoch": 0.7542301184433164, + "grad_norm": 7.5625, + "learning_rate": 4.81098290576162e-06, + "loss": 2.5029, + "step": 7132 + }, + { + "epoch": 0.7543358714043993, + "grad_norm": 8.25, + "learning_rate": 4.810917576184437e-06, + "loss": 2.7107, + "step": 7133 + }, + { + "epoch": 0.7544416243654822, + "grad_norm": 7.84375, + "learning_rate": 4.810852235763064e-06, + "loss": 2.7886, + "step": 7134 + }, + { + "epoch": 0.7545473773265652, + "grad_norm": 7.6875, + "learning_rate": 4.8107868844978055e-06, + "loss": 2.6259, + "step": 7135 + }, + { + "epoch": 0.754653130287648, + "grad_norm": 8.375, + "learning_rate": 4.810721522388969e-06, + "loss": 2.5417, + "step": 7136 + }, + { + "epoch": 0.754758883248731, + "grad_norm": 7.34375, + "learning_rate": 4.810656149436863e-06, + "loss": 2.4607, + "step": 7137 + }, + { + "epoch": 0.7548646362098139, + "grad_norm": 8.5625, + "learning_rate": 4.810590765641791e-06, + "loss": 2.4492, + "step": 7138 + }, + { + "epoch": 0.7549703891708968, + "grad_norm": 8.8125, + "learning_rate": 4.810525371004062e-06, + "loss": 2.6487, + "step": 7139 + }, + { + "epoch": 0.7550761421319797, + "grad_norm": 7.90625, + "learning_rate": 4.810459965523983e-06, + "loss": 2.5167, + "step": 7140 + }, + { + "epoch": 0.7551818950930627, + "grad_norm": 8.25, + "learning_rate": 4.810394549201861e-06, + "loss": 2.779, + "step": 7141 + }, + { + "epoch": 0.7552876480541455, + "grad_norm": 7.71875, + "learning_rate": 4.810329122038001e-06, + "loss": 2.6227, + "step": 7142 + }, + { + "epoch": 0.7553934010152284, + "grad_norm": 8.1875, + "learning_rate": 4.810263684032712e-06, + "loss": 2.5092, + "step": 7143 + }, + { + "epoch": 0.7554991539763113, + "grad_norm": 8.125, + "learning_rate": 4.8101982351862996e-06, + "loss": 2.6273, + "step": 7144 + }, + { + "epoch": 0.7556049069373942, + "grad_norm": 8.25, + "learning_rate": 4.8101327754990726e-06, + "loss": 2.4419, + "step": 7145 + }, + { + "epoch": 0.7557106598984772, + "grad_norm": 7.46875, + "learning_rate": 4.810067304971337e-06, + "loss": 2.7347, + "step": 7146 + }, + { + "epoch": 0.75581641285956, + "grad_norm": 8.0625, + "learning_rate": 4.810001823603399e-06, + "loss": 2.6351, + "step": 7147 + }, + { + "epoch": 0.755922165820643, + "grad_norm": 7.90625, + "learning_rate": 4.8099363313955695e-06, + "loss": 2.5793, + "step": 7148 + }, + { + "epoch": 0.7560279187817259, + "grad_norm": 8.625, + "learning_rate": 4.8098708283481525e-06, + "loss": 2.6053, + "step": 7149 + }, + { + "epoch": 0.7561336717428088, + "grad_norm": 8.3125, + "learning_rate": 4.8098053144614565e-06, + "loss": 2.8534, + "step": 7150 + }, + { + "epoch": 0.7562394247038917, + "grad_norm": 7.8125, + "learning_rate": 4.809739789735788e-06, + "loss": 2.5539, + "step": 7151 + }, + { + "epoch": 0.7563451776649747, + "grad_norm": 7.84375, + "learning_rate": 4.809674254171456e-06, + "loss": 2.5924, + "step": 7152 + }, + { + "epoch": 0.7564509306260575, + "grad_norm": 7.5, + "learning_rate": 4.809608707768768e-06, + "loss": 2.4237, + "step": 7153 + }, + { + "epoch": 0.7565566835871405, + "grad_norm": 8.0, + "learning_rate": 4.80954315052803e-06, + "loss": 2.5064, + "step": 7154 + }, + { + "epoch": 0.7566624365482234, + "grad_norm": 8.3125, + "learning_rate": 4.809477582449551e-06, + "loss": 2.7839, + "step": 7155 + }, + { + "epoch": 0.7567681895093062, + "grad_norm": 8.0625, + "learning_rate": 4.809412003533638e-06, + "loss": 2.5717, + "step": 7156 + }, + { + "epoch": 0.7568739424703892, + "grad_norm": 8.0, + "learning_rate": 4.809346413780599e-06, + "loss": 2.5128, + "step": 7157 + }, + { + "epoch": 0.756979695431472, + "grad_norm": 8.75, + "learning_rate": 4.809280813190742e-06, + "loss": 2.7009, + "step": 7158 + }, + { + "epoch": 0.757085448392555, + "grad_norm": 7.9375, + "learning_rate": 4.809215201764374e-06, + "loss": 2.6665, + "step": 7159 + }, + { + "epoch": 0.7571912013536379, + "grad_norm": 8.25, + "learning_rate": 4.809149579501804e-06, + "loss": 2.4368, + "step": 7160 + }, + { + "epoch": 0.7572969543147208, + "grad_norm": 7.8125, + "learning_rate": 4.80908394640334e-06, + "loss": 2.437, + "step": 7161 + }, + { + "epoch": 0.7574027072758037, + "grad_norm": 7.96875, + "learning_rate": 4.809018302469289e-06, + "loss": 2.6171, + "step": 7162 + }, + { + "epoch": 0.7575084602368867, + "grad_norm": 8.375, + "learning_rate": 4.808952647699959e-06, + "loss": 2.7857, + "step": 7163 + }, + { + "epoch": 0.7576142131979695, + "grad_norm": 8.375, + "learning_rate": 4.808886982095659e-06, + "loss": 2.7573, + "step": 7164 + }, + { + "epoch": 0.7577199661590525, + "grad_norm": 8.125, + "learning_rate": 4.808821305656697e-06, + "loss": 2.3833, + "step": 7165 + }, + { + "epoch": 0.7578257191201354, + "grad_norm": 8.0, + "learning_rate": 4.80875561838338e-06, + "loss": 2.432, + "step": 7166 + }, + { + "epoch": 0.7579314720812182, + "grad_norm": 7.84375, + "learning_rate": 4.808689920276018e-06, + "loss": 2.5853, + "step": 7167 + }, + { + "epoch": 0.7580372250423012, + "grad_norm": 8.375, + "learning_rate": 4.8086242113349175e-06, + "loss": 2.5826, + "step": 7168 + }, + { + "epoch": 0.7581429780033841, + "grad_norm": 7.5625, + "learning_rate": 4.808558491560388e-06, + "loss": 2.4772, + "step": 7169 + }, + { + "epoch": 0.758248730964467, + "grad_norm": 8.375, + "learning_rate": 4.808492760952738e-06, + "loss": 2.6249, + "step": 7170 + }, + { + "epoch": 0.7583544839255499, + "grad_norm": 7.5625, + "learning_rate": 4.808427019512276e-06, + "loss": 2.5814, + "step": 7171 + }, + { + "epoch": 0.7584602368866328, + "grad_norm": 7.78125, + "learning_rate": 4.8083612672393095e-06, + "loss": 2.5646, + "step": 7172 + }, + { + "epoch": 0.7585659898477157, + "grad_norm": 8.75, + "learning_rate": 4.808295504134148e-06, + "loss": 2.4884, + "step": 7173 + }, + { + "epoch": 0.7586717428087987, + "grad_norm": 8.1875, + "learning_rate": 4.8082297301970995e-06, + "loss": 2.6445, + "step": 7174 + }, + { + "epoch": 0.7587774957698815, + "grad_norm": 8.25, + "learning_rate": 4.808163945428473e-06, + "loss": 2.5774, + "step": 7175 + }, + { + "epoch": 0.7588832487309645, + "grad_norm": 8.3125, + "learning_rate": 4.8080981498285775e-06, + "loss": 2.5379, + "step": 7176 + }, + { + "epoch": 0.7589890016920474, + "grad_norm": 7.9375, + "learning_rate": 4.808032343397721e-06, + "loss": 2.6462, + "step": 7177 + }, + { + "epoch": 0.7590947546531303, + "grad_norm": 8.375, + "learning_rate": 4.807966526136212e-06, + "loss": 2.7401, + "step": 7178 + }, + { + "epoch": 0.7592005076142132, + "grad_norm": 7.4375, + "learning_rate": 4.807900698044361e-06, + "loss": 2.3836, + "step": 7179 + }, + { + "epoch": 0.7593062605752962, + "grad_norm": 9.4375, + "learning_rate": 4.807834859122475e-06, + "loss": 2.6646, + "step": 7180 + }, + { + "epoch": 0.759412013536379, + "grad_norm": 7.65625, + "learning_rate": 4.807769009370865e-06, + "loss": 2.5548, + "step": 7181 + }, + { + "epoch": 0.7595177664974619, + "grad_norm": 8.75, + "learning_rate": 4.8077031487898385e-06, + "loss": 2.8905, + "step": 7182 + }, + { + "epoch": 0.7596235194585449, + "grad_norm": 7.625, + "learning_rate": 4.8076372773797045e-06, + "loss": 2.5333, + "step": 7183 + }, + { + "epoch": 0.7597292724196277, + "grad_norm": 8.0625, + "learning_rate": 4.807571395140773e-06, + "loss": 2.6852, + "step": 7184 + }, + { + "epoch": 0.7598350253807107, + "grad_norm": 7.6875, + "learning_rate": 4.807505502073352e-06, + "loss": 2.6117, + "step": 7185 + }, + { + "epoch": 0.7599407783417935, + "grad_norm": 8.5, + "learning_rate": 4.807439598177752e-06, + "loss": 2.8139, + "step": 7186 + }, + { + "epoch": 0.7600465313028765, + "grad_norm": 8.0625, + "learning_rate": 4.807373683454282e-06, + "loss": 2.7477, + "step": 7187 + }, + { + "epoch": 0.7601522842639594, + "grad_norm": 7.9375, + "learning_rate": 4.80730775790325e-06, + "loss": 2.5241, + "step": 7188 + }, + { + "epoch": 0.7602580372250423, + "grad_norm": 8.0, + "learning_rate": 4.807241821524968e-06, + "loss": 2.5216, + "step": 7189 + }, + { + "epoch": 0.7603637901861252, + "grad_norm": 7.90625, + "learning_rate": 4.807175874319742e-06, + "loss": 2.7078, + "step": 7190 + }, + { + "epoch": 0.7604695431472082, + "grad_norm": 7.875, + "learning_rate": 4.8071099162878846e-06, + "loss": 2.6081, + "step": 7191 + }, + { + "epoch": 0.760575296108291, + "grad_norm": 8.6875, + "learning_rate": 4.8070439474297025e-06, + "loss": 2.463, + "step": 7192 + }, + { + "epoch": 0.760681049069374, + "grad_norm": 8.0625, + "learning_rate": 4.806977967745508e-06, + "loss": 2.5863, + "step": 7193 + }, + { + "epoch": 0.7607868020304569, + "grad_norm": 7.5625, + "learning_rate": 4.806911977235608e-06, + "loss": 2.6399, + "step": 7194 + }, + { + "epoch": 0.7608925549915397, + "grad_norm": 7.625, + "learning_rate": 4.806845975900315e-06, + "loss": 2.5582, + "step": 7195 + }, + { + "epoch": 0.7609983079526227, + "grad_norm": 7.625, + "learning_rate": 4.806779963739937e-06, + "loss": 2.5054, + "step": 7196 + }, + { + "epoch": 0.7611040609137056, + "grad_norm": 7.53125, + "learning_rate": 4.806713940754784e-06, + "loss": 2.5194, + "step": 7197 + }, + { + "epoch": 0.7612098138747885, + "grad_norm": 8.0, + "learning_rate": 4.806647906945166e-06, + "loss": 2.4804, + "step": 7198 + }, + { + "epoch": 0.7613155668358714, + "grad_norm": 8.1875, + "learning_rate": 4.8065818623113925e-06, + "loss": 2.7556, + "step": 7199 + }, + { + "epoch": 0.7614213197969543, + "grad_norm": 8.1875, + "learning_rate": 4.806515806853773e-06, + "loss": 2.4879, + "step": 7200 + }, + { + "epoch": 0.7615270727580372, + "grad_norm": 7.75, + "learning_rate": 4.806449740572619e-06, + "loss": 2.5816, + "step": 7201 + }, + { + "epoch": 0.7616328257191202, + "grad_norm": 8.125, + "learning_rate": 4.806383663468239e-06, + "loss": 2.5734, + "step": 7202 + }, + { + "epoch": 0.761738578680203, + "grad_norm": 7.90625, + "learning_rate": 4.806317575540945e-06, + "loss": 2.5747, + "step": 7203 + }, + { + "epoch": 0.761844331641286, + "grad_norm": 7.71875, + "learning_rate": 4.806251476791045e-06, + "loss": 2.4071, + "step": 7204 + }, + { + "epoch": 0.7619500846023689, + "grad_norm": 8.25, + "learning_rate": 4.80618536721885e-06, + "loss": 2.5942, + "step": 7205 + }, + { + "epoch": 0.7620558375634517, + "grad_norm": 8.1875, + "learning_rate": 4.806119246824671e-06, + "loss": 2.6798, + "step": 7206 + }, + { + "epoch": 0.7621615905245347, + "grad_norm": 7.71875, + "learning_rate": 4.806053115608817e-06, + "loss": 2.5833, + "step": 7207 + }, + { + "epoch": 0.7622673434856176, + "grad_norm": 7.78125, + "learning_rate": 4.8059869735715985e-06, + "loss": 2.4591, + "step": 7208 + }, + { + "epoch": 0.7623730964467005, + "grad_norm": 8.0625, + "learning_rate": 4.805920820713327e-06, + "loss": 2.5128, + "step": 7209 + }, + { + "epoch": 0.7624788494077834, + "grad_norm": 11.125, + "learning_rate": 4.805854657034311e-06, + "loss": 2.5461, + "step": 7210 + }, + { + "epoch": 0.7625846023688664, + "grad_norm": 7.53125, + "learning_rate": 4.805788482534863e-06, + "loss": 2.5799, + "step": 7211 + }, + { + "epoch": 0.7626903553299492, + "grad_norm": 7.78125, + "learning_rate": 4.805722297215293e-06, + "loss": 2.6542, + "step": 7212 + }, + { + "epoch": 0.7627961082910322, + "grad_norm": 8.25, + "learning_rate": 4.8056561010759106e-06, + "loss": 2.5348, + "step": 7213 + }, + { + "epoch": 0.762901861252115, + "grad_norm": 7.84375, + "learning_rate": 4.805589894117028e-06, + "loss": 2.4891, + "step": 7214 + }, + { + "epoch": 0.763007614213198, + "grad_norm": 7.84375, + "learning_rate": 4.805523676338954e-06, + "loss": 2.5501, + "step": 7215 + }, + { + "epoch": 0.7631133671742809, + "grad_norm": 7.65625, + "learning_rate": 4.805457447742e-06, + "loss": 2.5271, + "step": 7216 + }, + { + "epoch": 0.7632191201353637, + "grad_norm": 7.78125, + "learning_rate": 4.805391208326479e-06, + "loss": 2.5133, + "step": 7217 + }, + { + "epoch": 0.7633248730964467, + "grad_norm": 8.0, + "learning_rate": 4.805324958092699e-06, + "loss": 2.5041, + "step": 7218 + }, + { + "epoch": 0.7634306260575296, + "grad_norm": 7.75, + "learning_rate": 4.805258697040971e-06, + "loss": 2.5275, + "step": 7219 + }, + { + "epoch": 0.7635363790186125, + "grad_norm": 7.8125, + "learning_rate": 4.805192425171607e-06, + "loss": 2.6505, + "step": 7220 + }, + { + "epoch": 0.7636421319796954, + "grad_norm": 7.78125, + "learning_rate": 4.805126142484918e-06, + "loss": 2.6566, + "step": 7221 + }, + { + "epoch": 0.7637478849407784, + "grad_norm": 8.3125, + "learning_rate": 4.805059848981215e-06, + "loss": 2.7737, + "step": 7222 + }, + { + "epoch": 0.7638536379018612, + "grad_norm": 8.5625, + "learning_rate": 4.8049935446608086e-06, + "loss": 2.7359, + "step": 7223 + }, + { + "epoch": 0.7639593908629442, + "grad_norm": 7.9375, + "learning_rate": 4.80492722952401e-06, + "loss": 2.4683, + "step": 7224 + }, + { + "epoch": 0.7640651438240271, + "grad_norm": 7.53125, + "learning_rate": 4.804860903571131e-06, + "loss": 2.5624, + "step": 7225 + }, + { + "epoch": 0.76417089678511, + "grad_norm": 8.125, + "learning_rate": 4.8047945668024825e-06, + "loss": 2.4674, + "step": 7226 + }, + { + "epoch": 0.7642766497461929, + "grad_norm": 8.0625, + "learning_rate": 4.804728219218376e-06, + "loss": 2.7906, + "step": 7227 + }, + { + "epoch": 0.7643824027072758, + "grad_norm": 8.5, + "learning_rate": 4.8046618608191224e-06, + "loss": 2.5256, + "step": 7228 + }, + { + "epoch": 0.7644881556683587, + "grad_norm": 7.6875, + "learning_rate": 4.804595491605033e-06, + "loss": 2.6616, + "step": 7229 + }, + { + "epoch": 0.7645939086294417, + "grad_norm": 8.0, + "learning_rate": 4.804529111576419e-06, + "loss": 2.5498, + "step": 7230 + }, + { + "epoch": 0.7646996615905245, + "grad_norm": 8.0625, + "learning_rate": 4.804462720733592e-06, + "loss": 2.4789, + "step": 7231 + }, + { + "epoch": 0.7648054145516074, + "grad_norm": 8.3125, + "learning_rate": 4.804396319076866e-06, + "loss": 2.768, + "step": 7232 + }, + { + "epoch": 0.7649111675126904, + "grad_norm": 8.3125, + "learning_rate": 4.804329906606548e-06, + "loss": 2.7015, + "step": 7233 + }, + { + "epoch": 0.7650169204737732, + "grad_norm": 9.0, + "learning_rate": 4.804263483322955e-06, + "loss": 2.3765, + "step": 7234 + }, + { + "epoch": 0.7651226734348562, + "grad_norm": 7.90625, + "learning_rate": 4.8041970492263936e-06, + "loss": 2.527, + "step": 7235 + }, + { + "epoch": 0.7652284263959391, + "grad_norm": 7.9375, + "learning_rate": 4.804130604317179e-06, + "loss": 2.5585, + "step": 7236 + }, + { + "epoch": 0.765334179357022, + "grad_norm": 8.1875, + "learning_rate": 4.804064148595622e-06, + "loss": 2.7614, + "step": 7237 + }, + { + "epoch": 0.7654399323181049, + "grad_norm": 7.8125, + "learning_rate": 4.803997682062033e-06, + "loss": 2.4941, + "step": 7238 + }, + { + "epoch": 0.7655456852791879, + "grad_norm": 7.6875, + "learning_rate": 4.803931204716726e-06, + "loss": 2.5513, + "step": 7239 + }, + { + "epoch": 0.7656514382402707, + "grad_norm": 7.65625, + "learning_rate": 4.803864716560013e-06, + "loss": 2.4743, + "step": 7240 + }, + { + "epoch": 0.7657571912013537, + "grad_norm": 7.40625, + "learning_rate": 4.803798217592203e-06, + "loss": 2.4479, + "step": 7241 + }, + { + "epoch": 0.7658629441624365, + "grad_norm": 8.625, + "learning_rate": 4.803731707813612e-06, + "loss": 2.5198, + "step": 7242 + }, + { + "epoch": 0.7659686971235194, + "grad_norm": 7.6875, + "learning_rate": 4.80366518722455e-06, + "loss": 2.5786, + "step": 7243 + }, + { + "epoch": 0.7660744500846024, + "grad_norm": 7.65625, + "learning_rate": 4.8035986558253285e-06, + "loss": 2.4449, + "step": 7244 + }, + { + "epoch": 0.7661802030456852, + "grad_norm": 8.3125, + "learning_rate": 4.803532113616262e-06, + "loss": 2.5231, + "step": 7245 + }, + { + "epoch": 0.7662859560067682, + "grad_norm": 7.46875, + "learning_rate": 4.803465560597661e-06, + "loss": 2.54, + "step": 7246 + }, + { + "epoch": 0.7663917089678511, + "grad_norm": 8.0, + "learning_rate": 4.803398996769837e-06, + "loss": 2.5455, + "step": 7247 + }, + { + "epoch": 0.766497461928934, + "grad_norm": 7.90625, + "learning_rate": 4.803332422133105e-06, + "loss": 2.5777, + "step": 7248 + }, + { + "epoch": 0.7666032148900169, + "grad_norm": 7.6875, + "learning_rate": 4.803265836687775e-06, + "loss": 2.6029, + "step": 7249 + }, + { + "epoch": 0.7667089678510999, + "grad_norm": 7.96875, + "learning_rate": 4.803199240434161e-06, + "loss": 2.5745, + "step": 7250 + }, + { + "epoch": 0.7668147208121827, + "grad_norm": 7.8125, + "learning_rate": 4.803132633372575e-06, + "loss": 2.6095, + "step": 7251 + }, + { + "epoch": 0.7669204737732657, + "grad_norm": 7.90625, + "learning_rate": 4.803066015503329e-06, + "loss": 2.5718, + "step": 7252 + }, + { + "epoch": 0.7670262267343486, + "grad_norm": 8.25, + "learning_rate": 4.802999386826737e-06, + "loss": 2.7221, + "step": 7253 + }, + { + "epoch": 0.7671319796954315, + "grad_norm": 7.59375, + "learning_rate": 4.80293274734311e-06, + "loss": 2.6213, + "step": 7254 + }, + { + "epoch": 0.7672377326565144, + "grad_norm": 8.25, + "learning_rate": 4.802866097052762e-06, + "loss": 2.4964, + "step": 7255 + }, + { + "epoch": 0.7673434856175972, + "grad_norm": 8.125, + "learning_rate": 4.802799435956005e-06, + "loss": 2.6435, + "step": 7256 + }, + { + "epoch": 0.7674492385786802, + "grad_norm": 7.78125, + "learning_rate": 4.802732764053152e-06, + "loss": 2.6845, + "step": 7257 + }, + { + "epoch": 0.7675549915397631, + "grad_norm": 7.84375, + "learning_rate": 4.802666081344516e-06, + "loss": 2.7155, + "step": 7258 + }, + { + "epoch": 0.767660744500846, + "grad_norm": 8.1875, + "learning_rate": 4.802599387830409e-06, + "loss": 2.4926, + "step": 7259 + }, + { + "epoch": 0.7677664974619289, + "grad_norm": 8.125, + "learning_rate": 4.802532683511146e-06, + "loss": 2.649, + "step": 7260 + }, + { + "epoch": 0.7678722504230119, + "grad_norm": 9.625, + "learning_rate": 4.802465968387038e-06, + "loss": 2.6648, + "step": 7261 + }, + { + "epoch": 0.7679780033840947, + "grad_norm": 7.5625, + "learning_rate": 4.8023992424584e-06, + "loss": 2.6121, + "step": 7262 + }, + { + "epoch": 0.7680837563451777, + "grad_norm": 7.78125, + "learning_rate": 4.802332505725542e-06, + "loss": 2.6353, + "step": 7263 + }, + { + "epoch": 0.7681895093062606, + "grad_norm": 7.65625, + "learning_rate": 4.802265758188781e-06, + "loss": 2.6821, + "step": 7264 + }, + { + "epoch": 0.7682952622673435, + "grad_norm": 7.8125, + "learning_rate": 4.802198999848428e-06, + "loss": 2.6304, + "step": 7265 + }, + { + "epoch": 0.7684010152284264, + "grad_norm": 8.25, + "learning_rate": 4.802132230704797e-06, + "loss": 2.524, + "step": 7266 + }, + { + "epoch": 0.7685067681895094, + "grad_norm": 8.4375, + "learning_rate": 4.8020654507581995e-06, + "loss": 2.8622, + "step": 7267 + }, + { + "epoch": 0.7686125211505922, + "grad_norm": 8.4375, + "learning_rate": 4.801998660008952e-06, + "loss": 2.8445, + "step": 7268 + }, + { + "epoch": 0.7687182741116751, + "grad_norm": 8.0625, + "learning_rate": 4.8019318584573655e-06, + "loss": 2.7509, + "step": 7269 + }, + { + "epoch": 0.768824027072758, + "grad_norm": 8.875, + "learning_rate": 4.801865046103754e-06, + "loss": 2.6191, + "step": 7270 + }, + { + "epoch": 0.7689297800338409, + "grad_norm": 8.1875, + "learning_rate": 4.801798222948431e-06, + "loss": 2.7843, + "step": 7271 + }, + { + "epoch": 0.7690355329949239, + "grad_norm": 7.84375, + "learning_rate": 4.801731388991711e-06, + "loss": 2.3889, + "step": 7272 + }, + { + "epoch": 0.7691412859560067, + "grad_norm": 7.84375, + "learning_rate": 4.8016645442339076e-06, + "loss": 2.4487, + "step": 7273 + }, + { + "epoch": 0.7692470389170897, + "grad_norm": 7.46875, + "learning_rate": 4.8015976886753334e-06, + "loss": 2.613, + "step": 7274 + }, + { + "epoch": 0.7693527918781726, + "grad_norm": 8.625, + "learning_rate": 4.801530822316302e-06, + "loss": 2.6587, + "step": 7275 + }, + { + "epoch": 0.7694585448392555, + "grad_norm": 7.65625, + "learning_rate": 4.8014639451571285e-06, + "loss": 2.568, + "step": 7276 + }, + { + "epoch": 0.7695642978003384, + "grad_norm": 8.4375, + "learning_rate": 4.801397057198125e-06, + "loss": 2.4101, + "step": 7277 + }, + { + "epoch": 0.7696700507614214, + "grad_norm": 8.375, + "learning_rate": 4.801330158439608e-06, + "loss": 2.6765, + "step": 7278 + }, + { + "epoch": 0.7697758037225042, + "grad_norm": 8.4375, + "learning_rate": 4.801263248881888e-06, + "loss": 2.4966, + "step": 7279 + }, + { + "epoch": 0.7698815566835872, + "grad_norm": 9.0, + "learning_rate": 4.801196328525282e-06, + "loss": 2.6399, + "step": 7280 + }, + { + "epoch": 0.7699873096446701, + "grad_norm": 7.875, + "learning_rate": 4.801129397370102e-06, + "loss": 2.5296, + "step": 7281 + }, + { + "epoch": 0.7700930626057529, + "grad_norm": 8.5625, + "learning_rate": 4.801062455416663e-06, + "loss": 2.5213, + "step": 7282 + }, + { + "epoch": 0.7701988155668359, + "grad_norm": 10.8125, + "learning_rate": 4.800995502665279e-06, + "loss": 2.7652, + "step": 7283 + }, + { + "epoch": 0.7703045685279187, + "grad_norm": 7.9375, + "learning_rate": 4.800928539116264e-06, + "loss": 2.4925, + "step": 7284 + }, + { + "epoch": 0.7704103214890017, + "grad_norm": 8.375, + "learning_rate": 4.8008615647699334e-06, + "loss": 2.7592, + "step": 7285 + }, + { + "epoch": 0.7705160744500846, + "grad_norm": 8.125, + "learning_rate": 4.800794579626599e-06, + "loss": 2.4609, + "step": 7286 + }, + { + "epoch": 0.7706218274111675, + "grad_norm": 8.4375, + "learning_rate": 4.800727583686577e-06, + "loss": 2.5415, + "step": 7287 + }, + { + "epoch": 0.7707275803722504, + "grad_norm": 7.46875, + "learning_rate": 4.800660576950183e-06, + "loss": 2.3176, + "step": 7288 + }, + { + "epoch": 0.7708333333333334, + "grad_norm": 7.46875, + "learning_rate": 4.800593559417728e-06, + "loss": 2.4525, + "step": 7289 + }, + { + "epoch": 0.7709390862944162, + "grad_norm": 7.5625, + "learning_rate": 4.800526531089529e-06, + "loss": 2.5205, + "step": 7290 + }, + { + "epoch": 0.7710448392554992, + "grad_norm": 8.375, + "learning_rate": 4.8004594919659e-06, + "loss": 2.7262, + "step": 7291 + }, + { + "epoch": 0.7711505922165821, + "grad_norm": 7.59375, + "learning_rate": 4.800392442047155e-06, + "loss": 2.675, + "step": 7292 + }, + { + "epoch": 0.771256345177665, + "grad_norm": 7.84375, + "learning_rate": 4.80032538133361e-06, + "loss": 2.8103, + "step": 7293 + }, + { + "epoch": 0.7713620981387479, + "grad_norm": 7.65625, + "learning_rate": 4.800258309825578e-06, + "loss": 2.6253, + "step": 7294 + }, + { + "epoch": 0.7714678510998308, + "grad_norm": 7.40625, + "learning_rate": 4.800191227523375e-06, + "loss": 2.5776, + "step": 7295 + }, + { + "epoch": 0.7715736040609137, + "grad_norm": 7.96875, + "learning_rate": 4.800124134427315e-06, + "loss": 2.8175, + "step": 7296 + }, + { + "epoch": 0.7716793570219966, + "grad_norm": 7.59375, + "learning_rate": 4.800057030537714e-06, + "loss": 2.5857, + "step": 7297 + }, + { + "epoch": 0.7717851099830795, + "grad_norm": 7.78125, + "learning_rate": 4.799989915854885e-06, + "loss": 2.5871, + "step": 7298 + }, + { + "epoch": 0.7718908629441624, + "grad_norm": 8.625, + "learning_rate": 4.799922790379143e-06, + "loss": 2.5442, + "step": 7299 + }, + { + "epoch": 0.7719966159052454, + "grad_norm": 7.90625, + "learning_rate": 4.799855654110806e-06, + "loss": 2.6884, + "step": 7300 + }, + { + "epoch": 0.7721023688663282, + "grad_norm": 7.71875, + "learning_rate": 4.799788507050186e-06, + "loss": 2.4935, + "step": 7301 + }, + { + "epoch": 0.7722081218274112, + "grad_norm": 8.5, + "learning_rate": 4.7997213491975995e-06, + "loss": 2.5865, + "step": 7302 + }, + { + "epoch": 0.7723138747884941, + "grad_norm": 8.375, + "learning_rate": 4.799654180553362e-06, + "loss": 2.6547, + "step": 7303 + }, + { + "epoch": 0.772419627749577, + "grad_norm": 7.625, + "learning_rate": 4.799587001117786e-06, + "loss": 2.467, + "step": 7304 + }, + { + "epoch": 0.7725253807106599, + "grad_norm": 7.125, + "learning_rate": 4.79951981089119e-06, + "loss": 2.4322, + "step": 7305 + }, + { + "epoch": 0.7726311336717429, + "grad_norm": 7.59375, + "learning_rate": 4.799452609873888e-06, + "loss": 2.6739, + "step": 7306 + }, + { + "epoch": 0.7727368866328257, + "grad_norm": 8.4375, + "learning_rate": 4.7993853980661955e-06, + "loss": 2.5216, + "step": 7307 + }, + { + "epoch": 0.7728426395939086, + "grad_norm": 7.84375, + "learning_rate": 4.799318175468427e-06, + "loss": 2.5912, + "step": 7308 + }, + { + "epoch": 0.7729483925549916, + "grad_norm": 7.90625, + "learning_rate": 4.799250942080899e-06, + "loss": 2.6459, + "step": 7309 + }, + { + "epoch": 0.7730541455160744, + "grad_norm": 8.0, + "learning_rate": 4.7991836979039275e-06, + "loss": 2.7172, + "step": 7310 + }, + { + "epoch": 0.7731598984771574, + "grad_norm": 7.84375, + "learning_rate": 4.799116442937826e-06, + "loss": 2.7496, + "step": 7311 + }, + { + "epoch": 0.7732656514382402, + "grad_norm": 7.6875, + "learning_rate": 4.799049177182912e-06, + "loss": 2.6313, + "step": 7312 + }, + { + "epoch": 0.7733714043993232, + "grad_norm": 7.5625, + "learning_rate": 4.7989819006395e-06, + "loss": 2.5496, + "step": 7313 + }, + { + "epoch": 0.7734771573604061, + "grad_norm": 8.3125, + "learning_rate": 4.798914613307907e-06, + "loss": 2.6921, + "step": 7314 + }, + { + "epoch": 0.773582910321489, + "grad_norm": 8.5, + "learning_rate": 4.798847315188447e-06, + "loss": 2.6659, + "step": 7315 + }, + { + "epoch": 0.7736886632825719, + "grad_norm": 8.125, + "learning_rate": 4.798780006281438e-06, + "loss": 2.3238, + "step": 7316 + }, + { + "epoch": 0.7737944162436549, + "grad_norm": 7.84375, + "learning_rate": 4.798712686587194e-06, + "loss": 2.6605, + "step": 7317 + }, + { + "epoch": 0.7739001692047377, + "grad_norm": 7.5625, + "learning_rate": 4.798645356106031e-06, + "loss": 2.628, + "step": 7318 + }, + { + "epoch": 0.7740059221658206, + "grad_norm": 7.9375, + "learning_rate": 4.7985780148382655e-06, + "loss": 2.4414, + "step": 7319 + }, + { + "epoch": 0.7741116751269036, + "grad_norm": 7.875, + "learning_rate": 4.798510662784213e-06, + "loss": 2.7589, + "step": 7320 + }, + { + "epoch": 0.7742174280879864, + "grad_norm": 7.5625, + "learning_rate": 4.798443299944191e-06, + "loss": 2.6048, + "step": 7321 + }, + { + "epoch": 0.7743231810490694, + "grad_norm": 8.5, + "learning_rate": 4.798375926318515e-06, + "loss": 2.3972, + "step": 7322 + }, + { + "epoch": 0.7744289340101523, + "grad_norm": 7.90625, + "learning_rate": 4.798308541907501e-06, + "loss": 2.7258, + "step": 7323 + }, + { + "epoch": 0.7745346869712352, + "grad_norm": 7.9375, + "learning_rate": 4.798241146711463e-06, + "loss": 2.4995, + "step": 7324 + }, + { + "epoch": 0.7746404399323181, + "grad_norm": 7.71875, + "learning_rate": 4.79817374073072e-06, + "loss": 2.6777, + "step": 7325 + }, + { + "epoch": 0.774746192893401, + "grad_norm": 7.375, + "learning_rate": 4.798106323965589e-06, + "loss": 2.5563, + "step": 7326 + }, + { + "epoch": 0.7748519458544839, + "grad_norm": 7.3125, + "learning_rate": 4.798038896416384e-06, + "loss": 2.6327, + "step": 7327 + }, + { + "epoch": 0.7749576988155669, + "grad_norm": 8.3125, + "learning_rate": 4.797971458083422e-06, + "loss": 2.7511, + "step": 7328 + }, + { + "epoch": 0.7750634517766497, + "grad_norm": 7.53125, + "learning_rate": 4.79790400896702e-06, + "loss": 2.6739, + "step": 7329 + }, + { + "epoch": 0.7751692047377327, + "grad_norm": 7.90625, + "learning_rate": 4.797836549067494e-06, + "loss": 2.2809, + "step": 7330 + }, + { + "epoch": 0.7752749576988156, + "grad_norm": 7.8125, + "learning_rate": 4.797769078385162e-06, + "loss": 2.7605, + "step": 7331 + }, + { + "epoch": 0.7753807106598984, + "grad_norm": 8.0, + "learning_rate": 4.797701596920338e-06, + "loss": 2.614, + "step": 7332 + }, + { + "epoch": 0.7754864636209814, + "grad_norm": 7.5625, + "learning_rate": 4.797634104673341e-06, + "loss": 2.364, + "step": 7333 + }, + { + "epoch": 0.7755922165820643, + "grad_norm": 8.25, + "learning_rate": 4.797566601644487e-06, + "loss": 2.7164, + "step": 7334 + }, + { + "epoch": 0.7756979695431472, + "grad_norm": 8.1875, + "learning_rate": 4.797499087834092e-06, + "loss": 2.4603, + "step": 7335 + }, + { + "epoch": 0.7758037225042301, + "grad_norm": 7.90625, + "learning_rate": 4.797431563242473e-06, + "loss": 2.5379, + "step": 7336 + }, + { + "epoch": 0.7759094754653131, + "grad_norm": 7.8125, + "learning_rate": 4.797364027869949e-06, + "loss": 2.6754, + "step": 7337 + }, + { + "epoch": 0.7760152284263959, + "grad_norm": 7.9375, + "learning_rate": 4.797296481716833e-06, + "loss": 2.5534, + "step": 7338 + }, + { + "epoch": 0.7761209813874789, + "grad_norm": 8.125, + "learning_rate": 4.797228924783446e-06, + "loss": 2.5284, + "step": 7339 + }, + { + "epoch": 0.7762267343485617, + "grad_norm": 8.25, + "learning_rate": 4.797161357070102e-06, + "loss": 2.8509, + "step": 7340 + }, + { + "epoch": 0.7763324873096447, + "grad_norm": 8.5625, + "learning_rate": 4.79709377857712e-06, + "loss": 2.1987, + "step": 7341 + }, + { + "epoch": 0.7764382402707276, + "grad_norm": 8.8125, + "learning_rate": 4.797026189304817e-06, + "loss": 2.5803, + "step": 7342 + }, + { + "epoch": 0.7765439932318104, + "grad_norm": 7.5, + "learning_rate": 4.796958589253507e-06, + "loss": 2.6332, + "step": 7343 + }, + { + "epoch": 0.7766497461928934, + "grad_norm": 7.75, + "learning_rate": 4.796890978423512e-06, + "loss": 2.5742, + "step": 7344 + }, + { + "epoch": 0.7767554991539763, + "grad_norm": 8.8125, + "learning_rate": 4.796823356815146e-06, + "loss": 2.8795, + "step": 7345 + }, + { + "epoch": 0.7768612521150592, + "grad_norm": 7.78125, + "learning_rate": 4.7967557244287286e-06, + "loss": 2.5696, + "step": 7346 + }, + { + "epoch": 0.7769670050761421, + "grad_norm": 7.625, + "learning_rate": 4.7966880812645754e-06, + "loss": 2.6209, + "step": 7347 + }, + { + "epoch": 0.7770727580372251, + "grad_norm": 8.375, + "learning_rate": 4.796620427323004e-06, + "loss": 2.7126, + "step": 7348 + }, + { + "epoch": 0.7771785109983079, + "grad_norm": 7.84375, + "learning_rate": 4.796552762604332e-06, + "loss": 2.6249, + "step": 7349 + }, + { + "epoch": 0.7772842639593909, + "grad_norm": 8.0625, + "learning_rate": 4.796485087108877e-06, + "loss": 2.8597, + "step": 7350 + }, + { + "epoch": 0.7773900169204738, + "grad_norm": 8.0625, + "learning_rate": 4.796417400836958e-06, + "loss": 2.6862, + "step": 7351 + }, + { + "epoch": 0.7774957698815567, + "grad_norm": 7.96875, + "learning_rate": 4.796349703788891e-06, + "loss": 2.6188, + "step": 7352 + }, + { + "epoch": 0.7776015228426396, + "grad_norm": 8.0625, + "learning_rate": 4.7962819959649935e-06, + "loss": 2.2628, + "step": 7353 + }, + { + "epoch": 0.7777072758037225, + "grad_norm": 8.3125, + "learning_rate": 4.796214277365583e-06, + "loss": 2.7309, + "step": 7354 + }, + { + "epoch": 0.7778130287648054, + "grad_norm": 8.375, + "learning_rate": 4.7961465479909795e-06, + "loss": 2.5858, + "step": 7355 + }, + { + "epoch": 0.7779187817258884, + "grad_norm": 8.5625, + "learning_rate": 4.796078807841498e-06, + "loss": 2.9599, + "step": 7356 + }, + { + "epoch": 0.7780245346869712, + "grad_norm": 7.625, + "learning_rate": 4.7960110569174584e-06, + "loss": 2.309, + "step": 7357 + }, + { + "epoch": 0.7781302876480541, + "grad_norm": 7.71875, + "learning_rate": 4.795943295219178e-06, + "loss": 2.7419, + "step": 7358 + }, + { + "epoch": 0.7782360406091371, + "grad_norm": 8.5, + "learning_rate": 4.795875522746975e-06, + "loss": 2.6898, + "step": 7359 + }, + { + "epoch": 0.7783417935702199, + "grad_norm": 8.375, + "learning_rate": 4.795807739501167e-06, + "loss": 2.3511, + "step": 7360 + }, + { + "epoch": 0.7784475465313029, + "grad_norm": 8.75, + "learning_rate": 4.795739945482071e-06, + "loss": 2.6836, + "step": 7361 + }, + { + "epoch": 0.7785532994923858, + "grad_norm": 8.0625, + "learning_rate": 4.795672140690007e-06, + "loss": 2.7657, + "step": 7362 + }, + { + "epoch": 0.7786590524534687, + "grad_norm": 7.59375, + "learning_rate": 4.795604325125294e-06, + "loss": 2.5832, + "step": 7363 + }, + { + "epoch": 0.7787648054145516, + "grad_norm": 7.75, + "learning_rate": 4.795536498788247e-06, + "loss": 2.7345, + "step": 7364 + }, + { + "epoch": 0.7788705583756346, + "grad_norm": 8.0625, + "learning_rate": 4.795468661679187e-06, + "loss": 2.7017, + "step": 7365 + }, + { + "epoch": 0.7789763113367174, + "grad_norm": 7.6875, + "learning_rate": 4.795400813798431e-06, + "loss": 2.4225, + "step": 7366 + }, + { + "epoch": 0.7790820642978004, + "grad_norm": 8.4375, + "learning_rate": 4.795332955146298e-06, + "loss": 2.6112, + "step": 7367 + }, + { + "epoch": 0.7791878172588832, + "grad_norm": 8.4375, + "learning_rate": 4.795265085723106e-06, + "loss": 2.8495, + "step": 7368 + }, + { + "epoch": 0.7792935702199661, + "grad_norm": 7.9375, + "learning_rate": 4.795197205529174e-06, + "loss": 2.7609, + "step": 7369 + }, + { + "epoch": 0.7793993231810491, + "grad_norm": 7.75, + "learning_rate": 4.7951293145648205e-06, + "loss": 2.4934, + "step": 7370 + }, + { + "epoch": 0.7795050761421319, + "grad_norm": 7.9375, + "learning_rate": 4.795061412830363e-06, + "loss": 2.6394, + "step": 7371 + }, + { + "epoch": 0.7796108291032149, + "grad_norm": 8.375, + "learning_rate": 4.794993500326121e-06, + "loss": 2.632, + "step": 7372 + }, + { + "epoch": 0.7797165820642978, + "grad_norm": 7.59375, + "learning_rate": 4.794925577052414e-06, + "loss": 2.7846, + "step": 7373 + }, + { + "epoch": 0.7798223350253807, + "grad_norm": 7.90625, + "learning_rate": 4.794857643009559e-06, + "loss": 2.6125, + "step": 7374 + }, + { + "epoch": 0.7799280879864636, + "grad_norm": 7.75, + "learning_rate": 4.794789698197876e-06, + "loss": 2.6129, + "step": 7375 + }, + { + "epoch": 0.7800338409475466, + "grad_norm": 7.4375, + "learning_rate": 4.794721742617683e-06, + "loss": 2.6035, + "step": 7376 + }, + { + "epoch": 0.7801395939086294, + "grad_norm": 8.125, + "learning_rate": 4.7946537762693e-06, + "loss": 2.5838, + "step": 7377 + }, + { + "epoch": 0.7802453468697124, + "grad_norm": 7.71875, + "learning_rate": 4.794585799153045e-06, + "loss": 2.6267, + "step": 7378 + }, + { + "epoch": 0.7803510998307953, + "grad_norm": 8.375, + "learning_rate": 4.794517811269237e-06, + "loss": 2.7308, + "step": 7379 + }, + { + "epoch": 0.7804568527918782, + "grad_norm": 7.78125, + "learning_rate": 4.794449812618196e-06, + "loss": 2.3802, + "step": 7380 + }, + { + "epoch": 0.7805626057529611, + "grad_norm": 9.375, + "learning_rate": 4.79438180320024e-06, + "loss": 2.7593, + "step": 7381 + }, + { + "epoch": 0.7806683587140439, + "grad_norm": 7.875, + "learning_rate": 4.7943137830156884e-06, + "loss": 2.6285, + "step": 7382 + }, + { + "epoch": 0.7807741116751269, + "grad_norm": 8.4375, + "learning_rate": 4.794245752064861e-06, + "loss": 2.4074, + "step": 7383 + }, + { + "epoch": 0.7808798646362098, + "grad_norm": 7.84375, + "learning_rate": 4.794177710348076e-06, + "loss": 2.5203, + "step": 7384 + }, + { + "epoch": 0.7809856175972927, + "grad_norm": 8.1875, + "learning_rate": 4.794109657865654e-06, + "loss": 2.5079, + "step": 7385 + }, + { + "epoch": 0.7810913705583756, + "grad_norm": 7.75, + "learning_rate": 4.794041594617914e-06, + "loss": 2.4677, + "step": 7386 + }, + { + "epoch": 0.7811971235194586, + "grad_norm": 7.625, + "learning_rate": 4.793973520605174e-06, + "loss": 2.5566, + "step": 7387 + }, + { + "epoch": 0.7813028764805414, + "grad_norm": 7.65625, + "learning_rate": 4.793905435827755e-06, + "loss": 2.6518, + "step": 7388 + }, + { + "epoch": 0.7814086294416244, + "grad_norm": 8.3125, + "learning_rate": 4.793837340285975e-06, + "loss": 2.3747, + "step": 7389 + }, + { + "epoch": 0.7815143824027073, + "grad_norm": 8.5, + "learning_rate": 4.7937692339801554e-06, + "loss": 2.5727, + "step": 7390 + }, + { + "epoch": 0.7816201353637902, + "grad_norm": 8.5, + "learning_rate": 4.793701116910615e-06, + "loss": 2.7443, + "step": 7391 + }, + { + "epoch": 0.7817258883248731, + "grad_norm": 7.71875, + "learning_rate": 4.793632989077673e-06, + "loss": 2.6025, + "step": 7392 + }, + { + "epoch": 0.7818316412859561, + "grad_norm": 7.84375, + "learning_rate": 4.79356485048165e-06, + "loss": 2.6554, + "step": 7393 + }, + { + "epoch": 0.7819373942470389, + "grad_norm": 8.0625, + "learning_rate": 4.793496701122865e-06, + "loss": 2.6474, + "step": 7394 + }, + { + "epoch": 0.7820431472081218, + "grad_norm": 7.625, + "learning_rate": 4.793428541001638e-06, + "loss": 2.5222, + "step": 7395 + }, + { + "epoch": 0.7821489001692047, + "grad_norm": 7.625, + "learning_rate": 4.793360370118289e-06, + "loss": 2.6966, + "step": 7396 + }, + { + "epoch": 0.7822546531302876, + "grad_norm": 8.25, + "learning_rate": 4.793292188473137e-06, + "loss": 2.6635, + "step": 7397 + }, + { + "epoch": 0.7823604060913706, + "grad_norm": 8.0, + "learning_rate": 4.793223996066503e-06, + "loss": 2.4629, + "step": 7398 + }, + { + "epoch": 0.7824661590524534, + "grad_norm": 8.1875, + "learning_rate": 4.793155792898707e-06, + "loss": 2.7778, + "step": 7399 + }, + { + "epoch": 0.7825719120135364, + "grad_norm": 7.625, + "learning_rate": 4.793087578970068e-06, + "loss": 2.5231, + "step": 7400 + }, + { + "epoch": 0.7826776649746193, + "grad_norm": 8.0625, + "learning_rate": 4.793019354280908e-06, + "loss": 2.5194, + "step": 7401 + }, + { + "epoch": 0.7827834179357022, + "grad_norm": 7.6875, + "learning_rate": 4.7929511188315446e-06, + "loss": 2.5441, + "step": 7402 + }, + { + "epoch": 0.7828891708967851, + "grad_norm": 8.4375, + "learning_rate": 4.792882872622301e-06, + "loss": 2.7179, + "step": 7403 + }, + { + "epoch": 0.7829949238578681, + "grad_norm": 7.71875, + "learning_rate": 4.792814615653494e-06, + "loss": 2.7031, + "step": 7404 + }, + { + "epoch": 0.7831006768189509, + "grad_norm": 7.84375, + "learning_rate": 4.7927463479254466e-06, + "loss": 2.634, + "step": 7405 + }, + { + "epoch": 0.7832064297800339, + "grad_norm": 7.90625, + "learning_rate": 4.792678069438479e-06, + "loss": 2.4464, + "step": 7406 + }, + { + "epoch": 0.7833121827411168, + "grad_norm": 7.84375, + "learning_rate": 4.792609780192909e-06, + "loss": 2.6617, + "step": 7407 + }, + { + "epoch": 0.7834179357021996, + "grad_norm": 8.6875, + "learning_rate": 4.792541480189061e-06, + "loss": 2.6976, + "step": 7408 + }, + { + "epoch": 0.7835236886632826, + "grad_norm": 7.90625, + "learning_rate": 4.792473169427253e-06, + "loss": 2.6287, + "step": 7409 + }, + { + "epoch": 0.7836294416243654, + "grad_norm": 7.90625, + "learning_rate": 4.792404847907805e-06, + "loss": 2.6708, + "step": 7410 + }, + { + "epoch": 0.7837351945854484, + "grad_norm": 7.875, + "learning_rate": 4.79233651563104e-06, + "loss": 2.477, + "step": 7411 + }, + { + "epoch": 0.7838409475465313, + "grad_norm": 7.28125, + "learning_rate": 4.792268172597276e-06, + "loss": 2.479, + "step": 7412 + }, + { + "epoch": 0.7839467005076142, + "grad_norm": 7.6875, + "learning_rate": 4.792199818806835e-06, + "loss": 2.4866, + "step": 7413 + }, + { + "epoch": 0.7840524534686971, + "grad_norm": 7.5, + "learning_rate": 4.7921314542600385e-06, + "loss": 2.3863, + "step": 7414 + }, + { + "epoch": 0.7841582064297801, + "grad_norm": 8.0, + "learning_rate": 4.792063078957207e-06, + "loss": 2.2691, + "step": 7415 + }, + { + "epoch": 0.7842639593908629, + "grad_norm": 8.4375, + "learning_rate": 4.7919946928986594e-06, + "loss": 2.672, + "step": 7416 + }, + { + "epoch": 0.7843697123519459, + "grad_norm": 7.875, + "learning_rate": 4.791926296084719e-06, + "loss": 2.5197, + "step": 7417 + }, + { + "epoch": 0.7844754653130288, + "grad_norm": 7.375, + "learning_rate": 4.791857888515705e-06, + "loss": 2.7246, + "step": 7418 + }, + { + "epoch": 0.7845812182741116, + "grad_norm": 7.75, + "learning_rate": 4.791789470191941e-06, + "loss": 2.7036, + "step": 7419 + }, + { + "epoch": 0.7846869712351946, + "grad_norm": 7.65625, + "learning_rate": 4.791721041113745e-06, + "loss": 2.5524, + "step": 7420 + }, + { + "epoch": 0.7847927241962775, + "grad_norm": 7.96875, + "learning_rate": 4.79165260128144e-06, + "loss": 2.5343, + "step": 7421 + }, + { + "epoch": 0.7848984771573604, + "grad_norm": 7.59375, + "learning_rate": 4.791584150695347e-06, + "loss": 2.5724, + "step": 7422 + }, + { + "epoch": 0.7850042301184433, + "grad_norm": 8.5625, + "learning_rate": 4.7915156893557854e-06, + "loss": 2.7792, + "step": 7423 + }, + { + "epoch": 0.7851099830795262, + "grad_norm": 8.5, + "learning_rate": 4.791447217263079e-06, + "loss": 2.6392, + "step": 7424 + }, + { + "epoch": 0.7852157360406091, + "grad_norm": 7.65625, + "learning_rate": 4.791378734417547e-06, + "loss": 2.6074, + "step": 7425 + }, + { + "epoch": 0.7853214890016921, + "grad_norm": 7.78125, + "learning_rate": 4.791310240819512e-06, + "loss": 2.6023, + "step": 7426 + }, + { + "epoch": 0.7854272419627749, + "grad_norm": 7.40625, + "learning_rate": 4.791241736469296e-06, + "loss": 2.5326, + "step": 7427 + }, + { + "epoch": 0.7855329949238579, + "grad_norm": 8.125, + "learning_rate": 4.791173221367219e-06, + "loss": 2.4465, + "step": 7428 + }, + { + "epoch": 0.7856387478849408, + "grad_norm": 7.625, + "learning_rate": 4.791104695513603e-06, + "loss": 2.6972, + "step": 7429 + }, + { + "epoch": 0.7857445008460237, + "grad_norm": 8.25, + "learning_rate": 4.79103615890877e-06, + "loss": 2.7497, + "step": 7430 + }, + { + "epoch": 0.7858502538071066, + "grad_norm": 8.0, + "learning_rate": 4.790967611553041e-06, + "loss": 2.4311, + "step": 7431 + }, + { + "epoch": 0.7859560067681896, + "grad_norm": 10.3125, + "learning_rate": 4.790899053446739e-06, + "loss": 2.6675, + "step": 7432 + }, + { + "epoch": 0.7860617597292724, + "grad_norm": 8.6875, + "learning_rate": 4.790830484590183e-06, + "loss": 2.5153, + "step": 7433 + }, + { + "epoch": 0.7861675126903553, + "grad_norm": 7.40625, + "learning_rate": 4.790761904983698e-06, + "loss": 2.4215, + "step": 7434 + }, + { + "epoch": 0.7862732656514383, + "grad_norm": 7.78125, + "learning_rate": 4.790693314627604e-06, + "loss": 2.6392, + "step": 7435 + }, + { + "epoch": 0.7863790186125211, + "grad_norm": 8.0, + "learning_rate": 4.790624713522223e-06, + "loss": 2.7365, + "step": 7436 + }, + { + "epoch": 0.7864847715736041, + "grad_norm": 8.3125, + "learning_rate": 4.7905561016678756e-06, + "loss": 2.4685, + "step": 7437 + }, + { + "epoch": 0.7865905245346869, + "grad_norm": 7.65625, + "learning_rate": 4.790487479064887e-06, + "loss": 2.5742, + "step": 7438 + }, + { + "epoch": 0.7866962774957699, + "grad_norm": 7.625, + "learning_rate": 4.790418845713577e-06, + "loss": 2.6341, + "step": 7439 + }, + { + "epoch": 0.7868020304568528, + "grad_norm": 7.65625, + "learning_rate": 4.7903502016142674e-06, + "loss": 2.5832, + "step": 7440 + }, + { + "epoch": 0.7869077834179357, + "grad_norm": 9.3125, + "learning_rate": 4.790281546767282e-06, + "loss": 2.5307, + "step": 7441 + }, + { + "epoch": 0.7870135363790186, + "grad_norm": 7.5625, + "learning_rate": 4.790212881172941e-06, + "loss": 2.5384, + "step": 7442 + }, + { + "epoch": 0.7871192893401016, + "grad_norm": 10.75, + "learning_rate": 4.790144204831569e-06, + "loss": 2.783, + "step": 7443 + }, + { + "epoch": 0.7872250423011844, + "grad_norm": 8.375, + "learning_rate": 4.790075517743486e-06, + "loss": 2.6328, + "step": 7444 + }, + { + "epoch": 0.7873307952622673, + "grad_norm": 8.1875, + "learning_rate": 4.790006819909015e-06, + "loss": 2.6955, + "step": 7445 + }, + { + "epoch": 0.7874365482233503, + "grad_norm": 7.71875, + "learning_rate": 4.789938111328479e-06, + "loss": 2.5733, + "step": 7446 + }, + { + "epoch": 0.7875423011844331, + "grad_norm": 8.125, + "learning_rate": 4.7898693920022e-06, + "loss": 2.3941, + "step": 7447 + }, + { + "epoch": 0.7876480541455161, + "grad_norm": 8.9375, + "learning_rate": 4.7898006619305005e-06, + "loss": 2.588, + "step": 7448 + }, + { + "epoch": 0.787753807106599, + "grad_norm": 8.25, + "learning_rate": 4.7897319211137024e-06, + "loss": 2.6494, + "step": 7449 + }, + { + "epoch": 0.7878595600676819, + "grad_norm": 7.8125, + "learning_rate": 4.789663169552129e-06, + "loss": 2.5125, + "step": 7450 + }, + { + "epoch": 0.7879653130287648, + "grad_norm": 8.1875, + "learning_rate": 4.789594407246103e-06, + "loss": 2.4923, + "step": 7451 + }, + { + "epoch": 0.7880710659898477, + "grad_norm": 8.125, + "learning_rate": 4.789525634195948e-06, + "loss": 2.6204, + "step": 7452 + }, + { + "epoch": 0.7881768189509306, + "grad_norm": 7.5625, + "learning_rate": 4.789456850401984e-06, + "loss": 2.431, + "step": 7453 + }, + { + "epoch": 0.7882825719120136, + "grad_norm": 7.6875, + "learning_rate": 4.789388055864536e-06, + "loss": 2.6381, + "step": 7454 + }, + { + "epoch": 0.7883883248730964, + "grad_norm": 7.96875, + "learning_rate": 4.789319250583926e-06, + "loss": 2.1632, + "step": 7455 + }, + { + "epoch": 0.7884940778341794, + "grad_norm": 8.625, + "learning_rate": 4.789250434560477e-06, + "loss": 2.4425, + "step": 7456 + }, + { + "epoch": 0.7885998307952623, + "grad_norm": 8.375, + "learning_rate": 4.7891816077945115e-06, + "loss": 2.6149, + "step": 7457 + }, + { + "epoch": 0.7887055837563451, + "grad_norm": 8.125, + "learning_rate": 4.789112770286354e-06, + "loss": 2.5873, + "step": 7458 + }, + { + "epoch": 0.7888113367174281, + "grad_norm": 8.625, + "learning_rate": 4.789043922036326e-06, + "loss": 2.8171, + "step": 7459 + }, + { + "epoch": 0.788917089678511, + "grad_norm": 7.75, + "learning_rate": 4.7889750630447505e-06, + "loss": 2.4873, + "step": 7460 + }, + { + "epoch": 0.7890228426395939, + "grad_norm": 7.75, + "learning_rate": 4.788906193311952e-06, + "loss": 2.547, + "step": 7461 + }, + { + "epoch": 0.7891285956006768, + "grad_norm": 9.5625, + "learning_rate": 4.788837312838252e-06, + "loss": 2.8383, + "step": 7462 + }, + { + "epoch": 0.7892343485617598, + "grad_norm": 8.4375, + "learning_rate": 4.788768421623976e-06, + "loss": 2.6562, + "step": 7463 + }, + { + "epoch": 0.7893401015228426, + "grad_norm": 8.625, + "learning_rate": 4.788699519669445e-06, + "loss": 2.7999, + "step": 7464 + }, + { + "epoch": 0.7894458544839256, + "grad_norm": 8.125, + "learning_rate": 4.788630606974984e-06, + "loss": 2.5896, + "step": 7465 + }, + { + "epoch": 0.7895516074450084, + "grad_norm": 7.78125, + "learning_rate": 4.788561683540915e-06, + "loss": 2.5987, + "step": 7466 + }, + { + "epoch": 0.7896573604060914, + "grad_norm": 7.71875, + "learning_rate": 4.788492749367561e-06, + "loss": 2.6353, + "step": 7467 + }, + { + "epoch": 0.7897631133671743, + "grad_norm": 7.78125, + "learning_rate": 4.788423804455248e-06, + "loss": 2.4999, + "step": 7468 + }, + { + "epoch": 0.7898688663282571, + "grad_norm": 8.5625, + "learning_rate": 4.788354848804297e-06, + "loss": 2.4949, + "step": 7469 + }, + { + "epoch": 0.7899746192893401, + "grad_norm": 7.875, + "learning_rate": 4.788285882415033e-06, + "loss": 2.5239, + "step": 7470 + }, + { + "epoch": 0.790080372250423, + "grad_norm": 7.9375, + "learning_rate": 4.788216905287779e-06, + "loss": 2.656, + "step": 7471 + }, + { + "epoch": 0.7901861252115059, + "grad_norm": 8.0625, + "learning_rate": 4.788147917422859e-06, + "loss": 2.7491, + "step": 7472 + }, + { + "epoch": 0.7902918781725888, + "grad_norm": 8.3125, + "learning_rate": 4.788078918820597e-06, + "loss": 2.5764, + "step": 7473 + }, + { + "epoch": 0.7903976311336718, + "grad_norm": 7.75, + "learning_rate": 4.788009909481316e-06, + "loss": 2.6662, + "step": 7474 + }, + { + "epoch": 0.7905033840947546, + "grad_norm": 7.78125, + "learning_rate": 4.787940889405342e-06, + "loss": 2.5057, + "step": 7475 + }, + { + "epoch": 0.7906091370558376, + "grad_norm": 8.375, + "learning_rate": 4.787871858592995e-06, + "loss": 2.6545, + "step": 7476 + }, + { + "epoch": 0.7907148900169205, + "grad_norm": 8.0, + "learning_rate": 4.787802817044601e-06, + "loss": 2.5768, + "step": 7477 + }, + { + "epoch": 0.7908206429780034, + "grad_norm": 7.65625, + "learning_rate": 4.787733764760485e-06, + "loss": 2.4334, + "step": 7478 + }, + { + "epoch": 0.7909263959390863, + "grad_norm": 7.8125, + "learning_rate": 4.78766470174097e-06, + "loss": 2.6037, + "step": 7479 + }, + { + "epoch": 0.7910321489001692, + "grad_norm": 7.96875, + "learning_rate": 4.78759562798638e-06, + "loss": 2.573, + "step": 7480 + }, + { + "epoch": 0.7911379018612521, + "grad_norm": 8.125, + "learning_rate": 4.7875265434970395e-06, + "loss": 2.5994, + "step": 7481 + }, + { + "epoch": 0.791243654822335, + "grad_norm": 7.5, + "learning_rate": 4.787457448273272e-06, + "loss": 2.5409, + "step": 7482 + }, + { + "epoch": 0.7913494077834179, + "grad_norm": 8.0, + "learning_rate": 4.787388342315402e-06, + "loss": 2.7724, + "step": 7483 + }, + { + "epoch": 0.7914551607445008, + "grad_norm": 7.90625, + "learning_rate": 4.7873192256237555e-06, + "loss": 2.5678, + "step": 7484 + }, + { + "epoch": 0.7915609137055838, + "grad_norm": 7.90625, + "learning_rate": 4.7872500981986546e-06, + "loss": 2.5834, + "step": 7485 + }, + { + "epoch": 0.7916666666666666, + "grad_norm": 7.78125, + "learning_rate": 4.787180960040424e-06, + "loss": 2.5053, + "step": 7486 + }, + { + "epoch": 0.7917724196277496, + "grad_norm": 8.0625, + "learning_rate": 4.787111811149389e-06, + "loss": 2.5251, + "step": 7487 + }, + { + "epoch": 0.7918781725888325, + "grad_norm": 7.59375, + "learning_rate": 4.7870426515258735e-06, + "loss": 2.4775, + "step": 7488 + }, + { + "epoch": 0.7919839255499154, + "grad_norm": 7.4375, + "learning_rate": 4.786973481170202e-06, + "loss": 2.4098, + "step": 7489 + }, + { + "epoch": 0.7920896785109983, + "grad_norm": 8.4375, + "learning_rate": 4.7869043000827e-06, + "loss": 2.7418, + "step": 7490 + }, + { + "epoch": 0.7921954314720813, + "grad_norm": 7.625, + "learning_rate": 4.78683510826369e-06, + "loss": 2.4916, + "step": 7491 + }, + { + "epoch": 0.7923011844331641, + "grad_norm": 7.9375, + "learning_rate": 4.7867659057135e-06, + "loss": 2.2227, + "step": 7492 + }, + { + "epoch": 0.7924069373942471, + "grad_norm": 7.6875, + "learning_rate": 4.7866966924324515e-06, + "loss": 2.5422, + "step": 7493 + }, + { + "epoch": 0.7925126903553299, + "grad_norm": 7.625, + "learning_rate": 4.786627468420871e-06, + "loss": 2.7511, + "step": 7494 + }, + { + "epoch": 0.7926184433164128, + "grad_norm": 7.78125, + "learning_rate": 4.786558233679083e-06, + "loss": 2.7098, + "step": 7495 + }, + { + "epoch": 0.7927241962774958, + "grad_norm": 8.5625, + "learning_rate": 4.786488988207413e-06, + "loss": 2.7159, + "step": 7496 + }, + { + "epoch": 0.7928299492385786, + "grad_norm": 8.0625, + "learning_rate": 4.786419732006185e-06, + "loss": 2.6041, + "step": 7497 + }, + { + "epoch": 0.7929357021996616, + "grad_norm": 8.1875, + "learning_rate": 4.786350465075723e-06, + "loss": 2.6633, + "step": 7498 + }, + { + "epoch": 0.7930414551607445, + "grad_norm": 7.4375, + "learning_rate": 4.786281187416355e-06, + "loss": 2.7016, + "step": 7499 + }, + { + "epoch": 0.7931472081218274, + "grad_norm": 8.4375, + "learning_rate": 4.786211899028404e-06, + "loss": 2.8705, + "step": 7500 + }, + { + "epoch": 0.7932529610829103, + "grad_norm": 8.125, + "learning_rate": 4.786142599912195e-06, + "loss": 2.507, + "step": 7501 + }, + { + "epoch": 0.7933587140439933, + "grad_norm": 7.5, + "learning_rate": 4.786073290068055e-06, + "loss": 2.6105, + "step": 7502 + }, + { + "epoch": 0.7934644670050761, + "grad_norm": 8.8125, + "learning_rate": 4.786003969496307e-06, + "loss": 2.5813, + "step": 7503 + }, + { + "epoch": 0.7935702199661591, + "grad_norm": 7.875, + "learning_rate": 4.785934638197277e-06, + "loss": 2.5321, + "step": 7504 + }, + { + "epoch": 0.793675972927242, + "grad_norm": 8.3125, + "learning_rate": 4.785865296171292e-06, + "loss": 2.4217, + "step": 7505 + }, + { + "epoch": 0.7937817258883249, + "grad_norm": 8.75, + "learning_rate": 4.785795943418675e-06, + "loss": 2.4432, + "step": 7506 + }, + { + "epoch": 0.7938874788494078, + "grad_norm": 7.53125, + "learning_rate": 4.785726579939752e-06, + "loss": 2.6442, + "step": 7507 + }, + { + "epoch": 0.7939932318104906, + "grad_norm": 8.0625, + "learning_rate": 4.7856572057348505e-06, + "loss": 2.6421, + "step": 7508 + }, + { + "epoch": 0.7940989847715736, + "grad_norm": 8.125, + "learning_rate": 4.785587820804294e-06, + "loss": 2.3137, + "step": 7509 + }, + { + "epoch": 0.7942047377326565, + "grad_norm": 8.5, + "learning_rate": 4.785518425148408e-06, + "loss": 2.5171, + "step": 7510 + }, + { + "epoch": 0.7943104906937394, + "grad_norm": 8.0, + "learning_rate": 4.78544901876752e-06, + "loss": 2.5756, + "step": 7511 + }, + { + "epoch": 0.7944162436548223, + "grad_norm": 8.0, + "learning_rate": 4.785379601661953e-06, + "loss": 2.6801, + "step": 7512 + }, + { + "epoch": 0.7945219966159053, + "grad_norm": 7.71875, + "learning_rate": 4.7853101738320355e-06, + "loss": 2.5842, + "step": 7513 + }, + { + "epoch": 0.7946277495769881, + "grad_norm": 7.625, + "learning_rate": 4.785240735278091e-06, + "loss": 2.6862, + "step": 7514 + }, + { + "epoch": 0.7947335025380711, + "grad_norm": 7.65625, + "learning_rate": 4.785171286000447e-06, + "loss": 2.5579, + "step": 7515 + }, + { + "epoch": 0.794839255499154, + "grad_norm": 7.4375, + "learning_rate": 4.785101825999429e-06, + "loss": 2.6142, + "step": 7516 + }, + { + "epoch": 0.7949450084602369, + "grad_norm": 8.375, + "learning_rate": 4.785032355275362e-06, + "loss": 2.5684, + "step": 7517 + }, + { + "epoch": 0.7950507614213198, + "grad_norm": 7.75, + "learning_rate": 4.784962873828572e-06, + "loss": 2.5872, + "step": 7518 + }, + { + "epoch": 0.7951565143824028, + "grad_norm": 8.4375, + "learning_rate": 4.784893381659387e-06, + "loss": 2.8572, + "step": 7519 + }, + { + "epoch": 0.7952622673434856, + "grad_norm": 8.25, + "learning_rate": 4.784823878768132e-06, + "loss": 2.692, + "step": 7520 + }, + { + "epoch": 0.7953680203045685, + "grad_norm": 7.59375, + "learning_rate": 4.784754365155132e-06, + "loss": 2.6617, + "step": 7521 + }, + { + "epoch": 0.7954737732656514, + "grad_norm": 8.4375, + "learning_rate": 4.784684840820715e-06, + "loss": 2.6553, + "step": 7522 + }, + { + "epoch": 0.7955795262267343, + "grad_norm": 7.59375, + "learning_rate": 4.784615305765206e-06, + "loss": 2.656, + "step": 7523 + }, + { + "epoch": 0.7956852791878173, + "grad_norm": 7.65625, + "learning_rate": 4.784545759988932e-06, + "loss": 2.2741, + "step": 7524 + }, + { + "epoch": 0.7957910321489001, + "grad_norm": 7.90625, + "learning_rate": 4.784476203492219e-06, + "loss": 2.5864, + "step": 7525 + }, + { + "epoch": 0.7958967851099831, + "grad_norm": 7.53125, + "learning_rate": 4.7844066362753935e-06, + "loss": 2.5236, + "step": 7526 + }, + { + "epoch": 0.796002538071066, + "grad_norm": 7.90625, + "learning_rate": 4.784337058338782e-06, + "loss": 2.5886, + "step": 7527 + }, + { + "epoch": 0.7961082910321489, + "grad_norm": 8.375, + "learning_rate": 4.784267469682712e-06, + "loss": 2.5176, + "step": 7528 + }, + { + "epoch": 0.7962140439932318, + "grad_norm": 7.78125, + "learning_rate": 4.784197870307507e-06, + "loss": 2.4489, + "step": 7529 + }, + { + "epoch": 0.7963197969543148, + "grad_norm": 8.3125, + "learning_rate": 4.7841282602134964e-06, + "loss": 2.588, + "step": 7530 + }, + { + "epoch": 0.7964255499153976, + "grad_norm": 7.5625, + "learning_rate": 4.784058639401007e-06, + "loss": 2.4833, + "step": 7531 + }, + { + "epoch": 0.7965313028764806, + "grad_norm": 7.90625, + "learning_rate": 4.7839890078703635e-06, + "loss": 2.5594, + "step": 7532 + }, + { + "epoch": 0.7966370558375635, + "grad_norm": 7.46875, + "learning_rate": 4.783919365621894e-06, + "loss": 2.4252, + "step": 7533 + }, + { + "epoch": 0.7967428087986463, + "grad_norm": 8.125, + "learning_rate": 4.783849712655924e-06, + "loss": 2.6438, + "step": 7534 + }, + { + "epoch": 0.7968485617597293, + "grad_norm": 8.125, + "learning_rate": 4.783780048972783e-06, + "loss": 2.7249, + "step": 7535 + }, + { + "epoch": 0.7969543147208121, + "grad_norm": 7.875, + "learning_rate": 4.783710374572795e-06, + "loss": 2.5952, + "step": 7536 + }, + { + "epoch": 0.7970600676818951, + "grad_norm": 8.25, + "learning_rate": 4.783640689456289e-06, + "loss": 2.7126, + "step": 7537 + }, + { + "epoch": 0.797165820642978, + "grad_norm": 8.3125, + "learning_rate": 4.783570993623591e-06, + "loss": 2.6166, + "step": 7538 + }, + { + "epoch": 0.7972715736040609, + "grad_norm": 7.875, + "learning_rate": 4.783501287075029e-06, + "loss": 2.6128, + "step": 7539 + }, + { + "epoch": 0.7973773265651438, + "grad_norm": 7.84375, + "learning_rate": 4.783431569810928e-06, + "loss": 2.5577, + "step": 7540 + }, + { + "epoch": 0.7974830795262268, + "grad_norm": 8.0625, + "learning_rate": 4.7833618418316176e-06, + "loss": 2.4714, + "step": 7541 + }, + { + "epoch": 0.7975888324873096, + "grad_norm": 8.0, + "learning_rate": 4.783292103137424e-06, + "loss": 2.3724, + "step": 7542 + }, + { + "epoch": 0.7976945854483926, + "grad_norm": 7.96875, + "learning_rate": 4.783222353728673e-06, + "loss": 2.8025, + "step": 7543 + }, + { + "epoch": 0.7978003384094755, + "grad_norm": 8.4375, + "learning_rate": 4.783152593605694e-06, + "loss": 2.5752, + "step": 7544 + }, + { + "epoch": 0.7979060913705583, + "grad_norm": 7.84375, + "learning_rate": 4.783082822768814e-06, + "loss": 2.5417, + "step": 7545 + }, + { + "epoch": 0.7980118443316413, + "grad_norm": 8.0625, + "learning_rate": 4.783013041218361e-06, + "loss": 2.6551, + "step": 7546 + }, + { + "epoch": 0.7981175972927242, + "grad_norm": 7.9375, + "learning_rate": 4.78294324895466e-06, + "loss": 2.4611, + "step": 7547 + }, + { + "epoch": 0.7982233502538071, + "grad_norm": 7.9375, + "learning_rate": 4.78287344597804e-06, + "loss": 2.5271, + "step": 7548 + }, + { + "epoch": 0.79832910321489, + "grad_norm": 8.5625, + "learning_rate": 4.7828036322888296e-06, + "loss": 2.7704, + "step": 7549 + }, + { + "epoch": 0.7984348561759729, + "grad_norm": 7.59375, + "learning_rate": 4.7827338078873546e-06, + "loss": 2.6114, + "step": 7550 + }, + { + "epoch": 0.7985406091370558, + "grad_norm": 7.65625, + "learning_rate": 4.7826639727739435e-06, + "loss": 2.6423, + "step": 7551 + }, + { + "epoch": 0.7986463620981388, + "grad_norm": 7.84375, + "learning_rate": 4.782594126948924e-06, + "loss": 2.5854, + "step": 7552 + }, + { + "epoch": 0.7987521150592216, + "grad_norm": 7.6875, + "learning_rate": 4.782524270412624e-06, + "loss": 2.6352, + "step": 7553 + }, + { + "epoch": 0.7988578680203046, + "grad_norm": 8.1875, + "learning_rate": 4.7824544031653716e-06, + "loss": 2.4081, + "step": 7554 + }, + { + "epoch": 0.7989636209813875, + "grad_norm": 8.875, + "learning_rate": 4.782384525207493e-06, + "loss": 2.4557, + "step": 7555 + }, + { + "epoch": 0.7990693739424704, + "grad_norm": 8.25, + "learning_rate": 4.782314636539319e-06, + "loss": 2.5981, + "step": 7556 + }, + { + "epoch": 0.7991751269035533, + "grad_norm": 7.5, + "learning_rate": 4.782244737161175e-06, + "loss": 2.504, + "step": 7557 + }, + { + "epoch": 0.7992808798646363, + "grad_norm": 8.0, + "learning_rate": 4.7821748270733895e-06, + "loss": 2.6648, + "step": 7558 + }, + { + "epoch": 0.7993866328257191, + "grad_norm": 7.75, + "learning_rate": 4.7821049062762904e-06, + "loss": 2.6324, + "step": 7559 + }, + { + "epoch": 0.799492385786802, + "grad_norm": 8.375, + "learning_rate": 4.782034974770207e-06, + "loss": 2.5344, + "step": 7560 + }, + { + "epoch": 0.799598138747885, + "grad_norm": 8.4375, + "learning_rate": 4.781965032555467e-06, + "loss": 2.6506, + "step": 7561 + }, + { + "epoch": 0.7997038917089678, + "grad_norm": 7.375, + "learning_rate": 4.7818950796323986e-06, + "loss": 2.4207, + "step": 7562 + }, + { + "epoch": 0.7998096446700508, + "grad_norm": 7.40625, + "learning_rate": 4.781825116001329e-06, + "loss": 2.503, + "step": 7563 + }, + { + "epoch": 0.7999153976311336, + "grad_norm": 8.6875, + "learning_rate": 4.781755141662588e-06, + "loss": 2.5791, + "step": 7564 + }, + { + "epoch": 0.8000211505922166, + "grad_norm": 7.84375, + "learning_rate": 4.781685156616503e-06, + "loss": 2.5668, + "step": 7565 + }, + { + "epoch": 0.8001269035532995, + "grad_norm": 8.0, + "learning_rate": 4.781615160863403e-06, + "loss": 2.4561, + "step": 7566 + }, + { + "epoch": 0.8002326565143824, + "grad_norm": 7.75, + "learning_rate": 4.781545154403615e-06, + "loss": 2.4725, + "step": 7567 + }, + { + "epoch": 0.8003384094754653, + "grad_norm": 7.53125, + "learning_rate": 4.7814751372374706e-06, + "loss": 2.4441, + "step": 7568 + }, + { + "epoch": 0.8004441624365483, + "grad_norm": 8.5625, + "learning_rate": 4.781405109365296e-06, + "loss": 2.7568, + "step": 7569 + }, + { + "epoch": 0.8005499153976311, + "grad_norm": 7.96875, + "learning_rate": 4.7813350707874194e-06, + "loss": 2.6198, + "step": 7570 + }, + { + "epoch": 0.800655668358714, + "grad_norm": 8.625, + "learning_rate": 4.781265021504171e-06, + "loss": 2.4422, + "step": 7571 + }, + { + "epoch": 0.800761421319797, + "grad_norm": 8.25, + "learning_rate": 4.7811949615158785e-06, + "loss": 2.3904, + "step": 7572 + }, + { + "epoch": 0.8008671742808798, + "grad_norm": 7.75, + "learning_rate": 4.7811248908228715e-06, + "loss": 2.546, + "step": 7573 + }, + { + "epoch": 0.8009729272419628, + "grad_norm": 7.75, + "learning_rate": 4.781054809425478e-06, + "loss": 2.6737, + "step": 7574 + }, + { + "epoch": 0.8010786802030457, + "grad_norm": 7.8125, + "learning_rate": 4.780984717324027e-06, + "loss": 2.6765, + "step": 7575 + }, + { + "epoch": 0.8011844331641286, + "grad_norm": 8.25, + "learning_rate": 4.780914614518848e-06, + "loss": 2.5871, + "step": 7576 + }, + { + "epoch": 0.8012901861252115, + "grad_norm": 8.1875, + "learning_rate": 4.78084450101027e-06, + "loss": 2.589, + "step": 7577 + }, + { + "epoch": 0.8013959390862944, + "grad_norm": 7.78125, + "learning_rate": 4.780774376798621e-06, + "loss": 2.6377, + "step": 7578 + }, + { + "epoch": 0.8015016920473773, + "grad_norm": 7.46875, + "learning_rate": 4.780704241884231e-06, + "loss": 2.6842, + "step": 7579 + }, + { + "epoch": 0.8016074450084603, + "grad_norm": 7.9375, + "learning_rate": 4.780634096267428e-06, + "loss": 2.6792, + "step": 7580 + }, + { + "epoch": 0.8017131979695431, + "grad_norm": 8.625, + "learning_rate": 4.780563939948542e-06, + "loss": 2.5236, + "step": 7581 + }, + { + "epoch": 0.801818950930626, + "grad_norm": 7.75, + "learning_rate": 4.7804937729279035e-06, + "loss": 2.6981, + "step": 7582 + }, + { + "epoch": 0.801924703891709, + "grad_norm": 7.40625, + "learning_rate": 4.7804235952058385e-06, + "loss": 2.5114, + "step": 7583 + }, + { + "epoch": 0.8020304568527918, + "grad_norm": 8.0625, + "learning_rate": 4.780353406782679e-06, + "loss": 2.8012, + "step": 7584 + }, + { + "epoch": 0.8021362098138748, + "grad_norm": 8.0625, + "learning_rate": 4.780283207658755e-06, + "loss": 2.6226, + "step": 7585 + }, + { + "epoch": 0.8022419627749577, + "grad_norm": 7.75, + "learning_rate": 4.780212997834393e-06, + "loss": 2.5003, + "step": 7586 + }, + { + "epoch": 0.8023477157360406, + "grad_norm": 7.875, + "learning_rate": 4.780142777309924e-06, + "loss": 2.5559, + "step": 7587 + }, + { + "epoch": 0.8024534686971235, + "grad_norm": 8.125, + "learning_rate": 4.780072546085678e-06, + "loss": 2.5751, + "step": 7588 + }, + { + "epoch": 0.8025592216582065, + "grad_norm": 8.3125, + "learning_rate": 4.780002304161984e-06, + "loss": 2.4567, + "step": 7589 + }, + { + "epoch": 0.8026649746192893, + "grad_norm": 7.84375, + "learning_rate": 4.779932051539171e-06, + "loss": 2.5049, + "step": 7590 + }, + { + "epoch": 0.8027707275803723, + "grad_norm": 7.90625, + "learning_rate": 4.77986178821757e-06, + "loss": 2.3786, + "step": 7591 + }, + { + "epoch": 0.8028764805414551, + "grad_norm": 7.78125, + "learning_rate": 4.77979151419751e-06, + "loss": 2.545, + "step": 7592 + }, + { + "epoch": 0.8029822335025381, + "grad_norm": 7.5, + "learning_rate": 4.779721229479321e-06, + "loss": 2.5594, + "step": 7593 + }, + { + "epoch": 0.803087986463621, + "grad_norm": 7.375, + "learning_rate": 4.779650934063332e-06, + "loss": 2.5328, + "step": 7594 + }, + { + "epoch": 0.8031937394247038, + "grad_norm": 7.9375, + "learning_rate": 4.779580627949874e-06, + "loss": 2.6148, + "step": 7595 + }, + { + "epoch": 0.8032994923857868, + "grad_norm": 8.0625, + "learning_rate": 4.779510311139276e-06, + "loss": 2.7365, + "step": 7596 + }, + { + "epoch": 0.8034052453468697, + "grad_norm": 7.75, + "learning_rate": 4.77943998363187e-06, + "loss": 2.6154, + "step": 7597 + }, + { + "epoch": 0.8035109983079526, + "grad_norm": 7.8125, + "learning_rate": 4.779369645427982e-06, + "loss": 2.4614, + "step": 7598 + }, + { + "epoch": 0.8036167512690355, + "grad_norm": 7.9375, + "learning_rate": 4.7792992965279464e-06, + "loss": 2.404, + "step": 7599 + }, + { + "epoch": 0.8037225042301185, + "grad_norm": 8.625, + "learning_rate": 4.77922893693209e-06, + "loss": 2.4771, + "step": 7600 + }, + { + "epoch": 0.8038282571912013, + "grad_norm": 9.0625, + "learning_rate": 4.7791585666407454e-06, + "loss": 2.8288, + "step": 7601 + }, + { + "epoch": 0.8039340101522843, + "grad_norm": 8.1875, + "learning_rate": 4.779088185654242e-06, + "loss": 2.5014, + "step": 7602 + }, + { + "epoch": 0.8040397631133672, + "grad_norm": 8.1875, + "learning_rate": 4.779017793972909e-06, + "loss": 2.5654, + "step": 7603 + }, + { + "epoch": 0.8041455160744501, + "grad_norm": 8.0, + "learning_rate": 4.7789473915970775e-06, + "loss": 2.5687, + "step": 7604 + }, + { + "epoch": 0.804251269035533, + "grad_norm": 8.3125, + "learning_rate": 4.778876978527078e-06, + "loss": 2.7426, + "step": 7605 + }, + { + "epoch": 0.8043570219966159, + "grad_norm": 7.9375, + "learning_rate": 4.778806554763241e-06, + "loss": 2.4817, + "step": 7606 + }, + { + "epoch": 0.8044627749576988, + "grad_norm": 8.0625, + "learning_rate": 4.778736120305898e-06, + "loss": 2.5569, + "step": 7607 + }, + { + "epoch": 0.8045685279187818, + "grad_norm": 7.8125, + "learning_rate": 4.778665675155377e-06, + "loss": 2.5307, + "step": 7608 + }, + { + "epoch": 0.8046742808798646, + "grad_norm": 8.4375, + "learning_rate": 4.77859521931201e-06, + "loss": 2.9408, + "step": 7609 + }, + { + "epoch": 0.8047800338409475, + "grad_norm": 8.0625, + "learning_rate": 4.778524752776128e-06, + "loss": 2.7349, + "step": 7610 + }, + { + "epoch": 0.8048857868020305, + "grad_norm": 8.1875, + "learning_rate": 4.778454275548061e-06, + "loss": 2.497, + "step": 7611 + }, + { + "epoch": 0.8049915397631133, + "grad_norm": 8.125, + "learning_rate": 4.778383787628141e-06, + "loss": 2.5823, + "step": 7612 + }, + { + "epoch": 0.8050972927241963, + "grad_norm": 7.6875, + "learning_rate": 4.778313289016696e-06, + "loss": 2.5562, + "step": 7613 + }, + { + "epoch": 0.8052030456852792, + "grad_norm": 9.0, + "learning_rate": 4.778242779714059e-06, + "loss": 2.565, + "step": 7614 + }, + { + "epoch": 0.8053087986463621, + "grad_norm": 8.0, + "learning_rate": 4.7781722597205605e-06, + "loss": 2.5543, + "step": 7615 + }, + { + "epoch": 0.805414551607445, + "grad_norm": 7.84375, + "learning_rate": 4.7781017290365315e-06, + "loss": 2.6807, + "step": 7616 + }, + { + "epoch": 0.805520304568528, + "grad_norm": 8.25, + "learning_rate": 4.778031187662302e-06, + "loss": 2.6146, + "step": 7617 + }, + { + "epoch": 0.8056260575296108, + "grad_norm": 8.0, + "learning_rate": 4.7779606355982054e-06, + "loss": 2.7731, + "step": 7618 + }, + { + "epoch": 0.8057318104906938, + "grad_norm": 8.1875, + "learning_rate": 4.77789007284457e-06, + "loss": 2.4455, + "step": 7619 + }, + { + "epoch": 0.8058375634517766, + "grad_norm": 8.125, + "learning_rate": 4.777819499401729e-06, + "loss": 2.7489, + "step": 7620 + }, + { + "epoch": 0.8059433164128595, + "grad_norm": 8.375, + "learning_rate": 4.777748915270012e-06, + "loss": 2.492, + "step": 7621 + }, + { + "epoch": 0.8060490693739425, + "grad_norm": 8.625, + "learning_rate": 4.777678320449751e-06, + "loss": 2.8448, + "step": 7622 + }, + { + "epoch": 0.8061548223350253, + "grad_norm": 7.4375, + "learning_rate": 4.777607714941277e-06, + "loss": 2.668, + "step": 7623 + }, + { + "epoch": 0.8062605752961083, + "grad_norm": 8.1875, + "learning_rate": 4.777537098744922e-06, + "loss": 2.549, + "step": 7624 + }, + { + "epoch": 0.8063663282571912, + "grad_norm": 8.375, + "learning_rate": 4.777466471861016e-06, + "loss": 2.6006, + "step": 7625 + }, + { + "epoch": 0.8064720812182741, + "grad_norm": 7.65625, + "learning_rate": 4.777395834289892e-06, + "loss": 2.6401, + "step": 7626 + }, + { + "epoch": 0.806577834179357, + "grad_norm": 7.84375, + "learning_rate": 4.777325186031881e-06, + "loss": 2.5612, + "step": 7627 + }, + { + "epoch": 0.80668358714044, + "grad_norm": 7.625, + "learning_rate": 4.777254527087315e-06, + "loss": 2.5361, + "step": 7628 + }, + { + "epoch": 0.8067893401015228, + "grad_norm": 7.90625, + "learning_rate": 4.777183857456524e-06, + "loss": 2.6223, + "step": 7629 + }, + { + "epoch": 0.8068950930626058, + "grad_norm": 7.71875, + "learning_rate": 4.7771131771398394e-06, + "loss": 2.552, + "step": 7630 + }, + { + "epoch": 0.8070008460236887, + "grad_norm": 8.625, + "learning_rate": 4.777042486137595e-06, + "loss": 2.7546, + "step": 7631 + }, + { + "epoch": 0.8071065989847716, + "grad_norm": 7.78125, + "learning_rate": 4.776971784450122e-06, + "loss": 2.5486, + "step": 7632 + }, + { + "epoch": 0.8072123519458545, + "grad_norm": 8.0625, + "learning_rate": 4.776901072077751e-06, + "loss": 2.4352, + "step": 7633 + }, + { + "epoch": 0.8073181049069373, + "grad_norm": 8.3125, + "learning_rate": 4.776830349020815e-06, + "loss": 2.413, + "step": 7634 + }, + { + "epoch": 0.8074238578680203, + "grad_norm": 8.0625, + "learning_rate": 4.776759615279644e-06, + "loss": 2.5602, + "step": 7635 + }, + { + "epoch": 0.8075296108291032, + "grad_norm": 8.0625, + "learning_rate": 4.776688870854572e-06, + "loss": 2.4611, + "step": 7636 + }, + { + "epoch": 0.8076353637901861, + "grad_norm": 8.1875, + "learning_rate": 4.776618115745931e-06, + "loss": 2.5127, + "step": 7637 + }, + { + "epoch": 0.807741116751269, + "grad_norm": 8.4375, + "learning_rate": 4.776547349954052e-06, + "loss": 2.6477, + "step": 7638 + }, + { + "epoch": 0.807846869712352, + "grad_norm": 8.125, + "learning_rate": 4.776476573479268e-06, + "loss": 2.6345, + "step": 7639 + }, + { + "epoch": 0.8079526226734348, + "grad_norm": 7.9375, + "learning_rate": 4.776405786321908e-06, + "loss": 2.6098, + "step": 7640 + }, + { + "epoch": 0.8080583756345178, + "grad_norm": 8.125, + "learning_rate": 4.776334988482309e-06, + "loss": 2.7686, + "step": 7641 + }, + { + "epoch": 0.8081641285956007, + "grad_norm": 7.75, + "learning_rate": 4.7762641799608e-06, + "loss": 2.2991, + "step": 7642 + }, + { + "epoch": 0.8082698815566836, + "grad_norm": 8.6875, + "learning_rate": 4.776193360757715e-06, + "loss": 2.6274, + "step": 7643 + }, + { + "epoch": 0.8083756345177665, + "grad_norm": 8.1875, + "learning_rate": 4.776122530873384e-06, + "loss": 2.7137, + "step": 7644 + }, + { + "epoch": 0.8084813874788495, + "grad_norm": 8.125, + "learning_rate": 4.776051690308142e-06, + "loss": 2.7022, + "step": 7645 + }, + { + "epoch": 0.8085871404399323, + "grad_norm": 8.1875, + "learning_rate": 4.77598083906232e-06, + "loss": 2.6585, + "step": 7646 + }, + { + "epoch": 0.8086928934010152, + "grad_norm": 9.125, + "learning_rate": 4.775909977136251e-06, + "loss": 2.5039, + "step": 7647 + }, + { + "epoch": 0.8087986463620981, + "grad_norm": 8.3125, + "learning_rate": 4.775839104530267e-06, + "loss": 2.7521, + "step": 7648 + }, + { + "epoch": 0.808904399323181, + "grad_norm": 8.375, + "learning_rate": 4.775768221244701e-06, + "loss": 2.4486, + "step": 7649 + }, + { + "epoch": 0.809010152284264, + "grad_norm": 8.8125, + "learning_rate": 4.775697327279885e-06, + "loss": 2.8303, + "step": 7650 + }, + { + "epoch": 0.8091159052453468, + "grad_norm": 8.3125, + "learning_rate": 4.775626422636152e-06, + "loss": 2.7624, + "step": 7651 + }, + { + "epoch": 0.8092216582064298, + "grad_norm": 8.1875, + "learning_rate": 4.775555507313836e-06, + "loss": 2.6102, + "step": 7652 + }, + { + "epoch": 0.8093274111675127, + "grad_norm": 7.96875, + "learning_rate": 4.7754845813132685e-06, + "loss": 2.6871, + "step": 7653 + }, + { + "epoch": 0.8094331641285956, + "grad_norm": 7.9375, + "learning_rate": 4.775413644634782e-06, + "loss": 2.5688, + "step": 7654 + }, + { + "epoch": 0.8095389170896785, + "grad_norm": 7.5, + "learning_rate": 4.7753426972787094e-06, + "loss": 2.5643, + "step": 7655 + }, + { + "epoch": 0.8096446700507615, + "grad_norm": 8.0, + "learning_rate": 4.775271739245385e-06, + "loss": 2.8329, + "step": 7656 + }, + { + "epoch": 0.8097504230118443, + "grad_norm": 8.5, + "learning_rate": 4.775200770535141e-06, + "loss": 2.5825, + "step": 7657 + }, + { + "epoch": 0.8098561759729273, + "grad_norm": 8.1875, + "learning_rate": 4.77512979114831e-06, + "loss": 2.4614, + "step": 7658 + }, + { + "epoch": 0.8099619289340102, + "grad_norm": 7.90625, + "learning_rate": 4.775058801085225e-06, + "loss": 2.5462, + "step": 7659 + }, + { + "epoch": 0.810067681895093, + "grad_norm": 7.25, + "learning_rate": 4.77498780034622e-06, + "loss": 2.5057, + "step": 7660 + }, + { + "epoch": 0.810173434856176, + "grad_norm": 7.90625, + "learning_rate": 4.774916788931628e-06, + "loss": 2.6213, + "step": 7661 + }, + { + "epoch": 0.8102791878172588, + "grad_norm": 8.4375, + "learning_rate": 4.774845766841781e-06, + "loss": 2.6823, + "step": 7662 + }, + { + "epoch": 0.8103849407783418, + "grad_norm": 7.84375, + "learning_rate": 4.7747747340770136e-06, + "loss": 2.469, + "step": 7663 + }, + { + "epoch": 0.8104906937394247, + "grad_norm": 8.25, + "learning_rate": 4.774703690637659e-06, + "loss": 2.5106, + "step": 7664 + }, + { + "epoch": 0.8105964467005076, + "grad_norm": 8.0, + "learning_rate": 4.77463263652405e-06, + "loss": 2.5759, + "step": 7665 + }, + { + "epoch": 0.8107021996615905, + "grad_norm": 7.4375, + "learning_rate": 4.77456157173652e-06, + "loss": 2.5332, + "step": 7666 + }, + { + "epoch": 0.8108079526226735, + "grad_norm": 7.9375, + "learning_rate": 4.7744904962754035e-06, + "loss": 2.6851, + "step": 7667 + }, + { + "epoch": 0.8109137055837563, + "grad_norm": 7.78125, + "learning_rate": 4.774419410141034e-06, + "loss": 2.5438, + "step": 7668 + }, + { + "epoch": 0.8110194585448393, + "grad_norm": 7.84375, + "learning_rate": 4.774348313333743e-06, + "loss": 2.6834, + "step": 7669 + }, + { + "epoch": 0.8111252115059222, + "grad_norm": 7.65625, + "learning_rate": 4.7742772058538655e-06, + "loss": 2.5424, + "step": 7670 + }, + { + "epoch": 0.811230964467005, + "grad_norm": 8.0, + "learning_rate": 4.774206087701736e-06, + "loss": 2.6028, + "step": 7671 + }, + { + "epoch": 0.811336717428088, + "grad_norm": 8.6875, + "learning_rate": 4.774134958877688e-06, + "loss": 2.8527, + "step": 7672 + }, + { + "epoch": 0.811442470389171, + "grad_norm": 8.1875, + "learning_rate": 4.774063819382053e-06, + "loss": 2.281, + "step": 7673 + }, + { + "epoch": 0.8115482233502538, + "grad_norm": 9.6875, + "learning_rate": 4.7739926692151675e-06, + "loss": 2.6942, + "step": 7674 + }, + { + "epoch": 0.8116539763113367, + "grad_norm": 7.59375, + "learning_rate": 4.773921508377364e-06, + "loss": 2.5625, + "step": 7675 + }, + { + "epoch": 0.8117597292724196, + "grad_norm": 8.25, + "learning_rate": 4.7738503368689774e-06, + "loss": 2.5584, + "step": 7676 + }, + { + "epoch": 0.8118654822335025, + "grad_norm": 7.625, + "learning_rate": 4.7737791546903415e-06, + "loss": 2.646, + "step": 7677 + }, + { + "epoch": 0.8119712351945855, + "grad_norm": 8.5, + "learning_rate": 4.773707961841789e-06, + "loss": 2.4594, + "step": 7678 + }, + { + "epoch": 0.8120769881556683, + "grad_norm": 7.90625, + "learning_rate": 4.773636758323656e-06, + "loss": 2.5241, + "step": 7679 + }, + { + "epoch": 0.8121827411167513, + "grad_norm": 7.90625, + "learning_rate": 4.773565544136275e-06, + "loss": 2.5527, + "step": 7680 + }, + { + "epoch": 0.8122884940778342, + "grad_norm": 7.75, + "learning_rate": 4.77349431927998e-06, + "loss": 2.5992, + "step": 7681 + }, + { + "epoch": 0.812394247038917, + "grad_norm": 7.78125, + "learning_rate": 4.773423083755107e-06, + "loss": 2.4663, + "step": 7682 + }, + { + "epoch": 0.8125, + "grad_norm": 7.9375, + "learning_rate": 4.773351837561989e-06, + "loss": 2.4688, + "step": 7683 + }, + { + "epoch": 0.812605752961083, + "grad_norm": 7.4375, + "learning_rate": 4.77328058070096e-06, + "loss": 2.4943, + "step": 7684 + }, + { + "epoch": 0.8127115059221658, + "grad_norm": 8.1875, + "learning_rate": 4.7732093131723565e-06, + "loss": 2.6195, + "step": 7685 + }, + { + "epoch": 0.8128172588832487, + "grad_norm": 7.625, + "learning_rate": 4.77313803497651e-06, + "loss": 2.604, + "step": 7686 + }, + { + "epoch": 0.8129230118443317, + "grad_norm": 8.625, + "learning_rate": 4.773066746113757e-06, + "loss": 2.5957, + "step": 7687 + }, + { + "epoch": 0.8130287648054145, + "grad_norm": 7.65625, + "learning_rate": 4.772995446584431e-06, + "loss": 2.5138, + "step": 7688 + }, + { + "epoch": 0.8131345177664975, + "grad_norm": 8.4375, + "learning_rate": 4.772924136388868e-06, + "loss": 2.8436, + "step": 7689 + }, + { + "epoch": 0.8132402707275804, + "grad_norm": 7.9375, + "learning_rate": 4.772852815527401e-06, + "loss": 2.7112, + "step": 7690 + }, + { + "epoch": 0.8133460236886633, + "grad_norm": 8.0, + "learning_rate": 4.772781484000365e-06, + "loss": 2.5012, + "step": 7691 + }, + { + "epoch": 0.8134517766497462, + "grad_norm": 8.0, + "learning_rate": 4.772710141808095e-06, + "loss": 2.7867, + "step": 7692 + }, + { + "epoch": 0.813557529610829, + "grad_norm": 8.25, + "learning_rate": 4.772638788950926e-06, + "loss": 2.6901, + "step": 7693 + }, + { + "epoch": 0.813663282571912, + "grad_norm": 7.875, + "learning_rate": 4.772567425429193e-06, + "loss": 2.5696, + "step": 7694 + }, + { + "epoch": 0.813769035532995, + "grad_norm": 7.625, + "learning_rate": 4.77249605124323e-06, + "loss": 2.427, + "step": 7695 + }, + { + "epoch": 0.8138747884940778, + "grad_norm": 8.3125, + "learning_rate": 4.772424666393373e-06, + "loss": 2.5488, + "step": 7696 + }, + { + "epoch": 0.8139805414551607, + "grad_norm": 8.8125, + "learning_rate": 4.772353270879956e-06, + "loss": 2.5128, + "step": 7697 + }, + { + "epoch": 0.8140862944162437, + "grad_norm": 8.25, + "learning_rate": 4.772281864703315e-06, + "loss": 2.7196, + "step": 7698 + }, + { + "epoch": 0.8141920473773265, + "grad_norm": 8.25, + "learning_rate": 4.772210447863783e-06, + "loss": 2.4593, + "step": 7699 + }, + { + "epoch": 0.8142978003384095, + "grad_norm": 8.3125, + "learning_rate": 4.7721390203616976e-06, + "loss": 2.6786, + "step": 7700 + }, + { + "epoch": 0.8144035532994924, + "grad_norm": 9.0, + "learning_rate": 4.772067582197394e-06, + "loss": 2.7292, + "step": 7701 + }, + { + "epoch": 0.8145093062605753, + "grad_norm": 8.0625, + "learning_rate": 4.771996133371205e-06, + "loss": 2.4957, + "step": 7702 + }, + { + "epoch": 0.8146150592216582, + "grad_norm": 8.1875, + "learning_rate": 4.771924673883468e-06, + "loss": 2.6538, + "step": 7703 + }, + { + "epoch": 0.8147208121827412, + "grad_norm": 7.65625, + "learning_rate": 4.771853203734518e-06, + "loss": 2.5254, + "step": 7704 + }, + { + "epoch": 0.814826565143824, + "grad_norm": 8.5, + "learning_rate": 4.77178172292469e-06, + "loss": 2.792, + "step": 7705 + }, + { + "epoch": 0.814932318104907, + "grad_norm": 7.4375, + "learning_rate": 4.77171023145432e-06, + "loss": 2.5448, + "step": 7706 + }, + { + "epoch": 0.8150380710659898, + "grad_norm": 8.0625, + "learning_rate": 4.7716387293237415e-06, + "loss": 2.5706, + "step": 7707 + }, + { + "epoch": 0.8151438240270727, + "grad_norm": 7.78125, + "learning_rate": 4.771567216533293e-06, + "loss": 2.5314, + "step": 7708 + }, + { + "epoch": 0.8152495769881557, + "grad_norm": 8.3125, + "learning_rate": 4.771495693083307e-06, + "loss": 2.5147, + "step": 7709 + }, + { + "epoch": 0.8153553299492385, + "grad_norm": 8.375, + "learning_rate": 4.771424158974122e-06, + "loss": 2.6356, + "step": 7710 + }, + { + "epoch": 0.8154610829103215, + "grad_norm": 8.25, + "learning_rate": 4.771352614206072e-06, + "loss": 2.5851, + "step": 7711 + }, + { + "epoch": 0.8155668358714044, + "grad_norm": 8.25, + "learning_rate": 4.771281058779493e-06, + "loss": 2.6077, + "step": 7712 + }, + { + "epoch": 0.8156725888324873, + "grad_norm": 7.75, + "learning_rate": 4.7712094926947215e-06, + "loss": 2.6372, + "step": 7713 + }, + { + "epoch": 0.8157783417935702, + "grad_norm": 9.125, + "learning_rate": 4.7711379159520925e-06, + "loss": 2.5197, + "step": 7714 + }, + { + "epoch": 0.8158840947546532, + "grad_norm": 7.5625, + "learning_rate": 4.771066328551942e-06, + "loss": 2.5636, + "step": 7715 + }, + { + "epoch": 0.815989847715736, + "grad_norm": 8.1875, + "learning_rate": 4.770994730494606e-06, + "loss": 2.686, + "step": 7716 + }, + { + "epoch": 0.816095600676819, + "grad_norm": 8.3125, + "learning_rate": 4.7709231217804196e-06, + "loss": 2.6009, + "step": 7717 + }, + { + "epoch": 0.8162013536379019, + "grad_norm": 8.1875, + "learning_rate": 4.770851502409722e-06, + "loss": 2.6236, + "step": 7718 + }, + { + "epoch": 0.8163071065989848, + "grad_norm": 7.875, + "learning_rate": 4.7707798723828445e-06, + "loss": 2.7781, + "step": 7719 + }, + { + "epoch": 0.8164128595600677, + "grad_norm": 7.59375, + "learning_rate": 4.770708231700128e-06, + "loss": 2.5254, + "step": 7720 + }, + { + "epoch": 0.8165186125211505, + "grad_norm": 7.5625, + "learning_rate": 4.770636580361905e-06, + "loss": 2.6686, + "step": 7721 + }, + { + "epoch": 0.8166243654822335, + "grad_norm": 7.78125, + "learning_rate": 4.770564918368513e-06, + "loss": 2.2939, + "step": 7722 + }, + { + "epoch": 0.8167301184433164, + "grad_norm": 8.0, + "learning_rate": 4.77049324572029e-06, + "loss": 2.3029, + "step": 7723 + }, + { + "epoch": 0.8168358714043993, + "grad_norm": 8.125, + "learning_rate": 4.77042156241757e-06, + "loss": 2.4849, + "step": 7724 + }, + { + "epoch": 0.8169416243654822, + "grad_norm": 8.875, + "learning_rate": 4.770349868460689e-06, + "loss": 2.4252, + "step": 7725 + }, + { + "epoch": 0.8170473773265652, + "grad_norm": 7.59375, + "learning_rate": 4.770278163849986e-06, + "loss": 2.2911, + "step": 7726 + }, + { + "epoch": 0.817153130287648, + "grad_norm": 8.125, + "learning_rate": 4.7702064485857955e-06, + "loss": 2.7297, + "step": 7727 + }, + { + "epoch": 0.817258883248731, + "grad_norm": 8.4375, + "learning_rate": 4.770134722668456e-06, + "loss": 2.7594, + "step": 7728 + }, + { + "epoch": 0.8173646362098139, + "grad_norm": 8.25, + "learning_rate": 4.770062986098301e-06, + "loss": 2.5531, + "step": 7729 + }, + { + "epoch": 0.8174703891708968, + "grad_norm": 7.4375, + "learning_rate": 4.76999123887567e-06, + "loss": 2.4405, + "step": 7730 + }, + { + "epoch": 0.8175761421319797, + "grad_norm": 8.6875, + "learning_rate": 4.7699194810008975e-06, + "loss": 2.3074, + "step": 7731 + }, + { + "epoch": 0.8176818950930627, + "grad_norm": 8.125, + "learning_rate": 4.769847712474321e-06, + "loss": 2.7665, + "step": 7732 + }, + { + "epoch": 0.8177876480541455, + "grad_norm": 7.6875, + "learning_rate": 4.769775933296279e-06, + "loss": 2.3107, + "step": 7733 + }, + { + "epoch": 0.8178934010152284, + "grad_norm": 8.5, + "learning_rate": 4.769704143467106e-06, + "loss": 2.7474, + "step": 7734 + }, + { + "epoch": 0.8179991539763113, + "grad_norm": 7.9375, + "learning_rate": 4.76963234298714e-06, + "loss": 2.7688, + "step": 7735 + }, + { + "epoch": 0.8181049069373942, + "grad_norm": 8.5, + "learning_rate": 4.769560531856717e-06, + "loss": 2.765, + "step": 7736 + }, + { + "epoch": 0.8182106598984772, + "grad_norm": 7.75, + "learning_rate": 4.769488710076176e-06, + "loss": 2.5762, + "step": 7737 + }, + { + "epoch": 0.81831641285956, + "grad_norm": 7.78125, + "learning_rate": 4.769416877645851e-06, + "loss": 2.634, + "step": 7738 + }, + { + "epoch": 0.818422165820643, + "grad_norm": 8.5, + "learning_rate": 4.769345034566082e-06, + "loss": 2.4704, + "step": 7739 + }, + { + "epoch": 0.8185279187817259, + "grad_norm": 8.25, + "learning_rate": 4.769273180837204e-06, + "loss": 2.6123, + "step": 7740 + }, + { + "epoch": 0.8186336717428088, + "grad_norm": 7.875, + "learning_rate": 4.769201316459555e-06, + "loss": 2.554, + "step": 7741 + }, + { + "epoch": 0.8187394247038917, + "grad_norm": 7.96875, + "learning_rate": 4.7691294414334735e-06, + "loss": 2.7527, + "step": 7742 + }, + { + "epoch": 0.8188451776649747, + "grad_norm": 7.625, + "learning_rate": 4.769057555759295e-06, + "loss": 2.3915, + "step": 7743 + }, + { + "epoch": 0.8189509306260575, + "grad_norm": 7.84375, + "learning_rate": 4.768985659437357e-06, + "loss": 2.6291, + "step": 7744 + }, + { + "epoch": 0.8190566835871405, + "grad_norm": 8.375, + "learning_rate": 4.768913752467998e-06, + "loss": 2.6704, + "step": 7745 + }, + { + "epoch": 0.8191624365482234, + "grad_norm": 7.9375, + "learning_rate": 4.7688418348515535e-06, + "loss": 2.5998, + "step": 7746 + }, + { + "epoch": 0.8192681895093062, + "grad_norm": 8.375, + "learning_rate": 4.768769906588363e-06, + "loss": 2.6423, + "step": 7747 + }, + { + "epoch": 0.8193739424703892, + "grad_norm": 9.3125, + "learning_rate": 4.768697967678764e-06, + "loss": 2.4669, + "step": 7748 + }, + { + "epoch": 0.819479695431472, + "grad_norm": 8.1875, + "learning_rate": 4.768626018123092e-06, + "loss": 2.7865, + "step": 7749 + }, + { + "epoch": 0.819585448392555, + "grad_norm": 7.875, + "learning_rate": 4.768554057921686e-06, + "loss": 2.5367, + "step": 7750 + }, + { + "epoch": 0.8196912013536379, + "grad_norm": 8.125, + "learning_rate": 4.768482087074885e-06, + "loss": 2.7586, + "step": 7751 + }, + { + "epoch": 0.8197969543147208, + "grad_norm": 7.46875, + "learning_rate": 4.7684101055830234e-06, + "loss": 2.5414, + "step": 7752 + }, + { + "epoch": 0.8199027072758037, + "grad_norm": 8.4375, + "learning_rate": 4.7683381134464425e-06, + "loss": 2.7308, + "step": 7753 + }, + { + "epoch": 0.8200084602368867, + "grad_norm": 7.6875, + "learning_rate": 4.768266110665478e-06, + "loss": 2.5413, + "step": 7754 + }, + { + "epoch": 0.8201142131979695, + "grad_norm": 8.5625, + "learning_rate": 4.768194097240468e-06, + "loss": 2.5099, + "step": 7755 + }, + { + "epoch": 0.8202199661590525, + "grad_norm": 7.875, + "learning_rate": 4.768122073171751e-06, + "loss": 2.7053, + "step": 7756 + }, + { + "epoch": 0.8203257191201354, + "grad_norm": 7.90625, + "learning_rate": 4.768050038459665e-06, + "loss": 2.6297, + "step": 7757 + }, + { + "epoch": 0.8204314720812182, + "grad_norm": 8.1875, + "learning_rate": 4.767977993104548e-06, + "loss": 2.6312, + "step": 7758 + }, + { + "epoch": 0.8205372250423012, + "grad_norm": 8.3125, + "learning_rate": 4.767905937106737e-06, + "loss": 2.641, + "step": 7759 + }, + { + "epoch": 0.8206429780033841, + "grad_norm": 7.59375, + "learning_rate": 4.767833870466571e-06, + "loss": 2.5974, + "step": 7760 + }, + { + "epoch": 0.820748730964467, + "grad_norm": 7.9375, + "learning_rate": 4.767761793184388e-06, + "loss": 2.6024, + "step": 7761 + }, + { + "epoch": 0.8208544839255499, + "grad_norm": 8.0625, + "learning_rate": 4.767689705260527e-06, + "loss": 2.174, + "step": 7762 + }, + { + "epoch": 0.8209602368866328, + "grad_norm": 7.65625, + "learning_rate": 4.767617606695326e-06, + "loss": 2.5905, + "step": 7763 + }, + { + "epoch": 0.8210659898477157, + "grad_norm": 8.0625, + "learning_rate": 4.767545497489122e-06, + "loss": 2.8593, + "step": 7764 + }, + { + "epoch": 0.8211717428087987, + "grad_norm": 8.625, + "learning_rate": 4.767473377642255e-06, + "loss": 2.7309, + "step": 7765 + }, + { + "epoch": 0.8212774957698815, + "grad_norm": 7.625, + "learning_rate": 4.767401247155063e-06, + "loss": 2.5712, + "step": 7766 + }, + { + "epoch": 0.8213832487309645, + "grad_norm": 7.5625, + "learning_rate": 4.767329106027884e-06, + "loss": 2.5589, + "step": 7767 + }, + { + "epoch": 0.8214890016920474, + "grad_norm": 8.0, + "learning_rate": 4.767256954261057e-06, + "loss": 2.6086, + "step": 7768 + }, + { + "epoch": 0.8215947546531303, + "grad_norm": 8.0, + "learning_rate": 4.76718479185492e-06, + "loss": 2.5151, + "step": 7769 + }, + { + "epoch": 0.8217005076142132, + "grad_norm": 7.9375, + "learning_rate": 4.767112618809813e-06, + "loss": 2.6136, + "step": 7770 + }, + { + "epoch": 0.8218062605752962, + "grad_norm": 11.0, + "learning_rate": 4.767040435126072e-06, + "loss": 2.5009, + "step": 7771 + }, + { + "epoch": 0.821912013536379, + "grad_norm": 7.53125, + "learning_rate": 4.766968240804039e-06, + "loss": 2.6465, + "step": 7772 + }, + { + "epoch": 0.8220177664974619, + "grad_norm": 7.46875, + "learning_rate": 4.766896035844051e-06, + "loss": 2.5651, + "step": 7773 + }, + { + "epoch": 0.8221235194585449, + "grad_norm": 8.6875, + "learning_rate": 4.766823820246446e-06, + "loss": 2.6977, + "step": 7774 + }, + { + "epoch": 0.8222292724196277, + "grad_norm": 8.0625, + "learning_rate": 4.766751594011565e-06, + "loss": 2.7894, + "step": 7775 + }, + { + "epoch": 0.8223350253807107, + "grad_norm": 7.875, + "learning_rate": 4.766679357139745e-06, + "loss": 2.5305, + "step": 7776 + }, + { + "epoch": 0.8224407783417935, + "grad_norm": 8.0625, + "learning_rate": 4.766607109631326e-06, + "loss": 2.668, + "step": 7777 + }, + { + "epoch": 0.8225465313028765, + "grad_norm": 8.25, + "learning_rate": 4.766534851486647e-06, + "loss": 2.5897, + "step": 7778 + }, + { + "epoch": 0.8226522842639594, + "grad_norm": 7.6875, + "learning_rate": 4.766462582706048e-06, + "loss": 2.5683, + "step": 7779 + }, + { + "epoch": 0.8227580372250423, + "grad_norm": 8.125, + "learning_rate": 4.766390303289865e-06, + "loss": 2.6559, + "step": 7780 + }, + { + "epoch": 0.8228637901861252, + "grad_norm": 7.15625, + "learning_rate": 4.76631801323844e-06, + "loss": 2.5509, + "step": 7781 + }, + { + "epoch": 0.8229695431472082, + "grad_norm": 7.84375, + "learning_rate": 4.766245712552112e-06, + "loss": 2.5389, + "step": 7782 + }, + { + "epoch": 0.823075296108291, + "grad_norm": 7.75, + "learning_rate": 4.766173401231219e-06, + "loss": 2.5922, + "step": 7783 + }, + { + "epoch": 0.823181049069374, + "grad_norm": 8.0625, + "learning_rate": 4.766101079276102e-06, + "loss": 2.6842, + "step": 7784 + }, + { + "epoch": 0.8232868020304569, + "grad_norm": 8.0625, + "learning_rate": 4.766028746687099e-06, + "loss": 2.167, + "step": 7785 + }, + { + "epoch": 0.8233925549915397, + "grad_norm": 8.5, + "learning_rate": 4.76595640346455e-06, + "loss": 2.6158, + "step": 7786 + }, + { + "epoch": 0.8234983079526227, + "grad_norm": 8.5, + "learning_rate": 4.7658840496087935e-06, + "loss": 2.7323, + "step": 7787 + }, + { + "epoch": 0.8236040609137056, + "grad_norm": 7.78125, + "learning_rate": 4.76581168512017e-06, + "loss": 2.4702, + "step": 7788 + }, + { + "epoch": 0.8237098138747885, + "grad_norm": 7.4375, + "learning_rate": 4.7657393099990204e-06, + "loss": 2.5383, + "step": 7789 + }, + { + "epoch": 0.8238155668358714, + "grad_norm": 7.59375, + "learning_rate": 4.765666924245682e-06, + "loss": 2.5747, + "step": 7790 + }, + { + "epoch": 0.8239213197969543, + "grad_norm": 8.5, + "learning_rate": 4.765594527860496e-06, + "loss": 2.5813, + "step": 7791 + }, + { + "epoch": 0.8240270727580372, + "grad_norm": 8.25, + "learning_rate": 4.7655221208438e-06, + "loss": 2.5426, + "step": 7792 + }, + { + "epoch": 0.8241328257191202, + "grad_norm": 7.6875, + "learning_rate": 4.765449703195935e-06, + "loss": 2.5966, + "step": 7793 + }, + { + "epoch": 0.824238578680203, + "grad_norm": 7.46875, + "learning_rate": 4.765377274917243e-06, + "loss": 2.6549, + "step": 7794 + }, + { + "epoch": 0.824344331641286, + "grad_norm": 7.53125, + "learning_rate": 4.765304836008061e-06, + "loss": 2.5914, + "step": 7795 + }, + { + "epoch": 0.8244500846023689, + "grad_norm": 8.25, + "learning_rate": 4.76523238646873e-06, + "loss": 2.4699, + "step": 7796 + }, + { + "epoch": 0.8245558375634517, + "grad_norm": 7.84375, + "learning_rate": 4.76515992629959e-06, + "loss": 2.6792, + "step": 7797 + }, + { + "epoch": 0.8246615905245347, + "grad_norm": 9.0625, + "learning_rate": 4.7650874555009804e-06, + "loss": 2.4919, + "step": 7798 + }, + { + "epoch": 0.8247673434856176, + "grad_norm": 8.0625, + "learning_rate": 4.765014974073242e-06, + "loss": 2.6434, + "step": 7799 + }, + { + "epoch": 0.8248730964467005, + "grad_norm": 7.8125, + "learning_rate": 4.764942482016714e-06, + "loss": 2.6181, + "step": 7800 + }, + { + "epoch": 0.8249788494077834, + "grad_norm": 7.53125, + "learning_rate": 4.764869979331739e-06, + "loss": 2.5756, + "step": 7801 + }, + { + "epoch": 0.8250846023688664, + "grad_norm": 7.75, + "learning_rate": 4.764797466018654e-06, + "loss": 2.542, + "step": 7802 + }, + { + "epoch": 0.8251903553299492, + "grad_norm": 7.59375, + "learning_rate": 4.764724942077801e-06, + "loss": 2.6091, + "step": 7803 + }, + { + "epoch": 0.8252961082910322, + "grad_norm": 7.59375, + "learning_rate": 4.76465240750952e-06, + "loss": 2.5295, + "step": 7804 + }, + { + "epoch": 0.825401861252115, + "grad_norm": 7.75, + "learning_rate": 4.764579862314153e-06, + "loss": 2.7646, + "step": 7805 + }, + { + "epoch": 0.825507614213198, + "grad_norm": 7.90625, + "learning_rate": 4.764507306492037e-06, + "loss": 2.5623, + "step": 7806 + }, + { + "epoch": 0.8256133671742809, + "grad_norm": 8.1875, + "learning_rate": 4.764434740043515e-06, + "loss": 2.4645, + "step": 7807 + }, + { + "epoch": 0.8257191201353637, + "grad_norm": 7.96875, + "learning_rate": 4.764362162968927e-06, + "loss": 2.599, + "step": 7808 + }, + { + "epoch": 0.8258248730964467, + "grad_norm": 7.46875, + "learning_rate": 4.764289575268614e-06, + "loss": 2.7192, + "step": 7809 + }, + { + "epoch": 0.8259306260575296, + "grad_norm": 8.0, + "learning_rate": 4.764216976942914e-06, + "loss": 2.4606, + "step": 7810 + }, + { + "epoch": 0.8260363790186125, + "grad_norm": 8.125, + "learning_rate": 4.7641443679921716e-06, + "loss": 2.7625, + "step": 7811 + }, + { + "epoch": 0.8261421319796954, + "grad_norm": 8.3125, + "learning_rate": 4.764071748416725e-06, + "loss": 2.4743, + "step": 7812 + }, + { + "epoch": 0.8262478849407784, + "grad_norm": 7.6875, + "learning_rate": 4.763999118216917e-06, + "loss": 2.4875, + "step": 7813 + }, + { + "epoch": 0.8263536379018612, + "grad_norm": 8.5, + "learning_rate": 4.763926477393086e-06, + "loss": 2.6599, + "step": 7814 + }, + { + "epoch": 0.8264593908629442, + "grad_norm": 8.8125, + "learning_rate": 4.763853825945573e-06, + "loss": 2.6172, + "step": 7815 + }, + { + "epoch": 0.8265651438240271, + "grad_norm": 7.59375, + "learning_rate": 4.763781163874722e-06, + "loss": 2.5316, + "step": 7816 + }, + { + "epoch": 0.82667089678511, + "grad_norm": 8.375, + "learning_rate": 4.7637084911808706e-06, + "loss": 2.7444, + "step": 7817 + }, + { + "epoch": 0.8267766497461929, + "grad_norm": 7.875, + "learning_rate": 4.763635807864361e-06, + "loss": 2.6808, + "step": 7818 + }, + { + "epoch": 0.8268824027072758, + "grad_norm": 7.71875, + "learning_rate": 4.763563113925534e-06, + "loss": 2.524, + "step": 7819 + }, + { + "epoch": 0.8269881556683587, + "grad_norm": 7.6875, + "learning_rate": 4.763490409364732e-06, + "loss": 2.4833, + "step": 7820 + }, + { + "epoch": 0.8270939086294417, + "grad_norm": 8.1875, + "learning_rate": 4.763417694182295e-06, + "loss": 2.4945, + "step": 7821 + }, + { + "epoch": 0.8271996615905245, + "grad_norm": 7.8125, + "learning_rate": 4.763344968378565e-06, + "loss": 2.4563, + "step": 7822 + }, + { + "epoch": 0.8273054145516074, + "grad_norm": 8.375, + "learning_rate": 4.7632722319538814e-06, + "loss": 2.557, + "step": 7823 + }, + { + "epoch": 0.8274111675126904, + "grad_norm": 8.1875, + "learning_rate": 4.7631994849085875e-06, + "loss": 2.4253, + "step": 7824 + }, + { + "epoch": 0.8275169204737732, + "grad_norm": 8.125, + "learning_rate": 4.763126727243023e-06, + "loss": 2.6293, + "step": 7825 + }, + { + "epoch": 0.8276226734348562, + "grad_norm": 9.125, + "learning_rate": 4.763053958957533e-06, + "loss": 2.7808, + "step": 7826 + }, + { + "epoch": 0.8277284263959391, + "grad_norm": 8.1875, + "learning_rate": 4.762981180052454e-06, + "loss": 2.3399, + "step": 7827 + }, + { + "epoch": 0.827834179357022, + "grad_norm": 7.78125, + "learning_rate": 4.762908390528131e-06, + "loss": 2.5233, + "step": 7828 + }, + { + "epoch": 0.8279399323181049, + "grad_norm": 8.25, + "learning_rate": 4.7628355903849046e-06, + "loss": 2.6777, + "step": 7829 + }, + { + "epoch": 0.8280456852791879, + "grad_norm": 7.6875, + "learning_rate": 4.762762779623115e-06, + "loss": 2.7013, + "step": 7830 + }, + { + "epoch": 0.8281514382402707, + "grad_norm": 7.90625, + "learning_rate": 4.762689958243106e-06, + "loss": 2.4988, + "step": 7831 + }, + { + "epoch": 0.8282571912013537, + "grad_norm": 8.0625, + "learning_rate": 4.762617126245218e-06, + "loss": 2.6235, + "step": 7832 + }, + { + "epoch": 0.8283629441624365, + "grad_norm": 7.6875, + "learning_rate": 4.762544283629794e-06, + "loss": 2.6512, + "step": 7833 + }, + { + "epoch": 0.8284686971235194, + "grad_norm": 7.8125, + "learning_rate": 4.762471430397174e-06, + "loss": 2.6315, + "step": 7834 + }, + { + "epoch": 0.8285744500846024, + "grad_norm": 8.0, + "learning_rate": 4.762398566547701e-06, + "loss": 2.5617, + "step": 7835 + }, + { + "epoch": 0.8286802030456852, + "grad_norm": 8.0, + "learning_rate": 4.762325692081717e-06, + "loss": 2.5659, + "step": 7836 + }, + { + "epoch": 0.8287859560067682, + "grad_norm": 8.3125, + "learning_rate": 4.762252806999564e-06, + "loss": 2.6781, + "step": 7837 + }, + { + "epoch": 0.8288917089678511, + "grad_norm": 7.875, + "learning_rate": 4.762179911301584e-06, + "loss": 2.8104, + "step": 7838 + }, + { + "epoch": 0.828997461928934, + "grad_norm": 9.5, + "learning_rate": 4.762107004988119e-06, + "loss": 2.8368, + "step": 7839 + }, + { + "epoch": 0.8291032148900169, + "grad_norm": 8.4375, + "learning_rate": 4.76203408805951e-06, + "loss": 2.5923, + "step": 7840 + }, + { + "epoch": 0.8292089678510999, + "grad_norm": 8.125, + "learning_rate": 4.761961160516101e-06, + "loss": 2.3169, + "step": 7841 + }, + { + "epoch": 0.8293147208121827, + "grad_norm": 8.375, + "learning_rate": 4.761888222358234e-06, + "loss": 2.6193, + "step": 7842 + }, + { + "epoch": 0.8294204737732657, + "grad_norm": 7.8125, + "learning_rate": 4.761815273586249e-06, + "loss": 2.5431, + "step": 7843 + }, + { + "epoch": 0.8295262267343486, + "grad_norm": 9.5, + "learning_rate": 4.761742314200491e-06, + "loss": 2.8187, + "step": 7844 + }, + { + "epoch": 0.8296319796954315, + "grad_norm": 8.125, + "learning_rate": 4.761669344201301e-06, + "loss": 2.4736, + "step": 7845 + }, + { + "epoch": 0.8297377326565144, + "grad_norm": 7.75, + "learning_rate": 4.761596363589022e-06, + "loss": 2.5578, + "step": 7846 + }, + { + "epoch": 0.8298434856175972, + "grad_norm": 8.375, + "learning_rate": 4.761523372363996e-06, + "loss": 2.7503, + "step": 7847 + }, + { + "epoch": 0.8299492385786802, + "grad_norm": 8.5, + "learning_rate": 4.761450370526566e-06, + "loss": 2.5402, + "step": 7848 + }, + { + "epoch": 0.8300549915397631, + "grad_norm": 7.5, + "learning_rate": 4.761377358077074e-06, + "loss": 2.6942, + "step": 7849 + }, + { + "epoch": 0.830160744500846, + "grad_norm": 7.875, + "learning_rate": 4.761304335015863e-06, + "loss": 2.4533, + "step": 7850 + }, + { + "epoch": 0.8302664974619289, + "grad_norm": 8.25, + "learning_rate": 4.761231301343276e-06, + "loss": 2.6416, + "step": 7851 + }, + { + "epoch": 0.8303722504230119, + "grad_norm": 7.90625, + "learning_rate": 4.761158257059655e-06, + "loss": 2.6348, + "step": 7852 + }, + { + "epoch": 0.8304780033840947, + "grad_norm": 8.0, + "learning_rate": 4.761085202165343e-06, + "loss": 2.376, + "step": 7853 + }, + { + "epoch": 0.8305837563451777, + "grad_norm": 9.0, + "learning_rate": 4.761012136660683e-06, + "loss": 2.9342, + "step": 7854 + }, + { + "epoch": 0.8306895093062606, + "grad_norm": 8.1875, + "learning_rate": 4.760939060546018e-06, + "loss": 2.8129, + "step": 7855 + }, + { + "epoch": 0.8307952622673435, + "grad_norm": 7.6875, + "learning_rate": 4.76086597382169e-06, + "loss": 2.5083, + "step": 7856 + }, + { + "epoch": 0.8309010152284264, + "grad_norm": 7.96875, + "learning_rate": 4.760792876488044e-06, + "loss": 2.7905, + "step": 7857 + }, + { + "epoch": 0.8310067681895094, + "grad_norm": 8.125, + "learning_rate": 4.76071976854542e-06, + "loss": 2.6893, + "step": 7858 + }, + { + "epoch": 0.8311125211505922, + "grad_norm": 8.375, + "learning_rate": 4.760646649994164e-06, + "loss": 2.9399, + "step": 7859 + }, + { + "epoch": 0.8312182741116751, + "grad_norm": 7.90625, + "learning_rate": 4.7605735208346174e-06, + "loss": 2.5964, + "step": 7860 + }, + { + "epoch": 0.831324027072758, + "grad_norm": 8.0625, + "learning_rate": 4.760500381067123e-06, + "loss": 2.5659, + "step": 7861 + }, + { + "epoch": 0.8314297800338409, + "grad_norm": 8.125, + "learning_rate": 4.760427230692026e-06, + "loss": 2.8471, + "step": 7862 + }, + { + "epoch": 0.8315355329949239, + "grad_norm": 8.0, + "learning_rate": 4.760354069709668e-06, + "loss": 2.6065, + "step": 7863 + }, + { + "epoch": 0.8316412859560067, + "grad_norm": 8.0, + "learning_rate": 4.760280898120393e-06, + "loss": 2.6174, + "step": 7864 + }, + { + "epoch": 0.8317470389170897, + "grad_norm": 8.0625, + "learning_rate": 4.760207715924544e-06, + "loss": 2.5358, + "step": 7865 + }, + { + "epoch": 0.8318527918781726, + "grad_norm": 8.375, + "learning_rate": 4.760134523122465e-06, + "loss": 2.8334, + "step": 7866 + }, + { + "epoch": 0.8319585448392555, + "grad_norm": 8.6875, + "learning_rate": 4.760061319714498e-06, + "loss": 2.7597, + "step": 7867 + }, + { + "epoch": 0.8320642978003384, + "grad_norm": 7.53125, + "learning_rate": 4.759988105700989e-06, + "loss": 2.5151, + "step": 7868 + }, + { + "epoch": 0.8321700507614214, + "grad_norm": 8.1875, + "learning_rate": 4.759914881082279e-06, + "loss": 2.8362, + "step": 7869 + }, + { + "epoch": 0.8322758037225042, + "grad_norm": 8.4375, + "learning_rate": 4.7598416458587134e-06, + "loss": 2.6366, + "step": 7870 + }, + { + "epoch": 0.8323815566835872, + "grad_norm": 7.90625, + "learning_rate": 4.759768400030634e-06, + "loss": 2.6941, + "step": 7871 + }, + { + "epoch": 0.8324873096446701, + "grad_norm": 7.9375, + "learning_rate": 4.7596951435983875e-06, + "loss": 2.6469, + "step": 7872 + }, + { + "epoch": 0.8325930626057529, + "grad_norm": 7.8125, + "learning_rate": 4.759621876562315e-06, + "loss": 2.7197, + "step": 7873 + }, + { + "epoch": 0.8326988155668359, + "grad_norm": 8.375, + "learning_rate": 4.7595485989227605e-06, + "loss": 2.6242, + "step": 7874 + }, + { + "epoch": 0.8328045685279187, + "grad_norm": 8.0, + "learning_rate": 4.7594753106800696e-06, + "loss": 2.5492, + "step": 7875 + }, + { + "epoch": 0.8329103214890017, + "grad_norm": 7.75, + "learning_rate": 4.759402011834585e-06, + "loss": 2.5937, + "step": 7876 + }, + { + "epoch": 0.8330160744500846, + "grad_norm": 7.5625, + "learning_rate": 4.75932870238665e-06, + "loss": 2.5527, + "step": 7877 + }, + { + "epoch": 0.8331218274111675, + "grad_norm": 10.0625, + "learning_rate": 4.7592553823366105e-06, + "loss": 2.6808, + "step": 7878 + }, + { + "epoch": 0.8332275803722504, + "grad_norm": 8.375, + "learning_rate": 4.759182051684809e-06, + "loss": 2.4568, + "step": 7879 + }, + { + "epoch": 0.8333333333333334, + "grad_norm": 8.0, + "learning_rate": 4.75910871043159e-06, + "loss": 2.3417, + "step": 7880 + }, + { + "epoch": 0.8334390862944162, + "grad_norm": 8.0, + "learning_rate": 4.759035358577298e-06, + "loss": 2.6087, + "step": 7881 + }, + { + "epoch": 0.8335448392554992, + "grad_norm": 8.0625, + "learning_rate": 4.758961996122276e-06, + "loss": 2.5266, + "step": 7882 + }, + { + "epoch": 0.8336505922165821, + "grad_norm": 7.90625, + "learning_rate": 4.758888623066871e-06, + "loss": 2.4903, + "step": 7883 + }, + { + "epoch": 0.833756345177665, + "grad_norm": 8.1875, + "learning_rate": 4.758815239411424e-06, + "loss": 2.5309, + "step": 7884 + }, + { + "epoch": 0.8338620981387479, + "grad_norm": 7.78125, + "learning_rate": 4.758741845156283e-06, + "loss": 2.5454, + "step": 7885 + }, + { + "epoch": 0.8339678510998308, + "grad_norm": 8.125, + "learning_rate": 4.758668440301788e-06, + "loss": 2.6683, + "step": 7886 + }, + { + "epoch": 0.8340736040609137, + "grad_norm": 7.46875, + "learning_rate": 4.758595024848287e-06, + "loss": 2.5539, + "step": 7887 + }, + { + "epoch": 0.8341793570219966, + "grad_norm": 8.0, + "learning_rate": 4.758521598796123e-06, + "loss": 2.6096, + "step": 7888 + }, + { + "epoch": 0.8342851099830795, + "grad_norm": 7.71875, + "learning_rate": 4.758448162145641e-06, + "loss": 2.5056, + "step": 7889 + }, + { + "epoch": 0.8343908629441624, + "grad_norm": 7.8125, + "learning_rate": 4.7583747148971845e-06, + "loss": 2.4991, + "step": 7890 + }, + { + "epoch": 0.8344966159052454, + "grad_norm": 8.25, + "learning_rate": 4.7583012570511e-06, + "loss": 2.5744, + "step": 7891 + }, + { + "epoch": 0.8346023688663282, + "grad_norm": 7.875, + "learning_rate": 4.758227788607731e-06, + "loss": 2.6931, + "step": 7892 + }, + { + "epoch": 0.8347081218274112, + "grad_norm": 8.3125, + "learning_rate": 4.758154309567423e-06, + "loss": 2.7224, + "step": 7893 + }, + { + "epoch": 0.8348138747884941, + "grad_norm": 7.59375, + "learning_rate": 4.758080819930519e-06, + "loss": 2.6081, + "step": 7894 + }, + { + "epoch": 0.834919627749577, + "grad_norm": 8.125, + "learning_rate": 4.758007319697367e-06, + "loss": 2.3341, + "step": 7895 + }, + { + "epoch": 0.8350253807106599, + "grad_norm": 8.25, + "learning_rate": 4.75793380886831e-06, + "loss": 2.5259, + "step": 7896 + }, + { + "epoch": 0.8351311336717429, + "grad_norm": 8.25, + "learning_rate": 4.757860287443692e-06, + "loss": 2.5476, + "step": 7897 + }, + { + "epoch": 0.8352368866328257, + "grad_norm": 7.59375, + "learning_rate": 4.75778675542386e-06, + "loss": 2.6526, + "step": 7898 + }, + { + "epoch": 0.8353426395939086, + "grad_norm": 7.90625, + "learning_rate": 4.757713212809157e-06, + "loss": 2.6806, + "step": 7899 + }, + { + "epoch": 0.8354483925549916, + "grad_norm": 9.125, + "learning_rate": 4.7576396595999305e-06, + "loss": 2.6448, + "step": 7900 + }, + { + "epoch": 0.8355541455160744, + "grad_norm": 8.0, + "learning_rate": 4.757566095796524e-06, + "loss": 2.6546, + "step": 7901 + }, + { + "epoch": 0.8356598984771574, + "grad_norm": 8.3125, + "learning_rate": 4.757492521399284e-06, + "loss": 2.6546, + "step": 7902 + }, + { + "epoch": 0.8357656514382402, + "grad_norm": 8.0, + "learning_rate": 4.757418936408553e-06, + "loss": 2.6233, + "step": 7903 + }, + { + "epoch": 0.8358714043993232, + "grad_norm": 7.46875, + "learning_rate": 4.75734534082468e-06, + "loss": 2.4489, + "step": 7904 + }, + { + "epoch": 0.8359771573604061, + "grad_norm": 7.71875, + "learning_rate": 4.757271734648008e-06, + "loss": 2.5504, + "step": 7905 + }, + { + "epoch": 0.836082910321489, + "grad_norm": 7.6875, + "learning_rate": 4.757198117878882e-06, + "loss": 2.7193, + "step": 7906 + }, + { + "epoch": 0.8361886632825719, + "grad_norm": 8.1875, + "learning_rate": 4.757124490517649e-06, + "loss": 2.6026, + "step": 7907 + }, + { + "epoch": 0.8362944162436549, + "grad_norm": 7.96875, + "learning_rate": 4.757050852564654e-06, + "loss": 2.6012, + "step": 7908 + }, + { + "epoch": 0.8364001692047377, + "grad_norm": 8.125, + "learning_rate": 4.756977204020242e-06, + "loss": 2.4761, + "step": 7909 + }, + { + "epoch": 0.8365059221658206, + "grad_norm": 7.8125, + "learning_rate": 4.75690354488476e-06, + "loss": 2.3319, + "step": 7910 + }, + { + "epoch": 0.8366116751269036, + "grad_norm": 8.5625, + "learning_rate": 4.756829875158553e-06, + "loss": 2.8327, + "step": 7911 + }, + { + "epoch": 0.8367174280879864, + "grad_norm": 8.3125, + "learning_rate": 4.756756194841965e-06, + "loss": 2.8769, + "step": 7912 + }, + { + "epoch": 0.8368231810490694, + "grad_norm": 7.8125, + "learning_rate": 4.756682503935344e-06, + "loss": 2.6919, + "step": 7913 + }, + { + "epoch": 0.8369289340101523, + "grad_norm": 7.40625, + "learning_rate": 4.756608802439035e-06, + "loss": 2.1351, + "step": 7914 + }, + { + "epoch": 0.8370346869712352, + "grad_norm": 7.90625, + "learning_rate": 4.756535090353382e-06, + "loss": 2.7988, + "step": 7915 + }, + { + "epoch": 0.8371404399323181, + "grad_norm": 7.65625, + "learning_rate": 4.756461367678735e-06, + "loss": 2.5338, + "step": 7916 + }, + { + "epoch": 0.837246192893401, + "grad_norm": 7.90625, + "learning_rate": 4.756387634415437e-06, + "loss": 2.7016, + "step": 7917 + }, + { + "epoch": 0.8373519458544839, + "grad_norm": 7.9375, + "learning_rate": 4.756313890563834e-06, + "loss": 2.5289, + "step": 7918 + }, + { + "epoch": 0.8374576988155669, + "grad_norm": 7.75, + "learning_rate": 4.7562401361242735e-06, + "loss": 2.6826, + "step": 7919 + }, + { + "epoch": 0.8375634517766497, + "grad_norm": 7.8125, + "learning_rate": 4.756166371097101e-06, + "loss": 2.5091, + "step": 7920 + }, + { + "epoch": 0.8376692047377327, + "grad_norm": 7.84375, + "learning_rate": 4.756092595482662e-06, + "loss": 2.5023, + "step": 7921 + }, + { + "epoch": 0.8377749576988156, + "grad_norm": 8.0, + "learning_rate": 4.756018809281303e-06, + "loss": 2.6968, + "step": 7922 + }, + { + "epoch": 0.8378807106598984, + "grad_norm": 9.375, + "learning_rate": 4.755945012493371e-06, + "loss": 2.5233, + "step": 7923 + }, + { + "epoch": 0.8379864636209814, + "grad_norm": 7.71875, + "learning_rate": 4.7558712051192115e-06, + "loss": 2.4673, + "step": 7924 + }, + { + "epoch": 0.8380922165820643, + "grad_norm": 8.1875, + "learning_rate": 4.75579738715917e-06, + "loss": 2.5601, + "step": 7925 + }, + { + "epoch": 0.8381979695431472, + "grad_norm": 7.84375, + "learning_rate": 4.755723558613595e-06, + "loss": 2.6458, + "step": 7926 + }, + { + "epoch": 0.8383037225042301, + "grad_norm": 8.25, + "learning_rate": 4.755649719482832e-06, + "loss": 2.8545, + "step": 7927 + }, + { + "epoch": 0.8384094754653131, + "grad_norm": 8.0625, + "learning_rate": 4.7555758697672275e-06, + "loss": 2.5745, + "step": 7928 + }, + { + "epoch": 0.8385152284263959, + "grad_norm": 7.28125, + "learning_rate": 4.755502009467127e-06, + "loss": 2.3016, + "step": 7929 + }, + { + "epoch": 0.8386209813874789, + "grad_norm": 8.25, + "learning_rate": 4.755428138582879e-06, + "loss": 2.6089, + "step": 7930 + }, + { + "epoch": 0.8387267343485617, + "grad_norm": 7.90625, + "learning_rate": 4.755354257114829e-06, + "loss": 2.5769, + "step": 7931 + }, + { + "epoch": 0.8388324873096447, + "grad_norm": 8.125, + "learning_rate": 4.755280365063323e-06, + "loss": 2.6196, + "step": 7932 + }, + { + "epoch": 0.8389382402707276, + "grad_norm": 7.9375, + "learning_rate": 4.755206462428709e-06, + "loss": 2.4491, + "step": 7933 + }, + { + "epoch": 0.8390439932318104, + "grad_norm": 8.375, + "learning_rate": 4.7551325492113345e-06, + "loss": 2.5012, + "step": 7934 + }, + { + "epoch": 0.8391497461928934, + "grad_norm": 7.34375, + "learning_rate": 4.755058625411544e-06, + "loss": 2.4635, + "step": 7935 + }, + { + "epoch": 0.8392554991539763, + "grad_norm": 7.96875, + "learning_rate": 4.754984691029687e-06, + "loss": 2.5482, + "step": 7936 + }, + { + "epoch": 0.8393612521150592, + "grad_norm": 8.25, + "learning_rate": 4.754910746066108e-06, + "loss": 2.7747, + "step": 7937 + }, + { + "epoch": 0.8394670050761421, + "grad_norm": 8.875, + "learning_rate": 4.754836790521155e-06, + "loss": 2.6568, + "step": 7938 + }, + { + "epoch": 0.8395727580372251, + "grad_norm": 7.5625, + "learning_rate": 4.754762824395175e-06, + "loss": 2.4293, + "step": 7939 + }, + { + "epoch": 0.8396785109983079, + "grad_norm": 8.125, + "learning_rate": 4.754688847688516e-06, + "loss": 2.6477, + "step": 7940 + }, + { + "epoch": 0.8397842639593909, + "grad_norm": 7.8125, + "learning_rate": 4.754614860401524e-06, + "loss": 2.5629, + "step": 7941 + }, + { + "epoch": 0.8398900169204738, + "grad_norm": 8.125, + "learning_rate": 4.7545408625345466e-06, + "loss": 2.6686, + "step": 7942 + }, + { + "epoch": 0.8399957698815567, + "grad_norm": 8.125, + "learning_rate": 4.7544668540879316e-06, + "loss": 2.6415, + "step": 7943 + }, + { + "epoch": 0.8401015228426396, + "grad_norm": 8.6875, + "learning_rate": 4.754392835062025e-06, + "loss": 2.823, + "step": 7944 + }, + { + "epoch": 0.8402072758037225, + "grad_norm": 7.8125, + "learning_rate": 4.754318805457175e-06, + "loss": 2.6307, + "step": 7945 + }, + { + "epoch": 0.8403130287648054, + "grad_norm": 7.96875, + "learning_rate": 4.754244765273729e-06, + "loss": 2.5663, + "step": 7946 + }, + { + "epoch": 0.8404187817258884, + "grad_norm": 8.0, + "learning_rate": 4.754170714512034e-06, + "loss": 2.6818, + "step": 7947 + }, + { + "epoch": 0.8405245346869712, + "grad_norm": 7.09375, + "learning_rate": 4.754096653172438e-06, + "loss": 2.5663, + "step": 7948 + }, + { + "epoch": 0.8406302876480541, + "grad_norm": 7.875, + "learning_rate": 4.7540225812552875e-06, + "loss": 2.6814, + "step": 7949 + }, + { + "epoch": 0.8407360406091371, + "grad_norm": 7.71875, + "learning_rate": 4.753948498760932e-06, + "loss": 2.6212, + "step": 7950 + }, + { + "epoch": 0.8408417935702199, + "grad_norm": 7.53125, + "learning_rate": 4.753874405689718e-06, + "loss": 2.6294, + "step": 7951 + }, + { + "epoch": 0.8409475465313029, + "grad_norm": 8.5, + "learning_rate": 4.753800302041992e-06, + "loss": 2.6288, + "step": 7952 + }, + { + "epoch": 0.8410532994923858, + "grad_norm": 7.75, + "learning_rate": 4.753726187818104e-06, + "loss": 2.411, + "step": 7953 + }, + { + "epoch": 0.8411590524534687, + "grad_norm": 7.8125, + "learning_rate": 4.7536520630184e-06, + "loss": 2.6122, + "step": 7954 + }, + { + "epoch": 0.8412648054145516, + "grad_norm": 7.21875, + "learning_rate": 4.75357792764323e-06, + "loss": 2.5241, + "step": 7955 + }, + { + "epoch": 0.8413705583756346, + "grad_norm": 7.875, + "learning_rate": 4.753503781692939e-06, + "loss": 2.5461, + "step": 7956 + }, + { + "epoch": 0.8414763113367174, + "grad_norm": 7.78125, + "learning_rate": 4.753429625167877e-06, + "loss": 2.7801, + "step": 7957 + }, + { + "epoch": 0.8415820642978004, + "grad_norm": 7.34375, + "learning_rate": 4.753355458068391e-06, + "loss": 2.4065, + "step": 7958 + }, + { + "epoch": 0.8416878172588832, + "grad_norm": 7.78125, + "learning_rate": 4.753281280394831e-06, + "loss": 2.4947, + "step": 7959 + }, + { + "epoch": 0.8417935702199661, + "grad_norm": 8.25, + "learning_rate": 4.753207092147542e-06, + "loss": 2.1895, + "step": 7960 + }, + { + "epoch": 0.8418993231810491, + "grad_norm": 7.8125, + "learning_rate": 4.7531328933268735e-06, + "loss": 2.6328, + "step": 7961 + }, + { + "epoch": 0.8420050761421319, + "grad_norm": 8.125, + "learning_rate": 4.753058683933174e-06, + "loss": 2.5421, + "step": 7962 + }, + { + "epoch": 0.8421108291032149, + "grad_norm": 8.125, + "learning_rate": 4.7529844639667925e-06, + "loss": 2.6633, + "step": 7963 + }, + { + "epoch": 0.8422165820642978, + "grad_norm": 7.46875, + "learning_rate": 4.752910233428075e-06, + "loss": 2.4043, + "step": 7964 + }, + { + "epoch": 0.8423223350253807, + "grad_norm": 8.75, + "learning_rate": 4.752835992317372e-06, + "loss": 2.7125, + "step": 7965 + }, + { + "epoch": 0.8424280879864636, + "grad_norm": 10.625, + "learning_rate": 4.752761740635031e-06, + "loss": 2.6256, + "step": 7966 + }, + { + "epoch": 0.8425338409475466, + "grad_norm": 7.28125, + "learning_rate": 4.7526874783814e-06, + "loss": 2.4587, + "step": 7967 + }, + { + "epoch": 0.8426395939086294, + "grad_norm": 8.4375, + "learning_rate": 4.752613205556829e-06, + "loss": 2.4625, + "step": 7968 + }, + { + "epoch": 0.8427453468697124, + "grad_norm": 8.5, + "learning_rate": 4.752538922161665e-06, + "loss": 2.8309, + "step": 7969 + }, + { + "epoch": 0.8428510998307953, + "grad_norm": 8.4375, + "learning_rate": 4.7524646281962565e-06, + "loss": 2.5346, + "step": 7970 + }, + { + "epoch": 0.8429568527918782, + "grad_norm": 7.59375, + "learning_rate": 4.7523903236609535e-06, + "loss": 2.6233, + "step": 7971 + }, + { + "epoch": 0.8430626057529611, + "grad_norm": 8.0, + "learning_rate": 4.752316008556104e-06, + "loss": 2.7503, + "step": 7972 + }, + { + "epoch": 0.8431683587140439, + "grad_norm": 8.1875, + "learning_rate": 4.752241682882056e-06, + "loss": 2.1874, + "step": 7973 + }, + { + "epoch": 0.8432741116751269, + "grad_norm": 8.875, + "learning_rate": 4.752167346639159e-06, + "loss": 2.502, + "step": 7974 + }, + { + "epoch": 0.8433798646362098, + "grad_norm": 7.6875, + "learning_rate": 4.7520929998277624e-06, + "loss": 2.6604, + "step": 7975 + }, + { + "epoch": 0.8434856175972927, + "grad_norm": 8.75, + "learning_rate": 4.7520186424482135e-06, + "loss": 2.4858, + "step": 7976 + }, + { + "epoch": 0.8435913705583756, + "grad_norm": 7.875, + "learning_rate": 4.751944274500863e-06, + "loss": 2.6342, + "step": 7977 + }, + { + "epoch": 0.8436971235194586, + "grad_norm": 8.1875, + "learning_rate": 4.751869895986059e-06, + "loss": 2.5432, + "step": 7978 + }, + { + "epoch": 0.8438028764805414, + "grad_norm": 7.6875, + "learning_rate": 4.75179550690415e-06, + "loss": 2.5261, + "step": 7979 + }, + { + "epoch": 0.8439086294416244, + "grad_norm": 8.0, + "learning_rate": 4.751721107255486e-06, + "loss": 2.5367, + "step": 7980 + }, + { + "epoch": 0.8440143824027073, + "grad_norm": 7.625, + "learning_rate": 4.751646697040415e-06, + "loss": 2.4601, + "step": 7981 + }, + { + "epoch": 0.8441201353637902, + "grad_norm": 7.875, + "learning_rate": 4.751572276259287e-06, + "loss": 2.5706, + "step": 7982 + }, + { + "epoch": 0.8442258883248731, + "grad_norm": 8.125, + "learning_rate": 4.751497844912453e-06, + "loss": 2.6059, + "step": 7983 + }, + { + "epoch": 0.8443316412859561, + "grad_norm": 7.90625, + "learning_rate": 4.751423403000259e-06, + "loss": 2.6274, + "step": 7984 + }, + { + "epoch": 0.8444373942470389, + "grad_norm": 7.75, + "learning_rate": 4.751348950523055e-06, + "loss": 2.5086, + "step": 7985 + }, + { + "epoch": 0.8445431472081218, + "grad_norm": 9.875, + "learning_rate": 4.751274487481192e-06, + "loss": 2.4613, + "step": 7986 + }, + { + "epoch": 0.8446489001692047, + "grad_norm": 7.90625, + "learning_rate": 4.751200013875019e-06, + "loss": 2.6893, + "step": 7987 + }, + { + "epoch": 0.8447546531302876, + "grad_norm": 8.5, + "learning_rate": 4.7511255297048855e-06, + "loss": 2.9742, + "step": 7988 + }, + { + "epoch": 0.8448604060913706, + "grad_norm": 7.78125, + "learning_rate": 4.7510510349711395e-06, + "loss": 2.6698, + "step": 7989 + }, + { + "epoch": 0.8449661590524534, + "grad_norm": 8.0625, + "learning_rate": 4.750976529674131e-06, + "loss": 2.6752, + "step": 7990 + }, + { + "epoch": 0.8450719120135364, + "grad_norm": 7.59375, + "learning_rate": 4.750902013814211e-06, + "loss": 2.67, + "step": 7991 + }, + { + "epoch": 0.8451776649746193, + "grad_norm": 7.78125, + "learning_rate": 4.75082748739173e-06, + "loss": 2.5419, + "step": 7992 + }, + { + "epoch": 0.8452834179357022, + "grad_norm": 7.84375, + "learning_rate": 4.750752950407034e-06, + "loss": 2.678, + "step": 7993 + }, + { + "epoch": 0.8453891708967851, + "grad_norm": 8.0625, + "learning_rate": 4.750678402860476e-06, + "loss": 2.3733, + "step": 7994 + }, + { + "epoch": 0.8454949238578681, + "grad_norm": 8.75, + "learning_rate": 4.750603844752405e-06, + "loss": 2.791, + "step": 7995 + }, + { + "epoch": 0.8456006768189509, + "grad_norm": 7.75, + "learning_rate": 4.75052927608317e-06, + "loss": 2.8131, + "step": 7996 + }, + { + "epoch": 0.8457064297800339, + "grad_norm": 7.8125, + "learning_rate": 4.750454696853121e-06, + "loss": 2.5849, + "step": 7997 + }, + { + "epoch": 0.8458121827411168, + "grad_norm": 8.0625, + "learning_rate": 4.7503801070626095e-06, + "loss": 2.5498, + "step": 7998 + }, + { + "epoch": 0.8459179357021996, + "grad_norm": 7.84375, + "learning_rate": 4.750305506711984e-06, + "loss": 2.5759, + "step": 7999 + }, + { + "epoch": 0.8460236886632826, + "grad_norm": 8.5, + "learning_rate": 4.750230895801595e-06, + "loss": 2.6602, + "step": 8000 + }, + { + "epoch": 0.8461294416243654, + "grad_norm": 9.6875, + "learning_rate": 4.750156274331793e-06, + "loss": 2.8856, + "step": 8001 + }, + { + "epoch": 0.8462351945854484, + "grad_norm": 8.1875, + "learning_rate": 4.750081642302928e-06, + "loss": 2.5478, + "step": 8002 + }, + { + "epoch": 0.8463409475465313, + "grad_norm": 8.125, + "learning_rate": 4.75000699971535e-06, + "loss": 2.5779, + "step": 8003 + }, + { + "epoch": 0.8464467005076142, + "grad_norm": 8.25, + "learning_rate": 4.749932346569409e-06, + "loss": 2.616, + "step": 8004 + }, + { + "epoch": 0.8465524534686971, + "grad_norm": 8.1875, + "learning_rate": 4.749857682865456e-06, + "loss": 2.4602, + "step": 8005 + }, + { + "epoch": 0.8466582064297801, + "grad_norm": 7.5, + "learning_rate": 4.749783008603841e-06, + "loss": 2.777, + "step": 8006 + }, + { + "epoch": 0.8467639593908629, + "grad_norm": 8.0, + "learning_rate": 4.749708323784915e-06, + "loss": 2.5463, + "step": 8007 + }, + { + "epoch": 0.8468697123519459, + "grad_norm": 8.375, + "learning_rate": 4.749633628409028e-06, + "loss": 2.9128, + "step": 8008 + }, + { + "epoch": 0.8469754653130288, + "grad_norm": 8.375, + "learning_rate": 4.749558922476529e-06, + "loss": 2.611, + "step": 8009 + }, + { + "epoch": 0.8470812182741116, + "grad_norm": 7.75, + "learning_rate": 4.749484205987772e-06, + "loss": 2.7861, + "step": 8010 + }, + { + "epoch": 0.8471869712351946, + "grad_norm": 8.1875, + "learning_rate": 4.749409478943105e-06, + "loss": 2.3762, + "step": 8011 + }, + { + "epoch": 0.8472927241962775, + "grad_norm": 7.71875, + "learning_rate": 4.749334741342878e-06, + "loss": 2.5847, + "step": 8012 + }, + { + "epoch": 0.8473984771573604, + "grad_norm": 7.78125, + "learning_rate": 4.749259993187444e-06, + "loss": 2.9091, + "step": 8013 + }, + { + "epoch": 0.8475042301184433, + "grad_norm": 7.8125, + "learning_rate": 4.749185234477153e-06, + "loss": 2.4986, + "step": 8014 + }, + { + "epoch": 0.8476099830795262, + "grad_norm": 8.5625, + "learning_rate": 4.749110465212355e-06, + "loss": 2.7355, + "step": 8015 + }, + { + "epoch": 0.8477157360406091, + "grad_norm": 8.25, + "learning_rate": 4.749035685393401e-06, + "loss": 2.5489, + "step": 8016 + }, + { + "epoch": 0.8478214890016921, + "grad_norm": 8.0, + "learning_rate": 4.748960895020644e-06, + "loss": 2.5382, + "step": 8017 + }, + { + "epoch": 0.8479272419627749, + "grad_norm": 7.8125, + "learning_rate": 4.748886094094432e-06, + "loss": 2.5105, + "step": 8018 + }, + { + "epoch": 0.8480329949238579, + "grad_norm": 7.40625, + "learning_rate": 4.748811282615118e-06, + "loss": 2.5139, + "step": 8019 + }, + { + "epoch": 0.8481387478849408, + "grad_norm": 7.6875, + "learning_rate": 4.748736460583051e-06, + "loss": 2.6476, + "step": 8020 + }, + { + "epoch": 0.8482445008460237, + "grad_norm": 7.96875, + "learning_rate": 4.7486616279985855e-06, + "loss": 2.5647, + "step": 8021 + }, + { + "epoch": 0.8483502538071066, + "grad_norm": 7.75, + "learning_rate": 4.7485867848620695e-06, + "loss": 2.6469, + "step": 8022 + }, + { + "epoch": 0.8484560067681896, + "grad_norm": 7.46875, + "learning_rate": 4.748511931173856e-06, + "loss": 2.5992, + "step": 8023 + }, + { + "epoch": 0.8485617597292724, + "grad_norm": 8.4375, + "learning_rate": 4.748437066934295e-06, + "loss": 2.744, + "step": 8024 + }, + { + "epoch": 0.8486675126903553, + "grad_norm": 7.8125, + "learning_rate": 4.748362192143738e-06, + "loss": 2.541, + "step": 8025 + }, + { + "epoch": 0.8487732656514383, + "grad_norm": 7.46875, + "learning_rate": 4.748287306802538e-06, + "loss": 2.2577, + "step": 8026 + }, + { + "epoch": 0.8488790186125211, + "grad_norm": 7.71875, + "learning_rate": 4.748212410911044e-06, + "loss": 2.5797, + "step": 8027 + }, + { + "epoch": 0.8489847715736041, + "grad_norm": 7.84375, + "learning_rate": 4.7481375044696095e-06, + "loss": 2.4951, + "step": 8028 + }, + { + "epoch": 0.8490905245346869, + "grad_norm": 7.71875, + "learning_rate": 4.748062587478585e-06, + "loss": 2.5871, + "step": 8029 + }, + { + "epoch": 0.8491962774957699, + "grad_norm": 7.8125, + "learning_rate": 4.747987659938321e-06, + "loss": 2.3715, + "step": 8030 + }, + { + "epoch": 0.8493020304568528, + "grad_norm": 7.5, + "learning_rate": 4.747912721849172e-06, + "loss": 2.5391, + "step": 8031 + }, + { + "epoch": 0.8494077834179357, + "grad_norm": 8.25, + "learning_rate": 4.747837773211487e-06, + "loss": 2.7718, + "step": 8032 + }, + { + "epoch": 0.8495135363790186, + "grad_norm": 10.1875, + "learning_rate": 4.747762814025619e-06, + "loss": 2.2586, + "step": 8033 + }, + { + "epoch": 0.8496192893401016, + "grad_norm": 7.9375, + "learning_rate": 4.7476878442919196e-06, + "loss": 2.6216, + "step": 8034 + }, + { + "epoch": 0.8497250423011844, + "grad_norm": 7.6875, + "learning_rate": 4.7476128640107405e-06, + "loss": 2.518, + "step": 8035 + }, + { + "epoch": 0.8498307952622673, + "grad_norm": 8.375, + "learning_rate": 4.747537873182433e-06, + "loss": 2.572, + "step": 8036 + }, + { + "epoch": 0.8499365482233503, + "grad_norm": 7.96875, + "learning_rate": 4.74746287180735e-06, + "loss": 2.4252, + "step": 8037 + }, + { + "epoch": 0.8500423011844331, + "grad_norm": 7.84375, + "learning_rate": 4.747387859885843e-06, + "loss": 2.6744, + "step": 8038 + }, + { + "epoch": 0.8501480541455161, + "grad_norm": 7.8125, + "learning_rate": 4.747312837418263e-06, + "loss": 2.5523, + "step": 8039 + }, + { + "epoch": 0.850253807106599, + "grad_norm": 8.5, + "learning_rate": 4.7472378044049635e-06, + "loss": 2.7322, + "step": 8040 + }, + { + "epoch": 0.8503595600676819, + "grad_norm": 8.8125, + "learning_rate": 4.7471627608462965e-06, + "loss": 2.2045, + "step": 8041 + }, + { + "epoch": 0.8504653130287648, + "grad_norm": 8.1875, + "learning_rate": 4.747087706742613e-06, + "loss": 2.7496, + "step": 8042 + }, + { + "epoch": 0.8505710659898477, + "grad_norm": 8.0, + "learning_rate": 4.747012642094267e-06, + "loss": 2.539, + "step": 8043 + }, + { + "epoch": 0.8506768189509306, + "grad_norm": 7.5625, + "learning_rate": 4.7469375669016085e-06, + "loss": 2.4243, + "step": 8044 + }, + { + "epoch": 0.8507825719120136, + "grad_norm": 7.375, + "learning_rate": 4.746862481164991e-06, + "loss": 2.3953, + "step": 8045 + }, + { + "epoch": 0.8508883248730964, + "grad_norm": 8.125, + "learning_rate": 4.7467873848847666e-06, + "loss": 2.5262, + "step": 8046 + }, + { + "epoch": 0.8509940778341794, + "grad_norm": 8.3125, + "learning_rate": 4.746712278061289e-06, + "loss": 2.5336, + "step": 8047 + }, + { + "epoch": 0.8510998307952623, + "grad_norm": 7.78125, + "learning_rate": 4.746637160694909e-06, + "loss": 2.6373, + "step": 8048 + }, + { + "epoch": 0.8512055837563451, + "grad_norm": 8.0, + "learning_rate": 4.74656203278598e-06, + "loss": 2.7167, + "step": 8049 + }, + { + "epoch": 0.8513113367174281, + "grad_norm": 8.125, + "learning_rate": 4.746486894334854e-06, + "loss": 2.773, + "step": 8050 + }, + { + "epoch": 0.851417089678511, + "grad_norm": 7.84375, + "learning_rate": 4.746411745341883e-06, + "loss": 2.4294, + "step": 8051 + }, + { + "epoch": 0.8515228426395939, + "grad_norm": 7.78125, + "learning_rate": 4.746336585807421e-06, + "loss": 2.6099, + "step": 8052 + }, + { + "epoch": 0.8516285956006768, + "grad_norm": 8.5625, + "learning_rate": 4.7462614157318205e-06, + "loss": 2.6738, + "step": 8053 + }, + { + "epoch": 0.8517343485617598, + "grad_norm": 8.75, + "learning_rate": 4.746186235115433e-06, + "loss": 2.7709, + "step": 8054 + }, + { + "epoch": 0.8518401015228426, + "grad_norm": 8.75, + "learning_rate": 4.746111043958613e-06, + "loss": 2.5462, + "step": 8055 + }, + { + "epoch": 0.8519458544839256, + "grad_norm": 8.8125, + "learning_rate": 4.746035842261712e-06, + "loss": 2.5041, + "step": 8056 + }, + { + "epoch": 0.8520516074450084, + "grad_norm": 8.5625, + "learning_rate": 4.745960630025084e-06, + "loss": 2.7284, + "step": 8057 + }, + { + "epoch": 0.8521573604060914, + "grad_norm": 8.3125, + "learning_rate": 4.745885407249081e-06, + "loss": 2.5866, + "step": 8058 + }, + { + "epoch": 0.8522631133671743, + "grad_norm": 8.0, + "learning_rate": 4.745810173934056e-06, + "loss": 2.5667, + "step": 8059 + }, + { + "epoch": 0.8523688663282571, + "grad_norm": 7.84375, + "learning_rate": 4.745734930080363e-06, + "loss": 2.4016, + "step": 8060 + }, + { + "epoch": 0.8524746192893401, + "grad_norm": 8.3125, + "learning_rate": 4.745659675688354e-06, + "loss": 2.6439, + "step": 8061 + }, + { + "epoch": 0.852580372250423, + "grad_norm": 7.78125, + "learning_rate": 4.745584410758382e-06, + "loss": 2.4536, + "step": 8062 + }, + { + "epoch": 0.8526861252115059, + "grad_norm": 8.5625, + "learning_rate": 4.745509135290802e-06, + "loss": 2.7483, + "step": 8063 + }, + { + "epoch": 0.8527918781725888, + "grad_norm": 7.90625, + "learning_rate": 4.745433849285965e-06, + "loss": 2.8322, + "step": 8064 + }, + { + "epoch": 0.8528976311336718, + "grad_norm": 7.84375, + "learning_rate": 4.745358552744226e-06, + "loss": 2.4931, + "step": 8065 + }, + { + "epoch": 0.8530033840947546, + "grad_norm": 7.65625, + "learning_rate": 4.745283245665937e-06, + "loss": 2.402, + "step": 8066 + }, + { + "epoch": 0.8531091370558376, + "grad_norm": 9.5625, + "learning_rate": 4.745207928051453e-06, + "loss": 2.6469, + "step": 8067 + }, + { + "epoch": 0.8532148900169205, + "grad_norm": 7.75, + "learning_rate": 4.745132599901125e-06, + "loss": 2.7313, + "step": 8068 + }, + { + "epoch": 0.8533206429780034, + "grad_norm": 7.6875, + "learning_rate": 4.7450572612153085e-06, + "loss": 2.4632, + "step": 8069 + }, + { + "epoch": 0.8534263959390863, + "grad_norm": 8.0625, + "learning_rate": 4.744981911994358e-06, + "loss": 2.4717, + "step": 8070 + }, + { + "epoch": 0.8535321489001692, + "grad_norm": 7.6875, + "learning_rate": 4.744906552238624e-06, + "loss": 2.4783, + "step": 8071 + }, + { + "epoch": 0.8536379018612521, + "grad_norm": 8.125, + "learning_rate": 4.744831181948462e-06, + "loss": 2.7676, + "step": 8072 + }, + { + "epoch": 0.853743654822335, + "grad_norm": 8.25, + "learning_rate": 4.744755801124225e-06, + "loss": 2.5203, + "step": 8073 + }, + { + "epoch": 0.8538494077834179, + "grad_norm": 8.0625, + "learning_rate": 4.7446804097662676e-06, + "loss": 2.595, + "step": 8074 + }, + { + "epoch": 0.8539551607445008, + "grad_norm": 8.0625, + "learning_rate": 4.744605007874943e-06, + "loss": 2.5533, + "step": 8075 + }, + { + "epoch": 0.8540609137055838, + "grad_norm": 7.875, + "learning_rate": 4.744529595450605e-06, + "loss": 2.4792, + "step": 8076 + }, + { + "epoch": 0.8541666666666666, + "grad_norm": 8.0, + "learning_rate": 4.744454172493608e-06, + "loss": 2.5673, + "step": 8077 + }, + { + "epoch": 0.8542724196277496, + "grad_norm": 8.1875, + "learning_rate": 4.744378739004305e-06, + "loss": 2.3324, + "step": 8078 + }, + { + "epoch": 0.8543781725888325, + "grad_norm": 7.8125, + "learning_rate": 4.74430329498305e-06, + "loss": 2.6105, + "step": 8079 + }, + { + "epoch": 0.8544839255499154, + "grad_norm": 8.125, + "learning_rate": 4.744227840430199e-06, + "loss": 2.5728, + "step": 8080 + }, + { + "epoch": 0.8545896785109983, + "grad_norm": 8.375, + "learning_rate": 4.7441523753461036e-06, + "loss": 2.6806, + "step": 8081 + }, + { + "epoch": 0.8546954314720813, + "grad_norm": 8.0, + "learning_rate": 4.744076899731118e-06, + "loss": 2.5568, + "step": 8082 + }, + { + "epoch": 0.8548011844331641, + "grad_norm": 7.84375, + "learning_rate": 4.744001413585599e-06, + "loss": 2.5738, + "step": 8083 + }, + { + "epoch": 0.8549069373942471, + "grad_norm": 8.1875, + "learning_rate": 4.743925916909899e-06, + "loss": 2.5834, + "step": 8084 + }, + { + "epoch": 0.8550126903553299, + "grad_norm": 7.71875, + "learning_rate": 4.743850409704371e-06, + "loss": 2.6342, + "step": 8085 + }, + { + "epoch": 0.8551184433164128, + "grad_norm": 7.78125, + "learning_rate": 4.743774891969372e-06, + "loss": 2.5329, + "step": 8086 + }, + { + "epoch": 0.8552241962774958, + "grad_norm": 7.8125, + "learning_rate": 4.743699363705255e-06, + "loss": 2.6732, + "step": 8087 + }, + { + "epoch": 0.8553299492385786, + "grad_norm": 8.6875, + "learning_rate": 4.7436238249123746e-06, + "loss": 2.6415, + "step": 8088 + }, + { + "epoch": 0.8554357021996616, + "grad_norm": 7.5625, + "learning_rate": 4.7435482755910845e-06, + "loss": 2.5459, + "step": 8089 + }, + { + "epoch": 0.8555414551607445, + "grad_norm": 8.25, + "learning_rate": 4.743472715741741e-06, + "loss": 2.6578, + "step": 8090 + }, + { + "epoch": 0.8556472081218274, + "grad_norm": 7.59375, + "learning_rate": 4.743397145364696e-06, + "loss": 2.4911, + "step": 8091 + }, + { + "epoch": 0.8557529610829103, + "grad_norm": 7.9375, + "learning_rate": 4.743321564460307e-06, + "loss": 2.6671, + "step": 8092 + }, + { + "epoch": 0.8558587140439933, + "grad_norm": 8.125, + "learning_rate": 4.743245973028927e-06, + "loss": 2.4868, + "step": 8093 + }, + { + "epoch": 0.8559644670050761, + "grad_norm": 8.4375, + "learning_rate": 4.743170371070911e-06, + "loss": 2.5962, + "step": 8094 + }, + { + "epoch": 0.8560702199661591, + "grad_norm": 8.1875, + "learning_rate": 4.7430947585866146e-06, + "loss": 2.1466, + "step": 8095 + }, + { + "epoch": 0.856175972927242, + "grad_norm": 8.1875, + "learning_rate": 4.743019135576391e-06, + "loss": 2.3321, + "step": 8096 + }, + { + "epoch": 0.8562817258883249, + "grad_norm": 7.875, + "learning_rate": 4.742943502040596e-06, + "loss": 2.43, + "step": 8097 + }, + { + "epoch": 0.8563874788494078, + "grad_norm": 7.375, + "learning_rate": 4.742867857979585e-06, + "loss": 2.503, + "step": 8098 + }, + { + "epoch": 0.8564932318104906, + "grad_norm": 7.59375, + "learning_rate": 4.742792203393712e-06, + "loss": 2.6286, + "step": 8099 + }, + { + "epoch": 0.8565989847715736, + "grad_norm": 8.125, + "learning_rate": 4.7427165382833325e-06, + "loss": 2.6881, + "step": 8100 + }, + { + "epoch": 0.8567047377326565, + "grad_norm": 8.1875, + "learning_rate": 4.742640862648802e-06, + "loss": 2.6529, + "step": 8101 + }, + { + "epoch": 0.8568104906937394, + "grad_norm": 8.25, + "learning_rate": 4.7425651764904744e-06, + "loss": 2.5335, + "step": 8102 + }, + { + "epoch": 0.8569162436548223, + "grad_norm": 8.0625, + "learning_rate": 4.742489479808706e-06, + "loss": 2.6849, + "step": 8103 + }, + { + "epoch": 0.8570219966159053, + "grad_norm": 8.25, + "learning_rate": 4.7424137726038524e-06, + "loss": 2.5912, + "step": 8104 + }, + { + "epoch": 0.8571277495769881, + "grad_norm": 8.9375, + "learning_rate": 4.742338054876266e-06, + "loss": 2.6778, + "step": 8105 + }, + { + "epoch": 0.8572335025380711, + "grad_norm": 7.84375, + "learning_rate": 4.7422623266263066e-06, + "loss": 2.5646, + "step": 8106 + }, + { + "epoch": 0.857339255499154, + "grad_norm": 8.1875, + "learning_rate": 4.742186587854326e-06, + "loss": 2.6547, + "step": 8107 + }, + { + "epoch": 0.8574450084602369, + "grad_norm": 8.0625, + "learning_rate": 4.74211083856068e-06, + "loss": 2.4661, + "step": 8108 + }, + { + "epoch": 0.8575507614213198, + "grad_norm": 7.6875, + "learning_rate": 4.742035078745726e-06, + "loss": 2.558, + "step": 8109 + }, + { + "epoch": 0.8576565143824028, + "grad_norm": 7.5625, + "learning_rate": 4.741959308409818e-06, + "loss": 2.5978, + "step": 8110 + }, + { + "epoch": 0.8577622673434856, + "grad_norm": 7.9375, + "learning_rate": 4.741883527553312e-06, + "loss": 2.6199, + "step": 8111 + }, + { + "epoch": 0.8578680203045685, + "grad_norm": 8.5, + "learning_rate": 4.741807736176563e-06, + "loss": 2.5651, + "step": 8112 + }, + { + "epoch": 0.8579737732656514, + "grad_norm": 8.0625, + "learning_rate": 4.741731934279928e-06, + "loss": 2.6661, + "step": 8113 + }, + { + "epoch": 0.8580795262267343, + "grad_norm": 9.125, + "learning_rate": 4.741656121863761e-06, + "loss": 2.5584, + "step": 8114 + }, + { + "epoch": 0.8581852791878173, + "grad_norm": 7.9375, + "learning_rate": 4.74158029892842e-06, + "loss": 2.529, + "step": 8115 + }, + { + "epoch": 0.8582910321489001, + "grad_norm": 7.9375, + "learning_rate": 4.741504465474258e-06, + "loss": 2.6337, + "step": 8116 + }, + { + "epoch": 0.8583967851099831, + "grad_norm": 8.25, + "learning_rate": 4.741428621501633e-06, + "loss": 2.395, + "step": 8117 + }, + { + "epoch": 0.858502538071066, + "grad_norm": 8.4375, + "learning_rate": 4.741352767010899e-06, + "loss": 2.5854, + "step": 8118 + }, + { + "epoch": 0.8586082910321489, + "grad_norm": 7.65625, + "learning_rate": 4.741276902002415e-06, + "loss": 2.6025, + "step": 8119 + }, + { + "epoch": 0.8587140439932318, + "grad_norm": 9.5, + "learning_rate": 4.741201026476534e-06, + "loss": 2.627, + "step": 8120 + }, + { + "epoch": 0.8588197969543148, + "grad_norm": 7.71875, + "learning_rate": 4.7411251404336135e-06, + "loss": 2.3731, + "step": 8121 + }, + { + "epoch": 0.8589255499153976, + "grad_norm": 7.375, + "learning_rate": 4.741049243874009e-06, + "loss": 2.6386, + "step": 8122 + }, + { + "epoch": 0.8590313028764806, + "grad_norm": 8.0625, + "learning_rate": 4.740973336798077e-06, + "loss": 2.2407, + "step": 8123 + }, + { + "epoch": 0.8591370558375635, + "grad_norm": 8.25, + "learning_rate": 4.7408974192061744e-06, + "loss": 2.5214, + "step": 8124 + }, + { + "epoch": 0.8592428087986463, + "grad_norm": 8.25, + "learning_rate": 4.740821491098656e-06, + "loss": 2.5351, + "step": 8125 + }, + { + "epoch": 0.8593485617597293, + "grad_norm": 7.90625, + "learning_rate": 4.7407455524758795e-06, + "loss": 2.5633, + "step": 8126 + }, + { + "epoch": 0.8594543147208121, + "grad_norm": 8.3125, + "learning_rate": 4.740669603338199e-06, + "loss": 2.5745, + "step": 8127 + }, + { + "epoch": 0.8595600676818951, + "grad_norm": 8.0625, + "learning_rate": 4.740593643685974e-06, + "loss": 2.6296, + "step": 8128 + }, + { + "epoch": 0.859665820642978, + "grad_norm": 8.1875, + "learning_rate": 4.740517673519559e-06, + "loss": 2.5556, + "step": 8129 + }, + { + "epoch": 0.8597715736040609, + "grad_norm": 8.5, + "learning_rate": 4.74044169283931e-06, + "loss": 2.6619, + "step": 8130 + }, + { + "epoch": 0.8598773265651438, + "grad_norm": 7.6875, + "learning_rate": 4.740365701645586e-06, + "loss": 2.493, + "step": 8131 + }, + { + "epoch": 0.8599830795262268, + "grad_norm": 7.84375, + "learning_rate": 4.740289699938741e-06, + "loss": 2.6135, + "step": 8132 + }, + { + "epoch": 0.8600888324873096, + "grad_norm": 8.8125, + "learning_rate": 4.740213687719133e-06, + "loss": 2.642, + "step": 8133 + }, + { + "epoch": 0.8601945854483926, + "grad_norm": 7.53125, + "learning_rate": 4.740137664987119e-06, + "loss": 2.5331, + "step": 8134 + }, + { + "epoch": 0.8603003384094755, + "grad_norm": 7.78125, + "learning_rate": 4.740061631743054e-06, + "loss": 2.5017, + "step": 8135 + }, + { + "epoch": 0.8604060913705583, + "grad_norm": 7.84375, + "learning_rate": 4.739985587987296e-06, + "loss": 2.6366, + "step": 8136 + }, + { + "epoch": 0.8605118443316413, + "grad_norm": 8.0625, + "learning_rate": 4.739909533720203e-06, + "loss": 2.2804, + "step": 8137 + }, + { + "epoch": 0.8606175972927242, + "grad_norm": 8.6875, + "learning_rate": 4.739833468942129e-06, + "loss": 2.8565, + "step": 8138 + }, + { + "epoch": 0.8607233502538071, + "grad_norm": 7.96875, + "learning_rate": 4.739757393653433e-06, + "loss": 2.7965, + "step": 8139 + }, + { + "epoch": 0.86082910321489, + "grad_norm": 8.375, + "learning_rate": 4.739681307854472e-06, + "loss": 2.755, + "step": 8140 + }, + { + "epoch": 0.8609348561759729, + "grad_norm": 7.8125, + "learning_rate": 4.739605211545603e-06, + "loss": 2.4848, + "step": 8141 + }, + { + "epoch": 0.8610406091370558, + "grad_norm": 8.3125, + "learning_rate": 4.739529104727183e-06, + "loss": 2.5063, + "step": 8142 + }, + { + "epoch": 0.8611463620981388, + "grad_norm": 8.375, + "learning_rate": 4.739452987399567e-06, + "loss": 2.322, + "step": 8143 + }, + { + "epoch": 0.8612521150592216, + "grad_norm": 7.90625, + "learning_rate": 4.739376859563116e-06, + "loss": 2.6307, + "step": 8144 + }, + { + "epoch": 0.8613578680203046, + "grad_norm": 7.84375, + "learning_rate": 4.739300721218184e-06, + "loss": 2.7303, + "step": 8145 + }, + { + "epoch": 0.8614636209813875, + "grad_norm": 8.3125, + "learning_rate": 4.7392245723651306e-06, + "loss": 2.4961, + "step": 8146 + }, + { + "epoch": 0.8615693739424704, + "grad_norm": 7.53125, + "learning_rate": 4.739148413004311e-06, + "loss": 2.5376, + "step": 8147 + }, + { + "epoch": 0.8616751269035533, + "grad_norm": 7.40625, + "learning_rate": 4.739072243136085e-06, + "loss": 2.5458, + "step": 8148 + }, + { + "epoch": 0.8617808798646363, + "grad_norm": 8.125, + "learning_rate": 4.738996062760807e-06, + "loss": 2.5996, + "step": 8149 + }, + { + "epoch": 0.8618866328257191, + "grad_norm": 7.8125, + "learning_rate": 4.738919871878838e-06, + "loss": 2.2481, + "step": 8150 + }, + { + "epoch": 0.861992385786802, + "grad_norm": 8.1875, + "learning_rate": 4.7388436704905325e-06, + "loss": 2.7064, + "step": 8151 + }, + { + "epoch": 0.862098138747885, + "grad_norm": 7.75, + "learning_rate": 4.73876745859625e-06, + "loss": 2.7339, + "step": 8152 + }, + { + "epoch": 0.8622038917089678, + "grad_norm": 7.53125, + "learning_rate": 4.7386912361963465e-06, + "loss": 2.506, + "step": 8153 + }, + { + "epoch": 0.8623096446700508, + "grad_norm": 7.65625, + "learning_rate": 4.738615003291181e-06, + "loss": 2.6265, + "step": 8154 + }, + { + "epoch": 0.8624153976311336, + "grad_norm": 7.59375, + "learning_rate": 4.738538759881111e-06, + "loss": 2.3909, + "step": 8155 + }, + { + "epoch": 0.8625211505922166, + "grad_norm": 7.90625, + "learning_rate": 4.7384625059664945e-06, + "loss": 2.721, + "step": 8156 + }, + { + "epoch": 0.8626269035532995, + "grad_norm": 7.84375, + "learning_rate": 4.738386241547689e-06, + "loss": 2.43, + "step": 8157 + }, + { + "epoch": 0.8627326565143824, + "grad_norm": 7.75, + "learning_rate": 4.738309966625051e-06, + "loss": 2.6722, + "step": 8158 + }, + { + "epoch": 0.8628384094754653, + "grad_norm": 8.0625, + "learning_rate": 4.738233681198941e-06, + "loss": 2.6506, + "step": 8159 + }, + { + "epoch": 0.8629441624365483, + "grad_norm": 7.78125, + "learning_rate": 4.738157385269715e-06, + "loss": 2.4509, + "step": 8160 + }, + { + "epoch": 0.8630499153976311, + "grad_norm": 7.78125, + "learning_rate": 4.738081078837733e-06, + "loss": 2.4687, + "step": 8161 + }, + { + "epoch": 0.863155668358714, + "grad_norm": 8.0625, + "learning_rate": 4.7380047619033505e-06, + "loss": 2.6865, + "step": 8162 + }, + { + "epoch": 0.863261421319797, + "grad_norm": 8.625, + "learning_rate": 4.737928434466928e-06, + "loss": 2.6563, + "step": 8163 + }, + { + "epoch": 0.8633671742808798, + "grad_norm": 10.75, + "learning_rate": 4.737852096528822e-06, + "loss": 2.4103, + "step": 8164 + }, + { + "epoch": 0.8634729272419628, + "grad_norm": 7.96875, + "learning_rate": 4.737775748089392e-06, + "loss": 2.5411, + "step": 8165 + }, + { + "epoch": 0.8635786802030457, + "grad_norm": 8.0, + "learning_rate": 4.7376993891489946e-06, + "loss": 2.4929, + "step": 8166 + }, + { + "epoch": 0.8636844331641286, + "grad_norm": 7.9375, + "learning_rate": 4.73762301970799e-06, + "loss": 2.4626, + "step": 8167 + }, + { + "epoch": 0.8637901861252115, + "grad_norm": 7.375, + "learning_rate": 4.737546639766736e-06, + "loss": 2.4084, + "step": 8168 + }, + { + "epoch": 0.8638959390862944, + "grad_norm": 7.9375, + "learning_rate": 4.73747024932559e-06, + "loss": 2.5404, + "step": 8169 + }, + { + "epoch": 0.8640016920473773, + "grad_norm": 8.4375, + "learning_rate": 4.737393848384911e-06, + "loss": 2.698, + "step": 8170 + }, + { + "epoch": 0.8641074450084603, + "grad_norm": 8.0625, + "learning_rate": 4.737317436945058e-06, + "loss": 2.5777, + "step": 8171 + }, + { + "epoch": 0.8642131979695431, + "grad_norm": 7.6875, + "learning_rate": 4.7372410150063895e-06, + "loss": 2.4209, + "step": 8172 + }, + { + "epoch": 0.864318950930626, + "grad_norm": 8.3125, + "learning_rate": 4.7371645825692635e-06, + "loss": 2.7394, + "step": 8173 + }, + { + "epoch": 0.864424703891709, + "grad_norm": 8.4375, + "learning_rate": 4.73708813963404e-06, + "loss": 2.785, + "step": 8174 + }, + { + "epoch": 0.8645304568527918, + "grad_norm": 8.3125, + "learning_rate": 4.737011686201076e-06, + "loss": 2.5679, + "step": 8175 + }, + { + "epoch": 0.8646362098138748, + "grad_norm": 7.875, + "learning_rate": 4.736935222270732e-06, + "loss": 2.5122, + "step": 8176 + }, + { + "epoch": 0.8647419627749577, + "grad_norm": 7.71875, + "learning_rate": 4.736858747843365e-06, + "loss": 2.386, + "step": 8177 + }, + { + "epoch": 0.8648477157360406, + "grad_norm": 7.78125, + "learning_rate": 4.736782262919335e-06, + "loss": 2.7068, + "step": 8178 + }, + { + "epoch": 0.8649534686971235, + "grad_norm": 7.75, + "learning_rate": 4.736705767499e-06, + "loss": 2.6312, + "step": 8179 + }, + { + "epoch": 0.8650592216582065, + "grad_norm": 8.25, + "learning_rate": 4.73662926158272e-06, + "loss": 2.653, + "step": 8180 + }, + { + "epoch": 0.8651649746192893, + "grad_norm": 7.90625, + "learning_rate": 4.736552745170854e-06, + "loss": 2.3771, + "step": 8181 + }, + { + "epoch": 0.8652707275803723, + "grad_norm": 7.65625, + "learning_rate": 4.73647621826376e-06, + "loss": 2.5919, + "step": 8182 + }, + { + "epoch": 0.8653764805414551, + "grad_norm": 8.5625, + "learning_rate": 4.736399680861799e-06, + "loss": 2.6338, + "step": 8183 + }, + { + "epoch": 0.8654822335025381, + "grad_norm": 7.78125, + "learning_rate": 4.736323132965328e-06, + "loss": 2.5336, + "step": 8184 + }, + { + "epoch": 0.865587986463621, + "grad_norm": 8.4375, + "learning_rate": 4.736246574574708e-06, + "loss": 2.6803, + "step": 8185 + }, + { + "epoch": 0.8656937394247038, + "grad_norm": 7.5625, + "learning_rate": 4.736170005690296e-06, + "loss": 2.5027, + "step": 8186 + }, + { + "epoch": 0.8657994923857868, + "grad_norm": 8.0625, + "learning_rate": 4.7360934263124534e-06, + "loss": 2.5522, + "step": 8187 + }, + { + "epoch": 0.8659052453468697, + "grad_norm": 8.4375, + "learning_rate": 4.736016836441539e-06, + "loss": 2.7461, + "step": 8188 + }, + { + "epoch": 0.8660109983079526, + "grad_norm": 9.0625, + "learning_rate": 4.735940236077913e-06, + "loss": 2.5412, + "step": 8189 + }, + { + "epoch": 0.8661167512690355, + "grad_norm": 8.0, + "learning_rate": 4.735863625221932e-06, + "loss": 2.4243, + "step": 8190 + }, + { + "epoch": 0.8662225042301185, + "grad_norm": 8.25, + "learning_rate": 4.735787003873959e-06, + "loss": 2.3941, + "step": 8191 + }, + { + "epoch": 0.8663282571912013, + "grad_norm": 7.625, + "learning_rate": 4.7357103720343524e-06, + "loss": 2.446, + "step": 8192 + }, + { + "epoch": 0.8664340101522843, + "grad_norm": 7.4375, + "learning_rate": 4.73563372970347e-06, + "loss": 2.3086, + "step": 8193 + }, + { + "epoch": 0.8665397631133672, + "grad_norm": 8.75, + "learning_rate": 4.735557076881673e-06, + "loss": 2.6495, + "step": 8194 + }, + { + "epoch": 0.8666455160744501, + "grad_norm": 7.84375, + "learning_rate": 4.735480413569322e-06, + "loss": 2.5793, + "step": 8195 + }, + { + "epoch": 0.866751269035533, + "grad_norm": 7.875, + "learning_rate": 4.735403739766775e-06, + "loss": 2.632, + "step": 8196 + }, + { + "epoch": 0.8668570219966159, + "grad_norm": 9.4375, + "learning_rate": 4.7353270554743924e-06, + "loss": 2.6499, + "step": 8197 + }, + { + "epoch": 0.8669627749576988, + "grad_norm": 7.84375, + "learning_rate": 4.735250360692534e-06, + "loss": 2.4578, + "step": 8198 + }, + { + "epoch": 0.8670685279187818, + "grad_norm": 8.1875, + "learning_rate": 4.73517365542156e-06, + "loss": 2.5962, + "step": 8199 + }, + { + "epoch": 0.8671742808798646, + "grad_norm": 8.3125, + "learning_rate": 4.735096939661831e-06, + "loss": 2.5225, + "step": 8200 + }, + { + "epoch": 0.8672800338409475, + "grad_norm": 7.8125, + "learning_rate": 4.735020213413704e-06, + "loss": 2.5258, + "step": 8201 + }, + { + "epoch": 0.8673857868020305, + "grad_norm": 7.625, + "learning_rate": 4.734943476677544e-06, + "loss": 2.5847, + "step": 8202 + }, + { + "epoch": 0.8674915397631133, + "grad_norm": 7.6875, + "learning_rate": 4.7348667294537065e-06, + "loss": 2.6391, + "step": 8203 + }, + { + "epoch": 0.8675972927241963, + "grad_norm": 7.84375, + "learning_rate": 4.734789971742554e-06, + "loss": 2.7001, + "step": 8204 + }, + { + "epoch": 0.8677030456852792, + "grad_norm": 8.25, + "learning_rate": 4.734713203544447e-06, + "loss": 2.5846, + "step": 8205 + }, + { + "epoch": 0.8678087986463621, + "grad_norm": 7.96875, + "learning_rate": 4.734636424859743e-06, + "loss": 2.5269, + "step": 8206 + }, + { + "epoch": 0.867914551607445, + "grad_norm": 7.6875, + "learning_rate": 4.734559635688806e-06, + "loss": 2.5894, + "step": 8207 + }, + { + "epoch": 0.868020304568528, + "grad_norm": 8.125, + "learning_rate": 4.7344828360319934e-06, + "loss": 2.676, + "step": 8208 + }, + { + "epoch": 0.8681260575296108, + "grad_norm": 7.71875, + "learning_rate": 4.734406025889668e-06, + "loss": 2.6868, + "step": 8209 + }, + { + "epoch": 0.8682318104906938, + "grad_norm": 7.84375, + "learning_rate": 4.734329205262187e-06, + "loss": 2.8668, + "step": 8210 + }, + { + "epoch": 0.8683375634517766, + "grad_norm": 7.75, + "learning_rate": 4.734252374149914e-06, + "loss": 2.6279, + "step": 8211 + }, + { + "epoch": 0.8684433164128595, + "grad_norm": 8.5625, + "learning_rate": 4.734175532553208e-06, + "loss": 2.3341, + "step": 8212 + }, + { + "epoch": 0.8685490693739425, + "grad_norm": 8.0625, + "learning_rate": 4.7340986804724305e-06, + "loss": 2.4927, + "step": 8213 + }, + { + "epoch": 0.8686548223350253, + "grad_norm": 8.375, + "learning_rate": 4.734021817907941e-06, + "loss": 2.5551, + "step": 8214 + }, + { + "epoch": 0.8687605752961083, + "grad_norm": 8.25, + "learning_rate": 4.733944944860102e-06, + "loss": 2.5092, + "step": 8215 + }, + { + "epoch": 0.8688663282571912, + "grad_norm": 7.59375, + "learning_rate": 4.733868061329272e-06, + "loss": 2.5508, + "step": 8216 + }, + { + "epoch": 0.8689720812182741, + "grad_norm": 7.84375, + "learning_rate": 4.733791167315813e-06, + "loss": 2.5187, + "step": 8217 + }, + { + "epoch": 0.869077834179357, + "grad_norm": 7.90625, + "learning_rate": 4.733714262820086e-06, + "loss": 2.6067, + "step": 8218 + }, + { + "epoch": 0.86918358714044, + "grad_norm": 8.1875, + "learning_rate": 4.7336373478424504e-06, + "loss": 2.4418, + "step": 8219 + }, + { + "epoch": 0.8692893401015228, + "grad_norm": 7.875, + "learning_rate": 4.7335604223832685e-06, + "loss": 2.4492, + "step": 8220 + }, + { + "epoch": 0.8693950930626058, + "grad_norm": 8.25, + "learning_rate": 4.733483486442903e-06, + "loss": 2.6853, + "step": 8221 + }, + { + "epoch": 0.8695008460236887, + "grad_norm": 7.96875, + "learning_rate": 4.733406540021711e-06, + "loss": 2.3613, + "step": 8222 + }, + { + "epoch": 0.8696065989847716, + "grad_norm": 8.3125, + "learning_rate": 4.733329583120055e-06, + "loss": 2.5858, + "step": 8223 + }, + { + "epoch": 0.8697123519458545, + "grad_norm": 8.0, + "learning_rate": 4.733252615738298e-06, + "loss": 2.4803, + "step": 8224 + }, + { + "epoch": 0.8698181049069373, + "grad_norm": 8.125, + "learning_rate": 4.7331756378768005e-06, + "loss": 2.5913, + "step": 8225 + }, + { + "epoch": 0.8699238578680203, + "grad_norm": 7.21875, + "learning_rate": 4.733098649535922e-06, + "loss": 2.4188, + "step": 8226 + }, + { + "epoch": 0.8700296108291032, + "grad_norm": 8.3125, + "learning_rate": 4.733021650716025e-06, + "loss": 2.7148, + "step": 8227 + }, + { + "epoch": 0.8701353637901861, + "grad_norm": 8.25, + "learning_rate": 4.732944641417471e-06, + "loss": 2.6718, + "step": 8228 + }, + { + "epoch": 0.870241116751269, + "grad_norm": 8.0, + "learning_rate": 4.73286762164062e-06, + "loss": 2.4114, + "step": 8229 + }, + { + "epoch": 0.870346869712352, + "grad_norm": 7.75, + "learning_rate": 4.7327905913858355e-06, + "loss": 2.7003, + "step": 8230 + }, + { + "epoch": 0.8704526226734348, + "grad_norm": 8.4375, + "learning_rate": 4.732713550653477e-06, + "loss": 2.6396, + "step": 8231 + }, + { + "epoch": 0.8705583756345178, + "grad_norm": 7.78125, + "learning_rate": 4.732636499443908e-06, + "loss": 2.2638, + "step": 8232 + }, + { + "epoch": 0.8706641285956007, + "grad_norm": 7.8125, + "learning_rate": 4.7325594377574875e-06, + "loss": 2.5114, + "step": 8233 + }, + { + "epoch": 0.8707698815566836, + "grad_norm": 7.5, + "learning_rate": 4.732482365594581e-06, + "loss": 2.5719, + "step": 8234 + }, + { + "epoch": 0.8708756345177665, + "grad_norm": 7.96875, + "learning_rate": 4.7324052829555464e-06, + "loss": 2.5346, + "step": 8235 + }, + { + "epoch": 0.8709813874788495, + "grad_norm": 7.5, + "learning_rate": 4.732328189840747e-06, + "loss": 2.4623, + "step": 8236 + }, + { + "epoch": 0.8710871404399323, + "grad_norm": 8.6875, + "learning_rate": 4.732251086250544e-06, + "loss": 2.7171, + "step": 8237 + }, + { + "epoch": 0.8711928934010152, + "grad_norm": 8.4375, + "learning_rate": 4.7321739721853e-06, + "loss": 2.7395, + "step": 8238 + }, + { + "epoch": 0.8712986463620981, + "grad_norm": 8.3125, + "learning_rate": 4.7320968476453765e-06, + "loss": 2.7385, + "step": 8239 + }, + { + "epoch": 0.871404399323181, + "grad_norm": 8.375, + "learning_rate": 4.732019712631135e-06, + "loss": 2.3114, + "step": 8240 + }, + { + "epoch": 0.871510152284264, + "grad_norm": 7.59375, + "learning_rate": 4.731942567142939e-06, + "loss": 2.5231, + "step": 8241 + }, + { + "epoch": 0.8716159052453468, + "grad_norm": 7.78125, + "learning_rate": 4.731865411181148e-06, + "loss": 2.3994, + "step": 8242 + }, + { + "epoch": 0.8717216582064298, + "grad_norm": 7.28125, + "learning_rate": 4.731788244746126e-06, + "loss": 2.7203, + "step": 8243 + }, + { + "epoch": 0.8718274111675127, + "grad_norm": 8.375, + "learning_rate": 4.731711067838234e-06, + "loss": 2.7117, + "step": 8244 + }, + { + "epoch": 0.8719331641285956, + "grad_norm": 8.0625, + "learning_rate": 4.7316338804578355e-06, + "loss": 2.804, + "step": 8245 + }, + { + "epoch": 0.8720389170896785, + "grad_norm": 8.25, + "learning_rate": 4.731556682605291e-06, + "loss": 2.525, + "step": 8246 + }, + { + "epoch": 0.8721446700507615, + "grad_norm": 8.4375, + "learning_rate": 4.731479474280965e-06, + "loss": 2.4954, + "step": 8247 + }, + { + "epoch": 0.8722504230118443, + "grad_norm": 8.25, + "learning_rate": 4.731402255485218e-06, + "loss": 2.644, + "step": 8248 + }, + { + "epoch": 0.8723561759729273, + "grad_norm": 8.875, + "learning_rate": 4.731325026218412e-06, + "loss": 2.617, + "step": 8249 + }, + { + "epoch": 0.8724619289340102, + "grad_norm": 8.0, + "learning_rate": 4.7312477864809115e-06, + "loss": 2.3747, + "step": 8250 + }, + { + "epoch": 0.872567681895093, + "grad_norm": 8.1875, + "learning_rate": 4.731170536273076e-06, + "loss": 2.7771, + "step": 8251 + }, + { + "epoch": 0.872673434856176, + "grad_norm": 8.125, + "learning_rate": 4.731093275595271e-06, + "loss": 2.5008, + "step": 8252 + }, + { + "epoch": 0.8727791878172588, + "grad_norm": 7.75, + "learning_rate": 4.731016004447858e-06, + "loss": 2.4236, + "step": 8253 + }, + { + "epoch": 0.8728849407783418, + "grad_norm": 8.125, + "learning_rate": 4.730938722831199e-06, + "loss": 2.6665, + "step": 8254 + }, + { + "epoch": 0.8729906937394247, + "grad_norm": 8.1875, + "learning_rate": 4.730861430745657e-06, + "loss": 2.5291, + "step": 8255 + }, + { + "epoch": 0.8730964467005076, + "grad_norm": 7.9375, + "learning_rate": 4.730784128191594e-06, + "loss": 2.2296, + "step": 8256 + }, + { + "epoch": 0.8732021996615905, + "grad_norm": 8.1875, + "learning_rate": 4.730706815169374e-06, + "loss": 2.6333, + "step": 8257 + }, + { + "epoch": 0.8733079526226735, + "grad_norm": 8.25, + "learning_rate": 4.730629491679359e-06, + "loss": 2.543, + "step": 8258 + }, + { + "epoch": 0.8734137055837563, + "grad_norm": 8.0625, + "learning_rate": 4.730552157721912e-06, + "loss": 2.5257, + "step": 8259 + }, + { + "epoch": 0.8735194585448393, + "grad_norm": 7.1875, + "learning_rate": 4.7304748132973965e-06, + "loss": 2.5134, + "step": 8260 + }, + { + "epoch": 0.8736252115059222, + "grad_norm": 8.625, + "learning_rate": 4.730397458406175e-06, + "loss": 2.9156, + "step": 8261 + }, + { + "epoch": 0.873730964467005, + "grad_norm": 7.46875, + "learning_rate": 4.73032009304861e-06, + "loss": 2.279, + "step": 8262 + }, + { + "epoch": 0.873836717428088, + "grad_norm": 7.96875, + "learning_rate": 4.730242717225065e-06, + "loss": 2.5133, + "step": 8263 + }, + { + "epoch": 0.873942470389171, + "grad_norm": 7.96875, + "learning_rate": 4.730165330935903e-06, + "loss": 2.6084, + "step": 8264 + }, + { + "epoch": 0.8740482233502538, + "grad_norm": 8.0625, + "learning_rate": 4.730087934181487e-06, + "loss": 2.4147, + "step": 8265 + }, + { + "epoch": 0.8741539763113367, + "grad_norm": 8.3125, + "learning_rate": 4.730010526962181e-06, + "loss": 2.7243, + "step": 8266 + }, + { + "epoch": 0.8742597292724196, + "grad_norm": 7.4375, + "learning_rate": 4.729933109278348e-06, + "loss": 2.4529, + "step": 8267 + }, + { + "epoch": 0.8743654822335025, + "grad_norm": 7.6875, + "learning_rate": 4.72985568113035e-06, + "loss": 2.4262, + "step": 8268 + }, + { + "epoch": 0.8744712351945855, + "grad_norm": 8.125, + "learning_rate": 4.7297782425185515e-06, + "loss": 2.5557, + "step": 8269 + }, + { + "epoch": 0.8745769881556683, + "grad_norm": 8.0625, + "learning_rate": 4.729700793443316e-06, + "loss": 2.6227, + "step": 8270 + }, + { + "epoch": 0.8746827411167513, + "grad_norm": 7.90625, + "learning_rate": 4.729623333905006e-06, + "loss": 2.8245, + "step": 8271 + }, + { + "epoch": 0.8747884940778342, + "grad_norm": 7.75, + "learning_rate": 4.729545863903986e-06, + "loss": 2.53, + "step": 8272 + }, + { + "epoch": 0.874894247038917, + "grad_norm": 7.6875, + "learning_rate": 4.7294683834406185e-06, + "loss": 2.5863, + "step": 8273 + }, + { + "epoch": 0.875, + "grad_norm": 7.84375, + "learning_rate": 4.729390892515268e-06, + "loss": 2.3302, + "step": 8274 + }, + { + "epoch": 0.875105752961083, + "grad_norm": 7.71875, + "learning_rate": 4.729313391128298e-06, + "loss": 2.7796, + "step": 8275 + }, + { + "epoch": 0.8752115059221658, + "grad_norm": 7.84375, + "learning_rate": 4.729235879280073e-06, + "loss": 2.4569, + "step": 8276 + }, + { + "epoch": 0.8753172588832487, + "grad_norm": 7.8125, + "learning_rate": 4.729158356970954e-06, + "loss": 2.5556, + "step": 8277 + }, + { + "epoch": 0.8754230118443317, + "grad_norm": 7.65625, + "learning_rate": 4.729080824201308e-06, + "loss": 2.5754, + "step": 8278 + }, + { + "epoch": 0.8755287648054145, + "grad_norm": 8.125, + "learning_rate": 4.729003280971497e-06, + "loss": 2.5879, + "step": 8279 + }, + { + "epoch": 0.8756345177664975, + "grad_norm": 10.0, + "learning_rate": 4.7289257272818845e-06, + "loss": 2.3036, + "step": 8280 + }, + { + "epoch": 0.8757402707275804, + "grad_norm": 8.0, + "learning_rate": 4.728848163132835e-06, + "loss": 2.785, + "step": 8281 + }, + { + "epoch": 0.8758460236886633, + "grad_norm": 7.96875, + "learning_rate": 4.728770588524714e-06, + "loss": 2.5337, + "step": 8282 + }, + { + "epoch": 0.8759517766497462, + "grad_norm": 8.625, + "learning_rate": 4.728693003457883e-06, + "loss": 2.6352, + "step": 8283 + }, + { + "epoch": 0.876057529610829, + "grad_norm": 8.0, + "learning_rate": 4.728615407932708e-06, + "loss": 2.6963, + "step": 8284 + }, + { + "epoch": 0.876163282571912, + "grad_norm": 7.375, + "learning_rate": 4.728537801949552e-06, + "loss": 2.4612, + "step": 8285 + }, + { + "epoch": 0.876269035532995, + "grad_norm": 8.25, + "learning_rate": 4.72846018550878e-06, + "loss": 2.5252, + "step": 8286 + }, + { + "epoch": 0.8763747884940778, + "grad_norm": 8.0625, + "learning_rate": 4.728382558610755e-06, + "loss": 2.5039, + "step": 8287 + }, + { + "epoch": 0.8764805414551607, + "grad_norm": 7.8125, + "learning_rate": 4.728304921255842e-06, + "loss": 2.499, + "step": 8288 + }, + { + "epoch": 0.8765862944162437, + "grad_norm": 8.0, + "learning_rate": 4.728227273444406e-06, + "loss": 2.5323, + "step": 8289 + }, + { + "epoch": 0.8766920473773265, + "grad_norm": 8.0, + "learning_rate": 4.72814961517681e-06, + "loss": 2.6952, + "step": 8290 + }, + { + "epoch": 0.8767978003384095, + "grad_norm": 7.625, + "learning_rate": 4.7280719464534195e-06, + "loss": 2.6513, + "step": 8291 + }, + { + "epoch": 0.8769035532994924, + "grad_norm": 7.875, + "learning_rate": 4.727994267274599e-06, + "loss": 2.4587, + "step": 8292 + }, + { + "epoch": 0.8770093062605753, + "grad_norm": 8.625, + "learning_rate": 4.727916577640712e-06, + "loss": 2.8258, + "step": 8293 + }, + { + "epoch": 0.8771150592216582, + "grad_norm": 7.90625, + "learning_rate": 4.727838877552124e-06, + "loss": 2.4274, + "step": 8294 + }, + { + "epoch": 0.8772208121827412, + "grad_norm": 8.25, + "learning_rate": 4.7277611670092e-06, + "loss": 2.7143, + "step": 8295 + }, + { + "epoch": 0.877326565143824, + "grad_norm": 8.625, + "learning_rate": 4.727683446012302e-06, + "loss": 2.6181, + "step": 8296 + }, + { + "epoch": 0.877432318104907, + "grad_norm": 8.0, + "learning_rate": 4.7276057145617985e-06, + "loss": 2.4256, + "step": 8297 + }, + { + "epoch": 0.8775380710659898, + "grad_norm": 7.65625, + "learning_rate": 4.727527972658052e-06, + "loss": 2.512, + "step": 8298 + }, + { + "epoch": 0.8776438240270727, + "grad_norm": 7.9375, + "learning_rate": 4.727450220301427e-06, + "loss": 2.4147, + "step": 8299 + }, + { + "epoch": 0.8777495769881557, + "grad_norm": 8.3125, + "learning_rate": 4.72737245749229e-06, + "loss": 2.5187, + "step": 8300 + }, + { + "epoch": 0.8778553299492385, + "grad_norm": 8.125, + "learning_rate": 4.727294684231005e-06, + "loss": 2.6071, + "step": 8301 + }, + { + "epoch": 0.8779610829103215, + "grad_norm": 10.4375, + "learning_rate": 4.727216900517936e-06, + "loss": 2.5527, + "step": 8302 + }, + { + "epoch": 0.8780668358714044, + "grad_norm": 8.1875, + "learning_rate": 4.72713910635345e-06, + "loss": 2.6339, + "step": 8303 + }, + { + "epoch": 0.8781725888324873, + "grad_norm": 8.0, + "learning_rate": 4.727061301737911e-06, + "loss": 2.7693, + "step": 8304 + }, + { + "epoch": 0.8782783417935702, + "grad_norm": 8.375, + "learning_rate": 4.726983486671684e-06, + "loss": 2.7515, + "step": 8305 + }, + { + "epoch": 0.8783840947546532, + "grad_norm": 8.75, + "learning_rate": 4.726905661155134e-06, + "loss": 2.4666, + "step": 8306 + }, + { + "epoch": 0.878489847715736, + "grad_norm": 7.84375, + "learning_rate": 4.7268278251886264e-06, + "loss": 2.6421, + "step": 8307 + }, + { + "epoch": 0.878595600676819, + "grad_norm": 8.125, + "learning_rate": 4.726749978772527e-06, + "loss": 2.6244, + "step": 8308 + }, + { + "epoch": 0.8787013536379019, + "grad_norm": 8.125, + "learning_rate": 4.7266721219072e-06, + "loss": 2.7609, + "step": 8309 + }, + { + "epoch": 0.8788071065989848, + "grad_norm": 7.90625, + "learning_rate": 4.726594254593012e-06, + "loss": 2.6389, + "step": 8310 + }, + { + "epoch": 0.8789128595600677, + "grad_norm": 7.84375, + "learning_rate": 4.726516376830328e-06, + "loss": 2.4925, + "step": 8311 + }, + { + "epoch": 0.8790186125211505, + "grad_norm": 7.78125, + "learning_rate": 4.726438488619513e-06, + "loss": 2.728, + "step": 8312 + }, + { + "epoch": 0.8791243654822335, + "grad_norm": 7.65625, + "learning_rate": 4.726360589960933e-06, + "loss": 2.6465, + "step": 8313 + }, + { + "epoch": 0.8792301184433164, + "grad_norm": 16.625, + "learning_rate": 4.7262826808549525e-06, + "loss": 2.5119, + "step": 8314 + }, + { + "epoch": 0.8793358714043993, + "grad_norm": 7.96875, + "learning_rate": 4.726204761301939e-06, + "loss": 2.2896, + "step": 8315 + }, + { + "epoch": 0.8794416243654822, + "grad_norm": 7.3125, + "learning_rate": 4.726126831302256e-06, + "loss": 2.4767, + "step": 8316 + }, + { + "epoch": 0.8795473773265652, + "grad_norm": 7.65625, + "learning_rate": 4.726048890856271e-06, + "loss": 2.5017, + "step": 8317 + }, + { + "epoch": 0.879653130287648, + "grad_norm": 8.4375, + "learning_rate": 4.7259709399643486e-06, + "loss": 2.5607, + "step": 8318 + }, + { + "epoch": 0.879758883248731, + "grad_norm": 7.28125, + "learning_rate": 4.725892978626855e-06, + "loss": 2.5987, + "step": 8319 + }, + { + "epoch": 0.8798646362098139, + "grad_norm": 8.0, + "learning_rate": 4.7258150068441566e-06, + "loss": 2.719, + "step": 8320 + }, + { + "epoch": 0.8799703891708968, + "grad_norm": 8.125, + "learning_rate": 4.725737024616618e-06, + "loss": 2.7073, + "step": 8321 + }, + { + "epoch": 0.8800761421319797, + "grad_norm": 7.78125, + "learning_rate": 4.725659031944606e-06, + "loss": 2.6508, + "step": 8322 + }, + { + "epoch": 0.8801818950930627, + "grad_norm": 8.4375, + "learning_rate": 4.725581028828487e-06, + "loss": 2.702, + "step": 8323 + }, + { + "epoch": 0.8802876480541455, + "grad_norm": 8.1875, + "learning_rate": 4.725503015268626e-06, + "loss": 2.6809, + "step": 8324 + }, + { + "epoch": 0.8803934010152284, + "grad_norm": 8.3125, + "learning_rate": 4.725424991265388e-06, + "loss": 2.7824, + "step": 8325 + }, + { + "epoch": 0.8804991539763113, + "grad_norm": 7.78125, + "learning_rate": 4.725346956819143e-06, + "loss": 2.6258, + "step": 8326 + }, + { + "epoch": 0.8806049069373942, + "grad_norm": 7.9375, + "learning_rate": 4.725268911930254e-06, + "loss": 2.6163, + "step": 8327 + }, + { + "epoch": 0.8807106598984772, + "grad_norm": 8.3125, + "learning_rate": 4.725190856599088e-06, + "loss": 2.5667, + "step": 8328 + }, + { + "epoch": 0.88081641285956, + "grad_norm": 8.8125, + "learning_rate": 4.725112790826011e-06, + "loss": 2.67, + "step": 8329 + }, + { + "epoch": 0.880922165820643, + "grad_norm": 7.5625, + "learning_rate": 4.725034714611391e-06, + "loss": 2.5846, + "step": 8330 + }, + { + "epoch": 0.8810279187817259, + "grad_norm": 7.5625, + "learning_rate": 4.724956627955591e-06, + "loss": 2.569, + "step": 8331 + }, + { + "epoch": 0.8811336717428088, + "grad_norm": 7.625, + "learning_rate": 4.724878530858982e-06, + "loss": 2.5773, + "step": 8332 + }, + { + "epoch": 0.8812394247038917, + "grad_norm": 8.375, + "learning_rate": 4.724800423321926e-06, + "loss": 2.5377, + "step": 8333 + }, + { + "epoch": 0.8813451776649747, + "grad_norm": 8.125, + "learning_rate": 4.724722305344793e-06, + "loss": 2.7628, + "step": 8334 + }, + { + "epoch": 0.8814509306260575, + "grad_norm": 7.78125, + "learning_rate": 4.724644176927947e-06, + "loss": 2.53, + "step": 8335 + }, + { + "epoch": 0.8815566835871405, + "grad_norm": 8.3125, + "learning_rate": 4.724566038071756e-06, + "loss": 2.6774, + "step": 8336 + }, + { + "epoch": 0.8816624365482234, + "grad_norm": 8.8125, + "learning_rate": 4.7244878887765864e-06, + "loss": 2.8966, + "step": 8337 + }, + { + "epoch": 0.8817681895093062, + "grad_norm": 7.75, + "learning_rate": 4.724409729042806e-06, + "loss": 2.4399, + "step": 8338 + }, + { + "epoch": 0.8818739424703892, + "grad_norm": 8.0, + "learning_rate": 4.72433155887078e-06, + "loss": 2.7953, + "step": 8339 + }, + { + "epoch": 0.881979695431472, + "grad_norm": 7.78125, + "learning_rate": 4.724253378260875e-06, + "loss": 2.5505, + "step": 8340 + }, + { + "epoch": 0.882085448392555, + "grad_norm": 8.0, + "learning_rate": 4.7241751872134586e-06, + "loss": 2.5142, + "step": 8341 + }, + { + "epoch": 0.8821912013536379, + "grad_norm": 8.4375, + "learning_rate": 4.724096985728898e-06, + "loss": 2.8834, + "step": 8342 + }, + { + "epoch": 0.8822969543147208, + "grad_norm": 8.0625, + "learning_rate": 4.72401877380756e-06, + "loss": 2.5034, + "step": 8343 + }, + { + "epoch": 0.8824027072758037, + "grad_norm": 7.46875, + "learning_rate": 4.72394055144981e-06, + "loss": 2.4973, + "step": 8344 + }, + { + "epoch": 0.8825084602368867, + "grad_norm": 7.5625, + "learning_rate": 4.723862318656018e-06, + "loss": 2.4073, + "step": 8345 + }, + { + "epoch": 0.8826142131979695, + "grad_norm": 7.71875, + "learning_rate": 4.723784075426551e-06, + "loss": 2.622, + "step": 8346 + }, + { + "epoch": 0.8827199661590525, + "grad_norm": 8.0, + "learning_rate": 4.723705821761772e-06, + "loss": 2.7979, + "step": 8347 + }, + { + "epoch": 0.8828257191201354, + "grad_norm": 8.0625, + "learning_rate": 4.723627557662053e-06, + "loss": 2.4776, + "step": 8348 + }, + { + "epoch": 0.8829314720812182, + "grad_norm": 7.6875, + "learning_rate": 4.723549283127759e-06, + "loss": 2.5114, + "step": 8349 + }, + { + "epoch": 0.8830372250423012, + "grad_norm": 8.5625, + "learning_rate": 4.723470998159257e-06, + "loss": 2.757, + "step": 8350 + }, + { + "epoch": 0.8831429780033841, + "grad_norm": 7.59375, + "learning_rate": 4.723392702756916e-06, + "loss": 2.5343, + "step": 8351 + }, + { + "epoch": 0.883248730964467, + "grad_norm": 7.6875, + "learning_rate": 4.723314396921101e-06, + "loss": 2.5854, + "step": 8352 + }, + { + "epoch": 0.8833544839255499, + "grad_norm": 7.59375, + "learning_rate": 4.723236080652182e-06, + "loss": 2.6307, + "step": 8353 + }, + { + "epoch": 0.8834602368866328, + "grad_norm": 8.5, + "learning_rate": 4.723157753950525e-06, + "loss": 2.5635, + "step": 8354 + }, + { + "epoch": 0.8835659898477157, + "grad_norm": 7.71875, + "learning_rate": 4.723079416816497e-06, + "loss": 2.7062, + "step": 8355 + }, + { + "epoch": 0.8836717428087987, + "grad_norm": 8.375, + "learning_rate": 4.723001069250468e-06, + "loss": 2.4532, + "step": 8356 + }, + { + "epoch": 0.8837774957698815, + "grad_norm": 7.90625, + "learning_rate": 4.722922711252803e-06, + "loss": 2.5266, + "step": 8357 + }, + { + "epoch": 0.8838832487309645, + "grad_norm": 8.3125, + "learning_rate": 4.722844342823871e-06, + "loss": 2.2651, + "step": 8358 + }, + { + "epoch": 0.8839890016920474, + "grad_norm": 8.9375, + "learning_rate": 4.722765963964041e-06, + "loss": 2.2659, + "step": 8359 + }, + { + "epoch": 0.8840947546531303, + "grad_norm": 7.8125, + "learning_rate": 4.722687574673678e-06, + "loss": 2.615, + "step": 8360 + }, + { + "epoch": 0.8842005076142132, + "grad_norm": 8.4375, + "learning_rate": 4.722609174953151e-06, + "loss": 2.7598, + "step": 8361 + }, + { + "epoch": 0.8843062605752962, + "grad_norm": 7.40625, + "learning_rate": 4.722530764802829e-06, + "loss": 2.52, + "step": 8362 + }, + { + "epoch": 0.884412013536379, + "grad_norm": 7.65625, + "learning_rate": 4.722452344223079e-06, + "loss": 2.4736, + "step": 8363 + }, + { + "epoch": 0.8845177664974619, + "grad_norm": 8.25, + "learning_rate": 4.7223739132142684e-06, + "loss": 2.8847, + "step": 8364 + }, + { + "epoch": 0.8846235194585449, + "grad_norm": 7.90625, + "learning_rate": 4.722295471776766e-06, + "loss": 2.5704, + "step": 8365 + }, + { + "epoch": 0.8847292724196277, + "grad_norm": 7.625, + "learning_rate": 4.722217019910941e-06, + "loss": 2.5854, + "step": 8366 + }, + { + "epoch": 0.8848350253807107, + "grad_norm": 7.5, + "learning_rate": 4.72213855761716e-06, + "loss": 2.422, + "step": 8367 + }, + { + "epoch": 0.8849407783417935, + "grad_norm": 7.9375, + "learning_rate": 4.722060084895791e-06, + "loss": 2.3712, + "step": 8368 + }, + { + "epoch": 0.8850465313028765, + "grad_norm": 8.0625, + "learning_rate": 4.721981601747203e-06, + "loss": 2.7032, + "step": 8369 + }, + { + "epoch": 0.8851522842639594, + "grad_norm": 7.65625, + "learning_rate": 4.721903108171765e-06, + "loss": 2.4391, + "step": 8370 + }, + { + "epoch": 0.8852580372250423, + "grad_norm": 8.5625, + "learning_rate": 4.721824604169844e-06, + "loss": 2.4637, + "step": 8371 + }, + { + "epoch": 0.8853637901861252, + "grad_norm": 8.375, + "learning_rate": 4.7217460897418076e-06, + "loss": 2.5254, + "step": 8372 + }, + { + "epoch": 0.8854695431472082, + "grad_norm": 8.625, + "learning_rate": 4.7216675648880265e-06, + "loss": 2.5457, + "step": 8373 + }, + { + "epoch": 0.885575296108291, + "grad_norm": 7.5625, + "learning_rate": 4.721589029608868e-06, + "loss": 2.4714, + "step": 8374 + }, + { + "epoch": 0.885681049069374, + "grad_norm": 8.375, + "learning_rate": 4.721510483904701e-06, + "loss": 2.6119, + "step": 8375 + }, + { + "epoch": 0.8857868020304569, + "grad_norm": 8.5, + "learning_rate": 4.721431927775894e-06, + "loss": 2.786, + "step": 8376 + }, + { + "epoch": 0.8858925549915397, + "grad_norm": 8.75, + "learning_rate": 4.7213533612228145e-06, + "loss": 2.6307, + "step": 8377 + }, + { + "epoch": 0.8859983079526227, + "grad_norm": 8.3125, + "learning_rate": 4.721274784245833e-06, + "loss": 2.6697, + "step": 8378 + }, + { + "epoch": 0.8861040609137056, + "grad_norm": 8.1875, + "learning_rate": 4.721196196845317e-06, + "loss": 2.5754, + "step": 8379 + }, + { + "epoch": 0.8862098138747885, + "grad_norm": 8.125, + "learning_rate": 4.721117599021636e-06, + "loss": 2.4725, + "step": 8380 + }, + { + "epoch": 0.8863155668358714, + "grad_norm": 8.1875, + "learning_rate": 4.721038990775158e-06, + "loss": 2.5768, + "step": 8381 + }, + { + "epoch": 0.8864213197969543, + "grad_norm": 8.0, + "learning_rate": 4.7209603721062535e-06, + "loss": 2.7192, + "step": 8382 + }, + { + "epoch": 0.8865270727580372, + "grad_norm": 8.0625, + "learning_rate": 4.72088174301529e-06, + "loss": 2.6316, + "step": 8383 + }, + { + "epoch": 0.8866328257191202, + "grad_norm": 8.125, + "learning_rate": 4.720803103502636e-06, + "loss": 2.7877, + "step": 8384 + }, + { + "epoch": 0.886738578680203, + "grad_norm": 7.9375, + "learning_rate": 4.7207244535686615e-06, + "loss": 2.5043, + "step": 8385 + }, + { + "epoch": 0.886844331641286, + "grad_norm": 8.0, + "learning_rate": 4.720645793213736e-06, + "loss": 2.5537, + "step": 8386 + }, + { + "epoch": 0.8869500846023689, + "grad_norm": 8.0625, + "learning_rate": 4.720567122438228e-06, + "loss": 2.5092, + "step": 8387 + }, + { + "epoch": 0.8870558375634517, + "grad_norm": 7.6875, + "learning_rate": 4.720488441242505e-06, + "loss": 2.7079, + "step": 8388 + }, + { + "epoch": 0.8871615905245347, + "grad_norm": 8.1875, + "learning_rate": 4.720409749626939e-06, + "loss": 2.7081, + "step": 8389 + }, + { + "epoch": 0.8872673434856176, + "grad_norm": 8.0625, + "learning_rate": 4.7203310475919e-06, + "loss": 2.4711, + "step": 8390 + }, + { + "epoch": 0.8873730964467005, + "grad_norm": 7.59375, + "learning_rate": 4.720252335137753e-06, + "loss": 2.5287, + "step": 8391 + }, + { + "epoch": 0.8874788494077834, + "grad_norm": 8.125, + "learning_rate": 4.720173612264871e-06, + "loss": 2.5097, + "step": 8392 + }, + { + "epoch": 0.8875846023688664, + "grad_norm": 7.96875, + "learning_rate": 4.720094878973621e-06, + "loss": 2.538, + "step": 8393 + }, + { + "epoch": 0.8876903553299492, + "grad_norm": 7.9375, + "learning_rate": 4.720016135264375e-06, + "loss": 2.525, + "step": 8394 + }, + { + "epoch": 0.8877961082910322, + "grad_norm": 8.5, + "learning_rate": 4.719937381137502e-06, + "loss": 2.7441, + "step": 8395 + }, + { + "epoch": 0.887901861252115, + "grad_norm": 7.625, + "learning_rate": 4.719858616593369e-06, + "loss": 2.5853, + "step": 8396 + }, + { + "epoch": 0.888007614213198, + "grad_norm": 8.0, + "learning_rate": 4.719779841632348e-06, + "loss": 2.7661, + "step": 8397 + }, + { + "epoch": 0.8881133671742809, + "grad_norm": 7.875, + "learning_rate": 4.719701056254808e-06, + "loss": 2.5865, + "step": 8398 + }, + { + "epoch": 0.8882191201353637, + "grad_norm": 8.4375, + "learning_rate": 4.719622260461119e-06, + "loss": 2.587, + "step": 8399 + }, + { + "epoch": 0.8883248730964467, + "grad_norm": 8.125, + "learning_rate": 4.719543454251651e-06, + "loss": 2.5712, + "step": 8400 + }, + { + "epoch": 0.8884306260575296, + "grad_norm": 8.125, + "learning_rate": 4.719464637626773e-06, + "loss": 2.5337, + "step": 8401 + }, + { + "epoch": 0.8885363790186125, + "grad_norm": 8.1875, + "learning_rate": 4.719385810586854e-06, + "loss": 2.499, + "step": 8402 + }, + { + "epoch": 0.8886421319796954, + "grad_norm": 7.90625, + "learning_rate": 4.719306973132266e-06, + "loss": 2.5389, + "step": 8403 + }, + { + "epoch": 0.8887478849407784, + "grad_norm": 8.125, + "learning_rate": 4.719228125263379e-06, + "loss": 2.517, + "step": 8404 + }, + { + "epoch": 0.8888536379018612, + "grad_norm": 7.65625, + "learning_rate": 4.71914926698056e-06, + "loss": 2.4823, + "step": 8405 + }, + { + "epoch": 0.8889593908629442, + "grad_norm": 8.0625, + "learning_rate": 4.719070398284183e-06, + "loss": 2.46, + "step": 8406 + }, + { + "epoch": 0.8890651438240271, + "grad_norm": 7.5, + "learning_rate": 4.718991519174615e-06, + "loss": 2.4929, + "step": 8407 + }, + { + "epoch": 0.88917089678511, + "grad_norm": 8.25, + "learning_rate": 4.718912629652228e-06, + "loss": 2.6243, + "step": 8408 + }, + { + "epoch": 0.8892766497461929, + "grad_norm": 7.96875, + "learning_rate": 4.7188337297173915e-06, + "loss": 2.7488, + "step": 8409 + }, + { + "epoch": 0.8893824027072758, + "grad_norm": 8.0625, + "learning_rate": 4.718754819370475e-06, + "loss": 2.7872, + "step": 8410 + }, + { + "epoch": 0.8894881556683587, + "grad_norm": 7.65625, + "learning_rate": 4.718675898611851e-06, + "loss": 2.5635, + "step": 8411 + }, + { + "epoch": 0.8895939086294417, + "grad_norm": 7.8125, + "learning_rate": 4.7185969674418875e-06, + "loss": 2.2868, + "step": 8412 + }, + { + "epoch": 0.8896996615905245, + "grad_norm": 7.90625, + "learning_rate": 4.718518025860955e-06, + "loss": 2.5733, + "step": 8413 + }, + { + "epoch": 0.8898054145516074, + "grad_norm": 7.875, + "learning_rate": 4.718439073869426e-06, + "loss": 2.545, + "step": 8414 + }, + { + "epoch": 0.8899111675126904, + "grad_norm": 8.5, + "learning_rate": 4.718360111467669e-06, + "loss": 2.8836, + "step": 8415 + }, + { + "epoch": 0.8900169204737732, + "grad_norm": 8.1875, + "learning_rate": 4.718281138656056e-06, + "loss": 2.3911, + "step": 8416 + }, + { + "epoch": 0.8901226734348562, + "grad_norm": 9.4375, + "learning_rate": 4.718202155434957e-06, + "loss": 2.828, + "step": 8417 + }, + { + "epoch": 0.8902284263959391, + "grad_norm": 8.625, + "learning_rate": 4.718123161804742e-06, + "loss": 2.7978, + "step": 8418 + }, + { + "epoch": 0.890334179357022, + "grad_norm": 8.0625, + "learning_rate": 4.718044157765782e-06, + "loss": 2.5861, + "step": 8419 + }, + { + "epoch": 0.8904399323181049, + "grad_norm": 7.90625, + "learning_rate": 4.717965143318448e-06, + "loss": 2.595, + "step": 8420 + }, + { + "epoch": 0.8905456852791879, + "grad_norm": 8.0, + "learning_rate": 4.717886118463111e-06, + "loss": 2.7046, + "step": 8421 + }, + { + "epoch": 0.8906514382402707, + "grad_norm": 8.3125, + "learning_rate": 4.717807083200141e-06, + "loss": 2.7743, + "step": 8422 + }, + { + "epoch": 0.8907571912013537, + "grad_norm": 8.25, + "learning_rate": 4.7177280375299094e-06, + "loss": 2.6016, + "step": 8423 + }, + { + "epoch": 0.8908629441624365, + "grad_norm": 7.90625, + "learning_rate": 4.7176489814527875e-06, + "loss": 2.7192, + "step": 8424 + }, + { + "epoch": 0.8909686971235194, + "grad_norm": 8.0, + "learning_rate": 4.717569914969146e-06, + "loss": 2.9067, + "step": 8425 + }, + { + "epoch": 0.8910744500846024, + "grad_norm": 7.875, + "learning_rate": 4.717490838079356e-06, + "loss": 2.566, + "step": 8426 + }, + { + "epoch": 0.8911802030456852, + "grad_norm": 8.3125, + "learning_rate": 4.717411750783788e-06, + "loss": 2.7388, + "step": 8427 + }, + { + "epoch": 0.8912859560067682, + "grad_norm": 8.0625, + "learning_rate": 4.717332653082813e-06, + "loss": 2.559, + "step": 8428 + }, + { + "epoch": 0.8913917089678511, + "grad_norm": 8.25, + "learning_rate": 4.717253544976804e-06, + "loss": 2.3832, + "step": 8429 + }, + { + "epoch": 0.891497461928934, + "grad_norm": 8.125, + "learning_rate": 4.71717442646613e-06, + "loss": 2.5559, + "step": 8430 + }, + { + "epoch": 0.8916032148900169, + "grad_norm": 7.71875, + "learning_rate": 4.717095297551164e-06, + "loss": 2.4314, + "step": 8431 + }, + { + "epoch": 0.8917089678510999, + "grad_norm": 7.9375, + "learning_rate": 4.717016158232276e-06, + "loss": 2.6424, + "step": 8432 + }, + { + "epoch": 0.8918147208121827, + "grad_norm": 7.34375, + "learning_rate": 4.716937008509838e-06, + "loss": 2.523, + "step": 8433 + }, + { + "epoch": 0.8919204737732657, + "grad_norm": 7.8125, + "learning_rate": 4.716857848384221e-06, + "loss": 2.6056, + "step": 8434 + }, + { + "epoch": 0.8920262267343486, + "grad_norm": 7.625, + "learning_rate": 4.716778677855797e-06, + "loss": 2.3649, + "step": 8435 + }, + { + "epoch": 0.8921319796954315, + "grad_norm": 8.0625, + "learning_rate": 4.716699496924937e-06, + "loss": 2.7789, + "step": 8436 + }, + { + "epoch": 0.8922377326565144, + "grad_norm": 8.125, + "learning_rate": 4.7166203055920135e-06, + "loss": 2.7083, + "step": 8437 + }, + { + "epoch": 0.8923434856175972, + "grad_norm": 7.875, + "learning_rate": 4.7165411038573974e-06, + "loss": 2.3474, + "step": 8438 + }, + { + "epoch": 0.8924492385786802, + "grad_norm": 8.8125, + "learning_rate": 4.71646189172146e-06, + "loss": 2.5165, + "step": 8439 + }, + { + "epoch": 0.8925549915397631, + "grad_norm": 7.84375, + "learning_rate": 4.7163826691845735e-06, + "loss": 2.5103, + "step": 8440 + }, + { + "epoch": 0.892660744500846, + "grad_norm": 8.1875, + "learning_rate": 4.71630343624711e-06, + "loss": 2.5818, + "step": 8441 + }, + { + "epoch": 0.8927664974619289, + "grad_norm": 8.25, + "learning_rate": 4.716224192909441e-06, + "loss": 2.5191, + "step": 8442 + }, + { + "epoch": 0.8928722504230119, + "grad_norm": 7.78125, + "learning_rate": 4.7161449391719385e-06, + "loss": 2.6911, + "step": 8443 + }, + { + "epoch": 0.8929780033840947, + "grad_norm": 7.8125, + "learning_rate": 4.716065675034973e-06, + "loss": 2.5332, + "step": 8444 + }, + { + "epoch": 0.8930837563451777, + "grad_norm": 8.3125, + "learning_rate": 4.7159864004989185e-06, + "loss": 2.7147, + "step": 8445 + }, + { + "epoch": 0.8931895093062606, + "grad_norm": 7.53125, + "learning_rate": 4.715907115564146e-06, + "loss": 2.603, + "step": 8446 + }, + { + "epoch": 0.8932952622673435, + "grad_norm": 8.0, + "learning_rate": 4.715827820231027e-06, + "loss": 2.4384, + "step": 8447 + }, + { + "epoch": 0.8934010152284264, + "grad_norm": 8.3125, + "learning_rate": 4.715748514499935e-06, + "loss": 2.237, + "step": 8448 + }, + { + "epoch": 0.8935067681895094, + "grad_norm": 7.625, + "learning_rate": 4.715669198371241e-06, + "loss": 2.6779, + "step": 8449 + }, + { + "epoch": 0.8936125211505922, + "grad_norm": 8.125, + "learning_rate": 4.715589871845319e-06, + "loss": 2.3079, + "step": 8450 + }, + { + "epoch": 0.8937182741116751, + "grad_norm": 8.375, + "learning_rate": 4.715510534922539e-06, + "loss": 2.5901, + "step": 8451 + }, + { + "epoch": 0.893824027072758, + "grad_norm": 7.8125, + "learning_rate": 4.715431187603273e-06, + "loss": 2.5447, + "step": 8452 + }, + { + "epoch": 0.8939297800338409, + "grad_norm": 7.65625, + "learning_rate": 4.715351829887895e-06, + "loss": 2.4926, + "step": 8453 + }, + { + "epoch": 0.8940355329949239, + "grad_norm": 8.0625, + "learning_rate": 4.715272461776778e-06, + "loss": 2.7503, + "step": 8454 + }, + { + "epoch": 0.8941412859560067, + "grad_norm": 8.125, + "learning_rate": 4.715193083270293e-06, + "loss": 2.548, + "step": 8455 + }, + { + "epoch": 0.8942470389170897, + "grad_norm": 8.1875, + "learning_rate": 4.7151136943688124e-06, + "loss": 2.5631, + "step": 8456 + }, + { + "epoch": 0.8943527918781726, + "grad_norm": 8.0625, + "learning_rate": 4.715034295072709e-06, + "loss": 2.6553, + "step": 8457 + }, + { + "epoch": 0.8944585448392555, + "grad_norm": 7.9375, + "learning_rate": 4.714954885382356e-06, + "loss": 2.6355, + "step": 8458 + }, + { + "epoch": 0.8945642978003384, + "grad_norm": 8.4375, + "learning_rate": 4.714875465298126e-06, + "loss": 2.6839, + "step": 8459 + }, + { + "epoch": 0.8946700507614214, + "grad_norm": 8.6875, + "learning_rate": 4.71479603482039e-06, + "loss": 2.6224, + "step": 8460 + }, + { + "epoch": 0.8947758037225042, + "grad_norm": 8.0, + "learning_rate": 4.714716593949523e-06, + "loss": 2.6616, + "step": 8461 + }, + { + "epoch": 0.8948815566835872, + "grad_norm": 7.96875, + "learning_rate": 4.714637142685898e-06, + "loss": 2.5493, + "step": 8462 + }, + { + "epoch": 0.8949873096446701, + "grad_norm": 8.5, + "learning_rate": 4.714557681029884e-06, + "loss": 2.5343, + "step": 8463 + }, + { + "epoch": 0.8950930626057529, + "grad_norm": 7.78125, + "learning_rate": 4.714478208981859e-06, + "loss": 2.4851, + "step": 8464 + }, + { + "epoch": 0.8951988155668359, + "grad_norm": 7.90625, + "learning_rate": 4.714398726542192e-06, + "loss": 2.5442, + "step": 8465 + }, + { + "epoch": 0.8953045685279187, + "grad_norm": 7.78125, + "learning_rate": 4.7143192337112575e-06, + "loss": 2.5347, + "step": 8466 + }, + { + "epoch": 0.8954103214890017, + "grad_norm": 7.96875, + "learning_rate": 4.714239730489428e-06, + "loss": 2.2848, + "step": 8467 + }, + { + "epoch": 0.8955160744500846, + "grad_norm": 7.8125, + "learning_rate": 4.714160216877078e-06, + "loss": 2.6033, + "step": 8468 + }, + { + "epoch": 0.8956218274111675, + "grad_norm": 8.625, + "learning_rate": 4.7140806928745795e-06, + "loss": 2.6259, + "step": 8469 + }, + { + "epoch": 0.8957275803722504, + "grad_norm": 7.78125, + "learning_rate": 4.714001158482307e-06, + "loss": 2.677, + "step": 8470 + }, + { + "epoch": 0.8958333333333334, + "grad_norm": 7.8125, + "learning_rate": 4.71392161370063e-06, + "loss": 2.6617, + "step": 8471 + }, + { + "epoch": 0.8959390862944162, + "grad_norm": 8.5, + "learning_rate": 4.713842058529926e-06, + "loss": 2.9009, + "step": 8472 + }, + { + "epoch": 0.8960448392554992, + "grad_norm": 10.1875, + "learning_rate": 4.713762492970566e-06, + "loss": 2.5919, + "step": 8473 + }, + { + "epoch": 0.8961505922165821, + "grad_norm": 8.875, + "learning_rate": 4.713682917022925e-06, + "loss": 2.4247, + "step": 8474 + }, + { + "epoch": 0.896256345177665, + "grad_norm": 7.875, + "learning_rate": 4.713603330687374e-06, + "loss": 2.7122, + "step": 8475 + }, + { + "epoch": 0.8963620981387479, + "grad_norm": 7.71875, + "learning_rate": 4.713523733964289e-06, + "loss": 2.3877, + "step": 8476 + }, + { + "epoch": 0.8964678510998308, + "grad_norm": 8.4375, + "learning_rate": 4.713444126854042e-06, + "loss": 2.5212, + "step": 8477 + }, + { + "epoch": 0.8965736040609137, + "grad_norm": 7.5625, + "learning_rate": 4.713364509357007e-06, + "loss": 2.5661, + "step": 8478 + }, + { + "epoch": 0.8966793570219966, + "grad_norm": 8.0, + "learning_rate": 4.7132848814735574e-06, + "loss": 2.7453, + "step": 8479 + }, + { + "epoch": 0.8967851099830795, + "grad_norm": 8.0625, + "learning_rate": 4.713205243204068e-06, + "loss": 2.7096, + "step": 8480 + }, + { + "epoch": 0.8968908629441624, + "grad_norm": 8.125, + "learning_rate": 4.7131255945489105e-06, + "loss": 2.3203, + "step": 8481 + }, + { + "epoch": 0.8969966159052454, + "grad_norm": 8.3125, + "learning_rate": 4.71304593550846e-06, + "loss": 2.8081, + "step": 8482 + }, + { + "epoch": 0.8971023688663282, + "grad_norm": 8.8125, + "learning_rate": 4.7129662660830895e-06, + "loss": 2.6412, + "step": 8483 + }, + { + "epoch": 0.8972081218274112, + "grad_norm": 8.3125, + "learning_rate": 4.7128865862731745e-06, + "loss": 2.739, + "step": 8484 + }, + { + "epoch": 0.8973138747884941, + "grad_norm": 8.25, + "learning_rate": 4.712806896079087e-06, + "loss": 2.6768, + "step": 8485 + }, + { + "epoch": 0.897419627749577, + "grad_norm": 7.65625, + "learning_rate": 4.712727195501202e-06, + "loss": 2.535, + "step": 8486 + }, + { + "epoch": 0.8975253807106599, + "grad_norm": 8.6875, + "learning_rate": 4.712647484539894e-06, + "loss": 2.7744, + "step": 8487 + }, + { + "epoch": 0.8976311336717429, + "grad_norm": 7.625, + "learning_rate": 4.7125677631955355e-06, + "loss": 2.4331, + "step": 8488 + }, + { + "epoch": 0.8977368866328257, + "grad_norm": 8.375, + "learning_rate": 4.712488031468501e-06, + "loss": 2.457, + "step": 8489 + }, + { + "epoch": 0.8978426395939086, + "grad_norm": 7.875, + "learning_rate": 4.712408289359166e-06, + "loss": 2.4693, + "step": 8490 + }, + { + "epoch": 0.8979483925549916, + "grad_norm": 7.8125, + "learning_rate": 4.712328536867903e-06, + "loss": 2.7312, + "step": 8491 + }, + { + "epoch": 0.8980541455160744, + "grad_norm": 8.4375, + "learning_rate": 4.712248773995088e-06, + "loss": 2.6301, + "step": 8492 + }, + { + "epoch": 0.8981598984771574, + "grad_norm": 7.875, + "learning_rate": 4.712169000741093e-06, + "loss": 2.6647, + "step": 8493 + }, + { + "epoch": 0.8982656514382402, + "grad_norm": 7.84375, + "learning_rate": 4.712089217106294e-06, + "loss": 2.702, + "step": 8494 + }, + { + "epoch": 0.8983714043993232, + "grad_norm": 7.8125, + "learning_rate": 4.712009423091065e-06, + "loss": 2.6304, + "step": 8495 + }, + { + "epoch": 0.8984771573604061, + "grad_norm": 8.1875, + "learning_rate": 4.711929618695781e-06, + "loss": 2.6881, + "step": 8496 + }, + { + "epoch": 0.898582910321489, + "grad_norm": 7.78125, + "learning_rate": 4.711849803920815e-06, + "loss": 2.6567, + "step": 8497 + }, + { + "epoch": 0.8986886632825719, + "grad_norm": 7.9375, + "learning_rate": 4.711769978766543e-06, + "loss": 2.5451, + "step": 8498 + }, + { + "epoch": 0.8987944162436549, + "grad_norm": 7.75, + "learning_rate": 4.7116901432333394e-06, + "loss": 2.4298, + "step": 8499 + }, + { + "epoch": 0.8989001692047377, + "grad_norm": 8.3125, + "learning_rate": 4.711610297321578e-06, + "loss": 2.3786, + "step": 8500 + }, + { + "epoch": 0.8990059221658206, + "grad_norm": 7.75, + "learning_rate": 4.7115304410316345e-06, + "loss": 2.5494, + "step": 8501 + }, + { + "epoch": 0.8991116751269036, + "grad_norm": 8.0625, + "learning_rate": 4.711450574363883e-06, + "loss": 2.5846, + "step": 8502 + }, + { + "epoch": 0.8992174280879864, + "grad_norm": 7.71875, + "learning_rate": 4.711370697318698e-06, + "loss": 2.5319, + "step": 8503 + }, + { + "epoch": 0.8993231810490694, + "grad_norm": 7.34375, + "learning_rate": 4.711290809896455e-06, + "loss": 2.3943, + "step": 8504 + }, + { + "epoch": 0.8994289340101523, + "grad_norm": 8.375, + "learning_rate": 4.711210912097529e-06, + "loss": 2.5072, + "step": 8505 + }, + { + "epoch": 0.8995346869712352, + "grad_norm": 8.1875, + "learning_rate": 4.711131003922294e-06, + "loss": 2.7468, + "step": 8506 + }, + { + "epoch": 0.8996404399323181, + "grad_norm": 7.59375, + "learning_rate": 4.711051085371126e-06, + "loss": 2.5136, + "step": 8507 + }, + { + "epoch": 0.899746192893401, + "grad_norm": 9.5, + "learning_rate": 4.710971156444399e-06, + "loss": 2.594, + "step": 8508 + }, + { + "epoch": 0.8998519458544839, + "grad_norm": 8.5, + "learning_rate": 4.710891217142489e-06, + "loss": 2.5015, + "step": 8509 + }, + { + "epoch": 0.8999576988155669, + "grad_norm": 8.1875, + "learning_rate": 4.71081126746577e-06, + "loss": 2.6493, + "step": 8510 + }, + { + "epoch": 0.9000634517766497, + "grad_norm": 7.875, + "learning_rate": 4.710731307414619e-06, + "loss": 2.7348, + "step": 8511 + }, + { + "epoch": 0.9001692047377327, + "grad_norm": 7.96875, + "learning_rate": 4.710651336989409e-06, + "loss": 2.688, + "step": 8512 + }, + { + "epoch": 0.9002749576988156, + "grad_norm": 8.375, + "learning_rate": 4.710571356190517e-06, + "loss": 2.6292, + "step": 8513 + }, + { + "epoch": 0.9003807106598984, + "grad_norm": 8.0, + "learning_rate": 4.710491365018318e-06, + "loss": 2.5042, + "step": 8514 + }, + { + "epoch": 0.9004864636209814, + "grad_norm": 8.1875, + "learning_rate": 4.710411363473187e-06, + "loss": 2.7462, + "step": 8515 + }, + { + "epoch": 0.9005922165820643, + "grad_norm": 7.71875, + "learning_rate": 4.710331351555499e-06, + "loss": 2.6364, + "step": 8516 + }, + { + "epoch": 0.9006979695431472, + "grad_norm": 8.75, + "learning_rate": 4.710251329265631e-06, + "loss": 2.8396, + "step": 8517 + }, + { + "epoch": 0.9008037225042301, + "grad_norm": 7.59375, + "learning_rate": 4.710171296603957e-06, + "loss": 2.6118, + "step": 8518 + }, + { + "epoch": 0.9009094754653131, + "grad_norm": 7.40625, + "learning_rate": 4.710091253570852e-06, + "loss": 2.3704, + "step": 8519 + }, + { + "epoch": 0.9010152284263959, + "grad_norm": 7.9375, + "learning_rate": 4.710011200166693e-06, + "loss": 2.3908, + "step": 8520 + }, + { + "epoch": 0.9011209813874789, + "grad_norm": 8.0, + "learning_rate": 4.709931136391857e-06, + "loss": 2.6968, + "step": 8521 + }, + { + "epoch": 0.9012267343485617, + "grad_norm": 8.125, + "learning_rate": 4.709851062246716e-06, + "loss": 2.4923, + "step": 8522 + }, + { + "epoch": 0.9013324873096447, + "grad_norm": 8.25, + "learning_rate": 4.709770977731649e-06, + "loss": 2.6724, + "step": 8523 + }, + { + "epoch": 0.9014382402707276, + "grad_norm": 7.78125, + "learning_rate": 4.70969088284703e-06, + "loss": 2.2014, + "step": 8524 + }, + { + "epoch": 0.9015439932318104, + "grad_norm": 8.1875, + "learning_rate": 4.709610777593235e-06, + "loss": 2.5873, + "step": 8525 + }, + { + "epoch": 0.9016497461928934, + "grad_norm": 8.25, + "learning_rate": 4.709530661970641e-06, + "loss": 2.4935, + "step": 8526 + }, + { + "epoch": 0.9017554991539763, + "grad_norm": 8.3125, + "learning_rate": 4.709450535979623e-06, + "loss": 2.3859, + "step": 8527 + }, + { + "epoch": 0.9018612521150592, + "grad_norm": 8.0625, + "learning_rate": 4.7093703996205576e-06, + "loss": 2.5981, + "step": 8528 + }, + { + "epoch": 0.9019670050761421, + "grad_norm": 8.75, + "learning_rate": 4.70929025289382e-06, + "loss": 2.5669, + "step": 8529 + }, + { + "epoch": 0.9020727580372251, + "grad_norm": 7.46875, + "learning_rate": 4.709210095799787e-06, + "loss": 2.6073, + "step": 8530 + }, + { + "epoch": 0.9021785109983079, + "grad_norm": 7.96875, + "learning_rate": 4.709129928338835e-06, + "loss": 2.6904, + "step": 8531 + }, + { + "epoch": 0.9022842639593909, + "grad_norm": 8.375, + "learning_rate": 4.70904975051134e-06, + "loss": 2.602, + "step": 8532 + }, + { + "epoch": 0.9023900169204738, + "grad_norm": 8.125, + "learning_rate": 4.708969562317677e-06, + "loss": 2.7692, + "step": 8533 + }, + { + "epoch": 0.9024957698815567, + "grad_norm": 7.46875, + "learning_rate": 4.708889363758224e-06, + "loss": 2.3775, + "step": 8534 + }, + { + "epoch": 0.9026015228426396, + "grad_norm": 8.5, + "learning_rate": 4.7088091548333556e-06, + "loss": 2.5786, + "step": 8535 + }, + { + "epoch": 0.9027072758037225, + "grad_norm": 7.90625, + "learning_rate": 4.70872893554345e-06, + "loss": 2.66, + "step": 8536 + }, + { + "epoch": 0.9028130287648054, + "grad_norm": 8.25, + "learning_rate": 4.708648705888883e-06, + "loss": 2.8353, + "step": 8537 + }, + { + "epoch": 0.9029187817258884, + "grad_norm": 7.71875, + "learning_rate": 4.70856846587003e-06, + "loss": 2.3493, + "step": 8538 + }, + { + "epoch": 0.9030245346869712, + "grad_norm": 8.125, + "learning_rate": 4.708488215487269e-06, + "loss": 2.3835, + "step": 8539 + }, + { + "epoch": 0.9031302876480541, + "grad_norm": 8.1875, + "learning_rate": 4.708407954740976e-06, + "loss": 2.4501, + "step": 8540 + }, + { + "epoch": 0.9032360406091371, + "grad_norm": 8.25, + "learning_rate": 4.708327683631528e-06, + "loss": 2.3292, + "step": 8541 + }, + { + "epoch": 0.9033417935702199, + "grad_norm": 7.59375, + "learning_rate": 4.708247402159301e-06, + "loss": 2.6043, + "step": 8542 + }, + { + "epoch": 0.9034475465313029, + "grad_norm": 7.78125, + "learning_rate": 4.708167110324673e-06, + "loss": 2.5222, + "step": 8543 + }, + { + "epoch": 0.9035532994923858, + "grad_norm": 9.25, + "learning_rate": 4.708086808128019e-06, + "loss": 2.6538, + "step": 8544 + }, + { + "epoch": 0.9036590524534687, + "grad_norm": 7.84375, + "learning_rate": 4.708006495569717e-06, + "loss": 2.6139, + "step": 8545 + }, + { + "epoch": 0.9037648054145516, + "grad_norm": 8.125, + "learning_rate": 4.7079261726501434e-06, + "loss": 2.6878, + "step": 8546 + }, + { + "epoch": 0.9038705583756346, + "grad_norm": 7.75, + "learning_rate": 4.707845839369675e-06, + "loss": 2.5269, + "step": 8547 + }, + { + "epoch": 0.9039763113367174, + "grad_norm": 7.6875, + "learning_rate": 4.70776549572869e-06, + "loss": 2.533, + "step": 8548 + }, + { + "epoch": 0.9040820642978004, + "grad_norm": 7.625, + "learning_rate": 4.707685141727564e-06, + "loss": 2.3532, + "step": 8549 + }, + { + "epoch": 0.9041878172588832, + "grad_norm": 9.1875, + "learning_rate": 4.707604777366674e-06, + "loss": 2.6664, + "step": 8550 + }, + { + "epoch": 0.9042935702199661, + "grad_norm": 8.375, + "learning_rate": 4.7075244026463976e-06, + "loss": 2.504, + "step": 8551 + }, + { + "epoch": 0.9043993231810491, + "grad_norm": 8.0625, + "learning_rate": 4.707444017567113e-06, + "loss": 2.7761, + "step": 8552 + }, + { + "epoch": 0.9045050761421319, + "grad_norm": 9.1875, + "learning_rate": 4.707363622129196e-06, + "loss": 2.8966, + "step": 8553 + }, + { + "epoch": 0.9046108291032149, + "grad_norm": 8.25, + "learning_rate": 4.707283216333024e-06, + "loss": 2.5809, + "step": 8554 + }, + { + "epoch": 0.9047165820642978, + "grad_norm": 7.65625, + "learning_rate": 4.707202800178975e-06, + "loss": 2.5605, + "step": 8555 + }, + { + "epoch": 0.9048223350253807, + "grad_norm": 8.0625, + "learning_rate": 4.7071223736674255e-06, + "loss": 2.7074, + "step": 8556 + }, + { + "epoch": 0.9049280879864636, + "grad_norm": 8.5, + "learning_rate": 4.707041936798753e-06, + "loss": 2.7674, + "step": 8557 + }, + { + "epoch": 0.9050338409475466, + "grad_norm": 7.96875, + "learning_rate": 4.706961489573336e-06, + "loss": 2.621, + "step": 8558 + }, + { + "epoch": 0.9051395939086294, + "grad_norm": 7.65625, + "learning_rate": 4.706881031991552e-06, + "loss": 2.4612, + "step": 8559 + }, + { + "epoch": 0.9052453468697124, + "grad_norm": 8.25, + "learning_rate": 4.706800564053778e-06, + "loss": 2.4786, + "step": 8560 + }, + { + "epoch": 0.9053510998307953, + "grad_norm": 7.75, + "learning_rate": 4.70672008576039e-06, + "loss": 2.6849, + "step": 8561 + }, + { + "epoch": 0.9054568527918782, + "grad_norm": 8.5, + "learning_rate": 4.706639597111768e-06, + "loss": 2.7471, + "step": 8562 + }, + { + "epoch": 0.9055626057529611, + "grad_norm": 8.875, + "learning_rate": 4.7065590981082885e-06, + "loss": 2.697, + "step": 8563 + }, + { + "epoch": 0.9056683587140439, + "grad_norm": 7.71875, + "learning_rate": 4.706478588750329e-06, + "loss": 2.4726, + "step": 8564 + }, + { + "epoch": 0.9057741116751269, + "grad_norm": 7.71875, + "learning_rate": 4.706398069038269e-06, + "loss": 2.6333, + "step": 8565 + }, + { + "epoch": 0.9058798646362098, + "grad_norm": 8.1875, + "learning_rate": 4.706317538972486e-06, + "loss": 2.6589, + "step": 8566 + }, + { + "epoch": 0.9059856175972927, + "grad_norm": 8.3125, + "learning_rate": 4.706236998553356e-06, + "loss": 2.4659, + "step": 8567 + }, + { + "epoch": 0.9060913705583756, + "grad_norm": 9.375, + "learning_rate": 4.706156447781258e-06, + "loss": 2.8952, + "step": 8568 + }, + { + "epoch": 0.9061971235194586, + "grad_norm": 7.75, + "learning_rate": 4.7060758866565705e-06, + "loss": 2.6861, + "step": 8569 + }, + { + "epoch": 0.9063028764805414, + "grad_norm": 7.5625, + "learning_rate": 4.705995315179671e-06, + "loss": 2.6476, + "step": 8570 + }, + { + "epoch": 0.9064086294416244, + "grad_norm": 8.6875, + "learning_rate": 4.705914733350938e-06, + "loss": 2.5933, + "step": 8571 + }, + { + "epoch": 0.9065143824027073, + "grad_norm": 7.6875, + "learning_rate": 4.705834141170748e-06, + "loss": 2.504, + "step": 8572 + }, + { + "epoch": 0.9066201353637902, + "grad_norm": 7.6875, + "learning_rate": 4.705753538639482e-06, + "loss": 2.5534, + "step": 8573 + }, + { + "epoch": 0.9067258883248731, + "grad_norm": 7.71875, + "learning_rate": 4.7056729257575164e-06, + "loss": 2.4764, + "step": 8574 + }, + { + "epoch": 0.9068316412859561, + "grad_norm": 8.4375, + "learning_rate": 4.70559230252523e-06, + "loss": 2.7679, + "step": 8575 + }, + { + "epoch": 0.9069373942470389, + "grad_norm": 7.5625, + "learning_rate": 4.705511668943001e-06, + "loss": 2.5427, + "step": 8576 + }, + { + "epoch": 0.9070431472081218, + "grad_norm": 8.375, + "learning_rate": 4.705431025011206e-06, + "loss": 2.4865, + "step": 8577 + }, + { + "epoch": 0.9071489001692047, + "grad_norm": 11.5, + "learning_rate": 4.705350370730227e-06, + "loss": 2.5797, + "step": 8578 + }, + { + "epoch": 0.9072546531302876, + "grad_norm": 8.125, + "learning_rate": 4.705269706100441e-06, + "loss": 2.5493, + "step": 8579 + }, + { + "epoch": 0.9073604060913706, + "grad_norm": 7.71875, + "learning_rate": 4.705189031122225e-06, + "loss": 2.6612, + "step": 8580 + }, + { + "epoch": 0.9074661590524534, + "grad_norm": 7.875, + "learning_rate": 4.7051083457959596e-06, + "loss": 2.5579, + "step": 8581 + }, + { + "epoch": 0.9075719120135364, + "grad_norm": 8.0625, + "learning_rate": 4.705027650122022e-06, + "loss": 2.6969, + "step": 8582 + }, + { + "epoch": 0.9076776649746193, + "grad_norm": 8.375, + "learning_rate": 4.704946944100792e-06, + "loss": 2.6381, + "step": 8583 + }, + { + "epoch": 0.9077834179357022, + "grad_norm": 8.0625, + "learning_rate": 4.704866227732647e-06, + "loss": 2.6888, + "step": 8584 + }, + { + "epoch": 0.9078891708967851, + "grad_norm": 8.0, + "learning_rate": 4.704785501017967e-06, + "loss": 2.3936, + "step": 8585 + }, + { + "epoch": 0.9079949238578681, + "grad_norm": 7.90625, + "learning_rate": 4.704704763957131e-06, + "loss": 2.5033, + "step": 8586 + }, + { + "epoch": 0.9081006768189509, + "grad_norm": 8.4375, + "learning_rate": 4.704624016550516e-06, + "loss": 2.5409, + "step": 8587 + }, + { + "epoch": 0.9082064297800339, + "grad_norm": 7.9375, + "learning_rate": 4.7045432587985035e-06, + "loss": 2.6279, + "step": 8588 + }, + { + "epoch": 0.9083121827411168, + "grad_norm": 7.9375, + "learning_rate": 4.70446249070147e-06, + "loss": 2.5963, + "step": 8589 + }, + { + "epoch": 0.9084179357021996, + "grad_norm": 7.6875, + "learning_rate": 4.704381712259797e-06, + "loss": 2.4308, + "step": 8590 + }, + { + "epoch": 0.9085236886632826, + "grad_norm": 8.3125, + "learning_rate": 4.704300923473862e-06, + "loss": 2.6771, + "step": 8591 + }, + { + "epoch": 0.9086294416243654, + "grad_norm": 7.71875, + "learning_rate": 4.704220124344042e-06, + "loss": 2.4394, + "step": 8592 + }, + { + "epoch": 0.9087351945854484, + "grad_norm": 8.1875, + "learning_rate": 4.704139314870721e-06, + "loss": 2.6852, + "step": 8593 + }, + { + "epoch": 0.9088409475465313, + "grad_norm": 7.84375, + "learning_rate": 4.704058495054275e-06, + "loss": 2.3778, + "step": 8594 + }, + { + "epoch": 0.9089467005076142, + "grad_norm": 7.9375, + "learning_rate": 4.703977664895083e-06, + "loss": 2.3418, + "step": 8595 + }, + { + "epoch": 0.9090524534686971, + "grad_norm": 7.6875, + "learning_rate": 4.703896824393526e-06, + "loss": 2.6346, + "step": 8596 + }, + { + "epoch": 0.9091582064297801, + "grad_norm": 7.75, + "learning_rate": 4.703815973549982e-06, + "loss": 2.5702, + "step": 8597 + }, + { + "epoch": 0.9092639593908629, + "grad_norm": 8.25, + "learning_rate": 4.703735112364831e-06, + "loss": 2.7488, + "step": 8598 + }, + { + "epoch": 0.9093697123519459, + "grad_norm": 7.875, + "learning_rate": 4.703654240838453e-06, + "loss": 2.5338, + "step": 8599 + }, + { + "epoch": 0.9094754653130288, + "grad_norm": 8.0, + "learning_rate": 4.703573358971226e-06, + "loss": 2.6107, + "step": 8600 + }, + { + "epoch": 0.9095812182741116, + "grad_norm": 7.84375, + "learning_rate": 4.703492466763532e-06, + "loss": 2.5649, + "step": 8601 + }, + { + "epoch": 0.9096869712351946, + "grad_norm": 7.75, + "learning_rate": 4.703411564215748e-06, + "loss": 2.58, + "step": 8602 + }, + { + "epoch": 0.9097927241962775, + "grad_norm": 7.84375, + "learning_rate": 4.703330651328255e-06, + "loss": 2.4737, + "step": 8603 + }, + { + "epoch": 0.9098984771573604, + "grad_norm": 8.4375, + "learning_rate": 4.703249728101432e-06, + "loss": 2.703, + "step": 8604 + }, + { + "epoch": 0.9100042301184433, + "grad_norm": 7.65625, + "learning_rate": 4.70316879453566e-06, + "loss": 2.4437, + "step": 8605 + }, + { + "epoch": 0.9101099830795262, + "grad_norm": 7.4375, + "learning_rate": 4.703087850631316e-06, + "loss": 2.4539, + "step": 8606 + }, + { + "epoch": 0.9102157360406091, + "grad_norm": 8.0, + "learning_rate": 4.703006896388784e-06, + "loss": 2.6205, + "step": 8607 + }, + { + "epoch": 0.9103214890016921, + "grad_norm": 8.4375, + "learning_rate": 4.70292593180844e-06, + "loss": 2.658, + "step": 8608 + }, + { + "epoch": 0.9104272419627749, + "grad_norm": 8.3125, + "learning_rate": 4.702844956890666e-06, + "loss": 2.7318, + "step": 8609 + }, + { + "epoch": 0.9105329949238579, + "grad_norm": 7.75, + "learning_rate": 4.702763971635842e-06, + "loss": 2.6705, + "step": 8610 + }, + { + "epoch": 0.9106387478849408, + "grad_norm": 7.625, + "learning_rate": 4.702682976044348e-06, + "loss": 2.5464, + "step": 8611 + }, + { + "epoch": 0.9107445008460237, + "grad_norm": 8.1875, + "learning_rate": 4.7026019701165626e-06, + "loss": 2.5478, + "step": 8612 + }, + { + "epoch": 0.9108502538071066, + "grad_norm": 8.375, + "learning_rate": 4.702520953852867e-06, + "loss": 2.5554, + "step": 8613 + }, + { + "epoch": 0.9109560067681896, + "grad_norm": 8.25, + "learning_rate": 4.702439927253643e-06, + "loss": 2.6133, + "step": 8614 + }, + { + "epoch": 0.9110617597292724, + "grad_norm": 8.3125, + "learning_rate": 4.702358890319268e-06, + "loss": 2.6322, + "step": 8615 + }, + { + "epoch": 0.9111675126903553, + "grad_norm": 7.75, + "learning_rate": 4.702277843050122e-06, + "loss": 2.6116, + "step": 8616 + }, + { + "epoch": 0.9112732656514383, + "grad_norm": 8.375, + "learning_rate": 4.702196785446589e-06, + "loss": 2.6561, + "step": 8617 + }, + { + "epoch": 0.9113790186125211, + "grad_norm": 8.1875, + "learning_rate": 4.702115717509046e-06, + "loss": 2.7506, + "step": 8618 + }, + { + "epoch": 0.9114847715736041, + "grad_norm": 8.0, + "learning_rate": 4.702034639237875e-06, + "loss": 2.6059, + "step": 8619 + }, + { + "epoch": 0.9115905245346869, + "grad_norm": 7.875, + "learning_rate": 4.7019535506334564e-06, + "loss": 2.4954, + "step": 8620 + }, + { + "epoch": 0.9116962774957699, + "grad_norm": 7.8125, + "learning_rate": 4.7018724516961695e-06, + "loss": 2.6398, + "step": 8621 + }, + { + "epoch": 0.9118020304568528, + "grad_norm": 8.0625, + "learning_rate": 4.701791342426396e-06, + "loss": 2.6416, + "step": 8622 + }, + { + "epoch": 0.9119077834179357, + "grad_norm": 8.3125, + "learning_rate": 4.701710222824517e-06, + "loss": 2.7602, + "step": 8623 + }, + { + "epoch": 0.9120135363790186, + "grad_norm": 8.0625, + "learning_rate": 4.701629092890911e-06, + "loss": 2.2354, + "step": 8624 + }, + { + "epoch": 0.9121192893401016, + "grad_norm": 8.125, + "learning_rate": 4.701547952625961e-06, + "loss": 2.6756, + "step": 8625 + }, + { + "epoch": 0.9122250423011844, + "grad_norm": 8.25, + "learning_rate": 4.701466802030047e-06, + "loss": 2.6444, + "step": 8626 + }, + { + "epoch": 0.9123307952622673, + "grad_norm": 7.8125, + "learning_rate": 4.701385641103549e-06, + "loss": 2.4886, + "step": 8627 + }, + { + "epoch": 0.9124365482233503, + "grad_norm": 9.25, + "learning_rate": 4.701304469846849e-06, + "loss": 2.6847, + "step": 8628 + }, + { + "epoch": 0.9125423011844331, + "grad_norm": 8.0, + "learning_rate": 4.7012232882603286e-06, + "loss": 2.8172, + "step": 8629 + }, + { + "epoch": 0.9126480541455161, + "grad_norm": 7.9375, + "learning_rate": 4.7011420963443655e-06, + "loss": 2.5546, + "step": 8630 + }, + { + "epoch": 0.912753807106599, + "grad_norm": 7.90625, + "learning_rate": 4.701060894099344e-06, + "loss": 2.2138, + "step": 8631 + }, + { + "epoch": 0.9128595600676819, + "grad_norm": 7.9375, + "learning_rate": 4.700979681525643e-06, + "loss": 2.3535, + "step": 8632 + }, + { + "epoch": 0.9129653130287648, + "grad_norm": 7.6875, + "learning_rate": 4.700898458623645e-06, + "loss": 2.6843, + "step": 8633 + }, + { + "epoch": 0.9130710659898477, + "grad_norm": 7.96875, + "learning_rate": 4.7008172253937315e-06, + "loss": 2.7801, + "step": 8634 + }, + { + "epoch": 0.9131768189509306, + "grad_norm": 7.375, + "learning_rate": 4.700735981836282e-06, + "loss": 2.6078, + "step": 8635 + }, + { + "epoch": 0.9132825719120136, + "grad_norm": 7.75, + "learning_rate": 4.700654727951678e-06, + "loss": 2.4973, + "step": 8636 + }, + { + "epoch": 0.9133883248730964, + "grad_norm": 9.3125, + "learning_rate": 4.700573463740302e-06, + "loss": 2.7397, + "step": 8637 + }, + { + "epoch": 0.9134940778341794, + "grad_norm": 8.4375, + "learning_rate": 4.700492189202535e-06, + "loss": 2.772, + "step": 8638 + }, + { + "epoch": 0.9135998307952623, + "grad_norm": 8.75, + "learning_rate": 4.700410904338758e-06, + "loss": 2.7308, + "step": 8639 + }, + { + "epoch": 0.9137055837563451, + "grad_norm": 7.71875, + "learning_rate": 4.700329609149352e-06, + "loss": 2.5655, + "step": 8640 + }, + { + "epoch": 0.9138113367174281, + "grad_norm": 7.625, + "learning_rate": 4.7002483036347e-06, + "loss": 2.4455, + "step": 8641 + }, + { + "epoch": 0.913917089678511, + "grad_norm": 8.25, + "learning_rate": 4.700166987795183e-06, + "loss": 2.5849, + "step": 8642 + }, + { + "epoch": 0.9140228426395939, + "grad_norm": 8.0, + "learning_rate": 4.700085661631181e-06, + "loss": 2.6416, + "step": 8643 + }, + { + "epoch": 0.9141285956006768, + "grad_norm": 8.0625, + "learning_rate": 4.700004325143077e-06, + "loss": 2.6006, + "step": 8644 + }, + { + "epoch": 0.9142343485617598, + "grad_norm": 8.1875, + "learning_rate": 4.699922978331252e-06, + "loss": 2.4545, + "step": 8645 + }, + { + "epoch": 0.9143401015228426, + "grad_norm": 9.0625, + "learning_rate": 4.699841621196089e-06, + "loss": 2.5313, + "step": 8646 + }, + { + "epoch": 0.9144458544839256, + "grad_norm": 7.84375, + "learning_rate": 4.699760253737969e-06, + "loss": 2.6284, + "step": 8647 + }, + { + "epoch": 0.9145516074450084, + "grad_norm": 7.9375, + "learning_rate": 4.699678875957273e-06, + "loss": 2.564, + "step": 8648 + }, + { + "epoch": 0.9146573604060914, + "grad_norm": 8.5, + "learning_rate": 4.699597487854385e-06, + "loss": 2.5976, + "step": 8649 + }, + { + "epoch": 0.9147631133671743, + "grad_norm": 7.6875, + "learning_rate": 4.6995160894296844e-06, + "loss": 2.3444, + "step": 8650 + }, + { + "epoch": 0.9148688663282571, + "grad_norm": 8.5625, + "learning_rate": 4.699434680683555e-06, + "loss": 2.6325, + "step": 8651 + }, + { + "epoch": 0.9149746192893401, + "grad_norm": 8.5, + "learning_rate": 4.6993532616163774e-06, + "loss": 2.5973, + "step": 8652 + }, + { + "epoch": 0.915080372250423, + "grad_norm": 7.59375, + "learning_rate": 4.699271832228536e-06, + "loss": 2.3723, + "step": 8653 + }, + { + "epoch": 0.9151861252115059, + "grad_norm": 7.875, + "learning_rate": 4.69919039252041e-06, + "loss": 2.5557, + "step": 8654 + }, + { + "epoch": 0.9152918781725888, + "grad_norm": 8.0, + "learning_rate": 4.699108942492383e-06, + "loss": 2.5628, + "step": 8655 + }, + { + "epoch": 0.9153976311336718, + "grad_norm": 9.3125, + "learning_rate": 4.699027482144838e-06, + "loss": 2.6795, + "step": 8656 + }, + { + "epoch": 0.9155033840947546, + "grad_norm": 7.8125, + "learning_rate": 4.698946011478155e-06, + "loss": 2.5271, + "step": 8657 + }, + { + "epoch": 0.9156091370558376, + "grad_norm": 7.375, + "learning_rate": 4.69886453049272e-06, + "loss": 2.3895, + "step": 8658 + }, + { + "epoch": 0.9157148900169205, + "grad_norm": 8.3125, + "learning_rate": 4.698783039188911e-06, + "loss": 2.5128, + "step": 8659 + }, + { + "epoch": 0.9158206429780034, + "grad_norm": 8.5, + "learning_rate": 4.698701537567113e-06, + "loss": 2.7262, + "step": 8660 + }, + { + "epoch": 0.9159263959390863, + "grad_norm": 7.75, + "learning_rate": 4.6986200256277085e-06, + "loss": 2.5218, + "step": 8661 + }, + { + "epoch": 0.9160321489001692, + "grad_norm": 8.4375, + "learning_rate": 4.698538503371079e-06, + "loss": 2.5278, + "step": 8662 + }, + { + "epoch": 0.9161379018612521, + "grad_norm": 8.1875, + "learning_rate": 4.698456970797607e-06, + "loss": 2.5913, + "step": 8663 + }, + { + "epoch": 0.916243654822335, + "grad_norm": 8.0, + "learning_rate": 4.698375427907675e-06, + "loss": 2.6959, + "step": 8664 + }, + { + "epoch": 0.9163494077834179, + "grad_norm": 8.0, + "learning_rate": 4.6982938747016685e-06, + "loss": 2.6039, + "step": 8665 + }, + { + "epoch": 0.9164551607445008, + "grad_norm": 8.0, + "learning_rate": 4.698212311179966e-06, + "loss": 2.7261, + "step": 8666 + }, + { + "epoch": 0.9165609137055838, + "grad_norm": 8.125, + "learning_rate": 4.698130737342953e-06, + "loss": 2.6408, + "step": 8667 + }, + { + "epoch": 0.9166666666666666, + "grad_norm": 8.375, + "learning_rate": 4.698049153191011e-06, + "loss": 2.5735, + "step": 8668 + }, + { + "epoch": 0.9167724196277496, + "grad_norm": 8.5, + "learning_rate": 4.697967558724523e-06, + "loss": 2.4152, + "step": 8669 + }, + { + "epoch": 0.9168781725888325, + "grad_norm": 7.71875, + "learning_rate": 4.697885953943873e-06, + "loss": 2.5802, + "step": 8670 + }, + { + "epoch": 0.9169839255499154, + "grad_norm": 9.125, + "learning_rate": 4.697804338849443e-06, + "loss": 2.5925, + "step": 8671 + }, + { + "epoch": 0.9170896785109983, + "grad_norm": 7.8125, + "learning_rate": 4.6977227134416155e-06, + "loss": 2.5938, + "step": 8672 + }, + { + "epoch": 0.9171954314720813, + "grad_norm": 8.625, + "learning_rate": 4.697641077720775e-06, + "loss": 2.2457, + "step": 8673 + }, + { + "epoch": 0.9173011844331641, + "grad_norm": 8.125, + "learning_rate": 4.697559431687303e-06, + "loss": 2.701, + "step": 8674 + }, + { + "epoch": 0.9174069373942471, + "grad_norm": 8.375, + "learning_rate": 4.697477775341584e-06, + "loss": 2.7466, + "step": 8675 + }, + { + "epoch": 0.9175126903553299, + "grad_norm": 8.9375, + "learning_rate": 4.697396108684e-06, + "loss": 2.6274, + "step": 8676 + }, + { + "epoch": 0.9176184433164128, + "grad_norm": 8.0625, + "learning_rate": 4.697314431714936e-06, + "loss": 2.5944, + "step": 8677 + }, + { + "epoch": 0.9177241962774958, + "grad_norm": 8.1875, + "learning_rate": 4.697232744434772e-06, + "loss": 2.787, + "step": 8678 + }, + { + "epoch": 0.9178299492385786, + "grad_norm": 8.625, + "learning_rate": 4.697151046843895e-06, + "loss": 2.7361, + "step": 8679 + }, + { + "epoch": 0.9179357021996616, + "grad_norm": 8.75, + "learning_rate": 4.697069338942686e-06, + "loss": 2.5693, + "step": 8680 + }, + { + "epoch": 0.9180414551607445, + "grad_norm": 8.25, + "learning_rate": 4.696987620731529e-06, + "loss": 2.6274, + "step": 8681 + }, + { + "epoch": 0.9181472081218274, + "grad_norm": 7.28125, + "learning_rate": 4.696905892210808e-06, + "loss": 2.4166, + "step": 8682 + }, + { + "epoch": 0.9182529610829103, + "grad_norm": 8.25, + "learning_rate": 4.696824153380906e-06, + "loss": 2.7101, + "step": 8683 + }, + { + "epoch": 0.9183587140439933, + "grad_norm": 8.375, + "learning_rate": 4.696742404242207e-06, + "loss": 2.4414, + "step": 8684 + }, + { + "epoch": 0.9184644670050761, + "grad_norm": 9.25, + "learning_rate": 4.696660644795094e-06, + "loss": 2.684, + "step": 8685 + }, + { + "epoch": 0.9185702199661591, + "grad_norm": 8.4375, + "learning_rate": 4.696578875039951e-06, + "loss": 2.4691, + "step": 8686 + }, + { + "epoch": 0.918675972927242, + "grad_norm": 7.5, + "learning_rate": 4.696497094977162e-06, + "loss": 2.5755, + "step": 8687 + }, + { + "epoch": 0.9187817258883249, + "grad_norm": 8.5625, + "learning_rate": 4.69641530460711e-06, + "loss": 2.526, + "step": 8688 + }, + { + "epoch": 0.9188874788494078, + "grad_norm": 7.625, + "learning_rate": 4.696333503930181e-06, + "loss": 2.5157, + "step": 8689 + }, + { + "epoch": 0.9189932318104906, + "grad_norm": 8.5625, + "learning_rate": 4.696251692946755e-06, + "loss": 2.6101, + "step": 8690 + }, + { + "epoch": 0.9190989847715736, + "grad_norm": 7.875, + "learning_rate": 4.6961698716572185e-06, + "loss": 2.6546, + "step": 8691 + }, + { + "epoch": 0.9192047377326565, + "grad_norm": 7.84375, + "learning_rate": 4.696088040061955e-06, + "loss": 2.4902, + "step": 8692 + }, + { + "epoch": 0.9193104906937394, + "grad_norm": 8.4375, + "learning_rate": 4.696006198161349e-06, + "loss": 2.6881, + "step": 8693 + }, + { + "epoch": 0.9194162436548223, + "grad_norm": 7.96875, + "learning_rate": 4.695924345955783e-06, + "loss": 2.3685, + "step": 8694 + }, + { + "epoch": 0.9195219966159053, + "grad_norm": 8.75, + "learning_rate": 4.6958424834456436e-06, + "loss": 2.6575, + "step": 8695 + }, + { + "epoch": 0.9196277495769881, + "grad_norm": 8.1875, + "learning_rate": 4.695760610631313e-06, + "loss": 2.526, + "step": 8696 + }, + { + "epoch": 0.9197335025380711, + "grad_norm": 8.25, + "learning_rate": 4.695678727513176e-06, + "loss": 2.7739, + "step": 8697 + }, + { + "epoch": 0.919839255499154, + "grad_norm": 8.3125, + "learning_rate": 4.6955968340916165e-06, + "loss": 2.6663, + "step": 8698 + }, + { + "epoch": 0.9199450084602369, + "grad_norm": 7.96875, + "learning_rate": 4.695514930367019e-06, + "loss": 2.5535, + "step": 8699 + }, + { + "epoch": 0.9200507614213198, + "grad_norm": 8.3125, + "learning_rate": 4.695433016339768e-06, + "loss": 2.4216, + "step": 8700 + }, + { + "epoch": 0.9201565143824028, + "grad_norm": 8.0, + "learning_rate": 4.695351092010247e-06, + "loss": 2.5265, + "step": 8701 + }, + { + "epoch": 0.9202622673434856, + "grad_norm": 8.8125, + "learning_rate": 4.695269157378843e-06, + "loss": 2.7321, + "step": 8702 + }, + { + "epoch": 0.9203680203045685, + "grad_norm": 7.84375, + "learning_rate": 4.695187212445937e-06, + "loss": 2.6228, + "step": 8703 + }, + { + "epoch": 0.9204737732656514, + "grad_norm": 8.3125, + "learning_rate": 4.695105257211916e-06, + "loss": 2.5104, + "step": 8704 + }, + { + "epoch": 0.9205795262267343, + "grad_norm": 7.65625, + "learning_rate": 4.695023291677163e-06, + "loss": 2.5293, + "step": 8705 + }, + { + "epoch": 0.9206852791878173, + "grad_norm": 8.3125, + "learning_rate": 4.694941315842064e-06, + "loss": 2.5189, + "step": 8706 + }, + { + "epoch": 0.9207910321489001, + "grad_norm": 7.875, + "learning_rate": 4.694859329707003e-06, + "loss": 2.5111, + "step": 8707 + }, + { + "epoch": 0.9208967851099831, + "grad_norm": 8.6875, + "learning_rate": 4.6947773332723645e-06, + "loss": 2.2463, + "step": 8708 + }, + { + "epoch": 0.921002538071066, + "grad_norm": 8.25, + "learning_rate": 4.6946953265385346e-06, + "loss": 2.6673, + "step": 8709 + }, + { + "epoch": 0.9211082910321489, + "grad_norm": 8.25, + "learning_rate": 4.694613309505896e-06, + "loss": 2.6484, + "step": 8710 + }, + { + "epoch": 0.9212140439932318, + "grad_norm": 7.5, + "learning_rate": 4.6945312821748355e-06, + "loss": 2.533, + "step": 8711 + }, + { + "epoch": 0.9213197969543148, + "grad_norm": 8.25, + "learning_rate": 4.694449244545737e-06, + "loss": 2.534, + "step": 8712 + }, + { + "epoch": 0.9214255499153976, + "grad_norm": 8.5, + "learning_rate": 4.694367196618985e-06, + "loss": 2.7535, + "step": 8713 + }, + { + "epoch": 0.9215313028764806, + "grad_norm": 7.40625, + "learning_rate": 4.694285138394966e-06, + "loss": 2.577, + "step": 8714 + }, + { + "epoch": 0.9216370558375635, + "grad_norm": 7.9375, + "learning_rate": 4.694203069874064e-06, + "loss": 2.6586, + "step": 8715 + }, + { + "epoch": 0.9217428087986463, + "grad_norm": 7.78125, + "learning_rate": 4.694120991056665e-06, + "loss": 2.4904, + "step": 8716 + }, + { + "epoch": 0.9218485617597293, + "grad_norm": 7.84375, + "learning_rate": 4.694038901943153e-06, + "loss": 2.5748, + "step": 8717 + }, + { + "epoch": 0.9219543147208121, + "grad_norm": 7.75, + "learning_rate": 4.693956802533914e-06, + "loss": 2.5685, + "step": 8718 + }, + { + "epoch": 0.9220600676818951, + "grad_norm": 7.625, + "learning_rate": 4.693874692829332e-06, + "loss": 2.6399, + "step": 8719 + }, + { + "epoch": 0.922165820642978, + "grad_norm": 8.125, + "learning_rate": 4.693792572829795e-06, + "loss": 2.6876, + "step": 8720 + }, + { + "epoch": 0.9222715736040609, + "grad_norm": 8.25, + "learning_rate": 4.693710442535686e-06, + "loss": 2.4192, + "step": 8721 + }, + { + "epoch": 0.9223773265651438, + "grad_norm": 8.0, + "learning_rate": 4.693628301947391e-06, + "loss": 2.5501, + "step": 8722 + }, + { + "epoch": 0.9224830795262268, + "grad_norm": 8.0, + "learning_rate": 4.6935461510652955e-06, + "loss": 2.7348, + "step": 8723 + }, + { + "epoch": 0.9225888324873096, + "grad_norm": 7.78125, + "learning_rate": 4.693463989889785e-06, + "loss": 2.6504, + "step": 8724 + }, + { + "epoch": 0.9226945854483926, + "grad_norm": 8.8125, + "learning_rate": 4.6933818184212445e-06, + "loss": 2.707, + "step": 8725 + }, + { + "epoch": 0.9228003384094755, + "grad_norm": 7.875, + "learning_rate": 4.693299636660061e-06, + "loss": 2.6878, + "step": 8726 + }, + { + "epoch": 0.9229060913705583, + "grad_norm": 8.25, + "learning_rate": 4.693217444606619e-06, + "loss": 2.6546, + "step": 8727 + }, + { + "epoch": 0.9230118443316413, + "grad_norm": 7.71875, + "learning_rate": 4.693135242261305e-06, + "loss": 2.5139, + "step": 8728 + }, + { + "epoch": 0.9231175972927242, + "grad_norm": 8.3125, + "learning_rate": 4.693053029624504e-06, + "loss": 2.4028, + "step": 8729 + }, + { + "epoch": 0.9232233502538071, + "grad_norm": 7.9375, + "learning_rate": 4.6929708066966016e-06, + "loss": 2.6465, + "step": 8730 + }, + { + "epoch": 0.92332910321489, + "grad_norm": 8.1875, + "learning_rate": 4.692888573477985e-06, + "loss": 2.4719, + "step": 8731 + }, + { + "epoch": 0.9234348561759729, + "grad_norm": 7.875, + "learning_rate": 4.6928063299690385e-06, + "loss": 2.6133, + "step": 8732 + }, + { + "epoch": 0.9235406091370558, + "grad_norm": 8.375, + "learning_rate": 4.692724076170149e-06, + "loss": 2.8795, + "step": 8733 + }, + { + "epoch": 0.9236463620981388, + "grad_norm": 7.375, + "learning_rate": 4.692641812081702e-06, + "loss": 2.5511, + "step": 8734 + }, + { + "epoch": 0.9237521150592216, + "grad_norm": 8.75, + "learning_rate": 4.692559537704084e-06, + "loss": 2.4771, + "step": 8735 + }, + { + "epoch": 0.9238578680203046, + "grad_norm": 8.125, + "learning_rate": 4.69247725303768e-06, + "loss": 2.5254, + "step": 8736 + }, + { + "epoch": 0.9239636209813875, + "grad_norm": 7.65625, + "learning_rate": 4.692394958082879e-06, + "loss": 2.6226, + "step": 8737 + }, + { + "epoch": 0.9240693739424704, + "grad_norm": 8.1875, + "learning_rate": 4.692312652840063e-06, + "loss": 2.3934, + "step": 8738 + }, + { + "epoch": 0.9241751269035533, + "grad_norm": 8.375, + "learning_rate": 4.692230337309621e-06, + "loss": 2.7302, + "step": 8739 + }, + { + "epoch": 0.9242808798646363, + "grad_norm": 7.53125, + "learning_rate": 4.692148011491939e-06, + "loss": 2.4834, + "step": 8740 + }, + { + "epoch": 0.9243866328257191, + "grad_norm": 7.6875, + "learning_rate": 4.692065675387403e-06, + "loss": 2.5656, + "step": 8741 + }, + { + "epoch": 0.924492385786802, + "grad_norm": 8.125, + "learning_rate": 4.6919833289963985e-06, + "loss": 2.5311, + "step": 8742 + }, + { + "epoch": 0.924598138747885, + "grad_norm": 7.875, + "learning_rate": 4.691900972319313e-06, + "loss": 2.4954, + "step": 8743 + }, + { + "epoch": 0.9247038917089678, + "grad_norm": 7.6875, + "learning_rate": 4.691818605356533e-06, + "loss": 2.6415, + "step": 8744 + }, + { + "epoch": 0.9248096446700508, + "grad_norm": 8.1875, + "learning_rate": 4.691736228108445e-06, + "loss": 2.3306, + "step": 8745 + }, + { + "epoch": 0.9249153976311336, + "grad_norm": 7.78125, + "learning_rate": 4.691653840575434e-06, + "loss": 2.5145, + "step": 8746 + }, + { + "epoch": 0.9250211505922166, + "grad_norm": 8.5, + "learning_rate": 4.691571442757889e-06, + "loss": 2.5512, + "step": 8747 + }, + { + "epoch": 0.9251269035532995, + "grad_norm": 9.1875, + "learning_rate": 4.691489034656196e-06, + "loss": 2.7328, + "step": 8748 + }, + { + "epoch": 0.9252326565143824, + "grad_norm": 7.90625, + "learning_rate": 4.69140661627074e-06, + "loss": 2.5699, + "step": 8749 + }, + { + "epoch": 0.9253384094754653, + "grad_norm": 7.84375, + "learning_rate": 4.691324187601909e-06, + "loss": 2.5433, + "step": 8750 + }, + { + "epoch": 0.9254441624365483, + "grad_norm": 7.96875, + "learning_rate": 4.6912417486500905e-06, + "loss": 2.7285, + "step": 8751 + }, + { + "epoch": 0.9255499153976311, + "grad_norm": 7.625, + "learning_rate": 4.69115929941567e-06, + "loss": 2.5475, + "step": 8752 + }, + { + "epoch": 0.925655668358714, + "grad_norm": 7.5625, + "learning_rate": 4.691076839899037e-06, + "loss": 2.56, + "step": 8753 + }, + { + "epoch": 0.925761421319797, + "grad_norm": 7.59375, + "learning_rate": 4.690994370100574e-06, + "loss": 2.6032, + "step": 8754 + }, + { + "epoch": 0.9258671742808798, + "grad_norm": 8.25, + "learning_rate": 4.690911890020672e-06, + "loss": 2.6075, + "step": 8755 + }, + { + "epoch": 0.9259729272419628, + "grad_norm": 7.84375, + "learning_rate": 4.690829399659716e-06, + "loss": 2.565, + "step": 8756 + }, + { + "epoch": 0.9260786802030457, + "grad_norm": 7.8125, + "learning_rate": 4.690746899018094e-06, + "loss": 2.3571, + "step": 8757 + }, + { + "epoch": 0.9261844331641286, + "grad_norm": 8.125, + "learning_rate": 4.690664388096192e-06, + "loss": 2.625, + "step": 8758 + }, + { + "epoch": 0.9262901861252115, + "grad_norm": 7.71875, + "learning_rate": 4.6905818668943996e-06, + "loss": 2.5393, + "step": 8759 + }, + { + "epoch": 0.9263959390862944, + "grad_norm": 7.5625, + "learning_rate": 4.690499335413101e-06, + "loss": 2.4541, + "step": 8760 + }, + { + "epoch": 0.9265016920473773, + "grad_norm": 8.3125, + "learning_rate": 4.690416793652685e-06, + "loss": 2.811, + "step": 8761 + }, + { + "epoch": 0.9266074450084603, + "grad_norm": 7.75, + "learning_rate": 4.690334241613539e-06, + "loss": 2.6748, + "step": 8762 + }, + { + "epoch": 0.9267131979695431, + "grad_norm": 7.78125, + "learning_rate": 4.69025167929605e-06, + "loss": 2.4971, + "step": 8763 + }, + { + "epoch": 0.926818950930626, + "grad_norm": 8.1875, + "learning_rate": 4.690169106700606e-06, + "loss": 2.6059, + "step": 8764 + }, + { + "epoch": 0.926924703891709, + "grad_norm": 8.5, + "learning_rate": 4.690086523827595e-06, + "loss": 2.6566, + "step": 8765 + }, + { + "epoch": 0.9270304568527918, + "grad_norm": 8.0, + "learning_rate": 4.690003930677402e-06, + "loss": 2.8482, + "step": 8766 + }, + { + "epoch": 0.9271362098138748, + "grad_norm": 8.6875, + "learning_rate": 4.689921327250417e-06, + "loss": 2.714, + "step": 8767 + }, + { + "epoch": 0.9272419627749577, + "grad_norm": 8.5, + "learning_rate": 4.689838713547027e-06, + "loss": 2.5787, + "step": 8768 + }, + { + "epoch": 0.9273477157360406, + "grad_norm": 7.71875, + "learning_rate": 4.6897560895676195e-06, + "loss": 2.6835, + "step": 8769 + }, + { + "epoch": 0.9274534686971235, + "grad_norm": 7.5625, + "learning_rate": 4.689673455312582e-06, + "loss": 2.4909, + "step": 8770 + }, + { + "epoch": 0.9275592216582065, + "grad_norm": 7.6875, + "learning_rate": 4.689590810782303e-06, + "loss": 2.5768, + "step": 8771 + }, + { + "epoch": 0.9276649746192893, + "grad_norm": 8.0625, + "learning_rate": 4.689508155977169e-06, + "loss": 2.6002, + "step": 8772 + }, + { + "epoch": 0.9277707275803723, + "grad_norm": 8.25, + "learning_rate": 4.68942549089757e-06, + "loss": 2.6789, + "step": 8773 + }, + { + "epoch": 0.9278764805414551, + "grad_norm": 9.5625, + "learning_rate": 4.689342815543891e-06, + "loss": 2.8485, + "step": 8774 + }, + { + "epoch": 0.9279822335025381, + "grad_norm": 8.0625, + "learning_rate": 4.689260129916523e-06, + "loss": 2.4533, + "step": 8775 + }, + { + "epoch": 0.928087986463621, + "grad_norm": 7.53125, + "learning_rate": 4.689177434015851e-06, + "loss": 2.641, + "step": 8776 + }, + { + "epoch": 0.9281937394247038, + "grad_norm": 7.65625, + "learning_rate": 4.6890947278422664e-06, + "loss": 2.5171, + "step": 8777 + }, + { + "epoch": 0.9282994923857868, + "grad_norm": 8.4375, + "learning_rate": 4.689012011396155e-06, + "loss": 2.698, + "step": 8778 + }, + { + "epoch": 0.9284052453468697, + "grad_norm": 7.8125, + "learning_rate": 4.688929284677905e-06, + "loss": 2.5723, + "step": 8779 + }, + { + "epoch": 0.9285109983079526, + "grad_norm": 7.75, + "learning_rate": 4.688846547687905e-06, + "loss": 2.4839, + "step": 8780 + }, + { + "epoch": 0.9286167512690355, + "grad_norm": 8.125, + "learning_rate": 4.6887638004265445e-06, + "loss": 2.5233, + "step": 8781 + }, + { + "epoch": 0.9287225042301185, + "grad_norm": 7.90625, + "learning_rate": 4.6886810428942094e-06, + "loss": 2.5876, + "step": 8782 + }, + { + "epoch": 0.9288282571912013, + "grad_norm": 8.4375, + "learning_rate": 4.68859827509129e-06, + "loss": 2.6456, + "step": 8783 + }, + { + "epoch": 0.9289340101522843, + "grad_norm": 7.875, + "learning_rate": 4.688515497018174e-06, + "loss": 2.7325, + "step": 8784 + }, + { + "epoch": 0.9290397631133672, + "grad_norm": 8.375, + "learning_rate": 4.68843270867525e-06, + "loss": 2.7327, + "step": 8785 + }, + { + "epoch": 0.9291455160744501, + "grad_norm": 7.9375, + "learning_rate": 4.688349910062905e-06, + "loss": 2.4523, + "step": 8786 + }, + { + "epoch": 0.929251269035533, + "grad_norm": 13.8125, + "learning_rate": 4.688267101181531e-06, + "loss": 2.5414, + "step": 8787 + }, + { + "epoch": 0.9293570219966159, + "grad_norm": 8.25, + "learning_rate": 4.688184282031513e-06, + "loss": 2.4554, + "step": 8788 + }, + { + "epoch": 0.9294627749576988, + "grad_norm": 7.53125, + "learning_rate": 4.6881014526132415e-06, + "loss": 2.492, + "step": 8789 + }, + { + "epoch": 0.9295685279187818, + "grad_norm": 7.65625, + "learning_rate": 4.688018612927105e-06, + "loss": 2.4404, + "step": 8790 + }, + { + "epoch": 0.9296742808798646, + "grad_norm": 7.9375, + "learning_rate": 4.687935762973491e-06, + "loss": 2.523, + "step": 8791 + }, + { + "epoch": 0.9297800338409475, + "grad_norm": 8.0625, + "learning_rate": 4.6878529027527905e-06, + "loss": 2.5014, + "step": 8792 + }, + { + "epoch": 0.9298857868020305, + "grad_norm": 8.1875, + "learning_rate": 4.68777003226539e-06, + "loss": 2.8207, + "step": 8793 + }, + { + "epoch": 0.9299915397631133, + "grad_norm": 8.0625, + "learning_rate": 4.6876871515116804e-06, + "loss": 2.5439, + "step": 8794 + }, + { + "epoch": 0.9300972927241963, + "grad_norm": 8.25, + "learning_rate": 4.6876042604920495e-06, + "loss": 2.6609, + "step": 8795 + }, + { + "epoch": 0.9302030456852792, + "grad_norm": 8.3125, + "learning_rate": 4.687521359206887e-06, + "loss": 2.6929, + "step": 8796 + }, + { + "epoch": 0.9303087986463621, + "grad_norm": 8.9375, + "learning_rate": 4.687438447656581e-06, + "loss": 2.9007, + "step": 8797 + }, + { + "epoch": 0.930414551607445, + "grad_norm": 8.875, + "learning_rate": 4.68735552584152e-06, + "loss": 2.3278, + "step": 8798 + }, + { + "epoch": 0.930520304568528, + "grad_norm": 8.125, + "learning_rate": 4.687272593762096e-06, + "loss": 2.5657, + "step": 8799 + }, + { + "epoch": 0.9306260575296108, + "grad_norm": 8.6875, + "learning_rate": 4.687189651418695e-06, + "loss": 2.6818, + "step": 8800 + }, + { + "epoch": 0.9307318104906938, + "grad_norm": 7.4375, + "learning_rate": 4.687106698811707e-06, + "loss": 2.5592, + "step": 8801 + }, + { + "epoch": 0.9308375634517766, + "grad_norm": 7.8125, + "learning_rate": 4.687023735941523e-06, + "loss": 2.5666, + "step": 8802 + }, + { + "epoch": 0.9309433164128595, + "grad_norm": 8.0, + "learning_rate": 4.6869407628085305e-06, + "loss": 2.5299, + "step": 8803 + }, + { + "epoch": 0.9310490693739425, + "grad_norm": 7.96875, + "learning_rate": 4.68685777941312e-06, + "loss": 2.5453, + "step": 8804 + }, + { + "epoch": 0.9311548223350253, + "grad_norm": 8.4375, + "learning_rate": 4.68677478575568e-06, + "loss": 2.38, + "step": 8805 + }, + { + "epoch": 0.9312605752961083, + "grad_norm": 7.96875, + "learning_rate": 4.6866917818366e-06, + "loss": 2.4631, + "step": 8806 + }, + { + "epoch": 0.9313663282571912, + "grad_norm": 7.84375, + "learning_rate": 4.6866087676562696e-06, + "loss": 2.512, + "step": 8807 + }, + { + "epoch": 0.9314720812182741, + "grad_norm": 7.90625, + "learning_rate": 4.686525743215079e-06, + "loss": 2.5171, + "step": 8808 + }, + { + "epoch": 0.931577834179357, + "grad_norm": 7.375, + "learning_rate": 4.686442708513418e-06, + "loss": 2.55, + "step": 8809 + }, + { + "epoch": 0.93168358714044, + "grad_norm": 8.125, + "learning_rate": 4.686359663551675e-06, + "loss": 2.4298, + "step": 8810 + }, + { + "epoch": 0.9317893401015228, + "grad_norm": 7.78125, + "learning_rate": 4.6862766083302406e-06, + "loss": 2.6132, + "step": 8811 + }, + { + "epoch": 0.9318950930626058, + "grad_norm": 8.25, + "learning_rate": 4.6861935428495035e-06, + "loss": 2.5537, + "step": 8812 + }, + { + "epoch": 0.9320008460236887, + "grad_norm": 8.3125, + "learning_rate": 4.686110467109854e-06, + "loss": 2.7119, + "step": 8813 + }, + { + "epoch": 0.9321065989847716, + "grad_norm": 7.71875, + "learning_rate": 4.686027381111683e-06, + "loss": 2.5148, + "step": 8814 + }, + { + "epoch": 0.9322123519458545, + "grad_norm": 8.125, + "learning_rate": 4.685944284855379e-06, + "loss": 2.6153, + "step": 8815 + }, + { + "epoch": 0.9323181049069373, + "grad_norm": 7.65625, + "learning_rate": 4.685861178341333e-06, + "loss": 2.6196, + "step": 8816 + }, + { + "epoch": 0.9324238578680203, + "grad_norm": 7.46875, + "learning_rate": 4.685778061569934e-06, + "loss": 2.513, + "step": 8817 + }, + { + "epoch": 0.9325296108291032, + "grad_norm": 8.0, + "learning_rate": 4.685694934541573e-06, + "loss": 2.4807, + "step": 8818 + }, + { + "epoch": 0.9326353637901861, + "grad_norm": 7.75, + "learning_rate": 4.685611797256639e-06, + "loss": 2.6542, + "step": 8819 + }, + { + "epoch": 0.932741116751269, + "grad_norm": 8.5, + "learning_rate": 4.685528649715523e-06, + "loss": 2.8008, + "step": 8820 + }, + { + "epoch": 0.932846869712352, + "grad_norm": 8.375, + "learning_rate": 4.685445491918616e-06, + "loss": 2.6583, + "step": 8821 + }, + { + "epoch": 0.9329526226734348, + "grad_norm": 7.875, + "learning_rate": 4.6853623238663045e-06, + "loss": 2.5364, + "step": 8822 + }, + { + "epoch": 0.9330583756345178, + "grad_norm": 7.84375, + "learning_rate": 4.685279145558984e-06, + "loss": 2.5984, + "step": 8823 + }, + { + "epoch": 0.9331641285956007, + "grad_norm": 8.0625, + "learning_rate": 4.6851959569970406e-06, + "loss": 2.5826, + "step": 8824 + }, + { + "epoch": 0.9332698815566836, + "grad_norm": 8.1875, + "learning_rate": 4.685112758180867e-06, + "loss": 2.5007, + "step": 8825 + }, + { + "epoch": 0.9333756345177665, + "grad_norm": 7.84375, + "learning_rate": 4.685029549110853e-06, + "loss": 2.4752, + "step": 8826 + }, + { + "epoch": 0.9334813874788495, + "grad_norm": 8.9375, + "learning_rate": 4.684946329787389e-06, + "loss": 2.5676, + "step": 8827 + }, + { + "epoch": 0.9335871404399323, + "grad_norm": 8.0, + "learning_rate": 4.684863100210865e-06, + "loss": 2.749, + "step": 8828 + }, + { + "epoch": 0.9336928934010152, + "grad_norm": 7.90625, + "learning_rate": 4.684779860381672e-06, + "loss": 2.4021, + "step": 8829 + }, + { + "epoch": 0.9337986463620981, + "grad_norm": 7.59375, + "learning_rate": 4.684696610300201e-06, + "loss": 2.5903, + "step": 8830 + }, + { + "epoch": 0.933904399323181, + "grad_norm": 8.125, + "learning_rate": 4.684613349966842e-06, + "loss": 2.6141, + "step": 8831 + }, + { + "epoch": 0.934010152284264, + "grad_norm": 8.4375, + "learning_rate": 4.684530079381987e-06, + "loss": 2.5755, + "step": 8832 + }, + { + "epoch": 0.9341159052453468, + "grad_norm": 8.1875, + "learning_rate": 4.6844467985460255e-06, + "loss": 2.5662, + "step": 8833 + }, + { + "epoch": 0.9342216582064298, + "grad_norm": 8.875, + "learning_rate": 4.684363507459347e-06, + "loss": 2.8729, + "step": 8834 + }, + { + "epoch": 0.9343274111675127, + "grad_norm": 7.375, + "learning_rate": 4.684280206122346e-06, + "loss": 2.4415, + "step": 8835 + }, + { + "epoch": 0.9344331641285956, + "grad_norm": 8.125, + "learning_rate": 4.68419689453541e-06, + "loss": 2.5866, + "step": 8836 + }, + { + "epoch": 0.9345389170896785, + "grad_norm": 7.9375, + "learning_rate": 4.684113572698932e-06, + "loss": 2.685, + "step": 8837 + }, + { + "epoch": 0.9346446700507615, + "grad_norm": 8.8125, + "learning_rate": 4.684030240613302e-06, + "loss": 2.8368, + "step": 8838 + }, + { + "epoch": 0.9347504230118443, + "grad_norm": 7.71875, + "learning_rate": 4.683946898278911e-06, + "loss": 2.6278, + "step": 8839 + }, + { + "epoch": 0.9348561759729273, + "grad_norm": 7.75, + "learning_rate": 4.6838635456961505e-06, + "loss": 2.491, + "step": 8840 + }, + { + "epoch": 0.9349619289340102, + "grad_norm": 8.0625, + "learning_rate": 4.683780182865411e-06, + "loss": 2.5482, + "step": 8841 + }, + { + "epoch": 0.935067681895093, + "grad_norm": 8.1875, + "learning_rate": 4.683696809787086e-06, + "loss": 2.5359, + "step": 8842 + }, + { + "epoch": 0.935173434856176, + "grad_norm": 7.53125, + "learning_rate": 4.683613426461564e-06, + "loss": 2.5183, + "step": 8843 + }, + { + "epoch": 0.9352791878172588, + "grad_norm": 7.84375, + "learning_rate": 4.683530032889237e-06, + "loss": 2.6575, + "step": 8844 + }, + { + "epoch": 0.9353849407783418, + "grad_norm": 7.9375, + "learning_rate": 4.683446629070497e-06, + "loss": 2.5206, + "step": 8845 + }, + { + "epoch": 0.9354906937394247, + "grad_norm": 7.625, + "learning_rate": 4.683363215005734e-06, + "loss": 2.5702, + "step": 8846 + }, + { + "epoch": 0.9355964467005076, + "grad_norm": 8.0625, + "learning_rate": 4.683279790695342e-06, + "loss": 2.6631, + "step": 8847 + }, + { + "epoch": 0.9357021996615905, + "grad_norm": 8.0, + "learning_rate": 4.68319635613971e-06, + "loss": 2.6851, + "step": 8848 + }, + { + "epoch": 0.9358079526226735, + "grad_norm": 7.78125, + "learning_rate": 4.683112911339231e-06, + "loss": 2.4895, + "step": 8849 + }, + { + "epoch": 0.9359137055837563, + "grad_norm": 8.5625, + "learning_rate": 4.683029456294295e-06, + "loss": 2.5648, + "step": 8850 + }, + { + "epoch": 0.9360194585448393, + "grad_norm": 7.84375, + "learning_rate": 4.682945991005294e-06, + "loss": 2.504, + "step": 8851 + }, + { + "epoch": 0.9361252115059222, + "grad_norm": 7.8125, + "learning_rate": 4.682862515472622e-06, + "loss": 2.5456, + "step": 8852 + }, + { + "epoch": 0.936230964467005, + "grad_norm": 8.4375, + "learning_rate": 4.682779029696668e-06, + "loss": 2.5356, + "step": 8853 + }, + { + "epoch": 0.936336717428088, + "grad_norm": 8.4375, + "learning_rate": 4.682695533677825e-06, + "loss": 2.5089, + "step": 8854 + }, + { + "epoch": 0.936442470389171, + "grad_norm": 7.84375, + "learning_rate": 4.682612027416484e-06, + "loss": 2.7495, + "step": 8855 + }, + { + "epoch": 0.9365482233502538, + "grad_norm": 7.75, + "learning_rate": 4.682528510913039e-06, + "loss": 2.7404, + "step": 8856 + }, + { + "epoch": 0.9366539763113367, + "grad_norm": 7.53125, + "learning_rate": 4.6824449841678785e-06, + "loss": 2.5836, + "step": 8857 + }, + { + "epoch": 0.9367597292724196, + "grad_norm": 8.375, + "learning_rate": 4.682361447181397e-06, + "loss": 2.6105, + "step": 8858 + }, + { + "epoch": 0.9368654822335025, + "grad_norm": 8.25, + "learning_rate": 4.6822778999539866e-06, + "loss": 2.3349, + "step": 8859 + }, + { + "epoch": 0.9369712351945855, + "grad_norm": 7.8125, + "learning_rate": 4.6821943424860375e-06, + "loss": 2.5268, + "step": 8860 + }, + { + "epoch": 0.9370769881556683, + "grad_norm": 8.6875, + "learning_rate": 4.682110774777943e-06, + "loss": 2.5807, + "step": 8861 + }, + { + "epoch": 0.9371827411167513, + "grad_norm": 8.3125, + "learning_rate": 4.682027196830096e-06, + "loss": 2.2349, + "step": 8862 + }, + { + "epoch": 0.9372884940778342, + "grad_norm": 7.9375, + "learning_rate": 4.681943608642886e-06, + "loss": 2.7005, + "step": 8863 + }, + { + "epoch": 0.937394247038917, + "grad_norm": 8.25, + "learning_rate": 4.681860010216709e-06, + "loss": 2.7166, + "step": 8864 + }, + { + "epoch": 0.9375, + "grad_norm": 7.625, + "learning_rate": 4.681776401551954e-06, + "loss": 2.5726, + "step": 8865 + }, + { + "epoch": 0.937605752961083, + "grad_norm": 7.71875, + "learning_rate": 4.681692782649014e-06, + "loss": 2.4874, + "step": 8866 + }, + { + "epoch": 0.9377115059221658, + "grad_norm": 9.3125, + "learning_rate": 4.681609153508284e-06, + "loss": 2.4208, + "step": 8867 + }, + { + "epoch": 0.9378172588832487, + "grad_norm": 8.625, + "learning_rate": 4.681525514130153e-06, + "loss": 2.6859, + "step": 8868 + }, + { + "epoch": 0.9379230118443317, + "grad_norm": 7.75, + "learning_rate": 4.6814418645150155e-06, + "loss": 2.6631, + "step": 8869 + }, + { + "epoch": 0.9380287648054145, + "grad_norm": 8.0625, + "learning_rate": 4.681358204663263e-06, + "loss": 2.3305, + "step": 8870 + }, + { + "epoch": 0.9381345177664975, + "grad_norm": 7.71875, + "learning_rate": 4.68127453457529e-06, + "loss": 2.5904, + "step": 8871 + }, + { + "epoch": 0.9382402707275804, + "grad_norm": 8.0625, + "learning_rate": 4.681190854251487e-06, + "loss": 2.5218, + "step": 8872 + }, + { + "epoch": 0.9383460236886633, + "grad_norm": 7.6875, + "learning_rate": 4.681107163692247e-06, + "loss": 2.4921, + "step": 8873 + }, + { + "epoch": 0.9384517766497462, + "grad_norm": 7.96875, + "learning_rate": 4.681023462897964e-06, + "loss": 2.5181, + "step": 8874 + }, + { + "epoch": 0.938557529610829, + "grad_norm": 8.0625, + "learning_rate": 4.680939751869028e-06, + "loss": 2.6715, + "step": 8875 + }, + { + "epoch": 0.938663282571912, + "grad_norm": 7.8125, + "learning_rate": 4.6808560306058356e-06, + "loss": 2.4355, + "step": 8876 + }, + { + "epoch": 0.938769035532995, + "grad_norm": 7.90625, + "learning_rate": 4.6807722991087765e-06, + "loss": 2.4637, + "step": 8877 + }, + { + "epoch": 0.9388747884940778, + "grad_norm": 7.8125, + "learning_rate": 4.680688557378246e-06, + "loss": 2.5262, + "step": 8878 + }, + { + "epoch": 0.9389805414551607, + "grad_norm": 8.1875, + "learning_rate": 4.680604805414635e-06, + "loss": 2.6284, + "step": 8879 + }, + { + "epoch": 0.9390862944162437, + "grad_norm": 8.375, + "learning_rate": 4.680521043218339e-06, + "loss": 2.6721, + "step": 8880 + }, + { + "epoch": 0.9391920473773265, + "grad_norm": 7.84375, + "learning_rate": 4.680437270789748e-06, + "loss": 2.4562, + "step": 8881 + }, + { + "epoch": 0.9392978003384095, + "grad_norm": 7.71875, + "learning_rate": 4.680353488129257e-06, + "loss": 2.533, + "step": 8882 + }, + { + "epoch": 0.9394035532994924, + "grad_norm": 7.9375, + "learning_rate": 4.68026969523726e-06, + "loss": 2.6266, + "step": 8883 + }, + { + "epoch": 0.9395093062605753, + "grad_norm": 8.3125, + "learning_rate": 4.680185892114148e-06, + "loss": 2.8068, + "step": 8884 + }, + { + "epoch": 0.9396150592216582, + "grad_norm": 8.375, + "learning_rate": 4.680102078760315e-06, + "loss": 2.4362, + "step": 8885 + }, + { + "epoch": 0.9397208121827412, + "grad_norm": 8.3125, + "learning_rate": 4.680018255176155e-06, + "loss": 2.5825, + "step": 8886 + }, + { + "epoch": 0.939826565143824, + "grad_norm": 8.9375, + "learning_rate": 4.67993442136206e-06, + "loss": 2.8295, + "step": 8887 + }, + { + "epoch": 0.939932318104907, + "grad_norm": 8.0625, + "learning_rate": 4.6798505773184256e-06, + "loss": 2.5789, + "step": 8888 + }, + { + "epoch": 0.9400380710659898, + "grad_norm": 7.875, + "learning_rate": 4.679766723045643e-06, + "loss": 2.4148, + "step": 8889 + }, + { + "epoch": 0.9401438240270727, + "grad_norm": 9.0, + "learning_rate": 4.679682858544108e-06, + "loss": 2.7238, + "step": 8890 + }, + { + "epoch": 0.9402495769881557, + "grad_norm": 8.875, + "learning_rate": 4.6795989838142116e-06, + "loss": 2.9176, + "step": 8891 + }, + { + "epoch": 0.9403553299492385, + "grad_norm": 7.78125, + "learning_rate": 4.679515098856349e-06, + "loss": 2.6068, + "step": 8892 + }, + { + "epoch": 0.9404610829103215, + "grad_norm": 8.1875, + "learning_rate": 4.679431203670913e-06, + "loss": 2.5887, + "step": 8893 + }, + { + "epoch": 0.9405668358714044, + "grad_norm": 7.875, + "learning_rate": 4.679347298258299e-06, + "loss": 2.5757, + "step": 8894 + }, + { + "epoch": 0.9406725888324873, + "grad_norm": 8.4375, + "learning_rate": 4.679263382618898e-06, + "loss": 2.7253, + "step": 8895 + }, + { + "epoch": 0.9407783417935702, + "grad_norm": 7.75, + "learning_rate": 4.6791794567531055e-06, + "loss": 2.5063, + "step": 8896 + }, + { + "epoch": 0.9408840947546532, + "grad_norm": 8.875, + "learning_rate": 4.6790955206613155e-06, + "loss": 2.4683, + "step": 8897 + }, + { + "epoch": 0.940989847715736, + "grad_norm": 8.125, + "learning_rate": 4.679011574343922e-06, + "loss": 2.5736, + "step": 8898 + }, + { + "epoch": 0.941095600676819, + "grad_norm": 7.75, + "learning_rate": 4.678927617801318e-06, + "loss": 2.372, + "step": 8899 + }, + { + "epoch": 0.9412013536379019, + "grad_norm": 8.375, + "learning_rate": 4.678843651033897e-06, + "loss": 2.2303, + "step": 8900 + }, + { + "epoch": 0.9413071065989848, + "grad_norm": 8.0, + "learning_rate": 4.678759674042054e-06, + "loss": 2.5811, + "step": 8901 + }, + { + "epoch": 0.9414128595600677, + "grad_norm": 8.3125, + "learning_rate": 4.678675686826184e-06, + "loss": 2.3749, + "step": 8902 + }, + { + "epoch": 0.9415186125211505, + "grad_norm": 8.25, + "learning_rate": 4.678591689386679e-06, + "loss": 2.6136, + "step": 8903 + }, + { + "epoch": 0.9416243654822335, + "grad_norm": 8.125, + "learning_rate": 4.678507681723935e-06, + "loss": 2.3318, + "step": 8904 + }, + { + "epoch": 0.9417301184433164, + "grad_norm": 7.90625, + "learning_rate": 4.678423663838345e-06, + "loss": 2.5809, + "step": 8905 + }, + { + "epoch": 0.9418358714043993, + "grad_norm": 7.8125, + "learning_rate": 4.678339635730304e-06, + "loss": 2.5386, + "step": 8906 + }, + { + "epoch": 0.9419416243654822, + "grad_norm": 8.0, + "learning_rate": 4.678255597400206e-06, + "loss": 2.482, + "step": 8907 + }, + { + "epoch": 0.9420473773265652, + "grad_norm": 9.0, + "learning_rate": 4.678171548848444e-06, + "loss": 2.8132, + "step": 8908 + }, + { + "epoch": 0.942153130287648, + "grad_norm": 7.8125, + "learning_rate": 4.678087490075415e-06, + "loss": 2.6076, + "step": 8909 + }, + { + "epoch": 0.942258883248731, + "grad_norm": 8.125, + "learning_rate": 4.678003421081512e-06, + "loss": 2.5511, + "step": 8910 + }, + { + "epoch": 0.9423646362098139, + "grad_norm": 7.9375, + "learning_rate": 4.67791934186713e-06, + "loss": 2.5473, + "step": 8911 + }, + { + "epoch": 0.9424703891708968, + "grad_norm": 8.25, + "learning_rate": 4.6778352524326635e-06, + "loss": 2.6091, + "step": 8912 + }, + { + "epoch": 0.9425761421319797, + "grad_norm": 8.375, + "learning_rate": 4.677751152778506e-06, + "loss": 2.6623, + "step": 8913 + }, + { + "epoch": 0.9426818950930627, + "grad_norm": 9.0625, + "learning_rate": 4.677667042905053e-06, + "loss": 2.622, + "step": 8914 + }, + { + "epoch": 0.9427876480541455, + "grad_norm": 8.0625, + "learning_rate": 4.6775829228127e-06, + "loss": 2.5591, + "step": 8915 + }, + { + "epoch": 0.9428934010152284, + "grad_norm": 7.78125, + "learning_rate": 4.67749879250184e-06, + "loss": 2.6153, + "step": 8916 + }, + { + "epoch": 0.9429991539763113, + "grad_norm": 7.9375, + "learning_rate": 4.67741465197287e-06, + "loss": 2.6612, + "step": 8917 + }, + { + "epoch": 0.9431049069373942, + "grad_norm": 7.84375, + "learning_rate": 4.677330501226183e-06, + "loss": 2.5397, + "step": 8918 + }, + { + "epoch": 0.9432106598984772, + "grad_norm": 8.3125, + "learning_rate": 4.6772463402621735e-06, + "loss": 2.684, + "step": 8919 + }, + { + "epoch": 0.94331641285956, + "grad_norm": 8.25, + "learning_rate": 4.677162169081238e-06, + "loss": 2.4909, + "step": 8920 + }, + { + "epoch": 0.943422165820643, + "grad_norm": 8.25, + "learning_rate": 4.677077987683771e-06, + "loss": 2.5174, + "step": 8921 + }, + { + "epoch": 0.9435279187817259, + "grad_norm": 8.0625, + "learning_rate": 4.676993796070168e-06, + "loss": 2.4971, + "step": 8922 + }, + { + "epoch": 0.9436336717428088, + "grad_norm": 8.0625, + "learning_rate": 4.676909594240822e-06, + "loss": 2.5006, + "step": 8923 + }, + { + "epoch": 0.9437394247038917, + "grad_norm": 8.375, + "learning_rate": 4.676825382196132e-06, + "loss": 2.638, + "step": 8924 + }, + { + "epoch": 0.9438451776649747, + "grad_norm": 8.6875, + "learning_rate": 4.6767411599364885e-06, + "loss": 2.5441, + "step": 8925 + }, + { + "epoch": 0.9439509306260575, + "grad_norm": 8.0625, + "learning_rate": 4.67665692746229e-06, + "loss": 2.5183, + "step": 8926 + }, + { + "epoch": 0.9440566835871405, + "grad_norm": 8.1875, + "learning_rate": 4.6765726847739305e-06, + "loss": 2.6684, + "step": 8927 + }, + { + "epoch": 0.9441624365482234, + "grad_norm": 7.96875, + "learning_rate": 4.676488431871806e-06, + "loss": 2.6575, + "step": 8928 + }, + { + "epoch": 0.9442681895093062, + "grad_norm": 7.90625, + "learning_rate": 4.676404168756311e-06, + "loss": 2.6177, + "step": 8929 + }, + { + "epoch": 0.9443739424703892, + "grad_norm": 8.0, + "learning_rate": 4.676319895427841e-06, + "loss": 2.6518, + "step": 8930 + }, + { + "epoch": 0.944479695431472, + "grad_norm": 7.84375, + "learning_rate": 4.676235611886793e-06, + "loss": 2.3728, + "step": 8931 + }, + { + "epoch": 0.944585448392555, + "grad_norm": 7.75, + "learning_rate": 4.67615131813356e-06, + "loss": 2.6221, + "step": 8932 + }, + { + "epoch": 0.9446912013536379, + "grad_norm": 8.3125, + "learning_rate": 4.676067014168539e-06, + "loss": 2.6031, + "step": 8933 + }, + { + "epoch": 0.9447969543147208, + "grad_norm": 8.125, + "learning_rate": 4.675982699992126e-06, + "loss": 2.5918, + "step": 8934 + }, + { + "epoch": 0.9449027072758037, + "grad_norm": 7.90625, + "learning_rate": 4.6758983756047164e-06, + "loss": 2.2656, + "step": 8935 + }, + { + "epoch": 0.9450084602368867, + "grad_norm": 7.9375, + "learning_rate": 4.675814041006706e-06, + "loss": 2.5978, + "step": 8936 + }, + { + "epoch": 0.9451142131979695, + "grad_norm": 8.0625, + "learning_rate": 4.675729696198489e-06, + "loss": 2.5752, + "step": 8937 + }, + { + "epoch": 0.9452199661590525, + "grad_norm": 7.75, + "learning_rate": 4.675645341180463e-06, + "loss": 2.5866, + "step": 8938 + }, + { + "epoch": 0.9453257191201354, + "grad_norm": 8.1875, + "learning_rate": 4.675560975953023e-06, + "loss": 2.4958, + "step": 8939 + }, + { + "epoch": 0.9454314720812182, + "grad_norm": 8.25, + "learning_rate": 4.675476600516565e-06, + "loss": 2.7228, + "step": 8940 + }, + { + "epoch": 0.9455372250423012, + "grad_norm": 8.625, + "learning_rate": 4.675392214871486e-06, + "loss": 2.5877, + "step": 8941 + }, + { + "epoch": 0.9456429780033841, + "grad_norm": 8.1875, + "learning_rate": 4.67530781901818e-06, + "loss": 2.6636, + "step": 8942 + }, + { + "epoch": 0.945748730964467, + "grad_norm": 7.9375, + "learning_rate": 4.675223412957044e-06, + "loss": 2.6524, + "step": 8943 + }, + { + "epoch": 0.9458544839255499, + "grad_norm": 8.0625, + "learning_rate": 4.675138996688475e-06, + "loss": 2.7273, + "step": 8944 + }, + { + "epoch": 0.9459602368866328, + "grad_norm": 8.5625, + "learning_rate": 4.675054570212868e-06, + "loss": 2.6276, + "step": 8945 + }, + { + "epoch": 0.9460659898477157, + "grad_norm": 8.6875, + "learning_rate": 4.674970133530619e-06, + "loss": 2.524, + "step": 8946 + }, + { + "epoch": 0.9461717428087987, + "grad_norm": 7.8125, + "learning_rate": 4.6748856866421245e-06, + "loss": 2.5872, + "step": 8947 + }, + { + "epoch": 0.9462774957698815, + "grad_norm": 8.0625, + "learning_rate": 4.674801229547782e-06, + "loss": 2.5344, + "step": 8948 + }, + { + "epoch": 0.9463832487309645, + "grad_norm": 9.0625, + "learning_rate": 4.674716762247986e-06, + "loss": 2.6218, + "step": 8949 + }, + { + "epoch": 0.9464890016920474, + "grad_norm": 7.53125, + "learning_rate": 4.674632284743134e-06, + "loss": 2.5245, + "step": 8950 + }, + { + "epoch": 0.9465947546531303, + "grad_norm": 8.75, + "learning_rate": 4.674547797033622e-06, + "loss": 2.7274, + "step": 8951 + }, + { + "epoch": 0.9467005076142132, + "grad_norm": 8.1875, + "learning_rate": 4.6744632991198465e-06, + "loss": 2.459, + "step": 8952 + }, + { + "epoch": 0.9468062605752962, + "grad_norm": 7.875, + "learning_rate": 4.674378791002204e-06, + "loss": 2.5564, + "step": 8953 + }, + { + "epoch": 0.946912013536379, + "grad_norm": 8.4375, + "learning_rate": 4.674294272681091e-06, + "loss": 2.6692, + "step": 8954 + }, + { + "epoch": 0.9470177664974619, + "grad_norm": 8.0625, + "learning_rate": 4.674209744156905e-06, + "loss": 2.6229, + "step": 8955 + }, + { + "epoch": 0.9471235194585449, + "grad_norm": 7.96875, + "learning_rate": 4.674125205430041e-06, + "loss": 2.7187, + "step": 8956 + }, + { + "epoch": 0.9472292724196277, + "grad_norm": 7.625, + "learning_rate": 4.674040656500898e-06, + "loss": 2.443, + "step": 8957 + }, + { + "epoch": 0.9473350253807107, + "grad_norm": 7.71875, + "learning_rate": 4.67395609736987e-06, + "loss": 2.5297, + "step": 8958 + }, + { + "epoch": 0.9474407783417935, + "grad_norm": 8.125, + "learning_rate": 4.6738715280373545e-06, + "loss": 2.3977, + "step": 8959 + }, + { + "epoch": 0.9475465313028765, + "grad_norm": 7.5625, + "learning_rate": 4.6737869485037505e-06, + "loss": 2.3463, + "step": 8960 + }, + { + "epoch": 0.9476522842639594, + "grad_norm": 7.625, + "learning_rate": 4.673702358769453e-06, + "loss": 2.4543, + "step": 8961 + }, + { + "epoch": 0.9477580372250423, + "grad_norm": 8.3125, + "learning_rate": 4.673617758834859e-06, + "loss": 2.5269, + "step": 8962 + }, + { + "epoch": 0.9478637901861252, + "grad_norm": 7.84375, + "learning_rate": 4.6735331487003664e-06, + "loss": 2.5768, + "step": 8963 + }, + { + "epoch": 0.9479695431472082, + "grad_norm": 7.8125, + "learning_rate": 4.673448528366371e-06, + "loss": 2.6842, + "step": 8964 + }, + { + "epoch": 0.948075296108291, + "grad_norm": 8.25, + "learning_rate": 4.673363897833271e-06, + "loss": 2.6101, + "step": 8965 + }, + { + "epoch": 0.948181049069374, + "grad_norm": 8.1875, + "learning_rate": 4.673279257101463e-06, + "loss": 2.575, + "step": 8966 + }, + { + "epoch": 0.9482868020304569, + "grad_norm": 9.0, + "learning_rate": 4.673194606171344e-06, + "loss": 2.4307, + "step": 8967 + }, + { + "epoch": 0.9483925549915397, + "grad_norm": 8.75, + "learning_rate": 4.673109945043312e-06, + "loss": 2.4982, + "step": 8968 + }, + { + "epoch": 0.9484983079526227, + "grad_norm": 8.3125, + "learning_rate": 4.6730252737177635e-06, + "loss": 2.6669, + "step": 8969 + }, + { + "epoch": 0.9486040609137056, + "grad_norm": 7.96875, + "learning_rate": 4.672940592195097e-06, + "loss": 2.5742, + "step": 8970 + }, + { + "epoch": 0.9487098138747885, + "grad_norm": 8.375, + "learning_rate": 4.6728559004757076e-06, + "loss": 2.3592, + "step": 8971 + }, + { + "epoch": 0.9488155668358714, + "grad_norm": 8.1875, + "learning_rate": 4.672771198559995e-06, + "loss": 2.5809, + "step": 8972 + }, + { + "epoch": 0.9489213197969543, + "grad_norm": 8.625, + "learning_rate": 4.672686486448355e-06, + "loss": 2.7421, + "step": 8973 + }, + { + "epoch": 0.9490270727580372, + "grad_norm": 8.5625, + "learning_rate": 4.672601764141187e-06, + "loss": 2.6664, + "step": 8974 + }, + { + "epoch": 0.9491328257191202, + "grad_norm": 8.3125, + "learning_rate": 4.672517031638887e-06, + "loss": 2.5982, + "step": 8975 + }, + { + "epoch": 0.949238578680203, + "grad_norm": 8.375, + "learning_rate": 4.672432288941853e-06, + "loss": 2.5808, + "step": 8976 + }, + { + "epoch": 0.949344331641286, + "grad_norm": 7.84375, + "learning_rate": 4.672347536050483e-06, + "loss": 2.5065, + "step": 8977 + }, + { + "epoch": 0.9494500846023689, + "grad_norm": 7.625, + "learning_rate": 4.672262772965174e-06, + "loss": 2.562, + "step": 8978 + }, + { + "epoch": 0.9495558375634517, + "grad_norm": 8.0625, + "learning_rate": 4.672177999686325e-06, + "loss": 2.5154, + "step": 8979 + }, + { + "epoch": 0.9496615905245347, + "grad_norm": 8.4375, + "learning_rate": 4.672093216214332e-06, + "loss": 2.7369, + "step": 8980 + }, + { + "epoch": 0.9497673434856176, + "grad_norm": 7.625, + "learning_rate": 4.6720084225495945e-06, + "loss": 2.4929, + "step": 8981 + }, + { + "epoch": 0.9498730964467005, + "grad_norm": 8.6875, + "learning_rate": 4.67192361869251e-06, + "loss": 2.6049, + "step": 8982 + }, + { + "epoch": 0.9499788494077834, + "grad_norm": 7.8125, + "learning_rate": 4.671838804643476e-06, + "loss": 2.5227, + "step": 8983 + }, + { + "epoch": 0.9500846023688664, + "grad_norm": 7.96875, + "learning_rate": 4.671753980402891e-06, + "loss": 2.6274, + "step": 8984 + }, + { + "epoch": 0.9501903553299492, + "grad_norm": 8.0, + "learning_rate": 4.671669145971152e-06, + "loss": 2.5372, + "step": 8985 + }, + { + "epoch": 0.9502961082910322, + "grad_norm": 8.125, + "learning_rate": 4.671584301348658e-06, + "loss": 2.5921, + "step": 8986 + }, + { + "epoch": 0.950401861252115, + "grad_norm": 7.875, + "learning_rate": 4.671499446535808e-06, + "loss": 2.6296, + "step": 8987 + }, + { + "epoch": 0.950507614213198, + "grad_norm": 8.0625, + "learning_rate": 4.671414581532999e-06, + "loss": 2.7377, + "step": 8988 + }, + { + "epoch": 0.9506133671742809, + "grad_norm": 7.59375, + "learning_rate": 4.671329706340629e-06, + "loss": 2.694, + "step": 8989 + }, + { + "epoch": 0.9507191201353637, + "grad_norm": 7.78125, + "learning_rate": 4.671244820959097e-06, + "loss": 2.5216, + "step": 8990 + }, + { + "epoch": 0.9508248730964467, + "grad_norm": 7.5, + "learning_rate": 4.6711599253888005e-06, + "loss": 2.6088, + "step": 8991 + }, + { + "epoch": 0.9509306260575296, + "grad_norm": 8.5, + "learning_rate": 4.671075019630139e-06, + "loss": 2.5884, + "step": 8992 + }, + { + "epoch": 0.9510363790186125, + "grad_norm": 8.625, + "learning_rate": 4.670990103683511e-06, + "loss": 2.6896, + "step": 8993 + }, + { + "epoch": 0.9511421319796954, + "grad_norm": 7.71875, + "learning_rate": 4.670905177549313e-06, + "loss": 2.5599, + "step": 8994 + }, + { + "epoch": 0.9512478849407784, + "grad_norm": 8.125, + "learning_rate": 4.6708202412279445e-06, + "loss": 2.5225, + "step": 8995 + }, + { + "epoch": 0.9513536379018612, + "grad_norm": 8.6875, + "learning_rate": 4.6707352947198056e-06, + "loss": 2.5761, + "step": 8996 + }, + { + "epoch": 0.9514593908629442, + "grad_norm": 8.5625, + "learning_rate": 4.670650338025294e-06, + "loss": 2.6872, + "step": 8997 + }, + { + "epoch": 0.9515651438240271, + "grad_norm": 7.8125, + "learning_rate": 4.6705653711448076e-06, + "loss": 2.6486, + "step": 8998 + }, + { + "epoch": 0.95167089678511, + "grad_norm": 8.3125, + "learning_rate": 4.6704803940787455e-06, + "loss": 2.6839, + "step": 8999 + }, + { + "epoch": 0.9517766497461929, + "grad_norm": 7.6875, + "learning_rate": 4.670395406827507e-06, + "loss": 2.5535, + "step": 9000 + }, + { + "epoch": 0.9518824027072758, + "grad_norm": 7.625, + "learning_rate": 4.67031040939149e-06, + "loss": 2.6319, + "step": 9001 + }, + { + "epoch": 0.9519881556683587, + "grad_norm": 8.3125, + "learning_rate": 4.670225401771095e-06, + "loss": 2.6771, + "step": 9002 + }, + { + "epoch": 0.9520939086294417, + "grad_norm": 7.9375, + "learning_rate": 4.670140383966718e-06, + "loss": 2.3262, + "step": 9003 + }, + { + "epoch": 0.9521996615905245, + "grad_norm": 7.84375, + "learning_rate": 4.670055355978761e-06, + "loss": 2.4538, + "step": 9004 + }, + { + "epoch": 0.9523054145516074, + "grad_norm": 7.71875, + "learning_rate": 4.669970317807621e-06, + "loss": 2.5287, + "step": 9005 + }, + { + "epoch": 0.9524111675126904, + "grad_norm": 8.25, + "learning_rate": 4.669885269453699e-06, + "loss": 2.6049, + "step": 9006 + }, + { + "epoch": 0.9525169204737732, + "grad_norm": 7.875, + "learning_rate": 4.669800210917392e-06, + "loss": 2.4976, + "step": 9007 + }, + { + "epoch": 0.9526226734348562, + "grad_norm": 8.25, + "learning_rate": 4.6697151421991e-06, + "loss": 2.5365, + "step": 9008 + }, + { + "epoch": 0.9527284263959391, + "grad_norm": 8.1875, + "learning_rate": 4.6696300632992225e-06, + "loss": 2.5764, + "step": 9009 + }, + { + "epoch": 0.952834179357022, + "grad_norm": 8.5625, + "learning_rate": 4.669544974218159e-06, + "loss": 2.7441, + "step": 9010 + }, + { + "epoch": 0.9529399323181049, + "grad_norm": 8.0625, + "learning_rate": 4.669459874956307e-06, + "loss": 2.7248, + "step": 9011 + }, + { + "epoch": 0.9530456852791879, + "grad_norm": 7.59375, + "learning_rate": 4.669374765514067e-06, + "loss": 2.6191, + "step": 9012 + }, + { + "epoch": 0.9531514382402707, + "grad_norm": 7.9375, + "learning_rate": 4.669289645891839e-06, + "loss": 2.5704, + "step": 9013 + }, + { + "epoch": 0.9532571912013537, + "grad_norm": 8.0, + "learning_rate": 4.669204516090023e-06, + "loss": 2.3734, + "step": 9014 + }, + { + "epoch": 0.9533629441624365, + "grad_norm": 8.25, + "learning_rate": 4.6691193761090156e-06, + "loss": 2.7115, + "step": 9015 + }, + { + "epoch": 0.9534686971235194, + "grad_norm": 8.0625, + "learning_rate": 4.6690342259492195e-06, + "loss": 2.2315, + "step": 9016 + }, + { + "epoch": 0.9535744500846024, + "grad_norm": 7.6875, + "learning_rate": 4.6689490656110325e-06, + "loss": 2.4603, + "step": 9017 + }, + { + "epoch": 0.9536802030456852, + "grad_norm": 7.65625, + "learning_rate": 4.668863895094854e-06, + "loss": 2.4567, + "step": 9018 + }, + { + "epoch": 0.9537859560067682, + "grad_norm": 7.96875, + "learning_rate": 4.668778714401085e-06, + "loss": 2.6081, + "step": 9019 + }, + { + "epoch": 0.9538917089678511, + "grad_norm": 7.625, + "learning_rate": 4.668693523530124e-06, + "loss": 2.3977, + "step": 9020 + }, + { + "epoch": 0.953997461928934, + "grad_norm": 8.125, + "learning_rate": 4.668608322482372e-06, + "loss": 2.6653, + "step": 9021 + }, + { + "epoch": 0.9541032148900169, + "grad_norm": 8.25, + "learning_rate": 4.668523111258227e-06, + "loss": 2.6819, + "step": 9022 + }, + { + "epoch": 0.9542089678510999, + "grad_norm": 7.5625, + "learning_rate": 4.668437889858091e-06, + "loss": 2.4368, + "step": 9023 + }, + { + "epoch": 0.9543147208121827, + "grad_norm": 8.5625, + "learning_rate": 4.668352658282361e-06, + "loss": 2.8137, + "step": 9024 + }, + { + "epoch": 0.9544204737732657, + "grad_norm": 7.8125, + "learning_rate": 4.6682674165314406e-06, + "loss": 2.7675, + "step": 9025 + }, + { + "epoch": 0.9545262267343486, + "grad_norm": 8.1875, + "learning_rate": 4.668182164605728e-06, + "loss": 2.5832, + "step": 9026 + }, + { + "epoch": 0.9546319796954315, + "grad_norm": 7.78125, + "learning_rate": 4.668096902505622e-06, + "loss": 2.4603, + "step": 9027 + }, + { + "epoch": 0.9547377326565144, + "grad_norm": 8.125, + "learning_rate": 4.668011630231525e-06, + "loss": 2.5061, + "step": 9028 + }, + { + "epoch": 0.9548434856175972, + "grad_norm": 7.84375, + "learning_rate": 4.667926347783836e-06, + "loss": 2.4491, + "step": 9029 + }, + { + "epoch": 0.9549492385786802, + "grad_norm": 7.9375, + "learning_rate": 4.667841055162955e-06, + "loss": 2.4737, + "step": 9030 + }, + { + "epoch": 0.9550549915397631, + "grad_norm": 7.90625, + "learning_rate": 4.667755752369283e-06, + "loss": 2.5808, + "step": 9031 + }, + { + "epoch": 0.955160744500846, + "grad_norm": 8.1875, + "learning_rate": 4.667670439403219e-06, + "loss": 2.6422, + "step": 9032 + }, + { + "epoch": 0.9552664974619289, + "grad_norm": 8.25, + "learning_rate": 4.667585116265165e-06, + "loss": 2.1237, + "step": 9033 + }, + { + "epoch": 0.9553722504230119, + "grad_norm": 7.71875, + "learning_rate": 4.66749978295552e-06, + "loss": 2.4478, + "step": 9034 + }, + { + "epoch": 0.9554780033840947, + "grad_norm": 7.9375, + "learning_rate": 4.667414439474686e-06, + "loss": 2.6227, + "step": 9035 + }, + { + "epoch": 0.9555837563451777, + "grad_norm": 7.9375, + "learning_rate": 4.667329085823062e-06, + "loss": 2.4813, + "step": 9036 + }, + { + "epoch": 0.9556895093062606, + "grad_norm": 8.1875, + "learning_rate": 4.667243722001048e-06, + "loss": 2.8422, + "step": 9037 + }, + { + "epoch": 0.9557952622673435, + "grad_norm": 8.1875, + "learning_rate": 4.667158348009046e-06, + "loss": 2.7249, + "step": 9038 + }, + { + "epoch": 0.9559010152284264, + "grad_norm": 8.125, + "learning_rate": 4.6670729638474565e-06, + "loss": 2.5472, + "step": 9039 + }, + { + "epoch": 0.9560067681895094, + "grad_norm": 8.0, + "learning_rate": 4.66698756951668e-06, + "loss": 2.6652, + "step": 9040 + }, + { + "epoch": 0.9561125211505922, + "grad_norm": 7.78125, + "learning_rate": 4.666902165017117e-06, + "loss": 2.5199, + "step": 9041 + }, + { + "epoch": 0.9562182741116751, + "grad_norm": 9.0, + "learning_rate": 4.666816750349168e-06, + "loss": 2.6192, + "step": 9042 + }, + { + "epoch": 0.956324027072758, + "grad_norm": 8.75, + "learning_rate": 4.666731325513235e-06, + "loss": 2.519, + "step": 9043 + }, + { + "epoch": 0.9564297800338409, + "grad_norm": 8.375, + "learning_rate": 4.666645890509718e-06, + "loss": 2.4909, + "step": 9044 + }, + { + "epoch": 0.9565355329949239, + "grad_norm": 7.34375, + "learning_rate": 4.666560445339016e-06, + "loss": 2.5342, + "step": 9045 + }, + { + "epoch": 0.9566412859560067, + "grad_norm": 7.5, + "learning_rate": 4.666474990001534e-06, + "loss": 2.4589, + "step": 9046 + }, + { + "epoch": 0.9567470389170897, + "grad_norm": 7.65625, + "learning_rate": 4.666389524497671e-06, + "loss": 2.5059, + "step": 9047 + }, + { + "epoch": 0.9568527918781726, + "grad_norm": 8.5, + "learning_rate": 4.666304048827828e-06, + "loss": 2.6246, + "step": 9048 + }, + { + "epoch": 0.9569585448392555, + "grad_norm": 8.5625, + "learning_rate": 4.666218562992405e-06, + "loss": 2.6206, + "step": 9049 + }, + { + "epoch": 0.9570642978003384, + "grad_norm": 8.5, + "learning_rate": 4.666133066991805e-06, + "loss": 2.6062, + "step": 9050 + }, + { + "epoch": 0.9571700507614214, + "grad_norm": 7.625, + "learning_rate": 4.666047560826428e-06, + "loss": 2.551, + "step": 9051 + }, + { + "epoch": 0.9572758037225042, + "grad_norm": 8.6875, + "learning_rate": 4.665962044496675e-06, + "loss": 2.6988, + "step": 9052 + }, + { + "epoch": 0.9573815566835872, + "grad_norm": 8.25, + "learning_rate": 4.66587651800295e-06, + "loss": 2.6251, + "step": 9053 + }, + { + "epoch": 0.9574873096446701, + "grad_norm": 8.5625, + "learning_rate": 4.66579098134565e-06, + "loss": 2.9423, + "step": 9054 + }, + { + "epoch": 0.9575930626057529, + "grad_norm": 7.21875, + "learning_rate": 4.665705434525181e-06, + "loss": 2.5896, + "step": 9055 + }, + { + "epoch": 0.9576988155668359, + "grad_norm": 8.75, + "learning_rate": 4.66561987754194e-06, + "loss": 2.6597, + "step": 9056 + }, + { + "epoch": 0.9578045685279187, + "grad_norm": 7.84375, + "learning_rate": 4.6655343103963324e-06, + "loss": 2.5784, + "step": 9057 + }, + { + "epoch": 0.9579103214890017, + "grad_norm": 7.78125, + "learning_rate": 4.665448733088756e-06, + "loss": 2.6855, + "step": 9058 + }, + { + "epoch": 0.9580160744500846, + "grad_norm": 7.96875, + "learning_rate": 4.665363145619616e-06, + "loss": 2.4837, + "step": 9059 + }, + { + "epoch": 0.9581218274111675, + "grad_norm": 7.46875, + "learning_rate": 4.665277547989312e-06, + "loss": 2.3696, + "step": 9060 + }, + { + "epoch": 0.9582275803722504, + "grad_norm": 8.875, + "learning_rate": 4.665191940198246e-06, + "loss": 2.7556, + "step": 9061 + }, + { + "epoch": 0.9583333333333334, + "grad_norm": 7.9375, + "learning_rate": 4.66510632224682e-06, + "loss": 2.3956, + "step": 9062 + }, + { + "epoch": 0.9584390862944162, + "grad_norm": 8.3125, + "learning_rate": 4.665020694135435e-06, + "loss": 2.4342, + "step": 9063 + }, + { + "epoch": 0.9585448392554992, + "grad_norm": 8.3125, + "learning_rate": 4.6649350558644935e-06, + "loss": 2.5406, + "step": 9064 + }, + { + "epoch": 0.9586505922165821, + "grad_norm": 8.4375, + "learning_rate": 4.664849407434397e-06, + "loss": 2.6874, + "step": 9065 + }, + { + "epoch": 0.958756345177665, + "grad_norm": 7.84375, + "learning_rate": 4.664763748845548e-06, + "loss": 2.5266, + "step": 9066 + }, + { + "epoch": 0.9588620981387479, + "grad_norm": 7.9375, + "learning_rate": 4.664678080098348e-06, + "loss": 2.5961, + "step": 9067 + }, + { + "epoch": 0.9589678510998308, + "grad_norm": 7.78125, + "learning_rate": 4.664592401193199e-06, + "loss": 2.7281, + "step": 9068 + }, + { + "epoch": 0.9590736040609137, + "grad_norm": 8.125, + "learning_rate": 4.664506712130503e-06, + "loss": 2.3863, + "step": 9069 + }, + { + "epoch": 0.9591793570219966, + "grad_norm": 8.25, + "learning_rate": 4.664421012910662e-06, + "loss": 2.6277, + "step": 9070 + }, + { + "epoch": 0.9592851099830795, + "grad_norm": 8.0625, + "learning_rate": 4.66433530353408e-06, + "loss": 2.3128, + "step": 9071 + }, + { + "epoch": 0.9593908629441624, + "grad_norm": 7.84375, + "learning_rate": 4.664249584001156e-06, + "loss": 2.4402, + "step": 9072 + }, + { + "epoch": 0.9594966159052454, + "grad_norm": 7.75, + "learning_rate": 4.664163854312294e-06, + "loss": 2.6589, + "step": 9073 + }, + { + "epoch": 0.9596023688663282, + "grad_norm": 8.8125, + "learning_rate": 4.664078114467897e-06, + "loss": 2.5485, + "step": 9074 + }, + { + "epoch": 0.9597081218274112, + "grad_norm": 8.0, + "learning_rate": 4.663992364468366e-06, + "loss": 2.5793, + "step": 9075 + }, + { + "epoch": 0.9598138747884941, + "grad_norm": 8.0625, + "learning_rate": 4.663906604314104e-06, + "loss": 2.5658, + "step": 9076 + }, + { + "epoch": 0.959919627749577, + "grad_norm": 8.4375, + "learning_rate": 4.663820834005513e-06, + "loss": 2.5349, + "step": 9077 + }, + { + "epoch": 0.9600253807106599, + "grad_norm": 7.9375, + "learning_rate": 4.6637350535429955e-06, + "loss": 2.6269, + "step": 9078 + }, + { + "epoch": 0.9601311336717429, + "grad_norm": 8.75, + "learning_rate": 4.663649262926955e-06, + "loss": 2.7577, + "step": 9079 + }, + { + "epoch": 0.9602368866328257, + "grad_norm": 8.875, + "learning_rate": 4.663563462157793e-06, + "loss": 2.6108, + "step": 9080 + }, + { + "epoch": 0.9603426395939086, + "grad_norm": 7.71875, + "learning_rate": 4.663477651235913e-06, + "loss": 2.5156, + "step": 9081 + }, + { + "epoch": 0.9604483925549916, + "grad_norm": 8.25, + "learning_rate": 4.663391830161717e-06, + "loss": 2.4478, + "step": 9082 + }, + { + "epoch": 0.9605541455160744, + "grad_norm": 7.6875, + "learning_rate": 4.663305998935607e-06, + "loss": 2.4968, + "step": 9083 + }, + { + "epoch": 0.9606598984771574, + "grad_norm": 11.1875, + "learning_rate": 4.663220157557988e-06, + "loss": 2.4733, + "step": 9084 + }, + { + "epoch": 0.9607656514382402, + "grad_norm": 7.625, + "learning_rate": 4.663134306029261e-06, + "loss": 2.4998, + "step": 9085 + }, + { + "epoch": 0.9608714043993232, + "grad_norm": 8.375, + "learning_rate": 4.663048444349829e-06, + "loss": 2.5383, + "step": 9086 + }, + { + "epoch": 0.9609771573604061, + "grad_norm": 7.4375, + "learning_rate": 4.662962572520096e-06, + "loss": 2.5033, + "step": 9087 + }, + { + "epoch": 0.961082910321489, + "grad_norm": 8.0625, + "learning_rate": 4.662876690540464e-06, + "loss": 2.7328, + "step": 9088 + }, + { + "epoch": 0.9611886632825719, + "grad_norm": 7.90625, + "learning_rate": 4.6627907984113365e-06, + "loss": 2.6199, + "step": 9089 + }, + { + "epoch": 0.9612944162436549, + "grad_norm": 7.34375, + "learning_rate": 4.6627048961331165e-06, + "loss": 2.3623, + "step": 9090 + }, + { + "epoch": 0.9614001692047377, + "grad_norm": 7.8125, + "learning_rate": 4.662618983706206e-06, + "loss": 2.6205, + "step": 9091 + }, + { + "epoch": 0.9615059221658206, + "grad_norm": 8.1875, + "learning_rate": 4.66253306113101e-06, + "loss": 2.4974, + "step": 9092 + }, + { + "epoch": 0.9616116751269036, + "grad_norm": 7.46875, + "learning_rate": 4.6624471284079305e-06, + "loss": 2.5851, + "step": 9093 + }, + { + "epoch": 0.9617174280879864, + "grad_norm": 7.78125, + "learning_rate": 4.662361185537371e-06, + "loss": 2.7321, + "step": 9094 + }, + { + "epoch": 0.9618231810490694, + "grad_norm": 7.78125, + "learning_rate": 4.662275232519735e-06, + "loss": 2.6064, + "step": 9095 + }, + { + "epoch": 0.9619289340101523, + "grad_norm": 7.53125, + "learning_rate": 4.662189269355425e-06, + "loss": 2.3989, + "step": 9096 + }, + { + "epoch": 0.9620346869712352, + "grad_norm": 7.8125, + "learning_rate": 4.662103296044846e-06, + "loss": 2.5196, + "step": 9097 + }, + { + "epoch": 0.9621404399323181, + "grad_norm": 7.6875, + "learning_rate": 4.6620173125884005e-06, + "loss": 2.5926, + "step": 9098 + }, + { + "epoch": 0.962246192893401, + "grad_norm": 7.90625, + "learning_rate": 4.6619313189864915e-06, + "loss": 2.528, + "step": 9099 + }, + { + "epoch": 0.9623519458544839, + "grad_norm": 8.0, + "learning_rate": 4.6618453152395224e-06, + "loss": 2.4092, + "step": 9100 + }, + { + "epoch": 0.9624576988155669, + "grad_norm": 8.1875, + "learning_rate": 4.661759301347898e-06, + "loss": 2.6607, + "step": 9101 + }, + { + "epoch": 0.9625634517766497, + "grad_norm": 7.84375, + "learning_rate": 4.6616732773120225e-06, + "loss": 2.5503, + "step": 9102 + }, + { + "epoch": 0.9626692047377327, + "grad_norm": 7.625, + "learning_rate": 4.661587243132296e-06, + "loss": 2.6087, + "step": 9103 + }, + { + "epoch": 0.9627749576988156, + "grad_norm": 8.0, + "learning_rate": 4.661501198809126e-06, + "loss": 2.4381, + "step": 9104 + }, + { + "epoch": 0.9628807106598984, + "grad_norm": 8.0, + "learning_rate": 4.661415144342916e-06, + "loss": 2.6362, + "step": 9105 + }, + { + "epoch": 0.9629864636209814, + "grad_norm": 8.1875, + "learning_rate": 4.6613290797340674e-06, + "loss": 2.532, + "step": 9106 + }, + { + "epoch": 0.9630922165820643, + "grad_norm": 7.5625, + "learning_rate": 4.661243004982985e-06, + "loss": 2.3481, + "step": 9107 + }, + { + "epoch": 0.9631979695431472, + "grad_norm": 7.875, + "learning_rate": 4.661156920090074e-06, + "loss": 2.5198, + "step": 9108 + }, + { + "epoch": 0.9633037225042301, + "grad_norm": 8.25, + "learning_rate": 4.661070825055737e-06, + "loss": 2.6016, + "step": 9109 + }, + { + "epoch": 0.9634094754653131, + "grad_norm": 7.78125, + "learning_rate": 4.660984719880379e-06, + "loss": 2.6744, + "step": 9110 + }, + { + "epoch": 0.9635152284263959, + "grad_norm": 7.5625, + "learning_rate": 4.660898604564403e-06, + "loss": 2.4489, + "step": 9111 + }, + { + "epoch": 0.9636209813874789, + "grad_norm": 7.78125, + "learning_rate": 4.660812479108213e-06, + "loss": 2.5495, + "step": 9112 + }, + { + "epoch": 0.9637267343485617, + "grad_norm": 8.0625, + "learning_rate": 4.660726343512215e-06, + "loss": 2.4536, + "step": 9113 + }, + { + "epoch": 0.9638324873096447, + "grad_norm": 7.78125, + "learning_rate": 4.660640197776811e-06, + "loss": 2.5453, + "step": 9114 + }, + { + "epoch": 0.9639382402707276, + "grad_norm": 8.25, + "learning_rate": 4.660554041902407e-06, + "loss": 2.5844, + "step": 9115 + }, + { + "epoch": 0.9640439932318104, + "grad_norm": 7.75, + "learning_rate": 4.660467875889406e-06, + "loss": 2.3645, + "step": 9116 + }, + { + "epoch": 0.9641497461928934, + "grad_norm": 7.96875, + "learning_rate": 4.660381699738213e-06, + "loss": 2.5012, + "step": 9117 + }, + { + "epoch": 0.9642554991539763, + "grad_norm": 7.46875, + "learning_rate": 4.660295513449233e-06, + "loss": 2.0633, + "step": 9118 + }, + { + "epoch": 0.9643612521150592, + "grad_norm": 8.5625, + "learning_rate": 4.660209317022869e-06, + "loss": 2.5961, + "step": 9119 + }, + { + "epoch": 0.9644670050761421, + "grad_norm": 7.96875, + "learning_rate": 4.660123110459525e-06, + "loss": 2.6845, + "step": 9120 + }, + { + "epoch": 0.9645727580372251, + "grad_norm": 8.375, + "learning_rate": 4.660036893759608e-06, + "loss": 2.582, + "step": 9121 + }, + { + "epoch": 0.9646785109983079, + "grad_norm": 8.25, + "learning_rate": 4.6599506669235215e-06, + "loss": 2.6302, + "step": 9122 + }, + { + "epoch": 0.9647842639593909, + "grad_norm": 7.65625, + "learning_rate": 4.659864429951671e-06, + "loss": 2.5227, + "step": 9123 + }, + { + "epoch": 0.9648900169204738, + "grad_norm": 8.0625, + "learning_rate": 4.659778182844458e-06, + "loss": 2.2915, + "step": 9124 + }, + { + "epoch": 0.9649957698815567, + "grad_norm": 8.8125, + "learning_rate": 4.65969192560229e-06, + "loss": 2.4997, + "step": 9125 + }, + { + "epoch": 0.9651015228426396, + "grad_norm": 7.625, + "learning_rate": 4.659605658225571e-06, + "loss": 2.5499, + "step": 9126 + }, + { + "epoch": 0.9652072758037225, + "grad_norm": 8.625, + "learning_rate": 4.659519380714706e-06, + "loss": 2.745, + "step": 9127 + }, + { + "epoch": 0.9653130287648054, + "grad_norm": 7.53125, + "learning_rate": 4.659433093070099e-06, + "loss": 2.5945, + "step": 9128 + }, + { + "epoch": 0.9654187817258884, + "grad_norm": 8.25, + "learning_rate": 4.659346795292157e-06, + "loss": 2.2192, + "step": 9129 + }, + { + "epoch": 0.9655245346869712, + "grad_norm": 8.0, + "learning_rate": 4.659260487381283e-06, + "loss": 2.0819, + "step": 9130 + }, + { + "epoch": 0.9656302876480541, + "grad_norm": 8.9375, + "learning_rate": 4.659174169337882e-06, + "loss": 2.7056, + "step": 9131 + }, + { + "epoch": 0.9657360406091371, + "grad_norm": 8.1875, + "learning_rate": 4.65908784116236e-06, + "loss": 2.6005, + "step": 9132 + }, + { + "epoch": 0.9658417935702199, + "grad_norm": 8.125, + "learning_rate": 4.6590015028551225e-06, + "loss": 2.2175, + "step": 9133 + }, + { + "epoch": 0.9659475465313029, + "grad_norm": 8.125, + "learning_rate": 4.658915154416574e-06, + "loss": 2.3805, + "step": 9134 + }, + { + "epoch": 0.9660532994923858, + "grad_norm": 8.25, + "learning_rate": 4.658828795847119e-06, + "loss": 2.5083, + "step": 9135 + }, + { + "epoch": 0.9661590524534687, + "grad_norm": 7.6875, + "learning_rate": 4.658742427147163e-06, + "loss": 2.4818, + "step": 9136 + }, + { + "epoch": 0.9662648054145516, + "grad_norm": 7.9375, + "learning_rate": 4.658656048317113e-06, + "loss": 2.6215, + "step": 9137 + }, + { + "epoch": 0.9663705583756346, + "grad_norm": 8.0625, + "learning_rate": 4.6585696593573724e-06, + "loss": 2.5825, + "step": 9138 + }, + { + "epoch": 0.9664763113367174, + "grad_norm": 8.4375, + "learning_rate": 4.658483260268347e-06, + "loss": 2.3483, + "step": 9139 + }, + { + "epoch": 0.9665820642978004, + "grad_norm": 18.5, + "learning_rate": 4.658396851050443e-06, + "loss": 2.7408, + "step": 9140 + }, + { + "epoch": 0.9666878172588832, + "grad_norm": 8.0625, + "learning_rate": 4.658310431704065e-06, + "loss": 2.5657, + "step": 9141 + }, + { + "epoch": 0.9667935702199661, + "grad_norm": 7.78125, + "learning_rate": 4.658224002229619e-06, + "loss": 2.6623, + "step": 9142 + }, + { + "epoch": 0.9668993231810491, + "grad_norm": 8.1875, + "learning_rate": 4.658137562627509e-06, + "loss": 2.8151, + "step": 9143 + }, + { + "epoch": 0.9670050761421319, + "grad_norm": 8.875, + "learning_rate": 4.658051112898145e-06, + "loss": 2.5633, + "step": 9144 + }, + { + "epoch": 0.9671108291032149, + "grad_norm": 7.625, + "learning_rate": 4.657964653041927e-06, + "loss": 2.4281, + "step": 9145 + }, + { + "epoch": 0.9672165820642978, + "grad_norm": 7.96875, + "learning_rate": 4.657878183059264e-06, + "loss": 2.5294, + "step": 9146 + }, + { + "epoch": 0.9673223350253807, + "grad_norm": 8.5625, + "learning_rate": 4.657791702950563e-06, + "loss": 2.7975, + "step": 9147 + }, + { + "epoch": 0.9674280879864636, + "grad_norm": 7.75, + "learning_rate": 4.657705212716226e-06, + "loss": 2.5041, + "step": 9148 + }, + { + "epoch": 0.9675338409475466, + "grad_norm": 7.84375, + "learning_rate": 4.657618712356662e-06, + "loss": 2.6425, + "step": 9149 + }, + { + "epoch": 0.9676395939086294, + "grad_norm": 7.8125, + "learning_rate": 4.6575322018722755e-06, + "loss": 2.5078, + "step": 9150 + }, + { + "epoch": 0.9677453468697124, + "grad_norm": 7.6875, + "learning_rate": 4.657445681263473e-06, + "loss": 2.6051, + "step": 9151 + }, + { + "epoch": 0.9678510998307953, + "grad_norm": 8.75, + "learning_rate": 4.65735915053066e-06, + "loss": 2.5695, + "step": 9152 + }, + { + "epoch": 0.9679568527918782, + "grad_norm": 8.4375, + "learning_rate": 4.657272609674243e-06, + "loss": 2.5745, + "step": 9153 + }, + { + "epoch": 0.9680626057529611, + "grad_norm": 7.71875, + "learning_rate": 4.657186058694627e-06, + "loss": 2.4747, + "step": 9154 + }, + { + "epoch": 0.9681683587140439, + "grad_norm": 8.0625, + "learning_rate": 4.65709949759222e-06, + "loss": 2.6577, + "step": 9155 + }, + { + "epoch": 0.9682741116751269, + "grad_norm": 7.9375, + "learning_rate": 4.657012926367428e-06, + "loss": 2.5041, + "step": 9156 + }, + { + "epoch": 0.9683798646362098, + "grad_norm": 8.0625, + "learning_rate": 4.656926345020655e-06, + "loss": 2.7191, + "step": 9157 + }, + { + "epoch": 0.9684856175972927, + "grad_norm": 8.25, + "learning_rate": 4.65683975355231e-06, + "loss": 2.6847, + "step": 9158 + }, + { + "epoch": 0.9685913705583756, + "grad_norm": 8.3125, + "learning_rate": 4.656753151962797e-06, + "loss": 2.7222, + "step": 9159 + }, + { + "epoch": 0.9686971235194586, + "grad_norm": 7.90625, + "learning_rate": 4.656666540252524e-06, + "loss": 2.6244, + "step": 9160 + }, + { + "epoch": 0.9688028764805414, + "grad_norm": 8.3125, + "learning_rate": 4.656579918421898e-06, + "loss": 2.6494, + "step": 9161 + }, + { + "epoch": 0.9689086294416244, + "grad_norm": 8.625, + "learning_rate": 4.656493286471324e-06, + "loss": 2.6052, + "step": 9162 + }, + { + "epoch": 0.9690143824027073, + "grad_norm": 8.875, + "learning_rate": 4.656406644401208e-06, + "loss": 2.6362, + "step": 9163 + }, + { + "epoch": 0.9691201353637902, + "grad_norm": 8.25, + "learning_rate": 4.6563199922119584e-06, + "loss": 2.5611, + "step": 9164 + }, + { + "epoch": 0.9692258883248731, + "grad_norm": 8.5625, + "learning_rate": 4.656233329903981e-06, + "loss": 2.475, + "step": 9165 + }, + { + "epoch": 0.9693316412859561, + "grad_norm": 7.9375, + "learning_rate": 4.656146657477682e-06, + "loss": 2.3174, + "step": 9166 + }, + { + "epoch": 0.9694373942470389, + "grad_norm": 8.4375, + "learning_rate": 4.656059974933469e-06, + "loss": 2.8111, + "step": 9167 + }, + { + "epoch": 0.9695431472081218, + "grad_norm": 8.0625, + "learning_rate": 4.655973282271749e-06, + "loss": 2.4745, + "step": 9168 + }, + { + "epoch": 0.9696489001692047, + "grad_norm": 7.75, + "learning_rate": 4.655886579492926e-06, + "loss": 2.534, + "step": 9169 + }, + { + "epoch": 0.9697546531302876, + "grad_norm": 8.0, + "learning_rate": 4.65579986659741e-06, + "loss": 2.2073, + "step": 9170 + }, + { + "epoch": 0.9698604060913706, + "grad_norm": 8.0625, + "learning_rate": 4.655713143585608e-06, + "loss": 2.5851, + "step": 9171 + }, + { + "epoch": 0.9699661590524534, + "grad_norm": 8.0625, + "learning_rate": 4.655626410457925e-06, + "loss": 2.4492, + "step": 9172 + }, + { + "epoch": 0.9700719120135364, + "grad_norm": 8.375, + "learning_rate": 4.655539667214769e-06, + "loss": 2.4333, + "step": 9173 + }, + { + "epoch": 0.9701776649746193, + "grad_norm": 8.4375, + "learning_rate": 4.655452913856547e-06, + "loss": 2.4675, + "step": 9174 + }, + { + "epoch": 0.9702834179357022, + "grad_norm": 7.65625, + "learning_rate": 4.655366150383666e-06, + "loss": 2.5729, + "step": 9175 + }, + { + "epoch": 0.9703891708967851, + "grad_norm": 7.9375, + "learning_rate": 4.6552793767965336e-06, + "loss": 2.558, + "step": 9176 + }, + { + "epoch": 0.9704949238578681, + "grad_norm": 7.1875, + "learning_rate": 4.655192593095555e-06, + "loss": 2.5112, + "step": 9177 + }, + { + "epoch": 0.9706006768189509, + "grad_norm": 10.4375, + "learning_rate": 4.6551057992811405e-06, + "loss": 2.6579, + "step": 9178 + }, + { + "epoch": 0.9707064297800339, + "grad_norm": 7.9375, + "learning_rate": 4.6550189953536954e-06, + "loss": 2.7183, + "step": 9179 + }, + { + "epoch": 0.9708121827411168, + "grad_norm": 8.0, + "learning_rate": 4.654932181313627e-06, + "loss": 2.6843, + "step": 9180 + }, + { + "epoch": 0.9709179357021996, + "grad_norm": 8.6875, + "learning_rate": 4.654845357161344e-06, + "loss": 2.5663, + "step": 9181 + }, + { + "epoch": 0.9710236886632826, + "grad_norm": 8.0, + "learning_rate": 4.6547585228972535e-06, + "loss": 2.5246, + "step": 9182 + }, + { + "epoch": 0.9711294416243654, + "grad_norm": 7.71875, + "learning_rate": 4.654671678521762e-06, + "loss": 2.5469, + "step": 9183 + }, + { + "epoch": 0.9712351945854484, + "grad_norm": 8.0625, + "learning_rate": 4.654584824035277e-06, + "loss": 2.3017, + "step": 9184 + }, + { + "epoch": 0.9713409475465313, + "grad_norm": 8.3125, + "learning_rate": 4.6544979594382065e-06, + "loss": 2.6466, + "step": 9185 + }, + { + "epoch": 0.9714467005076142, + "grad_norm": 7.875, + "learning_rate": 4.65441108473096e-06, + "loss": 2.4571, + "step": 9186 + }, + { + "epoch": 0.9715524534686971, + "grad_norm": 8.1875, + "learning_rate": 4.654324199913941e-06, + "loss": 2.5837, + "step": 9187 + }, + { + "epoch": 0.9716582064297801, + "grad_norm": 7.78125, + "learning_rate": 4.654237304987561e-06, + "loss": 2.6294, + "step": 9188 + }, + { + "epoch": 0.9717639593908629, + "grad_norm": 8.25, + "learning_rate": 4.654150399952226e-06, + "loss": 2.598, + "step": 9189 + }, + { + "epoch": 0.9718697123519459, + "grad_norm": 8.0, + "learning_rate": 4.654063484808344e-06, + "loss": 2.4735, + "step": 9190 + }, + { + "epoch": 0.9719754653130288, + "grad_norm": 7.78125, + "learning_rate": 4.6539765595563245e-06, + "loss": 2.6533, + "step": 9191 + }, + { + "epoch": 0.9720812182741116, + "grad_norm": 7.96875, + "learning_rate": 4.653889624196573e-06, + "loss": 2.6412, + "step": 9192 + }, + { + "epoch": 0.9721869712351946, + "grad_norm": 7.90625, + "learning_rate": 4.653802678729498e-06, + "loss": 2.5139, + "step": 9193 + }, + { + "epoch": 0.9722927241962775, + "grad_norm": 7.71875, + "learning_rate": 4.653715723155507e-06, + "loss": 2.6161, + "step": 9194 + }, + { + "epoch": 0.9723984771573604, + "grad_norm": 9.4375, + "learning_rate": 4.653628757475011e-06, + "loss": 2.5246, + "step": 9195 + }, + { + "epoch": 0.9725042301184433, + "grad_norm": 8.125, + "learning_rate": 4.653541781688415e-06, + "loss": 2.7323, + "step": 9196 + }, + { + "epoch": 0.9726099830795262, + "grad_norm": 7.625, + "learning_rate": 4.653454795796129e-06, + "loss": 2.5617, + "step": 9197 + }, + { + "epoch": 0.9727157360406091, + "grad_norm": 7.71875, + "learning_rate": 4.65336779979856e-06, + "loss": 2.5693, + "step": 9198 + }, + { + "epoch": 0.9728214890016921, + "grad_norm": 9.5, + "learning_rate": 4.6532807936961155e-06, + "loss": 2.579, + "step": 9199 + }, + { + "epoch": 0.9729272419627749, + "grad_norm": 8.0, + "learning_rate": 4.653193777489206e-06, + "loss": 2.3959, + "step": 9200 + }, + { + "epoch": 0.9730329949238579, + "grad_norm": 8.25, + "learning_rate": 4.653106751178238e-06, + "loss": 2.5861, + "step": 9201 + }, + { + "epoch": 0.9731387478849408, + "grad_norm": 8.1875, + "learning_rate": 4.653019714763622e-06, + "loss": 2.6191, + "step": 9202 + }, + { + "epoch": 0.9732445008460237, + "grad_norm": 7.90625, + "learning_rate": 4.6529326682457635e-06, + "loss": 2.2674, + "step": 9203 + }, + { + "epoch": 0.9733502538071066, + "grad_norm": 8.8125, + "learning_rate": 4.652845611625073e-06, + "loss": 2.4593, + "step": 9204 + }, + { + "epoch": 0.9734560067681896, + "grad_norm": 7.71875, + "learning_rate": 4.652758544901959e-06, + "loss": 2.5653, + "step": 9205 + }, + { + "epoch": 0.9735617597292724, + "grad_norm": 13.9375, + "learning_rate": 4.652671468076828e-06, + "loss": 2.443, + "step": 9206 + }, + { + "epoch": 0.9736675126903553, + "grad_norm": 8.5625, + "learning_rate": 4.652584381150092e-06, + "loss": 2.7733, + "step": 9207 + }, + { + "epoch": 0.9737732656514383, + "grad_norm": 8.0625, + "learning_rate": 4.652497284122157e-06, + "loss": 2.6719, + "step": 9208 + }, + { + "epoch": 0.9738790186125211, + "grad_norm": 7.90625, + "learning_rate": 4.6524101769934335e-06, + "loss": 2.5872, + "step": 9209 + }, + { + "epoch": 0.9739847715736041, + "grad_norm": 8.5, + "learning_rate": 4.652323059764329e-06, + "loss": 2.6813, + "step": 9210 + }, + { + "epoch": 0.9740905245346869, + "grad_norm": 7.8125, + "learning_rate": 4.652235932435252e-06, + "loss": 2.4988, + "step": 9211 + }, + { + "epoch": 0.9741962774957699, + "grad_norm": 8.5625, + "learning_rate": 4.652148795006613e-06, + "loss": 2.6672, + "step": 9212 + }, + { + "epoch": 0.9743020304568528, + "grad_norm": 8.375, + "learning_rate": 4.652061647478819e-06, + "loss": 2.3703, + "step": 9213 + }, + { + "epoch": 0.9744077834179357, + "grad_norm": 7.65625, + "learning_rate": 4.651974489852279e-06, + "loss": 2.4219, + "step": 9214 + }, + { + "epoch": 0.9745135363790186, + "grad_norm": 7.78125, + "learning_rate": 4.651887322127405e-06, + "loss": 2.5811, + "step": 9215 + }, + { + "epoch": 0.9746192893401016, + "grad_norm": 8.25, + "learning_rate": 4.651800144304602e-06, + "loss": 2.5719, + "step": 9216 + }, + { + "epoch": 0.9747250423011844, + "grad_norm": 8.625, + "learning_rate": 4.651712956384283e-06, + "loss": 2.7232, + "step": 9217 + }, + { + "epoch": 0.9748307952622673, + "grad_norm": 8.125, + "learning_rate": 4.6516257583668535e-06, + "loss": 2.6173, + "step": 9218 + }, + { + "epoch": 0.9749365482233503, + "grad_norm": 7.8125, + "learning_rate": 4.651538550252725e-06, + "loss": 2.4951, + "step": 9219 + }, + { + "epoch": 0.9750423011844331, + "grad_norm": 8.375, + "learning_rate": 4.651451332042305e-06, + "loss": 2.7104, + "step": 9220 + }, + { + "epoch": 0.9751480541455161, + "grad_norm": 7.75, + "learning_rate": 4.651364103736005e-06, + "loss": 2.4704, + "step": 9221 + }, + { + "epoch": 0.975253807106599, + "grad_norm": 8.3125, + "learning_rate": 4.651276865334233e-06, + "loss": 2.4689, + "step": 9222 + }, + { + "epoch": 0.9753595600676819, + "grad_norm": 8.5625, + "learning_rate": 4.651189616837398e-06, + "loss": 2.8198, + "step": 9223 + }, + { + "epoch": 0.9754653130287648, + "grad_norm": 8.125, + "learning_rate": 4.65110235824591e-06, + "loss": 2.6674, + "step": 9224 + }, + { + "epoch": 0.9755710659898477, + "grad_norm": 7.8125, + "learning_rate": 4.651015089560179e-06, + "loss": 2.4875, + "step": 9225 + }, + { + "epoch": 0.9756768189509306, + "grad_norm": 7.90625, + "learning_rate": 4.650927810780613e-06, + "loss": 2.6226, + "step": 9226 + }, + { + "epoch": 0.9757825719120136, + "grad_norm": 8.8125, + "learning_rate": 4.650840521907623e-06, + "loss": 2.6556, + "step": 9227 + }, + { + "epoch": 0.9758883248730964, + "grad_norm": 7.90625, + "learning_rate": 4.650753222941619e-06, + "loss": 2.5785, + "step": 9228 + }, + { + "epoch": 0.9759940778341794, + "grad_norm": 7.6875, + "learning_rate": 4.650665913883009e-06, + "loss": 2.6014, + "step": 9229 + }, + { + "epoch": 0.9760998307952623, + "grad_norm": 8.6875, + "learning_rate": 4.650578594732203e-06, + "loss": 2.711, + "step": 9230 + }, + { + "epoch": 0.9762055837563451, + "grad_norm": 7.84375, + "learning_rate": 4.650491265489612e-06, + "loss": 2.5984, + "step": 9231 + }, + { + "epoch": 0.9763113367174281, + "grad_norm": 7.34375, + "learning_rate": 4.650403926155644e-06, + "loss": 2.4876, + "step": 9232 + }, + { + "epoch": 0.976417089678511, + "grad_norm": 7.90625, + "learning_rate": 4.65031657673071e-06, + "loss": 2.5694, + "step": 9233 + }, + { + "epoch": 0.9765228426395939, + "grad_norm": 8.0625, + "learning_rate": 4.6502292172152204e-06, + "loss": 2.5098, + "step": 9234 + }, + { + "epoch": 0.9766285956006768, + "grad_norm": 7.875, + "learning_rate": 4.650141847609585e-06, + "loss": 2.6415, + "step": 9235 + }, + { + "epoch": 0.9767343485617598, + "grad_norm": 8.375, + "learning_rate": 4.650054467914212e-06, + "loss": 2.6415, + "step": 9236 + }, + { + "epoch": 0.9768401015228426, + "grad_norm": 7.78125, + "learning_rate": 4.649967078129513e-06, + "loss": 2.4742, + "step": 9237 + }, + { + "epoch": 0.9769458544839256, + "grad_norm": 7.8125, + "learning_rate": 4.649879678255898e-06, + "loss": 2.6116, + "step": 9238 + }, + { + "epoch": 0.9770516074450084, + "grad_norm": 8.9375, + "learning_rate": 4.6497922682937775e-06, + "loss": 2.5568, + "step": 9239 + }, + { + "epoch": 0.9771573604060914, + "grad_norm": 7.59375, + "learning_rate": 4.6497048482435604e-06, + "loss": 2.4864, + "step": 9240 + }, + { + "epoch": 0.9772631133671743, + "grad_norm": 8.5625, + "learning_rate": 4.649617418105658e-06, + "loss": 2.6295, + "step": 9241 + }, + { + "epoch": 0.9773688663282571, + "grad_norm": 7.9375, + "learning_rate": 4.6495299778804805e-06, + "loss": 2.4322, + "step": 9242 + }, + { + "epoch": 0.9774746192893401, + "grad_norm": 7.90625, + "learning_rate": 4.6494425275684376e-06, + "loss": 2.5047, + "step": 9243 + }, + { + "epoch": 0.977580372250423, + "grad_norm": 8.0625, + "learning_rate": 4.64935506716994e-06, + "loss": 2.5292, + "step": 9244 + }, + { + "epoch": 0.9776861252115059, + "grad_norm": 7.875, + "learning_rate": 4.649267596685397e-06, + "loss": 2.5601, + "step": 9245 + }, + { + "epoch": 0.9777918781725888, + "grad_norm": 8.25, + "learning_rate": 4.649180116115222e-06, + "loss": 2.7002, + "step": 9246 + }, + { + "epoch": 0.9778976311336718, + "grad_norm": 7.75, + "learning_rate": 4.649092625459823e-06, + "loss": 2.61, + "step": 9247 + }, + { + "epoch": 0.9780033840947546, + "grad_norm": 7.78125, + "learning_rate": 4.6490051247196115e-06, + "loss": 2.2848, + "step": 9248 + }, + { + "epoch": 0.9781091370558376, + "grad_norm": 8.3125, + "learning_rate": 4.648917613894997e-06, + "loss": 2.5953, + "step": 9249 + }, + { + "epoch": 0.9782148900169205, + "grad_norm": 8.0, + "learning_rate": 4.648830092986392e-06, + "loss": 2.4935, + "step": 9250 + }, + { + "epoch": 0.9783206429780034, + "grad_norm": 8.5625, + "learning_rate": 4.6487425619942055e-06, + "loss": 2.652, + "step": 9251 + }, + { + "epoch": 0.9784263959390863, + "grad_norm": 8.8125, + "learning_rate": 4.6486550209188494e-06, + "loss": 2.8605, + "step": 9252 + }, + { + "epoch": 0.9785321489001692, + "grad_norm": 7.71875, + "learning_rate": 4.648567469760734e-06, + "loss": 2.4504, + "step": 9253 + }, + { + "epoch": 0.9786379018612521, + "grad_norm": 8.0, + "learning_rate": 4.6484799085202705e-06, + "loss": 2.5397, + "step": 9254 + }, + { + "epoch": 0.978743654822335, + "grad_norm": 7.6875, + "learning_rate": 4.648392337197869e-06, + "loss": 2.6071, + "step": 9255 + }, + { + "epoch": 0.9788494077834179, + "grad_norm": 8.625, + "learning_rate": 4.6483047557939405e-06, + "loss": 2.6593, + "step": 9256 + }, + { + "epoch": 0.9789551607445008, + "grad_norm": 8.125, + "learning_rate": 4.648217164308897e-06, + "loss": 2.7438, + "step": 9257 + }, + { + "epoch": 0.9790609137055838, + "grad_norm": 8.1875, + "learning_rate": 4.6481295627431485e-06, + "loss": 2.6189, + "step": 9258 + }, + { + "epoch": 0.9791666666666666, + "grad_norm": 7.8125, + "learning_rate": 4.648041951097108e-06, + "loss": 2.5985, + "step": 9259 + }, + { + "epoch": 0.9792724196277496, + "grad_norm": 8.875, + "learning_rate": 4.6479543293711834e-06, + "loss": 2.5248, + "step": 9260 + }, + { + "epoch": 0.9793781725888325, + "grad_norm": 8.25, + "learning_rate": 4.647866697565788e-06, + "loss": 2.8795, + "step": 9261 + }, + { + "epoch": 0.9794839255499154, + "grad_norm": 8.0625, + "learning_rate": 4.6477790556813335e-06, + "loss": 2.5842, + "step": 9262 + }, + { + "epoch": 0.9795896785109983, + "grad_norm": 8.6875, + "learning_rate": 4.6476914037182296e-06, + "loss": 2.5213, + "step": 9263 + }, + { + "epoch": 0.9796954314720813, + "grad_norm": 7.59375, + "learning_rate": 4.647603741676888e-06, + "loss": 2.5753, + "step": 9264 + }, + { + "epoch": 0.9798011844331641, + "grad_norm": 8.6875, + "learning_rate": 4.647516069557722e-06, + "loss": 2.6836, + "step": 9265 + }, + { + "epoch": 0.9799069373942471, + "grad_norm": 8.25, + "learning_rate": 4.64742838736114e-06, + "loss": 2.7515, + "step": 9266 + }, + { + "epoch": 0.9800126903553299, + "grad_norm": 8.3125, + "learning_rate": 4.647340695087555e-06, + "loss": 2.7842, + "step": 9267 + }, + { + "epoch": 0.9801184433164128, + "grad_norm": 8.0, + "learning_rate": 4.647252992737378e-06, + "loss": 2.5938, + "step": 9268 + }, + { + "epoch": 0.9802241962774958, + "grad_norm": 7.6875, + "learning_rate": 4.647165280311022e-06, + "loss": 2.6436, + "step": 9269 + }, + { + "epoch": 0.9803299492385786, + "grad_norm": 9.1875, + "learning_rate": 4.647077557808896e-06, + "loss": 2.9613, + "step": 9270 + }, + { + "epoch": 0.9804357021996616, + "grad_norm": 8.75, + "learning_rate": 4.646989825231414e-06, + "loss": 2.5311, + "step": 9271 + }, + { + "epoch": 0.9805414551607445, + "grad_norm": 7.84375, + "learning_rate": 4.646902082578987e-06, + "loss": 2.468, + "step": 9272 + }, + { + "epoch": 0.9806472081218274, + "grad_norm": 7.71875, + "learning_rate": 4.646814329852026e-06, + "loss": 2.4243, + "step": 9273 + }, + { + "epoch": 0.9807529610829103, + "grad_norm": 7.84375, + "learning_rate": 4.646726567050943e-06, + "loss": 2.336, + "step": 9274 + }, + { + "epoch": 0.9808587140439933, + "grad_norm": 7.6875, + "learning_rate": 4.646638794176151e-06, + "loss": 2.5401, + "step": 9275 + }, + { + "epoch": 0.9809644670050761, + "grad_norm": 8.3125, + "learning_rate": 4.646551011228062e-06, + "loss": 2.5932, + "step": 9276 + }, + { + "epoch": 0.9810702199661591, + "grad_norm": 8.5, + "learning_rate": 4.646463218207086e-06, + "loss": 2.6306, + "step": 9277 + }, + { + "epoch": 0.981175972927242, + "grad_norm": 7.75, + "learning_rate": 4.646375415113636e-06, + "loss": 2.6254, + "step": 9278 + }, + { + "epoch": 0.9812817258883249, + "grad_norm": 8.0625, + "learning_rate": 4.646287601948124e-06, + "loss": 2.6638, + "step": 9279 + }, + { + "epoch": 0.9813874788494078, + "grad_norm": 8.5, + "learning_rate": 4.646199778710962e-06, + "loss": 2.1293, + "step": 9280 + }, + { + "epoch": 0.9814932318104906, + "grad_norm": 8.0625, + "learning_rate": 4.6461119454025625e-06, + "loss": 2.3525, + "step": 9281 + }, + { + "epoch": 0.9815989847715736, + "grad_norm": 7.75, + "learning_rate": 4.646024102023337e-06, + "loss": 2.553, + "step": 9282 + }, + { + "epoch": 0.9817047377326565, + "grad_norm": 7.90625, + "learning_rate": 4.645936248573698e-06, + "loss": 2.5178, + "step": 9283 + }, + { + "epoch": 0.9818104906937394, + "grad_norm": 8.1875, + "learning_rate": 4.645848385054059e-06, + "loss": 2.4809, + "step": 9284 + }, + { + "epoch": 0.9819162436548223, + "grad_norm": 8.6875, + "learning_rate": 4.64576051146483e-06, + "loss": 2.741, + "step": 9285 + }, + { + "epoch": 0.9820219966159053, + "grad_norm": 7.9375, + "learning_rate": 4.645672627806425e-06, + "loss": 2.5382, + "step": 9286 + }, + { + "epoch": 0.9821277495769881, + "grad_norm": 7.53125, + "learning_rate": 4.645584734079255e-06, + "loss": 2.4558, + "step": 9287 + }, + { + "epoch": 0.9822335025380711, + "grad_norm": 8.25, + "learning_rate": 4.645496830283734e-06, + "loss": 2.7367, + "step": 9288 + }, + { + "epoch": 0.982339255499154, + "grad_norm": 8.0, + "learning_rate": 4.645408916420274e-06, + "loss": 2.5903, + "step": 9289 + }, + { + "epoch": 0.9824450084602369, + "grad_norm": 7.875, + "learning_rate": 4.645320992489287e-06, + "loss": 2.4353, + "step": 9290 + }, + { + "epoch": 0.9825507614213198, + "grad_norm": 8.6875, + "learning_rate": 4.645233058491186e-06, + "loss": 2.6242, + "step": 9291 + }, + { + "epoch": 0.9826565143824028, + "grad_norm": 8.0625, + "learning_rate": 4.645145114426385e-06, + "loss": 2.5346, + "step": 9292 + }, + { + "epoch": 0.9827622673434856, + "grad_norm": 8.375, + "learning_rate": 4.6450571602952946e-06, + "loss": 2.523, + "step": 9293 + }, + { + "epoch": 0.9828680203045685, + "grad_norm": 7.90625, + "learning_rate": 4.644969196098327e-06, + "loss": 2.5859, + "step": 9294 + }, + { + "epoch": 0.9829737732656514, + "grad_norm": 7.75, + "learning_rate": 4.644881221835898e-06, + "loss": 2.7049, + "step": 9295 + }, + { + "epoch": 0.9830795262267343, + "grad_norm": 9.3125, + "learning_rate": 4.644793237508417e-06, + "loss": 2.2064, + "step": 9296 + }, + { + "epoch": 0.9831852791878173, + "grad_norm": 10.0625, + "learning_rate": 4.6447052431162996e-06, + "loss": 2.6023, + "step": 9297 + }, + { + "epoch": 0.9832910321489001, + "grad_norm": 8.1875, + "learning_rate": 4.644617238659957e-06, + "loss": 2.3835, + "step": 9298 + }, + { + "epoch": 0.9833967851099831, + "grad_norm": 7.59375, + "learning_rate": 4.644529224139803e-06, + "loss": 2.5042, + "step": 9299 + }, + { + "epoch": 0.983502538071066, + "grad_norm": 8.0, + "learning_rate": 4.644441199556251e-06, + "loss": 2.5605, + "step": 9300 + }, + { + "epoch": 0.9836082910321489, + "grad_norm": 7.84375, + "learning_rate": 4.644353164909713e-06, + "loss": 2.6229, + "step": 9301 + }, + { + "epoch": 0.9837140439932318, + "grad_norm": 7.84375, + "learning_rate": 4.644265120200603e-06, + "loss": 2.4616, + "step": 9302 + }, + { + "epoch": 0.9838197969543148, + "grad_norm": 10.5625, + "learning_rate": 4.644177065429333e-06, + "loss": 2.5544, + "step": 9303 + }, + { + "epoch": 0.9839255499153976, + "grad_norm": 7.78125, + "learning_rate": 4.644089000596317e-06, + "loss": 2.5619, + "step": 9304 + }, + { + "epoch": 0.9840313028764806, + "grad_norm": 7.53125, + "learning_rate": 4.644000925701969e-06, + "loss": 2.4692, + "step": 9305 + }, + { + "epoch": 0.9841370558375635, + "grad_norm": 8.125, + "learning_rate": 4.643912840746702e-06, + "loss": 2.7471, + "step": 9306 + }, + { + "epoch": 0.9842428087986463, + "grad_norm": 8.4375, + "learning_rate": 4.643824745730927e-06, + "loss": 2.6194, + "step": 9307 + }, + { + "epoch": 0.9843485617597293, + "grad_norm": 7.625, + "learning_rate": 4.643736640655061e-06, + "loss": 2.4574, + "step": 9308 + }, + { + "epoch": 0.9844543147208121, + "grad_norm": 8.5, + "learning_rate": 4.643648525519514e-06, + "loss": 2.4425, + "step": 9309 + }, + { + "epoch": 0.9845600676818951, + "grad_norm": 9.0625, + "learning_rate": 4.643560400324703e-06, + "loss": 2.6144, + "step": 9310 + }, + { + "epoch": 0.984665820642978, + "grad_norm": 8.5625, + "learning_rate": 4.643472265071038e-06, + "loss": 2.7975, + "step": 9311 + }, + { + "epoch": 0.9847715736040609, + "grad_norm": 8.75, + "learning_rate": 4.6433841197589356e-06, + "loss": 2.5995, + "step": 9312 + }, + { + "epoch": 0.9848773265651438, + "grad_norm": 8.5, + "learning_rate": 4.643295964388808e-06, + "loss": 2.8212, + "step": 9313 + }, + { + "epoch": 0.9849830795262268, + "grad_norm": 7.4375, + "learning_rate": 4.643207798961069e-06, + "loss": 2.571, + "step": 9314 + }, + { + "epoch": 0.9850888324873096, + "grad_norm": 7.875, + "learning_rate": 4.643119623476132e-06, + "loss": 2.6258, + "step": 9315 + }, + { + "epoch": 0.9851945854483926, + "grad_norm": 9.25, + "learning_rate": 4.643031437934411e-06, + "loss": 2.6688, + "step": 9316 + }, + { + "epoch": 0.9853003384094755, + "grad_norm": 7.625, + "learning_rate": 4.64294324233632e-06, + "loss": 2.5437, + "step": 9317 + }, + { + "epoch": 0.9854060913705583, + "grad_norm": 7.6875, + "learning_rate": 4.642855036682273e-06, + "loss": 2.4406, + "step": 9318 + }, + { + "epoch": 0.9855118443316413, + "grad_norm": 7.46875, + "learning_rate": 4.642766820972684e-06, + "loss": 2.5133, + "step": 9319 + }, + { + "epoch": 0.9856175972927242, + "grad_norm": 7.71875, + "learning_rate": 4.642678595207966e-06, + "loss": 2.5907, + "step": 9320 + }, + { + "epoch": 0.9857233502538071, + "grad_norm": 7.875, + "learning_rate": 4.6425903593885345e-06, + "loss": 2.7231, + "step": 9321 + }, + { + "epoch": 0.98582910321489, + "grad_norm": 8.0625, + "learning_rate": 4.642502113514802e-06, + "loss": 2.5222, + "step": 9322 + }, + { + "epoch": 0.9859348561759729, + "grad_norm": 8.125, + "learning_rate": 4.642413857587183e-06, + "loss": 2.5107, + "step": 9323 + }, + { + "epoch": 0.9860406091370558, + "grad_norm": 7.65625, + "learning_rate": 4.642325591606093e-06, + "loss": 2.5449, + "step": 9324 + }, + { + "epoch": 0.9861463620981388, + "grad_norm": 8.4375, + "learning_rate": 4.642237315571945e-06, + "loss": 2.418, + "step": 9325 + }, + { + "epoch": 0.9862521150592216, + "grad_norm": 7.8125, + "learning_rate": 4.642149029485153e-06, + "loss": 2.5587, + "step": 9326 + }, + { + "epoch": 0.9863578680203046, + "grad_norm": 7.625, + "learning_rate": 4.642060733346132e-06, + "loss": 2.5688, + "step": 9327 + }, + { + "epoch": 0.9864636209813875, + "grad_norm": 9.125, + "learning_rate": 4.641972427155296e-06, + "loss": 2.8161, + "step": 9328 + }, + { + "epoch": 0.9865693739424704, + "grad_norm": 8.125, + "learning_rate": 4.64188411091306e-06, + "loss": 2.5542, + "step": 9329 + }, + { + "epoch": 0.9866751269035533, + "grad_norm": 8.375, + "learning_rate": 4.641795784619837e-06, + "loss": 2.296, + "step": 9330 + }, + { + "epoch": 0.9867808798646363, + "grad_norm": 8.375, + "learning_rate": 4.641707448276042e-06, + "loss": 2.6871, + "step": 9331 + }, + { + "epoch": 0.9868866328257191, + "grad_norm": 8.3125, + "learning_rate": 4.641619101882091e-06, + "loss": 2.8444, + "step": 9332 + }, + { + "epoch": 0.986992385786802, + "grad_norm": 7.75, + "learning_rate": 4.641530745438396e-06, + "loss": 2.5038, + "step": 9333 + }, + { + "epoch": 0.987098138747885, + "grad_norm": 7.84375, + "learning_rate": 4.641442378945375e-06, + "loss": 2.3645, + "step": 9334 + }, + { + "epoch": 0.9872038917089678, + "grad_norm": 8.0, + "learning_rate": 4.641354002403439e-06, + "loss": 2.588, + "step": 9335 + }, + { + "epoch": 0.9873096446700508, + "grad_norm": 8.5625, + "learning_rate": 4.641265615813005e-06, + "loss": 3.006, + "step": 9336 + }, + { + "epoch": 0.9874153976311336, + "grad_norm": 8.0, + "learning_rate": 4.641177219174487e-06, + "loss": 2.5593, + "step": 9337 + }, + { + "epoch": 0.9875211505922166, + "grad_norm": 7.84375, + "learning_rate": 4.641088812488301e-06, + "loss": 2.684, + "step": 9338 + }, + { + "epoch": 0.9876269035532995, + "grad_norm": 7.90625, + "learning_rate": 4.6410003957548605e-06, + "loss": 2.6366, + "step": 9339 + }, + { + "epoch": 0.9877326565143824, + "grad_norm": 8.125, + "learning_rate": 4.640911968974581e-06, + "loss": 2.7725, + "step": 9340 + }, + { + "epoch": 0.9878384094754653, + "grad_norm": 7.75, + "learning_rate": 4.640823532147877e-06, + "loss": 2.4287, + "step": 9341 + }, + { + "epoch": 0.9879441624365483, + "grad_norm": 8.125, + "learning_rate": 4.640735085275163e-06, + "loss": 2.736, + "step": 9342 + }, + { + "epoch": 0.9880499153976311, + "grad_norm": 8.375, + "learning_rate": 4.640646628356855e-06, + "loss": 2.7511, + "step": 9343 + }, + { + "epoch": 0.988155668358714, + "grad_norm": 7.90625, + "learning_rate": 4.640558161393368e-06, + "loss": 2.6387, + "step": 9344 + }, + { + "epoch": 0.988261421319797, + "grad_norm": 8.4375, + "learning_rate": 4.6404696843851175e-06, + "loss": 2.5745, + "step": 9345 + }, + { + "epoch": 0.9883671742808798, + "grad_norm": 8.75, + "learning_rate": 4.640381197332518e-06, + "loss": 2.6836, + "step": 9346 + }, + { + "epoch": 0.9884729272419628, + "grad_norm": 8.25, + "learning_rate": 4.640292700235985e-06, + "loss": 2.5678, + "step": 9347 + }, + { + "epoch": 0.9885786802030457, + "grad_norm": 7.5625, + "learning_rate": 4.640204193095933e-06, + "loss": 2.5632, + "step": 9348 + }, + { + "epoch": 0.9886844331641286, + "grad_norm": 8.125, + "learning_rate": 4.640115675912779e-06, + "loss": 2.6038, + "step": 9349 + }, + { + "epoch": 0.9887901861252115, + "grad_norm": 7.8125, + "learning_rate": 4.640027148686936e-06, + "loss": 2.5819, + "step": 9350 + }, + { + "epoch": 0.9888959390862944, + "grad_norm": 8.3125, + "learning_rate": 4.639938611418823e-06, + "loss": 2.515, + "step": 9351 + }, + { + "epoch": 0.9890016920473773, + "grad_norm": 8.375, + "learning_rate": 4.639850064108852e-06, + "loss": 2.6925, + "step": 9352 + }, + { + "epoch": 0.9891074450084603, + "grad_norm": 7.96875, + "learning_rate": 4.63976150675744e-06, + "loss": 2.6068, + "step": 9353 + }, + { + "epoch": 0.9892131979695431, + "grad_norm": 7.5, + "learning_rate": 4.639672939365002e-06, + "loss": 2.6105, + "step": 9354 + }, + { + "epoch": 0.989318950930626, + "grad_norm": 8.5625, + "learning_rate": 4.639584361931954e-06, + "loss": 2.1656, + "step": 9355 + }, + { + "epoch": 0.989424703891709, + "grad_norm": 8.125, + "learning_rate": 4.639495774458712e-06, + "loss": 2.4807, + "step": 9356 + }, + { + "epoch": 0.9895304568527918, + "grad_norm": 8.375, + "learning_rate": 4.639407176945692e-06, + "loss": 2.5252, + "step": 9357 + }, + { + "epoch": 0.9896362098138748, + "grad_norm": 7.84375, + "learning_rate": 4.639318569393308e-06, + "loss": 2.3802, + "step": 9358 + }, + { + "epoch": 0.9897419627749577, + "grad_norm": 7.5, + "learning_rate": 4.639229951801977e-06, + "loss": 2.6088, + "step": 9359 + }, + { + "epoch": 0.9898477157360406, + "grad_norm": 8.0625, + "learning_rate": 4.639141324172115e-06, + "loss": 2.5861, + "step": 9360 + }, + { + "epoch": 0.9899534686971235, + "grad_norm": 8.3125, + "learning_rate": 4.639052686504138e-06, + "loss": 2.6354, + "step": 9361 + }, + { + "epoch": 0.9900592216582065, + "grad_norm": 7.75, + "learning_rate": 4.638964038798462e-06, + "loss": 2.5082, + "step": 9362 + }, + { + "epoch": 0.9901649746192893, + "grad_norm": 8.9375, + "learning_rate": 4.638875381055501e-06, + "loss": 2.847, + "step": 9363 + }, + { + "epoch": 0.9902707275803723, + "grad_norm": 7.59375, + "learning_rate": 4.638786713275674e-06, + "loss": 2.4586, + "step": 9364 + }, + { + "epoch": 0.9903764805414551, + "grad_norm": 7.625, + "learning_rate": 4.638698035459395e-06, + "loss": 2.3947, + "step": 9365 + }, + { + "epoch": 0.9904822335025381, + "grad_norm": 8.125, + "learning_rate": 4.6386093476070804e-06, + "loss": 2.5773, + "step": 9366 + }, + { + "epoch": 0.990587986463621, + "grad_norm": 7.75, + "learning_rate": 4.6385206497191475e-06, + "loss": 2.7646, + "step": 9367 + }, + { + "epoch": 0.9906937394247038, + "grad_norm": 12.625, + "learning_rate": 4.638431941796012e-06, + "loss": 2.6677, + "step": 9368 + }, + { + "epoch": 0.9907994923857868, + "grad_norm": 8.3125, + "learning_rate": 4.638343223838089e-06, + "loss": 2.7365, + "step": 9369 + }, + { + "epoch": 0.9909052453468697, + "grad_norm": 8.0625, + "learning_rate": 4.638254495845796e-06, + "loss": 2.4374, + "step": 9370 + }, + { + "epoch": 0.9910109983079526, + "grad_norm": 7.65625, + "learning_rate": 4.63816575781955e-06, + "loss": 2.4882, + "step": 9371 + }, + { + "epoch": 0.9911167512690355, + "grad_norm": 7.75, + "learning_rate": 4.638077009759766e-06, + "loss": 2.6614, + "step": 9372 + }, + { + "epoch": 0.9912225042301185, + "grad_norm": 7.90625, + "learning_rate": 4.6379882516668616e-06, + "loss": 2.5777, + "step": 9373 + }, + { + "epoch": 0.9913282571912013, + "grad_norm": 8.3125, + "learning_rate": 4.637899483541252e-06, + "loss": 2.6702, + "step": 9374 + }, + { + "epoch": 0.9914340101522843, + "grad_norm": 8.1875, + "learning_rate": 4.637810705383355e-06, + "loss": 2.7393, + "step": 9375 + }, + { + "epoch": 0.9915397631133672, + "grad_norm": 7.6875, + "learning_rate": 4.637721917193587e-06, + "loss": 2.5242, + "step": 9376 + }, + { + "epoch": 0.9916455160744501, + "grad_norm": 8.5625, + "learning_rate": 4.637633118972364e-06, + "loss": 2.8494, + "step": 9377 + }, + { + "epoch": 0.991751269035533, + "grad_norm": 7.71875, + "learning_rate": 4.637544310720103e-06, + "loss": 2.2842, + "step": 9378 + }, + { + "epoch": 0.9918570219966159, + "grad_norm": 8.0625, + "learning_rate": 4.637455492437221e-06, + "loss": 2.5385, + "step": 9379 + }, + { + "epoch": 0.9919627749576988, + "grad_norm": 8.5, + "learning_rate": 4.6373666641241336e-06, + "loss": 2.8943, + "step": 9380 + }, + { + "epoch": 0.9920685279187818, + "grad_norm": 7.9375, + "learning_rate": 4.637277825781259e-06, + "loss": 2.5766, + "step": 9381 + }, + { + "epoch": 0.9921742808798646, + "grad_norm": 8.0625, + "learning_rate": 4.637188977409014e-06, + "loss": 2.6669, + "step": 9382 + }, + { + "epoch": 0.9922800338409475, + "grad_norm": 8.0, + "learning_rate": 4.637100119007816e-06, + "loss": 2.4469, + "step": 9383 + }, + { + "epoch": 0.9923857868020305, + "grad_norm": 7.96875, + "learning_rate": 4.63701125057808e-06, + "loss": 2.6441, + "step": 9384 + }, + { + "epoch": 0.9924915397631133, + "grad_norm": 7.71875, + "learning_rate": 4.636922372120224e-06, + "loss": 2.4586, + "step": 9385 + }, + { + "epoch": 0.9925972927241963, + "grad_norm": 7.8125, + "learning_rate": 4.636833483634666e-06, + "loss": 2.4397, + "step": 9386 + }, + { + "epoch": 0.9927030456852792, + "grad_norm": 7.875, + "learning_rate": 4.636744585121823e-06, + "loss": 2.7271, + "step": 9387 + }, + { + "epoch": 0.9928087986463621, + "grad_norm": 7.75, + "learning_rate": 4.63665567658211e-06, + "loss": 2.6803, + "step": 9388 + }, + { + "epoch": 0.992914551607445, + "grad_norm": 9.375, + "learning_rate": 4.636566758015947e-06, + "loss": 2.8465, + "step": 9389 + }, + { + "epoch": 0.993020304568528, + "grad_norm": 9.0625, + "learning_rate": 4.63647782942375e-06, + "loss": 2.5804, + "step": 9390 + }, + { + "epoch": 0.9931260575296108, + "grad_norm": 7.65625, + "learning_rate": 4.636388890805936e-06, + "loss": 2.4784, + "step": 9391 + }, + { + "epoch": 0.9932318104906938, + "grad_norm": 7.78125, + "learning_rate": 4.636299942162922e-06, + "loss": 2.5853, + "step": 9392 + }, + { + "epoch": 0.9933375634517766, + "grad_norm": 8.75, + "learning_rate": 4.636210983495127e-06, + "loss": 2.7257, + "step": 9393 + }, + { + "epoch": 0.9934433164128595, + "grad_norm": 8.5, + "learning_rate": 4.636122014802967e-06, + "loss": 2.4227, + "step": 9394 + }, + { + "epoch": 0.9935490693739425, + "grad_norm": 8.4375, + "learning_rate": 4.636033036086861e-06, + "loss": 2.5713, + "step": 9395 + }, + { + "epoch": 0.9936548223350253, + "grad_norm": 8.375, + "learning_rate": 4.6359440473472235e-06, + "loss": 2.5678, + "step": 9396 + }, + { + "epoch": 0.9937605752961083, + "grad_norm": 8.1875, + "learning_rate": 4.635855048584477e-06, + "loss": 2.6052, + "step": 9397 + }, + { + "epoch": 0.9938663282571912, + "grad_norm": 7.4375, + "learning_rate": 4.635766039799035e-06, + "loss": 2.4109, + "step": 9398 + }, + { + "epoch": 0.9939720812182741, + "grad_norm": 8.375, + "learning_rate": 4.635677020991315e-06, + "loss": 2.4453, + "step": 9399 + }, + { + "epoch": 0.994077834179357, + "grad_norm": 7.59375, + "learning_rate": 4.635587992161737e-06, + "loss": 2.3728, + "step": 9400 + }, + { + "epoch": 0.99418358714044, + "grad_norm": 8.75, + "learning_rate": 4.6354989533107195e-06, + "loss": 2.0839, + "step": 9401 + }, + { + "epoch": 0.9942893401015228, + "grad_norm": 7.59375, + "learning_rate": 4.635409904438678e-06, + "loss": 2.6225, + "step": 9402 + }, + { + "epoch": 0.9943950930626058, + "grad_norm": 7.625, + "learning_rate": 4.635320845546031e-06, + "loss": 2.4948, + "step": 9403 + }, + { + "epoch": 0.9945008460236887, + "grad_norm": 8.1875, + "learning_rate": 4.635231776633195e-06, + "loss": 2.5043, + "step": 9404 + }, + { + "epoch": 0.9946065989847716, + "grad_norm": 7.71875, + "learning_rate": 4.635142697700592e-06, + "loss": 2.5367, + "step": 9405 + }, + { + "epoch": 0.9947123519458545, + "grad_norm": 8.0625, + "learning_rate": 4.635053608748637e-06, + "loss": 2.4418, + "step": 9406 + }, + { + "epoch": 0.9948181049069373, + "grad_norm": 8.625, + "learning_rate": 4.634964509777748e-06, + "loss": 2.4321, + "step": 9407 + }, + { + "epoch": 0.9949238578680203, + "grad_norm": 7.75, + "learning_rate": 4.634875400788344e-06, + "loss": 2.6668, + "step": 9408 + }, + { + "epoch": 0.9950296108291032, + "grad_norm": 7.8125, + "learning_rate": 4.634786281780843e-06, + "loss": 2.3509, + "step": 9409 + }, + { + "epoch": 0.9951353637901861, + "grad_norm": 7.96875, + "learning_rate": 4.634697152755663e-06, + "loss": 2.597, + "step": 9410 + }, + { + "epoch": 0.995241116751269, + "grad_norm": 8.0625, + "learning_rate": 4.634608013713222e-06, + "loss": 2.3275, + "step": 9411 + }, + { + "epoch": 0.995346869712352, + "grad_norm": 9.25, + "learning_rate": 4.634518864653938e-06, + "loss": 2.5834, + "step": 9412 + }, + { + "epoch": 0.9954526226734348, + "grad_norm": 7.71875, + "learning_rate": 4.634429705578232e-06, + "loss": 2.5741, + "step": 9413 + }, + { + "epoch": 0.9955583756345178, + "grad_norm": 8.375, + "learning_rate": 4.634340536486519e-06, + "loss": 2.5923, + "step": 9414 + }, + { + "epoch": 0.9956641285956007, + "grad_norm": 8.375, + "learning_rate": 4.6342513573792195e-06, + "loss": 2.7394, + "step": 9415 + }, + { + "epoch": 0.9957698815566836, + "grad_norm": 8.25, + "learning_rate": 4.63416216825675e-06, + "loss": 2.6479, + "step": 9416 + }, + { + "epoch": 0.9958756345177665, + "grad_norm": 8.125, + "learning_rate": 4.634072969119531e-06, + "loss": 2.6215, + "step": 9417 + }, + { + "epoch": 0.9959813874788495, + "grad_norm": 7.75, + "learning_rate": 4.633983759967981e-06, + "loss": 2.2745, + "step": 9418 + }, + { + "epoch": 0.9960871404399323, + "grad_norm": 8.25, + "learning_rate": 4.633894540802517e-06, + "loss": 2.7128, + "step": 9419 + }, + { + "epoch": 0.9961928934010152, + "grad_norm": 7.6875, + "learning_rate": 4.6338053116235584e-06, + "loss": 2.2905, + "step": 9420 + }, + { + "epoch": 0.9962986463620981, + "grad_norm": 8.3125, + "learning_rate": 4.633716072431525e-06, + "loss": 2.7388, + "step": 9421 + }, + { + "epoch": 0.996404399323181, + "grad_norm": 8.25, + "learning_rate": 4.633626823226834e-06, + "loss": 2.6735, + "step": 9422 + }, + { + "epoch": 0.996510152284264, + "grad_norm": 7.78125, + "learning_rate": 4.633537564009905e-06, + "loss": 2.68, + "step": 9423 + }, + { + "epoch": 0.9966159052453468, + "grad_norm": 7.84375, + "learning_rate": 4.633448294781159e-06, + "loss": 2.4529, + "step": 9424 + }, + { + "epoch": 0.9967216582064298, + "grad_norm": 7.875, + "learning_rate": 4.633359015541011e-06, + "loss": 2.6659, + "step": 9425 + }, + { + "epoch": 0.9968274111675127, + "grad_norm": 7.71875, + "learning_rate": 4.633269726289881e-06, + "loss": 2.4323, + "step": 9426 + }, + { + "epoch": 0.9969331641285956, + "grad_norm": 8.125, + "learning_rate": 4.633180427028189e-06, + "loss": 2.5807, + "step": 9427 + }, + { + "epoch": 0.9970389170896785, + "grad_norm": 7.71875, + "learning_rate": 4.633091117756354e-06, + "loss": 2.6242, + "step": 9428 + }, + { + "epoch": 0.9971446700507615, + "grad_norm": 9.25, + "learning_rate": 4.633001798474795e-06, + "loss": 2.691, + "step": 9429 + }, + { + "epoch": 0.9972504230118443, + "grad_norm": 8.1875, + "learning_rate": 4.63291246918393e-06, + "loss": 2.4794, + "step": 9430 + }, + { + "epoch": 0.9973561759729273, + "grad_norm": 7.875, + "learning_rate": 4.63282312988418e-06, + "loss": 2.6492, + "step": 9431 + }, + { + "epoch": 0.9974619289340102, + "grad_norm": 8.25, + "learning_rate": 4.632733780575963e-06, + "loss": 2.5763, + "step": 9432 + }, + { + "epoch": 0.997567681895093, + "grad_norm": 7.53125, + "learning_rate": 4.632644421259699e-06, + "loss": 2.3162, + "step": 9433 + }, + { + "epoch": 0.997673434856176, + "grad_norm": 8.3125, + "learning_rate": 4.632555051935806e-06, + "loss": 2.8176, + "step": 9434 + }, + { + "epoch": 0.9977791878172588, + "grad_norm": 8.5625, + "learning_rate": 4.632465672604705e-06, + "loss": 2.7285, + "step": 9435 + }, + { + "epoch": 0.9978849407783418, + "grad_norm": 7.78125, + "learning_rate": 4.632376283266816e-06, + "loss": 2.4856, + "step": 9436 + }, + { + "epoch": 0.9979906937394247, + "grad_norm": 8.1875, + "learning_rate": 4.632286883922556e-06, + "loss": 2.5785, + "step": 9437 + }, + { + "epoch": 0.9980964467005076, + "grad_norm": 8.25, + "learning_rate": 4.6321974745723455e-06, + "loss": 2.8214, + "step": 9438 + }, + { + "epoch": 0.9982021996615905, + "grad_norm": 8.0, + "learning_rate": 4.632108055216604e-06, + "loss": 2.558, + "step": 9439 + }, + { + "epoch": 0.9983079526226735, + "grad_norm": 8.0625, + "learning_rate": 4.632018625855752e-06, + "loss": 2.3514, + "step": 9440 + }, + { + "epoch": 0.9984137055837563, + "grad_norm": 7.75, + "learning_rate": 4.631929186490209e-06, + "loss": 2.4037, + "step": 9441 + }, + { + "epoch": 0.9985194585448393, + "grad_norm": 8.375, + "learning_rate": 4.631839737120394e-06, + "loss": 2.5476, + "step": 9442 + }, + { + "epoch": 0.9986252115059222, + "grad_norm": 7.78125, + "learning_rate": 4.631750277746726e-06, + "loss": 2.6162, + "step": 9443 + }, + { + "epoch": 0.998730964467005, + "grad_norm": 7.90625, + "learning_rate": 4.631660808369627e-06, + "loss": 2.5674, + "step": 9444 + }, + { + "epoch": 0.998836717428088, + "grad_norm": 8.125, + "learning_rate": 4.631571328989515e-06, + "loss": 2.8488, + "step": 9445 + }, + { + "epoch": 0.998942470389171, + "grad_norm": 7.90625, + "learning_rate": 4.6314818396068105e-06, + "loss": 2.4199, + "step": 9446 + }, + { + "epoch": 0.9990482233502538, + "grad_norm": 8.1875, + "learning_rate": 4.6313923402219345e-06, + "loss": 2.5724, + "step": 9447 + }, + { + "epoch": 0.9991539763113367, + "grad_norm": 7.65625, + "learning_rate": 4.631302830835304e-06, + "loss": 2.4144, + "step": 9448 + }, + { + "epoch": 0.9992597292724196, + "grad_norm": 7.625, + "learning_rate": 4.631213311447342e-06, + "loss": 2.6532, + "step": 9449 + }, + { + "epoch": 0.9993654822335025, + "grad_norm": 8.25, + "learning_rate": 4.631123782058468e-06, + "loss": 2.6658, + "step": 9450 + }, + { + "epoch": 0.9994712351945855, + "grad_norm": 7.8125, + "learning_rate": 4.6310342426691015e-06, + "loss": 2.6964, + "step": 9451 + }, + { + "epoch": 0.9995769881556683, + "grad_norm": 8.25, + "learning_rate": 4.630944693279663e-06, + "loss": 2.6261, + "step": 9452 + }, + { + "epoch": 0.9996827411167513, + "grad_norm": 8.0625, + "learning_rate": 4.630855133890572e-06, + "loss": 2.5987, + "step": 9453 + }, + { + "epoch": 0.9997884940778342, + "grad_norm": 8.8125, + "learning_rate": 4.63076556450225e-06, + "loss": 2.6575, + "step": 9454 + }, + { + "epoch": 0.999894247038917, + "grad_norm": 8.5, + "learning_rate": 4.630675985115116e-06, + "loss": 2.4527, + "step": 9455 + }, + { + "epoch": 1.0, + "grad_norm": 7.90625, + "learning_rate": 4.630586395729592e-06, + "loss": 2.5878, + "step": 9456 + }, + { + "epoch": 1.0001057529610828, + "grad_norm": 8.25, + "learning_rate": 4.630496796346097e-06, + "loss": 2.6615, + "step": 9457 + }, + { + "epoch": 1.000211505922166, + "grad_norm": 7.625, + "learning_rate": 4.630407186965051e-06, + "loss": 2.3279, + "step": 9458 + }, + { + "epoch": 1.0003172588832487, + "grad_norm": 8.25, + "learning_rate": 4.630317567586876e-06, + "loss": 2.5474, + "step": 9459 + }, + { + "epoch": 1.0004230118443316, + "grad_norm": 7.65625, + "learning_rate": 4.630227938211992e-06, + "loss": 2.4571, + "step": 9460 + }, + { + "epoch": 1.0005287648054146, + "grad_norm": 7.6875, + "learning_rate": 4.63013829884082e-06, + "loss": 2.5773, + "step": 9461 + }, + { + "epoch": 1.0006345177664975, + "grad_norm": 8.6875, + "learning_rate": 4.63004864947378e-06, + "loss": 2.681, + "step": 9462 + }, + { + "epoch": 1.0007402707275803, + "grad_norm": 7.75, + "learning_rate": 4.629958990111292e-06, + "loss": 2.4796, + "step": 9463 + }, + { + "epoch": 1.0008460236886634, + "grad_norm": 8.125, + "learning_rate": 4.629869320753778e-06, + "loss": 2.6694, + "step": 9464 + }, + { + "epoch": 1.0009517766497462, + "grad_norm": 7.8125, + "learning_rate": 4.629779641401658e-06, + "loss": 2.7154, + "step": 9465 + }, + { + "epoch": 1.001057529610829, + "grad_norm": 8.125, + "learning_rate": 4.629689952055354e-06, + "loss": 2.434, + "step": 9466 + }, + { + "epoch": 1.0011632825719121, + "grad_norm": 7.59375, + "learning_rate": 4.629600252715286e-06, + "loss": 2.4791, + "step": 9467 + }, + { + "epoch": 1.001269035532995, + "grad_norm": 8.5, + "learning_rate": 4.629510543381874e-06, + "loss": 2.5011, + "step": 9468 + }, + { + "epoch": 1.0013747884940778, + "grad_norm": 8.5, + "learning_rate": 4.62942082405554e-06, + "loss": 2.4104, + "step": 9469 + }, + { + "epoch": 1.0014805414551609, + "grad_norm": 8.25, + "learning_rate": 4.629331094736706e-06, + "loss": 2.7315, + "step": 9470 + }, + { + "epoch": 1.0015862944162437, + "grad_norm": 7.375, + "learning_rate": 4.629241355425792e-06, + "loss": 2.4653, + "step": 9471 + }, + { + "epoch": 1.0016920473773265, + "grad_norm": 8.0, + "learning_rate": 4.629151606123219e-06, + "loss": 2.5533, + "step": 9472 + }, + { + "epoch": 1.0017978003384094, + "grad_norm": 8.0625, + "learning_rate": 4.629061846829408e-06, + "loss": 2.588, + "step": 9473 + }, + { + "epoch": 1.0019035532994924, + "grad_norm": 8.0, + "learning_rate": 4.628972077544781e-06, + "loss": 2.4671, + "step": 9474 + }, + { + "epoch": 1.0020093062605753, + "grad_norm": 9.0, + "learning_rate": 4.628882298269758e-06, + "loss": 2.5686, + "step": 9475 + }, + { + "epoch": 1.002115059221658, + "grad_norm": 8.4375, + "learning_rate": 4.6287925090047615e-06, + "loss": 2.7392, + "step": 9476 + }, + { + "epoch": 1.0022208121827412, + "grad_norm": 7.96875, + "learning_rate": 4.628702709750213e-06, + "loss": 2.6317, + "step": 9477 + }, + { + "epoch": 1.002326565143824, + "grad_norm": 7.8125, + "learning_rate": 4.628612900506532e-06, + "loss": 2.4455, + "step": 9478 + }, + { + "epoch": 1.0024323181049068, + "grad_norm": 7.59375, + "learning_rate": 4.628523081274142e-06, + "loss": 2.472, + "step": 9479 + }, + { + "epoch": 1.00253807106599, + "grad_norm": 7.78125, + "learning_rate": 4.628433252053463e-06, + "loss": 2.4253, + "step": 9480 + }, + { + "epoch": 1.0026438240270727, + "grad_norm": 7.875, + "learning_rate": 4.6283434128449185e-06, + "loss": 2.4606, + "step": 9481 + }, + { + "epoch": 1.0027495769881556, + "grad_norm": 8.125, + "learning_rate": 4.628253563648928e-06, + "loss": 2.5066, + "step": 9482 + }, + { + "epoch": 1.0028553299492386, + "grad_norm": 8.125, + "learning_rate": 4.628163704465914e-06, + "loss": 2.1431, + "step": 9483 + }, + { + "epoch": 1.0029610829103215, + "grad_norm": 8.9375, + "learning_rate": 4.628073835296299e-06, + "loss": 2.5151, + "step": 9484 + }, + { + "epoch": 1.0030668358714043, + "grad_norm": 7.96875, + "learning_rate": 4.627983956140503e-06, + "loss": 2.4119, + "step": 9485 + }, + { + "epoch": 1.0031725888324874, + "grad_norm": 7.78125, + "learning_rate": 4.627894066998948e-06, + "loss": 2.444, + "step": 9486 + }, + { + "epoch": 1.0032783417935702, + "grad_norm": 7.90625, + "learning_rate": 4.627804167872057e-06, + "loss": 2.3662, + "step": 9487 + }, + { + "epoch": 1.003384094754653, + "grad_norm": 8.3125, + "learning_rate": 4.627714258760252e-06, + "loss": 2.5356, + "step": 9488 + }, + { + "epoch": 1.0034898477157361, + "grad_norm": 7.71875, + "learning_rate": 4.627624339663953e-06, + "loss": 2.4205, + "step": 9489 + }, + { + "epoch": 1.003595600676819, + "grad_norm": 8.0625, + "learning_rate": 4.627534410583584e-06, + "loss": 2.5226, + "step": 9490 + }, + { + "epoch": 1.0037013536379018, + "grad_norm": 7.75, + "learning_rate": 4.627444471519566e-06, + "loss": 2.3437, + "step": 9491 + }, + { + "epoch": 1.0038071065989849, + "grad_norm": 8.375, + "learning_rate": 4.627354522472321e-06, + "loss": 2.6765, + "step": 9492 + }, + { + "epoch": 1.0039128595600677, + "grad_norm": 8.25, + "learning_rate": 4.627264563442271e-06, + "loss": 2.7626, + "step": 9493 + }, + { + "epoch": 1.0040186125211505, + "grad_norm": 7.75, + "learning_rate": 4.627174594429839e-06, + "loss": 2.5402, + "step": 9494 + }, + { + "epoch": 1.0041243654822336, + "grad_norm": 8.4375, + "learning_rate": 4.627084615435447e-06, + "loss": 2.736, + "step": 9495 + }, + { + "epoch": 1.0042301184433164, + "grad_norm": 7.78125, + "learning_rate": 4.626994626459516e-06, + "loss": 2.4372, + "step": 9496 + }, + { + "epoch": 1.0043358714043993, + "grad_norm": 8.0625, + "learning_rate": 4.626904627502469e-06, + "loss": 2.4646, + "step": 9497 + }, + { + "epoch": 1.0044416243654823, + "grad_norm": 8.625, + "learning_rate": 4.6268146185647285e-06, + "loss": 2.8457, + "step": 9498 + }, + { + "epoch": 1.0045473773265652, + "grad_norm": 7.875, + "learning_rate": 4.626724599646717e-06, + "loss": 2.5819, + "step": 9499 + }, + { + "epoch": 1.004653130287648, + "grad_norm": 7.90625, + "learning_rate": 4.626634570748857e-06, + "loss": 2.4855, + "step": 9500 + }, + { + "epoch": 1.0047588832487309, + "grad_norm": 8.1875, + "learning_rate": 4.626544531871571e-06, + "loss": 2.713, + "step": 9501 + }, + { + "epoch": 1.004864636209814, + "grad_norm": 8.1875, + "learning_rate": 4.626454483015281e-06, + "loss": 2.4935, + "step": 9502 + }, + { + "epoch": 1.0049703891708968, + "grad_norm": 7.5, + "learning_rate": 4.6263644241804095e-06, + "loss": 2.4667, + "step": 9503 + }, + { + "epoch": 1.0050761421319796, + "grad_norm": 8.0, + "learning_rate": 4.62627435536738e-06, + "loss": 2.5321, + "step": 9504 + }, + { + "epoch": 1.0051818950930627, + "grad_norm": 7.46875, + "learning_rate": 4.626184276576614e-06, + "loss": 2.3695, + "step": 9505 + }, + { + "epoch": 1.0052876480541455, + "grad_norm": 7.59375, + "learning_rate": 4.626094187808534e-06, + "loss": 2.5113, + "step": 9506 + }, + { + "epoch": 1.0053934010152283, + "grad_norm": 7.96875, + "learning_rate": 4.626004089063565e-06, + "loss": 2.4535, + "step": 9507 + }, + { + "epoch": 1.0054991539763114, + "grad_norm": 7.9375, + "learning_rate": 4.625913980342128e-06, + "loss": 2.5589, + "step": 9508 + }, + { + "epoch": 1.0056049069373942, + "grad_norm": 7.9375, + "learning_rate": 4.625823861644646e-06, + "loss": 2.5045, + "step": 9509 + }, + { + "epoch": 1.005710659898477, + "grad_norm": 7.9375, + "learning_rate": 4.625733732971541e-06, + "loss": 2.5227, + "step": 9510 + }, + { + "epoch": 1.0058164128595601, + "grad_norm": 7.9375, + "learning_rate": 4.6256435943232385e-06, + "loss": 2.5073, + "step": 9511 + }, + { + "epoch": 1.005922165820643, + "grad_norm": 7.9375, + "learning_rate": 4.62555344570016e-06, + "loss": 2.6519, + "step": 9512 + }, + { + "epoch": 1.0060279187817258, + "grad_norm": 8.0, + "learning_rate": 4.625463287102727e-06, + "loss": 2.5724, + "step": 9513 + }, + { + "epoch": 1.0061336717428089, + "grad_norm": 7.875, + "learning_rate": 4.625373118531366e-06, + "loss": 2.5085, + "step": 9514 + }, + { + "epoch": 1.0062394247038917, + "grad_norm": 7.90625, + "learning_rate": 4.6252829399864965e-06, + "loss": 2.4375, + "step": 9515 + }, + { + "epoch": 1.0063451776649746, + "grad_norm": 8.0625, + "learning_rate": 4.625192751468544e-06, + "loss": 2.489, + "step": 9516 + }, + { + "epoch": 1.0064509306260576, + "grad_norm": 8.1875, + "learning_rate": 4.625102552977932e-06, + "loss": 2.5311, + "step": 9517 + }, + { + "epoch": 1.0065566835871405, + "grad_norm": 8.0, + "learning_rate": 4.625012344515082e-06, + "loss": 2.4673, + "step": 9518 + }, + { + "epoch": 1.0066624365482233, + "grad_norm": 8.0625, + "learning_rate": 4.624922126080418e-06, + "loss": 2.6528, + "step": 9519 + }, + { + "epoch": 1.0067681895093064, + "grad_norm": 7.9375, + "learning_rate": 4.624831897674364e-06, + "loss": 2.4364, + "step": 9520 + }, + { + "epoch": 1.0068739424703892, + "grad_norm": 7.96875, + "learning_rate": 4.624741659297342e-06, + "loss": 2.6597, + "step": 9521 + }, + { + "epoch": 1.006979695431472, + "grad_norm": 8.0625, + "learning_rate": 4.624651410949777e-06, + "loss": 2.4302, + "step": 9522 + }, + { + "epoch": 1.007085448392555, + "grad_norm": 8.5625, + "learning_rate": 4.624561152632092e-06, + "loss": 2.625, + "step": 9523 + }, + { + "epoch": 1.007191201353638, + "grad_norm": 8.4375, + "learning_rate": 4.62447088434471e-06, + "loss": 2.2812, + "step": 9524 + }, + { + "epoch": 1.0072969543147208, + "grad_norm": 8.0625, + "learning_rate": 4.624380606088056e-06, + "loss": 2.5825, + "step": 9525 + }, + { + "epoch": 1.0074027072758038, + "grad_norm": 8.25, + "learning_rate": 4.624290317862552e-06, + "loss": 2.7296, + "step": 9526 + }, + { + "epoch": 1.0075084602368867, + "grad_norm": 7.65625, + "learning_rate": 4.624200019668622e-06, + "loss": 2.4759, + "step": 9527 + }, + { + "epoch": 1.0076142131979695, + "grad_norm": 7.96875, + "learning_rate": 4.6241097115066905e-06, + "loss": 2.4415, + "step": 9528 + }, + { + "epoch": 1.0077199661590523, + "grad_norm": 8.0625, + "learning_rate": 4.62401939337718e-06, + "loss": 2.5171, + "step": 9529 + }, + { + "epoch": 1.0078257191201354, + "grad_norm": 7.9375, + "learning_rate": 4.623929065280517e-06, + "loss": 2.505, + "step": 9530 + }, + { + "epoch": 1.0079314720812182, + "grad_norm": 8.25, + "learning_rate": 4.623838727217122e-06, + "loss": 2.6015, + "step": 9531 + }, + { + "epoch": 1.008037225042301, + "grad_norm": 8.1875, + "learning_rate": 4.623748379187422e-06, + "loss": 2.2823, + "step": 9532 + }, + { + "epoch": 1.0081429780033841, + "grad_norm": 8.125, + "learning_rate": 4.623658021191837e-06, + "loss": 2.675, + "step": 9533 + }, + { + "epoch": 1.008248730964467, + "grad_norm": 8.1875, + "learning_rate": 4.623567653230796e-06, + "loss": 2.7232, + "step": 9534 + }, + { + "epoch": 1.0083544839255498, + "grad_norm": 8.1875, + "learning_rate": 4.623477275304719e-06, + "loss": 2.5806, + "step": 9535 + }, + { + "epoch": 1.0084602368866329, + "grad_norm": 8.625, + "learning_rate": 4.623386887414032e-06, + "loss": 2.6171, + "step": 9536 + }, + { + "epoch": 1.0085659898477157, + "grad_norm": 7.90625, + "learning_rate": 4.623296489559158e-06, + "loss": 2.2692, + "step": 9537 + }, + { + "epoch": 1.0086717428087986, + "grad_norm": 8.5, + "learning_rate": 4.623206081740523e-06, + "loss": 2.7381, + "step": 9538 + }, + { + "epoch": 1.0087774957698816, + "grad_norm": 10.3125, + "learning_rate": 4.62311566395855e-06, + "loss": 2.5547, + "step": 9539 + }, + { + "epoch": 1.0088832487309645, + "grad_norm": 7.90625, + "learning_rate": 4.623025236213664e-06, + "loss": 2.4773, + "step": 9540 + }, + { + "epoch": 1.0089890016920473, + "grad_norm": 8.4375, + "learning_rate": 4.622934798506288e-06, + "loss": 2.5961, + "step": 9541 + }, + { + "epoch": 1.0090947546531304, + "grad_norm": 11.0, + "learning_rate": 4.622844350836848e-06, + "loss": 2.7403, + "step": 9542 + }, + { + "epoch": 1.0092005076142132, + "grad_norm": 7.875, + "learning_rate": 4.622753893205767e-06, + "loss": 2.5035, + "step": 9543 + }, + { + "epoch": 1.009306260575296, + "grad_norm": 8.0625, + "learning_rate": 4.622663425613471e-06, + "loss": 2.7234, + "step": 9544 + }, + { + "epoch": 1.009412013536379, + "grad_norm": 7.75, + "learning_rate": 4.6225729480603826e-06, + "loss": 2.4805, + "step": 9545 + }, + { + "epoch": 1.009517766497462, + "grad_norm": 7.78125, + "learning_rate": 4.622482460546928e-06, + "loss": 2.4584, + "step": 9546 + }, + { + "epoch": 1.0096235194585448, + "grad_norm": 8.4375, + "learning_rate": 4.6223919630735315e-06, + "loss": 2.5752, + "step": 9547 + }, + { + "epoch": 1.0097292724196278, + "grad_norm": 8.1875, + "learning_rate": 4.6223014556406174e-06, + "loss": 2.3554, + "step": 9548 + }, + { + "epoch": 1.0098350253807107, + "grad_norm": 7.9375, + "learning_rate": 4.62221093824861e-06, + "loss": 2.4803, + "step": 9549 + }, + { + "epoch": 1.0099407783417935, + "grad_norm": 7.84375, + "learning_rate": 4.622120410897935e-06, + "loss": 2.5268, + "step": 9550 + }, + { + "epoch": 1.0100465313028766, + "grad_norm": 8.25, + "learning_rate": 4.622029873589017e-06, + "loss": 2.6514, + "step": 9551 + }, + { + "epoch": 1.0101522842639594, + "grad_norm": 8.4375, + "learning_rate": 4.62193932632228e-06, + "loss": 2.4612, + "step": 9552 + }, + { + "epoch": 1.0102580372250423, + "grad_norm": 8.375, + "learning_rate": 4.6218487690981505e-06, + "loss": 2.4417, + "step": 9553 + }, + { + "epoch": 1.0103637901861253, + "grad_norm": 8.125, + "learning_rate": 4.621758201917053e-06, + "loss": 2.776, + "step": 9554 + }, + { + "epoch": 1.0104695431472082, + "grad_norm": 8.375, + "learning_rate": 4.6216676247794105e-06, + "loss": 2.6461, + "step": 9555 + }, + { + "epoch": 1.010575296108291, + "grad_norm": 7.84375, + "learning_rate": 4.621577037685651e-06, + "loss": 2.6231, + "step": 9556 + }, + { + "epoch": 1.0106810490693738, + "grad_norm": 7.5625, + "learning_rate": 4.621486440636197e-06, + "loss": 2.4221, + "step": 9557 + }, + { + "epoch": 1.010786802030457, + "grad_norm": 7.90625, + "learning_rate": 4.621395833631475e-06, + "loss": 2.4612, + "step": 9558 + }, + { + "epoch": 1.0108925549915397, + "grad_norm": 8.3125, + "learning_rate": 4.621305216671911e-06, + "loss": 2.7098, + "step": 9559 + }, + { + "epoch": 1.0109983079526226, + "grad_norm": 12.3125, + "learning_rate": 4.621214589757928e-06, + "loss": 2.5486, + "step": 9560 + }, + { + "epoch": 1.0111040609137056, + "grad_norm": 7.9375, + "learning_rate": 4.621123952889953e-06, + "loss": 2.5423, + "step": 9561 + }, + { + "epoch": 1.0112098138747885, + "grad_norm": 7.96875, + "learning_rate": 4.62103330606841e-06, + "loss": 2.6066, + "step": 9562 + }, + { + "epoch": 1.0113155668358713, + "grad_norm": 8.0625, + "learning_rate": 4.620942649293726e-06, + "loss": 2.5616, + "step": 9563 + }, + { + "epoch": 1.0114213197969544, + "grad_norm": 8.0, + "learning_rate": 4.620851982566326e-06, + "loss": 2.5295, + "step": 9564 + }, + { + "epoch": 1.0115270727580372, + "grad_norm": 8.25, + "learning_rate": 4.620761305886634e-06, + "loss": 2.5365, + "step": 9565 + }, + { + "epoch": 1.01163282571912, + "grad_norm": 8.375, + "learning_rate": 4.620670619255078e-06, + "loss": 2.5864, + "step": 9566 + }, + { + "epoch": 1.0117385786802031, + "grad_norm": 8.5, + "learning_rate": 4.62057992267208e-06, + "loss": 2.333, + "step": 9567 + }, + { + "epoch": 1.011844331641286, + "grad_norm": 7.6875, + "learning_rate": 4.62048921613807e-06, + "loss": 2.4111, + "step": 9568 + }, + { + "epoch": 1.0119500846023688, + "grad_norm": 8.125, + "learning_rate": 4.620398499653469e-06, + "loss": 2.5108, + "step": 9569 + }, + { + "epoch": 1.0120558375634519, + "grad_norm": 8.1875, + "learning_rate": 4.620307773218707e-06, + "loss": 2.433, + "step": 9570 + }, + { + "epoch": 1.0121615905245347, + "grad_norm": 8.4375, + "learning_rate": 4.620217036834207e-06, + "loss": 2.3205, + "step": 9571 + }, + { + "epoch": 1.0122673434856175, + "grad_norm": 8.25, + "learning_rate": 4.620126290500395e-06, + "loss": 2.61, + "step": 9572 + }, + { + "epoch": 1.0123730964467006, + "grad_norm": 7.84375, + "learning_rate": 4.620035534217699e-06, + "loss": 2.501, + "step": 9573 + }, + { + "epoch": 1.0124788494077834, + "grad_norm": 8.5, + "learning_rate": 4.619944767986542e-06, + "loss": 2.6412, + "step": 9574 + }, + { + "epoch": 1.0125846023688663, + "grad_norm": 8.875, + "learning_rate": 4.619853991807351e-06, + "loss": 2.5708, + "step": 9575 + }, + { + "epoch": 1.0126903553299493, + "grad_norm": 7.71875, + "learning_rate": 4.619763205680553e-06, + "loss": 2.6798, + "step": 9576 + }, + { + "epoch": 1.0127961082910322, + "grad_norm": 7.9375, + "learning_rate": 4.619672409606574e-06, + "loss": 2.4106, + "step": 9577 + }, + { + "epoch": 1.012901861252115, + "grad_norm": 7.90625, + "learning_rate": 4.619581603585838e-06, + "loss": 2.5505, + "step": 9578 + }, + { + "epoch": 1.013007614213198, + "grad_norm": 7.9375, + "learning_rate": 4.619490787618773e-06, + "loss": 2.5665, + "step": 9579 + }, + { + "epoch": 1.013113367174281, + "grad_norm": 7.875, + "learning_rate": 4.619399961705804e-06, + "loss": 2.3747, + "step": 9580 + }, + { + "epoch": 1.0132191201353637, + "grad_norm": 8.0625, + "learning_rate": 4.619309125847358e-06, + "loss": 2.7706, + "step": 9581 + }, + { + "epoch": 1.0133248730964468, + "grad_norm": 8.25, + "learning_rate": 4.619218280043861e-06, + "loss": 2.4388, + "step": 9582 + }, + { + "epoch": 1.0134306260575296, + "grad_norm": 7.9375, + "learning_rate": 4.61912742429574e-06, + "loss": 2.537, + "step": 9583 + }, + { + "epoch": 1.0135363790186125, + "grad_norm": 8.0625, + "learning_rate": 4.61903655860342e-06, + "loss": 2.5657, + "step": 9584 + }, + { + "epoch": 1.0136421319796953, + "grad_norm": 7.90625, + "learning_rate": 4.618945682967329e-06, + "loss": 2.3818, + "step": 9585 + }, + { + "epoch": 1.0137478849407784, + "grad_norm": 7.78125, + "learning_rate": 4.618854797387891e-06, + "loss": 2.3953, + "step": 9586 + }, + { + "epoch": 1.0138536379018612, + "grad_norm": 7.8125, + "learning_rate": 4.618763901865535e-06, + "loss": 2.5803, + "step": 9587 + }, + { + "epoch": 1.013959390862944, + "grad_norm": 8.25, + "learning_rate": 4.618672996400687e-06, + "loss": 2.4259, + "step": 9588 + }, + { + "epoch": 1.0140651438240271, + "grad_norm": 7.71875, + "learning_rate": 4.6185820809937724e-06, + "loss": 2.4831, + "step": 9589 + }, + { + "epoch": 1.01417089678511, + "grad_norm": 7.9375, + "learning_rate": 4.618491155645218e-06, + "loss": 2.3911, + "step": 9590 + }, + { + "epoch": 1.0142766497461928, + "grad_norm": 8.6875, + "learning_rate": 4.618400220355452e-06, + "loss": 2.4234, + "step": 9591 + }, + { + "epoch": 1.0143824027072759, + "grad_norm": 8.0, + "learning_rate": 4.6183092751248994e-06, + "loss": 2.3182, + "step": 9592 + }, + { + "epoch": 1.0144881556683587, + "grad_norm": 8.0, + "learning_rate": 4.618218319953989e-06, + "loss": 2.5152, + "step": 9593 + }, + { + "epoch": 1.0145939086294415, + "grad_norm": 8.25, + "learning_rate": 4.618127354843145e-06, + "loss": 2.4419, + "step": 9594 + }, + { + "epoch": 1.0146996615905246, + "grad_norm": 8.125, + "learning_rate": 4.618036379792796e-06, + "loss": 2.5585, + "step": 9595 + }, + { + "epoch": 1.0148054145516074, + "grad_norm": 7.90625, + "learning_rate": 4.617945394803369e-06, + "loss": 2.6454, + "step": 9596 + }, + { + "epoch": 1.0149111675126903, + "grad_norm": 8.0625, + "learning_rate": 4.61785439987529e-06, + "loss": 2.2104, + "step": 9597 + }, + { + "epoch": 1.0150169204737733, + "grad_norm": 8.0625, + "learning_rate": 4.617763395008987e-06, + "loss": 2.4156, + "step": 9598 + }, + { + "epoch": 1.0151226734348562, + "grad_norm": 7.9375, + "learning_rate": 4.617672380204886e-06, + "loss": 2.4832, + "step": 9599 + }, + { + "epoch": 1.015228426395939, + "grad_norm": 8.625, + "learning_rate": 4.617581355463415e-06, + "loss": 2.5335, + "step": 9600 + }, + { + "epoch": 1.015334179357022, + "grad_norm": 7.8125, + "learning_rate": 4.617490320785e-06, + "loss": 2.5977, + "step": 9601 + }, + { + "epoch": 1.015439932318105, + "grad_norm": 8.375, + "learning_rate": 4.61739927617007e-06, + "loss": 2.389, + "step": 9602 + }, + { + "epoch": 1.0155456852791878, + "grad_norm": 8.5625, + "learning_rate": 4.6173082216190505e-06, + "loss": 2.5395, + "step": 9603 + }, + { + "epoch": 1.0156514382402708, + "grad_norm": 8.125, + "learning_rate": 4.61721715713237e-06, + "loss": 2.4483, + "step": 9604 + }, + { + "epoch": 1.0157571912013537, + "grad_norm": 7.9375, + "learning_rate": 4.6171260827104546e-06, + "loss": 2.4927, + "step": 9605 + }, + { + "epoch": 1.0158629441624365, + "grad_norm": 7.65625, + "learning_rate": 4.617034998353733e-06, + "loss": 2.4393, + "step": 9606 + }, + { + "epoch": 1.0159686971235196, + "grad_norm": 7.9375, + "learning_rate": 4.616943904062632e-06, + "loss": 2.602, + "step": 9607 + }, + { + "epoch": 1.0160744500846024, + "grad_norm": 8.0625, + "learning_rate": 4.616852799837578e-06, + "loss": 2.4504, + "step": 9608 + }, + { + "epoch": 1.0161802030456852, + "grad_norm": 8.3125, + "learning_rate": 4.6167616856790005e-06, + "loss": 2.5238, + "step": 9609 + }, + { + "epoch": 1.0162859560067683, + "grad_norm": 8.375, + "learning_rate": 4.616670561587326e-06, + "loss": 2.3842, + "step": 9610 + }, + { + "epoch": 1.0163917089678511, + "grad_norm": 7.78125, + "learning_rate": 4.616579427562982e-06, + "loss": 2.5044, + "step": 9611 + }, + { + "epoch": 1.016497461928934, + "grad_norm": 8.0, + "learning_rate": 4.6164882836063965e-06, + "loss": 2.3158, + "step": 9612 + }, + { + "epoch": 1.0166032148900168, + "grad_norm": 8.3125, + "learning_rate": 4.616397129717998e-06, + "loss": 2.6486, + "step": 9613 + }, + { + "epoch": 1.0167089678510999, + "grad_norm": 8.375, + "learning_rate": 4.616305965898212e-06, + "loss": 2.6747, + "step": 9614 + }, + { + "epoch": 1.0168147208121827, + "grad_norm": 8.0, + "learning_rate": 4.616214792147468e-06, + "loss": 2.4405, + "step": 9615 + }, + { + "epoch": 1.0169204737732656, + "grad_norm": 7.21875, + "learning_rate": 4.616123608466194e-06, + "loss": 2.4971, + "step": 9616 + }, + { + "epoch": 1.0170262267343486, + "grad_norm": 8.4375, + "learning_rate": 4.616032414854817e-06, + "loss": 2.4954, + "step": 9617 + }, + { + "epoch": 1.0171319796954315, + "grad_norm": 7.96875, + "learning_rate": 4.615941211313765e-06, + "loss": 2.5738, + "step": 9618 + }, + { + "epoch": 1.0172377326565143, + "grad_norm": 8.0625, + "learning_rate": 4.615849997843467e-06, + "loss": 2.5495, + "step": 9619 + }, + { + "epoch": 1.0173434856175974, + "grad_norm": 7.75, + "learning_rate": 4.61575877444435e-06, + "loss": 2.7302, + "step": 9620 + }, + { + "epoch": 1.0174492385786802, + "grad_norm": 7.8125, + "learning_rate": 4.615667541116842e-06, + "loss": 2.5035, + "step": 9621 + }, + { + "epoch": 1.017554991539763, + "grad_norm": 8.0625, + "learning_rate": 4.615576297861372e-06, + "loss": 2.6272, + "step": 9622 + }, + { + "epoch": 1.017660744500846, + "grad_norm": 7.65625, + "learning_rate": 4.615485044678367e-06, + "loss": 2.4464, + "step": 9623 + }, + { + "epoch": 1.017766497461929, + "grad_norm": 8.25, + "learning_rate": 4.615393781568257e-06, + "loss": 2.5598, + "step": 9624 + }, + { + "epoch": 1.0178722504230118, + "grad_norm": 8.625, + "learning_rate": 4.615302508531468e-06, + "loss": 2.3708, + "step": 9625 + }, + { + "epoch": 1.0179780033840948, + "grad_norm": 8.1875, + "learning_rate": 4.615211225568431e-06, + "loss": 2.5471, + "step": 9626 + }, + { + "epoch": 1.0180837563451777, + "grad_norm": 8.5, + "learning_rate": 4.615119932679571e-06, + "loss": 2.6477, + "step": 9627 + }, + { + "epoch": 1.0181895093062605, + "grad_norm": 8.9375, + "learning_rate": 4.615028629865319e-06, + "loss": 2.9853, + "step": 9628 + }, + { + "epoch": 1.0182952622673436, + "grad_norm": 8.125, + "learning_rate": 4.614937317126103e-06, + "loss": 2.5881, + "step": 9629 + }, + { + "epoch": 1.0184010152284264, + "grad_norm": 8.0625, + "learning_rate": 4.6148459944623504e-06, + "loss": 2.5937, + "step": 9630 + }, + { + "epoch": 1.0185067681895092, + "grad_norm": 8.0, + "learning_rate": 4.614754661874491e-06, + "loss": 2.5702, + "step": 9631 + }, + { + "epoch": 1.0186125211505923, + "grad_norm": 8.375, + "learning_rate": 4.614663319362953e-06, + "loss": 2.5147, + "step": 9632 + }, + { + "epoch": 1.0187182741116751, + "grad_norm": 7.875, + "learning_rate": 4.614571966928165e-06, + "loss": 2.2769, + "step": 9633 + }, + { + "epoch": 1.018824027072758, + "grad_norm": 8.25, + "learning_rate": 4.614480604570555e-06, + "loss": 2.5799, + "step": 9634 + }, + { + "epoch": 1.018929780033841, + "grad_norm": 7.9375, + "learning_rate": 4.614389232290553e-06, + "loss": 2.669, + "step": 9635 + }, + { + "epoch": 1.0190355329949239, + "grad_norm": 7.9375, + "learning_rate": 4.6142978500885866e-06, + "loss": 2.4555, + "step": 9636 + }, + { + "epoch": 1.0191412859560067, + "grad_norm": 7.78125, + "learning_rate": 4.6142064579650854e-06, + "loss": 2.5026, + "step": 9637 + }, + { + "epoch": 1.0192470389170898, + "grad_norm": 8.4375, + "learning_rate": 4.614115055920477e-06, + "loss": 2.6852, + "step": 9638 + }, + { + "epoch": 1.0193527918781726, + "grad_norm": 8.6875, + "learning_rate": 4.614023643955193e-06, + "loss": 2.4321, + "step": 9639 + }, + { + "epoch": 1.0194585448392555, + "grad_norm": 8.5625, + "learning_rate": 4.613932222069659e-06, + "loss": 2.6356, + "step": 9640 + }, + { + "epoch": 1.0195642978003383, + "grad_norm": 8.0625, + "learning_rate": 4.613840790264307e-06, + "loss": 2.6594, + "step": 9641 + }, + { + "epoch": 1.0196700507614214, + "grad_norm": 7.84375, + "learning_rate": 4.613749348539564e-06, + "loss": 2.4653, + "step": 9642 + }, + { + "epoch": 1.0197758037225042, + "grad_norm": 8.125, + "learning_rate": 4.6136578968958594e-06, + "loss": 2.5492, + "step": 9643 + }, + { + "epoch": 1.019881556683587, + "grad_norm": 8.1875, + "learning_rate": 4.613566435333623e-06, + "loss": 2.6644, + "step": 9644 + }, + { + "epoch": 1.01998730964467, + "grad_norm": 8.125, + "learning_rate": 4.613474963853285e-06, + "loss": 2.5937, + "step": 9645 + }, + { + "epoch": 1.020093062605753, + "grad_norm": 8.625, + "learning_rate": 4.6133834824552715e-06, + "loss": 2.6964, + "step": 9646 + }, + { + "epoch": 1.0201988155668358, + "grad_norm": 8.5625, + "learning_rate": 4.613291991140015e-06, + "loss": 2.7309, + "step": 9647 + }, + { + "epoch": 1.0203045685279188, + "grad_norm": 8.5625, + "learning_rate": 4.613200489907943e-06, + "loss": 2.6103, + "step": 9648 + }, + { + "epoch": 1.0204103214890017, + "grad_norm": 7.71875, + "learning_rate": 4.613108978759485e-06, + "loss": 2.6249, + "step": 9649 + }, + { + "epoch": 1.0205160744500845, + "grad_norm": 8.8125, + "learning_rate": 4.613017457695072e-06, + "loss": 2.7769, + "step": 9650 + }, + { + "epoch": 1.0206218274111676, + "grad_norm": 7.84375, + "learning_rate": 4.61292592671513e-06, + "loss": 2.6616, + "step": 9651 + }, + { + "epoch": 1.0207275803722504, + "grad_norm": 7.875, + "learning_rate": 4.612834385820093e-06, + "loss": 2.4475, + "step": 9652 + }, + { + "epoch": 1.0208333333333333, + "grad_norm": 8.125, + "learning_rate": 4.6127428350103875e-06, + "loss": 2.4376, + "step": 9653 + }, + { + "epoch": 1.0209390862944163, + "grad_norm": 8.1875, + "learning_rate": 4.612651274286444e-06, + "loss": 2.4983, + "step": 9654 + }, + { + "epoch": 1.0210448392554992, + "grad_norm": 7.78125, + "learning_rate": 4.612559703648692e-06, + "loss": 2.6643, + "step": 9655 + }, + { + "epoch": 1.021150592216582, + "grad_norm": 7.9375, + "learning_rate": 4.612468123097561e-06, + "loss": 2.6883, + "step": 9656 + }, + { + "epoch": 1.021256345177665, + "grad_norm": 8.875, + "learning_rate": 4.6123765326334825e-06, + "loss": 2.6885, + "step": 9657 + }, + { + "epoch": 1.021362098138748, + "grad_norm": 8.125, + "learning_rate": 4.612284932256884e-06, + "loss": 2.5957, + "step": 9658 + }, + { + "epoch": 1.0214678510998307, + "grad_norm": 8.25, + "learning_rate": 4.612193321968195e-06, + "loss": 2.7402, + "step": 9659 + }, + { + "epoch": 1.0215736040609138, + "grad_norm": 11.0625, + "learning_rate": 4.6121017017678485e-06, + "loss": 2.6289, + "step": 9660 + }, + { + "epoch": 1.0216793570219966, + "grad_norm": 8.125, + "learning_rate": 4.612010071656271e-06, + "loss": 2.4294, + "step": 9661 + }, + { + "epoch": 1.0217851099830795, + "grad_norm": 8.1875, + "learning_rate": 4.611918431633895e-06, + "loss": 2.4461, + "step": 9662 + }, + { + "epoch": 1.0218908629441625, + "grad_norm": 9.1875, + "learning_rate": 4.611826781701149e-06, + "loss": 2.6427, + "step": 9663 + }, + { + "epoch": 1.0219966159052454, + "grad_norm": 8.9375, + "learning_rate": 4.6117351218584646e-06, + "loss": 2.6611, + "step": 9664 + }, + { + "epoch": 1.0221023688663282, + "grad_norm": 8.1875, + "learning_rate": 4.61164345210627e-06, + "loss": 2.5269, + "step": 9665 + }, + { + "epoch": 1.0222081218274113, + "grad_norm": 8.125, + "learning_rate": 4.611551772444996e-06, + "loss": 2.595, + "step": 9666 + }, + { + "epoch": 1.0223138747884941, + "grad_norm": 7.90625, + "learning_rate": 4.6114600828750745e-06, + "loss": 2.2262, + "step": 9667 + }, + { + "epoch": 1.022419627749577, + "grad_norm": 7.65625, + "learning_rate": 4.611368383396934e-06, + "loss": 2.4345, + "step": 9668 + }, + { + "epoch": 1.0225253807106598, + "grad_norm": 7.5625, + "learning_rate": 4.611276674011004e-06, + "loss": 2.4967, + "step": 9669 + }, + { + "epoch": 1.0226311336717429, + "grad_norm": 8.3125, + "learning_rate": 4.611184954717717e-06, + "loss": 2.6108, + "step": 9670 + }, + { + "epoch": 1.0227368866328257, + "grad_norm": 8.125, + "learning_rate": 4.6110932255175025e-06, + "loss": 2.4206, + "step": 9671 + }, + { + "epoch": 1.0228426395939085, + "grad_norm": 8.125, + "learning_rate": 4.611001486410791e-06, + "loss": 2.5728, + "step": 9672 + }, + { + "epoch": 1.0229483925549916, + "grad_norm": 8.0, + "learning_rate": 4.610909737398012e-06, + "loss": 2.6135, + "step": 9673 + }, + { + "epoch": 1.0230541455160744, + "grad_norm": 7.84375, + "learning_rate": 4.610817978479597e-06, + "loss": 2.6516, + "step": 9674 + }, + { + "epoch": 1.0231598984771573, + "grad_norm": 8.375, + "learning_rate": 4.610726209655977e-06, + "loss": 2.744, + "step": 9675 + }, + { + "epoch": 1.0232656514382403, + "grad_norm": 8.0, + "learning_rate": 4.6106344309275825e-06, + "loss": 2.621, + "step": 9676 + }, + { + "epoch": 1.0233714043993232, + "grad_norm": 8.0625, + "learning_rate": 4.610542642294844e-06, + "loss": 2.3998, + "step": 9677 + }, + { + "epoch": 1.023477157360406, + "grad_norm": 8.3125, + "learning_rate": 4.610450843758191e-06, + "loss": 2.6142, + "step": 9678 + }, + { + "epoch": 1.023582910321489, + "grad_norm": 7.78125, + "learning_rate": 4.610359035318056e-06, + "loss": 2.5523, + "step": 9679 + }, + { + "epoch": 1.023688663282572, + "grad_norm": 8.3125, + "learning_rate": 4.610267216974869e-06, + "loss": 2.674, + "step": 9680 + }, + { + "epoch": 1.0237944162436547, + "grad_norm": 8.75, + "learning_rate": 4.610175388729061e-06, + "loss": 2.461, + "step": 9681 + }, + { + "epoch": 1.0239001692047378, + "grad_norm": 8.4375, + "learning_rate": 4.610083550581063e-06, + "loss": 2.6856, + "step": 9682 + }, + { + "epoch": 1.0240059221658206, + "grad_norm": 8.5625, + "learning_rate": 4.609991702531306e-06, + "loss": 2.4937, + "step": 9683 + }, + { + "epoch": 1.0241116751269035, + "grad_norm": 7.875, + "learning_rate": 4.609899844580221e-06, + "loss": 2.4807, + "step": 9684 + }, + { + "epoch": 1.0242174280879865, + "grad_norm": 7.53125, + "learning_rate": 4.609807976728239e-06, + "loss": 2.55, + "step": 9685 + }, + { + "epoch": 1.0243231810490694, + "grad_norm": 9.125, + "learning_rate": 4.60971609897579e-06, + "loss": 2.5634, + "step": 9686 + }, + { + "epoch": 1.0244289340101522, + "grad_norm": 8.3125, + "learning_rate": 4.609624211323307e-06, + "loss": 2.6108, + "step": 9687 + }, + { + "epoch": 1.0245346869712353, + "grad_norm": 8.1875, + "learning_rate": 4.609532313771221e-06, + "loss": 2.5504, + "step": 9688 + }, + { + "epoch": 1.0246404399323181, + "grad_norm": 7.625, + "learning_rate": 4.609440406319961e-06, + "loss": 2.4505, + "step": 9689 + }, + { + "epoch": 1.024746192893401, + "grad_norm": 8.25, + "learning_rate": 4.609348488969962e-06, + "loss": 2.7775, + "step": 9690 + }, + { + "epoch": 1.024851945854484, + "grad_norm": 7.75, + "learning_rate": 4.609256561721652e-06, + "loss": 2.1795, + "step": 9691 + }, + { + "epoch": 1.0249576988155669, + "grad_norm": 8.125, + "learning_rate": 4.609164624575464e-06, + "loss": 2.2142, + "step": 9692 + }, + { + "epoch": 1.0250634517766497, + "grad_norm": 8.125, + "learning_rate": 4.609072677531829e-06, + "loss": 2.5788, + "step": 9693 + }, + { + "epoch": 1.0251692047377328, + "grad_norm": 7.8125, + "learning_rate": 4.608980720591177e-06, + "loss": 2.5318, + "step": 9694 + }, + { + "epoch": 1.0252749576988156, + "grad_norm": 8.5625, + "learning_rate": 4.6088887537539425e-06, + "loss": 2.7418, + "step": 9695 + }, + { + "epoch": 1.0253807106598984, + "grad_norm": 8.1875, + "learning_rate": 4.608796777020556e-06, + "loss": 2.6794, + "step": 9696 + }, + { + "epoch": 1.0254864636209813, + "grad_norm": 8.25, + "learning_rate": 4.608704790391448e-06, + "loss": 2.3922, + "step": 9697 + }, + { + "epoch": 1.0255922165820643, + "grad_norm": 8.125, + "learning_rate": 4.608612793867051e-06, + "loss": 2.4467, + "step": 9698 + }, + { + "epoch": 1.0256979695431472, + "grad_norm": 7.875, + "learning_rate": 4.608520787447797e-06, + "loss": 2.6276, + "step": 9699 + }, + { + "epoch": 1.02580372250423, + "grad_norm": 7.9375, + "learning_rate": 4.608428771134117e-06, + "loss": 2.3952, + "step": 9700 + }, + { + "epoch": 1.025909475465313, + "grad_norm": 7.90625, + "learning_rate": 4.608336744926442e-06, + "loss": 2.5158, + "step": 9701 + }, + { + "epoch": 1.026015228426396, + "grad_norm": 8.125, + "learning_rate": 4.608244708825206e-06, + "loss": 2.454, + "step": 9702 + }, + { + "epoch": 1.0261209813874788, + "grad_norm": 7.75, + "learning_rate": 4.608152662830839e-06, + "loss": 2.5931, + "step": 9703 + }, + { + "epoch": 1.0262267343485618, + "grad_norm": 7.65625, + "learning_rate": 4.608060606943775e-06, + "loss": 2.5801, + "step": 9704 + }, + { + "epoch": 1.0263324873096447, + "grad_norm": 7.9375, + "learning_rate": 4.6079685411644446e-06, + "loss": 2.6611, + "step": 9705 + }, + { + "epoch": 1.0264382402707275, + "grad_norm": 8.1875, + "learning_rate": 4.60787646549328e-06, + "loss": 2.632, + "step": 9706 + }, + { + "epoch": 1.0265439932318106, + "grad_norm": 8.5, + "learning_rate": 4.607784379930712e-06, + "loss": 2.5952, + "step": 9707 + }, + { + "epoch": 1.0266497461928934, + "grad_norm": 8.3125, + "learning_rate": 4.607692284477175e-06, + "loss": 2.494, + "step": 9708 + }, + { + "epoch": 1.0267554991539762, + "grad_norm": 8.625, + "learning_rate": 4.607600179133099e-06, + "loss": 2.5114, + "step": 9709 + }, + { + "epoch": 1.0268612521150593, + "grad_norm": 8.0625, + "learning_rate": 4.607508063898919e-06, + "loss": 2.5711, + "step": 9710 + }, + { + "epoch": 1.0269670050761421, + "grad_norm": 8.8125, + "learning_rate": 4.607415938775064e-06, + "loss": 2.6548, + "step": 9711 + }, + { + "epoch": 1.027072758037225, + "grad_norm": 8.4375, + "learning_rate": 4.60732380376197e-06, + "loss": 2.3499, + "step": 9712 + }, + { + "epoch": 1.027178510998308, + "grad_norm": 8.8125, + "learning_rate": 4.607231658860065e-06, + "loss": 2.277, + "step": 9713 + }, + { + "epoch": 1.0272842639593909, + "grad_norm": 8.5, + "learning_rate": 4.607139504069785e-06, + "loss": 2.8367, + "step": 9714 + }, + { + "epoch": 1.0273900169204737, + "grad_norm": 8.375, + "learning_rate": 4.6070473393915604e-06, + "loss": 2.6465, + "step": 9715 + }, + { + "epoch": 1.0274957698815568, + "grad_norm": 8.0625, + "learning_rate": 4.606955164825825e-06, + "loss": 2.4991, + "step": 9716 + }, + { + "epoch": 1.0276015228426396, + "grad_norm": 7.875, + "learning_rate": 4.60686298037301e-06, + "loss": 2.5787, + "step": 9717 + }, + { + "epoch": 1.0277072758037225, + "grad_norm": 8.625, + "learning_rate": 4.606770786033549e-06, + "loss": 2.3922, + "step": 9718 + }, + { + "epoch": 1.0278130287648055, + "grad_norm": 8.0, + "learning_rate": 4.606678581807874e-06, + "loss": 2.4759, + "step": 9719 + }, + { + "epoch": 1.0279187817258884, + "grad_norm": 8.0, + "learning_rate": 4.606586367696419e-06, + "loss": 2.5814, + "step": 9720 + }, + { + "epoch": 1.0280245346869712, + "grad_norm": 8.6875, + "learning_rate": 4.606494143699616e-06, + "loss": 2.7705, + "step": 9721 + }, + { + "epoch": 1.0281302876480543, + "grad_norm": 8.25, + "learning_rate": 4.6064019098178955e-06, + "loss": 2.6376, + "step": 9722 + }, + { + "epoch": 1.028236040609137, + "grad_norm": 8.9375, + "learning_rate": 4.606309666051694e-06, + "loss": 2.8502, + "step": 9723 + }, + { + "epoch": 1.02834179357022, + "grad_norm": 8.0625, + "learning_rate": 4.606217412401443e-06, + "loss": 2.5471, + "step": 9724 + }, + { + "epoch": 1.0284475465313028, + "grad_norm": 7.90625, + "learning_rate": 4.6061251488675735e-06, + "loss": 2.4344, + "step": 9725 + }, + { + "epoch": 1.0285532994923858, + "grad_norm": 8.3125, + "learning_rate": 4.606032875450521e-06, + "loss": 2.4647, + "step": 9726 + }, + { + "epoch": 1.0286590524534687, + "grad_norm": 7.875, + "learning_rate": 4.605940592150717e-06, + "loss": 2.6408, + "step": 9727 + }, + { + "epoch": 1.0287648054145515, + "grad_norm": 8.6875, + "learning_rate": 4.605848298968596e-06, + "loss": 2.8216, + "step": 9728 + }, + { + "epoch": 1.0288705583756346, + "grad_norm": 7.75, + "learning_rate": 4.605755995904589e-06, + "loss": 2.4727, + "step": 9729 + }, + { + "epoch": 1.0289763113367174, + "grad_norm": 8.0, + "learning_rate": 4.605663682959132e-06, + "loss": 2.3605, + "step": 9730 + }, + { + "epoch": 1.0290820642978002, + "grad_norm": 8.5625, + "learning_rate": 4.605571360132655e-06, + "loss": 2.7988, + "step": 9731 + }, + { + "epoch": 1.0291878172588833, + "grad_norm": 8.8125, + "learning_rate": 4.605479027425593e-06, + "loss": 2.7196, + "step": 9732 + }, + { + "epoch": 1.0292935702199661, + "grad_norm": 7.8125, + "learning_rate": 4.60538668483838e-06, + "loss": 2.5843, + "step": 9733 + }, + { + "epoch": 1.029399323181049, + "grad_norm": 7.71875, + "learning_rate": 4.605294332371448e-06, + "loss": 2.5335, + "step": 9734 + }, + { + "epoch": 1.029505076142132, + "grad_norm": 7.96875, + "learning_rate": 4.60520197002523e-06, + "loss": 2.5807, + "step": 9735 + }, + { + "epoch": 1.0296108291032149, + "grad_norm": 8.25, + "learning_rate": 4.605109597800161e-06, + "loss": 2.7332, + "step": 9736 + }, + { + "epoch": 1.0297165820642977, + "grad_norm": 7.84375, + "learning_rate": 4.605017215696673e-06, + "loss": 2.5472, + "step": 9737 + }, + { + "epoch": 1.0298223350253808, + "grad_norm": 7.625, + "learning_rate": 4.6049248237152e-06, + "loss": 2.5533, + "step": 9738 + }, + { + "epoch": 1.0299280879864636, + "grad_norm": 7.90625, + "learning_rate": 4.604832421856177e-06, + "loss": 2.6088, + "step": 9739 + }, + { + "epoch": 1.0300338409475465, + "grad_norm": 8.5625, + "learning_rate": 4.604740010120035e-06, + "loss": 2.5022, + "step": 9740 + }, + { + "epoch": 1.0301395939086295, + "grad_norm": 8.25, + "learning_rate": 4.6046475885072094e-06, + "loss": 2.4959, + "step": 9741 + }, + { + "epoch": 1.0302453468697124, + "grad_norm": 7.78125, + "learning_rate": 4.604555157018134e-06, + "loss": 2.4837, + "step": 9742 + }, + { + "epoch": 1.0303510998307952, + "grad_norm": 8.1875, + "learning_rate": 4.60446271565324e-06, + "loss": 2.6671, + "step": 9743 + }, + { + "epoch": 1.0304568527918783, + "grad_norm": 7.84375, + "learning_rate": 4.604370264412965e-06, + "loss": 2.591, + "step": 9744 + }, + { + "epoch": 1.030562605752961, + "grad_norm": 8.5, + "learning_rate": 4.604277803297741e-06, + "loss": 2.6175, + "step": 9745 + }, + { + "epoch": 1.030668358714044, + "grad_norm": 7.65625, + "learning_rate": 4.604185332308001e-06, + "loss": 2.368, + "step": 9746 + }, + { + "epoch": 1.030774111675127, + "grad_norm": 8.5, + "learning_rate": 4.60409285144418e-06, + "loss": 2.607, + "step": 9747 + }, + { + "epoch": 1.0308798646362098, + "grad_norm": 8.5625, + "learning_rate": 4.604000360706712e-06, + "loss": 2.6248, + "step": 9748 + }, + { + "epoch": 1.0309856175972927, + "grad_norm": 11.125, + "learning_rate": 4.603907860096031e-06, + "loss": 2.6406, + "step": 9749 + }, + { + "epoch": 1.0310913705583757, + "grad_norm": 8.4375, + "learning_rate": 4.60381534961257e-06, + "loss": 2.6505, + "step": 9750 + }, + { + "epoch": 1.0311971235194586, + "grad_norm": 8.1875, + "learning_rate": 4.603722829256765e-06, + "loss": 2.4474, + "step": 9751 + }, + { + "epoch": 1.0313028764805414, + "grad_norm": 7.8125, + "learning_rate": 4.603630299029049e-06, + "loss": 2.5263, + "step": 9752 + }, + { + "epoch": 1.0314086294416243, + "grad_norm": 8.0625, + "learning_rate": 4.603537758929857e-06, + "loss": 2.6927, + "step": 9753 + }, + { + "epoch": 1.0315143824027073, + "grad_norm": 8.125, + "learning_rate": 4.603445208959621e-06, + "loss": 2.4161, + "step": 9754 + }, + { + "epoch": 1.0316201353637902, + "grad_norm": 8.0, + "learning_rate": 4.603352649118778e-06, + "loss": 2.641, + "step": 9755 + }, + { + "epoch": 1.031725888324873, + "grad_norm": 7.9375, + "learning_rate": 4.603260079407761e-06, + "loss": 2.4996, + "step": 9756 + }, + { + "epoch": 1.031831641285956, + "grad_norm": 7.84375, + "learning_rate": 4.603167499827005e-06, + "loss": 2.6294, + "step": 9757 + }, + { + "epoch": 1.031937394247039, + "grad_norm": 8.0, + "learning_rate": 4.603074910376943e-06, + "loss": 2.5993, + "step": 9758 + }, + { + "epoch": 1.0320431472081217, + "grad_norm": 7.71875, + "learning_rate": 4.602982311058012e-06, + "loss": 2.5341, + "step": 9759 + }, + { + "epoch": 1.0321489001692048, + "grad_norm": 8.3125, + "learning_rate": 4.602889701870644e-06, + "loss": 2.6219, + "step": 9760 + }, + { + "epoch": 1.0322546531302876, + "grad_norm": 8.8125, + "learning_rate": 4.6027970828152756e-06, + "loss": 2.6182, + "step": 9761 + }, + { + "epoch": 1.0323604060913705, + "grad_norm": 8.6875, + "learning_rate": 4.602704453892339e-06, + "loss": 2.6174, + "step": 9762 + }, + { + "epoch": 1.0324661590524535, + "grad_norm": 9.125, + "learning_rate": 4.6026118151022716e-06, + "loss": 2.72, + "step": 9763 + }, + { + "epoch": 1.0325719120135364, + "grad_norm": 8.5, + "learning_rate": 4.602519166445507e-06, + "loss": 2.4359, + "step": 9764 + }, + { + "epoch": 1.0326776649746192, + "grad_norm": 8.3125, + "learning_rate": 4.60242650792248e-06, + "loss": 2.5113, + "step": 9765 + }, + { + "epoch": 1.0327834179357023, + "grad_norm": 7.5, + "learning_rate": 4.602333839533624e-06, + "loss": 2.5519, + "step": 9766 + }, + { + "epoch": 1.0328891708967851, + "grad_norm": 8.5625, + "learning_rate": 4.602241161279376e-06, + "loss": 2.4041, + "step": 9767 + }, + { + "epoch": 1.032994923857868, + "grad_norm": 8.0, + "learning_rate": 4.60214847316017e-06, + "loss": 2.5459, + "step": 9768 + }, + { + "epoch": 1.033100676818951, + "grad_norm": 8.3125, + "learning_rate": 4.602055775176441e-06, + "loss": 2.4857, + "step": 9769 + }, + { + "epoch": 1.0332064297800339, + "grad_norm": 7.84375, + "learning_rate": 4.6019630673286235e-06, + "loss": 2.5488, + "step": 9770 + }, + { + "epoch": 1.0333121827411167, + "grad_norm": 8.375, + "learning_rate": 4.601870349617153e-06, + "loss": 2.5252, + "step": 9771 + }, + { + "epoch": 1.0334179357021998, + "grad_norm": 8.5, + "learning_rate": 4.601777622042465e-06, + "loss": 2.5775, + "step": 9772 + }, + { + "epoch": 1.0335236886632826, + "grad_norm": 7.8125, + "learning_rate": 4.601684884604994e-06, + "loss": 2.5474, + "step": 9773 + }, + { + "epoch": 1.0336294416243654, + "grad_norm": 8.1875, + "learning_rate": 4.6015921373051765e-06, + "loss": 2.6269, + "step": 9774 + }, + { + "epoch": 1.0337351945854485, + "grad_norm": 8.1875, + "learning_rate": 4.601499380143446e-06, + "loss": 2.6486, + "step": 9775 + }, + { + "epoch": 1.0338409475465313, + "grad_norm": 8.125, + "learning_rate": 4.601406613120237e-06, + "loss": 2.5587, + "step": 9776 + }, + { + "epoch": 1.0339467005076142, + "grad_norm": 8.625, + "learning_rate": 4.601313836235988e-06, + "loss": 2.4617, + "step": 9777 + }, + { + "epoch": 1.0340524534686972, + "grad_norm": 8.1875, + "learning_rate": 4.601221049491131e-06, + "loss": 2.4961, + "step": 9778 + }, + { + "epoch": 1.03415820642978, + "grad_norm": 8.375, + "learning_rate": 4.601128252886104e-06, + "loss": 2.7204, + "step": 9779 + }, + { + "epoch": 1.034263959390863, + "grad_norm": 8.625, + "learning_rate": 4.601035446421341e-06, + "loss": 2.5109, + "step": 9780 + }, + { + "epoch": 1.0343697123519457, + "grad_norm": 8.5, + "learning_rate": 4.6009426300972785e-06, + "loss": 2.5358, + "step": 9781 + }, + { + "epoch": 1.0344754653130288, + "grad_norm": 8.3125, + "learning_rate": 4.600849803914351e-06, + "loss": 2.5648, + "step": 9782 + }, + { + "epoch": 1.0345812182741116, + "grad_norm": 7.6875, + "learning_rate": 4.600756967872995e-06, + "loss": 2.4992, + "step": 9783 + }, + { + "epoch": 1.0346869712351945, + "grad_norm": 7.78125, + "learning_rate": 4.600664121973646e-06, + "loss": 2.6683, + "step": 9784 + }, + { + "epoch": 1.0347927241962775, + "grad_norm": 8.5, + "learning_rate": 4.600571266216739e-06, + "loss": 2.5748, + "step": 9785 + }, + { + "epoch": 1.0348984771573604, + "grad_norm": 8.1875, + "learning_rate": 4.6004784006027095e-06, + "loss": 2.2963, + "step": 9786 + }, + { + "epoch": 1.0350042301184432, + "grad_norm": 7.875, + "learning_rate": 4.600385525131995e-06, + "loss": 2.4545, + "step": 9787 + }, + { + "epoch": 1.0351099830795263, + "grad_norm": 8.5, + "learning_rate": 4.600292639805029e-06, + "loss": 2.6244, + "step": 9788 + }, + { + "epoch": 1.0352157360406091, + "grad_norm": 8.8125, + "learning_rate": 4.60019974462225e-06, + "loss": 2.732, + "step": 9789 + }, + { + "epoch": 1.035321489001692, + "grad_norm": 7.625, + "learning_rate": 4.600106839584092e-06, + "loss": 2.4848, + "step": 9790 + }, + { + "epoch": 1.035427241962775, + "grad_norm": 8.5625, + "learning_rate": 4.600013924690992e-06, + "loss": 2.6892, + "step": 9791 + }, + { + "epoch": 1.0355329949238579, + "grad_norm": 8.1875, + "learning_rate": 4.599920999943385e-06, + "loss": 2.5992, + "step": 9792 + }, + { + "epoch": 1.0356387478849407, + "grad_norm": 7.78125, + "learning_rate": 4.5998280653417075e-06, + "loss": 2.574, + "step": 9793 + }, + { + "epoch": 1.0357445008460238, + "grad_norm": 7.65625, + "learning_rate": 4.599735120886395e-06, + "loss": 2.5664, + "step": 9794 + }, + { + "epoch": 1.0358502538071066, + "grad_norm": 7.9375, + "learning_rate": 4.5996421665778855e-06, + "loss": 2.546, + "step": 9795 + }, + { + "epoch": 1.0359560067681894, + "grad_norm": 8.0625, + "learning_rate": 4.599549202416613e-06, + "loss": 2.7807, + "step": 9796 + }, + { + "epoch": 1.0360617597292725, + "grad_norm": 8.125, + "learning_rate": 4.599456228403016e-06, + "loss": 2.5141, + "step": 9797 + }, + { + "epoch": 1.0361675126903553, + "grad_norm": 8.1875, + "learning_rate": 4.599363244537529e-06, + "loss": 2.3592, + "step": 9798 + }, + { + "epoch": 1.0362732656514382, + "grad_norm": 8.25, + "learning_rate": 4.599270250820589e-06, + "loss": 2.5261, + "step": 9799 + }, + { + "epoch": 1.0363790186125212, + "grad_norm": 8.9375, + "learning_rate": 4.599177247252632e-06, + "loss": 2.7525, + "step": 9800 + }, + { + "epoch": 1.036484771573604, + "grad_norm": 8.0, + "learning_rate": 4.599084233834095e-06, + "loss": 2.4626, + "step": 9801 + }, + { + "epoch": 1.036590524534687, + "grad_norm": 7.71875, + "learning_rate": 4.598991210565414e-06, + "loss": 2.3487, + "step": 9802 + }, + { + "epoch": 1.03669627749577, + "grad_norm": 8.3125, + "learning_rate": 4.598898177447026e-06, + "loss": 2.6935, + "step": 9803 + }, + { + "epoch": 1.0368020304568528, + "grad_norm": 7.9375, + "learning_rate": 4.598805134479367e-06, + "loss": 2.6879, + "step": 9804 + }, + { + "epoch": 1.0369077834179357, + "grad_norm": 8.3125, + "learning_rate": 4.598712081662874e-06, + "loss": 2.3259, + "step": 9805 + }, + { + "epoch": 1.0370135363790187, + "grad_norm": 8.625, + "learning_rate": 4.598619018997983e-06, + "loss": 2.6286, + "step": 9806 + }, + { + "epoch": 1.0371192893401016, + "grad_norm": 7.84375, + "learning_rate": 4.598525946485131e-06, + "loss": 2.5565, + "step": 9807 + }, + { + "epoch": 1.0372250423011844, + "grad_norm": 8.5, + "learning_rate": 4.598432864124757e-06, + "loss": 2.7609, + "step": 9808 + }, + { + "epoch": 1.0373307952622675, + "grad_norm": 8.0625, + "learning_rate": 4.598339771917294e-06, + "loss": 2.5986, + "step": 9809 + }, + { + "epoch": 1.0374365482233503, + "grad_norm": 7.875, + "learning_rate": 4.598246669863182e-06, + "loss": 2.5915, + "step": 9810 + }, + { + "epoch": 1.0375423011844331, + "grad_norm": 8.1875, + "learning_rate": 4.598153557962856e-06, + "loss": 2.6957, + "step": 9811 + }, + { + "epoch": 1.037648054145516, + "grad_norm": 7.96875, + "learning_rate": 4.5980604362167525e-06, + "loss": 2.5696, + "step": 9812 + }, + { + "epoch": 1.037753807106599, + "grad_norm": 8.25, + "learning_rate": 4.59796730462531e-06, + "loss": 2.635, + "step": 9813 + }, + { + "epoch": 1.0378595600676819, + "grad_norm": 8.375, + "learning_rate": 4.5978741631889655e-06, + "loss": 2.6968, + "step": 9814 + }, + { + "epoch": 1.0379653130287647, + "grad_norm": 7.96875, + "learning_rate": 4.597781011908155e-06, + "loss": 2.644, + "step": 9815 + }, + { + "epoch": 1.0380710659898478, + "grad_norm": 7.78125, + "learning_rate": 4.597687850783316e-06, + "loss": 2.3908, + "step": 9816 + }, + { + "epoch": 1.0381768189509306, + "grad_norm": 8.1875, + "learning_rate": 4.597594679814886e-06, + "loss": 2.6115, + "step": 9817 + }, + { + "epoch": 1.0382825719120135, + "grad_norm": 8.0625, + "learning_rate": 4.597501499003303e-06, + "loss": 2.6391, + "step": 9818 + }, + { + "epoch": 1.0383883248730965, + "grad_norm": 7.90625, + "learning_rate": 4.5974083083490024e-06, + "loss": 2.472, + "step": 9819 + }, + { + "epoch": 1.0384940778341794, + "grad_norm": 8.375, + "learning_rate": 4.5973151078524225e-06, + "loss": 1.8801, + "step": 9820 + }, + { + "epoch": 1.0385998307952622, + "grad_norm": 8.0625, + "learning_rate": 4.597221897514001e-06, + "loss": 2.6303, + "step": 9821 + }, + { + "epoch": 1.0387055837563453, + "grad_norm": 8.375, + "learning_rate": 4.597128677334174e-06, + "loss": 2.5693, + "step": 9822 + }, + { + "epoch": 1.038811336717428, + "grad_norm": 8.3125, + "learning_rate": 4.597035447313381e-06, + "loss": 2.6012, + "step": 9823 + }, + { + "epoch": 1.038917089678511, + "grad_norm": 8.125, + "learning_rate": 4.596942207452057e-06, + "loss": 2.4642, + "step": 9824 + }, + { + "epoch": 1.039022842639594, + "grad_norm": 7.9375, + "learning_rate": 4.596848957750642e-06, + "loss": 2.3655, + "step": 9825 + }, + { + "epoch": 1.0391285956006768, + "grad_norm": 8.125, + "learning_rate": 4.596755698209572e-06, + "loss": 2.3873, + "step": 9826 + }, + { + "epoch": 1.0392343485617597, + "grad_norm": 7.75, + "learning_rate": 4.596662428829285e-06, + "loss": 2.4233, + "step": 9827 + }, + { + "epoch": 1.0393401015228427, + "grad_norm": 7.71875, + "learning_rate": 4.596569149610219e-06, + "loss": 2.4527, + "step": 9828 + }, + { + "epoch": 1.0394458544839256, + "grad_norm": 8.5625, + "learning_rate": 4.59647586055281e-06, + "loss": 2.6241, + "step": 9829 + }, + { + "epoch": 1.0395516074450084, + "grad_norm": 8.625, + "learning_rate": 4.596382561657499e-06, + "loss": 2.5391, + "step": 9830 + }, + { + "epoch": 1.0396573604060915, + "grad_norm": 7.84375, + "learning_rate": 4.596289252924721e-06, + "loss": 2.4883, + "step": 9831 + }, + { + "epoch": 1.0397631133671743, + "grad_norm": 8.25, + "learning_rate": 4.596195934354915e-06, + "loss": 2.4644, + "step": 9832 + }, + { + "epoch": 1.0398688663282571, + "grad_norm": 8.0, + "learning_rate": 4.596102605948519e-06, + "loss": 2.5538, + "step": 9833 + }, + { + "epoch": 1.0399746192893402, + "grad_norm": 8.0, + "learning_rate": 4.596009267705971e-06, + "loss": 2.4465, + "step": 9834 + }, + { + "epoch": 1.040080372250423, + "grad_norm": 8.4375, + "learning_rate": 4.595915919627708e-06, + "loss": 2.6564, + "step": 9835 + }, + { + "epoch": 1.0401861252115059, + "grad_norm": 8.0625, + "learning_rate": 4.595822561714171e-06, + "loss": 2.673, + "step": 9836 + }, + { + "epoch": 1.0402918781725887, + "grad_norm": 8.625, + "learning_rate": 4.595729193965793e-06, + "loss": 2.5676, + "step": 9837 + }, + { + "epoch": 1.0403976311336718, + "grad_norm": 8.125, + "learning_rate": 4.5956358163830164e-06, + "loss": 2.7098, + "step": 9838 + }, + { + "epoch": 1.0405033840947546, + "grad_norm": 8.125, + "learning_rate": 4.595542428966278e-06, + "loss": 2.5629, + "step": 9839 + }, + { + "epoch": 1.0406091370558375, + "grad_norm": 8.3125, + "learning_rate": 4.595449031716016e-06, + "loss": 2.7327, + "step": 9840 + }, + { + "epoch": 1.0407148900169205, + "grad_norm": 8.0, + "learning_rate": 4.5953556246326684e-06, + "loss": 2.5679, + "step": 9841 + }, + { + "epoch": 1.0408206429780034, + "grad_norm": 8.125, + "learning_rate": 4.595262207716674e-06, + "loss": 2.7137, + "step": 9842 + }, + { + "epoch": 1.0409263959390862, + "grad_norm": 8.4375, + "learning_rate": 4.595168780968471e-06, + "loss": 2.5347, + "step": 9843 + }, + { + "epoch": 1.0410321489001693, + "grad_norm": 8.375, + "learning_rate": 4.595075344388497e-06, + "loss": 2.6896, + "step": 9844 + }, + { + "epoch": 1.041137901861252, + "grad_norm": 7.96875, + "learning_rate": 4.594981897977193e-06, + "loss": 2.5679, + "step": 9845 + }, + { + "epoch": 1.041243654822335, + "grad_norm": 8.5, + "learning_rate": 4.594888441734995e-06, + "loss": 2.5301, + "step": 9846 + }, + { + "epoch": 1.041349407783418, + "grad_norm": 7.71875, + "learning_rate": 4.594794975662341e-06, + "loss": 2.6247, + "step": 9847 + }, + { + "epoch": 1.0414551607445008, + "grad_norm": 8.0625, + "learning_rate": 4.594701499759673e-06, + "loss": 2.7065, + "step": 9848 + }, + { + "epoch": 1.0415609137055837, + "grad_norm": 8.5, + "learning_rate": 4.594608014027427e-06, + "loss": 2.6485, + "step": 9849 + }, + { + "epoch": 1.0416666666666667, + "grad_norm": 8.125, + "learning_rate": 4.594514518466041e-06, + "loss": 2.4967, + "step": 9850 + }, + { + "epoch": 1.0417724196277496, + "grad_norm": 8.625, + "learning_rate": 4.594421013075956e-06, + "loss": 2.5597, + "step": 9851 + }, + { + "epoch": 1.0418781725888324, + "grad_norm": 8.0625, + "learning_rate": 4.59432749785761e-06, + "loss": 2.2875, + "step": 9852 + }, + { + "epoch": 1.0419839255499155, + "grad_norm": 8.5, + "learning_rate": 4.5942339728114415e-06, + "loss": 2.5812, + "step": 9853 + }, + { + "epoch": 1.0420896785109983, + "grad_norm": 8.1875, + "learning_rate": 4.594140437937889e-06, + "loss": 2.6207, + "step": 9854 + }, + { + "epoch": 1.0421954314720812, + "grad_norm": 8.0625, + "learning_rate": 4.594046893237392e-06, + "loss": 2.583, + "step": 9855 + }, + { + "epoch": 1.0423011844331642, + "grad_norm": 8.3125, + "learning_rate": 4.59395333871039e-06, + "loss": 2.4798, + "step": 9856 + }, + { + "epoch": 1.042406937394247, + "grad_norm": 8.0, + "learning_rate": 4.59385977435732e-06, + "loss": 2.5757, + "step": 9857 + }, + { + "epoch": 1.04251269035533, + "grad_norm": 8.0, + "learning_rate": 4.593766200178623e-06, + "loss": 2.5515, + "step": 9858 + }, + { + "epoch": 1.042618443316413, + "grad_norm": 8.125, + "learning_rate": 4.593672616174738e-06, + "loss": 2.66, + "step": 9859 + }, + { + "epoch": 1.0427241962774958, + "grad_norm": 9.125, + "learning_rate": 4.593579022346104e-06, + "loss": 2.5511, + "step": 9860 + }, + { + "epoch": 1.0428299492385786, + "grad_norm": 8.0625, + "learning_rate": 4.59348541869316e-06, + "loss": 2.6089, + "step": 9861 + }, + { + "epoch": 1.0429357021996617, + "grad_norm": 8.0625, + "learning_rate": 4.593391805216344e-06, + "loss": 2.7456, + "step": 9862 + }, + { + "epoch": 1.0430414551607445, + "grad_norm": 8.4375, + "learning_rate": 4.593298181916097e-06, + "loss": 2.3896, + "step": 9863 + }, + { + "epoch": 1.0431472081218274, + "grad_norm": 8.1875, + "learning_rate": 4.593204548792858e-06, + "loss": 2.571, + "step": 9864 + }, + { + "epoch": 1.0432529610829104, + "grad_norm": 8.25, + "learning_rate": 4.593110905847066e-06, + "loss": 2.5523, + "step": 9865 + }, + { + "epoch": 1.0433587140439933, + "grad_norm": 7.78125, + "learning_rate": 4.593017253079161e-06, + "loss": 2.5883, + "step": 9866 + }, + { + "epoch": 1.0434644670050761, + "grad_norm": 7.875, + "learning_rate": 4.592923590489581e-06, + "loss": 2.5655, + "step": 9867 + }, + { + "epoch": 1.043570219966159, + "grad_norm": 7.59375, + "learning_rate": 4.592829918078767e-06, + "loss": 2.4626, + "step": 9868 + }, + { + "epoch": 1.043675972927242, + "grad_norm": 8.3125, + "learning_rate": 4.592736235847159e-06, + "loss": 2.3564, + "step": 9869 + }, + { + "epoch": 1.0437817258883249, + "grad_norm": 7.875, + "learning_rate": 4.592642543795194e-06, + "loss": 2.5684, + "step": 9870 + }, + { + "epoch": 1.0438874788494077, + "grad_norm": 8.125, + "learning_rate": 4.592548841923314e-06, + "loss": 2.5447, + "step": 9871 + }, + { + "epoch": 1.0439932318104908, + "grad_norm": 8.0, + "learning_rate": 4.592455130231959e-06, + "loss": 2.3882, + "step": 9872 + }, + { + "epoch": 1.0440989847715736, + "grad_norm": 8.3125, + "learning_rate": 4.592361408721567e-06, + "loss": 2.6984, + "step": 9873 + }, + { + "epoch": 1.0442047377326564, + "grad_norm": 9.25, + "learning_rate": 4.59226767739258e-06, + "loss": 2.7891, + "step": 9874 + }, + { + "epoch": 1.0443104906937395, + "grad_norm": 7.90625, + "learning_rate": 4.592173936245436e-06, + "loss": 2.6386, + "step": 9875 + }, + { + "epoch": 1.0444162436548223, + "grad_norm": 7.71875, + "learning_rate": 4.592080185280574e-06, + "loss": 2.3455, + "step": 9876 + }, + { + "epoch": 1.0445219966159052, + "grad_norm": 7.71875, + "learning_rate": 4.591986424498436e-06, + "loss": 2.4208, + "step": 9877 + }, + { + "epoch": 1.0446277495769882, + "grad_norm": 8.125, + "learning_rate": 4.591892653899461e-06, + "loss": 2.5658, + "step": 9878 + }, + { + "epoch": 1.044733502538071, + "grad_norm": 8.1875, + "learning_rate": 4.591798873484091e-06, + "loss": 2.7711, + "step": 9879 + }, + { + "epoch": 1.044839255499154, + "grad_norm": 9.1875, + "learning_rate": 4.5917050832527635e-06, + "loss": 2.4249, + "step": 9880 + }, + { + "epoch": 1.044945008460237, + "grad_norm": 8.625, + "learning_rate": 4.591611283205919e-06, + "loss": 2.1845, + "step": 9881 + }, + { + "epoch": 1.0450507614213198, + "grad_norm": 7.84375, + "learning_rate": 4.591517473343998e-06, + "loss": 2.593, + "step": 9882 + }, + { + "epoch": 1.0451565143824026, + "grad_norm": 8.25, + "learning_rate": 4.5914236536674416e-06, + "loss": 2.5456, + "step": 9883 + }, + { + "epoch": 1.0452622673434857, + "grad_norm": 7.96875, + "learning_rate": 4.591329824176689e-06, + "loss": 2.5159, + "step": 9884 + }, + { + "epoch": 1.0453680203045685, + "grad_norm": 8.25, + "learning_rate": 4.591235984872181e-06, + "loss": 2.6009, + "step": 9885 + }, + { + "epoch": 1.0454737732656514, + "grad_norm": 8.625, + "learning_rate": 4.591142135754357e-06, + "loss": 2.6605, + "step": 9886 + }, + { + "epoch": 1.0455795262267344, + "grad_norm": 7.71875, + "learning_rate": 4.59104827682366e-06, + "loss": 2.5642, + "step": 9887 + }, + { + "epoch": 1.0456852791878173, + "grad_norm": 7.84375, + "learning_rate": 4.590954408080528e-06, + "loss": 2.4849, + "step": 9888 + }, + { + "epoch": 1.0457910321489001, + "grad_norm": 8.5625, + "learning_rate": 4.590860529525401e-06, + "loss": 2.5721, + "step": 9889 + }, + { + "epoch": 1.0458967851099832, + "grad_norm": 8.625, + "learning_rate": 4.590766641158722e-06, + "loss": 2.492, + "step": 9890 + }, + { + "epoch": 1.046002538071066, + "grad_norm": 7.8125, + "learning_rate": 4.59067274298093e-06, + "loss": 2.5945, + "step": 9891 + }, + { + "epoch": 1.0461082910321489, + "grad_norm": 8.625, + "learning_rate": 4.590578834992465e-06, + "loss": 2.7946, + "step": 9892 + }, + { + "epoch": 1.0462140439932317, + "grad_norm": 8.375, + "learning_rate": 4.5904849171937695e-06, + "loss": 2.4706, + "step": 9893 + }, + { + "epoch": 1.0463197969543148, + "grad_norm": 8.4375, + "learning_rate": 4.590390989585282e-06, + "loss": 2.7369, + "step": 9894 + }, + { + "epoch": 1.0464255499153976, + "grad_norm": 8.375, + "learning_rate": 4.5902970521674455e-06, + "loss": 2.826, + "step": 9895 + }, + { + "epoch": 1.0465313028764804, + "grad_norm": 8.0625, + "learning_rate": 4.5902031049407e-06, + "loss": 2.6159, + "step": 9896 + }, + { + "epoch": 1.0466370558375635, + "grad_norm": 8.1875, + "learning_rate": 4.590109147905486e-06, + "loss": 2.6132, + "step": 9897 + }, + { + "epoch": 1.0467428087986463, + "grad_norm": 7.59375, + "learning_rate": 4.590015181062244e-06, + "loss": 2.5326, + "step": 9898 + }, + { + "epoch": 1.0468485617597292, + "grad_norm": 8.4375, + "learning_rate": 4.589921204411417e-06, + "loss": 2.5119, + "step": 9899 + }, + { + "epoch": 1.0469543147208122, + "grad_norm": 9.0625, + "learning_rate": 4.589827217953443e-06, + "loss": 2.6571, + "step": 9900 + }, + { + "epoch": 1.047060067681895, + "grad_norm": 8.6875, + "learning_rate": 4.589733221688765e-06, + "loss": 2.6717, + "step": 9901 + }, + { + "epoch": 1.047165820642978, + "grad_norm": 8.0, + "learning_rate": 4.589639215617824e-06, + "loss": 2.5046, + "step": 9902 + }, + { + "epoch": 1.047271573604061, + "grad_norm": 8.0625, + "learning_rate": 4.5895451997410615e-06, + "loss": 2.4497, + "step": 9903 + }, + { + "epoch": 1.0473773265651438, + "grad_norm": 8.375, + "learning_rate": 4.589451174058916e-06, + "loss": 2.5186, + "step": 9904 + }, + { + "epoch": 1.0474830795262267, + "grad_norm": 8.375, + "learning_rate": 4.589357138571832e-06, + "loss": 2.5834, + "step": 9905 + }, + { + "epoch": 1.0475888324873097, + "grad_norm": 8.1875, + "learning_rate": 4.5892630932802494e-06, + "loss": 2.3532, + "step": 9906 + }, + { + "epoch": 1.0476945854483926, + "grad_norm": 8.3125, + "learning_rate": 4.58916903818461e-06, + "loss": 2.5453, + "step": 9907 + }, + { + "epoch": 1.0478003384094754, + "grad_norm": 8.3125, + "learning_rate": 4.589074973285353e-06, + "loss": 2.5289, + "step": 9908 + }, + { + "epoch": 1.0479060913705585, + "grad_norm": 7.96875, + "learning_rate": 4.588980898582923e-06, + "loss": 2.4279, + "step": 9909 + }, + { + "epoch": 1.0480118443316413, + "grad_norm": 8.75, + "learning_rate": 4.58888681407776e-06, + "loss": 2.7898, + "step": 9910 + }, + { + "epoch": 1.0481175972927241, + "grad_norm": 8.0625, + "learning_rate": 4.588792719770305e-06, + "loss": 2.5384, + "step": 9911 + }, + { + "epoch": 1.0482233502538072, + "grad_norm": 8.125, + "learning_rate": 4.588698615661001e-06, + "loss": 2.6174, + "step": 9912 + }, + { + "epoch": 1.04832910321489, + "grad_norm": 7.84375, + "learning_rate": 4.588604501750288e-06, + "loss": 2.627, + "step": 9913 + }, + { + "epoch": 1.0484348561759729, + "grad_norm": 8.25, + "learning_rate": 4.588510378038607e-06, + "loss": 2.6939, + "step": 9914 + }, + { + "epoch": 1.048540609137056, + "grad_norm": 8.0625, + "learning_rate": 4.588416244526403e-06, + "loss": 2.6781, + "step": 9915 + }, + { + "epoch": 1.0486463620981388, + "grad_norm": 8.0625, + "learning_rate": 4.588322101214115e-06, + "loss": 2.5865, + "step": 9916 + }, + { + "epoch": 1.0487521150592216, + "grad_norm": 8.3125, + "learning_rate": 4.588227948102185e-06, + "loss": 2.6688, + "step": 9917 + }, + { + "epoch": 1.0488578680203047, + "grad_norm": 7.875, + "learning_rate": 4.588133785191055e-06, + "loss": 2.5023, + "step": 9918 + }, + { + "epoch": 1.0489636209813875, + "grad_norm": 9.125, + "learning_rate": 4.588039612481169e-06, + "loss": 2.6268, + "step": 9919 + }, + { + "epoch": 1.0490693739424704, + "grad_norm": 7.9375, + "learning_rate": 4.587945429972966e-06, + "loss": 2.487, + "step": 9920 + }, + { + "epoch": 1.0491751269035534, + "grad_norm": 8.0, + "learning_rate": 4.587851237666889e-06, + "loss": 2.5027, + "step": 9921 + }, + { + "epoch": 1.0492808798646363, + "grad_norm": 8.0625, + "learning_rate": 4.587757035563379e-06, + "loss": 2.3575, + "step": 9922 + }, + { + "epoch": 1.049386632825719, + "grad_norm": 8.0625, + "learning_rate": 4.587662823662881e-06, + "loss": 2.4138, + "step": 9923 + }, + { + "epoch": 1.049492385786802, + "grad_norm": 8.1875, + "learning_rate": 4.587568601965834e-06, + "loss": 2.5354, + "step": 9924 + }, + { + "epoch": 1.049598138747885, + "grad_norm": 7.9375, + "learning_rate": 4.587474370472682e-06, + "loss": 2.5309, + "step": 9925 + }, + { + "epoch": 1.0497038917089678, + "grad_norm": 8.0, + "learning_rate": 4.587380129183866e-06, + "loss": 2.5624, + "step": 9926 + }, + { + "epoch": 1.0498096446700507, + "grad_norm": 7.8125, + "learning_rate": 4.587285878099829e-06, + "loss": 2.3464, + "step": 9927 + }, + { + "epoch": 1.0499153976311337, + "grad_norm": 8.875, + "learning_rate": 4.587191617221013e-06, + "loss": 2.5519, + "step": 9928 + }, + { + "epoch": 1.0500211505922166, + "grad_norm": 8.125, + "learning_rate": 4.58709734654786e-06, + "loss": 2.6544, + "step": 9929 + }, + { + "epoch": 1.0501269035532994, + "grad_norm": 8.125, + "learning_rate": 4.587003066080813e-06, + "loss": 2.4817, + "step": 9930 + }, + { + "epoch": 1.0502326565143825, + "grad_norm": 8.0, + "learning_rate": 4.586908775820314e-06, + "loss": 2.4633, + "step": 9931 + }, + { + "epoch": 1.0503384094754653, + "grad_norm": 7.84375, + "learning_rate": 4.586814475766806e-06, + "loss": 2.6297, + "step": 9932 + }, + { + "epoch": 1.0504441624365481, + "grad_norm": 8.25, + "learning_rate": 4.5867201659207314e-06, + "loss": 2.596, + "step": 9933 + }, + { + "epoch": 1.0505499153976312, + "grad_norm": 8.125, + "learning_rate": 4.586625846282532e-06, + "loss": 2.5943, + "step": 9934 + }, + { + "epoch": 1.050655668358714, + "grad_norm": 7.71875, + "learning_rate": 4.586531516852652e-06, + "loss": 2.4301, + "step": 9935 + }, + { + "epoch": 1.0507614213197969, + "grad_norm": 8.5, + "learning_rate": 4.586437177631531e-06, + "loss": 2.6172, + "step": 9936 + }, + { + "epoch": 1.05086717428088, + "grad_norm": 8.5, + "learning_rate": 4.586342828619614e-06, + "loss": 2.4786, + "step": 9937 + }, + { + "epoch": 1.0509729272419628, + "grad_norm": 7.625, + "learning_rate": 4.586248469817345e-06, + "loss": 2.372, + "step": 9938 + }, + { + "epoch": 1.0510786802030456, + "grad_norm": 8.1875, + "learning_rate": 4.586154101225164e-06, + "loss": 2.6524, + "step": 9939 + }, + { + "epoch": 1.0511844331641287, + "grad_norm": 8.5, + "learning_rate": 4.586059722843515e-06, + "loss": 2.4635, + "step": 9940 + }, + { + "epoch": 1.0512901861252115, + "grad_norm": 8.1875, + "learning_rate": 4.585965334672842e-06, + "loss": 2.5525, + "step": 9941 + }, + { + "epoch": 1.0513959390862944, + "grad_norm": 8.5, + "learning_rate": 4.5858709367135855e-06, + "loss": 2.6685, + "step": 9942 + }, + { + "epoch": 1.0515016920473774, + "grad_norm": 8.3125, + "learning_rate": 4.58577652896619e-06, + "loss": 2.6149, + "step": 9943 + }, + { + "epoch": 1.0516074450084603, + "grad_norm": 8.0625, + "learning_rate": 4.585682111431098e-06, + "loss": 2.4208, + "step": 9944 + }, + { + "epoch": 1.051713197969543, + "grad_norm": 7.75, + "learning_rate": 4.5855876841087535e-06, + "loss": 2.5993, + "step": 9945 + }, + { + "epoch": 1.0518189509306262, + "grad_norm": 7.71875, + "learning_rate": 4.5854932469995995e-06, + "loss": 2.5098, + "step": 9946 + }, + { + "epoch": 1.051924703891709, + "grad_norm": 7.9375, + "learning_rate": 4.5853988001040775e-06, + "loss": 2.6262, + "step": 9947 + }, + { + "epoch": 1.0520304568527918, + "grad_norm": 8.0, + "learning_rate": 4.5853043434226315e-06, + "loss": 2.6015, + "step": 9948 + }, + { + "epoch": 1.0521362098138747, + "grad_norm": 7.90625, + "learning_rate": 4.585209876955706e-06, + "loss": 2.1045, + "step": 9949 + }, + { + "epoch": 1.0522419627749577, + "grad_norm": 8.5, + "learning_rate": 4.585115400703744e-06, + "loss": 2.3901, + "step": 9950 + }, + { + "epoch": 1.0523477157360406, + "grad_norm": 7.8125, + "learning_rate": 4.585020914667188e-06, + "loss": 2.4691, + "step": 9951 + }, + { + "epoch": 1.0524534686971234, + "grad_norm": 8.1875, + "learning_rate": 4.5849264188464805e-06, + "loss": 2.4997, + "step": 9952 + }, + { + "epoch": 1.0525592216582065, + "grad_norm": 8.0, + "learning_rate": 4.5848319132420665e-06, + "loss": 2.5756, + "step": 9953 + }, + { + "epoch": 1.0526649746192893, + "grad_norm": 7.9375, + "learning_rate": 4.584737397854389e-06, + "loss": 2.5956, + "step": 9954 + }, + { + "epoch": 1.0527707275803722, + "grad_norm": 8.3125, + "learning_rate": 4.5846428726838925e-06, + "loss": 2.591, + "step": 9955 + }, + { + "epoch": 1.0528764805414552, + "grad_norm": 8.375, + "learning_rate": 4.584548337731018e-06, + "loss": 2.7273, + "step": 9956 + }, + { + "epoch": 1.052982233502538, + "grad_norm": 8.4375, + "learning_rate": 4.5844537929962115e-06, + "loss": 2.5404, + "step": 9957 + }, + { + "epoch": 1.053087986463621, + "grad_norm": 9.0625, + "learning_rate": 4.5843592384799156e-06, + "loss": 2.7584, + "step": 9958 + }, + { + "epoch": 1.053193739424704, + "grad_norm": 8.6875, + "learning_rate": 4.584264674182574e-06, + "loss": 2.578, + "step": 9959 + }, + { + "epoch": 1.0532994923857868, + "grad_norm": 7.75, + "learning_rate": 4.584170100104631e-06, + "loss": 2.63, + "step": 9960 + }, + { + "epoch": 1.0534052453468696, + "grad_norm": 8.25, + "learning_rate": 4.58407551624653e-06, + "loss": 2.4101, + "step": 9961 + }, + { + "epoch": 1.0535109983079527, + "grad_norm": 8.0, + "learning_rate": 4.5839809226087165e-06, + "loss": 2.6066, + "step": 9962 + }, + { + "epoch": 1.0536167512690355, + "grad_norm": 8.1875, + "learning_rate": 4.5838863191916305e-06, + "loss": 2.6182, + "step": 9963 + }, + { + "epoch": 1.0537225042301184, + "grad_norm": 8.25, + "learning_rate": 4.583791705995719e-06, + "loss": 2.5004, + "step": 9964 + }, + { + "epoch": 1.0538282571912014, + "grad_norm": 8.125, + "learning_rate": 4.583697083021427e-06, + "loss": 2.3704, + "step": 9965 + }, + { + "epoch": 1.0539340101522843, + "grad_norm": 8.4375, + "learning_rate": 4.5836024502691945e-06, + "loss": 2.6055, + "step": 9966 + }, + { + "epoch": 1.0540397631133671, + "grad_norm": 8.125, + "learning_rate": 4.583507807739469e-06, + "loss": 2.5297, + "step": 9967 + }, + { + "epoch": 1.0541455160744502, + "grad_norm": 7.90625, + "learning_rate": 4.583413155432694e-06, + "loss": 2.5772, + "step": 9968 + }, + { + "epoch": 1.054251269035533, + "grad_norm": 8.5, + "learning_rate": 4.583318493349311e-06, + "loss": 2.6137, + "step": 9969 + }, + { + "epoch": 1.0543570219966159, + "grad_norm": 8.4375, + "learning_rate": 4.583223821489768e-06, + "loss": 2.6952, + "step": 9970 + }, + { + "epoch": 1.054462774957699, + "grad_norm": 8.5625, + "learning_rate": 4.583129139854507e-06, + "loss": 2.5988, + "step": 9971 + }, + { + "epoch": 1.0545685279187818, + "grad_norm": 8.0625, + "learning_rate": 4.583034448443974e-06, + "loss": 2.5426, + "step": 9972 + }, + { + "epoch": 1.0546742808798646, + "grad_norm": 8.4375, + "learning_rate": 4.582939747258611e-06, + "loss": 2.6101, + "step": 9973 + }, + { + "epoch": 1.0547800338409477, + "grad_norm": 7.8125, + "learning_rate": 4.582845036298865e-06, + "loss": 2.2293, + "step": 9974 + }, + { + "epoch": 1.0548857868020305, + "grad_norm": 7.78125, + "learning_rate": 4.5827503155651775e-06, + "loss": 2.7267, + "step": 9975 + }, + { + "epoch": 1.0549915397631133, + "grad_norm": 8.625, + "learning_rate": 4.582655585057996e-06, + "loss": 2.7103, + "step": 9976 + }, + { + "epoch": 1.0550972927241964, + "grad_norm": 8.125, + "learning_rate": 4.582560844777763e-06, + "loss": 2.5709, + "step": 9977 + }, + { + "epoch": 1.0552030456852792, + "grad_norm": 8.125, + "learning_rate": 4.582466094724924e-06, + "loss": 2.5242, + "step": 9978 + }, + { + "epoch": 1.055308798646362, + "grad_norm": 8.9375, + "learning_rate": 4.5823713348999225e-06, + "loss": 2.6957, + "step": 9979 + }, + { + "epoch": 1.055414551607445, + "grad_norm": 8.375, + "learning_rate": 4.5822765653032045e-06, + "loss": 2.5942, + "step": 9980 + }, + { + "epoch": 1.055520304568528, + "grad_norm": 8.75, + "learning_rate": 4.582181785935214e-06, + "loss": 2.721, + "step": 9981 + }, + { + "epoch": 1.0556260575296108, + "grad_norm": 8.25, + "learning_rate": 4.582086996796397e-06, + "loss": 2.6712, + "step": 9982 + }, + { + "epoch": 1.0557318104906936, + "grad_norm": 7.90625, + "learning_rate": 4.581992197887197e-06, + "loss": 2.5853, + "step": 9983 + }, + { + "epoch": 1.0558375634517767, + "grad_norm": 7.71875, + "learning_rate": 4.581897389208057e-06, + "loss": 2.3834, + "step": 9984 + }, + { + "epoch": 1.0559433164128595, + "grad_norm": 8.5625, + "learning_rate": 4.581802570759426e-06, + "loss": 2.5982, + "step": 9985 + }, + { + "epoch": 1.0560490693739424, + "grad_norm": 8.0, + "learning_rate": 4.581707742541747e-06, + "loss": 2.4362, + "step": 9986 + }, + { + "epoch": 1.0561548223350254, + "grad_norm": 8.0625, + "learning_rate": 4.581612904555464e-06, + "loss": 2.6144, + "step": 9987 + }, + { + "epoch": 1.0562605752961083, + "grad_norm": 7.84375, + "learning_rate": 4.581518056801023e-06, + "loss": 2.4355, + "step": 9988 + }, + { + "epoch": 1.0563663282571911, + "grad_norm": 10.5, + "learning_rate": 4.58142319927887e-06, + "loss": 2.2699, + "step": 9989 + }, + { + "epoch": 1.0564720812182742, + "grad_norm": 8.375, + "learning_rate": 4.581328331989449e-06, + "loss": 2.6218, + "step": 9990 + }, + { + "epoch": 1.056577834179357, + "grad_norm": 7.84375, + "learning_rate": 4.581233454933205e-06, + "loss": 2.3218, + "step": 9991 + }, + { + "epoch": 1.0566835871404399, + "grad_norm": 8.375, + "learning_rate": 4.581138568110584e-06, + "loss": 2.7884, + "step": 9992 + }, + { + "epoch": 1.056789340101523, + "grad_norm": 8.1875, + "learning_rate": 4.58104367152203e-06, + "loss": 2.6642, + "step": 9993 + }, + { + "epoch": 1.0568950930626058, + "grad_norm": 7.90625, + "learning_rate": 4.58094876516799e-06, + "loss": 2.448, + "step": 9994 + }, + { + "epoch": 1.0570008460236886, + "grad_norm": 7.96875, + "learning_rate": 4.580853849048909e-06, + "loss": 2.7294, + "step": 9995 + }, + { + "epoch": 1.0571065989847717, + "grad_norm": 8.25, + "learning_rate": 4.58075892316523e-06, + "loss": 2.5205, + "step": 9996 + }, + { + "epoch": 1.0572123519458545, + "grad_norm": 8.0, + "learning_rate": 4.5806639875174016e-06, + "loss": 2.5637, + "step": 9997 + }, + { + "epoch": 1.0573181049069373, + "grad_norm": 8.25, + "learning_rate": 4.580569042105869e-06, + "loss": 2.603, + "step": 9998 + }, + { + "epoch": 1.0574238578680204, + "grad_norm": 9.9375, + "learning_rate": 4.580474086931075e-06, + "loss": 2.6259, + "step": 9999 + }, + { + "epoch": 1.0575296108291032, + "grad_norm": 8.75, + "learning_rate": 4.580379121993468e-06, + "loss": 2.6532, + "step": 10000 + }, + { + "epoch": 1.057635363790186, + "grad_norm": 8.5625, + "learning_rate": 4.580284147293493e-06, + "loss": 2.4969, + "step": 10001 + }, + { + "epoch": 1.0577411167512691, + "grad_norm": 8.1875, + "learning_rate": 4.580189162831594e-06, + "loss": 2.5908, + "step": 10002 + }, + { + "epoch": 1.057846869712352, + "grad_norm": 8.375, + "learning_rate": 4.580094168608219e-06, + "loss": 2.5069, + "step": 10003 + }, + { + "epoch": 1.0579526226734348, + "grad_norm": 9.1875, + "learning_rate": 4.579999164623813e-06, + "loss": 2.7449, + "step": 10004 + }, + { + "epoch": 1.0580583756345177, + "grad_norm": 9.6875, + "learning_rate": 4.579904150878821e-06, + "loss": 2.5651, + "step": 10005 + }, + { + "epoch": 1.0581641285956007, + "grad_norm": 8.375, + "learning_rate": 4.57980912737369e-06, + "loss": 2.5109, + "step": 10006 + }, + { + "epoch": 1.0582698815566836, + "grad_norm": 8.3125, + "learning_rate": 4.579714094108865e-06, + "loss": 2.5332, + "step": 10007 + }, + { + "epoch": 1.0583756345177664, + "grad_norm": 8.0, + "learning_rate": 4.579619051084793e-06, + "loss": 2.6311, + "step": 10008 + }, + { + "epoch": 1.0584813874788495, + "grad_norm": 8.25, + "learning_rate": 4.579523998301918e-06, + "loss": 2.4415, + "step": 10009 + }, + { + "epoch": 1.0585871404399323, + "grad_norm": 8.5625, + "learning_rate": 4.579428935760688e-06, + "loss": 2.5947, + "step": 10010 + }, + { + "epoch": 1.0586928934010151, + "grad_norm": 7.84375, + "learning_rate": 4.579333863461548e-06, + "loss": 2.4736, + "step": 10011 + }, + { + "epoch": 1.0587986463620982, + "grad_norm": 8.0625, + "learning_rate": 4.579238781404945e-06, + "loss": 2.5395, + "step": 10012 + }, + { + "epoch": 1.058904399323181, + "grad_norm": 8.125, + "learning_rate": 4.579143689591326e-06, + "loss": 2.5641, + "step": 10013 + }, + { + "epoch": 1.0590101522842639, + "grad_norm": 8.1875, + "learning_rate": 4.5790485880211345e-06, + "loss": 2.2562, + "step": 10014 + }, + { + "epoch": 1.059115905245347, + "grad_norm": 7.8125, + "learning_rate": 4.5789534766948185e-06, + "loss": 2.6399, + "step": 10015 + }, + { + "epoch": 1.0592216582064298, + "grad_norm": 8.0, + "learning_rate": 4.578858355612824e-06, + "loss": 2.4222, + "step": 10016 + }, + { + "epoch": 1.0593274111675126, + "grad_norm": 8.25, + "learning_rate": 4.578763224775599e-06, + "loss": 2.7081, + "step": 10017 + }, + { + "epoch": 1.0594331641285957, + "grad_norm": 8.0625, + "learning_rate": 4.5786680841835865e-06, + "loss": 2.6092, + "step": 10018 + }, + { + "epoch": 1.0595389170896785, + "grad_norm": 8.125, + "learning_rate": 4.578572933837235e-06, + "loss": 2.5407, + "step": 10019 + }, + { + "epoch": 1.0596446700507614, + "grad_norm": 8.5, + "learning_rate": 4.5784777737369915e-06, + "loss": 2.6532, + "step": 10020 + }, + { + "epoch": 1.0597504230118444, + "grad_norm": 7.96875, + "learning_rate": 4.578382603883302e-06, + "loss": 2.7986, + "step": 10021 + }, + { + "epoch": 1.0598561759729273, + "grad_norm": 7.9375, + "learning_rate": 4.578287424276613e-06, + "loss": 2.2762, + "step": 10022 + }, + { + "epoch": 1.05996192893401, + "grad_norm": 8.375, + "learning_rate": 4.578192234917371e-06, + "loss": 2.5776, + "step": 10023 + }, + { + "epoch": 1.0600676818950932, + "grad_norm": 8.8125, + "learning_rate": 4.578097035806022e-06, + "loss": 2.7885, + "step": 10024 + }, + { + "epoch": 1.060173434856176, + "grad_norm": 8.625, + "learning_rate": 4.578001826943014e-06, + "loss": 2.6043, + "step": 10025 + }, + { + "epoch": 1.0602791878172588, + "grad_norm": 8.1875, + "learning_rate": 4.577906608328793e-06, + "loss": 2.6975, + "step": 10026 + }, + { + "epoch": 1.060384940778342, + "grad_norm": 9.6875, + "learning_rate": 4.577811379963807e-06, + "loss": 2.6092, + "step": 10027 + }, + { + "epoch": 1.0604906937394247, + "grad_norm": 8.5625, + "learning_rate": 4.577716141848501e-06, + "loss": 2.4986, + "step": 10028 + }, + { + "epoch": 1.0605964467005076, + "grad_norm": 8.0, + "learning_rate": 4.577620893983323e-06, + "loss": 2.3864, + "step": 10029 + }, + { + "epoch": 1.0607021996615906, + "grad_norm": 7.84375, + "learning_rate": 4.577525636368721e-06, + "loss": 2.5025, + "step": 10030 + }, + { + "epoch": 1.0608079526226735, + "grad_norm": 8.0625, + "learning_rate": 4.577430369005141e-06, + "loss": 2.5257, + "step": 10031 + }, + { + "epoch": 1.0609137055837563, + "grad_norm": 8.375, + "learning_rate": 4.577335091893028e-06, + "loss": 2.3201, + "step": 10032 + }, + { + "epoch": 1.0610194585448394, + "grad_norm": 11.125, + "learning_rate": 4.577239805032832e-06, + "loss": 2.3937, + "step": 10033 + }, + { + "epoch": 1.0611252115059222, + "grad_norm": 8.4375, + "learning_rate": 4.577144508424999e-06, + "loss": 2.5921, + "step": 10034 + }, + { + "epoch": 1.061230964467005, + "grad_norm": 8.125, + "learning_rate": 4.577049202069977e-06, + "loss": 2.4596, + "step": 10035 + }, + { + "epoch": 1.0613367174280879, + "grad_norm": 8.4375, + "learning_rate": 4.576953885968212e-06, + "loss": 2.6596, + "step": 10036 + }, + { + "epoch": 1.061442470389171, + "grad_norm": 8.625, + "learning_rate": 4.576858560120152e-06, + "loss": 2.553, + "step": 10037 + }, + { + "epoch": 1.0615482233502538, + "grad_norm": 7.9375, + "learning_rate": 4.576763224526244e-06, + "loss": 2.541, + "step": 10038 + }, + { + "epoch": 1.0616539763113366, + "grad_norm": 7.71875, + "learning_rate": 4.576667879186936e-06, + "loss": 2.4918, + "step": 10039 + }, + { + "epoch": 1.0617597292724197, + "grad_norm": 7.625, + "learning_rate": 4.576572524102674e-06, + "loss": 2.4876, + "step": 10040 + }, + { + "epoch": 1.0618654822335025, + "grad_norm": 9.0625, + "learning_rate": 4.5764771592739076e-06, + "loss": 2.6041, + "step": 10041 + }, + { + "epoch": 1.0619712351945854, + "grad_norm": 7.96875, + "learning_rate": 4.576381784701083e-06, + "loss": 2.6313, + "step": 10042 + }, + { + "epoch": 1.0620769881556684, + "grad_norm": 7.5625, + "learning_rate": 4.576286400384647e-06, + "loss": 2.4994, + "step": 10043 + }, + { + "epoch": 1.0621827411167513, + "grad_norm": 7.96875, + "learning_rate": 4.576191006325049e-06, + "loss": 2.5205, + "step": 10044 + }, + { + "epoch": 1.062288494077834, + "grad_norm": 7.875, + "learning_rate": 4.576095602522735e-06, + "loss": 2.5, + "step": 10045 + }, + { + "epoch": 1.0623942470389172, + "grad_norm": 8.25, + "learning_rate": 4.576000188978154e-06, + "loss": 2.5141, + "step": 10046 + }, + { + "epoch": 1.0625, + "grad_norm": 8.75, + "learning_rate": 4.575904765691753e-06, + "loss": 2.5717, + "step": 10047 + }, + { + "epoch": 1.0626057529610828, + "grad_norm": 8.125, + "learning_rate": 4.57580933266398e-06, + "loss": 2.557, + "step": 10048 + }, + { + "epoch": 1.062711505922166, + "grad_norm": 8.5625, + "learning_rate": 4.5757138898952824e-06, + "loss": 2.6013, + "step": 10049 + }, + { + "epoch": 1.0628172588832487, + "grad_norm": 8.1875, + "learning_rate": 4.575618437386109e-06, + "loss": 2.5097, + "step": 10050 + }, + { + "epoch": 1.0629230118443316, + "grad_norm": 7.90625, + "learning_rate": 4.5755229751369064e-06, + "loss": 2.6771, + "step": 10051 + }, + { + "epoch": 1.0630287648054146, + "grad_norm": 7.8125, + "learning_rate": 4.575427503148125e-06, + "loss": 2.4734, + "step": 10052 + }, + { + "epoch": 1.0631345177664975, + "grad_norm": 8.0, + "learning_rate": 4.57533202142021e-06, + "loss": 2.4577, + "step": 10053 + }, + { + "epoch": 1.0632402707275803, + "grad_norm": 8.0, + "learning_rate": 4.57523652995361e-06, + "loss": 2.5937, + "step": 10054 + }, + { + "epoch": 1.0633460236886634, + "grad_norm": 8.25, + "learning_rate": 4.575141028748774e-06, + "loss": 2.6274, + "step": 10055 + }, + { + "epoch": 1.0634517766497462, + "grad_norm": 8.375, + "learning_rate": 4.57504551780615e-06, + "loss": 2.4049, + "step": 10056 + }, + { + "epoch": 1.063557529610829, + "grad_norm": 8.0, + "learning_rate": 4.574949997126187e-06, + "loss": 2.6827, + "step": 10057 + }, + { + "epoch": 1.0636632825719121, + "grad_norm": 7.96875, + "learning_rate": 4.5748544667093315e-06, + "loss": 2.464, + "step": 10058 + }, + { + "epoch": 1.063769035532995, + "grad_norm": 8.125, + "learning_rate": 4.574758926556032e-06, + "loss": 2.488, + "step": 10059 + }, + { + "epoch": 1.0638747884940778, + "grad_norm": 8.25, + "learning_rate": 4.574663376666738e-06, + "loss": 2.5444, + "step": 10060 + }, + { + "epoch": 1.0639805414551606, + "grad_norm": 7.59375, + "learning_rate": 4.5745678170418975e-06, + "loss": 2.4175, + "step": 10061 + }, + { + "epoch": 1.0640862944162437, + "grad_norm": 7.8125, + "learning_rate": 4.57447224768196e-06, + "loss": 2.5501, + "step": 10062 + }, + { + "epoch": 1.0641920473773265, + "grad_norm": 8.25, + "learning_rate": 4.57437666858737e-06, + "loss": 2.4234, + "step": 10063 + }, + { + "epoch": 1.0642978003384094, + "grad_norm": 8.3125, + "learning_rate": 4.57428107975858e-06, + "loss": 2.5223, + "step": 10064 + }, + { + "epoch": 1.0644035532994924, + "grad_norm": 8.125, + "learning_rate": 4.574185481196038e-06, + "loss": 2.6739, + "step": 10065 + }, + { + "epoch": 1.0645093062605753, + "grad_norm": 7.8125, + "learning_rate": 4.574089872900191e-06, + "loss": 2.5662, + "step": 10066 + }, + { + "epoch": 1.064615059221658, + "grad_norm": 8.125, + "learning_rate": 4.573994254871488e-06, + "loss": 2.3643, + "step": 10067 + }, + { + "epoch": 1.0647208121827412, + "grad_norm": 8.125, + "learning_rate": 4.573898627110378e-06, + "loss": 2.5002, + "step": 10068 + }, + { + "epoch": 1.064826565143824, + "grad_norm": 8.0625, + "learning_rate": 4.573802989617311e-06, + "loss": 2.5001, + "step": 10069 + }, + { + "epoch": 1.0649323181049068, + "grad_norm": 8.1875, + "learning_rate": 4.573707342392734e-06, + "loss": 2.6878, + "step": 10070 + }, + { + "epoch": 1.06503807106599, + "grad_norm": 7.96875, + "learning_rate": 4.5736116854370974e-06, + "loss": 2.5078, + "step": 10071 + }, + { + "epoch": 1.0651438240270727, + "grad_norm": 8.9375, + "learning_rate": 4.573516018750848e-06, + "loss": 2.7755, + "step": 10072 + }, + { + "epoch": 1.0652495769881556, + "grad_norm": 8.25, + "learning_rate": 4.573420342334437e-06, + "loss": 2.5983, + "step": 10073 + }, + { + "epoch": 1.0653553299492386, + "grad_norm": 8.875, + "learning_rate": 4.573324656188312e-06, + "loss": 2.7775, + "step": 10074 + }, + { + "epoch": 1.0654610829103215, + "grad_norm": 7.9375, + "learning_rate": 4.5732289603129225e-06, + "loss": 2.4107, + "step": 10075 + }, + { + "epoch": 1.0655668358714043, + "grad_norm": 8.5625, + "learning_rate": 4.573133254708717e-06, + "loss": 2.6107, + "step": 10076 + }, + { + "epoch": 1.0656725888324874, + "grad_norm": 8.0, + "learning_rate": 4.573037539376145e-06, + "loss": 2.5138, + "step": 10077 + }, + { + "epoch": 1.0657783417935702, + "grad_norm": 7.8125, + "learning_rate": 4.572941814315657e-06, + "loss": 2.4394, + "step": 10078 + }, + { + "epoch": 1.065884094754653, + "grad_norm": 7.875, + "learning_rate": 4.5728460795276995e-06, + "loss": 2.2819, + "step": 10079 + }, + { + "epoch": 1.0659898477157361, + "grad_norm": 8.25, + "learning_rate": 4.572750335012724e-06, + "loss": 2.5517, + "step": 10080 + }, + { + "epoch": 1.066095600676819, + "grad_norm": 7.8125, + "learning_rate": 4.572654580771177e-06, + "loss": 2.5436, + "step": 10081 + }, + { + "epoch": 1.0662013536379018, + "grad_norm": 7.71875, + "learning_rate": 4.572558816803512e-06, + "loss": 2.3287, + "step": 10082 + }, + { + "epoch": 1.0663071065989849, + "grad_norm": 9.9375, + "learning_rate": 4.572463043110175e-06, + "loss": 2.6775, + "step": 10083 + }, + { + "epoch": 1.0664128595600677, + "grad_norm": 7.78125, + "learning_rate": 4.572367259691617e-06, + "loss": 2.4291, + "step": 10084 + }, + { + "epoch": 1.0665186125211505, + "grad_norm": 7.90625, + "learning_rate": 4.572271466548286e-06, + "loss": 2.4393, + "step": 10085 + }, + { + "epoch": 1.0666243654822336, + "grad_norm": 7.78125, + "learning_rate": 4.572175663680634e-06, + "loss": 2.5573, + "step": 10086 + }, + { + "epoch": 1.0667301184433164, + "grad_norm": 8.0, + "learning_rate": 4.572079851089108e-06, + "loss": 2.421, + "step": 10087 + }, + { + "epoch": 1.0668358714043993, + "grad_norm": 8.3125, + "learning_rate": 4.571984028774159e-06, + "loss": 2.6801, + "step": 10088 + }, + { + "epoch": 1.0669416243654823, + "grad_norm": 8.375, + "learning_rate": 4.571888196736237e-06, + "loss": 2.6846, + "step": 10089 + }, + { + "epoch": 1.0670473773265652, + "grad_norm": 7.96875, + "learning_rate": 4.5717923549757904e-06, + "loss": 2.7438, + "step": 10090 + }, + { + "epoch": 1.067153130287648, + "grad_norm": 7.90625, + "learning_rate": 4.571696503493269e-06, + "loss": 2.4741, + "step": 10091 + }, + { + "epoch": 1.0672588832487309, + "grad_norm": 8.4375, + "learning_rate": 4.5716006422891245e-06, + "loss": 2.5799, + "step": 10092 + }, + { + "epoch": 1.067364636209814, + "grad_norm": 8.5625, + "learning_rate": 4.571504771363805e-06, + "loss": 2.7285, + "step": 10093 + }, + { + "epoch": 1.0674703891708968, + "grad_norm": 8.5, + "learning_rate": 4.57140889071776e-06, + "loss": 2.3129, + "step": 10094 + }, + { + "epoch": 1.0675761421319796, + "grad_norm": 7.65625, + "learning_rate": 4.571313000351442e-06, + "loss": 2.544, + "step": 10095 + }, + { + "epoch": 1.0676818950930627, + "grad_norm": 8.9375, + "learning_rate": 4.5712171002652985e-06, + "loss": 2.4794, + "step": 10096 + }, + { + "epoch": 1.0677876480541455, + "grad_norm": 7.90625, + "learning_rate": 4.57112119045978e-06, + "loss": 2.5208, + "step": 10097 + }, + { + "epoch": 1.0678934010152283, + "grad_norm": 7.46875, + "learning_rate": 4.5710252709353365e-06, + "loss": 2.3993, + "step": 10098 + }, + { + "epoch": 1.0679991539763114, + "grad_norm": 7.65625, + "learning_rate": 4.570929341692418e-06, + "loss": 2.5513, + "step": 10099 + }, + { + "epoch": 1.0681049069373942, + "grad_norm": 8.125, + "learning_rate": 4.570833402731476e-06, + "loss": 2.7966, + "step": 10100 + }, + { + "epoch": 1.068210659898477, + "grad_norm": 8.4375, + "learning_rate": 4.57073745405296e-06, + "loss": 2.2803, + "step": 10101 + }, + { + "epoch": 1.0683164128595601, + "grad_norm": 8.0, + "learning_rate": 4.57064149565732e-06, + "loss": 2.5869, + "step": 10102 + }, + { + "epoch": 1.068422165820643, + "grad_norm": 8.3125, + "learning_rate": 4.570545527545005e-06, + "loss": 2.223, + "step": 10103 + }, + { + "epoch": 1.0685279187817258, + "grad_norm": 7.96875, + "learning_rate": 4.570449549716468e-06, + "loss": 2.5484, + "step": 10104 + }, + { + "epoch": 1.0686336717428089, + "grad_norm": 8.1875, + "learning_rate": 4.570353562172157e-06, + "loss": 2.472, + "step": 10105 + }, + { + "epoch": 1.0687394247038917, + "grad_norm": 7.75, + "learning_rate": 4.570257564912525e-06, + "loss": 2.4819, + "step": 10106 + }, + { + "epoch": 1.0688451776649746, + "grad_norm": 8.25, + "learning_rate": 4.57016155793802e-06, + "loss": 2.4675, + "step": 10107 + }, + { + "epoch": 1.0689509306260576, + "grad_norm": 8.25, + "learning_rate": 4.570065541249094e-06, + "loss": 2.2913, + "step": 10108 + }, + { + "epoch": 1.0690566835871405, + "grad_norm": 8.5625, + "learning_rate": 4.569969514846196e-06, + "loss": 2.6673, + "step": 10109 + }, + { + "epoch": 1.0691624365482233, + "grad_norm": 7.8125, + "learning_rate": 4.569873478729778e-06, + "loss": 2.1579, + "step": 10110 + }, + { + "epoch": 1.0692681895093064, + "grad_norm": 9.0625, + "learning_rate": 4.569777432900291e-06, + "loss": 2.6782, + "step": 10111 + }, + { + "epoch": 1.0693739424703892, + "grad_norm": 8.0, + "learning_rate": 4.569681377358184e-06, + "loss": 2.5473, + "step": 10112 + }, + { + "epoch": 1.069479695431472, + "grad_norm": 8.5625, + "learning_rate": 4.569585312103909e-06, + "loss": 2.6342, + "step": 10113 + }, + { + "epoch": 1.069585448392555, + "grad_norm": 8.125, + "learning_rate": 4.569489237137917e-06, + "loss": 2.5175, + "step": 10114 + }, + { + "epoch": 1.069691201353638, + "grad_norm": 9.0, + "learning_rate": 4.569393152460659e-06, + "loss": 2.7244, + "step": 10115 + }, + { + "epoch": 1.0697969543147208, + "grad_norm": 8.375, + "learning_rate": 4.5692970580725845e-06, + "loss": 2.5998, + "step": 10116 + }, + { + "epoch": 1.0699027072758036, + "grad_norm": 8.3125, + "learning_rate": 4.569200953974145e-06, + "loss": 2.608, + "step": 10117 + }, + { + "epoch": 1.0700084602368867, + "grad_norm": 7.59375, + "learning_rate": 4.569104840165792e-06, + "loss": 2.4163, + "step": 10118 + }, + { + "epoch": 1.0701142131979695, + "grad_norm": 7.90625, + "learning_rate": 4.5690087166479765e-06, + "loss": 2.6518, + "step": 10119 + }, + { + "epoch": 1.0702199661590523, + "grad_norm": 8.0625, + "learning_rate": 4.568912583421149e-06, + "loss": 2.5141, + "step": 10120 + }, + { + "epoch": 1.0703257191201354, + "grad_norm": 8.625, + "learning_rate": 4.56881644048576e-06, + "loss": 2.7077, + "step": 10121 + }, + { + "epoch": 1.0704314720812182, + "grad_norm": 8.125, + "learning_rate": 4.568720287842263e-06, + "loss": 2.6127, + "step": 10122 + }, + { + "epoch": 1.070537225042301, + "grad_norm": 8.0, + "learning_rate": 4.568624125491108e-06, + "loss": 2.5137, + "step": 10123 + }, + { + "epoch": 1.0706429780033841, + "grad_norm": 8.3125, + "learning_rate": 4.5685279534327445e-06, + "loss": 2.6099, + "step": 10124 + }, + { + "epoch": 1.070748730964467, + "grad_norm": 8.875, + "learning_rate": 4.568431771667627e-06, + "loss": 2.5862, + "step": 10125 + }, + { + "epoch": 1.0708544839255498, + "grad_norm": 8.8125, + "learning_rate": 4.568335580196204e-06, + "loss": 2.3247, + "step": 10126 + }, + { + "epoch": 1.0709602368866329, + "grad_norm": 8.0625, + "learning_rate": 4.568239379018928e-06, + "loss": 2.7096, + "step": 10127 + }, + { + "epoch": 1.0710659898477157, + "grad_norm": 8.5625, + "learning_rate": 4.568143168136251e-06, + "loss": 2.4183, + "step": 10128 + }, + { + "epoch": 1.0711717428087986, + "grad_norm": 7.8125, + "learning_rate": 4.5680469475486235e-06, + "loss": 2.3667, + "step": 10129 + }, + { + "epoch": 1.0712774957698816, + "grad_norm": 7.59375, + "learning_rate": 4.567950717256498e-06, + "loss": 2.542, + "step": 10130 + }, + { + "epoch": 1.0713832487309645, + "grad_norm": 8.0, + "learning_rate": 4.567854477260325e-06, + "loss": 2.5599, + "step": 10131 + }, + { + "epoch": 1.0714890016920473, + "grad_norm": 8.25, + "learning_rate": 4.567758227560557e-06, + "loss": 2.7335, + "step": 10132 + }, + { + "epoch": 1.0715947546531304, + "grad_norm": 7.90625, + "learning_rate": 4.567661968157645e-06, + "loss": 2.6013, + "step": 10133 + }, + { + "epoch": 1.0717005076142132, + "grad_norm": 7.4375, + "learning_rate": 4.567565699052041e-06, + "loss": 2.4336, + "step": 10134 + }, + { + "epoch": 1.071806260575296, + "grad_norm": 8.0, + "learning_rate": 4.567469420244197e-06, + "loss": 2.5692, + "step": 10135 + }, + { + "epoch": 1.071912013536379, + "grad_norm": 8.1875, + "learning_rate": 4.5673731317345645e-06, + "loss": 2.636, + "step": 10136 + }, + { + "epoch": 1.072017766497462, + "grad_norm": 7.625, + "learning_rate": 4.5672768335235955e-06, + "loss": 2.6528, + "step": 10137 + }, + { + "epoch": 1.0721235194585448, + "grad_norm": 8.4375, + "learning_rate": 4.567180525611742e-06, + "loss": 2.7037, + "step": 10138 + }, + { + "epoch": 1.0722292724196278, + "grad_norm": 8.0625, + "learning_rate": 4.5670842079994555e-06, + "loss": 2.7443, + "step": 10139 + }, + { + "epoch": 1.0723350253807107, + "grad_norm": 8.1875, + "learning_rate": 4.566987880687188e-06, + "loss": 2.4652, + "step": 10140 + }, + { + "epoch": 1.0724407783417935, + "grad_norm": 8.125, + "learning_rate": 4.5668915436753925e-06, + "loss": 2.6121, + "step": 10141 + }, + { + "epoch": 1.0725465313028766, + "grad_norm": 8.1875, + "learning_rate": 4.56679519696452e-06, + "loss": 2.7924, + "step": 10142 + }, + { + "epoch": 1.0726522842639594, + "grad_norm": 8.75, + "learning_rate": 4.5666988405550224e-06, + "loss": 2.5693, + "step": 10143 + }, + { + "epoch": 1.0727580372250423, + "grad_norm": 8.1875, + "learning_rate": 4.566602474447353e-06, + "loss": 2.6094, + "step": 10144 + }, + { + "epoch": 1.0728637901861253, + "grad_norm": 8.3125, + "learning_rate": 4.566506098641963e-06, + "loss": 2.4732, + "step": 10145 + }, + { + "epoch": 1.0729695431472082, + "grad_norm": 7.96875, + "learning_rate": 4.5664097131393056e-06, + "loss": 2.7321, + "step": 10146 + }, + { + "epoch": 1.073075296108291, + "grad_norm": 9.9375, + "learning_rate": 4.566313317939832e-06, + "loss": 2.7174, + "step": 10147 + }, + { + "epoch": 1.0731810490693738, + "grad_norm": 8.6875, + "learning_rate": 4.566216913043995e-06, + "loss": 2.7113, + "step": 10148 + }, + { + "epoch": 1.073286802030457, + "grad_norm": 8.1875, + "learning_rate": 4.566120498452248e-06, + "loss": 2.488, + "step": 10149 + }, + { + "epoch": 1.0733925549915397, + "grad_norm": 7.71875, + "learning_rate": 4.566024074165042e-06, + "loss": 2.3731, + "step": 10150 + }, + { + "epoch": 1.0734983079526226, + "grad_norm": 8.0625, + "learning_rate": 4.565927640182829e-06, + "loss": 2.4292, + "step": 10151 + }, + { + "epoch": 1.0736040609137056, + "grad_norm": 8.5, + "learning_rate": 4.565831196506064e-06, + "loss": 2.7094, + "step": 10152 + }, + { + "epoch": 1.0737098138747885, + "grad_norm": 8.0, + "learning_rate": 4.565734743135197e-06, + "loss": 2.5393, + "step": 10153 + }, + { + "epoch": 1.0738155668358713, + "grad_norm": 7.90625, + "learning_rate": 4.565638280070682e-06, + "loss": 2.5468, + "step": 10154 + }, + { + "epoch": 1.0739213197969544, + "grad_norm": 8.25, + "learning_rate": 4.565541807312972e-06, + "loss": 2.3706, + "step": 10155 + }, + { + "epoch": 1.0740270727580372, + "grad_norm": 7.9375, + "learning_rate": 4.5654453248625184e-06, + "loss": 2.546, + "step": 10156 + }, + { + "epoch": 1.07413282571912, + "grad_norm": 8.4375, + "learning_rate": 4.565348832719775e-06, + "loss": 2.6035, + "step": 10157 + }, + { + "epoch": 1.0742385786802031, + "grad_norm": 7.65625, + "learning_rate": 4.565252330885193e-06, + "loss": 2.495, + "step": 10158 + }, + { + "epoch": 1.074344331641286, + "grad_norm": 7.65625, + "learning_rate": 4.565155819359228e-06, + "loss": 2.5466, + "step": 10159 + }, + { + "epoch": 1.0744500846023688, + "grad_norm": 8.0, + "learning_rate": 4.565059298142331e-06, + "loss": 2.6001, + "step": 10160 + }, + { + "epoch": 1.0745558375634519, + "grad_norm": 8.3125, + "learning_rate": 4.5649627672349556e-06, + "loss": 2.7061, + "step": 10161 + }, + { + "epoch": 1.0746615905245347, + "grad_norm": 8.3125, + "learning_rate": 4.564866226637554e-06, + "loss": 2.5964, + "step": 10162 + }, + { + "epoch": 1.0747673434856175, + "grad_norm": 7.875, + "learning_rate": 4.56476967635058e-06, + "loss": 2.5967, + "step": 10163 + }, + { + "epoch": 1.0748730964467006, + "grad_norm": 8.3125, + "learning_rate": 4.564673116374486e-06, + "loss": 2.6408, + "step": 10164 + }, + { + "epoch": 1.0749788494077834, + "grad_norm": 8.5, + "learning_rate": 4.564576546709726e-06, + "loss": 2.0088, + "step": 10165 + }, + { + "epoch": 1.0750846023688663, + "grad_norm": 8.3125, + "learning_rate": 4.564479967356752e-06, + "loss": 2.7277, + "step": 10166 + }, + { + "epoch": 1.0751903553299493, + "grad_norm": 8.5625, + "learning_rate": 4.564383378316018e-06, + "loss": 2.8612, + "step": 10167 + }, + { + "epoch": 1.0752961082910322, + "grad_norm": 8.0, + "learning_rate": 4.564286779587977e-06, + "loss": 2.4148, + "step": 10168 + }, + { + "epoch": 1.075401861252115, + "grad_norm": 9.0625, + "learning_rate": 4.564190171173083e-06, + "loss": 2.6713, + "step": 10169 + }, + { + "epoch": 1.075507614213198, + "grad_norm": 8.0625, + "learning_rate": 4.5640935530717884e-06, + "loss": 2.6122, + "step": 10170 + }, + { + "epoch": 1.075613367174281, + "grad_norm": 8.1875, + "learning_rate": 4.563996925284547e-06, + "loss": 2.662, + "step": 10171 + }, + { + "epoch": 1.0757191201353637, + "grad_norm": 8.1875, + "learning_rate": 4.563900287811812e-06, + "loss": 2.4158, + "step": 10172 + }, + { + "epoch": 1.0758248730964466, + "grad_norm": 8.1875, + "learning_rate": 4.563803640654037e-06, + "loss": 2.5477, + "step": 10173 + }, + { + "epoch": 1.0759306260575296, + "grad_norm": 8.1875, + "learning_rate": 4.563706983811676e-06, + "loss": 2.607, + "step": 10174 + }, + { + "epoch": 1.0760363790186125, + "grad_norm": 8.0, + "learning_rate": 4.563610317285182e-06, + "loss": 2.5715, + "step": 10175 + }, + { + "epoch": 1.0761421319796955, + "grad_norm": 7.9375, + "learning_rate": 4.563513641075008e-06, + "loss": 2.4303, + "step": 10176 + }, + { + "epoch": 1.0762478849407784, + "grad_norm": 7.9375, + "learning_rate": 4.56341695518161e-06, + "loss": 2.4857, + "step": 10177 + }, + { + "epoch": 1.0763536379018612, + "grad_norm": 8.0, + "learning_rate": 4.563320259605439e-06, + "loss": 2.4631, + "step": 10178 + }, + { + "epoch": 1.076459390862944, + "grad_norm": 8.75, + "learning_rate": 4.56322355434695e-06, + "loss": 2.7157, + "step": 10179 + }, + { + "epoch": 1.0765651438240271, + "grad_norm": 7.75, + "learning_rate": 4.563126839406596e-06, + "loss": 2.5098, + "step": 10180 + }, + { + "epoch": 1.07667089678511, + "grad_norm": 8.1875, + "learning_rate": 4.5630301147848336e-06, + "loss": 2.6593, + "step": 10181 + }, + { + "epoch": 1.0767766497461928, + "grad_norm": 9.3125, + "learning_rate": 4.562933380482113e-06, + "loss": 2.806, + "step": 10182 + }, + { + "epoch": 1.0768824027072759, + "grad_norm": 8.25, + "learning_rate": 4.56283663649889e-06, + "loss": 2.4699, + "step": 10183 + }, + { + "epoch": 1.0769881556683587, + "grad_norm": 7.84375, + "learning_rate": 4.562739882835618e-06, + "loss": 2.3822, + "step": 10184 + }, + { + "epoch": 1.0770939086294415, + "grad_norm": 7.65625, + "learning_rate": 4.5626431194927525e-06, + "loss": 2.5895, + "step": 10185 + }, + { + "epoch": 1.0771996615905246, + "grad_norm": 7.90625, + "learning_rate": 4.562546346470746e-06, + "loss": 2.5151, + "step": 10186 + }, + { + "epoch": 1.0773054145516074, + "grad_norm": 8.125, + "learning_rate": 4.562449563770053e-06, + "loss": 2.6111, + "step": 10187 + }, + { + "epoch": 1.0774111675126903, + "grad_norm": 8.3125, + "learning_rate": 4.562352771391127e-06, + "loss": 2.4696, + "step": 10188 + }, + { + "epoch": 1.0775169204737733, + "grad_norm": 8.0, + "learning_rate": 4.562255969334424e-06, + "loss": 2.5573, + "step": 10189 + }, + { + "epoch": 1.0776226734348562, + "grad_norm": 8.1875, + "learning_rate": 4.5621591576003965e-06, + "loss": 2.6442, + "step": 10190 + }, + { + "epoch": 1.077728426395939, + "grad_norm": 8.75, + "learning_rate": 4.562062336189499e-06, + "loss": 2.6928, + "step": 10191 + }, + { + "epoch": 1.077834179357022, + "grad_norm": 8.3125, + "learning_rate": 4.561965505102187e-06, + "loss": 2.7559, + "step": 10192 + }, + { + "epoch": 1.077939932318105, + "grad_norm": 7.46875, + "learning_rate": 4.561868664338914e-06, + "loss": 2.4356, + "step": 10193 + }, + { + "epoch": 1.0780456852791878, + "grad_norm": 8.0625, + "learning_rate": 4.5617718139001345e-06, + "loss": 2.6182, + "step": 10194 + }, + { + "epoch": 1.0781514382402708, + "grad_norm": 7.71875, + "learning_rate": 4.561674953786304e-06, + "loss": 2.5467, + "step": 10195 + }, + { + "epoch": 1.0782571912013537, + "grad_norm": 8.25, + "learning_rate": 4.561578083997875e-06, + "loss": 2.4866, + "step": 10196 + }, + { + "epoch": 1.0783629441624365, + "grad_norm": 8.6875, + "learning_rate": 4.561481204535304e-06, + "loss": 2.6619, + "step": 10197 + }, + { + "epoch": 1.0784686971235196, + "grad_norm": 8.5, + "learning_rate": 4.561384315399045e-06, + "loss": 2.779, + "step": 10198 + }, + { + "epoch": 1.0785744500846024, + "grad_norm": 8.8125, + "learning_rate": 4.561287416589552e-06, + "loss": 2.9961, + "step": 10199 + }, + { + "epoch": 1.0786802030456852, + "grad_norm": 8.3125, + "learning_rate": 4.56119050810728e-06, + "loss": 2.5311, + "step": 10200 + }, + { + "epoch": 1.0787859560067683, + "grad_norm": 8.3125, + "learning_rate": 4.561093589952684e-06, + "loss": 2.5951, + "step": 10201 + }, + { + "epoch": 1.0788917089678511, + "grad_norm": 8.5, + "learning_rate": 4.560996662126219e-06, + "loss": 2.6694, + "step": 10202 + }, + { + "epoch": 1.078997461928934, + "grad_norm": 8.25, + "learning_rate": 4.560899724628339e-06, + "loss": 2.5686, + "step": 10203 + }, + { + "epoch": 1.0791032148900168, + "grad_norm": 8.8125, + "learning_rate": 4.5608027774595e-06, + "loss": 2.5875, + "step": 10204 + }, + { + "epoch": 1.0792089678510999, + "grad_norm": 8.25, + "learning_rate": 4.560705820620156e-06, + "loss": 2.6573, + "step": 10205 + }, + { + "epoch": 1.0793147208121827, + "grad_norm": 7.8125, + "learning_rate": 4.560608854110763e-06, + "loss": 2.6649, + "step": 10206 + }, + { + "epoch": 1.0794204737732656, + "grad_norm": 8.875, + "learning_rate": 4.560511877931775e-06, + "loss": 2.2898, + "step": 10207 + }, + { + "epoch": 1.0795262267343486, + "grad_norm": 8.1875, + "learning_rate": 4.560414892083648e-06, + "loss": 2.5675, + "step": 10208 + }, + { + "epoch": 1.0796319796954315, + "grad_norm": 7.90625, + "learning_rate": 4.560317896566836e-06, + "loss": 2.3146, + "step": 10209 + }, + { + "epoch": 1.0797377326565143, + "grad_norm": 8.0625, + "learning_rate": 4.560220891381795e-06, + "loss": 2.4894, + "step": 10210 + }, + { + "epoch": 1.0798434856175974, + "grad_norm": 8.9375, + "learning_rate": 4.56012387652898e-06, + "loss": 2.485, + "step": 10211 + }, + { + "epoch": 1.0799492385786802, + "grad_norm": 8.625, + "learning_rate": 4.5600268520088455e-06, + "loss": 2.6757, + "step": 10212 + }, + { + "epoch": 1.080054991539763, + "grad_norm": 7.8125, + "learning_rate": 4.559929817821848e-06, + "loss": 2.4836, + "step": 10213 + }, + { + "epoch": 1.080160744500846, + "grad_norm": 8.1875, + "learning_rate": 4.5598327739684426e-06, + "loss": 2.7791, + "step": 10214 + }, + { + "epoch": 1.080266497461929, + "grad_norm": 8.8125, + "learning_rate": 4.559735720449085e-06, + "loss": 2.688, + "step": 10215 + }, + { + "epoch": 1.0803722504230118, + "grad_norm": 8.0, + "learning_rate": 4.55963865726423e-06, + "loss": 2.4941, + "step": 10216 + }, + { + "epoch": 1.0804780033840948, + "grad_norm": 8.1875, + "learning_rate": 4.5595415844143315e-06, + "loss": 2.4457, + "step": 10217 + }, + { + "epoch": 1.0805837563451777, + "grad_norm": 8.0, + "learning_rate": 4.559444501899848e-06, + "loss": 2.6521, + "step": 10218 + }, + { + "epoch": 1.0806895093062605, + "grad_norm": 8.0625, + "learning_rate": 4.559347409721233e-06, + "loss": 2.5458, + "step": 10219 + }, + { + "epoch": 1.0807952622673436, + "grad_norm": 7.875, + "learning_rate": 4.559250307878943e-06, + "loss": 2.6208, + "step": 10220 + }, + { + "epoch": 1.0809010152284264, + "grad_norm": 7.96875, + "learning_rate": 4.559153196373434e-06, + "loss": 2.6822, + "step": 10221 + }, + { + "epoch": 1.0810067681895092, + "grad_norm": 8.4375, + "learning_rate": 4.559056075205161e-06, + "loss": 2.3509, + "step": 10222 + }, + { + "epoch": 1.0811125211505923, + "grad_norm": 7.875, + "learning_rate": 4.55895894437458e-06, + "loss": 2.3216, + "step": 10223 + }, + { + "epoch": 1.0812182741116751, + "grad_norm": 8.0, + "learning_rate": 4.558861803882147e-06, + "loss": 2.4079, + "step": 10224 + }, + { + "epoch": 1.081324027072758, + "grad_norm": 7.90625, + "learning_rate": 4.558764653728317e-06, + "loss": 2.4839, + "step": 10225 + }, + { + "epoch": 1.081429780033841, + "grad_norm": 9.125, + "learning_rate": 4.558667493913547e-06, + "loss": 2.5725, + "step": 10226 + }, + { + "epoch": 1.0815355329949239, + "grad_norm": 8.1875, + "learning_rate": 4.558570324438292e-06, + "loss": 2.6627, + "step": 10227 + }, + { + "epoch": 1.0816412859560067, + "grad_norm": 8.4375, + "learning_rate": 4.558473145303008e-06, + "loss": 2.5542, + "step": 10228 + }, + { + "epoch": 1.0817470389170896, + "grad_norm": 8.3125, + "learning_rate": 4.558375956508153e-06, + "loss": 2.4977, + "step": 10229 + }, + { + "epoch": 1.0818527918781726, + "grad_norm": 7.9375, + "learning_rate": 4.558278758054181e-06, + "loss": 2.4963, + "step": 10230 + }, + { + "epoch": 1.0819585448392555, + "grad_norm": 8.375, + "learning_rate": 4.558181549941547e-06, + "loss": 2.2768, + "step": 10231 + }, + { + "epoch": 1.0820642978003385, + "grad_norm": 7.9375, + "learning_rate": 4.55808433217071e-06, + "loss": 2.5355, + "step": 10232 + }, + { + "epoch": 1.0821700507614214, + "grad_norm": 7.875, + "learning_rate": 4.557987104742125e-06, + "loss": 2.4958, + "step": 10233 + }, + { + "epoch": 1.0822758037225042, + "grad_norm": 8.125, + "learning_rate": 4.557889867656248e-06, + "loss": 2.5748, + "step": 10234 + }, + { + "epoch": 1.082381556683587, + "grad_norm": 8.125, + "learning_rate": 4.557792620913536e-06, + "loss": 2.3058, + "step": 10235 + }, + { + "epoch": 1.08248730964467, + "grad_norm": 8.9375, + "learning_rate": 4.557695364514444e-06, + "loss": 2.7855, + "step": 10236 + }, + { + "epoch": 1.082593062605753, + "grad_norm": 8.0, + "learning_rate": 4.55759809845943e-06, + "loss": 2.6492, + "step": 10237 + }, + { + "epoch": 1.0826988155668358, + "grad_norm": 8.25, + "learning_rate": 4.557500822748949e-06, + "loss": 2.6084, + "step": 10238 + }, + { + "epoch": 1.0828045685279188, + "grad_norm": 7.78125, + "learning_rate": 4.557403537383458e-06, + "loss": 2.4451, + "step": 10239 + }, + { + "epoch": 1.0829103214890017, + "grad_norm": 8.4375, + "learning_rate": 4.557306242363414e-06, + "loss": 2.5261, + "step": 10240 + }, + { + "epoch": 1.0830160744500845, + "grad_norm": 8.875, + "learning_rate": 4.557208937689274e-06, + "loss": 2.4616, + "step": 10241 + }, + { + "epoch": 1.0831218274111676, + "grad_norm": 8.0625, + "learning_rate": 4.557111623361493e-06, + "loss": 2.6401, + "step": 10242 + }, + { + "epoch": 1.0832275803722504, + "grad_norm": 8.375, + "learning_rate": 4.557014299380528e-06, + "loss": 2.5903, + "step": 10243 + }, + { + "epoch": 1.0833333333333333, + "grad_norm": 8.0, + "learning_rate": 4.556916965746837e-06, + "loss": 2.5979, + "step": 10244 + }, + { + "epoch": 1.0834390862944163, + "grad_norm": 8.125, + "learning_rate": 4.556819622460875e-06, + "loss": 2.5273, + "step": 10245 + }, + { + "epoch": 1.0835448392554992, + "grad_norm": 8.0, + "learning_rate": 4.556722269523101e-06, + "loss": 2.3043, + "step": 10246 + }, + { + "epoch": 1.083650592216582, + "grad_norm": 7.96875, + "learning_rate": 4.55662490693397e-06, + "loss": 2.5926, + "step": 10247 + }, + { + "epoch": 1.083756345177665, + "grad_norm": 8.1875, + "learning_rate": 4.556527534693939e-06, + "loss": 2.3932, + "step": 10248 + }, + { + "epoch": 1.083862098138748, + "grad_norm": 8.375, + "learning_rate": 4.556430152803466e-06, + "loss": 2.6328, + "step": 10249 + }, + { + "epoch": 1.0839678510998307, + "grad_norm": 7.78125, + "learning_rate": 4.556332761263007e-06, + "loss": 2.5348, + "step": 10250 + }, + { + "epoch": 1.0840736040609138, + "grad_norm": 8.125, + "learning_rate": 4.556235360073019e-06, + "loss": 2.5368, + "step": 10251 + }, + { + "epoch": 1.0841793570219966, + "grad_norm": 8.125, + "learning_rate": 4.5561379492339605e-06, + "loss": 2.4049, + "step": 10252 + }, + { + "epoch": 1.0842851099830795, + "grad_norm": 8.1875, + "learning_rate": 4.5560405287462866e-06, + "loss": 2.5098, + "step": 10253 + }, + { + "epoch": 1.0843908629441625, + "grad_norm": 8.6875, + "learning_rate": 4.555943098610456e-06, + "loss": 2.6974, + "step": 10254 + }, + { + "epoch": 1.0844966159052454, + "grad_norm": 8.25, + "learning_rate": 4.555845658826925e-06, + "loss": 2.4403, + "step": 10255 + }, + { + "epoch": 1.0846023688663282, + "grad_norm": 8.0625, + "learning_rate": 4.5557482093961515e-06, + "loss": 2.3533, + "step": 10256 + }, + { + "epoch": 1.0847081218274113, + "grad_norm": 13.625, + "learning_rate": 4.555650750318592e-06, + "loss": 2.5077, + "step": 10257 + }, + { + "epoch": 1.0848138747884941, + "grad_norm": 7.96875, + "learning_rate": 4.555553281594704e-06, + "loss": 2.5621, + "step": 10258 + }, + { + "epoch": 1.084919627749577, + "grad_norm": 8.125, + "learning_rate": 4.555455803224945e-06, + "loss": 2.5851, + "step": 10259 + }, + { + "epoch": 1.0850253807106598, + "grad_norm": 7.90625, + "learning_rate": 4.555358315209774e-06, + "loss": 2.4226, + "step": 10260 + }, + { + "epoch": 1.0851311336717429, + "grad_norm": 8.9375, + "learning_rate": 4.5552608175496455e-06, + "loss": 2.4463, + "step": 10261 + }, + { + "epoch": 1.0852368866328257, + "grad_norm": 7.78125, + "learning_rate": 4.555163310245019e-06, + "loss": 2.6008, + "step": 10262 + }, + { + "epoch": 1.0853426395939085, + "grad_norm": 8.0625, + "learning_rate": 4.555065793296352e-06, + "loss": 2.4041, + "step": 10263 + }, + { + "epoch": 1.0854483925549916, + "grad_norm": 8.5, + "learning_rate": 4.5549682667041015e-06, + "loss": 2.875, + "step": 10264 + }, + { + "epoch": 1.0855541455160744, + "grad_norm": 7.9375, + "learning_rate": 4.554870730468725e-06, + "loss": 2.5133, + "step": 10265 + }, + { + "epoch": 1.0856598984771573, + "grad_norm": 9.0625, + "learning_rate": 4.55477318459068e-06, + "loss": 2.7042, + "step": 10266 + }, + { + "epoch": 1.0857656514382403, + "grad_norm": 8.625, + "learning_rate": 4.554675629070427e-06, + "loss": 2.4549, + "step": 10267 + }, + { + "epoch": 1.0858714043993232, + "grad_norm": 8.125, + "learning_rate": 4.55457806390842e-06, + "loss": 2.6061, + "step": 10268 + }, + { + "epoch": 1.085977157360406, + "grad_norm": 8.3125, + "learning_rate": 4.5544804891051185e-06, + "loss": 2.5013, + "step": 10269 + }, + { + "epoch": 1.086082910321489, + "grad_norm": 7.84375, + "learning_rate": 4.554382904660981e-06, + "loss": 2.5583, + "step": 10270 + }, + { + "epoch": 1.086188663282572, + "grad_norm": 8.125, + "learning_rate": 4.5542853105764636e-06, + "loss": 2.4872, + "step": 10271 + }, + { + "epoch": 1.0862944162436547, + "grad_norm": 8.125, + "learning_rate": 4.554187706852027e-06, + "loss": 2.5295, + "step": 10272 + }, + { + "epoch": 1.0864001692047378, + "grad_norm": 8.4375, + "learning_rate": 4.554090093488127e-06, + "loss": 2.5611, + "step": 10273 + }, + { + "epoch": 1.0865059221658206, + "grad_norm": 8.0625, + "learning_rate": 4.553992470485222e-06, + "loss": 2.6662, + "step": 10274 + }, + { + "epoch": 1.0866116751269035, + "grad_norm": 8.6875, + "learning_rate": 4.55389483784377e-06, + "loss": 2.5567, + "step": 10275 + }, + { + "epoch": 1.0867174280879865, + "grad_norm": 7.9375, + "learning_rate": 4.55379719556423e-06, + "loss": 2.418, + "step": 10276 + }, + { + "epoch": 1.0868231810490694, + "grad_norm": 7.875, + "learning_rate": 4.55369954364706e-06, + "loss": 2.5877, + "step": 10277 + }, + { + "epoch": 1.0869289340101522, + "grad_norm": 8.625, + "learning_rate": 4.553601882092718e-06, + "loss": 2.8257, + "step": 10278 + }, + { + "epoch": 1.0870346869712353, + "grad_norm": 8.1875, + "learning_rate": 4.553504210901661e-06, + "loss": 2.6928, + "step": 10279 + }, + { + "epoch": 1.0871404399323181, + "grad_norm": 8.125, + "learning_rate": 4.55340653007435e-06, + "loss": 2.4665, + "step": 10280 + }, + { + "epoch": 1.087246192893401, + "grad_norm": 8.0, + "learning_rate": 4.5533088396112415e-06, + "loss": 2.1117, + "step": 10281 + }, + { + "epoch": 1.087351945854484, + "grad_norm": 8.1875, + "learning_rate": 4.553211139512794e-06, + "loss": 2.5217, + "step": 10282 + }, + { + "epoch": 1.0874576988155669, + "grad_norm": 8.3125, + "learning_rate": 4.553113429779468e-06, + "loss": 2.6729, + "step": 10283 + }, + { + "epoch": 1.0875634517766497, + "grad_norm": 8.1875, + "learning_rate": 4.553015710411718e-06, + "loss": 2.5825, + "step": 10284 + }, + { + "epoch": 1.0876692047377325, + "grad_norm": 8.4375, + "learning_rate": 4.552917981410006e-06, + "loss": 2.3869, + "step": 10285 + }, + { + "epoch": 1.0877749576988156, + "grad_norm": 7.8125, + "learning_rate": 4.5528202427747906e-06, + "loss": 2.3754, + "step": 10286 + }, + { + "epoch": 1.0878807106598984, + "grad_norm": 8.3125, + "learning_rate": 4.5527224945065275e-06, + "loss": 2.525, + "step": 10287 + }, + { + "epoch": 1.0879864636209815, + "grad_norm": 8.125, + "learning_rate": 4.552624736605679e-06, + "loss": 2.5565, + "step": 10288 + }, + { + "epoch": 1.0880922165820643, + "grad_norm": 8.1875, + "learning_rate": 4.552526969072701e-06, + "loss": 2.7527, + "step": 10289 + }, + { + "epoch": 1.0881979695431472, + "grad_norm": 8.1875, + "learning_rate": 4.5524291919080536e-06, + "loss": 2.6067, + "step": 10290 + }, + { + "epoch": 1.08830372250423, + "grad_norm": 7.59375, + "learning_rate": 4.552331405112195e-06, + "loss": 2.4789, + "step": 10291 + }, + { + "epoch": 1.088409475465313, + "grad_norm": 8.625, + "learning_rate": 4.552233608685585e-06, + "loss": 2.9294, + "step": 10292 + }, + { + "epoch": 1.088515228426396, + "grad_norm": 8.4375, + "learning_rate": 4.552135802628682e-06, + "loss": 2.5422, + "step": 10293 + }, + { + "epoch": 1.0886209813874788, + "grad_norm": 8.5625, + "learning_rate": 4.552037986941945e-06, + "loss": 2.5445, + "step": 10294 + }, + { + "epoch": 1.0887267343485618, + "grad_norm": 7.9375, + "learning_rate": 4.551940161625833e-06, + "loss": 2.6221, + "step": 10295 + }, + { + "epoch": 1.0888324873096447, + "grad_norm": 8.1875, + "learning_rate": 4.551842326680805e-06, + "loss": 2.7456, + "step": 10296 + }, + { + "epoch": 1.0889382402707275, + "grad_norm": 9.0, + "learning_rate": 4.55174448210732e-06, + "loss": 2.743, + "step": 10297 + }, + { + "epoch": 1.0890439932318106, + "grad_norm": 8.0625, + "learning_rate": 4.551646627905838e-06, + "loss": 2.4516, + "step": 10298 + }, + { + "epoch": 1.0891497461928934, + "grad_norm": 8.625, + "learning_rate": 4.551548764076816e-06, + "loss": 2.4865, + "step": 10299 + }, + { + "epoch": 1.0892554991539762, + "grad_norm": 7.84375, + "learning_rate": 4.551450890620716e-06, + "loss": 2.3081, + "step": 10300 + }, + { + "epoch": 1.0893612521150593, + "grad_norm": 8.0, + "learning_rate": 4.551353007537995e-06, + "loss": 2.4722, + "step": 10301 + }, + { + "epoch": 1.0894670050761421, + "grad_norm": 8.125, + "learning_rate": 4.551255114829114e-06, + "loss": 2.5703, + "step": 10302 + }, + { + "epoch": 1.089572758037225, + "grad_norm": 8.375, + "learning_rate": 4.551157212494532e-06, + "loss": 2.6987, + "step": 10303 + }, + { + "epoch": 1.089678510998308, + "grad_norm": 8.0, + "learning_rate": 4.551059300534707e-06, + "loss": 2.5575, + "step": 10304 + }, + { + "epoch": 1.0897842639593909, + "grad_norm": 8.5, + "learning_rate": 4.5509613789501e-06, + "loss": 2.8411, + "step": 10305 + }, + { + "epoch": 1.0898900169204737, + "grad_norm": 8.0625, + "learning_rate": 4.550863447741171e-06, + "loss": 2.4759, + "step": 10306 + }, + { + "epoch": 1.0899957698815568, + "grad_norm": 8.0625, + "learning_rate": 4.550765506908378e-06, + "loss": 2.5289, + "step": 10307 + }, + { + "epoch": 1.0901015228426396, + "grad_norm": 8.125, + "learning_rate": 4.550667556452181e-06, + "loss": 2.4704, + "step": 10308 + }, + { + "epoch": 1.0902072758037225, + "grad_norm": 8.4375, + "learning_rate": 4.55056959637304e-06, + "loss": 2.6489, + "step": 10309 + }, + { + "epoch": 1.0903130287648055, + "grad_norm": 8.25, + "learning_rate": 4.550471626671414e-06, + "loss": 2.5098, + "step": 10310 + }, + { + "epoch": 1.0904187817258884, + "grad_norm": 8.9375, + "learning_rate": 4.550373647347763e-06, + "loss": 2.3235, + "step": 10311 + }, + { + "epoch": 1.0905245346869712, + "grad_norm": 8.125, + "learning_rate": 4.550275658402549e-06, + "loss": 2.5838, + "step": 10312 + }, + { + "epoch": 1.0906302876480543, + "grad_norm": 8.375, + "learning_rate": 4.550177659836228e-06, + "loss": 2.7985, + "step": 10313 + }, + { + "epoch": 1.090736040609137, + "grad_norm": 7.6875, + "learning_rate": 4.5500796516492624e-06, + "loss": 2.5246, + "step": 10314 + }, + { + "epoch": 1.09084179357022, + "grad_norm": 8.8125, + "learning_rate": 4.549981633842112e-06, + "loss": 2.4709, + "step": 10315 + }, + { + "epoch": 1.0909475465313028, + "grad_norm": 8.0, + "learning_rate": 4.549883606415235e-06, + "loss": 2.6178, + "step": 10316 + }, + { + "epoch": 1.0910532994923858, + "grad_norm": 8.75, + "learning_rate": 4.549785569369093e-06, + "loss": 2.5405, + "step": 10317 + }, + { + "epoch": 1.0911590524534687, + "grad_norm": 7.96875, + "learning_rate": 4.549687522704146e-06, + "loss": 2.1381, + "step": 10318 + }, + { + "epoch": 1.0912648054145515, + "grad_norm": 8.0625, + "learning_rate": 4.549589466420854e-06, + "loss": 2.5473, + "step": 10319 + }, + { + "epoch": 1.0913705583756346, + "grad_norm": 8.1875, + "learning_rate": 4.549491400519676e-06, + "loss": 2.7136, + "step": 10320 + }, + { + "epoch": 1.0914763113367174, + "grad_norm": 8.0, + "learning_rate": 4.549393325001074e-06, + "loss": 2.4605, + "step": 10321 + }, + { + "epoch": 1.0915820642978002, + "grad_norm": 8.25, + "learning_rate": 4.549295239865507e-06, + "loss": 2.4186, + "step": 10322 + }, + { + "epoch": 1.0916878172588833, + "grad_norm": 7.9375, + "learning_rate": 4.549197145113435e-06, + "loss": 2.6, + "step": 10323 + }, + { + "epoch": 1.0917935702199661, + "grad_norm": 7.6875, + "learning_rate": 4.54909904074532e-06, + "loss": 2.4753, + "step": 10324 + }, + { + "epoch": 1.091899323181049, + "grad_norm": 7.90625, + "learning_rate": 4.54900092676162e-06, + "loss": 2.4498, + "step": 10325 + }, + { + "epoch": 1.092005076142132, + "grad_norm": 8.1875, + "learning_rate": 4.548902803162798e-06, + "loss": 2.4714, + "step": 10326 + }, + { + "epoch": 1.0921108291032149, + "grad_norm": 8.125, + "learning_rate": 4.5488046699493125e-06, + "loss": 2.8138, + "step": 10327 + }, + { + "epoch": 1.0922165820642977, + "grad_norm": 7.875, + "learning_rate": 4.5487065271216244e-06, + "loss": 2.5262, + "step": 10328 + }, + { + "epoch": 1.0923223350253808, + "grad_norm": 8.0, + "learning_rate": 4.548608374680195e-06, + "loss": 2.489, + "step": 10329 + }, + { + "epoch": 1.0924280879864636, + "grad_norm": 7.625, + "learning_rate": 4.548510212625483e-06, + "loss": 2.5205, + "step": 10330 + }, + { + "epoch": 1.0925338409475465, + "grad_norm": 7.84375, + "learning_rate": 4.548412040957951e-06, + "loss": 2.4485, + "step": 10331 + }, + { + "epoch": 1.0926395939086295, + "grad_norm": 8.4375, + "learning_rate": 4.54831385967806e-06, + "loss": 2.7241, + "step": 10332 + }, + { + "epoch": 1.0927453468697124, + "grad_norm": 8.0, + "learning_rate": 4.548215668786269e-06, + "loss": 2.6839, + "step": 10333 + }, + { + "epoch": 1.0928510998307952, + "grad_norm": 8.625, + "learning_rate": 4.548117468283039e-06, + "loss": 2.5092, + "step": 10334 + }, + { + "epoch": 1.0929568527918783, + "grad_norm": 8.1875, + "learning_rate": 4.548019258168833e-06, + "loss": 2.6539, + "step": 10335 + }, + { + "epoch": 1.093062605752961, + "grad_norm": 8.0, + "learning_rate": 4.547921038444109e-06, + "loss": 2.4, + "step": 10336 + }, + { + "epoch": 1.093168358714044, + "grad_norm": 8.0, + "learning_rate": 4.54782280910933e-06, + "loss": 2.75, + "step": 10337 + }, + { + "epoch": 1.093274111675127, + "grad_norm": 8.0, + "learning_rate": 4.547724570164955e-06, + "loss": 2.589, + "step": 10338 + }, + { + "epoch": 1.0933798646362098, + "grad_norm": 8.5, + "learning_rate": 4.547626321611447e-06, + "loss": 2.359, + "step": 10339 + }, + { + "epoch": 1.0934856175972927, + "grad_norm": 8.3125, + "learning_rate": 4.547528063449266e-06, + "loss": 2.4758, + "step": 10340 + }, + { + "epoch": 1.0935913705583755, + "grad_norm": 7.8125, + "learning_rate": 4.547429795678873e-06, + "loss": 2.5951, + "step": 10341 + }, + { + "epoch": 1.0936971235194586, + "grad_norm": 8.75, + "learning_rate": 4.547331518300729e-06, + "loss": 2.5775, + "step": 10342 + }, + { + "epoch": 1.0938028764805414, + "grad_norm": 8.0625, + "learning_rate": 4.5472332313152955e-06, + "loss": 2.4713, + "step": 10343 + }, + { + "epoch": 1.0939086294416245, + "grad_norm": 8.4375, + "learning_rate": 4.547134934723034e-06, + "loss": 2.6949, + "step": 10344 + }, + { + "epoch": 1.0940143824027073, + "grad_norm": 7.71875, + "learning_rate": 4.547036628524406e-06, + "loss": 2.5178, + "step": 10345 + }, + { + "epoch": 1.0941201353637902, + "grad_norm": 7.875, + "learning_rate": 4.546938312719871e-06, + "loss": 2.5663, + "step": 10346 + }, + { + "epoch": 1.094225888324873, + "grad_norm": 8.5625, + "learning_rate": 4.5468399873098935e-06, + "loss": 2.5115, + "step": 10347 + }, + { + "epoch": 1.094331641285956, + "grad_norm": 8.125, + "learning_rate": 4.546741652294932e-06, + "loss": 2.3693, + "step": 10348 + }, + { + "epoch": 1.094437394247039, + "grad_norm": 7.9375, + "learning_rate": 4.546643307675449e-06, + "loss": 2.376, + "step": 10349 + }, + { + "epoch": 1.0945431472081217, + "grad_norm": 7.71875, + "learning_rate": 4.546544953451906e-06, + "loss": 2.4226, + "step": 10350 + }, + { + "epoch": 1.0946489001692048, + "grad_norm": 8.1875, + "learning_rate": 4.5464465896247644e-06, + "loss": 2.8367, + "step": 10351 + }, + { + "epoch": 1.0947546531302876, + "grad_norm": 7.75, + "learning_rate": 4.546348216194486e-06, + "loss": 2.4763, + "step": 10352 + }, + { + "epoch": 1.0948604060913705, + "grad_norm": 8.9375, + "learning_rate": 4.546249833161533e-06, + "loss": 2.5686, + "step": 10353 + }, + { + "epoch": 1.0949661590524535, + "grad_norm": 8.4375, + "learning_rate": 4.546151440526366e-06, + "loss": 2.5613, + "step": 10354 + }, + { + "epoch": 1.0950719120135364, + "grad_norm": 8.1875, + "learning_rate": 4.546053038289447e-06, + "loss": 2.5578, + "step": 10355 + }, + { + "epoch": 1.0951776649746192, + "grad_norm": 8.125, + "learning_rate": 4.545954626451238e-06, + "loss": 2.3879, + "step": 10356 + }, + { + "epoch": 1.0952834179357023, + "grad_norm": 8.25, + "learning_rate": 4.545856205012201e-06, + "loss": 2.5639, + "step": 10357 + }, + { + "epoch": 1.0953891708967851, + "grad_norm": 8.1875, + "learning_rate": 4.545757773972797e-06, + "loss": 2.6496, + "step": 10358 + }, + { + "epoch": 1.095494923857868, + "grad_norm": 7.59375, + "learning_rate": 4.545659333333489e-06, + "loss": 2.446, + "step": 10359 + }, + { + "epoch": 1.095600676818951, + "grad_norm": 8.6875, + "learning_rate": 4.545560883094737e-06, + "loss": 2.7393, + "step": 10360 + }, + { + "epoch": 1.0957064297800339, + "grad_norm": 8.375, + "learning_rate": 4.545462423257006e-06, + "loss": 2.374, + "step": 10361 + }, + { + "epoch": 1.0958121827411167, + "grad_norm": 8.0, + "learning_rate": 4.545363953820756e-06, + "loss": 2.4402, + "step": 10362 + }, + { + "epoch": 1.0959179357021998, + "grad_norm": 8.375, + "learning_rate": 4.545265474786449e-06, + "loss": 2.4852, + "step": 10363 + }, + { + "epoch": 1.0960236886632826, + "grad_norm": 8.1875, + "learning_rate": 4.545166986154548e-06, + "loss": 2.5423, + "step": 10364 + }, + { + "epoch": 1.0961294416243654, + "grad_norm": 8.125, + "learning_rate": 4.545068487925515e-06, + "loss": 2.5931, + "step": 10365 + }, + { + "epoch": 1.0962351945854485, + "grad_norm": 7.4375, + "learning_rate": 4.544969980099812e-06, + "loss": 2.4061, + "step": 10366 + }, + { + "epoch": 1.0963409475465313, + "grad_norm": 8.0, + "learning_rate": 4.544871462677901e-06, + "loss": 2.5987, + "step": 10367 + }, + { + "epoch": 1.0964467005076142, + "grad_norm": 8.4375, + "learning_rate": 4.544772935660244e-06, + "loss": 2.5342, + "step": 10368 + }, + { + "epoch": 1.0965524534686972, + "grad_norm": 8.1875, + "learning_rate": 4.544674399047305e-06, + "loss": 2.5642, + "step": 10369 + }, + { + "epoch": 1.09665820642978, + "grad_norm": 7.78125, + "learning_rate": 4.544575852839545e-06, + "loss": 2.5249, + "step": 10370 + }, + { + "epoch": 1.096763959390863, + "grad_norm": 7.90625, + "learning_rate": 4.5444772970374266e-06, + "loss": 2.6762, + "step": 10371 + }, + { + "epoch": 1.0968697123519457, + "grad_norm": 8.9375, + "learning_rate": 4.544378731641412e-06, + "loss": 2.5353, + "step": 10372 + }, + { + "epoch": 1.0969754653130288, + "grad_norm": 8.1875, + "learning_rate": 4.544280156651965e-06, + "loss": 2.4218, + "step": 10373 + }, + { + "epoch": 1.0970812182741116, + "grad_norm": 8.375, + "learning_rate": 4.544181572069546e-06, + "loss": 2.508, + "step": 10374 + }, + { + "epoch": 1.0971869712351945, + "grad_norm": 8.0625, + "learning_rate": 4.544082977894619e-06, + "loss": 2.5198, + "step": 10375 + }, + { + "epoch": 1.0972927241962775, + "grad_norm": 8.8125, + "learning_rate": 4.543984374127648e-06, + "loss": 2.5609, + "step": 10376 + }, + { + "epoch": 1.0973984771573604, + "grad_norm": 8.875, + "learning_rate": 4.543885760769093e-06, + "loss": 2.3194, + "step": 10377 + }, + { + "epoch": 1.0975042301184432, + "grad_norm": 8.375, + "learning_rate": 4.543787137819419e-06, + "loss": 2.7489, + "step": 10378 + }, + { + "epoch": 1.0976099830795263, + "grad_norm": 7.78125, + "learning_rate": 4.5436885052790875e-06, + "loss": 2.6415, + "step": 10379 + }, + { + "epoch": 1.0977157360406091, + "grad_norm": 8.1875, + "learning_rate": 4.543589863148562e-06, + "loss": 2.6403, + "step": 10380 + }, + { + "epoch": 1.097821489001692, + "grad_norm": 7.84375, + "learning_rate": 4.543491211428304e-06, + "loss": 2.6518, + "step": 10381 + }, + { + "epoch": 1.097927241962775, + "grad_norm": 7.875, + "learning_rate": 4.5433925501187784e-06, + "loss": 2.516, + "step": 10382 + }, + { + "epoch": 1.0980329949238579, + "grad_norm": 8.1875, + "learning_rate": 4.543293879220448e-06, + "loss": 2.6219, + "step": 10383 + }, + { + "epoch": 1.0981387478849407, + "grad_norm": 7.8125, + "learning_rate": 4.543195198733773e-06, + "loss": 2.6612, + "step": 10384 + }, + { + "epoch": 1.0982445008460238, + "grad_norm": 7.625, + "learning_rate": 4.54309650865922e-06, + "loss": 2.4924, + "step": 10385 + }, + { + "epoch": 1.0983502538071066, + "grad_norm": 8.3125, + "learning_rate": 4.54299780899725e-06, + "loss": 2.7076, + "step": 10386 + }, + { + "epoch": 1.0984560067681894, + "grad_norm": 8.3125, + "learning_rate": 4.542899099748328e-06, + "loss": 2.5429, + "step": 10387 + }, + { + "epoch": 1.0985617597292725, + "grad_norm": 8.25, + "learning_rate": 4.542800380912915e-06, + "loss": 2.4932, + "step": 10388 + }, + { + "epoch": 1.0986675126903553, + "grad_norm": 8.5, + "learning_rate": 4.542701652491476e-06, + "loss": 2.5528, + "step": 10389 + }, + { + "epoch": 1.0987732656514382, + "grad_norm": 8.0625, + "learning_rate": 4.542602914484473e-06, + "loss": 2.4747, + "step": 10390 + }, + { + "epoch": 1.0988790186125212, + "grad_norm": 8.0625, + "learning_rate": 4.54250416689237e-06, + "loss": 2.5874, + "step": 10391 + }, + { + "epoch": 1.098984771573604, + "grad_norm": 8.25, + "learning_rate": 4.54240540971563e-06, + "loss": 2.6135, + "step": 10392 + }, + { + "epoch": 1.099090524534687, + "grad_norm": 7.84375, + "learning_rate": 4.542306642954718e-06, + "loss": 2.5437, + "step": 10393 + }, + { + "epoch": 1.09919627749577, + "grad_norm": 8.0625, + "learning_rate": 4.542207866610095e-06, + "loss": 2.6122, + "step": 10394 + }, + { + "epoch": 1.0993020304568528, + "grad_norm": 9.3125, + "learning_rate": 4.542109080682226e-06, + "loss": 2.7913, + "step": 10395 + }, + { + "epoch": 1.0994077834179357, + "grad_norm": 8.0625, + "learning_rate": 4.542010285171574e-06, + "loss": 2.5382, + "step": 10396 + }, + { + "epoch": 1.0995135363790185, + "grad_norm": 9.125, + "learning_rate": 4.541911480078603e-06, + "loss": 2.5077, + "step": 10397 + }, + { + "epoch": 1.0996192893401016, + "grad_norm": 7.78125, + "learning_rate": 4.541812665403776e-06, + "loss": 2.6335, + "step": 10398 + }, + { + "epoch": 1.0997250423011844, + "grad_norm": 8.3125, + "learning_rate": 4.541713841147559e-06, + "loss": 2.7621, + "step": 10399 + }, + { + "epoch": 1.0998307952622675, + "grad_norm": 8.375, + "learning_rate": 4.541615007310413e-06, + "loss": 2.7728, + "step": 10400 + }, + { + "epoch": 1.0999365482233503, + "grad_norm": 8.8125, + "learning_rate": 4.541516163892802e-06, + "loss": 2.5305, + "step": 10401 + }, + { + "epoch": 1.1000423011844331, + "grad_norm": 8.25, + "learning_rate": 4.541417310895192e-06, + "loss": 2.5608, + "step": 10402 + }, + { + "epoch": 1.100148054145516, + "grad_norm": 9.5625, + "learning_rate": 4.541318448318045e-06, + "loss": 2.5883, + "step": 10403 + }, + { + "epoch": 1.100253807106599, + "grad_norm": 8.375, + "learning_rate": 4.541219576161825e-06, + "loss": 2.3253, + "step": 10404 + }, + { + "epoch": 1.1003595600676819, + "grad_norm": 8.1875, + "learning_rate": 4.541120694426996e-06, + "loss": 2.6165, + "step": 10405 + }, + { + "epoch": 1.1004653130287647, + "grad_norm": 8.375, + "learning_rate": 4.541021803114023e-06, + "loss": 2.5281, + "step": 10406 + }, + { + "epoch": 1.1005710659898478, + "grad_norm": 8.25, + "learning_rate": 4.540922902223369e-06, + "loss": 2.5512, + "step": 10407 + }, + { + "epoch": 1.1006768189509306, + "grad_norm": 7.9375, + "learning_rate": 4.540823991755499e-06, + "loss": 2.6001, + "step": 10408 + }, + { + "epoch": 1.1007825719120135, + "grad_norm": 8.25, + "learning_rate": 4.540725071710876e-06, + "loss": 2.5879, + "step": 10409 + }, + { + "epoch": 1.1008883248730965, + "grad_norm": 8.0, + "learning_rate": 4.540626142089966e-06, + "loss": 2.4898, + "step": 10410 + }, + { + "epoch": 1.1009940778341794, + "grad_norm": 8.375, + "learning_rate": 4.540527202893231e-06, + "loss": 2.6458, + "step": 10411 + }, + { + "epoch": 1.1010998307952622, + "grad_norm": 7.9375, + "learning_rate": 4.5404282541211375e-06, + "loss": 2.5408, + "step": 10412 + }, + { + "epoch": 1.1012055837563453, + "grad_norm": 8.25, + "learning_rate": 4.540329295774148e-06, + "loss": 2.5593, + "step": 10413 + }, + { + "epoch": 1.101311336717428, + "grad_norm": 7.8125, + "learning_rate": 4.540230327852727e-06, + "loss": 2.5346, + "step": 10414 + }, + { + "epoch": 1.101417089678511, + "grad_norm": 8.0625, + "learning_rate": 4.540131350357341e-06, + "loss": 2.5901, + "step": 10415 + }, + { + "epoch": 1.101522842639594, + "grad_norm": 8.625, + "learning_rate": 4.540032363288452e-06, + "loss": 2.6996, + "step": 10416 + }, + { + "epoch": 1.1016285956006768, + "grad_norm": 7.65625, + "learning_rate": 4.5399333666465255e-06, + "loss": 2.5728, + "step": 10417 + }, + { + "epoch": 1.1017343485617597, + "grad_norm": 7.96875, + "learning_rate": 4.539834360432026e-06, + "loss": 2.5085, + "step": 10418 + }, + { + "epoch": 1.1018401015228427, + "grad_norm": 8.875, + "learning_rate": 4.539735344645418e-06, + "loss": 2.6425, + "step": 10419 + }, + { + "epoch": 1.1019458544839256, + "grad_norm": 7.96875, + "learning_rate": 4.539636319287168e-06, + "loss": 2.4539, + "step": 10420 + }, + { + "epoch": 1.1020516074450084, + "grad_norm": 7.96875, + "learning_rate": 4.539537284357736e-06, + "loss": 2.4198, + "step": 10421 + }, + { + "epoch": 1.1021573604060915, + "grad_norm": 8.0625, + "learning_rate": 4.539438239857591e-06, + "loss": 2.5442, + "step": 10422 + }, + { + "epoch": 1.1022631133671743, + "grad_norm": 9.875, + "learning_rate": 4.539339185787197e-06, + "loss": 2.7071, + "step": 10423 + }, + { + "epoch": 1.1023688663282571, + "grad_norm": 7.8125, + "learning_rate": 4.539240122147017e-06, + "loss": 2.5056, + "step": 10424 + }, + { + "epoch": 1.1024746192893402, + "grad_norm": 8.5625, + "learning_rate": 4.539141048937519e-06, + "loss": 2.3784, + "step": 10425 + }, + { + "epoch": 1.102580372250423, + "grad_norm": 8.1875, + "learning_rate": 4.539041966159164e-06, + "loss": 2.546, + "step": 10426 + }, + { + "epoch": 1.1026861252115059, + "grad_norm": 8.8125, + "learning_rate": 4.5389428738124205e-06, + "loss": 2.8851, + "step": 10427 + }, + { + "epoch": 1.1027918781725887, + "grad_norm": 7.84375, + "learning_rate": 4.538843771897751e-06, + "loss": 2.3459, + "step": 10428 + }, + { + "epoch": 1.1028976311336718, + "grad_norm": 8.125, + "learning_rate": 4.538744660415623e-06, + "loss": 2.7161, + "step": 10429 + }, + { + "epoch": 1.1030033840947546, + "grad_norm": 8.5, + "learning_rate": 4.538645539366499e-06, + "loss": 2.2249, + "step": 10430 + }, + { + "epoch": 1.1031091370558375, + "grad_norm": 8.125, + "learning_rate": 4.538546408750844e-06, + "loss": 2.5973, + "step": 10431 + }, + { + "epoch": 1.1032148900169205, + "grad_norm": 9.5625, + "learning_rate": 4.538447268569126e-06, + "loss": 2.7565, + "step": 10432 + }, + { + "epoch": 1.1033206429780034, + "grad_norm": 8.0625, + "learning_rate": 4.53834811882181e-06, + "loss": 2.3716, + "step": 10433 + }, + { + "epoch": 1.1034263959390862, + "grad_norm": 8.0, + "learning_rate": 4.538248959509357e-06, + "loss": 2.5182, + "step": 10434 + }, + { + "epoch": 1.1035321489001693, + "grad_norm": 8.125, + "learning_rate": 4.538149790632237e-06, + "loss": 2.5911, + "step": 10435 + }, + { + "epoch": 1.103637901861252, + "grad_norm": 8.125, + "learning_rate": 4.5380506121909125e-06, + "loss": 2.5277, + "step": 10436 + }, + { + "epoch": 1.103743654822335, + "grad_norm": 8.8125, + "learning_rate": 4.537951424185851e-06, + "loss": 2.6828, + "step": 10437 + }, + { + "epoch": 1.103849407783418, + "grad_norm": 8.5625, + "learning_rate": 4.537852226617517e-06, + "loss": 2.6534, + "step": 10438 + }, + { + "epoch": 1.1039551607445008, + "grad_norm": 8.6875, + "learning_rate": 4.537753019486376e-06, + "loss": 2.478, + "step": 10439 + }, + { + "epoch": 1.1040609137055837, + "grad_norm": 8.3125, + "learning_rate": 4.537653802792893e-06, + "loss": 2.6547, + "step": 10440 + }, + { + "epoch": 1.1041666666666667, + "grad_norm": 7.875, + "learning_rate": 4.537554576537534e-06, + "loss": 2.6165, + "step": 10441 + }, + { + "epoch": 1.1042724196277496, + "grad_norm": 7.84375, + "learning_rate": 4.537455340720765e-06, + "loss": 2.3967, + "step": 10442 + }, + { + "epoch": 1.1043781725888324, + "grad_norm": 7.78125, + "learning_rate": 4.537356095343051e-06, + "loss": 2.3256, + "step": 10443 + }, + { + "epoch": 1.1044839255499155, + "grad_norm": 8.8125, + "learning_rate": 4.537256840404859e-06, + "loss": 2.5941, + "step": 10444 + }, + { + "epoch": 1.1045896785109983, + "grad_norm": 7.5625, + "learning_rate": 4.537157575906653e-06, + "loss": 2.4366, + "step": 10445 + }, + { + "epoch": 1.1046954314720812, + "grad_norm": 8.5, + "learning_rate": 4.5370583018489e-06, + "loss": 2.6898, + "step": 10446 + }, + { + "epoch": 1.1048011844331642, + "grad_norm": 8.25, + "learning_rate": 4.536959018232065e-06, + "loss": 2.3888, + "step": 10447 + }, + { + "epoch": 1.104906937394247, + "grad_norm": 8.0625, + "learning_rate": 4.536859725056615e-06, + "loss": 2.5607, + "step": 10448 + }, + { + "epoch": 1.10501269035533, + "grad_norm": 8.5, + "learning_rate": 4.536760422323015e-06, + "loss": 2.6249, + "step": 10449 + }, + { + "epoch": 1.105118443316413, + "grad_norm": 8.125, + "learning_rate": 4.536661110031732e-06, + "loss": 2.6655, + "step": 10450 + }, + { + "epoch": 1.1052241962774958, + "grad_norm": 8.5, + "learning_rate": 4.536561788183232e-06, + "loss": 2.4189, + "step": 10451 + }, + { + "epoch": 1.1053299492385786, + "grad_norm": 8.625, + "learning_rate": 4.5364624567779784e-06, + "loss": 2.7062, + "step": 10452 + }, + { + "epoch": 1.1054357021996615, + "grad_norm": 8.6875, + "learning_rate": 4.53636311581644e-06, + "loss": 2.682, + "step": 10453 + }, + { + "epoch": 1.1055414551607445, + "grad_norm": 9.875, + "learning_rate": 4.536263765299083e-06, + "loss": 2.638, + "step": 10454 + }, + { + "epoch": 1.1056472081218274, + "grad_norm": 8.0625, + "learning_rate": 4.536164405226373e-06, + "loss": 2.5813, + "step": 10455 + }, + { + "epoch": 1.1057529610829104, + "grad_norm": 7.78125, + "learning_rate": 4.536065035598776e-06, + "loss": 2.5438, + "step": 10456 + }, + { + "epoch": 1.1058587140439933, + "grad_norm": 8.3125, + "learning_rate": 4.535965656416757e-06, + "loss": 2.4732, + "step": 10457 + }, + { + "epoch": 1.1059644670050761, + "grad_norm": 7.78125, + "learning_rate": 4.535866267680786e-06, + "loss": 2.5525, + "step": 10458 + }, + { + "epoch": 1.106070219966159, + "grad_norm": 8.375, + "learning_rate": 4.535766869391327e-06, + "loss": 2.6591, + "step": 10459 + }, + { + "epoch": 1.106175972927242, + "grad_norm": 8.3125, + "learning_rate": 4.535667461548846e-06, + "loss": 2.2752, + "step": 10460 + }, + { + "epoch": 1.1062817258883249, + "grad_norm": 8.0, + "learning_rate": 4.53556804415381e-06, + "loss": 2.5238, + "step": 10461 + }, + { + "epoch": 1.1063874788494077, + "grad_norm": 9.0, + "learning_rate": 4.535468617206686e-06, + "loss": 2.902, + "step": 10462 + }, + { + "epoch": 1.1064932318104908, + "grad_norm": 8.375, + "learning_rate": 4.53536918070794e-06, + "loss": 2.6224, + "step": 10463 + }, + { + "epoch": 1.1065989847715736, + "grad_norm": 11.75, + "learning_rate": 4.53526973465804e-06, + "loss": 2.3078, + "step": 10464 + }, + { + "epoch": 1.1067047377326564, + "grad_norm": 7.75, + "learning_rate": 4.5351702790574505e-06, + "loss": 2.5993, + "step": 10465 + }, + { + "epoch": 1.1068104906937395, + "grad_norm": 8.4375, + "learning_rate": 4.535070813906639e-06, + "loss": 2.5421, + "step": 10466 + }, + { + "epoch": 1.1069162436548223, + "grad_norm": 8.0625, + "learning_rate": 4.534971339206073e-06, + "loss": 2.3953, + "step": 10467 + }, + { + "epoch": 1.1070219966159052, + "grad_norm": 8.8125, + "learning_rate": 4.534871854956219e-06, + "loss": 2.3484, + "step": 10468 + }, + { + "epoch": 1.1071277495769882, + "grad_norm": 7.9375, + "learning_rate": 4.534772361157543e-06, + "loss": 2.3219, + "step": 10469 + }, + { + "epoch": 1.107233502538071, + "grad_norm": 8.375, + "learning_rate": 4.534672857810512e-06, + "loss": 2.2495, + "step": 10470 + }, + { + "epoch": 1.107339255499154, + "grad_norm": 7.65625, + "learning_rate": 4.534573344915595e-06, + "loss": 2.4539, + "step": 10471 + }, + { + "epoch": 1.107445008460237, + "grad_norm": 8.0, + "learning_rate": 4.5344738224732564e-06, + "loss": 2.4824, + "step": 10472 + }, + { + "epoch": 1.1075507614213198, + "grad_norm": 8.4375, + "learning_rate": 4.534374290483965e-06, + "loss": 2.6706, + "step": 10473 + }, + { + "epoch": 1.1076565143824026, + "grad_norm": 8.375, + "learning_rate": 4.534274748948187e-06, + "loss": 2.6805, + "step": 10474 + }, + { + "epoch": 1.1077622673434857, + "grad_norm": 9.0, + "learning_rate": 4.534175197866389e-06, + "loss": 2.4112, + "step": 10475 + }, + { + "epoch": 1.1078680203045685, + "grad_norm": 8.4375, + "learning_rate": 4.534075637239039e-06, + "loss": 2.5653, + "step": 10476 + }, + { + "epoch": 1.1079737732656514, + "grad_norm": 8.0625, + "learning_rate": 4.533976067066604e-06, + "loss": 2.5367, + "step": 10477 + }, + { + "epoch": 1.1080795262267344, + "grad_norm": 8.75, + "learning_rate": 4.53387648734955e-06, + "loss": 2.829, + "step": 10478 + }, + { + "epoch": 1.1081852791878173, + "grad_norm": 8.3125, + "learning_rate": 4.533776898088347e-06, + "loss": 2.5712, + "step": 10479 + }, + { + "epoch": 1.1082910321489001, + "grad_norm": 8.5, + "learning_rate": 4.53367729928346e-06, + "loss": 2.6729, + "step": 10480 + }, + { + "epoch": 1.1083967851099832, + "grad_norm": 7.875, + "learning_rate": 4.533577690935358e-06, + "loss": 2.5035, + "step": 10481 + }, + { + "epoch": 1.108502538071066, + "grad_norm": 7.78125, + "learning_rate": 4.533478073044507e-06, + "loss": 2.3624, + "step": 10482 + }, + { + "epoch": 1.1086082910321489, + "grad_norm": 8.0, + "learning_rate": 4.533378445611375e-06, + "loss": 2.5511, + "step": 10483 + }, + { + "epoch": 1.1087140439932317, + "grad_norm": 7.96875, + "learning_rate": 4.53327880863643e-06, + "loss": 2.3023, + "step": 10484 + }, + { + "epoch": 1.1088197969543148, + "grad_norm": 8.1875, + "learning_rate": 4.5331791621201386e-06, + "loss": 2.5567, + "step": 10485 + }, + { + "epoch": 1.1089255499153976, + "grad_norm": 8.9375, + "learning_rate": 4.533079506062969e-06, + "loss": 2.5243, + "step": 10486 + }, + { + "epoch": 1.1090313028764804, + "grad_norm": 8.125, + "learning_rate": 4.532979840465389e-06, + "loss": 2.6201, + "step": 10487 + }, + { + "epoch": 1.1091370558375635, + "grad_norm": 7.75, + "learning_rate": 4.532880165327866e-06, + "loss": 2.5322, + "step": 10488 + }, + { + "epoch": 1.1092428087986463, + "grad_norm": 8.5625, + "learning_rate": 4.532780480650868e-06, + "loss": 2.6466, + "step": 10489 + }, + { + "epoch": 1.1093485617597292, + "grad_norm": 7.96875, + "learning_rate": 4.532680786434862e-06, + "loss": 2.398, + "step": 10490 + }, + { + "epoch": 1.1094543147208122, + "grad_norm": 7.875, + "learning_rate": 4.532581082680318e-06, + "loss": 2.3679, + "step": 10491 + }, + { + "epoch": 1.109560067681895, + "grad_norm": 9.25, + "learning_rate": 4.5324813693877e-06, + "loss": 2.5611, + "step": 10492 + }, + { + "epoch": 1.109665820642978, + "grad_norm": 7.71875, + "learning_rate": 4.5323816465574795e-06, + "loss": 2.5033, + "step": 10493 + }, + { + "epoch": 1.109771573604061, + "grad_norm": 8.0625, + "learning_rate": 4.532281914190123e-06, + "loss": 2.5998, + "step": 10494 + }, + { + "epoch": 1.1098773265651438, + "grad_norm": 8.25, + "learning_rate": 4.532182172286098e-06, + "loss": 2.6823, + "step": 10495 + }, + { + "epoch": 1.1099830795262267, + "grad_norm": 8.5, + "learning_rate": 4.532082420845875e-06, + "loss": 2.6436, + "step": 10496 + }, + { + "epoch": 1.1100888324873097, + "grad_norm": 8.625, + "learning_rate": 4.531982659869918e-06, + "loss": 2.7405, + "step": 10497 + }, + { + "epoch": 1.1101945854483926, + "grad_norm": 7.53125, + "learning_rate": 4.531882889358699e-06, + "loss": 2.4362, + "step": 10498 + }, + { + "epoch": 1.1103003384094754, + "grad_norm": 7.9375, + "learning_rate": 4.531783109312684e-06, + "loss": 2.3525, + "step": 10499 + }, + { + "epoch": 1.1104060913705585, + "grad_norm": 8.5, + "learning_rate": 4.531683319732342e-06, + "loss": 2.6275, + "step": 10500 + }, + { + "epoch": 1.1105118443316413, + "grad_norm": 8.0625, + "learning_rate": 4.531583520618141e-06, + "loss": 2.6804, + "step": 10501 + }, + { + "epoch": 1.1106175972927241, + "grad_norm": 8.1875, + "learning_rate": 4.5314837119705494e-06, + "loss": 2.6389, + "step": 10502 + }, + { + "epoch": 1.1107233502538072, + "grad_norm": 7.84375, + "learning_rate": 4.5313838937900355e-06, + "loss": 2.5109, + "step": 10503 + }, + { + "epoch": 1.11082910321489, + "grad_norm": 8.1875, + "learning_rate": 4.531284066077069e-06, + "loss": 2.6607, + "step": 10504 + }, + { + "epoch": 1.1109348561759729, + "grad_norm": 8.1875, + "learning_rate": 4.531184228832115e-06, + "loss": 2.641, + "step": 10505 + }, + { + "epoch": 1.111040609137056, + "grad_norm": 8.125, + "learning_rate": 4.531084382055646e-06, + "loss": 2.5602, + "step": 10506 + }, + { + "epoch": 1.1111463620981388, + "grad_norm": 8.4375, + "learning_rate": 4.5309845257481274e-06, + "loss": 2.628, + "step": 10507 + }, + { + "epoch": 1.1112521150592216, + "grad_norm": 8.75, + "learning_rate": 4.53088465991003e-06, + "loss": 2.6026, + "step": 10508 + }, + { + "epoch": 1.1113578680203045, + "grad_norm": 7.9375, + "learning_rate": 4.5307847845418205e-06, + "loss": 2.4018, + "step": 10509 + }, + { + "epoch": 1.1114636209813875, + "grad_norm": 9.3125, + "learning_rate": 4.5306848996439696e-06, + "loss": 2.6574, + "step": 10510 + }, + { + "epoch": 1.1115693739424704, + "grad_norm": 8.25, + "learning_rate": 4.530585005216945e-06, + "loss": 2.6269, + "step": 10511 + }, + { + "epoch": 1.1116751269035534, + "grad_norm": 9.125, + "learning_rate": 4.530485101261215e-06, + "loss": 2.376, + "step": 10512 + }, + { + "epoch": 1.1117808798646363, + "grad_norm": 8.3125, + "learning_rate": 4.530385187777249e-06, + "loss": 2.6311, + "step": 10513 + }, + { + "epoch": 1.111886632825719, + "grad_norm": 7.90625, + "learning_rate": 4.530285264765515e-06, + "loss": 2.5357, + "step": 10514 + }, + { + "epoch": 1.111992385786802, + "grad_norm": 7.875, + "learning_rate": 4.530185332226483e-06, + "loss": 2.5647, + "step": 10515 + }, + { + "epoch": 1.112098138747885, + "grad_norm": 8.5625, + "learning_rate": 4.530085390160623e-06, + "loss": 2.6355, + "step": 10516 + }, + { + "epoch": 1.1122038917089678, + "grad_norm": 8.0, + "learning_rate": 4.5299854385684e-06, + "loss": 2.4163, + "step": 10517 + }, + { + "epoch": 1.1123096446700507, + "grad_norm": 9.5, + "learning_rate": 4.529885477450288e-06, + "loss": 2.7909, + "step": 10518 + }, + { + "epoch": 1.1124153976311337, + "grad_norm": 8.75, + "learning_rate": 4.529785506806752e-06, + "loss": 2.6466, + "step": 10519 + }, + { + "epoch": 1.1125211505922166, + "grad_norm": 8.3125, + "learning_rate": 4.529685526638263e-06, + "loss": 2.7564, + "step": 10520 + }, + { + "epoch": 1.1126269035532994, + "grad_norm": 7.90625, + "learning_rate": 4.52958553694529e-06, + "loss": 2.5989, + "step": 10521 + }, + { + "epoch": 1.1127326565143825, + "grad_norm": 8.625, + "learning_rate": 4.529485537728303e-06, + "loss": 2.5687, + "step": 10522 + }, + { + "epoch": 1.1128384094754653, + "grad_norm": 8.25, + "learning_rate": 4.529385528987769e-06, + "loss": 2.5108, + "step": 10523 + }, + { + "epoch": 1.1129441624365481, + "grad_norm": 8.4375, + "learning_rate": 4.52928551072416e-06, + "loss": 2.5593, + "step": 10524 + }, + { + "epoch": 1.1130499153976312, + "grad_norm": 7.8125, + "learning_rate": 4.5291854829379436e-06, + "loss": 2.5468, + "step": 10525 + }, + { + "epoch": 1.113155668358714, + "grad_norm": 8.0625, + "learning_rate": 4.5290854456295894e-06, + "loss": 2.4951, + "step": 10526 + }, + { + "epoch": 1.1132614213197969, + "grad_norm": 8.5, + "learning_rate": 4.528985398799567e-06, + "loss": 2.6462, + "step": 10527 + }, + { + "epoch": 1.11336717428088, + "grad_norm": 8.0625, + "learning_rate": 4.528885342448347e-06, + "loss": 2.5347, + "step": 10528 + }, + { + "epoch": 1.1134729272419628, + "grad_norm": 8.125, + "learning_rate": 4.528785276576396e-06, + "loss": 2.673, + "step": 10529 + }, + { + "epoch": 1.1135786802030456, + "grad_norm": 8.4375, + "learning_rate": 4.528685201184187e-06, + "loss": 2.6374, + "step": 10530 + }, + { + "epoch": 1.1136844331641287, + "grad_norm": 8.375, + "learning_rate": 4.528585116272187e-06, + "loss": 2.4494, + "step": 10531 + }, + { + "epoch": 1.1137901861252115, + "grad_norm": 7.875, + "learning_rate": 4.528485021840868e-06, + "loss": 2.548, + "step": 10532 + }, + { + "epoch": 1.1138959390862944, + "grad_norm": 8.125, + "learning_rate": 4.5283849178906974e-06, + "loss": 2.5619, + "step": 10533 + }, + { + "epoch": 1.1140016920473774, + "grad_norm": 8.125, + "learning_rate": 4.528284804422146e-06, + "loss": 2.4312, + "step": 10534 + }, + { + "epoch": 1.1141074450084603, + "grad_norm": 8.0625, + "learning_rate": 4.528184681435684e-06, + "loss": 2.5131, + "step": 10535 + }, + { + "epoch": 1.114213197969543, + "grad_norm": 8.1875, + "learning_rate": 4.5280845489317805e-06, + "loss": 2.514, + "step": 10536 + }, + { + "epoch": 1.1143189509306262, + "grad_norm": 8.4375, + "learning_rate": 4.527984406910906e-06, + "loss": 2.3913, + "step": 10537 + }, + { + "epoch": 1.114424703891709, + "grad_norm": 8.3125, + "learning_rate": 4.527884255373529e-06, + "loss": 2.6725, + "step": 10538 + }, + { + "epoch": 1.1145304568527918, + "grad_norm": 7.90625, + "learning_rate": 4.527784094320123e-06, + "loss": 2.373, + "step": 10539 + }, + { + "epoch": 1.1146362098138747, + "grad_norm": 8.4375, + "learning_rate": 4.527683923751153e-06, + "loss": 2.5702, + "step": 10540 + }, + { + "epoch": 1.1147419627749577, + "grad_norm": 9.375, + "learning_rate": 4.527583743667093e-06, + "loss": 2.4841, + "step": 10541 + }, + { + "epoch": 1.1148477157360406, + "grad_norm": 8.25, + "learning_rate": 4.527483554068411e-06, + "loss": 2.5293, + "step": 10542 + }, + { + "epoch": 1.1149534686971234, + "grad_norm": 7.78125, + "learning_rate": 4.527383354955578e-06, + "loss": 2.5726, + "step": 10543 + }, + { + "epoch": 1.1150592216582065, + "grad_norm": 8.125, + "learning_rate": 4.527283146329064e-06, + "loss": 2.4616, + "step": 10544 + }, + { + "epoch": 1.1151649746192893, + "grad_norm": 7.90625, + "learning_rate": 4.52718292818934e-06, + "loss": 2.4611, + "step": 10545 + }, + { + "epoch": 1.1152707275803722, + "grad_norm": 8.3125, + "learning_rate": 4.5270827005368755e-06, + "loss": 2.4854, + "step": 10546 + }, + { + "epoch": 1.1153764805414552, + "grad_norm": 8.5, + "learning_rate": 4.526982463372141e-06, + "loss": 2.6013, + "step": 10547 + }, + { + "epoch": 1.115482233502538, + "grad_norm": 8.3125, + "learning_rate": 4.526882216695606e-06, + "loss": 2.5008, + "step": 10548 + }, + { + "epoch": 1.115587986463621, + "grad_norm": 8.625, + "learning_rate": 4.526781960507743e-06, + "loss": 2.6688, + "step": 10549 + }, + { + "epoch": 1.115693739424704, + "grad_norm": 8.0625, + "learning_rate": 4.52668169480902e-06, + "loss": 2.4273, + "step": 10550 + }, + { + "epoch": 1.1157994923857868, + "grad_norm": 10.0, + "learning_rate": 4.526581419599909e-06, + "loss": 2.4978, + "step": 10551 + }, + { + "epoch": 1.1159052453468696, + "grad_norm": 8.0625, + "learning_rate": 4.5264811348808805e-06, + "loss": 2.2524, + "step": 10552 + }, + { + "epoch": 1.1160109983079527, + "grad_norm": 8.0625, + "learning_rate": 4.526380840652405e-06, + "loss": 2.7023, + "step": 10553 + }, + { + "epoch": 1.1161167512690355, + "grad_norm": 8.5625, + "learning_rate": 4.526280536914952e-06, + "loss": 2.6094, + "step": 10554 + }, + { + "epoch": 1.1162225042301184, + "grad_norm": 8.5625, + "learning_rate": 4.526180223668994e-06, + "loss": 2.3031, + "step": 10555 + }, + { + "epoch": 1.1163282571912014, + "grad_norm": 8.625, + "learning_rate": 4.5260799009150015e-06, + "loss": 2.4094, + "step": 10556 + }, + { + "epoch": 1.1164340101522843, + "grad_norm": 8.25, + "learning_rate": 4.5259795686534435e-06, + "loss": 2.7429, + "step": 10557 + }, + { + "epoch": 1.1165397631133671, + "grad_norm": 7.84375, + "learning_rate": 4.525879226884793e-06, + "loss": 2.6412, + "step": 10558 + }, + { + "epoch": 1.1166455160744502, + "grad_norm": 8.5, + "learning_rate": 4.525778875609519e-06, + "loss": 2.6061, + "step": 10559 + }, + { + "epoch": 1.116751269035533, + "grad_norm": 8.1875, + "learning_rate": 4.525678514828094e-06, + "loss": 2.4928, + "step": 10560 + }, + { + "epoch": 1.1168570219966159, + "grad_norm": 8.9375, + "learning_rate": 4.525578144540987e-06, + "loss": 2.6386, + "step": 10561 + }, + { + "epoch": 1.116962774957699, + "grad_norm": 9.125, + "learning_rate": 4.525477764748671e-06, + "loss": 2.6821, + "step": 10562 + }, + { + "epoch": 1.1170685279187818, + "grad_norm": 8.5625, + "learning_rate": 4.5253773754516165e-06, + "loss": 2.2429, + "step": 10563 + }, + { + "epoch": 1.1171742808798646, + "grad_norm": 8.75, + "learning_rate": 4.5252769766502935e-06, + "loss": 2.5267, + "step": 10564 + }, + { + "epoch": 1.1172800338409477, + "grad_norm": 7.96875, + "learning_rate": 4.525176568345174e-06, + "loss": 2.2113, + "step": 10565 + }, + { + "epoch": 1.1173857868020305, + "grad_norm": 7.625, + "learning_rate": 4.52507615053673e-06, + "loss": 2.4268, + "step": 10566 + }, + { + "epoch": 1.1174915397631133, + "grad_norm": 8.625, + "learning_rate": 4.524975723225431e-06, + "loss": 2.7148, + "step": 10567 + }, + { + "epoch": 1.1175972927241964, + "grad_norm": 8.0, + "learning_rate": 4.5248752864117495e-06, + "loss": 2.653, + "step": 10568 + }, + { + "epoch": 1.1177030456852792, + "grad_norm": 8.125, + "learning_rate": 4.5247748400961565e-06, + "loss": 2.516, + "step": 10569 + }, + { + "epoch": 1.117808798646362, + "grad_norm": 8.5, + "learning_rate": 4.524674384279123e-06, + "loss": 2.409, + "step": 10570 + }, + { + "epoch": 1.117914551607445, + "grad_norm": 8.0625, + "learning_rate": 4.524573918961122e-06, + "loss": 2.5786, + "step": 10571 + }, + { + "epoch": 1.118020304568528, + "grad_norm": 8.375, + "learning_rate": 4.524473444142622e-06, + "loss": 2.7065, + "step": 10572 + }, + { + "epoch": 1.1181260575296108, + "grad_norm": 8.3125, + "learning_rate": 4.524372959824096e-06, + "loss": 2.5544, + "step": 10573 + }, + { + "epoch": 1.1182318104906936, + "grad_norm": 8.375, + "learning_rate": 4.524272466006016e-06, + "loss": 2.4202, + "step": 10574 + }, + { + "epoch": 1.1183375634517767, + "grad_norm": 7.71875, + "learning_rate": 4.524171962688854e-06, + "loss": 2.5378, + "step": 10575 + }, + { + "epoch": 1.1184433164128595, + "grad_norm": 8.625, + "learning_rate": 4.524071449873081e-06, + "loss": 2.5069, + "step": 10576 + }, + { + "epoch": 1.1185490693739424, + "grad_norm": 7.96875, + "learning_rate": 4.523970927559167e-06, + "loss": 2.3966, + "step": 10577 + }, + { + "epoch": 1.1186548223350254, + "grad_norm": 8.8125, + "learning_rate": 4.523870395747586e-06, + "loss": 2.661, + "step": 10578 + }, + { + "epoch": 1.1187605752961083, + "grad_norm": 8.25, + "learning_rate": 4.5237698544388096e-06, + "loss": 2.4587, + "step": 10579 + }, + { + "epoch": 1.1188663282571911, + "grad_norm": 8.1875, + "learning_rate": 4.523669303633308e-06, + "loss": 2.6, + "step": 10580 + }, + { + "epoch": 1.1189720812182742, + "grad_norm": 8.875, + "learning_rate": 4.523568743331554e-06, + "loss": 2.5918, + "step": 10581 + }, + { + "epoch": 1.119077834179357, + "grad_norm": 7.78125, + "learning_rate": 4.52346817353402e-06, + "loss": 2.4007, + "step": 10582 + }, + { + "epoch": 1.1191835871404399, + "grad_norm": 8.5625, + "learning_rate": 4.523367594241178e-06, + "loss": 2.4869, + "step": 10583 + }, + { + "epoch": 1.119289340101523, + "grad_norm": 7.84375, + "learning_rate": 4.523267005453499e-06, + "loss": 2.6602, + "step": 10584 + }, + { + "epoch": 1.1193950930626058, + "grad_norm": 7.6875, + "learning_rate": 4.523166407171455e-06, + "loss": 2.5413, + "step": 10585 + }, + { + "epoch": 1.1195008460236886, + "grad_norm": 8.0625, + "learning_rate": 4.523065799395519e-06, + "loss": 2.4284, + "step": 10586 + }, + { + "epoch": 1.1196065989847717, + "grad_norm": 8.0, + "learning_rate": 4.522965182126162e-06, + "loss": 2.526, + "step": 10587 + }, + { + "epoch": 1.1197123519458545, + "grad_norm": 8.25, + "learning_rate": 4.522864555363856e-06, + "loss": 2.511, + "step": 10588 + }, + { + "epoch": 1.1198181049069373, + "grad_norm": 8.4375, + "learning_rate": 4.522763919109075e-06, + "loss": 2.6989, + "step": 10589 + }, + { + "epoch": 1.1199238578680204, + "grad_norm": 8.25, + "learning_rate": 4.522663273362291e-06, + "loss": 2.2049, + "step": 10590 + }, + { + "epoch": 1.1200296108291032, + "grad_norm": 8.5, + "learning_rate": 4.5225626181239745e-06, + "loss": 2.7972, + "step": 10591 + }, + { + "epoch": 1.120135363790186, + "grad_norm": 7.78125, + "learning_rate": 4.522461953394599e-06, + "loss": 2.4281, + "step": 10592 + }, + { + "epoch": 1.1202411167512691, + "grad_norm": 7.84375, + "learning_rate": 4.522361279174637e-06, + "loss": 2.5367, + "step": 10593 + }, + { + "epoch": 1.120346869712352, + "grad_norm": 9.1875, + "learning_rate": 4.52226059546456e-06, + "loss": 2.3742, + "step": 10594 + }, + { + "epoch": 1.1204526226734348, + "grad_norm": 7.8125, + "learning_rate": 4.522159902264842e-06, + "loss": 2.6145, + "step": 10595 + }, + { + "epoch": 1.1205583756345177, + "grad_norm": 7.71875, + "learning_rate": 4.522059199575953e-06, + "loss": 2.5854, + "step": 10596 + }, + { + "epoch": 1.1206641285956007, + "grad_norm": 8.3125, + "learning_rate": 4.521958487398369e-06, + "loss": 2.4659, + "step": 10597 + }, + { + "epoch": 1.1207698815566836, + "grad_norm": 8.625, + "learning_rate": 4.5218577657325605e-06, + "loss": 2.6904, + "step": 10598 + }, + { + "epoch": 1.1208756345177664, + "grad_norm": 7.875, + "learning_rate": 4.521757034579e-06, + "loss": 2.3303, + "step": 10599 + }, + { + "epoch": 1.1209813874788495, + "grad_norm": 8.3125, + "learning_rate": 4.52165629393816e-06, + "loss": 2.4991, + "step": 10600 + }, + { + "epoch": 1.1210871404399323, + "grad_norm": 7.75, + "learning_rate": 4.521555543810514e-06, + "loss": 2.3969, + "step": 10601 + }, + { + "epoch": 1.1211928934010151, + "grad_norm": 8.6875, + "learning_rate": 4.521454784196535e-06, + "loss": 2.6957, + "step": 10602 + }, + { + "epoch": 1.1212986463620982, + "grad_norm": 8.5, + "learning_rate": 4.521354015096696e-06, + "loss": 2.5208, + "step": 10603 + }, + { + "epoch": 1.121404399323181, + "grad_norm": 8.625, + "learning_rate": 4.5212532365114686e-06, + "loss": 2.5713, + "step": 10604 + }, + { + "epoch": 1.1215101522842639, + "grad_norm": 7.90625, + "learning_rate": 4.5211524484413265e-06, + "loss": 2.479, + "step": 10605 + }, + { + "epoch": 1.121615905245347, + "grad_norm": 8.125, + "learning_rate": 4.521051650886743e-06, + "loss": 2.6538, + "step": 10606 + }, + { + "epoch": 1.1217216582064298, + "grad_norm": 7.71875, + "learning_rate": 4.52095084384819e-06, + "loss": 2.4909, + "step": 10607 + }, + { + "epoch": 1.1218274111675126, + "grad_norm": 8.875, + "learning_rate": 4.520850027326142e-06, + "loss": 2.6818, + "step": 10608 + }, + { + "epoch": 1.1219331641285957, + "grad_norm": 8.25, + "learning_rate": 4.520749201321071e-06, + "loss": 2.5352, + "step": 10609 + }, + { + "epoch": 1.1220389170896785, + "grad_norm": 10.0, + "learning_rate": 4.52064836583345e-06, + "loss": 2.5083, + "step": 10610 + }, + { + "epoch": 1.1221446700507614, + "grad_norm": 8.3125, + "learning_rate": 4.520547520863753e-06, + "loss": 2.3655, + "step": 10611 + }, + { + "epoch": 1.1222504230118444, + "grad_norm": 8.25, + "learning_rate": 4.520446666412453e-06, + "loss": 2.7882, + "step": 10612 + }, + { + "epoch": 1.1223561759729273, + "grad_norm": 7.84375, + "learning_rate": 4.5203458024800225e-06, + "loss": 2.685, + "step": 10613 + }, + { + "epoch": 1.12246192893401, + "grad_norm": 8.0625, + "learning_rate": 4.520244929066936e-06, + "loss": 2.5693, + "step": 10614 + }, + { + "epoch": 1.1225676818950932, + "grad_norm": 8.3125, + "learning_rate": 4.520144046173666e-06, + "loss": 2.668, + "step": 10615 + }, + { + "epoch": 1.122673434856176, + "grad_norm": 8.0625, + "learning_rate": 4.520043153800688e-06, + "loss": 2.5267, + "step": 10616 + }, + { + "epoch": 1.1227791878172588, + "grad_norm": 8.0625, + "learning_rate": 4.519942251948472e-06, + "loss": 2.5312, + "step": 10617 + }, + { + "epoch": 1.122884940778342, + "grad_norm": 7.75, + "learning_rate": 4.519841340617492e-06, + "loss": 2.3689, + "step": 10618 + }, + { + "epoch": 1.1229906937394247, + "grad_norm": 8.0625, + "learning_rate": 4.519740419808225e-06, + "loss": 2.6067, + "step": 10619 + }, + { + "epoch": 1.1230964467005076, + "grad_norm": 9.0, + "learning_rate": 4.519639489521141e-06, + "loss": 2.3053, + "step": 10620 + }, + { + "epoch": 1.1232021996615906, + "grad_norm": 8.5, + "learning_rate": 4.519538549756714e-06, + "loss": 2.3622, + "step": 10621 + }, + { + "epoch": 1.1233079526226735, + "grad_norm": 7.875, + "learning_rate": 4.519437600515421e-06, + "loss": 2.4207, + "step": 10622 + }, + { + "epoch": 1.1234137055837563, + "grad_norm": 9.5, + "learning_rate": 4.519336641797732e-06, + "loss": 3.0015, + "step": 10623 + }, + { + "epoch": 1.1235194585448394, + "grad_norm": 7.71875, + "learning_rate": 4.519235673604121e-06, + "loss": 2.5673, + "step": 10624 + }, + { + "epoch": 1.1236252115059222, + "grad_norm": 8.125, + "learning_rate": 4.519134695935063e-06, + "loss": 2.3463, + "step": 10625 + }, + { + "epoch": 1.123730964467005, + "grad_norm": 8.0625, + "learning_rate": 4.519033708791033e-06, + "loss": 2.6161, + "step": 10626 + }, + { + "epoch": 1.1238367174280879, + "grad_norm": 7.84375, + "learning_rate": 4.518932712172503e-06, + "loss": 2.5408, + "step": 10627 + }, + { + "epoch": 1.123942470389171, + "grad_norm": 8.25, + "learning_rate": 4.5188317060799475e-06, + "loss": 2.5201, + "step": 10628 + }, + { + "epoch": 1.1240482233502538, + "grad_norm": 7.53125, + "learning_rate": 4.5187306905138404e-06, + "loss": 2.3058, + "step": 10629 + }, + { + "epoch": 1.1241539763113366, + "grad_norm": 7.90625, + "learning_rate": 4.518629665474656e-06, + "loss": 2.3859, + "step": 10630 + }, + { + "epoch": 1.1242597292724197, + "grad_norm": 8.4375, + "learning_rate": 4.518528630962868e-06, + "loss": 2.4494, + "step": 10631 + }, + { + "epoch": 1.1243654822335025, + "grad_norm": 7.84375, + "learning_rate": 4.51842758697895e-06, + "loss": 2.3186, + "step": 10632 + }, + { + "epoch": 1.1244712351945854, + "grad_norm": 8.4375, + "learning_rate": 4.518326533523378e-06, + "loss": 2.45, + "step": 10633 + }, + { + "epoch": 1.1245769881556684, + "grad_norm": 8.5, + "learning_rate": 4.518225470596624e-06, + "loss": 2.5492, + "step": 10634 + }, + { + "epoch": 1.1246827411167513, + "grad_norm": 8.3125, + "learning_rate": 4.518124398199164e-06, + "loss": 2.5871, + "step": 10635 + }, + { + "epoch": 1.124788494077834, + "grad_norm": 8.1875, + "learning_rate": 4.518023316331472e-06, + "loss": 2.6189, + "step": 10636 + }, + { + "epoch": 1.1248942470389172, + "grad_norm": 8.125, + "learning_rate": 4.517922224994022e-06, + "loss": 2.5621, + "step": 10637 + }, + { + "epoch": 1.125, + "grad_norm": 8.9375, + "learning_rate": 4.517821124187287e-06, + "loss": 2.4626, + "step": 10638 + }, + { + "epoch": 1.1251057529610828, + "grad_norm": 7.6875, + "learning_rate": 4.5177200139117425e-06, + "loss": 2.3533, + "step": 10639 + }, + { + "epoch": 1.125211505922166, + "grad_norm": 8.125, + "learning_rate": 4.517618894167865e-06, + "loss": 2.5882, + "step": 10640 + }, + { + "epoch": 1.1253172588832487, + "grad_norm": 8.125, + "learning_rate": 4.5175177649561266e-06, + "loss": 2.3835, + "step": 10641 + }, + { + "epoch": 1.1254230118443316, + "grad_norm": 7.8125, + "learning_rate": 4.517416626277001e-06, + "loss": 2.4509, + "step": 10642 + }, + { + "epoch": 1.1255287648054146, + "grad_norm": 8.375, + "learning_rate": 4.517315478130966e-06, + "loss": 2.7092, + "step": 10643 + }, + { + "epoch": 1.1256345177664975, + "grad_norm": 7.75, + "learning_rate": 4.517214320518493e-06, + "loss": 2.6886, + "step": 10644 + }, + { + "epoch": 1.1257402707275803, + "grad_norm": 8.4375, + "learning_rate": 4.51711315344006e-06, + "loss": 2.2393, + "step": 10645 + }, + { + "epoch": 1.1258460236886634, + "grad_norm": 7.9375, + "learning_rate": 4.517011976896139e-06, + "loss": 2.4822, + "step": 10646 + }, + { + "epoch": 1.1259517766497462, + "grad_norm": 8.1875, + "learning_rate": 4.5169107908872055e-06, + "loss": 2.5102, + "step": 10647 + }, + { + "epoch": 1.126057529610829, + "grad_norm": 8.3125, + "learning_rate": 4.516809595413736e-06, + "loss": 2.658, + "step": 10648 + }, + { + "epoch": 1.1261632825719121, + "grad_norm": 8.25, + "learning_rate": 4.516708390476202e-06, + "loss": 2.5825, + "step": 10649 + }, + { + "epoch": 1.126269035532995, + "grad_norm": 8.3125, + "learning_rate": 4.516607176075081e-06, + "loss": 2.6019, + "step": 10650 + }, + { + "epoch": 1.1263747884940778, + "grad_norm": 7.9375, + "learning_rate": 4.516505952210848e-06, + "loss": 2.5958, + "step": 10651 + }, + { + "epoch": 1.1264805414551606, + "grad_norm": 8.5, + "learning_rate": 4.516404718883976e-06, + "loss": 2.7104, + "step": 10652 + }, + { + "epoch": 1.1265862944162437, + "grad_norm": 8.1875, + "learning_rate": 4.516303476094942e-06, + "loss": 2.5213, + "step": 10653 + }, + { + "epoch": 1.1266920473773265, + "grad_norm": 7.9375, + "learning_rate": 4.51620222384422e-06, + "loss": 2.6147, + "step": 10654 + }, + { + "epoch": 1.1267978003384096, + "grad_norm": 8.25, + "learning_rate": 4.516100962132287e-06, + "loss": 2.501, + "step": 10655 + }, + { + "epoch": 1.1269035532994924, + "grad_norm": 8.25, + "learning_rate": 4.515999690959615e-06, + "loss": 2.4602, + "step": 10656 + }, + { + "epoch": 1.1270093062605753, + "grad_norm": 8.0625, + "learning_rate": 4.515898410326682e-06, + "loss": 2.5718, + "step": 10657 + }, + { + "epoch": 1.127115059221658, + "grad_norm": 8.1875, + "learning_rate": 4.515797120233962e-06, + "loss": 2.4611, + "step": 10658 + }, + { + "epoch": 1.1272208121827412, + "grad_norm": 8.375, + "learning_rate": 4.5156958206819305e-06, + "loss": 2.515, + "step": 10659 + }, + { + "epoch": 1.127326565143824, + "grad_norm": 8.25, + "learning_rate": 4.515594511671063e-06, + "loss": 2.626, + "step": 10660 + }, + { + "epoch": 1.1274323181049068, + "grad_norm": 7.625, + "learning_rate": 4.515493193201834e-06, + "loss": 2.3875, + "step": 10661 + }, + { + "epoch": 1.12753807106599, + "grad_norm": 8.0, + "learning_rate": 4.515391865274721e-06, + "loss": 2.5942, + "step": 10662 + }, + { + "epoch": 1.1276438240270727, + "grad_norm": 8.5625, + "learning_rate": 4.5152905278901976e-06, + "loss": 2.686, + "step": 10663 + }, + { + "epoch": 1.1277495769881556, + "grad_norm": 8.4375, + "learning_rate": 4.51518918104874e-06, + "loss": 2.6679, + "step": 10664 + }, + { + "epoch": 1.1278553299492386, + "grad_norm": 8.3125, + "learning_rate": 4.515087824750824e-06, + "loss": 2.5751, + "step": 10665 + }, + { + "epoch": 1.1279610829103215, + "grad_norm": 8.75, + "learning_rate": 4.514986458996925e-06, + "loss": 2.6489, + "step": 10666 + }, + { + "epoch": 1.1280668358714043, + "grad_norm": 8.3125, + "learning_rate": 4.514885083787518e-06, + "loss": 2.5984, + "step": 10667 + }, + { + "epoch": 1.1281725888324874, + "grad_norm": 8.375, + "learning_rate": 4.51478369912308e-06, + "loss": 2.4401, + "step": 10668 + }, + { + "epoch": 1.1282783417935702, + "grad_norm": 13.25, + "learning_rate": 4.514682305004086e-06, + "loss": 2.6781, + "step": 10669 + }, + { + "epoch": 1.128384094754653, + "grad_norm": 8.0625, + "learning_rate": 4.514580901431012e-06, + "loss": 2.5582, + "step": 10670 + }, + { + "epoch": 1.1284898477157361, + "grad_norm": 8.5, + "learning_rate": 4.5144794884043336e-06, + "loss": 2.6359, + "step": 10671 + }, + { + "epoch": 1.128595600676819, + "grad_norm": 8.0, + "learning_rate": 4.514378065924527e-06, + "loss": 2.4948, + "step": 10672 + }, + { + "epoch": 1.1287013536379018, + "grad_norm": 8.5, + "learning_rate": 4.514276633992067e-06, + "loss": 2.4793, + "step": 10673 + }, + { + "epoch": 1.1288071065989849, + "grad_norm": 8.3125, + "learning_rate": 4.514175192607432e-06, + "loss": 2.4008, + "step": 10674 + }, + { + "epoch": 1.1289128595600677, + "grad_norm": 8.1875, + "learning_rate": 4.514073741771096e-06, + "loss": 2.6579, + "step": 10675 + }, + { + "epoch": 1.1290186125211505, + "grad_norm": 8.1875, + "learning_rate": 4.5139722814835356e-06, + "loss": 2.7156, + "step": 10676 + }, + { + "epoch": 1.1291243654822334, + "grad_norm": 8.1875, + "learning_rate": 4.5138708117452266e-06, + "loss": 2.4808, + "step": 10677 + }, + { + "epoch": 1.1292301184433164, + "grad_norm": 9.0, + "learning_rate": 4.513769332556646e-06, + "loss": 2.627, + "step": 10678 + }, + { + "epoch": 1.1293358714043993, + "grad_norm": 7.9375, + "learning_rate": 4.51366784391827e-06, + "loss": 2.3752, + "step": 10679 + }, + { + "epoch": 1.1294416243654823, + "grad_norm": 8.9375, + "learning_rate": 4.513566345830574e-06, + "loss": 2.477, + "step": 10680 + }, + { + "epoch": 1.1295473773265652, + "grad_norm": 8.5625, + "learning_rate": 4.5134648382940346e-06, + "loss": 2.6067, + "step": 10681 + }, + { + "epoch": 1.129653130287648, + "grad_norm": 8.1875, + "learning_rate": 4.513363321309127e-06, + "loss": 2.6123, + "step": 10682 + }, + { + "epoch": 1.1297588832487309, + "grad_norm": 8.125, + "learning_rate": 4.513261794876331e-06, + "loss": 2.5269, + "step": 10683 + }, + { + "epoch": 1.129864636209814, + "grad_norm": 7.84375, + "learning_rate": 4.51316025899612e-06, + "loss": 2.5085, + "step": 10684 + }, + { + "epoch": 1.1299703891708968, + "grad_norm": 7.84375, + "learning_rate": 4.513058713668971e-06, + "loss": 2.6753, + "step": 10685 + }, + { + "epoch": 1.1300761421319796, + "grad_norm": 8.0625, + "learning_rate": 4.51295715889536e-06, + "loss": 2.4752, + "step": 10686 + }, + { + "epoch": 1.1301818950930627, + "grad_norm": 8.125, + "learning_rate": 4.5128555946757655e-06, + "loss": 2.4507, + "step": 10687 + }, + { + "epoch": 1.1302876480541455, + "grad_norm": 8.125, + "learning_rate": 4.512754021010663e-06, + "loss": 2.7225, + "step": 10688 + }, + { + "epoch": 1.1303934010152283, + "grad_norm": 8.1875, + "learning_rate": 4.5126524379005275e-06, + "loss": 2.4346, + "step": 10689 + }, + { + "epoch": 1.1304991539763114, + "grad_norm": 8.9375, + "learning_rate": 4.5125508453458386e-06, + "loss": 2.5278, + "step": 10690 + }, + { + "epoch": 1.1306049069373942, + "grad_norm": 8.5, + "learning_rate": 4.512449243347071e-06, + "loss": 2.462, + "step": 10691 + }, + { + "epoch": 1.130710659898477, + "grad_norm": 8.3125, + "learning_rate": 4.512347631904703e-06, + "loss": 2.6712, + "step": 10692 + }, + { + "epoch": 1.1308164128595601, + "grad_norm": 8.5625, + "learning_rate": 4.51224601101921e-06, + "loss": 2.6605, + "step": 10693 + }, + { + "epoch": 1.130922165820643, + "grad_norm": 7.8125, + "learning_rate": 4.512144380691069e-06, + "loss": 2.4823, + "step": 10694 + }, + { + "epoch": 1.1310279187817258, + "grad_norm": 8.375, + "learning_rate": 4.5120427409207585e-06, + "loss": 2.6618, + "step": 10695 + }, + { + "epoch": 1.1311336717428089, + "grad_norm": 8.0, + "learning_rate": 4.511941091708754e-06, + "loss": 2.6233, + "step": 10696 + }, + { + "epoch": 1.1312394247038917, + "grad_norm": 8.125, + "learning_rate": 4.511839433055532e-06, + "loss": 2.6041, + "step": 10697 + }, + { + "epoch": 1.1313451776649746, + "grad_norm": 8.1875, + "learning_rate": 4.511737764961571e-06, + "loss": 2.6677, + "step": 10698 + }, + { + "epoch": 1.1314509306260576, + "grad_norm": 8.625, + "learning_rate": 4.511636087427348e-06, + "loss": 2.7213, + "step": 10699 + }, + { + "epoch": 1.1315566835871405, + "grad_norm": 8.4375, + "learning_rate": 4.511534400453338e-06, + "loss": 2.5644, + "step": 10700 + }, + { + "epoch": 1.1316624365482233, + "grad_norm": 8.3125, + "learning_rate": 4.51143270404002e-06, + "loss": 2.5152, + "step": 10701 + }, + { + "epoch": 1.1317681895093064, + "grad_norm": 8.0625, + "learning_rate": 4.511330998187873e-06, + "loss": 2.434, + "step": 10702 + }, + { + "epoch": 1.1318739424703892, + "grad_norm": 8.0625, + "learning_rate": 4.511229282897371e-06, + "loss": 2.6265, + "step": 10703 + }, + { + "epoch": 1.131979695431472, + "grad_norm": 7.71875, + "learning_rate": 4.511127558168992e-06, + "loss": 2.6034, + "step": 10704 + }, + { + "epoch": 1.132085448392555, + "grad_norm": 8.25, + "learning_rate": 4.511025824003215e-06, + "loss": 2.7524, + "step": 10705 + }, + { + "epoch": 1.132191201353638, + "grad_norm": 7.96875, + "learning_rate": 4.510924080400515e-06, + "loss": 2.5867, + "step": 10706 + }, + { + "epoch": 1.1322969543147208, + "grad_norm": 7.96875, + "learning_rate": 4.510822327361372e-06, + "loss": 2.5624, + "step": 10707 + }, + { + "epoch": 1.1324027072758036, + "grad_norm": 8.625, + "learning_rate": 4.51072056488626e-06, + "loss": 2.5696, + "step": 10708 + }, + { + "epoch": 1.1325084602368867, + "grad_norm": 8.1875, + "learning_rate": 4.510618792975661e-06, + "loss": 2.5071, + "step": 10709 + }, + { + "epoch": 1.1326142131979695, + "grad_norm": 7.90625, + "learning_rate": 4.5105170116300505e-06, + "loss": 2.6228, + "step": 10710 + }, + { + "epoch": 1.1327199661590526, + "grad_norm": 8.5, + "learning_rate": 4.510415220849904e-06, + "loss": 2.2264, + "step": 10711 + }, + { + "epoch": 1.1328257191201354, + "grad_norm": 8.0625, + "learning_rate": 4.510313420635702e-06, + "loss": 2.461, + "step": 10712 + }, + { + "epoch": 1.1329314720812182, + "grad_norm": 8.375, + "learning_rate": 4.5102116109879225e-06, + "loss": 2.5394, + "step": 10713 + }, + { + "epoch": 1.133037225042301, + "grad_norm": 8.125, + "learning_rate": 4.51010979190704e-06, + "loss": 2.5228, + "step": 10714 + }, + { + "epoch": 1.1331429780033841, + "grad_norm": 7.6875, + "learning_rate": 4.510007963393536e-06, + "loss": 2.5632, + "step": 10715 + }, + { + "epoch": 1.133248730964467, + "grad_norm": 8.0625, + "learning_rate": 4.509906125447886e-06, + "loss": 2.3529, + "step": 10716 + }, + { + "epoch": 1.1333544839255498, + "grad_norm": 8.4375, + "learning_rate": 4.509804278070569e-06, + "loss": 2.5778, + "step": 10717 + }, + { + "epoch": 1.1334602368866329, + "grad_norm": 8.875, + "learning_rate": 4.509702421262061e-06, + "loss": 2.7676, + "step": 10718 + }, + { + "epoch": 1.1335659898477157, + "grad_norm": 8.0625, + "learning_rate": 4.509600555022843e-06, + "loss": 2.4397, + "step": 10719 + }, + { + "epoch": 1.1336717428087986, + "grad_norm": 7.6875, + "learning_rate": 4.509498679353391e-06, + "loss": 2.2985, + "step": 10720 + }, + { + "epoch": 1.1337774957698816, + "grad_norm": 8.125, + "learning_rate": 4.509396794254184e-06, + "loss": 2.6295, + "step": 10721 + }, + { + "epoch": 1.1338832487309645, + "grad_norm": 8.125, + "learning_rate": 4.509294899725699e-06, + "loss": 2.5434, + "step": 10722 + }, + { + "epoch": 1.1339890016920473, + "grad_norm": 7.78125, + "learning_rate": 4.509192995768416e-06, + "loss": 2.3482, + "step": 10723 + }, + { + "epoch": 1.1340947546531304, + "grad_norm": 7.84375, + "learning_rate": 4.509091082382811e-06, + "loss": 2.7381, + "step": 10724 + }, + { + "epoch": 1.1342005076142132, + "grad_norm": 10.25, + "learning_rate": 4.508989159569363e-06, + "loss": 2.4111, + "step": 10725 + }, + { + "epoch": 1.134306260575296, + "grad_norm": 7.59375, + "learning_rate": 4.508887227328551e-06, + "loss": 2.5365, + "step": 10726 + }, + { + "epoch": 1.134412013536379, + "grad_norm": 7.84375, + "learning_rate": 4.508785285660852e-06, + "loss": 2.5083, + "step": 10727 + }, + { + "epoch": 1.134517766497462, + "grad_norm": 8.25, + "learning_rate": 4.508683334566747e-06, + "loss": 2.5124, + "step": 10728 + }, + { + "epoch": 1.1346235194585448, + "grad_norm": 8.5, + "learning_rate": 4.508581374046711e-06, + "loss": 2.6718, + "step": 10729 + }, + { + "epoch": 1.1347292724196278, + "grad_norm": 7.8125, + "learning_rate": 4.508479404101225e-06, + "loss": 2.2769, + "step": 10730 + }, + { + "epoch": 1.1348350253807107, + "grad_norm": 7.625, + "learning_rate": 4.508377424730767e-06, + "loss": 2.4936, + "step": 10731 + }, + { + "epoch": 1.1349407783417935, + "grad_norm": 10.25, + "learning_rate": 4.5082754359358135e-06, + "loss": 2.8869, + "step": 10732 + }, + { + "epoch": 1.1350465313028764, + "grad_norm": 7.65625, + "learning_rate": 4.508173437716846e-06, + "loss": 2.5969, + "step": 10733 + }, + { + "epoch": 1.1351522842639594, + "grad_norm": 9.4375, + "learning_rate": 4.5080714300743415e-06, + "loss": 2.7362, + "step": 10734 + }, + { + "epoch": 1.1352580372250423, + "grad_norm": 7.6875, + "learning_rate": 4.507969413008779e-06, + "loss": 2.4541, + "step": 10735 + }, + { + "epoch": 1.1353637901861253, + "grad_norm": 7.8125, + "learning_rate": 4.507867386520637e-06, + "loss": 2.4327, + "step": 10736 + }, + { + "epoch": 1.1354695431472082, + "grad_norm": 7.84375, + "learning_rate": 4.507765350610396e-06, + "loss": 2.666, + "step": 10737 + }, + { + "epoch": 1.135575296108291, + "grad_norm": 8.4375, + "learning_rate": 4.507663305278532e-06, + "loss": 2.5288, + "step": 10738 + }, + { + "epoch": 1.1356810490693738, + "grad_norm": 7.90625, + "learning_rate": 4.507561250525525e-06, + "loss": 2.4055, + "step": 10739 + }, + { + "epoch": 1.135786802030457, + "grad_norm": 8.625, + "learning_rate": 4.5074591863518545e-06, + "loss": 2.3349, + "step": 10740 + }, + { + "epoch": 1.1358925549915397, + "grad_norm": 8.25, + "learning_rate": 4.507357112757999e-06, + "loss": 2.6444, + "step": 10741 + }, + { + "epoch": 1.1359983079526226, + "grad_norm": 7.875, + "learning_rate": 4.507255029744438e-06, + "loss": 2.4568, + "step": 10742 + }, + { + "epoch": 1.1361040609137056, + "grad_norm": 8.125, + "learning_rate": 4.50715293731165e-06, + "loss": 2.4332, + "step": 10743 + }, + { + "epoch": 1.1362098138747885, + "grad_norm": 8.6875, + "learning_rate": 4.507050835460114e-06, + "loss": 2.6116, + "step": 10744 + }, + { + "epoch": 1.1363155668358713, + "grad_norm": 8.0, + "learning_rate": 4.506948724190309e-06, + "loss": 2.3708, + "step": 10745 + }, + { + "epoch": 1.1364213197969544, + "grad_norm": 7.9375, + "learning_rate": 4.5068466035027144e-06, + "loss": 2.5852, + "step": 10746 + }, + { + "epoch": 1.1365270727580372, + "grad_norm": 8.1875, + "learning_rate": 4.50674447339781e-06, + "loss": 2.5835, + "step": 10747 + }, + { + "epoch": 1.13663282571912, + "grad_norm": 8.1875, + "learning_rate": 4.506642333876075e-06, + "loss": 2.396, + "step": 10748 + }, + { + "epoch": 1.1367385786802031, + "grad_norm": 8.4375, + "learning_rate": 4.5065401849379865e-06, + "loss": 2.3487, + "step": 10749 + }, + { + "epoch": 1.136844331641286, + "grad_norm": 8.6875, + "learning_rate": 4.506438026584027e-06, + "loss": 2.5592, + "step": 10750 + }, + { + "epoch": 1.1369500846023688, + "grad_norm": 8.25, + "learning_rate": 4.506335858814674e-06, + "loss": 2.5013, + "step": 10751 + }, + { + "epoch": 1.1370558375634519, + "grad_norm": 9.1875, + "learning_rate": 4.5062336816304075e-06, + "loss": 2.6382, + "step": 10752 + }, + { + "epoch": 1.1371615905245347, + "grad_norm": 8.0625, + "learning_rate": 4.506131495031707e-06, + "loss": 2.6543, + "step": 10753 + }, + { + "epoch": 1.1372673434856175, + "grad_norm": 8.375, + "learning_rate": 4.506029299019052e-06, + "loss": 2.363, + "step": 10754 + }, + { + "epoch": 1.1373730964467006, + "grad_norm": 8.3125, + "learning_rate": 4.505927093592921e-06, + "loss": 2.5956, + "step": 10755 + }, + { + "epoch": 1.1374788494077834, + "grad_norm": 8.25, + "learning_rate": 4.505824878753795e-06, + "loss": 2.4135, + "step": 10756 + }, + { + "epoch": 1.1375846023688663, + "grad_norm": 8.3125, + "learning_rate": 4.5057226545021535e-06, + "loss": 2.4519, + "step": 10757 + }, + { + "epoch": 1.1376903553299493, + "grad_norm": 8.875, + "learning_rate": 4.505620420838476e-06, + "loss": 2.7339, + "step": 10758 + }, + { + "epoch": 1.1377961082910322, + "grad_norm": 8.25, + "learning_rate": 4.5055181777632416e-06, + "loss": 2.4046, + "step": 10759 + }, + { + "epoch": 1.137901861252115, + "grad_norm": 8.8125, + "learning_rate": 4.50541592527693e-06, + "loss": 2.709, + "step": 10760 + }, + { + "epoch": 1.138007614213198, + "grad_norm": 8.0, + "learning_rate": 4.505313663380023e-06, + "loss": 2.5414, + "step": 10761 + }, + { + "epoch": 1.138113367174281, + "grad_norm": 8.1875, + "learning_rate": 4.505211392072999e-06, + "loss": 2.6047, + "step": 10762 + }, + { + "epoch": 1.1382191201353637, + "grad_norm": 8.0625, + "learning_rate": 4.5051091113563374e-06, + "loss": 2.4194, + "step": 10763 + }, + { + "epoch": 1.1383248730964466, + "grad_norm": 8.125, + "learning_rate": 4.5050068212305185e-06, + "loss": 2.4991, + "step": 10764 + }, + { + "epoch": 1.1384306260575296, + "grad_norm": 7.625, + "learning_rate": 4.504904521696024e-06, + "loss": 2.4414, + "step": 10765 + }, + { + "epoch": 1.1385363790186125, + "grad_norm": 8.5625, + "learning_rate": 4.504802212753331e-06, + "loss": 2.5726, + "step": 10766 + }, + { + "epoch": 1.1386421319796955, + "grad_norm": 8.6875, + "learning_rate": 4.504699894402923e-06, + "loss": 2.5876, + "step": 10767 + }, + { + "epoch": 1.1387478849407784, + "grad_norm": 8.0625, + "learning_rate": 4.504597566645277e-06, + "loss": 2.5408, + "step": 10768 + }, + { + "epoch": 1.1388536379018612, + "grad_norm": 8.375, + "learning_rate": 4.504495229480874e-06, + "loss": 2.633, + "step": 10769 + }, + { + "epoch": 1.138959390862944, + "grad_norm": 7.78125, + "learning_rate": 4.504392882910196e-06, + "loss": 2.2924, + "step": 10770 + }, + { + "epoch": 1.1390651438240271, + "grad_norm": 7.59375, + "learning_rate": 4.504290526933721e-06, + "loss": 2.4479, + "step": 10771 + }, + { + "epoch": 1.13917089678511, + "grad_norm": 8.375, + "learning_rate": 4.50418816155193e-06, + "loss": 2.4759, + "step": 10772 + }, + { + "epoch": 1.1392766497461928, + "grad_norm": 7.75, + "learning_rate": 4.504085786765304e-06, + "loss": 2.4949, + "step": 10773 + }, + { + "epoch": 1.1393824027072759, + "grad_norm": 9.0625, + "learning_rate": 4.503983402574324e-06, + "loss": 2.5921, + "step": 10774 + }, + { + "epoch": 1.1394881556683587, + "grad_norm": 8.1875, + "learning_rate": 4.503881008979468e-06, + "loss": 2.6846, + "step": 10775 + }, + { + "epoch": 1.1395939086294415, + "grad_norm": 8.0625, + "learning_rate": 4.503778605981219e-06, + "loss": 2.5781, + "step": 10776 + }, + { + "epoch": 1.1396996615905246, + "grad_norm": 8.125, + "learning_rate": 4.503676193580057e-06, + "loss": 2.5442, + "step": 10777 + }, + { + "epoch": 1.1398054145516074, + "grad_norm": 8.375, + "learning_rate": 4.503573771776461e-06, + "loss": 2.7234, + "step": 10778 + }, + { + "epoch": 1.1399111675126903, + "grad_norm": 8.0625, + "learning_rate": 4.503471340570913e-06, + "loss": 2.5067, + "step": 10779 + }, + { + "epoch": 1.1400169204737733, + "grad_norm": 7.8125, + "learning_rate": 4.503368899963894e-06, + "loss": 2.3666, + "step": 10780 + }, + { + "epoch": 1.1401226734348562, + "grad_norm": 7.9375, + "learning_rate": 4.503266449955883e-06, + "loss": 2.4586, + "step": 10781 + }, + { + "epoch": 1.140228426395939, + "grad_norm": 7.34375, + "learning_rate": 4.503163990547362e-06, + "loss": 2.4997, + "step": 10782 + }, + { + "epoch": 1.140334179357022, + "grad_norm": 7.8125, + "learning_rate": 4.503061521738813e-06, + "loss": 2.1755, + "step": 10783 + }, + { + "epoch": 1.140439932318105, + "grad_norm": 8.5625, + "learning_rate": 4.502959043530715e-06, + "loss": 2.6336, + "step": 10784 + }, + { + "epoch": 1.1405456852791878, + "grad_norm": 7.84375, + "learning_rate": 4.5028565559235486e-06, + "loss": 2.6088, + "step": 10785 + }, + { + "epoch": 1.1406514382402708, + "grad_norm": 8.5625, + "learning_rate": 4.502754058917795e-06, + "loss": 2.4987, + "step": 10786 + }, + { + "epoch": 1.1407571912013537, + "grad_norm": 8.375, + "learning_rate": 4.502651552513938e-06, + "loss": 2.4325, + "step": 10787 + }, + { + "epoch": 1.1408629441624365, + "grad_norm": 7.53125, + "learning_rate": 4.502549036712454e-06, + "loss": 2.2794, + "step": 10788 + }, + { + "epoch": 1.1409686971235193, + "grad_norm": 8.1875, + "learning_rate": 4.502446511513828e-06, + "loss": 2.6904, + "step": 10789 + }, + { + "epoch": 1.1410744500846024, + "grad_norm": 7.90625, + "learning_rate": 4.502343976918538e-06, + "loss": 2.511, + "step": 10790 + }, + { + "epoch": 1.1411802030456852, + "grad_norm": 8.4375, + "learning_rate": 4.5022414329270685e-06, + "loss": 2.6439, + "step": 10791 + }, + { + "epoch": 1.1412859560067683, + "grad_norm": 7.875, + "learning_rate": 4.502138879539898e-06, + "loss": 2.4986, + "step": 10792 + }, + { + "epoch": 1.1413917089678511, + "grad_norm": 8.4375, + "learning_rate": 4.502036316757508e-06, + "loss": 2.7425, + "step": 10793 + }, + { + "epoch": 1.141497461928934, + "grad_norm": 8.9375, + "learning_rate": 4.50193374458038e-06, + "loss": 2.5599, + "step": 10794 + }, + { + "epoch": 1.1416032148900168, + "grad_norm": 8.3125, + "learning_rate": 4.5018311630089975e-06, + "loss": 2.5155, + "step": 10795 + }, + { + "epoch": 1.1417089678510999, + "grad_norm": 8.125, + "learning_rate": 4.5017285720438385e-06, + "loss": 2.464, + "step": 10796 + }, + { + "epoch": 1.1418147208121827, + "grad_norm": 8.3125, + "learning_rate": 4.501625971685386e-06, + "loss": 2.6921, + "step": 10797 + }, + { + "epoch": 1.1419204737732658, + "grad_norm": 8.875, + "learning_rate": 4.501523361934123e-06, + "loss": 2.0591, + "step": 10798 + }, + { + "epoch": 1.1420262267343486, + "grad_norm": 8.375, + "learning_rate": 4.501420742790527e-06, + "loss": 2.7481, + "step": 10799 + }, + { + "epoch": 1.1421319796954315, + "grad_norm": 8.5, + "learning_rate": 4.501318114255085e-06, + "loss": 2.522, + "step": 10800 + }, + { + "epoch": 1.1422377326565143, + "grad_norm": 9.0, + "learning_rate": 4.5012154763282736e-06, + "loss": 2.524, + "step": 10801 + }, + { + "epoch": 1.1423434856175974, + "grad_norm": 7.6875, + "learning_rate": 4.501112829010577e-06, + "loss": 2.501, + "step": 10802 + }, + { + "epoch": 1.1424492385786802, + "grad_norm": 8.375, + "learning_rate": 4.501010172302476e-06, + "loss": 2.4642, + "step": 10803 + }, + { + "epoch": 1.142554991539763, + "grad_norm": 8.25, + "learning_rate": 4.500907506204453e-06, + "loss": 2.4459, + "step": 10804 + }, + { + "epoch": 1.142660744500846, + "grad_norm": 7.9375, + "learning_rate": 4.50080483071699e-06, + "loss": 2.599, + "step": 10805 + }, + { + "epoch": 1.142766497461929, + "grad_norm": 9.9375, + "learning_rate": 4.500702145840566e-06, + "loss": 2.8889, + "step": 10806 + }, + { + "epoch": 1.1428722504230118, + "grad_norm": 8.1875, + "learning_rate": 4.500599451575667e-06, + "loss": 2.5584, + "step": 10807 + }, + { + "epoch": 1.1429780033840948, + "grad_norm": 8.0625, + "learning_rate": 4.500496747922773e-06, + "loss": 2.5946, + "step": 10808 + }, + { + "epoch": 1.1430837563451777, + "grad_norm": 7.75, + "learning_rate": 4.500394034882365e-06, + "loss": 2.5879, + "step": 10809 + }, + { + "epoch": 1.1431895093062605, + "grad_norm": 8.125, + "learning_rate": 4.500291312454927e-06, + "loss": 2.4381, + "step": 10810 + }, + { + "epoch": 1.1432952622673436, + "grad_norm": 7.84375, + "learning_rate": 4.500188580640939e-06, + "loss": 2.4469, + "step": 10811 + }, + { + "epoch": 1.1434010152284264, + "grad_norm": 8.1875, + "learning_rate": 4.500085839440884e-06, + "loss": 2.5945, + "step": 10812 + }, + { + "epoch": 1.1435067681895092, + "grad_norm": 7.875, + "learning_rate": 4.4999830888552455e-06, + "loss": 2.5575, + "step": 10813 + }, + { + "epoch": 1.1436125211505923, + "grad_norm": 7.84375, + "learning_rate": 4.499880328884503e-06, + "loss": 2.4425, + "step": 10814 + }, + { + "epoch": 1.1437182741116751, + "grad_norm": 7.84375, + "learning_rate": 4.499777559529141e-06, + "loss": 2.405, + "step": 10815 + }, + { + "epoch": 1.143824027072758, + "grad_norm": 8.6875, + "learning_rate": 4.49967478078964e-06, + "loss": 2.5364, + "step": 10816 + }, + { + "epoch": 1.143929780033841, + "grad_norm": 8.0, + "learning_rate": 4.499571992666483e-06, + "loss": 2.5458, + "step": 10817 + }, + { + "epoch": 1.1440355329949239, + "grad_norm": 9.5625, + "learning_rate": 4.499469195160154e-06, + "loss": 2.7107, + "step": 10818 + }, + { + "epoch": 1.1441412859560067, + "grad_norm": 8.75, + "learning_rate": 4.499366388271131e-06, + "loss": 2.6153, + "step": 10819 + }, + { + "epoch": 1.1442470389170896, + "grad_norm": 8.1875, + "learning_rate": 4.499263571999901e-06, + "loss": 2.5609, + "step": 10820 + }, + { + "epoch": 1.1443527918781726, + "grad_norm": 8.125, + "learning_rate": 4.499160746346945e-06, + "loss": 2.6617, + "step": 10821 + }, + { + "epoch": 1.1444585448392555, + "grad_norm": 8.125, + "learning_rate": 4.499057911312745e-06, + "loss": 2.3868, + "step": 10822 + }, + { + "epoch": 1.1445642978003385, + "grad_norm": 8.0625, + "learning_rate": 4.498955066897784e-06, + "loss": 2.5263, + "step": 10823 + }, + { + "epoch": 1.1446700507614214, + "grad_norm": 8.75, + "learning_rate": 4.498852213102544e-06, + "loss": 2.6521, + "step": 10824 + }, + { + "epoch": 1.1447758037225042, + "grad_norm": 8.3125, + "learning_rate": 4.498749349927508e-06, + "loss": 2.517, + "step": 10825 + }, + { + "epoch": 1.144881556683587, + "grad_norm": 8.1875, + "learning_rate": 4.498646477373159e-06, + "loss": 2.5228, + "step": 10826 + }, + { + "epoch": 1.14498730964467, + "grad_norm": 8.0625, + "learning_rate": 4.4985435954399795e-06, + "loss": 2.2329, + "step": 10827 + }, + { + "epoch": 1.145093062605753, + "grad_norm": 7.9375, + "learning_rate": 4.498440704128452e-06, + "loss": 2.6305, + "step": 10828 + }, + { + "epoch": 1.1451988155668358, + "grad_norm": 7.8125, + "learning_rate": 4.498337803439059e-06, + "loss": 2.6171, + "step": 10829 + }, + { + "epoch": 1.1453045685279188, + "grad_norm": 7.96875, + "learning_rate": 4.498234893372285e-06, + "loss": 2.6456, + "step": 10830 + }, + { + "epoch": 1.1454103214890017, + "grad_norm": 8.6875, + "learning_rate": 4.498131973928612e-06, + "loss": 2.5482, + "step": 10831 + }, + { + "epoch": 1.1455160744500845, + "grad_norm": 8.375, + "learning_rate": 4.498029045108523e-06, + "loss": 2.6174, + "step": 10832 + }, + { + "epoch": 1.1456218274111676, + "grad_norm": 7.875, + "learning_rate": 4.4979261069125e-06, + "loss": 2.3916, + "step": 10833 + }, + { + "epoch": 1.1457275803722504, + "grad_norm": 8.5, + "learning_rate": 4.497823159341027e-06, + "loss": 2.4739, + "step": 10834 + }, + { + "epoch": 1.1458333333333333, + "grad_norm": 8.8125, + "learning_rate": 4.497720202394587e-06, + "loss": 2.7047, + "step": 10835 + }, + { + "epoch": 1.1459390862944163, + "grad_norm": 8.1875, + "learning_rate": 4.497617236073663e-06, + "loss": 2.6662, + "step": 10836 + }, + { + "epoch": 1.1460448392554992, + "grad_norm": 8.875, + "learning_rate": 4.49751426037874e-06, + "loss": 2.4853, + "step": 10837 + }, + { + "epoch": 1.146150592216582, + "grad_norm": 8.4375, + "learning_rate": 4.4974112753102985e-06, + "loss": 2.7388, + "step": 10838 + }, + { + "epoch": 1.146256345177665, + "grad_norm": 8.125, + "learning_rate": 4.497308280868822e-06, + "loss": 2.5398, + "step": 10839 + }, + { + "epoch": 1.146362098138748, + "grad_norm": 8.625, + "learning_rate": 4.497205277054795e-06, + "loss": 2.5506, + "step": 10840 + }, + { + "epoch": 1.1464678510998307, + "grad_norm": 8.375, + "learning_rate": 4.497102263868701e-06, + "loss": 2.6033, + "step": 10841 + }, + { + "epoch": 1.1465736040609138, + "grad_norm": 8.25, + "learning_rate": 4.496999241311022e-06, + "loss": 2.7216, + "step": 10842 + }, + { + "epoch": 1.1466793570219966, + "grad_norm": 8.3125, + "learning_rate": 4.496896209382243e-06, + "loss": 2.5292, + "step": 10843 + }, + { + "epoch": 1.1467851099830795, + "grad_norm": 8.0625, + "learning_rate": 4.496793168082847e-06, + "loss": 2.5215, + "step": 10844 + }, + { + "epoch": 1.1468908629441623, + "grad_norm": 7.875, + "learning_rate": 4.496690117413317e-06, + "loss": 2.1754, + "step": 10845 + }, + { + "epoch": 1.1469966159052454, + "grad_norm": 8.6875, + "learning_rate": 4.496587057374136e-06, + "loss": 2.4359, + "step": 10846 + }, + { + "epoch": 1.1471023688663282, + "grad_norm": 8.4375, + "learning_rate": 4.496483987965789e-06, + "loss": 2.5951, + "step": 10847 + }, + { + "epoch": 1.1472081218274113, + "grad_norm": 8.375, + "learning_rate": 4.49638090918876e-06, + "loss": 2.4439, + "step": 10848 + }, + { + "epoch": 1.1473138747884941, + "grad_norm": 8.0, + "learning_rate": 4.496277821043532e-06, + "loss": 2.5263, + "step": 10849 + }, + { + "epoch": 1.147419627749577, + "grad_norm": 8.0625, + "learning_rate": 4.4961747235305875e-06, + "loss": 2.6397, + "step": 10850 + }, + { + "epoch": 1.1475253807106598, + "grad_norm": 8.375, + "learning_rate": 4.4960716166504124e-06, + "loss": 2.6666, + "step": 10851 + }, + { + "epoch": 1.1476311336717429, + "grad_norm": 7.96875, + "learning_rate": 4.4959685004034895e-06, + "loss": 2.6458, + "step": 10852 + }, + { + "epoch": 1.1477368866328257, + "grad_norm": 7.9375, + "learning_rate": 4.495865374790302e-06, + "loss": 2.5624, + "step": 10853 + }, + { + "epoch": 1.1478426395939088, + "grad_norm": 8.25, + "learning_rate": 4.495762239811335e-06, + "loss": 2.122, + "step": 10854 + }, + { + "epoch": 1.1479483925549916, + "grad_norm": 8.3125, + "learning_rate": 4.495659095467071e-06, + "loss": 2.683, + "step": 10855 + }, + { + "epoch": 1.1480541455160744, + "grad_norm": 8.3125, + "learning_rate": 4.495555941757996e-06, + "loss": 2.5423, + "step": 10856 + }, + { + "epoch": 1.1481598984771573, + "grad_norm": 8.1875, + "learning_rate": 4.495452778684593e-06, + "loss": 2.7305, + "step": 10857 + }, + { + "epoch": 1.1482656514382403, + "grad_norm": 8.125, + "learning_rate": 4.495349606247347e-06, + "loss": 2.5469, + "step": 10858 + }, + { + "epoch": 1.1483714043993232, + "grad_norm": 7.9375, + "learning_rate": 4.495246424446741e-06, + "loss": 2.4737, + "step": 10859 + }, + { + "epoch": 1.148477157360406, + "grad_norm": 7.9375, + "learning_rate": 4.495143233283259e-06, + "loss": 2.4499, + "step": 10860 + }, + { + "epoch": 1.148582910321489, + "grad_norm": 7.90625, + "learning_rate": 4.495040032757386e-06, + "loss": 2.5634, + "step": 10861 + }, + { + "epoch": 1.148688663282572, + "grad_norm": 9.3125, + "learning_rate": 4.494936822869606e-06, + "loss": 2.4847, + "step": 10862 + }, + { + "epoch": 1.1487944162436547, + "grad_norm": 8.3125, + "learning_rate": 4.494833603620404e-06, + "loss": 2.4363, + "step": 10863 + }, + { + "epoch": 1.1489001692047378, + "grad_norm": 8.0625, + "learning_rate": 4.494730375010264e-06, + "loss": 2.5547, + "step": 10864 + }, + { + "epoch": 1.1490059221658206, + "grad_norm": 8.75, + "learning_rate": 4.494627137039669e-06, + "loss": 2.667, + "step": 10865 + }, + { + "epoch": 1.1491116751269035, + "grad_norm": 8.9375, + "learning_rate": 4.494523889709105e-06, + "loss": 2.5191, + "step": 10866 + }, + { + "epoch": 1.1492174280879865, + "grad_norm": 8.5, + "learning_rate": 4.494420633019057e-06, + "loss": 2.6347, + "step": 10867 + }, + { + "epoch": 1.1493231810490694, + "grad_norm": 8.3125, + "learning_rate": 4.494317366970007e-06, + "loss": 2.5859, + "step": 10868 + }, + { + "epoch": 1.1494289340101522, + "grad_norm": 8.875, + "learning_rate": 4.494214091562442e-06, + "loss": 2.5908, + "step": 10869 + }, + { + "epoch": 1.1495346869712353, + "grad_norm": 10.0, + "learning_rate": 4.4941108067968466e-06, + "loss": 2.7249, + "step": 10870 + }, + { + "epoch": 1.1496404399323181, + "grad_norm": 9.0, + "learning_rate": 4.4940075126737036e-06, + "loss": 2.5652, + "step": 10871 + }, + { + "epoch": 1.149746192893401, + "grad_norm": 7.9375, + "learning_rate": 4.493904209193499e-06, + "loss": 2.7535, + "step": 10872 + }, + { + "epoch": 1.149851945854484, + "grad_norm": 8.4375, + "learning_rate": 4.4938008963567185e-06, + "loss": 2.6207, + "step": 10873 + }, + { + "epoch": 1.1499576988155669, + "grad_norm": 8.5625, + "learning_rate": 4.4936975741638456e-06, + "loss": 2.7201, + "step": 10874 + }, + { + "epoch": 1.1500634517766497, + "grad_norm": 8.6875, + "learning_rate": 4.493594242615365e-06, + "loss": 2.4983, + "step": 10875 + }, + { + "epoch": 1.1501692047377325, + "grad_norm": 8.3125, + "learning_rate": 4.493490901711762e-06, + "loss": 2.6397, + "step": 10876 + }, + { + "epoch": 1.1502749576988156, + "grad_norm": 8.4375, + "learning_rate": 4.49338755145352e-06, + "loss": 2.5861, + "step": 10877 + }, + { + "epoch": 1.1503807106598984, + "grad_norm": 8.9375, + "learning_rate": 4.493284191841129e-06, + "loss": 2.5795, + "step": 10878 + }, + { + "epoch": 1.1504864636209815, + "grad_norm": 8.0625, + "learning_rate": 4.493180822875067e-06, + "loss": 2.71, + "step": 10879 + }, + { + "epoch": 1.1505922165820643, + "grad_norm": 7.78125, + "learning_rate": 4.493077444555825e-06, + "loss": 2.4665, + "step": 10880 + }, + { + "epoch": 1.1506979695431472, + "grad_norm": 8.6875, + "learning_rate": 4.492974056883884e-06, + "loss": 2.689, + "step": 10881 + }, + { + "epoch": 1.15080372250423, + "grad_norm": 8.25, + "learning_rate": 4.492870659859732e-06, + "loss": 2.5434, + "step": 10882 + }, + { + "epoch": 1.150909475465313, + "grad_norm": 8.1875, + "learning_rate": 4.492767253483853e-06, + "loss": 2.4976, + "step": 10883 + }, + { + "epoch": 1.151015228426396, + "grad_norm": 8.375, + "learning_rate": 4.492663837756731e-06, + "loss": 2.5515, + "step": 10884 + }, + { + "epoch": 1.1511209813874788, + "grad_norm": 8.0625, + "learning_rate": 4.492560412678854e-06, + "loss": 2.6434, + "step": 10885 + }, + { + "epoch": 1.1512267343485618, + "grad_norm": 7.5625, + "learning_rate": 4.492456978250706e-06, + "loss": 2.6256, + "step": 10886 + }, + { + "epoch": 1.1513324873096447, + "grad_norm": 7.9375, + "learning_rate": 4.492353534472771e-06, + "loss": 2.3641, + "step": 10887 + }, + { + "epoch": 1.1514382402707275, + "grad_norm": 7.78125, + "learning_rate": 4.492250081345536e-06, + "loss": 2.5052, + "step": 10888 + }, + { + "epoch": 1.1515439932318106, + "grad_norm": 7.84375, + "learning_rate": 4.492146618869487e-06, + "loss": 2.3945, + "step": 10889 + }, + { + "epoch": 1.1516497461928934, + "grad_norm": 8.0625, + "learning_rate": 4.492043147045108e-06, + "loss": 2.5442, + "step": 10890 + }, + { + "epoch": 1.1517554991539762, + "grad_norm": 8.9375, + "learning_rate": 4.491939665872886e-06, + "loss": 2.5359, + "step": 10891 + }, + { + "epoch": 1.1518612521150593, + "grad_norm": 8.1875, + "learning_rate": 4.4918361753533055e-06, + "loss": 2.5841, + "step": 10892 + }, + { + "epoch": 1.1519670050761421, + "grad_norm": 8.0625, + "learning_rate": 4.491732675486851e-06, + "loss": 2.618, + "step": 10893 + }, + { + "epoch": 1.152072758037225, + "grad_norm": 8.4375, + "learning_rate": 4.491629166274012e-06, + "loss": 2.7555, + "step": 10894 + }, + { + "epoch": 1.152178510998308, + "grad_norm": 8.875, + "learning_rate": 4.49152564771527e-06, + "loss": 2.6859, + "step": 10895 + }, + { + "epoch": 1.1522842639593909, + "grad_norm": 8.375, + "learning_rate": 4.491422119811114e-06, + "loss": 2.5706, + "step": 10896 + }, + { + "epoch": 1.1523900169204737, + "grad_norm": 8.1875, + "learning_rate": 4.491318582562028e-06, + "loss": 2.707, + "step": 10897 + }, + { + "epoch": 1.1524957698815568, + "grad_norm": 8.3125, + "learning_rate": 4.491215035968498e-06, + "loss": 2.6108, + "step": 10898 + }, + { + "epoch": 1.1526015228426396, + "grad_norm": 8.625, + "learning_rate": 4.491111480031011e-06, + "loss": 2.7565, + "step": 10899 + }, + { + "epoch": 1.1527072758037225, + "grad_norm": 8.3125, + "learning_rate": 4.491007914750051e-06, + "loss": 2.3818, + "step": 10900 + }, + { + "epoch": 1.1528130287648053, + "grad_norm": 7.875, + "learning_rate": 4.490904340126105e-06, + "loss": 2.4936, + "step": 10901 + }, + { + "epoch": 1.1529187817258884, + "grad_norm": 8.8125, + "learning_rate": 4.490800756159661e-06, + "loss": 2.7894, + "step": 10902 + }, + { + "epoch": 1.1530245346869712, + "grad_norm": 7.75, + "learning_rate": 4.490697162851202e-06, + "loss": 2.6174, + "step": 10903 + }, + { + "epoch": 1.1531302876480543, + "grad_norm": 7.75, + "learning_rate": 4.490593560201215e-06, + "loss": 2.5036, + "step": 10904 + }, + { + "epoch": 1.153236040609137, + "grad_norm": 8.5, + "learning_rate": 4.490489948210187e-06, + "loss": 2.5816, + "step": 10905 + }, + { + "epoch": 1.15334179357022, + "grad_norm": 7.59375, + "learning_rate": 4.490386326878604e-06, + "loss": 2.5508, + "step": 10906 + }, + { + "epoch": 1.1534475465313028, + "grad_norm": 8.1875, + "learning_rate": 4.490282696206952e-06, + "loss": 2.6959, + "step": 10907 + }, + { + "epoch": 1.1535532994923858, + "grad_norm": 7.84375, + "learning_rate": 4.490179056195717e-06, + "loss": 2.5719, + "step": 10908 + }, + { + "epoch": 1.1536590524534687, + "grad_norm": 8.3125, + "learning_rate": 4.490075406845386e-06, + "loss": 2.6429, + "step": 10909 + }, + { + "epoch": 1.1537648054145517, + "grad_norm": 8.125, + "learning_rate": 4.489971748156445e-06, + "loss": 2.3975, + "step": 10910 + }, + { + "epoch": 1.1538705583756346, + "grad_norm": 8.0625, + "learning_rate": 4.48986808012938e-06, + "loss": 2.5246, + "step": 10911 + }, + { + "epoch": 1.1539763113367174, + "grad_norm": 8.125, + "learning_rate": 4.489764402764678e-06, + "loss": 2.6009, + "step": 10912 + }, + { + "epoch": 1.1540820642978002, + "grad_norm": 7.78125, + "learning_rate": 4.489660716062826e-06, + "loss": 2.3653, + "step": 10913 + }, + { + "epoch": 1.1541878172588833, + "grad_norm": 8.1875, + "learning_rate": 4.4895570200243095e-06, + "loss": 2.5685, + "step": 10914 + }, + { + "epoch": 1.1542935702199661, + "grad_norm": 7.96875, + "learning_rate": 4.4894533146496165e-06, + "loss": 2.5752, + "step": 10915 + }, + { + "epoch": 1.154399323181049, + "grad_norm": 8.375, + "learning_rate": 4.489349599939232e-06, + "loss": 2.5265, + "step": 10916 + }, + { + "epoch": 1.154505076142132, + "grad_norm": 7.6875, + "learning_rate": 4.489245875893643e-06, + "loss": 2.3994, + "step": 10917 + }, + { + "epoch": 1.1546108291032149, + "grad_norm": 8.1875, + "learning_rate": 4.489142142513337e-06, + "loss": 2.5459, + "step": 10918 + }, + { + "epoch": 1.1547165820642977, + "grad_norm": 8.25, + "learning_rate": 4.489038399798801e-06, + "loss": 2.6628, + "step": 10919 + }, + { + "epoch": 1.1548223350253808, + "grad_norm": 8.1875, + "learning_rate": 4.488934647750521e-06, + "loss": 2.5556, + "step": 10920 + }, + { + "epoch": 1.1549280879864636, + "grad_norm": 9.9375, + "learning_rate": 4.488830886368983e-06, + "loss": 2.9889, + "step": 10921 + }, + { + "epoch": 1.1550338409475465, + "grad_norm": 8.4375, + "learning_rate": 4.488727115654676e-06, + "loss": 2.5442, + "step": 10922 + }, + { + "epoch": 1.1551395939086295, + "grad_norm": 8.375, + "learning_rate": 4.488623335608087e-06, + "loss": 2.7915, + "step": 10923 + }, + { + "epoch": 1.1552453468697124, + "grad_norm": 8.0625, + "learning_rate": 4.488519546229701e-06, + "loss": 2.8035, + "step": 10924 + }, + { + "epoch": 1.1553510998307952, + "grad_norm": 8.4375, + "learning_rate": 4.488415747520006e-06, + "loss": 2.671, + "step": 10925 + }, + { + "epoch": 1.1554568527918783, + "grad_norm": 8.4375, + "learning_rate": 4.488311939479489e-06, + "loss": 2.6433, + "step": 10926 + }, + { + "epoch": 1.155562605752961, + "grad_norm": 8.9375, + "learning_rate": 4.488208122108638e-06, + "loss": 1.9617, + "step": 10927 + }, + { + "epoch": 1.155668358714044, + "grad_norm": 8.375, + "learning_rate": 4.488104295407939e-06, + "loss": 2.2975, + "step": 10928 + }, + { + "epoch": 1.155774111675127, + "grad_norm": 8.0, + "learning_rate": 4.488000459377879e-06, + "loss": 2.4872, + "step": 10929 + }, + { + "epoch": 1.1558798646362098, + "grad_norm": 7.9375, + "learning_rate": 4.487896614018947e-06, + "loss": 2.5011, + "step": 10930 + }, + { + "epoch": 1.1559856175972927, + "grad_norm": 8.9375, + "learning_rate": 4.487792759331629e-06, + "loss": 2.7018, + "step": 10931 + }, + { + "epoch": 1.1560913705583755, + "grad_norm": 8.375, + "learning_rate": 4.487688895316412e-06, + "loss": 2.5346, + "step": 10932 + }, + { + "epoch": 1.1561971235194586, + "grad_norm": 8.4375, + "learning_rate": 4.4875850219737844e-06, + "loss": 2.607, + "step": 10933 + }, + { + "epoch": 1.1563028764805414, + "grad_norm": 8.5625, + "learning_rate": 4.487481139304233e-06, + "loss": 2.4172, + "step": 10934 + }, + { + "epoch": 1.1564086294416245, + "grad_norm": 8.4375, + "learning_rate": 4.487377247308245e-06, + "loss": 2.5844, + "step": 10935 + }, + { + "epoch": 1.1565143824027073, + "grad_norm": 8.0, + "learning_rate": 4.487273345986309e-06, + "loss": 2.528, + "step": 10936 + }, + { + "epoch": 1.1566201353637902, + "grad_norm": 8.4375, + "learning_rate": 4.4871694353389125e-06, + "loss": 2.7618, + "step": 10937 + }, + { + "epoch": 1.156725888324873, + "grad_norm": 8.25, + "learning_rate": 4.487065515366542e-06, + "loss": 2.5639, + "step": 10938 + }, + { + "epoch": 1.156831641285956, + "grad_norm": 8.3125, + "learning_rate": 4.486961586069684e-06, + "loss": 2.6908, + "step": 10939 + }, + { + "epoch": 1.156937394247039, + "grad_norm": 8.3125, + "learning_rate": 4.4868576474488304e-06, + "loss": 2.5886, + "step": 10940 + }, + { + "epoch": 1.1570431472081217, + "grad_norm": 8.375, + "learning_rate": 4.4867536995044655e-06, + "loss": 2.5451, + "step": 10941 + }, + { + "epoch": 1.1571489001692048, + "grad_norm": 8.125, + "learning_rate": 4.486649742237078e-06, + "loss": 2.3789, + "step": 10942 + }, + { + "epoch": 1.1572546531302876, + "grad_norm": 8.375, + "learning_rate": 4.486545775647157e-06, + "loss": 2.3844, + "step": 10943 + }, + { + "epoch": 1.1573604060913705, + "grad_norm": 7.875, + "learning_rate": 4.486441799735187e-06, + "loss": 2.558, + "step": 10944 + }, + { + "epoch": 1.1574661590524535, + "grad_norm": 7.96875, + "learning_rate": 4.48633781450166e-06, + "loss": 2.5398, + "step": 10945 + }, + { + "epoch": 1.1575719120135364, + "grad_norm": 8.1875, + "learning_rate": 4.48623381994706e-06, + "loss": 2.5463, + "step": 10946 + }, + { + "epoch": 1.1576776649746192, + "grad_norm": 8.0, + "learning_rate": 4.486129816071879e-06, + "loss": 2.5958, + "step": 10947 + }, + { + "epoch": 1.1577834179357023, + "grad_norm": 7.96875, + "learning_rate": 4.486025802876602e-06, + "loss": 2.696, + "step": 10948 + }, + { + "epoch": 1.1578891708967851, + "grad_norm": 8.75, + "learning_rate": 4.485921780361718e-06, + "loss": 2.6534, + "step": 10949 + }, + { + "epoch": 1.157994923857868, + "grad_norm": 8.5625, + "learning_rate": 4.485817748527716e-06, + "loss": 2.6557, + "step": 10950 + }, + { + "epoch": 1.158100676818951, + "grad_norm": 8.375, + "learning_rate": 4.485713707375083e-06, + "loss": 2.6247, + "step": 10951 + }, + { + "epoch": 1.1582064297800339, + "grad_norm": 8.25, + "learning_rate": 4.485609656904309e-06, + "loss": 2.6859, + "step": 10952 + }, + { + "epoch": 1.1583121827411167, + "grad_norm": 7.875, + "learning_rate": 4.485505597115879e-06, + "loss": 2.6135, + "step": 10953 + }, + { + "epoch": 1.1584179357021998, + "grad_norm": 8.375, + "learning_rate": 4.4854015280102846e-06, + "loss": 2.2868, + "step": 10954 + }, + { + "epoch": 1.1585236886632826, + "grad_norm": 8.25, + "learning_rate": 4.4852974495880116e-06, + "loss": 2.6075, + "step": 10955 + }, + { + "epoch": 1.1586294416243654, + "grad_norm": 8.4375, + "learning_rate": 4.485193361849551e-06, + "loss": 2.5058, + "step": 10956 + }, + { + "epoch": 1.1587351945854483, + "grad_norm": 8.4375, + "learning_rate": 4.485089264795389e-06, + "loss": 2.5407, + "step": 10957 + }, + { + "epoch": 1.1588409475465313, + "grad_norm": 14.9375, + "learning_rate": 4.484985158426015e-06, + "loss": 2.4312, + "step": 10958 + }, + { + "epoch": 1.1589467005076142, + "grad_norm": 8.1875, + "learning_rate": 4.4848810427419175e-06, + "loss": 2.4541, + "step": 10959 + }, + { + "epoch": 1.1590524534686972, + "grad_norm": 8.375, + "learning_rate": 4.4847769177435855e-06, + "loss": 2.6297, + "step": 10960 + }, + { + "epoch": 1.15915820642978, + "grad_norm": 8.125, + "learning_rate": 4.484672783431507e-06, + "loss": 2.4519, + "step": 10961 + }, + { + "epoch": 1.159263959390863, + "grad_norm": 8.125, + "learning_rate": 4.484568639806171e-06, + "loss": 2.4888, + "step": 10962 + }, + { + "epoch": 1.1593697123519457, + "grad_norm": 8.375, + "learning_rate": 4.484464486868066e-06, + "loss": 2.4739, + "step": 10963 + }, + { + "epoch": 1.1594754653130288, + "grad_norm": 7.96875, + "learning_rate": 4.4843603246176804e-06, + "loss": 2.4899, + "step": 10964 + }, + { + "epoch": 1.1595812182741116, + "grad_norm": 8.0625, + "learning_rate": 4.484256153055503e-06, + "loss": 2.5146, + "step": 10965 + }, + { + "epoch": 1.1596869712351947, + "grad_norm": 7.75, + "learning_rate": 4.484151972182024e-06, + "loss": 2.7483, + "step": 10966 + }, + { + "epoch": 1.1597927241962775, + "grad_norm": 7.71875, + "learning_rate": 4.4840477819977305e-06, + "loss": 2.309, + "step": 10967 + }, + { + "epoch": 1.1598984771573604, + "grad_norm": 8.4375, + "learning_rate": 4.4839435825031125e-06, + "loss": 2.4943, + "step": 10968 + }, + { + "epoch": 1.1600042301184432, + "grad_norm": 7.90625, + "learning_rate": 4.483839373698659e-06, + "loss": 2.2384, + "step": 10969 + }, + { + "epoch": 1.1601099830795263, + "grad_norm": 8.625, + "learning_rate": 4.483735155584858e-06, + "loss": 2.7311, + "step": 10970 + }, + { + "epoch": 1.1602157360406091, + "grad_norm": 8.125, + "learning_rate": 4.4836309281622e-06, + "loss": 2.2931, + "step": 10971 + }, + { + "epoch": 1.160321489001692, + "grad_norm": 8.75, + "learning_rate": 4.483526691431173e-06, + "loss": 2.6187, + "step": 10972 + }, + { + "epoch": 1.160427241962775, + "grad_norm": 8.625, + "learning_rate": 4.483422445392266e-06, + "loss": 2.5124, + "step": 10973 + }, + { + "epoch": 1.1605329949238579, + "grad_norm": 8.125, + "learning_rate": 4.4833181900459685e-06, + "loss": 2.6054, + "step": 10974 + }, + { + "epoch": 1.1606387478849407, + "grad_norm": 8.0, + "learning_rate": 4.483213925392771e-06, + "loss": 2.5824, + "step": 10975 + }, + { + "epoch": 1.1607445008460238, + "grad_norm": 8.5, + "learning_rate": 4.4831096514331615e-06, + "loss": 2.8985, + "step": 10976 + }, + { + "epoch": 1.1608502538071066, + "grad_norm": 8.4375, + "learning_rate": 4.483005368167629e-06, + "loss": 2.5851, + "step": 10977 + }, + { + "epoch": 1.1609560067681894, + "grad_norm": 7.96875, + "learning_rate": 4.482901075596663e-06, + "loss": 2.3427, + "step": 10978 + }, + { + "epoch": 1.1610617597292725, + "grad_norm": 7.78125, + "learning_rate": 4.482796773720754e-06, + "loss": 2.5187, + "step": 10979 + }, + { + "epoch": 1.1611675126903553, + "grad_norm": 8.25, + "learning_rate": 4.48269246254039e-06, + "loss": 2.5497, + "step": 10980 + }, + { + "epoch": 1.1612732656514382, + "grad_norm": 8.125, + "learning_rate": 4.482588142056062e-06, + "loss": 2.392, + "step": 10981 + }, + { + "epoch": 1.1613790186125212, + "grad_norm": 8.6875, + "learning_rate": 4.482483812268258e-06, + "loss": 2.5999, + "step": 10982 + }, + { + "epoch": 1.161484771573604, + "grad_norm": 7.78125, + "learning_rate": 4.482379473177469e-06, + "loss": 2.5416, + "step": 10983 + }, + { + "epoch": 1.161590524534687, + "grad_norm": 8.1875, + "learning_rate": 4.4822751247841825e-06, + "loss": 2.4425, + "step": 10984 + }, + { + "epoch": 1.16169627749577, + "grad_norm": 8.25, + "learning_rate": 4.482170767088891e-06, + "loss": 2.6116, + "step": 10985 + }, + { + "epoch": 1.1618020304568528, + "grad_norm": 8.9375, + "learning_rate": 4.4820664000920826e-06, + "loss": 2.5023, + "step": 10986 + }, + { + "epoch": 1.1619077834179357, + "grad_norm": 8.5, + "learning_rate": 4.481962023794246e-06, + "loss": 2.5714, + "step": 10987 + }, + { + "epoch": 1.1620135363790185, + "grad_norm": 8.125, + "learning_rate": 4.481857638195873e-06, + "loss": 2.6228, + "step": 10988 + }, + { + "epoch": 1.1621192893401016, + "grad_norm": 8.5625, + "learning_rate": 4.481753243297453e-06, + "loss": 2.7283, + "step": 10989 + }, + { + "epoch": 1.1622250423011844, + "grad_norm": 8.5, + "learning_rate": 4.481648839099475e-06, + "loss": 2.6742, + "step": 10990 + }, + { + "epoch": 1.1623307952622675, + "grad_norm": 7.875, + "learning_rate": 4.48154442560243e-06, + "loss": 2.3396, + "step": 10991 + }, + { + "epoch": 1.1624365482233503, + "grad_norm": 8.0, + "learning_rate": 4.4814400028068065e-06, + "loss": 2.4874, + "step": 10992 + }, + { + "epoch": 1.1625423011844331, + "grad_norm": 8.1875, + "learning_rate": 4.4813355707130965e-06, + "loss": 2.578, + "step": 10993 + }, + { + "epoch": 1.162648054145516, + "grad_norm": 8.4375, + "learning_rate": 4.481231129321789e-06, + "loss": 2.6606, + "step": 10994 + }, + { + "epoch": 1.162753807106599, + "grad_norm": 8.25, + "learning_rate": 4.481126678633373e-06, + "loss": 2.4479, + "step": 10995 + }, + { + "epoch": 1.1628595600676819, + "grad_norm": 7.8125, + "learning_rate": 4.481022218648341e-06, + "loss": 2.701, + "step": 10996 + }, + { + "epoch": 1.1629653130287647, + "grad_norm": 8.25, + "learning_rate": 4.480917749367181e-06, + "loss": 2.6373, + "step": 10997 + }, + { + "epoch": 1.1630710659898478, + "grad_norm": 7.90625, + "learning_rate": 4.480813270790385e-06, + "loss": 2.4004, + "step": 10998 + }, + { + "epoch": 1.1631768189509306, + "grad_norm": 8.1875, + "learning_rate": 4.480708782918441e-06, + "loss": 2.6041, + "step": 10999 + }, + { + "epoch": 1.1632825719120135, + "grad_norm": 8.375, + "learning_rate": 4.480604285751843e-06, + "loss": 2.7152, + "step": 11000 + }, + { + "epoch": 1.1633883248730965, + "grad_norm": 8.375, + "learning_rate": 4.480499779291077e-06, + "loss": 2.6576, + "step": 11001 + }, + { + "epoch": 1.1634940778341794, + "grad_norm": 8.4375, + "learning_rate": 4.480395263536637e-06, + "loss": 2.5694, + "step": 11002 + }, + { + "epoch": 1.1635998307952622, + "grad_norm": 8.1875, + "learning_rate": 4.480290738489011e-06, + "loss": 2.5492, + "step": 11003 + }, + { + "epoch": 1.1637055837563453, + "grad_norm": 8.1875, + "learning_rate": 4.480186204148691e-06, + "loss": 2.4402, + "step": 11004 + }, + { + "epoch": 1.163811336717428, + "grad_norm": 8.25, + "learning_rate": 4.480081660516167e-06, + "loss": 2.5597, + "step": 11005 + }, + { + "epoch": 1.163917089678511, + "grad_norm": 7.78125, + "learning_rate": 4.479977107591929e-06, + "loss": 2.3489, + "step": 11006 + }, + { + "epoch": 1.164022842639594, + "grad_norm": 10.0625, + "learning_rate": 4.479872545376469e-06, + "loss": 2.6932, + "step": 11007 + }, + { + "epoch": 1.1641285956006768, + "grad_norm": 9.3125, + "learning_rate": 4.479767973870277e-06, + "loss": 2.7655, + "step": 11008 + }, + { + "epoch": 1.1642343485617597, + "grad_norm": 8.3125, + "learning_rate": 4.479663393073843e-06, + "loss": 2.5283, + "step": 11009 + }, + { + "epoch": 1.1643401015228427, + "grad_norm": 8.625, + "learning_rate": 4.479558802987658e-06, + "loss": 2.7582, + "step": 11010 + }, + { + "epoch": 1.1644458544839256, + "grad_norm": 8.375, + "learning_rate": 4.4794542036122134e-06, + "loss": 2.7434, + "step": 11011 + }, + { + "epoch": 1.1645516074450084, + "grad_norm": 8.0, + "learning_rate": 4.479349594948001e-06, + "loss": 2.6679, + "step": 11012 + }, + { + "epoch": 1.1646573604060912, + "grad_norm": 8.0, + "learning_rate": 4.479244976995509e-06, + "loss": 2.495, + "step": 11013 + }, + { + "epoch": 1.1647631133671743, + "grad_norm": 8.4375, + "learning_rate": 4.4791403497552296e-06, + "loss": 2.8211, + "step": 11014 + }, + { + "epoch": 1.1648688663282571, + "grad_norm": 7.90625, + "learning_rate": 4.4790357132276554e-06, + "loss": 2.5341, + "step": 11015 + }, + { + "epoch": 1.1649746192893402, + "grad_norm": 8.1875, + "learning_rate": 4.478931067413276e-06, + "loss": 2.6625, + "step": 11016 + }, + { + "epoch": 1.165080372250423, + "grad_norm": 7.625, + "learning_rate": 4.478826412312581e-06, + "loss": 2.6802, + "step": 11017 + }, + { + "epoch": 1.1651861252115059, + "grad_norm": 7.875, + "learning_rate": 4.478721747926064e-06, + "loss": 2.156, + "step": 11018 + }, + { + "epoch": 1.1652918781725887, + "grad_norm": 8.0, + "learning_rate": 4.478617074254215e-06, + "loss": 2.4835, + "step": 11019 + }, + { + "epoch": 1.1653976311336718, + "grad_norm": 8.0625, + "learning_rate": 4.478512391297525e-06, + "loss": 2.4305, + "step": 11020 + }, + { + "epoch": 1.1655033840947546, + "grad_norm": 7.875, + "learning_rate": 4.4784076990564855e-06, + "loss": 2.6019, + "step": 11021 + }, + { + "epoch": 1.1656091370558377, + "grad_norm": 8.625, + "learning_rate": 4.478302997531588e-06, + "loss": 2.8135, + "step": 11022 + }, + { + "epoch": 1.1657148900169205, + "grad_norm": 8.4375, + "learning_rate": 4.4781982867233235e-06, + "loss": 2.6387, + "step": 11023 + }, + { + "epoch": 1.1658206429780034, + "grad_norm": 8.75, + "learning_rate": 4.478093566632183e-06, + "loss": 2.4154, + "step": 11024 + }, + { + "epoch": 1.1659263959390862, + "grad_norm": 10.625, + "learning_rate": 4.47798883725866e-06, + "loss": 2.453, + "step": 11025 + }, + { + "epoch": 1.1660321489001693, + "grad_norm": 8.4375, + "learning_rate": 4.477884098603244e-06, + "loss": 2.5979, + "step": 11026 + }, + { + "epoch": 1.166137901861252, + "grad_norm": 8.625, + "learning_rate": 4.477779350666426e-06, + "loss": 2.6922, + "step": 11027 + }, + { + "epoch": 1.166243654822335, + "grad_norm": 7.875, + "learning_rate": 4.477674593448698e-06, + "loss": 2.6737, + "step": 11028 + }, + { + "epoch": 1.166349407783418, + "grad_norm": 8.25, + "learning_rate": 4.477569826950553e-06, + "loss": 2.5315, + "step": 11029 + }, + { + "epoch": 1.1664551607445008, + "grad_norm": 8.1875, + "learning_rate": 4.477465051172481e-06, + "loss": 2.3734, + "step": 11030 + }, + { + "epoch": 1.1665609137055837, + "grad_norm": 8.5625, + "learning_rate": 4.477360266114974e-06, + "loss": 2.4604, + "step": 11031 + }, + { + "epoch": 1.1666666666666667, + "grad_norm": 9.0, + "learning_rate": 4.477255471778524e-06, + "loss": 2.05, + "step": 11032 + }, + { + "epoch": 1.1667724196277496, + "grad_norm": 8.0, + "learning_rate": 4.477150668163624e-06, + "loss": 2.6139, + "step": 11033 + }, + { + "epoch": 1.1668781725888324, + "grad_norm": 8.1875, + "learning_rate": 4.477045855270764e-06, + "loss": 2.7101, + "step": 11034 + }, + { + "epoch": 1.1669839255499155, + "grad_norm": 8.875, + "learning_rate": 4.476941033100436e-06, + "loss": 2.4587, + "step": 11035 + }, + { + "epoch": 1.1670896785109983, + "grad_norm": 7.84375, + "learning_rate": 4.476836201653132e-06, + "loss": 2.3327, + "step": 11036 + }, + { + "epoch": 1.1671954314720812, + "grad_norm": 9.125, + "learning_rate": 4.476731360929345e-06, + "loss": 2.7236, + "step": 11037 + }, + { + "epoch": 1.1673011844331642, + "grad_norm": 8.1875, + "learning_rate": 4.476626510929566e-06, + "loss": 2.675, + "step": 11038 + }, + { + "epoch": 1.167406937394247, + "grad_norm": 8.0, + "learning_rate": 4.4765216516542865e-06, + "loss": 2.3255, + "step": 11039 + }, + { + "epoch": 1.16751269035533, + "grad_norm": 8.375, + "learning_rate": 4.476416783103999e-06, + "loss": 2.7764, + "step": 11040 + }, + { + "epoch": 1.167618443316413, + "grad_norm": 8.125, + "learning_rate": 4.4763119052791974e-06, + "loss": 2.5118, + "step": 11041 + }, + { + "epoch": 1.1677241962774958, + "grad_norm": 8.1875, + "learning_rate": 4.476207018180372e-06, + "loss": 2.5854, + "step": 11042 + }, + { + "epoch": 1.1678299492385786, + "grad_norm": 7.6875, + "learning_rate": 4.476102121808014e-06, + "loss": 2.3714, + "step": 11043 + }, + { + "epoch": 1.1679357021996615, + "grad_norm": 8.4375, + "learning_rate": 4.475997216162618e-06, + "loss": 2.721, + "step": 11044 + }, + { + "epoch": 1.1680414551607445, + "grad_norm": 8.0, + "learning_rate": 4.4758923012446745e-06, + "loss": 2.4828, + "step": 11045 + }, + { + "epoch": 1.1681472081218274, + "grad_norm": 7.90625, + "learning_rate": 4.4757873770546775e-06, + "loss": 2.4104, + "step": 11046 + }, + { + "epoch": 1.1682529610829104, + "grad_norm": 7.96875, + "learning_rate": 4.475682443593118e-06, + "loss": 2.4822, + "step": 11047 + }, + { + "epoch": 1.1683587140439933, + "grad_norm": 9.0, + "learning_rate": 4.475577500860488e-06, + "loss": 2.7752, + "step": 11048 + }, + { + "epoch": 1.1684644670050761, + "grad_norm": 11.625, + "learning_rate": 4.475472548857282e-06, + "loss": 2.4027, + "step": 11049 + }, + { + "epoch": 1.168570219966159, + "grad_norm": 8.1875, + "learning_rate": 4.475367587583991e-06, + "loss": 2.5247, + "step": 11050 + }, + { + "epoch": 1.168675972927242, + "grad_norm": 8.6875, + "learning_rate": 4.475262617041107e-06, + "loss": 2.6238, + "step": 11051 + }, + { + "epoch": 1.1687817258883249, + "grad_norm": 8.25, + "learning_rate": 4.4751576372291244e-06, + "loss": 2.4785, + "step": 11052 + }, + { + "epoch": 1.1688874788494077, + "grad_norm": 8.125, + "learning_rate": 4.475052648148535e-06, + "loss": 2.5291, + "step": 11053 + }, + { + "epoch": 1.1689932318104908, + "grad_norm": 8.125, + "learning_rate": 4.47494764979983e-06, + "loss": 2.442, + "step": 11054 + }, + { + "epoch": 1.1690989847715736, + "grad_norm": 9.375, + "learning_rate": 4.474842642183504e-06, + "loss": 2.2492, + "step": 11055 + }, + { + "epoch": 1.1692047377326564, + "grad_norm": 8.0, + "learning_rate": 4.47473762530005e-06, + "loss": 2.6498, + "step": 11056 + }, + { + "epoch": 1.1693104906937395, + "grad_norm": 8.5, + "learning_rate": 4.474632599149958e-06, + "loss": 2.5487, + "step": 11057 + }, + { + "epoch": 1.1694162436548223, + "grad_norm": 8.6875, + "learning_rate": 4.474527563733725e-06, + "loss": 2.7897, + "step": 11058 + }, + { + "epoch": 1.1695219966159052, + "grad_norm": 8.3125, + "learning_rate": 4.47442251905184e-06, + "loss": 2.6544, + "step": 11059 + }, + { + "epoch": 1.1696277495769882, + "grad_norm": 8.5, + "learning_rate": 4.474317465104799e-06, + "loss": 2.5615, + "step": 11060 + }, + { + "epoch": 1.169733502538071, + "grad_norm": 8.0625, + "learning_rate": 4.474212401893094e-06, + "loss": 2.4357, + "step": 11061 + }, + { + "epoch": 1.169839255499154, + "grad_norm": 8.4375, + "learning_rate": 4.474107329417215e-06, + "loss": 2.6403, + "step": 11062 + }, + { + "epoch": 1.169945008460237, + "grad_norm": 8.125, + "learning_rate": 4.4740022476776595e-06, + "loss": 2.7125, + "step": 11063 + }, + { + "epoch": 1.1700507614213198, + "grad_norm": 8.1875, + "learning_rate": 4.473897156674919e-06, + "loss": 2.7692, + "step": 11064 + }, + { + "epoch": 1.1701565143824026, + "grad_norm": 8.0625, + "learning_rate": 4.473792056409486e-06, + "loss": 2.4633, + "step": 11065 + }, + { + "epoch": 1.1702622673434857, + "grad_norm": 8.0, + "learning_rate": 4.4736869468818535e-06, + "loss": 2.615, + "step": 11066 + }, + { + "epoch": 1.1703680203045685, + "grad_norm": 8.0625, + "learning_rate": 4.473581828092516e-06, + "loss": 2.5301, + "step": 11067 + }, + { + "epoch": 1.1704737732656514, + "grad_norm": 7.8125, + "learning_rate": 4.4734767000419665e-06, + "loss": 2.5296, + "step": 11068 + }, + { + "epoch": 1.1705795262267342, + "grad_norm": 8.5625, + "learning_rate": 4.473371562730697e-06, + "loss": 2.6466, + "step": 11069 + }, + { + "epoch": 1.1706852791878173, + "grad_norm": 8.0, + "learning_rate": 4.473266416159203e-06, + "loss": 2.318, + "step": 11070 + }, + { + "epoch": 1.1707910321489001, + "grad_norm": 7.8125, + "learning_rate": 4.4731612603279765e-06, + "loss": 2.514, + "step": 11071 + }, + { + "epoch": 1.1708967851099832, + "grad_norm": 7.875, + "learning_rate": 4.473056095237511e-06, + "loss": 2.5199, + "step": 11072 + }, + { + "epoch": 1.171002538071066, + "grad_norm": 8.375, + "learning_rate": 4.4729509208883e-06, + "loss": 2.6579, + "step": 11073 + }, + { + "epoch": 1.1711082910321489, + "grad_norm": 7.875, + "learning_rate": 4.472845737280837e-06, + "loss": 2.4851, + "step": 11074 + }, + { + "epoch": 1.1712140439932317, + "grad_norm": 7.65625, + "learning_rate": 4.472740544415616e-06, + "loss": 2.3692, + "step": 11075 + }, + { + "epoch": 1.1713197969543148, + "grad_norm": 8.125, + "learning_rate": 4.472635342293131e-06, + "loss": 2.6334, + "step": 11076 + }, + { + "epoch": 1.1714255499153976, + "grad_norm": 8.1875, + "learning_rate": 4.472530130913875e-06, + "loss": 2.4793, + "step": 11077 + }, + { + "epoch": 1.1715313028764807, + "grad_norm": 8.1875, + "learning_rate": 4.4724249102783415e-06, + "loss": 2.5406, + "step": 11078 + }, + { + "epoch": 1.1716370558375635, + "grad_norm": 8.625, + "learning_rate": 4.472319680387024e-06, + "loss": 2.6547, + "step": 11079 + }, + { + "epoch": 1.1717428087986463, + "grad_norm": 8.375, + "learning_rate": 4.472214441240418e-06, + "loss": 2.7321, + "step": 11080 + }, + { + "epoch": 1.1718485617597292, + "grad_norm": 8.0, + "learning_rate": 4.472109192839016e-06, + "loss": 2.449, + "step": 11081 + }, + { + "epoch": 1.1719543147208122, + "grad_norm": 8.125, + "learning_rate": 4.472003935183311e-06, + "loss": 2.7167, + "step": 11082 + }, + { + "epoch": 1.172060067681895, + "grad_norm": 8.25, + "learning_rate": 4.471898668273799e-06, + "loss": 2.3867, + "step": 11083 + }, + { + "epoch": 1.172165820642978, + "grad_norm": 7.96875, + "learning_rate": 4.471793392110974e-06, + "loss": 2.582, + "step": 11084 + }, + { + "epoch": 1.172271573604061, + "grad_norm": 8.1875, + "learning_rate": 4.471688106695327e-06, + "loss": 2.3919, + "step": 11085 + }, + { + "epoch": 1.1723773265651438, + "grad_norm": 8.9375, + "learning_rate": 4.471582812027355e-06, + "loss": 2.7983, + "step": 11086 + }, + { + "epoch": 1.1724830795262267, + "grad_norm": 8.125, + "learning_rate": 4.4714775081075525e-06, + "loss": 2.3934, + "step": 11087 + }, + { + "epoch": 1.1725888324873097, + "grad_norm": 8.9375, + "learning_rate": 4.471372194936411e-06, + "loss": 2.4612, + "step": 11088 + }, + { + "epoch": 1.1726945854483926, + "grad_norm": 8.125, + "learning_rate": 4.471266872514426e-06, + "loss": 2.5371, + "step": 11089 + }, + { + "epoch": 1.1728003384094754, + "grad_norm": 8.375, + "learning_rate": 4.471161540842092e-06, + "loss": 2.6271, + "step": 11090 + }, + { + "epoch": 1.1729060913705585, + "grad_norm": 7.9375, + "learning_rate": 4.4710561999199035e-06, + "loss": 2.5275, + "step": 11091 + }, + { + "epoch": 1.1730118443316413, + "grad_norm": 8.4375, + "learning_rate": 4.4709508497483536e-06, + "loss": 2.6749, + "step": 11092 + }, + { + "epoch": 1.1731175972927241, + "grad_norm": 8.625, + "learning_rate": 4.470845490327938e-06, + "loss": 2.884, + "step": 11093 + }, + { + "epoch": 1.1732233502538072, + "grad_norm": 8.25, + "learning_rate": 4.4707401216591504e-06, + "loss": 2.7117, + "step": 11094 + }, + { + "epoch": 1.17332910321489, + "grad_norm": 7.75, + "learning_rate": 4.470634743742486e-06, + "loss": 2.3884, + "step": 11095 + }, + { + "epoch": 1.1734348561759729, + "grad_norm": 8.0625, + "learning_rate": 4.470529356578438e-06, + "loss": 2.4178, + "step": 11096 + }, + { + "epoch": 1.173540609137056, + "grad_norm": 7.875, + "learning_rate": 4.470423960167502e-06, + "loss": 2.543, + "step": 11097 + }, + { + "epoch": 1.1736463620981388, + "grad_norm": 7.375, + "learning_rate": 4.470318554510171e-06, + "loss": 2.3464, + "step": 11098 + }, + { + "epoch": 1.1737521150592216, + "grad_norm": 7.8125, + "learning_rate": 4.470213139606943e-06, + "loss": 2.5966, + "step": 11099 + }, + { + "epoch": 1.1738578680203045, + "grad_norm": 7.9375, + "learning_rate": 4.470107715458308e-06, + "loss": 2.4167, + "step": 11100 + }, + { + "epoch": 1.1739636209813875, + "grad_norm": 8.4375, + "learning_rate": 4.470002282064765e-06, + "loss": 2.6571, + "step": 11101 + }, + { + "epoch": 1.1740693739424704, + "grad_norm": 7.8125, + "learning_rate": 4.4698968394268075e-06, + "loss": 2.4493, + "step": 11102 + }, + { + "epoch": 1.1741751269035534, + "grad_norm": 8.4375, + "learning_rate": 4.469791387544929e-06, + "loss": 2.482, + "step": 11103 + }, + { + "epoch": 1.1742808798646363, + "grad_norm": 8.25, + "learning_rate": 4.469685926419625e-06, + "loss": 2.2802, + "step": 11104 + }, + { + "epoch": 1.174386632825719, + "grad_norm": 8.25, + "learning_rate": 4.46958045605139e-06, + "loss": 2.5396, + "step": 11105 + }, + { + "epoch": 1.174492385786802, + "grad_norm": 7.6875, + "learning_rate": 4.4694749764407206e-06, + "loss": 2.5669, + "step": 11106 + }, + { + "epoch": 1.174598138747885, + "grad_norm": 8.0, + "learning_rate": 4.46936948758811e-06, + "loss": 2.4654, + "step": 11107 + }, + { + "epoch": 1.1747038917089678, + "grad_norm": 8.9375, + "learning_rate": 4.469263989494053e-06, + "loss": 2.6826, + "step": 11108 + }, + { + "epoch": 1.1748096446700507, + "grad_norm": 8.0625, + "learning_rate": 4.469158482159047e-06, + "loss": 2.6133, + "step": 11109 + }, + { + "epoch": 1.1749153976311337, + "grad_norm": 8.5, + "learning_rate": 4.469052965583585e-06, + "loss": 2.7832, + "step": 11110 + }, + { + "epoch": 1.1750211505922166, + "grad_norm": 22.0, + "learning_rate": 4.468947439768163e-06, + "loss": 2.4372, + "step": 11111 + }, + { + "epoch": 1.1751269035532994, + "grad_norm": 9.6875, + "learning_rate": 4.468841904713276e-06, + "loss": 2.6868, + "step": 11112 + }, + { + "epoch": 1.1752326565143825, + "grad_norm": 8.4375, + "learning_rate": 4.468736360419418e-06, + "loss": 2.634, + "step": 11113 + }, + { + "epoch": 1.1753384094754653, + "grad_norm": 7.78125, + "learning_rate": 4.468630806887087e-06, + "loss": 2.6338, + "step": 11114 + }, + { + "epoch": 1.1754441624365481, + "grad_norm": 8.5, + "learning_rate": 4.4685252441167755e-06, + "loss": 2.5824, + "step": 11115 + }, + { + "epoch": 1.1755499153976312, + "grad_norm": 8.1875, + "learning_rate": 4.4684196721089814e-06, + "loss": 2.7544, + "step": 11116 + }, + { + "epoch": 1.175655668358714, + "grad_norm": 7.875, + "learning_rate": 4.468314090864197e-06, + "loss": 2.5635, + "step": 11117 + }, + { + "epoch": 1.1757614213197969, + "grad_norm": 8.25, + "learning_rate": 4.468208500382921e-06, + "loss": 2.406, + "step": 11118 + }, + { + "epoch": 1.17586717428088, + "grad_norm": 8.125, + "learning_rate": 4.468102900665647e-06, + "loss": 2.6252, + "step": 11119 + }, + { + "epoch": 1.1759729272419628, + "grad_norm": 8.0, + "learning_rate": 4.467997291712871e-06, + "loss": 2.1986, + "step": 11120 + }, + { + "epoch": 1.1760786802030456, + "grad_norm": 7.71875, + "learning_rate": 4.46789167352509e-06, + "loss": 2.5082, + "step": 11121 + }, + { + "epoch": 1.1761844331641287, + "grad_norm": 8.5625, + "learning_rate": 4.467786046102797e-06, + "loss": 2.3792, + "step": 11122 + }, + { + "epoch": 1.1762901861252115, + "grad_norm": 8.75, + "learning_rate": 4.467680409446489e-06, + "loss": 2.7804, + "step": 11123 + }, + { + "epoch": 1.1763959390862944, + "grad_norm": 8.0625, + "learning_rate": 4.467574763556661e-06, + "loss": 2.334, + "step": 11124 + }, + { + "epoch": 1.1765016920473774, + "grad_norm": 8.5625, + "learning_rate": 4.46746910843381e-06, + "loss": 2.6666, + "step": 11125 + }, + { + "epoch": 1.1766074450084603, + "grad_norm": 7.84375, + "learning_rate": 4.467363444078431e-06, + "loss": 2.514, + "step": 11126 + }, + { + "epoch": 1.176713197969543, + "grad_norm": 8.3125, + "learning_rate": 4.467257770491019e-06, + "loss": 2.6391, + "step": 11127 + }, + { + "epoch": 1.1768189509306262, + "grad_norm": 8.5, + "learning_rate": 4.467152087672073e-06, + "loss": 2.5577, + "step": 11128 + }, + { + "epoch": 1.176924703891709, + "grad_norm": 8.0, + "learning_rate": 4.467046395622085e-06, + "loss": 2.5049, + "step": 11129 + }, + { + "epoch": 1.1770304568527918, + "grad_norm": 8.0, + "learning_rate": 4.466940694341554e-06, + "loss": 2.4766, + "step": 11130 + }, + { + "epoch": 1.1771362098138747, + "grad_norm": 7.875, + "learning_rate": 4.466834983830974e-06, + "loss": 2.4192, + "step": 11131 + }, + { + "epoch": 1.1772419627749577, + "grad_norm": 8.25, + "learning_rate": 4.466729264090842e-06, + "loss": 2.3855, + "step": 11132 + }, + { + "epoch": 1.1773477157360406, + "grad_norm": 7.875, + "learning_rate": 4.4666235351216535e-06, + "loss": 2.4219, + "step": 11133 + }, + { + "epoch": 1.1774534686971236, + "grad_norm": 8.4375, + "learning_rate": 4.466517796923906e-06, + "loss": 2.5162, + "step": 11134 + }, + { + "epoch": 1.1775592216582065, + "grad_norm": 8.5625, + "learning_rate": 4.466412049498094e-06, + "loss": 2.506, + "step": 11135 + }, + { + "epoch": 1.1776649746192893, + "grad_norm": 8.375, + "learning_rate": 4.466306292844714e-06, + "loss": 2.5122, + "step": 11136 + }, + { + "epoch": 1.1777707275803722, + "grad_norm": 8.0, + "learning_rate": 4.466200526964264e-06, + "loss": 2.5608, + "step": 11137 + }, + { + "epoch": 1.1778764805414552, + "grad_norm": 7.84375, + "learning_rate": 4.4660947518572384e-06, + "loss": 2.5654, + "step": 11138 + }, + { + "epoch": 1.177982233502538, + "grad_norm": 8.375, + "learning_rate": 4.465988967524135e-06, + "loss": 2.5816, + "step": 11139 + }, + { + "epoch": 1.178087986463621, + "grad_norm": 8.625, + "learning_rate": 4.465883173965449e-06, + "loss": 2.472, + "step": 11140 + }, + { + "epoch": 1.178193739424704, + "grad_norm": 8.5, + "learning_rate": 4.4657773711816766e-06, + "loss": 2.6225, + "step": 11141 + }, + { + "epoch": 1.1782994923857868, + "grad_norm": 7.75, + "learning_rate": 4.465671559173316e-06, + "loss": 2.4042, + "step": 11142 + }, + { + "epoch": 1.1784052453468696, + "grad_norm": 8.3125, + "learning_rate": 4.465565737940862e-06, + "loss": 2.4659, + "step": 11143 + }, + { + "epoch": 1.1785109983079527, + "grad_norm": 8.625, + "learning_rate": 4.465459907484811e-06, + "loss": 2.6729, + "step": 11144 + }, + { + "epoch": 1.1786167512690355, + "grad_norm": 8.8125, + "learning_rate": 4.465354067805663e-06, + "loss": 2.5511, + "step": 11145 + }, + { + "epoch": 1.1787225042301184, + "grad_norm": 8.375, + "learning_rate": 4.465248218903911e-06, + "loss": 2.4748, + "step": 11146 + }, + { + "epoch": 1.1788282571912014, + "grad_norm": 8.3125, + "learning_rate": 4.4651423607800514e-06, + "loss": 2.7118, + "step": 11147 + }, + { + "epoch": 1.1789340101522843, + "grad_norm": 8.625, + "learning_rate": 4.4650364934345835e-06, + "loss": 2.6863, + "step": 11148 + }, + { + "epoch": 1.1790397631133671, + "grad_norm": 9.1875, + "learning_rate": 4.464930616868003e-06, + "loss": 2.7102, + "step": 11149 + }, + { + "epoch": 1.1791455160744502, + "grad_norm": 8.5625, + "learning_rate": 4.464824731080808e-06, + "loss": 2.5354, + "step": 11150 + }, + { + "epoch": 1.179251269035533, + "grad_norm": 8.25, + "learning_rate": 4.464718836073493e-06, + "loss": 2.6457, + "step": 11151 + }, + { + "epoch": 1.1793570219966159, + "grad_norm": 8.3125, + "learning_rate": 4.464612931846556e-06, + "loss": 2.5215, + "step": 11152 + }, + { + "epoch": 1.179462774957699, + "grad_norm": 9.0, + "learning_rate": 4.464507018400493e-06, + "loss": 2.6571, + "step": 11153 + }, + { + "epoch": 1.1795685279187818, + "grad_norm": 8.0625, + "learning_rate": 4.464401095735803e-06, + "loss": 2.4454, + "step": 11154 + }, + { + "epoch": 1.1796742808798646, + "grad_norm": 8.5625, + "learning_rate": 4.464295163852983e-06, + "loss": 2.3007, + "step": 11155 + }, + { + "epoch": 1.1797800338409474, + "grad_norm": 10.75, + "learning_rate": 4.464189222752528e-06, + "loss": 2.4907, + "step": 11156 + }, + { + "epoch": 1.1798857868020305, + "grad_norm": 8.5, + "learning_rate": 4.464083272434937e-06, + "loss": 2.7344, + "step": 11157 + }, + { + "epoch": 1.1799915397631133, + "grad_norm": 8.375, + "learning_rate": 4.463977312900706e-06, + "loss": 2.4464, + "step": 11158 + }, + { + "epoch": 1.1800972927241964, + "grad_norm": 7.71875, + "learning_rate": 4.463871344150332e-06, + "loss": 2.5176, + "step": 11159 + }, + { + "epoch": 1.1802030456852792, + "grad_norm": 8.8125, + "learning_rate": 4.463765366184315e-06, + "loss": 2.6336, + "step": 11160 + }, + { + "epoch": 1.180308798646362, + "grad_norm": 8.25, + "learning_rate": 4.463659379003149e-06, + "loss": 2.5868, + "step": 11161 + }, + { + "epoch": 1.180414551607445, + "grad_norm": 8.625, + "learning_rate": 4.4635533826073315e-06, + "loss": 2.6061, + "step": 11162 + }, + { + "epoch": 1.180520304568528, + "grad_norm": 8.25, + "learning_rate": 4.463447376997363e-06, + "loss": 2.4876, + "step": 11163 + }, + { + "epoch": 1.1806260575296108, + "grad_norm": 8.4375, + "learning_rate": 4.463341362173738e-06, + "loss": 2.5164, + "step": 11164 + }, + { + "epoch": 1.1807318104906936, + "grad_norm": 8.3125, + "learning_rate": 4.463235338136955e-06, + "loss": 2.5657, + "step": 11165 + }, + { + "epoch": 1.1808375634517767, + "grad_norm": 8.25, + "learning_rate": 4.463129304887513e-06, + "loss": 2.7129, + "step": 11166 + }, + { + "epoch": 1.1809433164128595, + "grad_norm": 8.0625, + "learning_rate": 4.463023262425906e-06, + "loss": 2.6281, + "step": 11167 + }, + { + "epoch": 1.1810490693739424, + "grad_norm": 8.375, + "learning_rate": 4.462917210752634e-06, + "loss": 2.4654, + "step": 11168 + }, + { + "epoch": 1.1811548223350254, + "grad_norm": 8.0625, + "learning_rate": 4.462811149868195e-06, + "loss": 2.4703, + "step": 11169 + }, + { + "epoch": 1.1812605752961083, + "grad_norm": 7.875, + "learning_rate": 4.462705079773086e-06, + "loss": 2.5804, + "step": 11170 + }, + { + "epoch": 1.1813663282571911, + "grad_norm": 8.3125, + "learning_rate": 4.4625990004678045e-06, + "loss": 2.6386, + "step": 11171 + }, + { + "epoch": 1.1814720812182742, + "grad_norm": 8.0, + "learning_rate": 4.462492911952849e-06, + "loss": 2.534, + "step": 11172 + }, + { + "epoch": 1.181577834179357, + "grad_norm": 8.0, + "learning_rate": 4.462386814228716e-06, + "loss": 2.6382, + "step": 11173 + }, + { + "epoch": 1.1816835871404399, + "grad_norm": 8.4375, + "learning_rate": 4.462280707295905e-06, + "loss": 2.6289, + "step": 11174 + }, + { + "epoch": 1.181789340101523, + "grad_norm": 8.125, + "learning_rate": 4.462174591154913e-06, + "loss": 2.6501, + "step": 11175 + }, + { + "epoch": 1.1818950930626058, + "grad_norm": 7.875, + "learning_rate": 4.4620684658062384e-06, + "loss": 2.4561, + "step": 11176 + }, + { + "epoch": 1.1820008460236886, + "grad_norm": 8.375, + "learning_rate": 4.461962331250378e-06, + "loss": 2.6012, + "step": 11177 + }, + { + "epoch": 1.1821065989847717, + "grad_norm": 8.0, + "learning_rate": 4.461856187487832e-06, + "loss": 2.462, + "step": 11178 + }, + { + "epoch": 1.1822123519458545, + "grad_norm": 9.1875, + "learning_rate": 4.4617500345190965e-06, + "loss": 2.7409, + "step": 11179 + }, + { + "epoch": 1.1823181049069373, + "grad_norm": 8.375, + "learning_rate": 4.461643872344671e-06, + "loss": 2.5563, + "step": 11180 + }, + { + "epoch": 1.1824238578680204, + "grad_norm": 7.96875, + "learning_rate": 4.461537700965053e-06, + "loss": 2.5368, + "step": 11181 + }, + { + "epoch": 1.1825296108291032, + "grad_norm": 8.0, + "learning_rate": 4.46143152038074e-06, + "loss": 2.6046, + "step": 11182 + }, + { + "epoch": 1.182635363790186, + "grad_norm": 8.5, + "learning_rate": 4.461325330592232e-06, + "loss": 2.3932, + "step": 11183 + }, + { + "epoch": 1.1827411167512691, + "grad_norm": 7.84375, + "learning_rate": 4.461219131600026e-06, + "loss": 2.4022, + "step": 11184 + }, + { + "epoch": 1.182846869712352, + "grad_norm": 8.4375, + "learning_rate": 4.461112923404622e-06, + "loss": 2.6267, + "step": 11185 + }, + { + "epoch": 1.1829526226734348, + "grad_norm": 8.3125, + "learning_rate": 4.461006706006515e-06, + "loss": 2.7623, + "step": 11186 + }, + { + "epoch": 1.1830583756345177, + "grad_norm": 7.96875, + "learning_rate": 4.460900479406206e-06, + "loss": 2.6973, + "step": 11187 + }, + { + "epoch": 1.1831641285956007, + "grad_norm": 8.625, + "learning_rate": 4.460794243604194e-06, + "loss": 2.5704, + "step": 11188 + }, + { + "epoch": 1.1832698815566836, + "grad_norm": 8.125, + "learning_rate": 4.460687998600975e-06, + "loss": 2.4093, + "step": 11189 + }, + { + "epoch": 1.1833756345177666, + "grad_norm": 7.625, + "learning_rate": 4.460581744397051e-06, + "loss": 2.4209, + "step": 11190 + }, + { + "epoch": 1.1834813874788495, + "grad_norm": 8.0, + "learning_rate": 4.460475480992917e-06, + "loss": 2.6177, + "step": 11191 + }, + { + "epoch": 1.1835871404399323, + "grad_norm": 9.3125, + "learning_rate": 4.460369208389075e-06, + "loss": 2.7739, + "step": 11192 + }, + { + "epoch": 1.1836928934010151, + "grad_norm": 8.5625, + "learning_rate": 4.460262926586021e-06, + "loss": 2.4347, + "step": 11193 + }, + { + "epoch": 1.1837986463620982, + "grad_norm": 8.5, + "learning_rate": 4.460156635584255e-06, + "loss": 2.5628, + "step": 11194 + }, + { + "epoch": 1.183904399323181, + "grad_norm": 8.1875, + "learning_rate": 4.460050335384276e-06, + "loss": 2.5119, + "step": 11195 + }, + { + "epoch": 1.1840101522842639, + "grad_norm": 7.78125, + "learning_rate": 4.459944025986583e-06, + "loss": 2.5484, + "step": 11196 + }, + { + "epoch": 1.184115905245347, + "grad_norm": 8.0625, + "learning_rate": 4.459837707391673e-06, + "loss": 2.8296, + "step": 11197 + }, + { + "epoch": 1.1842216582064298, + "grad_norm": 7.9375, + "learning_rate": 4.459731379600046e-06, + "loss": 2.5882, + "step": 11198 + }, + { + "epoch": 1.1843274111675126, + "grad_norm": 7.96875, + "learning_rate": 4.459625042612202e-06, + "loss": 2.2022, + "step": 11199 + }, + { + "epoch": 1.1844331641285957, + "grad_norm": 8.4375, + "learning_rate": 4.459518696428639e-06, + "loss": 2.606, + "step": 11200 + }, + { + "epoch": 1.1845389170896785, + "grad_norm": 8.0, + "learning_rate": 4.459412341049856e-06, + "loss": 2.5301, + "step": 11201 + }, + { + "epoch": 1.1846446700507614, + "grad_norm": 8.125, + "learning_rate": 4.459305976476352e-06, + "loss": 2.4212, + "step": 11202 + }, + { + "epoch": 1.1847504230118444, + "grad_norm": 8.0625, + "learning_rate": 4.459199602708627e-06, + "loss": 2.5624, + "step": 11203 + }, + { + "epoch": 1.1848561759729273, + "grad_norm": 8.0, + "learning_rate": 4.459093219747179e-06, + "loss": 2.4532, + "step": 11204 + }, + { + "epoch": 1.18496192893401, + "grad_norm": 8.25, + "learning_rate": 4.458986827592508e-06, + "loss": 2.5494, + "step": 11205 + }, + { + "epoch": 1.1850676818950932, + "grad_norm": 7.6875, + "learning_rate": 4.458880426245113e-06, + "loss": 2.466, + "step": 11206 + }, + { + "epoch": 1.185173434856176, + "grad_norm": 7.875, + "learning_rate": 4.458774015705494e-06, + "loss": 2.574, + "step": 11207 + }, + { + "epoch": 1.1852791878172588, + "grad_norm": 8.4375, + "learning_rate": 4.4586675959741485e-06, + "loss": 2.5837, + "step": 11208 + }, + { + "epoch": 1.185384940778342, + "grad_norm": 8.25, + "learning_rate": 4.458561167051577e-06, + "loss": 2.5477, + "step": 11209 + }, + { + "epoch": 1.1854906937394247, + "grad_norm": 8.1875, + "learning_rate": 4.458454728938279e-06, + "loss": 2.6292, + "step": 11210 + }, + { + "epoch": 1.1855964467005076, + "grad_norm": 8.75, + "learning_rate": 4.458348281634755e-06, + "loss": 2.2412, + "step": 11211 + }, + { + "epoch": 1.1857021996615904, + "grad_norm": 7.6875, + "learning_rate": 4.458241825141503e-06, + "loss": 2.3512, + "step": 11212 + }, + { + "epoch": 1.1858079526226735, + "grad_norm": 8.9375, + "learning_rate": 4.458135359459023e-06, + "loss": 2.4459, + "step": 11213 + }, + { + "epoch": 1.1859137055837563, + "grad_norm": 7.96875, + "learning_rate": 4.458028884587814e-06, + "loss": 2.6251, + "step": 11214 + }, + { + "epoch": 1.1860194585448394, + "grad_norm": 8.375, + "learning_rate": 4.4579224005283775e-06, + "loss": 2.4516, + "step": 11215 + }, + { + "epoch": 1.1861252115059222, + "grad_norm": 8.0625, + "learning_rate": 4.457815907281211e-06, + "loss": 2.1749, + "step": 11216 + }, + { + "epoch": 1.186230964467005, + "grad_norm": 8.0, + "learning_rate": 4.457709404846815e-06, + "loss": 2.7143, + "step": 11217 + }, + { + "epoch": 1.1863367174280879, + "grad_norm": 8.0625, + "learning_rate": 4.45760289322569e-06, + "loss": 2.5996, + "step": 11218 + }, + { + "epoch": 1.186442470389171, + "grad_norm": 8.125, + "learning_rate": 4.457496372418335e-06, + "loss": 2.4937, + "step": 11219 + }, + { + "epoch": 1.1865482233502538, + "grad_norm": 9.125, + "learning_rate": 4.4573898424252504e-06, + "loss": 2.5401, + "step": 11220 + }, + { + "epoch": 1.1866539763113366, + "grad_norm": 8.75, + "learning_rate": 4.457283303246935e-06, + "loss": 2.5636, + "step": 11221 + }, + { + "epoch": 1.1867597292724197, + "grad_norm": 7.65625, + "learning_rate": 4.45717675488389e-06, + "loss": 2.5243, + "step": 11222 + }, + { + "epoch": 1.1868654822335025, + "grad_norm": 8.3125, + "learning_rate": 4.457070197336615e-06, + "loss": 2.4758, + "step": 11223 + }, + { + "epoch": 1.1869712351945854, + "grad_norm": 8.5, + "learning_rate": 4.456963630605611e-06, + "loss": 2.6605, + "step": 11224 + }, + { + "epoch": 1.1870769881556684, + "grad_norm": 8.75, + "learning_rate": 4.456857054691376e-06, + "loss": 2.6675, + "step": 11225 + }, + { + "epoch": 1.1871827411167513, + "grad_norm": 8.4375, + "learning_rate": 4.4567504695944105e-06, + "loss": 2.3823, + "step": 11226 + }, + { + "epoch": 1.187288494077834, + "grad_norm": 8.4375, + "learning_rate": 4.456643875315216e-06, + "loss": 2.411, + "step": 11227 + }, + { + "epoch": 1.1873942470389172, + "grad_norm": 7.71875, + "learning_rate": 4.456537271854293e-06, + "loss": 2.4881, + "step": 11228 + }, + { + "epoch": 1.1875, + "grad_norm": 8.3125, + "learning_rate": 4.45643065921214e-06, + "loss": 2.4988, + "step": 11229 + }, + { + "epoch": 1.1876057529610828, + "grad_norm": 8.8125, + "learning_rate": 4.456324037389257e-06, + "loss": 2.5352, + "step": 11230 + }, + { + "epoch": 1.187711505922166, + "grad_norm": 7.6875, + "learning_rate": 4.456217406386147e-06, + "loss": 2.467, + "step": 11231 + }, + { + "epoch": 1.1878172588832487, + "grad_norm": 8.4375, + "learning_rate": 4.456110766203307e-06, + "loss": 2.519, + "step": 11232 + }, + { + "epoch": 1.1879230118443316, + "grad_norm": 8.0, + "learning_rate": 4.456004116841241e-06, + "loss": 2.5689, + "step": 11233 + }, + { + "epoch": 1.1880287648054146, + "grad_norm": 8.1875, + "learning_rate": 4.4558974583004475e-06, + "loss": 2.4198, + "step": 11234 + }, + { + "epoch": 1.1881345177664975, + "grad_norm": 8.4375, + "learning_rate": 4.455790790581426e-06, + "loss": 2.517, + "step": 11235 + }, + { + "epoch": 1.1882402707275803, + "grad_norm": 8.0625, + "learning_rate": 4.4556841136846794e-06, + "loss": 2.2624, + "step": 11236 + }, + { + "epoch": 1.1883460236886634, + "grad_norm": 7.625, + "learning_rate": 4.455577427610707e-06, + "loss": 2.5166, + "step": 11237 + }, + { + "epoch": 1.1884517766497462, + "grad_norm": 8.0625, + "learning_rate": 4.4554707323600086e-06, + "loss": 2.5067, + "step": 11238 + }, + { + "epoch": 1.188557529610829, + "grad_norm": 9.0625, + "learning_rate": 4.455364027933087e-06, + "loss": 2.6109, + "step": 11239 + }, + { + "epoch": 1.1886632825719121, + "grad_norm": 8.4375, + "learning_rate": 4.45525731433044e-06, + "loss": 2.6474, + "step": 11240 + }, + { + "epoch": 1.188769035532995, + "grad_norm": 9.0, + "learning_rate": 4.455150591552571e-06, + "loss": 2.5905, + "step": 11241 + }, + { + "epoch": 1.1888747884940778, + "grad_norm": 7.6875, + "learning_rate": 4.455043859599981e-06, + "loss": 2.612, + "step": 11242 + }, + { + "epoch": 1.1889805414551606, + "grad_norm": 8.3125, + "learning_rate": 4.454937118473168e-06, + "loss": 2.4947, + "step": 11243 + }, + { + "epoch": 1.1890862944162437, + "grad_norm": 8.4375, + "learning_rate": 4.454830368172636e-06, + "loss": 2.4459, + "step": 11244 + }, + { + "epoch": 1.1891920473773265, + "grad_norm": 7.875, + "learning_rate": 4.4547236086988835e-06, + "loss": 2.5596, + "step": 11245 + }, + { + "epoch": 1.1892978003384096, + "grad_norm": 8.0, + "learning_rate": 4.454616840052413e-06, + "loss": 2.4326, + "step": 11246 + }, + { + "epoch": 1.1894035532994924, + "grad_norm": 7.96875, + "learning_rate": 4.454510062233726e-06, + "loss": 2.652, + "step": 11247 + }, + { + "epoch": 1.1895093062605753, + "grad_norm": 7.8125, + "learning_rate": 4.4544032752433214e-06, + "loss": 2.6509, + "step": 11248 + }, + { + "epoch": 1.189615059221658, + "grad_norm": 7.96875, + "learning_rate": 4.4542964790817026e-06, + "loss": 2.4606, + "step": 11249 + }, + { + "epoch": 1.1897208121827412, + "grad_norm": 7.90625, + "learning_rate": 4.454189673749368e-06, + "loss": 2.465, + "step": 11250 + }, + { + "epoch": 1.189826565143824, + "grad_norm": 7.8125, + "learning_rate": 4.454082859246822e-06, + "loss": 2.5181, + "step": 11251 + }, + { + "epoch": 1.1899323181049068, + "grad_norm": 8.5, + "learning_rate": 4.453976035574564e-06, + "loss": 2.5736, + "step": 11252 + }, + { + "epoch": 1.19003807106599, + "grad_norm": 8.25, + "learning_rate": 4.453869202733096e-06, + "loss": 2.6612, + "step": 11253 + }, + { + "epoch": 1.1901438240270727, + "grad_norm": 8.125, + "learning_rate": 4.453762360722918e-06, + "loss": 2.5132, + "step": 11254 + }, + { + "epoch": 1.1902495769881556, + "grad_norm": 8.25, + "learning_rate": 4.453655509544533e-06, + "loss": 2.6388, + "step": 11255 + }, + { + "epoch": 1.1903553299492386, + "grad_norm": 7.84375, + "learning_rate": 4.453548649198442e-06, + "loss": 2.6413, + "step": 11256 + }, + { + "epoch": 1.1904610829103215, + "grad_norm": 7.84375, + "learning_rate": 4.453441779685145e-06, + "loss": 2.5109, + "step": 11257 + }, + { + "epoch": 1.1905668358714043, + "grad_norm": 7.75, + "learning_rate": 4.453334901005145e-06, + "loss": 2.5548, + "step": 11258 + }, + { + "epoch": 1.1906725888324874, + "grad_norm": 8.125, + "learning_rate": 4.453228013158943e-06, + "loss": 2.5677, + "step": 11259 + }, + { + "epoch": 1.1907783417935702, + "grad_norm": 7.84375, + "learning_rate": 4.453121116147042e-06, + "loss": 2.4264, + "step": 11260 + }, + { + "epoch": 1.190884094754653, + "grad_norm": 8.25, + "learning_rate": 4.453014209969941e-06, + "loss": 2.6423, + "step": 11261 + }, + { + "epoch": 1.1909898477157361, + "grad_norm": 7.84375, + "learning_rate": 4.452907294628144e-06, + "loss": 2.4655, + "step": 11262 + }, + { + "epoch": 1.191095600676819, + "grad_norm": 8.6875, + "learning_rate": 4.4528003701221526e-06, + "loss": 2.5137, + "step": 11263 + }, + { + "epoch": 1.1912013536379018, + "grad_norm": 9.25, + "learning_rate": 4.452693436452466e-06, + "loss": 2.5782, + "step": 11264 + }, + { + "epoch": 1.1913071065989849, + "grad_norm": 8.8125, + "learning_rate": 4.452586493619588e-06, + "loss": 2.7997, + "step": 11265 + }, + { + "epoch": 1.1914128595600677, + "grad_norm": 8.5625, + "learning_rate": 4.4524795416240205e-06, + "loss": 2.5483, + "step": 11266 + }, + { + "epoch": 1.1915186125211505, + "grad_norm": 9.1875, + "learning_rate": 4.452372580466265e-06, + "loss": 2.8496, + "step": 11267 + }, + { + "epoch": 1.1916243654822334, + "grad_norm": 8.375, + "learning_rate": 4.452265610146824e-06, + "loss": 2.6322, + "step": 11268 + }, + { + "epoch": 1.1917301184433164, + "grad_norm": 8.0625, + "learning_rate": 4.452158630666199e-06, + "loss": 2.5316, + "step": 11269 + }, + { + "epoch": 1.1918358714043993, + "grad_norm": 8.1875, + "learning_rate": 4.452051642024891e-06, + "loss": 2.4591, + "step": 11270 + }, + { + "epoch": 1.1919416243654823, + "grad_norm": 8.6875, + "learning_rate": 4.4519446442234035e-06, + "loss": 2.6835, + "step": 11271 + }, + { + "epoch": 1.1920473773265652, + "grad_norm": 8.9375, + "learning_rate": 4.451837637262238e-06, + "loss": 2.5822, + "step": 11272 + }, + { + "epoch": 1.192153130287648, + "grad_norm": 7.8125, + "learning_rate": 4.451730621141897e-06, + "loss": 2.363, + "step": 11273 + }, + { + "epoch": 1.1922588832487309, + "grad_norm": 7.90625, + "learning_rate": 4.451623595862883e-06, + "loss": 2.5457, + "step": 11274 + }, + { + "epoch": 1.192364636209814, + "grad_norm": 8.5, + "learning_rate": 4.451516561425696e-06, + "loss": 2.5506, + "step": 11275 + }, + { + "epoch": 1.1924703891708968, + "grad_norm": 7.90625, + "learning_rate": 4.451409517830841e-06, + "loss": 2.4718, + "step": 11276 + }, + { + "epoch": 1.1925761421319796, + "grad_norm": 7.875, + "learning_rate": 4.451302465078819e-06, + "loss": 2.5995, + "step": 11277 + }, + { + "epoch": 1.1926818950930627, + "grad_norm": 8.875, + "learning_rate": 4.451195403170132e-06, + "loss": 2.517, + "step": 11278 + }, + { + "epoch": 1.1927876480541455, + "grad_norm": 9.375, + "learning_rate": 4.451088332105284e-06, + "loss": 2.5077, + "step": 11279 + }, + { + "epoch": 1.1928934010152283, + "grad_norm": 8.0625, + "learning_rate": 4.450981251884776e-06, + "loss": 2.6711, + "step": 11280 + }, + { + "epoch": 1.1929991539763114, + "grad_norm": 8.25, + "learning_rate": 4.450874162509111e-06, + "loss": 2.5768, + "step": 11281 + }, + { + "epoch": 1.1931049069373942, + "grad_norm": 8.375, + "learning_rate": 4.450767063978792e-06, + "loss": 2.7591, + "step": 11282 + }, + { + "epoch": 1.193210659898477, + "grad_norm": 8.9375, + "learning_rate": 4.45065995629432e-06, + "loss": 2.6829, + "step": 11283 + }, + { + "epoch": 1.1933164128595601, + "grad_norm": 8.4375, + "learning_rate": 4.450552839456199e-06, + "loss": 2.6508, + "step": 11284 + }, + { + "epoch": 1.193422165820643, + "grad_norm": 7.96875, + "learning_rate": 4.450445713464931e-06, + "loss": 2.3897, + "step": 11285 + }, + { + "epoch": 1.1935279187817258, + "grad_norm": 8.4375, + "learning_rate": 4.45033857832102e-06, + "loss": 2.4877, + "step": 11286 + }, + { + "epoch": 1.1936336717428089, + "grad_norm": 8.1875, + "learning_rate": 4.450231434024967e-06, + "loss": 2.1641, + "step": 11287 + }, + { + "epoch": 1.1937394247038917, + "grad_norm": 8.0625, + "learning_rate": 4.450124280577275e-06, + "loss": 2.4732, + "step": 11288 + }, + { + "epoch": 1.1938451776649746, + "grad_norm": 7.59375, + "learning_rate": 4.4500171179784475e-06, + "loss": 2.5171, + "step": 11289 + }, + { + "epoch": 1.1939509306260576, + "grad_norm": 8.0, + "learning_rate": 4.4499099462289876e-06, + "loss": 2.4746, + "step": 11290 + }, + { + "epoch": 1.1940566835871405, + "grad_norm": 8.1875, + "learning_rate": 4.4498027653293975e-06, + "loss": 2.5046, + "step": 11291 + }, + { + "epoch": 1.1941624365482233, + "grad_norm": 8.25, + "learning_rate": 4.44969557528018e-06, + "loss": 2.5001, + "step": 11292 + }, + { + "epoch": 1.1942681895093064, + "grad_norm": 8.5, + "learning_rate": 4.449588376081839e-06, + "loss": 2.7637, + "step": 11293 + }, + { + "epoch": 1.1943739424703892, + "grad_norm": 8.1875, + "learning_rate": 4.4494811677348785e-06, + "loss": 2.6102, + "step": 11294 + }, + { + "epoch": 1.194479695431472, + "grad_norm": 8.5625, + "learning_rate": 4.449373950239799e-06, + "loss": 2.6479, + "step": 11295 + }, + { + "epoch": 1.194585448392555, + "grad_norm": 8.25, + "learning_rate": 4.4492667235971044e-06, + "loss": 2.6513, + "step": 11296 + }, + { + "epoch": 1.194691201353638, + "grad_norm": 8.1875, + "learning_rate": 4.449159487807298e-06, + "loss": 2.6654, + "step": 11297 + }, + { + "epoch": 1.1947969543147208, + "grad_norm": 8.4375, + "learning_rate": 4.449052242870885e-06, + "loss": 2.5559, + "step": 11298 + }, + { + "epoch": 1.1949027072758036, + "grad_norm": 8.5, + "learning_rate": 4.448944988788367e-06, + "loss": 2.6521, + "step": 11299 + }, + { + "epoch": 1.1950084602368867, + "grad_norm": 8.1875, + "learning_rate": 4.448837725560245e-06, + "loss": 2.603, + "step": 11300 + }, + { + "epoch": 1.1951142131979695, + "grad_norm": 8.5, + "learning_rate": 4.448730453187026e-06, + "loss": 2.5299, + "step": 11301 + }, + { + "epoch": 1.1952199661590526, + "grad_norm": 7.96875, + "learning_rate": 4.4486231716692115e-06, + "loss": 2.5049, + "step": 11302 + }, + { + "epoch": 1.1953257191201354, + "grad_norm": 8.0625, + "learning_rate": 4.448515881007306e-06, + "loss": 2.6226, + "step": 11303 + }, + { + "epoch": 1.1954314720812182, + "grad_norm": 8.6875, + "learning_rate": 4.448408581201812e-06, + "loss": 2.6084, + "step": 11304 + }, + { + "epoch": 1.195537225042301, + "grad_norm": 8.125, + "learning_rate": 4.448301272253234e-06, + "loss": 2.672, + "step": 11305 + }, + { + "epoch": 1.1956429780033841, + "grad_norm": 8.125, + "learning_rate": 4.448193954162075e-06, + "loss": 2.7105, + "step": 11306 + }, + { + "epoch": 1.195748730964467, + "grad_norm": 9.0, + "learning_rate": 4.448086626928837e-06, + "loss": 2.659, + "step": 11307 + }, + { + "epoch": 1.1958544839255498, + "grad_norm": 8.125, + "learning_rate": 4.447979290554026e-06, + "loss": 2.486, + "step": 11308 + }, + { + "epoch": 1.1959602368866329, + "grad_norm": 7.90625, + "learning_rate": 4.447871945038146e-06, + "loss": 2.3701, + "step": 11309 + }, + { + "epoch": 1.1960659898477157, + "grad_norm": 8.75, + "learning_rate": 4.447764590381698e-06, + "loss": 2.5401, + "step": 11310 + }, + { + "epoch": 1.1961717428087986, + "grad_norm": 8.6875, + "learning_rate": 4.447657226585188e-06, + "loss": 2.3477, + "step": 11311 + }, + { + "epoch": 1.1962774957698816, + "grad_norm": 8.875, + "learning_rate": 4.44754985364912e-06, + "loss": 2.4624, + "step": 11312 + }, + { + "epoch": 1.1963832487309645, + "grad_norm": 9.125, + "learning_rate": 4.4474424715739965e-06, + "loss": 2.7738, + "step": 11313 + }, + { + "epoch": 1.1964890016920473, + "grad_norm": 9.1875, + "learning_rate": 4.447335080360322e-06, + "loss": 2.7187, + "step": 11314 + }, + { + "epoch": 1.1965947546531304, + "grad_norm": 7.6875, + "learning_rate": 4.4472276800086e-06, + "loss": 2.6406, + "step": 11315 + }, + { + "epoch": 1.1967005076142132, + "grad_norm": 7.96875, + "learning_rate": 4.447120270519335e-06, + "loss": 2.385, + "step": 11316 + }, + { + "epoch": 1.196806260575296, + "grad_norm": 8.5625, + "learning_rate": 4.447012851893031e-06, + "loss": 2.5516, + "step": 11317 + }, + { + "epoch": 1.196912013536379, + "grad_norm": 7.875, + "learning_rate": 4.446905424130192e-06, + "loss": 2.5374, + "step": 11318 + }, + { + "epoch": 1.197017766497462, + "grad_norm": 8.125, + "learning_rate": 4.446797987231322e-06, + "loss": 2.5045, + "step": 11319 + }, + { + "epoch": 1.1971235194585448, + "grad_norm": 9.0625, + "learning_rate": 4.4466905411969254e-06, + "loss": 2.6249, + "step": 11320 + }, + { + "epoch": 1.1972292724196278, + "grad_norm": 8.5625, + "learning_rate": 4.446583086027506e-06, + "loss": 2.4269, + "step": 11321 + }, + { + "epoch": 1.1973350253807107, + "grad_norm": 7.9375, + "learning_rate": 4.4464756217235695e-06, + "loss": 2.5244, + "step": 11322 + }, + { + "epoch": 1.1974407783417935, + "grad_norm": 7.9375, + "learning_rate": 4.446368148285617e-06, + "loss": 2.4479, + "step": 11323 + }, + { + "epoch": 1.1975465313028764, + "grad_norm": 9.125, + "learning_rate": 4.446260665714156e-06, + "loss": 2.9445, + "step": 11324 + }, + { + "epoch": 1.1976522842639594, + "grad_norm": 8.25, + "learning_rate": 4.44615317400969e-06, + "loss": 2.5701, + "step": 11325 + }, + { + "epoch": 1.1977580372250423, + "grad_norm": 8.3125, + "learning_rate": 4.446045673172722e-06, + "loss": 2.5382, + "step": 11326 + }, + { + "epoch": 1.1978637901861253, + "grad_norm": 8.5, + "learning_rate": 4.445938163203759e-06, + "loss": 2.6229, + "step": 11327 + }, + { + "epoch": 1.1979695431472082, + "grad_norm": 8.25, + "learning_rate": 4.445830644103303e-06, + "loss": 2.6109, + "step": 11328 + }, + { + "epoch": 1.198075296108291, + "grad_norm": 8.25, + "learning_rate": 4.44572311587186e-06, + "loss": 2.3359, + "step": 11329 + }, + { + "epoch": 1.1981810490693738, + "grad_norm": 8.375, + "learning_rate": 4.445615578509934e-06, + "loss": 2.4291, + "step": 11330 + }, + { + "epoch": 1.198286802030457, + "grad_norm": 8.125, + "learning_rate": 4.44550803201803e-06, + "loss": 2.5567, + "step": 11331 + }, + { + "epoch": 1.1983925549915397, + "grad_norm": 7.84375, + "learning_rate": 4.445400476396652e-06, + "loss": 2.497, + "step": 11332 + }, + { + "epoch": 1.1984983079526226, + "grad_norm": 8.1875, + "learning_rate": 4.445292911646306e-06, + "loss": 2.6066, + "step": 11333 + }, + { + "epoch": 1.1986040609137056, + "grad_norm": 7.875, + "learning_rate": 4.445185337767496e-06, + "loss": 2.443, + "step": 11334 + }, + { + "epoch": 1.1987098138747885, + "grad_norm": 8.875, + "learning_rate": 4.445077754760726e-06, + "loss": 2.6234, + "step": 11335 + }, + { + "epoch": 1.1988155668358713, + "grad_norm": 8.1875, + "learning_rate": 4.444970162626502e-06, + "loss": 2.52, + "step": 11336 + }, + { + "epoch": 1.1989213197969544, + "grad_norm": 8.375, + "learning_rate": 4.444862561365328e-06, + "loss": 2.5132, + "step": 11337 + }, + { + "epoch": 1.1990270727580372, + "grad_norm": 8.625, + "learning_rate": 4.44475495097771e-06, + "loss": 2.5503, + "step": 11338 + }, + { + "epoch": 1.19913282571912, + "grad_norm": 8.75, + "learning_rate": 4.4446473314641525e-06, + "loss": 2.6899, + "step": 11339 + }, + { + "epoch": 1.1992385786802031, + "grad_norm": 8.6875, + "learning_rate": 4.44453970282516e-06, + "loss": 2.7328, + "step": 11340 + }, + { + "epoch": 1.199344331641286, + "grad_norm": 8.125, + "learning_rate": 4.444432065061238e-06, + "loss": 2.4055, + "step": 11341 + }, + { + "epoch": 1.1994500846023688, + "grad_norm": 7.84375, + "learning_rate": 4.4443244181728915e-06, + "loss": 2.4566, + "step": 11342 + }, + { + "epoch": 1.1995558375634519, + "grad_norm": 8.125, + "learning_rate": 4.444216762160626e-06, + "loss": 2.4815, + "step": 11343 + }, + { + "epoch": 1.1996615905245347, + "grad_norm": 8.125, + "learning_rate": 4.444109097024947e-06, + "loss": 2.6396, + "step": 11344 + }, + { + "epoch": 1.1997673434856175, + "grad_norm": 7.96875, + "learning_rate": 4.444001422766358e-06, + "loss": 2.5374, + "step": 11345 + }, + { + "epoch": 1.1998730964467006, + "grad_norm": 8.3125, + "learning_rate": 4.4438937393853654e-06, + "loss": 2.6628, + "step": 11346 + }, + { + "epoch": 1.1999788494077834, + "grad_norm": 8.3125, + "learning_rate": 4.443786046882475e-06, + "loss": 2.4909, + "step": 11347 + }, + { + "epoch": 1.2000846023688663, + "grad_norm": 8.125, + "learning_rate": 4.443678345258191e-06, + "loss": 2.6276, + "step": 11348 + }, + { + "epoch": 1.2001903553299493, + "grad_norm": 8.125, + "learning_rate": 4.44357063451302e-06, + "loss": 2.5897, + "step": 11349 + }, + { + "epoch": 1.2002961082910322, + "grad_norm": 8.625, + "learning_rate": 4.443462914647468e-06, + "loss": 2.5504, + "step": 11350 + }, + { + "epoch": 1.200401861252115, + "grad_norm": 9.0, + "learning_rate": 4.443355185662037e-06, + "loss": 2.5399, + "step": 11351 + }, + { + "epoch": 1.200507614213198, + "grad_norm": 8.0, + "learning_rate": 4.443247447557237e-06, + "loss": 2.7624, + "step": 11352 + }, + { + "epoch": 1.200613367174281, + "grad_norm": 7.9375, + "learning_rate": 4.443139700333571e-06, + "loss": 2.3529, + "step": 11353 + }, + { + "epoch": 1.2007191201353637, + "grad_norm": 7.875, + "learning_rate": 4.443031943991544e-06, + "loss": 2.5469, + "step": 11354 + }, + { + "epoch": 1.2008248730964466, + "grad_norm": 8.375, + "learning_rate": 4.442924178531664e-06, + "loss": 2.3104, + "step": 11355 + }, + { + "epoch": 1.2009306260575296, + "grad_norm": 9.375, + "learning_rate": 4.442816403954435e-06, + "loss": 2.4913, + "step": 11356 + }, + { + "epoch": 1.2010363790186125, + "grad_norm": 8.25, + "learning_rate": 4.442708620260363e-06, + "loss": 2.557, + "step": 11357 + }, + { + "epoch": 1.2011421319796955, + "grad_norm": 8.0, + "learning_rate": 4.442600827449954e-06, + "loss": 2.4921, + "step": 11358 + }, + { + "epoch": 1.2012478849407784, + "grad_norm": 7.8125, + "learning_rate": 4.442493025523714e-06, + "loss": 2.5901, + "step": 11359 + }, + { + "epoch": 1.2013536379018612, + "grad_norm": 8.25, + "learning_rate": 4.442385214482149e-06, + "loss": 2.4905, + "step": 11360 + }, + { + "epoch": 1.201459390862944, + "grad_norm": 8.4375, + "learning_rate": 4.442277394325764e-06, + "loss": 2.5834, + "step": 11361 + }, + { + "epoch": 1.2015651438240271, + "grad_norm": 8.4375, + "learning_rate": 4.4421695650550665e-06, + "loss": 2.6368, + "step": 11362 + }, + { + "epoch": 1.20167089678511, + "grad_norm": 8.4375, + "learning_rate": 4.442061726670561e-06, + "loss": 2.4345, + "step": 11363 + }, + { + "epoch": 1.2017766497461928, + "grad_norm": 8.125, + "learning_rate": 4.441953879172753e-06, + "loss": 2.4956, + "step": 11364 + }, + { + "epoch": 1.2018824027072759, + "grad_norm": 8.1875, + "learning_rate": 4.441846022562152e-06, + "loss": 2.5199, + "step": 11365 + }, + { + "epoch": 1.2019881556683587, + "grad_norm": 9.8125, + "learning_rate": 4.441738156839261e-06, + "loss": 2.5745, + "step": 11366 + }, + { + "epoch": 1.2020939086294415, + "grad_norm": 8.3125, + "learning_rate": 4.441630282004585e-06, + "loss": 2.7442, + "step": 11367 + }, + { + "epoch": 1.2021996615905246, + "grad_norm": 9.25, + "learning_rate": 4.441522398058634e-06, + "loss": 2.6141, + "step": 11368 + }, + { + "epoch": 1.2023054145516074, + "grad_norm": 7.6875, + "learning_rate": 4.441414505001912e-06, + "loss": 2.3362, + "step": 11369 + }, + { + "epoch": 1.2024111675126903, + "grad_norm": 8.4375, + "learning_rate": 4.4413066028349265e-06, + "loss": 2.6116, + "step": 11370 + }, + { + "epoch": 1.2025169204737733, + "grad_norm": 8.0625, + "learning_rate": 4.441198691558182e-06, + "loss": 2.6463, + "step": 11371 + }, + { + "epoch": 1.2026226734348562, + "grad_norm": 7.71875, + "learning_rate": 4.441090771172186e-06, + "loss": 2.3556, + "step": 11372 + }, + { + "epoch": 1.202728426395939, + "grad_norm": 8.3125, + "learning_rate": 4.440982841677446e-06, + "loss": 2.5954, + "step": 11373 + }, + { + "epoch": 1.202834179357022, + "grad_norm": 7.96875, + "learning_rate": 4.440874903074466e-06, + "loss": 2.6219, + "step": 11374 + }, + { + "epoch": 1.202939932318105, + "grad_norm": 7.875, + "learning_rate": 4.4407669553637545e-06, + "loss": 2.4279, + "step": 11375 + }, + { + "epoch": 1.2030456852791878, + "grad_norm": 8.5, + "learning_rate": 4.440658998545818e-06, + "loss": 2.537, + "step": 11376 + }, + { + "epoch": 1.2031514382402708, + "grad_norm": 8.75, + "learning_rate": 4.440551032621162e-06, + "loss": 2.6175, + "step": 11377 + }, + { + "epoch": 1.2032571912013537, + "grad_norm": 8.125, + "learning_rate": 4.4404430575902935e-06, + "loss": 2.4604, + "step": 11378 + }, + { + "epoch": 1.2033629441624365, + "grad_norm": 8.5, + "learning_rate": 4.440335073453719e-06, + "loss": 2.6841, + "step": 11379 + }, + { + "epoch": 1.2034686971235193, + "grad_norm": 8.8125, + "learning_rate": 4.4402270802119465e-06, + "loss": 2.6603, + "step": 11380 + }, + { + "epoch": 1.2035744500846024, + "grad_norm": 8.4375, + "learning_rate": 4.440119077865481e-06, + "loss": 2.6454, + "step": 11381 + }, + { + "epoch": 1.2036802030456852, + "grad_norm": 8.6875, + "learning_rate": 4.440011066414831e-06, + "loss": 2.6257, + "step": 11382 + }, + { + "epoch": 1.2037859560067683, + "grad_norm": 7.90625, + "learning_rate": 4.439903045860502e-06, + "loss": 2.5913, + "step": 11383 + }, + { + "epoch": 1.2038917089678511, + "grad_norm": 8.375, + "learning_rate": 4.4397950162030016e-06, + "loss": 2.5611, + "step": 11384 + }, + { + "epoch": 1.203997461928934, + "grad_norm": 8.125, + "learning_rate": 4.439686977442836e-06, + "loss": 2.4754, + "step": 11385 + }, + { + "epoch": 1.2041032148900168, + "grad_norm": 8.4375, + "learning_rate": 4.439578929580512e-06, + "loss": 2.4569, + "step": 11386 + }, + { + "epoch": 1.2042089678510999, + "grad_norm": 7.78125, + "learning_rate": 4.439470872616539e-06, + "loss": 2.3311, + "step": 11387 + }, + { + "epoch": 1.2043147208121827, + "grad_norm": 8.8125, + "learning_rate": 4.439362806551422e-06, + "loss": 2.4105, + "step": 11388 + }, + { + "epoch": 1.2044204737732658, + "grad_norm": 8.5, + "learning_rate": 4.439254731385668e-06, + "loss": 2.5355, + "step": 11389 + }, + { + "epoch": 1.2045262267343486, + "grad_norm": 8.375, + "learning_rate": 4.439146647119786e-06, + "loss": 2.6462, + "step": 11390 + }, + { + "epoch": 1.2046319796954315, + "grad_norm": 7.8125, + "learning_rate": 4.43903855375428e-06, + "loss": 2.4513, + "step": 11391 + }, + { + "epoch": 1.2047377326565143, + "grad_norm": 8.4375, + "learning_rate": 4.43893045128966e-06, + "loss": 2.5953, + "step": 11392 + }, + { + "epoch": 1.2048434856175974, + "grad_norm": 7.96875, + "learning_rate": 4.438822339726433e-06, + "loss": 2.5838, + "step": 11393 + }, + { + "epoch": 1.2049492385786802, + "grad_norm": 9.375, + "learning_rate": 4.438714219065104e-06, + "loss": 2.5596, + "step": 11394 + }, + { + "epoch": 1.205054991539763, + "grad_norm": 8.125, + "learning_rate": 4.438606089306184e-06, + "loss": 2.4766, + "step": 11395 + }, + { + "epoch": 1.205160744500846, + "grad_norm": 8.0625, + "learning_rate": 4.438497950450178e-06, + "loss": 2.5821, + "step": 11396 + }, + { + "epoch": 1.205266497461929, + "grad_norm": 9.0, + "learning_rate": 4.4383898024975935e-06, + "loss": 2.8284, + "step": 11397 + }, + { + "epoch": 1.2053722504230118, + "grad_norm": 8.5625, + "learning_rate": 4.438281645448938e-06, + "loss": 2.4159, + "step": 11398 + }, + { + "epoch": 1.2054780033840948, + "grad_norm": 8.9375, + "learning_rate": 4.43817347930472e-06, + "loss": 2.5908, + "step": 11399 + }, + { + "epoch": 1.2055837563451777, + "grad_norm": 8.375, + "learning_rate": 4.438065304065447e-06, + "loss": 2.6623, + "step": 11400 + }, + { + "epoch": 1.2056895093062605, + "grad_norm": 8.5625, + "learning_rate": 4.437957119731626e-06, + "loss": 2.5613, + "step": 11401 + }, + { + "epoch": 1.2057952622673436, + "grad_norm": 8.1875, + "learning_rate": 4.437848926303764e-06, + "loss": 2.5514, + "step": 11402 + }, + { + "epoch": 1.2059010152284264, + "grad_norm": 8.1875, + "learning_rate": 4.43774072378237e-06, + "loss": 2.4476, + "step": 11403 + }, + { + "epoch": 1.2060067681895092, + "grad_norm": 8.0625, + "learning_rate": 4.437632512167952e-06, + "loss": 2.3102, + "step": 11404 + }, + { + "epoch": 1.2061125211505923, + "grad_norm": 8.8125, + "learning_rate": 4.437524291461016e-06, + "loss": 2.4946, + "step": 11405 + }, + { + "epoch": 1.2062182741116751, + "grad_norm": 8.1875, + "learning_rate": 4.437416061662071e-06, + "loss": 2.6596, + "step": 11406 + }, + { + "epoch": 1.206324027072758, + "grad_norm": 8.25, + "learning_rate": 4.437307822771626e-06, + "loss": 2.6022, + "step": 11407 + }, + { + "epoch": 1.206429780033841, + "grad_norm": 8.75, + "learning_rate": 4.437199574790186e-06, + "loss": 2.5071, + "step": 11408 + }, + { + "epoch": 1.2065355329949239, + "grad_norm": 8.25, + "learning_rate": 4.437091317718262e-06, + "loss": 2.515, + "step": 11409 + }, + { + "epoch": 1.2066412859560067, + "grad_norm": 8.4375, + "learning_rate": 4.4369830515563605e-06, + "loss": 2.5384, + "step": 11410 + }, + { + "epoch": 1.2067470389170896, + "grad_norm": 9.9375, + "learning_rate": 4.436874776304989e-06, + "loss": 2.5263, + "step": 11411 + }, + { + "epoch": 1.2068527918781726, + "grad_norm": 8.5, + "learning_rate": 4.436766491964657e-06, + "loss": 2.636, + "step": 11412 + }, + { + "epoch": 1.2069585448392555, + "grad_norm": 8.8125, + "learning_rate": 4.436658198535872e-06, + "loss": 2.7949, + "step": 11413 + }, + { + "epoch": 1.2070642978003385, + "grad_norm": 7.9375, + "learning_rate": 4.4365498960191425e-06, + "loss": 2.3558, + "step": 11414 + }, + { + "epoch": 1.2071700507614214, + "grad_norm": 7.875, + "learning_rate": 4.4364415844149755e-06, + "loss": 2.5296, + "step": 11415 + }, + { + "epoch": 1.2072758037225042, + "grad_norm": 8.625, + "learning_rate": 4.4363332637238805e-06, + "loss": 2.4798, + "step": 11416 + }, + { + "epoch": 1.207381556683587, + "grad_norm": 8.1875, + "learning_rate": 4.436224933946365e-06, + "loss": 2.5891, + "step": 11417 + }, + { + "epoch": 1.20748730964467, + "grad_norm": 7.9375, + "learning_rate": 4.436116595082938e-06, + "loss": 2.4732, + "step": 11418 + }, + { + "epoch": 1.207593062605753, + "grad_norm": 8.0625, + "learning_rate": 4.4360082471341085e-06, + "loss": 2.4657, + "step": 11419 + }, + { + "epoch": 1.2076988155668358, + "grad_norm": 8.0625, + "learning_rate": 4.435899890100383e-06, + "loss": 2.6642, + "step": 11420 + }, + { + "epoch": 1.2078045685279188, + "grad_norm": 8.0, + "learning_rate": 4.435791523982272e-06, + "loss": 2.6068, + "step": 11421 + }, + { + "epoch": 1.2079103214890017, + "grad_norm": 8.4375, + "learning_rate": 4.435683148780282e-06, + "loss": 2.6323, + "step": 11422 + }, + { + "epoch": 1.2080160744500845, + "grad_norm": 7.78125, + "learning_rate": 4.435574764494923e-06, + "loss": 2.4687, + "step": 11423 + }, + { + "epoch": 1.2081218274111676, + "grad_norm": 8.125, + "learning_rate": 4.435466371126704e-06, + "loss": 2.501, + "step": 11424 + }, + { + "epoch": 1.2082275803722504, + "grad_norm": 7.9375, + "learning_rate": 4.435357968676131e-06, + "loss": 2.3721, + "step": 11425 + }, + { + "epoch": 1.2083333333333333, + "grad_norm": 8.25, + "learning_rate": 4.435249557143717e-06, + "loss": 2.596, + "step": 11426 + }, + { + "epoch": 1.2084390862944163, + "grad_norm": 7.90625, + "learning_rate": 4.435141136529968e-06, + "loss": 2.6073, + "step": 11427 + }, + { + "epoch": 1.2085448392554992, + "grad_norm": 7.875, + "learning_rate": 4.435032706835392e-06, + "loss": 2.5477, + "step": 11428 + }, + { + "epoch": 1.208650592216582, + "grad_norm": 8.25, + "learning_rate": 4.4349242680604985e-06, + "loss": 2.7047, + "step": 11429 + }, + { + "epoch": 1.208756345177665, + "grad_norm": 8.5, + "learning_rate": 4.434815820205798e-06, + "loss": 2.6386, + "step": 11430 + }, + { + "epoch": 1.208862098138748, + "grad_norm": 8.3125, + "learning_rate": 4.434707363271797e-06, + "loss": 2.6504, + "step": 11431 + }, + { + "epoch": 1.2089678510998307, + "grad_norm": 8.0625, + "learning_rate": 4.434598897259007e-06, + "loss": 2.6063, + "step": 11432 + }, + { + "epoch": 1.2090736040609138, + "grad_norm": 8.875, + "learning_rate": 4.434490422167935e-06, + "loss": 2.487, + "step": 11433 + }, + { + "epoch": 1.2091793570219966, + "grad_norm": 8.0625, + "learning_rate": 4.43438193799909e-06, + "loss": 2.7019, + "step": 11434 + }, + { + "epoch": 1.2092851099830795, + "grad_norm": 9.0, + "learning_rate": 4.4342734447529825e-06, + "loss": 2.3356, + "step": 11435 + }, + { + "epoch": 1.2093908629441623, + "grad_norm": 8.9375, + "learning_rate": 4.434164942430121e-06, + "loss": 2.5301, + "step": 11436 + }, + { + "epoch": 1.2094966159052454, + "grad_norm": 8.8125, + "learning_rate": 4.434056431031014e-06, + "loss": 2.4934, + "step": 11437 + }, + { + "epoch": 1.2096023688663282, + "grad_norm": 7.875, + "learning_rate": 4.4339479105561715e-06, + "loss": 2.4251, + "step": 11438 + }, + { + "epoch": 1.2097081218274113, + "grad_norm": 8.5, + "learning_rate": 4.4338393810061014e-06, + "loss": 2.3974, + "step": 11439 + }, + { + "epoch": 1.2098138747884941, + "grad_norm": 7.96875, + "learning_rate": 4.433730842381315e-06, + "loss": 2.4347, + "step": 11440 + }, + { + "epoch": 1.209919627749577, + "grad_norm": 8.5, + "learning_rate": 4.433622294682321e-06, + "loss": 2.4138, + "step": 11441 + }, + { + "epoch": 1.2100253807106598, + "grad_norm": 8.25, + "learning_rate": 4.433513737909627e-06, + "loss": 2.5293, + "step": 11442 + }, + { + "epoch": 1.2101311336717429, + "grad_norm": 8.375, + "learning_rate": 4.433405172063745e-06, + "loss": 2.3993, + "step": 11443 + }, + { + "epoch": 1.2102368866328257, + "grad_norm": 8.25, + "learning_rate": 4.433296597145182e-06, + "loss": 2.451, + "step": 11444 + }, + { + "epoch": 1.2103426395939088, + "grad_norm": 8.375, + "learning_rate": 4.4331880131544495e-06, + "loss": 2.655, + "step": 11445 + }, + { + "epoch": 1.2104483925549916, + "grad_norm": 8.375, + "learning_rate": 4.433079420092057e-06, + "loss": 2.664, + "step": 11446 + }, + { + "epoch": 1.2105541455160744, + "grad_norm": 7.90625, + "learning_rate": 4.432970817958513e-06, + "loss": 2.4687, + "step": 11447 + }, + { + "epoch": 1.2106598984771573, + "grad_norm": 8.4375, + "learning_rate": 4.432862206754327e-06, + "loss": 2.4821, + "step": 11448 + }, + { + "epoch": 1.2107656514382403, + "grad_norm": 8.5, + "learning_rate": 4.4327535864800095e-06, + "loss": 2.4563, + "step": 11449 + }, + { + "epoch": 1.2108714043993232, + "grad_norm": 7.96875, + "learning_rate": 4.43264495713607e-06, + "loss": 2.3303, + "step": 11450 + }, + { + "epoch": 1.210977157360406, + "grad_norm": 8.75, + "learning_rate": 4.432536318723017e-06, + "loss": 2.4915, + "step": 11451 + }, + { + "epoch": 1.211082910321489, + "grad_norm": 8.0, + "learning_rate": 4.432427671241363e-06, + "loss": 2.5189, + "step": 11452 + }, + { + "epoch": 1.211188663282572, + "grad_norm": 8.625, + "learning_rate": 4.432319014691615e-06, + "loss": 2.3546, + "step": 11453 + }, + { + "epoch": 1.2112944162436547, + "grad_norm": 8.125, + "learning_rate": 4.4322103490742855e-06, + "loss": 2.537, + "step": 11454 + }, + { + "epoch": 1.2114001692047378, + "grad_norm": 8.6875, + "learning_rate": 4.432101674389883e-06, + "loss": 2.3947, + "step": 11455 + }, + { + "epoch": 1.2115059221658206, + "grad_norm": 8.375, + "learning_rate": 4.4319929906389155e-06, + "loss": 2.6885, + "step": 11456 + }, + { + "epoch": 1.2116116751269035, + "grad_norm": 8.5, + "learning_rate": 4.431884297821896e-06, + "loss": 2.7745, + "step": 11457 + }, + { + "epoch": 1.2117174280879865, + "grad_norm": 8.25, + "learning_rate": 4.431775595939335e-06, + "loss": 2.4215, + "step": 11458 + }, + { + "epoch": 1.2118231810490694, + "grad_norm": 8.25, + "learning_rate": 4.43166688499174e-06, + "loss": 2.1208, + "step": 11459 + }, + { + "epoch": 1.2119289340101522, + "grad_norm": 8.375, + "learning_rate": 4.431558164979624e-06, + "loss": 2.6447, + "step": 11460 + }, + { + "epoch": 1.2120346869712353, + "grad_norm": 7.875, + "learning_rate": 4.431449435903493e-06, + "loss": 2.557, + "step": 11461 + }, + { + "epoch": 1.2121404399323181, + "grad_norm": 8.8125, + "learning_rate": 4.431340697763861e-06, + "loss": 2.5379, + "step": 11462 + }, + { + "epoch": 1.212246192893401, + "grad_norm": 8.3125, + "learning_rate": 4.431231950561237e-06, + "loss": 2.6327, + "step": 11463 + }, + { + "epoch": 1.212351945854484, + "grad_norm": 7.9375, + "learning_rate": 4.431123194296131e-06, + "loss": 2.4307, + "step": 11464 + }, + { + "epoch": 1.2124576988155669, + "grad_norm": 8.4375, + "learning_rate": 4.431014428969054e-06, + "loss": 2.7767, + "step": 11465 + }, + { + "epoch": 1.2125634517766497, + "grad_norm": 8.25, + "learning_rate": 4.430905654580516e-06, + "loss": 2.6282, + "step": 11466 + }, + { + "epoch": 1.2126692047377325, + "grad_norm": 8.5625, + "learning_rate": 4.430796871131028e-06, + "loss": 2.8369, + "step": 11467 + }, + { + "epoch": 1.2127749576988156, + "grad_norm": 8.1875, + "learning_rate": 4.430688078621099e-06, + "loss": 2.6841, + "step": 11468 + }, + { + "epoch": 1.2128807106598984, + "grad_norm": 8.3125, + "learning_rate": 4.43057927705124e-06, + "loss": 2.7202, + "step": 11469 + }, + { + "epoch": 1.2129864636209815, + "grad_norm": 8.0625, + "learning_rate": 4.4304704664219634e-06, + "loss": 2.4722, + "step": 11470 + }, + { + "epoch": 1.2130922165820643, + "grad_norm": 9.25, + "learning_rate": 4.430361646733777e-06, + "loss": 2.4677, + "step": 11471 + }, + { + "epoch": 1.2131979695431472, + "grad_norm": 8.125, + "learning_rate": 4.430252817987194e-06, + "loss": 2.3144, + "step": 11472 + }, + { + "epoch": 1.21330372250423, + "grad_norm": 8.125, + "learning_rate": 4.430143980182724e-06, + "loss": 2.6451, + "step": 11473 + }, + { + "epoch": 1.213409475465313, + "grad_norm": 8.3125, + "learning_rate": 4.4300351333208775e-06, + "loss": 2.5276, + "step": 11474 + }, + { + "epoch": 1.213515228426396, + "grad_norm": 7.96875, + "learning_rate": 4.429926277402165e-06, + "loss": 2.3825, + "step": 11475 + }, + { + "epoch": 1.2136209813874788, + "grad_norm": 8.8125, + "learning_rate": 4.429817412427099e-06, + "loss": 2.5746, + "step": 11476 + }, + { + "epoch": 1.2137267343485618, + "grad_norm": 7.96875, + "learning_rate": 4.429708538396188e-06, + "loss": 2.5712, + "step": 11477 + }, + { + "epoch": 1.2138324873096447, + "grad_norm": 7.9375, + "learning_rate": 4.429599655309944e-06, + "loss": 2.6355, + "step": 11478 + }, + { + "epoch": 1.2139382402707275, + "grad_norm": 8.25, + "learning_rate": 4.429490763168879e-06, + "loss": 2.4783, + "step": 11479 + }, + { + "epoch": 1.2140439932318106, + "grad_norm": 8.625, + "learning_rate": 4.429381861973502e-06, + "loss": 2.2357, + "step": 11480 + }, + { + "epoch": 1.2141497461928934, + "grad_norm": 7.875, + "learning_rate": 4.429272951724326e-06, + "loss": 2.4544, + "step": 11481 + }, + { + "epoch": 1.2142554991539762, + "grad_norm": 8.0625, + "learning_rate": 4.4291640324218595e-06, + "loss": 2.2603, + "step": 11482 + }, + { + "epoch": 1.2143612521150593, + "grad_norm": 7.84375, + "learning_rate": 4.429055104066616e-06, + "loss": 2.334, + "step": 11483 + }, + { + "epoch": 1.2144670050761421, + "grad_norm": 8.5625, + "learning_rate": 4.428946166659107e-06, + "loss": 2.6634, + "step": 11484 + }, + { + "epoch": 1.214572758037225, + "grad_norm": 7.9375, + "learning_rate": 4.428837220199841e-06, + "loss": 2.5319, + "step": 11485 + }, + { + "epoch": 1.214678510998308, + "grad_norm": 8.8125, + "learning_rate": 4.428728264689331e-06, + "loss": 2.3433, + "step": 11486 + }, + { + "epoch": 1.2147842639593909, + "grad_norm": 8.375, + "learning_rate": 4.428619300128089e-06, + "loss": 2.6709, + "step": 11487 + }, + { + "epoch": 1.2148900169204737, + "grad_norm": 8.1875, + "learning_rate": 4.4285103265166245e-06, + "loss": 2.3033, + "step": 11488 + }, + { + "epoch": 1.2149957698815568, + "grad_norm": 8.625, + "learning_rate": 4.428401343855451e-06, + "loss": 2.6193, + "step": 11489 + }, + { + "epoch": 1.2151015228426396, + "grad_norm": 8.3125, + "learning_rate": 4.428292352145077e-06, + "loss": 2.629, + "step": 11490 + }, + { + "epoch": 1.2152072758037225, + "grad_norm": 9.6875, + "learning_rate": 4.428183351386016e-06, + "loss": 2.641, + "step": 11491 + }, + { + "epoch": 1.2153130287648053, + "grad_norm": 8.125, + "learning_rate": 4.428074341578779e-06, + "loss": 2.5442, + "step": 11492 + }, + { + "epoch": 1.2154187817258884, + "grad_norm": 11.3125, + "learning_rate": 4.4279653227238794e-06, + "loss": 2.2782, + "step": 11493 + }, + { + "epoch": 1.2155245346869712, + "grad_norm": 8.1875, + "learning_rate": 4.427856294821826e-06, + "loss": 2.5344, + "step": 11494 + }, + { + "epoch": 1.2156302876480543, + "grad_norm": 8.3125, + "learning_rate": 4.4277472578731315e-06, + "loss": 2.7304, + "step": 11495 + }, + { + "epoch": 1.215736040609137, + "grad_norm": 8.3125, + "learning_rate": 4.427638211878307e-06, + "loss": 2.2101, + "step": 11496 + }, + { + "epoch": 1.21584179357022, + "grad_norm": 8.125, + "learning_rate": 4.427529156837865e-06, + "loss": 2.4066, + "step": 11497 + }, + { + "epoch": 1.2159475465313028, + "grad_norm": 7.8125, + "learning_rate": 4.427420092752317e-06, + "loss": 2.5078, + "step": 11498 + }, + { + "epoch": 1.2160532994923858, + "grad_norm": 8.6875, + "learning_rate": 4.4273110196221756e-06, + "loss": 2.548, + "step": 11499 + }, + { + "epoch": 1.2161590524534687, + "grad_norm": 7.90625, + "learning_rate": 4.427201937447951e-06, + "loss": 2.4623, + "step": 11500 + }, + { + "epoch": 1.2162648054145517, + "grad_norm": 8.375, + "learning_rate": 4.427092846230156e-06, + "loss": 2.6679, + "step": 11501 + }, + { + "epoch": 1.2163705583756346, + "grad_norm": 8.3125, + "learning_rate": 4.426983745969302e-06, + "loss": 2.673, + "step": 11502 + }, + { + "epoch": 1.2164763113367174, + "grad_norm": 8.5, + "learning_rate": 4.426874636665902e-06, + "loss": 2.6064, + "step": 11503 + }, + { + "epoch": 1.2165820642978002, + "grad_norm": 8.3125, + "learning_rate": 4.4267655183204675e-06, + "loss": 2.6178, + "step": 11504 + }, + { + "epoch": 1.2166878172588833, + "grad_norm": 8.25, + "learning_rate": 4.426656390933511e-06, + "loss": 2.5266, + "step": 11505 + }, + { + "epoch": 1.2167935702199661, + "grad_norm": 7.875, + "learning_rate": 4.426547254505542e-06, + "loss": 2.5001, + "step": 11506 + }, + { + "epoch": 1.216899323181049, + "grad_norm": 8.125, + "learning_rate": 4.426438109037076e-06, + "loss": 2.6123, + "step": 11507 + }, + { + "epoch": 1.217005076142132, + "grad_norm": 8.25, + "learning_rate": 4.426328954528624e-06, + "loss": 2.4255, + "step": 11508 + }, + { + "epoch": 1.2171108291032149, + "grad_norm": 8.0, + "learning_rate": 4.426219790980698e-06, + "loss": 2.4273, + "step": 11509 + }, + { + "epoch": 1.2172165820642977, + "grad_norm": 8.625, + "learning_rate": 4.426110618393809e-06, + "loss": 2.4367, + "step": 11510 + }, + { + "epoch": 1.2173223350253808, + "grad_norm": 8.0625, + "learning_rate": 4.4260014367684725e-06, + "loss": 2.421, + "step": 11511 + }, + { + "epoch": 1.2174280879864636, + "grad_norm": 12.5625, + "learning_rate": 4.425892246105198e-06, + "loss": 2.3739, + "step": 11512 + }, + { + "epoch": 1.2175338409475465, + "grad_norm": 8.25, + "learning_rate": 4.4257830464044995e-06, + "loss": 2.5521, + "step": 11513 + }, + { + "epoch": 1.2176395939086295, + "grad_norm": 8.375, + "learning_rate": 4.425673837666888e-06, + "loss": 2.4989, + "step": 11514 + }, + { + "epoch": 1.2177453468697124, + "grad_norm": 8.125, + "learning_rate": 4.425564619892878e-06, + "loss": 2.5624, + "step": 11515 + }, + { + "epoch": 1.2178510998307952, + "grad_norm": 7.78125, + "learning_rate": 4.425455393082979e-06, + "loss": 2.6033, + "step": 11516 + }, + { + "epoch": 1.2179568527918783, + "grad_norm": 8.9375, + "learning_rate": 4.4253461572377065e-06, + "loss": 2.6683, + "step": 11517 + }, + { + "epoch": 1.218062605752961, + "grad_norm": 8.25, + "learning_rate": 4.425236912357572e-06, + "loss": 2.6227, + "step": 11518 + }, + { + "epoch": 1.218168358714044, + "grad_norm": 8.25, + "learning_rate": 4.4251276584430875e-06, + "loss": 2.8126, + "step": 11519 + }, + { + "epoch": 1.218274111675127, + "grad_norm": 8.5625, + "learning_rate": 4.425018395494767e-06, + "loss": 2.5659, + "step": 11520 + }, + { + "epoch": 1.2183798646362098, + "grad_norm": 7.875, + "learning_rate": 4.4249091235131215e-06, + "loss": 2.5017, + "step": 11521 + }, + { + "epoch": 1.2184856175972927, + "grad_norm": 8.3125, + "learning_rate": 4.424799842498665e-06, + "loss": 2.6809, + "step": 11522 + }, + { + "epoch": 1.2185913705583755, + "grad_norm": 8.0, + "learning_rate": 4.4246905524519105e-06, + "loss": 2.4221, + "step": 11523 + }, + { + "epoch": 1.2186971235194586, + "grad_norm": 8.375, + "learning_rate": 4.42458125337337e-06, + "loss": 2.6353, + "step": 11524 + }, + { + "epoch": 1.2188028764805414, + "grad_norm": 7.8125, + "learning_rate": 4.424471945263557e-06, + "loss": 2.5075, + "step": 11525 + }, + { + "epoch": 1.2189086294416245, + "grad_norm": 7.96875, + "learning_rate": 4.424362628122983e-06, + "loss": 2.6403, + "step": 11526 + }, + { + "epoch": 1.2190143824027073, + "grad_norm": 8.0, + "learning_rate": 4.424253301952164e-06, + "loss": 2.6237, + "step": 11527 + }, + { + "epoch": 1.2191201353637902, + "grad_norm": 8.125, + "learning_rate": 4.424143966751611e-06, + "loss": 2.6195, + "step": 11528 + }, + { + "epoch": 1.219225888324873, + "grad_norm": 8.375, + "learning_rate": 4.424034622521837e-06, + "loss": 2.7373, + "step": 11529 + }, + { + "epoch": 1.219331641285956, + "grad_norm": 8.1875, + "learning_rate": 4.423925269263355e-06, + "loss": 2.5264, + "step": 11530 + }, + { + "epoch": 1.219437394247039, + "grad_norm": 8.3125, + "learning_rate": 4.423815906976679e-06, + "loss": 2.6083, + "step": 11531 + }, + { + "epoch": 1.2195431472081217, + "grad_norm": 8.625, + "learning_rate": 4.423706535662322e-06, + "loss": 2.7166, + "step": 11532 + }, + { + "epoch": 1.2196489001692048, + "grad_norm": 8.375, + "learning_rate": 4.423597155320797e-06, + "loss": 2.6716, + "step": 11533 + }, + { + "epoch": 1.2197546531302876, + "grad_norm": 8.125, + "learning_rate": 4.423487765952617e-06, + "loss": 2.3792, + "step": 11534 + }, + { + "epoch": 1.2198604060913705, + "grad_norm": 8.5625, + "learning_rate": 4.423378367558295e-06, + "loss": 2.7613, + "step": 11535 + }, + { + "epoch": 1.2199661590524535, + "grad_norm": 8.0625, + "learning_rate": 4.423268960138346e-06, + "loss": 2.5729, + "step": 11536 + }, + { + "epoch": 1.2200719120135364, + "grad_norm": 8.25, + "learning_rate": 4.423159543693282e-06, + "loss": 2.3331, + "step": 11537 + }, + { + "epoch": 1.2201776649746192, + "grad_norm": 8.25, + "learning_rate": 4.4230501182236164e-06, + "loss": 2.288, + "step": 11538 + }, + { + "epoch": 1.2202834179357023, + "grad_norm": 8.375, + "learning_rate": 4.422940683729864e-06, + "loss": 2.5174, + "step": 11539 + }, + { + "epoch": 1.2203891708967851, + "grad_norm": 8.4375, + "learning_rate": 4.422831240212536e-06, + "loss": 2.6485, + "step": 11540 + }, + { + "epoch": 1.220494923857868, + "grad_norm": 8.6875, + "learning_rate": 4.422721787672148e-06, + "loss": 2.398, + "step": 11541 + }, + { + "epoch": 1.220600676818951, + "grad_norm": 8.875, + "learning_rate": 4.422612326109214e-06, + "loss": 2.5781, + "step": 11542 + }, + { + "epoch": 1.2207064297800339, + "grad_norm": 8.1875, + "learning_rate": 4.422502855524246e-06, + "loss": 2.563, + "step": 11543 + }, + { + "epoch": 1.2208121827411167, + "grad_norm": 8.25, + "learning_rate": 4.422393375917757e-06, + "loss": 2.4744, + "step": 11544 + }, + { + "epoch": 1.2209179357021998, + "grad_norm": 8.125, + "learning_rate": 4.422283887290264e-06, + "loss": 2.2082, + "step": 11545 + }, + { + "epoch": 1.2210236886632826, + "grad_norm": 8.125, + "learning_rate": 4.422174389642279e-06, + "loss": 2.7518, + "step": 11546 + }, + { + "epoch": 1.2211294416243654, + "grad_norm": 8.3125, + "learning_rate": 4.422064882974314e-06, + "loss": 2.458, + "step": 11547 + }, + { + "epoch": 1.2212351945854483, + "grad_norm": 8.125, + "learning_rate": 4.4219553672868856e-06, + "loss": 2.5923, + "step": 11548 + }, + { + "epoch": 1.2213409475465313, + "grad_norm": 7.4375, + "learning_rate": 4.421845842580507e-06, + "loss": 2.5823, + "step": 11549 + }, + { + "epoch": 1.2214467005076142, + "grad_norm": 8.375, + "learning_rate": 4.421736308855692e-06, + "loss": 2.6059, + "step": 11550 + }, + { + "epoch": 1.2215524534686972, + "grad_norm": 8.1875, + "learning_rate": 4.4216267661129544e-06, + "loss": 2.6101, + "step": 11551 + }, + { + "epoch": 1.22165820642978, + "grad_norm": 8.6875, + "learning_rate": 4.421517214352807e-06, + "loss": 2.5248, + "step": 11552 + }, + { + "epoch": 1.221763959390863, + "grad_norm": 8.25, + "learning_rate": 4.421407653575767e-06, + "loss": 2.5823, + "step": 11553 + }, + { + "epoch": 1.2218697123519457, + "grad_norm": 8.125, + "learning_rate": 4.421298083782345e-06, + "loss": 2.6291, + "step": 11554 + }, + { + "epoch": 1.2219754653130288, + "grad_norm": 9.0, + "learning_rate": 4.421188504973059e-06, + "loss": 2.5695, + "step": 11555 + }, + { + "epoch": 1.2220812182741116, + "grad_norm": 8.75, + "learning_rate": 4.421078917148419e-06, + "loss": 2.6553, + "step": 11556 + }, + { + "epoch": 1.2221869712351947, + "grad_norm": 8.9375, + "learning_rate": 4.420969320308942e-06, + "loss": 2.6778, + "step": 11557 + }, + { + "epoch": 1.2222927241962775, + "grad_norm": 8.0625, + "learning_rate": 4.420859714455143e-06, + "loss": 2.4162, + "step": 11558 + }, + { + "epoch": 1.2223984771573604, + "grad_norm": 8.625, + "learning_rate": 4.420750099587533e-06, + "loss": 2.424, + "step": 11559 + }, + { + "epoch": 1.2225042301184432, + "grad_norm": 9.3125, + "learning_rate": 4.420640475706629e-06, + "loss": 2.7519, + "step": 11560 + }, + { + "epoch": 1.2226099830795263, + "grad_norm": 7.96875, + "learning_rate": 4.420530842812945e-06, + "loss": 2.4065, + "step": 11561 + }, + { + "epoch": 1.2227157360406091, + "grad_norm": 8.0, + "learning_rate": 4.420421200906994e-06, + "loss": 2.4711, + "step": 11562 + }, + { + "epoch": 1.222821489001692, + "grad_norm": 7.8125, + "learning_rate": 4.420311549989294e-06, + "loss": 2.5215, + "step": 11563 + }, + { + "epoch": 1.222927241962775, + "grad_norm": 7.84375, + "learning_rate": 4.420201890060356e-06, + "loss": 2.5413, + "step": 11564 + }, + { + "epoch": 1.2230329949238579, + "grad_norm": 8.8125, + "learning_rate": 4.420092221120696e-06, + "loss": 2.6156, + "step": 11565 + }, + { + "epoch": 1.2231387478849407, + "grad_norm": 8.1875, + "learning_rate": 4.419982543170829e-06, + "loss": 2.5661, + "step": 11566 + }, + { + "epoch": 1.2232445008460238, + "grad_norm": 8.125, + "learning_rate": 4.419872856211268e-06, + "loss": 2.3787, + "step": 11567 + }, + { + "epoch": 1.2233502538071066, + "grad_norm": 8.875, + "learning_rate": 4.41976316024253e-06, + "loss": 2.8006, + "step": 11568 + }, + { + "epoch": 1.2234560067681894, + "grad_norm": 8.5, + "learning_rate": 4.4196534552651286e-06, + "loss": 2.6996, + "step": 11569 + }, + { + "epoch": 1.2235617597292725, + "grad_norm": 8.1875, + "learning_rate": 4.419543741279578e-06, + "loss": 2.5802, + "step": 11570 + }, + { + "epoch": 1.2236675126903553, + "grad_norm": 8.3125, + "learning_rate": 4.419434018286394e-06, + "loss": 2.6713, + "step": 11571 + }, + { + "epoch": 1.2237732656514382, + "grad_norm": 8.4375, + "learning_rate": 4.419324286286091e-06, + "loss": 2.6097, + "step": 11572 + }, + { + "epoch": 1.2238790186125212, + "grad_norm": 8.25, + "learning_rate": 4.4192145452791834e-06, + "loss": 2.4408, + "step": 11573 + }, + { + "epoch": 1.223984771573604, + "grad_norm": 8.5, + "learning_rate": 4.419104795266188e-06, + "loss": 2.5429, + "step": 11574 + }, + { + "epoch": 1.224090524534687, + "grad_norm": 7.9375, + "learning_rate": 4.418995036247619e-06, + "loss": 2.3224, + "step": 11575 + }, + { + "epoch": 1.22419627749577, + "grad_norm": 8.5625, + "learning_rate": 4.41888526822399e-06, + "loss": 2.7048, + "step": 11576 + }, + { + "epoch": 1.2243020304568528, + "grad_norm": 8.4375, + "learning_rate": 4.418775491195817e-06, + "loss": 2.5414, + "step": 11577 + }, + { + "epoch": 1.2244077834179357, + "grad_norm": 8.25, + "learning_rate": 4.418665705163616e-06, + "loss": 2.5326, + "step": 11578 + }, + { + "epoch": 1.2245135363790185, + "grad_norm": 8.25, + "learning_rate": 4.418555910127903e-06, + "loss": 2.6059, + "step": 11579 + }, + { + "epoch": 1.2246192893401016, + "grad_norm": 8.1875, + "learning_rate": 4.418446106089189e-06, + "loss": 2.5115, + "step": 11580 + }, + { + "epoch": 1.2247250423011844, + "grad_norm": 8.0625, + "learning_rate": 4.4183362930479945e-06, + "loss": 2.3823, + "step": 11581 + }, + { + "epoch": 1.2248307952622675, + "grad_norm": 8.5, + "learning_rate": 4.418226471004831e-06, + "loss": 2.3674, + "step": 11582 + }, + { + "epoch": 1.2249365482233503, + "grad_norm": 8.25, + "learning_rate": 4.418116639960215e-06, + "loss": 2.6193, + "step": 11583 + }, + { + "epoch": 1.2250423011844331, + "grad_norm": 7.78125, + "learning_rate": 4.418006799914664e-06, + "loss": 2.579, + "step": 11584 + }, + { + "epoch": 1.225148054145516, + "grad_norm": 8.5, + "learning_rate": 4.417896950868689e-06, + "loss": 2.3602, + "step": 11585 + }, + { + "epoch": 1.225253807106599, + "grad_norm": 8.625, + "learning_rate": 4.417787092822809e-06, + "loss": 2.6982, + "step": 11586 + }, + { + "epoch": 1.2253595600676819, + "grad_norm": 7.71875, + "learning_rate": 4.417677225777539e-06, + "loss": 2.4705, + "step": 11587 + }, + { + "epoch": 1.2254653130287647, + "grad_norm": 8.4375, + "learning_rate": 4.417567349733394e-06, + "loss": 2.4605, + "step": 11588 + }, + { + "epoch": 1.2255710659898478, + "grad_norm": 8.375, + "learning_rate": 4.417457464690889e-06, + "loss": 2.7154, + "step": 11589 + }, + { + "epoch": 1.2256768189509306, + "grad_norm": 8.0625, + "learning_rate": 4.417347570650541e-06, + "loss": 2.5052, + "step": 11590 + }, + { + "epoch": 1.2257825719120135, + "grad_norm": 8.375, + "learning_rate": 4.417237667612865e-06, + "loss": 2.512, + "step": 11591 + }, + { + "epoch": 1.2258883248730965, + "grad_norm": 8.1875, + "learning_rate": 4.4171277555783766e-06, + "loss": 2.524, + "step": 11592 + }, + { + "epoch": 1.2259940778341794, + "grad_norm": 8.125, + "learning_rate": 4.417017834547591e-06, + "loss": 2.6474, + "step": 11593 + }, + { + "epoch": 1.2260998307952622, + "grad_norm": 8.25, + "learning_rate": 4.416907904521026e-06, + "loss": 2.5759, + "step": 11594 + }, + { + "epoch": 1.2262055837563453, + "grad_norm": 8.75, + "learning_rate": 4.416797965499195e-06, + "loss": 2.5819, + "step": 11595 + }, + { + "epoch": 1.226311336717428, + "grad_norm": 9.0, + "learning_rate": 4.416688017482616e-06, + "loss": 2.5386, + "step": 11596 + }, + { + "epoch": 1.226417089678511, + "grad_norm": 8.0625, + "learning_rate": 4.416578060471803e-06, + "loss": 2.5011, + "step": 11597 + }, + { + "epoch": 1.226522842639594, + "grad_norm": 8.4375, + "learning_rate": 4.416468094467275e-06, + "loss": 2.7417, + "step": 11598 + }, + { + "epoch": 1.2266285956006768, + "grad_norm": 8.4375, + "learning_rate": 4.416358119469543e-06, + "loss": 2.311, + "step": 11599 + }, + { + "epoch": 1.2267343485617597, + "grad_norm": 8.0, + "learning_rate": 4.416248135479128e-06, + "loss": 2.4801, + "step": 11600 + }, + { + "epoch": 1.2268401015228427, + "grad_norm": 8.0625, + "learning_rate": 4.416138142496545e-06, + "loss": 2.4361, + "step": 11601 + }, + { + "epoch": 1.2269458544839256, + "grad_norm": 8.0, + "learning_rate": 4.416028140522308e-06, + "loss": 2.5136, + "step": 11602 + }, + { + "epoch": 1.2270516074450084, + "grad_norm": 8.4375, + "learning_rate": 4.4159181295569355e-06, + "loss": 2.5266, + "step": 11603 + }, + { + "epoch": 1.2271573604060912, + "grad_norm": 7.75, + "learning_rate": 4.415808109600942e-06, + "loss": 2.4331, + "step": 11604 + }, + { + "epoch": 1.2272631133671743, + "grad_norm": 8.625, + "learning_rate": 4.415698080654845e-06, + "loss": 2.5282, + "step": 11605 + }, + { + "epoch": 1.2273688663282571, + "grad_norm": 8.25, + "learning_rate": 4.415588042719161e-06, + "loss": 2.5032, + "step": 11606 + }, + { + "epoch": 1.2274746192893402, + "grad_norm": 7.71875, + "learning_rate": 4.4154779957944045e-06, + "loss": 2.5599, + "step": 11607 + }, + { + "epoch": 1.227580372250423, + "grad_norm": 7.8125, + "learning_rate": 4.415367939881094e-06, + "loss": 2.3601, + "step": 11608 + }, + { + "epoch": 1.2276861252115059, + "grad_norm": 8.25, + "learning_rate": 4.4152578749797446e-06, + "loss": 2.5379, + "step": 11609 + }, + { + "epoch": 1.2277918781725887, + "grad_norm": 7.90625, + "learning_rate": 4.4151478010908735e-06, + "loss": 2.6038, + "step": 11610 + }, + { + "epoch": 1.2278976311336718, + "grad_norm": 8.0625, + "learning_rate": 4.415037718214997e-06, + "loss": 2.6112, + "step": 11611 + }, + { + "epoch": 1.2280033840947546, + "grad_norm": 8.125, + "learning_rate": 4.414927626352632e-06, + "loss": 2.6089, + "step": 11612 + }, + { + "epoch": 1.2281091370558377, + "grad_norm": 8.3125, + "learning_rate": 4.414817525504295e-06, + "loss": 2.7184, + "step": 11613 + }, + { + "epoch": 1.2282148900169205, + "grad_norm": 8.8125, + "learning_rate": 4.414707415670502e-06, + "loss": 2.9538, + "step": 11614 + }, + { + "epoch": 1.2283206429780034, + "grad_norm": 8.0, + "learning_rate": 4.41459729685177e-06, + "loss": 2.3832, + "step": 11615 + }, + { + "epoch": 1.2284263959390862, + "grad_norm": 8.1875, + "learning_rate": 4.4144871690486165e-06, + "loss": 2.6727, + "step": 11616 + }, + { + "epoch": 1.2285321489001693, + "grad_norm": 8.0, + "learning_rate": 4.414377032261557e-06, + "loss": 2.5613, + "step": 11617 + }, + { + "epoch": 1.228637901861252, + "grad_norm": 8.25, + "learning_rate": 4.4142668864911095e-06, + "loss": 2.3062, + "step": 11618 + }, + { + "epoch": 1.228743654822335, + "grad_norm": 8.125, + "learning_rate": 4.414156731737791e-06, + "loss": 2.367, + "step": 11619 + }, + { + "epoch": 1.228849407783418, + "grad_norm": 8.0625, + "learning_rate": 4.4140465680021175e-06, + "loss": 2.5858, + "step": 11620 + }, + { + "epoch": 1.2289551607445008, + "grad_norm": 8.375, + "learning_rate": 4.413936395284605e-06, + "loss": 2.4498, + "step": 11621 + }, + { + "epoch": 1.2290609137055837, + "grad_norm": 8.5625, + "learning_rate": 4.413826213585773e-06, + "loss": 2.4463, + "step": 11622 + }, + { + "epoch": 1.2291666666666667, + "grad_norm": 8.3125, + "learning_rate": 4.413716022906137e-06, + "loss": 2.4697, + "step": 11623 + }, + { + "epoch": 1.2292724196277496, + "grad_norm": 7.8125, + "learning_rate": 4.413605823246215e-06, + "loss": 2.6728, + "step": 11624 + }, + { + "epoch": 1.2293781725888324, + "grad_norm": 8.25, + "learning_rate": 4.413495614606523e-06, + "loss": 2.6968, + "step": 11625 + }, + { + "epoch": 1.2294839255499155, + "grad_norm": 8.375, + "learning_rate": 4.4133853969875785e-06, + "loss": 2.4184, + "step": 11626 + }, + { + "epoch": 1.2295896785109983, + "grad_norm": 8.8125, + "learning_rate": 4.413275170389899e-06, + "loss": 2.5557, + "step": 11627 + }, + { + "epoch": 1.2296954314720812, + "grad_norm": 8.625, + "learning_rate": 4.413164934814002e-06, + "loss": 2.5398, + "step": 11628 + }, + { + "epoch": 1.2298011844331642, + "grad_norm": 8.5, + "learning_rate": 4.413054690260404e-06, + "loss": 2.603, + "step": 11629 + }, + { + "epoch": 1.229906937394247, + "grad_norm": 7.875, + "learning_rate": 4.412944436729623e-06, + "loss": 2.4707, + "step": 11630 + }, + { + "epoch": 1.23001269035533, + "grad_norm": 8.75, + "learning_rate": 4.412834174222175e-06, + "loss": 2.795, + "step": 11631 + }, + { + "epoch": 1.230118443316413, + "grad_norm": 9.25, + "learning_rate": 4.41272390273858e-06, + "loss": 2.4356, + "step": 11632 + }, + { + "epoch": 1.2302241962774958, + "grad_norm": 7.90625, + "learning_rate": 4.4126136222793536e-06, + "loss": 2.546, + "step": 11633 + }, + { + "epoch": 1.2303299492385786, + "grad_norm": 7.90625, + "learning_rate": 4.412503332845013e-06, + "loss": 2.6702, + "step": 11634 + }, + { + "epoch": 1.2304357021996615, + "grad_norm": 7.90625, + "learning_rate": 4.412393034436077e-06, + "loss": 2.7961, + "step": 11635 + }, + { + "epoch": 1.2305414551607445, + "grad_norm": 7.84375, + "learning_rate": 4.412282727053063e-06, + "loss": 2.5344, + "step": 11636 + }, + { + "epoch": 1.2306472081218274, + "grad_norm": 8.375, + "learning_rate": 4.4121724106964885e-06, + "loss": 2.6571, + "step": 11637 + }, + { + "epoch": 1.2307529610829104, + "grad_norm": 8.6875, + "learning_rate": 4.41206208536687e-06, + "loss": 2.6169, + "step": 11638 + }, + { + "epoch": 1.2308587140439933, + "grad_norm": 8.3125, + "learning_rate": 4.411951751064726e-06, + "loss": 2.5732, + "step": 11639 + }, + { + "epoch": 1.2309644670050761, + "grad_norm": 8.0625, + "learning_rate": 4.4118414077905755e-06, + "loss": 2.5758, + "step": 11640 + }, + { + "epoch": 1.231070219966159, + "grad_norm": 8.3125, + "learning_rate": 4.4117310555449346e-06, + "loss": 2.4775, + "step": 11641 + }, + { + "epoch": 1.231175972927242, + "grad_norm": 7.8125, + "learning_rate": 4.411620694328321e-06, + "loss": 2.5771, + "step": 11642 + }, + { + "epoch": 1.2312817258883249, + "grad_norm": 8.375, + "learning_rate": 4.411510324141255e-06, + "loss": 2.7507, + "step": 11643 + }, + { + "epoch": 1.2313874788494077, + "grad_norm": 7.9375, + "learning_rate": 4.411399944984252e-06, + "loss": 2.6258, + "step": 11644 + }, + { + "epoch": 1.2314932318104908, + "grad_norm": 8.25, + "learning_rate": 4.41128955685783e-06, + "loss": 2.4861, + "step": 11645 + }, + { + "epoch": 1.2315989847715736, + "grad_norm": 7.6875, + "learning_rate": 4.411179159762509e-06, + "loss": 2.6316, + "step": 11646 + }, + { + "epoch": 1.2317047377326564, + "grad_norm": 8.0625, + "learning_rate": 4.411068753698805e-06, + "loss": 2.5213, + "step": 11647 + }, + { + "epoch": 1.2318104906937395, + "grad_norm": 8.3125, + "learning_rate": 4.4109583386672376e-06, + "loss": 2.4708, + "step": 11648 + }, + { + "epoch": 1.2319162436548223, + "grad_norm": 8.4375, + "learning_rate": 4.4108479146683235e-06, + "loss": 2.4662, + "step": 11649 + }, + { + "epoch": 1.2320219966159052, + "grad_norm": 7.96875, + "learning_rate": 4.4107374817025815e-06, + "loss": 2.5376, + "step": 11650 + }, + { + "epoch": 1.2321277495769882, + "grad_norm": 8.125, + "learning_rate": 4.410627039770531e-06, + "loss": 2.3446, + "step": 11651 + }, + { + "epoch": 1.232233502538071, + "grad_norm": 8.25, + "learning_rate": 4.410516588872689e-06, + "loss": 2.7208, + "step": 11652 + }, + { + "epoch": 1.232339255499154, + "grad_norm": 8.25, + "learning_rate": 4.410406129009574e-06, + "loss": 2.4241, + "step": 11653 + }, + { + "epoch": 1.232445008460237, + "grad_norm": 8.125, + "learning_rate": 4.410295660181704e-06, + "loss": 2.5871, + "step": 11654 + }, + { + "epoch": 1.2325507614213198, + "grad_norm": 8.375, + "learning_rate": 4.4101851823895974e-06, + "loss": 2.7142, + "step": 11655 + }, + { + "epoch": 1.2326565143824026, + "grad_norm": 8.25, + "learning_rate": 4.4100746956337735e-06, + "loss": 2.5495, + "step": 11656 + }, + { + "epoch": 1.2327622673434857, + "grad_norm": 9.1875, + "learning_rate": 4.40996419991475e-06, + "loss": 2.7333, + "step": 11657 + }, + { + "epoch": 1.2328680203045685, + "grad_norm": 8.125, + "learning_rate": 4.409853695233046e-06, + "loss": 2.5799, + "step": 11658 + }, + { + "epoch": 1.2329737732656514, + "grad_norm": 7.84375, + "learning_rate": 4.40974318158918e-06, + "loss": 2.4648, + "step": 11659 + }, + { + "epoch": 1.2330795262267342, + "grad_norm": 7.96875, + "learning_rate": 4.409632658983669e-06, + "loss": 2.3677, + "step": 11660 + }, + { + "epoch": 1.2331852791878173, + "grad_norm": 8.5, + "learning_rate": 4.409522127417034e-06, + "loss": 2.5345, + "step": 11661 + }, + { + "epoch": 1.2332910321489001, + "grad_norm": 7.625, + "learning_rate": 4.409411586889793e-06, + "loss": 2.3335, + "step": 11662 + }, + { + "epoch": 1.2333967851099832, + "grad_norm": 7.65625, + "learning_rate": 4.409301037402464e-06, + "loss": 2.4838, + "step": 11663 + }, + { + "epoch": 1.233502538071066, + "grad_norm": 7.90625, + "learning_rate": 4.4091904789555654e-06, + "loss": 2.5238, + "step": 11664 + }, + { + "epoch": 1.2336082910321489, + "grad_norm": 8.3125, + "learning_rate": 4.409079911549617e-06, + "loss": 2.6879, + "step": 11665 + }, + { + "epoch": 1.2337140439932317, + "grad_norm": 8.25, + "learning_rate": 4.408969335185138e-06, + "loss": 2.4763, + "step": 11666 + }, + { + "epoch": 1.2338197969543148, + "grad_norm": 7.96875, + "learning_rate": 4.408858749862647e-06, + "loss": 2.4301, + "step": 11667 + }, + { + "epoch": 1.2339255499153976, + "grad_norm": 8.375, + "learning_rate": 4.408748155582661e-06, + "loss": 2.5998, + "step": 11668 + }, + { + "epoch": 1.2340313028764807, + "grad_norm": 8.9375, + "learning_rate": 4.4086375523457015e-06, + "loss": 2.6306, + "step": 11669 + }, + { + "epoch": 1.2341370558375635, + "grad_norm": 9.8125, + "learning_rate": 4.4085269401522865e-06, + "loss": 2.7231, + "step": 11670 + }, + { + "epoch": 1.2342428087986463, + "grad_norm": 8.5, + "learning_rate": 4.4084163190029346e-06, + "loss": 2.7082, + "step": 11671 + }, + { + "epoch": 1.2343485617597292, + "grad_norm": 9.0, + "learning_rate": 4.408305688898166e-06, + "loss": 2.57, + "step": 11672 + }, + { + "epoch": 1.2344543147208122, + "grad_norm": 7.78125, + "learning_rate": 4.4081950498385e-06, + "loss": 2.4932, + "step": 11673 + }, + { + "epoch": 1.234560067681895, + "grad_norm": 8.5, + "learning_rate": 4.408084401824454e-06, + "loss": 2.7113, + "step": 11674 + }, + { + "epoch": 1.234665820642978, + "grad_norm": 8.5625, + "learning_rate": 4.407973744856548e-06, + "loss": 2.7299, + "step": 11675 + }, + { + "epoch": 1.234771573604061, + "grad_norm": 8.0625, + "learning_rate": 4.407863078935302e-06, + "loss": 2.5167, + "step": 11676 + }, + { + "epoch": 1.2348773265651438, + "grad_norm": 9.0, + "learning_rate": 4.4077524040612355e-06, + "loss": 2.3477, + "step": 11677 + }, + { + "epoch": 1.2349830795262267, + "grad_norm": 8.0625, + "learning_rate": 4.407641720234867e-06, + "loss": 2.6134, + "step": 11678 + }, + { + "epoch": 1.2350888324873097, + "grad_norm": 8.125, + "learning_rate": 4.407531027456715e-06, + "loss": 2.5097, + "step": 11679 + }, + { + "epoch": 1.2351945854483926, + "grad_norm": 7.71875, + "learning_rate": 4.407420325727301e-06, + "loss": 2.5333, + "step": 11680 + }, + { + "epoch": 1.2353003384094754, + "grad_norm": 8.25, + "learning_rate": 4.407309615047143e-06, + "loss": 2.5381, + "step": 11681 + }, + { + "epoch": 1.2354060913705585, + "grad_norm": 8.0625, + "learning_rate": 4.407198895416761e-06, + "loss": 2.6018, + "step": 11682 + }, + { + "epoch": 1.2355118443316413, + "grad_norm": 8.0625, + "learning_rate": 4.407088166836676e-06, + "loss": 2.5873, + "step": 11683 + }, + { + "epoch": 1.2356175972927241, + "grad_norm": 8.125, + "learning_rate": 4.406977429307405e-06, + "loss": 2.5341, + "step": 11684 + }, + { + "epoch": 1.2357233502538072, + "grad_norm": 8.25, + "learning_rate": 4.406866682829469e-06, + "loss": 2.4823, + "step": 11685 + }, + { + "epoch": 1.23582910321489, + "grad_norm": 7.90625, + "learning_rate": 4.406755927403387e-06, + "loss": 2.6336, + "step": 11686 + }, + { + "epoch": 1.2359348561759729, + "grad_norm": 9.0625, + "learning_rate": 4.40664516302968e-06, + "loss": 2.7489, + "step": 11687 + }, + { + "epoch": 1.236040609137056, + "grad_norm": 8.625, + "learning_rate": 4.406534389708867e-06, + "loss": 2.4184, + "step": 11688 + }, + { + "epoch": 1.2361463620981388, + "grad_norm": 8.375, + "learning_rate": 4.406423607441468e-06, + "loss": 2.712, + "step": 11689 + }, + { + "epoch": 1.2362521150592216, + "grad_norm": 7.625, + "learning_rate": 4.406312816228002e-06, + "loss": 2.41, + "step": 11690 + }, + { + "epoch": 1.2363578680203045, + "grad_norm": 8.5625, + "learning_rate": 4.40620201606899e-06, + "loss": 2.41, + "step": 11691 + }, + { + "epoch": 1.2364636209813875, + "grad_norm": 12.8125, + "learning_rate": 4.406091206964952e-06, + "loss": 2.5757, + "step": 11692 + }, + { + "epoch": 1.2365693739424704, + "grad_norm": 8.75, + "learning_rate": 4.405980388916407e-06, + "loss": 2.4328, + "step": 11693 + }, + { + "epoch": 1.2366751269035534, + "grad_norm": 8.25, + "learning_rate": 4.405869561923875e-06, + "loss": 2.5573, + "step": 11694 + }, + { + "epoch": 1.2367808798646363, + "grad_norm": 8.25, + "learning_rate": 4.405758725987878e-06, + "loss": 2.5056, + "step": 11695 + }, + { + "epoch": 1.236886632825719, + "grad_norm": 8.25, + "learning_rate": 4.405647881108934e-06, + "loss": 2.3343, + "step": 11696 + }, + { + "epoch": 1.236992385786802, + "grad_norm": 8.4375, + "learning_rate": 4.405537027287564e-06, + "loss": 2.4289, + "step": 11697 + }, + { + "epoch": 1.237098138747885, + "grad_norm": 7.9375, + "learning_rate": 4.405426164524288e-06, + "loss": 2.5228, + "step": 11698 + }, + { + "epoch": 1.2372038917089678, + "grad_norm": 8.4375, + "learning_rate": 4.4053152928196255e-06, + "loss": 2.6487, + "step": 11699 + }, + { + "epoch": 1.2373096446700507, + "grad_norm": 9.0625, + "learning_rate": 4.405204412174099e-06, + "loss": 2.8662, + "step": 11700 + }, + { + "epoch": 1.2374153976311337, + "grad_norm": 8.0625, + "learning_rate": 4.405093522588227e-06, + "loss": 2.8107, + "step": 11701 + }, + { + "epoch": 1.2375211505922166, + "grad_norm": 8.4375, + "learning_rate": 4.4049826240625295e-06, + "loss": 2.6957, + "step": 11702 + }, + { + "epoch": 1.2376269035532994, + "grad_norm": 7.96875, + "learning_rate": 4.404871716597528e-06, + "loss": 2.6167, + "step": 11703 + }, + { + "epoch": 1.2377326565143825, + "grad_norm": 8.375, + "learning_rate": 4.4047608001937425e-06, + "loss": 2.4595, + "step": 11704 + }, + { + "epoch": 1.2378384094754653, + "grad_norm": 8.3125, + "learning_rate": 4.404649874851694e-06, + "loss": 2.5327, + "step": 11705 + }, + { + "epoch": 1.2379441624365481, + "grad_norm": 8.125, + "learning_rate": 4.404538940571902e-06, + "loss": 2.5759, + "step": 11706 + }, + { + "epoch": 1.2380499153976312, + "grad_norm": 8.0, + "learning_rate": 4.4044279973548885e-06, + "loss": 2.5284, + "step": 11707 + }, + { + "epoch": 1.238155668358714, + "grad_norm": 8.1875, + "learning_rate": 4.404317045201173e-06, + "loss": 2.6362, + "step": 11708 + }, + { + "epoch": 1.2382614213197969, + "grad_norm": 9.8125, + "learning_rate": 4.404206084111277e-06, + "loss": 2.7953, + "step": 11709 + }, + { + "epoch": 1.23836717428088, + "grad_norm": 7.78125, + "learning_rate": 4.4040951140857195e-06, + "loss": 2.6969, + "step": 11710 + }, + { + "epoch": 1.2384729272419628, + "grad_norm": 8.5, + "learning_rate": 4.403984135125023e-06, + "loss": 2.7108, + "step": 11711 + }, + { + "epoch": 1.2385786802030456, + "grad_norm": 8.125, + "learning_rate": 4.403873147229707e-06, + "loss": 2.7526, + "step": 11712 + }, + { + "epoch": 1.2386844331641287, + "grad_norm": 9.125, + "learning_rate": 4.403762150400294e-06, + "loss": 2.7222, + "step": 11713 + }, + { + "epoch": 1.2387901861252115, + "grad_norm": 7.84375, + "learning_rate": 4.403651144637303e-06, + "loss": 2.4599, + "step": 11714 + }, + { + "epoch": 1.2388959390862944, + "grad_norm": 8.0625, + "learning_rate": 4.4035401299412565e-06, + "loss": 2.55, + "step": 11715 + }, + { + "epoch": 1.2390016920473774, + "grad_norm": 7.90625, + "learning_rate": 4.403429106312674e-06, + "loss": 2.6443, + "step": 11716 + }, + { + "epoch": 1.2391074450084603, + "grad_norm": 8.0625, + "learning_rate": 4.403318073752078e-06, + "loss": 2.58, + "step": 11717 + }, + { + "epoch": 1.239213197969543, + "grad_norm": 8.5625, + "learning_rate": 4.403207032259988e-06, + "loss": 2.6622, + "step": 11718 + }, + { + "epoch": 1.2393189509306262, + "grad_norm": 8.5, + "learning_rate": 4.403095981836925e-06, + "loss": 2.7549, + "step": 11719 + }, + { + "epoch": 1.239424703891709, + "grad_norm": 8.125, + "learning_rate": 4.402984922483413e-06, + "loss": 2.5091, + "step": 11720 + }, + { + "epoch": 1.2395304568527918, + "grad_norm": 8.375, + "learning_rate": 4.402873854199969e-06, + "loss": 2.6759, + "step": 11721 + }, + { + "epoch": 1.2396362098138747, + "grad_norm": 8.625, + "learning_rate": 4.402762776987117e-06, + "loss": 2.6803, + "step": 11722 + }, + { + "epoch": 1.2397419627749577, + "grad_norm": 8.6875, + "learning_rate": 4.402651690845379e-06, + "loss": 2.771, + "step": 11723 + }, + { + "epoch": 1.2398477157360406, + "grad_norm": 8.5, + "learning_rate": 4.402540595775273e-06, + "loss": 2.8381, + "step": 11724 + }, + { + "epoch": 1.2399534686971236, + "grad_norm": 7.625, + "learning_rate": 4.402429491777323e-06, + "loss": 2.3181, + "step": 11725 + }, + { + "epoch": 1.2400592216582065, + "grad_norm": 8.5, + "learning_rate": 4.402318378852049e-06, + "loss": 2.4851, + "step": 11726 + }, + { + "epoch": 1.2401649746192893, + "grad_norm": 8.0, + "learning_rate": 4.4022072569999716e-06, + "loss": 2.5229, + "step": 11727 + }, + { + "epoch": 1.2402707275803722, + "grad_norm": 7.875, + "learning_rate": 4.4020961262216155e-06, + "loss": 2.5959, + "step": 11728 + }, + { + "epoch": 1.2403764805414552, + "grad_norm": 8.0625, + "learning_rate": 4.4019849865175e-06, + "loss": 2.6418, + "step": 11729 + }, + { + "epoch": 1.240482233502538, + "grad_norm": 8.5, + "learning_rate": 4.401873837888145e-06, + "loss": 2.3127, + "step": 11730 + }, + { + "epoch": 1.240587986463621, + "grad_norm": 8.1875, + "learning_rate": 4.401762680334077e-06, + "loss": 2.467, + "step": 11731 + }, + { + "epoch": 1.240693739424704, + "grad_norm": 8.375, + "learning_rate": 4.401651513855812e-06, + "loss": 2.4954, + "step": 11732 + }, + { + "epoch": 1.2407994923857868, + "grad_norm": 8.625, + "learning_rate": 4.401540338453875e-06, + "loss": 2.6496, + "step": 11733 + }, + { + "epoch": 1.2409052453468696, + "grad_norm": 8.375, + "learning_rate": 4.401429154128787e-06, + "loss": 2.6482, + "step": 11734 + }, + { + "epoch": 1.2410109983079527, + "grad_norm": 7.8125, + "learning_rate": 4.401317960881069e-06, + "loss": 2.3378, + "step": 11735 + }, + { + "epoch": 1.2411167512690355, + "grad_norm": 8.5625, + "learning_rate": 4.401206758711244e-06, + "loss": 2.427, + "step": 11736 + }, + { + "epoch": 1.2412225042301184, + "grad_norm": 8.0625, + "learning_rate": 4.4010955476198334e-06, + "loss": 2.3564, + "step": 11737 + }, + { + "epoch": 1.2413282571912014, + "grad_norm": 7.84375, + "learning_rate": 4.4009843276073585e-06, + "loss": 2.4338, + "step": 11738 + }, + { + "epoch": 1.2414340101522843, + "grad_norm": 8.25, + "learning_rate": 4.400873098674342e-06, + "loss": 2.6043, + "step": 11739 + }, + { + "epoch": 1.2415397631133671, + "grad_norm": 7.875, + "learning_rate": 4.400761860821306e-06, + "loss": 2.3197, + "step": 11740 + }, + { + "epoch": 1.2416455160744502, + "grad_norm": 8.5625, + "learning_rate": 4.40065061404877e-06, + "loss": 2.583, + "step": 11741 + }, + { + "epoch": 1.241751269035533, + "grad_norm": 8.5625, + "learning_rate": 4.40053935835726e-06, + "loss": 2.5549, + "step": 11742 + }, + { + "epoch": 1.2418570219966159, + "grad_norm": 8.25, + "learning_rate": 4.400428093747295e-06, + "loss": 2.527, + "step": 11743 + }, + { + "epoch": 1.241962774957699, + "grad_norm": 8.1875, + "learning_rate": 4.400316820219399e-06, + "loss": 2.5, + "step": 11744 + }, + { + "epoch": 1.2420685279187818, + "grad_norm": 8.3125, + "learning_rate": 4.400205537774092e-06, + "loss": 2.4573, + "step": 11745 + }, + { + "epoch": 1.2421742808798646, + "grad_norm": 8.4375, + "learning_rate": 4.400094246411899e-06, + "loss": 2.4867, + "step": 11746 + }, + { + "epoch": 1.2422800338409474, + "grad_norm": 8.3125, + "learning_rate": 4.399982946133341e-06, + "loss": 2.4579, + "step": 11747 + }, + { + "epoch": 1.2423857868020305, + "grad_norm": 8.25, + "learning_rate": 4.399871636938939e-06, + "loss": 2.599, + "step": 11748 + }, + { + "epoch": 1.2424915397631133, + "grad_norm": 8.0625, + "learning_rate": 4.399760318829217e-06, + "loss": 2.4654, + "step": 11749 + }, + { + "epoch": 1.2425972927241964, + "grad_norm": 7.84375, + "learning_rate": 4.399648991804696e-06, + "loss": 2.4815, + "step": 11750 + }, + { + "epoch": 1.2427030456852792, + "grad_norm": 8.4375, + "learning_rate": 4.3995376558659005e-06, + "loss": 2.6459, + "step": 11751 + }, + { + "epoch": 1.242808798646362, + "grad_norm": 8.6875, + "learning_rate": 4.399426311013351e-06, + "loss": 2.4868, + "step": 11752 + }, + { + "epoch": 1.242914551607445, + "grad_norm": 8.6875, + "learning_rate": 4.39931495724757e-06, + "loss": 2.7581, + "step": 11753 + }, + { + "epoch": 1.243020304568528, + "grad_norm": 8.25, + "learning_rate": 4.399203594569082e-06, + "loss": 2.6466, + "step": 11754 + }, + { + "epoch": 1.2431260575296108, + "grad_norm": 8.125, + "learning_rate": 4.399092222978407e-06, + "loss": 2.5534, + "step": 11755 + }, + { + "epoch": 1.2432318104906936, + "grad_norm": 8.4375, + "learning_rate": 4.39898084247607e-06, + "loss": 2.5149, + "step": 11756 + }, + { + "epoch": 1.2433375634517767, + "grad_norm": 9.25, + "learning_rate": 4.398869453062592e-06, + "loss": 2.7535, + "step": 11757 + }, + { + "epoch": 1.2434433164128595, + "grad_norm": 8.0625, + "learning_rate": 4.3987580547384965e-06, + "loss": 2.7339, + "step": 11758 + }, + { + "epoch": 1.2435490693739424, + "grad_norm": 8.125, + "learning_rate": 4.398646647504306e-06, + "loss": 2.603, + "step": 11759 + }, + { + "epoch": 1.2436548223350254, + "grad_norm": 8.0625, + "learning_rate": 4.398535231360544e-06, + "loss": 2.4713, + "step": 11760 + }, + { + "epoch": 1.2437605752961083, + "grad_norm": 8.4375, + "learning_rate": 4.398423806307731e-06, + "loss": 2.5366, + "step": 11761 + }, + { + "epoch": 1.2438663282571911, + "grad_norm": 8.6875, + "learning_rate": 4.398312372346393e-06, + "loss": 2.5892, + "step": 11762 + }, + { + "epoch": 1.2439720812182742, + "grad_norm": 8.875, + "learning_rate": 4.39820092947705e-06, + "loss": 2.5616, + "step": 11763 + }, + { + "epoch": 1.244077834179357, + "grad_norm": 8.1875, + "learning_rate": 4.398089477700228e-06, + "loss": 2.6221, + "step": 11764 + }, + { + "epoch": 1.2441835871404399, + "grad_norm": 9.5, + "learning_rate": 4.397978017016447e-06, + "loss": 2.4713, + "step": 11765 + }, + { + "epoch": 1.244289340101523, + "grad_norm": 8.25, + "learning_rate": 4.397866547426232e-06, + "loss": 2.4086, + "step": 11766 + }, + { + "epoch": 1.2443950930626058, + "grad_norm": 8.0625, + "learning_rate": 4.397755068930105e-06, + "loss": 2.1706, + "step": 11767 + }, + { + "epoch": 1.2445008460236886, + "grad_norm": 8.375, + "learning_rate": 4.39764358152859e-06, + "loss": 2.6433, + "step": 11768 + }, + { + "epoch": 1.2446065989847717, + "grad_norm": 8.125, + "learning_rate": 4.397532085222211e-06, + "loss": 2.5451, + "step": 11769 + }, + { + "epoch": 1.2447123519458545, + "grad_norm": 8.625, + "learning_rate": 4.397420580011488e-06, + "loss": 2.5135, + "step": 11770 + }, + { + "epoch": 1.2448181049069373, + "grad_norm": 9.0625, + "learning_rate": 4.397309065896947e-06, + "loss": 2.62, + "step": 11771 + }, + { + "epoch": 1.2449238578680204, + "grad_norm": 8.1875, + "learning_rate": 4.397197542879111e-06, + "loss": 2.6428, + "step": 11772 + }, + { + "epoch": 1.2450296108291032, + "grad_norm": 7.9375, + "learning_rate": 4.397086010958502e-06, + "loss": 2.4214, + "step": 11773 + }, + { + "epoch": 1.245135363790186, + "grad_norm": 8.875, + "learning_rate": 4.396974470135645e-06, + "loss": 2.637, + "step": 11774 + }, + { + "epoch": 1.2452411167512691, + "grad_norm": 8.625, + "learning_rate": 4.396862920411062e-06, + "loss": 2.7233, + "step": 11775 + }, + { + "epoch": 1.245346869712352, + "grad_norm": 9.25, + "learning_rate": 4.3967513617852775e-06, + "loss": 2.823, + "step": 11776 + }, + { + "epoch": 1.2454526226734348, + "grad_norm": 8.0625, + "learning_rate": 4.396639794258815e-06, + "loss": 2.3863, + "step": 11777 + }, + { + "epoch": 1.2455583756345177, + "grad_norm": 8.375, + "learning_rate": 4.396528217832197e-06, + "loss": 2.6088, + "step": 11778 + }, + { + "epoch": 1.2456641285956007, + "grad_norm": 8.375, + "learning_rate": 4.396416632505947e-06, + "loss": 2.3093, + "step": 11779 + }, + { + "epoch": 1.2457698815566836, + "grad_norm": 8.125, + "learning_rate": 4.39630503828059e-06, + "loss": 2.6987, + "step": 11780 + }, + { + "epoch": 1.2458756345177666, + "grad_norm": 8.25, + "learning_rate": 4.396193435156651e-06, + "loss": 2.4583, + "step": 11781 + }, + { + "epoch": 1.2459813874788495, + "grad_norm": 7.9375, + "learning_rate": 4.396081823134649e-06, + "loss": 2.5279, + "step": 11782 + }, + { + "epoch": 1.2460871404399323, + "grad_norm": 8.375, + "learning_rate": 4.395970202215112e-06, + "loss": 2.6126, + "step": 11783 + }, + { + "epoch": 1.2461928934010151, + "grad_norm": 8.0, + "learning_rate": 4.395858572398562e-06, + "loss": 2.4429, + "step": 11784 + }, + { + "epoch": 1.2462986463620982, + "grad_norm": 8.1875, + "learning_rate": 4.3957469336855216e-06, + "loss": 2.5985, + "step": 11785 + }, + { + "epoch": 1.246404399323181, + "grad_norm": 8.75, + "learning_rate": 4.395635286076519e-06, + "loss": 2.5635, + "step": 11786 + }, + { + "epoch": 1.2465101522842639, + "grad_norm": 8.25, + "learning_rate": 4.395523629572073e-06, + "loss": 2.5153, + "step": 11787 + }, + { + "epoch": 1.246615905245347, + "grad_norm": 8.5625, + "learning_rate": 4.39541196417271e-06, + "loss": 2.4963, + "step": 11788 + }, + { + "epoch": 1.2467216582064298, + "grad_norm": 8.875, + "learning_rate": 4.395300289878954e-06, + "loss": 2.5882, + "step": 11789 + }, + { + "epoch": 1.2468274111675126, + "grad_norm": 8.5, + "learning_rate": 4.39518860669133e-06, + "loss": 2.6203, + "step": 11790 + }, + { + "epoch": 1.2469331641285957, + "grad_norm": 8.0625, + "learning_rate": 4.39507691461036e-06, + "loss": 2.5065, + "step": 11791 + }, + { + "epoch": 1.2470389170896785, + "grad_norm": 8.25, + "learning_rate": 4.394965213636569e-06, + "loss": 2.4837, + "step": 11792 + }, + { + "epoch": 1.2471446700507614, + "grad_norm": 7.84375, + "learning_rate": 4.394853503770482e-06, + "loss": 2.6374, + "step": 11793 + }, + { + "epoch": 1.2472504230118444, + "grad_norm": 8.375, + "learning_rate": 4.394741785012622e-06, + "loss": 2.5753, + "step": 11794 + }, + { + "epoch": 1.2473561759729273, + "grad_norm": 8.25, + "learning_rate": 4.394630057363514e-06, + "loss": 2.5002, + "step": 11795 + }, + { + "epoch": 1.24746192893401, + "grad_norm": 8.4375, + "learning_rate": 4.3945183208236815e-06, + "loss": 2.4994, + "step": 11796 + }, + { + "epoch": 1.2475676818950932, + "grad_norm": 8.1875, + "learning_rate": 4.3944065753936506e-06, + "loss": 2.5702, + "step": 11797 + }, + { + "epoch": 1.247673434856176, + "grad_norm": 8.25, + "learning_rate": 4.394294821073944e-06, + "loss": 2.5947, + "step": 11798 + }, + { + "epoch": 1.2477791878172588, + "grad_norm": 8.375, + "learning_rate": 4.394183057865085e-06, + "loss": 2.5825, + "step": 11799 + }, + { + "epoch": 1.247884940778342, + "grad_norm": 7.9375, + "learning_rate": 4.394071285767601e-06, + "loss": 2.4515, + "step": 11800 + }, + { + "epoch": 1.2479906937394247, + "grad_norm": 10.9375, + "learning_rate": 4.393959504782015e-06, + "loss": 2.5495, + "step": 11801 + }, + { + "epoch": 1.2480964467005076, + "grad_norm": 8.25, + "learning_rate": 4.393847714908851e-06, + "loss": 2.5167, + "step": 11802 + }, + { + "epoch": 1.2482021996615904, + "grad_norm": 8.0, + "learning_rate": 4.393735916148635e-06, + "loss": 2.5316, + "step": 11803 + }, + { + "epoch": 1.2483079526226735, + "grad_norm": 8.0625, + "learning_rate": 4.39362410850189e-06, + "loss": 2.5174, + "step": 11804 + }, + { + "epoch": 1.2484137055837563, + "grad_norm": 9.0, + "learning_rate": 4.393512291969142e-06, + "loss": 2.6344, + "step": 11805 + }, + { + "epoch": 1.2485194585448394, + "grad_norm": 7.96875, + "learning_rate": 4.393400466550915e-06, + "loss": 2.6634, + "step": 11806 + }, + { + "epoch": 1.2486252115059222, + "grad_norm": 8.1875, + "learning_rate": 4.393288632247735e-06, + "loss": 2.5896, + "step": 11807 + }, + { + "epoch": 1.248730964467005, + "grad_norm": 7.96875, + "learning_rate": 4.393176789060125e-06, + "loss": 2.6299, + "step": 11808 + }, + { + "epoch": 1.2488367174280879, + "grad_norm": 8.6875, + "learning_rate": 4.3930649369886105e-06, + "loss": 2.6254, + "step": 11809 + }, + { + "epoch": 1.248942470389171, + "grad_norm": 8.0, + "learning_rate": 4.392953076033717e-06, + "loss": 2.5899, + "step": 11810 + }, + { + "epoch": 1.2490482233502538, + "grad_norm": 8.125, + "learning_rate": 4.392841206195968e-06, + "loss": 2.6042, + "step": 11811 + }, + { + "epoch": 1.2491539763113366, + "grad_norm": 8.625, + "learning_rate": 4.392729327475891e-06, + "loss": 2.5093, + "step": 11812 + }, + { + "epoch": 1.2492597292724197, + "grad_norm": 7.9375, + "learning_rate": 4.392617439874007e-06, + "loss": 2.2672, + "step": 11813 + }, + { + "epoch": 1.2493654822335025, + "grad_norm": 8.0, + "learning_rate": 4.392505543390845e-06, + "loss": 2.5325, + "step": 11814 + }, + { + "epoch": 1.2494712351945854, + "grad_norm": 7.875, + "learning_rate": 4.3923936380269285e-06, + "loss": 2.4227, + "step": 11815 + }, + { + "epoch": 1.2495769881556684, + "grad_norm": 8.4375, + "learning_rate": 4.392281723782782e-06, + "loss": 2.4983, + "step": 11816 + }, + { + "epoch": 1.2496827411167513, + "grad_norm": 7.75, + "learning_rate": 4.392169800658932e-06, + "loss": 2.4969, + "step": 11817 + }, + { + "epoch": 1.249788494077834, + "grad_norm": 8.375, + "learning_rate": 4.392057868655902e-06, + "loss": 2.7931, + "step": 11818 + }, + { + "epoch": 1.2498942470389172, + "grad_norm": 8.1875, + "learning_rate": 4.391945927774219e-06, + "loss": 2.7403, + "step": 11819 + }, + { + "epoch": 1.25, + "grad_norm": 7.84375, + "learning_rate": 4.3918339780144074e-06, + "loss": 2.6777, + "step": 11820 + }, + { + "epoch": 1.2501057529610828, + "grad_norm": 8.0, + "learning_rate": 4.3917220193769925e-06, + "loss": 2.5991, + "step": 11821 + }, + { + "epoch": 1.250211505922166, + "grad_norm": 8.25, + "learning_rate": 4.3916100518625e-06, + "loss": 2.572, + "step": 11822 + }, + { + "epoch": 1.2503172588832487, + "grad_norm": 8.625, + "learning_rate": 4.391498075471455e-06, + "loss": 2.5225, + "step": 11823 + }, + { + "epoch": 1.2504230118443316, + "grad_norm": 8.3125, + "learning_rate": 4.391386090204383e-06, + "loss": 2.3599, + "step": 11824 + }, + { + "epoch": 1.2505287648054146, + "grad_norm": 7.65625, + "learning_rate": 4.39127409606181e-06, + "loss": 2.5969, + "step": 11825 + }, + { + "epoch": 1.2506345177664975, + "grad_norm": 8.1875, + "learning_rate": 4.391162093044262e-06, + "loss": 2.504, + "step": 11826 + }, + { + "epoch": 1.2507402707275803, + "grad_norm": 8.4375, + "learning_rate": 4.3910500811522625e-06, + "loss": 2.6626, + "step": 11827 + }, + { + "epoch": 1.2508460236886632, + "grad_norm": 7.96875, + "learning_rate": 4.390938060386339e-06, + "loss": 2.4971, + "step": 11828 + }, + { + "epoch": 1.2509517766497462, + "grad_norm": 7.875, + "learning_rate": 4.390826030747016e-06, + "loss": 2.4419, + "step": 11829 + }, + { + "epoch": 1.251057529610829, + "grad_norm": 8.125, + "learning_rate": 4.39071399223482e-06, + "loss": 2.4794, + "step": 11830 + }, + { + "epoch": 1.2511632825719121, + "grad_norm": 8.1875, + "learning_rate": 4.3906019448502765e-06, + "loss": 2.5482, + "step": 11831 + }, + { + "epoch": 1.251269035532995, + "grad_norm": 7.90625, + "learning_rate": 4.390489888593911e-06, + "loss": 2.4345, + "step": 11832 + }, + { + "epoch": 1.2513747884940778, + "grad_norm": 8.4375, + "learning_rate": 4.39037782346625e-06, + "loss": 2.5372, + "step": 11833 + }, + { + "epoch": 1.2514805414551606, + "grad_norm": 8.125, + "learning_rate": 4.390265749467819e-06, + "loss": 2.7108, + "step": 11834 + }, + { + "epoch": 1.2515862944162437, + "grad_norm": 8.625, + "learning_rate": 4.390153666599144e-06, + "loss": 2.7259, + "step": 11835 + }, + { + "epoch": 1.2516920473773265, + "grad_norm": 8.25, + "learning_rate": 4.390041574860751e-06, + "loss": 2.6591, + "step": 11836 + }, + { + "epoch": 1.2517978003384096, + "grad_norm": 8.75, + "learning_rate": 4.389929474253166e-06, + "loss": 2.5669, + "step": 11837 + }, + { + "epoch": 1.2519035532994924, + "grad_norm": 10.0, + "learning_rate": 4.3898173647769145e-06, + "loss": 2.5164, + "step": 11838 + }, + { + "epoch": 1.2520093062605753, + "grad_norm": 8.375, + "learning_rate": 4.389705246432523e-06, + "loss": 2.7142, + "step": 11839 + }, + { + "epoch": 1.252115059221658, + "grad_norm": 8.1875, + "learning_rate": 4.389593119220518e-06, + "loss": 2.5585, + "step": 11840 + }, + { + "epoch": 1.2522208121827412, + "grad_norm": 8.4375, + "learning_rate": 4.389480983141425e-06, + "loss": 2.6203, + "step": 11841 + }, + { + "epoch": 1.252326565143824, + "grad_norm": 8.5625, + "learning_rate": 4.389368838195771e-06, + "loss": 2.7315, + "step": 11842 + }, + { + "epoch": 1.252432318104907, + "grad_norm": 7.875, + "learning_rate": 4.389256684384081e-06, + "loss": 2.5598, + "step": 11843 + }, + { + "epoch": 1.25253807106599, + "grad_norm": 8.6875, + "learning_rate": 4.389144521706883e-06, + "loss": 2.5991, + "step": 11844 + }, + { + "epoch": 1.2526438240270727, + "grad_norm": 8.9375, + "learning_rate": 4.389032350164702e-06, + "loss": 2.6608, + "step": 11845 + }, + { + "epoch": 1.2527495769881556, + "grad_norm": 8.625, + "learning_rate": 4.388920169758064e-06, + "loss": 2.6028, + "step": 11846 + }, + { + "epoch": 1.2528553299492386, + "grad_norm": 7.5625, + "learning_rate": 4.388807980487497e-06, + "loss": 2.4768, + "step": 11847 + }, + { + "epoch": 1.2529610829103215, + "grad_norm": 8.1875, + "learning_rate": 4.388695782353526e-06, + "loss": 2.5729, + "step": 11848 + }, + { + "epoch": 1.2530668358714043, + "grad_norm": 8.875, + "learning_rate": 4.388583575356679e-06, + "loss": 2.8467, + "step": 11849 + }, + { + "epoch": 1.2531725888324874, + "grad_norm": 9.0, + "learning_rate": 4.3884713594974815e-06, + "loss": 2.4858, + "step": 11850 + }, + { + "epoch": 1.2532783417935702, + "grad_norm": 8.5, + "learning_rate": 4.388359134776459e-06, + "loss": 2.6538, + "step": 11851 + }, + { + "epoch": 1.253384094754653, + "grad_norm": 8.0625, + "learning_rate": 4.388246901194141e-06, + "loss": 2.5237, + "step": 11852 + }, + { + "epoch": 1.253489847715736, + "grad_norm": 8.1875, + "learning_rate": 4.388134658751052e-06, + "loss": 2.6668, + "step": 11853 + }, + { + "epoch": 1.253595600676819, + "grad_norm": 8.375, + "learning_rate": 4.388022407447719e-06, + "loss": 2.6812, + "step": 11854 + }, + { + "epoch": 1.2537013536379018, + "grad_norm": 7.9375, + "learning_rate": 4.38791014728467e-06, + "loss": 2.5815, + "step": 11855 + }, + { + "epoch": 1.2538071065989849, + "grad_norm": 8.5625, + "learning_rate": 4.38779787826243e-06, + "loss": 2.6089, + "step": 11856 + }, + { + "epoch": 1.2539128595600677, + "grad_norm": 8.3125, + "learning_rate": 4.387685600381526e-06, + "loss": 2.4344, + "step": 11857 + }, + { + "epoch": 1.2540186125211505, + "grad_norm": 8.3125, + "learning_rate": 4.3875733136424866e-06, + "loss": 2.5069, + "step": 11858 + }, + { + "epoch": 1.2541243654822334, + "grad_norm": 8.375, + "learning_rate": 4.387461018045838e-06, + "loss": 2.4296, + "step": 11859 + }, + { + "epoch": 1.2542301184433164, + "grad_norm": 8.0625, + "learning_rate": 4.387348713592105e-06, + "loss": 2.6136, + "step": 11860 + }, + { + "epoch": 1.2543358714043993, + "grad_norm": 8.1875, + "learning_rate": 4.387236400281818e-06, + "loss": 2.5631, + "step": 11861 + }, + { + "epoch": 1.2544416243654823, + "grad_norm": 7.8125, + "learning_rate": 4.387124078115502e-06, + "loss": 2.6554, + "step": 11862 + }, + { + "epoch": 1.2545473773265652, + "grad_norm": 8.1875, + "learning_rate": 4.387011747093684e-06, + "loss": 2.5594, + "step": 11863 + }, + { + "epoch": 1.254653130287648, + "grad_norm": 8.1875, + "learning_rate": 4.3868994072168925e-06, + "loss": 2.3316, + "step": 11864 + }, + { + "epoch": 1.2547588832487309, + "grad_norm": 8.0, + "learning_rate": 4.3867870584856535e-06, + "loss": 2.7, + "step": 11865 + }, + { + "epoch": 1.254864636209814, + "grad_norm": 8.3125, + "learning_rate": 4.386674700900494e-06, + "loss": 2.5636, + "step": 11866 + }, + { + "epoch": 1.2549703891708968, + "grad_norm": 8.5, + "learning_rate": 4.386562334461942e-06, + "loss": 2.6403, + "step": 11867 + }, + { + "epoch": 1.2550761421319798, + "grad_norm": 8.625, + "learning_rate": 4.386449959170524e-06, + "loss": 2.403, + "step": 11868 + }, + { + "epoch": 1.2551818950930627, + "grad_norm": 9.3125, + "learning_rate": 4.386337575026769e-06, + "loss": 2.4987, + "step": 11869 + }, + { + "epoch": 1.2552876480541455, + "grad_norm": 8.0625, + "learning_rate": 4.386225182031203e-06, + "loss": 2.5815, + "step": 11870 + }, + { + "epoch": 1.2553934010152283, + "grad_norm": 8.4375, + "learning_rate": 4.386112780184353e-06, + "loss": 2.7182, + "step": 11871 + }, + { + "epoch": 1.2554991539763114, + "grad_norm": 8.1875, + "learning_rate": 4.3860003694867475e-06, + "loss": 2.7655, + "step": 11872 + }, + { + "epoch": 1.2556049069373942, + "grad_norm": 8.0, + "learning_rate": 4.385887949938913e-06, + "loss": 2.6004, + "step": 11873 + }, + { + "epoch": 1.255710659898477, + "grad_norm": 8.0, + "learning_rate": 4.385775521541379e-06, + "loss": 2.3587, + "step": 11874 + }, + { + "epoch": 1.2558164128595601, + "grad_norm": 8.125, + "learning_rate": 4.3856630842946715e-06, + "loss": 2.5011, + "step": 11875 + }, + { + "epoch": 1.255922165820643, + "grad_norm": 8.625, + "learning_rate": 4.385550638199318e-06, + "loss": 2.5748, + "step": 11876 + }, + { + "epoch": 1.2560279187817258, + "grad_norm": 8.1875, + "learning_rate": 4.385438183255846e-06, + "loss": 2.6075, + "step": 11877 + }, + { + "epoch": 1.2561336717428089, + "grad_norm": 8.75, + "learning_rate": 4.385325719464784e-06, + "loss": 2.771, + "step": 11878 + }, + { + "epoch": 1.2562394247038917, + "grad_norm": 8.25, + "learning_rate": 4.385213246826659e-06, + "loss": 2.5922, + "step": 11879 + }, + { + "epoch": 1.2563451776649746, + "grad_norm": 8.375, + "learning_rate": 4.3851007653420004e-06, + "loss": 2.5529, + "step": 11880 + }, + { + "epoch": 1.2564509306260576, + "grad_norm": 9.0625, + "learning_rate": 4.384988275011334e-06, + "loss": 2.5939, + "step": 11881 + }, + { + "epoch": 1.2565566835871405, + "grad_norm": 8.375, + "learning_rate": 4.384875775835189e-06, + "loss": 2.8396, + "step": 11882 + }, + { + "epoch": 1.2566624365482233, + "grad_norm": 8.125, + "learning_rate": 4.3847632678140925e-06, + "loss": 2.791, + "step": 11883 + }, + { + "epoch": 1.2567681895093061, + "grad_norm": 7.71875, + "learning_rate": 4.3846507509485736e-06, + "loss": 2.445, + "step": 11884 + }, + { + "epoch": 1.2568739424703892, + "grad_norm": 8.1875, + "learning_rate": 4.384538225239159e-06, + "loss": 2.3914, + "step": 11885 + }, + { + "epoch": 1.256979695431472, + "grad_norm": 8.375, + "learning_rate": 4.384425690686377e-06, + "loss": 2.4425, + "step": 11886 + }, + { + "epoch": 1.257085448392555, + "grad_norm": 8.5, + "learning_rate": 4.384313147290756e-06, + "loss": 2.5653, + "step": 11887 + }, + { + "epoch": 1.257191201353638, + "grad_norm": 8.6875, + "learning_rate": 4.384200595052824e-06, + "loss": 2.5397, + "step": 11888 + }, + { + "epoch": 1.2572969543147208, + "grad_norm": 7.90625, + "learning_rate": 4.38408803397311e-06, + "loss": 2.4522, + "step": 11889 + }, + { + "epoch": 1.2574027072758036, + "grad_norm": 8.0625, + "learning_rate": 4.383975464052141e-06, + "loss": 2.4638, + "step": 11890 + }, + { + "epoch": 1.2575084602368867, + "grad_norm": 7.96875, + "learning_rate": 4.3838628852904455e-06, + "loss": 2.6362, + "step": 11891 + }, + { + "epoch": 1.2576142131979695, + "grad_norm": 8.5625, + "learning_rate": 4.383750297688551e-06, + "loss": 2.773, + "step": 11892 + }, + { + "epoch": 1.2577199661590526, + "grad_norm": 8.375, + "learning_rate": 4.383637701246989e-06, + "loss": 2.6269, + "step": 11893 + }, + { + "epoch": 1.2578257191201354, + "grad_norm": 8.0, + "learning_rate": 4.383525095966284e-06, + "loss": 2.4308, + "step": 11894 + }, + { + "epoch": 1.2579314720812182, + "grad_norm": 8.6875, + "learning_rate": 4.383412481846966e-06, + "loss": 2.6521, + "step": 11895 + }, + { + "epoch": 1.258037225042301, + "grad_norm": 7.75, + "learning_rate": 4.383299858889565e-06, + "loss": 2.5387, + "step": 11896 + }, + { + "epoch": 1.2581429780033841, + "grad_norm": 8.3125, + "learning_rate": 4.383187227094606e-06, + "loss": 2.4812, + "step": 11897 + }, + { + "epoch": 1.258248730964467, + "grad_norm": 7.96875, + "learning_rate": 4.383074586462621e-06, + "loss": 2.5583, + "step": 11898 + }, + { + "epoch": 1.25835448392555, + "grad_norm": 7.8125, + "learning_rate": 4.3829619369941365e-06, + "loss": 2.6149, + "step": 11899 + }, + { + "epoch": 1.2584602368866329, + "grad_norm": 7.96875, + "learning_rate": 4.382849278689682e-06, + "loss": 2.477, + "step": 11900 + }, + { + "epoch": 1.2585659898477157, + "grad_norm": 8.4375, + "learning_rate": 4.382736611549786e-06, + "loss": 2.5113, + "step": 11901 + }, + { + "epoch": 1.2586717428087986, + "grad_norm": 8.1875, + "learning_rate": 4.3826239355749764e-06, + "loss": 2.4501, + "step": 11902 + }, + { + "epoch": 1.2587774957698816, + "grad_norm": 8.0, + "learning_rate": 4.382511250765784e-06, + "loss": 2.462, + "step": 11903 + }, + { + "epoch": 1.2588832487309645, + "grad_norm": 8.4375, + "learning_rate": 4.382398557122735e-06, + "loss": 2.7285, + "step": 11904 + }, + { + "epoch": 1.2589890016920473, + "grad_norm": 8.0, + "learning_rate": 4.38228585464636e-06, + "loss": 2.3936, + "step": 11905 + }, + { + "epoch": 1.2590947546531304, + "grad_norm": 8.625, + "learning_rate": 4.3821731433371875e-06, + "loss": 2.4366, + "step": 11906 + }, + { + "epoch": 1.2592005076142132, + "grad_norm": 8.5, + "learning_rate": 4.382060423195745e-06, + "loss": 2.6511, + "step": 11907 + }, + { + "epoch": 1.259306260575296, + "grad_norm": 7.5625, + "learning_rate": 4.381947694222564e-06, + "loss": 2.2757, + "step": 11908 + }, + { + "epoch": 1.2594120135363789, + "grad_norm": 8.5625, + "learning_rate": 4.381834956418172e-06, + "loss": 2.5083, + "step": 11909 + }, + { + "epoch": 1.259517766497462, + "grad_norm": 8.5625, + "learning_rate": 4.3817222097830975e-06, + "loss": 2.7871, + "step": 11910 + }, + { + "epoch": 1.2596235194585448, + "grad_norm": 8.4375, + "learning_rate": 4.381609454317871e-06, + "loss": 2.4736, + "step": 11911 + }, + { + "epoch": 1.2597292724196278, + "grad_norm": 8.4375, + "learning_rate": 4.38149669002302e-06, + "loss": 2.4695, + "step": 11912 + }, + { + "epoch": 1.2598350253807107, + "grad_norm": 8.625, + "learning_rate": 4.3813839168990744e-06, + "loss": 2.787, + "step": 11913 + }, + { + "epoch": 1.2599407783417935, + "grad_norm": 8.0625, + "learning_rate": 4.381271134946565e-06, + "loss": 2.4908, + "step": 11914 + }, + { + "epoch": 1.2600465313028764, + "grad_norm": 8.75, + "learning_rate": 4.381158344166019e-06, + "loss": 2.5833, + "step": 11915 + }, + { + "epoch": 1.2601522842639594, + "grad_norm": 8.5625, + "learning_rate": 4.381045544557966e-06, + "loss": 2.5741, + "step": 11916 + }, + { + "epoch": 1.2602580372250423, + "grad_norm": 20.875, + "learning_rate": 4.380932736122935e-06, + "loss": 2.6654, + "step": 11917 + }, + { + "epoch": 1.2603637901861253, + "grad_norm": 8.625, + "learning_rate": 4.380819918861456e-06, + "loss": 2.7313, + "step": 11918 + }, + { + "epoch": 1.2604695431472082, + "grad_norm": 7.9375, + "learning_rate": 4.3807070927740595e-06, + "loss": 2.3576, + "step": 11919 + }, + { + "epoch": 1.260575296108291, + "grad_norm": 8.9375, + "learning_rate": 4.380594257861274e-06, + "loss": 2.9113, + "step": 11920 + }, + { + "epoch": 1.2606810490693738, + "grad_norm": 8.25, + "learning_rate": 4.380481414123627e-06, + "loss": 2.5176, + "step": 11921 + }, + { + "epoch": 1.260786802030457, + "grad_norm": 8.4375, + "learning_rate": 4.380368561561651e-06, + "loss": 2.5171, + "step": 11922 + }, + { + "epoch": 1.2608925549915397, + "grad_norm": 8.5625, + "learning_rate": 4.3802557001758746e-06, + "loss": 2.8735, + "step": 11923 + }, + { + "epoch": 1.2609983079526228, + "grad_norm": 7.75, + "learning_rate": 4.380142829966827e-06, + "loss": 2.5496, + "step": 11924 + }, + { + "epoch": 1.2611040609137056, + "grad_norm": 8.0625, + "learning_rate": 4.380029950935037e-06, + "loss": 2.6502, + "step": 11925 + }, + { + "epoch": 1.2612098138747885, + "grad_norm": 9.1875, + "learning_rate": 4.3799170630810364e-06, + "loss": 2.6789, + "step": 11926 + }, + { + "epoch": 1.2613155668358713, + "grad_norm": 7.78125, + "learning_rate": 4.379804166405354e-06, + "loss": 2.6317, + "step": 11927 + }, + { + "epoch": 1.2614213197969544, + "grad_norm": 7.84375, + "learning_rate": 4.379691260908519e-06, + "loss": 2.4559, + "step": 11928 + }, + { + "epoch": 1.2615270727580372, + "grad_norm": 8.8125, + "learning_rate": 4.379578346591062e-06, + "loss": 2.5301, + "step": 11929 + }, + { + "epoch": 1.26163282571912, + "grad_norm": 7.6875, + "learning_rate": 4.379465423453511e-06, + "loss": 2.5794, + "step": 11930 + }, + { + "epoch": 1.2617385786802031, + "grad_norm": 8.0, + "learning_rate": 4.3793524914964e-06, + "loss": 2.5999, + "step": 11931 + }, + { + "epoch": 1.261844331641286, + "grad_norm": 8.5, + "learning_rate": 4.379239550720255e-06, + "loss": 2.546, + "step": 11932 + }, + { + "epoch": 1.2619500846023688, + "grad_norm": 7.96875, + "learning_rate": 4.379126601125607e-06, + "loss": 2.5918, + "step": 11933 + }, + { + "epoch": 1.2620558375634519, + "grad_norm": 8.0625, + "learning_rate": 4.3790136427129875e-06, + "loss": 2.8196, + "step": 11934 + }, + { + "epoch": 1.2621615905245347, + "grad_norm": 8.1875, + "learning_rate": 4.378900675482925e-06, + "loss": 2.4991, + "step": 11935 + }, + { + "epoch": 1.2622673434856175, + "grad_norm": 7.96875, + "learning_rate": 4.37878769943595e-06, + "loss": 2.4057, + "step": 11936 + }, + { + "epoch": 1.2623730964467006, + "grad_norm": 8.8125, + "learning_rate": 4.378674714572593e-06, + "loss": 2.3849, + "step": 11937 + }, + { + "epoch": 1.2624788494077834, + "grad_norm": 8.0625, + "learning_rate": 4.3785617208933825e-06, + "loss": 2.508, + "step": 11938 + }, + { + "epoch": 1.2625846023688663, + "grad_norm": 8.0, + "learning_rate": 4.378448718398852e-06, + "loss": 2.5928, + "step": 11939 + }, + { + "epoch": 1.262690355329949, + "grad_norm": 8.6875, + "learning_rate": 4.378335707089529e-06, + "loss": 2.3754, + "step": 11940 + }, + { + "epoch": 1.2627961082910322, + "grad_norm": 8.875, + "learning_rate": 4.378222686965945e-06, + "loss": 2.792, + "step": 11941 + }, + { + "epoch": 1.262901861252115, + "grad_norm": 8.4375, + "learning_rate": 4.37810965802863e-06, + "loss": 2.4814, + "step": 11942 + }, + { + "epoch": 1.263007614213198, + "grad_norm": 7.96875, + "learning_rate": 4.377996620278114e-06, + "loss": 2.5173, + "step": 11943 + }, + { + "epoch": 1.263113367174281, + "grad_norm": 8.0625, + "learning_rate": 4.377883573714929e-06, + "loss": 2.3211, + "step": 11944 + }, + { + "epoch": 1.2632191201353637, + "grad_norm": 7.90625, + "learning_rate": 4.377770518339604e-06, + "loss": 2.5833, + "step": 11945 + }, + { + "epoch": 1.2633248730964466, + "grad_norm": 8.875, + "learning_rate": 4.37765745415267e-06, + "loss": 2.6908, + "step": 11946 + }, + { + "epoch": 1.2634306260575296, + "grad_norm": 8.3125, + "learning_rate": 4.377544381154657e-06, + "loss": 2.527, + "step": 11947 + }, + { + "epoch": 1.2635363790186125, + "grad_norm": 8.75, + "learning_rate": 4.377431299346097e-06, + "loss": 2.4922, + "step": 11948 + }, + { + "epoch": 1.2636421319796955, + "grad_norm": 8.25, + "learning_rate": 4.377318208727519e-06, + "loss": 2.5059, + "step": 11949 + }, + { + "epoch": 1.2637478849407784, + "grad_norm": 8.5, + "learning_rate": 4.377205109299454e-06, + "loss": 2.5951, + "step": 11950 + }, + { + "epoch": 1.2638536379018612, + "grad_norm": 8.0, + "learning_rate": 4.377092001062434e-06, + "loss": 2.5307, + "step": 11951 + }, + { + "epoch": 1.263959390862944, + "grad_norm": 8.4375, + "learning_rate": 4.376978884016989e-06, + "loss": 2.1987, + "step": 11952 + }, + { + "epoch": 1.2640651438240271, + "grad_norm": 8.125, + "learning_rate": 4.376865758163649e-06, + "loss": 2.4533, + "step": 11953 + }, + { + "epoch": 1.26417089678511, + "grad_norm": 8.125, + "learning_rate": 4.376752623502947e-06, + "loss": 2.4769, + "step": 11954 + }, + { + "epoch": 1.264276649746193, + "grad_norm": 8.3125, + "learning_rate": 4.376639480035411e-06, + "loss": 2.497, + "step": 11955 + }, + { + "epoch": 1.2643824027072759, + "grad_norm": 8.0625, + "learning_rate": 4.376526327761575e-06, + "loss": 2.5351, + "step": 11956 + }, + { + "epoch": 1.2644881556683587, + "grad_norm": 8.4375, + "learning_rate": 4.376413166681966e-06, + "loss": 2.5573, + "step": 11957 + }, + { + "epoch": 1.2645939086294415, + "grad_norm": 8.5, + "learning_rate": 4.37629999679712e-06, + "loss": 2.5069, + "step": 11958 + }, + { + "epoch": 1.2646996615905246, + "grad_norm": 8.1875, + "learning_rate": 4.376186818107564e-06, + "loss": 2.3566, + "step": 11959 + }, + { + "epoch": 1.2648054145516074, + "grad_norm": 8.75, + "learning_rate": 4.376073630613831e-06, + "loss": 2.5518, + "step": 11960 + }, + { + "epoch": 1.2649111675126903, + "grad_norm": 8.1875, + "learning_rate": 4.375960434316452e-06, + "loss": 2.6367, + "step": 11961 + }, + { + "epoch": 1.2650169204737733, + "grad_norm": 8.25, + "learning_rate": 4.375847229215957e-06, + "loss": 2.4146, + "step": 11962 + }, + { + "epoch": 1.2651226734348562, + "grad_norm": 7.71875, + "learning_rate": 4.375734015312879e-06, + "loss": 2.4935, + "step": 11963 + }, + { + "epoch": 1.265228426395939, + "grad_norm": 8.5625, + "learning_rate": 4.375620792607749e-06, + "loss": 2.7758, + "step": 11964 + }, + { + "epoch": 1.2653341793570219, + "grad_norm": 10.375, + "learning_rate": 4.3755075611010964e-06, + "loss": 2.6745, + "step": 11965 + }, + { + "epoch": 1.265439932318105, + "grad_norm": 8.3125, + "learning_rate": 4.375394320793455e-06, + "loss": 2.7525, + "step": 11966 + }, + { + "epoch": 1.2655456852791878, + "grad_norm": 8.0625, + "learning_rate": 4.375281071685354e-06, + "loss": 2.6163, + "step": 11967 + }, + { + "epoch": 1.2656514382402708, + "grad_norm": 7.6875, + "learning_rate": 4.3751678137773265e-06, + "loss": 2.428, + "step": 11968 + }, + { + "epoch": 1.2657571912013537, + "grad_norm": 8.3125, + "learning_rate": 4.375054547069903e-06, + "loss": 2.4077, + "step": 11969 + }, + { + "epoch": 1.2658629441624365, + "grad_norm": 8.3125, + "learning_rate": 4.3749412715636154e-06, + "loss": 2.7343, + "step": 11970 + }, + { + "epoch": 1.2659686971235193, + "grad_norm": 8.1875, + "learning_rate": 4.3748279872589964e-06, + "loss": 2.5708, + "step": 11971 + }, + { + "epoch": 1.2660744500846024, + "grad_norm": 8.75, + "learning_rate": 4.374714694156574e-06, + "loss": 2.7013, + "step": 11972 + }, + { + "epoch": 1.2661802030456852, + "grad_norm": 8.0625, + "learning_rate": 4.374601392256884e-06, + "loss": 2.5886, + "step": 11973 + }, + { + "epoch": 1.2662859560067683, + "grad_norm": 8.0625, + "learning_rate": 4.374488081560456e-06, + "loss": 2.4634, + "step": 11974 + }, + { + "epoch": 1.2663917089678511, + "grad_norm": 8.1875, + "learning_rate": 4.3743747620678225e-06, + "loss": 2.7032, + "step": 11975 + }, + { + "epoch": 1.266497461928934, + "grad_norm": 8.875, + "learning_rate": 4.374261433779515e-06, + "loss": 2.4381, + "step": 11976 + }, + { + "epoch": 1.2666032148900168, + "grad_norm": 8.5625, + "learning_rate": 4.374148096696065e-06, + "loss": 2.5843, + "step": 11977 + }, + { + "epoch": 1.2667089678510999, + "grad_norm": 7.8125, + "learning_rate": 4.3740347508180035e-06, + "loss": 2.3881, + "step": 11978 + }, + { + "epoch": 1.2668147208121827, + "grad_norm": 8.5625, + "learning_rate": 4.373921396145864e-06, + "loss": 2.2651, + "step": 11979 + }, + { + "epoch": 1.2669204737732658, + "grad_norm": 8.1875, + "learning_rate": 4.373808032680178e-06, + "loss": 2.703, + "step": 11980 + }, + { + "epoch": 1.2670262267343486, + "grad_norm": 8.3125, + "learning_rate": 4.373694660421477e-06, + "loss": 2.6648, + "step": 11981 + }, + { + "epoch": 1.2671319796954315, + "grad_norm": 8.1875, + "learning_rate": 4.3735812793702935e-06, + "loss": 2.6616, + "step": 11982 + }, + { + "epoch": 1.2672377326565143, + "grad_norm": 8.25, + "learning_rate": 4.373467889527159e-06, + "loss": 2.5122, + "step": 11983 + }, + { + "epoch": 1.2673434856175974, + "grad_norm": 8.3125, + "learning_rate": 4.373354490892606e-06, + "loss": 2.6086, + "step": 11984 + }, + { + "epoch": 1.2674492385786802, + "grad_norm": 8.125, + "learning_rate": 4.373241083467167e-06, + "loss": 2.5625, + "step": 11985 + }, + { + "epoch": 1.267554991539763, + "grad_norm": 8.125, + "learning_rate": 4.373127667251373e-06, + "loss": 2.4314, + "step": 11986 + }, + { + "epoch": 1.267660744500846, + "grad_norm": 8.0, + "learning_rate": 4.373014242245757e-06, + "loss": 2.4517, + "step": 11987 + }, + { + "epoch": 1.267766497461929, + "grad_norm": 8.5625, + "learning_rate": 4.3729008084508516e-06, + "loss": 2.7364, + "step": 11988 + }, + { + "epoch": 1.2678722504230118, + "grad_norm": 7.625, + "learning_rate": 4.372787365867189e-06, + "loss": 2.3174, + "step": 11989 + }, + { + "epoch": 1.2679780033840948, + "grad_norm": 8.0, + "learning_rate": 4.372673914495301e-06, + "loss": 2.6358, + "step": 11990 + }, + { + "epoch": 1.2680837563451777, + "grad_norm": 7.84375, + "learning_rate": 4.37256045433572e-06, + "loss": 2.5172, + "step": 11991 + }, + { + "epoch": 1.2681895093062605, + "grad_norm": 8.9375, + "learning_rate": 4.372446985388978e-06, + "loss": 2.4633, + "step": 11992 + }, + { + "epoch": 1.2682952622673436, + "grad_norm": 8.125, + "learning_rate": 4.37233350765561e-06, + "loss": 2.6389, + "step": 11993 + }, + { + "epoch": 1.2684010152284264, + "grad_norm": 7.96875, + "learning_rate": 4.372220021136144e-06, + "loss": 2.3342, + "step": 11994 + }, + { + "epoch": 1.2685067681895092, + "grad_norm": 8.375, + "learning_rate": 4.372106525831117e-06, + "loss": 2.6111, + "step": 11995 + }, + { + "epoch": 1.268612521150592, + "grad_norm": 7.875, + "learning_rate": 4.37199302174106e-06, + "loss": 2.5935, + "step": 11996 + }, + { + "epoch": 1.2687182741116751, + "grad_norm": 7.9375, + "learning_rate": 4.371879508866505e-06, + "loss": 2.6528, + "step": 11997 + }, + { + "epoch": 1.268824027072758, + "grad_norm": 8.25, + "learning_rate": 4.3717659872079845e-06, + "loss": 2.694, + "step": 11998 + }, + { + "epoch": 1.268929780033841, + "grad_norm": 8.4375, + "learning_rate": 4.3716524567660315e-06, + "loss": 2.6373, + "step": 11999 + }, + { + "epoch": 1.2690355329949239, + "grad_norm": 8.3125, + "learning_rate": 4.3715389175411795e-06, + "loss": 2.6513, + "step": 12000 + }, + { + "epoch": 1.2691412859560067, + "grad_norm": 8.625, + "learning_rate": 4.3714253695339605e-06, + "loss": 2.5757, + "step": 12001 + }, + { + "epoch": 1.2692470389170896, + "grad_norm": 8.25, + "learning_rate": 4.371311812744908e-06, + "loss": 2.6747, + "step": 12002 + }, + { + "epoch": 1.2693527918781726, + "grad_norm": 7.875, + "learning_rate": 4.371198247174554e-06, + "loss": 2.5623, + "step": 12003 + }, + { + "epoch": 1.2694585448392555, + "grad_norm": 8.3125, + "learning_rate": 4.371084672823433e-06, + "loss": 2.5765, + "step": 12004 + }, + { + "epoch": 1.2695642978003385, + "grad_norm": 8.875, + "learning_rate": 4.370971089692076e-06, + "loss": 2.8245, + "step": 12005 + }, + { + "epoch": 1.2696700507614214, + "grad_norm": 8.1875, + "learning_rate": 4.370857497781017e-06, + "loss": 2.5127, + "step": 12006 + }, + { + "epoch": 1.2697758037225042, + "grad_norm": 8.5, + "learning_rate": 4.370743897090789e-06, + "loss": 2.4349, + "step": 12007 + }, + { + "epoch": 1.269881556683587, + "grad_norm": 8.125, + "learning_rate": 4.370630287621925e-06, + "loss": 2.6045, + "step": 12008 + }, + { + "epoch": 1.26998730964467, + "grad_norm": 8.1875, + "learning_rate": 4.370516669374958e-06, + "loss": 2.5034, + "step": 12009 + }, + { + "epoch": 1.270093062605753, + "grad_norm": 7.90625, + "learning_rate": 4.3704030423504204e-06, + "loss": 2.5188, + "step": 12010 + }, + { + "epoch": 1.270198815566836, + "grad_norm": 9.5, + "learning_rate": 4.370289406548848e-06, + "loss": 2.8212, + "step": 12011 + }, + { + "epoch": 1.2703045685279188, + "grad_norm": 8.5, + "learning_rate": 4.3701757619707715e-06, + "loss": 2.4442, + "step": 12012 + }, + { + "epoch": 1.2704103214890017, + "grad_norm": 8.3125, + "learning_rate": 4.370062108616725e-06, + "loss": 2.6479, + "step": 12013 + }, + { + "epoch": 1.2705160744500845, + "grad_norm": 8.875, + "learning_rate": 4.369948446487242e-06, + "loss": 2.7401, + "step": 12014 + }, + { + "epoch": 1.2706218274111676, + "grad_norm": 7.96875, + "learning_rate": 4.369834775582857e-06, + "loss": 2.5091, + "step": 12015 + }, + { + "epoch": 1.2707275803722504, + "grad_norm": 8.25, + "learning_rate": 4.3697210959041e-06, + "loss": 2.605, + "step": 12016 + }, + { + "epoch": 1.2708333333333333, + "grad_norm": 8.4375, + "learning_rate": 4.3696074074515075e-06, + "loss": 2.5814, + "step": 12017 + }, + { + "epoch": 1.2709390862944163, + "grad_norm": 7.9375, + "learning_rate": 4.369493710225612e-06, + "loss": 2.5187, + "step": 12018 + }, + { + "epoch": 1.2710448392554992, + "grad_norm": 8.125, + "learning_rate": 4.369380004226946e-06, + "loss": 2.2382, + "step": 12019 + }, + { + "epoch": 1.271150592216582, + "grad_norm": 8.0, + "learning_rate": 4.369266289456046e-06, + "loss": 2.3502, + "step": 12020 + }, + { + "epoch": 1.2712563451776648, + "grad_norm": 8.0625, + "learning_rate": 4.369152565913443e-06, + "loss": 2.4672, + "step": 12021 + }, + { + "epoch": 1.271362098138748, + "grad_norm": 8.3125, + "learning_rate": 4.369038833599671e-06, + "loss": 2.6619, + "step": 12022 + }, + { + "epoch": 1.2714678510998307, + "grad_norm": 8.125, + "learning_rate": 4.368925092515265e-06, + "loss": 2.5808, + "step": 12023 + }, + { + "epoch": 1.2715736040609138, + "grad_norm": 8.5, + "learning_rate": 4.368811342660758e-06, + "loss": 2.455, + "step": 12024 + }, + { + "epoch": 1.2716793570219966, + "grad_norm": 8.25, + "learning_rate": 4.368697584036683e-06, + "loss": 2.5793, + "step": 12025 + }, + { + "epoch": 1.2717851099830795, + "grad_norm": 8.1875, + "learning_rate": 4.368583816643575e-06, + "loss": 2.5658, + "step": 12026 + }, + { + "epoch": 1.2718908629441623, + "grad_norm": 8.125, + "learning_rate": 4.368470040481967e-06, + "loss": 2.6086, + "step": 12027 + }, + { + "epoch": 1.2719966159052454, + "grad_norm": 8.875, + "learning_rate": 4.368356255552394e-06, + "loss": 2.2927, + "step": 12028 + }, + { + "epoch": 1.2721023688663282, + "grad_norm": 7.5625, + "learning_rate": 4.368242461855389e-06, + "loss": 2.4239, + "step": 12029 + }, + { + "epoch": 1.2722081218274113, + "grad_norm": 8.375, + "learning_rate": 4.3681286593914855e-06, + "loss": 2.6609, + "step": 12030 + }, + { + "epoch": 1.2723138747884941, + "grad_norm": 7.9375, + "learning_rate": 4.368014848161219e-06, + "loss": 2.5589, + "step": 12031 + }, + { + "epoch": 1.272419627749577, + "grad_norm": 8.75, + "learning_rate": 4.367901028165122e-06, + "loss": 2.5773, + "step": 12032 + }, + { + "epoch": 1.2725253807106598, + "grad_norm": 8.1875, + "learning_rate": 4.36778719940373e-06, + "loss": 2.2596, + "step": 12033 + }, + { + "epoch": 1.2726311336717429, + "grad_norm": 7.75, + "learning_rate": 4.367673361877577e-06, + "loss": 2.5199, + "step": 12034 + }, + { + "epoch": 1.2727368866328257, + "grad_norm": 8.3125, + "learning_rate": 4.367559515587196e-06, + "loss": 2.5194, + "step": 12035 + }, + { + "epoch": 1.2728426395939088, + "grad_norm": 8.5, + "learning_rate": 4.3674456605331225e-06, + "loss": 2.7662, + "step": 12036 + }, + { + "epoch": 1.2729483925549916, + "grad_norm": 8.25, + "learning_rate": 4.36733179671589e-06, + "loss": 2.3751, + "step": 12037 + }, + { + "epoch": 1.2730541455160744, + "grad_norm": 8.0625, + "learning_rate": 4.367217924136032e-06, + "loss": 2.6738, + "step": 12038 + }, + { + "epoch": 1.2731598984771573, + "grad_norm": 8.9375, + "learning_rate": 4.367104042794086e-06, + "loss": 2.6883, + "step": 12039 + }, + { + "epoch": 1.2732656514382403, + "grad_norm": 9.0625, + "learning_rate": 4.366990152690583e-06, + "loss": 2.4522, + "step": 12040 + }, + { + "epoch": 1.2733714043993232, + "grad_norm": 8.5, + "learning_rate": 4.366876253826059e-06, + "loss": 2.5493, + "step": 12041 + }, + { + "epoch": 1.273477157360406, + "grad_norm": 8.25, + "learning_rate": 4.366762346201049e-06, + "loss": 2.686, + "step": 12042 + }, + { + "epoch": 1.273582910321489, + "grad_norm": 8.25, + "learning_rate": 4.366648429816086e-06, + "loss": 2.5966, + "step": 12043 + }, + { + "epoch": 1.273688663282572, + "grad_norm": 7.65625, + "learning_rate": 4.366534504671706e-06, + "loss": 2.4865, + "step": 12044 + }, + { + "epoch": 1.2737944162436547, + "grad_norm": 8.1875, + "learning_rate": 4.366420570768442e-06, + "loss": 2.556, + "step": 12045 + }, + { + "epoch": 1.2739001692047378, + "grad_norm": 8.0625, + "learning_rate": 4.36630662810683e-06, + "loss": 2.6156, + "step": 12046 + }, + { + "epoch": 1.2740059221658206, + "grad_norm": 8.75, + "learning_rate": 4.366192676687405e-06, + "loss": 2.5224, + "step": 12047 + }, + { + "epoch": 1.2741116751269035, + "grad_norm": 8.5, + "learning_rate": 4.3660787165107e-06, + "loss": 2.6891, + "step": 12048 + }, + { + "epoch": 1.2742174280879865, + "grad_norm": 9.5625, + "learning_rate": 4.365964747577252e-06, + "loss": 2.7582, + "step": 12049 + }, + { + "epoch": 1.2743231810490694, + "grad_norm": 9.875, + "learning_rate": 4.365850769887593e-06, + "loss": 2.8469, + "step": 12050 + }, + { + "epoch": 1.2744289340101522, + "grad_norm": 7.96875, + "learning_rate": 4.36573678344226e-06, + "loss": 2.5964, + "step": 12051 + }, + { + "epoch": 1.274534686971235, + "grad_norm": 8.125, + "learning_rate": 4.365622788241788e-06, + "loss": 2.5727, + "step": 12052 + }, + { + "epoch": 1.2746404399323181, + "grad_norm": 8.375, + "learning_rate": 4.365508784286711e-06, + "loss": 2.4974, + "step": 12053 + }, + { + "epoch": 1.274746192893401, + "grad_norm": 8.75, + "learning_rate": 4.365394771577564e-06, + "loss": 2.5481, + "step": 12054 + }, + { + "epoch": 1.274851945854484, + "grad_norm": 8.25, + "learning_rate": 4.3652807501148824e-06, + "loss": 2.549, + "step": 12055 + }, + { + "epoch": 1.2749576988155669, + "grad_norm": 7.84375, + "learning_rate": 4.365166719899201e-06, + "loss": 2.3851, + "step": 12056 + }, + { + "epoch": 1.2750634517766497, + "grad_norm": 8.75, + "learning_rate": 4.3650526809310555e-06, + "loss": 2.6087, + "step": 12057 + }, + { + "epoch": 1.2751692047377325, + "grad_norm": 8.0, + "learning_rate": 4.36493863321098e-06, + "loss": 2.4053, + "step": 12058 + }, + { + "epoch": 1.2752749576988156, + "grad_norm": 8.25, + "learning_rate": 4.36482457673951e-06, + "loss": 2.6636, + "step": 12059 + }, + { + "epoch": 1.2753807106598984, + "grad_norm": 7.75, + "learning_rate": 4.364710511517182e-06, + "loss": 2.2616, + "step": 12060 + }, + { + "epoch": 1.2754864636209815, + "grad_norm": 7.875, + "learning_rate": 4.36459643754453e-06, + "loss": 2.7768, + "step": 12061 + }, + { + "epoch": 1.2755922165820643, + "grad_norm": 8.0, + "learning_rate": 4.364482354822088e-06, + "loss": 2.6429, + "step": 12062 + }, + { + "epoch": 1.2756979695431472, + "grad_norm": 7.8125, + "learning_rate": 4.3643682633503945e-06, + "loss": 2.4923, + "step": 12063 + }, + { + "epoch": 1.27580372250423, + "grad_norm": 8.125, + "learning_rate": 4.364254163129982e-06, + "loss": 2.4185, + "step": 12064 + }, + { + "epoch": 1.275909475465313, + "grad_norm": 8.125, + "learning_rate": 4.364140054161388e-06, + "loss": 2.4536, + "step": 12065 + }, + { + "epoch": 1.276015228426396, + "grad_norm": 7.65625, + "learning_rate": 4.364025936445147e-06, + "loss": 2.496, + "step": 12066 + }, + { + "epoch": 1.276120981387479, + "grad_norm": 8.875, + "learning_rate": 4.363911809981795e-06, + "loss": 2.5783, + "step": 12067 + }, + { + "epoch": 1.2762267343485618, + "grad_norm": 7.875, + "learning_rate": 4.363797674771867e-06, + "loss": 2.3564, + "step": 12068 + }, + { + "epoch": 1.2763324873096447, + "grad_norm": 7.8125, + "learning_rate": 4.363683530815898e-06, + "loss": 2.4084, + "step": 12069 + }, + { + "epoch": 1.2764382402707275, + "grad_norm": 8.3125, + "learning_rate": 4.363569378114425e-06, + "loss": 2.6275, + "step": 12070 + }, + { + "epoch": 1.2765439932318106, + "grad_norm": 8.125, + "learning_rate": 4.363455216667984e-06, + "loss": 2.5642, + "step": 12071 + }, + { + "epoch": 1.2766497461928934, + "grad_norm": 9.1875, + "learning_rate": 4.363341046477108e-06, + "loss": 3.0033, + "step": 12072 + }, + { + "epoch": 1.2767554991539762, + "grad_norm": 7.625, + "learning_rate": 4.363226867542335e-06, + "loss": 2.4986, + "step": 12073 + }, + { + "epoch": 1.2768612521150593, + "grad_norm": 7.96875, + "learning_rate": 4.363112679864201e-06, + "loss": 2.4773, + "step": 12074 + }, + { + "epoch": 1.2769670050761421, + "grad_norm": 7.84375, + "learning_rate": 4.362998483443241e-06, + "loss": 2.4695, + "step": 12075 + }, + { + "epoch": 1.277072758037225, + "grad_norm": 8.3125, + "learning_rate": 4.362884278279991e-06, + "loss": 2.5918, + "step": 12076 + }, + { + "epoch": 1.2771785109983078, + "grad_norm": 8.0625, + "learning_rate": 4.362770064374987e-06, + "loss": 2.3944, + "step": 12077 + }, + { + "epoch": 1.2772842639593909, + "grad_norm": 8.5625, + "learning_rate": 4.362655841728764e-06, + "loss": 2.5238, + "step": 12078 + }, + { + "epoch": 1.2773900169204737, + "grad_norm": 7.875, + "learning_rate": 4.362541610341861e-06, + "loss": 2.4379, + "step": 12079 + }, + { + "epoch": 1.2774957698815568, + "grad_norm": 8.625, + "learning_rate": 4.36242737021481e-06, + "loss": 2.6301, + "step": 12080 + }, + { + "epoch": 1.2776015228426396, + "grad_norm": 8.75, + "learning_rate": 4.36231312134815e-06, + "loss": 2.761, + "step": 12081 + }, + { + "epoch": 1.2777072758037225, + "grad_norm": 9.1875, + "learning_rate": 4.362198863742416e-06, + "loss": 2.5185, + "step": 12082 + }, + { + "epoch": 1.2778130287648053, + "grad_norm": 8.625, + "learning_rate": 4.3620845973981445e-06, + "loss": 2.8243, + "step": 12083 + }, + { + "epoch": 1.2779187817258884, + "grad_norm": 9.125, + "learning_rate": 4.361970322315872e-06, + "loss": 2.4361, + "step": 12084 + }, + { + "epoch": 1.2780245346869712, + "grad_norm": 8.8125, + "learning_rate": 4.361856038496133e-06, + "loss": 2.9735, + "step": 12085 + }, + { + "epoch": 1.2781302876480543, + "grad_norm": 8.3125, + "learning_rate": 4.3617417459394665e-06, + "loss": 2.5059, + "step": 12086 + }, + { + "epoch": 1.278236040609137, + "grad_norm": 8.4375, + "learning_rate": 4.361627444646406e-06, + "loss": 2.4726, + "step": 12087 + }, + { + "epoch": 1.27834179357022, + "grad_norm": 8.4375, + "learning_rate": 4.3615131346174905e-06, + "loss": 2.7354, + "step": 12088 + }, + { + "epoch": 1.2784475465313028, + "grad_norm": 8.0, + "learning_rate": 4.361398815853254e-06, + "loss": 2.1551, + "step": 12089 + }, + { + "epoch": 1.2785532994923858, + "grad_norm": 8.375, + "learning_rate": 4.361284488354236e-06, + "loss": 2.5776, + "step": 12090 + }, + { + "epoch": 1.2786590524534687, + "grad_norm": 8.0625, + "learning_rate": 4.361170152120969e-06, + "loss": 2.3154, + "step": 12091 + }, + { + "epoch": 1.2787648054145517, + "grad_norm": 8.1875, + "learning_rate": 4.361055807153993e-06, + "loss": 2.6761, + "step": 12092 + }, + { + "epoch": 1.2788705583756346, + "grad_norm": 8.3125, + "learning_rate": 4.360941453453843e-06, + "loss": 2.5945, + "step": 12093 + }, + { + "epoch": 1.2789763113367174, + "grad_norm": 7.78125, + "learning_rate": 4.360827091021056e-06, + "loss": 2.3309, + "step": 12094 + }, + { + "epoch": 1.2790820642978002, + "grad_norm": 7.71875, + "learning_rate": 4.360712719856168e-06, + "loss": 2.5341, + "step": 12095 + }, + { + "epoch": 1.2791878172588833, + "grad_norm": 8.0, + "learning_rate": 4.3605983399597165e-06, + "loss": 2.5199, + "step": 12096 + }, + { + "epoch": 1.2792935702199661, + "grad_norm": 7.96875, + "learning_rate": 4.360483951332238e-06, + "loss": 2.4928, + "step": 12097 + }, + { + "epoch": 1.279399323181049, + "grad_norm": 8.6875, + "learning_rate": 4.360369553974269e-06, + "loss": 2.7131, + "step": 12098 + }, + { + "epoch": 1.279505076142132, + "grad_norm": 9.0, + "learning_rate": 4.360255147886347e-06, + "loss": 2.7178, + "step": 12099 + }, + { + "epoch": 1.2796108291032149, + "grad_norm": 8.3125, + "learning_rate": 4.360140733069008e-06, + "loss": 2.6051, + "step": 12100 + }, + { + "epoch": 1.2797165820642977, + "grad_norm": 8.375, + "learning_rate": 4.360026309522789e-06, + "loss": 2.6686, + "step": 12101 + }, + { + "epoch": 1.2798223350253808, + "grad_norm": 8.75, + "learning_rate": 4.359911877248227e-06, + "loss": 2.6447, + "step": 12102 + }, + { + "epoch": 1.2799280879864636, + "grad_norm": 8.1875, + "learning_rate": 4.35979743624586e-06, + "loss": 2.6167, + "step": 12103 + }, + { + "epoch": 1.2800338409475465, + "grad_norm": 8.0625, + "learning_rate": 4.359682986516225e-06, + "loss": 2.5265, + "step": 12104 + }, + { + "epoch": 1.2801395939086295, + "grad_norm": 8.4375, + "learning_rate": 4.359568528059858e-06, + "loss": 2.8135, + "step": 12105 + }, + { + "epoch": 1.2802453468697124, + "grad_norm": 8.5, + "learning_rate": 4.359454060877295e-06, + "loss": 2.6147, + "step": 12106 + }, + { + "epoch": 1.2803510998307952, + "grad_norm": 8.3125, + "learning_rate": 4.3593395849690755e-06, + "loss": 2.5652, + "step": 12107 + }, + { + "epoch": 1.280456852791878, + "grad_norm": 8.4375, + "learning_rate": 4.3592251003357364e-06, + "loss": 2.6821, + "step": 12108 + }, + { + "epoch": 1.280562605752961, + "grad_norm": 8.8125, + "learning_rate": 4.359110606977813e-06, + "loss": 2.385, + "step": 12109 + }, + { + "epoch": 1.280668358714044, + "grad_norm": 8.3125, + "learning_rate": 4.358996104895845e-06, + "loss": 2.4126, + "step": 12110 + }, + { + "epoch": 1.280774111675127, + "grad_norm": 8.1875, + "learning_rate": 4.358881594090368e-06, + "loss": 2.5215, + "step": 12111 + }, + { + "epoch": 1.2808798646362098, + "grad_norm": 7.71875, + "learning_rate": 4.35876707456192e-06, + "loss": 2.4335, + "step": 12112 + }, + { + "epoch": 1.2809856175972927, + "grad_norm": 8.3125, + "learning_rate": 4.358652546311039e-06, + "loss": 2.4895, + "step": 12113 + }, + { + "epoch": 1.2810913705583755, + "grad_norm": 7.84375, + "learning_rate": 4.358538009338261e-06, + "loss": 2.4244, + "step": 12114 + }, + { + "epoch": 1.2811971235194586, + "grad_norm": 7.96875, + "learning_rate": 4.358423463644124e-06, + "loss": 2.642, + "step": 12115 + }, + { + "epoch": 1.2813028764805414, + "grad_norm": 8.4375, + "learning_rate": 4.358308909229167e-06, + "loss": 2.4236, + "step": 12116 + }, + { + "epoch": 1.2814086294416245, + "grad_norm": 8.0625, + "learning_rate": 4.358194346093925e-06, + "loss": 2.5635, + "step": 12117 + }, + { + "epoch": 1.2815143824027073, + "grad_norm": 8.1875, + "learning_rate": 4.358079774238937e-06, + "loss": 2.6891, + "step": 12118 + }, + { + "epoch": 1.2816201353637902, + "grad_norm": 8.5, + "learning_rate": 4.357965193664741e-06, + "loss": 2.6291, + "step": 12119 + }, + { + "epoch": 1.281725888324873, + "grad_norm": 8.375, + "learning_rate": 4.357850604371875e-06, + "loss": 2.7146, + "step": 12120 + }, + { + "epoch": 1.281831641285956, + "grad_norm": 8.5625, + "learning_rate": 4.357736006360874e-06, + "loss": 2.4999, + "step": 12121 + }, + { + "epoch": 1.281937394247039, + "grad_norm": 8.0, + "learning_rate": 4.357621399632279e-06, + "loss": 2.5993, + "step": 12122 + }, + { + "epoch": 1.282043147208122, + "grad_norm": 7.875, + "learning_rate": 4.357506784186627e-06, + "loss": 2.5147, + "step": 12123 + }, + { + "epoch": 1.2821489001692048, + "grad_norm": 7.8125, + "learning_rate": 4.357392160024454e-06, + "loss": 2.551, + "step": 12124 + }, + { + "epoch": 1.2822546531302876, + "grad_norm": 8.375, + "learning_rate": 4.3572775271463005e-06, + "loss": 2.6522, + "step": 12125 + }, + { + "epoch": 1.2823604060913705, + "grad_norm": 8.3125, + "learning_rate": 4.357162885552703e-06, + "loss": 2.6001, + "step": 12126 + }, + { + "epoch": 1.2824661590524535, + "grad_norm": 8.6875, + "learning_rate": 4.357048235244199e-06, + "loss": 2.4363, + "step": 12127 + }, + { + "epoch": 1.2825719120135364, + "grad_norm": 8.1875, + "learning_rate": 4.356933576221328e-06, + "loss": 2.575, + "step": 12128 + }, + { + "epoch": 1.2826776649746192, + "grad_norm": 8.0625, + "learning_rate": 4.356818908484626e-06, + "loss": 2.5118, + "step": 12129 + }, + { + "epoch": 1.2827834179357023, + "grad_norm": 8.1875, + "learning_rate": 4.356704232034633e-06, + "loss": 2.5622, + "step": 12130 + }, + { + "epoch": 1.2828891708967851, + "grad_norm": 7.96875, + "learning_rate": 4.356589546871886e-06, + "loss": 2.5266, + "step": 12131 + }, + { + "epoch": 1.282994923857868, + "grad_norm": 8.4375, + "learning_rate": 4.3564748529969234e-06, + "loss": 2.5202, + "step": 12132 + }, + { + "epoch": 1.283100676818951, + "grad_norm": 8.125, + "learning_rate": 4.356360150410284e-06, + "loss": 2.4956, + "step": 12133 + }, + { + "epoch": 1.2832064297800339, + "grad_norm": 8.0625, + "learning_rate": 4.3562454391125055e-06, + "loss": 2.6571, + "step": 12134 + }, + { + "epoch": 1.2833121827411167, + "grad_norm": 8.5625, + "learning_rate": 4.356130719104127e-06, + "loss": 2.7765, + "step": 12135 + }, + { + "epoch": 1.2834179357021998, + "grad_norm": 7.84375, + "learning_rate": 4.3560159903856855e-06, + "loss": 2.436, + "step": 12136 + }, + { + "epoch": 1.2835236886632826, + "grad_norm": 7.9375, + "learning_rate": 4.35590125295772e-06, + "loss": 2.4933, + "step": 12137 + }, + { + "epoch": 1.2836294416243654, + "grad_norm": 8.4375, + "learning_rate": 4.35578650682077e-06, + "loss": 2.6435, + "step": 12138 + }, + { + "epoch": 1.2837351945854483, + "grad_norm": 8.5625, + "learning_rate": 4.355671751975371e-06, + "loss": 2.6768, + "step": 12139 + }, + { + "epoch": 1.2838409475465313, + "grad_norm": 10.0, + "learning_rate": 4.355556988422065e-06, + "loss": 2.5977, + "step": 12140 + }, + { + "epoch": 1.2839467005076142, + "grad_norm": 8.25, + "learning_rate": 4.355442216161388e-06, + "loss": 2.6948, + "step": 12141 + }, + { + "epoch": 1.2840524534686972, + "grad_norm": 8.0625, + "learning_rate": 4.35532743519388e-06, + "loss": 2.5736, + "step": 12142 + }, + { + "epoch": 1.28415820642978, + "grad_norm": 8.25, + "learning_rate": 4.355212645520079e-06, + "loss": 2.6895, + "step": 12143 + }, + { + "epoch": 1.284263959390863, + "grad_norm": 8.125, + "learning_rate": 4.355097847140524e-06, + "loss": 2.3511, + "step": 12144 + }, + { + "epoch": 1.2843697123519457, + "grad_norm": 8.875, + "learning_rate": 4.3549830400557535e-06, + "loss": 2.529, + "step": 12145 + }, + { + "epoch": 1.2844754653130288, + "grad_norm": 8.25, + "learning_rate": 4.354868224266307e-06, + "loss": 2.5232, + "step": 12146 + }, + { + "epoch": 1.2845812182741116, + "grad_norm": 8.4375, + "learning_rate": 4.354753399772721e-06, + "loss": 2.5201, + "step": 12147 + }, + { + "epoch": 1.2846869712351947, + "grad_norm": 7.96875, + "learning_rate": 4.354638566575536e-06, + "loss": 2.528, + "step": 12148 + }, + { + "epoch": 1.2847927241962775, + "grad_norm": 7.625, + "learning_rate": 4.354523724675292e-06, + "loss": 2.3857, + "step": 12149 + }, + { + "epoch": 1.2848984771573604, + "grad_norm": 7.75, + "learning_rate": 4.3544088740725255e-06, + "loss": 2.5012, + "step": 12150 + }, + { + "epoch": 1.2850042301184432, + "grad_norm": 8.1875, + "learning_rate": 4.354294014767777e-06, + "loss": 2.562, + "step": 12151 + }, + { + "epoch": 1.2851099830795263, + "grad_norm": 8.3125, + "learning_rate": 4.354179146761586e-06, + "loss": 2.5447, + "step": 12152 + }, + { + "epoch": 1.2852157360406091, + "grad_norm": 8.25, + "learning_rate": 4.3540642700544885e-06, + "loss": 2.6587, + "step": 12153 + }, + { + "epoch": 1.2853214890016922, + "grad_norm": 9.25, + "learning_rate": 4.353949384647027e-06, + "loss": 2.4966, + "step": 12154 + }, + { + "epoch": 1.285427241962775, + "grad_norm": 8.625, + "learning_rate": 4.35383449053974e-06, + "loss": 2.5487, + "step": 12155 + }, + { + "epoch": 1.2855329949238579, + "grad_norm": 8.625, + "learning_rate": 4.353719587733164e-06, + "loss": 2.5412, + "step": 12156 + }, + { + "epoch": 1.2856387478849407, + "grad_norm": 7.96875, + "learning_rate": 4.353604676227841e-06, + "loss": 2.4162, + "step": 12157 + }, + { + "epoch": 1.2857445008460238, + "grad_norm": 8.75, + "learning_rate": 4.35348975602431e-06, + "loss": 2.5928, + "step": 12158 + }, + { + "epoch": 1.2858502538071066, + "grad_norm": 8.3125, + "learning_rate": 4.353374827123109e-06, + "loss": 2.3236, + "step": 12159 + }, + { + "epoch": 1.2859560067681894, + "grad_norm": 8.625, + "learning_rate": 4.353259889524778e-06, + "loss": 2.6029, + "step": 12160 + }, + { + "epoch": 1.2860617597292725, + "grad_norm": 7.9375, + "learning_rate": 4.353144943229857e-06, + "loss": 2.4948, + "step": 12161 + }, + { + "epoch": 1.2861675126903553, + "grad_norm": 8.6875, + "learning_rate": 4.353029988238884e-06, + "loss": 2.616, + "step": 12162 + }, + { + "epoch": 1.2862732656514382, + "grad_norm": 8.5, + "learning_rate": 4.352915024552399e-06, + "loss": 2.6273, + "step": 12163 + }, + { + "epoch": 1.286379018612521, + "grad_norm": 8.0625, + "learning_rate": 4.352800052170942e-06, + "loss": 2.5356, + "step": 12164 + }, + { + "epoch": 1.286484771573604, + "grad_norm": 8.3125, + "learning_rate": 4.352685071095052e-06, + "loss": 2.481, + "step": 12165 + }, + { + "epoch": 1.286590524534687, + "grad_norm": 8.0625, + "learning_rate": 4.352570081325269e-06, + "loss": 2.6017, + "step": 12166 + }, + { + "epoch": 1.28669627749577, + "grad_norm": 7.9375, + "learning_rate": 4.352455082862132e-06, + "loss": 2.6249, + "step": 12167 + }, + { + "epoch": 1.2868020304568528, + "grad_norm": 8.3125, + "learning_rate": 4.35234007570618e-06, + "loss": 2.5322, + "step": 12168 + }, + { + "epoch": 1.2869077834179357, + "grad_norm": 8.0, + "learning_rate": 4.3522250598579546e-06, + "loss": 2.737, + "step": 12169 + }, + { + "epoch": 1.2870135363790185, + "grad_norm": 8.3125, + "learning_rate": 4.352110035317994e-06, + "loss": 2.4411, + "step": 12170 + }, + { + "epoch": 1.2871192893401016, + "grad_norm": 8.5, + "learning_rate": 4.35199500208684e-06, + "loss": 2.4438, + "step": 12171 + }, + { + "epoch": 1.2872250423011844, + "grad_norm": 8.5, + "learning_rate": 4.351879960165029e-06, + "loss": 2.703, + "step": 12172 + }, + { + "epoch": 1.2873307952622675, + "grad_norm": 7.9375, + "learning_rate": 4.351764909553103e-06, + "loss": 2.5568, + "step": 12173 + }, + { + "epoch": 1.2874365482233503, + "grad_norm": 8.25, + "learning_rate": 4.351649850251603e-06, + "loss": 2.7742, + "step": 12174 + }, + { + "epoch": 1.2875423011844331, + "grad_norm": 8.4375, + "learning_rate": 4.3515347822610665e-06, + "loss": 2.3984, + "step": 12175 + }, + { + "epoch": 1.287648054145516, + "grad_norm": 8.125, + "learning_rate": 4.3514197055820355e-06, + "loss": 2.4178, + "step": 12176 + }, + { + "epoch": 1.287753807106599, + "grad_norm": 8.1875, + "learning_rate": 4.351304620215047e-06, + "loss": 2.5603, + "step": 12177 + }, + { + "epoch": 1.2878595600676819, + "grad_norm": 8.125, + "learning_rate": 4.351189526160644e-06, + "loss": 2.5198, + "step": 12178 + }, + { + "epoch": 1.287965313028765, + "grad_norm": 7.875, + "learning_rate": 4.351074423419367e-06, + "loss": 2.4959, + "step": 12179 + }, + { + "epoch": 1.2880710659898478, + "grad_norm": 7.8125, + "learning_rate": 4.3509593119917534e-06, + "loss": 2.6169, + "step": 12180 + }, + { + "epoch": 1.2881768189509306, + "grad_norm": 8.3125, + "learning_rate": 4.350844191878345e-06, + "loss": 2.6931, + "step": 12181 + }, + { + "epoch": 1.2882825719120135, + "grad_norm": 9.3125, + "learning_rate": 4.350729063079682e-06, + "loss": 2.6118, + "step": 12182 + }, + { + "epoch": 1.2883883248730965, + "grad_norm": 8.0, + "learning_rate": 4.350613925596304e-06, + "loss": 2.429, + "step": 12183 + }, + { + "epoch": 1.2884940778341794, + "grad_norm": 8.9375, + "learning_rate": 4.350498779428752e-06, + "loss": 2.4881, + "step": 12184 + }, + { + "epoch": 1.2885998307952622, + "grad_norm": 8.25, + "learning_rate": 4.350383624577567e-06, + "loss": 2.6139, + "step": 12185 + }, + { + "epoch": 1.2887055837563453, + "grad_norm": 7.875, + "learning_rate": 4.350268461043287e-06, + "loss": 2.5268, + "step": 12186 + }, + { + "epoch": 1.288811336717428, + "grad_norm": 8.3125, + "learning_rate": 4.3501532888264545e-06, + "loss": 2.579, + "step": 12187 + }, + { + "epoch": 1.288917089678511, + "grad_norm": 8.5, + "learning_rate": 4.350038107927609e-06, + "loss": 2.6746, + "step": 12188 + }, + { + "epoch": 1.289022842639594, + "grad_norm": 8.0625, + "learning_rate": 4.349922918347291e-06, + "loss": 2.6673, + "step": 12189 + }, + { + "epoch": 1.2891285956006768, + "grad_norm": 8.0625, + "learning_rate": 4.349807720086042e-06, + "loss": 2.5496, + "step": 12190 + }, + { + "epoch": 1.2892343485617597, + "grad_norm": 8.8125, + "learning_rate": 4.3496925131444025e-06, + "loss": 2.5818, + "step": 12191 + }, + { + "epoch": 1.2893401015228427, + "grad_norm": 7.90625, + "learning_rate": 4.3495772975229115e-06, + "loss": 2.5899, + "step": 12192 + }, + { + "epoch": 1.2894458544839256, + "grad_norm": 8.5625, + "learning_rate": 4.349462073222111e-06, + "loss": 2.6708, + "step": 12193 + }, + { + "epoch": 1.2895516074450084, + "grad_norm": 7.6875, + "learning_rate": 4.349346840242541e-06, + "loss": 2.4058, + "step": 12194 + }, + { + "epoch": 1.2896573604060912, + "grad_norm": 8.0625, + "learning_rate": 4.3492315985847435e-06, + "loss": 2.6671, + "step": 12195 + }, + { + "epoch": 1.2897631133671743, + "grad_norm": 7.8125, + "learning_rate": 4.349116348249257e-06, + "loss": 2.5876, + "step": 12196 + }, + { + "epoch": 1.2898688663282571, + "grad_norm": 8.625, + "learning_rate": 4.349001089236625e-06, + "loss": 2.6961, + "step": 12197 + }, + { + "epoch": 1.2899746192893402, + "grad_norm": 8.0, + "learning_rate": 4.3488858215473875e-06, + "loss": 2.5045, + "step": 12198 + }, + { + "epoch": 1.290080372250423, + "grad_norm": 8.625, + "learning_rate": 4.348770545182084e-06, + "loss": 2.8922, + "step": 12199 + }, + { + "epoch": 1.2901861252115059, + "grad_norm": 8.125, + "learning_rate": 4.348655260141257e-06, + "loss": 2.4095, + "step": 12200 + }, + { + "epoch": 1.2902918781725887, + "grad_norm": 8.125, + "learning_rate": 4.3485399664254455e-06, + "loss": 2.3951, + "step": 12201 + }, + { + "epoch": 1.2903976311336718, + "grad_norm": 7.6875, + "learning_rate": 4.348424664035194e-06, + "loss": 2.6077, + "step": 12202 + }, + { + "epoch": 1.2905033840947546, + "grad_norm": 8.3125, + "learning_rate": 4.348309352971041e-06, + "loss": 2.6006, + "step": 12203 + }, + { + "epoch": 1.2906091370558377, + "grad_norm": 8.3125, + "learning_rate": 4.348194033233527e-06, + "loss": 2.5164, + "step": 12204 + }, + { + "epoch": 1.2907148900169205, + "grad_norm": 7.875, + "learning_rate": 4.3480787048231955e-06, + "loss": 2.4584, + "step": 12205 + }, + { + "epoch": 1.2908206429780034, + "grad_norm": 8.375, + "learning_rate": 4.347963367740586e-06, + "loss": 2.7127, + "step": 12206 + }, + { + "epoch": 1.2909263959390862, + "grad_norm": 8.5, + "learning_rate": 4.34784802198624e-06, + "loss": 2.6339, + "step": 12207 + }, + { + "epoch": 1.2910321489001693, + "grad_norm": 7.90625, + "learning_rate": 4.347732667560699e-06, + "loss": 2.5291, + "step": 12208 + }, + { + "epoch": 1.291137901861252, + "grad_norm": 8.6875, + "learning_rate": 4.347617304464505e-06, + "loss": 2.7276, + "step": 12209 + }, + { + "epoch": 1.2912436548223352, + "grad_norm": 8.625, + "learning_rate": 4.347501932698198e-06, + "loss": 2.5073, + "step": 12210 + }, + { + "epoch": 1.291349407783418, + "grad_norm": 8.0, + "learning_rate": 4.347386552262321e-06, + "loss": 2.5279, + "step": 12211 + }, + { + "epoch": 1.2914551607445008, + "grad_norm": 8.25, + "learning_rate": 4.347271163157413e-06, + "loss": 2.6387, + "step": 12212 + }, + { + "epoch": 1.2915609137055837, + "grad_norm": 8.0625, + "learning_rate": 4.347155765384018e-06, + "loss": 2.5622, + "step": 12213 + }, + { + "epoch": 1.2916666666666667, + "grad_norm": 7.6875, + "learning_rate": 4.347040358942676e-06, + "loss": 2.3912, + "step": 12214 + }, + { + "epoch": 1.2917724196277496, + "grad_norm": 8.5625, + "learning_rate": 4.346924943833929e-06, + "loss": 2.2975, + "step": 12215 + }, + { + "epoch": 1.2918781725888324, + "grad_norm": 8.4375, + "learning_rate": 4.3468095200583185e-06, + "loss": 2.58, + "step": 12216 + }, + { + "epoch": 1.2919839255499155, + "grad_norm": 7.84375, + "learning_rate": 4.346694087616386e-06, + "loss": 2.4435, + "step": 12217 + }, + { + "epoch": 1.2920896785109983, + "grad_norm": 8.25, + "learning_rate": 4.346578646508674e-06, + "loss": 2.8265, + "step": 12218 + }, + { + "epoch": 1.2921954314720812, + "grad_norm": 8.5625, + "learning_rate": 4.346463196735723e-06, + "loss": 2.6932, + "step": 12219 + }, + { + "epoch": 1.292301184433164, + "grad_norm": 8.125, + "learning_rate": 4.346347738298076e-06, + "loss": 2.4593, + "step": 12220 + }, + { + "epoch": 1.292406937394247, + "grad_norm": 8.1875, + "learning_rate": 4.346232271196274e-06, + "loss": 2.6472, + "step": 12221 + }, + { + "epoch": 1.29251269035533, + "grad_norm": 9.0625, + "learning_rate": 4.346116795430859e-06, + "loss": 2.3963, + "step": 12222 + }, + { + "epoch": 1.292618443316413, + "grad_norm": 8.1875, + "learning_rate": 4.346001311002373e-06, + "loss": 2.5179, + "step": 12223 + }, + { + "epoch": 1.2927241962774958, + "grad_norm": 7.9375, + "learning_rate": 4.345885817911358e-06, + "loss": 2.4367, + "step": 12224 + }, + { + "epoch": 1.2928299492385786, + "grad_norm": 8.375, + "learning_rate": 4.3457703161583565e-06, + "loss": 2.3521, + "step": 12225 + }, + { + "epoch": 1.2929357021996615, + "grad_norm": 8.375, + "learning_rate": 4.345654805743908e-06, + "loss": 2.7944, + "step": 12226 + }, + { + "epoch": 1.2930414551607445, + "grad_norm": 7.9375, + "learning_rate": 4.345539286668558e-06, + "loss": 2.3894, + "step": 12227 + }, + { + "epoch": 1.2931472081218274, + "grad_norm": 8.0, + "learning_rate": 4.345423758932847e-06, + "loss": 2.5375, + "step": 12228 + }, + { + "epoch": 1.2932529610829104, + "grad_norm": 8.0625, + "learning_rate": 4.3453082225373155e-06, + "loss": 2.3537, + "step": 12229 + }, + { + "epoch": 1.2933587140439933, + "grad_norm": 8.0625, + "learning_rate": 4.345192677482508e-06, + "loss": 2.3225, + "step": 12230 + }, + { + "epoch": 1.2934644670050761, + "grad_norm": 8.0625, + "learning_rate": 4.3450771237689656e-06, + "loss": 2.5043, + "step": 12231 + }, + { + "epoch": 1.293570219966159, + "grad_norm": 7.59375, + "learning_rate": 4.344961561397231e-06, + "loss": 2.4934, + "step": 12232 + }, + { + "epoch": 1.293675972927242, + "grad_norm": 8.5625, + "learning_rate": 4.3448459903678466e-06, + "loss": 2.4115, + "step": 12233 + }, + { + "epoch": 1.2937817258883249, + "grad_norm": 9.75, + "learning_rate": 4.344730410681355e-06, + "loss": 2.5796, + "step": 12234 + }, + { + "epoch": 1.293887478849408, + "grad_norm": 7.71875, + "learning_rate": 4.344614822338297e-06, + "loss": 2.3665, + "step": 12235 + }, + { + "epoch": 1.2939932318104908, + "grad_norm": 8.125, + "learning_rate": 4.344499225339217e-06, + "loss": 2.4456, + "step": 12236 + }, + { + "epoch": 1.2940989847715736, + "grad_norm": 8.6875, + "learning_rate": 4.344383619684655e-06, + "loss": 2.4813, + "step": 12237 + }, + { + "epoch": 1.2942047377326564, + "grad_norm": 8.3125, + "learning_rate": 4.344268005375157e-06, + "loss": 2.5181, + "step": 12238 + }, + { + "epoch": 1.2943104906937395, + "grad_norm": 8.1875, + "learning_rate": 4.344152382411262e-06, + "loss": 2.5452, + "step": 12239 + }, + { + "epoch": 1.2944162436548223, + "grad_norm": 7.84375, + "learning_rate": 4.344036750793515e-06, + "loss": 2.4438, + "step": 12240 + }, + { + "epoch": 1.2945219966159052, + "grad_norm": 8.375, + "learning_rate": 4.343921110522457e-06, + "loss": 2.3234, + "step": 12241 + }, + { + "epoch": 1.2946277495769882, + "grad_norm": 7.75, + "learning_rate": 4.343805461598632e-06, + "loss": 2.4671, + "step": 12242 + }, + { + "epoch": 1.294733502538071, + "grad_norm": 7.875, + "learning_rate": 4.343689804022581e-06, + "loss": 2.4093, + "step": 12243 + }, + { + "epoch": 1.294839255499154, + "grad_norm": 8.0625, + "learning_rate": 4.34357413779485e-06, + "loss": 2.4614, + "step": 12244 + }, + { + "epoch": 1.294945008460237, + "grad_norm": 8.9375, + "learning_rate": 4.343458462915977e-06, + "loss": 2.6871, + "step": 12245 + }, + { + "epoch": 1.2950507614213198, + "grad_norm": 8.3125, + "learning_rate": 4.3433427793865094e-06, + "loss": 2.391, + "step": 12246 + }, + { + "epoch": 1.2951565143824026, + "grad_norm": 7.6875, + "learning_rate": 4.343227087206987e-06, + "loss": 2.2321, + "step": 12247 + }, + { + "epoch": 1.2952622673434857, + "grad_norm": 8.1875, + "learning_rate": 4.343111386377954e-06, + "loss": 2.7136, + "step": 12248 + }, + { + "epoch": 1.2953680203045685, + "grad_norm": 8.875, + "learning_rate": 4.342995676899953e-06, + "loss": 2.5102, + "step": 12249 + }, + { + "epoch": 1.2954737732656514, + "grad_norm": 8.125, + "learning_rate": 4.342879958773527e-06, + "loss": 2.5736, + "step": 12250 + }, + { + "epoch": 1.2955795262267342, + "grad_norm": 8.4375, + "learning_rate": 4.342764231999219e-06, + "loss": 2.8296, + "step": 12251 + }, + { + "epoch": 1.2956852791878173, + "grad_norm": 8.125, + "learning_rate": 4.342648496577572e-06, + "loss": 2.4297, + "step": 12252 + }, + { + "epoch": 1.2957910321489001, + "grad_norm": 8.625, + "learning_rate": 4.34253275250913e-06, + "loss": 2.6737, + "step": 12253 + }, + { + "epoch": 1.2958967851099832, + "grad_norm": 8.0625, + "learning_rate": 4.342416999794434e-06, + "loss": 2.4879, + "step": 12254 + }, + { + "epoch": 1.296002538071066, + "grad_norm": 9.5, + "learning_rate": 4.34230123843403e-06, + "loss": 2.6827, + "step": 12255 + }, + { + "epoch": 1.2961082910321489, + "grad_norm": 8.125, + "learning_rate": 4.342185468428459e-06, + "loss": 2.4708, + "step": 12256 + }, + { + "epoch": 1.2962140439932317, + "grad_norm": 7.9375, + "learning_rate": 4.342069689778265e-06, + "loss": 2.4603, + "step": 12257 + }, + { + "epoch": 1.2963197969543148, + "grad_norm": 8.25, + "learning_rate": 4.341953902483992e-06, + "loss": 2.601, + "step": 12258 + }, + { + "epoch": 1.2964255499153976, + "grad_norm": 7.78125, + "learning_rate": 4.341838106546182e-06, + "loss": 2.6123, + "step": 12259 + }, + { + "epoch": 1.2965313028764807, + "grad_norm": 8.125, + "learning_rate": 4.3417223019653785e-06, + "loss": 2.5075, + "step": 12260 + }, + { + "epoch": 1.2966370558375635, + "grad_norm": 8.0625, + "learning_rate": 4.341606488742126e-06, + "loss": 2.6248, + "step": 12261 + }, + { + "epoch": 1.2967428087986463, + "grad_norm": 8.0625, + "learning_rate": 4.3414906668769684e-06, + "loss": 2.4129, + "step": 12262 + }, + { + "epoch": 1.2968485617597292, + "grad_norm": 8.875, + "learning_rate": 4.3413748363704476e-06, + "loss": 2.2594, + "step": 12263 + }, + { + "epoch": 1.2969543147208122, + "grad_norm": 8.4375, + "learning_rate": 4.341258997223107e-06, + "loss": 2.4235, + "step": 12264 + }, + { + "epoch": 1.297060067681895, + "grad_norm": 8.6875, + "learning_rate": 4.341143149435491e-06, + "loss": 2.6781, + "step": 12265 + }, + { + "epoch": 1.2971658206429781, + "grad_norm": 8.375, + "learning_rate": 4.3410272930081445e-06, + "loss": 2.376, + "step": 12266 + }, + { + "epoch": 1.297271573604061, + "grad_norm": 8.25, + "learning_rate": 4.3409114279416095e-06, + "loss": 2.646, + "step": 12267 + }, + { + "epoch": 1.2973773265651438, + "grad_norm": 8.0625, + "learning_rate": 4.3407955542364295e-06, + "loss": 2.4442, + "step": 12268 + }, + { + "epoch": 1.2974830795262267, + "grad_norm": 8.5625, + "learning_rate": 4.340679671893149e-06, + "loss": 2.6758, + "step": 12269 + }, + { + "epoch": 1.2975888324873097, + "grad_norm": 8.5625, + "learning_rate": 4.340563780912312e-06, + "loss": 2.5539, + "step": 12270 + }, + { + "epoch": 1.2976945854483926, + "grad_norm": 8.5, + "learning_rate": 4.340447881294462e-06, + "loss": 2.7264, + "step": 12271 + }, + { + "epoch": 1.2978003384094754, + "grad_norm": 8.0625, + "learning_rate": 4.340331973040141e-06, + "loss": 2.5653, + "step": 12272 + }, + { + "epoch": 1.2979060913705585, + "grad_norm": 8.3125, + "learning_rate": 4.340216056149897e-06, + "loss": 2.4763, + "step": 12273 + }, + { + "epoch": 1.2980118443316413, + "grad_norm": 7.8125, + "learning_rate": 4.3401001306242705e-06, + "loss": 2.5009, + "step": 12274 + }, + { + "epoch": 1.2981175972927241, + "grad_norm": 7.9375, + "learning_rate": 4.339984196463808e-06, + "loss": 2.4362, + "step": 12275 + }, + { + "epoch": 1.298223350253807, + "grad_norm": 8.1875, + "learning_rate": 4.339868253669051e-06, + "loss": 2.6347, + "step": 12276 + }, + { + "epoch": 1.29832910321489, + "grad_norm": 8.1875, + "learning_rate": 4.339752302240545e-06, + "loss": 2.5979, + "step": 12277 + }, + { + "epoch": 1.2984348561759729, + "grad_norm": 7.875, + "learning_rate": 4.339636342178833e-06, + "loss": 2.6194, + "step": 12278 + }, + { + "epoch": 1.298540609137056, + "grad_norm": 7.8125, + "learning_rate": 4.339520373484461e-06, + "loss": 2.6137, + "step": 12279 + }, + { + "epoch": 1.2986463620981388, + "grad_norm": 8.3125, + "learning_rate": 4.339404396157972e-06, + "loss": 2.542, + "step": 12280 + }, + { + "epoch": 1.2987521150592216, + "grad_norm": 8.3125, + "learning_rate": 4.339288410199911e-06, + "loss": 2.5805, + "step": 12281 + }, + { + "epoch": 1.2988578680203045, + "grad_norm": 7.65625, + "learning_rate": 4.339172415610822e-06, + "loss": 2.4039, + "step": 12282 + }, + { + "epoch": 1.2989636209813875, + "grad_norm": 8.375, + "learning_rate": 4.3390564123912474e-06, + "loss": 2.6612, + "step": 12283 + }, + { + "epoch": 1.2990693739424704, + "grad_norm": 8.125, + "learning_rate": 4.338940400541734e-06, + "loss": 2.5598, + "step": 12284 + }, + { + "epoch": 1.2991751269035534, + "grad_norm": 8.3125, + "learning_rate": 4.338824380062826e-06, + "loss": 2.5436, + "step": 12285 + }, + { + "epoch": 1.2992808798646363, + "grad_norm": 8.875, + "learning_rate": 4.3387083509550666e-06, + "loss": 2.7962, + "step": 12286 + }, + { + "epoch": 1.299386632825719, + "grad_norm": 8.375, + "learning_rate": 4.338592313219001e-06, + "loss": 2.6598, + "step": 12287 + }, + { + "epoch": 1.299492385786802, + "grad_norm": 8.25, + "learning_rate": 4.338476266855174e-06, + "loss": 2.572, + "step": 12288 + }, + { + "epoch": 1.299598138747885, + "grad_norm": 8.0625, + "learning_rate": 4.33836021186413e-06, + "loss": 2.6841, + "step": 12289 + }, + { + "epoch": 1.2997038917089678, + "grad_norm": 8.0625, + "learning_rate": 4.338244148246412e-06, + "loss": 2.4937, + "step": 12290 + }, + { + "epoch": 1.299809644670051, + "grad_norm": 9.0625, + "learning_rate": 4.338128076002567e-06, + "loss": 2.7845, + "step": 12291 + }, + { + "epoch": 1.2999153976311337, + "grad_norm": 8.3125, + "learning_rate": 4.3380119951331386e-06, + "loss": 2.3827, + "step": 12292 + }, + { + "epoch": 1.3000211505922166, + "grad_norm": 8.1875, + "learning_rate": 4.3378959056386725e-06, + "loss": 2.5139, + "step": 12293 + }, + { + "epoch": 1.3001269035532994, + "grad_norm": 8.375, + "learning_rate": 4.337779807519712e-06, + "loss": 2.8036, + "step": 12294 + }, + { + "epoch": 1.3002326565143825, + "grad_norm": 8.0625, + "learning_rate": 4.337663700776802e-06, + "loss": 2.4825, + "step": 12295 + }, + { + "epoch": 1.3003384094754653, + "grad_norm": 8.8125, + "learning_rate": 4.337547585410488e-06, + "loss": 2.6535, + "step": 12296 + }, + { + "epoch": 1.3004441624365481, + "grad_norm": 8.25, + "learning_rate": 4.337431461421314e-06, + "loss": 2.5294, + "step": 12297 + }, + { + "epoch": 1.3005499153976312, + "grad_norm": 8.25, + "learning_rate": 4.337315328809827e-06, + "loss": 2.4626, + "step": 12298 + }, + { + "epoch": 1.300655668358714, + "grad_norm": 8.3125, + "learning_rate": 4.33719918757657e-06, + "loss": 2.7671, + "step": 12299 + }, + { + "epoch": 1.3007614213197969, + "grad_norm": 8.6875, + "learning_rate": 4.3370830377220885e-06, + "loss": 2.6017, + "step": 12300 + }, + { + "epoch": 1.30086717428088, + "grad_norm": 8.375, + "learning_rate": 4.336966879246928e-06, + "loss": 2.5249, + "step": 12301 + }, + { + "epoch": 1.3009729272419628, + "grad_norm": 8.0, + "learning_rate": 4.336850712151633e-06, + "loss": 2.5084, + "step": 12302 + }, + { + "epoch": 1.3010786802030456, + "grad_norm": 8.6875, + "learning_rate": 4.336734536436748e-06, + "loss": 2.621, + "step": 12303 + }, + { + "epoch": 1.3011844331641287, + "grad_norm": 7.4375, + "learning_rate": 4.33661835210282e-06, + "loss": 2.4704, + "step": 12304 + }, + { + "epoch": 1.3012901861252115, + "grad_norm": 8.9375, + "learning_rate": 4.336502159150394e-06, + "loss": 2.2784, + "step": 12305 + }, + { + "epoch": 1.3013959390862944, + "grad_norm": 8.1875, + "learning_rate": 4.336385957580013e-06, + "loss": 2.4801, + "step": 12306 + }, + { + "epoch": 1.3015016920473772, + "grad_norm": 8.0625, + "learning_rate": 4.3362697473922245e-06, + "loss": 2.5954, + "step": 12307 + }, + { + "epoch": 1.3016074450084603, + "grad_norm": 7.65625, + "learning_rate": 4.336153528587573e-06, + "loss": 2.4636, + "step": 12308 + }, + { + "epoch": 1.301713197969543, + "grad_norm": 8.875, + "learning_rate": 4.336037301166604e-06, + "loss": 2.4875, + "step": 12309 + }, + { + "epoch": 1.3018189509306262, + "grad_norm": 8.0625, + "learning_rate": 4.335921065129863e-06, + "loss": 2.355, + "step": 12310 + }, + { + "epoch": 1.301924703891709, + "grad_norm": 9.3125, + "learning_rate": 4.3358048204778945e-06, + "loss": 2.75, + "step": 12311 + }, + { + "epoch": 1.3020304568527918, + "grad_norm": 8.1875, + "learning_rate": 4.335688567211246e-06, + "loss": 2.4089, + "step": 12312 + }, + { + "epoch": 1.3021362098138747, + "grad_norm": 8.375, + "learning_rate": 4.335572305330461e-06, + "loss": 2.3676, + "step": 12313 + }, + { + "epoch": 1.3022419627749577, + "grad_norm": 8.8125, + "learning_rate": 4.335456034836086e-06, + "loss": 2.3744, + "step": 12314 + }, + { + "epoch": 1.3023477157360406, + "grad_norm": 8.0625, + "learning_rate": 4.335339755728666e-06, + "loss": 2.5977, + "step": 12315 + }, + { + "epoch": 1.3024534686971236, + "grad_norm": 8.375, + "learning_rate": 4.335223468008748e-06, + "loss": 2.5847, + "step": 12316 + }, + { + "epoch": 1.3025592216582065, + "grad_norm": 7.65625, + "learning_rate": 4.335107171676877e-06, + "loss": 2.3996, + "step": 12317 + }, + { + "epoch": 1.3026649746192893, + "grad_norm": 7.96875, + "learning_rate": 4.334990866733598e-06, + "loss": 2.6391, + "step": 12318 + }, + { + "epoch": 1.3027707275803722, + "grad_norm": 8.0, + "learning_rate": 4.334874553179458e-06, + "loss": 2.5823, + "step": 12319 + }, + { + "epoch": 1.3028764805414552, + "grad_norm": 8.4375, + "learning_rate": 4.334758231015e-06, + "loss": 2.7092, + "step": 12320 + }, + { + "epoch": 1.302982233502538, + "grad_norm": 7.875, + "learning_rate": 4.334641900240775e-06, + "loss": 2.4981, + "step": 12321 + }, + { + "epoch": 1.3030879864636211, + "grad_norm": 7.90625, + "learning_rate": 4.3345255608573245e-06, + "loss": 2.5858, + "step": 12322 + }, + { + "epoch": 1.303193739424704, + "grad_norm": 8.1875, + "learning_rate": 4.334409212865195e-06, + "loss": 2.6354, + "step": 12323 + }, + { + "epoch": 1.3032994923857868, + "grad_norm": 8.125, + "learning_rate": 4.334292856264935e-06, + "loss": 2.5551, + "step": 12324 + }, + { + "epoch": 1.3034052453468696, + "grad_norm": 8.1875, + "learning_rate": 4.334176491057088e-06, + "loss": 2.7033, + "step": 12325 + }, + { + "epoch": 1.3035109983079527, + "grad_norm": 7.8125, + "learning_rate": 4.3340601172422e-06, + "loss": 2.5502, + "step": 12326 + }, + { + "epoch": 1.3036167512690355, + "grad_norm": 8.625, + "learning_rate": 4.3339437348208196e-06, + "loss": 2.6802, + "step": 12327 + }, + { + "epoch": 1.3037225042301184, + "grad_norm": 7.78125, + "learning_rate": 4.333827343793491e-06, + "loss": 2.5689, + "step": 12328 + }, + { + "epoch": 1.3038282571912014, + "grad_norm": 7.90625, + "learning_rate": 4.33371094416076e-06, + "loss": 2.53, + "step": 12329 + }, + { + "epoch": 1.3039340101522843, + "grad_norm": 8.4375, + "learning_rate": 4.333594535923173e-06, + "loss": 2.5395, + "step": 12330 + }, + { + "epoch": 1.3040397631133671, + "grad_norm": 8.25, + "learning_rate": 4.333478119081278e-06, + "loss": 2.4834, + "step": 12331 + }, + { + "epoch": 1.30414551607445, + "grad_norm": 8.125, + "learning_rate": 4.33336169363562e-06, + "loss": 2.5543, + "step": 12332 + }, + { + "epoch": 1.304251269035533, + "grad_norm": 7.6875, + "learning_rate": 4.333245259586745e-06, + "loss": 2.4081, + "step": 12333 + }, + { + "epoch": 1.3043570219966159, + "grad_norm": 8.0, + "learning_rate": 4.333128816935199e-06, + "loss": 2.5452, + "step": 12334 + }, + { + "epoch": 1.304462774957699, + "grad_norm": 8.5, + "learning_rate": 4.33301236568153e-06, + "loss": 2.6021, + "step": 12335 + }, + { + "epoch": 1.3045685279187818, + "grad_norm": 7.9375, + "learning_rate": 4.332895905826284e-06, + "loss": 2.5402, + "step": 12336 + }, + { + "epoch": 1.3046742808798646, + "grad_norm": 8.8125, + "learning_rate": 4.332779437370007e-06, + "loss": 2.6817, + "step": 12337 + }, + { + "epoch": 1.3047800338409474, + "grad_norm": 8.25, + "learning_rate": 4.332662960313245e-06, + "loss": 2.4647, + "step": 12338 + }, + { + "epoch": 1.3048857868020305, + "grad_norm": 7.8125, + "learning_rate": 4.332546474656546e-06, + "loss": 2.5434, + "step": 12339 + }, + { + "epoch": 1.3049915397631133, + "grad_norm": 8.125, + "learning_rate": 4.332429980400455e-06, + "loss": 2.662, + "step": 12340 + }, + { + "epoch": 1.3050972927241964, + "grad_norm": 10.1875, + "learning_rate": 4.33231347754552e-06, + "loss": 2.5151, + "step": 12341 + }, + { + "epoch": 1.3052030456852792, + "grad_norm": 8.4375, + "learning_rate": 4.332196966092288e-06, + "loss": 2.1871, + "step": 12342 + }, + { + "epoch": 1.305308798646362, + "grad_norm": 7.84375, + "learning_rate": 4.332080446041305e-06, + "loss": 2.445, + "step": 12343 + }, + { + "epoch": 1.305414551607445, + "grad_norm": 7.96875, + "learning_rate": 4.331963917393116e-06, + "loss": 2.3984, + "step": 12344 + }, + { + "epoch": 1.305520304568528, + "grad_norm": 7.78125, + "learning_rate": 4.331847380148271e-06, + "loss": 2.5087, + "step": 12345 + }, + { + "epoch": 1.3056260575296108, + "grad_norm": 8.5, + "learning_rate": 4.331730834307315e-06, + "loss": 2.5211, + "step": 12346 + }, + { + "epoch": 1.3057318104906939, + "grad_norm": 8.0625, + "learning_rate": 4.331614279870796e-06, + "loss": 2.4851, + "step": 12347 + }, + { + "epoch": 1.3058375634517767, + "grad_norm": 7.75, + "learning_rate": 4.3314977168392594e-06, + "loss": 2.556, + "step": 12348 + }, + { + "epoch": 1.3059433164128595, + "grad_norm": 8.875, + "learning_rate": 4.3313811452132535e-06, + "loss": 2.3433, + "step": 12349 + }, + { + "epoch": 1.3060490693739424, + "grad_norm": 7.75, + "learning_rate": 4.331264564993324e-06, + "loss": 2.2999, + "step": 12350 + }, + { + "epoch": 1.3061548223350254, + "grad_norm": 8.25, + "learning_rate": 4.33114797618002e-06, + "loss": 2.457, + "step": 12351 + }, + { + "epoch": 1.3062605752961083, + "grad_norm": 8.4375, + "learning_rate": 4.331031378773888e-06, + "loss": 2.4829, + "step": 12352 + }, + { + "epoch": 1.3063663282571911, + "grad_norm": 8.1875, + "learning_rate": 4.330914772775473e-06, + "loss": 2.5849, + "step": 12353 + }, + { + "epoch": 1.3064720812182742, + "grad_norm": 8.25, + "learning_rate": 4.3307981581853244e-06, + "loss": 2.5574, + "step": 12354 + }, + { + "epoch": 1.306577834179357, + "grad_norm": 7.71875, + "learning_rate": 4.330681535003989e-06, + "loss": 2.6138, + "step": 12355 + }, + { + "epoch": 1.3066835871404399, + "grad_norm": 8.5, + "learning_rate": 4.330564903232014e-06, + "loss": 2.3695, + "step": 12356 + }, + { + "epoch": 1.306789340101523, + "grad_norm": 8.375, + "learning_rate": 4.330448262869946e-06, + "loss": 2.4735, + "step": 12357 + }, + { + "epoch": 1.3068950930626058, + "grad_norm": 7.9375, + "learning_rate": 4.330331613918334e-06, + "loss": 2.5179, + "step": 12358 + }, + { + "epoch": 1.3070008460236886, + "grad_norm": 9.0625, + "learning_rate": 4.3302149563777234e-06, + "loss": 2.643, + "step": 12359 + }, + { + "epoch": 1.3071065989847717, + "grad_norm": 7.96875, + "learning_rate": 4.330098290248662e-06, + "loss": 2.5799, + "step": 12360 + }, + { + "epoch": 1.3072123519458545, + "grad_norm": 8.0, + "learning_rate": 4.3299816155316995e-06, + "loss": 2.5955, + "step": 12361 + }, + { + "epoch": 1.3073181049069373, + "grad_norm": 8.25, + "learning_rate": 4.329864932227381e-06, + "loss": 2.5595, + "step": 12362 + }, + { + "epoch": 1.3074238578680202, + "grad_norm": 8.625, + "learning_rate": 4.329748240336254e-06, + "loss": 2.1819, + "step": 12363 + }, + { + "epoch": 1.3075296108291032, + "grad_norm": 7.71875, + "learning_rate": 4.329631539858868e-06, + "loss": 2.4197, + "step": 12364 + }, + { + "epoch": 1.307635363790186, + "grad_norm": 8.3125, + "learning_rate": 4.329514830795769e-06, + "loss": 2.3825, + "step": 12365 + }, + { + "epoch": 1.3077411167512691, + "grad_norm": 7.90625, + "learning_rate": 4.329398113147505e-06, + "loss": 2.5216, + "step": 12366 + }, + { + "epoch": 1.307846869712352, + "grad_norm": 8.25, + "learning_rate": 4.329281386914623e-06, + "loss": 2.5203, + "step": 12367 + }, + { + "epoch": 1.3079526226734348, + "grad_norm": 8.1875, + "learning_rate": 4.329164652097672e-06, + "loss": 2.5846, + "step": 12368 + }, + { + "epoch": 1.3080583756345177, + "grad_norm": 7.96875, + "learning_rate": 4.3290479086972006e-06, + "loss": 2.6663, + "step": 12369 + }, + { + "epoch": 1.3081641285956007, + "grad_norm": 8.625, + "learning_rate": 4.328931156713755e-06, + "loss": 2.6907, + "step": 12370 + }, + { + "epoch": 1.3082698815566836, + "grad_norm": 8.25, + "learning_rate": 4.3288143961478836e-06, + "loss": 2.5576, + "step": 12371 + }, + { + "epoch": 1.3083756345177666, + "grad_norm": 8.75, + "learning_rate": 4.3286976270001335e-06, + "loss": 2.5933, + "step": 12372 + }, + { + "epoch": 1.3084813874788495, + "grad_norm": 8.0625, + "learning_rate": 4.328580849271054e-06, + "loss": 2.489, + "step": 12373 + }, + { + "epoch": 1.3085871404399323, + "grad_norm": 8.0625, + "learning_rate": 4.328464062961191e-06, + "loss": 2.4725, + "step": 12374 + }, + { + "epoch": 1.3086928934010151, + "grad_norm": 7.90625, + "learning_rate": 4.328347268071096e-06, + "loss": 2.5218, + "step": 12375 + }, + { + "epoch": 1.3087986463620982, + "grad_norm": 8.25, + "learning_rate": 4.328230464601314e-06, + "loss": 2.4624, + "step": 12376 + }, + { + "epoch": 1.308904399323181, + "grad_norm": 8.1875, + "learning_rate": 4.3281136525523935e-06, + "loss": 2.5299, + "step": 12377 + }, + { + "epoch": 1.309010152284264, + "grad_norm": 8.1875, + "learning_rate": 4.3279968319248845e-06, + "loss": 2.5348, + "step": 12378 + }, + { + "epoch": 1.309115905245347, + "grad_norm": 8.3125, + "learning_rate": 4.327880002719333e-06, + "loss": 2.7351, + "step": 12379 + }, + { + "epoch": 1.3092216582064298, + "grad_norm": 8.75, + "learning_rate": 4.327763164936289e-06, + "loss": 2.5732, + "step": 12380 + }, + { + "epoch": 1.3093274111675126, + "grad_norm": 7.875, + "learning_rate": 4.3276463185763004e-06, + "loss": 2.3499, + "step": 12381 + }, + { + "epoch": 1.3094331641285957, + "grad_norm": 8.875, + "learning_rate": 4.327529463639914e-06, + "loss": 2.5378, + "step": 12382 + }, + { + "epoch": 1.3095389170896785, + "grad_norm": 8.0, + "learning_rate": 4.32741260012768e-06, + "loss": 2.4621, + "step": 12383 + }, + { + "epoch": 1.3096446700507614, + "grad_norm": 7.875, + "learning_rate": 4.327295728040146e-06, + "loss": 2.5846, + "step": 12384 + }, + { + "epoch": 1.3097504230118444, + "grad_norm": 8.125, + "learning_rate": 4.3271788473778596e-06, + "loss": 2.7514, + "step": 12385 + }, + { + "epoch": 1.3098561759729273, + "grad_norm": 8.4375, + "learning_rate": 4.327061958141371e-06, + "loss": 2.6006, + "step": 12386 + }, + { + "epoch": 1.30996192893401, + "grad_norm": 8.1875, + "learning_rate": 4.326945060331228e-06, + "loss": 2.5714, + "step": 12387 + }, + { + "epoch": 1.310067681895093, + "grad_norm": 9.0, + "learning_rate": 4.326828153947979e-06, + "loss": 2.5863, + "step": 12388 + }, + { + "epoch": 1.310173434856176, + "grad_norm": 8.0625, + "learning_rate": 4.326711238992172e-06, + "loss": 2.4732, + "step": 12389 + }, + { + "epoch": 1.3102791878172588, + "grad_norm": 8.9375, + "learning_rate": 4.326594315464356e-06, + "loss": 2.5222, + "step": 12390 + }, + { + "epoch": 1.310384940778342, + "grad_norm": 8.6875, + "learning_rate": 4.326477383365081e-06, + "loss": 2.5262, + "step": 12391 + }, + { + "epoch": 1.3104906937394247, + "grad_norm": 7.9375, + "learning_rate": 4.326360442694895e-06, + "loss": 2.2201, + "step": 12392 + }, + { + "epoch": 1.3105964467005076, + "grad_norm": 8.0, + "learning_rate": 4.326243493454345e-06, + "loss": 2.6127, + "step": 12393 + }, + { + "epoch": 1.3107021996615904, + "grad_norm": 8.125, + "learning_rate": 4.3261265356439816e-06, + "loss": 2.5678, + "step": 12394 + }, + { + "epoch": 1.3108079526226735, + "grad_norm": 7.8125, + "learning_rate": 4.326009569264353e-06, + "loss": 2.4346, + "step": 12395 + }, + { + "epoch": 1.3109137055837563, + "grad_norm": 9.25, + "learning_rate": 4.325892594316009e-06, + "loss": 2.6489, + "step": 12396 + }, + { + "epoch": 1.3110194585448394, + "grad_norm": 9.125, + "learning_rate": 4.325775610799498e-06, + "loss": 2.593, + "step": 12397 + }, + { + "epoch": 1.3111252115059222, + "grad_norm": 7.84375, + "learning_rate": 4.325658618715368e-06, + "loss": 2.5691, + "step": 12398 + }, + { + "epoch": 1.311230964467005, + "grad_norm": 8.25, + "learning_rate": 4.32554161806417e-06, + "loss": 2.5565, + "step": 12399 + }, + { + "epoch": 1.3113367174280879, + "grad_norm": 8.625, + "learning_rate": 4.32542460884645e-06, + "loss": 2.6632, + "step": 12400 + }, + { + "epoch": 1.311442470389171, + "grad_norm": 8.3125, + "learning_rate": 4.325307591062759e-06, + "loss": 2.5952, + "step": 12401 + }, + { + "epoch": 1.3115482233502538, + "grad_norm": 8.375, + "learning_rate": 4.325190564713648e-06, + "loss": 2.4769, + "step": 12402 + }, + { + "epoch": 1.3116539763113368, + "grad_norm": 7.875, + "learning_rate": 4.325073529799663e-06, + "loss": 2.6127, + "step": 12403 + }, + { + "epoch": 1.3117597292724197, + "grad_norm": 8.1875, + "learning_rate": 4.324956486321354e-06, + "loss": 2.5785, + "step": 12404 + }, + { + "epoch": 1.3118654822335025, + "grad_norm": 8.5625, + "learning_rate": 4.324839434279271e-06, + "loss": 2.5806, + "step": 12405 + }, + { + "epoch": 1.3119712351945854, + "grad_norm": 8.125, + "learning_rate": 4.324722373673963e-06, + "loss": 2.572, + "step": 12406 + }, + { + "epoch": 1.3120769881556684, + "grad_norm": 8.0, + "learning_rate": 4.324605304505979e-06, + "loss": 2.4499, + "step": 12407 + }, + { + "epoch": 1.3121827411167513, + "grad_norm": 8.25, + "learning_rate": 4.324488226775868e-06, + "loss": 2.4555, + "step": 12408 + }, + { + "epoch": 1.312288494077834, + "grad_norm": 8.5, + "learning_rate": 4.3243711404841805e-06, + "loss": 2.4602, + "step": 12409 + }, + { + "epoch": 1.3123942470389172, + "grad_norm": 8.875, + "learning_rate": 4.324254045631465e-06, + "loss": 2.6975, + "step": 12410 + }, + { + "epoch": 1.3125, + "grad_norm": 8.0625, + "learning_rate": 4.324136942218272e-06, + "loss": 2.5949, + "step": 12411 + }, + { + "epoch": 1.3126057529610828, + "grad_norm": 7.84375, + "learning_rate": 4.32401983024515e-06, + "loss": 2.4956, + "step": 12412 + }, + { + "epoch": 1.312711505922166, + "grad_norm": 7.96875, + "learning_rate": 4.323902709712648e-06, + "loss": 2.6391, + "step": 12413 + }, + { + "epoch": 1.3128172588832487, + "grad_norm": 8.3125, + "learning_rate": 4.323785580621318e-06, + "loss": 2.6386, + "step": 12414 + }, + { + "epoch": 1.3129230118443316, + "grad_norm": 7.9375, + "learning_rate": 4.323668442971708e-06, + "loss": 2.5044, + "step": 12415 + }, + { + "epoch": 1.3130287648054146, + "grad_norm": 8.25, + "learning_rate": 4.3235512967643675e-06, + "loss": 2.5101, + "step": 12416 + }, + { + "epoch": 1.3131345177664975, + "grad_norm": 8.5, + "learning_rate": 4.323434141999846e-06, + "loss": 2.3674, + "step": 12417 + }, + { + "epoch": 1.3132402707275803, + "grad_norm": 7.90625, + "learning_rate": 4.323316978678695e-06, + "loss": 2.513, + "step": 12418 + }, + { + "epoch": 1.3133460236886632, + "grad_norm": 8.4375, + "learning_rate": 4.3231998068014625e-06, + "loss": 2.7133, + "step": 12419 + }, + { + "epoch": 1.3134517766497462, + "grad_norm": 8.5625, + "learning_rate": 4.323082626368699e-06, + "loss": 2.5806, + "step": 12420 + }, + { + "epoch": 1.313557529610829, + "grad_norm": 8.1875, + "learning_rate": 4.322965437380955e-06, + "loss": 2.5617, + "step": 12421 + }, + { + "epoch": 1.3136632825719121, + "grad_norm": 8.3125, + "learning_rate": 4.322848239838778e-06, + "loss": 2.5641, + "step": 12422 + }, + { + "epoch": 1.313769035532995, + "grad_norm": 8.125, + "learning_rate": 4.322731033742721e-06, + "loss": 2.4556, + "step": 12423 + }, + { + "epoch": 1.3138747884940778, + "grad_norm": 8.1875, + "learning_rate": 4.322613819093333e-06, + "loss": 2.8056, + "step": 12424 + }, + { + "epoch": 1.3139805414551606, + "grad_norm": 8.625, + "learning_rate": 4.3224965958911635e-06, + "loss": 2.5418, + "step": 12425 + }, + { + "epoch": 1.3140862944162437, + "grad_norm": 8.1875, + "learning_rate": 4.322379364136763e-06, + "loss": 2.579, + "step": 12426 + }, + { + "epoch": 1.3141920473773265, + "grad_norm": 8.625, + "learning_rate": 4.32226212383068e-06, + "loss": 2.5592, + "step": 12427 + }, + { + "epoch": 1.3142978003384096, + "grad_norm": 9.5, + "learning_rate": 4.322144874973468e-06, + "loss": 2.679, + "step": 12428 + }, + { + "epoch": 1.3144035532994924, + "grad_norm": 8.625, + "learning_rate": 4.322027617565675e-06, + "loss": 2.4879, + "step": 12429 + }, + { + "epoch": 1.3145093062605753, + "grad_norm": 9.0, + "learning_rate": 4.321910351607851e-06, + "loss": 2.7541, + "step": 12430 + }, + { + "epoch": 1.314615059221658, + "grad_norm": 8.5625, + "learning_rate": 4.321793077100547e-06, + "loss": 2.7102, + "step": 12431 + }, + { + "epoch": 1.3147208121827412, + "grad_norm": 8.0, + "learning_rate": 4.321675794044313e-06, + "loss": 2.4654, + "step": 12432 + }, + { + "epoch": 1.314826565143824, + "grad_norm": 9.375, + "learning_rate": 4.3215585024397e-06, + "loss": 2.4638, + "step": 12433 + }, + { + "epoch": 1.314932318104907, + "grad_norm": 8.0, + "learning_rate": 4.321441202287258e-06, + "loss": 2.4281, + "step": 12434 + }, + { + "epoch": 1.31503807106599, + "grad_norm": 8.5625, + "learning_rate": 4.3213238935875366e-06, + "loss": 2.3913, + "step": 12435 + }, + { + "epoch": 1.3151438240270727, + "grad_norm": 7.625, + "learning_rate": 4.321206576341088e-06, + "loss": 2.4553, + "step": 12436 + }, + { + "epoch": 1.3152495769881556, + "grad_norm": 7.9375, + "learning_rate": 4.321089250548461e-06, + "loss": 2.4687, + "step": 12437 + }, + { + "epoch": 1.3153553299492386, + "grad_norm": 8.0625, + "learning_rate": 4.320971916210207e-06, + "loss": 2.6083, + "step": 12438 + }, + { + "epoch": 1.3154610829103215, + "grad_norm": 7.625, + "learning_rate": 4.3208545733268775e-06, + "loss": 2.6839, + "step": 12439 + }, + { + "epoch": 1.3155668358714043, + "grad_norm": 8.25, + "learning_rate": 4.3207372218990215e-06, + "loss": 2.699, + "step": 12440 + }, + { + "epoch": 1.3156725888324874, + "grad_norm": 8.4375, + "learning_rate": 4.320619861927191e-06, + "loss": 2.6773, + "step": 12441 + }, + { + "epoch": 1.3157783417935702, + "grad_norm": 8.5625, + "learning_rate": 4.320502493411935e-06, + "loss": 2.3024, + "step": 12442 + }, + { + "epoch": 1.315884094754653, + "grad_norm": 8.75, + "learning_rate": 4.320385116353806e-06, + "loss": 2.6802, + "step": 12443 + }, + { + "epoch": 1.315989847715736, + "grad_norm": 7.6875, + "learning_rate": 4.320267730753353e-06, + "loss": 2.2025, + "step": 12444 + }, + { + "epoch": 1.316095600676819, + "grad_norm": 9.0, + "learning_rate": 4.32015033661113e-06, + "loss": 2.8814, + "step": 12445 + }, + { + "epoch": 1.3162013536379018, + "grad_norm": 8.1875, + "learning_rate": 4.320032933927685e-06, + "loss": 2.5474, + "step": 12446 + }, + { + "epoch": 1.3163071065989849, + "grad_norm": 8.25, + "learning_rate": 4.319915522703569e-06, + "loss": 2.5535, + "step": 12447 + }, + { + "epoch": 1.3164128595600677, + "grad_norm": 8.3125, + "learning_rate": 4.319798102939336e-06, + "loss": 2.5749, + "step": 12448 + }, + { + "epoch": 1.3165186125211505, + "grad_norm": 8.0625, + "learning_rate": 4.3196806746355325e-06, + "loss": 2.3699, + "step": 12449 + }, + { + "epoch": 1.3166243654822334, + "grad_norm": 7.875, + "learning_rate": 4.319563237792712e-06, + "loss": 2.6044, + "step": 12450 + }, + { + "epoch": 1.3167301184433164, + "grad_norm": 7.6875, + "learning_rate": 4.3194457924114265e-06, + "loss": 2.5911, + "step": 12451 + }, + { + "epoch": 1.3168358714043993, + "grad_norm": 8.375, + "learning_rate": 4.319328338492226e-06, + "loss": 2.5972, + "step": 12452 + }, + { + "epoch": 1.3169416243654823, + "grad_norm": 8.4375, + "learning_rate": 4.319210876035661e-06, + "loss": 2.4056, + "step": 12453 + }, + { + "epoch": 1.3170473773265652, + "grad_norm": 8.125, + "learning_rate": 4.319093405042284e-06, + "loss": 2.4921, + "step": 12454 + }, + { + "epoch": 1.317153130287648, + "grad_norm": 8.0625, + "learning_rate": 4.318975925512645e-06, + "loss": 2.4873, + "step": 12455 + }, + { + "epoch": 1.3172588832487309, + "grad_norm": 7.875, + "learning_rate": 4.318858437447296e-06, + "loss": 2.5948, + "step": 12456 + }, + { + "epoch": 1.317364636209814, + "grad_norm": 8.25, + "learning_rate": 4.318740940846789e-06, + "loss": 2.6085, + "step": 12457 + }, + { + "epoch": 1.3174703891708968, + "grad_norm": 9.4375, + "learning_rate": 4.318623435711674e-06, + "loss": 2.3571, + "step": 12458 + }, + { + "epoch": 1.3175761421319798, + "grad_norm": 8.3125, + "learning_rate": 4.318505922042503e-06, + "loss": 2.573, + "step": 12459 + }, + { + "epoch": 1.3176818950930627, + "grad_norm": 8.1875, + "learning_rate": 4.318388399839827e-06, + "loss": 2.3919, + "step": 12460 + }, + { + "epoch": 1.3177876480541455, + "grad_norm": 8.375, + "learning_rate": 4.3182708691041985e-06, + "loss": 2.4328, + "step": 12461 + }, + { + "epoch": 1.3178934010152283, + "grad_norm": 7.96875, + "learning_rate": 4.318153329836168e-06, + "loss": 2.4559, + "step": 12462 + }, + { + "epoch": 1.3179991539763114, + "grad_norm": 8.75, + "learning_rate": 4.318035782036288e-06, + "loss": 2.7279, + "step": 12463 + }, + { + "epoch": 1.3181049069373942, + "grad_norm": 8.75, + "learning_rate": 4.31791822570511e-06, + "loss": 2.5208, + "step": 12464 + }, + { + "epoch": 1.318210659898477, + "grad_norm": 8.1875, + "learning_rate": 4.317800660843185e-06, + "loss": 2.6605, + "step": 12465 + }, + { + "epoch": 1.3183164128595601, + "grad_norm": 8.3125, + "learning_rate": 4.317683087451065e-06, + "loss": 2.7753, + "step": 12466 + }, + { + "epoch": 1.318422165820643, + "grad_norm": 8.4375, + "learning_rate": 4.317565505529301e-06, + "loss": 2.2379, + "step": 12467 + }, + { + "epoch": 1.3185279187817258, + "grad_norm": 7.96875, + "learning_rate": 4.317447915078445e-06, + "loss": 2.4804, + "step": 12468 + }, + { + "epoch": 1.3186336717428089, + "grad_norm": 7.8125, + "learning_rate": 4.317330316099051e-06, + "loss": 2.5171, + "step": 12469 + }, + { + "epoch": 1.3187394247038917, + "grad_norm": 7.8125, + "learning_rate": 4.317212708591668e-06, + "loss": 2.4346, + "step": 12470 + }, + { + "epoch": 1.3188451776649746, + "grad_norm": 7.8125, + "learning_rate": 4.317095092556849e-06, + "loss": 2.4465, + "step": 12471 + }, + { + "epoch": 1.3189509306260576, + "grad_norm": 8.0, + "learning_rate": 4.316977467995146e-06, + "loss": 2.5189, + "step": 12472 + }, + { + "epoch": 1.3190566835871405, + "grad_norm": 8.3125, + "learning_rate": 4.316859834907111e-06, + "loss": 2.4372, + "step": 12473 + }, + { + "epoch": 1.3191624365482233, + "grad_norm": 8.4375, + "learning_rate": 4.316742193293296e-06, + "loss": 2.5207, + "step": 12474 + }, + { + "epoch": 1.3192681895093061, + "grad_norm": 8.1875, + "learning_rate": 4.316624543154252e-06, + "loss": 2.4543, + "step": 12475 + }, + { + "epoch": 1.3193739424703892, + "grad_norm": 8.1875, + "learning_rate": 4.316506884490533e-06, + "loss": 2.3542, + "step": 12476 + }, + { + "epoch": 1.319479695431472, + "grad_norm": 8.25, + "learning_rate": 4.31638921730269e-06, + "loss": 2.6106, + "step": 12477 + }, + { + "epoch": 1.319585448392555, + "grad_norm": 8.0, + "learning_rate": 4.316271541591275e-06, + "loss": 2.6278, + "step": 12478 + }, + { + "epoch": 1.319691201353638, + "grad_norm": 8.25, + "learning_rate": 4.316153857356841e-06, + "loss": 2.6607, + "step": 12479 + }, + { + "epoch": 1.3197969543147208, + "grad_norm": 8.6875, + "learning_rate": 4.316036164599939e-06, + "loss": 2.6897, + "step": 12480 + }, + { + "epoch": 1.3199027072758036, + "grad_norm": 8.5, + "learning_rate": 4.315918463321122e-06, + "loss": 2.6378, + "step": 12481 + }, + { + "epoch": 1.3200084602368867, + "grad_norm": 9.0625, + "learning_rate": 4.315800753520942e-06, + "loss": 2.5867, + "step": 12482 + }, + { + "epoch": 1.3201142131979695, + "grad_norm": 8.875, + "learning_rate": 4.3156830351999536e-06, + "loss": 2.6898, + "step": 12483 + }, + { + "epoch": 1.3202199661590526, + "grad_norm": 8.0, + "learning_rate": 4.3155653083587055e-06, + "loss": 2.5168, + "step": 12484 + }, + { + "epoch": 1.3203257191201354, + "grad_norm": 8.25, + "learning_rate": 4.315447572997753e-06, + "loss": 2.6225, + "step": 12485 + }, + { + "epoch": 1.3204314720812182, + "grad_norm": 7.6875, + "learning_rate": 4.3153298291176475e-06, + "loss": 2.5314, + "step": 12486 + }, + { + "epoch": 1.320537225042301, + "grad_norm": 8.5, + "learning_rate": 4.3152120767189395e-06, + "loss": 2.6072, + "step": 12487 + }, + { + "epoch": 1.3206429780033841, + "grad_norm": 8.1875, + "learning_rate": 4.315094315802186e-06, + "loss": 2.5444, + "step": 12488 + }, + { + "epoch": 1.320748730964467, + "grad_norm": 8.1875, + "learning_rate": 4.314976546367936e-06, + "loss": 2.544, + "step": 12489 + }, + { + "epoch": 1.32085448392555, + "grad_norm": 9.0625, + "learning_rate": 4.314858768416744e-06, + "loss": 2.5334, + "step": 12490 + }, + { + "epoch": 1.3209602368866329, + "grad_norm": 8.1875, + "learning_rate": 4.314740981949162e-06, + "loss": 2.5172, + "step": 12491 + }, + { + "epoch": 1.3210659898477157, + "grad_norm": 7.84375, + "learning_rate": 4.314623186965743e-06, + "loss": 2.6272, + "step": 12492 + }, + { + "epoch": 1.3211717428087986, + "grad_norm": 8.5, + "learning_rate": 4.314505383467039e-06, + "loss": 2.7492, + "step": 12493 + }, + { + "epoch": 1.3212774957698816, + "grad_norm": 8.5625, + "learning_rate": 4.3143875714536036e-06, + "loss": 2.5981, + "step": 12494 + }, + { + "epoch": 1.3213832487309645, + "grad_norm": 8.1875, + "learning_rate": 4.314269750925989e-06, + "loss": 2.483, + "step": 12495 + }, + { + "epoch": 1.3214890016920473, + "grad_norm": 8.3125, + "learning_rate": 4.3141519218847485e-06, + "loss": 2.5465, + "step": 12496 + }, + { + "epoch": 1.3215947546531304, + "grad_norm": 8.1875, + "learning_rate": 4.314034084330436e-06, + "loss": 2.5443, + "step": 12497 + }, + { + "epoch": 1.3217005076142132, + "grad_norm": 8.125, + "learning_rate": 4.313916238263603e-06, + "loss": 2.5226, + "step": 12498 + }, + { + "epoch": 1.321806260575296, + "grad_norm": 8.5, + "learning_rate": 4.313798383684802e-06, + "loss": 2.5898, + "step": 12499 + }, + { + "epoch": 1.3219120135363789, + "grad_norm": 7.8125, + "learning_rate": 4.313680520594588e-06, + "loss": 2.5642, + "step": 12500 + }, + { + "epoch": 1.322017766497462, + "grad_norm": 8.0625, + "learning_rate": 4.313562648993513e-06, + "loss": 2.5807, + "step": 12501 + }, + { + "epoch": 1.3221235194585448, + "grad_norm": 8.625, + "learning_rate": 4.31344476888213e-06, + "loss": 2.7619, + "step": 12502 + }, + { + "epoch": 1.3222292724196278, + "grad_norm": 9.3125, + "learning_rate": 4.313326880260992e-06, + "loss": 2.577, + "step": 12503 + }, + { + "epoch": 1.3223350253807107, + "grad_norm": 8.125, + "learning_rate": 4.313208983130653e-06, + "loss": 2.7508, + "step": 12504 + }, + { + "epoch": 1.3224407783417935, + "grad_norm": 9.0625, + "learning_rate": 4.313091077491665e-06, + "loss": 2.6221, + "step": 12505 + }, + { + "epoch": 1.3225465313028764, + "grad_norm": 8.3125, + "learning_rate": 4.312973163344583e-06, + "loss": 2.6818, + "step": 12506 + }, + { + "epoch": 1.3226522842639594, + "grad_norm": 8.6875, + "learning_rate": 4.31285524068996e-06, + "loss": 2.954, + "step": 12507 + }, + { + "epoch": 1.3227580372250423, + "grad_norm": 7.96875, + "learning_rate": 4.312737309528348e-06, + "loss": 2.6022, + "step": 12508 + }, + { + "epoch": 1.3228637901861253, + "grad_norm": 8.3125, + "learning_rate": 4.312619369860301e-06, + "loss": 2.6879, + "step": 12509 + }, + { + "epoch": 1.3229695431472082, + "grad_norm": 9.0625, + "learning_rate": 4.3125014216863734e-06, + "loss": 2.8883, + "step": 12510 + }, + { + "epoch": 1.323075296108291, + "grad_norm": 8.75, + "learning_rate": 4.312383465007117e-06, + "loss": 2.665, + "step": 12511 + }, + { + "epoch": 1.3231810490693738, + "grad_norm": 7.71875, + "learning_rate": 4.312265499823087e-06, + "loss": 2.4559, + "step": 12512 + }, + { + "epoch": 1.323286802030457, + "grad_norm": 7.9375, + "learning_rate": 4.312147526134835e-06, + "loss": 2.2671, + "step": 12513 + }, + { + "epoch": 1.3233925549915397, + "grad_norm": 7.71875, + "learning_rate": 4.312029543942918e-06, + "loss": 2.3707, + "step": 12514 + }, + { + "epoch": 1.3234983079526228, + "grad_norm": 8.25, + "learning_rate": 4.311911553247885e-06, + "loss": 2.7846, + "step": 12515 + }, + { + "epoch": 1.3236040609137056, + "grad_norm": 8.625, + "learning_rate": 4.3117935540502935e-06, + "loss": 2.5386, + "step": 12516 + }, + { + "epoch": 1.3237098138747885, + "grad_norm": 8.375, + "learning_rate": 4.311675546350695e-06, + "loss": 2.4039, + "step": 12517 + }, + { + "epoch": 1.3238155668358713, + "grad_norm": 8.375, + "learning_rate": 4.311557530149645e-06, + "loss": 2.6108, + "step": 12518 + }, + { + "epoch": 1.3239213197969544, + "grad_norm": 10.125, + "learning_rate": 4.311439505447696e-06, + "loss": 2.1019, + "step": 12519 + }, + { + "epoch": 1.3240270727580372, + "grad_norm": 7.96875, + "learning_rate": 4.311321472245402e-06, + "loss": 2.4107, + "step": 12520 + }, + { + "epoch": 1.32413282571912, + "grad_norm": 8.3125, + "learning_rate": 4.3112034305433186e-06, + "loss": 2.5483, + "step": 12521 + }, + { + "epoch": 1.3242385786802031, + "grad_norm": 8.375, + "learning_rate": 4.311085380341996e-06, + "loss": 2.5075, + "step": 12522 + }, + { + "epoch": 1.324344331641286, + "grad_norm": 8.125, + "learning_rate": 4.310967321641992e-06, + "loss": 2.6185, + "step": 12523 + }, + { + "epoch": 1.3244500846023688, + "grad_norm": 9.4375, + "learning_rate": 4.3108492544438584e-06, + "loss": 2.303, + "step": 12524 + }, + { + "epoch": 1.3245558375634519, + "grad_norm": 7.71875, + "learning_rate": 4.310731178748151e-06, + "loss": 2.5228, + "step": 12525 + }, + { + "epoch": 1.3246615905245347, + "grad_norm": 8.375, + "learning_rate": 4.310613094555422e-06, + "loss": 2.6035, + "step": 12526 + }, + { + "epoch": 1.3247673434856175, + "grad_norm": 9.0, + "learning_rate": 4.310495001866225e-06, + "loss": 2.2911, + "step": 12527 + }, + { + "epoch": 1.3248730964467006, + "grad_norm": 8.0625, + "learning_rate": 4.310376900681117e-06, + "loss": 2.4914, + "step": 12528 + }, + { + "epoch": 1.3249788494077834, + "grad_norm": 8.3125, + "learning_rate": 4.31025879100065e-06, + "loss": 2.6313, + "step": 12529 + }, + { + "epoch": 1.3250846023688663, + "grad_norm": 8.0625, + "learning_rate": 4.310140672825379e-06, + "loss": 2.6146, + "step": 12530 + }, + { + "epoch": 1.325190355329949, + "grad_norm": 8.4375, + "learning_rate": 4.310022546155858e-06, + "loss": 2.8456, + "step": 12531 + }, + { + "epoch": 1.3252961082910322, + "grad_norm": 8.25, + "learning_rate": 4.309904410992641e-06, + "loss": 2.7055, + "step": 12532 + }, + { + "epoch": 1.325401861252115, + "grad_norm": 8.3125, + "learning_rate": 4.309786267336284e-06, + "loss": 2.3378, + "step": 12533 + }, + { + "epoch": 1.325507614213198, + "grad_norm": 8.1875, + "learning_rate": 4.30966811518734e-06, + "loss": 2.4977, + "step": 12534 + }, + { + "epoch": 1.325613367174281, + "grad_norm": 8.3125, + "learning_rate": 4.309549954546362e-06, + "loss": 2.5383, + "step": 12535 + }, + { + "epoch": 1.3257191201353637, + "grad_norm": 8.0625, + "learning_rate": 4.309431785413908e-06, + "loss": 2.4524, + "step": 12536 + }, + { + "epoch": 1.3258248730964466, + "grad_norm": 8.1875, + "learning_rate": 4.30931360779053e-06, + "loss": 2.5096, + "step": 12537 + }, + { + "epoch": 1.3259306260575296, + "grad_norm": 8.4375, + "learning_rate": 4.309195421676784e-06, + "loss": 2.5879, + "step": 12538 + }, + { + "epoch": 1.3260363790186125, + "grad_norm": 8.75, + "learning_rate": 4.309077227073223e-06, + "loss": 2.4949, + "step": 12539 + }, + { + "epoch": 1.3261421319796955, + "grad_norm": 8.25, + "learning_rate": 4.308959023980402e-06, + "loss": 2.3405, + "step": 12540 + }, + { + "epoch": 1.3262478849407784, + "grad_norm": 7.90625, + "learning_rate": 4.308840812398877e-06, + "loss": 2.6357, + "step": 12541 + }, + { + "epoch": 1.3263536379018612, + "grad_norm": 8.0625, + "learning_rate": 4.308722592329202e-06, + "loss": 2.7088, + "step": 12542 + }, + { + "epoch": 1.326459390862944, + "grad_norm": 8.1875, + "learning_rate": 4.308604363771931e-06, + "loss": 2.6144, + "step": 12543 + }, + { + "epoch": 1.3265651438240271, + "grad_norm": 7.78125, + "learning_rate": 4.308486126727619e-06, + "loss": 2.3074, + "step": 12544 + }, + { + "epoch": 1.32667089678511, + "grad_norm": 8.25, + "learning_rate": 4.3083678811968225e-06, + "loss": 2.5494, + "step": 12545 + }, + { + "epoch": 1.326776649746193, + "grad_norm": 8.0625, + "learning_rate": 4.308249627180095e-06, + "loss": 2.469, + "step": 12546 + }, + { + "epoch": 1.3268824027072759, + "grad_norm": 8.375, + "learning_rate": 4.308131364677991e-06, + "loss": 2.5046, + "step": 12547 + }, + { + "epoch": 1.3269881556683587, + "grad_norm": 8.125, + "learning_rate": 4.3080130936910654e-06, + "loss": 2.5389, + "step": 12548 + }, + { + "epoch": 1.3270939086294415, + "grad_norm": 8.3125, + "learning_rate": 4.307894814219874e-06, + "loss": 2.6462, + "step": 12549 + }, + { + "epoch": 1.3271996615905246, + "grad_norm": 7.9375, + "learning_rate": 4.3077765262649725e-06, + "loss": 2.3301, + "step": 12550 + }, + { + "epoch": 1.3273054145516074, + "grad_norm": 8.3125, + "learning_rate": 4.307658229826915e-06, + "loss": 2.533, + "step": 12551 + }, + { + "epoch": 1.3274111675126903, + "grad_norm": 8.3125, + "learning_rate": 4.307539924906257e-06, + "loss": 2.4245, + "step": 12552 + }, + { + "epoch": 1.3275169204737733, + "grad_norm": 8.25, + "learning_rate": 4.307421611503552e-06, + "loss": 2.6585, + "step": 12553 + }, + { + "epoch": 1.3276226734348562, + "grad_norm": 8.3125, + "learning_rate": 4.307303289619358e-06, + "loss": 2.5647, + "step": 12554 + }, + { + "epoch": 1.327728426395939, + "grad_norm": 7.90625, + "learning_rate": 4.307184959254228e-06, + "loss": 2.3229, + "step": 12555 + }, + { + "epoch": 1.3278341793570219, + "grad_norm": 7.96875, + "learning_rate": 4.307066620408718e-06, + "loss": 2.4, + "step": 12556 + }, + { + "epoch": 1.327939932318105, + "grad_norm": 8.6875, + "learning_rate": 4.306948273083384e-06, + "loss": 2.5787, + "step": 12557 + }, + { + "epoch": 1.3280456852791878, + "grad_norm": 8.1875, + "learning_rate": 4.3068299172787806e-06, + "loss": 2.4427, + "step": 12558 + }, + { + "epoch": 1.3281514382402708, + "grad_norm": 7.9375, + "learning_rate": 4.306711552995464e-06, + "loss": 2.4879, + "step": 12559 + }, + { + "epoch": 1.3282571912013537, + "grad_norm": 8.4375, + "learning_rate": 4.306593180233988e-06, + "loss": 2.7654, + "step": 12560 + }, + { + "epoch": 1.3283629441624365, + "grad_norm": 8.5, + "learning_rate": 4.306474798994909e-06, + "loss": 2.6879, + "step": 12561 + }, + { + "epoch": 1.3284686971235193, + "grad_norm": 8.1875, + "learning_rate": 4.306356409278783e-06, + "loss": 2.377, + "step": 12562 + }, + { + "epoch": 1.3285744500846024, + "grad_norm": 7.96875, + "learning_rate": 4.306238011086165e-06, + "loss": 2.4518, + "step": 12563 + }, + { + "epoch": 1.3286802030456852, + "grad_norm": 8.0625, + "learning_rate": 4.3061196044176115e-06, + "loss": 2.6108, + "step": 12564 + }, + { + "epoch": 1.3287859560067683, + "grad_norm": 8.5, + "learning_rate": 4.306001189273676e-06, + "loss": 2.3061, + "step": 12565 + }, + { + "epoch": 1.3288917089678511, + "grad_norm": 8.125, + "learning_rate": 4.305882765654916e-06, + "loss": 2.6137, + "step": 12566 + }, + { + "epoch": 1.328997461928934, + "grad_norm": 8.125, + "learning_rate": 4.305764333561887e-06, + "loss": 2.5324, + "step": 12567 + }, + { + "epoch": 1.3291032148900168, + "grad_norm": 8.5, + "learning_rate": 4.305645892995145e-06, + "loss": 2.4666, + "step": 12568 + }, + { + "epoch": 1.3292089678510999, + "grad_norm": 8.75, + "learning_rate": 4.305527443955245e-06, + "loss": 2.7613, + "step": 12569 + }, + { + "epoch": 1.3293147208121827, + "grad_norm": 8.375, + "learning_rate": 4.305408986442743e-06, + "loss": 2.2032, + "step": 12570 + }, + { + "epoch": 1.3294204737732658, + "grad_norm": 8.0625, + "learning_rate": 4.305290520458195e-06, + "loss": 2.6043, + "step": 12571 + }, + { + "epoch": 1.3295262267343486, + "grad_norm": 9.5, + "learning_rate": 4.305172046002158e-06, + "loss": 2.5042, + "step": 12572 + }, + { + "epoch": 1.3296319796954315, + "grad_norm": 8.3125, + "learning_rate": 4.305053563075185e-06, + "loss": 2.5794, + "step": 12573 + }, + { + "epoch": 1.3297377326565143, + "grad_norm": 8.25, + "learning_rate": 4.304935071677835e-06, + "loss": 2.4721, + "step": 12574 + }, + { + "epoch": 1.3298434856175974, + "grad_norm": 8.625, + "learning_rate": 4.304816571810662e-06, + "loss": 2.466, + "step": 12575 + }, + { + "epoch": 1.3299492385786802, + "grad_norm": 8.375, + "learning_rate": 4.304698063474224e-06, + "loss": 2.5308, + "step": 12576 + }, + { + "epoch": 1.330054991539763, + "grad_norm": 8.375, + "learning_rate": 4.304579546669075e-06, + "loss": 2.4276, + "step": 12577 + }, + { + "epoch": 1.330160744500846, + "grad_norm": 8.5625, + "learning_rate": 4.304461021395774e-06, + "loss": 2.4755, + "step": 12578 + }, + { + "epoch": 1.330266497461929, + "grad_norm": 8.1875, + "learning_rate": 4.3043424876548734e-06, + "loss": 2.5072, + "step": 12579 + }, + { + "epoch": 1.3303722504230118, + "grad_norm": 7.625, + "learning_rate": 4.304223945446933e-06, + "loss": 2.5161, + "step": 12580 + }, + { + "epoch": 1.3304780033840948, + "grad_norm": 8.125, + "learning_rate": 4.304105394772507e-06, + "loss": 2.5125, + "step": 12581 + }, + { + "epoch": 1.3305837563451777, + "grad_norm": 8.5, + "learning_rate": 4.303986835632151e-06, + "loss": 2.5999, + "step": 12582 + }, + { + "epoch": 1.3306895093062605, + "grad_norm": 8.0625, + "learning_rate": 4.303868268026424e-06, + "loss": 2.5418, + "step": 12583 + }, + { + "epoch": 1.3307952622673436, + "grad_norm": 8.0625, + "learning_rate": 4.303749691955881e-06, + "loss": 2.5175, + "step": 12584 + }, + { + "epoch": 1.3309010152284264, + "grad_norm": 7.96875, + "learning_rate": 4.303631107421078e-06, + "loss": 2.5909, + "step": 12585 + }, + { + "epoch": 1.3310067681895092, + "grad_norm": 8.75, + "learning_rate": 4.303512514422572e-06, + "loss": 2.5438, + "step": 12586 + }, + { + "epoch": 1.331112521150592, + "grad_norm": 8.3125, + "learning_rate": 4.303393912960919e-06, + "loss": 2.4454, + "step": 12587 + }, + { + "epoch": 1.3312182741116751, + "grad_norm": 8.375, + "learning_rate": 4.303275303036677e-06, + "loss": 2.6024, + "step": 12588 + }, + { + "epoch": 1.331324027072758, + "grad_norm": 8.0, + "learning_rate": 4.3031566846504e-06, + "loss": 2.426, + "step": 12589 + }, + { + "epoch": 1.331429780033841, + "grad_norm": 8.0625, + "learning_rate": 4.303038057802648e-06, + "loss": 2.5306, + "step": 12590 + }, + { + "epoch": 1.3315355329949239, + "grad_norm": 8.1875, + "learning_rate": 4.302919422493974e-06, + "loss": 2.5317, + "step": 12591 + }, + { + "epoch": 1.3316412859560067, + "grad_norm": 8.375, + "learning_rate": 4.302800778724938e-06, + "loss": 2.5734, + "step": 12592 + }, + { + "epoch": 1.3317470389170896, + "grad_norm": 8.75, + "learning_rate": 4.3026821264960946e-06, + "loss": 2.659, + "step": 12593 + }, + { + "epoch": 1.3318527918781726, + "grad_norm": 8.4375, + "learning_rate": 4.302563465808001e-06, + "loss": 2.4826, + "step": 12594 + }, + { + "epoch": 1.3319585448392555, + "grad_norm": 7.875, + "learning_rate": 4.302444796661214e-06, + "loss": 2.5005, + "step": 12595 + }, + { + "epoch": 1.3320642978003385, + "grad_norm": 8.125, + "learning_rate": 4.302326119056291e-06, + "loss": 2.6573, + "step": 12596 + }, + { + "epoch": 1.3321700507614214, + "grad_norm": 8.375, + "learning_rate": 4.302207432993789e-06, + "loss": 2.378, + "step": 12597 + }, + { + "epoch": 1.3322758037225042, + "grad_norm": 8.75, + "learning_rate": 4.302088738474265e-06, + "loss": 2.7045, + "step": 12598 + }, + { + "epoch": 1.332381556683587, + "grad_norm": 8.1875, + "learning_rate": 4.301970035498275e-06, + "loss": 2.5898, + "step": 12599 + }, + { + "epoch": 1.33248730964467, + "grad_norm": 8.0625, + "learning_rate": 4.301851324066376e-06, + "loss": 2.6751, + "step": 12600 + }, + { + "epoch": 1.332593062605753, + "grad_norm": 8.375, + "learning_rate": 4.301732604179127e-06, + "loss": 2.5328, + "step": 12601 + }, + { + "epoch": 1.332698815566836, + "grad_norm": 8.75, + "learning_rate": 4.301613875837083e-06, + "loss": 2.5905, + "step": 12602 + }, + { + "epoch": 1.3328045685279188, + "grad_norm": 8.0625, + "learning_rate": 4.301495139040802e-06, + "loss": 2.5363, + "step": 12603 + }, + { + "epoch": 1.3329103214890017, + "grad_norm": 8.125, + "learning_rate": 4.301376393790842e-06, + "loss": 2.3626, + "step": 12604 + }, + { + "epoch": 1.3330160744500845, + "grad_norm": 8.875, + "learning_rate": 4.301257640087758e-06, + "loss": 2.5876, + "step": 12605 + }, + { + "epoch": 1.3331218274111676, + "grad_norm": 9.0625, + "learning_rate": 4.301138877932109e-06, + "loss": 2.7042, + "step": 12606 + }, + { + "epoch": 1.3332275803722504, + "grad_norm": 8.375, + "learning_rate": 4.301020107324452e-06, + "loss": 2.5467, + "step": 12607 + }, + { + "epoch": 1.3333333333333333, + "grad_norm": 8.125, + "learning_rate": 4.300901328265344e-06, + "loss": 2.1629, + "step": 12608 + }, + { + "epoch": 1.3334390862944163, + "grad_norm": 8.25, + "learning_rate": 4.300782540755343e-06, + "loss": 2.5468, + "step": 12609 + }, + { + "epoch": 1.3335448392554992, + "grad_norm": 8.8125, + "learning_rate": 4.300663744795005e-06, + "loss": 2.6817, + "step": 12610 + }, + { + "epoch": 1.333650592216582, + "grad_norm": 8.0, + "learning_rate": 4.300544940384889e-06, + "loss": 2.485, + "step": 12611 + }, + { + "epoch": 1.3337563451776648, + "grad_norm": 8.75, + "learning_rate": 4.3004261275255526e-06, + "loss": 2.3059, + "step": 12612 + }, + { + "epoch": 1.333862098138748, + "grad_norm": 7.84375, + "learning_rate": 4.300307306217553e-06, + "loss": 2.5524, + "step": 12613 + }, + { + "epoch": 1.3339678510998307, + "grad_norm": 8.875, + "learning_rate": 4.300188476461447e-06, + "loss": 2.408, + "step": 12614 + }, + { + "epoch": 1.3340736040609138, + "grad_norm": 7.84375, + "learning_rate": 4.300069638257792e-06, + "loss": 2.5882, + "step": 12615 + }, + { + "epoch": 1.3341793570219966, + "grad_norm": 8.5, + "learning_rate": 4.299950791607147e-06, + "loss": 2.7345, + "step": 12616 + }, + { + "epoch": 1.3342851099830795, + "grad_norm": 8.875, + "learning_rate": 4.299831936510069e-06, + "loss": 2.6876, + "step": 12617 + }, + { + "epoch": 1.3343908629441623, + "grad_norm": 8.0625, + "learning_rate": 4.299713072967116e-06, + "loss": 2.5688, + "step": 12618 + }, + { + "epoch": 1.3344966159052454, + "grad_norm": 7.90625, + "learning_rate": 4.299594200978845e-06, + "loss": 2.4499, + "step": 12619 + }, + { + "epoch": 1.3346023688663282, + "grad_norm": 9.375, + "learning_rate": 4.299475320545816e-06, + "loss": 2.3528, + "step": 12620 + }, + { + "epoch": 1.3347081218274113, + "grad_norm": 7.625, + "learning_rate": 4.299356431668583e-06, + "loss": 2.5346, + "step": 12621 + }, + { + "epoch": 1.3348138747884941, + "grad_norm": 8.25, + "learning_rate": 4.299237534347708e-06, + "loss": 2.3234, + "step": 12622 + }, + { + "epoch": 1.334919627749577, + "grad_norm": 8.0, + "learning_rate": 4.2991186285837474e-06, + "loss": 2.4092, + "step": 12623 + }, + { + "epoch": 1.3350253807106598, + "grad_norm": 8.5, + "learning_rate": 4.298999714377258e-06, + "loss": 2.6422, + "step": 12624 + }, + { + "epoch": 1.3351311336717429, + "grad_norm": 8.5625, + "learning_rate": 4.298880791728798e-06, + "loss": 2.3637, + "step": 12625 + }, + { + "epoch": 1.3352368866328257, + "grad_norm": 8.4375, + "learning_rate": 4.298761860638927e-06, + "loss": 2.6081, + "step": 12626 + }, + { + "epoch": 1.3353426395939088, + "grad_norm": 8.5625, + "learning_rate": 4.2986429211082025e-06, + "loss": 2.6467, + "step": 12627 + }, + { + "epoch": 1.3354483925549916, + "grad_norm": 8.375, + "learning_rate": 4.298523973137182e-06, + "loss": 2.4205, + "step": 12628 + }, + { + "epoch": 1.3355541455160744, + "grad_norm": 8.0625, + "learning_rate": 4.298405016726424e-06, + "loss": 2.5058, + "step": 12629 + }, + { + "epoch": 1.3356598984771573, + "grad_norm": 8.1875, + "learning_rate": 4.298286051876487e-06, + "loss": 2.6392, + "step": 12630 + }, + { + "epoch": 1.3357656514382403, + "grad_norm": 8.125, + "learning_rate": 4.298167078587929e-06, + "loss": 2.6599, + "step": 12631 + }, + { + "epoch": 1.3358714043993232, + "grad_norm": 8.5625, + "learning_rate": 4.298048096861308e-06, + "loss": 2.7687, + "step": 12632 + }, + { + "epoch": 1.335977157360406, + "grad_norm": 7.875, + "learning_rate": 4.297929106697183e-06, + "loss": 2.5565, + "step": 12633 + }, + { + "epoch": 1.336082910321489, + "grad_norm": 8.125, + "learning_rate": 4.297810108096112e-06, + "loss": 2.595, + "step": 12634 + }, + { + "epoch": 1.336188663282572, + "grad_norm": 8.4375, + "learning_rate": 4.297691101058653e-06, + "loss": 2.5151, + "step": 12635 + }, + { + "epoch": 1.3362944162436547, + "grad_norm": 7.78125, + "learning_rate": 4.297572085585365e-06, + "loss": 2.4928, + "step": 12636 + }, + { + "epoch": 1.3364001692047378, + "grad_norm": 8.125, + "learning_rate": 4.297453061676807e-06, + "loss": 2.5827, + "step": 12637 + }, + { + "epoch": 1.3365059221658206, + "grad_norm": 8.0, + "learning_rate": 4.297334029333536e-06, + "loss": 2.3652, + "step": 12638 + }, + { + "epoch": 1.3366116751269035, + "grad_norm": 8.0625, + "learning_rate": 4.297214988556113e-06, + "loss": 2.4479, + "step": 12639 + }, + { + "epoch": 1.3367174280879865, + "grad_norm": 8.5625, + "learning_rate": 4.297095939345094e-06, + "loss": 2.6463, + "step": 12640 + }, + { + "epoch": 1.3368231810490694, + "grad_norm": 7.46875, + "learning_rate": 4.296976881701039e-06, + "loss": 2.4174, + "step": 12641 + }, + { + "epoch": 1.3369289340101522, + "grad_norm": 8.125, + "learning_rate": 4.296857815624506e-06, + "loss": 2.3215, + "step": 12642 + }, + { + "epoch": 1.337034686971235, + "grad_norm": 7.875, + "learning_rate": 4.296738741116054e-06, + "loss": 2.4453, + "step": 12643 + }, + { + "epoch": 1.3371404399323181, + "grad_norm": 8.3125, + "learning_rate": 4.2966196581762434e-06, + "loss": 2.5612, + "step": 12644 + }, + { + "epoch": 1.337246192893401, + "grad_norm": 7.96875, + "learning_rate": 4.29650056680563e-06, + "loss": 2.4455, + "step": 12645 + }, + { + "epoch": 1.337351945854484, + "grad_norm": 8.5625, + "learning_rate": 4.296381467004775e-06, + "loss": 2.4229, + "step": 12646 + }, + { + "epoch": 1.3374576988155669, + "grad_norm": 9.4375, + "learning_rate": 4.2962623587742354e-06, + "loss": 2.6053, + "step": 12647 + }, + { + "epoch": 1.3375634517766497, + "grad_norm": 8.25, + "learning_rate": 4.296143242114572e-06, + "loss": 2.6086, + "step": 12648 + }, + { + "epoch": 1.3376692047377325, + "grad_norm": 8.5625, + "learning_rate": 4.296024117026344e-06, + "loss": 2.4858, + "step": 12649 + }, + { + "epoch": 1.3377749576988156, + "grad_norm": 8.4375, + "learning_rate": 4.295904983510108e-06, + "loss": 2.5337, + "step": 12650 + }, + { + "epoch": 1.3378807106598984, + "grad_norm": 8.5625, + "learning_rate": 4.295785841566425e-06, + "loss": 2.5378, + "step": 12651 + }, + { + "epoch": 1.3379864636209815, + "grad_norm": 8.375, + "learning_rate": 4.295666691195853e-06, + "loss": 2.3862, + "step": 12652 + }, + { + "epoch": 1.3380922165820643, + "grad_norm": 8.6875, + "learning_rate": 4.295547532398952e-06, + "loss": 2.5684, + "step": 12653 + }, + { + "epoch": 1.3381979695431472, + "grad_norm": 8.125, + "learning_rate": 4.295428365176281e-06, + "loss": 2.5376, + "step": 12654 + }, + { + "epoch": 1.33830372250423, + "grad_norm": 8.8125, + "learning_rate": 4.295309189528398e-06, + "loss": 2.7963, + "step": 12655 + }, + { + "epoch": 1.338409475465313, + "grad_norm": 8.5, + "learning_rate": 4.295190005455864e-06, + "loss": 2.6471, + "step": 12656 + }, + { + "epoch": 1.338515228426396, + "grad_norm": 8.9375, + "learning_rate": 4.295070812959238e-06, + "loss": 2.5263, + "step": 12657 + }, + { + "epoch": 1.338620981387479, + "grad_norm": 8.625, + "learning_rate": 4.294951612039078e-06, + "loss": 2.8207, + "step": 12658 + }, + { + "epoch": 1.3387267343485618, + "grad_norm": 8.75, + "learning_rate": 4.294832402695945e-06, + "loss": 2.554, + "step": 12659 + }, + { + "epoch": 1.3388324873096447, + "grad_norm": 8.25, + "learning_rate": 4.2947131849303955e-06, + "loss": 2.7072, + "step": 12660 + }, + { + "epoch": 1.3389382402707275, + "grad_norm": 8.5625, + "learning_rate": 4.294593958742993e-06, + "loss": 2.8, + "step": 12661 + }, + { + "epoch": 1.3390439932318106, + "grad_norm": 8.75, + "learning_rate": 4.294474724134294e-06, + "loss": 2.5945, + "step": 12662 + }, + { + "epoch": 1.3391497461928934, + "grad_norm": 8.25, + "learning_rate": 4.294355481104861e-06, + "loss": 2.6004, + "step": 12663 + }, + { + "epoch": 1.3392554991539762, + "grad_norm": 8.3125, + "learning_rate": 4.29423622965525e-06, + "loss": 2.5209, + "step": 12664 + }, + { + "epoch": 1.3393612521150593, + "grad_norm": 8.375, + "learning_rate": 4.294116969786022e-06, + "loss": 2.6903, + "step": 12665 + }, + { + "epoch": 1.3394670050761421, + "grad_norm": 8.3125, + "learning_rate": 4.293997701497738e-06, + "loss": 2.5336, + "step": 12666 + }, + { + "epoch": 1.339572758037225, + "grad_norm": 8.0625, + "learning_rate": 4.2938784247909545e-06, + "loss": 2.4843, + "step": 12667 + }, + { + "epoch": 1.3396785109983078, + "grad_norm": 7.65625, + "learning_rate": 4.293759139666235e-06, + "loss": 2.3881, + "step": 12668 + }, + { + "epoch": 1.3397842639593909, + "grad_norm": 9.5625, + "learning_rate": 4.293639846124137e-06, + "loss": 2.3526, + "step": 12669 + }, + { + "epoch": 1.3398900169204737, + "grad_norm": 10.25, + "learning_rate": 4.29352054416522e-06, + "loss": 2.2314, + "step": 12670 + }, + { + "epoch": 1.3399957698815568, + "grad_norm": 8.4375, + "learning_rate": 4.293401233790045e-06, + "loss": 2.641, + "step": 12671 + }, + { + "epoch": 1.3401015228426396, + "grad_norm": 8.375, + "learning_rate": 4.293281914999172e-06, + "loss": 2.4991, + "step": 12672 + }, + { + "epoch": 1.3402072758037225, + "grad_norm": 7.8125, + "learning_rate": 4.2931625877931605e-06, + "loss": 2.5087, + "step": 12673 + }, + { + "epoch": 1.3403130287648053, + "grad_norm": 8.1875, + "learning_rate": 4.29304325217257e-06, + "loss": 2.4457, + "step": 12674 + }, + { + "epoch": 1.3404187817258884, + "grad_norm": 7.75, + "learning_rate": 4.292923908137961e-06, + "loss": 2.5106, + "step": 12675 + }, + { + "epoch": 1.3405245346869712, + "grad_norm": 9.5, + "learning_rate": 4.292804555689893e-06, + "loss": 2.7535, + "step": 12676 + }, + { + "epoch": 1.3406302876480543, + "grad_norm": 8.0625, + "learning_rate": 4.2926851948289275e-06, + "loss": 2.6226, + "step": 12677 + }, + { + "epoch": 1.340736040609137, + "grad_norm": 7.78125, + "learning_rate": 4.2925658255556225e-06, + "loss": 2.5958, + "step": 12678 + }, + { + "epoch": 1.34084179357022, + "grad_norm": 8.5625, + "learning_rate": 4.29244644787054e-06, + "loss": 2.5566, + "step": 12679 + }, + { + "epoch": 1.3409475465313028, + "grad_norm": 8.8125, + "learning_rate": 4.292327061774238e-06, + "loss": 2.6698, + "step": 12680 + }, + { + "epoch": 1.3410532994923858, + "grad_norm": 8.125, + "learning_rate": 4.292207667267279e-06, + "loss": 2.4878, + "step": 12681 + }, + { + "epoch": 1.3411590524534687, + "grad_norm": 8.125, + "learning_rate": 4.2920882643502224e-06, + "loss": 2.5043, + "step": 12682 + }, + { + "epoch": 1.3412648054145517, + "grad_norm": 7.9375, + "learning_rate": 4.291968853023629e-06, + "loss": 2.4176, + "step": 12683 + }, + { + "epoch": 1.3413705583756346, + "grad_norm": 7.9375, + "learning_rate": 4.2918494332880575e-06, + "loss": 2.4348, + "step": 12684 + }, + { + "epoch": 1.3414763113367174, + "grad_norm": 8.4375, + "learning_rate": 4.29173000514407e-06, + "loss": 2.6659, + "step": 12685 + }, + { + "epoch": 1.3415820642978002, + "grad_norm": 7.90625, + "learning_rate": 4.291610568592227e-06, + "loss": 2.4975, + "step": 12686 + }, + { + "epoch": 1.3416878172588833, + "grad_norm": 8.625, + "learning_rate": 4.291491123633088e-06, + "loss": 2.7105, + "step": 12687 + }, + { + "epoch": 1.3417935702199661, + "grad_norm": 9.6875, + "learning_rate": 4.2913716702672135e-06, + "loss": 2.795, + "step": 12688 + }, + { + "epoch": 1.341899323181049, + "grad_norm": 8.875, + "learning_rate": 4.291252208495165e-06, + "loss": 2.7148, + "step": 12689 + }, + { + "epoch": 1.342005076142132, + "grad_norm": 8.3125, + "learning_rate": 4.291132738317502e-06, + "loss": 2.685, + "step": 12690 + }, + { + "epoch": 1.3421108291032149, + "grad_norm": 8.25, + "learning_rate": 4.291013259734786e-06, + "loss": 2.5103, + "step": 12691 + }, + { + "epoch": 1.3422165820642977, + "grad_norm": 8.3125, + "learning_rate": 4.290893772747577e-06, + "loss": 2.7357, + "step": 12692 + }, + { + "epoch": 1.3423223350253808, + "grad_norm": 8.6875, + "learning_rate": 4.290774277356436e-06, + "loss": 2.478, + "step": 12693 + }, + { + "epoch": 1.3424280879864636, + "grad_norm": 8.1875, + "learning_rate": 4.290654773561924e-06, + "loss": 2.6713, + "step": 12694 + }, + { + "epoch": 1.3425338409475465, + "grad_norm": 7.84375, + "learning_rate": 4.290535261364601e-06, + "loss": 2.3498, + "step": 12695 + }, + { + "epoch": 1.3426395939086295, + "grad_norm": 8.375, + "learning_rate": 4.290415740765028e-06, + "loss": 2.6613, + "step": 12696 + }, + { + "epoch": 1.3427453468697124, + "grad_norm": 8.125, + "learning_rate": 4.290296211763767e-06, + "loss": 2.1394, + "step": 12697 + }, + { + "epoch": 1.3428510998307952, + "grad_norm": 8.875, + "learning_rate": 4.290176674361378e-06, + "loss": 2.6279, + "step": 12698 + }, + { + "epoch": 1.342956852791878, + "grad_norm": 8.5, + "learning_rate": 4.290057128558421e-06, + "loss": 2.6351, + "step": 12699 + }, + { + "epoch": 1.343062605752961, + "grad_norm": 8.25, + "learning_rate": 4.289937574355459e-06, + "loss": 2.7306, + "step": 12700 + }, + { + "epoch": 1.343168358714044, + "grad_norm": 8.0625, + "learning_rate": 4.289818011753052e-06, + "loss": 2.416, + "step": 12701 + }, + { + "epoch": 1.343274111675127, + "grad_norm": 8.625, + "learning_rate": 4.289698440751761e-06, + "loss": 2.6939, + "step": 12702 + }, + { + "epoch": 1.3433798646362098, + "grad_norm": 8.1875, + "learning_rate": 4.289578861352147e-06, + "loss": 2.5991, + "step": 12703 + }, + { + "epoch": 1.3434856175972927, + "grad_norm": 8.4375, + "learning_rate": 4.289459273554772e-06, + "loss": 2.487, + "step": 12704 + }, + { + "epoch": 1.3435913705583755, + "grad_norm": 8.4375, + "learning_rate": 4.289339677360196e-06, + "loss": 2.5405, + "step": 12705 + }, + { + "epoch": 1.3436971235194586, + "grad_norm": 8.3125, + "learning_rate": 4.28922007276898e-06, + "loss": 2.5836, + "step": 12706 + }, + { + "epoch": 1.3438028764805414, + "grad_norm": 8.6875, + "learning_rate": 4.289100459781687e-06, + "loss": 2.3574, + "step": 12707 + }, + { + "epoch": 1.3439086294416245, + "grad_norm": 7.75, + "learning_rate": 4.288980838398876e-06, + "loss": 2.4148, + "step": 12708 + }, + { + "epoch": 1.3440143824027073, + "grad_norm": 8.25, + "learning_rate": 4.288861208621111e-06, + "loss": 2.5233, + "step": 12709 + }, + { + "epoch": 1.3441201353637902, + "grad_norm": 8.25, + "learning_rate": 4.288741570448952e-06, + "loss": 2.5488, + "step": 12710 + }, + { + "epoch": 1.344225888324873, + "grad_norm": 8.375, + "learning_rate": 4.28862192388296e-06, + "loss": 2.615, + "step": 12711 + }, + { + "epoch": 1.344331641285956, + "grad_norm": 8.25, + "learning_rate": 4.288502268923696e-06, + "loss": 2.556, + "step": 12712 + }, + { + "epoch": 1.344437394247039, + "grad_norm": 8.0625, + "learning_rate": 4.2883826055717225e-06, + "loss": 2.3986, + "step": 12713 + }, + { + "epoch": 1.344543147208122, + "grad_norm": 8.25, + "learning_rate": 4.288262933827601e-06, + "loss": 2.6173, + "step": 12714 + }, + { + "epoch": 1.3446489001692048, + "grad_norm": 8.875, + "learning_rate": 4.288143253691894e-06, + "loss": 2.5048, + "step": 12715 + }, + { + "epoch": 1.3447546531302876, + "grad_norm": 8.75, + "learning_rate": 4.288023565165161e-06, + "loss": 2.6455, + "step": 12716 + }, + { + "epoch": 1.3448604060913705, + "grad_norm": 8.0625, + "learning_rate": 4.287903868247965e-06, + "loss": 2.5028, + "step": 12717 + }, + { + "epoch": 1.3449661590524535, + "grad_norm": 8.625, + "learning_rate": 4.287784162940868e-06, + "loss": 2.7706, + "step": 12718 + }, + { + "epoch": 1.3450719120135364, + "grad_norm": 8.0625, + "learning_rate": 4.28766444924443e-06, + "loss": 2.562, + "step": 12719 + }, + { + "epoch": 1.3451776649746192, + "grad_norm": 8.8125, + "learning_rate": 4.287544727159214e-06, + "loss": 2.501, + "step": 12720 + }, + { + "epoch": 1.3452834179357023, + "grad_norm": 8.125, + "learning_rate": 4.287424996685782e-06, + "loss": 2.6499, + "step": 12721 + }, + { + "epoch": 1.3453891708967851, + "grad_norm": 8.125, + "learning_rate": 4.287305257824695e-06, + "loss": 2.4991, + "step": 12722 + }, + { + "epoch": 1.345494923857868, + "grad_norm": 8.5, + "learning_rate": 4.287185510576517e-06, + "loss": 2.6008, + "step": 12723 + }, + { + "epoch": 1.345600676818951, + "grad_norm": 8.5625, + "learning_rate": 4.287065754941807e-06, + "loss": 2.6996, + "step": 12724 + }, + { + "epoch": 1.3457064297800339, + "grad_norm": 7.96875, + "learning_rate": 4.286945990921129e-06, + "loss": 2.4028, + "step": 12725 + }, + { + "epoch": 1.3458121827411167, + "grad_norm": 8.5, + "learning_rate": 4.286826218515044e-06, + "loss": 2.5551, + "step": 12726 + }, + { + "epoch": 1.3459179357021998, + "grad_norm": 8.125, + "learning_rate": 4.286706437724113e-06, + "loss": 2.4799, + "step": 12727 + }, + { + "epoch": 1.3460236886632826, + "grad_norm": 7.9375, + "learning_rate": 4.286586648548901e-06, + "loss": 2.5293, + "step": 12728 + }, + { + "epoch": 1.3461294416243654, + "grad_norm": 8.6875, + "learning_rate": 4.286466850989968e-06, + "loss": 2.6121, + "step": 12729 + }, + { + "epoch": 1.3462351945854483, + "grad_norm": 8.125, + "learning_rate": 4.286347045047877e-06, + "loss": 2.6469, + "step": 12730 + }, + { + "epoch": 1.3463409475465313, + "grad_norm": 8.3125, + "learning_rate": 4.28622723072319e-06, + "loss": 2.4203, + "step": 12731 + }, + { + "epoch": 1.3464467005076142, + "grad_norm": 8.75, + "learning_rate": 4.286107408016469e-06, + "loss": 2.6732, + "step": 12732 + }, + { + "epoch": 1.3465524534686972, + "grad_norm": 8.0, + "learning_rate": 4.285987576928276e-06, + "loss": 2.4137, + "step": 12733 + }, + { + "epoch": 1.34665820642978, + "grad_norm": 8.4375, + "learning_rate": 4.285867737459174e-06, + "loss": 2.6446, + "step": 12734 + }, + { + "epoch": 1.346763959390863, + "grad_norm": 7.84375, + "learning_rate": 4.285747889609725e-06, + "loss": 2.5534, + "step": 12735 + }, + { + "epoch": 1.3468697123519457, + "grad_norm": 8.0625, + "learning_rate": 4.285628033380492e-06, + "loss": 2.5371, + "step": 12736 + }, + { + "epoch": 1.3469754653130288, + "grad_norm": 8.1875, + "learning_rate": 4.2855081687720365e-06, + "loss": 2.6766, + "step": 12737 + }, + { + "epoch": 1.3470812182741116, + "grad_norm": 7.9375, + "learning_rate": 4.285388295784921e-06, + "loss": 2.5616, + "step": 12738 + }, + { + "epoch": 1.3471869712351947, + "grad_norm": 8.375, + "learning_rate": 4.28526841441971e-06, + "loss": 2.7285, + "step": 12739 + }, + { + "epoch": 1.3472927241962775, + "grad_norm": 8.125, + "learning_rate": 4.285148524676963e-06, + "loss": 2.6426, + "step": 12740 + }, + { + "epoch": 1.3473984771573604, + "grad_norm": 8.375, + "learning_rate": 4.285028626557245e-06, + "loss": 2.624, + "step": 12741 + }, + { + "epoch": 1.3475042301184432, + "grad_norm": 7.8125, + "learning_rate": 4.2849087200611165e-06, + "loss": 2.4028, + "step": 12742 + }, + { + "epoch": 1.3476099830795263, + "grad_norm": 8.125, + "learning_rate": 4.2847888051891424e-06, + "loss": 2.62, + "step": 12743 + }, + { + "epoch": 1.3477157360406091, + "grad_norm": 8.375, + "learning_rate": 4.2846688819418845e-06, + "loss": 2.5739, + "step": 12744 + }, + { + "epoch": 1.3478214890016922, + "grad_norm": 7.90625, + "learning_rate": 4.284548950319904e-06, + "loss": 2.5593, + "step": 12745 + }, + { + "epoch": 1.347927241962775, + "grad_norm": 8.375, + "learning_rate": 4.2844290103237665e-06, + "loss": 2.4097, + "step": 12746 + }, + { + "epoch": 1.3480329949238579, + "grad_norm": 8.125, + "learning_rate": 4.284309061954033e-06, + "loss": 2.6425, + "step": 12747 + }, + { + "epoch": 1.3481387478849407, + "grad_norm": 9.1875, + "learning_rate": 4.284189105211267e-06, + "loss": 2.6742, + "step": 12748 + }, + { + "epoch": 1.3482445008460238, + "grad_norm": 8.0, + "learning_rate": 4.284069140096032e-06, + "loss": 2.6738, + "step": 12749 + }, + { + "epoch": 1.3483502538071066, + "grad_norm": 8.375, + "learning_rate": 4.283949166608889e-06, + "loss": 2.5423, + "step": 12750 + }, + { + "epoch": 1.3484560067681894, + "grad_norm": 8.0625, + "learning_rate": 4.283829184750402e-06, + "loss": 2.4985, + "step": 12751 + }, + { + "epoch": 1.3485617597292725, + "grad_norm": 8.3125, + "learning_rate": 4.283709194521135e-06, + "loss": 2.5299, + "step": 12752 + }, + { + "epoch": 1.3486675126903553, + "grad_norm": 8.375, + "learning_rate": 4.2835891959216504e-06, + "loss": 2.5855, + "step": 12753 + }, + { + "epoch": 1.3487732656514382, + "grad_norm": 8.125, + "learning_rate": 4.283469188952511e-06, + "loss": 2.4927, + "step": 12754 + }, + { + "epoch": 1.348879018612521, + "grad_norm": 8.875, + "learning_rate": 4.28334917361428e-06, + "loss": 2.3411, + "step": 12755 + }, + { + "epoch": 1.348984771573604, + "grad_norm": 8.75, + "learning_rate": 4.283229149907519e-06, + "loss": 2.7634, + "step": 12756 + }, + { + "epoch": 1.349090524534687, + "grad_norm": 7.96875, + "learning_rate": 4.283109117832796e-06, + "loss": 2.5983, + "step": 12757 + }, + { + "epoch": 1.34919627749577, + "grad_norm": 8.25, + "learning_rate": 4.282989077390669e-06, + "loss": 2.5352, + "step": 12758 + }, + { + "epoch": 1.3493020304568528, + "grad_norm": 8.375, + "learning_rate": 4.282869028581704e-06, + "loss": 2.564, + "step": 12759 + }, + { + "epoch": 1.3494077834179357, + "grad_norm": 8.0625, + "learning_rate": 4.282748971406464e-06, + "loss": 2.3849, + "step": 12760 + }, + { + "epoch": 1.3495135363790185, + "grad_norm": 8.625, + "learning_rate": 4.282628905865512e-06, + "loss": 2.2851, + "step": 12761 + }, + { + "epoch": 1.3496192893401016, + "grad_norm": 7.5, + "learning_rate": 4.282508831959412e-06, + "loss": 2.3983, + "step": 12762 + }, + { + "epoch": 1.3497250423011844, + "grad_norm": 8.3125, + "learning_rate": 4.282388749688727e-06, + "loss": 2.6231, + "step": 12763 + }, + { + "epoch": 1.3498307952622675, + "grad_norm": 8.0625, + "learning_rate": 4.282268659054021e-06, + "loss": 2.3223, + "step": 12764 + }, + { + "epoch": 1.3499365482233503, + "grad_norm": 7.84375, + "learning_rate": 4.282148560055857e-06, + "loss": 2.4257, + "step": 12765 + }, + { + "epoch": 1.3500423011844331, + "grad_norm": 8.625, + "learning_rate": 4.2820284526947975e-06, + "loss": 2.64, + "step": 12766 + }, + { + "epoch": 1.350148054145516, + "grad_norm": 8.3125, + "learning_rate": 4.281908336971407e-06, + "loss": 2.6346, + "step": 12767 + }, + { + "epoch": 1.350253807106599, + "grad_norm": 8.0625, + "learning_rate": 4.281788212886252e-06, + "loss": 2.5843, + "step": 12768 + }, + { + "epoch": 1.3503595600676819, + "grad_norm": 8.4375, + "learning_rate": 4.281668080439892e-06, + "loss": 2.3755, + "step": 12769 + }, + { + "epoch": 1.350465313028765, + "grad_norm": 8.25, + "learning_rate": 4.281547939632892e-06, + "loss": 2.5769, + "step": 12770 + }, + { + "epoch": 1.3505710659898478, + "grad_norm": 8.4375, + "learning_rate": 4.281427790465816e-06, + "loss": 2.5718, + "step": 12771 + }, + { + "epoch": 1.3506768189509306, + "grad_norm": 8.3125, + "learning_rate": 4.281307632939229e-06, + "loss": 2.5456, + "step": 12772 + }, + { + "epoch": 1.3507825719120135, + "grad_norm": 8.5625, + "learning_rate": 4.2811874670536936e-06, + "loss": 2.5588, + "step": 12773 + }, + { + "epoch": 1.3508883248730965, + "grad_norm": 8.75, + "learning_rate": 4.2810672928097736e-06, + "loss": 2.8222, + "step": 12774 + }, + { + "epoch": 1.3509940778341794, + "grad_norm": 8.3125, + "learning_rate": 4.280947110208034e-06, + "loss": 2.4343, + "step": 12775 + }, + { + "epoch": 1.3510998307952622, + "grad_norm": 10.0, + "learning_rate": 4.280826919249037e-06, + "loss": 2.5958, + "step": 12776 + }, + { + "epoch": 1.3512055837563453, + "grad_norm": 8.9375, + "learning_rate": 4.280706719933347e-06, + "loss": 2.6935, + "step": 12777 + }, + { + "epoch": 1.351311336717428, + "grad_norm": 8.6875, + "learning_rate": 4.280586512261529e-06, + "loss": 2.7273, + "step": 12778 + }, + { + "epoch": 1.351417089678511, + "grad_norm": 8.375, + "learning_rate": 4.280466296234148e-06, + "loss": 2.3563, + "step": 12779 + }, + { + "epoch": 1.351522842639594, + "grad_norm": 8.125, + "learning_rate": 4.2803460718517664e-06, + "loss": 2.646, + "step": 12780 + }, + { + "epoch": 1.3516285956006768, + "grad_norm": 7.875, + "learning_rate": 4.280225839114949e-06, + "loss": 2.635, + "step": 12781 + }, + { + "epoch": 1.3517343485617597, + "grad_norm": 8.625, + "learning_rate": 4.280105598024259e-06, + "loss": 2.593, + "step": 12782 + }, + { + "epoch": 1.3518401015228427, + "grad_norm": 8.9375, + "learning_rate": 4.279985348580262e-06, + "loss": 2.7093, + "step": 12783 + }, + { + "epoch": 1.3519458544839256, + "grad_norm": 8.125, + "learning_rate": 4.279865090783521e-06, + "loss": 2.5566, + "step": 12784 + }, + { + "epoch": 1.3520516074450084, + "grad_norm": 7.5625, + "learning_rate": 4.279744824634602e-06, + "loss": 2.3439, + "step": 12785 + }, + { + "epoch": 1.3521573604060912, + "grad_norm": 9.0, + "learning_rate": 4.279624550134068e-06, + "loss": 2.6377, + "step": 12786 + }, + { + "epoch": 1.3522631133671743, + "grad_norm": 9.3125, + "learning_rate": 4.279504267282484e-06, + "loss": 2.9979, + "step": 12787 + }, + { + "epoch": 1.3523688663282571, + "grad_norm": 8.1875, + "learning_rate": 4.279383976080414e-06, + "loss": 2.4241, + "step": 12788 + }, + { + "epoch": 1.3524746192893402, + "grad_norm": 8.375, + "learning_rate": 4.279263676528424e-06, + "loss": 2.5757, + "step": 12789 + }, + { + "epoch": 1.352580372250423, + "grad_norm": 8.1875, + "learning_rate": 4.2791433686270754e-06, + "loss": 2.5016, + "step": 12790 + }, + { + "epoch": 1.3526861252115059, + "grad_norm": 8.125, + "learning_rate": 4.279023052376936e-06, + "loss": 2.4824, + "step": 12791 + }, + { + "epoch": 1.3527918781725887, + "grad_norm": 8.0, + "learning_rate": 4.278902727778567e-06, + "loss": 2.2074, + "step": 12792 + }, + { + "epoch": 1.3528976311336718, + "grad_norm": 8.625, + "learning_rate": 4.278782394832538e-06, + "loss": 2.5029, + "step": 12793 + }, + { + "epoch": 1.3530033840947546, + "grad_norm": 7.90625, + "learning_rate": 4.278662053539409e-06, + "loss": 2.4384, + "step": 12794 + }, + { + "epoch": 1.3531091370558377, + "grad_norm": 8.375, + "learning_rate": 4.278541703899747e-06, + "loss": 2.4705, + "step": 12795 + }, + { + "epoch": 1.3532148900169205, + "grad_norm": 9.125, + "learning_rate": 4.278421345914116e-06, + "loss": 2.4753, + "step": 12796 + }, + { + "epoch": 1.3533206429780034, + "grad_norm": 8.9375, + "learning_rate": 4.27830097958308e-06, + "loss": 2.3962, + "step": 12797 + }, + { + "epoch": 1.3534263959390862, + "grad_norm": 8.25, + "learning_rate": 4.278180604907206e-06, + "loss": 2.5454, + "step": 12798 + }, + { + "epoch": 1.3535321489001693, + "grad_norm": 8.125, + "learning_rate": 4.278060221887057e-06, + "loss": 2.603, + "step": 12799 + }, + { + "epoch": 1.353637901861252, + "grad_norm": 8.3125, + "learning_rate": 4.277939830523199e-06, + "loss": 2.5626, + "step": 12800 + }, + { + "epoch": 1.3537436548223352, + "grad_norm": 8.25, + "learning_rate": 4.277819430816196e-06, + "loss": 2.7996, + "step": 12801 + }, + { + "epoch": 1.353849407783418, + "grad_norm": 9.1875, + "learning_rate": 4.277699022766614e-06, + "loss": 2.6642, + "step": 12802 + }, + { + "epoch": 1.3539551607445008, + "grad_norm": 7.59375, + "learning_rate": 4.277578606375017e-06, + "loss": 2.4741, + "step": 12803 + }, + { + "epoch": 1.3540609137055837, + "grad_norm": 9.125, + "learning_rate": 4.277458181641971e-06, + "loss": 2.7845, + "step": 12804 + }, + { + "epoch": 1.3541666666666667, + "grad_norm": 8.375, + "learning_rate": 4.277337748568041e-06, + "loss": 2.441, + "step": 12805 + }, + { + "epoch": 1.3542724196277496, + "grad_norm": 8.5625, + "learning_rate": 4.277217307153791e-06, + "loss": 2.5036, + "step": 12806 + }, + { + "epoch": 1.3543781725888324, + "grad_norm": 8.25, + "learning_rate": 4.2770968573997886e-06, + "loss": 2.6293, + "step": 12807 + }, + { + "epoch": 1.3544839255499155, + "grad_norm": 8.375, + "learning_rate": 4.276976399306596e-06, + "loss": 2.5015, + "step": 12808 + }, + { + "epoch": 1.3545896785109983, + "grad_norm": 7.84375, + "learning_rate": 4.27685593287478e-06, + "loss": 2.6057, + "step": 12809 + }, + { + "epoch": 1.3546954314720812, + "grad_norm": 8.4375, + "learning_rate": 4.276735458104907e-06, + "loss": 2.5073, + "step": 12810 + }, + { + "epoch": 1.354801184433164, + "grad_norm": 8.6875, + "learning_rate": 4.276614974997541e-06, + "loss": 2.5424, + "step": 12811 + }, + { + "epoch": 1.354906937394247, + "grad_norm": 8.375, + "learning_rate": 4.276494483553246e-06, + "loss": 2.5963, + "step": 12812 + }, + { + "epoch": 1.35501269035533, + "grad_norm": 7.875, + "learning_rate": 4.27637398377259e-06, + "loss": 2.5865, + "step": 12813 + }, + { + "epoch": 1.355118443316413, + "grad_norm": 8.0, + "learning_rate": 4.276253475656137e-06, + "loss": 2.4549, + "step": 12814 + }, + { + "epoch": 1.3552241962774958, + "grad_norm": 8.6875, + "learning_rate": 4.276132959204453e-06, + "loss": 2.4319, + "step": 12815 + }, + { + "epoch": 1.3553299492385786, + "grad_norm": 8.375, + "learning_rate": 4.276012434418104e-06, + "loss": 2.5532, + "step": 12816 + }, + { + "epoch": 1.3554357021996615, + "grad_norm": 7.96875, + "learning_rate": 4.275891901297654e-06, + "loss": 2.5477, + "step": 12817 + }, + { + "epoch": 1.3555414551607445, + "grad_norm": 8.25, + "learning_rate": 4.275771359843671e-06, + "loss": 2.7275, + "step": 12818 + }, + { + "epoch": 1.3556472081218274, + "grad_norm": 7.65625, + "learning_rate": 4.275650810056719e-06, + "loss": 2.5064, + "step": 12819 + }, + { + "epoch": 1.3557529610829104, + "grad_norm": 8.1875, + "learning_rate": 4.275530251937363e-06, + "loss": 2.5466, + "step": 12820 + }, + { + "epoch": 1.3558587140439933, + "grad_norm": 7.71875, + "learning_rate": 4.27540968548617e-06, + "loss": 2.5098, + "step": 12821 + }, + { + "epoch": 1.3559644670050761, + "grad_norm": 8.125, + "learning_rate": 4.275289110703706e-06, + "loss": 2.5345, + "step": 12822 + }, + { + "epoch": 1.356070219966159, + "grad_norm": 8.25, + "learning_rate": 4.275168527590535e-06, + "loss": 2.7642, + "step": 12823 + }, + { + "epoch": 1.356175972927242, + "grad_norm": 7.5625, + "learning_rate": 4.275047936147225e-06, + "loss": 2.6013, + "step": 12824 + }, + { + "epoch": 1.3562817258883249, + "grad_norm": 8.25, + "learning_rate": 4.2749273363743415e-06, + "loss": 2.6769, + "step": 12825 + }, + { + "epoch": 1.356387478849408, + "grad_norm": 7.96875, + "learning_rate": 4.274806728272449e-06, + "loss": 2.4666, + "step": 12826 + }, + { + "epoch": 1.3564932318104908, + "grad_norm": 8.125, + "learning_rate": 4.274686111842115e-06, + "loss": 2.2682, + "step": 12827 + }, + { + "epoch": 1.3565989847715736, + "grad_norm": 8.0, + "learning_rate": 4.274565487083905e-06, + "loss": 2.4802, + "step": 12828 + }, + { + "epoch": 1.3567047377326564, + "grad_norm": 8.125, + "learning_rate": 4.274444853998384e-06, + "loss": 2.3379, + "step": 12829 + }, + { + "epoch": 1.3568104906937395, + "grad_norm": 8.25, + "learning_rate": 4.274324212586121e-06, + "loss": 2.5304, + "step": 12830 + }, + { + "epoch": 1.3569162436548223, + "grad_norm": 7.96875, + "learning_rate": 4.274203562847679e-06, + "loss": 2.4924, + "step": 12831 + }, + { + "epoch": 1.3570219966159052, + "grad_norm": 8.1875, + "learning_rate": 4.274082904783624e-06, + "loss": 2.4959, + "step": 12832 + }, + { + "epoch": 1.3571277495769882, + "grad_norm": 8.25, + "learning_rate": 4.273962238394525e-06, + "loss": 2.6025, + "step": 12833 + }, + { + "epoch": 1.357233502538071, + "grad_norm": 8.0625, + "learning_rate": 4.2738415636809465e-06, + "loss": 2.5409, + "step": 12834 + }, + { + "epoch": 1.357339255499154, + "grad_norm": 8.3125, + "learning_rate": 4.273720880643455e-06, + "loss": 2.6375, + "step": 12835 + }, + { + "epoch": 1.357445008460237, + "grad_norm": 8.625, + "learning_rate": 4.2736001892826166e-06, + "loss": 2.6719, + "step": 12836 + }, + { + "epoch": 1.3575507614213198, + "grad_norm": 8.3125, + "learning_rate": 4.273479489598997e-06, + "loss": 2.5291, + "step": 12837 + }, + { + "epoch": 1.3576565143824026, + "grad_norm": 8.5625, + "learning_rate": 4.273358781593165e-06, + "loss": 2.669, + "step": 12838 + }, + { + "epoch": 1.3577622673434857, + "grad_norm": 8.1875, + "learning_rate": 4.273238065265686e-06, + "loss": 2.5934, + "step": 12839 + }, + { + "epoch": 1.3578680203045685, + "grad_norm": 8.125, + "learning_rate": 4.273117340617124e-06, + "loss": 2.5678, + "step": 12840 + }, + { + "epoch": 1.3579737732656514, + "grad_norm": 8.25, + "learning_rate": 4.272996607648049e-06, + "loss": 2.5489, + "step": 12841 + }, + { + "epoch": 1.3580795262267342, + "grad_norm": 8.1875, + "learning_rate": 4.272875866359026e-06, + "loss": 2.5055, + "step": 12842 + }, + { + "epoch": 1.3581852791878173, + "grad_norm": 7.9375, + "learning_rate": 4.272755116750621e-06, + "loss": 2.6027, + "step": 12843 + }, + { + "epoch": 1.3582910321489001, + "grad_norm": 8.1875, + "learning_rate": 4.2726343588234015e-06, + "loss": 2.6931, + "step": 12844 + }, + { + "epoch": 1.3583967851099832, + "grad_norm": 8.5625, + "learning_rate": 4.2725135925779335e-06, + "loss": 2.4018, + "step": 12845 + }, + { + "epoch": 1.358502538071066, + "grad_norm": 8.125, + "learning_rate": 4.272392818014785e-06, + "loss": 2.4663, + "step": 12846 + }, + { + "epoch": 1.3586082910321489, + "grad_norm": 8.3125, + "learning_rate": 4.272272035134522e-06, + "loss": 2.5428, + "step": 12847 + }, + { + "epoch": 1.3587140439932317, + "grad_norm": 8.1875, + "learning_rate": 4.27215124393771e-06, + "loss": 2.4989, + "step": 12848 + }, + { + "epoch": 1.3588197969543148, + "grad_norm": 8.375, + "learning_rate": 4.272030444424918e-06, + "loss": 2.6731, + "step": 12849 + }, + { + "epoch": 1.3589255499153976, + "grad_norm": 8.3125, + "learning_rate": 4.271909636596711e-06, + "loss": 2.4871, + "step": 12850 + }, + { + "epoch": 1.3590313028764807, + "grad_norm": 8.5625, + "learning_rate": 4.271788820453657e-06, + "loss": 2.6733, + "step": 12851 + }, + { + "epoch": 1.3591370558375635, + "grad_norm": 8.0625, + "learning_rate": 4.2716679959963235e-06, + "loss": 2.367, + "step": 12852 + }, + { + "epoch": 1.3592428087986463, + "grad_norm": 8.4375, + "learning_rate": 4.271547163225276e-06, + "loss": 2.5948, + "step": 12853 + }, + { + "epoch": 1.3593485617597292, + "grad_norm": 7.9375, + "learning_rate": 4.2714263221410825e-06, + "loss": 2.5815, + "step": 12854 + }, + { + "epoch": 1.3594543147208122, + "grad_norm": 8.1875, + "learning_rate": 4.2713054727443095e-06, + "loss": 2.6565, + "step": 12855 + }, + { + "epoch": 1.359560067681895, + "grad_norm": 7.875, + "learning_rate": 4.271184615035524e-06, + "loss": 2.5812, + "step": 12856 + }, + { + "epoch": 1.3596658206429781, + "grad_norm": 8.375, + "learning_rate": 4.271063749015294e-06, + "loss": 2.7032, + "step": 12857 + }, + { + "epoch": 1.359771573604061, + "grad_norm": 8.125, + "learning_rate": 4.270942874684186e-06, + "loss": 2.5373, + "step": 12858 + }, + { + "epoch": 1.3598773265651438, + "grad_norm": 8.375, + "learning_rate": 4.2708219920427675e-06, + "loss": 2.8381, + "step": 12859 + }, + { + "epoch": 1.3599830795262267, + "grad_norm": 7.78125, + "learning_rate": 4.270701101091606e-06, + "loss": 2.448, + "step": 12860 + }, + { + "epoch": 1.3600888324873097, + "grad_norm": 8.125, + "learning_rate": 4.270580201831268e-06, + "loss": 2.3505, + "step": 12861 + }, + { + "epoch": 1.3601945854483926, + "grad_norm": 8.25, + "learning_rate": 4.270459294262321e-06, + "loss": 2.6468, + "step": 12862 + }, + { + "epoch": 1.3603003384094754, + "grad_norm": 8.6875, + "learning_rate": 4.270338378385334e-06, + "loss": 2.7126, + "step": 12863 + }, + { + "epoch": 1.3604060913705585, + "grad_norm": 8.375, + "learning_rate": 4.2702174542008715e-06, + "loss": 2.5236, + "step": 12864 + }, + { + "epoch": 1.3605118443316413, + "grad_norm": 7.71875, + "learning_rate": 4.270096521709503e-06, + "loss": 2.5562, + "step": 12865 + }, + { + "epoch": 1.3606175972927241, + "grad_norm": 8.125, + "learning_rate": 4.269975580911795e-06, + "loss": 2.588, + "step": 12866 + }, + { + "epoch": 1.360723350253807, + "grad_norm": 8.4375, + "learning_rate": 4.269854631808316e-06, + "loss": 2.5645, + "step": 12867 + }, + { + "epoch": 1.36082910321489, + "grad_norm": 8.875, + "learning_rate": 4.2697336743996335e-06, + "loss": 2.5012, + "step": 12868 + }, + { + "epoch": 1.3609348561759729, + "grad_norm": 7.90625, + "learning_rate": 4.269612708686314e-06, + "loss": 2.5032, + "step": 12869 + }, + { + "epoch": 1.361040609137056, + "grad_norm": 8.4375, + "learning_rate": 4.269491734668926e-06, + "loss": 2.3829, + "step": 12870 + }, + { + "epoch": 1.3611463620981388, + "grad_norm": 8.0, + "learning_rate": 4.2693707523480365e-06, + "loss": 2.3108, + "step": 12871 + }, + { + "epoch": 1.3612521150592216, + "grad_norm": 7.96875, + "learning_rate": 4.269249761724214e-06, + "loss": 2.5755, + "step": 12872 + }, + { + "epoch": 1.3613578680203045, + "grad_norm": 8.3125, + "learning_rate": 4.269128762798026e-06, + "loss": 2.5803, + "step": 12873 + }, + { + "epoch": 1.3614636209813875, + "grad_norm": 8.3125, + "learning_rate": 4.26900775557004e-06, + "loss": 2.5933, + "step": 12874 + }, + { + "epoch": 1.3615693739424704, + "grad_norm": 8.75, + "learning_rate": 4.268886740040825e-06, + "loss": 2.795, + "step": 12875 + }, + { + "epoch": 1.3616751269035534, + "grad_norm": 8.125, + "learning_rate": 4.268765716210948e-06, + "loss": 2.5561, + "step": 12876 + }, + { + "epoch": 1.3617808798646363, + "grad_norm": 8.75, + "learning_rate": 4.2686446840809755e-06, + "loss": 2.6673, + "step": 12877 + }, + { + "epoch": 1.361886632825719, + "grad_norm": 7.8125, + "learning_rate": 4.268523643651478e-06, + "loss": 2.4308, + "step": 12878 + }, + { + "epoch": 1.361992385786802, + "grad_norm": 8.5, + "learning_rate": 4.268402594923022e-06, + "loss": 2.4857, + "step": 12879 + }, + { + "epoch": 1.362098138747885, + "grad_norm": 10.1875, + "learning_rate": 4.268281537896176e-06, + "loss": 2.2575, + "step": 12880 + }, + { + "epoch": 1.3622038917089678, + "grad_norm": 8.125, + "learning_rate": 4.268160472571507e-06, + "loss": 2.579, + "step": 12881 + }, + { + "epoch": 1.362309644670051, + "grad_norm": 8.1875, + "learning_rate": 4.268039398949586e-06, + "loss": 2.6203, + "step": 12882 + }, + { + "epoch": 1.3624153976311337, + "grad_norm": 7.96875, + "learning_rate": 4.267918317030978e-06, + "loss": 2.5387, + "step": 12883 + }, + { + "epoch": 1.3625211505922166, + "grad_norm": 8.125, + "learning_rate": 4.2677972268162525e-06, + "loss": 2.5404, + "step": 12884 + }, + { + "epoch": 1.3626269035532994, + "grad_norm": 8.0625, + "learning_rate": 4.2676761283059776e-06, + "loss": 2.539, + "step": 12885 + }, + { + "epoch": 1.3627326565143825, + "grad_norm": 8.125, + "learning_rate": 4.267555021500722e-06, + "loss": 2.4351, + "step": 12886 + }, + { + "epoch": 1.3628384094754653, + "grad_norm": 8.375, + "learning_rate": 4.267433906401053e-06, + "loss": 2.5932, + "step": 12887 + }, + { + "epoch": 1.3629441624365481, + "grad_norm": 8.4375, + "learning_rate": 4.267312783007539e-06, + "loss": 2.4209, + "step": 12888 + }, + { + "epoch": 1.3630499153976312, + "grad_norm": 8.6875, + "learning_rate": 4.26719165132075e-06, + "loss": 2.7879, + "step": 12889 + }, + { + "epoch": 1.363155668358714, + "grad_norm": 7.96875, + "learning_rate": 4.267070511341253e-06, + "loss": 2.6164, + "step": 12890 + }, + { + "epoch": 1.3632614213197969, + "grad_norm": 8.25, + "learning_rate": 4.266949363069617e-06, + "loss": 2.6133, + "step": 12891 + }, + { + "epoch": 1.36336717428088, + "grad_norm": 8.0, + "learning_rate": 4.26682820650641e-06, + "loss": 2.5915, + "step": 12892 + }, + { + "epoch": 1.3634729272419628, + "grad_norm": 8.9375, + "learning_rate": 4.266707041652201e-06, + "loss": 2.3215, + "step": 12893 + }, + { + "epoch": 1.3635786802030456, + "grad_norm": 8.8125, + "learning_rate": 4.266585868507558e-06, + "loss": 2.7722, + "step": 12894 + }, + { + "epoch": 1.3636844331641287, + "grad_norm": 8.3125, + "learning_rate": 4.266464687073051e-06, + "loss": 2.4367, + "step": 12895 + }, + { + "epoch": 1.3637901861252115, + "grad_norm": 8.0, + "learning_rate": 4.266343497349246e-06, + "loss": 2.5506, + "step": 12896 + }, + { + "epoch": 1.3638959390862944, + "grad_norm": 8.0625, + "learning_rate": 4.266222299336715e-06, + "loss": 2.4813, + "step": 12897 + }, + { + "epoch": 1.3640016920473772, + "grad_norm": 8.8125, + "learning_rate": 4.2661010930360245e-06, + "loss": 2.6877, + "step": 12898 + }, + { + "epoch": 1.3641074450084603, + "grad_norm": 8.3125, + "learning_rate": 4.265979878447744e-06, + "loss": 2.4841, + "step": 12899 + }, + { + "epoch": 1.364213197969543, + "grad_norm": 8.8125, + "learning_rate": 4.2658586555724415e-06, + "loss": 2.5715, + "step": 12900 + }, + { + "epoch": 1.3643189509306262, + "grad_norm": 8.25, + "learning_rate": 4.265737424410687e-06, + "loss": 2.6988, + "step": 12901 + }, + { + "epoch": 1.364424703891709, + "grad_norm": 8.4375, + "learning_rate": 4.26561618496305e-06, + "loss": 2.4767, + "step": 12902 + }, + { + "epoch": 1.3645304568527918, + "grad_norm": 9.5, + "learning_rate": 4.265494937230097e-06, + "loss": 2.6069, + "step": 12903 + }, + { + "epoch": 1.3646362098138747, + "grad_norm": 8.125, + "learning_rate": 4.265373681212399e-06, + "loss": 2.3908, + "step": 12904 + }, + { + "epoch": 1.3647419627749577, + "grad_norm": 8.0, + "learning_rate": 4.265252416910524e-06, + "loss": 2.4868, + "step": 12905 + }, + { + "epoch": 1.3648477157360406, + "grad_norm": 8.1875, + "learning_rate": 4.265131144325041e-06, + "loss": 2.3571, + "step": 12906 + }, + { + "epoch": 1.3649534686971236, + "grad_norm": 8.5, + "learning_rate": 4.26500986345652e-06, + "loss": 2.7183, + "step": 12907 + }, + { + "epoch": 1.3650592216582065, + "grad_norm": 7.9375, + "learning_rate": 4.26488857430553e-06, + "loss": 2.5277, + "step": 12908 + }, + { + "epoch": 1.3651649746192893, + "grad_norm": 8.4375, + "learning_rate": 4.264767276872639e-06, + "loss": 2.7075, + "step": 12909 + }, + { + "epoch": 1.3652707275803722, + "grad_norm": 8.0625, + "learning_rate": 4.264645971158416e-06, + "loss": 2.503, + "step": 12910 + }, + { + "epoch": 1.3653764805414552, + "grad_norm": 8.375, + "learning_rate": 4.264524657163434e-06, + "loss": 2.7625, + "step": 12911 + }, + { + "epoch": 1.365482233502538, + "grad_norm": 8.1875, + "learning_rate": 4.264403334888257e-06, + "loss": 2.6131, + "step": 12912 + }, + { + "epoch": 1.3655879864636211, + "grad_norm": 8.5, + "learning_rate": 4.264282004333458e-06, + "loss": 2.7509, + "step": 12913 + }, + { + "epoch": 1.365693739424704, + "grad_norm": 8.25, + "learning_rate": 4.2641606654996034e-06, + "loss": 2.7001, + "step": 12914 + }, + { + "epoch": 1.3657994923857868, + "grad_norm": 7.8125, + "learning_rate": 4.264039318387267e-06, + "loss": 2.3294, + "step": 12915 + }, + { + "epoch": 1.3659052453468696, + "grad_norm": 8.625, + "learning_rate": 4.263917962997013e-06, + "loss": 2.3426, + "step": 12916 + }, + { + "epoch": 1.3660109983079527, + "grad_norm": 8.25, + "learning_rate": 4.263796599329415e-06, + "loss": 2.4941, + "step": 12917 + }, + { + "epoch": 1.3661167512690355, + "grad_norm": 8.25, + "learning_rate": 4.26367522738504e-06, + "loss": 2.5905, + "step": 12918 + }, + { + "epoch": 1.3662225042301184, + "grad_norm": 9.0, + "learning_rate": 4.263553847164459e-06, + "loss": 2.371, + "step": 12919 + }, + { + "epoch": 1.3663282571912014, + "grad_norm": 8.125, + "learning_rate": 4.2634324586682416e-06, + "loss": 2.6922, + "step": 12920 + }, + { + "epoch": 1.3664340101522843, + "grad_norm": 7.8125, + "learning_rate": 4.263311061896956e-06, + "loss": 2.508, + "step": 12921 + }, + { + "epoch": 1.3665397631133671, + "grad_norm": 8.5625, + "learning_rate": 4.263189656851173e-06, + "loss": 2.6183, + "step": 12922 + }, + { + "epoch": 1.36664551607445, + "grad_norm": 7.96875, + "learning_rate": 4.2630682435314615e-06, + "loss": 2.4686, + "step": 12923 + }, + { + "epoch": 1.366751269035533, + "grad_norm": 7.625, + "learning_rate": 4.262946821938393e-06, + "loss": 2.4732, + "step": 12924 + }, + { + "epoch": 1.3668570219966159, + "grad_norm": 8.6875, + "learning_rate": 4.2628253920725355e-06, + "loss": 2.8499, + "step": 12925 + }, + { + "epoch": 1.366962774957699, + "grad_norm": 8.3125, + "learning_rate": 4.2627039539344596e-06, + "loss": 2.6458, + "step": 12926 + }, + { + "epoch": 1.3670685279187818, + "grad_norm": 7.9375, + "learning_rate": 4.262582507524734e-06, + "loss": 2.3829, + "step": 12927 + }, + { + "epoch": 1.3671742808798646, + "grad_norm": 8.25, + "learning_rate": 4.262461052843931e-06, + "loss": 2.4346, + "step": 12928 + }, + { + "epoch": 1.3672800338409474, + "grad_norm": 7.96875, + "learning_rate": 4.262339589892618e-06, + "loss": 2.5285, + "step": 12929 + }, + { + "epoch": 1.3673857868020305, + "grad_norm": 8.4375, + "learning_rate": 4.262218118671367e-06, + "loss": 2.551, + "step": 12930 + }, + { + "epoch": 1.3674915397631133, + "grad_norm": 8.375, + "learning_rate": 4.262096639180746e-06, + "loss": 2.6044, + "step": 12931 + }, + { + "epoch": 1.3675972927241964, + "grad_norm": 8.1875, + "learning_rate": 4.261975151421327e-06, + "loss": 2.6157, + "step": 12932 + }, + { + "epoch": 1.3677030456852792, + "grad_norm": 7.84375, + "learning_rate": 4.261853655393679e-06, + "loss": 2.4319, + "step": 12933 + }, + { + "epoch": 1.367808798646362, + "grad_norm": 8.1875, + "learning_rate": 4.261732151098372e-06, + "loss": 2.662, + "step": 12934 + }, + { + "epoch": 1.367914551607445, + "grad_norm": 8.375, + "learning_rate": 4.261610638535978e-06, + "loss": 2.3383, + "step": 12935 + }, + { + "epoch": 1.368020304568528, + "grad_norm": 8.25, + "learning_rate": 4.261489117707064e-06, + "loss": 2.6406, + "step": 12936 + }, + { + "epoch": 1.3681260575296108, + "grad_norm": 7.9375, + "learning_rate": 4.2613675886122025e-06, + "loss": 2.2953, + "step": 12937 + }, + { + "epoch": 1.3682318104906939, + "grad_norm": 8.8125, + "learning_rate": 4.261246051251964e-06, + "loss": 2.8009, + "step": 12938 + }, + { + "epoch": 1.3683375634517767, + "grad_norm": 8.0, + "learning_rate": 4.261124505626917e-06, + "loss": 2.4227, + "step": 12939 + }, + { + "epoch": 1.3684433164128595, + "grad_norm": 7.96875, + "learning_rate": 4.261002951737633e-06, + "loss": 2.6758, + "step": 12940 + }, + { + "epoch": 1.3685490693739424, + "grad_norm": 8.125, + "learning_rate": 4.260881389584683e-06, + "loss": 2.3737, + "step": 12941 + }, + { + "epoch": 1.3686548223350254, + "grad_norm": 9.0, + "learning_rate": 4.260759819168637e-06, + "loss": 2.8003, + "step": 12942 + }, + { + "epoch": 1.3687605752961083, + "grad_norm": 8.375, + "learning_rate": 4.260638240490065e-06, + "loss": 2.8277, + "step": 12943 + }, + { + "epoch": 1.3688663282571911, + "grad_norm": 7.90625, + "learning_rate": 4.260516653549538e-06, + "loss": 2.544, + "step": 12944 + }, + { + "epoch": 1.3689720812182742, + "grad_norm": 8.375, + "learning_rate": 4.260395058347626e-06, + "loss": 2.6305, + "step": 12945 + }, + { + "epoch": 1.369077834179357, + "grad_norm": 7.9375, + "learning_rate": 4.2602734548849e-06, + "loss": 2.3968, + "step": 12946 + }, + { + "epoch": 1.3691835871404399, + "grad_norm": 8.75, + "learning_rate": 4.26015184316193e-06, + "loss": 2.8636, + "step": 12947 + }, + { + "epoch": 1.369289340101523, + "grad_norm": 7.84375, + "learning_rate": 4.260030223179289e-06, + "loss": 2.4935, + "step": 12948 + }, + { + "epoch": 1.3693950930626058, + "grad_norm": 8.8125, + "learning_rate": 4.2599085949375445e-06, + "loss": 2.6962, + "step": 12949 + }, + { + "epoch": 1.3695008460236886, + "grad_norm": 8.0625, + "learning_rate": 4.2597869584372695e-06, + "loss": 2.6315, + "step": 12950 + }, + { + "epoch": 1.3696065989847717, + "grad_norm": 8.4375, + "learning_rate": 4.259665313679033e-06, + "loss": 2.733, + "step": 12951 + }, + { + "epoch": 1.3697123519458545, + "grad_norm": 8.1875, + "learning_rate": 4.259543660663408e-06, + "loss": 2.308, + "step": 12952 + }, + { + "epoch": 1.3698181049069373, + "grad_norm": 8.3125, + "learning_rate": 4.259421999390963e-06, + "loss": 2.7065, + "step": 12953 + }, + { + "epoch": 1.3699238578680202, + "grad_norm": 8.6875, + "learning_rate": 4.259300329862272e-06, + "loss": 2.3156, + "step": 12954 + }, + { + "epoch": 1.3700296108291032, + "grad_norm": 8.4375, + "learning_rate": 4.259178652077903e-06, + "loss": 2.495, + "step": 12955 + }, + { + "epoch": 1.370135363790186, + "grad_norm": 8.5, + "learning_rate": 4.259056966038429e-06, + "loss": 2.5537, + "step": 12956 + }, + { + "epoch": 1.3702411167512691, + "grad_norm": 8.3125, + "learning_rate": 4.258935271744419e-06, + "loss": 2.501, + "step": 12957 + }, + { + "epoch": 1.370346869712352, + "grad_norm": 8.5625, + "learning_rate": 4.258813569196445e-06, + "loss": 2.6107, + "step": 12958 + }, + { + "epoch": 1.3704526226734348, + "grad_norm": 8.1875, + "learning_rate": 4.2586918583950785e-06, + "loss": 2.5283, + "step": 12959 + }, + { + "epoch": 1.3705583756345177, + "grad_norm": 8.375, + "learning_rate": 4.2585701393408905e-06, + "loss": 2.4197, + "step": 12960 + }, + { + "epoch": 1.3706641285956007, + "grad_norm": 8.5, + "learning_rate": 4.258448412034452e-06, + "loss": 2.7313, + "step": 12961 + }, + { + "epoch": 1.3707698815566836, + "grad_norm": 9.1875, + "learning_rate": 4.258326676476334e-06, + "loss": 2.9388, + "step": 12962 + }, + { + "epoch": 1.3708756345177666, + "grad_norm": 8.125, + "learning_rate": 4.258204932667109e-06, + "loss": 2.5496, + "step": 12963 + }, + { + "epoch": 1.3709813874788495, + "grad_norm": 8.25, + "learning_rate": 4.258083180607347e-06, + "loss": 2.5114, + "step": 12964 + }, + { + "epoch": 1.3710871404399323, + "grad_norm": 8.0625, + "learning_rate": 4.257961420297619e-06, + "loss": 2.4429, + "step": 12965 + }, + { + "epoch": 1.3711928934010151, + "grad_norm": 8.0625, + "learning_rate": 4.257839651738498e-06, + "loss": 2.5258, + "step": 12966 + }, + { + "epoch": 1.3712986463620982, + "grad_norm": 8.0625, + "learning_rate": 4.257717874930553e-06, + "loss": 2.5701, + "step": 12967 + }, + { + "epoch": 1.371404399323181, + "grad_norm": 8.4375, + "learning_rate": 4.257596089874358e-06, + "loss": 2.5574, + "step": 12968 + }, + { + "epoch": 1.371510152284264, + "grad_norm": 8.0625, + "learning_rate": 4.257474296570483e-06, + "loss": 2.5319, + "step": 12969 + }, + { + "epoch": 1.371615905245347, + "grad_norm": 7.78125, + "learning_rate": 4.2573524950195e-06, + "loss": 2.4111, + "step": 12970 + }, + { + "epoch": 1.3717216582064298, + "grad_norm": 8.25, + "learning_rate": 4.257230685221981e-06, + "loss": 2.3806, + "step": 12971 + }, + { + "epoch": 1.3718274111675126, + "grad_norm": 7.9375, + "learning_rate": 4.2571088671784965e-06, + "loss": 2.5933, + "step": 12972 + }, + { + "epoch": 1.3719331641285957, + "grad_norm": 8.75, + "learning_rate": 4.256987040889619e-06, + "loss": 2.4707, + "step": 12973 + }, + { + "epoch": 1.3720389170896785, + "grad_norm": 8.125, + "learning_rate": 4.25686520635592e-06, + "loss": 2.5132, + "step": 12974 + }, + { + "epoch": 1.3721446700507614, + "grad_norm": 7.96875, + "learning_rate": 4.25674336357797e-06, + "loss": 2.645, + "step": 12975 + }, + { + "epoch": 1.3722504230118444, + "grad_norm": 7.96875, + "learning_rate": 4.256621512556342e-06, + "loss": 2.461, + "step": 12976 + }, + { + "epoch": 1.3723561759729273, + "grad_norm": 8.25, + "learning_rate": 4.256499653291609e-06, + "loss": 2.6148, + "step": 12977 + }, + { + "epoch": 1.37246192893401, + "grad_norm": 7.96875, + "learning_rate": 4.256377785784341e-06, + "loss": 2.5194, + "step": 12978 + }, + { + "epoch": 1.372567681895093, + "grad_norm": 7.78125, + "learning_rate": 4.256255910035111e-06, + "loss": 2.4166, + "step": 12979 + }, + { + "epoch": 1.372673434856176, + "grad_norm": 7.78125, + "learning_rate": 4.25613402604449e-06, + "loss": 2.4072, + "step": 12980 + }, + { + "epoch": 1.3727791878172588, + "grad_norm": 10.6875, + "learning_rate": 4.2560121338130486e-06, + "loss": 2.7108, + "step": 12981 + }, + { + "epoch": 1.372884940778342, + "grad_norm": 8.375, + "learning_rate": 4.255890233341362e-06, + "loss": 2.4741, + "step": 12982 + }, + { + "epoch": 1.3729906937394247, + "grad_norm": 8.4375, + "learning_rate": 4.255768324630001e-06, + "loss": 2.7087, + "step": 12983 + }, + { + "epoch": 1.3730964467005076, + "grad_norm": 8.125, + "learning_rate": 4.255646407679536e-06, + "loss": 2.4894, + "step": 12984 + }, + { + "epoch": 1.3732021996615904, + "grad_norm": 8.5, + "learning_rate": 4.255524482490541e-06, + "loss": 2.256, + "step": 12985 + }, + { + "epoch": 1.3733079526226735, + "grad_norm": 8.5, + "learning_rate": 4.255402549063588e-06, + "loss": 2.7254, + "step": 12986 + }, + { + "epoch": 1.3734137055837563, + "grad_norm": 7.90625, + "learning_rate": 4.255280607399248e-06, + "loss": 2.5323, + "step": 12987 + }, + { + "epoch": 1.3735194585448394, + "grad_norm": 8.9375, + "learning_rate": 4.255158657498095e-06, + "loss": 2.895, + "step": 12988 + }, + { + "epoch": 1.3736252115059222, + "grad_norm": 8.5, + "learning_rate": 4.2550366993607e-06, + "loss": 2.5393, + "step": 12989 + }, + { + "epoch": 1.373730964467005, + "grad_norm": 8.25, + "learning_rate": 4.254914732987635e-06, + "loss": 2.5703, + "step": 12990 + }, + { + "epoch": 1.3738367174280879, + "grad_norm": 8.25, + "learning_rate": 4.254792758379474e-06, + "loss": 2.5278, + "step": 12991 + }, + { + "epoch": 1.373942470389171, + "grad_norm": 9.0, + "learning_rate": 4.2546707755367865e-06, + "loss": 2.5294, + "step": 12992 + }, + { + "epoch": 1.3740482233502538, + "grad_norm": 8.1875, + "learning_rate": 4.254548784460148e-06, + "loss": 2.4985, + "step": 12993 + }, + { + "epoch": 1.3741539763113368, + "grad_norm": 7.9375, + "learning_rate": 4.254426785150129e-06, + "loss": 2.428, + "step": 12994 + }, + { + "epoch": 1.3742597292724197, + "grad_norm": 7.875, + "learning_rate": 4.254304777607303e-06, + "loss": 2.3428, + "step": 12995 + }, + { + "epoch": 1.3743654822335025, + "grad_norm": 8.5625, + "learning_rate": 4.254182761832242e-06, + "loss": 2.7115, + "step": 12996 + }, + { + "epoch": 1.3744712351945854, + "grad_norm": 8.6875, + "learning_rate": 4.25406073782552e-06, + "loss": 2.5253, + "step": 12997 + }, + { + "epoch": 1.3745769881556684, + "grad_norm": 8.25, + "learning_rate": 4.253938705587706e-06, + "loss": 2.4061, + "step": 12998 + }, + { + "epoch": 1.3746827411167513, + "grad_norm": 7.59375, + "learning_rate": 4.253816665119377e-06, + "loss": 2.3623, + "step": 12999 + }, + { + "epoch": 1.374788494077834, + "grad_norm": 8.5625, + "learning_rate": 4.253694616421102e-06, + "loss": 2.69, + "step": 13000 + }, + { + "epoch": 1.3748942470389172, + "grad_norm": 8.625, + "learning_rate": 4.253572559493457e-06, + "loss": 2.3119, + "step": 13001 + }, + { + "epoch": 1.375, + "grad_norm": 8.875, + "learning_rate": 4.253450494337012e-06, + "loss": 2.5663, + "step": 13002 + }, + { + "epoch": 1.3751057529610828, + "grad_norm": 7.5, + "learning_rate": 4.253328420952341e-06, + "loss": 2.3662, + "step": 13003 + }, + { + "epoch": 1.375211505922166, + "grad_norm": 8.125, + "learning_rate": 4.253206339340017e-06, + "loss": 2.6284, + "step": 13004 + }, + { + "epoch": 1.3753172588832487, + "grad_norm": 8.3125, + "learning_rate": 4.253084249500613e-06, + "loss": 2.7452, + "step": 13005 + }, + { + "epoch": 1.3754230118443316, + "grad_norm": 8.5625, + "learning_rate": 4.252962151434702e-06, + "loss": 2.7183, + "step": 13006 + }, + { + "epoch": 1.3755287648054146, + "grad_norm": 8.3125, + "learning_rate": 4.252840045142856e-06, + "loss": 2.6326, + "step": 13007 + }, + { + "epoch": 1.3756345177664975, + "grad_norm": 8.3125, + "learning_rate": 4.252717930625649e-06, + "loss": 2.601, + "step": 13008 + }, + { + "epoch": 1.3757402707275803, + "grad_norm": 8.0625, + "learning_rate": 4.252595807883653e-06, + "loss": 2.7339, + "step": 13009 + }, + { + "epoch": 1.3758460236886632, + "grad_norm": 7.78125, + "learning_rate": 4.252473676917441e-06, + "loss": 2.527, + "step": 13010 + }, + { + "epoch": 1.3759517766497462, + "grad_norm": 8.75, + "learning_rate": 4.252351537727588e-06, + "loss": 2.0127, + "step": 13011 + }, + { + "epoch": 1.376057529610829, + "grad_norm": 9.0625, + "learning_rate": 4.252229390314666e-06, + "loss": 2.6196, + "step": 13012 + }, + { + "epoch": 1.3761632825719121, + "grad_norm": 8.375, + "learning_rate": 4.252107234679248e-06, + "loss": 2.7015, + "step": 13013 + }, + { + "epoch": 1.376269035532995, + "grad_norm": 8.0, + "learning_rate": 4.251985070821906e-06, + "loss": 2.4348, + "step": 13014 + }, + { + "epoch": 1.3763747884940778, + "grad_norm": 8.0625, + "learning_rate": 4.251862898743216e-06, + "loss": 2.6213, + "step": 13015 + }, + { + "epoch": 1.3764805414551606, + "grad_norm": 8.625, + "learning_rate": 4.2517407184437495e-06, + "loss": 2.8116, + "step": 13016 + }, + { + "epoch": 1.3765862944162437, + "grad_norm": 8.8125, + "learning_rate": 4.251618529924081e-06, + "loss": 2.5864, + "step": 13017 + }, + { + "epoch": 1.3766920473773265, + "grad_norm": 8.1875, + "learning_rate": 4.251496333184781e-06, + "loss": 2.1777, + "step": 13018 + }, + { + "epoch": 1.3767978003384096, + "grad_norm": 8.125, + "learning_rate": 4.251374128226427e-06, + "loss": 2.5432, + "step": 13019 + }, + { + "epoch": 1.3769035532994924, + "grad_norm": 8.6875, + "learning_rate": 4.251251915049589e-06, + "loss": 2.4653, + "step": 13020 + }, + { + "epoch": 1.3770093062605753, + "grad_norm": 7.9375, + "learning_rate": 4.2511296936548444e-06, + "loss": 2.4649, + "step": 13021 + }, + { + "epoch": 1.377115059221658, + "grad_norm": 7.78125, + "learning_rate": 4.251007464042762e-06, + "loss": 2.53, + "step": 13022 + }, + { + "epoch": 1.3772208121827412, + "grad_norm": 8.25, + "learning_rate": 4.250885226213919e-06, + "loss": 2.7915, + "step": 13023 + }, + { + "epoch": 1.377326565143824, + "grad_norm": 8.0, + "learning_rate": 4.250762980168887e-06, + "loss": 2.4696, + "step": 13024 + }, + { + "epoch": 1.377432318104907, + "grad_norm": 8.375, + "learning_rate": 4.25064072590824e-06, + "loss": 2.4658, + "step": 13025 + }, + { + "epoch": 1.37753807106599, + "grad_norm": 8.1875, + "learning_rate": 4.2505184634325534e-06, + "loss": 2.4426, + "step": 13026 + }, + { + "epoch": 1.3776438240270727, + "grad_norm": 8.125, + "learning_rate": 4.250396192742399e-06, + "loss": 2.4496, + "step": 13027 + }, + { + "epoch": 1.3777495769881556, + "grad_norm": 8.125, + "learning_rate": 4.250273913838351e-06, + "loss": 2.6462, + "step": 13028 + }, + { + "epoch": 1.3778553299492386, + "grad_norm": 8.625, + "learning_rate": 4.250151626720983e-06, + "loss": 2.6315, + "step": 13029 + }, + { + "epoch": 1.3779610829103215, + "grad_norm": 8.125, + "learning_rate": 4.25002933139087e-06, + "loss": 2.6152, + "step": 13030 + }, + { + "epoch": 1.3780668358714043, + "grad_norm": 8.125, + "learning_rate": 4.249907027848585e-06, + "loss": 2.2794, + "step": 13031 + }, + { + "epoch": 1.3781725888324874, + "grad_norm": 8.3125, + "learning_rate": 4.249784716094701e-06, + "loss": 2.5604, + "step": 13032 + }, + { + "epoch": 1.3782783417935702, + "grad_norm": 8.375, + "learning_rate": 4.249662396129794e-06, + "loss": 2.4546, + "step": 13033 + }, + { + "epoch": 1.378384094754653, + "grad_norm": 8.0625, + "learning_rate": 4.249540067954436e-06, + "loss": 2.6292, + "step": 13034 + }, + { + "epoch": 1.378489847715736, + "grad_norm": 8.0, + "learning_rate": 4.249417731569203e-06, + "loss": 2.6648, + "step": 13035 + }, + { + "epoch": 1.378595600676819, + "grad_norm": 8.3125, + "learning_rate": 4.249295386974668e-06, + "loss": 2.4586, + "step": 13036 + }, + { + "epoch": 1.3787013536379018, + "grad_norm": 8.625, + "learning_rate": 4.249173034171404e-06, + "loss": 2.4208, + "step": 13037 + }, + { + "epoch": 1.3788071065989849, + "grad_norm": 8.375, + "learning_rate": 4.249050673159988e-06, + "loss": 2.6425, + "step": 13038 + }, + { + "epoch": 1.3789128595600677, + "grad_norm": 9.0625, + "learning_rate": 4.248928303940991e-06, + "loss": 2.6443, + "step": 13039 + }, + { + "epoch": 1.3790186125211505, + "grad_norm": 8.25, + "learning_rate": 4.2488059265149905e-06, + "loss": 2.4299, + "step": 13040 + }, + { + "epoch": 1.3791243654822334, + "grad_norm": 8.625, + "learning_rate": 4.248683540882558e-06, + "loss": 2.7039, + "step": 13041 + }, + { + "epoch": 1.3792301184433164, + "grad_norm": 9.0625, + "learning_rate": 4.2485611470442685e-06, + "loss": 2.6769, + "step": 13042 + }, + { + "epoch": 1.3793358714043993, + "grad_norm": 8.125, + "learning_rate": 4.248438745000697e-06, + "loss": 2.6173, + "step": 13043 + }, + { + "epoch": 1.3794416243654823, + "grad_norm": 8.125, + "learning_rate": 4.248316334752418e-06, + "loss": 2.6339, + "step": 13044 + }, + { + "epoch": 1.3795473773265652, + "grad_norm": 8.3125, + "learning_rate": 4.248193916300005e-06, + "loss": 2.4332, + "step": 13045 + }, + { + "epoch": 1.379653130287648, + "grad_norm": 8.5, + "learning_rate": 4.2480714896440325e-06, + "loss": 2.6563, + "step": 13046 + }, + { + "epoch": 1.3797588832487309, + "grad_norm": 8.1875, + "learning_rate": 4.247949054785076e-06, + "loss": 2.2506, + "step": 13047 + }, + { + "epoch": 1.379864636209814, + "grad_norm": 8.1875, + "learning_rate": 4.24782661172371e-06, + "loss": 2.6845, + "step": 13048 + }, + { + "epoch": 1.3799703891708968, + "grad_norm": 7.65625, + "learning_rate": 4.247704160460507e-06, + "loss": 2.3756, + "step": 13049 + }, + { + "epoch": 1.3800761421319798, + "grad_norm": 8.625, + "learning_rate": 4.2475817009960446e-06, + "loss": 2.6329, + "step": 13050 + }, + { + "epoch": 1.3801818950930627, + "grad_norm": 7.96875, + "learning_rate": 4.247459233330896e-06, + "loss": 2.3975, + "step": 13051 + }, + { + "epoch": 1.3802876480541455, + "grad_norm": 8.1875, + "learning_rate": 4.247336757465635e-06, + "loss": 2.6988, + "step": 13052 + }, + { + "epoch": 1.3803934010152283, + "grad_norm": 8.4375, + "learning_rate": 4.247214273400838e-06, + "loss": 2.5625, + "step": 13053 + }, + { + "epoch": 1.3804991539763114, + "grad_norm": 9.0625, + "learning_rate": 4.247091781137079e-06, + "loss": 2.3455, + "step": 13054 + }, + { + "epoch": 1.3806049069373942, + "grad_norm": 8.5, + "learning_rate": 4.246969280674932e-06, + "loss": 2.4434, + "step": 13055 + }, + { + "epoch": 1.380710659898477, + "grad_norm": 7.96875, + "learning_rate": 4.246846772014973e-06, + "loss": 2.5255, + "step": 13056 + }, + { + "epoch": 1.3808164128595601, + "grad_norm": 11.6875, + "learning_rate": 4.246724255157777e-06, + "loss": 2.9686, + "step": 13057 + }, + { + "epoch": 1.380922165820643, + "grad_norm": 9.0, + "learning_rate": 4.246601730103918e-06, + "loss": 2.4521, + "step": 13058 + }, + { + "epoch": 1.3810279187817258, + "grad_norm": 8.1875, + "learning_rate": 4.246479196853971e-06, + "loss": 2.708, + "step": 13059 + }, + { + "epoch": 1.3811336717428089, + "grad_norm": 8.25, + "learning_rate": 4.246356655408513e-06, + "loss": 2.3096, + "step": 13060 + }, + { + "epoch": 1.3812394247038917, + "grad_norm": 7.9375, + "learning_rate": 4.246234105768115e-06, + "loss": 2.504, + "step": 13061 + }, + { + "epoch": 1.3813451776649746, + "grad_norm": 8.4375, + "learning_rate": 4.246111547933357e-06, + "loss": 2.6434, + "step": 13062 + }, + { + "epoch": 1.3814509306260576, + "grad_norm": 7.84375, + "learning_rate": 4.24598898190481e-06, + "loss": 2.532, + "step": 13063 + }, + { + "epoch": 1.3815566835871405, + "grad_norm": 8.9375, + "learning_rate": 4.245866407683052e-06, + "loss": 2.6742, + "step": 13064 + }, + { + "epoch": 1.3816624365482233, + "grad_norm": 8.125, + "learning_rate": 4.245743825268656e-06, + "loss": 2.3591, + "step": 13065 + }, + { + "epoch": 1.3817681895093061, + "grad_norm": 9.0, + "learning_rate": 4.245621234662199e-06, + "loss": 2.4691, + "step": 13066 + }, + { + "epoch": 1.3818739424703892, + "grad_norm": 8.0625, + "learning_rate": 4.2454986358642545e-06, + "loss": 2.4193, + "step": 13067 + }, + { + "epoch": 1.381979695431472, + "grad_norm": 7.8125, + "learning_rate": 4.2453760288754e-06, + "loss": 2.4847, + "step": 13068 + }, + { + "epoch": 1.382085448392555, + "grad_norm": 9.1875, + "learning_rate": 4.245253413696209e-06, + "loss": 2.5379, + "step": 13069 + }, + { + "epoch": 1.382191201353638, + "grad_norm": 7.90625, + "learning_rate": 4.2451307903272574e-06, + "loss": 2.3642, + "step": 13070 + }, + { + "epoch": 1.3822969543147208, + "grad_norm": 8.5625, + "learning_rate": 4.2450081587691205e-06, + "loss": 2.6099, + "step": 13071 + }, + { + "epoch": 1.3824027072758036, + "grad_norm": 8.5625, + "learning_rate": 4.244885519022374e-06, + "loss": 2.5358, + "step": 13072 + }, + { + "epoch": 1.3825084602368867, + "grad_norm": 8.5, + "learning_rate": 4.2447628710875935e-06, + "loss": 2.7113, + "step": 13073 + }, + { + "epoch": 1.3826142131979695, + "grad_norm": 8.25, + "learning_rate": 4.2446402149653545e-06, + "loss": 2.6001, + "step": 13074 + }, + { + "epoch": 1.3827199661590526, + "grad_norm": 8.125, + "learning_rate": 4.244517550656233e-06, + "loss": 2.6787, + "step": 13075 + }, + { + "epoch": 1.3828257191201354, + "grad_norm": 8.4375, + "learning_rate": 4.244394878160804e-06, + "loss": 2.6319, + "step": 13076 + }, + { + "epoch": 1.3829314720812182, + "grad_norm": 8.25, + "learning_rate": 4.2442721974796434e-06, + "loss": 2.4749, + "step": 13077 + }, + { + "epoch": 1.383037225042301, + "grad_norm": 7.84375, + "learning_rate": 4.244149508613326e-06, + "loss": 2.5217, + "step": 13078 + }, + { + "epoch": 1.3831429780033841, + "grad_norm": 8.8125, + "learning_rate": 4.2440268115624285e-06, + "loss": 2.4834, + "step": 13079 + }, + { + "epoch": 1.383248730964467, + "grad_norm": 8.375, + "learning_rate": 4.243904106327527e-06, + "loss": 2.6152, + "step": 13080 + }, + { + "epoch": 1.38335448392555, + "grad_norm": 8.6875, + "learning_rate": 4.243781392909196e-06, + "loss": 2.6719, + "step": 13081 + }, + { + "epoch": 1.3834602368866329, + "grad_norm": 8.375, + "learning_rate": 4.243658671308013e-06, + "loss": 2.3037, + "step": 13082 + }, + { + "epoch": 1.3835659898477157, + "grad_norm": 8.375, + "learning_rate": 4.243535941524552e-06, + "loss": 2.5121, + "step": 13083 + }, + { + "epoch": 1.3836717428087986, + "grad_norm": 8.625, + "learning_rate": 4.24341320355939e-06, + "loss": 2.6852, + "step": 13084 + }, + { + "epoch": 1.3837774957698816, + "grad_norm": 7.78125, + "learning_rate": 4.243290457413104e-06, + "loss": 2.3632, + "step": 13085 + }, + { + "epoch": 1.3838832487309645, + "grad_norm": 8.8125, + "learning_rate": 4.243167703086268e-06, + "loss": 2.4186, + "step": 13086 + }, + { + "epoch": 1.3839890016920473, + "grad_norm": 9.1875, + "learning_rate": 4.243044940579459e-06, + "loss": 2.8422, + "step": 13087 + }, + { + "epoch": 1.3840947546531304, + "grad_norm": 8.0625, + "learning_rate": 4.242922169893253e-06, + "loss": 2.4805, + "step": 13088 + }, + { + "epoch": 1.3842005076142132, + "grad_norm": 9.8125, + "learning_rate": 4.242799391028226e-06, + "loss": 2.5485, + "step": 13089 + }, + { + "epoch": 1.384306260575296, + "grad_norm": 8.3125, + "learning_rate": 4.2426766039849545e-06, + "loss": 2.3657, + "step": 13090 + }, + { + "epoch": 1.3844120135363789, + "grad_norm": 7.78125, + "learning_rate": 4.2425538087640135e-06, + "loss": 2.4738, + "step": 13091 + }, + { + "epoch": 1.384517766497462, + "grad_norm": 8.1875, + "learning_rate": 4.242431005365981e-06, + "loss": 2.4433, + "step": 13092 + }, + { + "epoch": 1.3846235194585448, + "grad_norm": 8.0625, + "learning_rate": 4.242308193791433e-06, + "loss": 2.5521, + "step": 13093 + }, + { + "epoch": 1.3847292724196278, + "grad_norm": 8.4375, + "learning_rate": 4.242185374040945e-06, + "loss": 2.1523, + "step": 13094 + }, + { + "epoch": 1.3848350253807107, + "grad_norm": 8.75, + "learning_rate": 4.242062546115093e-06, + "loss": 2.5552, + "step": 13095 + }, + { + "epoch": 1.3849407783417935, + "grad_norm": 8.0625, + "learning_rate": 4.241939710014453e-06, + "loss": 2.4868, + "step": 13096 + }, + { + "epoch": 1.3850465313028764, + "grad_norm": 8.4375, + "learning_rate": 4.241816865739604e-06, + "loss": 2.5246, + "step": 13097 + }, + { + "epoch": 1.3851522842639594, + "grad_norm": 8.5, + "learning_rate": 4.241694013291121e-06, + "loss": 2.6266, + "step": 13098 + }, + { + "epoch": 1.3852580372250423, + "grad_norm": 8.3125, + "learning_rate": 4.241571152669579e-06, + "loss": 2.6549, + "step": 13099 + }, + { + "epoch": 1.3853637901861253, + "grad_norm": 7.96875, + "learning_rate": 4.2414482838755564e-06, + "loss": 2.5542, + "step": 13100 + }, + { + "epoch": 1.3854695431472082, + "grad_norm": 8.3125, + "learning_rate": 4.241325406909629e-06, + "loss": 2.5302, + "step": 13101 + }, + { + "epoch": 1.385575296108291, + "grad_norm": 7.875, + "learning_rate": 4.241202521772374e-06, + "loss": 2.5529, + "step": 13102 + }, + { + "epoch": 1.3856810490693738, + "grad_norm": 7.875, + "learning_rate": 4.241079628464368e-06, + "loss": 2.6656, + "step": 13103 + }, + { + "epoch": 1.385786802030457, + "grad_norm": 7.78125, + "learning_rate": 4.2409567269861866e-06, + "loss": 2.4726, + "step": 13104 + }, + { + "epoch": 1.3858925549915397, + "grad_norm": 8.5625, + "learning_rate": 4.240833817338408e-06, + "loss": 2.8958, + "step": 13105 + }, + { + "epoch": 1.3859983079526228, + "grad_norm": 8.6875, + "learning_rate": 4.2407108995216086e-06, + "loss": 2.8177, + "step": 13106 + }, + { + "epoch": 1.3861040609137056, + "grad_norm": 8.625, + "learning_rate": 4.240587973536363e-06, + "loss": 2.3098, + "step": 13107 + }, + { + "epoch": 1.3862098138747885, + "grad_norm": 7.84375, + "learning_rate": 4.240465039383251e-06, + "loss": 2.4878, + "step": 13108 + }, + { + "epoch": 1.3863155668358713, + "grad_norm": 8.1875, + "learning_rate": 4.2403420970628494e-06, + "loss": 2.6187, + "step": 13109 + }, + { + "epoch": 1.3864213197969544, + "grad_norm": 8.3125, + "learning_rate": 4.240219146575734e-06, + "loss": 2.6237, + "step": 13110 + }, + { + "epoch": 1.3865270727580372, + "grad_norm": 8.4375, + "learning_rate": 4.240096187922481e-06, + "loss": 2.6379, + "step": 13111 + }, + { + "epoch": 1.38663282571912, + "grad_norm": 8.1875, + "learning_rate": 4.239973221103668e-06, + "loss": 2.517, + "step": 13112 + }, + { + "epoch": 1.3867385786802031, + "grad_norm": 8.375, + "learning_rate": 4.239850246119873e-06, + "loss": 2.487, + "step": 13113 + }, + { + "epoch": 1.386844331641286, + "grad_norm": 8.25, + "learning_rate": 4.2397272629716716e-06, + "loss": 2.4211, + "step": 13114 + }, + { + "epoch": 1.3869500846023688, + "grad_norm": 8.1875, + "learning_rate": 4.239604271659642e-06, + "loss": 2.5573, + "step": 13115 + }, + { + "epoch": 1.3870558375634519, + "grad_norm": 8.1875, + "learning_rate": 4.2394812721843614e-06, + "loss": 2.5238, + "step": 13116 + }, + { + "epoch": 1.3871615905245347, + "grad_norm": 7.84375, + "learning_rate": 4.239358264546406e-06, + "loss": 2.408, + "step": 13117 + }, + { + "epoch": 1.3872673434856175, + "grad_norm": 8.3125, + "learning_rate": 4.2392352487463546e-06, + "loss": 2.2799, + "step": 13118 + }, + { + "epoch": 1.3873730964467006, + "grad_norm": 8.25, + "learning_rate": 4.239112224784783e-06, + "loss": 2.6133, + "step": 13119 + }, + { + "epoch": 1.3874788494077834, + "grad_norm": 7.9375, + "learning_rate": 4.238989192662268e-06, + "loss": 2.5042, + "step": 13120 + }, + { + "epoch": 1.3875846023688663, + "grad_norm": 8.0625, + "learning_rate": 4.23886615237939e-06, + "loss": 2.7136, + "step": 13121 + }, + { + "epoch": 1.387690355329949, + "grad_norm": 8.5625, + "learning_rate": 4.2387431039367235e-06, + "loss": 2.6658, + "step": 13122 + }, + { + "epoch": 1.3877961082910322, + "grad_norm": 8.6875, + "learning_rate": 4.238620047334846e-06, + "loss": 2.315, + "step": 13123 + }, + { + "epoch": 1.387901861252115, + "grad_norm": 8.125, + "learning_rate": 4.238496982574336e-06, + "loss": 2.7171, + "step": 13124 + }, + { + "epoch": 1.388007614213198, + "grad_norm": 8.25, + "learning_rate": 4.2383739096557704e-06, + "loss": 2.6517, + "step": 13125 + }, + { + "epoch": 1.388113367174281, + "grad_norm": 8.75, + "learning_rate": 4.238250828579728e-06, + "loss": 2.4197, + "step": 13126 + }, + { + "epoch": 1.3882191201353637, + "grad_norm": 8.25, + "learning_rate": 4.238127739346785e-06, + "loss": 2.2592, + "step": 13127 + }, + { + "epoch": 1.3883248730964466, + "grad_norm": 8.0625, + "learning_rate": 4.238004641957519e-06, + "loss": 2.5206, + "step": 13128 + }, + { + "epoch": 1.3884306260575296, + "grad_norm": 8.5, + "learning_rate": 4.237881536412508e-06, + "loss": 2.7203, + "step": 13129 + }, + { + "epoch": 1.3885363790186125, + "grad_norm": 7.90625, + "learning_rate": 4.237758422712331e-06, + "loss": 2.5023, + "step": 13130 + }, + { + "epoch": 1.3886421319796955, + "grad_norm": 8.6875, + "learning_rate": 4.237635300857562e-06, + "loss": 2.6394, + "step": 13131 + }, + { + "epoch": 1.3887478849407784, + "grad_norm": 8.5, + "learning_rate": 4.237512170848783e-06, + "loss": 2.5826, + "step": 13132 + }, + { + "epoch": 1.3888536379018612, + "grad_norm": 7.90625, + "learning_rate": 4.23738903268657e-06, + "loss": 2.485, + "step": 13133 + }, + { + "epoch": 1.388959390862944, + "grad_norm": 8.3125, + "learning_rate": 4.2372658863714995e-06, + "loss": 2.631, + "step": 13134 + }, + { + "epoch": 1.3890651438240271, + "grad_norm": 8.625, + "learning_rate": 4.237142731904152e-06, + "loss": 2.6019, + "step": 13135 + }, + { + "epoch": 1.38917089678511, + "grad_norm": 8.75, + "learning_rate": 4.237019569285103e-06, + "loss": 2.5649, + "step": 13136 + }, + { + "epoch": 1.389276649746193, + "grad_norm": 8.25, + "learning_rate": 4.236896398514933e-06, + "loss": 2.4989, + "step": 13137 + }, + { + "epoch": 1.3893824027072759, + "grad_norm": 8.5, + "learning_rate": 4.236773219594218e-06, + "loss": 2.5822, + "step": 13138 + }, + { + "epoch": 1.3894881556683587, + "grad_norm": 8.0625, + "learning_rate": 4.236650032523537e-06, + "loss": 2.5595, + "step": 13139 + }, + { + "epoch": 1.3895939086294415, + "grad_norm": 8.125, + "learning_rate": 4.236526837303467e-06, + "loss": 2.5436, + "step": 13140 + }, + { + "epoch": 1.3896996615905246, + "grad_norm": 7.75, + "learning_rate": 4.236403633934587e-06, + "loss": 2.3856, + "step": 13141 + }, + { + "epoch": 1.3898054145516074, + "grad_norm": 8.4375, + "learning_rate": 4.2362804224174745e-06, + "loss": 2.7281, + "step": 13142 + }, + { + "epoch": 1.3899111675126903, + "grad_norm": 7.9375, + "learning_rate": 4.236157202752709e-06, + "loss": 2.5964, + "step": 13143 + }, + { + "epoch": 1.3900169204737733, + "grad_norm": 8.4375, + "learning_rate": 4.236033974940867e-06, + "loss": 2.7385, + "step": 13144 + }, + { + "epoch": 1.3901226734348562, + "grad_norm": 7.75, + "learning_rate": 4.235910738982528e-06, + "loss": 2.4971, + "step": 13145 + }, + { + "epoch": 1.390228426395939, + "grad_norm": 8.5625, + "learning_rate": 4.23578749487827e-06, + "loss": 2.5773, + "step": 13146 + }, + { + "epoch": 1.3903341793570219, + "grad_norm": 8.0625, + "learning_rate": 4.23566424262867e-06, + "loss": 2.5249, + "step": 13147 + }, + { + "epoch": 1.390439932318105, + "grad_norm": 8.375, + "learning_rate": 4.235540982234309e-06, + "loss": 2.6607, + "step": 13148 + }, + { + "epoch": 1.3905456852791878, + "grad_norm": 8.1875, + "learning_rate": 4.235417713695763e-06, + "loss": 2.6771, + "step": 13149 + }, + { + "epoch": 1.3906514382402708, + "grad_norm": 8.1875, + "learning_rate": 4.235294437013613e-06, + "loss": 2.5221, + "step": 13150 + }, + { + "epoch": 1.3907571912013537, + "grad_norm": 8.3125, + "learning_rate": 4.2351711521884344e-06, + "loss": 2.4938, + "step": 13151 + }, + { + "epoch": 1.3908629441624365, + "grad_norm": 8.75, + "learning_rate": 4.2350478592208064e-06, + "loss": 2.571, + "step": 13152 + }, + { + "epoch": 1.3909686971235193, + "grad_norm": 8.3125, + "learning_rate": 4.2349245581113105e-06, + "loss": 2.2089, + "step": 13153 + }, + { + "epoch": 1.3910744500846024, + "grad_norm": 8.25, + "learning_rate": 4.234801248860522e-06, + "loss": 2.3572, + "step": 13154 + }, + { + "epoch": 1.3911802030456852, + "grad_norm": 8.5, + "learning_rate": 4.234677931469022e-06, + "loss": 2.681, + "step": 13155 + }, + { + "epoch": 1.3912859560067683, + "grad_norm": 8.5625, + "learning_rate": 4.2345546059373864e-06, + "loss": 2.5496, + "step": 13156 + }, + { + "epoch": 1.3913917089678511, + "grad_norm": 7.625, + "learning_rate": 4.234431272266197e-06, + "loss": 2.5064, + "step": 13157 + }, + { + "epoch": 1.391497461928934, + "grad_norm": 8.5625, + "learning_rate": 4.234307930456029e-06, + "loss": 2.6487, + "step": 13158 + }, + { + "epoch": 1.3916032148900168, + "grad_norm": 8.3125, + "learning_rate": 4.234184580507465e-06, + "loss": 2.525, + "step": 13159 + }, + { + "epoch": 1.3917089678510999, + "grad_norm": 8.25, + "learning_rate": 4.234061222421081e-06, + "loss": 2.6457, + "step": 13160 + }, + { + "epoch": 1.3918147208121827, + "grad_norm": 8.125, + "learning_rate": 4.2339378561974564e-06, + "loss": 2.5553, + "step": 13161 + }, + { + "epoch": 1.3919204737732658, + "grad_norm": 8.9375, + "learning_rate": 4.233814481837172e-06, + "loss": 2.5048, + "step": 13162 + }, + { + "epoch": 1.3920262267343486, + "grad_norm": 9.4375, + "learning_rate": 4.233691099340804e-06, + "loss": 2.685, + "step": 13163 + }, + { + "epoch": 1.3921319796954315, + "grad_norm": 8.3125, + "learning_rate": 4.233567708708934e-06, + "loss": 2.6829, + "step": 13164 + }, + { + "epoch": 1.3922377326565143, + "grad_norm": 8.3125, + "learning_rate": 4.23344430994214e-06, + "loss": 2.6386, + "step": 13165 + }, + { + "epoch": 1.3923434856175974, + "grad_norm": 8.1875, + "learning_rate": 4.233320903041e-06, + "loss": 2.6756, + "step": 13166 + }, + { + "epoch": 1.3924492385786802, + "grad_norm": 9.0, + "learning_rate": 4.233197488006093e-06, + "loss": 2.3133, + "step": 13167 + }, + { + "epoch": 1.392554991539763, + "grad_norm": 8.125, + "learning_rate": 4.233074064838e-06, + "loss": 2.4141, + "step": 13168 + }, + { + "epoch": 1.392660744500846, + "grad_norm": 9.125, + "learning_rate": 4.232950633537299e-06, + "loss": 2.626, + "step": 13169 + }, + { + "epoch": 1.392766497461929, + "grad_norm": 9.1875, + "learning_rate": 4.232827194104569e-06, + "loss": 2.5172, + "step": 13170 + }, + { + "epoch": 1.3928722504230118, + "grad_norm": 8.125, + "learning_rate": 4.232703746540391e-06, + "loss": 2.5664, + "step": 13171 + }, + { + "epoch": 1.3929780033840948, + "grad_norm": 8.25, + "learning_rate": 4.2325802908453425e-06, + "loss": 2.6676, + "step": 13172 + }, + { + "epoch": 1.3930837563451777, + "grad_norm": 8.9375, + "learning_rate": 4.232456827020003e-06, + "loss": 2.6871, + "step": 13173 + }, + { + "epoch": 1.3931895093062605, + "grad_norm": 8.5625, + "learning_rate": 4.232333355064952e-06, + "loss": 2.5949, + "step": 13174 + }, + { + "epoch": 1.3932952622673436, + "grad_norm": 8.75, + "learning_rate": 4.232209874980769e-06, + "loss": 2.8003, + "step": 13175 + }, + { + "epoch": 1.3934010152284264, + "grad_norm": 8.3125, + "learning_rate": 4.232086386768034e-06, + "loss": 2.4889, + "step": 13176 + }, + { + "epoch": 1.3935067681895092, + "grad_norm": 7.5625, + "learning_rate": 4.231962890427326e-06, + "loss": 2.4121, + "step": 13177 + }, + { + "epoch": 1.393612521150592, + "grad_norm": 8.1875, + "learning_rate": 4.231839385959224e-06, + "loss": 2.7681, + "step": 13178 + }, + { + "epoch": 1.3937182741116751, + "grad_norm": 8.25, + "learning_rate": 4.231715873364308e-06, + "loss": 2.4561, + "step": 13179 + }, + { + "epoch": 1.393824027072758, + "grad_norm": 7.96875, + "learning_rate": 4.2315923526431585e-06, + "loss": 2.4888, + "step": 13180 + }, + { + "epoch": 1.393929780033841, + "grad_norm": 8.6875, + "learning_rate": 4.231468823796353e-06, + "loss": 2.4625, + "step": 13181 + }, + { + "epoch": 1.3940355329949239, + "grad_norm": 8.5625, + "learning_rate": 4.2313452868244745e-06, + "loss": 2.6555, + "step": 13182 + }, + { + "epoch": 1.3941412859560067, + "grad_norm": 8.125, + "learning_rate": 4.231221741728099e-06, + "loss": 2.5402, + "step": 13183 + }, + { + "epoch": 1.3942470389170896, + "grad_norm": 8.3125, + "learning_rate": 4.231098188507809e-06, + "loss": 2.7484, + "step": 13184 + }, + { + "epoch": 1.3943527918781726, + "grad_norm": 8.3125, + "learning_rate": 4.2309746271641814e-06, + "loss": 2.5444, + "step": 13185 + }, + { + "epoch": 1.3944585448392555, + "grad_norm": 8.375, + "learning_rate": 4.230851057697799e-06, + "loss": 2.7266, + "step": 13186 + }, + { + "epoch": 1.3945642978003385, + "grad_norm": 8.375, + "learning_rate": 4.230727480109241e-06, + "loss": 2.5719, + "step": 13187 + }, + { + "epoch": 1.3946700507614214, + "grad_norm": 7.75, + "learning_rate": 4.230603894399086e-06, + "loss": 2.4581, + "step": 13188 + }, + { + "epoch": 1.3947758037225042, + "grad_norm": 8.3125, + "learning_rate": 4.230480300567914e-06, + "loss": 2.52, + "step": 13189 + }, + { + "epoch": 1.394881556683587, + "grad_norm": 8.0625, + "learning_rate": 4.230356698616307e-06, + "loss": 2.6646, + "step": 13190 + }, + { + "epoch": 1.39498730964467, + "grad_norm": 8.0, + "learning_rate": 4.230233088544843e-06, + "loss": 2.1107, + "step": 13191 + }, + { + "epoch": 1.395093062605753, + "grad_norm": 8.6875, + "learning_rate": 4.230109470354103e-06, + "loss": 2.3627, + "step": 13192 + }, + { + "epoch": 1.395198815566836, + "grad_norm": 8.4375, + "learning_rate": 4.2299858440446675e-06, + "loss": 2.6408, + "step": 13193 + }, + { + "epoch": 1.3953045685279188, + "grad_norm": 7.75, + "learning_rate": 4.229862209617115e-06, + "loss": 2.4001, + "step": 13194 + }, + { + "epoch": 1.3954103214890017, + "grad_norm": 8.5625, + "learning_rate": 4.229738567072027e-06, + "loss": 2.3557, + "step": 13195 + }, + { + "epoch": 1.3955160744500845, + "grad_norm": 8.0625, + "learning_rate": 4.229614916409984e-06, + "loss": 2.6602, + "step": 13196 + }, + { + "epoch": 1.3956218274111676, + "grad_norm": 8.3125, + "learning_rate": 4.229491257631564e-06, + "loss": 2.5025, + "step": 13197 + }, + { + "epoch": 1.3957275803722504, + "grad_norm": 7.9375, + "learning_rate": 4.22936759073735e-06, + "loss": 2.6584, + "step": 13198 + }, + { + "epoch": 1.3958333333333333, + "grad_norm": 8.1875, + "learning_rate": 4.229243915727921e-06, + "loss": 2.6062, + "step": 13199 + }, + { + "epoch": 1.3959390862944163, + "grad_norm": 7.9375, + "learning_rate": 4.229120232603858e-06, + "loss": 2.5719, + "step": 13200 + }, + { + "epoch": 1.3960448392554992, + "grad_norm": 8.1875, + "learning_rate": 4.22899654136574e-06, + "loss": 2.6008, + "step": 13201 + }, + { + "epoch": 1.396150592216582, + "grad_norm": 8.5625, + "learning_rate": 4.228872842014149e-06, + "loss": 2.6034, + "step": 13202 + }, + { + "epoch": 1.3962563451776648, + "grad_norm": 8.0, + "learning_rate": 4.228749134549665e-06, + "loss": 2.539, + "step": 13203 + }, + { + "epoch": 1.396362098138748, + "grad_norm": 8.0, + "learning_rate": 4.2286254189728675e-06, + "loss": 2.3293, + "step": 13204 + }, + { + "epoch": 1.3964678510998307, + "grad_norm": 7.53125, + "learning_rate": 4.228501695284338e-06, + "loss": 2.6004, + "step": 13205 + }, + { + "epoch": 1.3965736040609138, + "grad_norm": 8.3125, + "learning_rate": 4.228377963484658e-06, + "loss": 2.4516, + "step": 13206 + }, + { + "epoch": 1.3966793570219966, + "grad_norm": 8.1875, + "learning_rate": 4.228254223574406e-06, + "loss": 2.4791, + "step": 13207 + }, + { + "epoch": 1.3967851099830795, + "grad_norm": 8.125, + "learning_rate": 4.2281304755541644e-06, + "loss": 2.5653, + "step": 13208 + }, + { + "epoch": 1.3968908629441623, + "grad_norm": 8.1875, + "learning_rate": 4.228006719424513e-06, + "loss": 2.4204, + "step": 13209 + }, + { + "epoch": 1.3969966159052454, + "grad_norm": 7.75, + "learning_rate": 4.227882955186034e-06, + "loss": 2.5031, + "step": 13210 + }, + { + "epoch": 1.3971023688663282, + "grad_norm": 12.75, + "learning_rate": 4.2277591828393046e-06, + "loss": 2.3878, + "step": 13211 + }, + { + "epoch": 1.3972081218274113, + "grad_norm": 8.1875, + "learning_rate": 4.22763540238491e-06, + "loss": 2.4207, + "step": 13212 + }, + { + "epoch": 1.3973138747884941, + "grad_norm": 9.25, + "learning_rate": 4.227511613823428e-06, + "loss": 2.6277, + "step": 13213 + }, + { + "epoch": 1.397419627749577, + "grad_norm": 8.6875, + "learning_rate": 4.22738781715544e-06, + "loss": 2.432, + "step": 13214 + }, + { + "epoch": 1.3975253807106598, + "grad_norm": 9.125, + "learning_rate": 4.227264012381529e-06, + "loss": 2.7557, + "step": 13215 + }, + { + "epoch": 1.3976311336717429, + "grad_norm": 7.875, + "learning_rate": 4.227140199502273e-06, + "loss": 2.4848, + "step": 13216 + }, + { + "epoch": 1.3977368866328257, + "grad_norm": 8.0, + "learning_rate": 4.227016378518256e-06, + "loss": 2.3599, + "step": 13217 + }, + { + "epoch": 1.3978426395939088, + "grad_norm": 8.0, + "learning_rate": 4.2268925494300565e-06, + "loss": 2.2758, + "step": 13218 + }, + { + "epoch": 1.3979483925549916, + "grad_norm": 8.0, + "learning_rate": 4.226768712238256e-06, + "loss": 2.7686, + "step": 13219 + }, + { + "epoch": 1.3980541455160744, + "grad_norm": 8.5, + "learning_rate": 4.2266448669434365e-06, + "loss": 2.6238, + "step": 13220 + }, + { + "epoch": 1.3981598984771573, + "grad_norm": 7.875, + "learning_rate": 4.2265210135461795e-06, + "loss": 2.5437, + "step": 13221 + }, + { + "epoch": 1.3982656514382403, + "grad_norm": 8.375, + "learning_rate": 4.226397152047065e-06, + "loss": 2.6052, + "step": 13222 + }, + { + "epoch": 1.3983714043993232, + "grad_norm": 8.5625, + "learning_rate": 4.226273282446674e-06, + "loss": 2.6942, + "step": 13223 + }, + { + "epoch": 1.398477157360406, + "grad_norm": 8.3125, + "learning_rate": 4.226149404745591e-06, + "loss": 2.6033, + "step": 13224 + }, + { + "epoch": 1.398582910321489, + "grad_norm": 7.96875, + "learning_rate": 4.226025518944392e-06, + "loss": 2.5355, + "step": 13225 + }, + { + "epoch": 1.398688663282572, + "grad_norm": 8.125, + "learning_rate": 4.225901625043663e-06, + "loss": 2.5314, + "step": 13226 + }, + { + "epoch": 1.3987944162436547, + "grad_norm": 8.375, + "learning_rate": 4.225777723043982e-06, + "loss": 2.5544, + "step": 13227 + }, + { + "epoch": 1.3989001692047378, + "grad_norm": 7.96875, + "learning_rate": 4.225653812945933e-06, + "loss": 2.4756, + "step": 13228 + }, + { + "epoch": 1.3990059221658206, + "grad_norm": 8.4375, + "learning_rate": 4.225529894750097e-06, + "loss": 2.1635, + "step": 13229 + }, + { + "epoch": 1.3991116751269035, + "grad_norm": 8.4375, + "learning_rate": 4.225405968457054e-06, + "loss": 2.3374, + "step": 13230 + }, + { + "epoch": 1.3992174280879865, + "grad_norm": 8.375, + "learning_rate": 4.225282034067386e-06, + "loss": 2.5094, + "step": 13231 + }, + { + "epoch": 1.3993231810490694, + "grad_norm": 8.375, + "learning_rate": 4.225158091581676e-06, + "loss": 2.6686, + "step": 13232 + }, + { + "epoch": 1.3994289340101522, + "grad_norm": 8.3125, + "learning_rate": 4.225034141000504e-06, + "loss": 2.5278, + "step": 13233 + }, + { + "epoch": 1.399534686971235, + "grad_norm": 22.375, + "learning_rate": 4.224910182324453e-06, + "loss": 2.8423, + "step": 13234 + }, + { + "epoch": 1.3996404399323181, + "grad_norm": 7.9375, + "learning_rate": 4.224786215554103e-06, + "loss": 2.5427, + "step": 13235 + }, + { + "epoch": 1.399746192893401, + "grad_norm": 7.90625, + "learning_rate": 4.224662240690037e-06, + "loss": 2.533, + "step": 13236 + }, + { + "epoch": 1.399851945854484, + "grad_norm": 8.25, + "learning_rate": 4.224538257732838e-06, + "loss": 2.4623, + "step": 13237 + }, + { + "epoch": 1.3999576988155669, + "grad_norm": 8.4375, + "learning_rate": 4.224414266683084e-06, + "loss": 2.3547, + "step": 13238 + }, + { + "epoch": 1.4000634517766497, + "grad_norm": 8.0625, + "learning_rate": 4.224290267541361e-06, + "loss": 2.467, + "step": 13239 + }, + { + "epoch": 1.4001692047377325, + "grad_norm": 8.25, + "learning_rate": 4.224166260308248e-06, + "loss": 2.5473, + "step": 13240 + }, + { + "epoch": 1.4002749576988156, + "grad_norm": 8.5625, + "learning_rate": 4.224042244984327e-06, + "loss": 2.4974, + "step": 13241 + }, + { + "epoch": 1.4003807106598984, + "grad_norm": 8.3125, + "learning_rate": 4.223918221570182e-06, + "loss": 2.5258, + "step": 13242 + }, + { + "epoch": 1.4004864636209815, + "grad_norm": 9.5625, + "learning_rate": 4.223794190066393e-06, + "loss": 2.7786, + "step": 13243 + }, + { + "epoch": 1.4005922165820643, + "grad_norm": 9.125, + "learning_rate": 4.223670150473544e-06, + "loss": 2.6418, + "step": 13244 + }, + { + "epoch": 1.4006979695431472, + "grad_norm": 8.75, + "learning_rate": 4.2235461027922156e-06, + "loss": 2.4591, + "step": 13245 + }, + { + "epoch": 1.40080372250423, + "grad_norm": 8.125, + "learning_rate": 4.22342204702299e-06, + "loss": 2.6246, + "step": 13246 + }, + { + "epoch": 1.400909475465313, + "grad_norm": 9.125, + "learning_rate": 4.22329798316645e-06, + "loss": 2.4144, + "step": 13247 + }, + { + "epoch": 1.401015228426396, + "grad_norm": 7.875, + "learning_rate": 4.2231739112231775e-06, + "loss": 2.3982, + "step": 13248 + }, + { + "epoch": 1.401120981387479, + "grad_norm": 8.0625, + "learning_rate": 4.223049831193754e-06, + "loss": 2.5468, + "step": 13249 + }, + { + "epoch": 1.4012267343485618, + "grad_norm": 8.8125, + "learning_rate": 4.222925743078762e-06, + "loss": 2.6738, + "step": 13250 + }, + { + "epoch": 1.4013324873096447, + "grad_norm": 8.75, + "learning_rate": 4.222801646878785e-06, + "loss": 2.6472, + "step": 13251 + }, + { + "epoch": 1.4014382402707275, + "grad_norm": 8.0, + "learning_rate": 4.222677542594405e-06, + "loss": 2.5645, + "step": 13252 + }, + { + "epoch": 1.4015439932318106, + "grad_norm": 8.4375, + "learning_rate": 4.2225534302262026e-06, + "loss": 2.5447, + "step": 13253 + }, + { + "epoch": 1.4016497461928934, + "grad_norm": 8.5, + "learning_rate": 4.2224293097747615e-06, + "loss": 2.198, + "step": 13254 + }, + { + "epoch": 1.4017554991539762, + "grad_norm": 8.125, + "learning_rate": 4.222305181240664e-06, + "loss": 2.6556, + "step": 13255 + }, + { + "epoch": 1.4018612521150593, + "grad_norm": 8.75, + "learning_rate": 4.222181044624494e-06, + "loss": 2.6574, + "step": 13256 + }, + { + "epoch": 1.4019670050761421, + "grad_norm": 9.0, + "learning_rate": 4.222056899926831e-06, + "loss": 2.6063, + "step": 13257 + }, + { + "epoch": 1.402072758037225, + "grad_norm": 8.5625, + "learning_rate": 4.22193274714826e-06, + "loss": 2.1999, + "step": 13258 + }, + { + "epoch": 1.4021785109983078, + "grad_norm": 8.125, + "learning_rate": 4.221808586289363e-06, + "loss": 2.5874, + "step": 13259 + }, + { + "epoch": 1.4022842639593909, + "grad_norm": 8.5, + "learning_rate": 4.221684417350723e-06, + "loss": 2.7864, + "step": 13260 + }, + { + "epoch": 1.4023900169204737, + "grad_norm": 8.4375, + "learning_rate": 4.221560240332921e-06, + "loss": 2.5207, + "step": 13261 + }, + { + "epoch": 1.4024957698815568, + "grad_norm": 8.3125, + "learning_rate": 4.2214360552365415e-06, + "loss": 2.4018, + "step": 13262 + }, + { + "epoch": 1.4026015228426396, + "grad_norm": 8.0625, + "learning_rate": 4.221311862062168e-06, + "loss": 2.4606, + "step": 13263 + }, + { + "epoch": 1.4027072758037225, + "grad_norm": 9.0, + "learning_rate": 4.221187660810379e-06, + "loss": 2.5675, + "step": 13264 + }, + { + "epoch": 1.4028130287648053, + "grad_norm": 8.25, + "learning_rate": 4.2210634514817625e-06, + "loss": 2.6801, + "step": 13265 + }, + { + "epoch": 1.4029187817258884, + "grad_norm": 8.125, + "learning_rate": 4.220939234076898e-06, + "loss": 2.4934, + "step": 13266 + }, + { + "epoch": 1.4030245346869712, + "grad_norm": 8.875, + "learning_rate": 4.220815008596371e-06, + "loss": 2.3089, + "step": 13267 + }, + { + "epoch": 1.4031302876480543, + "grad_norm": 8.125, + "learning_rate": 4.220690775040761e-06, + "loss": 2.4168, + "step": 13268 + }, + { + "epoch": 1.403236040609137, + "grad_norm": 8.25, + "learning_rate": 4.220566533410654e-06, + "loss": 2.572, + "step": 13269 + }, + { + "epoch": 1.40334179357022, + "grad_norm": 8.3125, + "learning_rate": 4.220442283706632e-06, + "loss": 2.7408, + "step": 13270 + }, + { + "epoch": 1.4034475465313028, + "grad_norm": 8.5, + "learning_rate": 4.220318025929278e-06, + "loss": 2.4463, + "step": 13271 + }, + { + "epoch": 1.4035532994923858, + "grad_norm": 8.1875, + "learning_rate": 4.220193760079174e-06, + "loss": 2.5865, + "step": 13272 + }, + { + "epoch": 1.4036590524534687, + "grad_norm": 8.125, + "learning_rate": 4.220069486156906e-06, + "loss": 2.4607, + "step": 13273 + }, + { + "epoch": 1.4037648054145517, + "grad_norm": 8.1875, + "learning_rate": 4.219945204163054e-06, + "loss": 2.5407, + "step": 13274 + }, + { + "epoch": 1.4038705583756346, + "grad_norm": 7.6875, + "learning_rate": 4.219820914098203e-06, + "loss": 2.4202, + "step": 13275 + }, + { + "epoch": 1.4039763113367174, + "grad_norm": 7.625, + "learning_rate": 4.219696615962936e-06, + "loss": 2.5282, + "step": 13276 + }, + { + "epoch": 1.4040820642978002, + "grad_norm": 8.5625, + "learning_rate": 4.219572309757836e-06, + "loss": 2.699, + "step": 13277 + }, + { + "epoch": 1.4041878172588833, + "grad_norm": 7.96875, + "learning_rate": 4.219447995483487e-06, + "loss": 2.5666, + "step": 13278 + }, + { + "epoch": 1.4042935702199661, + "grad_norm": 8.4375, + "learning_rate": 4.2193236731404705e-06, + "loss": 2.4744, + "step": 13279 + }, + { + "epoch": 1.404399323181049, + "grad_norm": 9.0625, + "learning_rate": 4.219199342729372e-06, + "loss": 2.5937, + "step": 13280 + }, + { + "epoch": 1.404505076142132, + "grad_norm": 8.25, + "learning_rate": 4.219075004250775e-06, + "loss": 2.5545, + "step": 13281 + }, + { + "epoch": 1.4046108291032149, + "grad_norm": 8.6875, + "learning_rate": 4.218950657705261e-06, + "loss": 2.2534, + "step": 13282 + }, + { + "epoch": 1.4047165820642977, + "grad_norm": 8.9375, + "learning_rate": 4.218826303093414e-06, + "loss": 2.4653, + "step": 13283 + }, + { + "epoch": 1.4048223350253808, + "grad_norm": 8.25, + "learning_rate": 4.218701940415819e-06, + "loss": 2.4684, + "step": 13284 + }, + { + "epoch": 1.4049280879864636, + "grad_norm": 8.4375, + "learning_rate": 4.218577569673058e-06, + "loss": 2.5916, + "step": 13285 + }, + { + "epoch": 1.4050338409475465, + "grad_norm": 8.1875, + "learning_rate": 4.218453190865716e-06, + "loss": 2.5789, + "step": 13286 + }, + { + "epoch": 1.4051395939086295, + "grad_norm": 8.75, + "learning_rate": 4.218328803994376e-06, + "loss": 2.4469, + "step": 13287 + }, + { + "epoch": 1.4052453468697124, + "grad_norm": 9.0, + "learning_rate": 4.218204409059622e-06, + "loss": 2.4123, + "step": 13288 + }, + { + "epoch": 1.4053510998307952, + "grad_norm": 9.125, + "learning_rate": 4.218080006062037e-06, + "loss": 2.7284, + "step": 13289 + }, + { + "epoch": 1.405456852791878, + "grad_norm": 8.75, + "learning_rate": 4.217955595002205e-06, + "loss": 2.7707, + "step": 13290 + }, + { + "epoch": 1.405562605752961, + "grad_norm": 8.1875, + "learning_rate": 4.21783117588071e-06, + "loss": 2.6808, + "step": 13291 + }, + { + "epoch": 1.405668358714044, + "grad_norm": 8.4375, + "learning_rate": 4.217706748698136e-06, + "loss": 2.7024, + "step": 13292 + }, + { + "epoch": 1.405774111675127, + "grad_norm": 8.375, + "learning_rate": 4.217582313455067e-06, + "loss": 2.5473, + "step": 13293 + }, + { + "epoch": 1.4058798646362098, + "grad_norm": 9.0, + "learning_rate": 4.217457870152086e-06, + "loss": 2.3684, + "step": 13294 + }, + { + "epoch": 1.4059856175972927, + "grad_norm": 8.3125, + "learning_rate": 4.217333418789779e-06, + "loss": 2.6431, + "step": 13295 + }, + { + "epoch": 1.4060913705583755, + "grad_norm": 8.125, + "learning_rate": 4.217208959368727e-06, + "loss": 2.1879, + "step": 13296 + }, + { + "epoch": 1.4061971235194586, + "grad_norm": 8.75, + "learning_rate": 4.217084491889517e-06, + "loss": 2.7234, + "step": 13297 + }, + { + "epoch": 1.4063028764805414, + "grad_norm": 7.96875, + "learning_rate": 4.216960016352731e-06, + "loss": 2.4204, + "step": 13298 + }, + { + "epoch": 1.4064086294416245, + "grad_norm": 8.3125, + "learning_rate": 4.216835532758955e-06, + "loss": 2.6199, + "step": 13299 + }, + { + "epoch": 1.4065143824027073, + "grad_norm": 8.3125, + "learning_rate": 4.21671104110877e-06, + "loss": 2.7186, + "step": 13300 + }, + { + "epoch": 1.4066201353637902, + "grad_norm": 8.25, + "learning_rate": 4.216586541402763e-06, + "loss": 2.6447, + "step": 13301 + }, + { + "epoch": 1.406725888324873, + "grad_norm": 8.4375, + "learning_rate": 4.216462033641518e-06, + "loss": 2.7167, + "step": 13302 + }, + { + "epoch": 1.406831641285956, + "grad_norm": 8.375, + "learning_rate": 4.2163375178256185e-06, + "loss": 2.5845, + "step": 13303 + }, + { + "epoch": 1.406937394247039, + "grad_norm": 7.9375, + "learning_rate": 4.216212993955648e-06, + "loss": 2.5527, + "step": 13304 + }, + { + "epoch": 1.407043147208122, + "grad_norm": 8.5, + "learning_rate": 4.216088462032193e-06, + "loss": 2.5922, + "step": 13305 + }, + { + "epoch": 1.4071489001692048, + "grad_norm": 8.0625, + "learning_rate": 4.2159639220558366e-06, + "loss": 2.4671, + "step": 13306 + }, + { + "epoch": 1.4072546531302876, + "grad_norm": 7.875, + "learning_rate": 4.215839374027163e-06, + "loss": 2.4757, + "step": 13307 + }, + { + "epoch": 1.4073604060913705, + "grad_norm": 8.0625, + "learning_rate": 4.2157148179467575e-06, + "loss": 2.4226, + "step": 13308 + }, + { + "epoch": 1.4074661590524535, + "grad_norm": 8.125, + "learning_rate": 4.215590253815204e-06, + "loss": 2.6496, + "step": 13309 + }, + { + "epoch": 1.4075719120135364, + "grad_norm": 8.375, + "learning_rate": 4.215465681633086e-06, + "loss": 2.6295, + "step": 13310 + }, + { + "epoch": 1.4076776649746192, + "grad_norm": 9.0625, + "learning_rate": 4.2153411014009905e-06, + "loss": 2.5058, + "step": 13311 + }, + { + "epoch": 1.4077834179357023, + "grad_norm": 7.90625, + "learning_rate": 4.2152165131195e-06, + "loss": 2.6971, + "step": 13312 + }, + { + "epoch": 1.4078891708967851, + "grad_norm": 8.25, + "learning_rate": 4.2150919167892e-06, + "loss": 2.5815, + "step": 13313 + }, + { + "epoch": 1.407994923857868, + "grad_norm": 7.875, + "learning_rate": 4.214967312410675e-06, + "loss": 2.4889, + "step": 13314 + }, + { + "epoch": 1.408100676818951, + "grad_norm": 9.0625, + "learning_rate": 4.21484269998451e-06, + "loss": 2.4888, + "step": 13315 + }, + { + "epoch": 1.4082064297800339, + "grad_norm": 8.0, + "learning_rate": 4.214718079511289e-06, + "loss": 2.5102, + "step": 13316 + }, + { + "epoch": 1.4083121827411167, + "grad_norm": 8.375, + "learning_rate": 4.214593450991598e-06, + "loss": 2.5327, + "step": 13317 + }, + { + "epoch": 1.4084179357021998, + "grad_norm": 8.875, + "learning_rate": 4.214468814426022e-06, + "loss": 2.55, + "step": 13318 + }, + { + "epoch": 1.4085236886632826, + "grad_norm": 8.0625, + "learning_rate": 4.214344169815143e-06, + "loss": 2.5461, + "step": 13319 + }, + { + "epoch": 1.4086294416243654, + "grad_norm": 8.0625, + "learning_rate": 4.214219517159549e-06, + "loss": 2.3935, + "step": 13320 + }, + { + "epoch": 1.4087351945854483, + "grad_norm": 7.6875, + "learning_rate": 4.214094856459824e-06, + "loss": 2.518, + "step": 13321 + }, + { + "epoch": 1.4088409475465313, + "grad_norm": 8.0, + "learning_rate": 4.213970187716553e-06, + "loss": 2.5325, + "step": 13322 + }, + { + "epoch": 1.4089467005076142, + "grad_norm": 8.5625, + "learning_rate": 4.21384551093032e-06, + "loss": 2.4669, + "step": 13323 + }, + { + "epoch": 1.4090524534686972, + "grad_norm": 8.1875, + "learning_rate": 4.213720826101712e-06, + "loss": 2.4832, + "step": 13324 + }, + { + "epoch": 1.40915820642978, + "grad_norm": 7.96875, + "learning_rate": 4.213596133231312e-06, + "loss": 2.5415, + "step": 13325 + }, + { + "epoch": 1.409263959390863, + "grad_norm": 8.1875, + "learning_rate": 4.213471432319708e-06, + "loss": 2.5508, + "step": 13326 + }, + { + "epoch": 1.4093697123519457, + "grad_norm": 8.375, + "learning_rate": 4.213346723367482e-06, + "loss": 2.6446, + "step": 13327 + }, + { + "epoch": 1.4094754653130288, + "grad_norm": 8.0625, + "learning_rate": 4.2132220063752205e-06, + "loss": 2.4212, + "step": 13328 + }, + { + "epoch": 1.4095812182741116, + "grad_norm": 8.1875, + "learning_rate": 4.213097281343509e-06, + "loss": 2.7444, + "step": 13329 + }, + { + "epoch": 1.4096869712351947, + "grad_norm": 8.25, + "learning_rate": 4.212972548272932e-06, + "loss": 2.459, + "step": 13330 + }, + { + "epoch": 1.4097927241962775, + "grad_norm": 8.4375, + "learning_rate": 4.212847807164076e-06, + "loss": 2.6667, + "step": 13331 + }, + { + "epoch": 1.4098984771573604, + "grad_norm": 8.625, + "learning_rate": 4.212723058017527e-06, + "loss": 2.39, + "step": 13332 + }, + { + "epoch": 1.4100042301184432, + "grad_norm": 8.75, + "learning_rate": 4.212598300833868e-06, + "loss": 2.5457, + "step": 13333 + }, + { + "epoch": 1.4101099830795263, + "grad_norm": 8.3125, + "learning_rate": 4.212473535613686e-06, + "loss": 2.5608, + "step": 13334 + }, + { + "epoch": 1.4102157360406091, + "grad_norm": 8.4375, + "learning_rate": 4.212348762357566e-06, + "loss": 2.5119, + "step": 13335 + }, + { + "epoch": 1.4103214890016922, + "grad_norm": 8.0, + "learning_rate": 4.212223981066093e-06, + "loss": 2.3614, + "step": 13336 + }, + { + "epoch": 1.410427241962775, + "grad_norm": 7.9375, + "learning_rate": 4.212099191739853e-06, + "loss": 2.6408, + "step": 13337 + }, + { + "epoch": 1.4105329949238579, + "grad_norm": 8.125, + "learning_rate": 4.2119743943794315e-06, + "loss": 2.4447, + "step": 13338 + }, + { + "epoch": 1.4106387478849407, + "grad_norm": 8.375, + "learning_rate": 4.211849588985415e-06, + "loss": 2.6412, + "step": 13339 + }, + { + "epoch": 1.4107445008460238, + "grad_norm": 7.9375, + "learning_rate": 4.211724775558389e-06, + "loss": 2.465, + "step": 13340 + }, + { + "epoch": 1.4108502538071066, + "grad_norm": 8.5, + "learning_rate": 4.211599954098939e-06, + "loss": 2.594, + "step": 13341 + }, + { + "epoch": 1.4109560067681894, + "grad_norm": 8.25, + "learning_rate": 4.211475124607649e-06, + "loss": 2.4106, + "step": 13342 + }, + { + "epoch": 1.4110617597292725, + "grad_norm": 8.125, + "learning_rate": 4.211350287085106e-06, + "loss": 2.5337, + "step": 13343 + }, + { + "epoch": 1.4111675126903553, + "grad_norm": 11.5625, + "learning_rate": 4.211225441531898e-06, + "loss": 2.4419, + "step": 13344 + }, + { + "epoch": 1.4112732656514382, + "grad_norm": 8.25, + "learning_rate": 4.211100587948607e-06, + "loss": 2.7895, + "step": 13345 + }, + { + "epoch": 1.411379018612521, + "grad_norm": 8.25, + "learning_rate": 4.210975726335822e-06, + "loss": 2.4899, + "step": 13346 + }, + { + "epoch": 1.411484771573604, + "grad_norm": 8.625, + "learning_rate": 4.210850856694127e-06, + "loss": 2.7023, + "step": 13347 + }, + { + "epoch": 1.411590524534687, + "grad_norm": 9.5, + "learning_rate": 4.210725979024109e-06, + "loss": 2.4572, + "step": 13348 + }, + { + "epoch": 1.41169627749577, + "grad_norm": 8.6875, + "learning_rate": 4.210601093326353e-06, + "loss": 2.6246, + "step": 13349 + }, + { + "epoch": 1.4118020304568528, + "grad_norm": 8.375, + "learning_rate": 4.210476199601446e-06, + "loss": 2.6239, + "step": 13350 + }, + { + "epoch": 1.4119077834179357, + "grad_norm": 8.375, + "learning_rate": 4.210351297849974e-06, + "loss": 2.44, + "step": 13351 + }, + { + "epoch": 1.4120135363790185, + "grad_norm": 8.0, + "learning_rate": 4.2102263880725225e-06, + "loss": 2.4721, + "step": 13352 + }, + { + "epoch": 1.4121192893401016, + "grad_norm": 8.1875, + "learning_rate": 4.2101014702696785e-06, + "loss": 2.5202, + "step": 13353 + }, + { + "epoch": 1.4122250423011844, + "grad_norm": 9.0, + "learning_rate": 4.2099765444420275e-06, + "loss": 2.497, + "step": 13354 + }, + { + "epoch": 1.4123307952622675, + "grad_norm": 8.0625, + "learning_rate": 4.209851610590156e-06, + "loss": 2.7051, + "step": 13355 + }, + { + "epoch": 1.4124365482233503, + "grad_norm": 8.25, + "learning_rate": 4.20972666871465e-06, + "loss": 2.6282, + "step": 13356 + }, + { + "epoch": 1.4125423011844331, + "grad_norm": 8.375, + "learning_rate": 4.209601718816097e-06, + "loss": 2.8037, + "step": 13357 + }, + { + "epoch": 1.412648054145516, + "grad_norm": 8.0625, + "learning_rate": 4.2094767608950805e-06, + "loss": 2.6868, + "step": 13358 + }, + { + "epoch": 1.412753807106599, + "grad_norm": 8.3125, + "learning_rate": 4.20935179495219e-06, + "loss": 2.5046, + "step": 13359 + }, + { + "epoch": 1.4128595600676819, + "grad_norm": 8.5625, + "learning_rate": 4.20922682098801e-06, + "loss": 2.7779, + "step": 13360 + }, + { + "epoch": 1.412965313028765, + "grad_norm": 8.0625, + "learning_rate": 4.209101839003128e-06, + "loss": 2.5099, + "step": 13361 + }, + { + "epoch": 1.4130710659898478, + "grad_norm": 8.0, + "learning_rate": 4.20897684899813e-06, + "loss": 2.5827, + "step": 13362 + }, + { + "epoch": 1.4131768189509306, + "grad_norm": 8.5625, + "learning_rate": 4.208851850973603e-06, + "loss": 2.7462, + "step": 13363 + }, + { + "epoch": 1.4132825719120135, + "grad_norm": 8.25, + "learning_rate": 4.208726844930132e-06, + "loss": 2.7134, + "step": 13364 + }, + { + "epoch": 1.4133883248730965, + "grad_norm": 8.875, + "learning_rate": 4.208601830868305e-06, + "loss": 2.6177, + "step": 13365 + }, + { + "epoch": 1.4134940778341794, + "grad_norm": 7.46875, + "learning_rate": 4.20847680878871e-06, + "loss": 2.4135, + "step": 13366 + }, + { + "epoch": 1.4135998307952622, + "grad_norm": 8.75, + "learning_rate": 4.208351778691931e-06, + "loss": 2.5694, + "step": 13367 + }, + { + "epoch": 1.4137055837563453, + "grad_norm": 7.9375, + "learning_rate": 4.208226740578555e-06, + "loss": 2.4013, + "step": 13368 + }, + { + "epoch": 1.413811336717428, + "grad_norm": 8.25, + "learning_rate": 4.2081016944491706e-06, + "loss": 2.5314, + "step": 13369 + }, + { + "epoch": 1.413917089678511, + "grad_norm": 8.4375, + "learning_rate": 4.207976640304363e-06, + "loss": 2.537, + "step": 13370 + }, + { + "epoch": 1.414022842639594, + "grad_norm": 8.125, + "learning_rate": 4.20785157814472e-06, + "loss": 2.3501, + "step": 13371 + }, + { + "epoch": 1.4141285956006768, + "grad_norm": 8.0625, + "learning_rate": 4.207726507970828e-06, + "loss": 2.5604, + "step": 13372 + }, + { + "epoch": 1.4142343485617597, + "grad_norm": 8.875, + "learning_rate": 4.2076014297832736e-06, + "loss": 2.5065, + "step": 13373 + }, + { + "epoch": 1.4143401015228427, + "grad_norm": 8.375, + "learning_rate": 4.207476343582644e-06, + "loss": 2.5427, + "step": 13374 + }, + { + "epoch": 1.4144458544839256, + "grad_norm": 8.375, + "learning_rate": 4.207351249369527e-06, + "loss": 2.5195, + "step": 13375 + }, + { + "epoch": 1.4145516074450084, + "grad_norm": 8.4375, + "learning_rate": 4.2072261471445074e-06, + "loss": 2.3898, + "step": 13376 + }, + { + "epoch": 1.4146573604060912, + "grad_norm": 8.5, + "learning_rate": 4.207101036908176e-06, + "loss": 2.6332, + "step": 13377 + }, + { + "epoch": 1.4147631133671743, + "grad_norm": 8.75, + "learning_rate": 4.206975918661115e-06, + "loss": 2.6329, + "step": 13378 + }, + { + "epoch": 1.4148688663282571, + "grad_norm": 8.125, + "learning_rate": 4.206850792403916e-06, + "loss": 2.3363, + "step": 13379 + }, + { + "epoch": 1.4149746192893402, + "grad_norm": 8.625, + "learning_rate": 4.206725658137164e-06, + "loss": 2.6459, + "step": 13380 + }, + { + "epoch": 1.415080372250423, + "grad_norm": 8.25, + "learning_rate": 4.206600515861446e-06, + "loss": 2.5554, + "step": 13381 + }, + { + "epoch": 1.4151861252115059, + "grad_norm": 8.25, + "learning_rate": 4.206475365577349e-06, + "loss": 2.251, + "step": 13382 + }, + { + "epoch": 1.4152918781725887, + "grad_norm": 7.875, + "learning_rate": 4.206350207285463e-06, + "loss": 2.511, + "step": 13383 + }, + { + "epoch": 1.4153976311336718, + "grad_norm": 8.375, + "learning_rate": 4.206225040986372e-06, + "loss": 2.3829, + "step": 13384 + }, + { + "epoch": 1.4155033840947546, + "grad_norm": 7.875, + "learning_rate": 4.206099866680665e-06, + "loss": 2.5332, + "step": 13385 + }, + { + "epoch": 1.4156091370558377, + "grad_norm": 8.8125, + "learning_rate": 4.2059746843689296e-06, + "loss": 2.4692, + "step": 13386 + }, + { + "epoch": 1.4157148900169205, + "grad_norm": 7.75, + "learning_rate": 4.205849494051753e-06, + "loss": 2.4886, + "step": 13387 + }, + { + "epoch": 1.4158206429780034, + "grad_norm": 8.5, + "learning_rate": 4.205724295729721e-06, + "loss": 2.2589, + "step": 13388 + }, + { + "epoch": 1.4159263959390862, + "grad_norm": 8.25, + "learning_rate": 4.205599089403424e-06, + "loss": 2.4778, + "step": 13389 + }, + { + "epoch": 1.4160321489001693, + "grad_norm": 8.5, + "learning_rate": 4.205473875073446e-06, + "loss": 2.6185, + "step": 13390 + }, + { + "epoch": 1.416137901861252, + "grad_norm": 7.46875, + "learning_rate": 4.205348652740378e-06, + "loss": 2.5546, + "step": 13391 + }, + { + "epoch": 1.4162436548223352, + "grad_norm": 8.3125, + "learning_rate": 4.205223422404806e-06, + "loss": 2.7934, + "step": 13392 + }, + { + "epoch": 1.416349407783418, + "grad_norm": 8.0625, + "learning_rate": 4.205098184067318e-06, + "loss": 2.4625, + "step": 13393 + }, + { + "epoch": 1.4164551607445008, + "grad_norm": 8.125, + "learning_rate": 4.204972937728502e-06, + "loss": 2.4849, + "step": 13394 + }, + { + "epoch": 1.4165609137055837, + "grad_norm": 8.125, + "learning_rate": 4.204847683388945e-06, + "loss": 2.5743, + "step": 13395 + }, + { + "epoch": 1.4166666666666667, + "grad_norm": 8.6875, + "learning_rate": 4.204722421049235e-06, + "loss": 2.6132, + "step": 13396 + }, + { + "epoch": 1.4167724196277496, + "grad_norm": 8.4375, + "learning_rate": 4.20459715070996e-06, + "loss": 2.6314, + "step": 13397 + }, + { + "epoch": 1.4168781725888324, + "grad_norm": 7.84375, + "learning_rate": 4.2044718723717075e-06, + "loss": 2.5058, + "step": 13398 + }, + { + "epoch": 1.4169839255499155, + "grad_norm": 8.125, + "learning_rate": 4.204346586035065e-06, + "loss": 2.4085, + "step": 13399 + }, + { + "epoch": 1.4170896785109983, + "grad_norm": 8.625, + "learning_rate": 4.204221291700622e-06, + "loss": 2.4183, + "step": 13400 + }, + { + "epoch": 1.4171954314720812, + "grad_norm": 8.3125, + "learning_rate": 4.204095989368966e-06, + "loss": 2.3547, + "step": 13401 + }, + { + "epoch": 1.417301184433164, + "grad_norm": 8.125, + "learning_rate": 4.203970679040683e-06, + "loss": 2.7165, + "step": 13402 + }, + { + "epoch": 1.417406937394247, + "grad_norm": 8.5, + "learning_rate": 4.203845360716363e-06, + "loss": 2.7496, + "step": 13403 + }, + { + "epoch": 1.41751269035533, + "grad_norm": 7.9375, + "learning_rate": 4.203720034396594e-06, + "loss": 2.6946, + "step": 13404 + }, + { + "epoch": 1.417618443316413, + "grad_norm": 8.0625, + "learning_rate": 4.203594700081963e-06, + "loss": 2.3919, + "step": 13405 + }, + { + "epoch": 1.4177241962774958, + "grad_norm": 8.5, + "learning_rate": 4.203469357773059e-06, + "loss": 2.4449, + "step": 13406 + }, + { + "epoch": 1.4178299492385786, + "grad_norm": 8.1875, + "learning_rate": 4.203344007470471e-06, + "loss": 2.4751, + "step": 13407 + }, + { + "epoch": 1.4179357021996615, + "grad_norm": 8.3125, + "learning_rate": 4.203218649174785e-06, + "loss": 2.5286, + "step": 13408 + }, + { + "epoch": 1.4180414551607445, + "grad_norm": 8.9375, + "learning_rate": 4.2030932828865915e-06, + "loss": 2.7897, + "step": 13409 + }, + { + "epoch": 1.4181472081218274, + "grad_norm": 8.5, + "learning_rate": 4.202967908606477e-06, + "loss": 2.6335, + "step": 13410 + }, + { + "epoch": 1.4182529610829104, + "grad_norm": 8.6875, + "learning_rate": 4.2028425263350305e-06, + "loss": 2.5358, + "step": 13411 + }, + { + "epoch": 1.4183587140439933, + "grad_norm": 8.5625, + "learning_rate": 4.202717136072841e-06, + "loss": 2.4468, + "step": 13412 + }, + { + "epoch": 1.4184644670050761, + "grad_norm": 7.90625, + "learning_rate": 4.2025917378204965e-06, + "loss": 2.5308, + "step": 13413 + }, + { + "epoch": 1.418570219966159, + "grad_norm": 7.90625, + "learning_rate": 4.2024663315785844e-06, + "loss": 2.5481, + "step": 13414 + }, + { + "epoch": 1.418675972927242, + "grad_norm": 8.0, + "learning_rate": 4.202340917347695e-06, + "loss": 2.3826, + "step": 13415 + }, + { + "epoch": 1.4187817258883249, + "grad_norm": 8.0625, + "learning_rate": 4.2022154951284155e-06, + "loss": 2.4915, + "step": 13416 + }, + { + "epoch": 1.418887478849408, + "grad_norm": 8.125, + "learning_rate": 4.202090064921335e-06, + "loss": 2.4961, + "step": 13417 + }, + { + "epoch": 1.4189932318104908, + "grad_norm": 7.75, + "learning_rate": 4.201964626727041e-06, + "loss": 2.5116, + "step": 13418 + }, + { + "epoch": 1.4190989847715736, + "grad_norm": 8.4375, + "learning_rate": 4.201839180546125e-06, + "loss": 2.7502, + "step": 13419 + }, + { + "epoch": 1.4192047377326564, + "grad_norm": 8.5, + "learning_rate": 4.201713726379172e-06, + "loss": 2.6958, + "step": 13420 + }, + { + "epoch": 1.4193104906937395, + "grad_norm": 8.1875, + "learning_rate": 4.201588264226774e-06, + "loss": 2.387, + "step": 13421 + }, + { + "epoch": 1.4194162436548223, + "grad_norm": 8.5, + "learning_rate": 4.201462794089517e-06, + "loss": 2.6554, + "step": 13422 + }, + { + "epoch": 1.4195219966159052, + "grad_norm": 8.3125, + "learning_rate": 4.201337315967991e-06, + "loss": 2.6858, + "step": 13423 + }, + { + "epoch": 1.4196277495769882, + "grad_norm": 8.3125, + "learning_rate": 4.201211829862785e-06, + "loss": 2.5569, + "step": 13424 + }, + { + "epoch": 1.419733502538071, + "grad_norm": 8.0625, + "learning_rate": 4.201086335774488e-06, + "loss": 2.6652, + "step": 13425 + }, + { + "epoch": 1.419839255499154, + "grad_norm": 8.3125, + "learning_rate": 4.2009608337036885e-06, + "loss": 2.5346, + "step": 13426 + }, + { + "epoch": 1.419945008460237, + "grad_norm": 9.6875, + "learning_rate": 4.2008353236509745e-06, + "loss": 2.6942, + "step": 13427 + }, + { + "epoch": 1.4200507614213198, + "grad_norm": 8.1875, + "learning_rate": 4.2007098056169375e-06, + "loss": 2.4652, + "step": 13428 + }, + { + "epoch": 1.4201565143824026, + "grad_norm": 8.125, + "learning_rate": 4.200584279602164e-06, + "loss": 2.5999, + "step": 13429 + }, + { + "epoch": 1.4202622673434857, + "grad_norm": 8.6875, + "learning_rate": 4.200458745607244e-06, + "loss": 2.5133, + "step": 13430 + }, + { + "epoch": 1.4203680203045685, + "grad_norm": 8.25, + "learning_rate": 4.200333203632767e-06, + "loss": 2.608, + "step": 13431 + }, + { + "epoch": 1.4204737732656514, + "grad_norm": 8.125, + "learning_rate": 4.200207653679321e-06, + "loss": 2.5052, + "step": 13432 + }, + { + "epoch": 1.4205795262267342, + "grad_norm": 8.1875, + "learning_rate": 4.200082095747496e-06, + "loss": 2.4828, + "step": 13433 + }, + { + "epoch": 1.4206852791878173, + "grad_norm": 8.25, + "learning_rate": 4.199956529837882e-06, + "loss": 2.4877, + "step": 13434 + }, + { + "epoch": 1.4207910321489001, + "grad_norm": 7.71875, + "learning_rate": 4.199830955951066e-06, + "loss": 2.5151, + "step": 13435 + }, + { + "epoch": 1.4208967851099832, + "grad_norm": 8.0625, + "learning_rate": 4.19970537408764e-06, + "loss": 2.7052, + "step": 13436 + }, + { + "epoch": 1.421002538071066, + "grad_norm": 8.125, + "learning_rate": 4.199579784248191e-06, + "loss": 2.4341, + "step": 13437 + }, + { + "epoch": 1.4211082910321489, + "grad_norm": 8.5, + "learning_rate": 4.199454186433309e-06, + "loss": 2.5579, + "step": 13438 + }, + { + "epoch": 1.4212140439932317, + "grad_norm": 8.375, + "learning_rate": 4.199328580643584e-06, + "loss": 2.4675, + "step": 13439 + }, + { + "epoch": 1.4213197969543148, + "grad_norm": 7.46875, + "learning_rate": 4.199202966879604e-06, + "loss": 2.3499, + "step": 13440 + }, + { + "epoch": 1.4214255499153976, + "grad_norm": 8.375, + "learning_rate": 4.199077345141961e-06, + "loss": 2.4006, + "step": 13441 + }, + { + "epoch": 1.4215313028764807, + "grad_norm": 10.0, + "learning_rate": 4.198951715431242e-06, + "loss": 2.5806, + "step": 13442 + }, + { + "epoch": 1.4216370558375635, + "grad_norm": 8.5, + "learning_rate": 4.198826077748037e-06, + "loss": 2.7331, + "step": 13443 + }, + { + "epoch": 1.4217428087986463, + "grad_norm": 8.4375, + "learning_rate": 4.198700432092936e-06, + "loss": 2.5689, + "step": 13444 + }, + { + "epoch": 1.4218485617597292, + "grad_norm": 7.875, + "learning_rate": 4.198574778466529e-06, + "loss": 2.3493, + "step": 13445 + }, + { + "epoch": 1.4219543147208122, + "grad_norm": 7.875, + "learning_rate": 4.198449116869405e-06, + "loss": 2.577, + "step": 13446 + }, + { + "epoch": 1.422060067681895, + "grad_norm": 8.125, + "learning_rate": 4.1983234473021554e-06, + "loss": 2.54, + "step": 13447 + }, + { + "epoch": 1.4221658206429781, + "grad_norm": 8.25, + "learning_rate": 4.198197769765367e-06, + "loss": 2.4528, + "step": 13448 + }, + { + "epoch": 1.422271573604061, + "grad_norm": 8.9375, + "learning_rate": 4.198072084259631e-06, + "loss": 2.7341, + "step": 13449 + }, + { + "epoch": 1.4223773265651438, + "grad_norm": 8.75, + "learning_rate": 4.1979463907855375e-06, + "loss": 2.5971, + "step": 13450 + }, + { + "epoch": 1.4224830795262267, + "grad_norm": 8.0625, + "learning_rate": 4.197820689343677e-06, + "loss": 2.4778, + "step": 13451 + }, + { + "epoch": 1.4225888324873097, + "grad_norm": 8.1875, + "learning_rate": 4.197694979934637e-06, + "loss": 2.5926, + "step": 13452 + }, + { + "epoch": 1.4226945854483926, + "grad_norm": 8.375, + "learning_rate": 4.1975692625590095e-06, + "loss": 2.3953, + "step": 13453 + }, + { + "epoch": 1.4228003384094754, + "grad_norm": 8.3125, + "learning_rate": 4.197443537217383e-06, + "loss": 2.3516, + "step": 13454 + }, + { + "epoch": 1.4229060913705585, + "grad_norm": 8.375, + "learning_rate": 4.1973178039103494e-06, + "loss": 2.4719, + "step": 13455 + }, + { + "epoch": 1.4230118443316413, + "grad_norm": 8.0625, + "learning_rate": 4.197192062638496e-06, + "loss": 2.3161, + "step": 13456 + }, + { + "epoch": 1.4231175972927241, + "grad_norm": 8.0, + "learning_rate": 4.1970663134024144e-06, + "loss": 2.5134, + "step": 13457 + }, + { + "epoch": 1.423223350253807, + "grad_norm": 8.1875, + "learning_rate": 4.196940556202696e-06, + "loss": 2.5816, + "step": 13458 + }, + { + "epoch": 1.42332910321489, + "grad_norm": 8.0, + "learning_rate": 4.196814791039929e-06, + "loss": 2.4746, + "step": 13459 + }, + { + "epoch": 1.4234348561759729, + "grad_norm": 8.25, + "learning_rate": 4.196689017914704e-06, + "loss": 2.5301, + "step": 13460 + }, + { + "epoch": 1.423540609137056, + "grad_norm": 8.1875, + "learning_rate": 4.196563236827612e-06, + "loss": 2.5764, + "step": 13461 + }, + { + "epoch": 1.4236463620981388, + "grad_norm": 8.5625, + "learning_rate": 4.1964374477792425e-06, + "loss": 2.5407, + "step": 13462 + }, + { + "epoch": 1.4237521150592216, + "grad_norm": 8.1875, + "learning_rate": 4.196311650770185e-06, + "loss": 2.7493, + "step": 13463 + }, + { + "epoch": 1.4238578680203045, + "grad_norm": 7.8125, + "learning_rate": 4.196185845801032e-06, + "loss": 2.5681, + "step": 13464 + }, + { + "epoch": 1.4239636209813875, + "grad_norm": 8.0, + "learning_rate": 4.196060032872371e-06, + "loss": 2.5453, + "step": 13465 + }, + { + "epoch": 1.4240693739424704, + "grad_norm": 8.0625, + "learning_rate": 4.195934211984795e-06, + "loss": 2.5693, + "step": 13466 + }, + { + "epoch": 1.4241751269035534, + "grad_norm": 8.3125, + "learning_rate": 4.195808383138893e-06, + "loss": 2.5584, + "step": 13467 + }, + { + "epoch": 1.4242808798646363, + "grad_norm": 8.5625, + "learning_rate": 4.195682546335256e-06, + "loss": 2.5281, + "step": 13468 + }, + { + "epoch": 1.424386632825719, + "grad_norm": 8.6875, + "learning_rate": 4.195556701574474e-06, + "loss": 2.5929, + "step": 13469 + }, + { + "epoch": 1.424492385786802, + "grad_norm": 8.375, + "learning_rate": 4.195430848857139e-06, + "loss": 2.4696, + "step": 13470 + }, + { + "epoch": 1.424598138747885, + "grad_norm": 8.125, + "learning_rate": 4.1953049881838385e-06, + "loss": 2.7297, + "step": 13471 + }, + { + "epoch": 1.4247038917089678, + "grad_norm": 7.75, + "learning_rate": 4.195179119555166e-06, + "loss": 2.4753, + "step": 13472 + }, + { + "epoch": 1.424809644670051, + "grad_norm": 8.8125, + "learning_rate": 4.195053242971711e-06, + "loss": 2.6001, + "step": 13473 + }, + { + "epoch": 1.4249153976311337, + "grad_norm": 8.8125, + "learning_rate": 4.194927358434065e-06, + "loss": 2.6834, + "step": 13474 + }, + { + "epoch": 1.4250211505922166, + "grad_norm": 8.75, + "learning_rate": 4.194801465942819e-06, + "loss": 2.7273, + "step": 13475 + }, + { + "epoch": 1.4251269035532994, + "grad_norm": 7.96875, + "learning_rate": 4.19467556549856e-06, + "loss": 2.546, + "step": 13476 + }, + { + "epoch": 1.4252326565143825, + "grad_norm": 8.5, + "learning_rate": 4.194549657101884e-06, + "loss": 2.5751, + "step": 13477 + }, + { + "epoch": 1.4253384094754653, + "grad_norm": 8.0, + "learning_rate": 4.194423740753379e-06, + "loss": 2.4534, + "step": 13478 + }, + { + "epoch": 1.4254441624365481, + "grad_norm": 7.84375, + "learning_rate": 4.1942978164536365e-06, + "loss": 2.4935, + "step": 13479 + }, + { + "epoch": 1.4255499153976312, + "grad_norm": 9.0625, + "learning_rate": 4.194171884203246e-06, + "loss": 2.7956, + "step": 13480 + }, + { + "epoch": 1.425655668358714, + "grad_norm": 8.4375, + "learning_rate": 4.194045944002801e-06, + "loss": 2.5421, + "step": 13481 + }, + { + "epoch": 1.4257614213197969, + "grad_norm": 8.25, + "learning_rate": 4.193919995852891e-06, + "loss": 2.5583, + "step": 13482 + }, + { + "epoch": 1.42586717428088, + "grad_norm": 9.1875, + "learning_rate": 4.193794039754107e-06, + "loss": 2.657, + "step": 13483 + }, + { + "epoch": 1.4259729272419628, + "grad_norm": 8.0625, + "learning_rate": 4.19366807570704e-06, + "loss": 2.4276, + "step": 13484 + }, + { + "epoch": 1.4260786802030456, + "grad_norm": 9.125, + "learning_rate": 4.193542103712282e-06, + "loss": 2.5794, + "step": 13485 + }, + { + "epoch": 1.4261844331641287, + "grad_norm": 8.3125, + "learning_rate": 4.193416123770424e-06, + "loss": 2.6463, + "step": 13486 + }, + { + "epoch": 1.4262901861252115, + "grad_norm": 8.125, + "learning_rate": 4.193290135882056e-06, + "loss": 2.5792, + "step": 13487 + }, + { + "epoch": 1.4263959390862944, + "grad_norm": 7.875, + "learning_rate": 4.19316414004777e-06, + "loss": 2.4334, + "step": 13488 + }, + { + "epoch": 1.4265016920473772, + "grad_norm": 8.4375, + "learning_rate": 4.193038136268157e-06, + "loss": 2.4799, + "step": 13489 + }, + { + "epoch": 1.4266074450084603, + "grad_norm": 8.4375, + "learning_rate": 4.192912124543808e-06, + "loss": 2.5491, + "step": 13490 + }, + { + "epoch": 1.426713197969543, + "grad_norm": 8.5, + "learning_rate": 4.1927861048753165e-06, + "loss": 2.2911, + "step": 13491 + }, + { + "epoch": 1.4268189509306262, + "grad_norm": 8.0625, + "learning_rate": 4.192660077263271e-06, + "loss": 2.4517, + "step": 13492 + }, + { + "epoch": 1.426924703891709, + "grad_norm": 8.375, + "learning_rate": 4.192534041708265e-06, + "loss": 2.4233, + "step": 13493 + }, + { + "epoch": 1.4270304568527918, + "grad_norm": 8.0625, + "learning_rate": 4.192407998210887e-06, + "loss": 2.5884, + "step": 13494 + }, + { + "epoch": 1.4271362098138747, + "grad_norm": 8.375, + "learning_rate": 4.192281946771732e-06, + "loss": 2.7648, + "step": 13495 + }, + { + "epoch": 1.4272419627749577, + "grad_norm": 8.3125, + "learning_rate": 4.192155887391389e-06, + "loss": 2.197, + "step": 13496 + }, + { + "epoch": 1.4273477157360406, + "grad_norm": 8.375, + "learning_rate": 4.192029820070451e-06, + "loss": 2.5571, + "step": 13497 + }, + { + "epoch": 1.4274534686971236, + "grad_norm": 8.0625, + "learning_rate": 4.191903744809509e-06, + "loss": 2.514, + "step": 13498 + }, + { + "epoch": 1.4275592216582065, + "grad_norm": 8.875, + "learning_rate": 4.191777661609155e-06, + "loss": 2.3671, + "step": 13499 + }, + { + "epoch": 1.4276649746192893, + "grad_norm": 8.125, + "learning_rate": 4.19165157046998e-06, + "loss": 2.4878, + "step": 13500 + }, + { + "epoch": 1.4277707275803722, + "grad_norm": 8.6875, + "learning_rate": 4.191525471392576e-06, + "loss": 2.4482, + "step": 13501 + }, + { + "epoch": 1.4278764805414552, + "grad_norm": 7.78125, + "learning_rate": 4.191399364377535e-06, + "loss": 2.3105, + "step": 13502 + }, + { + "epoch": 1.427982233502538, + "grad_norm": 8.25, + "learning_rate": 4.191273249425448e-06, + "loss": 2.6474, + "step": 13503 + }, + { + "epoch": 1.4280879864636211, + "grad_norm": 8.0, + "learning_rate": 4.191147126536908e-06, + "loss": 2.395, + "step": 13504 + }, + { + "epoch": 1.428193739424704, + "grad_norm": 8.125, + "learning_rate": 4.191020995712507e-06, + "loss": 2.5235, + "step": 13505 + }, + { + "epoch": 1.4282994923857868, + "grad_norm": 8.3125, + "learning_rate": 4.190894856952834e-06, + "loss": 2.6154, + "step": 13506 + }, + { + "epoch": 1.4284052453468696, + "grad_norm": 8.625, + "learning_rate": 4.190768710258484e-06, + "loss": 2.5789, + "step": 13507 + }, + { + "epoch": 1.4285109983079527, + "grad_norm": 8.5, + "learning_rate": 4.190642555630048e-06, + "loss": 2.509, + "step": 13508 + }, + { + "epoch": 1.4286167512690355, + "grad_norm": 8.75, + "learning_rate": 4.190516393068118e-06, + "loss": 2.3977, + "step": 13509 + }, + { + "epoch": 1.4287225042301184, + "grad_norm": 9.875, + "learning_rate": 4.190390222573286e-06, + "loss": 2.6624, + "step": 13510 + }, + { + "epoch": 1.4288282571912014, + "grad_norm": 7.96875, + "learning_rate": 4.190264044146144e-06, + "loss": 2.4588, + "step": 13511 + }, + { + "epoch": 1.4289340101522843, + "grad_norm": 8.375, + "learning_rate": 4.1901378577872844e-06, + "loss": 2.7639, + "step": 13512 + }, + { + "epoch": 1.4290397631133671, + "grad_norm": 9.5625, + "learning_rate": 4.190011663497298e-06, + "loss": 2.4955, + "step": 13513 + }, + { + "epoch": 1.42914551607445, + "grad_norm": 8.5625, + "learning_rate": 4.189885461276779e-06, + "loss": 2.7689, + "step": 13514 + }, + { + "epoch": 1.429251269035533, + "grad_norm": 8.125, + "learning_rate": 4.189759251126319e-06, + "loss": 2.2384, + "step": 13515 + }, + { + "epoch": 1.4293570219966159, + "grad_norm": 8.625, + "learning_rate": 4.189633033046509e-06, + "loss": 2.6922, + "step": 13516 + }, + { + "epoch": 1.429462774957699, + "grad_norm": 8.0625, + "learning_rate": 4.189506807037943e-06, + "loss": 2.4363, + "step": 13517 + }, + { + "epoch": 1.4295685279187818, + "grad_norm": 8.5, + "learning_rate": 4.189380573101212e-06, + "loss": 2.7853, + "step": 13518 + }, + { + "epoch": 1.4296742808798646, + "grad_norm": 8.5, + "learning_rate": 4.189254331236909e-06, + "loss": 2.3202, + "step": 13519 + }, + { + "epoch": 1.4297800338409474, + "grad_norm": 8.125, + "learning_rate": 4.1891280814456265e-06, + "loss": 2.4263, + "step": 13520 + }, + { + "epoch": 1.4298857868020305, + "grad_norm": 8.3125, + "learning_rate": 4.189001823727956e-06, + "loss": 2.6505, + "step": 13521 + }, + { + "epoch": 1.4299915397631133, + "grad_norm": 8.375, + "learning_rate": 4.188875558084492e-06, + "loss": 2.308, + "step": 13522 + }, + { + "epoch": 1.4300972927241964, + "grad_norm": 7.6875, + "learning_rate": 4.188749284515825e-06, + "loss": 2.3762, + "step": 13523 + }, + { + "epoch": 1.4302030456852792, + "grad_norm": 8.4375, + "learning_rate": 4.188623003022548e-06, + "loss": 2.4971, + "step": 13524 + }, + { + "epoch": 1.430308798646362, + "grad_norm": 8.9375, + "learning_rate": 4.188496713605254e-06, + "loss": 2.6935, + "step": 13525 + }, + { + "epoch": 1.430414551607445, + "grad_norm": 8.8125, + "learning_rate": 4.188370416264536e-06, + "loss": 2.628, + "step": 13526 + }, + { + "epoch": 1.430520304568528, + "grad_norm": 8.0, + "learning_rate": 4.1882441110009855e-06, + "loss": 2.3406, + "step": 13527 + }, + { + "epoch": 1.4306260575296108, + "grad_norm": 8.375, + "learning_rate": 4.188117797815196e-06, + "loss": 2.5018, + "step": 13528 + }, + { + "epoch": 1.4307318104906939, + "grad_norm": 8.125, + "learning_rate": 4.187991476707761e-06, + "loss": 2.5092, + "step": 13529 + }, + { + "epoch": 1.4308375634517767, + "grad_norm": 9.0625, + "learning_rate": 4.1878651476792705e-06, + "loss": 2.6486, + "step": 13530 + }, + { + "epoch": 1.4309433164128595, + "grad_norm": 8.25, + "learning_rate": 4.187738810730322e-06, + "loss": 2.7037, + "step": 13531 + }, + { + "epoch": 1.4310490693739424, + "grad_norm": 8.0625, + "learning_rate": 4.187612465861503e-06, + "loss": 2.547, + "step": 13532 + }, + { + "epoch": 1.4311548223350254, + "grad_norm": 7.9375, + "learning_rate": 4.18748611307341e-06, + "loss": 2.6146, + "step": 13533 + }, + { + "epoch": 1.4312605752961083, + "grad_norm": 8.125, + "learning_rate": 4.1873597523666344e-06, + "loss": 2.5582, + "step": 13534 + }, + { + "epoch": 1.4313663282571911, + "grad_norm": 8.6875, + "learning_rate": 4.18723338374177e-06, + "loss": 2.3422, + "step": 13535 + }, + { + "epoch": 1.4314720812182742, + "grad_norm": 8.1875, + "learning_rate": 4.187107007199409e-06, + "loss": 2.6688, + "step": 13536 + }, + { + "epoch": 1.431577834179357, + "grad_norm": 8.5, + "learning_rate": 4.186980622740144e-06, + "loss": 2.2893, + "step": 13537 + }, + { + "epoch": 1.4316835871404399, + "grad_norm": 8.25, + "learning_rate": 4.1868542303645705e-06, + "loss": 2.4161, + "step": 13538 + }, + { + "epoch": 1.431789340101523, + "grad_norm": 9.125, + "learning_rate": 4.18672783007328e-06, + "loss": 2.4678, + "step": 13539 + }, + { + "epoch": 1.4318950930626058, + "grad_norm": 8.25, + "learning_rate": 4.186601421866864e-06, + "loss": 2.6509, + "step": 13540 + }, + { + "epoch": 1.4320008460236886, + "grad_norm": 8.3125, + "learning_rate": 4.186475005745919e-06, + "loss": 2.5306, + "step": 13541 + }, + { + "epoch": 1.4321065989847717, + "grad_norm": 8.3125, + "learning_rate": 4.186348581711035e-06, + "loss": 2.7555, + "step": 13542 + }, + { + "epoch": 1.4322123519458545, + "grad_norm": 9.0, + "learning_rate": 4.186222149762809e-06, + "loss": 2.3561, + "step": 13543 + }, + { + "epoch": 1.4323181049069373, + "grad_norm": 8.25, + "learning_rate": 4.186095709901829e-06, + "loss": 2.5056, + "step": 13544 + }, + { + "epoch": 1.4324238578680202, + "grad_norm": 8.3125, + "learning_rate": 4.185969262128694e-06, + "loss": 2.6365, + "step": 13545 + }, + { + "epoch": 1.4325296108291032, + "grad_norm": 8.125, + "learning_rate": 4.185842806443994e-06, + "loss": 2.5467, + "step": 13546 + }, + { + "epoch": 1.432635363790186, + "grad_norm": 9.0, + "learning_rate": 4.185716342848324e-06, + "loss": 2.581, + "step": 13547 + }, + { + "epoch": 1.4327411167512691, + "grad_norm": 8.5625, + "learning_rate": 4.1855898713422746e-06, + "loss": 2.6422, + "step": 13548 + }, + { + "epoch": 1.432846869712352, + "grad_norm": 8.0625, + "learning_rate": 4.185463391926443e-06, + "loss": 2.6029, + "step": 13549 + }, + { + "epoch": 1.4329526226734348, + "grad_norm": 7.75, + "learning_rate": 4.18533690460142e-06, + "loss": 2.494, + "step": 13550 + }, + { + "epoch": 1.4330583756345177, + "grad_norm": 8.5625, + "learning_rate": 4.185210409367801e-06, + "loss": 2.3594, + "step": 13551 + }, + { + "epoch": 1.4331641285956007, + "grad_norm": 8.375, + "learning_rate": 4.185083906226178e-06, + "loss": 2.317, + "step": 13552 + }, + { + "epoch": 1.4332698815566836, + "grad_norm": 8.8125, + "learning_rate": 4.184957395177146e-06, + "loss": 2.6396, + "step": 13553 + }, + { + "epoch": 1.4333756345177666, + "grad_norm": 8.3125, + "learning_rate": 4.184830876221298e-06, + "loss": 2.5989, + "step": 13554 + }, + { + "epoch": 1.4334813874788495, + "grad_norm": 8.375, + "learning_rate": 4.184704349359229e-06, + "loss": 2.7911, + "step": 13555 + }, + { + "epoch": 1.4335871404399323, + "grad_norm": 8.9375, + "learning_rate": 4.18457781459153e-06, + "loss": 2.599, + "step": 13556 + }, + { + "epoch": 1.4336928934010151, + "grad_norm": 8.3125, + "learning_rate": 4.184451271918796e-06, + "loss": 2.4464, + "step": 13557 + }, + { + "epoch": 1.4337986463620982, + "grad_norm": 8.3125, + "learning_rate": 4.184324721341622e-06, + "loss": 2.6238, + "step": 13558 + }, + { + "epoch": 1.433904399323181, + "grad_norm": 8.9375, + "learning_rate": 4.1841981628606e-06, + "loss": 2.7207, + "step": 13559 + }, + { + "epoch": 1.434010152284264, + "grad_norm": 8.75, + "learning_rate": 4.184071596476326e-06, + "loss": 2.5484, + "step": 13560 + }, + { + "epoch": 1.434115905245347, + "grad_norm": 8.1875, + "learning_rate": 4.183945022189391e-06, + "loss": 2.5709, + "step": 13561 + }, + { + "epoch": 1.4342216582064298, + "grad_norm": 8.4375, + "learning_rate": 4.183818440000393e-06, + "loss": 2.5828, + "step": 13562 + }, + { + "epoch": 1.4343274111675126, + "grad_norm": 8.625, + "learning_rate": 4.1836918499099224e-06, + "loss": 2.7089, + "step": 13563 + }, + { + "epoch": 1.4344331641285957, + "grad_norm": 9.0, + "learning_rate": 4.1835652519185745e-06, + "loss": 2.6968, + "step": 13564 + }, + { + "epoch": 1.4345389170896785, + "grad_norm": 8.0, + "learning_rate": 4.1834386460269435e-06, + "loss": 2.3769, + "step": 13565 + }, + { + "epoch": 1.4346446700507614, + "grad_norm": 8.3125, + "learning_rate": 4.183312032235625e-06, + "loss": 2.704, + "step": 13566 + }, + { + "epoch": 1.4347504230118444, + "grad_norm": 8.3125, + "learning_rate": 4.18318541054521e-06, + "loss": 2.5533, + "step": 13567 + }, + { + "epoch": 1.4348561759729273, + "grad_norm": 8.0, + "learning_rate": 4.183058780956294e-06, + "loss": 2.6281, + "step": 13568 + }, + { + "epoch": 1.43496192893401, + "grad_norm": 8.1875, + "learning_rate": 4.182932143469472e-06, + "loss": 2.619, + "step": 13569 + }, + { + "epoch": 1.435067681895093, + "grad_norm": 8.9375, + "learning_rate": 4.182805498085338e-06, + "loss": 2.629, + "step": 13570 + }, + { + "epoch": 1.435173434856176, + "grad_norm": 8.5, + "learning_rate": 4.182678844804485e-06, + "loss": 2.582, + "step": 13571 + }, + { + "epoch": 1.4352791878172588, + "grad_norm": 7.78125, + "learning_rate": 4.182552183627509e-06, + "loss": 2.4839, + "step": 13572 + }, + { + "epoch": 1.435384940778342, + "grad_norm": 8.5, + "learning_rate": 4.182425514555005e-06, + "loss": 2.7109, + "step": 13573 + }, + { + "epoch": 1.4354906937394247, + "grad_norm": 9.1875, + "learning_rate": 4.182298837587565e-06, + "loss": 2.7269, + "step": 13574 + }, + { + "epoch": 1.4355964467005076, + "grad_norm": 8.0625, + "learning_rate": 4.182172152725784e-06, + "loss": 2.6463, + "step": 13575 + }, + { + "epoch": 1.4357021996615904, + "grad_norm": 8.3125, + "learning_rate": 4.182045459970258e-06, + "loss": 2.6228, + "step": 13576 + }, + { + "epoch": 1.4358079526226735, + "grad_norm": 8.0625, + "learning_rate": 4.181918759321581e-06, + "loss": 2.5044, + "step": 13577 + }, + { + "epoch": 1.4359137055837563, + "grad_norm": 8.0625, + "learning_rate": 4.181792050780345e-06, + "loss": 2.7765, + "step": 13578 + }, + { + "epoch": 1.4360194585448394, + "grad_norm": 7.8125, + "learning_rate": 4.181665334347149e-06, + "loss": 2.5696, + "step": 13579 + }, + { + "epoch": 1.4361252115059222, + "grad_norm": 8.125, + "learning_rate": 4.181538610022584e-06, + "loss": 2.5847, + "step": 13580 + }, + { + "epoch": 1.436230964467005, + "grad_norm": 8.8125, + "learning_rate": 4.181411877807246e-06, + "loss": 2.4109, + "step": 13581 + }, + { + "epoch": 1.4363367174280879, + "grad_norm": 8.375, + "learning_rate": 4.181285137701732e-06, + "loss": 2.4499, + "step": 13582 + }, + { + "epoch": 1.436442470389171, + "grad_norm": 8.25, + "learning_rate": 4.181158389706632e-06, + "loss": 2.7757, + "step": 13583 + }, + { + "epoch": 1.4365482233502538, + "grad_norm": 8.125, + "learning_rate": 4.1810316338225445e-06, + "loss": 2.4363, + "step": 13584 + }, + { + "epoch": 1.4366539763113368, + "grad_norm": 8.25, + "learning_rate": 4.180904870050062e-06, + "loss": 2.5615, + "step": 13585 + }, + { + "epoch": 1.4367597292724197, + "grad_norm": 7.53125, + "learning_rate": 4.1807780983897815e-06, + "loss": 2.4338, + "step": 13586 + }, + { + "epoch": 1.4368654822335025, + "grad_norm": 8.1875, + "learning_rate": 4.180651318842296e-06, + "loss": 2.3019, + "step": 13587 + }, + { + "epoch": 1.4369712351945854, + "grad_norm": 8.625, + "learning_rate": 4.180524531408201e-06, + "loss": 2.5695, + "step": 13588 + }, + { + "epoch": 1.4370769881556684, + "grad_norm": 8.375, + "learning_rate": 4.180397736088092e-06, + "loss": 2.502, + "step": 13589 + }, + { + "epoch": 1.4371827411167513, + "grad_norm": 8.3125, + "learning_rate": 4.180270932882564e-06, + "loss": 2.5414, + "step": 13590 + }, + { + "epoch": 1.437288494077834, + "grad_norm": 7.9375, + "learning_rate": 4.180144121792212e-06, + "loss": 2.4004, + "step": 13591 + }, + { + "epoch": 1.4373942470389172, + "grad_norm": 8.25, + "learning_rate": 4.180017302817631e-06, + "loss": 2.6475, + "step": 13592 + }, + { + "epoch": 1.4375, + "grad_norm": 8.6875, + "learning_rate": 4.179890475959415e-06, + "loss": 2.5931, + "step": 13593 + }, + { + "epoch": 1.4376057529610828, + "grad_norm": 8.375, + "learning_rate": 4.17976364121816e-06, + "loss": 2.6624, + "step": 13594 + }, + { + "epoch": 1.437711505922166, + "grad_norm": 8.0625, + "learning_rate": 4.179636798594462e-06, + "loss": 2.3982, + "step": 13595 + }, + { + "epoch": 1.4378172588832487, + "grad_norm": 8.625, + "learning_rate": 4.179509948088914e-06, + "loss": 2.6347, + "step": 13596 + }, + { + "epoch": 1.4379230118443316, + "grad_norm": 8.9375, + "learning_rate": 4.179383089702115e-06, + "loss": 2.7355, + "step": 13597 + }, + { + "epoch": 1.4380287648054146, + "grad_norm": 8.3125, + "learning_rate": 4.179256223434656e-06, + "loss": 2.7108, + "step": 13598 + }, + { + "epoch": 1.4381345177664975, + "grad_norm": 8.25, + "learning_rate": 4.179129349287136e-06, + "loss": 2.4952, + "step": 13599 + }, + { + "epoch": 1.4382402707275803, + "grad_norm": 8.125, + "learning_rate": 4.179002467260147e-06, + "loss": 2.5674, + "step": 13600 + }, + { + "epoch": 1.4383460236886632, + "grad_norm": 8.875, + "learning_rate": 4.178875577354287e-06, + "loss": 2.6661, + "step": 13601 + }, + { + "epoch": 1.4384517766497462, + "grad_norm": 8.6875, + "learning_rate": 4.17874867957015e-06, + "loss": 2.753, + "step": 13602 + }, + { + "epoch": 1.438557529610829, + "grad_norm": 8.25, + "learning_rate": 4.178621773908332e-06, + "loss": 2.6076, + "step": 13603 + }, + { + "epoch": 1.4386632825719121, + "grad_norm": 7.6875, + "learning_rate": 4.1784948603694295e-06, + "loss": 2.5314, + "step": 13604 + }, + { + "epoch": 1.438769035532995, + "grad_norm": 8.75, + "learning_rate": 4.178367938954037e-06, + "loss": 2.5535, + "step": 13605 + }, + { + "epoch": 1.4388747884940778, + "grad_norm": 7.71875, + "learning_rate": 4.17824100966275e-06, + "loss": 2.4175, + "step": 13606 + }, + { + "epoch": 1.4389805414551606, + "grad_norm": 8.3125, + "learning_rate": 4.178114072496164e-06, + "loss": 2.6447, + "step": 13607 + }, + { + "epoch": 1.4390862944162437, + "grad_norm": 8.25, + "learning_rate": 4.177987127454875e-06, + "loss": 2.5027, + "step": 13608 + }, + { + "epoch": 1.4391920473773265, + "grad_norm": 8.75, + "learning_rate": 4.177860174539479e-06, + "loss": 2.7164, + "step": 13609 + }, + { + "epoch": 1.4392978003384096, + "grad_norm": 7.96875, + "learning_rate": 4.177733213750571e-06, + "loss": 2.4553, + "step": 13610 + }, + { + "epoch": 1.4394035532994924, + "grad_norm": 7.90625, + "learning_rate": 4.177606245088748e-06, + "loss": 2.4252, + "step": 13611 + }, + { + "epoch": 1.4395093062605753, + "grad_norm": 7.84375, + "learning_rate": 4.177479268554604e-06, + "loss": 2.6044, + "step": 13612 + }, + { + "epoch": 1.439615059221658, + "grad_norm": 8.1875, + "learning_rate": 4.177352284148736e-06, + "loss": 2.161, + "step": 13613 + }, + { + "epoch": 1.4397208121827412, + "grad_norm": 8.5625, + "learning_rate": 4.17722529187174e-06, + "loss": 2.5406, + "step": 13614 + }, + { + "epoch": 1.439826565143824, + "grad_norm": 9.3125, + "learning_rate": 4.177098291724212e-06, + "loss": 2.7437, + "step": 13615 + }, + { + "epoch": 1.439932318104907, + "grad_norm": 7.90625, + "learning_rate": 4.1769712837067475e-06, + "loss": 2.4546, + "step": 13616 + }, + { + "epoch": 1.44003807106599, + "grad_norm": 8.125, + "learning_rate": 4.1768442678199424e-06, + "loss": 2.3911, + "step": 13617 + }, + { + "epoch": 1.4401438240270727, + "grad_norm": 8.1875, + "learning_rate": 4.176717244064393e-06, + "loss": 2.5621, + "step": 13618 + }, + { + "epoch": 1.4402495769881556, + "grad_norm": 8.6875, + "learning_rate": 4.176590212440696e-06, + "loss": 2.7303, + "step": 13619 + }, + { + "epoch": 1.4403553299492386, + "grad_norm": 8.625, + "learning_rate": 4.1764631729494456e-06, + "loss": 2.5165, + "step": 13620 + }, + { + "epoch": 1.4404610829103215, + "grad_norm": 7.875, + "learning_rate": 4.176336125591239e-06, + "loss": 2.5988, + "step": 13621 + }, + { + "epoch": 1.4405668358714043, + "grad_norm": 8.25, + "learning_rate": 4.176209070366674e-06, + "loss": 2.4829, + "step": 13622 + }, + { + "epoch": 1.4406725888324874, + "grad_norm": 7.65625, + "learning_rate": 4.176082007276344e-06, + "loss": 2.3991, + "step": 13623 + }, + { + "epoch": 1.4407783417935702, + "grad_norm": 8.4375, + "learning_rate": 4.175954936320847e-06, + "loss": 2.5883, + "step": 13624 + }, + { + "epoch": 1.440884094754653, + "grad_norm": 8.0625, + "learning_rate": 4.17582785750078e-06, + "loss": 2.3374, + "step": 13625 + }, + { + "epoch": 1.440989847715736, + "grad_norm": 8.25, + "learning_rate": 4.175700770816738e-06, + "loss": 2.7393, + "step": 13626 + }, + { + "epoch": 1.441095600676819, + "grad_norm": 8.5, + "learning_rate": 4.175573676269317e-06, + "loss": 2.5904, + "step": 13627 + }, + { + "epoch": 1.4412013536379018, + "grad_norm": 8.75, + "learning_rate": 4.1754465738591135e-06, + "loss": 2.5779, + "step": 13628 + }, + { + "epoch": 1.4413071065989849, + "grad_norm": 8.375, + "learning_rate": 4.175319463586726e-06, + "loss": 2.5487, + "step": 13629 + }, + { + "epoch": 1.4414128595600677, + "grad_norm": 8.6875, + "learning_rate": 4.1751923454527475e-06, + "loss": 2.6058, + "step": 13630 + }, + { + "epoch": 1.4415186125211505, + "grad_norm": 8.625, + "learning_rate": 4.1750652194577775e-06, + "loss": 2.5987, + "step": 13631 + }, + { + "epoch": 1.4416243654822334, + "grad_norm": 7.59375, + "learning_rate": 4.174938085602412e-06, + "loss": 2.4277, + "step": 13632 + }, + { + "epoch": 1.4417301184433164, + "grad_norm": 8.0, + "learning_rate": 4.1748109438872465e-06, + "loss": 2.5511, + "step": 13633 + }, + { + "epoch": 1.4418358714043993, + "grad_norm": 8.0625, + "learning_rate": 4.174683794312878e-06, + "loss": 2.5018, + "step": 13634 + }, + { + "epoch": 1.4419416243654823, + "grad_norm": 7.96875, + "learning_rate": 4.174556636879904e-06, + "loss": 2.6365, + "step": 13635 + }, + { + "epoch": 1.4420473773265652, + "grad_norm": 8.125, + "learning_rate": 4.17442947158892e-06, + "loss": 2.6285, + "step": 13636 + }, + { + "epoch": 1.442153130287648, + "grad_norm": 8.375, + "learning_rate": 4.174302298440524e-06, + "loss": 2.6957, + "step": 13637 + }, + { + "epoch": 1.4422588832487309, + "grad_norm": 7.96875, + "learning_rate": 4.1741751174353114e-06, + "loss": 2.6363, + "step": 13638 + }, + { + "epoch": 1.442364636209814, + "grad_norm": 7.90625, + "learning_rate": 4.17404792857388e-06, + "loss": 2.4019, + "step": 13639 + }, + { + "epoch": 1.4424703891708968, + "grad_norm": 8.5625, + "learning_rate": 4.173920731856826e-06, + "loss": 2.7692, + "step": 13640 + }, + { + "epoch": 1.4425761421319798, + "grad_norm": 7.71875, + "learning_rate": 4.173793527284748e-06, + "loss": 2.3735, + "step": 13641 + }, + { + "epoch": 1.4426818950930627, + "grad_norm": 8.0, + "learning_rate": 4.173666314858239e-06, + "loss": 2.5339, + "step": 13642 + }, + { + "epoch": 1.4427876480541455, + "grad_norm": 7.75, + "learning_rate": 4.173539094577901e-06, + "loss": 2.5098, + "step": 13643 + }, + { + "epoch": 1.4428934010152283, + "grad_norm": 8.1875, + "learning_rate": 4.173411866444327e-06, + "loss": 2.4249, + "step": 13644 + }, + { + "epoch": 1.4429991539763114, + "grad_norm": 8.875, + "learning_rate": 4.173284630458115e-06, + "loss": 2.5381, + "step": 13645 + }, + { + "epoch": 1.4431049069373942, + "grad_norm": 8.375, + "learning_rate": 4.173157386619864e-06, + "loss": 2.4732, + "step": 13646 + }, + { + "epoch": 1.443210659898477, + "grad_norm": 8.4375, + "learning_rate": 4.173030134930169e-06, + "loss": 2.7901, + "step": 13647 + }, + { + "epoch": 1.4433164128595601, + "grad_norm": 8.0625, + "learning_rate": 4.172902875389628e-06, + "loss": 2.6575, + "step": 13648 + }, + { + "epoch": 1.443422165820643, + "grad_norm": 8.5625, + "learning_rate": 4.1727756079988376e-06, + "loss": 2.3792, + "step": 13649 + }, + { + "epoch": 1.4435279187817258, + "grad_norm": 9.125, + "learning_rate": 4.172648332758397e-06, + "loss": 2.5702, + "step": 13650 + }, + { + "epoch": 1.4436336717428089, + "grad_norm": 8.625, + "learning_rate": 4.172521049668901e-06, + "loss": 2.6074, + "step": 13651 + }, + { + "epoch": 1.4437394247038917, + "grad_norm": 8.5625, + "learning_rate": 4.172393758730947e-06, + "loss": 2.5151, + "step": 13652 + }, + { + "epoch": 1.4438451776649746, + "grad_norm": 8.3125, + "learning_rate": 4.172266459945133e-06, + "loss": 2.5966, + "step": 13653 + }, + { + "epoch": 1.4439509306260576, + "grad_norm": 8.375, + "learning_rate": 4.172139153312057e-06, + "loss": 2.1857, + "step": 13654 + }, + { + "epoch": 1.4440566835871405, + "grad_norm": 8.0625, + "learning_rate": 4.172011838832316e-06, + "loss": 2.4952, + "step": 13655 + }, + { + "epoch": 1.4441624365482233, + "grad_norm": 8.0, + "learning_rate": 4.171884516506507e-06, + "loss": 2.4685, + "step": 13656 + }, + { + "epoch": 1.4442681895093061, + "grad_norm": 16.125, + "learning_rate": 4.171757186335228e-06, + "loss": 2.393, + "step": 13657 + }, + { + "epoch": 1.4443739424703892, + "grad_norm": 7.8125, + "learning_rate": 4.1716298483190755e-06, + "loss": 2.5592, + "step": 13658 + }, + { + "epoch": 1.444479695431472, + "grad_norm": 7.90625, + "learning_rate": 4.171502502458649e-06, + "loss": 2.5614, + "step": 13659 + }, + { + "epoch": 1.444585448392555, + "grad_norm": 7.96875, + "learning_rate": 4.171375148754544e-06, + "loss": 2.4875, + "step": 13660 + }, + { + "epoch": 1.444691201353638, + "grad_norm": 8.9375, + "learning_rate": 4.1712477872073585e-06, + "loss": 2.7027, + "step": 13661 + }, + { + "epoch": 1.4447969543147208, + "grad_norm": 7.90625, + "learning_rate": 4.171120417817691e-06, + "loss": 2.6393, + "step": 13662 + }, + { + "epoch": 1.4449027072758036, + "grad_norm": 8.9375, + "learning_rate": 4.17099304058614e-06, + "loss": 2.3787, + "step": 13663 + }, + { + "epoch": 1.4450084602368867, + "grad_norm": 8.4375, + "learning_rate": 4.170865655513301e-06, + "loss": 2.6708, + "step": 13664 + }, + { + "epoch": 1.4451142131979695, + "grad_norm": 8.25, + "learning_rate": 4.170738262599773e-06, + "loss": 2.6259, + "step": 13665 + }, + { + "epoch": 1.4452199661590526, + "grad_norm": 8.5625, + "learning_rate": 4.1706108618461536e-06, + "loss": 2.1667, + "step": 13666 + }, + { + "epoch": 1.4453257191201354, + "grad_norm": 8.0, + "learning_rate": 4.170483453253041e-06, + "loss": 2.5285, + "step": 13667 + }, + { + "epoch": 1.4454314720812182, + "grad_norm": 8.25, + "learning_rate": 4.1703560368210325e-06, + "loss": 2.474, + "step": 13668 + }, + { + "epoch": 1.445537225042301, + "grad_norm": 9.8125, + "learning_rate": 4.170228612550726e-06, + "loss": 2.6885, + "step": 13669 + }, + { + "epoch": 1.4456429780033841, + "grad_norm": 8.5625, + "learning_rate": 4.17010118044272e-06, + "loss": 2.6066, + "step": 13670 + }, + { + "epoch": 1.445748730964467, + "grad_norm": 8.125, + "learning_rate": 4.169973740497613e-06, + "loss": 2.7613, + "step": 13671 + }, + { + "epoch": 1.44585448392555, + "grad_norm": 8.0625, + "learning_rate": 4.169846292716001e-06, + "loss": 2.4447, + "step": 13672 + }, + { + "epoch": 1.4459602368866329, + "grad_norm": 8.25, + "learning_rate": 4.169718837098484e-06, + "loss": 2.651, + "step": 13673 + }, + { + "epoch": 1.4460659898477157, + "grad_norm": 7.875, + "learning_rate": 4.169591373645659e-06, + "loss": 2.6477, + "step": 13674 + }, + { + "epoch": 1.4461717428087986, + "grad_norm": 8.375, + "learning_rate": 4.169463902358125e-06, + "loss": 2.6909, + "step": 13675 + }, + { + "epoch": 1.4462774957698816, + "grad_norm": 8.0, + "learning_rate": 4.1693364232364795e-06, + "loss": 2.5301, + "step": 13676 + }, + { + "epoch": 1.4463832487309645, + "grad_norm": 7.9375, + "learning_rate": 4.169208936281321e-06, + "loss": 2.5822, + "step": 13677 + }, + { + "epoch": 1.4464890016920473, + "grad_norm": 8.125, + "learning_rate": 4.169081441493247e-06, + "loss": 2.8321, + "step": 13678 + }, + { + "epoch": 1.4465947546531304, + "grad_norm": 8.25, + "learning_rate": 4.168953938872857e-06, + "loss": 2.6139, + "step": 13679 + }, + { + "epoch": 1.4467005076142132, + "grad_norm": 7.84375, + "learning_rate": 4.168826428420749e-06, + "loss": 2.4864, + "step": 13680 + }, + { + "epoch": 1.446806260575296, + "grad_norm": 8.625, + "learning_rate": 4.168698910137521e-06, + "loss": 2.6084, + "step": 13681 + }, + { + "epoch": 1.4469120135363789, + "grad_norm": 7.84375, + "learning_rate": 4.168571384023771e-06, + "loss": 2.3847, + "step": 13682 + }, + { + "epoch": 1.447017766497462, + "grad_norm": 8.125, + "learning_rate": 4.168443850080097e-06, + "loss": 2.4456, + "step": 13683 + }, + { + "epoch": 1.4471235194585448, + "grad_norm": 8.25, + "learning_rate": 4.1683163083071e-06, + "loss": 2.5515, + "step": 13684 + }, + { + "epoch": 1.4472292724196278, + "grad_norm": 8.4375, + "learning_rate": 4.168188758705377e-06, + "loss": 2.5337, + "step": 13685 + }, + { + "epoch": 1.4473350253807107, + "grad_norm": 8.1875, + "learning_rate": 4.168061201275526e-06, + "loss": 2.5491, + "step": 13686 + }, + { + "epoch": 1.4474407783417935, + "grad_norm": 8.0, + "learning_rate": 4.167933636018145e-06, + "loss": 2.4961, + "step": 13687 + }, + { + "epoch": 1.4475465313028764, + "grad_norm": 8.5, + "learning_rate": 4.167806062933835e-06, + "loss": 2.4669, + "step": 13688 + }, + { + "epoch": 1.4476522842639594, + "grad_norm": 7.90625, + "learning_rate": 4.167678482023193e-06, + "loss": 2.6707, + "step": 13689 + }, + { + "epoch": 1.4477580372250423, + "grad_norm": 8.5625, + "learning_rate": 4.167550893286817e-06, + "loss": 2.5097, + "step": 13690 + }, + { + "epoch": 1.4478637901861253, + "grad_norm": 8.75, + "learning_rate": 4.1674232967253075e-06, + "loss": 2.4587, + "step": 13691 + }, + { + "epoch": 1.4479695431472082, + "grad_norm": 8.75, + "learning_rate": 4.167295692339262e-06, + "loss": 2.6794, + "step": 13692 + }, + { + "epoch": 1.448075296108291, + "grad_norm": 8.3125, + "learning_rate": 4.16716808012928e-06, + "loss": 2.449, + "step": 13693 + }, + { + "epoch": 1.4481810490693738, + "grad_norm": 8.4375, + "learning_rate": 4.16704046009596e-06, + "loss": 2.7048, + "step": 13694 + }, + { + "epoch": 1.448286802030457, + "grad_norm": 8.0, + "learning_rate": 4.166912832239901e-06, + "loss": 2.6545, + "step": 13695 + }, + { + "epoch": 1.4483925549915397, + "grad_norm": 8.5, + "learning_rate": 4.166785196561702e-06, + "loss": 2.6207, + "step": 13696 + }, + { + "epoch": 1.4484983079526228, + "grad_norm": 8.5625, + "learning_rate": 4.166657553061961e-06, + "loss": 2.6097, + "step": 13697 + }, + { + "epoch": 1.4486040609137056, + "grad_norm": 8.3125, + "learning_rate": 4.166529901741279e-06, + "loss": 2.6424, + "step": 13698 + }, + { + "epoch": 1.4487098138747885, + "grad_norm": 8.75, + "learning_rate": 4.166402242600252e-06, + "loss": 2.5404, + "step": 13699 + }, + { + "epoch": 1.4488155668358713, + "grad_norm": 8.25, + "learning_rate": 4.166274575639482e-06, + "loss": 2.6107, + "step": 13700 + }, + { + "epoch": 1.4489213197969544, + "grad_norm": 8.5, + "learning_rate": 4.166146900859567e-06, + "loss": 2.5822, + "step": 13701 + }, + { + "epoch": 1.4490270727580372, + "grad_norm": 8.375, + "learning_rate": 4.166019218261106e-06, + "loss": 2.6599, + "step": 13702 + }, + { + "epoch": 1.44913282571912, + "grad_norm": 7.78125, + "learning_rate": 4.165891527844697e-06, + "loss": 2.6317, + "step": 13703 + }, + { + "epoch": 1.4492385786802031, + "grad_norm": 8.0625, + "learning_rate": 4.165763829610941e-06, + "loss": 2.6194, + "step": 13704 + }, + { + "epoch": 1.449344331641286, + "grad_norm": 7.34375, + "learning_rate": 4.165636123560438e-06, + "loss": 2.2879, + "step": 13705 + }, + { + "epoch": 1.4494500846023688, + "grad_norm": 7.90625, + "learning_rate": 4.165508409693785e-06, + "loss": 2.529, + "step": 13706 + }, + { + "epoch": 1.4495558375634519, + "grad_norm": 8.1875, + "learning_rate": 4.165380688011582e-06, + "loss": 2.4501, + "step": 13707 + }, + { + "epoch": 1.4496615905245347, + "grad_norm": 7.84375, + "learning_rate": 4.165252958514428e-06, + "loss": 2.5621, + "step": 13708 + }, + { + "epoch": 1.4497673434856175, + "grad_norm": 8.625, + "learning_rate": 4.165125221202924e-06, + "loss": 2.7053, + "step": 13709 + }, + { + "epoch": 1.4498730964467006, + "grad_norm": 8.3125, + "learning_rate": 4.1649974760776675e-06, + "loss": 2.1861, + "step": 13710 + }, + { + "epoch": 1.4499788494077834, + "grad_norm": 8.3125, + "learning_rate": 4.164869723139259e-06, + "loss": 2.5203, + "step": 13711 + }, + { + "epoch": 1.4500846023688663, + "grad_norm": 7.96875, + "learning_rate": 4.164741962388298e-06, + "loss": 2.6034, + "step": 13712 + }, + { + "epoch": 1.450190355329949, + "grad_norm": 8.375, + "learning_rate": 4.164614193825384e-06, + "loss": 2.4653, + "step": 13713 + }, + { + "epoch": 1.4502961082910322, + "grad_norm": 7.9375, + "learning_rate": 4.164486417451116e-06, + "loss": 2.5397, + "step": 13714 + }, + { + "epoch": 1.450401861252115, + "grad_norm": 8.0625, + "learning_rate": 4.164358633266094e-06, + "loss": 2.4448, + "step": 13715 + }, + { + "epoch": 1.450507614213198, + "grad_norm": 8.1875, + "learning_rate": 4.164230841270918e-06, + "loss": 2.6674, + "step": 13716 + }, + { + "epoch": 1.450613367174281, + "grad_norm": 8.375, + "learning_rate": 4.1641030414661874e-06, + "loss": 2.6429, + "step": 13717 + }, + { + "epoch": 1.4507191201353637, + "grad_norm": 8.625, + "learning_rate": 4.163975233852501e-06, + "loss": 2.5679, + "step": 13718 + }, + { + "epoch": 1.4508248730964466, + "grad_norm": 7.65625, + "learning_rate": 4.16384741843046e-06, + "loss": 2.5148, + "step": 13719 + }, + { + "epoch": 1.4509306260575296, + "grad_norm": 8.875, + "learning_rate": 4.1637195952006635e-06, + "loss": 2.6475, + "step": 13720 + }, + { + "epoch": 1.4510363790186125, + "grad_norm": 8.0, + "learning_rate": 4.163591764163712e-06, + "loss": 2.4146, + "step": 13721 + }, + { + "epoch": 1.4511421319796955, + "grad_norm": 8.3125, + "learning_rate": 4.163463925320203e-06, + "loss": 2.5639, + "step": 13722 + }, + { + "epoch": 1.4512478849407784, + "grad_norm": 8.625, + "learning_rate": 4.163336078670739e-06, + "loss": 2.6248, + "step": 13723 + }, + { + "epoch": 1.4513536379018612, + "grad_norm": 8.3125, + "learning_rate": 4.16320822421592e-06, + "loss": 2.5467, + "step": 13724 + }, + { + "epoch": 1.451459390862944, + "grad_norm": 8.1875, + "learning_rate": 4.1630803619563445e-06, + "loss": 2.7595, + "step": 13725 + }, + { + "epoch": 1.4515651438240271, + "grad_norm": 7.875, + "learning_rate": 4.162952491892612e-06, + "loss": 2.5269, + "step": 13726 + }, + { + "epoch": 1.45167089678511, + "grad_norm": 8.6875, + "learning_rate": 4.1628246140253245e-06, + "loss": 2.5913, + "step": 13727 + }, + { + "epoch": 1.451776649746193, + "grad_norm": 7.40625, + "learning_rate": 4.162696728355081e-06, + "loss": 2.3478, + "step": 13728 + }, + { + "epoch": 1.4518824027072759, + "grad_norm": 8.1875, + "learning_rate": 4.162568834882482e-06, + "loss": 2.5357, + "step": 13729 + }, + { + "epoch": 1.4519881556683587, + "grad_norm": 9.1875, + "learning_rate": 4.162440933608128e-06, + "loss": 2.7951, + "step": 13730 + }, + { + "epoch": 1.4520939086294415, + "grad_norm": 8.6875, + "learning_rate": 4.1623130245326175e-06, + "loss": 2.3441, + "step": 13731 + }, + { + "epoch": 1.4521996615905246, + "grad_norm": 8.75, + "learning_rate": 4.162185107656552e-06, + "loss": 2.2399, + "step": 13732 + }, + { + "epoch": 1.4523054145516074, + "grad_norm": 8.5, + "learning_rate": 4.1620571829805315e-06, + "loss": 2.4871, + "step": 13733 + }, + { + "epoch": 1.4524111675126903, + "grad_norm": 8.6875, + "learning_rate": 4.161929250505157e-06, + "loss": 2.5704, + "step": 13734 + }, + { + "epoch": 1.4525169204737733, + "grad_norm": 7.9375, + "learning_rate": 4.161801310231028e-06, + "loss": 2.5602, + "step": 13735 + }, + { + "epoch": 1.4526226734348562, + "grad_norm": 8.3125, + "learning_rate": 4.161673362158745e-06, + "loss": 2.6423, + "step": 13736 + }, + { + "epoch": 1.452728426395939, + "grad_norm": 8.3125, + "learning_rate": 4.161545406288908e-06, + "loss": 2.3657, + "step": 13737 + }, + { + "epoch": 1.4528341793570219, + "grad_norm": 8.25, + "learning_rate": 4.161417442622119e-06, + "loss": 2.4692, + "step": 13738 + }, + { + "epoch": 1.452939932318105, + "grad_norm": 8.125, + "learning_rate": 4.161289471158977e-06, + "loss": 2.6549, + "step": 13739 + }, + { + "epoch": 1.4530456852791878, + "grad_norm": 8.25, + "learning_rate": 4.161161491900083e-06, + "loss": 2.4828, + "step": 13740 + }, + { + "epoch": 1.4531514382402708, + "grad_norm": 8.8125, + "learning_rate": 4.161033504846037e-06, + "loss": 2.5382, + "step": 13741 + }, + { + "epoch": 1.4532571912013537, + "grad_norm": 8.1875, + "learning_rate": 4.16090550999744e-06, + "loss": 2.7986, + "step": 13742 + }, + { + "epoch": 1.4533629441624365, + "grad_norm": 8.75, + "learning_rate": 4.160777507354894e-06, + "loss": 2.8791, + "step": 13743 + }, + { + "epoch": 1.4534686971235193, + "grad_norm": 8.3125, + "learning_rate": 4.160649496918998e-06, + "loss": 2.2851, + "step": 13744 + }, + { + "epoch": 1.4535744500846024, + "grad_norm": 8.3125, + "learning_rate": 4.160521478690353e-06, + "loss": 2.7037, + "step": 13745 + }, + { + "epoch": 1.4536802030456852, + "grad_norm": 8.0625, + "learning_rate": 4.160393452669559e-06, + "loss": 2.6707, + "step": 13746 + }, + { + "epoch": 1.4537859560067683, + "grad_norm": 9.1875, + "learning_rate": 4.160265418857218e-06, + "loss": 2.4825, + "step": 13747 + }, + { + "epoch": 1.4538917089678511, + "grad_norm": 8.4375, + "learning_rate": 4.160137377253931e-06, + "loss": 2.5876, + "step": 13748 + }, + { + "epoch": 1.453997461928934, + "grad_norm": 8.5625, + "learning_rate": 4.160009327860298e-06, + "loss": 2.5998, + "step": 13749 + }, + { + "epoch": 1.4541032148900168, + "grad_norm": 7.84375, + "learning_rate": 4.15988127067692e-06, + "loss": 2.5183, + "step": 13750 + }, + { + "epoch": 1.4542089678510999, + "grad_norm": 8.75, + "learning_rate": 4.159753205704398e-06, + "loss": 2.6454, + "step": 13751 + }, + { + "epoch": 1.4543147208121827, + "grad_norm": 8.125, + "learning_rate": 4.159625132943333e-06, + "loss": 2.6433, + "step": 13752 + }, + { + "epoch": 1.4544204737732658, + "grad_norm": 8.625, + "learning_rate": 4.159497052394326e-06, + "loss": 2.4894, + "step": 13753 + }, + { + "epoch": 1.4545262267343486, + "grad_norm": 9.0625, + "learning_rate": 4.1593689640579786e-06, + "loss": 2.6522, + "step": 13754 + }, + { + "epoch": 1.4546319796954315, + "grad_norm": 8.25, + "learning_rate": 4.15924086793489e-06, + "loss": 2.5114, + "step": 13755 + }, + { + "epoch": 1.4547377326565143, + "grad_norm": 8.8125, + "learning_rate": 4.159112764025663e-06, + "loss": 2.7139, + "step": 13756 + }, + { + "epoch": 1.4548434856175974, + "grad_norm": 7.96875, + "learning_rate": 4.1589846523308995e-06, + "loss": 2.2055, + "step": 13757 + }, + { + "epoch": 1.4549492385786802, + "grad_norm": 8.25, + "learning_rate": 4.158856532851199e-06, + "loss": 2.3482, + "step": 13758 + }, + { + "epoch": 1.455054991539763, + "grad_norm": 7.9375, + "learning_rate": 4.158728405587163e-06, + "loss": 2.5725, + "step": 13759 + }, + { + "epoch": 1.455160744500846, + "grad_norm": 8.0625, + "learning_rate": 4.158600270539393e-06, + "loss": 2.735, + "step": 13760 + }, + { + "epoch": 1.455266497461929, + "grad_norm": 8.1875, + "learning_rate": 4.158472127708491e-06, + "loss": 2.6627, + "step": 13761 + }, + { + "epoch": 1.4553722504230118, + "grad_norm": 8.375, + "learning_rate": 4.158343977095056e-06, + "loss": 2.6573, + "step": 13762 + }, + { + "epoch": 1.4554780033840948, + "grad_norm": 9.5, + "learning_rate": 4.158215818699692e-06, + "loss": 2.7262, + "step": 13763 + }, + { + "epoch": 1.4555837563451777, + "grad_norm": 7.9375, + "learning_rate": 4.158087652522999e-06, + "loss": 2.5073, + "step": 13764 + }, + { + "epoch": 1.4556895093062605, + "grad_norm": 9.0, + "learning_rate": 4.15795947856558e-06, + "loss": 2.7579, + "step": 13765 + }, + { + "epoch": 1.4557952622673436, + "grad_norm": 8.8125, + "learning_rate": 4.157831296828033e-06, + "loss": 2.58, + "step": 13766 + }, + { + "epoch": 1.4559010152284264, + "grad_norm": 8.625, + "learning_rate": 4.157703107310964e-06, + "loss": 2.5632, + "step": 13767 + }, + { + "epoch": 1.4560067681895092, + "grad_norm": 8.625, + "learning_rate": 4.157574910014971e-06, + "loss": 2.335, + "step": 13768 + }, + { + "epoch": 1.456112521150592, + "grad_norm": 8.0625, + "learning_rate": 4.157446704940657e-06, + "loss": 2.4915, + "step": 13769 + }, + { + "epoch": 1.4562182741116751, + "grad_norm": 8.3125, + "learning_rate": 4.157318492088625e-06, + "loss": 2.5465, + "step": 13770 + }, + { + "epoch": 1.456324027072758, + "grad_norm": 8.125, + "learning_rate": 4.157190271459473e-06, + "loss": 2.5158, + "step": 13771 + }, + { + "epoch": 1.456429780033841, + "grad_norm": 8.25, + "learning_rate": 4.157062043053805e-06, + "loss": 2.3021, + "step": 13772 + }, + { + "epoch": 1.4565355329949239, + "grad_norm": 8.0625, + "learning_rate": 4.156933806872223e-06, + "loss": 2.4, + "step": 13773 + }, + { + "epoch": 1.4566412859560067, + "grad_norm": 8.125, + "learning_rate": 4.1568055629153294e-06, + "loss": 2.5038, + "step": 13774 + }, + { + "epoch": 1.4567470389170896, + "grad_norm": 7.875, + "learning_rate": 4.156677311183724e-06, + "loss": 2.2886, + "step": 13775 + }, + { + "epoch": 1.4568527918781726, + "grad_norm": 8.3125, + "learning_rate": 4.156549051678009e-06, + "loss": 2.6454, + "step": 13776 + }, + { + "epoch": 1.4569585448392555, + "grad_norm": 8.0625, + "learning_rate": 4.156420784398788e-06, + "loss": 2.5813, + "step": 13777 + }, + { + "epoch": 1.4570642978003385, + "grad_norm": 8.875, + "learning_rate": 4.156292509346661e-06, + "loss": 2.7323, + "step": 13778 + }, + { + "epoch": 1.4571700507614214, + "grad_norm": 8.25, + "learning_rate": 4.15616422652223e-06, + "loss": 2.5497, + "step": 13779 + }, + { + "epoch": 1.4572758037225042, + "grad_norm": 8.6875, + "learning_rate": 4.1560359359260985e-06, + "loss": 2.6449, + "step": 13780 + }, + { + "epoch": 1.457381556683587, + "grad_norm": 8.1875, + "learning_rate": 4.155907637558867e-06, + "loss": 2.4246, + "step": 13781 + }, + { + "epoch": 1.45748730964467, + "grad_norm": 9.25, + "learning_rate": 4.1557793314211395e-06, + "loss": 2.7945, + "step": 13782 + }, + { + "epoch": 1.457593062605753, + "grad_norm": 8.25, + "learning_rate": 4.155651017513516e-06, + "loss": 2.6743, + "step": 13783 + }, + { + "epoch": 1.457698815566836, + "grad_norm": 8.3125, + "learning_rate": 4.155522695836599e-06, + "loss": 2.4942, + "step": 13784 + }, + { + "epoch": 1.4578045685279188, + "grad_norm": 8.25, + "learning_rate": 4.155394366390991e-06, + "loss": 2.7777, + "step": 13785 + }, + { + "epoch": 1.4579103214890017, + "grad_norm": 8.125, + "learning_rate": 4.155266029177295e-06, + "loss": 2.4395, + "step": 13786 + }, + { + "epoch": 1.4580160744500845, + "grad_norm": 8.75, + "learning_rate": 4.155137684196114e-06, + "loss": 2.7198, + "step": 13787 + }, + { + "epoch": 1.4581218274111676, + "grad_norm": 8.5, + "learning_rate": 4.155009331448046e-06, + "loss": 2.5262, + "step": 13788 + }, + { + "epoch": 1.4582275803722504, + "grad_norm": 8.1875, + "learning_rate": 4.154880970933697e-06, + "loss": 2.6835, + "step": 13789 + }, + { + "epoch": 1.4583333333333333, + "grad_norm": 8.3125, + "learning_rate": 4.154752602653668e-06, + "loss": 2.4068, + "step": 13790 + }, + { + "epoch": 1.4584390862944163, + "grad_norm": 7.875, + "learning_rate": 4.154624226608563e-06, + "loss": 2.2279, + "step": 13791 + }, + { + "epoch": 1.4585448392554992, + "grad_norm": 7.84375, + "learning_rate": 4.154495842798982e-06, + "loss": 2.4617, + "step": 13792 + }, + { + "epoch": 1.458650592216582, + "grad_norm": 8.75, + "learning_rate": 4.1543674512255295e-06, + "loss": 2.3903, + "step": 13793 + }, + { + "epoch": 1.4587563451776648, + "grad_norm": 8.0625, + "learning_rate": 4.154239051888807e-06, + "loss": 2.5658, + "step": 13794 + }, + { + "epoch": 1.458862098138748, + "grad_norm": 8.25, + "learning_rate": 4.154110644789418e-06, + "loss": 2.5769, + "step": 13795 + }, + { + "epoch": 1.4589678510998307, + "grad_norm": 8.4375, + "learning_rate": 4.153982229927963e-06, + "loss": 2.4581, + "step": 13796 + }, + { + "epoch": 1.4590736040609138, + "grad_norm": 8.8125, + "learning_rate": 4.153853807305046e-06, + "loss": 2.7469, + "step": 13797 + }, + { + "epoch": 1.4591793570219966, + "grad_norm": 8.375, + "learning_rate": 4.153725376921269e-06, + "loss": 2.5357, + "step": 13798 + }, + { + "epoch": 1.4592851099830795, + "grad_norm": 8.125, + "learning_rate": 4.153596938777236e-06, + "loss": 2.5226, + "step": 13799 + }, + { + "epoch": 1.4593908629441623, + "grad_norm": 9.1875, + "learning_rate": 4.153468492873549e-06, + "loss": 2.343, + "step": 13800 + }, + { + "epoch": 1.4594966159052454, + "grad_norm": 8.4375, + "learning_rate": 4.15334003921081e-06, + "loss": 2.5677, + "step": 13801 + }, + { + "epoch": 1.4596023688663282, + "grad_norm": 8.3125, + "learning_rate": 4.153211577789623e-06, + "loss": 2.5742, + "step": 13802 + }, + { + "epoch": 1.4597081218274113, + "grad_norm": 8.3125, + "learning_rate": 4.15308310861059e-06, + "loss": 2.7311, + "step": 13803 + }, + { + "epoch": 1.4598138747884941, + "grad_norm": 8.1875, + "learning_rate": 4.152954631674314e-06, + "loss": 2.5516, + "step": 13804 + }, + { + "epoch": 1.459919627749577, + "grad_norm": 8.375, + "learning_rate": 4.152826146981398e-06, + "loss": 2.633, + "step": 13805 + }, + { + "epoch": 1.4600253807106598, + "grad_norm": 8.375, + "learning_rate": 4.152697654532445e-06, + "loss": 2.4522, + "step": 13806 + }, + { + "epoch": 1.4601311336717429, + "grad_norm": 8.75, + "learning_rate": 4.152569154328056e-06, + "loss": 2.6092, + "step": 13807 + }, + { + "epoch": 1.4602368866328257, + "grad_norm": 9.375, + "learning_rate": 4.1524406463688385e-06, + "loss": 2.6292, + "step": 13808 + }, + { + "epoch": 1.4603426395939088, + "grad_norm": 7.96875, + "learning_rate": 4.1523121306553915e-06, + "loss": 2.4908, + "step": 13809 + }, + { + "epoch": 1.4604483925549916, + "grad_norm": 8.375, + "learning_rate": 4.15218360718832e-06, + "loss": 2.6451, + "step": 13810 + }, + { + "epoch": 1.4605541455160744, + "grad_norm": 8.1875, + "learning_rate": 4.152055075968227e-06, + "loss": 2.5135, + "step": 13811 + }, + { + "epoch": 1.4606598984771573, + "grad_norm": 7.875, + "learning_rate": 4.151926536995714e-06, + "loss": 2.4047, + "step": 13812 + }, + { + "epoch": 1.4607656514382403, + "grad_norm": 8.25, + "learning_rate": 4.1517979902713855e-06, + "loss": 2.6429, + "step": 13813 + }, + { + "epoch": 1.4608714043993232, + "grad_norm": 8.0625, + "learning_rate": 4.151669435795846e-06, + "loss": 2.4827, + "step": 13814 + }, + { + "epoch": 1.460977157360406, + "grad_norm": 8.375, + "learning_rate": 4.151540873569696e-06, + "loss": 2.5587, + "step": 13815 + }, + { + "epoch": 1.461082910321489, + "grad_norm": 8.625, + "learning_rate": 4.15141230359354e-06, + "loss": 2.2893, + "step": 13816 + }, + { + "epoch": 1.461188663282572, + "grad_norm": 8.8125, + "learning_rate": 4.1512837258679815e-06, + "loss": 2.7011, + "step": 13817 + }, + { + "epoch": 1.4612944162436547, + "grad_norm": 9.3125, + "learning_rate": 4.151155140393624e-06, + "loss": 2.4614, + "step": 13818 + }, + { + "epoch": 1.4614001692047378, + "grad_norm": 8.375, + "learning_rate": 4.151026547171072e-06, + "loss": 2.6982, + "step": 13819 + }, + { + "epoch": 1.4615059221658206, + "grad_norm": 8.125, + "learning_rate": 4.150897946200927e-06, + "loss": 2.6607, + "step": 13820 + }, + { + "epoch": 1.4616116751269035, + "grad_norm": 8.0, + "learning_rate": 4.150769337483792e-06, + "loss": 2.3901, + "step": 13821 + }, + { + "epoch": 1.4617174280879865, + "grad_norm": 8.3125, + "learning_rate": 4.150640721020272e-06, + "loss": 2.6456, + "step": 13822 + }, + { + "epoch": 1.4618231810490694, + "grad_norm": 7.875, + "learning_rate": 4.1505120968109715e-06, + "loss": 2.4162, + "step": 13823 + }, + { + "epoch": 1.4619289340101522, + "grad_norm": 7.53125, + "learning_rate": 4.150383464856492e-06, + "loss": 2.5145, + "step": 13824 + }, + { + "epoch": 1.462034686971235, + "grad_norm": 8.8125, + "learning_rate": 4.150254825157437e-06, + "loss": 2.3698, + "step": 13825 + }, + { + "epoch": 1.4621404399323181, + "grad_norm": 7.8125, + "learning_rate": 4.150126177714412e-06, + "loss": 2.5605, + "step": 13826 + }, + { + "epoch": 1.462246192893401, + "grad_norm": 8.4375, + "learning_rate": 4.14999752252802e-06, + "loss": 2.5191, + "step": 13827 + }, + { + "epoch": 1.462351945854484, + "grad_norm": 8.0, + "learning_rate": 4.149868859598864e-06, + "loss": 2.3101, + "step": 13828 + }, + { + "epoch": 1.4624576988155669, + "grad_norm": 8.625, + "learning_rate": 4.149740188927548e-06, + "loss": 2.5717, + "step": 13829 + }, + { + "epoch": 1.4625634517766497, + "grad_norm": 8.1875, + "learning_rate": 4.149611510514676e-06, + "loss": 2.518, + "step": 13830 + }, + { + "epoch": 1.4626692047377325, + "grad_norm": 8.5, + "learning_rate": 4.149482824360852e-06, + "loss": 2.6064, + "step": 13831 + }, + { + "epoch": 1.4627749576988156, + "grad_norm": 8.25, + "learning_rate": 4.14935413046668e-06, + "loss": 2.1389, + "step": 13832 + }, + { + "epoch": 1.4628807106598984, + "grad_norm": 7.46875, + "learning_rate": 4.149225428832763e-06, + "loss": 2.4395, + "step": 13833 + }, + { + "epoch": 1.4629864636209815, + "grad_norm": 8.0, + "learning_rate": 4.149096719459706e-06, + "loss": 2.4531, + "step": 13834 + }, + { + "epoch": 1.4630922165820643, + "grad_norm": 8.25, + "learning_rate": 4.148968002348113e-06, + "loss": 2.4783, + "step": 13835 + }, + { + "epoch": 1.4631979695431472, + "grad_norm": 8.25, + "learning_rate": 4.148839277498587e-06, + "loss": 2.4747, + "step": 13836 + }, + { + "epoch": 1.46330372250423, + "grad_norm": 9.0, + "learning_rate": 4.148710544911733e-06, + "loss": 2.4938, + "step": 13837 + }, + { + "epoch": 1.463409475465313, + "grad_norm": 8.4375, + "learning_rate": 4.148581804588154e-06, + "loss": 2.6989, + "step": 13838 + }, + { + "epoch": 1.463515228426396, + "grad_norm": 8.5, + "learning_rate": 4.1484530565284555e-06, + "loss": 2.4274, + "step": 13839 + }, + { + "epoch": 1.463620981387479, + "grad_norm": 8.75, + "learning_rate": 4.148324300733241e-06, + "loss": 2.582, + "step": 13840 + }, + { + "epoch": 1.4637267343485618, + "grad_norm": 8.5625, + "learning_rate": 4.148195537203115e-06, + "loss": 2.7298, + "step": 13841 + }, + { + "epoch": 1.4638324873096447, + "grad_norm": 8.125, + "learning_rate": 4.14806676593868e-06, + "loss": 2.5744, + "step": 13842 + }, + { + "epoch": 1.4639382402707275, + "grad_norm": 8.5625, + "learning_rate": 4.147937986940543e-06, + "loss": 2.6098, + "step": 13843 + }, + { + "epoch": 1.4640439932318106, + "grad_norm": 8.0625, + "learning_rate": 4.147809200209306e-06, + "loss": 2.4723, + "step": 13844 + }, + { + "epoch": 1.4641497461928934, + "grad_norm": 7.875, + "learning_rate": 4.147680405745575e-06, + "loss": 2.4593, + "step": 13845 + }, + { + "epoch": 1.4642554991539762, + "grad_norm": 8.25, + "learning_rate": 4.147551603549954e-06, + "loss": 2.4359, + "step": 13846 + }, + { + "epoch": 1.4643612521150593, + "grad_norm": 8.375, + "learning_rate": 4.147422793623047e-06, + "loss": 2.5228, + "step": 13847 + }, + { + "epoch": 1.4644670050761421, + "grad_norm": 8.3125, + "learning_rate": 4.147293975965458e-06, + "loss": 2.6048, + "step": 13848 + }, + { + "epoch": 1.464572758037225, + "grad_norm": 8.875, + "learning_rate": 4.147165150577793e-06, + "loss": 2.6135, + "step": 13849 + }, + { + "epoch": 1.4646785109983078, + "grad_norm": 9.3125, + "learning_rate": 4.147036317460655e-06, + "loss": 2.6769, + "step": 13850 + }, + { + "epoch": 1.4647842639593909, + "grad_norm": 8.4375, + "learning_rate": 4.146907476614649e-06, + "loss": 2.2754, + "step": 13851 + }, + { + "epoch": 1.4648900169204737, + "grad_norm": 8.375, + "learning_rate": 4.14677862804038e-06, + "loss": 2.4745, + "step": 13852 + }, + { + "epoch": 1.4649957698815568, + "grad_norm": 8.8125, + "learning_rate": 4.146649771738453e-06, + "loss": 2.6111, + "step": 13853 + }, + { + "epoch": 1.4651015228426396, + "grad_norm": 9.0, + "learning_rate": 4.146520907709471e-06, + "loss": 2.5819, + "step": 13854 + }, + { + "epoch": 1.4652072758037225, + "grad_norm": 8.875, + "learning_rate": 4.146392035954041e-06, + "loss": 2.6866, + "step": 13855 + }, + { + "epoch": 1.4653130287648053, + "grad_norm": 7.8125, + "learning_rate": 4.1462631564727665e-06, + "loss": 2.3833, + "step": 13856 + }, + { + "epoch": 1.4654187817258884, + "grad_norm": 8.25, + "learning_rate": 4.146134269266252e-06, + "loss": 2.5527, + "step": 13857 + }, + { + "epoch": 1.4655245346869712, + "grad_norm": 8.625, + "learning_rate": 4.1460053743351015e-06, + "loss": 2.7549, + "step": 13858 + }, + { + "epoch": 1.4656302876480543, + "grad_norm": 9.125, + "learning_rate": 4.145876471679922e-06, + "loss": 2.2698, + "step": 13859 + }, + { + "epoch": 1.465736040609137, + "grad_norm": 7.59375, + "learning_rate": 4.145747561301317e-06, + "loss": 2.2601, + "step": 13860 + }, + { + "epoch": 1.46584179357022, + "grad_norm": 7.90625, + "learning_rate": 4.145618643199893e-06, + "loss": 2.5995, + "step": 13861 + }, + { + "epoch": 1.4659475465313028, + "grad_norm": 7.96875, + "learning_rate": 4.145489717376252e-06, + "loss": 2.5751, + "step": 13862 + }, + { + "epoch": 1.4660532994923858, + "grad_norm": 8.75, + "learning_rate": 4.145360783831001e-06, + "loss": 2.7456, + "step": 13863 + }, + { + "epoch": 1.4661590524534687, + "grad_norm": 8.25, + "learning_rate": 4.1452318425647455e-06, + "loss": 2.3745, + "step": 13864 + }, + { + "epoch": 1.4662648054145517, + "grad_norm": 8.5625, + "learning_rate": 4.1451028935780895e-06, + "loss": 2.4578, + "step": 13865 + }, + { + "epoch": 1.4663705583756346, + "grad_norm": 8.5625, + "learning_rate": 4.144973936871639e-06, + "loss": 2.6364, + "step": 13866 + }, + { + "epoch": 1.4664763113367174, + "grad_norm": 8.3125, + "learning_rate": 4.144844972445997e-06, + "loss": 2.589, + "step": 13867 + }, + { + "epoch": 1.4665820642978002, + "grad_norm": 8.625, + "learning_rate": 4.144716000301772e-06, + "loss": 2.6198, + "step": 13868 + }, + { + "epoch": 1.4666878172588833, + "grad_norm": 8.875, + "learning_rate": 4.144587020439567e-06, + "loss": 2.4516, + "step": 13869 + }, + { + "epoch": 1.4667935702199661, + "grad_norm": 8.5, + "learning_rate": 4.144458032859987e-06, + "loss": 2.5597, + "step": 13870 + }, + { + "epoch": 1.466899323181049, + "grad_norm": 8.0625, + "learning_rate": 4.144329037563639e-06, + "loss": 2.1848, + "step": 13871 + }, + { + "epoch": 1.467005076142132, + "grad_norm": 8.1875, + "learning_rate": 4.144200034551127e-06, + "loss": 2.6041, + "step": 13872 + }, + { + "epoch": 1.4671108291032149, + "grad_norm": 8.125, + "learning_rate": 4.144071023823056e-06, + "loss": 2.5547, + "step": 13873 + }, + { + "epoch": 1.4672165820642977, + "grad_norm": 8.4375, + "learning_rate": 4.143942005380032e-06, + "loss": 2.6036, + "step": 13874 + }, + { + "epoch": 1.4673223350253808, + "grad_norm": 8.5, + "learning_rate": 4.143812979222662e-06, + "loss": 2.562, + "step": 13875 + }, + { + "epoch": 1.4674280879864636, + "grad_norm": 8.8125, + "learning_rate": 4.1436839453515495e-06, + "loss": 2.4137, + "step": 13876 + }, + { + "epoch": 1.4675338409475465, + "grad_norm": 8.0, + "learning_rate": 4.1435549037673e-06, + "loss": 2.5883, + "step": 13877 + }, + { + "epoch": 1.4676395939086295, + "grad_norm": 8.3125, + "learning_rate": 4.14342585447052e-06, + "loss": 2.5508, + "step": 13878 + }, + { + "epoch": 1.4677453468697124, + "grad_norm": 8.1875, + "learning_rate": 4.143296797461814e-06, + "loss": 2.5322, + "step": 13879 + }, + { + "epoch": 1.4678510998307952, + "grad_norm": 8.25, + "learning_rate": 4.143167732741789e-06, + "loss": 2.5792, + "step": 13880 + }, + { + "epoch": 1.467956852791878, + "grad_norm": 7.9375, + "learning_rate": 4.143038660311049e-06, + "loss": 2.3882, + "step": 13881 + }, + { + "epoch": 1.468062605752961, + "grad_norm": 8.25, + "learning_rate": 4.1429095801702005e-06, + "loss": 2.5733, + "step": 13882 + }, + { + "epoch": 1.468168358714044, + "grad_norm": 8.125, + "learning_rate": 4.14278049231985e-06, + "loss": 2.4305, + "step": 13883 + }, + { + "epoch": 1.468274111675127, + "grad_norm": 8.6875, + "learning_rate": 4.142651396760602e-06, + "loss": 2.5131, + "step": 13884 + }, + { + "epoch": 1.4683798646362098, + "grad_norm": 8.4375, + "learning_rate": 4.142522293493063e-06, + "loss": 2.611, + "step": 13885 + }, + { + "epoch": 1.4684856175972927, + "grad_norm": 8.5, + "learning_rate": 4.142393182517839e-06, + "loss": 2.6847, + "step": 13886 + }, + { + "epoch": 1.4685913705583755, + "grad_norm": 8.625, + "learning_rate": 4.142264063835535e-06, + "loss": 2.5612, + "step": 13887 + }, + { + "epoch": 1.4686971235194586, + "grad_norm": 7.75, + "learning_rate": 4.142134937446758e-06, + "loss": 2.5394, + "step": 13888 + }, + { + "epoch": 1.4688028764805414, + "grad_norm": 7.75, + "learning_rate": 4.142005803352113e-06, + "loss": 2.4623, + "step": 13889 + }, + { + "epoch": 1.4689086294416245, + "grad_norm": 8.1875, + "learning_rate": 4.141876661552205e-06, + "loss": 2.6012, + "step": 13890 + }, + { + "epoch": 1.4690143824027073, + "grad_norm": 7.96875, + "learning_rate": 4.141747512047644e-06, + "loss": 2.7871, + "step": 13891 + }, + { + "epoch": 1.4691201353637902, + "grad_norm": 8.0, + "learning_rate": 4.141618354839032e-06, + "loss": 2.2963, + "step": 13892 + }, + { + "epoch": 1.469225888324873, + "grad_norm": 8.0625, + "learning_rate": 4.141489189926976e-06, + "loss": 2.523, + "step": 13893 + }, + { + "epoch": 1.469331641285956, + "grad_norm": 7.90625, + "learning_rate": 4.141360017312083e-06, + "loss": 2.632, + "step": 13894 + }, + { + "epoch": 1.469437394247039, + "grad_norm": 8.5, + "learning_rate": 4.141230836994959e-06, + "loss": 2.5819, + "step": 13895 + }, + { + "epoch": 1.469543147208122, + "grad_norm": 8.375, + "learning_rate": 4.141101648976209e-06, + "loss": 2.5543, + "step": 13896 + }, + { + "epoch": 1.4696489001692048, + "grad_norm": 7.9375, + "learning_rate": 4.140972453256441e-06, + "loss": 2.5611, + "step": 13897 + }, + { + "epoch": 1.4697546531302876, + "grad_norm": 8.8125, + "learning_rate": 4.14084324983626e-06, + "loss": 2.6887, + "step": 13898 + }, + { + "epoch": 1.4698604060913705, + "grad_norm": 9.0625, + "learning_rate": 4.140714038716273e-06, + "loss": 2.5551, + "step": 13899 + }, + { + "epoch": 1.4699661590524535, + "grad_norm": 7.75, + "learning_rate": 4.1405848198970866e-06, + "loss": 2.2354, + "step": 13900 + }, + { + "epoch": 1.4700719120135364, + "grad_norm": 8.3125, + "learning_rate": 4.140455593379305e-06, + "loss": 2.4804, + "step": 13901 + }, + { + "epoch": 1.4701776649746192, + "grad_norm": 8.0625, + "learning_rate": 4.140326359163538e-06, + "loss": 2.5776, + "step": 13902 + }, + { + "epoch": 1.4702834179357023, + "grad_norm": 8.375, + "learning_rate": 4.14019711725039e-06, + "loss": 2.4354, + "step": 13903 + }, + { + "epoch": 1.4703891708967851, + "grad_norm": 8.5625, + "learning_rate": 4.140067867640466e-06, + "loss": 2.3909, + "step": 13904 + }, + { + "epoch": 1.470494923857868, + "grad_norm": 8.0625, + "learning_rate": 4.1399386103343765e-06, + "loss": 2.5783, + "step": 13905 + }, + { + "epoch": 1.470600676818951, + "grad_norm": 8.8125, + "learning_rate": 4.139809345332724e-06, + "loss": 2.4277, + "step": 13906 + }, + { + "epoch": 1.4707064297800339, + "grad_norm": 8.25, + "learning_rate": 4.1396800726361184e-06, + "loss": 2.6528, + "step": 13907 + }, + { + "epoch": 1.4708121827411167, + "grad_norm": 8.5, + "learning_rate": 4.139550792245164e-06, + "loss": 2.805, + "step": 13908 + }, + { + "epoch": 1.4709179357021998, + "grad_norm": 8.6875, + "learning_rate": 4.139421504160468e-06, + "loss": 2.6992, + "step": 13909 + }, + { + "epoch": 1.4710236886632826, + "grad_norm": 8.3125, + "learning_rate": 4.139292208382638e-06, + "loss": 2.54, + "step": 13910 + }, + { + "epoch": 1.4711294416243654, + "grad_norm": 9.125, + "learning_rate": 4.139162904912281e-06, + "loss": 2.4773, + "step": 13911 + }, + { + "epoch": 1.4712351945854483, + "grad_norm": 8.25, + "learning_rate": 4.139033593750001e-06, + "loss": 2.4642, + "step": 13912 + }, + { + "epoch": 1.4713409475465313, + "grad_norm": 8.5625, + "learning_rate": 4.138904274896407e-06, + "loss": 2.5787, + "step": 13913 + }, + { + "epoch": 1.4714467005076142, + "grad_norm": 8.75, + "learning_rate": 4.138774948352106e-06, + "loss": 2.7272, + "step": 13914 + }, + { + "epoch": 1.4715524534686972, + "grad_norm": 8.5, + "learning_rate": 4.138645614117705e-06, + "loss": 2.6823, + "step": 13915 + }, + { + "epoch": 1.47165820642978, + "grad_norm": 8.75, + "learning_rate": 4.13851627219381e-06, + "loss": 2.6654, + "step": 13916 + }, + { + "epoch": 1.471763959390863, + "grad_norm": 8.5625, + "learning_rate": 4.138386922581027e-06, + "loss": 2.6086, + "step": 13917 + }, + { + "epoch": 1.4718697123519457, + "grad_norm": 8.3125, + "learning_rate": 4.138257565279966e-06, + "loss": 2.5522, + "step": 13918 + }, + { + "epoch": 1.4719754653130288, + "grad_norm": 8.25, + "learning_rate": 4.138128200291231e-06, + "loss": 2.5586, + "step": 13919 + }, + { + "epoch": 1.4720812182741116, + "grad_norm": 8.125, + "learning_rate": 4.137998827615431e-06, + "loss": 2.6461, + "step": 13920 + }, + { + "epoch": 1.4721869712351947, + "grad_norm": 8.1875, + "learning_rate": 4.137869447253173e-06, + "loss": 2.6154, + "step": 13921 + }, + { + "epoch": 1.4722927241962775, + "grad_norm": 8.0, + "learning_rate": 4.137740059205062e-06, + "loss": 2.5481, + "step": 13922 + }, + { + "epoch": 1.4723984771573604, + "grad_norm": 8.3125, + "learning_rate": 4.137610663471708e-06, + "loss": 2.5818, + "step": 13923 + }, + { + "epoch": 1.4725042301184432, + "grad_norm": 8.5, + "learning_rate": 4.137481260053716e-06, + "loss": 2.6706, + "step": 13924 + }, + { + "epoch": 1.4726099830795263, + "grad_norm": 8.0625, + "learning_rate": 4.137351848951695e-06, + "loss": 2.4495, + "step": 13925 + }, + { + "epoch": 1.4727157360406091, + "grad_norm": 11.9375, + "learning_rate": 4.137222430166251e-06, + "loss": 2.5753, + "step": 13926 + }, + { + "epoch": 1.4728214890016922, + "grad_norm": 8.0625, + "learning_rate": 4.1370930036979914e-06, + "loss": 2.6605, + "step": 13927 + }, + { + "epoch": 1.472927241962775, + "grad_norm": 8.625, + "learning_rate": 4.1369635695475245e-06, + "loss": 2.741, + "step": 13928 + }, + { + "epoch": 1.4730329949238579, + "grad_norm": 9.1875, + "learning_rate": 4.136834127715456e-06, + "loss": 2.6966, + "step": 13929 + }, + { + "epoch": 1.4731387478849407, + "grad_norm": 8.375, + "learning_rate": 4.136704678202396e-06, + "loss": 2.7233, + "step": 13930 + }, + { + "epoch": 1.4732445008460238, + "grad_norm": 8.625, + "learning_rate": 4.136575221008949e-06, + "loss": 2.7533, + "step": 13931 + }, + { + "epoch": 1.4733502538071066, + "grad_norm": 8.375, + "learning_rate": 4.136445756135725e-06, + "loss": 2.5766, + "step": 13932 + }, + { + "epoch": 1.4734560067681894, + "grad_norm": 8.375, + "learning_rate": 4.136316283583329e-06, + "loss": 2.6542, + "step": 13933 + }, + { + "epoch": 1.4735617597292725, + "grad_norm": 8.8125, + "learning_rate": 4.136186803352371e-06, + "loss": 2.7517, + "step": 13934 + }, + { + "epoch": 1.4736675126903553, + "grad_norm": 9.5, + "learning_rate": 4.136057315443457e-06, + "loss": 2.6654, + "step": 13935 + }, + { + "epoch": 1.4737732656514382, + "grad_norm": 8.3125, + "learning_rate": 4.1359278198571955e-06, + "loss": 2.6008, + "step": 13936 + }, + { + "epoch": 1.473879018612521, + "grad_norm": 7.75, + "learning_rate": 4.135798316594194e-06, + "loss": 2.3996, + "step": 13937 + }, + { + "epoch": 1.473984771573604, + "grad_norm": 8.4375, + "learning_rate": 4.135668805655059e-06, + "loss": 2.5222, + "step": 13938 + }, + { + "epoch": 1.474090524534687, + "grad_norm": 7.90625, + "learning_rate": 4.1355392870404e-06, + "loss": 2.4068, + "step": 13939 + }, + { + "epoch": 1.47419627749577, + "grad_norm": 8.3125, + "learning_rate": 4.135409760750824e-06, + "loss": 2.3198, + "step": 13940 + }, + { + "epoch": 1.4743020304568528, + "grad_norm": 8.5, + "learning_rate": 4.135280226786939e-06, + "loss": 2.6725, + "step": 13941 + }, + { + "epoch": 1.4744077834179357, + "grad_norm": 8.125, + "learning_rate": 4.135150685149353e-06, + "loss": 2.6224, + "step": 13942 + }, + { + "epoch": 1.4745135363790185, + "grad_norm": 8.8125, + "learning_rate": 4.135021135838672e-06, + "loss": 2.5174, + "step": 13943 + }, + { + "epoch": 1.4746192893401016, + "grad_norm": 8.5, + "learning_rate": 4.134891578855506e-06, + "loss": 2.4748, + "step": 13944 + }, + { + "epoch": 1.4747250423011844, + "grad_norm": 7.9375, + "learning_rate": 4.134762014200464e-06, + "loss": 2.4846, + "step": 13945 + }, + { + "epoch": 1.4748307952622675, + "grad_norm": 8.125, + "learning_rate": 4.134632441874151e-06, + "loss": 2.3095, + "step": 13946 + }, + { + "epoch": 1.4749365482233503, + "grad_norm": 8.6875, + "learning_rate": 4.134502861877177e-06, + "loss": 2.4346, + "step": 13947 + }, + { + "epoch": 1.4750423011844331, + "grad_norm": 8.6875, + "learning_rate": 4.13437327421015e-06, + "loss": 2.7706, + "step": 13948 + }, + { + "epoch": 1.475148054145516, + "grad_norm": 8.125, + "learning_rate": 4.134243678873677e-06, + "loss": 2.58, + "step": 13949 + }, + { + "epoch": 1.475253807106599, + "grad_norm": 8.125, + "learning_rate": 4.1341140758683674e-06, + "loss": 2.3461, + "step": 13950 + }, + { + "epoch": 1.4753595600676819, + "grad_norm": 8.1875, + "learning_rate": 4.133984465194828e-06, + "loss": 2.5553, + "step": 13951 + }, + { + "epoch": 1.475465313028765, + "grad_norm": 8.375, + "learning_rate": 4.133854846853668e-06, + "loss": 2.6, + "step": 13952 + }, + { + "epoch": 1.4755710659898478, + "grad_norm": 8.5, + "learning_rate": 4.133725220845496e-06, + "loss": 2.4878, + "step": 13953 + }, + { + "epoch": 1.4756768189509306, + "grad_norm": 8.5625, + "learning_rate": 4.1335955871709184e-06, + "loss": 2.3735, + "step": 13954 + }, + { + "epoch": 1.4757825719120135, + "grad_norm": 8.375, + "learning_rate": 4.133465945830546e-06, + "loss": 2.7572, + "step": 13955 + }, + { + "epoch": 1.4758883248730965, + "grad_norm": 7.8125, + "learning_rate": 4.133336296824986e-06, + "loss": 2.6148, + "step": 13956 + }, + { + "epoch": 1.4759940778341794, + "grad_norm": 9.875, + "learning_rate": 4.133206640154846e-06, + "loss": 2.5982, + "step": 13957 + }, + { + "epoch": 1.4760998307952622, + "grad_norm": 8.4375, + "learning_rate": 4.133076975820735e-06, + "loss": 2.6188, + "step": 13958 + }, + { + "epoch": 1.4762055837563453, + "grad_norm": 8.1875, + "learning_rate": 4.1329473038232635e-06, + "loss": 2.5714, + "step": 13959 + }, + { + "epoch": 1.476311336717428, + "grad_norm": 8.75, + "learning_rate": 4.132817624163037e-06, + "loss": 2.4779, + "step": 13960 + }, + { + "epoch": 1.476417089678511, + "grad_norm": 7.78125, + "learning_rate": 4.132687936840665e-06, + "loss": 2.4562, + "step": 13961 + }, + { + "epoch": 1.476522842639594, + "grad_norm": 8.125, + "learning_rate": 4.1325582418567565e-06, + "loss": 2.5866, + "step": 13962 + }, + { + "epoch": 1.4766285956006768, + "grad_norm": 8.125, + "learning_rate": 4.13242853921192e-06, + "loss": 2.474, + "step": 13963 + }, + { + "epoch": 1.4767343485617597, + "grad_norm": 8.4375, + "learning_rate": 4.132298828906763e-06, + "loss": 2.6117, + "step": 13964 + }, + { + "epoch": 1.4768401015228427, + "grad_norm": 8.3125, + "learning_rate": 4.132169110941896e-06, + "loss": 2.5765, + "step": 13965 + }, + { + "epoch": 1.4769458544839256, + "grad_norm": 8.5, + "learning_rate": 4.1320393853179275e-06, + "loss": 2.4166, + "step": 13966 + }, + { + "epoch": 1.4770516074450084, + "grad_norm": 8.9375, + "learning_rate": 4.1319096520354645e-06, + "loss": 2.6483, + "step": 13967 + }, + { + "epoch": 1.4771573604060912, + "grad_norm": 8.1875, + "learning_rate": 4.131779911095117e-06, + "loss": 2.5386, + "step": 13968 + }, + { + "epoch": 1.4772631133671743, + "grad_norm": 8.0, + "learning_rate": 4.131650162497495e-06, + "loss": 2.6409, + "step": 13969 + }, + { + "epoch": 1.4773688663282571, + "grad_norm": 8.625, + "learning_rate": 4.131520406243205e-06, + "loss": 2.5802, + "step": 13970 + }, + { + "epoch": 1.4774746192893402, + "grad_norm": 7.875, + "learning_rate": 4.131390642332857e-06, + "loss": 2.4111, + "step": 13971 + }, + { + "epoch": 1.477580372250423, + "grad_norm": 8.625, + "learning_rate": 4.131260870767061e-06, + "loss": 2.6763, + "step": 13972 + }, + { + "epoch": 1.4776861252115059, + "grad_norm": 7.65625, + "learning_rate": 4.131131091546424e-06, + "loss": 2.5243, + "step": 13973 + }, + { + "epoch": 1.4777918781725887, + "grad_norm": 8.625, + "learning_rate": 4.131001304671556e-06, + "loss": 2.1932, + "step": 13974 + }, + { + "epoch": 1.4778976311336718, + "grad_norm": 8.375, + "learning_rate": 4.1308715101430655e-06, + "loss": 2.2355, + "step": 13975 + }, + { + "epoch": 1.4780033840947546, + "grad_norm": 8.5, + "learning_rate": 4.130741707961563e-06, + "loss": 2.7341, + "step": 13976 + }, + { + "epoch": 1.4781091370558377, + "grad_norm": 8.0625, + "learning_rate": 4.130611898127656e-06, + "loss": 2.3442, + "step": 13977 + }, + { + "epoch": 1.4782148900169205, + "grad_norm": 7.875, + "learning_rate": 4.130482080641954e-06, + "loss": 2.4743, + "step": 13978 + }, + { + "epoch": 1.4783206429780034, + "grad_norm": 8.0625, + "learning_rate": 4.130352255505067e-06, + "loss": 2.6321, + "step": 13979 + }, + { + "epoch": 1.4784263959390862, + "grad_norm": 8.5, + "learning_rate": 4.130222422717604e-06, + "loss": 2.5191, + "step": 13980 + }, + { + "epoch": 1.4785321489001693, + "grad_norm": 7.9375, + "learning_rate": 4.130092582280173e-06, + "loss": 2.415, + "step": 13981 + }, + { + "epoch": 1.478637901861252, + "grad_norm": 8.0, + "learning_rate": 4.129962734193384e-06, + "loss": 2.5294, + "step": 13982 + }, + { + "epoch": 1.4787436548223352, + "grad_norm": 8.3125, + "learning_rate": 4.1298328784578474e-06, + "loss": 2.4826, + "step": 13983 + }, + { + "epoch": 1.478849407783418, + "grad_norm": 8.25, + "learning_rate": 4.129703015074171e-06, + "loss": 2.5654, + "step": 13984 + }, + { + "epoch": 1.4789551607445008, + "grad_norm": 9.0, + "learning_rate": 4.129573144042965e-06, + "loss": 2.8361, + "step": 13985 + }, + { + "epoch": 1.4790609137055837, + "grad_norm": 8.125, + "learning_rate": 4.1294432653648395e-06, + "loss": 2.489, + "step": 13986 + }, + { + "epoch": 1.4791666666666667, + "grad_norm": 8.4375, + "learning_rate": 4.1293133790404026e-06, + "loss": 2.8471, + "step": 13987 + }, + { + "epoch": 1.4792724196277496, + "grad_norm": 8.0625, + "learning_rate": 4.129183485070265e-06, + "loss": 2.4843, + "step": 13988 + }, + { + "epoch": 1.4793781725888324, + "grad_norm": 7.84375, + "learning_rate": 4.129053583455035e-06, + "loss": 2.4553, + "step": 13989 + }, + { + "epoch": 1.4794839255499155, + "grad_norm": 8.5625, + "learning_rate": 4.128923674195322e-06, + "loss": 2.483, + "step": 13990 + }, + { + "epoch": 1.4795896785109983, + "grad_norm": 9.375, + "learning_rate": 4.128793757291738e-06, + "loss": 2.3906, + "step": 13991 + }, + { + "epoch": 1.4796954314720812, + "grad_norm": 8.375, + "learning_rate": 4.12866383274489e-06, + "loss": 2.5708, + "step": 13992 + }, + { + "epoch": 1.479801184433164, + "grad_norm": 8.0625, + "learning_rate": 4.128533900555389e-06, + "loss": 2.4539, + "step": 13993 + }, + { + "epoch": 1.479906937394247, + "grad_norm": 8.3125, + "learning_rate": 4.128403960723844e-06, + "loss": 2.4612, + "step": 13994 + }, + { + "epoch": 1.48001269035533, + "grad_norm": 7.96875, + "learning_rate": 4.128274013250866e-06, + "loss": 2.5871, + "step": 13995 + }, + { + "epoch": 1.480118443316413, + "grad_norm": 8.0, + "learning_rate": 4.128144058137063e-06, + "loss": 2.4393, + "step": 13996 + }, + { + "epoch": 1.4802241962774958, + "grad_norm": 8.0, + "learning_rate": 4.128014095383047e-06, + "loss": 2.6626, + "step": 13997 + }, + { + "epoch": 1.4803299492385786, + "grad_norm": 7.8125, + "learning_rate": 4.127884124989426e-06, + "loss": 2.5116, + "step": 13998 + }, + { + "epoch": 1.4804357021996615, + "grad_norm": 8.25, + "learning_rate": 4.127754146956811e-06, + "loss": 2.6533, + "step": 13999 + }, + { + "epoch": 1.4805414551607445, + "grad_norm": 7.53125, + "learning_rate": 4.127624161285812e-06, + "loss": 2.4126, + "step": 14000 + }, + { + "epoch": 1.4806472081218274, + "grad_norm": 8.1875, + "learning_rate": 4.127494167977036e-06, + "loss": 2.6752, + "step": 14001 + }, + { + "epoch": 1.4807529610829104, + "grad_norm": 8.625, + "learning_rate": 4.127364167031099e-06, + "loss": 2.5724, + "step": 14002 + }, + { + "epoch": 1.4808587140439933, + "grad_norm": 8.6875, + "learning_rate": 4.127234158448604e-06, + "loss": 2.6847, + "step": 14003 + }, + { + "epoch": 1.4809644670050761, + "grad_norm": 8.25, + "learning_rate": 4.1271041422301675e-06, + "loss": 2.5552, + "step": 14004 + }, + { + "epoch": 1.481070219966159, + "grad_norm": 8.125, + "learning_rate": 4.126974118376396e-06, + "loss": 2.3488, + "step": 14005 + }, + { + "epoch": 1.481175972927242, + "grad_norm": 7.75, + "learning_rate": 4.126844086887899e-06, + "loss": 2.6481, + "step": 14006 + }, + { + "epoch": 1.4812817258883249, + "grad_norm": 8.3125, + "learning_rate": 4.12671404776529e-06, + "loss": 2.7777, + "step": 14007 + }, + { + "epoch": 1.481387478849408, + "grad_norm": 8.625, + "learning_rate": 4.126584001009175e-06, + "loss": 2.7198, + "step": 14008 + }, + { + "epoch": 1.4814932318104908, + "grad_norm": 8.25, + "learning_rate": 4.1264539466201685e-06, + "loss": 2.545, + "step": 14009 + }, + { + "epoch": 1.4815989847715736, + "grad_norm": 8.125, + "learning_rate": 4.126323884598878e-06, + "loss": 2.1671, + "step": 14010 + }, + { + "epoch": 1.4817047377326564, + "grad_norm": 8.0, + "learning_rate": 4.126193814945915e-06, + "loss": 2.4416, + "step": 14011 + }, + { + "epoch": 1.4818104906937395, + "grad_norm": 8.5625, + "learning_rate": 4.12606373766189e-06, + "loss": 2.6569, + "step": 14012 + }, + { + "epoch": 1.4819162436548223, + "grad_norm": 7.71875, + "learning_rate": 4.125933652747411e-06, + "loss": 2.5277, + "step": 14013 + }, + { + "epoch": 1.4820219966159052, + "grad_norm": 8.5625, + "learning_rate": 4.125803560203092e-06, + "loss": 2.5333, + "step": 14014 + }, + { + "epoch": 1.4821277495769882, + "grad_norm": 8.375, + "learning_rate": 4.125673460029542e-06, + "loss": 2.3616, + "step": 14015 + }, + { + "epoch": 1.482233502538071, + "grad_norm": 8.5625, + "learning_rate": 4.12554335222737e-06, + "loss": 2.4961, + "step": 14016 + }, + { + "epoch": 1.482339255499154, + "grad_norm": 9.0, + "learning_rate": 4.125413236797189e-06, + "loss": 2.6684, + "step": 14017 + }, + { + "epoch": 1.482445008460237, + "grad_norm": 8.25, + "learning_rate": 4.125283113739608e-06, + "loss": 2.5453, + "step": 14018 + }, + { + "epoch": 1.4825507614213198, + "grad_norm": 8.5625, + "learning_rate": 4.125152983055238e-06, + "loss": 2.4578, + "step": 14019 + }, + { + "epoch": 1.4826565143824026, + "grad_norm": 7.84375, + "learning_rate": 4.12502284474469e-06, + "loss": 2.4781, + "step": 14020 + }, + { + "epoch": 1.4827622673434857, + "grad_norm": 8.875, + "learning_rate": 4.124892698808574e-06, + "loss": 2.4426, + "step": 14021 + }, + { + "epoch": 1.4828680203045685, + "grad_norm": 8.6875, + "learning_rate": 4.124762545247501e-06, + "loss": 2.4184, + "step": 14022 + }, + { + "epoch": 1.4829737732656514, + "grad_norm": 8.5625, + "learning_rate": 4.1246323840620825e-06, + "loss": 2.5304, + "step": 14023 + }, + { + "epoch": 1.4830795262267342, + "grad_norm": 7.90625, + "learning_rate": 4.124502215252929e-06, + "loss": 2.4075, + "step": 14024 + }, + { + "epoch": 1.4831852791878173, + "grad_norm": 7.8125, + "learning_rate": 4.124372038820651e-06, + "loss": 2.3848, + "step": 14025 + }, + { + "epoch": 1.4832910321489001, + "grad_norm": 8.5, + "learning_rate": 4.124241854765858e-06, + "loss": 2.5212, + "step": 14026 + }, + { + "epoch": 1.4833967851099832, + "grad_norm": 9.0, + "learning_rate": 4.1241116630891634e-06, + "loss": 2.7196, + "step": 14027 + }, + { + "epoch": 1.483502538071066, + "grad_norm": 7.6875, + "learning_rate": 4.123981463791176e-06, + "loss": 2.4089, + "step": 14028 + }, + { + "epoch": 1.4836082910321489, + "grad_norm": 8.125, + "learning_rate": 4.123851256872509e-06, + "loss": 2.5196, + "step": 14029 + }, + { + "epoch": 1.4837140439932317, + "grad_norm": 8.25, + "learning_rate": 4.1237210423337715e-06, + "loss": 2.6616, + "step": 14030 + }, + { + "epoch": 1.4838197969543148, + "grad_norm": 7.96875, + "learning_rate": 4.1235908201755755e-06, + "loss": 2.4987, + "step": 14031 + }, + { + "epoch": 1.4839255499153976, + "grad_norm": 8.5625, + "learning_rate": 4.1234605903985315e-06, + "loss": 2.4933, + "step": 14032 + }, + { + "epoch": 1.4840313028764807, + "grad_norm": 8.0625, + "learning_rate": 4.123330353003251e-06, + "loss": 2.5625, + "step": 14033 + }, + { + "epoch": 1.4841370558375635, + "grad_norm": 7.8125, + "learning_rate": 4.123200107990345e-06, + "loss": 2.4772, + "step": 14034 + }, + { + "epoch": 1.4842428087986463, + "grad_norm": 8.5625, + "learning_rate": 4.123069855360425e-06, + "loss": 2.6211, + "step": 14035 + }, + { + "epoch": 1.4843485617597292, + "grad_norm": 9.25, + "learning_rate": 4.122939595114102e-06, + "loss": 2.6989, + "step": 14036 + }, + { + "epoch": 1.4844543147208122, + "grad_norm": 8.0625, + "learning_rate": 4.1228093272519874e-06, + "loss": 2.4423, + "step": 14037 + }, + { + "epoch": 1.484560067681895, + "grad_norm": 8.625, + "learning_rate": 4.122679051774692e-06, + "loss": 2.6541, + "step": 14038 + }, + { + "epoch": 1.4846658206429781, + "grad_norm": 8.25, + "learning_rate": 4.122548768682828e-06, + "loss": 2.5943, + "step": 14039 + }, + { + "epoch": 1.484771573604061, + "grad_norm": 8.25, + "learning_rate": 4.122418477977005e-06, + "loss": 2.6375, + "step": 14040 + }, + { + "epoch": 1.4848773265651438, + "grad_norm": 9.4375, + "learning_rate": 4.122288179657836e-06, + "loss": 2.8306, + "step": 14041 + }, + { + "epoch": 1.4849830795262267, + "grad_norm": 8.5625, + "learning_rate": 4.1221578737259336e-06, + "loss": 2.7739, + "step": 14042 + }, + { + "epoch": 1.4850888324873097, + "grad_norm": 8.125, + "learning_rate": 4.122027560181906e-06, + "loss": 2.5339, + "step": 14043 + }, + { + "epoch": 1.4851945854483926, + "grad_norm": 8.5625, + "learning_rate": 4.121897239026367e-06, + "loss": 2.5464, + "step": 14044 + }, + { + "epoch": 1.4853003384094754, + "grad_norm": 8.625, + "learning_rate": 4.121766910259928e-06, + "loss": 2.5217, + "step": 14045 + }, + { + "epoch": 1.4854060913705585, + "grad_norm": 8.4375, + "learning_rate": 4.1216365738832e-06, + "loss": 2.6851, + "step": 14046 + }, + { + "epoch": 1.4855118443316413, + "grad_norm": 8.5625, + "learning_rate": 4.121506229896794e-06, + "loss": 2.65, + "step": 14047 + }, + { + "epoch": 1.4856175972927241, + "grad_norm": 8.0625, + "learning_rate": 4.1213758783013235e-06, + "loss": 2.6227, + "step": 14048 + }, + { + "epoch": 1.485723350253807, + "grad_norm": 7.9375, + "learning_rate": 4.121245519097399e-06, + "loss": 2.6012, + "step": 14049 + }, + { + "epoch": 1.48582910321489, + "grad_norm": 8.6875, + "learning_rate": 4.1211151522856315e-06, + "loss": 2.6998, + "step": 14050 + }, + { + "epoch": 1.4859348561759729, + "grad_norm": 8.6875, + "learning_rate": 4.120984777866634e-06, + "loss": 2.7537, + "step": 14051 + }, + { + "epoch": 1.486040609137056, + "grad_norm": 8.0, + "learning_rate": 4.120854395841019e-06, + "loss": 2.4574, + "step": 14052 + }, + { + "epoch": 1.4861463620981388, + "grad_norm": 8.6875, + "learning_rate": 4.120724006209395e-06, + "loss": 2.5166, + "step": 14053 + }, + { + "epoch": 1.4862521150592216, + "grad_norm": 7.9375, + "learning_rate": 4.120593608972377e-06, + "loss": 2.4723, + "step": 14054 + }, + { + "epoch": 1.4863578680203045, + "grad_norm": 8.0625, + "learning_rate": 4.120463204130577e-06, + "loss": 2.3436, + "step": 14055 + }, + { + "epoch": 1.4864636209813875, + "grad_norm": 8.125, + "learning_rate": 4.1203327916846044e-06, + "loss": 2.5502, + "step": 14056 + }, + { + "epoch": 1.4865693739424704, + "grad_norm": 8.1875, + "learning_rate": 4.120202371635074e-06, + "loss": 2.7193, + "step": 14057 + }, + { + "epoch": 1.4866751269035534, + "grad_norm": 7.78125, + "learning_rate": 4.1200719439825965e-06, + "loss": 2.5367, + "step": 14058 + }, + { + "epoch": 1.4867808798646363, + "grad_norm": 8.125, + "learning_rate": 4.119941508727784e-06, + "loss": 2.3523, + "step": 14059 + }, + { + "epoch": 1.486886632825719, + "grad_norm": 8.9375, + "learning_rate": 4.119811065871248e-06, + "loss": 2.9283, + "step": 14060 + }, + { + "epoch": 1.486992385786802, + "grad_norm": 8.5625, + "learning_rate": 4.119680615413601e-06, + "loss": 2.4866, + "step": 14061 + }, + { + "epoch": 1.487098138747885, + "grad_norm": 8.3125, + "learning_rate": 4.1195501573554555e-06, + "loss": 2.4293, + "step": 14062 + }, + { + "epoch": 1.4872038917089678, + "grad_norm": 8.25, + "learning_rate": 4.119419691697423e-06, + "loss": 2.5563, + "step": 14063 + }, + { + "epoch": 1.487309644670051, + "grad_norm": 8.375, + "learning_rate": 4.119289218440117e-06, + "loss": 2.5011, + "step": 14064 + }, + { + "epoch": 1.4874153976311337, + "grad_norm": 8.0625, + "learning_rate": 4.119158737584148e-06, + "loss": 2.4535, + "step": 14065 + }, + { + "epoch": 1.4875211505922166, + "grad_norm": 8.125, + "learning_rate": 4.119028249130131e-06, + "loss": 2.5979, + "step": 14066 + }, + { + "epoch": 1.4876269035532994, + "grad_norm": 8.25, + "learning_rate": 4.118897753078675e-06, + "loss": 2.4962, + "step": 14067 + }, + { + "epoch": 1.4877326565143825, + "grad_norm": 8.5, + "learning_rate": 4.118767249430394e-06, + "loss": 2.6727, + "step": 14068 + }, + { + "epoch": 1.4878384094754653, + "grad_norm": 8.1875, + "learning_rate": 4.118636738185902e-06, + "loss": 2.4786, + "step": 14069 + }, + { + "epoch": 1.4879441624365481, + "grad_norm": 7.875, + "learning_rate": 4.118506219345808e-06, + "loss": 2.5106, + "step": 14070 + }, + { + "epoch": 1.4880499153976312, + "grad_norm": 8.0625, + "learning_rate": 4.118375692910726e-06, + "loss": 2.4334, + "step": 14071 + }, + { + "epoch": 1.488155668358714, + "grad_norm": 7.90625, + "learning_rate": 4.11824515888127e-06, + "loss": 2.5509, + "step": 14072 + }, + { + "epoch": 1.4882614213197969, + "grad_norm": 8.6875, + "learning_rate": 4.118114617258051e-06, + "loss": 2.5398, + "step": 14073 + }, + { + "epoch": 1.48836717428088, + "grad_norm": 8.6875, + "learning_rate": 4.117984068041681e-06, + "loss": 2.8572, + "step": 14074 + }, + { + "epoch": 1.4884729272419628, + "grad_norm": 8.3125, + "learning_rate": 4.117853511232775e-06, + "loss": 2.7189, + "step": 14075 + }, + { + "epoch": 1.4885786802030456, + "grad_norm": 8.75, + "learning_rate": 4.1177229468319435e-06, + "loss": 2.6437, + "step": 14076 + }, + { + "epoch": 1.4886844331641287, + "grad_norm": 8.1875, + "learning_rate": 4.1175923748398e-06, + "loss": 2.5662, + "step": 14077 + }, + { + "epoch": 1.4887901861252115, + "grad_norm": 8.4375, + "learning_rate": 4.117461795256957e-06, + "loss": 2.5061, + "step": 14078 + }, + { + "epoch": 1.4888959390862944, + "grad_norm": 8.125, + "learning_rate": 4.1173312080840274e-06, + "loss": 2.6945, + "step": 14079 + }, + { + "epoch": 1.4890016920473772, + "grad_norm": 7.875, + "learning_rate": 4.117200613321623e-06, + "loss": 2.5182, + "step": 14080 + }, + { + "epoch": 1.4891074450084603, + "grad_norm": 8.125, + "learning_rate": 4.117070010970359e-06, + "loss": 2.4132, + "step": 14081 + }, + { + "epoch": 1.489213197969543, + "grad_norm": 7.6875, + "learning_rate": 4.116939401030845e-06, + "loss": 2.4653, + "step": 14082 + }, + { + "epoch": 1.4893189509306262, + "grad_norm": 8.0625, + "learning_rate": 4.116808783503698e-06, + "loss": 2.2718, + "step": 14083 + }, + { + "epoch": 1.489424703891709, + "grad_norm": 8.4375, + "learning_rate": 4.116678158389526e-06, + "loss": 2.5093, + "step": 14084 + }, + { + "epoch": 1.4895304568527918, + "grad_norm": 8.25, + "learning_rate": 4.116547525688947e-06, + "loss": 2.55, + "step": 14085 + }, + { + "epoch": 1.4896362098138747, + "grad_norm": 8.5, + "learning_rate": 4.11641688540257e-06, + "loss": 2.8363, + "step": 14086 + }, + { + "epoch": 1.4897419627749577, + "grad_norm": 8.5625, + "learning_rate": 4.116286237531011e-06, + "loss": 2.1649, + "step": 14087 + }, + { + "epoch": 1.4898477157360406, + "grad_norm": 8.25, + "learning_rate": 4.11615558207488e-06, + "loss": 2.703, + "step": 14088 + }, + { + "epoch": 1.4899534686971236, + "grad_norm": 7.78125, + "learning_rate": 4.116024919034793e-06, + "loss": 2.4878, + "step": 14089 + }, + { + "epoch": 1.4900592216582065, + "grad_norm": 8.5, + "learning_rate": 4.115894248411362e-06, + "loss": 2.4508, + "step": 14090 + }, + { + "epoch": 1.4901649746192893, + "grad_norm": 8.5625, + "learning_rate": 4.1157635702052e-06, + "loss": 2.6069, + "step": 14091 + }, + { + "epoch": 1.4902707275803722, + "grad_norm": 9.0625, + "learning_rate": 4.115632884416919e-06, + "loss": 2.7342, + "step": 14092 + }, + { + "epoch": 1.4903764805414552, + "grad_norm": 8.375, + "learning_rate": 4.115502191047136e-06, + "loss": 2.7325, + "step": 14093 + }, + { + "epoch": 1.490482233502538, + "grad_norm": 8.25, + "learning_rate": 4.11537149009646e-06, + "loss": 2.4061, + "step": 14094 + }, + { + "epoch": 1.4905879864636211, + "grad_norm": 8.0, + "learning_rate": 4.1152407815655084e-06, + "loss": 2.556, + "step": 14095 + }, + { + "epoch": 1.490693739424704, + "grad_norm": 8.3125, + "learning_rate": 4.11511006545489e-06, + "loss": 2.2656, + "step": 14096 + }, + { + "epoch": 1.4907994923857868, + "grad_norm": 7.84375, + "learning_rate": 4.114979341765223e-06, + "loss": 2.4686, + "step": 14097 + }, + { + "epoch": 1.4909052453468696, + "grad_norm": 8.625, + "learning_rate": 4.114848610497117e-06, + "loss": 2.5459, + "step": 14098 + }, + { + "epoch": 1.4910109983079527, + "grad_norm": 8.0625, + "learning_rate": 4.114717871651186e-06, + "loss": 2.5631, + "step": 14099 + }, + { + "epoch": 1.4911167512690355, + "grad_norm": 8.6875, + "learning_rate": 4.114587125228046e-06, + "loss": 2.543, + "step": 14100 + }, + { + "epoch": 1.4912225042301184, + "grad_norm": 8.125, + "learning_rate": 4.114456371228309e-06, + "loss": 2.4214, + "step": 14101 + }, + { + "epoch": 1.4913282571912014, + "grad_norm": 8.6875, + "learning_rate": 4.114325609652588e-06, + "loss": 2.7136, + "step": 14102 + }, + { + "epoch": 1.4914340101522843, + "grad_norm": 8.25, + "learning_rate": 4.114194840501497e-06, + "loss": 2.5107, + "step": 14103 + }, + { + "epoch": 1.4915397631133671, + "grad_norm": 8.1875, + "learning_rate": 4.11406406377565e-06, + "loss": 2.6306, + "step": 14104 + }, + { + "epoch": 1.49164551607445, + "grad_norm": 8.25, + "learning_rate": 4.1139332794756605e-06, + "loss": 2.5399, + "step": 14105 + }, + { + "epoch": 1.491751269035533, + "grad_norm": 9.125, + "learning_rate": 4.113802487602143e-06, + "loss": 2.4255, + "step": 14106 + }, + { + "epoch": 1.4918570219966159, + "grad_norm": 9.625, + "learning_rate": 4.113671688155709e-06, + "loss": 2.6816, + "step": 14107 + }, + { + "epoch": 1.491962774957699, + "grad_norm": 8.5, + "learning_rate": 4.113540881136975e-06, + "loss": 2.5573, + "step": 14108 + }, + { + "epoch": 1.4920685279187818, + "grad_norm": 9.25, + "learning_rate": 4.113410066546552e-06, + "loss": 2.671, + "step": 14109 + }, + { + "epoch": 1.4921742808798646, + "grad_norm": 8.0, + "learning_rate": 4.113279244385057e-06, + "loss": 2.3539, + "step": 14110 + }, + { + "epoch": 1.4922800338409474, + "grad_norm": 7.9375, + "learning_rate": 4.1131484146531015e-06, + "loss": 2.6256, + "step": 14111 + }, + { + "epoch": 1.4923857868020305, + "grad_norm": 8.375, + "learning_rate": 4.1130175773513e-06, + "loss": 2.4789, + "step": 14112 + }, + { + "epoch": 1.4924915397631133, + "grad_norm": 8.6875, + "learning_rate": 4.112886732480268e-06, + "loss": 2.4145, + "step": 14113 + }, + { + "epoch": 1.4925972927241964, + "grad_norm": 8.8125, + "learning_rate": 4.112755880040617e-06, + "loss": 2.9091, + "step": 14114 + }, + { + "epoch": 1.4927030456852792, + "grad_norm": 8.5, + "learning_rate": 4.112625020032963e-06, + "loss": 2.5783, + "step": 14115 + }, + { + "epoch": 1.492808798646362, + "grad_norm": 7.59375, + "learning_rate": 4.112494152457919e-06, + "loss": 2.4951, + "step": 14116 + }, + { + "epoch": 1.492914551607445, + "grad_norm": 9.1875, + "learning_rate": 4.1123632773160994e-06, + "loss": 2.8628, + "step": 14117 + }, + { + "epoch": 1.493020304568528, + "grad_norm": 8.0625, + "learning_rate": 4.112232394608118e-06, + "loss": 2.5317, + "step": 14118 + }, + { + "epoch": 1.4931260575296108, + "grad_norm": 8.75, + "learning_rate": 4.112101504334589e-06, + "loss": 2.4362, + "step": 14119 + }, + { + "epoch": 1.4932318104906939, + "grad_norm": 8.6875, + "learning_rate": 4.1119706064961284e-06, + "loss": 2.5435, + "step": 14120 + }, + { + "epoch": 1.4933375634517767, + "grad_norm": 9.0625, + "learning_rate": 4.111839701093348e-06, + "loss": 2.3302, + "step": 14121 + }, + { + "epoch": 1.4934433164128595, + "grad_norm": 8.625, + "learning_rate": 4.111708788126864e-06, + "loss": 2.5773, + "step": 14122 + }, + { + "epoch": 1.4935490693739424, + "grad_norm": 8.0625, + "learning_rate": 4.111577867597289e-06, + "loss": 2.4472, + "step": 14123 + }, + { + "epoch": 1.4936548223350254, + "grad_norm": 8.1875, + "learning_rate": 4.1114469395052385e-06, + "loss": 2.5888, + "step": 14124 + }, + { + "epoch": 1.4937605752961083, + "grad_norm": 9.0, + "learning_rate": 4.111316003851325e-06, + "loss": 2.514, + "step": 14125 + }, + { + "epoch": 1.4938663282571911, + "grad_norm": 8.0625, + "learning_rate": 4.111185060636167e-06, + "loss": 2.5705, + "step": 14126 + }, + { + "epoch": 1.4939720812182742, + "grad_norm": 9.1875, + "learning_rate": 4.111054109860375e-06, + "loss": 2.448, + "step": 14127 + }, + { + "epoch": 1.494077834179357, + "grad_norm": 8.375, + "learning_rate": 4.110923151524566e-06, + "loss": 2.5028, + "step": 14128 + }, + { + "epoch": 1.4941835871404399, + "grad_norm": 8.8125, + "learning_rate": 4.110792185629353e-06, + "loss": 2.7075, + "step": 14129 + }, + { + "epoch": 1.494289340101523, + "grad_norm": 8.0, + "learning_rate": 4.110661212175351e-06, + "loss": 2.7432, + "step": 14130 + }, + { + "epoch": 1.4943950930626058, + "grad_norm": 8.375, + "learning_rate": 4.110530231163175e-06, + "loss": 2.5493, + "step": 14131 + }, + { + "epoch": 1.4945008460236886, + "grad_norm": 8.6875, + "learning_rate": 4.110399242593438e-06, + "loss": 2.678, + "step": 14132 + }, + { + "epoch": 1.4946065989847717, + "grad_norm": 8.9375, + "learning_rate": 4.110268246466757e-06, + "loss": 2.413, + "step": 14133 + }, + { + "epoch": 1.4947123519458545, + "grad_norm": 8.125, + "learning_rate": 4.110137242783745e-06, + "loss": 2.3439, + "step": 14134 + }, + { + "epoch": 1.4948181049069373, + "grad_norm": 8.6875, + "learning_rate": 4.110006231545018e-06, + "loss": 2.2715, + "step": 14135 + }, + { + "epoch": 1.4949238578680202, + "grad_norm": 8.8125, + "learning_rate": 4.109875212751191e-06, + "loss": 2.5149, + "step": 14136 + }, + { + "epoch": 1.4950296108291032, + "grad_norm": 7.96875, + "learning_rate": 4.1097441864028775e-06, + "loss": 2.3722, + "step": 14137 + }, + { + "epoch": 1.495135363790186, + "grad_norm": 8.25, + "learning_rate": 4.109613152500692e-06, + "loss": 2.3998, + "step": 14138 + }, + { + "epoch": 1.4952411167512691, + "grad_norm": 8.4375, + "learning_rate": 4.109482111045251e-06, + "loss": 2.4909, + "step": 14139 + }, + { + "epoch": 1.495346869712352, + "grad_norm": 8.1875, + "learning_rate": 4.109351062037169e-06, + "loss": 2.451, + "step": 14140 + }, + { + "epoch": 1.4954526226734348, + "grad_norm": 8.0, + "learning_rate": 4.10922000547706e-06, + "loss": 2.3129, + "step": 14141 + }, + { + "epoch": 1.4955583756345177, + "grad_norm": 8.1875, + "learning_rate": 4.109088941365541e-06, + "loss": 2.6504, + "step": 14142 + }, + { + "epoch": 1.4956641285956007, + "grad_norm": 8.75, + "learning_rate": 4.108957869703224e-06, + "loss": 2.6115, + "step": 14143 + }, + { + "epoch": 1.4957698815566836, + "grad_norm": 8.25, + "learning_rate": 4.108826790490726e-06, + "loss": 2.7464, + "step": 14144 + }, + { + "epoch": 1.4958756345177666, + "grad_norm": 8.3125, + "learning_rate": 4.108695703728663e-06, + "loss": 2.5595, + "step": 14145 + }, + { + "epoch": 1.4959813874788495, + "grad_norm": 11.25, + "learning_rate": 4.108564609417648e-06, + "loss": 2.4932, + "step": 14146 + }, + { + "epoch": 1.4960871404399323, + "grad_norm": 8.125, + "learning_rate": 4.108433507558298e-06, + "loss": 2.6756, + "step": 14147 + }, + { + "epoch": 1.4961928934010151, + "grad_norm": 8.0625, + "learning_rate": 4.108302398151227e-06, + "loss": 2.6393, + "step": 14148 + }, + { + "epoch": 1.4962986463620982, + "grad_norm": 7.875, + "learning_rate": 4.108171281197051e-06, + "loss": 2.4611, + "step": 14149 + }, + { + "epoch": 1.496404399323181, + "grad_norm": 8.0, + "learning_rate": 4.108040156696384e-06, + "loss": 2.4498, + "step": 14150 + }, + { + "epoch": 1.496510152284264, + "grad_norm": 8.0, + "learning_rate": 4.107909024649842e-06, + "loss": 2.5409, + "step": 14151 + }, + { + "epoch": 1.496615905245347, + "grad_norm": 8.5625, + "learning_rate": 4.107777885058041e-06, + "loss": 2.7008, + "step": 14152 + }, + { + "epoch": 1.4967216582064298, + "grad_norm": 8.4375, + "learning_rate": 4.107646737921597e-06, + "loss": 2.7605, + "step": 14153 + }, + { + "epoch": 1.4968274111675126, + "grad_norm": 9.4375, + "learning_rate": 4.107515583241123e-06, + "loss": 2.5332, + "step": 14154 + }, + { + "epoch": 1.4969331641285957, + "grad_norm": 7.875, + "learning_rate": 4.107384421017236e-06, + "loss": 2.5214, + "step": 14155 + }, + { + "epoch": 1.4970389170896785, + "grad_norm": 8.25, + "learning_rate": 4.107253251250552e-06, + "loss": 2.5309, + "step": 14156 + }, + { + "epoch": 1.4971446700507614, + "grad_norm": 8.4375, + "learning_rate": 4.1071220739416854e-06, + "loss": 2.5139, + "step": 14157 + }, + { + "epoch": 1.4972504230118444, + "grad_norm": 8.5, + "learning_rate": 4.106990889091253e-06, + "loss": 2.5833, + "step": 14158 + }, + { + "epoch": 1.4973561759729273, + "grad_norm": 8.375, + "learning_rate": 4.106859696699868e-06, + "loss": 2.5111, + "step": 14159 + }, + { + "epoch": 1.49746192893401, + "grad_norm": 8.4375, + "learning_rate": 4.106728496768149e-06, + "loss": 2.5262, + "step": 14160 + }, + { + "epoch": 1.497567681895093, + "grad_norm": 8.5625, + "learning_rate": 4.1065972892967084e-06, + "loss": 2.3853, + "step": 14161 + }, + { + "epoch": 1.497673434856176, + "grad_norm": 8.75, + "learning_rate": 4.106466074286166e-06, + "loss": 2.4556, + "step": 14162 + }, + { + "epoch": 1.4977791878172588, + "grad_norm": 8.625, + "learning_rate": 4.106334851737135e-06, + "loss": 2.6481, + "step": 14163 + }, + { + "epoch": 1.497884940778342, + "grad_norm": 8.0, + "learning_rate": 4.106203621650231e-06, + "loss": 2.5679, + "step": 14164 + }, + { + "epoch": 1.4979906937394247, + "grad_norm": 8.0625, + "learning_rate": 4.106072384026069e-06, + "loss": 2.5214, + "step": 14165 + }, + { + "epoch": 1.4980964467005076, + "grad_norm": 8.625, + "learning_rate": 4.105941138865268e-06, + "loss": 2.5241, + "step": 14166 + }, + { + "epoch": 1.4982021996615904, + "grad_norm": 8.125, + "learning_rate": 4.105809886168442e-06, + "loss": 2.4244, + "step": 14167 + }, + { + "epoch": 1.4983079526226735, + "grad_norm": 8.6875, + "learning_rate": 4.105678625936206e-06, + "loss": 2.6453, + "step": 14168 + }, + { + "epoch": 1.4984137055837563, + "grad_norm": 8.125, + "learning_rate": 4.105547358169177e-06, + "loss": 2.5709, + "step": 14169 + }, + { + "epoch": 1.4985194585448394, + "grad_norm": 8.125, + "learning_rate": 4.105416082867972e-06, + "loss": 2.6203, + "step": 14170 + }, + { + "epoch": 1.4986252115059222, + "grad_norm": 8.875, + "learning_rate": 4.105284800033205e-06, + "loss": 2.7252, + "step": 14171 + }, + { + "epoch": 1.498730964467005, + "grad_norm": 7.75, + "learning_rate": 4.105153509665493e-06, + "loss": 2.462, + "step": 14172 + }, + { + "epoch": 1.4988367174280879, + "grad_norm": 8.6875, + "learning_rate": 4.1050222117654525e-06, + "loss": 2.3716, + "step": 14173 + }, + { + "epoch": 1.498942470389171, + "grad_norm": 7.875, + "learning_rate": 4.1048909063337e-06, + "loss": 2.5733, + "step": 14174 + }, + { + "epoch": 1.4990482233502538, + "grad_norm": 9.3125, + "learning_rate": 4.104759593370848e-06, + "loss": 2.6706, + "step": 14175 + }, + { + "epoch": 1.4991539763113368, + "grad_norm": 7.875, + "learning_rate": 4.104628272877517e-06, + "loss": 2.5376, + "step": 14176 + }, + { + "epoch": 1.4992597292724197, + "grad_norm": 8.0625, + "learning_rate": 4.1044969448543235e-06, + "loss": 2.5366, + "step": 14177 + }, + { + "epoch": 1.4993654822335025, + "grad_norm": 8.375, + "learning_rate": 4.10436560930188e-06, + "loss": 2.5823, + "step": 14178 + }, + { + "epoch": 1.4994712351945854, + "grad_norm": 8.3125, + "learning_rate": 4.104234266220805e-06, + "loss": 2.7272, + "step": 14179 + }, + { + "epoch": 1.4995769881556684, + "grad_norm": 8.6875, + "learning_rate": 4.104102915611716e-06, + "loss": 2.6746, + "step": 14180 + }, + { + "epoch": 1.4996827411167513, + "grad_norm": 8.625, + "learning_rate": 4.103971557475227e-06, + "loss": 2.4296, + "step": 14181 + }, + { + "epoch": 1.499788494077834, + "grad_norm": 8.3125, + "learning_rate": 4.103840191811957e-06, + "loss": 2.6651, + "step": 14182 + }, + { + "epoch": 1.4998942470389172, + "grad_norm": 9.0625, + "learning_rate": 4.103708818622519e-06, + "loss": 2.7019, + "step": 14183 + }, + { + "epoch": 1.5, + "grad_norm": 8.0625, + "learning_rate": 4.1035774379075325e-06, + "loss": 2.4663, + "step": 14184 + }, + { + "epoch": 1.5001057529610828, + "grad_norm": 8.4375, + "learning_rate": 4.103446049667612e-06, + "loss": 2.2766, + "step": 14185 + }, + { + "epoch": 1.5002115059221657, + "grad_norm": 7.8125, + "learning_rate": 4.103314653903376e-06, + "loss": 2.5432, + "step": 14186 + }, + { + "epoch": 1.5003172588832487, + "grad_norm": 8.4375, + "learning_rate": 4.103183250615439e-06, + "loss": 2.672, + "step": 14187 + }, + { + "epoch": 1.5004230118443318, + "grad_norm": 8.375, + "learning_rate": 4.10305183980442e-06, + "loss": 2.7279, + "step": 14188 + }, + { + "epoch": 1.5005287648054146, + "grad_norm": 8.125, + "learning_rate": 4.102920421470933e-06, + "loss": 2.3306, + "step": 14189 + }, + { + "epoch": 1.5006345177664975, + "grad_norm": 8.4375, + "learning_rate": 4.102788995615596e-06, + "loss": 2.4334, + "step": 14190 + }, + { + "epoch": 1.5007402707275803, + "grad_norm": 8.25, + "learning_rate": 4.102657562239027e-06, + "loss": 2.6254, + "step": 14191 + }, + { + "epoch": 1.5008460236886632, + "grad_norm": 8.1875, + "learning_rate": 4.102526121341841e-06, + "loss": 2.4852, + "step": 14192 + }, + { + "epoch": 1.5009517766497462, + "grad_norm": 8.125, + "learning_rate": 4.102394672924656e-06, + "loss": 2.6142, + "step": 14193 + }, + { + "epoch": 1.501057529610829, + "grad_norm": 10.875, + "learning_rate": 4.102263216988087e-06, + "loss": 2.3484, + "step": 14194 + }, + { + "epoch": 1.5011632825719121, + "grad_norm": 7.96875, + "learning_rate": 4.102131753532752e-06, + "loss": 2.4832, + "step": 14195 + }, + { + "epoch": 1.501269035532995, + "grad_norm": 8.5625, + "learning_rate": 4.1020002825592686e-06, + "loss": 2.8022, + "step": 14196 + }, + { + "epoch": 1.5013747884940778, + "grad_norm": 8.5625, + "learning_rate": 4.101868804068253e-06, + "loss": 2.6477, + "step": 14197 + }, + { + "epoch": 1.5014805414551606, + "grad_norm": 7.90625, + "learning_rate": 4.1017373180603216e-06, + "loss": 2.48, + "step": 14198 + }, + { + "epoch": 1.5015862944162437, + "grad_norm": 8.375, + "learning_rate": 4.101605824536092e-06, + "loss": 2.6284, + "step": 14199 + }, + { + "epoch": 1.5016920473773265, + "grad_norm": 7.6875, + "learning_rate": 4.101474323496183e-06, + "loss": 2.1416, + "step": 14200 + }, + { + "epoch": 1.5017978003384096, + "grad_norm": 8.125, + "learning_rate": 4.101342814941208e-06, + "loss": 2.4529, + "step": 14201 + }, + { + "epoch": 1.5019035532994924, + "grad_norm": 8.1875, + "learning_rate": 4.101211298871787e-06, + "loss": 2.6483, + "step": 14202 + }, + { + "epoch": 1.5020093062605753, + "grad_norm": 8.3125, + "learning_rate": 4.101079775288537e-06, + "loss": 2.5467, + "step": 14203 + }, + { + "epoch": 1.502115059221658, + "grad_norm": 8.125, + "learning_rate": 4.100948244192073e-06, + "loss": 2.5648, + "step": 14204 + }, + { + "epoch": 1.5022208121827412, + "grad_norm": 8.1875, + "learning_rate": 4.100816705583014e-06, + "loss": 2.7635, + "step": 14205 + }, + { + "epoch": 1.502326565143824, + "grad_norm": 8.6875, + "learning_rate": 4.100685159461978e-06, + "loss": 2.7434, + "step": 14206 + }, + { + "epoch": 1.502432318104907, + "grad_norm": 8.25, + "learning_rate": 4.100553605829579e-06, + "loss": 2.3713, + "step": 14207 + }, + { + "epoch": 1.50253807106599, + "grad_norm": 8.125, + "learning_rate": 4.1004220446864385e-06, + "loss": 2.5283, + "step": 14208 + }, + { + "epoch": 1.5026438240270727, + "grad_norm": 8.0, + "learning_rate": 4.100290476033172e-06, + "loss": 2.4039, + "step": 14209 + }, + { + "epoch": 1.5027495769881556, + "grad_norm": 8.4375, + "learning_rate": 4.100158899870395e-06, + "loss": 2.5979, + "step": 14210 + }, + { + "epoch": 1.5028553299492384, + "grad_norm": 8.625, + "learning_rate": 4.1000273161987284e-06, + "loss": 2.6084, + "step": 14211 + }, + { + "epoch": 1.5029610829103215, + "grad_norm": 8.9375, + "learning_rate": 4.099895725018788e-06, + "loss": 2.2793, + "step": 14212 + }, + { + "epoch": 1.5030668358714045, + "grad_norm": 8.1875, + "learning_rate": 4.0997641263311905e-06, + "loss": 2.5455, + "step": 14213 + }, + { + "epoch": 1.5031725888324874, + "grad_norm": 8.125, + "learning_rate": 4.0996325201365535e-06, + "loss": 2.5481, + "step": 14214 + }, + { + "epoch": 1.5032783417935702, + "grad_norm": 8.0625, + "learning_rate": 4.099500906435497e-06, + "loss": 2.6915, + "step": 14215 + }, + { + "epoch": 1.503384094754653, + "grad_norm": 8.75, + "learning_rate": 4.099369285228636e-06, + "loss": 2.6304, + "step": 14216 + }, + { + "epoch": 1.503489847715736, + "grad_norm": 8.25, + "learning_rate": 4.099237656516589e-06, + "loss": 2.5307, + "step": 14217 + }, + { + "epoch": 1.503595600676819, + "grad_norm": 7.875, + "learning_rate": 4.099106020299975e-06, + "loss": 2.502, + "step": 14218 + }, + { + "epoch": 1.503701353637902, + "grad_norm": 8.25, + "learning_rate": 4.098974376579409e-06, + "loss": 2.6574, + "step": 14219 + }, + { + "epoch": 1.5038071065989849, + "grad_norm": 8.0, + "learning_rate": 4.098842725355511e-06, + "loss": 2.4732, + "step": 14220 + }, + { + "epoch": 1.5039128595600677, + "grad_norm": 8.5625, + "learning_rate": 4.098711066628897e-06, + "loss": 2.7751, + "step": 14221 + }, + { + "epoch": 1.5040186125211505, + "grad_norm": 8.75, + "learning_rate": 4.0985794004001865e-06, + "loss": 2.4994, + "step": 14222 + }, + { + "epoch": 1.5041243654822334, + "grad_norm": 8.0625, + "learning_rate": 4.098447726669996e-06, + "loss": 2.5034, + "step": 14223 + }, + { + "epoch": 1.5042301184433164, + "grad_norm": 8.0, + "learning_rate": 4.098316045438945e-06, + "loss": 2.0849, + "step": 14224 + }, + { + "epoch": 1.5043358714043993, + "grad_norm": 7.96875, + "learning_rate": 4.098184356707651e-06, + "loss": 2.3197, + "step": 14225 + }, + { + "epoch": 1.5044416243654823, + "grad_norm": 8.3125, + "learning_rate": 4.09805266047673e-06, + "loss": 2.5804, + "step": 14226 + }, + { + "epoch": 1.5045473773265652, + "grad_norm": 8.375, + "learning_rate": 4.097920956746802e-06, + "loss": 2.5018, + "step": 14227 + }, + { + "epoch": 1.504653130287648, + "grad_norm": 8.0, + "learning_rate": 4.097789245518485e-06, + "loss": 2.3866, + "step": 14228 + }, + { + "epoch": 1.5047588832487309, + "grad_norm": 8.5625, + "learning_rate": 4.097657526792396e-06, + "loss": 2.7525, + "step": 14229 + }, + { + "epoch": 1.504864636209814, + "grad_norm": 8.0, + "learning_rate": 4.097525800569153e-06, + "loss": 2.501, + "step": 14230 + }, + { + "epoch": 1.5049703891708968, + "grad_norm": 8.5, + "learning_rate": 4.097394066849376e-06, + "loss": 2.5264, + "step": 14231 + }, + { + "epoch": 1.5050761421319798, + "grad_norm": 8.125, + "learning_rate": 4.097262325633682e-06, + "loss": 2.1998, + "step": 14232 + }, + { + "epoch": 1.5051818950930627, + "grad_norm": 8.1875, + "learning_rate": 4.0971305769226875e-06, + "loss": 2.4805, + "step": 14233 + }, + { + "epoch": 1.5052876480541455, + "grad_norm": 7.9375, + "learning_rate": 4.0969988207170145e-06, + "loss": 2.4363, + "step": 14234 + }, + { + "epoch": 1.5053934010152283, + "grad_norm": 8.5625, + "learning_rate": 4.096867057017278e-06, + "loss": 2.5188, + "step": 14235 + }, + { + "epoch": 1.5054991539763112, + "grad_norm": 7.625, + "learning_rate": 4.096735285824098e-06, + "loss": 2.5467, + "step": 14236 + }, + { + "epoch": 1.5056049069373942, + "grad_norm": 8.25, + "learning_rate": 4.096603507138092e-06, + "loss": 2.6126, + "step": 14237 + }, + { + "epoch": 1.5057106598984773, + "grad_norm": 9.0, + "learning_rate": 4.0964717209598785e-06, + "loss": 2.7403, + "step": 14238 + }, + { + "epoch": 1.5058164128595601, + "grad_norm": 8.5625, + "learning_rate": 4.096339927290076e-06, + "loss": 2.4733, + "step": 14239 + }, + { + "epoch": 1.505922165820643, + "grad_norm": 8.4375, + "learning_rate": 4.096208126129304e-06, + "loss": 2.4061, + "step": 14240 + }, + { + "epoch": 1.5060279187817258, + "grad_norm": 8.5625, + "learning_rate": 4.096076317478179e-06, + "loss": 2.5491, + "step": 14241 + }, + { + "epoch": 1.5061336717428087, + "grad_norm": 7.96875, + "learning_rate": 4.095944501337322e-06, + "loss": 2.6108, + "step": 14242 + }, + { + "epoch": 1.5062394247038917, + "grad_norm": 8.5625, + "learning_rate": 4.0958126777073495e-06, + "loss": 2.4725, + "step": 14243 + }, + { + "epoch": 1.5063451776649748, + "grad_norm": 9.875, + "learning_rate": 4.095680846588881e-06, + "loss": 2.7642, + "step": 14244 + }, + { + "epoch": 1.5064509306260576, + "grad_norm": 8.75, + "learning_rate": 4.095549007982534e-06, + "loss": 2.7091, + "step": 14245 + }, + { + "epoch": 1.5065566835871405, + "grad_norm": 9.25, + "learning_rate": 4.0954171618889286e-06, + "loss": 2.2283, + "step": 14246 + }, + { + "epoch": 1.5066624365482233, + "grad_norm": 8.625, + "learning_rate": 4.095285308308683e-06, + "loss": 2.3651, + "step": 14247 + }, + { + "epoch": 1.5067681895093061, + "grad_norm": 8.1875, + "learning_rate": 4.095153447242416e-06, + "loss": 2.6518, + "step": 14248 + }, + { + "epoch": 1.5068739424703892, + "grad_norm": 9.375, + "learning_rate": 4.095021578690747e-06, + "loss": 2.797, + "step": 14249 + }, + { + "epoch": 1.506979695431472, + "grad_norm": 9.0625, + "learning_rate": 4.094889702654292e-06, + "loss": 2.7664, + "step": 14250 + }, + { + "epoch": 1.507085448392555, + "grad_norm": 8.125, + "learning_rate": 4.094757819133673e-06, + "loss": 2.6531, + "step": 14251 + }, + { + "epoch": 1.507191201353638, + "grad_norm": 7.78125, + "learning_rate": 4.0946259281295086e-06, + "loss": 2.6565, + "step": 14252 + }, + { + "epoch": 1.5072969543147208, + "grad_norm": 7.96875, + "learning_rate": 4.094494029642417e-06, + "loss": 2.6681, + "step": 14253 + }, + { + "epoch": 1.5074027072758036, + "grad_norm": 8.1875, + "learning_rate": 4.094362123673016e-06, + "loss": 2.5587, + "step": 14254 + }, + { + "epoch": 1.5075084602368867, + "grad_norm": 8.25, + "learning_rate": 4.094230210221926e-06, + "loss": 2.601, + "step": 14255 + }, + { + "epoch": 1.5076142131979695, + "grad_norm": 8.25, + "learning_rate": 4.094098289289765e-06, + "loss": 2.5148, + "step": 14256 + }, + { + "epoch": 1.5077199661590526, + "grad_norm": 8.125, + "learning_rate": 4.0939663608771535e-06, + "loss": 2.3398, + "step": 14257 + }, + { + "epoch": 1.5078257191201354, + "grad_norm": 8.5625, + "learning_rate": 4.09383442498471e-06, + "loss": 2.5553, + "step": 14258 + }, + { + "epoch": 1.5079314720812182, + "grad_norm": 8.875, + "learning_rate": 4.093702481613052e-06, + "loss": 2.5577, + "step": 14259 + }, + { + "epoch": 1.508037225042301, + "grad_norm": 8.3125, + "learning_rate": 4.093570530762801e-06, + "loss": 2.3115, + "step": 14260 + }, + { + "epoch": 1.5081429780033841, + "grad_norm": 7.71875, + "learning_rate": 4.093438572434576e-06, + "loss": 2.5294, + "step": 14261 + }, + { + "epoch": 1.508248730964467, + "grad_norm": 8.375, + "learning_rate": 4.093306606628995e-06, + "loss": 2.572, + "step": 14262 + }, + { + "epoch": 1.50835448392555, + "grad_norm": 9.0, + "learning_rate": 4.093174633346677e-06, + "loss": 2.674, + "step": 14263 + }, + { + "epoch": 1.5084602368866329, + "grad_norm": 8.25, + "learning_rate": 4.093042652588243e-06, + "loss": 2.6509, + "step": 14264 + }, + { + "epoch": 1.5085659898477157, + "grad_norm": 8.125, + "learning_rate": 4.0929106643543104e-06, + "loss": 2.5102, + "step": 14265 + }, + { + "epoch": 1.5086717428087986, + "grad_norm": 9.0, + "learning_rate": 4.0927786686455e-06, + "loss": 2.6445, + "step": 14266 + }, + { + "epoch": 1.5087774957698814, + "grad_norm": 8.4375, + "learning_rate": 4.092646665462431e-06, + "loss": 2.533, + "step": 14267 + }, + { + "epoch": 1.5088832487309645, + "grad_norm": 8.0, + "learning_rate": 4.092514654805722e-06, + "loss": 2.4591, + "step": 14268 + }, + { + "epoch": 1.5089890016920475, + "grad_norm": 8.0625, + "learning_rate": 4.092382636675993e-06, + "loss": 2.5507, + "step": 14269 + }, + { + "epoch": 1.5090947546531304, + "grad_norm": 8.125, + "learning_rate": 4.092250611073863e-06, + "loss": 2.6449, + "step": 14270 + }, + { + "epoch": 1.5092005076142132, + "grad_norm": 8.5625, + "learning_rate": 4.092118577999954e-06, + "loss": 2.7352, + "step": 14271 + }, + { + "epoch": 1.509306260575296, + "grad_norm": 8.4375, + "learning_rate": 4.091986537454883e-06, + "loss": 2.2311, + "step": 14272 + }, + { + "epoch": 1.5094120135363789, + "grad_norm": 7.875, + "learning_rate": 4.09185448943927e-06, + "loss": 2.5282, + "step": 14273 + }, + { + "epoch": 1.509517766497462, + "grad_norm": 8.0, + "learning_rate": 4.091722433953734e-06, + "loss": 2.7569, + "step": 14274 + }, + { + "epoch": 1.509623519458545, + "grad_norm": 8.3125, + "learning_rate": 4.091590370998897e-06, + "loss": 2.5944, + "step": 14275 + }, + { + "epoch": 1.5097292724196278, + "grad_norm": 8.125, + "learning_rate": 4.091458300575376e-06, + "loss": 2.6531, + "step": 14276 + }, + { + "epoch": 1.5098350253807107, + "grad_norm": 7.90625, + "learning_rate": 4.091326222683793e-06, + "loss": 2.4103, + "step": 14277 + }, + { + "epoch": 1.5099407783417935, + "grad_norm": 8.0, + "learning_rate": 4.091194137324767e-06, + "loss": 2.4717, + "step": 14278 + }, + { + "epoch": 1.5100465313028764, + "grad_norm": 8.75, + "learning_rate": 4.091062044498918e-06, + "loss": 2.6027, + "step": 14279 + }, + { + "epoch": 1.5101522842639594, + "grad_norm": 7.8125, + "learning_rate": 4.090929944206864e-06, + "loss": 2.3943, + "step": 14280 + }, + { + "epoch": 1.5102580372250423, + "grad_norm": 8.375, + "learning_rate": 4.090797836449228e-06, + "loss": 2.5409, + "step": 14281 + }, + { + "epoch": 1.5103637901861253, + "grad_norm": 8.125, + "learning_rate": 4.090665721226628e-06, + "loss": 2.3804, + "step": 14282 + }, + { + "epoch": 1.5104695431472082, + "grad_norm": 8.5, + "learning_rate": 4.090533598539685e-06, + "loss": 2.3707, + "step": 14283 + }, + { + "epoch": 1.510575296108291, + "grad_norm": 8.4375, + "learning_rate": 4.090401468389016e-06, + "loss": 2.5717, + "step": 14284 + }, + { + "epoch": 1.5106810490693738, + "grad_norm": 9.0625, + "learning_rate": 4.090269330775245e-06, + "loss": 2.7058, + "step": 14285 + }, + { + "epoch": 1.510786802030457, + "grad_norm": 8.75, + "learning_rate": 4.09013718569899e-06, + "loss": 2.5264, + "step": 14286 + }, + { + "epoch": 1.5108925549915397, + "grad_norm": 8.1875, + "learning_rate": 4.090005033160872e-06, + "loss": 2.593, + "step": 14287 + }, + { + "epoch": 1.5109983079526228, + "grad_norm": 8.125, + "learning_rate": 4.08987287316151e-06, + "loss": 2.4832, + "step": 14288 + }, + { + "epoch": 1.5111040609137056, + "grad_norm": 8.5, + "learning_rate": 4.089740705701526e-06, + "loss": 2.5133, + "step": 14289 + }, + { + "epoch": 1.5112098138747885, + "grad_norm": 8.375, + "learning_rate": 4.089608530781538e-06, + "loss": 2.6195, + "step": 14290 + }, + { + "epoch": 1.5113155668358713, + "grad_norm": 8.5, + "learning_rate": 4.089476348402168e-06, + "loss": 2.7579, + "step": 14291 + }, + { + "epoch": 1.5114213197969542, + "grad_norm": 7.71875, + "learning_rate": 4.0893441585640344e-06, + "loss": 2.4905, + "step": 14292 + }, + { + "epoch": 1.5115270727580372, + "grad_norm": 8.125, + "learning_rate": 4.089211961267759e-06, + "loss": 2.4347, + "step": 14293 + }, + { + "epoch": 1.5116328257191203, + "grad_norm": 8.0625, + "learning_rate": 4.089079756513963e-06, + "loss": 2.4695, + "step": 14294 + }, + { + "epoch": 1.5117385786802031, + "grad_norm": 7.71875, + "learning_rate": 4.088947544303265e-06, + "loss": 2.5242, + "step": 14295 + }, + { + "epoch": 1.511844331641286, + "grad_norm": 8.0625, + "learning_rate": 4.088815324636286e-06, + "loss": 2.5689, + "step": 14296 + }, + { + "epoch": 1.5119500846023688, + "grad_norm": 8.625, + "learning_rate": 4.088683097513646e-06, + "loss": 2.4068, + "step": 14297 + }, + { + "epoch": 1.5120558375634516, + "grad_norm": 8.1875, + "learning_rate": 4.088550862935967e-06, + "loss": 2.6095, + "step": 14298 + }, + { + "epoch": 1.5121615905245347, + "grad_norm": 8.0, + "learning_rate": 4.088418620903868e-06, + "loss": 2.5182, + "step": 14299 + }, + { + "epoch": 1.5122673434856178, + "grad_norm": 8.125, + "learning_rate": 4.08828637141797e-06, + "loss": 2.4602, + "step": 14300 + }, + { + "epoch": 1.5123730964467006, + "grad_norm": 8.75, + "learning_rate": 4.088154114478893e-06, + "loss": 2.3346, + "step": 14301 + }, + { + "epoch": 1.5124788494077834, + "grad_norm": 7.84375, + "learning_rate": 4.088021850087259e-06, + "loss": 2.4594, + "step": 14302 + }, + { + "epoch": 1.5125846023688663, + "grad_norm": 8.375, + "learning_rate": 4.0878895782436875e-06, + "loss": 2.6857, + "step": 14303 + }, + { + "epoch": 1.512690355329949, + "grad_norm": 8.6875, + "learning_rate": 4.0877572989488005e-06, + "loss": 2.5822, + "step": 14304 + }, + { + "epoch": 1.5127961082910322, + "grad_norm": 8.4375, + "learning_rate": 4.087625012203217e-06, + "loss": 2.7002, + "step": 14305 + }, + { + "epoch": 1.512901861252115, + "grad_norm": 8.25, + "learning_rate": 4.087492718007559e-06, + "loss": 2.6919, + "step": 14306 + }, + { + "epoch": 1.513007614213198, + "grad_norm": 8.1875, + "learning_rate": 4.087360416362447e-06, + "loss": 2.4506, + "step": 14307 + }, + { + "epoch": 1.513113367174281, + "grad_norm": 12.5625, + "learning_rate": 4.087228107268502e-06, + "loss": 2.5019, + "step": 14308 + }, + { + "epoch": 1.5132191201353637, + "grad_norm": 8.75, + "learning_rate": 4.087095790726344e-06, + "loss": 2.5954, + "step": 14309 + }, + { + "epoch": 1.5133248730964466, + "grad_norm": 8.1875, + "learning_rate": 4.0869634667365945e-06, + "loss": 2.5523, + "step": 14310 + }, + { + "epoch": 1.5134306260575296, + "grad_norm": 8.125, + "learning_rate": 4.086831135299875e-06, + "loss": 2.5339, + "step": 14311 + }, + { + "epoch": 1.5135363790186125, + "grad_norm": 8.3125, + "learning_rate": 4.086698796416806e-06, + "loss": 2.3403, + "step": 14312 + }, + { + "epoch": 1.5136421319796955, + "grad_norm": 8.125, + "learning_rate": 4.086566450088008e-06, + "loss": 2.4278, + "step": 14313 + }, + { + "epoch": 1.5137478849407784, + "grad_norm": 8.3125, + "learning_rate": 4.086434096314102e-06, + "loss": 2.5431, + "step": 14314 + }, + { + "epoch": 1.5138536379018612, + "grad_norm": 8.0625, + "learning_rate": 4.0863017350957115e-06, + "loss": 2.4344, + "step": 14315 + }, + { + "epoch": 1.513959390862944, + "grad_norm": 8.3125, + "learning_rate": 4.086169366433455e-06, + "loss": 2.7025, + "step": 14316 + }, + { + "epoch": 1.5140651438240271, + "grad_norm": 8.0625, + "learning_rate": 4.086036990327954e-06, + "loss": 2.4517, + "step": 14317 + }, + { + "epoch": 1.51417089678511, + "grad_norm": 8.125, + "learning_rate": 4.0859046067798295e-06, + "loss": 2.6407, + "step": 14318 + }, + { + "epoch": 1.514276649746193, + "grad_norm": 8.5625, + "learning_rate": 4.085772215789704e-06, + "loss": 2.5221, + "step": 14319 + }, + { + "epoch": 1.5143824027072759, + "grad_norm": 8.125, + "learning_rate": 4.085639817358198e-06, + "loss": 2.3074, + "step": 14320 + }, + { + "epoch": 1.5144881556683587, + "grad_norm": 7.90625, + "learning_rate": 4.085507411485933e-06, + "loss": 2.6091, + "step": 14321 + }, + { + "epoch": 1.5145939086294415, + "grad_norm": 7.90625, + "learning_rate": 4.08537499817353e-06, + "loss": 2.4848, + "step": 14322 + }, + { + "epoch": 1.5146996615905244, + "grad_norm": 8.1875, + "learning_rate": 4.085242577421611e-06, + "loss": 2.4209, + "step": 14323 + }, + { + "epoch": 1.5148054145516074, + "grad_norm": 8.0625, + "learning_rate": 4.085110149230796e-06, + "loss": 2.7011, + "step": 14324 + }, + { + "epoch": 1.5149111675126905, + "grad_norm": 7.9375, + "learning_rate": 4.0849777136017074e-06, + "loss": 2.5371, + "step": 14325 + }, + { + "epoch": 1.5150169204737733, + "grad_norm": 8.4375, + "learning_rate": 4.084845270534967e-06, + "loss": 2.5979, + "step": 14326 + }, + { + "epoch": 1.5151226734348562, + "grad_norm": 8.25, + "learning_rate": 4.084712820031197e-06, + "loss": 2.6606, + "step": 14327 + }, + { + "epoch": 1.515228426395939, + "grad_norm": 8.4375, + "learning_rate": 4.084580362091017e-06, + "loss": 2.7838, + "step": 14328 + }, + { + "epoch": 1.5153341793570219, + "grad_norm": 8.625, + "learning_rate": 4.084447896715048e-06, + "loss": 2.7556, + "step": 14329 + }, + { + "epoch": 1.515439932318105, + "grad_norm": 8.375, + "learning_rate": 4.084315423903914e-06, + "loss": 2.7848, + "step": 14330 + }, + { + "epoch": 1.515545685279188, + "grad_norm": 8.5, + "learning_rate": 4.084182943658236e-06, + "loss": 2.6549, + "step": 14331 + }, + { + "epoch": 1.5156514382402708, + "grad_norm": 8.5, + "learning_rate": 4.0840504559786356e-06, + "loss": 2.4949, + "step": 14332 + }, + { + "epoch": 1.5157571912013537, + "grad_norm": 8.125, + "learning_rate": 4.0839179608657345e-06, + "loss": 2.5011, + "step": 14333 + }, + { + "epoch": 1.5158629441624365, + "grad_norm": 7.90625, + "learning_rate": 4.083785458320153e-06, + "loss": 2.5171, + "step": 14334 + }, + { + "epoch": 1.5159686971235193, + "grad_norm": 8.3125, + "learning_rate": 4.083652948342515e-06, + "loss": 2.4226, + "step": 14335 + }, + { + "epoch": 1.5160744500846024, + "grad_norm": 7.96875, + "learning_rate": 4.083520430933442e-06, + "loss": 2.4668, + "step": 14336 + }, + { + "epoch": 1.5161802030456852, + "grad_norm": 8.25, + "learning_rate": 4.083387906093554e-06, + "loss": 2.5589, + "step": 14337 + }, + { + "epoch": 1.5162859560067683, + "grad_norm": 8.25, + "learning_rate": 4.0832553738234746e-06, + "loss": 2.5371, + "step": 14338 + }, + { + "epoch": 1.5163917089678511, + "grad_norm": 8.25, + "learning_rate": 4.083122834123825e-06, + "loss": 2.5594, + "step": 14339 + }, + { + "epoch": 1.516497461928934, + "grad_norm": 8.1875, + "learning_rate": 4.0829902869952275e-06, + "loss": 2.6497, + "step": 14340 + }, + { + "epoch": 1.5166032148900168, + "grad_norm": 7.90625, + "learning_rate": 4.082857732438305e-06, + "loss": 2.5018, + "step": 14341 + }, + { + "epoch": 1.5167089678510999, + "grad_norm": 7.8125, + "learning_rate": 4.082725170453677e-06, + "loss": 2.4955, + "step": 14342 + }, + { + "epoch": 1.5168147208121827, + "grad_norm": 9.0, + "learning_rate": 4.082592601041968e-06, + "loss": 2.6203, + "step": 14343 + }, + { + "epoch": 1.5169204737732658, + "grad_norm": 8.75, + "learning_rate": 4.0824600242038e-06, + "loss": 2.5544, + "step": 14344 + }, + { + "epoch": 1.5170262267343486, + "grad_norm": 8.0, + "learning_rate": 4.082327439939792e-06, + "loss": 2.5117, + "step": 14345 + }, + { + "epoch": 1.5171319796954315, + "grad_norm": 8.6875, + "learning_rate": 4.08219484825057e-06, + "loss": 2.5786, + "step": 14346 + }, + { + "epoch": 1.5172377326565143, + "grad_norm": 8.4375, + "learning_rate": 4.082062249136754e-06, + "loss": 2.479, + "step": 14347 + }, + { + "epoch": 1.5173434856175971, + "grad_norm": 8.375, + "learning_rate": 4.081929642598967e-06, + "loss": 2.7267, + "step": 14348 + }, + { + "epoch": 1.5174492385786802, + "grad_norm": 7.875, + "learning_rate": 4.081797028637831e-06, + "loss": 2.561, + "step": 14349 + }, + { + "epoch": 1.5175549915397633, + "grad_norm": 8.6875, + "learning_rate": 4.0816644072539705e-06, + "loss": 2.6608, + "step": 14350 + }, + { + "epoch": 1.517660744500846, + "grad_norm": 8.25, + "learning_rate": 4.081531778448004e-06, + "loss": 2.3381, + "step": 14351 + }, + { + "epoch": 1.517766497461929, + "grad_norm": 8.0, + "learning_rate": 4.081399142220556e-06, + "loss": 2.6033, + "step": 14352 + }, + { + "epoch": 1.5178722504230118, + "grad_norm": 7.84375, + "learning_rate": 4.081266498572248e-06, + "loss": 2.498, + "step": 14353 + }, + { + "epoch": 1.5179780033840946, + "grad_norm": 8.0, + "learning_rate": 4.0811338475037045e-06, + "loss": 2.5246, + "step": 14354 + }, + { + "epoch": 1.5180837563451777, + "grad_norm": 8.5, + "learning_rate": 4.081001189015545e-06, + "loss": 2.6218, + "step": 14355 + }, + { + "epoch": 1.5181895093062607, + "grad_norm": 8.625, + "learning_rate": 4.0808685231083936e-06, + "loss": 2.705, + "step": 14356 + }, + { + "epoch": 1.5182952622673436, + "grad_norm": 8.1875, + "learning_rate": 4.080735849782874e-06, + "loss": 2.1696, + "step": 14357 + }, + { + "epoch": 1.5184010152284264, + "grad_norm": 8.3125, + "learning_rate": 4.080603169039607e-06, + "loss": 2.4841, + "step": 14358 + }, + { + "epoch": 1.5185067681895092, + "grad_norm": 8.5, + "learning_rate": 4.080470480879215e-06, + "loss": 2.5651, + "step": 14359 + }, + { + "epoch": 1.518612521150592, + "grad_norm": 8.125, + "learning_rate": 4.080337785302323e-06, + "loss": 2.5458, + "step": 14360 + }, + { + "epoch": 1.5187182741116751, + "grad_norm": 8.75, + "learning_rate": 4.080205082309551e-06, + "loss": 2.3058, + "step": 14361 + }, + { + "epoch": 1.518824027072758, + "grad_norm": 7.9375, + "learning_rate": 4.0800723719015225e-06, + "loss": 2.6766, + "step": 14362 + }, + { + "epoch": 1.518929780033841, + "grad_norm": 9.125, + "learning_rate": 4.079939654078862e-06, + "loss": 2.5437, + "step": 14363 + }, + { + "epoch": 1.5190355329949239, + "grad_norm": 8.4375, + "learning_rate": 4.07980692884219e-06, + "loss": 2.3917, + "step": 14364 + }, + { + "epoch": 1.5191412859560067, + "grad_norm": 7.9375, + "learning_rate": 4.079674196192129e-06, + "loss": 2.5183, + "step": 14365 + }, + { + "epoch": 1.5192470389170896, + "grad_norm": 8.0625, + "learning_rate": 4.079541456129305e-06, + "loss": 2.5247, + "step": 14366 + }, + { + "epoch": 1.5193527918781726, + "grad_norm": 8.6875, + "learning_rate": 4.079408708654337e-06, + "loss": 2.6719, + "step": 14367 + }, + { + "epoch": 1.5194585448392555, + "grad_norm": 8.125, + "learning_rate": 4.079275953767852e-06, + "loss": 2.4726, + "step": 14368 + }, + { + "epoch": 1.5195642978003385, + "grad_norm": 8.3125, + "learning_rate": 4.07914319147047e-06, + "loss": 2.754, + "step": 14369 + }, + { + "epoch": 1.5196700507614214, + "grad_norm": 8.0, + "learning_rate": 4.079010421762815e-06, + "loss": 2.4421, + "step": 14370 + }, + { + "epoch": 1.5197758037225042, + "grad_norm": 8.125, + "learning_rate": 4.078877644645509e-06, + "loss": 2.3309, + "step": 14371 + }, + { + "epoch": 1.519881556683587, + "grad_norm": 8.5, + "learning_rate": 4.078744860119177e-06, + "loss": 2.5087, + "step": 14372 + }, + { + "epoch": 1.51998730964467, + "grad_norm": 8.1875, + "learning_rate": 4.07861206818444e-06, + "loss": 2.55, + "step": 14373 + }, + { + "epoch": 1.520093062605753, + "grad_norm": 7.90625, + "learning_rate": 4.078479268841923e-06, + "loss": 2.5781, + "step": 14374 + }, + { + "epoch": 1.520198815566836, + "grad_norm": 8.0, + "learning_rate": 4.078346462092248e-06, + "loss": 2.5727, + "step": 14375 + }, + { + "epoch": 1.5203045685279188, + "grad_norm": 7.84375, + "learning_rate": 4.078213647936039e-06, + "loss": 2.521, + "step": 14376 + }, + { + "epoch": 1.5204103214890017, + "grad_norm": 8.5, + "learning_rate": 4.078080826373919e-06, + "loss": 2.5723, + "step": 14377 + }, + { + "epoch": 1.5205160744500845, + "grad_norm": 8.0, + "learning_rate": 4.0779479974065105e-06, + "loss": 2.5975, + "step": 14378 + }, + { + "epoch": 1.5206218274111674, + "grad_norm": 8.875, + "learning_rate": 4.077815161034438e-06, + "loss": 2.5053, + "step": 14379 + }, + { + "epoch": 1.5207275803722504, + "grad_norm": 9.0, + "learning_rate": 4.077682317258324e-06, + "loss": 2.6853, + "step": 14380 + }, + { + "epoch": 1.5208333333333335, + "grad_norm": 8.875, + "learning_rate": 4.077549466078792e-06, + "loss": 2.6181, + "step": 14381 + }, + { + "epoch": 1.5209390862944163, + "grad_norm": 8.375, + "learning_rate": 4.077416607496465e-06, + "loss": 2.5302, + "step": 14382 + }, + { + "epoch": 1.5210448392554992, + "grad_norm": 8.25, + "learning_rate": 4.077283741511968e-06, + "loss": 2.5419, + "step": 14383 + }, + { + "epoch": 1.521150592216582, + "grad_norm": 7.625, + "learning_rate": 4.077150868125923e-06, + "loss": 2.439, + "step": 14384 + }, + { + "epoch": 1.5212563451776648, + "grad_norm": 8.5625, + "learning_rate": 4.077017987338954e-06, + "loss": 2.3988, + "step": 14385 + }, + { + "epoch": 1.521362098138748, + "grad_norm": 8.3125, + "learning_rate": 4.076885099151685e-06, + "loss": 2.6363, + "step": 14386 + }, + { + "epoch": 1.521467851099831, + "grad_norm": 8.4375, + "learning_rate": 4.076752203564739e-06, + "loss": 2.4374, + "step": 14387 + }, + { + "epoch": 1.5215736040609138, + "grad_norm": 8.1875, + "learning_rate": 4.076619300578739e-06, + "loss": 2.5518, + "step": 14388 + }, + { + "epoch": 1.5216793570219966, + "grad_norm": 8.125, + "learning_rate": 4.07648639019431e-06, + "loss": 2.3507, + "step": 14389 + }, + { + "epoch": 1.5217851099830795, + "grad_norm": 9.0625, + "learning_rate": 4.076353472412076e-06, + "loss": 2.7726, + "step": 14390 + }, + { + "epoch": 1.5218908629441623, + "grad_norm": 8.5625, + "learning_rate": 4.076220547232658e-06, + "loss": 2.6516, + "step": 14391 + }, + { + "epoch": 1.5219966159052454, + "grad_norm": 8.3125, + "learning_rate": 4.076087614656683e-06, + "loss": 2.6231, + "step": 14392 + }, + { + "epoch": 1.5221023688663282, + "grad_norm": 8.375, + "learning_rate": 4.075954674684774e-06, + "loss": 2.5071, + "step": 14393 + }, + { + "epoch": 1.5222081218274113, + "grad_norm": 8.375, + "learning_rate": 4.075821727317552e-06, + "loss": 2.4803, + "step": 14394 + }, + { + "epoch": 1.5223138747884941, + "grad_norm": 7.9375, + "learning_rate": 4.075688772555645e-06, + "loss": 2.6418, + "step": 14395 + }, + { + "epoch": 1.522419627749577, + "grad_norm": 8.25, + "learning_rate": 4.075555810399674e-06, + "loss": 2.7008, + "step": 14396 + }, + { + "epoch": 1.5225253807106598, + "grad_norm": 8.5, + "learning_rate": 4.075422840850265e-06, + "loss": 2.5906, + "step": 14397 + }, + { + "epoch": 1.5226311336717429, + "grad_norm": 9.9375, + "learning_rate": 4.0752898639080404e-06, + "loss": 2.4284, + "step": 14398 + }, + { + "epoch": 1.5227368866328257, + "grad_norm": 8.4375, + "learning_rate": 4.075156879573624e-06, + "loss": 2.4115, + "step": 14399 + }, + { + "epoch": 1.5228426395939088, + "grad_norm": 8.0625, + "learning_rate": 4.075023887847641e-06, + "loss": 2.1748, + "step": 14400 + }, + { + "epoch": 1.5229483925549916, + "grad_norm": 8.3125, + "learning_rate": 4.0748908887307156e-06, + "loss": 2.6571, + "step": 14401 + }, + { + "epoch": 1.5230541455160744, + "grad_norm": 8.25, + "learning_rate": 4.074757882223471e-06, + "loss": 2.4705, + "step": 14402 + }, + { + "epoch": 1.5231598984771573, + "grad_norm": 8.0625, + "learning_rate": 4.074624868326532e-06, + "loss": 2.5101, + "step": 14403 + }, + { + "epoch": 1.52326565143824, + "grad_norm": 8.25, + "learning_rate": 4.074491847040522e-06, + "loss": 2.6257, + "step": 14404 + }, + { + "epoch": 1.5233714043993232, + "grad_norm": 9.4375, + "learning_rate": 4.074358818366066e-06, + "loss": 2.7704, + "step": 14405 + }, + { + "epoch": 1.5234771573604062, + "grad_norm": 8.8125, + "learning_rate": 4.074225782303789e-06, + "loss": 2.6347, + "step": 14406 + }, + { + "epoch": 1.523582910321489, + "grad_norm": 8.125, + "learning_rate": 4.0740927388543126e-06, + "loss": 2.579, + "step": 14407 + }, + { + "epoch": 1.523688663282572, + "grad_norm": 8.6875, + "learning_rate": 4.073959688018263e-06, + "loss": 2.6537, + "step": 14408 + }, + { + "epoch": 1.5237944162436547, + "grad_norm": 8.25, + "learning_rate": 4.073826629796265e-06, + "loss": 2.6375, + "step": 14409 + }, + { + "epoch": 1.5239001692047376, + "grad_norm": 8.0625, + "learning_rate": 4.073693564188942e-06, + "loss": 2.5064, + "step": 14410 + }, + { + "epoch": 1.5240059221658206, + "grad_norm": 8.625, + "learning_rate": 4.073560491196918e-06, + "loss": 2.6368, + "step": 14411 + }, + { + "epoch": 1.5241116751269037, + "grad_norm": 8.375, + "learning_rate": 4.07342741082082e-06, + "loss": 2.5727, + "step": 14412 + }, + { + "epoch": 1.5242174280879865, + "grad_norm": 8.25, + "learning_rate": 4.073294323061269e-06, + "loss": 2.616, + "step": 14413 + }, + { + "epoch": 1.5243231810490694, + "grad_norm": 8.25, + "learning_rate": 4.073161227918892e-06, + "loss": 2.5923, + "step": 14414 + }, + { + "epoch": 1.5244289340101522, + "grad_norm": 8.5625, + "learning_rate": 4.0730281253943125e-06, + "loss": 2.6516, + "step": 14415 + }, + { + "epoch": 1.524534686971235, + "grad_norm": 8.125, + "learning_rate": 4.072895015488155e-06, + "loss": 2.5302, + "step": 14416 + }, + { + "epoch": 1.5246404399323181, + "grad_norm": 8.25, + "learning_rate": 4.072761898201046e-06, + "loss": 2.5347, + "step": 14417 + }, + { + "epoch": 1.524746192893401, + "grad_norm": 8.5, + "learning_rate": 4.072628773533607e-06, + "loss": 2.3473, + "step": 14418 + }, + { + "epoch": 1.524851945854484, + "grad_norm": 8.1875, + "learning_rate": 4.072495641486465e-06, + "loss": 2.5887, + "step": 14419 + }, + { + "epoch": 1.5249576988155669, + "grad_norm": 8.3125, + "learning_rate": 4.072362502060243e-06, + "loss": 2.705, + "step": 14420 + }, + { + "epoch": 1.5250634517766497, + "grad_norm": 8.1875, + "learning_rate": 4.072229355255568e-06, + "loss": 2.4404, + "step": 14421 + }, + { + "epoch": 1.5251692047377325, + "grad_norm": 7.96875, + "learning_rate": 4.072096201073064e-06, + "loss": 2.5525, + "step": 14422 + }, + { + "epoch": 1.5252749576988156, + "grad_norm": 8.375, + "learning_rate": 4.071963039513355e-06, + "loss": 2.5776, + "step": 14423 + }, + { + "epoch": 1.5253807106598984, + "grad_norm": 8.0, + "learning_rate": 4.071829870577066e-06, + "loss": 2.4999, + "step": 14424 + }, + { + "epoch": 1.5254864636209815, + "grad_norm": 8.75, + "learning_rate": 4.071696694264823e-06, + "loss": 2.7581, + "step": 14425 + }, + { + "epoch": 1.5255922165820643, + "grad_norm": 8.25, + "learning_rate": 4.071563510577249e-06, + "loss": 2.5624, + "step": 14426 + }, + { + "epoch": 1.5256979695431472, + "grad_norm": 8.1875, + "learning_rate": 4.071430319514971e-06, + "loss": 2.4369, + "step": 14427 + }, + { + "epoch": 1.52580372250423, + "grad_norm": 8.875, + "learning_rate": 4.0712971210786135e-06, + "loss": 2.4202, + "step": 14428 + }, + { + "epoch": 1.525909475465313, + "grad_norm": 8.1875, + "learning_rate": 4.0711639152688006e-06, + "loss": 2.3788, + "step": 14429 + }, + { + "epoch": 1.526015228426396, + "grad_norm": 7.65625, + "learning_rate": 4.071030702086159e-06, + "loss": 2.4774, + "step": 14430 + }, + { + "epoch": 1.526120981387479, + "grad_norm": 7.78125, + "learning_rate": 4.0708974815313115e-06, + "loss": 2.3802, + "step": 14431 + }, + { + "epoch": 1.5262267343485618, + "grad_norm": 8.75, + "learning_rate": 4.070764253604886e-06, + "loss": 2.8282, + "step": 14432 + }, + { + "epoch": 1.5263324873096447, + "grad_norm": 9.1875, + "learning_rate": 4.070631018307505e-06, + "loss": 2.7073, + "step": 14433 + }, + { + "epoch": 1.5264382402707275, + "grad_norm": 8.4375, + "learning_rate": 4.070497775639796e-06, + "loss": 2.7382, + "step": 14434 + }, + { + "epoch": 1.5265439932318103, + "grad_norm": 7.65625, + "learning_rate": 4.070364525602382e-06, + "loss": 2.5169, + "step": 14435 + }, + { + "epoch": 1.5266497461928934, + "grad_norm": 8.25, + "learning_rate": 4.070231268195891e-06, + "loss": 2.5617, + "step": 14436 + }, + { + "epoch": 1.5267554991539765, + "grad_norm": 8.375, + "learning_rate": 4.070098003420946e-06, + "loss": 2.7704, + "step": 14437 + }, + { + "epoch": 1.5268612521150593, + "grad_norm": 8.625, + "learning_rate": 4.069964731278173e-06, + "loss": 2.5882, + "step": 14438 + }, + { + "epoch": 1.5269670050761421, + "grad_norm": 8.4375, + "learning_rate": 4.069831451768198e-06, + "loss": 2.5589, + "step": 14439 + }, + { + "epoch": 1.527072758037225, + "grad_norm": 8.0625, + "learning_rate": 4.069698164891647e-06, + "loss": 2.4199, + "step": 14440 + }, + { + "epoch": 1.5271785109983078, + "grad_norm": 7.75, + "learning_rate": 4.069564870649143e-06, + "loss": 2.4946, + "step": 14441 + }, + { + "epoch": 1.5272842639593909, + "grad_norm": 9.25, + "learning_rate": 4.069431569041313e-06, + "loss": 2.7386, + "step": 14442 + }, + { + "epoch": 1.527390016920474, + "grad_norm": 9.0625, + "learning_rate": 4.0692982600687835e-06, + "loss": 2.7588, + "step": 14443 + }, + { + "epoch": 1.5274957698815568, + "grad_norm": 8.375, + "learning_rate": 4.069164943732179e-06, + "loss": 2.4857, + "step": 14444 + }, + { + "epoch": 1.5276015228426396, + "grad_norm": 7.90625, + "learning_rate": 4.069031620032125e-06, + "loss": 2.5352, + "step": 14445 + }, + { + "epoch": 1.5277072758037225, + "grad_norm": 7.96875, + "learning_rate": 4.068898288969247e-06, + "loss": 2.5723, + "step": 14446 + }, + { + "epoch": 1.5278130287648053, + "grad_norm": 8.3125, + "learning_rate": 4.06876495054417e-06, + "loss": 2.6126, + "step": 14447 + }, + { + "epoch": 1.5279187817258884, + "grad_norm": 8.625, + "learning_rate": 4.068631604757523e-06, + "loss": 2.5828, + "step": 14448 + }, + { + "epoch": 1.5280245346869712, + "grad_norm": 8.1875, + "learning_rate": 4.068498251609928e-06, + "loss": 2.5921, + "step": 14449 + }, + { + "epoch": 1.5281302876480543, + "grad_norm": 8.4375, + "learning_rate": 4.068364891102012e-06, + "loss": 2.4348, + "step": 14450 + }, + { + "epoch": 1.528236040609137, + "grad_norm": 7.78125, + "learning_rate": 4.068231523234402e-06, + "loss": 2.3806, + "step": 14451 + }, + { + "epoch": 1.52834179357022, + "grad_norm": 8.3125, + "learning_rate": 4.068098148007723e-06, + "loss": 2.7281, + "step": 14452 + }, + { + "epoch": 1.5284475465313028, + "grad_norm": 8.6875, + "learning_rate": 4.0679647654226e-06, + "loss": 2.5091, + "step": 14453 + }, + { + "epoch": 1.5285532994923858, + "grad_norm": 8.4375, + "learning_rate": 4.06783137547966e-06, + "loss": 2.6766, + "step": 14454 + }, + { + "epoch": 1.5286590524534687, + "grad_norm": 8.3125, + "learning_rate": 4.067697978179527e-06, + "loss": 2.48, + "step": 14455 + }, + { + "epoch": 1.5287648054145517, + "grad_norm": 8.375, + "learning_rate": 4.06756457352283e-06, + "loss": 2.3949, + "step": 14456 + }, + { + "epoch": 1.5288705583756346, + "grad_norm": 7.96875, + "learning_rate": 4.0674311615101934e-06, + "loss": 2.6649, + "step": 14457 + }, + { + "epoch": 1.5289763113367174, + "grad_norm": 8.1875, + "learning_rate": 4.067297742142244e-06, + "loss": 2.5747, + "step": 14458 + }, + { + "epoch": 1.5290820642978002, + "grad_norm": 8.0625, + "learning_rate": 4.067164315419606e-06, + "loss": 2.5681, + "step": 14459 + }, + { + "epoch": 1.529187817258883, + "grad_norm": 7.9375, + "learning_rate": 4.0670308813429085e-06, + "loss": 2.2732, + "step": 14460 + }, + { + "epoch": 1.5292935702199661, + "grad_norm": 8.375, + "learning_rate": 4.066897439912775e-06, + "loss": 2.5352, + "step": 14461 + }, + { + "epoch": 1.5293993231810492, + "grad_norm": 7.9375, + "learning_rate": 4.066763991129832e-06, + "loss": 2.5164, + "step": 14462 + }, + { + "epoch": 1.529505076142132, + "grad_norm": 8.0625, + "learning_rate": 4.0666305349947065e-06, + "loss": 2.605, + "step": 14463 + }, + { + "epoch": 1.5296108291032149, + "grad_norm": 7.84375, + "learning_rate": 4.066497071508026e-06, + "loss": 2.4467, + "step": 14464 + }, + { + "epoch": 1.5297165820642977, + "grad_norm": 8.125, + "learning_rate": 4.0663636006704145e-06, + "loss": 2.7238, + "step": 14465 + }, + { + "epoch": 1.5298223350253806, + "grad_norm": 8.125, + "learning_rate": 4.066230122482499e-06, + "loss": 2.7288, + "step": 14466 + }, + { + "epoch": 1.5299280879864636, + "grad_norm": 8.0625, + "learning_rate": 4.066096636944906e-06, + "loss": 2.658, + "step": 14467 + }, + { + "epoch": 1.5300338409475467, + "grad_norm": 7.84375, + "learning_rate": 4.065963144058263e-06, + "loss": 2.5507, + "step": 14468 + }, + { + "epoch": 1.5301395939086295, + "grad_norm": 7.71875, + "learning_rate": 4.065829643823195e-06, + "loss": 2.4714, + "step": 14469 + }, + { + "epoch": 1.5302453468697124, + "grad_norm": 7.6875, + "learning_rate": 4.065696136240328e-06, + "loss": 2.5557, + "step": 14470 + }, + { + "epoch": 1.5303510998307952, + "grad_norm": 8.625, + "learning_rate": 4.0655626213102905e-06, + "loss": 2.353, + "step": 14471 + }, + { + "epoch": 1.530456852791878, + "grad_norm": 7.875, + "learning_rate": 4.0654290990337075e-06, + "loss": 2.4101, + "step": 14472 + }, + { + "epoch": 1.530562605752961, + "grad_norm": 7.90625, + "learning_rate": 4.065295569411206e-06, + "loss": 2.5222, + "step": 14473 + }, + { + "epoch": 1.530668358714044, + "grad_norm": 8.25, + "learning_rate": 4.065162032443412e-06, + "loss": 2.6187, + "step": 14474 + }, + { + "epoch": 1.530774111675127, + "grad_norm": 8.1875, + "learning_rate": 4.065028488130954e-06, + "loss": 2.4384, + "step": 14475 + }, + { + "epoch": 1.5308798646362098, + "grad_norm": 8.1875, + "learning_rate": 4.064894936474456e-06, + "loss": 2.5689, + "step": 14476 + }, + { + "epoch": 1.5309856175972927, + "grad_norm": 9.125, + "learning_rate": 4.064761377474547e-06, + "loss": 2.6004, + "step": 14477 + }, + { + "epoch": 1.5310913705583755, + "grad_norm": 8.3125, + "learning_rate": 4.064627811131853e-06, + "loss": 2.6169, + "step": 14478 + }, + { + "epoch": 1.5311971235194586, + "grad_norm": 8.125, + "learning_rate": 4.0644942374469996e-06, + "loss": 2.5217, + "step": 14479 + }, + { + "epoch": 1.5313028764805414, + "grad_norm": 8.75, + "learning_rate": 4.064360656420615e-06, + "loss": 2.5311, + "step": 14480 + }, + { + "epoch": 1.5314086294416245, + "grad_norm": 8.5, + "learning_rate": 4.064227068053326e-06, + "loss": 2.5894, + "step": 14481 + }, + { + "epoch": 1.5315143824027073, + "grad_norm": 8.1875, + "learning_rate": 4.0640934723457595e-06, + "loss": 2.604, + "step": 14482 + }, + { + "epoch": 1.5316201353637902, + "grad_norm": 7.8125, + "learning_rate": 4.063959869298541e-06, + "loss": 2.5409, + "step": 14483 + }, + { + "epoch": 1.531725888324873, + "grad_norm": 8.75, + "learning_rate": 4.063826258912299e-06, + "loss": 2.6801, + "step": 14484 + }, + { + "epoch": 1.531831641285956, + "grad_norm": 8.75, + "learning_rate": 4.06369264118766e-06, + "loss": 2.5884, + "step": 14485 + }, + { + "epoch": 1.531937394247039, + "grad_norm": 7.84375, + "learning_rate": 4.063559016125251e-06, + "loss": 2.4388, + "step": 14486 + }, + { + "epoch": 1.532043147208122, + "grad_norm": 8.1875, + "learning_rate": 4.063425383725699e-06, + "loss": 2.221, + "step": 14487 + }, + { + "epoch": 1.5321489001692048, + "grad_norm": 8.25, + "learning_rate": 4.06329174398963e-06, + "loss": 2.375, + "step": 14488 + }, + { + "epoch": 1.5322546531302876, + "grad_norm": 8.125, + "learning_rate": 4.063158096917673e-06, + "loss": 2.6344, + "step": 14489 + }, + { + "epoch": 1.5323604060913705, + "grad_norm": 8.4375, + "learning_rate": 4.0630244425104545e-06, + "loss": 2.5168, + "step": 14490 + }, + { + "epoch": 1.5324661590524533, + "grad_norm": 8.3125, + "learning_rate": 4.062890780768602e-06, + "loss": 2.5515, + "step": 14491 + }, + { + "epoch": 1.5325719120135364, + "grad_norm": 8.75, + "learning_rate": 4.062757111692742e-06, + "loss": 2.5785, + "step": 14492 + }, + { + "epoch": 1.5326776649746194, + "grad_norm": 8.25, + "learning_rate": 4.062623435283501e-06, + "loss": 2.4932, + "step": 14493 + }, + { + "epoch": 1.5327834179357023, + "grad_norm": 8.1875, + "learning_rate": 4.0624897515415084e-06, + "loss": 2.4024, + "step": 14494 + }, + { + "epoch": 1.5328891708967851, + "grad_norm": 8.25, + "learning_rate": 4.062356060467391e-06, + "loss": 2.5377, + "step": 14495 + }, + { + "epoch": 1.532994923857868, + "grad_norm": 8.1875, + "learning_rate": 4.062222362061774e-06, + "loss": 2.6524, + "step": 14496 + }, + { + "epoch": 1.5331006768189508, + "grad_norm": 8.5, + "learning_rate": 4.062088656325287e-06, + "loss": 2.711, + "step": 14497 + }, + { + "epoch": 1.5332064297800339, + "grad_norm": 7.90625, + "learning_rate": 4.061954943258557e-06, + "loss": 2.5457, + "step": 14498 + }, + { + "epoch": 1.533312182741117, + "grad_norm": 8.5, + "learning_rate": 4.061821222862211e-06, + "loss": 2.6618, + "step": 14499 + }, + { + "epoch": 1.5334179357021998, + "grad_norm": 7.96875, + "learning_rate": 4.061687495136878e-06, + "loss": 2.5158, + "step": 14500 + }, + { + "epoch": 1.5335236886632826, + "grad_norm": 8.9375, + "learning_rate": 4.061553760083183e-06, + "loss": 2.631, + "step": 14501 + }, + { + "epoch": 1.5336294416243654, + "grad_norm": 8.5, + "learning_rate": 4.061420017701755e-06, + "loss": 2.5592, + "step": 14502 + }, + { + "epoch": 1.5337351945854483, + "grad_norm": 8.0625, + "learning_rate": 4.061286267993222e-06, + "loss": 2.5934, + "step": 14503 + }, + { + "epoch": 1.5338409475465313, + "grad_norm": 8.5, + "learning_rate": 4.06115251095821e-06, + "loss": 2.6874, + "step": 14504 + }, + { + "epoch": 1.5339467005076142, + "grad_norm": 8.5, + "learning_rate": 4.061018746597349e-06, + "loss": 2.5165, + "step": 14505 + }, + { + "epoch": 1.5340524534686972, + "grad_norm": 8.5, + "learning_rate": 4.060884974911265e-06, + "loss": 2.2881, + "step": 14506 + }, + { + "epoch": 1.53415820642978, + "grad_norm": 8.6875, + "learning_rate": 4.060751195900585e-06, + "loss": 2.6107, + "step": 14507 + }, + { + "epoch": 1.534263959390863, + "grad_norm": 8.75, + "learning_rate": 4.06061740956594e-06, + "loss": 2.7144, + "step": 14508 + }, + { + "epoch": 1.5343697123519457, + "grad_norm": 8.0, + "learning_rate": 4.060483615907954e-06, + "loss": 2.2304, + "step": 14509 + }, + { + "epoch": 1.5344754653130288, + "grad_norm": 8.3125, + "learning_rate": 4.060349814927258e-06, + "loss": 2.5622, + "step": 14510 + }, + { + "epoch": 1.5345812182741116, + "grad_norm": 8.0, + "learning_rate": 4.060216006624478e-06, + "loss": 2.5412, + "step": 14511 + }, + { + "epoch": 1.5346869712351947, + "grad_norm": 7.96875, + "learning_rate": 4.060082191000242e-06, + "loss": 2.5166, + "step": 14512 + }, + { + "epoch": 1.5347927241962775, + "grad_norm": 8.25, + "learning_rate": 4.059948368055179e-06, + "loss": 2.7635, + "step": 14513 + }, + { + "epoch": 1.5348984771573604, + "grad_norm": 8.625, + "learning_rate": 4.059814537789916e-06, + "loss": 2.7528, + "step": 14514 + }, + { + "epoch": 1.5350042301184432, + "grad_norm": 8.3125, + "learning_rate": 4.059680700205081e-06, + "loss": 2.4865, + "step": 14515 + }, + { + "epoch": 1.535109983079526, + "grad_norm": 8.1875, + "learning_rate": 4.059546855301302e-06, + "loss": 2.4145, + "step": 14516 + }, + { + "epoch": 1.5352157360406091, + "grad_norm": 8.625, + "learning_rate": 4.059413003079208e-06, + "loss": 2.6051, + "step": 14517 + }, + { + "epoch": 1.5353214890016922, + "grad_norm": 8.3125, + "learning_rate": 4.059279143539427e-06, + "loss": 2.5641, + "step": 14518 + }, + { + "epoch": 1.535427241962775, + "grad_norm": 7.90625, + "learning_rate": 4.059145276682586e-06, + "loss": 2.5098, + "step": 14519 + }, + { + "epoch": 1.5355329949238579, + "grad_norm": 8.5625, + "learning_rate": 4.059011402509314e-06, + "loss": 2.4775, + "step": 14520 + }, + { + "epoch": 1.5356387478849407, + "grad_norm": 8.5, + "learning_rate": 4.058877521020239e-06, + "loss": 2.5707, + "step": 14521 + }, + { + "epoch": 1.5357445008460235, + "grad_norm": 8.0625, + "learning_rate": 4.0587436322159905e-06, + "loss": 2.7158, + "step": 14522 + }, + { + "epoch": 1.5358502538071066, + "grad_norm": 8.1875, + "learning_rate": 4.058609736097195e-06, + "loss": 2.7295, + "step": 14523 + }, + { + "epoch": 1.5359560067681897, + "grad_norm": 8.5625, + "learning_rate": 4.058475832664481e-06, + "loss": 2.4548, + "step": 14524 + }, + { + "epoch": 1.5360617597292725, + "grad_norm": 8.875, + "learning_rate": 4.058341921918477e-06, + "loss": 2.767, + "step": 14525 + }, + { + "epoch": 1.5361675126903553, + "grad_norm": 8.1875, + "learning_rate": 4.058208003859813e-06, + "loss": 2.4408, + "step": 14526 + }, + { + "epoch": 1.5362732656514382, + "grad_norm": 8.375, + "learning_rate": 4.0580740784891145e-06, + "loss": 2.2672, + "step": 14527 + }, + { + "epoch": 1.536379018612521, + "grad_norm": 8.1875, + "learning_rate": 4.057940145807013e-06, + "loss": 2.4954, + "step": 14528 + }, + { + "epoch": 1.536484771573604, + "grad_norm": 8.375, + "learning_rate": 4.057806205814135e-06, + "loss": 2.4104, + "step": 14529 + }, + { + "epoch": 1.536590524534687, + "grad_norm": 8.25, + "learning_rate": 4.057672258511109e-06, + "loss": 2.5174, + "step": 14530 + }, + { + "epoch": 1.53669627749577, + "grad_norm": 8.3125, + "learning_rate": 4.057538303898566e-06, + "loss": 2.627, + "step": 14531 + }, + { + "epoch": 1.5368020304568528, + "grad_norm": 8.3125, + "learning_rate": 4.057404341977131e-06, + "loss": 2.6317, + "step": 14532 + }, + { + "epoch": 1.5369077834179357, + "grad_norm": 8.9375, + "learning_rate": 4.057270372747435e-06, + "loss": 2.7236, + "step": 14533 + }, + { + "epoch": 1.5370135363790185, + "grad_norm": 7.9375, + "learning_rate": 4.057136396210106e-06, + "loss": 2.4219, + "step": 14534 + }, + { + "epoch": 1.5371192893401016, + "grad_norm": 8.4375, + "learning_rate": 4.057002412365773e-06, + "loss": 2.4849, + "step": 14535 + }, + { + "epoch": 1.5372250423011844, + "grad_norm": 7.90625, + "learning_rate": 4.056868421215063e-06, + "loss": 2.5741, + "step": 14536 + }, + { + "epoch": 1.5373307952622675, + "grad_norm": 7.96875, + "learning_rate": 4.056734422758608e-06, + "loss": 2.6205, + "step": 14537 + }, + { + "epoch": 1.5374365482233503, + "grad_norm": 9.125, + "learning_rate": 4.056600416997034e-06, + "loss": 2.615, + "step": 14538 + }, + { + "epoch": 1.5375423011844331, + "grad_norm": 8.0625, + "learning_rate": 4.056466403930972e-06, + "loss": 2.5504, + "step": 14539 + }, + { + "epoch": 1.537648054145516, + "grad_norm": 7.75, + "learning_rate": 4.056332383561049e-06, + "loss": 2.5986, + "step": 14540 + }, + { + "epoch": 1.537753807106599, + "grad_norm": 8.375, + "learning_rate": 4.056198355887894e-06, + "loss": 2.3693, + "step": 14541 + }, + { + "epoch": 1.5378595600676819, + "grad_norm": 8.25, + "learning_rate": 4.056064320912138e-06, + "loss": 2.4456, + "step": 14542 + }, + { + "epoch": 1.537965313028765, + "grad_norm": 8.0, + "learning_rate": 4.055930278634408e-06, + "loss": 2.5887, + "step": 14543 + }, + { + "epoch": 1.5380710659898478, + "grad_norm": 8.375, + "learning_rate": 4.055796229055332e-06, + "loss": 2.4754, + "step": 14544 + }, + { + "epoch": 1.5381768189509306, + "grad_norm": 8.4375, + "learning_rate": 4.055662172175543e-06, + "loss": 2.2149, + "step": 14545 + }, + { + "epoch": 1.5382825719120135, + "grad_norm": 7.90625, + "learning_rate": 4.055528107995667e-06, + "loss": 2.5121, + "step": 14546 + }, + { + "epoch": 1.5383883248730963, + "grad_norm": 8.1875, + "learning_rate": 4.055394036516333e-06, + "loss": 2.4336, + "step": 14547 + }, + { + "epoch": 1.5384940778341794, + "grad_norm": 8.1875, + "learning_rate": 4.055259957738171e-06, + "loss": 2.6977, + "step": 14548 + }, + { + "epoch": 1.5385998307952624, + "grad_norm": 8.5625, + "learning_rate": 4.0551258716618104e-06, + "loss": 2.2682, + "step": 14549 + }, + { + "epoch": 1.5387055837563453, + "grad_norm": 7.78125, + "learning_rate": 4.05499177828788e-06, + "loss": 2.6378, + "step": 14550 + }, + { + "epoch": 1.538811336717428, + "grad_norm": 8.6875, + "learning_rate": 4.05485767761701e-06, + "loss": 2.5998, + "step": 14551 + }, + { + "epoch": 1.538917089678511, + "grad_norm": 8.3125, + "learning_rate": 4.054723569649827e-06, + "loss": 2.6795, + "step": 14552 + }, + { + "epoch": 1.5390228426395938, + "grad_norm": 8.125, + "learning_rate": 4.054589454386964e-06, + "loss": 2.4631, + "step": 14553 + }, + { + "epoch": 1.5391285956006768, + "grad_norm": 9.25, + "learning_rate": 4.054455331829048e-06, + "loss": 2.6313, + "step": 14554 + }, + { + "epoch": 1.53923434856176, + "grad_norm": 8.0, + "learning_rate": 4.054321201976709e-06, + "loss": 2.4954, + "step": 14555 + }, + { + "epoch": 1.5393401015228427, + "grad_norm": 8.8125, + "learning_rate": 4.054187064830576e-06, + "loss": 2.5431, + "step": 14556 + }, + { + "epoch": 1.5394458544839256, + "grad_norm": 8.0, + "learning_rate": 4.054052920391279e-06, + "loss": 2.4623, + "step": 14557 + }, + { + "epoch": 1.5395516074450084, + "grad_norm": 8.5, + "learning_rate": 4.053918768659446e-06, + "loss": 2.4085, + "step": 14558 + }, + { + "epoch": 1.5396573604060912, + "grad_norm": 8.5625, + "learning_rate": 4.053784609635709e-06, + "loss": 2.4689, + "step": 14559 + }, + { + "epoch": 1.5397631133671743, + "grad_norm": 8.875, + "learning_rate": 4.053650443320696e-06, + "loss": 2.3759, + "step": 14560 + }, + { + "epoch": 1.5398688663282571, + "grad_norm": 8.125, + "learning_rate": 4.053516269715036e-06, + "loss": 2.4305, + "step": 14561 + }, + { + "epoch": 1.5399746192893402, + "grad_norm": 8.25, + "learning_rate": 4.053382088819361e-06, + "loss": 2.6268, + "step": 14562 + }, + { + "epoch": 1.540080372250423, + "grad_norm": 8.75, + "learning_rate": 4.053247900634298e-06, + "loss": 2.6985, + "step": 14563 + }, + { + "epoch": 1.5401861252115059, + "grad_norm": 8.625, + "learning_rate": 4.053113705160478e-06, + "loss": 2.5817, + "step": 14564 + }, + { + "epoch": 1.5402918781725887, + "grad_norm": 7.96875, + "learning_rate": 4.052979502398532e-06, + "loss": 2.5354, + "step": 14565 + }, + { + "epoch": 1.5403976311336718, + "grad_norm": 8.8125, + "learning_rate": 4.052845292349087e-06, + "loss": 2.5158, + "step": 14566 + }, + { + "epoch": 1.5405033840947546, + "grad_norm": 8.5, + "learning_rate": 4.052711075012774e-06, + "loss": 2.7993, + "step": 14567 + }, + { + "epoch": 1.5406091370558377, + "grad_norm": 7.96875, + "learning_rate": 4.052576850390224e-06, + "loss": 2.6196, + "step": 14568 + }, + { + "epoch": 1.5407148900169205, + "grad_norm": 8.0625, + "learning_rate": 4.052442618482066e-06, + "loss": 2.4916, + "step": 14569 + }, + { + "epoch": 1.5408206429780034, + "grad_norm": 8.375, + "learning_rate": 4.052308379288928e-06, + "loss": 2.5823, + "step": 14570 + }, + { + "epoch": 1.5409263959390862, + "grad_norm": 8.25, + "learning_rate": 4.052174132811443e-06, + "loss": 2.5486, + "step": 14571 + }, + { + "epoch": 1.541032148900169, + "grad_norm": 8.375, + "learning_rate": 4.052039879050239e-06, + "loss": 2.6658, + "step": 14572 + }, + { + "epoch": 1.541137901861252, + "grad_norm": 7.96875, + "learning_rate": 4.051905618005947e-06, + "loss": 2.6188, + "step": 14573 + }, + { + "epoch": 1.5412436548223352, + "grad_norm": 8.875, + "learning_rate": 4.051771349679197e-06, + "loss": 2.3178, + "step": 14574 + }, + { + "epoch": 1.541349407783418, + "grad_norm": 9.0625, + "learning_rate": 4.051637074070618e-06, + "loss": 2.7922, + "step": 14575 + }, + { + "epoch": 1.5414551607445008, + "grad_norm": 8.8125, + "learning_rate": 4.05150279118084e-06, + "loss": 2.592, + "step": 14576 + }, + { + "epoch": 1.5415609137055837, + "grad_norm": 8.6875, + "learning_rate": 4.0513685010104954e-06, + "loss": 2.6647, + "step": 14577 + }, + { + "epoch": 1.5416666666666665, + "grad_norm": 8.125, + "learning_rate": 4.051234203560212e-06, + "loss": 2.6295, + "step": 14578 + }, + { + "epoch": 1.5417724196277496, + "grad_norm": 8.125, + "learning_rate": 4.051099898830622e-06, + "loss": 2.377, + "step": 14579 + }, + { + "epoch": 1.5418781725888326, + "grad_norm": 8.3125, + "learning_rate": 4.050965586822354e-06, + "loss": 2.4918, + "step": 14580 + }, + { + "epoch": 1.5419839255499155, + "grad_norm": 7.9375, + "learning_rate": 4.050831267536038e-06, + "loss": 2.4167, + "step": 14581 + }, + { + "epoch": 1.5420896785109983, + "grad_norm": 8.25, + "learning_rate": 4.050696940972306e-06, + "loss": 2.2473, + "step": 14582 + }, + { + "epoch": 1.5421954314720812, + "grad_norm": 8.375, + "learning_rate": 4.050562607131788e-06, + "loss": 2.5606, + "step": 14583 + }, + { + "epoch": 1.542301184433164, + "grad_norm": 7.8125, + "learning_rate": 4.050428266015114e-06, + "loss": 2.3759, + "step": 14584 + }, + { + "epoch": 1.542406937394247, + "grad_norm": 8.25, + "learning_rate": 4.050293917622913e-06, + "loss": 2.5864, + "step": 14585 + }, + { + "epoch": 1.54251269035533, + "grad_norm": 8.5625, + "learning_rate": 4.0501595619558165e-06, + "loss": 2.3545, + "step": 14586 + }, + { + "epoch": 1.542618443316413, + "grad_norm": 8.4375, + "learning_rate": 4.050025199014455e-06, + "loss": 2.5151, + "step": 14587 + }, + { + "epoch": 1.5427241962774958, + "grad_norm": 7.78125, + "learning_rate": 4.049890828799461e-06, + "loss": 2.4994, + "step": 14588 + }, + { + "epoch": 1.5428299492385786, + "grad_norm": 7.84375, + "learning_rate": 4.049756451311462e-06, + "loss": 2.4935, + "step": 14589 + }, + { + "epoch": 1.5429357021996615, + "grad_norm": 7.90625, + "learning_rate": 4.04962206655109e-06, + "loss": 2.5148, + "step": 14590 + }, + { + "epoch": 1.5430414551607445, + "grad_norm": 7.9375, + "learning_rate": 4.049487674518976e-06, + "loss": 2.4393, + "step": 14591 + }, + { + "epoch": 1.5431472081218274, + "grad_norm": 8.6875, + "learning_rate": 4.049353275215749e-06, + "loss": 2.5046, + "step": 14592 + }, + { + "epoch": 1.5432529610829104, + "grad_norm": 8.75, + "learning_rate": 4.049218868642041e-06, + "loss": 2.5033, + "step": 14593 + }, + { + "epoch": 1.5433587140439933, + "grad_norm": 8.25, + "learning_rate": 4.049084454798483e-06, + "loss": 2.5671, + "step": 14594 + }, + { + "epoch": 1.5434644670050761, + "grad_norm": 8.8125, + "learning_rate": 4.0489500336857045e-06, + "loss": 2.6013, + "step": 14595 + }, + { + "epoch": 1.543570219966159, + "grad_norm": 8.375, + "learning_rate": 4.048815605304338e-06, + "loss": 2.6159, + "step": 14596 + }, + { + "epoch": 1.543675972927242, + "grad_norm": 8.375, + "learning_rate": 4.048681169655012e-06, + "loss": 2.7346, + "step": 14597 + }, + { + "epoch": 1.5437817258883249, + "grad_norm": 8.5, + "learning_rate": 4.04854672673836e-06, + "loss": 2.2859, + "step": 14598 + }, + { + "epoch": 1.543887478849408, + "grad_norm": 7.875, + "learning_rate": 4.048412276555011e-06, + "loss": 2.5131, + "step": 14599 + }, + { + "epoch": 1.5439932318104908, + "grad_norm": 8.1875, + "learning_rate": 4.048277819105596e-06, + "loss": 2.5384, + "step": 14600 + }, + { + "epoch": 1.5440989847715736, + "grad_norm": 8.125, + "learning_rate": 4.048143354390747e-06, + "loss": 2.438, + "step": 14601 + }, + { + "epoch": 1.5442047377326564, + "grad_norm": 8.5, + "learning_rate": 4.048008882411094e-06, + "loss": 2.7286, + "step": 14602 + }, + { + "epoch": 1.5443104906937393, + "grad_norm": 8.1875, + "learning_rate": 4.047874403167268e-06, + "loss": 2.5769, + "step": 14603 + }, + { + "epoch": 1.5444162436548223, + "grad_norm": 8.75, + "learning_rate": 4.047739916659902e-06, + "loss": 2.5014, + "step": 14604 + }, + { + "epoch": 1.5445219966159054, + "grad_norm": 8.375, + "learning_rate": 4.047605422889625e-06, + "loss": 2.5591, + "step": 14605 + }, + { + "epoch": 1.5446277495769882, + "grad_norm": 8.4375, + "learning_rate": 4.047470921857068e-06, + "loss": 2.5997, + "step": 14606 + }, + { + "epoch": 1.544733502538071, + "grad_norm": 8.25, + "learning_rate": 4.047336413562864e-06, + "loss": 2.7264, + "step": 14607 + }, + { + "epoch": 1.544839255499154, + "grad_norm": 8.125, + "learning_rate": 4.047201898007642e-06, + "loss": 2.5427, + "step": 14608 + }, + { + "epoch": 1.5449450084602367, + "grad_norm": 8.25, + "learning_rate": 4.047067375192034e-06, + "loss": 2.6435, + "step": 14609 + }, + { + "epoch": 1.5450507614213198, + "grad_norm": 8.4375, + "learning_rate": 4.046932845116673e-06, + "loss": 2.4715, + "step": 14610 + }, + { + "epoch": 1.5451565143824029, + "grad_norm": 8.1875, + "learning_rate": 4.046798307782187e-06, + "loss": 2.6123, + "step": 14611 + }, + { + "epoch": 1.5452622673434857, + "grad_norm": 8.8125, + "learning_rate": 4.04666376318921e-06, + "loss": 2.4051, + "step": 14612 + }, + { + "epoch": 1.5453680203045685, + "grad_norm": 8.0625, + "learning_rate": 4.046529211338372e-06, + "loss": 2.4795, + "step": 14613 + }, + { + "epoch": 1.5454737732656514, + "grad_norm": 8.0, + "learning_rate": 4.046394652230307e-06, + "loss": 2.5535, + "step": 14614 + }, + { + "epoch": 1.5455795262267342, + "grad_norm": 7.90625, + "learning_rate": 4.046260085865642e-06, + "loss": 2.4212, + "step": 14615 + }, + { + "epoch": 1.5456852791878173, + "grad_norm": 8.625, + "learning_rate": 4.046125512245011e-06, + "loss": 2.6862, + "step": 14616 + }, + { + "epoch": 1.5457910321489001, + "grad_norm": 8.0625, + "learning_rate": 4.045990931369047e-06, + "loss": 2.6017, + "step": 14617 + }, + { + "epoch": 1.5458967851099832, + "grad_norm": 8.125, + "learning_rate": 4.045856343238379e-06, + "loss": 2.441, + "step": 14618 + }, + { + "epoch": 1.546002538071066, + "grad_norm": 8.4375, + "learning_rate": 4.045721747853639e-06, + "loss": 2.4781, + "step": 14619 + }, + { + "epoch": 1.5461082910321489, + "grad_norm": 8.5625, + "learning_rate": 4.045587145215459e-06, + "loss": 2.5102, + "step": 14620 + }, + { + "epoch": 1.5462140439932317, + "grad_norm": 8.75, + "learning_rate": 4.045452535324471e-06, + "loss": 2.8658, + "step": 14621 + }, + { + "epoch": 1.5463197969543148, + "grad_norm": 7.875, + "learning_rate": 4.045317918181307e-06, + "loss": 2.6234, + "step": 14622 + }, + { + "epoch": 1.5464255499153976, + "grad_norm": 8.5, + "learning_rate": 4.045183293786597e-06, + "loss": 2.5221, + "step": 14623 + }, + { + "epoch": 1.5465313028764807, + "grad_norm": 8.4375, + "learning_rate": 4.045048662140974e-06, + "loss": 2.7021, + "step": 14624 + }, + { + "epoch": 1.5466370558375635, + "grad_norm": 7.78125, + "learning_rate": 4.044914023245069e-06, + "loss": 2.4085, + "step": 14625 + }, + { + "epoch": 1.5467428087986463, + "grad_norm": 7.96875, + "learning_rate": 4.044779377099516e-06, + "loss": 2.4888, + "step": 14626 + }, + { + "epoch": 1.5468485617597292, + "grad_norm": 8.0, + "learning_rate": 4.044644723704944e-06, + "loss": 2.3686, + "step": 14627 + }, + { + "epoch": 1.546954314720812, + "grad_norm": 8.0, + "learning_rate": 4.044510063061986e-06, + "loss": 2.4598, + "step": 14628 + }, + { + "epoch": 1.547060067681895, + "grad_norm": 7.875, + "learning_rate": 4.044375395171274e-06, + "loss": 2.3263, + "step": 14629 + }, + { + "epoch": 1.5471658206429781, + "grad_norm": 8.125, + "learning_rate": 4.04424072003344e-06, + "loss": 2.7198, + "step": 14630 + }, + { + "epoch": 1.547271573604061, + "grad_norm": 8.375, + "learning_rate": 4.044106037649116e-06, + "loss": 2.5524, + "step": 14631 + }, + { + "epoch": 1.5473773265651438, + "grad_norm": 8.1875, + "learning_rate": 4.043971348018933e-06, + "loss": 2.3961, + "step": 14632 + }, + { + "epoch": 1.5474830795262267, + "grad_norm": 8.5, + "learning_rate": 4.043836651143525e-06, + "loss": 2.4208, + "step": 14633 + }, + { + "epoch": 1.5475888324873095, + "grad_norm": 8.3125, + "learning_rate": 4.043701947023522e-06, + "loss": 2.6684, + "step": 14634 + }, + { + "epoch": 1.5476945854483926, + "grad_norm": 7.8125, + "learning_rate": 4.043567235659559e-06, + "loss": 2.4776, + "step": 14635 + }, + { + "epoch": 1.5478003384094756, + "grad_norm": 8.875, + "learning_rate": 4.043432517052265e-06, + "loss": 2.8205, + "step": 14636 + }, + { + "epoch": 1.5479060913705585, + "grad_norm": 8.125, + "learning_rate": 4.043297791202273e-06, + "loss": 2.4057, + "step": 14637 + }, + { + "epoch": 1.5480118443316413, + "grad_norm": 8.4375, + "learning_rate": 4.043163058110217e-06, + "loss": 2.6785, + "step": 14638 + }, + { + "epoch": 1.5481175972927241, + "grad_norm": 8.1875, + "learning_rate": 4.043028317776727e-06, + "loss": 2.2387, + "step": 14639 + }, + { + "epoch": 1.548223350253807, + "grad_norm": 8.4375, + "learning_rate": 4.042893570202436e-06, + "loss": 2.4403, + "step": 14640 + }, + { + "epoch": 1.54832910321489, + "grad_norm": 8.375, + "learning_rate": 4.042758815387976e-06, + "loss": 2.4421, + "step": 14641 + }, + { + "epoch": 1.5484348561759729, + "grad_norm": 7.96875, + "learning_rate": 4.042624053333981e-06, + "loss": 2.5738, + "step": 14642 + }, + { + "epoch": 1.548540609137056, + "grad_norm": 8.625, + "learning_rate": 4.042489284041082e-06, + "loss": 2.8075, + "step": 14643 + }, + { + "epoch": 1.5486463620981388, + "grad_norm": 7.8125, + "learning_rate": 4.042354507509911e-06, + "loss": 2.5436, + "step": 14644 + }, + { + "epoch": 1.5487521150592216, + "grad_norm": 8.0625, + "learning_rate": 4.042219723741102e-06, + "loss": 2.6739, + "step": 14645 + }, + { + "epoch": 1.5488578680203045, + "grad_norm": 8.1875, + "learning_rate": 4.042084932735286e-06, + "loss": 2.6191, + "step": 14646 + }, + { + "epoch": 1.5489636209813875, + "grad_norm": 8.0625, + "learning_rate": 4.041950134493096e-06, + "loss": 2.4133, + "step": 14647 + }, + { + "epoch": 1.5490693739424704, + "grad_norm": 8.375, + "learning_rate": 4.041815329015165e-06, + "loss": 2.7638, + "step": 14648 + }, + { + "epoch": 1.5491751269035534, + "grad_norm": 8.4375, + "learning_rate": 4.0416805163021255e-06, + "loss": 2.5502, + "step": 14649 + }, + { + "epoch": 1.5492808798646363, + "grad_norm": 7.96875, + "learning_rate": 4.0415456963546094e-06, + "loss": 2.5041, + "step": 14650 + }, + { + "epoch": 1.549386632825719, + "grad_norm": 8.3125, + "learning_rate": 4.04141086917325e-06, + "loss": 2.4809, + "step": 14651 + }, + { + "epoch": 1.549492385786802, + "grad_norm": 8.125, + "learning_rate": 4.04127603475868e-06, + "loss": 2.6242, + "step": 14652 + }, + { + "epoch": 1.549598138747885, + "grad_norm": 8.3125, + "learning_rate": 4.041141193111531e-06, + "loss": 2.5522, + "step": 14653 + }, + { + "epoch": 1.5497038917089678, + "grad_norm": 8.375, + "learning_rate": 4.041006344232438e-06, + "loss": 2.4708, + "step": 14654 + }, + { + "epoch": 1.549809644670051, + "grad_norm": 9.0625, + "learning_rate": 4.040871488122031e-06, + "loss": 2.7083, + "step": 14655 + }, + { + "epoch": 1.5499153976311337, + "grad_norm": 8.0625, + "learning_rate": 4.040736624780946e-06, + "loss": 2.6602, + "step": 14656 + }, + { + "epoch": 1.5500211505922166, + "grad_norm": 8.4375, + "learning_rate": 4.040601754209813e-06, + "loss": 2.4519, + "step": 14657 + }, + { + "epoch": 1.5501269035532994, + "grad_norm": 8.8125, + "learning_rate": 4.040466876409266e-06, + "loss": 2.3457, + "step": 14658 + }, + { + "epoch": 1.5502326565143822, + "grad_norm": 8.5625, + "learning_rate": 4.040331991379939e-06, + "loss": 2.5519, + "step": 14659 + }, + { + "epoch": 1.5503384094754653, + "grad_norm": 8.8125, + "learning_rate": 4.040197099122463e-06, + "loss": 2.5921, + "step": 14660 + }, + { + "epoch": 1.5504441624365484, + "grad_norm": 8.8125, + "learning_rate": 4.040062199637472e-06, + "loss": 2.3719, + "step": 14661 + }, + { + "epoch": 1.5505499153976312, + "grad_norm": 8.0, + "learning_rate": 4.0399272929256e-06, + "loss": 2.3819, + "step": 14662 + }, + { + "epoch": 1.550655668358714, + "grad_norm": 8.0, + "learning_rate": 4.039792378987477e-06, + "loss": 2.6722, + "step": 14663 + }, + { + "epoch": 1.5507614213197969, + "grad_norm": 8.625, + "learning_rate": 4.039657457823739e-06, + "loss": 2.4306, + "step": 14664 + }, + { + "epoch": 1.5508671742808797, + "grad_norm": 8.0, + "learning_rate": 4.039522529435019e-06, + "loss": 2.5414, + "step": 14665 + }, + { + "epoch": 1.5509729272419628, + "grad_norm": 7.90625, + "learning_rate": 4.039387593821949e-06, + "loss": 2.4029, + "step": 14666 + }, + { + "epoch": 1.5510786802030458, + "grad_norm": 8.25, + "learning_rate": 4.039252650985162e-06, + "loss": 2.2986, + "step": 14667 + }, + { + "epoch": 1.5511844331641287, + "grad_norm": 7.96875, + "learning_rate": 4.039117700925293e-06, + "loss": 2.4376, + "step": 14668 + }, + { + "epoch": 1.5512901861252115, + "grad_norm": 8.375, + "learning_rate": 4.038982743642974e-06, + "loss": 2.3506, + "step": 14669 + }, + { + "epoch": 1.5513959390862944, + "grad_norm": 8.5625, + "learning_rate": 4.038847779138837e-06, + "loss": 2.5146, + "step": 14670 + }, + { + "epoch": 1.5515016920473772, + "grad_norm": 8.0625, + "learning_rate": 4.0387128074135185e-06, + "loss": 2.5173, + "step": 14671 + }, + { + "epoch": 1.5516074450084603, + "grad_norm": 8.0625, + "learning_rate": 4.038577828467649e-06, + "loss": 2.4743, + "step": 14672 + }, + { + "epoch": 1.551713197969543, + "grad_norm": 8.1875, + "learning_rate": 4.0384428423018626e-06, + "loss": 2.524, + "step": 14673 + }, + { + "epoch": 1.5518189509306262, + "grad_norm": 8.375, + "learning_rate": 4.038307848916794e-06, + "loss": 2.6338, + "step": 14674 + }, + { + "epoch": 1.551924703891709, + "grad_norm": 8.3125, + "learning_rate": 4.0381728483130756e-06, + "loss": 2.5688, + "step": 14675 + }, + { + "epoch": 1.5520304568527918, + "grad_norm": 8.375, + "learning_rate": 4.038037840491341e-06, + "loss": 2.5283, + "step": 14676 + }, + { + "epoch": 1.5521362098138747, + "grad_norm": 7.8125, + "learning_rate": 4.0379028254522244e-06, + "loss": 2.6487, + "step": 14677 + }, + { + "epoch": 1.5522419627749577, + "grad_norm": 7.90625, + "learning_rate": 4.037767803196358e-06, + "loss": 2.6054, + "step": 14678 + }, + { + "epoch": 1.5523477157360406, + "grad_norm": 8.0625, + "learning_rate": 4.037632773724377e-06, + "loss": 2.4064, + "step": 14679 + }, + { + "epoch": 1.5524534686971236, + "grad_norm": 7.90625, + "learning_rate": 4.037497737036914e-06, + "loss": 2.4451, + "step": 14680 + }, + { + "epoch": 1.5525592216582065, + "grad_norm": 8.3125, + "learning_rate": 4.0373626931346035e-06, + "loss": 2.1652, + "step": 14681 + }, + { + "epoch": 1.5526649746192893, + "grad_norm": 8.25, + "learning_rate": 4.0372276420180776e-06, + "loss": 2.3414, + "step": 14682 + }, + { + "epoch": 1.5527707275803722, + "grad_norm": 8.5, + "learning_rate": 4.037092583687971e-06, + "loss": 2.47, + "step": 14683 + }, + { + "epoch": 1.552876480541455, + "grad_norm": 8.3125, + "learning_rate": 4.036957518144918e-06, + "loss": 2.5921, + "step": 14684 + }, + { + "epoch": 1.552982233502538, + "grad_norm": 8.875, + "learning_rate": 4.036822445389552e-06, + "loss": 2.6036, + "step": 14685 + }, + { + "epoch": 1.5530879864636211, + "grad_norm": 8.375, + "learning_rate": 4.036687365422507e-06, + "loss": 2.3703, + "step": 14686 + }, + { + "epoch": 1.553193739424704, + "grad_norm": 8.6875, + "learning_rate": 4.036552278244417e-06, + "loss": 2.464, + "step": 14687 + }, + { + "epoch": 1.5532994923857868, + "grad_norm": 8.4375, + "learning_rate": 4.036417183855914e-06, + "loss": 2.5947, + "step": 14688 + }, + { + "epoch": 1.5534052453468696, + "grad_norm": 8.25, + "learning_rate": 4.036282082257635e-06, + "loss": 2.7312, + "step": 14689 + }, + { + "epoch": 1.5535109983079525, + "grad_norm": 8.125, + "learning_rate": 4.036146973450213e-06, + "loss": 2.6274, + "step": 14690 + }, + { + "epoch": 1.5536167512690355, + "grad_norm": 8.1875, + "learning_rate": 4.036011857434279e-06, + "loss": 2.4731, + "step": 14691 + }, + { + "epoch": 1.5537225042301186, + "grad_norm": 9.0, + "learning_rate": 4.035876734210472e-06, + "loss": 2.7264, + "step": 14692 + }, + { + "epoch": 1.5538282571912014, + "grad_norm": 8.75, + "learning_rate": 4.035741603779423e-06, + "loss": 2.7426, + "step": 14693 + }, + { + "epoch": 1.5539340101522843, + "grad_norm": 8.3125, + "learning_rate": 4.035606466141766e-06, + "loss": 2.5784, + "step": 14694 + }, + { + "epoch": 1.5540397631133671, + "grad_norm": 8.5, + "learning_rate": 4.035471321298136e-06, + "loss": 2.7361, + "step": 14695 + }, + { + "epoch": 1.55414551607445, + "grad_norm": 7.8125, + "learning_rate": 4.035336169249168e-06, + "loss": 2.4183, + "step": 14696 + }, + { + "epoch": 1.554251269035533, + "grad_norm": 8.875, + "learning_rate": 4.0352010099954955e-06, + "loss": 2.7451, + "step": 14697 + }, + { + "epoch": 1.5543570219966159, + "grad_norm": 8.5, + "learning_rate": 4.035065843537751e-06, + "loss": 2.4647, + "step": 14698 + }, + { + "epoch": 1.554462774957699, + "grad_norm": 8.375, + "learning_rate": 4.034930669876571e-06, + "loss": 2.5311, + "step": 14699 + }, + { + "epoch": 1.5545685279187818, + "grad_norm": 7.875, + "learning_rate": 4.034795489012589e-06, + "loss": 2.2539, + "step": 14700 + }, + { + "epoch": 1.5546742808798646, + "grad_norm": 8.3125, + "learning_rate": 4.03466030094644e-06, + "loss": 2.3598, + "step": 14701 + }, + { + "epoch": 1.5547800338409474, + "grad_norm": 8.875, + "learning_rate": 4.034525105678758e-06, + "loss": 2.5693, + "step": 14702 + }, + { + "epoch": 1.5548857868020305, + "grad_norm": 8.4375, + "learning_rate": 4.034389903210177e-06, + "loss": 2.3919, + "step": 14703 + }, + { + "epoch": 1.5549915397631133, + "grad_norm": 8.0, + "learning_rate": 4.0342546935413315e-06, + "loss": 2.1371, + "step": 14704 + }, + { + "epoch": 1.5550972927241964, + "grad_norm": 8.0, + "learning_rate": 4.034119476672857e-06, + "loss": 2.5664, + "step": 14705 + }, + { + "epoch": 1.5552030456852792, + "grad_norm": 7.71875, + "learning_rate": 4.033984252605386e-06, + "loss": 2.6215, + "step": 14706 + }, + { + "epoch": 1.555308798646362, + "grad_norm": 8.4375, + "learning_rate": 4.033849021339555e-06, + "loss": 2.6128, + "step": 14707 + }, + { + "epoch": 1.555414551607445, + "grad_norm": 8.0, + "learning_rate": 4.0337137828759975e-06, + "loss": 2.6547, + "step": 14708 + }, + { + "epoch": 1.555520304568528, + "grad_norm": 8.4375, + "learning_rate": 4.033578537215348e-06, + "loss": 2.6537, + "step": 14709 + }, + { + "epoch": 1.5556260575296108, + "grad_norm": 8.9375, + "learning_rate": 4.0334432843582425e-06, + "loss": 2.095, + "step": 14710 + }, + { + "epoch": 1.5557318104906939, + "grad_norm": 7.96875, + "learning_rate": 4.033308024305315e-06, + "loss": 2.4819, + "step": 14711 + }, + { + "epoch": 1.5558375634517767, + "grad_norm": 8.3125, + "learning_rate": 4.0331727570572e-06, + "loss": 2.5372, + "step": 14712 + }, + { + "epoch": 1.5559433164128595, + "grad_norm": 7.96875, + "learning_rate": 4.033037482614532e-06, + "loss": 2.5151, + "step": 14713 + }, + { + "epoch": 1.5560490693739424, + "grad_norm": 8.1875, + "learning_rate": 4.0329022009779464e-06, + "loss": 2.4336, + "step": 14714 + }, + { + "epoch": 1.5561548223350252, + "grad_norm": 8.625, + "learning_rate": 4.0327669121480774e-06, + "loss": 2.6567, + "step": 14715 + }, + { + "epoch": 1.5562605752961083, + "grad_norm": 8.6875, + "learning_rate": 4.0326316161255596e-06, + "loss": 2.6837, + "step": 14716 + }, + { + "epoch": 1.5563663282571913, + "grad_norm": 8.8125, + "learning_rate": 4.03249631291103e-06, + "loss": 2.6092, + "step": 14717 + }, + { + "epoch": 1.5564720812182742, + "grad_norm": 7.8125, + "learning_rate": 4.03236100250512e-06, + "loss": 2.548, + "step": 14718 + }, + { + "epoch": 1.556577834179357, + "grad_norm": 8.8125, + "learning_rate": 4.032225684908467e-06, + "loss": 2.6351, + "step": 14719 + }, + { + "epoch": 1.5566835871404399, + "grad_norm": 8.125, + "learning_rate": 4.0320903601217064e-06, + "loss": 2.6674, + "step": 14720 + }, + { + "epoch": 1.5567893401015227, + "grad_norm": 8.25, + "learning_rate": 4.0319550281454724e-06, + "loss": 2.75, + "step": 14721 + }, + { + "epoch": 1.5568950930626058, + "grad_norm": 7.96875, + "learning_rate": 4.0318196889803995e-06, + "loss": 2.4154, + "step": 14722 + }, + { + "epoch": 1.5570008460236888, + "grad_norm": 8.1875, + "learning_rate": 4.031684342627124e-06, + "loss": 2.2681, + "step": 14723 + }, + { + "epoch": 1.5571065989847717, + "grad_norm": 8.25, + "learning_rate": 4.031548989086279e-06, + "loss": 2.5144, + "step": 14724 + }, + { + "epoch": 1.5572123519458545, + "grad_norm": 8.125, + "learning_rate": 4.0314136283585016e-06, + "loss": 2.6412, + "step": 14725 + }, + { + "epoch": 1.5573181049069373, + "grad_norm": 8.375, + "learning_rate": 4.0312782604444265e-06, + "loss": 2.3866, + "step": 14726 + }, + { + "epoch": 1.5574238578680202, + "grad_norm": 8.3125, + "learning_rate": 4.031142885344689e-06, + "loss": 2.6481, + "step": 14727 + }, + { + "epoch": 1.5575296108291032, + "grad_norm": 9.1875, + "learning_rate": 4.031007503059924e-06, + "loss": 2.8741, + "step": 14728 + }, + { + "epoch": 1.557635363790186, + "grad_norm": 8.625, + "learning_rate": 4.030872113590766e-06, + "loss": 2.5451, + "step": 14729 + }, + { + "epoch": 1.5577411167512691, + "grad_norm": 7.96875, + "learning_rate": 4.030736716937853e-06, + "loss": 2.3117, + "step": 14730 + }, + { + "epoch": 1.557846869712352, + "grad_norm": 8.375, + "learning_rate": 4.030601313101818e-06, + "loss": 2.3952, + "step": 14731 + }, + { + "epoch": 1.5579526226734348, + "grad_norm": 8.0, + "learning_rate": 4.0304659020832955e-06, + "loss": 2.6384, + "step": 14732 + }, + { + "epoch": 1.5580583756345177, + "grad_norm": 8.0625, + "learning_rate": 4.030330483882924e-06, + "loss": 2.6239, + "step": 14733 + }, + { + "epoch": 1.5581641285956007, + "grad_norm": 8.9375, + "learning_rate": 4.030195058501337e-06, + "loss": 2.5659, + "step": 14734 + }, + { + "epoch": 1.5582698815566836, + "grad_norm": 8.4375, + "learning_rate": 4.030059625939171e-06, + "loss": 2.59, + "step": 14735 + }, + { + "epoch": 1.5583756345177666, + "grad_norm": 8.1875, + "learning_rate": 4.029924186197061e-06, + "loss": 2.6522, + "step": 14736 + }, + { + "epoch": 1.5584813874788495, + "grad_norm": 8.1875, + "learning_rate": 4.029788739275641e-06, + "loss": 2.6375, + "step": 14737 + }, + { + "epoch": 1.5585871404399323, + "grad_norm": 8.25, + "learning_rate": 4.02965328517555e-06, + "loss": 2.7766, + "step": 14738 + }, + { + "epoch": 1.5586928934010151, + "grad_norm": 7.75, + "learning_rate": 4.0295178238974205e-06, + "loss": 2.5425, + "step": 14739 + }, + { + "epoch": 1.558798646362098, + "grad_norm": 8.0625, + "learning_rate": 4.02938235544189e-06, + "loss": 2.5404, + "step": 14740 + }, + { + "epoch": 1.558904399323181, + "grad_norm": 7.71875, + "learning_rate": 4.029246879809593e-06, + "loss": 2.5739, + "step": 14741 + }, + { + "epoch": 1.559010152284264, + "grad_norm": 7.84375, + "learning_rate": 4.029111397001166e-06, + "loss": 2.4895, + "step": 14742 + }, + { + "epoch": 1.559115905245347, + "grad_norm": 8.5625, + "learning_rate": 4.0289759070172454e-06, + "loss": 2.6791, + "step": 14743 + }, + { + "epoch": 1.5592216582064298, + "grad_norm": 8.25, + "learning_rate": 4.028840409858465e-06, + "loss": 2.4302, + "step": 14744 + }, + { + "epoch": 1.5593274111675126, + "grad_norm": 8.0, + "learning_rate": 4.028704905525463e-06, + "loss": 2.611, + "step": 14745 + }, + { + "epoch": 1.5594331641285955, + "grad_norm": 8.375, + "learning_rate": 4.028569394018873e-06, + "loss": 2.3393, + "step": 14746 + }, + { + "epoch": 1.5595389170896785, + "grad_norm": 8.4375, + "learning_rate": 4.028433875339333e-06, + "loss": 2.7132, + "step": 14747 + }, + { + "epoch": 1.5596446700507616, + "grad_norm": 8.0625, + "learning_rate": 4.0282983494874775e-06, + "loss": 2.4373, + "step": 14748 + }, + { + "epoch": 1.5597504230118444, + "grad_norm": 7.78125, + "learning_rate": 4.028162816463943e-06, + "loss": 2.434, + "step": 14749 + }, + { + "epoch": 1.5598561759729273, + "grad_norm": 7.75, + "learning_rate": 4.0280272762693654e-06, + "loss": 2.5497, + "step": 14750 + }, + { + "epoch": 1.55996192893401, + "grad_norm": 8.125, + "learning_rate": 4.027891728904381e-06, + "loss": 2.5694, + "step": 14751 + }, + { + "epoch": 1.560067681895093, + "grad_norm": 7.84375, + "learning_rate": 4.027756174369625e-06, + "loss": 2.5504, + "step": 14752 + }, + { + "epoch": 1.560173434856176, + "grad_norm": 8.375, + "learning_rate": 4.027620612665733e-06, + "loss": 2.5171, + "step": 14753 + }, + { + "epoch": 1.5602791878172588, + "grad_norm": 7.75, + "learning_rate": 4.027485043793345e-06, + "loss": 2.4869, + "step": 14754 + }, + { + "epoch": 1.560384940778342, + "grad_norm": 8.1875, + "learning_rate": 4.027349467753092e-06, + "loss": 2.6193, + "step": 14755 + }, + { + "epoch": 1.5604906937394247, + "grad_norm": 8.6875, + "learning_rate": 4.027213884545615e-06, + "loss": 2.7002, + "step": 14756 + }, + { + "epoch": 1.5605964467005076, + "grad_norm": 8.4375, + "learning_rate": 4.0270782941715455e-06, + "loss": 2.5885, + "step": 14757 + }, + { + "epoch": 1.5607021996615904, + "grad_norm": 8.5, + "learning_rate": 4.026942696631523e-06, + "loss": 2.2623, + "step": 14758 + }, + { + "epoch": 1.5608079526226735, + "grad_norm": 8.6875, + "learning_rate": 4.026807091926183e-06, + "loss": 2.6447, + "step": 14759 + }, + { + "epoch": 1.5609137055837563, + "grad_norm": 8.0, + "learning_rate": 4.0266714800561615e-06, + "loss": 2.68, + "step": 14760 + }, + { + "epoch": 1.5610194585448394, + "grad_norm": 7.71875, + "learning_rate": 4.026535861022096e-06, + "loss": 2.3924, + "step": 14761 + }, + { + "epoch": 1.5611252115059222, + "grad_norm": 8.1875, + "learning_rate": 4.02640023482462e-06, + "loss": 2.4968, + "step": 14762 + }, + { + "epoch": 1.561230964467005, + "grad_norm": 8.1875, + "learning_rate": 4.0262646014643745e-06, + "loss": 2.6078, + "step": 14763 + }, + { + "epoch": 1.5613367174280879, + "grad_norm": 8.5, + "learning_rate": 4.026128960941992e-06, + "loss": 2.5383, + "step": 14764 + }, + { + "epoch": 1.561442470389171, + "grad_norm": 8.1875, + "learning_rate": 4.025993313258111e-06, + "loss": 2.5362, + "step": 14765 + }, + { + "epoch": 1.5615482233502538, + "grad_norm": 8.1875, + "learning_rate": 4.0258576584133675e-06, + "loss": 2.769, + "step": 14766 + }, + { + "epoch": 1.5616539763113368, + "grad_norm": 8.6875, + "learning_rate": 4.025721996408398e-06, + "loss": 2.9474, + "step": 14767 + }, + { + "epoch": 1.5617597292724197, + "grad_norm": 8.5625, + "learning_rate": 4.025586327243839e-06, + "loss": 2.5624, + "step": 14768 + }, + { + "epoch": 1.5618654822335025, + "grad_norm": 8.3125, + "learning_rate": 4.025450650920327e-06, + "loss": 2.5482, + "step": 14769 + }, + { + "epoch": 1.5619712351945854, + "grad_norm": 8.625, + "learning_rate": 4.0253149674385e-06, + "loss": 2.5848, + "step": 14770 + }, + { + "epoch": 1.5620769881556682, + "grad_norm": 8.375, + "learning_rate": 4.025179276798992e-06, + "loss": 2.5627, + "step": 14771 + }, + { + "epoch": 1.5621827411167513, + "grad_norm": 7.78125, + "learning_rate": 4.025043579002442e-06, + "loss": 2.4159, + "step": 14772 + }, + { + "epoch": 1.5622884940778343, + "grad_norm": 7.6875, + "learning_rate": 4.024907874049488e-06, + "loss": 2.3635, + "step": 14773 + }, + { + "epoch": 1.5623942470389172, + "grad_norm": 8.1875, + "learning_rate": 4.024772161940763e-06, + "loss": 2.6129, + "step": 14774 + }, + { + "epoch": 1.5625, + "grad_norm": 8.8125, + "learning_rate": 4.024636442676907e-06, + "loss": 2.4938, + "step": 14775 + }, + { + "epoch": 1.5626057529610828, + "grad_norm": 8.4375, + "learning_rate": 4.024500716258555e-06, + "loss": 2.592, + "step": 14776 + }, + { + "epoch": 1.5627115059221657, + "grad_norm": 8.875, + "learning_rate": 4.024364982686344e-06, + "loss": 2.6045, + "step": 14777 + }, + { + "epoch": 1.5628172588832487, + "grad_norm": 8.1875, + "learning_rate": 4.024229241960913e-06, + "loss": 2.6964, + "step": 14778 + }, + { + "epoch": 1.5629230118443318, + "grad_norm": 8.375, + "learning_rate": 4.024093494082897e-06, + "loss": 2.5446, + "step": 14779 + }, + { + "epoch": 1.5630287648054146, + "grad_norm": 8.0, + "learning_rate": 4.023957739052933e-06, + "loss": 2.4616, + "step": 14780 + }, + { + "epoch": 1.5631345177664975, + "grad_norm": 8.375, + "learning_rate": 4.023821976871659e-06, + "loss": 2.3771, + "step": 14781 + }, + { + "epoch": 1.5632402707275803, + "grad_norm": 8.6875, + "learning_rate": 4.023686207539712e-06, + "loss": 2.411, + "step": 14782 + }, + { + "epoch": 1.5633460236886632, + "grad_norm": 8.5, + "learning_rate": 4.023550431057729e-06, + "loss": 2.4535, + "step": 14783 + }, + { + "epoch": 1.5634517766497462, + "grad_norm": 8.0625, + "learning_rate": 4.023414647426346e-06, + "loss": 2.5805, + "step": 14784 + }, + { + "epoch": 1.563557529610829, + "grad_norm": 8.375, + "learning_rate": 4.023278856646202e-06, + "loss": 2.7734, + "step": 14785 + }, + { + "epoch": 1.5636632825719121, + "grad_norm": 8.3125, + "learning_rate": 4.023143058717933e-06, + "loss": 2.4054, + "step": 14786 + }, + { + "epoch": 1.563769035532995, + "grad_norm": 8.4375, + "learning_rate": 4.023007253642177e-06, + "loss": 2.7382, + "step": 14787 + }, + { + "epoch": 1.5638747884940778, + "grad_norm": 8.5625, + "learning_rate": 4.02287144141957e-06, + "loss": 2.6755, + "step": 14788 + }, + { + "epoch": 1.5639805414551606, + "grad_norm": 8.125, + "learning_rate": 4.022735622050751e-06, + "loss": 2.5873, + "step": 14789 + }, + { + "epoch": 1.5640862944162437, + "grad_norm": 8.1875, + "learning_rate": 4.022599795536356e-06, + "loss": 2.614, + "step": 14790 + }, + { + "epoch": 1.5641920473773265, + "grad_norm": 8.375, + "learning_rate": 4.022463961877023e-06, + "loss": 2.6631, + "step": 14791 + }, + { + "epoch": 1.5642978003384096, + "grad_norm": 7.9375, + "learning_rate": 4.022328121073389e-06, + "loss": 2.41, + "step": 14792 + }, + { + "epoch": 1.5644035532994924, + "grad_norm": 8.1875, + "learning_rate": 4.022192273126092e-06, + "loss": 2.6238, + "step": 14793 + }, + { + "epoch": 1.5645093062605753, + "grad_norm": 7.90625, + "learning_rate": 4.0220564180357694e-06, + "loss": 2.6751, + "step": 14794 + }, + { + "epoch": 1.564615059221658, + "grad_norm": 8.4375, + "learning_rate": 4.021920555803058e-06, + "loss": 2.705, + "step": 14795 + }, + { + "epoch": 1.5647208121827412, + "grad_norm": 8.375, + "learning_rate": 4.021784686428597e-06, + "loss": 2.6418, + "step": 14796 + }, + { + "epoch": 1.564826565143824, + "grad_norm": 7.96875, + "learning_rate": 4.021648809913021e-06, + "loss": 2.6192, + "step": 14797 + }, + { + "epoch": 1.564932318104907, + "grad_norm": 8.3125, + "learning_rate": 4.02151292625697e-06, + "loss": 2.6408, + "step": 14798 + }, + { + "epoch": 1.56503807106599, + "grad_norm": 7.96875, + "learning_rate": 4.021377035461082e-06, + "loss": 2.3796, + "step": 14799 + }, + { + "epoch": 1.5651438240270727, + "grad_norm": 8.5, + "learning_rate": 4.021241137525994e-06, + "loss": 2.6208, + "step": 14800 + }, + { + "epoch": 1.5652495769881556, + "grad_norm": 9.0625, + "learning_rate": 4.021105232452343e-06, + "loss": 2.7909, + "step": 14801 + }, + { + "epoch": 1.5653553299492384, + "grad_norm": 9.25, + "learning_rate": 4.020969320240767e-06, + "loss": 2.6378, + "step": 14802 + }, + { + "epoch": 1.5654610829103215, + "grad_norm": 8.8125, + "learning_rate": 4.020833400891905e-06, + "loss": 2.2339, + "step": 14803 + }, + { + "epoch": 1.5655668358714045, + "grad_norm": 8.3125, + "learning_rate": 4.020697474406393e-06, + "loss": 2.4084, + "step": 14804 + }, + { + "epoch": 1.5656725888324874, + "grad_norm": 8.125, + "learning_rate": 4.02056154078487e-06, + "loss": 2.4899, + "step": 14805 + }, + { + "epoch": 1.5657783417935702, + "grad_norm": 8.4375, + "learning_rate": 4.020425600027973e-06, + "loss": 2.7678, + "step": 14806 + }, + { + "epoch": 1.565884094754653, + "grad_norm": 8.375, + "learning_rate": 4.020289652136341e-06, + "loss": 2.7577, + "step": 14807 + }, + { + "epoch": 1.565989847715736, + "grad_norm": 8.25, + "learning_rate": 4.020153697110612e-06, + "loss": 2.5929, + "step": 14808 + }, + { + "epoch": 1.566095600676819, + "grad_norm": 9.0, + "learning_rate": 4.020017734951423e-06, + "loss": 2.2346, + "step": 14809 + }, + { + "epoch": 1.566201353637902, + "grad_norm": 8.125, + "learning_rate": 4.019881765659411e-06, + "loss": 2.6432, + "step": 14810 + }, + { + "epoch": 1.5663071065989849, + "grad_norm": 7.9375, + "learning_rate": 4.019745789235217e-06, + "loss": 2.6417, + "step": 14811 + }, + { + "epoch": 1.5664128595600677, + "grad_norm": 8.3125, + "learning_rate": 4.0196098056794775e-06, + "loss": 2.6748, + "step": 14812 + }, + { + "epoch": 1.5665186125211505, + "grad_norm": 9.25, + "learning_rate": 4.0194738149928305e-06, + "loss": 2.575, + "step": 14813 + }, + { + "epoch": 1.5666243654822334, + "grad_norm": 8.3125, + "learning_rate": 4.0193378171759136e-06, + "loss": 2.5033, + "step": 14814 + }, + { + "epoch": 1.5667301184433164, + "grad_norm": 8.25, + "learning_rate": 4.019201812229366e-06, + "loss": 2.7018, + "step": 14815 + }, + { + "epoch": 1.5668358714043993, + "grad_norm": 8.1875, + "learning_rate": 4.019065800153826e-06, + "loss": 2.5424, + "step": 14816 + }, + { + "epoch": 1.5669416243654823, + "grad_norm": 8.4375, + "learning_rate": 4.018929780949932e-06, + "loss": 2.6018, + "step": 14817 + }, + { + "epoch": 1.5670473773265652, + "grad_norm": 8.3125, + "learning_rate": 4.018793754618321e-06, + "loss": 2.6082, + "step": 14818 + }, + { + "epoch": 1.567153130287648, + "grad_norm": 8.4375, + "learning_rate": 4.018657721159632e-06, + "loss": 2.5273, + "step": 14819 + }, + { + "epoch": 1.5672588832487309, + "grad_norm": 7.75, + "learning_rate": 4.018521680574504e-06, + "loss": 2.5465, + "step": 14820 + }, + { + "epoch": 1.567364636209814, + "grad_norm": 9.5625, + "learning_rate": 4.018385632863575e-06, + "loss": 2.8572, + "step": 14821 + }, + { + "epoch": 1.5674703891708968, + "grad_norm": 8.5625, + "learning_rate": 4.018249578027482e-06, + "loss": 2.5016, + "step": 14822 + }, + { + "epoch": 1.5675761421319798, + "grad_norm": 7.96875, + "learning_rate": 4.018113516066866e-06, + "loss": 2.5393, + "step": 14823 + }, + { + "epoch": 1.5676818950930627, + "grad_norm": 8.5, + "learning_rate": 4.017977446982363e-06, + "loss": 2.6048, + "step": 14824 + }, + { + "epoch": 1.5677876480541455, + "grad_norm": 8.25, + "learning_rate": 4.017841370774613e-06, + "loss": 2.4699, + "step": 14825 + }, + { + "epoch": 1.5678934010152283, + "grad_norm": 8.4375, + "learning_rate": 4.017705287444255e-06, + "loss": 2.4782, + "step": 14826 + }, + { + "epoch": 1.5679991539763112, + "grad_norm": 8.3125, + "learning_rate": 4.017569196991926e-06, + "loss": 2.6696, + "step": 14827 + }, + { + "epoch": 1.5681049069373942, + "grad_norm": 8.25, + "learning_rate": 4.017433099418266e-06, + "loss": 2.7375, + "step": 14828 + }, + { + "epoch": 1.5682106598984773, + "grad_norm": 8.25, + "learning_rate": 4.017296994723913e-06, + "loss": 2.708, + "step": 14829 + }, + { + "epoch": 1.5683164128595601, + "grad_norm": 8.375, + "learning_rate": 4.0171608829095055e-06, + "loss": 2.1811, + "step": 14830 + }, + { + "epoch": 1.568422165820643, + "grad_norm": 8.1875, + "learning_rate": 4.017024763975683e-06, + "loss": 2.5375, + "step": 14831 + }, + { + "epoch": 1.5685279187817258, + "grad_norm": 8.75, + "learning_rate": 4.016888637923083e-06, + "loss": 2.6238, + "step": 14832 + }, + { + "epoch": 1.5686336717428087, + "grad_norm": 8.3125, + "learning_rate": 4.016752504752345e-06, + "loss": 2.7859, + "step": 14833 + }, + { + "epoch": 1.5687394247038917, + "grad_norm": 8.25, + "learning_rate": 4.0166163644641074e-06, + "loss": 2.5148, + "step": 14834 + }, + { + "epoch": 1.5688451776649748, + "grad_norm": 7.96875, + "learning_rate": 4.01648021705901e-06, + "loss": 2.4283, + "step": 14835 + }, + { + "epoch": 1.5689509306260576, + "grad_norm": 8.0, + "learning_rate": 4.016344062537692e-06, + "loss": 2.4342, + "step": 14836 + }, + { + "epoch": 1.5690566835871405, + "grad_norm": 8.375, + "learning_rate": 4.016207900900791e-06, + "loss": 2.4495, + "step": 14837 + }, + { + "epoch": 1.5691624365482233, + "grad_norm": 8.5625, + "learning_rate": 4.0160717321489464e-06, + "loss": 2.2349, + "step": 14838 + }, + { + "epoch": 1.5692681895093061, + "grad_norm": 8.0625, + "learning_rate": 4.015935556282797e-06, + "loss": 2.47, + "step": 14839 + }, + { + "epoch": 1.5693739424703892, + "grad_norm": 8.5, + "learning_rate": 4.015799373302982e-06, + "loss": 2.6991, + "step": 14840 + }, + { + "epoch": 1.569479695431472, + "grad_norm": 7.65625, + "learning_rate": 4.01566318321014e-06, + "loss": 2.397, + "step": 14841 + }, + { + "epoch": 1.569585448392555, + "grad_norm": 8.6875, + "learning_rate": 4.0155269860049104e-06, + "loss": 2.5969, + "step": 14842 + }, + { + "epoch": 1.569691201353638, + "grad_norm": 8.4375, + "learning_rate": 4.015390781687933e-06, + "loss": 2.4993, + "step": 14843 + }, + { + "epoch": 1.5697969543147208, + "grad_norm": 8.625, + "learning_rate": 4.015254570259848e-06, + "loss": 2.6161, + "step": 14844 + }, + { + "epoch": 1.5699027072758036, + "grad_norm": 7.875, + "learning_rate": 4.015118351721291e-06, + "loss": 2.4253, + "step": 14845 + }, + { + "epoch": 1.5700084602368867, + "grad_norm": 8.1875, + "learning_rate": 4.0149821260729035e-06, + "loss": 2.6651, + "step": 14846 + }, + { + "epoch": 1.5701142131979695, + "grad_norm": 8.6875, + "learning_rate": 4.014845893315325e-06, + "loss": 2.58, + "step": 14847 + }, + { + "epoch": 1.5702199661590526, + "grad_norm": 8.0625, + "learning_rate": 4.014709653449194e-06, + "loss": 2.6752, + "step": 14848 + }, + { + "epoch": 1.5703257191201354, + "grad_norm": 8.0, + "learning_rate": 4.014573406475151e-06, + "loss": 2.5253, + "step": 14849 + }, + { + "epoch": 1.5704314720812182, + "grad_norm": 8.3125, + "learning_rate": 4.014437152393834e-06, + "loss": 2.6865, + "step": 14850 + }, + { + "epoch": 1.570537225042301, + "grad_norm": 7.96875, + "learning_rate": 4.014300891205882e-06, + "loss": 2.5105, + "step": 14851 + }, + { + "epoch": 1.5706429780033841, + "grad_norm": 8.75, + "learning_rate": 4.014164622911935e-06, + "loss": 2.7098, + "step": 14852 + }, + { + "epoch": 1.570748730964467, + "grad_norm": 8.75, + "learning_rate": 4.014028347512634e-06, + "loss": 2.6678, + "step": 14853 + }, + { + "epoch": 1.57085448392555, + "grad_norm": 8.5, + "learning_rate": 4.013892065008617e-06, + "loss": 2.5196, + "step": 14854 + }, + { + "epoch": 1.5709602368866329, + "grad_norm": 8.4375, + "learning_rate": 4.013755775400523e-06, + "loss": 2.5458, + "step": 14855 + }, + { + "epoch": 1.5710659898477157, + "grad_norm": 8.4375, + "learning_rate": 4.013619478688992e-06, + "loss": 2.4562, + "step": 14856 + }, + { + "epoch": 1.5711717428087986, + "grad_norm": 7.875, + "learning_rate": 4.013483174874665e-06, + "loss": 2.5716, + "step": 14857 + }, + { + "epoch": 1.5712774957698814, + "grad_norm": 7.84375, + "learning_rate": 4.01334686395818e-06, + "loss": 2.5387, + "step": 14858 + }, + { + "epoch": 1.5713832487309645, + "grad_norm": 8.1875, + "learning_rate": 4.0132105459401765e-06, + "loss": 2.5491, + "step": 14859 + }, + { + "epoch": 1.5714890016920475, + "grad_norm": 8.75, + "learning_rate": 4.013074220821296e-06, + "loss": 2.6824, + "step": 14860 + }, + { + "epoch": 1.5715947546531304, + "grad_norm": 8.1875, + "learning_rate": 4.0129378886021756e-06, + "loss": 2.2199, + "step": 14861 + }, + { + "epoch": 1.5717005076142132, + "grad_norm": 9.5, + "learning_rate": 4.012801549283457e-06, + "loss": 2.6221, + "step": 14862 + }, + { + "epoch": 1.571806260575296, + "grad_norm": 9.3125, + "learning_rate": 4.01266520286578e-06, + "loss": 2.585, + "step": 14863 + }, + { + "epoch": 1.5719120135363789, + "grad_norm": 7.96875, + "learning_rate": 4.012528849349784e-06, + "loss": 2.5137, + "step": 14864 + }, + { + "epoch": 1.572017766497462, + "grad_norm": 7.9375, + "learning_rate": 4.012392488736108e-06, + "loss": 2.4136, + "step": 14865 + }, + { + "epoch": 1.572123519458545, + "grad_norm": 8.25, + "learning_rate": 4.0122561210253924e-06, + "loss": 2.4391, + "step": 14866 + }, + { + "epoch": 1.5722292724196278, + "grad_norm": 8.0, + "learning_rate": 4.012119746218277e-06, + "loss": 2.3216, + "step": 14867 + }, + { + "epoch": 1.5723350253807107, + "grad_norm": 8.6875, + "learning_rate": 4.011983364315403e-06, + "loss": 2.9869, + "step": 14868 + }, + { + "epoch": 1.5724407783417935, + "grad_norm": 8.4375, + "learning_rate": 4.01184697531741e-06, + "loss": 2.6074, + "step": 14869 + }, + { + "epoch": 1.5725465313028764, + "grad_norm": 8.75, + "learning_rate": 4.011710579224936e-06, + "loss": 2.3951, + "step": 14870 + }, + { + "epoch": 1.5726522842639594, + "grad_norm": 8.3125, + "learning_rate": 4.011574176038623e-06, + "loss": 2.514, + "step": 14871 + }, + { + "epoch": 1.5727580372250423, + "grad_norm": 8.3125, + "learning_rate": 4.011437765759111e-06, + "loss": 2.6395, + "step": 14872 + }, + { + "epoch": 1.5728637901861253, + "grad_norm": 8.25, + "learning_rate": 4.011301348387039e-06, + "loss": 2.6335, + "step": 14873 + }, + { + "epoch": 1.5729695431472082, + "grad_norm": 8.625, + "learning_rate": 4.011164923923049e-06, + "loss": 2.3781, + "step": 14874 + }, + { + "epoch": 1.573075296108291, + "grad_norm": 8.5625, + "learning_rate": 4.01102849236778e-06, + "loss": 2.3293, + "step": 14875 + }, + { + "epoch": 1.5731810490693738, + "grad_norm": 8.4375, + "learning_rate": 4.010892053721871e-06, + "loss": 2.4519, + "step": 14876 + }, + { + "epoch": 1.573286802030457, + "grad_norm": 7.875, + "learning_rate": 4.010755607985964e-06, + "loss": 2.4212, + "step": 14877 + }, + { + "epoch": 1.5733925549915397, + "grad_norm": 8.25, + "learning_rate": 4.0106191551606986e-06, + "loss": 2.5244, + "step": 14878 + }, + { + "epoch": 1.5734983079526228, + "grad_norm": 8.8125, + "learning_rate": 4.0104826952467165e-06, + "loss": 2.6725, + "step": 14879 + }, + { + "epoch": 1.5736040609137056, + "grad_norm": 9.25, + "learning_rate": 4.010346228244656e-06, + "loss": 2.6205, + "step": 14880 + }, + { + "epoch": 1.5737098138747885, + "grad_norm": 8.6875, + "learning_rate": 4.010209754155158e-06, + "loss": 2.2573, + "step": 14881 + }, + { + "epoch": 1.5738155668358713, + "grad_norm": 8.0, + "learning_rate": 4.010073272978865e-06, + "loss": 2.5422, + "step": 14882 + }, + { + "epoch": 1.5739213197969542, + "grad_norm": 7.84375, + "learning_rate": 4.009936784716414e-06, + "loss": 2.5965, + "step": 14883 + }, + { + "epoch": 1.5740270727580372, + "grad_norm": 8.4375, + "learning_rate": 4.009800289368447e-06, + "loss": 2.4508, + "step": 14884 + }, + { + "epoch": 1.5741328257191203, + "grad_norm": 8.3125, + "learning_rate": 4.009663786935606e-06, + "loss": 2.6515, + "step": 14885 + }, + { + "epoch": 1.5742385786802031, + "grad_norm": 7.59375, + "learning_rate": 4.009527277418529e-06, + "loss": 2.3988, + "step": 14886 + }, + { + "epoch": 1.574344331641286, + "grad_norm": 8.6875, + "learning_rate": 4.00939076081786e-06, + "loss": 2.5567, + "step": 14887 + }, + { + "epoch": 1.5744500846023688, + "grad_norm": 7.875, + "learning_rate": 4.009254237134235e-06, + "loss": 2.4441, + "step": 14888 + }, + { + "epoch": 1.5745558375634516, + "grad_norm": 9.0625, + "learning_rate": 4.009117706368298e-06, + "loss": 2.5278, + "step": 14889 + }, + { + "epoch": 1.5746615905245347, + "grad_norm": 8.125, + "learning_rate": 4.00898116852069e-06, + "loss": 2.4931, + "step": 14890 + }, + { + "epoch": 1.5747673434856178, + "grad_norm": 8.25, + "learning_rate": 4.008844623592049e-06, + "loss": 2.6548, + "step": 14891 + }, + { + "epoch": 1.5748730964467006, + "grad_norm": 8.3125, + "learning_rate": 4.008708071583018e-06, + "loss": 2.4769, + "step": 14892 + }, + { + "epoch": 1.5749788494077834, + "grad_norm": 8.75, + "learning_rate": 4.008571512494237e-06, + "loss": 2.8012, + "step": 14893 + }, + { + "epoch": 1.5750846023688663, + "grad_norm": 7.75, + "learning_rate": 4.0084349463263465e-06, + "loss": 2.5488, + "step": 14894 + }, + { + "epoch": 1.575190355329949, + "grad_norm": 7.78125, + "learning_rate": 4.008298373079988e-06, + "loss": 2.1382, + "step": 14895 + }, + { + "epoch": 1.5752961082910322, + "grad_norm": 8.5, + "learning_rate": 4.008161792755802e-06, + "loss": 2.7298, + "step": 14896 + }, + { + "epoch": 1.575401861252115, + "grad_norm": 8.6875, + "learning_rate": 4.008025205354429e-06, + "loss": 2.6645, + "step": 14897 + }, + { + "epoch": 1.575507614213198, + "grad_norm": 8.3125, + "learning_rate": 4.0078886108765116e-06, + "loss": 2.4083, + "step": 14898 + }, + { + "epoch": 1.575613367174281, + "grad_norm": 8.3125, + "learning_rate": 4.00775200932269e-06, + "loss": 2.5805, + "step": 14899 + }, + { + "epoch": 1.5757191201353637, + "grad_norm": 8.25, + "learning_rate": 4.007615400693603e-06, + "loss": 2.5913, + "step": 14900 + }, + { + "epoch": 1.5758248730964466, + "grad_norm": 7.84375, + "learning_rate": 4.007478784989895e-06, + "loss": 2.5853, + "step": 14901 + }, + { + "epoch": 1.5759306260575296, + "grad_norm": 9.3125, + "learning_rate": 4.007342162212205e-06, + "loss": 2.2925, + "step": 14902 + }, + { + "epoch": 1.5760363790186125, + "grad_norm": 8.25, + "learning_rate": 4.0072055323611745e-06, + "loss": 2.5806, + "step": 14903 + }, + { + "epoch": 1.5761421319796955, + "grad_norm": 7.78125, + "learning_rate": 4.0070688954374456e-06, + "loss": 2.5845, + "step": 14904 + }, + { + "epoch": 1.5762478849407784, + "grad_norm": 8.75, + "learning_rate": 4.006932251441658e-06, + "loss": 2.6246, + "step": 14905 + }, + { + "epoch": 1.5763536379018612, + "grad_norm": 8.1875, + "learning_rate": 4.006795600374454e-06, + "loss": 2.628, + "step": 14906 + }, + { + "epoch": 1.576459390862944, + "grad_norm": 8.1875, + "learning_rate": 4.006658942236474e-06, + "loss": 2.4473, + "step": 14907 + }, + { + "epoch": 1.5765651438240271, + "grad_norm": 7.96875, + "learning_rate": 4.006522277028361e-06, + "loss": 2.5298, + "step": 14908 + }, + { + "epoch": 1.57667089678511, + "grad_norm": 8.0625, + "learning_rate": 4.006385604750754e-06, + "loss": 2.602, + "step": 14909 + }, + { + "epoch": 1.576776649746193, + "grad_norm": 8.75, + "learning_rate": 4.006248925404297e-06, + "loss": 2.6035, + "step": 14910 + }, + { + "epoch": 1.5768824027072759, + "grad_norm": 8.25, + "learning_rate": 4.006112238989628e-06, + "loss": 2.5279, + "step": 14911 + }, + { + "epoch": 1.5769881556683587, + "grad_norm": 8.3125, + "learning_rate": 4.0059755455073905e-06, + "loss": 2.6008, + "step": 14912 + }, + { + "epoch": 1.5770939086294415, + "grad_norm": 8.125, + "learning_rate": 4.005838844958226e-06, + "loss": 2.6001, + "step": 14913 + }, + { + "epoch": 1.5771996615905244, + "grad_norm": 8.4375, + "learning_rate": 4.0057021373427765e-06, + "loss": 2.6898, + "step": 14914 + }, + { + "epoch": 1.5773054145516074, + "grad_norm": 8.3125, + "learning_rate": 4.005565422661681e-06, + "loss": 2.6196, + "step": 14915 + }, + { + "epoch": 1.5774111675126905, + "grad_norm": 8.125, + "learning_rate": 4.005428700915584e-06, + "loss": 2.4457, + "step": 14916 + }, + { + "epoch": 1.5775169204737733, + "grad_norm": 8.0625, + "learning_rate": 4.005291972105126e-06, + "loss": 2.7205, + "step": 14917 + }, + { + "epoch": 1.5776226734348562, + "grad_norm": 8.625, + "learning_rate": 4.005155236230948e-06, + "loss": 2.6114, + "step": 14918 + }, + { + "epoch": 1.577728426395939, + "grad_norm": 8.0, + "learning_rate": 4.0050184932936916e-06, + "loss": 2.3285, + "step": 14919 + }, + { + "epoch": 1.5778341793570219, + "grad_norm": 8.625, + "learning_rate": 4.004881743294e-06, + "loss": 2.5887, + "step": 14920 + }, + { + "epoch": 1.577939932318105, + "grad_norm": 7.84375, + "learning_rate": 4.0047449862325125e-06, + "loss": 2.4708, + "step": 14921 + }, + { + "epoch": 1.578045685279188, + "grad_norm": 8.375, + "learning_rate": 4.004608222109872e-06, + "loss": 2.5225, + "step": 14922 + }, + { + "epoch": 1.5781514382402708, + "grad_norm": 8.5625, + "learning_rate": 4.004471450926723e-06, + "loss": 2.7045, + "step": 14923 + }, + { + "epoch": 1.5782571912013537, + "grad_norm": 8.125, + "learning_rate": 4.0043346726837025e-06, + "loss": 2.4257, + "step": 14924 + }, + { + "epoch": 1.5783629441624365, + "grad_norm": 7.875, + "learning_rate": 4.0041978873814556e-06, + "loss": 2.6194, + "step": 14925 + }, + { + "epoch": 1.5784686971235193, + "grad_norm": 8.0, + "learning_rate": 4.004061095020623e-06, + "loss": 2.4012, + "step": 14926 + }, + { + "epoch": 1.5785744500846024, + "grad_norm": 7.9375, + "learning_rate": 4.003924295601846e-06, + "loss": 2.4031, + "step": 14927 + }, + { + "epoch": 1.5786802030456852, + "grad_norm": 7.65625, + "learning_rate": 4.003787489125769e-06, + "loss": 2.3963, + "step": 14928 + }, + { + "epoch": 1.5787859560067683, + "grad_norm": 8.375, + "learning_rate": 4.003650675593031e-06, + "loss": 2.6975, + "step": 14929 + }, + { + "epoch": 1.5788917089678511, + "grad_norm": 8.25, + "learning_rate": 4.003513855004275e-06, + "loss": 2.6949, + "step": 14930 + }, + { + "epoch": 1.578997461928934, + "grad_norm": 7.65625, + "learning_rate": 4.003377027360144e-06, + "loss": 2.3845, + "step": 14931 + }, + { + "epoch": 1.5791032148900168, + "grad_norm": 8.3125, + "learning_rate": 4.00324019266128e-06, + "loss": 2.7297, + "step": 14932 + }, + { + "epoch": 1.5792089678510999, + "grad_norm": 8.5625, + "learning_rate": 4.003103350908324e-06, + "loss": 2.6025, + "step": 14933 + }, + { + "epoch": 1.5793147208121827, + "grad_norm": 7.90625, + "learning_rate": 4.0029665021019195e-06, + "loss": 2.2369, + "step": 14934 + }, + { + "epoch": 1.5794204737732658, + "grad_norm": 8.125, + "learning_rate": 4.0028296462427065e-06, + "loss": 2.5988, + "step": 14935 + }, + { + "epoch": 1.5795262267343486, + "grad_norm": 8.1875, + "learning_rate": 4.00269278333133e-06, + "loss": 2.5434, + "step": 14936 + }, + { + "epoch": 1.5796319796954315, + "grad_norm": 8.0625, + "learning_rate": 4.00255591336843e-06, + "loss": 2.4604, + "step": 14937 + }, + { + "epoch": 1.5797377326565143, + "grad_norm": 8.375, + "learning_rate": 4.00241903635465e-06, + "loss": 2.3362, + "step": 14938 + }, + { + "epoch": 1.5798434856175971, + "grad_norm": 8.3125, + "learning_rate": 4.002282152290632e-06, + "loss": 2.3755, + "step": 14939 + }, + { + "epoch": 1.5799492385786802, + "grad_norm": 8.1875, + "learning_rate": 4.002145261177018e-06, + "loss": 2.4552, + "step": 14940 + }, + { + "epoch": 1.5800549915397633, + "grad_norm": 8.25, + "learning_rate": 4.002008363014451e-06, + "loss": 2.6108, + "step": 14941 + }, + { + "epoch": 1.580160744500846, + "grad_norm": 8.0625, + "learning_rate": 4.001871457803572e-06, + "loss": 2.5364, + "step": 14942 + }, + { + "epoch": 1.580266497461929, + "grad_norm": 8.0, + "learning_rate": 4.0017345455450255e-06, + "loss": 2.4608, + "step": 14943 + }, + { + "epoch": 1.5803722504230118, + "grad_norm": 7.75, + "learning_rate": 4.001597626239453e-06, + "loss": 2.5251, + "step": 14944 + }, + { + "epoch": 1.5804780033840946, + "grad_norm": 7.875, + "learning_rate": 4.001460699887497e-06, + "loss": 2.5127, + "step": 14945 + }, + { + "epoch": 1.5805837563451777, + "grad_norm": 8.1875, + "learning_rate": 4.0013237664897995e-06, + "loss": 2.305, + "step": 14946 + }, + { + "epoch": 1.5806895093062607, + "grad_norm": 8.375, + "learning_rate": 4.001186826047005e-06, + "loss": 2.4144, + "step": 14947 + }, + { + "epoch": 1.5807952622673436, + "grad_norm": 9.125, + "learning_rate": 4.001049878559752e-06, + "loss": 2.6823, + "step": 14948 + }, + { + "epoch": 1.5809010152284264, + "grad_norm": 8.5, + "learning_rate": 4.000912924028687e-06, + "loss": 2.6711, + "step": 14949 + }, + { + "epoch": 1.5810067681895092, + "grad_norm": 7.78125, + "learning_rate": 4.000775962454452e-06, + "loss": 2.4805, + "step": 14950 + }, + { + "epoch": 1.581112521150592, + "grad_norm": 8.375, + "learning_rate": 4.000638993837689e-06, + "loss": 2.5563, + "step": 14951 + }, + { + "epoch": 1.5812182741116751, + "grad_norm": 8.25, + "learning_rate": 4.000502018179042e-06, + "loss": 2.6861, + "step": 14952 + }, + { + "epoch": 1.581324027072758, + "grad_norm": 8.6875, + "learning_rate": 4.0003650354791515e-06, + "loss": 2.7472, + "step": 14953 + }, + { + "epoch": 1.581429780033841, + "grad_norm": 8.4375, + "learning_rate": 4.000228045738661e-06, + "loss": 2.627, + "step": 14954 + }, + { + "epoch": 1.5815355329949239, + "grad_norm": 7.90625, + "learning_rate": 4.000091048958215e-06, + "loss": 2.5357, + "step": 14955 + }, + { + "epoch": 1.5816412859560067, + "grad_norm": 8.0, + "learning_rate": 3.999954045138455e-06, + "loss": 2.6122, + "step": 14956 + }, + { + "epoch": 1.5817470389170896, + "grad_norm": 8.375, + "learning_rate": 3.999817034280023e-06, + "loss": 2.6135, + "step": 14957 + }, + { + "epoch": 1.5818527918781726, + "grad_norm": 7.75, + "learning_rate": 3.999680016383565e-06, + "loss": 2.4132, + "step": 14958 + }, + { + "epoch": 1.5819585448392555, + "grad_norm": 8.1875, + "learning_rate": 3.9995429914497205e-06, + "loss": 2.5317, + "step": 14959 + }, + { + "epoch": 1.5820642978003385, + "grad_norm": 8.9375, + "learning_rate": 3.999405959479135e-06, + "loss": 2.4195, + "step": 14960 + }, + { + "epoch": 1.5821700507614214, + "grad_norm": 8.3125, + "learning_rate": 3.99926892047245e-06, + "loss": 2.4545, + "step": 14961 + }, + { + "epoch": 1.5822758037225042, + "grad_norm": 9.25, + "learning_rate": 3.999131874430309e-06, + "loss": 2.5952, + "step": 14962 + }, + { + "epoch": 1.582381556683587, + "grad_norm": 8.6875, + "learning_rate": 3.998994821353356e-06, + "loss": 2.6647, + "step": 14963 + }, + { + "epoch": 1.58248730964467, + "grad_norm": 8.375, + "learning_rate": 3.998857761242232e-06, + "loss": 2.4788, + "step": 14964 + }, + { + "epoch": 1.582593062605753, + "grad_norm": 8.9375, + "learning_rate": 3.9987206940975835e-06, + "loss": 2.9407, + "step": 14965 + }, + { + "epoch": 1.582698815566836, + "grad_norm": 8.0625, + "learning_rate": 3.99858361992005e-06, + "loss": 2.6016, + "step": 14966 + }, + { + "epoch": 1.5828045685279188, + "grad_norm": 8.3125, + "learning_rate": 3.998446538710277e-06, + "loss": 2.3118, + "step": 14967 + }, + { + "epoch": 1.5829103214890017, + "grad_norm": 8.375, + "learning_rate": 3.998309450468908e-06, + "loss": 2.7365, + "step": 14968 + }, + { + "epoch": 1.5830160744500845, + "grad_norm": 9.875, + "learning_rate": 3.998172355196584e-06, + "loss": 2.5771, + "step": 14969 + }, + { + "epoch": 1.5831218274111674, + "grad_norm": 8.3125, + "learning_rate": 3.998035252893951e-06, + "loss": 2.7201, + "step": 14970 + }, + { + "epoch": 1.5832275803722504, + "grad_norm": 10.9375, + "learning_rate": 3.9978981435616515e-06, + "loss": 2.5187, + "step": 14971 + }, + { + "epoch": 1.5833333333333335, + "grad_norm": 8.375, + "learning_rate": 3.997761027200328e-06, + "loss": 2.3901, + "step": 14972 + }, + { + "epoch": 1.5834390862944163, + "grad_norm": 8.3125, + "learning_rate": 3.9976239038106246e-06, + "loss": 2.7167, + "step": 14973 + }, + { + "epoch": 1.5835448392554992, + "grad_norm": 8.125, + "learning_rate": 3.997486773393185e-06, + "loss": 2.5199, + "step": 14974 + }, + { + "epoch": 1.583650592216582, + "grad_norm": 8.625, + "learning_rate": 3.997349635948652e-06, + "loss": 2.8168, + "step": 14975 + }, + { + "epoch": 1.5837563451776648, + "grad_norm": 7.84375, + "learning_rate": 3.997212491477669e-06, + "loss": 2.5203, + "step": 14976 + }, + { + "epoch": 1.583862098138748, + "grad_norm": 8.125, + "learning_rate": 3.997075339980882e-06, + "loss": 2.5679, + "step": 14977 + }, + { + "epoch": 1.583967851099831, + "grad_norm": 9.0, + "learning_rate": 3.996938181458931e-06, + "loss": 2.8308, + "step": 14978 + }, + { + "epoch": 1.5840736040609138, + "grad_norm": 8.1875, + "learning_rate": 3.996801015912462e-06, + "loss": 2.5169, + "step": 14979 + }, + { + "epoch": 1.5841793570219966, + "grad_norm": 8.375, + "learning_rate": 3.996663843342118e-06, + "loss": 2.4431, + "step": 14980 + }, + { + "epoch": 1.5842851099830795, + "grad_norm": 7.78125, + "learning_rate": 3.996526663748542e-06, + "loss": 2.4725, + "step": 14981 + }, + { + "epoch": 1.5843908629441623, + "grad_norm": 7.96875, + "learning_rate": 3.99638947713238e-06, + "loss": 2.542, + "step": 14982 + }, + { + "epoch": 1.5844966159052454, + "grad_norm": 8.1875, + "learning_rate": 3.996252283494273e-06, + "loss": 2.7754, + "step": 14983 + }, + { + "epoch": 1.5846023688663282, + "grad_norm": 7.8125, + "learning_rate": 3.996115082834866e-06, + "loss": 2.468, + "step": 14984 + }, + { + "epoch": 1.5847081218274113, + "grad_norm": 7.90625, + "learning_rate": 3.995977875154803e-06, + "loss": 2.52, + "step": 14985 + }, + { + "epoch": 1.5848138747884941, + "grad_norm": 8.5, + "learning_rate": 3.9958406604547276e-06, + "loss": 2.2474, + "step": 14986 + }, + { + "epoch": 1.584919627749577, + "grad_norm": 8.0625, + "learning_rate": 3.995703438735284e-06, + "loss": 2.5701, + "step": 14987 + }, + { + "epoch": 1.5850253807106598, + "grad_norm": 7.9375, + "learning_rate": 3.995566209997116e-06, + "loss": 2.5911, + "step": 14988 + }, + { + "epoch": 1.5851311336717429, + "grad_norm": 8.3125, + "learning_rate": 3.995428974240867e-06, + "loss": 2.5775, + "step": 14989 + }, + { + "epoch": 1.5852368866328257, + "grad_norm": 8.125, + "learning_rate": 3.995291731467182e-06, + "loss": 2.4156, + "step": 14990 + }, + { + "epoch": 1.5853426395939088, + "grad_norm": 8.625, + "learning_rate": 3.995154481676704e-06, + "loss": 2.3066, + "step": 14991 + }, + { + "epoch": 1.5854483925549916, + "grad_norm": 8.125, + "learning_rate": 3.995017224870077e-06, + "loss": 2.4025, + "step": 14992 + }, + { + "epoch": 1.5855541455160744, + "grad_norm": 8.25, + "learning_rate": 3.994879961047946e-06, + "loss": 2.6099, + "step": 14993 + }, + { + "epoch": 1.5856598984771573, + "grad_norm": 10.5625, + "learning_rate": 3.994742690210956e-06, + "loss": 2.7382, + "step": 14994 + }, + { + "epoch": 1.58576565143824, + "grad_norm": 8.0625, + "learning_rate": 3.994605412359748e-06, + "loss": 2.4129, + "step": 14995 + }, + { + "epoch": 1.5858714043993232, + "grad_norm": 8.6875, + "learning_rate": 3.994468127494969e-06, + "loss": 2.594, + "step": 14996 + }, + { + "epoch": 1.5859771573604062, + "grad_norm": 8.75, + "learning_rate": 3.994330835617263e-06, + "loss": 2.5345, + "step": 14997 + }, + { + "epoch": 1.586082910321489, + "grad_norm": 8.4375, + "learning_rate": 3.994193536727272e-06, + "loss": 2.708, + "step": 14998 + }, + { + "epoch": 1.586188663282572, + "grad_norm": 8.3125, + "learning_rate": 3.994056230825642e-06, + "loss": 2.4738, + "step": 14999 + }, + { + "epoch": 1.5862944162436547, + "grad_norm": 8.5625, + "learning_rate": 3.993918917913019e-06, + "loss": 2.4718, + "step": 15000 + }, + { + "epoch": 1.5864001692047376, + "grad_norm": 8.0625, + "learning_rate": 3.993781597990043e-06, + "loss": 2.631, + "step": 15001 + }, + { + "epoch": 1.5865059221658206, + "grad_norm": 8.1875, + "learning_rate": 3.993644271057363e-06, + "loss": 2.3816, + "step": 15002 + }, + { + "epoch": 1.5866116751269037, + "grad_norm": 8.375, + "learning_rate": 3.99350693711562e-06, + "loss": 2.6151, + "step": 15003 + }, + { + "epoch": 1.5867174280879865, + "grad_norm": 9.125, + "learning_rate": 3.99336959616546e-06, + "loss": 2.4684, + "step": 15004 + }, + { + "epoch": 1.5868231810490694, + "grad_norm": 8.25, + "learning_rate": 3.993232248207528e-06, + "loss": 2.4916, + "step": 15005 + }, + { + "epoch": 1.5869289340101522, + "grad_norm": 8.375, + "learning_rate": 3.993094893242467e-06, + "loss": 2.5007, + "step": 15006 + }, + { + "epoch": 1.587034686971235, + "grad_norm": 8.8125, + "learning_rate": 3.9929575312709226e-06, + "loss": 2.5585, + "step": 15007 + }, + { + "epoch": 1.5871404399323181, + "grad_norm": 9.0625, + "learning_rate": 3.9928201622935395e-06, + "loss": 2.562, + "step": 15008 + }, + { + "epoch": 1.587246192893401, + "grad_norm": 8.375, + "learning_rate": 3.992682786310962e-06, + "loss": 2.549, + "step": 15009 + }, + { + "epoch": 1.587351945854484, + "grad_norm": 8.8125, + "learning_rate": 3.992545403323833e-06, + "loss": 2.5689, + "step": 15010 + }, + { + "epoch": 1.5874576988155669, + "grad_norm": 8.4375, + "learning_rate": 3.9924080133328e-06, + "loss": 2.5824, + "step": 15011 + }, + { + "epoch": 1.5875634517766497, + "grad_norm": 8.75, + "learning_rate": 3.992270616338507e-06, + "loss": 2.2315, + "step": 15012 + }, + { + "epoch": 1.5876692047377325, + "grad_norm": 8.3125, + "learning_rate": 3.992133212341598e-06, + "loss": 2.5161, + "step": 15013 + }, + { + "epoch": 1.5877749576988156, + "grad_norm": 8.125, + "learning_rate": 3.9919958013427176e-06, + "loss": 2.4405, + "step": 15014 + }, + { + "epoch": 1.5878807106598984, + "grad_norm": 8.1875, + "learning_rate": 3.9918583833425115e-06, + "loss": 2.353, + "step": 15015 + }, + { + "epoch": 1.5879864636209815, + "grad_norm": 8.5625, + "learning_rate": 3.991720958341625e-06, + "loss": 2.4644, + "step": 15016 + }, + { + "epoch": 1.5880922165820643, + "grad_norm": 8.25, + "learning_rate": 3.991583526340701e-06, + "loss": 2.5095, + "step": 15017 + }, + { + "epoch": 1.5881979695431472, + "grad_norm": 8.25, + "learning_rate": 3.991446087340386e-06, + "loss": 2.4094, + "step": 15018 + }, + { + "epoch": 1.58830372250423, + "grad_norm": 8.0625, + "learning_rate": 3.9913086413413235e-06, + "loss": 2.5697, + "step": 15019 + }, + { + "epoch": 1.588409475465313, + "grad_norm": 8.4375, + "learning_rate": 3.991171188344161e-06, + "loss": 2.5156, + "step": 15020 + }, + { + "epoch": 1.588515228426396, + "grad_norm": 8.1875, + "learning_rate": 3.991033728349541e-06, + "loss": 2.593, + "step": 15021 + }, + { + "epoch": 1.588620981387479, + "grad_norm": 9.625, + "learning_rate": 3.9908962613581095e-06, + "loss": 2.7769, + "step": 15022 + }, + { + "epoch": 1.5887267343485618, + "grad_norm": 7.8125, + "learning_rate": 3.990758787370512e-06, + "loss": 2.5749, + "step": 15023 + }, + { + "epoch": 1.5888324873096447, + "grad_norm": 7.59375, + "learning_rate": 3.9906213063873934e-06, + "loss": 2.483, + "step": 15024 + }, + { + "epoch": 1.5889382402707275, + "grad_norm": 7.78125, + "learning_rate": 3.990483818409399e-06, + "loss": 2.5119, + "step": 15025 + }, + { + "epoch": 1.5890439932318103, + "grad_norm": 7.9375, + "learning_rate": 3.9903463234371725e-06, + "loss": 2.6798, + "step": 15026 + }, + { + "epoch": 1.5891497461928934, + "grad_norm": 8.0625, + "learning_rate": 3.990208821471361e-06, + "loss": 2.4121, + "step": 15027 + }, + { + "epoch": 1.5892554991539765, + "grad_norm": 8.625, + "learning_rate": 3.990071312512608e-06, + "loss": 2.52, + "step": 15028 + }, + { + "epoch": 1.5893612521150593, + "grad_norm": 8.0625, + "learning_rate": 3.989933796561561e-06, + "loss": 2.5447, + "step": 15029 + }, + { + "epoch": 1.5894670050761421, + "grad_norm": 8.6875, + "learning_rate": 3.989796273618863e-06, + "loss": 2.6799, + "step": 15030 + }, + { + "epoch": 1.589572758037225, + "grad_norm": 8.0, + "learning_rate": 3.989658743685161e-06, + "loss": 2.5606, + "step": 15031 + }, + { + "epoch": 1.5896785109983078, + "grad_norm": 7.875, + "learning_rate": 3.9895212067611e-06, + "loss": 2.348, + "step": 15032 + }, + { + "epoch": 1.5897842639593909, + "grad_norm": 8.0, + "learning_rate": 3.989383662847325e-06, + "loss": 2.6699, + "step": 15033 + }, + { + "epoch": 1.589890016920474, + "grad_norm": 9.0625, + "learning_rate": 3.9892461119444815e-06, + "loss": 2.3653, + "step": 15034 + }, + { + "epoch": 1.5899957698815568, + "grad_norm": 8.625, + "learning_rate": 3.989108554053215e-06, + "loss": 2.8997, + "step": 15035 + }, + { + "epoch": 1.5901015228426396, + "grad_norm": 7.96875, + "learning_rate": 3.988970989174171e-06, + "loss": 2.2453, + "step": 15036 + }, + { + "epoch": 1.5902072758037225, + "grad_norm": 8.1875, + "learning_rate": 3.988833417307995e-06, + "loss": 2.4274, + "step": 15037 + }, + { + "epoch": 1.5903130287648053, + "grad_norm": 8.25, + "learning_rate": 3.9886958384553335e-06, + "loss": 2.7764, + "step": 15038 + }, + { + "epoch": 1.5904187817258884, + "grad_norm": 8.9375, + "learning_rate": 3.98855825261683e-06, + "loss": 2.6493, + "step": 15039 + }, + { + "epoch": 1.5905245346869712, + "grad_norm": 8.375, + "learning_rate": 3.9884206597931315e-06, + "loss": 2.4771, + "step": 15040 + }, + { + "epoch": 1.5906302876480543, + "grad_norm": 8.125, + "learning_rate": 3.988283059984885e-06, + "loss": 2.3892, + "step": 15041 + }, + { + "epoch": 1.590736040609137, + "grad_norm": 8.5625, + "learning_rate": 3.988145453192734e-06, + "loss": 2.5307, + "step": 15042 + }, + { + "epoch": 1.59084179357022, + "grad_norm": 8.5, + "learning_rate": 3.988007839417325e-06, + "loss": 2.4997, + "step": 15043 + }, + { + "epoch": 1.5909475465313028, + "grad_norm": 8.25, + "learning_rate": 3.987870218659305e-06, + "loss": 2.6266, + "step": 15044 + }, + { + "epoch": 1.5910532994923858, + "grad_norm": 8.375, + "learning_rate": 3.987732590919316e-06, + "loss": 2.5048, + "step": 15045 + }, + { + "epoch": 1.5911590524534687, + "grad_norm": 9.1875, + "learning_rate": 3.987594956198008e-06, + "loss": 2.4888, + "step": 15046 + }, + { + "epoch": 1.5912648054145517, + "grad_norm": 9.0, + "learning_rate": 3.9874573144960245e-06, + "loss": 2.2569, + "step": 15047 + }, + { + "epoch": 1.5913705583756346, + "grad_norm": 8.125, + "learning_rate": 3.9873196658140135e-06, + "loss": 2.4183, + "step": 15048 + }, + { + "epoch": 1.5914763113367174, + "grad_norm": 7.96875, + "learning_rate": 3.987182010152619e-06, + "loss": 2.4802, + "step": 15049 + }, + { + "epoch": 1.5915820642978002, + "grad_norm": 9.1875, + "learning_rate": 3.987044347512487e-06, + "loss": 2.6389, + "step": 15050 + }, + { + "epoch": 1.591687817258883, + "grad_norm": 8.3125, + "learning_rate": 3.986906677894265e-06, + "loss": 2.4635, + "step": 15051 + }, + { + "epoch": 1.5917935702199661, + "grad_norm": 8.125, + "learning_rate": 3.986769001298597e-06, + "loss": 2.4957, + "step": 15052 + }, + { + "epoch": 1.5918993231810492, + "grad_norm": 8.625, + "learning_rate": 3.986631317726131e-06, + "loss": 2.3754, + "step": 15053 + }, + { + "epoch": 1.592005076142132, + "grad_norm": 8.0625, + "learning_rate": 3.9864936271775116e-06, + "loss": 2.54, + "step": 15054 + }, + { + "epoch": 1.5921108291032149, + "grad_norm": 9.0625, + "learning_rate": 3.986355929653386e-06, + "loss": 2.5545, + "step": 15055 + }, + { + "epoch": 1.5922165820642977, + "grad_norm": 8.1875, + "learning_rate": 3.986218225154399e-06, + "loss": 2.507, + "step": 15056 + }, + { + "epoch": 1.5923223350253806, + "grad_norm": 8.25, + "learning_rate": 3.986080513681199e-06, + "loss": 2.5634, + "step": 15057 + }, + { + "epoch": 1.5924280879864636, + "grad_norm": 7.9375, + "learning_rate": 3.985942795234431e-06, + "loss": 2.4664, + "step": 15058 + }, + { + "epoch": 1.5925338409475467, + "grad_norm": 8.3125, + "learning_rate": 3.98580506981474e-06, + "loss": 2.4194, + "step": 15059 + }, + { + "epoch": 1.5926395939086295, + "grad_norm": 8.3125, + "learning_rate": 3.985667337422775e-06, + "loss": 2.594, + "step": 15060 + }, + { + "epoch": 1.5927453468697124, + "grad_norm": 8.5, + "learning_rate": 3.985529598059179e-06, + "loss": 2.5144, + "step": 15061 + }, + { + "epoch": 1.5928510998307952, + "grad_norm": 8.0625, + "learning_rate": 3.985391851724601e-06, + "loss": 2.547, + "step": 15062 + }, + { + "epoch": 1.592956852791878, + "grad_norm": 8.6875, + "learning_rate": 3.985254098419687e-06, + "loss": 2.5154, + "step": 15063 + }, + { + "epoch": 1.593062605752961, + "grad_norm": 7.8125, + "learning_rate": 3.985116338145082e-06, + "loss": 2.5675, + "step": 15064 + }, + { + "epoch": 1.593168358714044, + "grad_norm": 8.4375, + "learning_rate": 3.9849785709014344e-06, + "loss": 2.5903, + "step": 15065 + }, + { + "epoch": 1.593274111675127, + "grad_norm": 8.625, + "learning_rate": 3.98484079668939e-06, + "loss": 2.5519, + "step": 15066 + }, + { + "epoch": 1.5933798646362098, + "grad_norm": 8.0625, + "learning_rate": 3.984703015509594e-06, + "loss": 2.5708, + "step": 15067 + }, + { + "epoch": 1.5934856175972927, + "grad_norm": 8.4375, + "learning_rate": 3.9845652273626945e-06, + "loss": 2.5865, + "step": 15068 + }, + { + "epoch": 1.5935913705583755, + "grad_norm": 8.3125, + "learning_rate": 3.984427432249337e-06, + "loss": 2.5182, + "step": 15069 + }, + { + "epoch": 1.5936971235194586, + "grad_norm": 8.375, + "learning_rate": 3.98428963017017e-06, + "loss": 2.4295, + "step": 15070 + }, + { + "epoch": 1.5938028764805414, + "grad_norm": 8.5, + "learning_rate": 3.984151821125837e-06, + "loss": 2.6926, + "step": 15071 + }, + { + "epoch": 1.5939086294416245, + "grad_norm": 8.625, + "learning_rate": 3.984014005116987e-06, + "loss": 2.6833, + "step": 15072 + }, + { + "epoch": 1.5940143824027073, + "grad_norm": 8.9375, + "learning_rate": 3.983876182144267e-06, + "loss": 2.583, + "step": 15073 + }, + { + "epoch": 1.5941201353637902, + "grad_norm": 7.84375, + "learning_rate": 3.9837383522083225e-06, + "loss": 2.396, + "step": 15074 + }, + { + "epoch": 1.594225888324873, + "grad_norm": 8.0625, + "learning_rate": 3.9836005153098e-06, + "loss": 2.4828, + "step": 15075 + }, + { + "epoch": 1.594331641285956, + "grad_norm": 8.75, + "learning_rate": 3.983462671449348e-06, + "loss": 2.4422, + "step": 15076 + }, + { + "epoch": 1.594437394247039, + "grad_norm": 7.90625, + "learning_rate": 3.983324820627613e-06, + "loss": 2.3917, + "step": 15077 + }, + { + "epoch": 1.594543147208122, + "grad_norm": 8.0, + "learning_rate": 3.98318696284524e-06, + "loss": 2.453, + "step": 15078 + }, + { + "epoch": 1.5946489001692048, + "grad_norm": 8.25, + "learning_rate": 3.983049098102878e-06, + "loss": 2.6124, + "step": 15079 + }, + { + "epoch": 1.5947546531302876, + "grad_norm": 8.125, + "learning_rate": 3.982911226401173e-06, + "loss": 2.718, + "step": 15080 + }, + { + "epoch": 1.5948604060913705, + "grad_norm": 9.25, + "learning_rate": 3.982773347740771e-06, + "loss": 2.8325, + "step": 15081 + }, + { + "epoch": 1.5949661590524533, + "grad_norm": 8.6875, + "learning_rate": 3.982635462122322e-06, + "loss": 2.6565, + "step": 15082 + }, + { + "epoch": 1.5950719120135364, + "grad_norm": 8.1875, + "learning_rate": 3.98249756954647e-06, + "loss": 2.4741, + "step": 15083 + }, + { + "epoch": 1.5951776649746194, + "grad_norm": 8.125, + "learning_rate": 3.982359670013863e-06, + "loss": 2.5247, + "step": 15084 + }, + { + "epoch": 1.5952834179357023, + "grad_norm": 9.5, + "learning_rate": 3.9822217635251494e-06, + "loss": 2.5949, + "step": 15085 + }, + { + "epoch": 1.5953891708967851, + "grad_norm": 8.1875, + "learning_rate": 3.982083850080975e-06, + "loss": 2.5806, + "step": 15086 + }, + { + "epoch": 1.595494923857868, + "grad_norm": 8.1875, + "learning_rate": 3.981945929681986e-06, + "loss": 2.4569, + "step": 15087 + }, + { + "epoch": 1.5956006768189508, + "grad_norm": 7.875, + "learning_rate": 3.981808002328832e-06, + "loss": 2.55, + "step": 15088 + }, + { + "epoch": 1.5957064297800339, + "grad_norm": 8.0, + "learning_rate": 3.981670068022159e-06, + "loss": 2.6202, + "step": 15089 + }, + { + "epoch": 1.595812182741117, + "grad_norm": 8.4375, + "learning_rate": 3.981532126762614e-06, + "loss": 2.5269, + "step": 15090 + }, + { + "epoch": 1.5959179357021998, + "grad_norm": 8.375, + "learning_rate": 3.9813941785508445e-06, + "loss": 2.6596, + "step": 15091 + }, + { + "epoch": 1.5960236886632826, + "grad_norm": 8.25, + "learning_rate": 3.9812562233874985e-06, + "loss": 2.6405, + "step": 15092 + }, + { + "epoch": 1.5961294416243654, + "grad_norm": 8.125, + "learning_rate": 3.981118261273222e-06, + "loss": 2.4828, + "step": 15093 + }, + { + "epoch": 1.5962351945854483, + "grad_norm": 8.375, + "learning_rate": 3.980980292208665e-06, + "loss": 2.3882, + "step": 15094 + }, + { + "epoch": 1.5963409475465313, + "grad_norm": 8.25, + "learning_rate": 3.980842316194472e-06, + "loss": 2.621, + "step": 15095 + }, + { + "epoch": 1.5964467005076142, + "grad_norm": 8.8125, + "learning_rate": 3.980704333231291e-06, + "loss": 2.4376, + "step": 15096 + }, + { + "epoch": 1.5965524534686972, + "grad_norm": 8.25, + "learning_rate": 3.980566343319771e-06, + "loss": 2.6266, + "step": 15097 + }, + { + "epoch": 1.59665820642978, + "grad_norm": 8.1875, + "learning_rate": 3.980428346460558e-06, + "loss": 2.6902, + "step": 15098 + }, + { + "epoch": 1.596763959390863, + "grad_norm": 9.375, + "learning_rate": 3.980290342654301e-06, + "loss": 2.5946, + "step": 15099 + }, + { + "epoch": 1.5968697123519457, + "grad_norm": 8.0, + "learning_rate": 3.9801523319016475e-06, + "loss": 2.447, + "step": 15100 + }, + { + "epoch": 1.5969754653130288, + "grad_norm": 8.8125, + "learning_rate": 3.980014314203243e-06, + "loss": 2.7258, + "step": 15101 + }, + { + "epoch": 1.5970812182741116, + "grad_norm": 8.375, + "learning_rate": 3.979876289559738e-06, + "loss": 2.6379, + "step": 15102 + }, + { + "epoch": 1.5971869712351947, + "grad_norm": 8.25, + "learning_rate": 3.979738257971778e-06, + "loss": 2.6435, + "step": 15103 + }, + { + "epoch": 1.5972927241962775, + "grad_norm": 8.6875, + "learning_rate": 3.979600219440012e-06, + "loss": 2.5086, + "step": 15104 + }, + { + "epoch": 1.5973984771573604, + "grad_norm": 8.4375, + "learning_rate": 3.979462173965086e-06, + "loss": 2.6552, + "step": 15105 + }, + { + "epoch": 1.5975042301184432, + "grad_norm": 8.375, + "learning_rate": 3.979324121547651e-06, + "loss": 2.567, + "step": 15106 + }, + { + "epoch": 1.597609983079526, + "grad_norm": 8.4375, + "learning_rate": 3.979186062188351e-06, + "loss": 2.5703, + "step": 15107 + }, + { + "epoch": 1.5977157360406091, + "grad_norm": 8.375, + "learning_rate": 3.979047995887837e-06, + "loss": 2.4562, + "step": 15108 + }, + { + "epoch": 1.5978214890016922, + "grad_norm": 8.1875, + "learning_rate": 3.978909922646756e-06, + "loss": 2.5268, + "step": 15109 + }, + { + "epoch": 1.597927241962775, + "grad_norm": 7.90625, + "learning_rate": 3.978771842465754e-06, + "loss": 2.5173, + "step": 15110 + }, + { + "epoch": 1.5980329949238579, + "grad_norm": 8.1875, + "learning_rate": 3.978633755345482e-06, + "loss": 2.4452, + "step": 15111 + }, + { + "epoch": 1.5981387478849407, + "grad_norm": 7.8125, + "learning_rate": 3.9784956612865855e-06, + "loss": 2.4627, + "step": 15112 + }, + { + "epoch": 1.5982445008460235, + "grad_norm": 8.125, + "learning_rate": 3.978357560289714e-06, + "loss": 2.629, + "step": 15113 + }, + { + "epoch": 1.5983502538071066, + "grad_norm": 8.4375, + "learning_rate": 3.978219452355516e-06, + "loss": 2.5596, + "step": 15114 + }, + { + "epoch": 1.5984560067681897, + "grad_norm": 8.0, + "learning_rate": 3.978081337484637e-06, + "loss": 2.5312, + "step": 15115 + }, + { + "epoch": 1.5985617597292725, + "grad_norm": 8.6875, + "learning_rate": 3.977943215677728e-06, + "loss": 2.4881, + "step": 15116 + }, + { + "epoch": 1.5986675126903553, + "grad_norm": 8.0625, + "learning_rate": 3.9778050869354355e-06, + "loss": 2.4465, + "step": 15117 + }, + { + "epoch": 1.5987732656514382, + "grad_norm": 8.25, + "learning_rate": 3.977666951258408e-06, + "loss": 2.6092, + "step": 15118 + }, + { + "epoch": 1.598879018612521, + "grad_norm": 8.3125, + "learning_rate": 3.977528808647295e-06, + "loss": 2.6217, + "step": 15119 + }, + { + "epoch": 1.598984771573604, + "grad_norm": 8.4375, + "learning_rate": 3.977390659102742e-06, + "loss": 2.504, + "step": 15120 + }, + { + "epoch": 1.599090524534687, + "grad_norm": 7.75, + "learning_rate": 3.977252502625399e-06, + "loss": 2.6782, + "step": 15121 + }, + { + "epoch": 1.59919627749577, + "grad_norm": 8.9375, + "learning_rate": 3.9771143392159155e-06, + "loss": 2.4937, + "step": 15122 + }, + { + "epoch": 1.5993020304568528, + "grad_norm": 8.4375, + "learning_rate": 3.976976168874938e-06, + "loss": 2.4118, + "step": 15123 + }, + { + "epoch": 1.5994077834179357, + "grad_norm": 8.1875, + "learning_rate": 3.9768379916031155e-06, + "loss": 2.5934, + "step": 15124 + }, + { + "epoch": 1.5995135363790185, + "grad_norm": 9.25, + "learning_rate": 3.976699807401096e-06, + "loss": 2.5394, + "step": 15125 + }, + { + "epoch": 1.5996192893401016, + "grad_norm": 8.6875, + "learning_rate": 3.9765616162695274e-06, + "loss": 2.6851, + "step": 15126 + }, + { + "epoch": 1.5997250423011844, + "grad_norm": 8.5, + "learning_rate": 3.97642341820906e-06, + "loss": 2.644, + "step": 15127 + }, + { + "epoch": 1.5998307952622675, + "grad_norm": 8.0, + "learning_rate": 3.976285213220341e-06, + "loss": 2.4594, + "step": 15128 + }, + { + "epoch": 1.5999365482233503, + "grad_norm": 8.375, + "learning_rate": 3.976147001304019e-06, + "loss": 2.7214, + "step": 15129 + }, + { + "epoch": 1.6000423011844331, + "grad_norm": 8.5625, + "learning_rate": 3.976008782460744e-06, + "loss": 2.4847, + "step": 15130 + }, + { + "epoch": 1.600148054145516, + "grad_norm": 8.375, + "learning_rate": 3.975870556691163e-06, + "loss": 2.5022, + "step": 15131 + }, + { + "epoch": 1.600253807106599, + "grad_norm": 8.125, + "learning_rate": 3.975732323995925e-06, + "loss": 2.572, + "step": 15132 + }, + { + "epoch": 1.6003595600676819, + "grad_norm": 8.1875, + "learning_rate": 3.975594084375678e-06, + "loss": 2.7049, + "step": 15133 + }, + { + "epoch": 1.600465313028765, + "grad_norm": 8.8125, + "learning_rate": 3.975455837831072e-06, + "loss": 2.5529, + "step": 15134 + }, + { + "epoch": 1.6005710659898478, + "grad_norm": 8.3125, + "learning_rate": 3.9753175843627555e-06, + "loss": 2.7275, + "step": 15135 + }, + { + "epoch": 1.6006768189509306, + "grad_norm": 8.25, + "learning_rate": 3.975179323971377e-06, + "loss": 2.426, + "step": 15136 + }, + { + "epoch": 1.6007825719120135, + "grad_norm": 7.8125, + "learning_rate": 3.975041056657585e-06, + "loss": 2.4491, + "step": 15137 + }, + { + "epoch": 1.6008883248730963, + "grad_norm": 8.6875, + "learning_rate": 3.974902782422028e-06, + "loss": 2.6813, + "step": 15138 + }, + { + "epoch": 1.6009940778341794, + "grad_norm": 8.25, + "learning_rate": 3.974764501265356e-06, + "loss": 2.6067, + "step": 15139 + }, + { + "epoch": 1.6010998307952624, + "grad_norm": 8.0, + "learning_rate": 3.974626213188218e-06, + "loss": 2.5088, + "step": 15140 + }, + { + "epoch": 1.6012055837563453, + "grad_norm": 7.90625, + "learning_rate": 3.974487918191262e-06, + "loss": 2.5176, + "step": 15141 + }, + { + "epoch": 1.601311336717428, + "grad_norm": 7.96875, + "learning_rate": 3.974349616275136e-06, + "loss": 2.5085, + "step": 15142 + }, + { + "epoch": 1.601417089678511, + "grad_norm": 8.4375, + "learning_rate": 3.974211307440491e-06, + "loss": 2.5785, + "step": 15143 + }, + { + "epoch": 1.6015228426395938, + "grad_norm": 8.4375, + "learning_rate": 3.974072991687975e-06, + "loss": 2.7249, + "step": 15144 + }, + { + "epoch": 1.6016285956006768, + "grad_norm": 8.125, + "learning_rate": 3.973934669018239e-06, + "loss": 2.5083, + "step": 15145 + }, + { + "epoch": 1.60173434856176, + "grad_norm": 8.125, + "learning_rate": 3.973796339431929e-06, + "loss": 2.5508, + "step": 15146 + }, + { + "epoch": 1.6018401015228427, + "grad_norm": 8.3125, + "learning_rate": 3.973658002929695e-06, + "loss": 2.5505, + "step": 15147 + }, + { + "epoch": 1.6019458544839256, + "grad_norm": 8.5625, + "learning_rate": 3.973519659512188e-06, + "loss": 2.5375, + "step": 15148 + }, + { + "epoch": 1.6020516074450084, + "grad_norm": 8.4375, + "learning_rate": 3.973381309180055e-06, + "loss": 2.6108, + "step": 15149 + }, + { + "epoch": 1.6021573604060912, + "grad_norm": 8.4375, + "learning_rate": 3.973242951933947e-06, + "loss": 2.6987, + "step": 15150 + }, + { + "epoch": 1.6022631133671743, + "grad_norm": 8.875, + "learning_rate": 3.973104587774511e-06, + "loss": 2.7357, + "step": 15151 + }, + { + "epoch": 1.6023688663282571, + "grad_norm": 8.5625, + "learning_rate": 3.972966216702399e-06, + "loss": 2.4536, + "step": 15152 + }, + { + "epoch": 1.6024746192893402, + "grad_norm": 8.625, + "learning_rate": 3.9728278387182575e-06, + "loss": 2.4284, + "step": 15153 + }, + { + "epoch": 1.602580372250423, + "grad_norm": 8.1875, + "learning_rate": 3.972689453822738e-06, + "loss": 2.7741, + "step": 15154 + }, + { + "epoch": 1.6026861252115059, + "grad_norm": 7.6875, + "learning_rate": 3.97255106201649e-06, + "loss": 2.3809, + "step": 15155 + }, + { + "epoch": 1.6027918781725887, + "grad_norm": 8.8125, + "learning_rate": 3.9724126633001615e-06, + "loss": 2.5485, + "step": 15156 + }, + { + "epoch": 1.6028976311336718, + "grad_norm": 8.4375, + "learning_rate": 3.972274257674402e-06, + "loss": 2.8641, + "step": 15157 + }, + { + "epoch": 1.6030033840947546, + "grad_norm": 8.0, + "learning_rate": 3.972135845139863e-06, + "loss": 2.5437, + "step": 15158 + }, + { + "epoch": 1.6031091370558377, + "grad_norm": 8.75, + "learning_rate": 3.971997425697192e-06, + "loss": 2.7286, + "step": 15159 + }, + { + "epoch": 1.6032148900169205, + "grad_norm": 8.8125, + "learning_rate": 3.971858999347038e-06, + "loss": 2.9833, + "step": 15160 + }, + { + "epoch": 1.6033206429780034, + "grad_norm": 8.5625, + "learning_rate": 3.971720566090053e-06, + "loss": 2.698, + "step": 15161 + }, + { + "epoch": 1.6034263959390862, + "grad_norm": 8.375, + "learning_rate": 3.971582125926884e-06, + "loss": 2.5172, + "step": 15162 + }, + { + "epoch": 1.603532148900169, + "grad_norm": 7.9375, + "learning_rate": 3.9714436788581825e-06, + "loss": 2.7088, + "step": 15163 + }, + { + "epoch": 1.603637901861252, + "grad_norm": 8.5, + "learning_rate": 3.971305224884599e-06, + "loss": 2.5902, + "step": 15164 + }, + { + "epoch": 1.6037436548223352, + "grad_norm": 8.25, + "learning_rate": 3.97116676400678e-06, + "loss": 2.3619, + "step": 15165 + }, + { + "epoch": 1.603849407783418, + "grad_norm": 8.3125, + "learning_rate": 3.971028296225378e-06, + "loss": 2.533, + "step": 15166 + }, + { + "epoch": 1.6039551607445008, + "grad_norm": 8.5625, + "learning_rate": 3.970889821541041e-06, + "loss": 2.7585, + "step": 15167 + }, + { + "epoch": 1.6040609137055837, + "grad_norm": 8.25, + "learning_rate": 3.9707513399544206e-06, + "loss": 2.5797, + "step": 15168 + }, + { + "epoch": 1.6041666666666665, + "grad_norm": 8.5, + "learning_rate": 3.9706128514661644e-06, + "loss": 2.6136, + "step": 15169 + }, + { + "epoch": 1.6042724196277496, + "grad_norm": 8.125, + "learning_rate": 3.970474356076925e-06, + "loss": 2.5101, + "step": 15170 + }, + { + "epoch": 1.6043781725888326, + "grad_norm": 8.375, + "learning_rate": 3.9703358537873504e-06, + "loss": 2.2814, + "step": 15171 + }, + { + "epoch": 1.6044839255499155, + "grad_norm": 8.75, + "learning_rate": 3.970197344598091e-06, + "loss": 2.4561, + "step": 15172 + }, + { + "epoch": 1.6045896785109983, + "grad_norm": 8.1875, + "learning_rate": 3.9700588285097966e-06, + "loss": 2.3875, + "step": 15173 + }, + { + "epoch": 1.6046954314720812, + "grad_norm": 8.6875, + "learning_rate": 3.9699203055231176e-06, + "loss": 2.7597, + "step": 15174 + }, + { + "epoch": 1.604801184433164, + "grad_norm": 8.125, + "learning_rate": 3.969781775638703e-06, + "loss": 2.4558, + "step": 15175 + }, + { + "epoch": 1.604906937394247, + "grad_norm": 8.4375, + "learning_rate": 3.9696432388572046e-06, + "loss": 2.3065, + "step": 15176 + }, + { + "epoch": 1.60501269035533, + "grad_norm": 8.3125, + "learning_rate": 3.969504695179271e-06, + "loss": 2.601, + "step": 15177 + }, + { + "epoch": 1.605118443316413, + "grad_norm": 8.1875, + "learning_rate": 3.969366144605553e-06, + "loss": 2.6857, + "step": 15178 + }, + { + "epoch": 1.6052241962774958, + "grad_norm": 8.0625, + "learning_rate": 3.9692275871367e-06, + "loss": 2.6124, + "step": 15179 + }, + { + "epoch": 1.6053299492385786, + "grad_norm": 8.1875, + "learning_rate": 3.969089022773364e-06, + "loss": 2.6672, + "step": 15180 + }, + { + "epoch": 1.6054357021996615, + "grad_norm": 8.25, + "learning_rate": 3.968950451516193e-06, + "loss": 2.5396, + "step": 15181 + }, + { + "epoch": 1.6055414551607445, + "grad_norm": 8.4375, + "learning_rate": 3.96881187336584e-06, + "loss": 2.6796, + "step": 15182 + }, + { + "epoch": 1.6056472081218274, + "grad_norm": 8.25, + "learning_rate": 3.968673288322952e-06, + "loss": 2.512, + "step": 15183 + }, + { + "epoch": 1.6057529610829104, + "grad_norm": 7.8125, + "learning_rate": 3.968534696388181e-06, + "loss": 2.5167, + "step": 15184 + }, + { + "epoch": 1.6058587140439933, + "grad_norm": 8.625, + "learning_rate": 3.968396097562178e-06, + "loss": 2.6701, + "step": 15185 + }, + { + "epoch": 1.6059644670050761, + "grad_norm": 8.125, + "learning_rate": 3.968257491845592e-06, + "loss": 2.523, + "step": 15186 + }, + { + "epoch": 1.606070219966159, + "grad_norm": 7.71875, + "learning_rate": 3.968118879239075e-06, + "loss": 2.5715, + "step": 15187 + }, + { + "epoch": 1.606175972927242, + "grad_norm": 8.0, + "learning_rate": 3.967980259743275e-06, + "loss": 2.4564, + "step": 15188 + }, + { + "epoch": 1.6062817258883249, + "grad_norm": 8.1875, + "learning_rate": 3.967841633358846e-06, + "loss": 2.6469, + "step": 15189 + }, + { + "epoch": 1.606387478849408, + "grad_norm": 8.8125, + "learning_rate": 3.967703000086435e-06, + "loss": 2.5324, + "step": 15190 + }, + { + "epoch": 1.6064932318104908, + "grad_norm": 8.3125, + "learning_rate": 3.967564359926695e-06, + "loss": 2.5698, + "step": 15191 + }, + { + "epoch": 1.6065989847715736, + "grad_norm": 8.4375, + "learning_rate": 3.967425712880274e-06, + "loss": 2.6229, + "step": 15192 + }, + { + "epoch": 1.6067047377326564, + "grad_norm": 8.1875, + "learning_rate": 3.967287058947826e-06, + "loss": 2.6061, + "step": 15193 + }, + { + "epoch": 1.6068104906937393, + "grad_norm": 7.78125, + "learning_rate": 3.967148398129999e-06, + "loss": 2.3619, + "step": 15194 + }, + { + "epoch": 1.6069162436548223, + "grad_norm": 8.25, + "learning_rate": 3.967009730427446e-06, + "loss": 2.5491, + "step": 15195 + }, + { + "epoch": 1.6070219966159054, + "grad_norm": 8.4375, + "learning_rate": 3.9668710558408155e-06, + "loss": 2.6767, + "step": 15196 + }, + { + "epoch": 1.6071277495769882, + "grad_norm": 8.0625, + "learning_rate": 3.966732374370759e-06, + "loss": 2.602, + "step": 15197 + }, + { + "epoch": 1.607233502538071, + "grad_norm": 7.875, + "learning_rate": 3.966593686017926e-06, + "loss": 2.5126, + "step": 15198 + }, + { + "epoch": 1.607339255499154, + "grad_norm": 8.5625, + "learning_rate": 3.966454990782971e-06, + "loss": 2.7101, + "step": 15199 + }, + { + "epoch": 1.6074450084602367, + "grad_norm": 8.5, + "learning_rate": 3.9663162886665415e-06, + "loss": 2.602, + "step": 15200 + }, + { + "epoch": 1.6075507614213198, + "grad_norm": 7.6875, + "learning_rate": 3.966177579669289e-06, + "loss": 2.3493, + "step": 15201 + }, + { + "epoch": 1.6076565143824029, + "grad_norm": 8.0625, + "learning_rate": 3.966038863791865e-06, + "loss": 2.3954, + "step": 15202 + }, + { + "epoch": 1.6077622673434857, + "grad_norm": 7.78125, + "learning_rate": 3.965900141034921e-06, + "loss": 2.5601, + "step": 15203 + }, + { + "epoch": 1.6078680203045685, + "grad_norm": 7.96875, + "learning_rate": 3.965761411399106e-06, + "loss": 2.5671, + "step": 15204 + }, + { + "epoch": 1.6079737732656514, + "grad_norm": 8.375, + "learning_rate": 3.965622674885072e-06, + "loss": 2.5567, + "step": 15205 + }, + { + "epoch": 1.6080795262267342, + "grad_norm": 8.5625, + "learning_rate": 3.965483931493471e-06, + "loss": 2.6102, + "step": 15206 + }, + { + "epoch": 1.6081852791878173, + "grad_norm": 8.625, + "learning_rate": 3.965345181224954e-06, + "loss": 2.5771, + "step": 15207 + }, + { + "epoch": 1.6082910321489001, + "grad_norm": 7.75, + "learning_rate": 3.96520642408017e-06, + "loss": 2.5934, + "step": 15208 + }, + { + "epoch": 1.6083967851099832, + "grad_norm": 8.1875, + "learning_rate": 3.965067660059771e-06, + "loss": 2.4087, + "step": 15209 + }, + { + "epoch": 1.608502538071066, + "grad_norm": 8.625, + "learning_rate": 3.96492888916441e-06, + "loss": 2.7205, + "step": 15210 + }, + { + "epoch": 1.6086082910321489, + "grad_norm": 7.9375, + "learning_rate": 3.964790111394738e-06, + "loss": 2.5002, + "step": 15211 + }, + { + "epoch": 1.6087140439932317, + "grad_norm": 8.8125, + "learning_rate": 3.964651326751402e-06, + "loss": 2.8185, + "step": 15212 + }, + { + "epoch": 1.6088197969543148, + "grad_norm": 8.1875, + "learning_rate": 3.964512535235059e-06, + "loss": 2.6031, + "step": 15213 + }, + { + "epoch": 1.6089255499153976, + "grad_norm": 8.0625, + "learning_rate": 3.9643737368463554e-06, + "loss": 2.4994, + "step": 15214 + }, + { + "epoch": 1.6090313028764807, + "grad_norm": 7.96875, + "learning_rate": 3.964234931585946e-06, + "loss": 2.4695, + "step": 15215 + }, + { + "epoch": 1.6091370558375635, + "grad_norm": 8.375, + "learning_rate": 3.964096119454481e-06, + "loss": 2.5871, + "step": 15216 + }, + { + "epoch": 1.6092428087986463, + "grad_norm": 8.125, + "learning_rate": 3.963957300452612e-06, + "loss": 2.3392, + "step": 15217 + }, + { + "epoch": 1.6093485617597292, + "grad_norm": 8.25, + "learning_rate": 3.963818474580989e-06, + "loss": 2.5119, + "step": 15218 + }, + { + "epoch": 1.609454314720812, + "grad_norm": 8.375, + "learning_rate": 3.963679641840266e-06, + "loss": 2.4209, + "step": 15219 + }, + { + "epoch": 1.609560067681895, + "grad_norm": 8.0625, + "learning_rate": 3.9635408022310924e-06, + "loss": 2.5031, + "step": 15220 + }, + { + "epoch": 1.6096658206429781, + "grad_norm": 8.3125, + "learning_rate": 3.963401955754121e-06, + "loss": 2.4361, + "step": 15221 + }, + { + "epoch": 1.609771573604061, + "grad_norm": 8.5, + "learning_rate": 3.963263102410001e-06, + "loss": 2.586, + "step": 15222 + }, + { + "epoch": 1.6098773265651438, + "grad_norm": 8.25, + "learning_rate": 3.963124242199386e-06, + "loss": 2.6036, + "step": 15223 + }, + { + "epoch": 1.6099830795262267, + "grad_norm": 8.4375, + "learning_rate": 3.962985375122928e-06, + "loss": 2.7271, + "step": 15224 + }, + { + "epoch": 1.6100888324873095, + "grad_norm": 7.78125, + "learning_rate": 3.962846501181279e-06, + "loss": 2.491, + "step": 15225 + }, + { + "epoch": 1.6101945854483926, + "grad_norm": 7.90625, + "learning_rate": 3.962707620375088e-06, + "loss": 2.4895, + "step": 15226 + }, + { + "epoch": 1.6103003384094756, + "grad_norm": 8.3125, + "learning_rate": 3.96256873270501e-06, + "loss": 2.6414, + "step": 15227 + }, + { + "epoch": 1.6104060913705585, + "grad_norm": 7.9375, + "learning_rate": 3.962429838171694e-06, + "loss": 2.5508, + "step": 15228 + }, + { + "epoch": 1.6105118443316413, + "grad_norm": 8.9375, + "learning_rate": 3.962290936775793e-06, + "loss": 2.4662, + "step": 15229 + }, + { + "epoch": 1.6106175972927241, + "grad_norm": 8.625, + "learning_rate": 3.962152028517959e-06, + "loss": 2.3073, + "step": 15230 + }, + { + "epoch": 1.610723350253807, + "grad_norm": 7.875, + "learning_rate": 3.962013113398842e-06, + "loss": 2.5338, + "step": 15231 + }, + { + "epoch": 1.61082910321489, + "grad_norm": 8.0, + "learning_rate": 3.961874191419097e-06, + "loss": 2.4457, + "step": 15232 + }, + { + "epoch": 1.6109348561759729, + "grad_norm": 8.5625, + "learning_rate": 3.961735262579375e-06, + "loss": 2.6229, + "step": 15233 + }, + { + "epoch": 1.611040609137056, + "grad_norm": 8.25, + "learning_rate": 3.961596326880326e-06, + "loss": 2.5667, + "step": 15234 + }, + { + "epoch": 1.6111463620981388, + "grad_norm": 7.90625, + "learning_rate": 3.961457384322603e-06, + "loss": 2.4425, + "step": 15235 + }, + { + "epoch": 1.6112521150592216, + "grad_norm": 8.5, + "learning_rate": 3.961318434906859e-06, + "loss": 2.5234, + "step": 15236 + }, + { + "epoch": 1.6113578680203045, + "grad_norm": 8.0625, + "learning_rate": 3.961179478633745e-06, + "loss": 2.5094, + "step": 15237 + }, + { + "epoch": 1.6114636209813875, + "grad_norm": 8.6875, + "learning_rate": 3.961040515503913e-06, + "loss": 2.6122, + "step": 15238 + }, + { + "epoch": 1.6115693739424704, + "grad_norm": 8.75, + "learning_rate": 3.960901545518015e-06, + "loss": 2.4654, + "step": 15239 + }, + { + "epoch": 1.6116751269035534, + "grad_norm": 8.25, + "learning_rate": 3.960762568676705e-06, + "loss": 2.3878, + "step": 15240 + }, + { + "epoch": 1.6117808798646363, + "grad_norm": 7.8125, + "learning_rate": 3.960623584980633e-06, + "loss": 2.5132, + "step": 15241 + }, + { + "epoch": 1.611886632825719, + "grad_norm": 8.1875, + "learning_rate": 3.960484594430451e-06, + "loss": 2.587, + "step": 15242 + }, + { + "epoch": 1.611992385786802, + "grad_norm": 8.125, + "learning_rate": 3.960345597026814e-06, + "loss": 2.6279, + "step": 15243 + }, + { + "epoch": 1.612098138747885, + "grad_norm": 8.25, + "learning_rate": 3.9602065927703705e-06, + "loss": 2.5801, + "step": 15244 + }, + { + "epoch": 1.6122038917089678, + "grad_norm": 8.25, + "learning_rate": 3.9600675816617754e-06, + "loss": 2.557, + "step": 15245 + }, + { + "epoch": 1.612309644670051, + "grad_norm": 8.4375, + "learning_rate": 3.959928563701681e-06, + "loss": 2.6467, + "step": 15246 + }, + { + "epoch": 1.6124153976311337, + "grad_norm": 7.90625, + "learning_rate": 3.959789538890737e-06, + "loss": 2.6117, + "step": 15247 + }, + { + "epoch": 1.6125211505922166, + "grad_norm": 8.4375, + "learning_rate": 3.959650507229599e-06, + "loss": 2.6283, + "step": 15248 + }, + { + "epoch": 1.6126269035532994, + "grad_norm": 8.9375, + "learning_rate": 3.959511468718919e-06, + "loss": 2.7015, + "step": 15249 + }, + { + "epoch": 1.6127326565143822, + "grad_norm": 8.0, + "learning_rate": 3.959372423359346e-06, + "loss": 2.4391, + "step": 15250 + }, + { + "epoch": 1.6128384094754653, + "grad_norm": 8.5625, + "learning_rate": 3.959233371151538e-06, + "loss": 2.4642, + "step": 15251 + }, + { + "epoch": 1.6129441624365484, + "grad_norm": 8.875, + "learning_rate": 3.959094312096144e-06, + "loss": 2.5114, + "step": 15252 + }, + { + "epoch": 1.6130499153976312, + "grad_norm": 8.375, + "learning_rate": 3.958955246193815e-06, + "loss": 2.6079, + "step": 15253 + }, + { + "epoch": 1.613155668358714, + "grad_norm": 8.1875, + "learning_rate": 3.958816173445208e-06, + "loss": 2.746, + "step": 15254 + }, + { + "epoch": 1.6132614213197969, + "grad_norm": 8.6875, + "learning_rate": 3.958677093850972e-06, + "loss": 2.4743, + "step": 15255 + }, + { + "epoch": 1.6133671742808797, + "grad_norm": 8.0625, + "learning_rate": 3.958538007411762e-06, + "loss": 2.5227, + "step": 15256 + }, + { + "epoch": 1.6134729272419628, + "grad_norm": 8.5625, + "learning_rate": 3.958398914128228e-06, + "loss": 2.1309, + "step": 15257 + }, + { + "epoch": 1.6135786802030458, + "grad_norm": 8.3125, + "learning_rate": 3.958259814001027e-06, + "loss": 2.518, + "step": 15258 + }, + { + "epoch": 1.6136844331641287, + "grad_norm": 9.625, + "learning_rate": 3.958120707030807e-06, + "loss": 2.6676, + "step": 15259 + }, + { + "epoch": 1.6137901861252115, + "grad_norm": 9.1875, + "learning_rate": 3.957981593218223e-06, + "loss": 2.6526, + "step": 15260 + }, + { + "epoch": 1.6138959390862944, + "grad_norm": 9.9375, + "learning_rate": 3.957842472563929e-06, + "loss": 2.4886, + "step": 15261 + }, + { + "epoch": 1.6140016920473772, + "grad_norm": 8.9375, + "learning_rate": 3.957703345068575e-06, + "loss": 2.8018, + "step": 15262 + }, + { + "epoch": 1.6141074450084603, + "grad_norm": 8.4375, + "learning_rate": 3.957564210732817e-06, + "loss": 2.7952, + "step": 15263 + }, + { + "epoch": 1.614213197969543, + "grad_norm": 8.3125, + "learning_rate": 3.957425069557305e-06, + "loss": 2.616, + "step": 15264 + }, + { + "epoch": 1.6143189509306262, + "grad_norm": 7.875, + "learning_rate": 3.957285921542694e-06, + "loss": 2.552, + "step": 15265 + }, + { + "epoch": 1.614424703891709, + "grad_norm": 8.625, + "learning_rate": 3.957146766689635e-06, + "loss": 2.656, + "step": 15266 + }, + { + "epoch": 1.6145304568527918, + "grad_norm": 8.8125, + "learning_rate": 3.957007604998784e-06, + "loss": 2.5141, + "step": 15267 + }, + { + "epoch": 1.6146362098138747, + "grad_norm": 8.25, + "learning_rate": 3.956868436470791e-06, + "loss": 2.5077, + "step": 15268 + }, + { + "epoch": 1.6147419627749577, + "grad_norm": 9.25, + "learning_rate": 3.95672926110631e-06, + "loss": 2.6143, + "step": 15269 + }, + { + "epoch": 1.6148477157360406, + "grad_norm": 8.1875, + "learning_rate": 3.956590078905995e-06, + "loss": 2.6127, + "step": 15270 + }, + { + "epoch": 1.6149534686971236, + "grad_norm": 8.3125, + "learning_rate": 3.956450889870498e-06, + "loss": 2.6431, + "step": 15271 + }, + { + "epoch": 1.6150592216582065, + "grad_norm": 8.25, + "learning_rate": 3.9563116940004735e-06, + "loss": 2.5406, + "step": 15272 + }, + { + "epoch": 1.6151649746192893, + "grad_norm": 8.8125, + "learning_rate": 3.956172491296572e-06, + "loss": 2.5064, + "step": 15273 + }, + { + "epoch": 1.6152707275803722, + "grad_norm": 7.9375, + "learning_rate": 3.95603328175945e-06, + "loss": 2.5556, + "step": 15274 + }, + { + "epoch": 1.615376480541455, + "grad_norm": 8.625, + "learning_rate": 3.955894065389759e-06, + "loss": 2.61, + "step": 15275 + }, + { + "epoch": 1.615482233502538, + "grad_norm": 8.25, + "learning_rate": 3.955754842188153e-06, + "loss": 2.6127, + "step": 15276 + }, + { + "epoch": 1.6155879864636211, + "grad_norm": 7.9375, + "learning_rate": 3.955615612155285e-06, + "loss": 2.4708, + "step": 15277 + }, + { + "epoch": 1.615693739424704, + "grad_norm": 8.4375, + "learning_rate": 3.955476375291808e-06, + "loss": 2.4913, + "step": 15278 + }, + { + "epoch": 1.6157994923857868, + "grad_norm": 8.3125, + "learning_rate": 3.955337131598376e-06, + "loss": 2.5908, + "step": 15279 + }, + { + "epoch": 1.6159052453468696, + "grad_norm": 8.5, + "learning_rate": 3.955197881075641e-06, + "loss": 2.5249, + "step": 15280 + }, + { + "epoch": 1.6160109983079525, + "grad_norm": 8.0625, + "learning_rate": 3.955058623724258e-06, + "loss": 2.4676, + "step": 15281 + }, + { + "epoch": 1.6161167512690355, + "grad_norm": 8.1875, + "learning_rate": 3.95491935954488e-06, + "loss": 2.1161, + "step": 15282 + }, + { + "epoch": 1.6162225042301186, + "grad_norm": 7.9375, + "learning_rate": 3.95478008853816e-06, + "loss": 2.5608, + "step": 15283 + }, + { + "epoch": 1.6163282571912014, + "grad_norm": 8.3125, + "learning_rate": 3.954640810704753e-06, + "loss": 2.6044, + "step": 15284 + }, + { + "epoch": 1.6164340101522843, + "grad_norm": 7.9375, + "learning_rate": 3.954501526045311e-06, + "loss": 2.4716, + "step": 15285 + }, + { + "epoch": 1.6165397631133671, + "grad_norm": 8.0625, + "learning_rate": 3.954362234560488e-06, + "loss": 2.4427, + "step": 15286 + }, + { + "epoch": 1.61664551607445, + "grad_norm": 8.625, + "learning_rate": 3.954222936250939e-06, + "loss": 2.5148, + "step": 15287 + }, + { + "epoch": 1.616751269035533, + "grad_norm": 8.4375, + "learning_rate": 3.9540836311173145e-06, + "loss": 2.223, + "step": 15288 + }, + { + "epoch": 1.6168570219966159, + "grad_norm": 8.4375, + "learning_rate": 3.953944319160272e-06, + "loss": 2.838, + "step": 15289 + }, + { + "epoch": 1.616962774957699, + "grad_norm": 8.3125, + "learning_rate": 3.953805000380463e-06, + "loss": 2.6011, + "step": 15290 + }, + { + "epoch": 1.6170685279187818, + "grad_norm": 8.0625, + "learning_rate": 3.953665674778542e-06, + "loss": 2.4612, + "step": 15291 + }, + { + "epoch": 1.6171742808798646, + "grad_norm": 8.25, + "learning_rate": 3.953526342355162e-06, + "loss": 2.6798, + "step": 15292 + }, + { + "epoch": 1.6172800338409474, + "grad_norm": 8.4375, + "learning_rate": 3.953387003110977e-06, + "loss": 2.4042, + "step": 15293 + }, + { + "epoch": 1.6173857868020305, + "grad_norm": 8.5625, + "learning_rate": 3.953247657046642e-06, + "loss": 2.5201, + "step": 15294 + }, + { + "epoch": 1.6174915397631133, + "grad_norm": 8.3125, + "learning_rate": 3.953108304162809e-06, + "loss": 2.7943, + "step": 15295 + }, + { + "epoch": 1.6175972927241964, + "grad_norm": 8.0625, + "learning_rate": 3.952968944460135e-06, + "loss": 2.4212, + "step": 15296 + }, + { + "epoch": 1.6177030456852792, + "grad_norm": 8.375, + "learning_rate": 3.95282957793927e-06, + "loss": 2.4987, + "step": 15297 + }, + { + "epoch": 1.617808798646362, + "grad_norm": 8.5625, + "learning_rate": 3.952690204600871e-06, + "loss": 2.809, + "step": 15298 + }, + { + "epoch": 1.617914551607445, + "grad_norm": 8.125, + "learning_rate": 3.952550824445591e-06, + "loss": 2.7276, + "step": 15299 + }, + { + "epoch": 1.618020304568528, + "grad_norm": 8.125, + "learning_rate": 3.952411437474084e-06, + "loss": 2.5211, + "step": 15300 + }, + { + "epoch": 1.6181260575296108, + "grad_norm": 8.5, + "learning_rate": 3.952272043687004e-06, + "loss": 2.4501, + "step": 15301 + }, + { + "epoch": 1.6182318104906939, + "grad_norm": 8.125, + "learning_rate": 3.952132643085005e-06, + "loss": 2.5822, + "step": 15302 + }, + { + "epoch": 1.6183375634517767, + "grad_norm": 8.6875, + "learning_rate": 3.9519932356687416e-06, + "loss": 2.4842, + "step": 15303 + }, + { + "epoch": 1.6184433164128595, + "grad_norm": 8.3125, + "learning_rate": 3.9518538214388676e-06, + "loss": 2.4864, + "step": 15304 + }, + { + "epoch": 1.6185490693739424, + "grad_norm": 8.0625, + "learning_rate": 3.951714400396038e-06, + "loss": 2.3171, + "step": 15305 + }, + { + "epoch": 1.6186548223350252, + "grad_norm": 8.0, + "learning_rate": 3.951574972540906e-06, + "loss": 2.325, + "step": 15306 + }, + { + "epoch": 1.6187605752961083, + "grad_norm": 8.1875, + "learning_rate": 3.951435537874127e-06, + "loss": 2.5766, + "step": 15307 + }, + { + "epoch": 1.6188663282571913, + "grad_norm": 7.875, + "learning_rate": 3.951296096396353e-06, + "loss": 2.4669, + "step": 15308 + }, + { + "epoch": 1.6189720812182742, + "grad_norm": 8.375, + "learning_rate": 3.951156648108241e-06, + "loss": 2.4059, + "step": 15309 + }, + { + "epoch": 1.619077834179357, + "grad_norm": 8.3125, + "learning_rate": 3.951017193010445e-06, + "loss": 2.6863, + "step": 15310 + }, + { + "epoch": 1.6191835871404399, + "grad_norm": 7.78125, + "learning_rate": 3.950877731103618e-06, + "loss": 2.4741, + "step": 15311 + }, + { + "epoch": 1.6192893401015227, + "grad_norm": 8.3125, + "learning_rate": 3.950738262388415e-06, + "loss": 2.5706, + "step": 15312 + }, + { + "epoch": 1.6193950930626058, + "grad_norm": 8.75, + "learning_rate": 3.950598786865492e-06, + "loss": 2.5879, + "step": 15313 + }, + { + "epoch": 1.6195008460236888, + "grad_norm": 7.8125, + "learning_rate": 3.9504593045355e-06, + "loss": 2.3503, + "step": 15314 + }, + { + "epoch": 1.6196065989847717, + "grad_norm": 8.4375, + "learning_rate": 3.950319815399097e-06, + "loss": 2.7026, + "step": 15315 + }, + { + "epoch": 1.6197123519458545, + "grad_norm": 7.84375, + "learning_rate": 3.950180319456936e-06, + "loss": 2.4742, + "step": 15316 + }, + { + "epoch": 1.6198181049069373, + "grad_norm": 8.125, + "learning_rate": 3.950040816709672e-06, + "loss": 2.5019, + "step": 15317 + }, + { + "epoch": 1.6199238578680202, + "grad_norm": 8.0625, + "learning_rate": 3.94990130715796e-06, + "loss": 2.5811, + "step": 15318 + }, + { + "epoch": 1.6200296108291032, + "grad_norm": 8.5, + "learning_rate": 3.949761790802453e-06, + "loss": 2.6683, + "step": 15319 + }, + { + "epoch": 1.620135363790186, + "grad_norm": 8.0625, + "learning_rate": 3.9496222676438076e-06, + "loss": 2.5274, + "step": 15320 + }, + { + "epoch": 1.6202411167512691, + "grad_norm": 8.125, + "learning_rate": 3.9494827376826785e-06, + "loss": 2.4854, + "step": 15321 + }, + { + "epoch": 1.620346869712352, + "grad_norm": 8.4375, + "learning_rate": 3.949343200919718e-06, + "loss": 2.5078, + "step": 15322 + }, + { + "epoch": 1.6204526226734348, + "grad_norm": 8.375, + "learning_rate": 3.949203657355584e-06, + "loss": 2.5444, + "step": 15323 + }, + { + "epoch": 1.6205583756345177, + "grad_norm": 8.875, + "learning_rate": 3.9490641069909295e-06, + "loss": 2.4965, + "step": 15324 + }, + { + "epoch": 1.6206641285956007, + "grad_norm": 8.5625, + "learning_rate": 3.94892454982641e-06, + "loss": 2.627, + "step": 15325 + }, + { + "epoch": 1.6207698815566836, + "grad_norm": 8.375, + "learning_rate": 3.94878498586268e-06, + "loss": 2.5795, + "step": 15326 + }, + { + "epoch": 1.6208756345177666, + "grad_norm": 8.3125, + "learning_rate": 3.9486454151003944e-06, + "loss": 2.622, + "step": 15327 + }, + { + "epoch": 1.6209813874788495, + "grad_norm": 7.9375, + "learning_rate": 3.948505837540208e-06, + "loss": 2.5226, + "step": 15328 + }, + { + "epoch": 1.6210871404399323, + "grad_norm": 8.25, + "learning_rate": 3.9483662531827775e-06, + "loss": 2.5378, + "step": 15329 + }, + { + "epoch": 1.6211928934010151, + "grad_norm": 8.9375, + "learning_rate": 3.948226662028754e-06, + "loss": 2.5862, + "step": 15330 + }, + { + "epoch": 1.621298646362098, + "grad_norm": 8.4375, + "learning_rate": 3.948087064078797e-06, + "loss": 2.6344, + "step": 15331 + }, + { + "epoch": 1.621404399323181, + "grad_norm": 8.125, + "learning_rate": 3.94794745933356e-06, + "loss": 2.5705, + "step": 15332 + }, + { + "epoch": 1.621510152284264, + "grad_norm": 9.25, + "learning_rate": 3.947807847793698e-06, + "loss": 2.6316, + "step": 15333 + }, + { + "epoch": 1.621615905245347, + "grad_norm": 7.90625, + "learning_rate": 3.9476682294598645e-06, + "loss": 2.4955, + "step": 15334 + }, + { + "epoch": 1.6217216582064298, + "grad_norm": 7.625, + "learning_rate": 3.947528604332717e-06, + "loss": 2.5301, + "step": 15335 + }, + { + "epoch": 1.6218274111675126, + "grad_norm": 8.5625, + "learning_rate": 3.947388972412909e-06, + "loss": 2.4916, + "step": 15336 + }, + { + "epoch": 1.6219331641285955, + "grad_norm": 8.25, + "learning_rate": 3.947249333701096e-06, + "loss": 2.316, + "step": 15337 + }, + { + "epoch": 1.6220389170896785, + "grad_norm": 8.25, + "learning_rate": 3.947109688197936e-06, + "loss": 2.2148, + "step": 15338 + }, + { + "epoch": 1.6221446700507616, + "grad_norm": 8.0, + "learning_rate": 3.94697003590408e-06, + "loss": 2.5242, + "step": 15339 + }, + { + "epoch": 1.6222504230118444, + "grad_norm": 8.875, + "learning_rate": 3.946830376820186e-06, + "loss": 2.4292, + "step": 15340 + }, + { + "epoch": 1.6223561759729273, + "grad_norm": 9.0625, + "learning_rate": 3.9466907109469085e-06, + "loss": 2.8003, + "step": 15341 + }, + { + "epoch": 1.62246192893401, + "grad_norm": 8.5625, + "learning_rate": 3.946551038284904e-06, + "loss": 2.4683, + "step": 15342 + }, + { + "epoch": 1.622567681895093, + "grad_norm": 8.625, + "learning_rate": 3.946411358834827e-06, + "loss": 2.6617, + "step": 15343 + }, + { + "epoch": 1.622673434856176, + "grad_norm": 7.9375, + "learning_rate": 3.946271672597333e-06, + "loss": 2.5203, + "step": 15344 + }, + { + "epoch": 1.6227791878172588, + "grad_norm": 8.5, + "learning_rate": 3.9461319795730765e-06, + "loss": 2.6809, + "step": 15345 + }, + { + "epoch": 1.622884940778342, + "grad_norm": 7.78125, + "learning_rate": 3.9459922797627155e-06, + "loss": 2.5855, + "step": 15346 + }, + { + "epoch": 1.6229906937394247, + "grad_norm": 7.9375, + "learning_rate": 3.945852573166903e-06, + "loss": 2.5358, + "step": 15347 + }, + { + "epoch": 1.6230964467005076, + "grad_norm": 7.96875, + "learning_rate": 3.945712859786297e-06, + "loss": 2.5359, + "step": 15348 + }, + { + "epoch": 1.6232021996615904, + "grad_norm": 8.25, + "learning_rate": 3.945573139621551e-06, + "loss": 2.5122, + "step": 15349 + }, + { + "epoch": 1.6233079526226735, + "grad_norm": 8.5, + "learning_rate": 3.945433412673321e-06, + "loss": 2.6901, + "step": 15350 + }, + { + "epoch": 1.6234137055837563, + "grad_norm": 8.3125, + "learning_rate": 3.945293678942264e-06, + "loss": 2.4284, + "step": 15351 + }, + { + "epoch": 1.6235194585448394, + "grad_norm": 8.75, + "learning_rate": 3.945153938429034e-06, + "loss": 2.5139, + "step": 15352 + }, + { + "epoch": 1.6236252115059222, + "grad_norm": 8.8125, + "learning_rate": 3.9450141911342884e-06, + "loss": 2.847, + "step": 15353 + }, + { + "epoch": 1.623730964467005, + "grad_norm": 8.3125, + "learning_rate": 3.944874437058682e-06, + "loss": 2.7192, + "step": 15354 + }, + { + "epoch": 1.6238367174280879, + "grad_norm": 7.96875, + "learning_rate": 3.94473467620287e-06, + "loss": 2.5286, + "step": 15355 + }, + { + "epoch": 1.623942470389171, + "grad_norm": 8.3125, + "learning_rate": 3.9445949085675104e-06, + "loss": 2.3845, + "step": 15356 + }, + { + "epoch": 1.6240482233502538, + "grad_norm": 8.1875, + "learning_rate": 3.944455134153257e-06, + "loss": 2.6651, + "step": 15357 + }, + { + "epoch": 1.6241539763113368, + "grad_norm": 8.625, + "learning_rate": 3.944315352960767e-06, + "loss": 2.5762, + "step": 15358 + }, + { + "epoch": 1.6242597292724197, + "grad_norm": 8.6875, + "learning_rate": 3.944175564990696e-06, + "loss": 2.4766, + "step": 15359 + }, + { + "epoch": 1.6243654822335025, + "grad_norm": 8.625, + "learning_rate": 3.944035770243698e-06, + "loss": 2.532, + "step": 15360 + }, + { + "epoch": 1.6244712351945854, + "grad_norm": 8.1875, + "learning_rate": 3.943895968720431e-06, + "loss": 2.458, + "step": 15361 + }, + { + "epoch": 1.6245769881556682, + "grad_norm": 8.4375, + "learning_rate": 3.943756160421553e-06, + "loss": 2.6236, + "step": 15362 + }, + { + "epoch": 1.6246827411167513, + "grad_norm": 8.0625, + "learning_rate": 3.9436163453477165e-06, + "loss": 2.5648, + "step": 15363 + }, + { + "epoch": 1.6247884940778343, + "grad_norm": 8.625, + "learning_rate": 3.943476523499578e-06, + "loss": 2.3436, + "step": 15364 + }, + { + "epoch": 1.6248942470389172, + "grad_norm": 8.25, + "learning_rate": 3.943336694877795e-06, + "loss": 2.5237, + "step": 15365 + }, + { + "epoch": 1.625, + "grad_norm": 8.3125, + "learning_rate": 3.9431968594830234e-06, + "loss": 2.3228, + "step": 15366 + }, + { + "epoch": 1.6251057529610828, + "grad_norm": 8.5, + "learning_rate": 3.943057017315919e-06, + "loss": 2.6409, + "step": 15367 + }, + { + "epoch": 1.6252115059221657, + "grad_norm": 7.84375, + "learning_rate": 3.9429171683771385e-06, + "loss": 2.3202, + "step": 15368 + }, + { + "epoch": 1.6253172588832487, + "grad_norm": 8.125, + "learning_rate": 3.942777312667338e-06, + "loss": 2.5715, + "step": 15369 + }, + { + "epoch": 1.6254230118443318, + "grad_norm": 8.25, + "learning_rate": 3.942637450187173e-06, + "loss": 2.6036, + "step": 15370 + }, + { + "epoch": 1.6255287648054146, + "grad_norm": 8.375, + "learning_rate": 3.942497580937301e-06, + "loss": 2.6496, + "step": 15371 + }, + { + "epoch": 1.6256345177664975, + "grad_norm": 8.0625, + "learning_rate": 3.942357704918378e-06, + "loss": 2.6056, + "step": 15372 + }, + { + "epoch": 1.6257402707275803, + "grad_norm": 8.375, + "learning_rate": 3.9422178221310595e-06, + "loss": 2.6195, + "step": 15373 + }, + { + "epoch": 1.6258460236886632, + "grad_norm": 8.875, + "learning_rate": 3.942077932576003e-06, + "loss": 2.8539, + "step": 15374 + }, + { + "epoch": 1.6259517766497462, + "grad_norm": 8.625, + "learning_rate": 3.941938036253865e-06, + "loss": 2.5572, + "step": 15375 + }, + { + "epoch": 1.626057529610829, + "grad_norm": 8.5625, + "learning_rate": 3.941798133165301e-06, + "loss": 2.6724, + "step": 15376 + }, + { + "epoch": 1.6261632825719121, + "grad_norm": 7.90625, + "learning_rate": 3.941658223310968e-06, + "loss": 2.3571, + "step": 15377 + }, + { + "epoch": 1.626269035532995, + "grad_norm": 8.0625, + "learning_rate": 3.941518306691523e-06, + "loss": 2.5689, + "step": 15378 + }, + { + "epoch": 1.6263747884940778, + "grad_norm": 8.1875, + "learning_rate": 3.941378383307621e-06, + "loss": 2.7009, + "step": 15379 + }, + { + "epoch": 1.6264805414551606, + "grad_norm": 7.875, + "learning_rate": 3.941238453159922e-06, + "loss": 2.3919, + "step": 15380 + }, + { + "epoch": 1.6265862944162437, + "grad_norm": 8.125, + "learning_rate": 3.941098516249078e-06, + "loss": 2.7323, + "step": 15381 + }, + { + "epoch": 1.6266920473773265, + "grad_norm": 8.375, + "learning_rate": 3.94095857257575e-06, + "loss": 2.7174, + "step": 15382 + }, + { + "epoch": 1.6267978003384096, + "grad_norm": 8.125, + "learning_rate": 3.940818622140591e-06, + "loss": 2.5125, + "step": 15383 + }, + { + "epoch": 1.6269035532994924, + "grad_norm": 8.5, + "learning_rate": 3.94067866494426e-06, + "loss": 2.8578, + "step": 15384 + }, + { + "epoch": 1.6270093062605753, + "grad_norm": 8.25, + "learning_rate": 3.940538700987414e-06, + "loss": 2.5315, + "step": 15385 + }, + { + "epoch": 1.627115059221658, + "grad_norm": 8.625, + "learning_rate": 3.940398730270708e-06, + "loss": 2.7176, + "step": 15386 + }, + { + "epoch": 1.6272208121827412, + "grad_norm": 8.1875, + "learning_rate": 3.940258752794801e-06, + "loss": 2.4804, + "step": 15387 + }, + { + "epoch": 1.627326565143824, + "grad_norm": 7.8125, + "learning_rate": 3.940118768560348e-06, + "loss": 2.4412, + "step": 15388 + }, + { + "epoch": 1.627432318104907, + "grad_norm": 8.5625, + "learning_rate": 3.939978777568007e-06, + "loss": 2.5241, + "step": 15389 + }, + { + "epoch": 1.62753807106599, + "grad_norm": 8.25, + "learning_rate": 3.939838779818434e-06, + "loss": 2.6688, + "step": 15390 + }, + { + "epoch": 1.6276438240270727, + "grad_norm": 8.0, + "learning_rate": 3.939698775312286e-06, + "loss": 2.6521, + "step": 15391 + }, + { + "epoch": 1.6277495769881556, + "grad_norm": 8.5, + "learning_rate": 3.939558764050222e-06, + "loss": 2.4733, + "step": 15392 + }, + { + "epoch": 1.6278553299492384, + "grad_norm": 8.3125, + "learning_rate": 3.939418746032896e-06, + "loss": 2.6062, + "step": 15393 + }, + { + "epoch": 1.6279610829103215, + "grad_norm": 8.4375, + "learning_rate": 3.939278721260968e-06, + "loss": 2.5854, + "step": 15394 + }, + { + "epoch": 1.6280668358714045, + "grad_norm": 8.1875, + "learning_rate": 3.939138689735093e-06, + "loss": 2.3621, + "step": 15395 + }, + { + "epoch": 1.6281725888324874, + "grad_norm": 7.96875, + "learning_rate": 3.938998651455928e-06, + "loss": 2.448, + "step": 15396 + }, + { + "epoch": 1.6282783417935702, + "grad_norm": 8.4375, + "learning_rate": 3.938858606424132e-06, + "loss": 2.3906, + "step": 15397 + }, + { + "epoch": 1.628384094754653, + "grad_norm": 8.125, + "learning_rate": 3.93871855464036e-06, + "loss": 2.6041, + "step": 15398 + }, + { + "epoch": 1.628489847715736, + "grad_norm": 8.3125, + "learning_rate": 3.938578496105271e-06, + "loss": 2.501, + "step": 15399 + }, + { + "epoch": 1.628595600676819, + "grad_norm": 7.875, + "learning_rate": 3.938438430819521e-06, + "loss": 2.4527, + "step": 15400 + }, + { + "epoch": 1.628701353637902, + "grad_norm": 8.25, + "learning_rate": 3.9382983587837674e-06, + "loss": 2.4859, + "step": 15401 + }, + { + "epoch": 1.6288071065989849, + "grad_norm": 7.53125, + "learning_rate": 3.938158279998669e-06, + "loss": 2.4994, + "step": 15402 + }, + { + "epoch": 1.6289128595600677, + "grad_norm": 7.84375, + "learning_rate": 3.938018194464881e-06, + "loss": 2.1697, + "step": 15403 + }, + { + "epoch": 1.6290186125211505, + "grad_norm": 8.4375, + "learning_rate": 3.937878102183062e-06, + "loss": 2.6637, + "step": 15404 + }, + { + "epoch": 1.6291243654822334, + "grad_norm": 8.625, + "learning_rate": 3.937738003153869e-06, + "loss": 2.5808, + "step": 15405 + }, + { + "epoch": 1.6292301184433164, + "grad_norm": 8.3125, + "learning_rate": 3.93759789737796e-06, + "loss": 2.3893, + "step": 15406 + }, + { + "epoch": 1.6293358714043993, + "grad_norm": 8.75, + "learning_rate": 3.937457784855991e-06, + "loss": 2.6173, + "step": 15407 + }, + { + "epoch": 1.6294416243654823, + "grad_norm": 8.125, + "learning_rate": 3.937317665588621e-06, + "loss": 2.4237, + "step": 15408 + }, + { + "epoch": 1.6295473773265652, + "grad_norm": 8.4375, + "learning_rate": 3.937177539576508e-06, + "loss": 2.327, + "step": 15409 + }, + { + "epoch": 1.629653130287648, + "grad_norm": 8.0625, + "learning_rate": 3.937037406820308e-06, + "loss": 2.4901, + "step": 15410 + }, + { + "epoch": 1.6297588832487309, + "grad_norm": 8.4375, + "learning_rate": 3.93689726732068e-06, + "loss": 2.4673, + "step": 15411 + }, + { + "epoch": 1.629864636209814, + "grad_norm": 8.3125, + "learning_rate": 3.936757121078279e-06, + "loss": 2.5294, + "step": 15412 + }, + { + "epoch": 1.6299703891708968, + "grad_norm": 8.1875, + "learning_rate": 3.936616968093766e-06, + "loss": 2.6174, + "step": 15413 + }, + { + "epoch": 1.6300761421319798, + "grad_norm": 8.5, + "learning_rate": 3.936476808367796e-06, + "loss": 2.5971, + "step": 15414 + }, + { + "epoch": 1.6301818950930627, + "grad_norm": 8.0625, + "learning_rate": 3.936336641901029e-06, + "loss": 2.6742, + "step": 15415 + }, + { + "epoch": 1.6302876480541455, + "grad_norm": 8.25, + "learning_rate": 3.936196468694121e-06, + "loss": 2.6024, + "step": 15416 + }, + { + "epoch": 1.6303934010152283, + "grad_norm": 8.6875, + "learning_rate": 3.9360562887477295e-06, + "loss": 2.8594, + "step": 15417 + }, + { + "epoch": 1.6304991539763112, + "grad_norm": 8.25, + "learning_rate": 3.935916102062514e-06, + "loss": 2.3307, + "step": 15418 + }, + { + "epoch": 1.6306049069373942, + "grad_norm": 8.25, + "learning_rate": 3.935775908639131e-06, + "loss": 2.1212, + "step": 15419 + }, + { + "epoch": 1.6307106598984773, + "grad_norm": 8.875, + "learning_rate": 3.9356357084782395e-06, + "loss": 2.6762, + "step": 15420 + }, + { + "epoch": 1.6308164128595601, + "grad_norm": 8.125, + "learning_rate": 3.935495501580497e-06, + "loss": 2.5244, + "step": 15421 + }, + { + "epoch": 1.630922165820643, + "grad_norm": 8.3125, + "learning_rate": 3.935355287946561e-06, + "loss": 2.5414, + "step": 15422 + }, + { + "epoch": 1.6310279187817258, + "grad_norm": 9.0625, + "learning_rate": 3.935215067577089e-06, + "loss": 2.4742, + "step": 15423 + }, + { + "epoch": 1.6311336717428087, + "grad_norm": 8.125, + "learning_rate": 3.93507484047274e-06, + "loss": 2.4117, + "step": 15424 + }, + { + "epoch": 1.6312394247038917, + "grad_norm": 8.5, + "learning_rate": 3.934934606634172e-06, + "loss": 2.5654, + "step": 15425 + }, + { + "epoch": 1.6313451776649748, + "grad_norm": 8.0, + "learning_rate": 3.934794366062043e-06, + "loss": 2.3662, + "step": 15426 + }, + { + "epoch": 1.6314509306260576, + "grad_norm": 7.9375, + "learning_rate": 3.93465411875701e-06, + "loss": 2.6016, + "step": 15427 + }, + { + "epoch": 1.6315566835871405, + "grad_norm": 8.375, + "learning_rate": 3.934513864719732e-06, + "loss": 2.5836, + "step": 15428 + }, + { + "epoch": 1.6316624365482233, + "grad_norm": 8.0, + "learning_rate": 3.934373603950868e-06, + "loss": 2.3904, + "step": 15429 + }, + { + "epoch": 1.6317681895093061, + "grad_norm": 8.25, + "learning_rate": 3.934233336451074e-06, + "loss": 2.5376, + "step": 15430 + }, + { + "epoch": 1.6318739424703892, + "grad_norm": 8.125, + "learning_rate": 3.934093062221011e-06, + "loss": 2.4542, + "step": 15431 + }, + { + "epoch": 1.631979695431472, + "grad_norm": 8.8125, + "learning_rate": 3.9339527812613356e-06, + "loss": 2.5594, + "step": 15432 + }, + { + "epoch": 1.632085448392555, + "grad_norm": 8.3125, + "learning_rate": 3.933812493572706e-06, + "loss": 2.5919, + "step": 15433 + }, + { + "epoch": 1.632191201353638, + "grad_norm": 8.125, + "learning_rate": 3.93367219915578e-06, + "loss": 2.4726, + "step": 15434 + }, + { + "epoch": 1.6322969543147208, + "grad_norm": 8.5625, + "learning_rate": 3.933531898011217e-06, + "loss": 2.452, + "step": 15435 + }, + { + "epoch": 1.6324027072758036, + "grad_norm": 8.0625, + "learning_rate": 3.933391590139675e-06, + "loss": 2.449, + "step": 15436 + }, + { + "epoch": 1.6325084602368867, + "grad_norm": 8.6875, + "learning_rate": 3.933251275541814e-06, + "loss": 2.5635, + "step": 15437 + }, + { + "epoch": 1.6326142131979695, + "grad_norm": 8.1875, + "learning_rate": 3.93311095421829e-06, + "loss": 2.5161, + "step": 15438 + }, + { + "epoch": 1.6327199661590526, + "grad_norm": 8.5625, + "learning_rate": 3.932970626169762e-06, + "loss": 2.4292, + "step": 15439 + }, + { + "epoch": 1.6328257191201354, + "grad_norm": 8.875, + "learning_rate": 3.932830291396889e-06, + "loss": 2.674, + "step": 15440 + }, + { + "epoch": 1.6329314720812182, + "grad_norm": 8.375, + "learning_rate": 3.93268994990033e-06, + "loss": 2.6831, + "step": 15441 + }, + { + "epoch": 1.633037225042301, + "grad_norm": 7.96875, + "learning_rate": 3.9325496016807425e-06, + "loss": 2.5328, + "step": 15442 + }, + { + "epoch": 1.6331429780033841, + "grad_norm": 9.0625, + "learning_rate": 3.932409246738785e-06, + "loss": 2.6637, + "step": 15443 + }, + { + "epoch": 1.633248730964467, + "grad_norm": 7.9375, + "learning_rate": 3.9322688850751176e-06, + "loss": 2.4453, + "step": 15444 + }, + { + "epoch": 1.63335448392555, + "grad_norm": 8.75, + "learning_rate": 3.9321285166903985e-06, + "loss": 2.8528, + "step": 15445 + }, + { + "epoch": 1.6334602368866329, + "grad_norm": 8.375, + "learning_rate": 3.931988141585285e-06, + "loss": 2.7039, + "step": 15446 + }, + { + "epoch": 1.6335659898477157, + "grad_norm": 8.0, + "learning_rate": 3.931847759760438e-06, + "loss": 2.4119, + "step": 15447 + }, + { + "epoch": 1.6336717428087986, + "grad_norm": 8.3125, + "learning_rate": 3.931707371216514e-06, + "loss": 2.5145, + "step": 15448 + }, + { + "epoch": 1.6337774957698814, + "grad_norm": 9.3125, + "learning_rate": 3.931566975954173e-06, + "loss": 2.449, + "step": 15449 + }, + { + "epoch": 1.6338832487309645, + "grad_norm": 7.96875, + "learning_rate": 3.931426573974074e-06, + "loss": 2.6518, + "step": 15450 + }, + { + "epoch": 1.6339890016920475, + "grad_norm": 8.0625, + "learning_rate": 3.931286165276876e-06, + "loss": 2.4695, + "step": 15451 + }, + { + "epoch": 1.6340947546531304, + "grad_norm": 8.8125, + "learning_rate": 3.931145749863237e-06, + "loss": 2.5745, + "step": 15452 + }, + { + "epoch": 1.6342005076142132, + "grad_norm": 8.5625, + "learning_rate": 3.931005327733816e-06, + "loss": 2.4783, + "step": 15453 + }, + { + "epoch": 1.634306260575296, + "grad_norm": 8.9375, + "learning_rate": 3.930864898889272e-06, + "loss": 2.6248, + "step": 15454 + }, + { + "epoch": 1.6344120135363789, + "grad_norm": 8.4375, + "learning_rate": 3.930724463330265e-06, + "loss": 2.6663, + "step": 15455 + }, + { + "epoch": 1.634517766497462, + "grad_norm": 8.375, + "learning_rate": 3.930584021057453e-06, + "loss": 2.5846, + "step": 15456 + }, + { + "epoch": 1.634623519458545, + "grad_norm": 8.6875, + "learning_rate": 3.930443572071495e-06, + "loss": 2.6399, + "step": 15457 + }, + { + "epoch": 1.6347292724196278, + "grad_norm": 8.0625, + "learning_rate": 3.930303116373051e-06, + "loss": 2.5463, + "step": 15458 + }, + { + "epoch": 1.6348350253807107, + "grad_norm": 7.875, + "learning_rate": 3.930162653962779e-06, + "loss": 2.51, + "step": 15459 + }, + { + "epoch": 1.6349407783417935, + "grad_norm": 9.0625, + "learning_rate": 3.930022184841338e-06, + "loss": 2.6741, + "step": 15460 + }, + { + "epoch": 1.6350465313028764, + "grad_norm": 8.125, + "learning_rate": 3.929881709009388e-06, + "loss": 2.7393, + "step": 15461 + }, + { + "epoch": 1.6351522842639594, + "grad_norm": 8.4375, + "learning_rate": 3.9297412264675885e-06, + "loss": 2.5696, + "step": 15462 + }, + { + "epoch": 1.6352580372250423, + "grad_norm": 8.4375, + "learning_rate": 3.929600737216598e-06, + "loss": 2.6283, + "step": 15463 + }, + { + "epoch": 1.6353637901861253, + "grad_norm": 8.125, + "learning_rate": 3.929460241257076e-06, + "loss": 2.5456, + "step": 15464 + }, + { + "epoch": 1.6354695431472082, + "grad_norm": 8.5625, + "learning_rate": 3.9293197385896815e-06, + "loss": 2.5772, + "step": 15465 + }, + { + "epoch": 1.635575296108291, + "grad_norm": 8.4375, + "learning_rate": 3.929179229215075e-06, + "loss": 2.684, + "step": 15466 + }, + { + "epoch": 1.6356810490693738, + "grad_norm": 8.25, + "learning_rate": 3.929038713133913e-06, + "loss": 2.8508, + "step": 15467 + }, + { + "epoch": 1.635786802030457, + "grad_norm": 9.3125, + "learning_rate": 3.9288981903468585e-06, + "loss": 2.6265, + "step": 15468 + }, + { + "epoch": 1.6358925549915397, + "grad_norm": 8.0625, + "learning_rate": 3.928757660854568e-06, + "loss": 2.5503, + "step": 15469 + }, + { + "epoch": 1.6359983079526228, + "grad_norm": 9.0, + "learning_rate": 3.928617124657702e-06, + "loss": 2.6636, + "step": 15470 + }, + { + "epoch": 1.6361040609137056, + "grad_norm": 9.3125, + "learning_rate": 3.928476581756921e-06, + "loss": 2.6903, + "step": 15471 + }, + { + "epoch": 1.6362098138747885, + "grad_norm": 8.1875, + "learning_rate": 3.928336032152883e-06, + "loss": 2.5492, + "step": 15472 + }, + { + "epoch": 1.6363155668358713, + "grad_norm": 8.0, + "learning_rate": 3.928195475846248e-06, + "loss": 2.5095, + "step": 15473 + }, + { + "epoch": 1.6364213197969542, + "grad_norm": 8.8125, + "learning_rate": 3.928054912837676e-06, + "loss": 2.4025, + "step": 15474 + }, + { + "epoch": 1.6365270727580372, + "grad_norm": 8.1875, + "learning_rate": 3.927914343127828e-06, + "loss": 2.5123, + "step": 15475 + }, + { + "epoch": 1.6366328257191203, + "grad_norm": 8.5, + "learning_rate": 3.9277737667173605e-06, + "loss": 2.4705, + "step": 15476 + }, + { + "epoch": 1.6367385786802031, + "grad_norm": 8.1875, + "learning_rate": 3.927633183606935e-06, + "loss": 2.4125, + "step": 15477 + }, + { + "epoch": 1.636844331641286, + "grad_norm": 8.5625, + "learning_rate": 3.92749259379721e-06, + "loss": 2.5674, + "step": 15478 + }, + { + "epoch": 1.6369500846023688, + "grad_norm": 8.3125, + "learning_rate": 3.927351997288846e-06, + "loss": 2.5068, + "step": 15479 + }, + { + "epoch": 1.6370558375634516, + "grad_norm": 8.6875, + "learning_rate": 3.927211394082504e-06, + "loss": 2.7235, + "step": 15480 + }, + { + "epoch": 1.6371615905245347, + "grad_norm": 7.96875, + "learning_rate": 3.9270707841788425e-06, + "loss": 2.4607, + "step": 15481 + }, + { + "epoch": 1.6372673434856178, + "grad_norm": 8.25, + "learning_rate": 3.926930167578521e-06, + "loss": 2.3644, + "step": 15482 + }, + { + "epoch": 1.6373730964467006, + "grad_norm": 8.0625, + "learning_rate": 3.9267895442822005e-06, + "loss": 2.1725, + "step": 15483 + }, + { + "epoch": 1.6374788494077834, + "grad_norm": 7.8125, + "learning_rate": 3.926648914290539e-06, + "loss": 2.2967, + "step": 15484 + }, + { + "epoch": 1.6375846023688663, + "grad_norm": 8.1875, + "learning_rate": 3.926508277604199e-06, + "loss": 2.3625, + "step": 15485 + }, + { + "epoch": 1.637690355329949, + "grad_norm": 8.625, + "learning_rate": 3.926367634223838e-06, + "loss": 2.598, + "step": 15486 + }, + { + "epoch": 1.6377961082910322, + "grad_norm": 8.125, + "learning_rate": 3.9262269841501165e-06, + "loss": 2.689, + "step": 15487 + }, + { + "epoch": 1.637901861252115, + "grad_norm": 9.0625, + "learning_rate": 3.926086327383696e-06, + "loss": 2.6445, + "step": 15488 + }, + { + "epoch": 1.638007614213198, + "grad_norm": 9.25, + "learning_rate": 3.925945663925235e-06, + "loss": 2.4836, + "step": 15489 + }, + { + "epoch": 1.638113367174281, + "grad_norm": 8.5625, + "learning_rate": 3.9258049937753955e-06, + "loss": 2.5328, + "step": 15490 + }, + { + "epoch": 1.6382191201353637, + "grad_norm": 7.96875, + "learning_rate": 3.9256643169348365e-06, + "loss": 2.5382, + "step": 15491 + }, + { + "epoch": 1.6383248730964466, + "grad_norm": 8.0625, + "learning_rate": 3.925523633404216e-06, + "loss": 2.5157, + "step": 15492 + }, + { + "epoch": 1.6384306260575296, + "grad_norm": 8.3125, + "learning_rate": 3.925382943184197e-06, + "loss": 2.5435, + "step": 15493 + }, + { + "epoch": 1.6385363790186125, + "grad_norm": 8.25, + "learning_rate": 3.925242246275439e-06, + "loss": 2.6896, + "step": 15494 + }, + { + "epoch": 1.6386421319796955, + "grad_norm": 8.4375, + "learning_rate": 3.925101542678602e-06, + "loss": 2.2189, + "step": 15495 + }, + { + "epoch": 1.6387478849407784, + "grad_norm": 8.625, + "learning_rate": 3.924960832394346e-06, + "loss": 2.5297, + "step": 15496 + }, + { + "epoch": 1.6388536379018612, + "grad_norm": 8.3125, + "learning_rate": 3.924820115423332e-06, + "loss": 2.2506, + "step": 15497 + }, + { + "epoch": 1.638959390862944, + "grad_norm": 8.25, + "learning_rate": 3.92467939176622e-06, + "loss": 2.3141, + "step": 15498 + }, + { + "epoch": 1.6390651438240271, + "grad_norm": 8.3125, + "learning_rate": 3.9245386614236705e-06, + "loss": 2.4188, + "step": 15499 + }, + { + "epoch": 1.63917089678511, + "grad_norm": 8.1875, + "learning_rate": 3.924397924396343e-06, + "loss": 2.5833, + "step": 15500 + }, + { + "epoch": 1.639276649746193, + "grad_norm": 8.5, + "learning_rate": 3.924257180684899e-06, + "loss": 2.6531, + "step": 15501 + }, + { + "epoch": 1.6393824027072759, + "grad_norm": 8.1875, + "learning_rate": 3.924116430289999e-06, + "loss": 2.5179, + "step": 15502 + }, + { + "epoch": 1.6394881556683587, + "grad_norm": 8.5625, + "learning_rate": 3.923975673212302e-06, + "loss": 2.5478, + "step": 15503 + }, + { + "epoch": 1.6395939086294415, + "grad_norm": 7.8125, + "learning_rate": 3.92383490945247e-06, + "loss": 2.5579, + "step": 15504 + }, + { + "epoch": 1.6396996615905244, + "grad_norm": 8.4375, + "learning_rate": 3.9236941390111624e-06, + "loss": 2.7045, + "step": 15505 + }, + { + "epoch": 1.6398054145516074, + "grad_norm": 8.5, + "learning_rate": 3.923553361889042e-06, + "loss": 2.4766, + "step": 15506 + }, + { + "epoch": 1.6399111675126905, + "grad_norm": 8.0625, + "learning_rate": 3.923412578086767e-06, + "loss": 2.4153, + "step": 15507 + }, + { + "epoch": 1.6400169204737733, + "grad_norm": 8.1875, + "learning_rate": 3.923271787604999e-06, + "loss": 2.4194, + "step": 15508 + }, + { + "epoch": 1.6401226734348562, + "grad_norm": 8.3125, + "learning_rate": 3.923130990444398e-06, + "loss": 2.4732, + "step": 15509 + }, + { + "epoch": 1.640228426395939, + "grad_norm": 8.75, + "learning_rate": 3.922990186605626e-06, + "loss": 2.5633, + "step": 15510 + }, + { + "epoch": 1.6403341793570219, + "grad_norm": 8.1875, + "learning_rate": 3.922849376089343e-06, + "loss": 2.5406, + "step": 15511 + }, + { + "epoch": 1.640439932318105, + "grad_norm": 8.1875, + "learning_rate": 3.92270855889621e-06, + "loss": 2.5376, + "step": 15512 + }, + { + "epoch": 1.640545685279188, + "grad_norm": 8.75, + "learning_rate": 3.922567735026887e-06, + "loss": 2.705, + "step": 15513 + }, + { + "epoch": 1.6406514382402708, + "grad_norm": 8.3125, + "learning_rate": 3.922426904482036e-06, + "loss": 2.6785, + "step": 15514 + }, + { + "epoch": 1.6407571912013537, + "grad_norm": 8.125, + "learning_rate": 3.922286067262318e-06, + "loss": 2.5382, + "step": 15515 + }, + { + "epoch": 1.6408629441624365, + "grad_norm": 9.0625, + "learning_rate": 3.922145223368391e-06, + "loss": 2.7584, + "step": 15516 + }, + { + "epoch": 1.6409686971235193, + "grad_norm": 8.4375, + "learning_rate": 3.92200437280092e-06, + "loss": 2.6054, + "step": 15517 + }, + { + "epoch": 1.6410744500846024, + "grad_norm": 8.1875, + "learning_rate": 3.9218635155605634e-06, + "loss": 2.755, + "step": 15518 + }, + { + "epoch": 1.6411802030456852, + "grad_norm": 8.875, + "learning_rate": 3.921722651647982e-06, + "loss": 2.5938, + "step": 15519 + }, + { + "epoch": 1.6412859560067683, + "grad_norm": 9.0, + "learning_rate": 3.921581781063839e-06, + "loss": 2.5418, + "step": 15520 + }, + { + "epoch": 1.6413917089678511, + "grad_norm": 8.5, + "learning_rate": 3.921440903808795e-06, + "loss": 2.7768, + "step": 15521 + }, + { + "epoch": 1.641497461928934, + "grad_norm": 8.5625, + "learning_rate": 3.921300019883508e-06, + "loss": 2.6385, + "step": 15522 + }, + { + "epoch": 1.6416032148900168, + "grad_norm": 8.6875, + "learning_rate": 3.921159129288642e-06, + "loss": 2.8956, + "step": 15523 + }, + { + "epoch": 1.6417089678510999, + "grad_norm": 8.125, + "learning_rate": 3.9210182320248565e-06, + "loss": 2.3873, + "step": 15524 + }, + { + "epoch": 1.6418147208121827, + "grad_norm": 8.0, + "learning_rate": 3.920877328092815e-06, + "loss": 2.4888, + "step": 15525 + }, + { + "epoch": 1.6419204737732658, + "grad_norm": 8.25, + "learning_rate": 3.920736417493177e-06, + "loss": 2.5541, + "step": 15526 + }, + { + "epoch": 1.6420262267343486, + "grad_norm": 8.6875, + "learning_rate": 3.920595500226605e-06, + "loss": 2.6071, + "step": 15527 + }, + { + "epoch": 1.6421319796954315, + "grad_norm": 8.1875, + "learning_rate": 3.920454576293757e-06, + "loss": 2.6, + "step": 15528 + }, + { + "epoch": 1.6422377326565143, + "grad_norm": 8.125, + "learning_rate": 3.920313645695299e-06, + "loss": 2.4263, + "step": 15529 + }, + { + "epoch": 1.6423434856175971, + "grad_norm": 8.125, + "learning_rate": 3.920172708431888e-06, + "loss": 2.5879, + "step": 15530 + }, + { + "epoch": 1.6424492385786802, + "grad_norm": 8.375, + "learning_rate": 3.920031764504189e-06, + "loss": 2.6435, + "step": 15531 + }, + { + "epoch": 1.6425549915397633, + "grad_norm": 8.375, + "learning_rate": 3.919890813912861e-06, + "loss": 2.6129, + "step": 15532 + }, + { + "epoch": 1.642660744500846, + "grad_norm": 8.4375, + "learning_rate": 3.919749856658566e-06, + "loss": 2.5754, + "step": 15533 + }, + { + "epoch": 1.642766497461929, + "grad_norm": 8.375, + "learning_rate": 3.9196088927419655e-06, + "loss": 2.7608, + "step": 15534 + }, + { + "epoch": 1.6428722504230118, + "grad_norm": 7.75, + "learning_rate": 3.919467922163721e-06, + "loss": 2.6029, + "step": 15535 + }, + { + "epoch": 1.6429780033840946, + "grad_norm": 9.0, + "learning_rate": 3.919326944924494e-06, + "loss": 2.5063, + "step": 15536 + }, + { + "epoch": 1.6430837563451777, + "grad_norm": 8.1875, + "learning_rate": 3.9191859610249476e-06, + "loss": 2.5025, + "step": 15537 + }, + { + "epoch": 1.6431895093062607, + "grad_norm": 8.75, + "learning_rate": 3.919044970465741e-06, + "loss": 2.7783, + "step": 15538 + }, + { + "epoch": 1.6432952622673436, + "grad_norm": 10.375, + "learning_rate": 3.918903973247535e-06, + "loss": 2.4925, + "step": 15539 + }, + { + "epoch": 1.6434010152284264, + "grad_norm": 8.5625, + "learning_rate": 3.918762969370995e-06, + "loss": 2.7143, + "step": 15540 + }, + { + "epoch": 1.6435067681895092, + "grad_norm": 12.375, + "learning_rate": 3.91862195883678e-06, + "loss": 2.4843, + "step": 15541 + }, + { + "epoch": 1.643612521150592, + "grad_norm": 8.5625, + "learning_rate": 3.918480941645553e-06, + "loss": 2.6167, + "step": 15542 + }, + { + "epoch": 1.6437182741116751, + "grad_norm": 8.1875, + "learning_rate": 3.918339917797974e-06, + "loss": 2.3075, + "step": 15543 + }, + { + "epoch": 1.643824027072758, + "grad_norm": 8.25, + "learning_rate": 3.918198887294706e-06, + "loss": 2.5166, + "step": 15544 + }, + { + "epoch": 1.643929780033841, + "grad_norm": 8.125, + "learning_rate": 3.9180578501364104e-06, + "loss": 2.4027, + "step": 15545 + }, + { + "epoch": 1.6440355329949239, + "grad_norm": 9.1875, + "learning_rate": 3.917916806323751e-06, + "loss": 2.7664, + "step": 15546 + }, + { + "epoch": 1.6441412859560067, + "grad_norm": 8.375, + "learning_rate": 3.9177757558573865e-06, + "loss": 2.829, + "step": 15547 + }, + { + "epoch": 1.6442470389170896, + "grad_norm": 8.0, + "learning_rate": 3.917634698737981e-06, + "loss": 2.5784, + "step": 15548 + }, + { + "epoch": 1.6443527918781726, + "grad_norm": 8.375, + "learning_rate": 3.917493634966195e-06, + "loss": 2.6031, + "step": 15549 + }, + { + "epoch": 1.6444585448392555, + "grad_norm": 7.9375, + "learning_rate": 3.917352564542691e-06, + "loss": 2.4853, + "step": 15550 + }, + { + "epoch": 1.6445642978003385, + "grad_norm": 8.9375, + "learning_rate": 3.917211487468131e-06, + "loss": 2.4173, + "step": 15551 + }, + { + "epoch": 1.6446700507614214, + "grad_norm": 8.125, + "learning_rate": 3.917070403743178e-06, + "loss": 2.5594, + "step": 15552 + }, + { + "epoch": 1.6447758037225042, + "grad_norm": 8.6875, + "learning_rate": 3.916929313368493e-06, + "loss": 2.6431, + "step": 15553 + }, + { + "epoch": 1.644881556683587, + "grad_norm": 8.1875, + "learning_rate": 3.9167882163447364e-06, + "loss": 2.62, + "step": 15554 + }, + { + "epoch": 1.64498730964467, + "grad_norm": 8.6875, + "learning_rate": 3.916647112672574e-06, + "loss": 2.678, + "step": 15555 + }, + { + "epoch": 1.645093062605753, + "grad_norm": 8.0625, + "learning_rate": 3.9165060023526665e-06, + "loss": 2.5849, + "step": 15556 + }, + { + "epoch": 1.645198815566836, + "grad_norm": 8.3125, + "learning_rate": 3.916364885385674e-06, + "loss": 2.3782, + "step": 15557 + }, + { + "epoch": 1.6453045685279188, + "grad_norm": 8.3125, + "learning_rate": 3.916223761772261e-06, + "loss": 2.6667, + "step": 15558 + }, + { + "epoch": 1.6454103214890017, + "grad_norm": 8.25, + "learning_rate": 3.916082631513089e-06, + "loss": 2.2454, + "step": 15559 + }, + { + "epoch": 1.6455160744500845, + "grad_norm": 8.5625, + "learning_rate": 3.915941494608821e-06, + "loss": 2.5338, + "step": 15560 + }, + { + "epoch": 1.6456218274111674, + "grad_norm": 8.1875, + "learning_rate": 3.91580035106012e-06, + "loss": 2.6268, + "step": 15561 + }, + { + "epoch": 1.6457275803722504, + "grad_norm": 7.96875, + "learning_rate": 3.9156592008676445e-06, + "loss": 2.587, + "step": 15562 + }, + { + "epoch": 1.6458333333333335, + "grad_norm": 8.5, + "learning_rate": 3.91551804403206e-06, + "loss": 2.4981, + "step": 15563 + }, + { + "epoch": 1.6459390862944163, + "grad_norm": 7.9375, + "learning_rate": 3.915376880554029e-06, + "loss": 2.4947, + "step": 15564 + }, + { + "epoch": 1.6460448392554992, + "grad_norm": 8.75, + "learning_rate": 3.915235710434212e-06, + "loss": 2.3959, + "step": 15565 + }, + { + "epoch": 1.646150592216582, + "grad_norm": 7.875, + "learning_rate": 3.915094533673274e-06, + "loss": 2.4436, + "step": 15566 + }, + { + "epoch": 1.6462563451776648, + "grad_norm": 8.1875, + "learning_rate": 3.914953350271875e-06, + "loss": 2.5449, + "step": 15567 + }, + { + "epoch": 1.646362098138748, + "grad_norm": 8.25, + "learning_rate": 3.914812160230679e-06, + "loss": 2.5186, + "step": 15568 + }, + { + "epoch": 1.646467851099831, + "grad_norm": 8.4375, + "learning_rate": 3.914670963550348e-06, + "loss": 2.6797, + "step": 15569 + }, + { + "epoch": 1.6465736040609138, + "grad_norm": 8.375, + "learning_rate": 3.914529760231545e-06, + "loss": 2.851, + "step": 15570 + }, + { + "epoch": 1.6466793570219966, + "grad_norm": 8.125, + "learning_rate": 3.914388550274932e-06, + "loss": 2.7047, + "step": 15571 + }, + { + "epoch": 1.6467851099830795, + "grad_norm": 7.78125, + "learning_rate": 3.914247333681172e-06, + "loss": 2.5116, + "step": 15572 + }, + { + "epoch": 1.6468908629441623, + "grad_norm": 8.3125, + "learning_rate": 3.914106110450928e-06, + "loss": 2.4773, + "step": 15573 + }, + { + "epoch": 1.6469966159052454, + "grad_norm": 8.4375, + "learning_rate": 3.9139648805848625e-06, + "loss": 2.6778, + "step": 15574 + }, + { + "epoch": 1.6471023688663282, + "grad_norm": 8.4375, + "learning_rate": 3.913823644083639e-06, + "loss": 2.5148, + "step": 15575 + }, + { + "epoch": 1.6472081218274113, + "grad_norm": 8.375, + "learning_rate": 3.913682400947917e-06, + "loss": 2.4507, + "step": 15576 + }, + { + "epoch": 1.6473138747884941, + "grad_norm": 8.1875, + "learning_rate": 3.913541151178364e-06, + "loss": 2.6191, + "step": 15577 + }, + { + "epoch": 1.647419627749577, + "grad_norm": 8.125, + "learning_rate": 3.913399894775639e-06, + "loss": 2.6268, + "step": 15578 + }, + { + "epoch": 1.6475253807106598, + "grad_norm": 8.3125, + "learning_rate": 3.913258631740407e-06, + "loss": 2.522, + "step": 15579 + }, + { + "epoch": 1.6476311336717429, + "grad_norm": 8.375, + "learning_rate": 3.91311736207333e-06, + "loss": 2.4361, + "step": 15580 + }, + { + "epoch": 1.6477368866328257, + "grad_norm": 7.90625, + "learning_rate": 3.912976085775071e-06, + "loss": 2.612, + "step": 15581 + }, + { + "epoch": 1.6478426395939088, + "grad_norm": 7.8125, + "learning_rate": 3.912834802846293e-06, + "loss": 2.4631, + "step": 15582 + }, + { + "epoch": 1.6479483925549916, + "grad_norm": 8.0625, + "learning_rate": 3.9126935132876595e-06, + "loss": 2.5121, + "step": 15583 + }, + { + "epoch": 1.6480541455160744, + "grad_norm": 8.5, + "learning_rate": 3.912552217099833e-06, + "loss": 2.4569, + "step": 15584 + }, + { + "epoch": 1.6481598984771573, + "grad_norm": 8.75, + "learning_rate": 3.912410914283477e-06, + "loss": 2.6876, + "step": 15585 + }, + { + "epoch": 1.64826565143824, + "grad_norm": 8.4375, + "learning_rate": 3.912269604839254e-06, + "loss": 2.5913, + "step": 15586 + }, + { + "epoch": 1.6483714043993232, + "grad_norm": 8.375, + "learning_rate": 3.912128288767827e-06, + "loss": 2.6181, + "step": 15587 + }, + { + "epoch": 1.6484771573604062, + "grad_norm": 8.0625, + "learning_rate": 3.91198696606986e-06, + "loss": 2.4036, + "step": 15588 + }, + { + "epoch": 1.648582910321489, + "grad_norm": 8.25, + "learning_rate": 3.911845636746015e-06, + "loss": 2.5284, + "step": 15589 + }, + { + "epoch": 1.648688663282572, + "grad_norm": 8.0625, + "learning_rate": 3.9117043007969565e-06, + "loss": 2.5627, + "step": 15590 + }, + { + "epoch": 1.6487944162436547, + "grad_norm": 8.125, + "learning_rate": 3.911562958223347e-06, + "loss": 2.6419, + "step": 15591 + }, + { + "epoch": 1.6489001692047376, + "grad_norm": 8.5625, + "learning_rate": 3.911421609025849e-06, + "loss": 2.7029, + "step": 15592 + }, + { + "epoch": 1.6490059221658206, + "grad_norm": 8.4375, + "learning_rate": 3.911280253205128e-06, + "loss": 2.6292, + "step": 15593 + }, + { + "epoch": 1.6491116751269037, + "grad_norm": 8.0, + "learning_rate": 3.911138890761845e-06, + "loss": 2.5278, + "step": 15594 + }, + { + "epoch": 1.6492174280879865, + "grad_norm": 8.625, + "learning_rate": 3.910997521696665e-06, + "loss": 2.4866, + "step": 15595 + }, + { + "epoch": 1.6493231810490694, + "grad_norm": 8.75, + "learning_rate": 3.91085614601025e-06, + "loss": 2.0798, + "step": 15596 + }, + { + "epoch": 1.6494289340101522, + "grad_norm": 8.25, + "learning_rate": 3.910714763703265e-06, + "loss": 2.2522, + "step": 15597 + }, + { + "epoch": 1.649534686971235, + "grad_norm": 8.5625, + "learning_rate": 3.910573374776372e-06, + "loss": 2.5535, + "step": 15598 + }, + { + "epoch": 1.6496404399323181, + "grad_norm": 8.0625, + "learning_rate": 3.9104319792302345e-06, + "loss": 2.5257, + "step": 15599 + }, + { + "epoch": 1.649746192893401, + "grad_norm": 8.8125, + "learning_rate": 3.910290577065518e-06, + "loss": 2.6263, + "step": 15600 + }, + { + "epoch": 1.649851945854484, + "grad_norm": 8.6875, + "learning_rate": 3.910149168282883e-06, + "loss": 2.5067, + "step": 15601 + }, + { + "epoch": 1.6499576988155669, + "grad_norm": 8.125, + "learning_rate": 3.9100077528829954e-06, + "loss": 2.4602, + "step": 15602 + }, + { + "epoch": 1.6500634517766497, + "grad_norm": 8.5, + "learning_rate": 3.909866330866518e-06, + "loss": 2.4793, + "step": 15603 + }, + { + "epoch": 1.6501692047377325, + "grad_norm": 8.375, + "learning_rate": 3.909724902234116e-06, + "loss": 2.4995, + "step": 15604 + }, + { + "epoch": 1.6502749576988156, + "grad_norm": 7.96875, + "learning_rate": 3.909583466986449e-06, + "loss": 2.5227, + "step": 15605 + }, + { + "epoch": 1.6503807106598984, + "grad_norm": 8.0, + "learning_rate": 3.909442025124185e-06, + "loss": 2.4652, + "step": 15606 + }, + { + "epoch": 1.6504864636209815, + "grad_norm": 7.9375, + "learning_rate": 3.909300576647986e-06, + "loss": 2.6113, + "step": 15607 + }, + { + "epoch": 1.6505922165820643, + "grad_norm": 8.5, + "learning_rate": 3.909159121558515e-06, + "loss": 2.4863, + "step": 15608 + }, + { + "epoch": 1.6506979695431472, + "grad_norm": 8.875, + "learning_rate": 3.909017659856437e-06, + "loss": 2.5313, + "step": 15609 + }, + { + "epoch": 1.65080372250423, + "grad_norm": 8.25, + "learning_rate": 3.908876191542416e-06, + "loss": 2.2338, + "step": 15610 + }, + { + "epoch": 1.650909475465313, + "grad_norm": 8.0, + "learning_rate": 3.908734716617113e-06, + "loss": 2.3887, + "step": 15611 + }, + { + "epoch": 1.651015228426396, + "grad_norm": 8.3125, + "learning_rate": 3.908593235081196e-06, + "loss": 2.4117, + "step": 15612 + }, + { + "epoch": 1.651120981387479, + "grad_norm": 8.0, + "learning_rate": 3.908451746935327e-06, + "loss": 2.5181, + "step": 15613 + }, + { + "epoch": 1.6512267343485618, + "grad_norm": 8.0625, + "learning_rate": 3.90831025218017e-06, + "loss": 2.545, + "step": 15614 + }, + { + "epoch": 1.6513324873096447, + "grad_norm": 7.84375, + "learning_rate": 3.908168750816389e-06, + "loss": 2.3972, + "step": 15615 + }, + { + "epoch": 1.6514382402707275, + "grad_norm": 8.0625, + "learning_rate": 3.908027242844648e-06, + "loss": 2.5297, + "step": 15616 + }, + { + "epoch": 1.6515439932318103, + "grad_norm": 8.25, + "learning_rate": 3.90788572826561e-06, + "loss": 2.5661, + "step": 15617 + }, + { + "epoch": 1.6516497461928934, + "grad_norm": 7.9375, + "learning_rate": 3.907744207079941e-06, + "loss": 2.5003, + "step": 15618 + }, + { + "epoch": 1.6517554991539765, + "grad_norm": 8.1875, + "learning_rate": 3.9076026792883035e-06, + "loss": 2.463, + "step": 15619 + }, + { + "epoch": 1.6518612521150593, + "grad_norm": 8.8125, + "learning_rate": 3.907461144891362e-06, + "loss": 2.6631, + "step": 15620 + }, + { + "epoch": 1.6519670050761421, + "grad_norm": 8.375, + "learning_rate": 3.907319603889782e-06, + "loss": 2.5796, + "step": 15621 + }, + { + "epoch": 1.652072758037225, + "grad_norm": 8.1875, + "learning_rate": 3.9071780562842275e-06, + "loss": 2.4318, + "step": 15622 + }, + { + "epoch": 1.6521785109983078, + "grad_norm": 8.1875, + "learning_rate": 3.90703650207536e-06, + "loss": 2.4685, + "step": 15623 + }, + { + "epoch": 1.6522842639593909, + "grad_norm": 8.4375, + "learning_rate": 3.9068949412638465e-06, + "loss": 2.2622, + "step": 15624 + }, + { + "epoch": 1.652390016920474, + "grad_norm": 8.0625, + "learning_rate": 3.9067533738503505e-06, + "loss": 2.5187, + "step": 15625 + }, + { + "epoch": 1.6524957698815568, + "grad_norm": 8.375, + "learning_rate": 3.906611799835536e-06, + "loss": 2.5687, + "step": 15626 + }, + { + "epoch": 1.6526015228426396, + "grad_norm": 8.4375, + "learning_rate": 3.906470219220068e-06, + "loss": 2.6193, + "step": 15627 + }, + { + "epoch": 1.6527072758037225, + "grad_norm": 8.75, + "learning_rate": 3.90632863200461e-06, + "loss": 2.6666, + "step": 15628 + }, + { + "epoch": 1.6528130287648053, + "grad_norm": 8.625, + "learning_rate": 3.906187038189827e-06, + "loss": 2.5533, + "step": 15629 + }, + { + "epoch": 1.6529187817258884, + "grad_norm": 8.6875, + "learning_rate": 3.906045437776384e-06, + "loss": 2.46, + "step": 15630 + }, + { + "epoch": 1.6530245346869712, + "grad_norm": 8.5625, + "learning_rate": 3.905903830764944e-06, + "loss": 2.6452, + "step": 15631 + }, + { + "epoch": 1.6531302876480543, + "grad_norm": 8.25, + "learning_rate": 3.905762217156173e-06, + "loss": 2.5475, + "step": 15632 + }, + { + "epoch": 1.653236040609137, + "grad_norm": 8.4375, + "learning_rate": 3.9056205969507344e-06, + "loss": 2.6826, + "step": 15633 + }, + { + "epoch": 1.65334179357022, + "grad_norm": 8.0625, + "learning_rate": 3.905478970149294e-06, + "loss": 2.6379, + "step": 15634 + }, + { + "epoch": 1.6534475465313028, + "grad_norm": 8.3125, + "learning_rate": 3.905337336752514e-06, + "loss": 2.569, + "step": 15635 + }, + { + "epoch": 1.6535532994923858, + "grad_norm": 8.625, + "learning_rate": 3.905195696761062e-06, + "loss": 2.8343, + "step": 15636 + }, + { + "epoch": 1.6536590524534687, + "grad_norm": 8.4375, + "learning_rate": 3.905054050175601e-06, + "loss": 2.6128, + "step": 15637 + }, + { + "epoch": 1.6537648054145517, + "grad_norm": 8.3125, + "learning_rate": 3.904912396996795e-06, + "loss": 2.7773, + "step": 15638 + }, + { + "epoch": 1.6538705583756346, + "grad_norm": 8.0625, + "learning_rate": 3.904770737225311e-06, + "loss": 2.4628, + "step": 15639 + }, + { + "epoch": 1.6539763113367174, + "grad_norm": 8.1875, + "learning_rate": 3.904629070861812e-06, + "loss": 2.5643, + "step": 15640 + }, + { + "epoch": 1.6540820642978002, + "grad_norm": 8.5, + "learning_rate": 3.9044873979069635e-06, + "loss": 2.4169, + "step": 15641 + }, + { + "epoch": 1.654187817258883, + "grad_norm": 8.1875, + "learning_rate": 3.90434571836143e-06, + "loss": 2.4199, + "step": 15642 + }, + { + "epoch": 1.6542935702199661, + "grad_norm": 8.375, + "learning_rate": 3.904204032225876e-06, + "loss": 2.8291, + "step": 15643 + }, + { + "epoch": 1.6543993231810492, + "grad_norm": 7.84375, + "learning_rate": 3.904062339500967e-06, + "loss": 2.4611, + "step": 15644 + }, + { + "epoch": 1.654505076142132, + "grad_norm": 7.96875, + "learning_rate": 3.903920640187367e-06, + "loss": 2.4009, + "step": 15645 + }, + { + "epoch": 1.6546108291032149, + "grad_norm": 8.0, + "learning_rate": 3.903778934285742e-06, + "loss": 2.5675, + "step": 15646 + }, + { + "epoch": 1.6547165820642977, + "grad_norm": 8.0625, + "learning_rate": 3.903637221796758e-06, + "loss": 2.3725, + "step": 15647 + }, + { + "epoch": 1.6548223350253806, + "grad_norm": 8.375, + "learning_rate": 3.9034955027210766e-06, + "loss": 2.3979, + "step": 15648 + }, + { + "epoch": 1.6549280879864636, + "grad_norm": 8.0, + "learning_rate": 3.903353777059366e-06, + "loss": 2.3255, + "step": 15649 + }, + { + "epoch": 1.6550338409475467, + "grad_norm": 8.5625, + "learning_rate": 3.9032120448122904e-06, + "loss": 2.3374, + "step": 15650 + }, + { + "epoch": 1.6551395939086295, + "grad_norm": 8.6875, + "learning_rate": 3.903070305980513e-06, + "loss": 2.6947, + "step": 15651 + }, + { + "epoch": 1.6552453468697124, + "grad_norm": 8.25, + "learning_rate": 3.902928560564701e-06, + "loss": 2.5947, + "step": 15652 + }, + { + "epoch": 1.6553510998307952, + "grad_norm": 8.0625, + "learning_rate": 3.90278680856552e-06, + "loss": 2.5429, + "step": 15653 + }, + { + "epoch": 1.655456852791878, + "grad_norm": 8.4375, + "learning_rate": 3.902645049983635e-06, + "loss": 2.6044, + "step": 15654 + }, + { + "epoch": 1.655562605752961, + "grad_norm": 8.75, + "learning_rate": 3.902503284819708e-06, + "loss": 2.6557, + "step": 15655 + }, + { + "epoch": 1.655668358714044, + "grad_norm": 8.3125, + "learning_rate": 3.902361513074407e-06, + "loss": 2.6124, + "step": 15656 + }, + { + "epoch": 1.655774111675127, + "grad_norm": 8.5, + "learning_rate": 3.902219734748399e-06, + "loss": 2.7141, + "step": 15657 + }, + { + "epoch": 1.6558798646362098, + "grad_norm": 8.25, + "learning_rate": 3.902077949842345e-06, + "loss": 2.6705, + "step": 15658 + }, + { + "epoch": 1.6559856175972927, + "grad_norm": 8.4375, + "learning_rate": 3.901936158356914e-06, + "loss": 2.5638, + "step": 15659 + }, + { + "epoch": 1.6560913705583755, + "grad_norm": 8.25, + "learning_rate": 3.90179436029277e-06, + "loss": 2.5666, + "step": 15660 + }, + { + "epoch": 1.6561971235194586, + "grad_norm": 8.6875, + "learning_rate": 3.901652555650578e-06, + "loss": 2.625, + "step": 15661 + }, + { + "epoch": 1.6563028764805414, + "grad_norm": 8.25, + "learning_rate": 3.901510744431004e-06, + "loss": 2.5482, + "step": 15662 + }, + { + "epoch": 1.6564086294416245, + "grad_norm": 7.875, + "learning_rate": 3.901368926634713e-06, + "loss": 2.4471, + "step": 15663 + }, + { + "epoch": 1.6565143824027073, + "grad_norm": 8.125, + "learning_rate": 3.9012271022623704e-06, + "loss": 2.4582, + "step": 15664 + }, + { + "epoch": 1.6566201353637902, + "grad_norm": 8.625, + "learning_rate": 3.901085271314643e-06, + "loss": 2.6544, + "step": 15665 + }, + { + "epoch": 1.656725888324873, + "grad_norm": 8.375, + "learning_rate": 3.900943433792195e-06, + "loss": 2.3973, + "step": 15666 + }, + { + "epoch": 1.656831641285956, + "grad_norm": 8.25, + "learning_rate": 3.9008015896956924e-06, + "loss": 2.6653, + "step": 15667 + }, + { + "epoch": 1.656937394247039, + "grad_norm": 8.0625, + "learning_rate": 3.900659739025801e-06, + "loss": 2.4671, + "step": 15668 + }, + { + "epoch": 1.657043147208122, + "grad_norm": 8.875, + "learning_rate": 3.900517881783186e-06, + "loss": 2.5758, + "step": 15669 + }, + { + "epoch": 1.6571489001692048, + "grad_norm": 8.4375, + "learning_rate": 3.900376017968513e-06, + "loss": 2.4534, + "step": 15670 + }, + { + "epoch": 1.6572546531302876, + "grad_norm": 8.25, + "learning_rate": 3.900234147582449e-06, + "loss": 2.3153, + "step": 15671 + }, + { + "epoch": 1.6573604060913705, + "grad_norm": 7.875, + "learning_rate": 3.9000922706256595e-06, + "loss": 2.4485, + "step": 15672 + }, + { + "epoch": 1.6574661590524533, + "grad_norm": 8.4375, + "learning_rate": 3.899950387098808e-06, + "loss": 2.6386, + "step": 15673 + }, + { + "epoch": 1.6575719120135364, + "grad_norm": 8.375, + "learning_rate": 3.899808497002562e-06, + "loss": 2.6132, + "step": 15674 + }, + { + "epoch": 1.6576776649746194, + "grad_norm": 8.0, + "learning_rate": 3.8996666003375885e-06, + "loss": 2.2567, + "step": 15675 + }, + { + "epoch": 1.6577834179357023, + "grad_norm": 7.6875, + "learning_rate": 3.899524697104551e-06, + "loss": 2.4512, + "step": 15676 + }, + { + "epoch": 1.6578891708967851, + "grad_norm": 8.5, + "learning_rate": 3.899382787304117e-06, + "loss": 2.5182, + "step": 15677 + }, + { + "epoch": 1.657994923857868, + "grad_norm": 9.0, + "learning_rate": 3.8992408709369515e-06, + "loss": 2.556, + "step": 15678 + }, + { + "epoch": 1.6581006768189508, + "grad_norm": 8.875, + "learning_rate": 3.8990989480037215e-06, + "loss": 2.5222, + "step": 15679 + }, + { + "epoch": 1.6582064297800339, + "grad_norm": 8.5625, + "learning_rate": 3.898957018505091e-06, + "loss": 2.7123, + "step": 15680 + }, + { + "epoch": 1.658312182741117, + "grad_norm": 8.625, + "learning_rate": 3.898815082441728e-06, + "loss": 2.6373, + "step": 15681 + }, + { + "epoch": 1.6584179357021998, + "grad_norm": 9.0625, + "learning_rate": 3.898673139814298e-06, + "loss": 2.8433, + "step": 15682 + }, + { + "epoch": 1.6585236886632826, + "grad_norm": 8.4375, + "learning_rate": 3.898531190623468e-06, + "loss": 2.5021, + "step": 15683 + }, + { + "epoch": 1.6586294416243654, + "grad_norm": 8.25, + "learning_rate": 3.898389234869902e-06, + "loss": 2.3426, + "step": 15684 + }, + { + "epoch": 1.6587351945854483, + "grad_norm": 8.25, + "learning_rate": 3.898247272554268e-06, + "loss": 2.5824, + "step": 15685 + }, + { + "epoch": 1.6588409475465313, + "grad_norm": 8.375, + "learning_rate": 3.89810530367723e-06, + "loss": 2.634, + "step": 15686 + }, + { + "epoch": 1.6589467005076142, + "grad_norm": 8.25, + "learning_rate": 3.897963328239457e-06, + "loss": 2.567, + "step": 15687 + }, + { + "epoch": 1.6590524534686972, + "grad_norm": 8.5, + "learning_rate": 3.897821346241614e-06, + "loss": 2.4519, + "step": 15688 + }, + { + "epoch": 1.65915820642978, + "grad_norm": 8.375, + "learning_rate": 3.897679357684366e-06, + "loss": 2.5221, + "step": 15689 + }, + { + "epoch": 1.659263959390863, + "grad_norm": 8.8125, + "learning_rate": 3.897537362568381e-06, + "loss": 2.7078, + "step": 15690 + }, + { + "epoch": 1.6593697123519457, + "grad_norm": 8.375, + "learning_rate": 3.897395360894324e-06, + "loss": 2.7664, + "step": 15691 + }, + { + "epoch": 1.6594754653130288, + "grad_norm": 8.1875, + "learning_rate": 3.897253352662863e-06, + "loss": 2.3692, + "step": 15692 + }, + { + "epoch": 1.6595812182741116, + "grad_norm": 8.0625, + "learning_rate": 3.897111337874663e-06, + "loss": 2.4574, + "step": 15693 + }, + { + "epoch": 1.6596869712351947, + "grad_norm": 8.0, + "learning_rate": 3.896969316530391e-06, + "loss": 2.4453, + "step": 15694 + }, + { + "epoch": 1.6597927241962775, + "grad_norm": 7.78125, + "learning_rate": 3.896827288630714e-06, + "loss": 2.4335, + "step": 15695 + }, + { + "epoch": 1.6598984771573604, + "grad_norm": 8.5, + "learning_rate": 3.896685254176297e-06, + "loss": 2.251, + "step": 15696 + }, + { + "epoch": 1.6600042301184432, + "grad_norm": 8.0625, + "learning_rate": 3.896543213167807e-06, + "loss": 2.5131, + "step": 15697 + }, + { + "epoch": 1.660109983079526, + "grad_norm": 8.1875, + "learning_rate": 3.896401165605912e-06, + "loss": 2.4135, + "step": 15698 + }, + { + "epoch": 1.6602157360406091, + "grad_norm": 8.125, + "learning_rate": 3.8962591114912764e-06, + "loss": 2.3864, + "step": 15699 + }, + { + "epoch": 1.6603214890016922, + "grad_norm": 8.5625, + "learning_rate": 3.8961170508245684e-06, + "loss": 2.593, + "step": 15700 + }, + { + "epoch": 1.660427241962775, + "grad_norm": 8.5, + "learning_rate": 3.8959749836064545e-06, + "loss": 2.5594, + "step": 15701 + }, + { + "epoch": 1.6605329949238579, + "grad_norm": 7.8125, + "learning_rate": 3.8958329098376e-06, + "loss": 2.4505, + "step": 15702 + }, + { + "epoch": 1.6606387478849407, + "grad_norm": 8.0625, + "learning_rate": 3.895690829518673e-06, + "loss": 2.6849, + "step": 15703 + }, + { + "epoch": 1.6607445008460235, + "grad_norm": 8.0625, + "learning_rate": 3.895548742650339e-06, + "loss": 2.245, + "step": 15704 + }, + { + "epoch": 1.6608502538071066, + "grad_norm": 9.125, + "learning_rate": 3.895406649233267e-06, + "loss": 3.0613, + "step": 15705 + }, + { + "epoch": 1.6609560067681897, + "grad_norm": 8.25, + "learning_rate": 3.895264549268122e-06, + "loss": 2.4878, + "step": 15706 + }, + { + "epoch": 1.6610617597292725, + "grad_norm": 8.875, + "learning_rate": 3.89512244275557e-06, + "loss": 2.763, + "step": 15707 + }, + { + "epoch": 1.6611675126903553, + "grad_norm": 8.0625, + "learning_rate": 3.89498032969628e-06, + "loss": 2.4158, + "step": 15708 + }, + { + "epoch": 1.6612732656514382, + "grad_norm": 9.125, + "learning_rate": 3.894838210090916e-06, + "loss": 2.7353, + "step": 15709 + }, + { + "epoch": 1.661379018612521, + "grad_norm": 8.125, + "learning_rate": 3.894696083940149e-06, + "loss": 2.4822, + "step": 15710 + }, + { + "epoch": 1.661484771573604, + "grad_norm": 7.71875, + "learning_rate": 3.894553951244643e-06, + "loss": 2.4871, + "step": 15711 + }, + { + "epoch": 1.661590524534687, + "grad_norm": 8.5, + "learning_rate": 3.894411812005066e-06, + "loss": 2.663, + "step": 15712 + }, + { + "epoch": 1.66169627749577, + "grad_norm": 8.625, + "learning_rate": 3.8942696662220836e-06, + "loss": 2.5406, + "step": 15713 + }, + { + "epoch": 1.6618020304568528, + "grad_norm": 8.125, + "learning_rate": 3.894127513896365e-06, + "loss": 2.5593, + "step": 15714 + }, + { + "epoch": 1.6619077834179357, + "grad_norm": 8.0, + "learning_rate": 3.893985355028575e-06, + "loss": 2.4291, + "step": 15715 + }, + { + "epoch": 1.6620135363790185, + "grad_norm": 8.1875, + "learning_rate": 3.893843189619383e-06, + "loss": 2.4949, + "step": 15716 + }, + { + "epoch": 1.6621192893401016, + "grad_norm": 8.5625, + "learning_rate": 3.893701017669454e-06, + "loss": 2.5917, + "step": 15717 + }, + { + "epoch": 1.6622250423011844, + "grad_norm": 8.5625, + "learning_rate": 3.893558839179456e-06, + "loss": 2.6662, + "step": 15718 + }, + { + "epoch": 1.6623307952622675, + "grad_norm": 8.1875, + "learning_rate": 3.893416654150057e-06, + "loss": 2.5643, + "step": 15719 + }, + { + "epoch": 1.6624365482233503, + "grad_norm": 7.5625, + "learning_rate": 3.893274462581924e-06, + "loss": 2.3054, + "step": 15720 + }, + { + "epoch": 1.6625423011844331, + "grad_norm": 8.4375, + "learning_rate": 3.893132264475723e-06, + "loss": 2.2366, + "step": 15721 + }, + { + "epoch": 1.662648054145516, + "grad_norm": 8.625, + "learning_rate": 3.892990059832122e-06, + "loss": 2.6158, + "step": 15722 + }, + { + "epoch": 1.662753807106599, + "grad_norm": 8.3125, + "learning_rate": 3.892847848651789e-06, + "loss": 2.6183, + "step": 15723 + }, + { + "epoch": 1.6628595600676819, + "grad_norm": 8.375, + "learning_rate": 3.892705630935391e-06, + "loss": 2.5187, + "step": 15724 + }, + { + "epoch": 1.662965313028765, + "grad_norm": 7.84375, + "learning_rate": 3.892563406683594e-06, + "loss": 2.7204, + "step": 15725 + }, + { + "epoch": 1.6630710659898478, + "grad_norm": 8.1875, + "learning_rate": 3.892421175897067e-06, + "loss": 2.3993, + "step": 15726 + }, + { + "epoch": 1.6631768189509306, + "grad_norm": 8.3125, + "learning_rate": 3.892278938576476e-06, + "loss": 2.6628, + "step": 15727 + }, + { + "epoch": 1.6632825719120135, + "grad_norm": 8.4375, + "learning_rate": 3.89213669472249e-06, + "loss": 2.8261, + "step": 15728 + }, + { + "epoch": 1.6633883248730963, + "grad_norm": 8.0625, + "learning_rate": 3.891994444335776e-06, + "loss": 2.6115, + "step": 15729 + }, + { + "epoch": 1.6634940778341794, + "grad_norm": 8.25, + "learning_rate": 3.891852187417001e-06, + "loss": 2.5629, + "step": 15730 + }, + { + "epoch": 1.6635998307952624, + "grad_norm": 8.0625, + "learning_rate": 3.891709923966834e-06, + "loss": 2.5885, + "step": 15731 + }, + { + "epoch": 1.6637055837563453, + "grad_norm": 8.4375, + "learning_rate": 3.89156765398594e-06, + "loss": 2.4864, + "step": 15732 + }, + { + "epoch": 1.663811336717428, + "grad_norm": 8.375, + "learning_rate": 3.8914253774749896e-06, + "loss": 2.6748, + "step": 15733 + }, + { + "epoch": 1.663917089678511, + "grad_norm": 8.75, + "learning_rate": 3.891283094434648e-06, + "loss": 2.6319, + "step": 15734 + }, + { + "epoch": 1.6640228426395938, + "grad_norm": 8.375, + "learning_rate": 3.891140804865584e-06, + "loss": 2.6698, + "step": 15735 + }, + { + "epoch": 1.6641285956006768, + "grad_norm": 8.5625, + "learning_rate": 3.890998508768465e-06, + "loss": 2.4224, + "step": 15736 + }, + { + "epoch": 1.66423434856176, + "grad_norm": 7.9375, + "learning_rate": 3.89085620614396e-06, + "loss": 2.599, + "step": 15737 + }, + { + "epoch": 1.6643401015228427, + "grad_norm": 8.5, + "learning_rate": 3.890713896992735e-06, + "loss": 2.582, + "step": 15738 + }, + { + "epoch": 1.6644458544839256, + "grad_norm": 8.1875, + "learning_rate": 3.8905715813154586e-06, + "loss": 2.5452, + "step": 15739 + }, + { + "epoch": 1.6645516074450084, + "grad_norm": 8.0, + "learning_rate": 3.890429259112798e-06, + "loss": 2.3613, + "step": 15740 + }, + { + "epoch": 1.6646573604060912, + "grad_norm": 9.0, + "learning_rate": 3.890286930385422e-06, + "loss": 2.7115, + "step": 15741 + }, + { + "epoch": 1.6647631133671743, + "grad_norm": 9.0625, + "learning_rate": 3.8901445951339976e-06, + "loss": 2.6167, + "step": 15742 + }, + { + "epoch": 1.6648688663282571, + "grad_norm": 8.3125, + "learning_rate": 3.890002253359193e-06, + "loss": 2.6582, + "step": 15743 + }, + { + "epoch": 1.6649746192893402, + "grad_norm": 9.3125, + "learning_rate": 3.889859905061677e-06, + "loss": 2.8206, + "step": 15744 + }, + { + "epoch": 1.665080372250423, + "grad_norm": 8.9375, + "learning_rate": 3.889717550242117e-06, + "loss": 2.3729, + "step": 15745 + }, + { + "epoch": 1.6651861252115059, + "grad_norm": 8.875, + "learning_rate": 3.88957518890118e-06, + "loss": 2.6834, + "step": 15746 + }, + { + "epoch": 1.6652918781725887, + "grad_norm": 8.0625, + "learning_rate": 3.889432821039536e-06, + "loss": 2.3995, + "step": 15747 + }, + { + "epoch": 1.6653976311336718, + "grad_norm": 8.0625, + "learning_rate": 3.889290446657851e-06, + "loss": 2.6416, + "step": 15748 + }, + { + "epoch": 1.6655033840947546, + "grad_norm": 9.125, + "learning_rate": 3.889148065756795e-06, + "loss": 2.3874, + "step": 15749 + }, + { + "epoch": 1.6656091370558377, + "grad_norm": 7.8125, + "learning_rate": 3.889005678337035e-06, + "loss": 2.3387, + "step": 15750 + }, + { + "epoch": 1.6657148900169205, + "grad_norm": 8.125, + "learning_rate": 3.888863284399239e-06, + "loss": 2.6468, + "step": 15751 + }, + { + "epoch": 1.6658206429780034, + "grad_norm": 10.125, + "learning_rate": 3.888720883944077e-06, + "loss": 2.634, + "step": 15752 + }, + { + "epoch": 1.6659263959390862, + "grad_norm": 7.90625, + "learning_rate": 3.888578476972214e-06, + "loss": 2.5903, + "step": 15753 + }, + { + "epoch": 1.666032148900169, + "grad_norm": 8.0, + "learning_rate": 3.888436063484321e-06, + "loss": 2.4886, + "step": 15754 + }, + { + "epoch": 1.666137901861252, + "grad_norm": 8.875, + "learning_rate": 3.888293643481065e-06, + "loss": 2.4622, + "step": 15755 + }, + { + "epoch": 1.6662436548223352, + "grad_norm": 10.4375, + "learning_rate": 3.888151216963115e-06, + "loss": 2.2605, + "step": 15756 + }, + { + "epoch": 1.666349407783418, + "grad_norm": 9.3125, + "learning_rate": 3.88800878393114e-06, + "loss": 2.8209, + "step": 15757 + }, + { + "epoch": 1.6664551607445008, + "grad_norm": 8.5625, + "learning_rate": 3.887866344385807e-06, + "loss": 2.6006, + "step": 15758 + }, + { + "epoch": 1.6665609137055837, + "grad_norm": 8.0625, + "learning_rate": 3.887723898327784e-06, + "loss": 2.6654, + "step": 15759 + }, + { + "epoch": 1.6666666666666665, + "grad_norm": 8.5, + "learning_rate": 3.887581445757741e-06, + "loss": 2.5311, + "step": 15760 + }, + { + "epoch": 1.6667724196277496, + "grad_norm": 8.3125, + "learning_rate": 3.887438986676345e-06, + "loss": 2.5698, + "step": 15761 + }, + { + "epoch": 1.6668781725888326, + "grad_norm": 8.5625, + "learning_rate": 3.887296521084266e-06, + "loss": 2.5796, + "step": 15762 + }, + { + "epoch": 1.6669839255499155, + "grad_norm": 8.625, + "learning_rate": 3.887154048982171e-06, + "loss": 2.8058, + "step": 15763 + }, + { + "epoch": 1.6670896785109983, + "grad_norm": 8.0625, + "learning_rate": 3.887011570370731e-06, + "loss": 2.453, + "step": 15764 + }, + { + "epoch": 1.6671954314720812, + "grad_norm": 7.96875, + "learning_rate": 3.886869085250612e-06, + "loss": 2.5052, + "step": 15765 + }, + { + "epoch": 1.667301184433164, + "grad_norm": 8.5, + "learning_rate": 3.8867265936224835e-06, + "loss": 2.7073, + "step": 15766 + }, + { + "epoch": 1.667406937394247, + "grad_norm": 8.125, + "learning_rate": 3.886584095487014e-06, + "loss": 2.4132, + "step": 15767 + }, + { + "epoch": 1.66751269035533, + "grad_norm": 8.5625, + "learning_rate": 3.886441590844873e-06, + "loss": 2.7495, + "step": 15768 + }, + { + "epoch": 1.667618443316413, + "grad_norm": 8.375, + "learning_rate": 3.886299079696728e-06, + "loss": 2.6264, + "step": 15769 + }, + { + "epoch": 1.6677241962774958, + "grad_norm": 8.375, + "learning_rate": 3.886156562043249e-06, + "loss": 2.5862, + "step": 15770 + }, + { + "epoch": 1.6678299492385786, + "grad_norm": 8.4375, + "learning_rate": 3.886014037885104e-06, + "loss": 2.6124, + "step": 15771 + }, + { + "epoch": 1.6679357021996615, + "grad_norm": 8.875, + "learning_rate": 3.8858715072229615e-06, + "loss": 2.9861, + "step": 15772 + }, + { + "epoch": 1.6680414551607445, + "grad_norm": 8.125, + "learning_rate": 3.885728970057492e-06, + "loss": 2.4935, + "step": 15773 + }, + { + "epoch": 1.6681472081218274, + "grad_norm": 7.96875, + "learning_rate": 3.885586426389363e-06, + "loss": 2.4712, + "step": 15774 + }, + { + "epoch": 1.6682529610829104, + "grad_norm": 8.3125, + "learning_rate": 3.885443876219242e-06, + "loss": 2.5693, + "step": 15775 + }, + { + "epoch": 1.6683587140439933, + "grad_norm": 8.0, + "learning_rate": 3.885301319547801e-06, + "loss": 2.455, + "step": 15776 + }, + { + "epoch": 1.6684644670050761, + "grad_norm": 7.875, + "learning_rate": 3.885158756375708e-06, + "loss": 2.4931, + "step": 15777 + }, + { + "epoch": 1.668570219966159, + "grad_norm": 8.3125, + "learning_rate": 3.8850161867036296e-06, + "loss": 2.7607, + "step": 15778 + }, + { + "epoch": 1.668675972927242, + "grad_norm": 8.1875, + "learning_rate": 3.884873610532237e-06, + "loss": 2.1373, + "step": 15779 + }, + { + "epoch": 1.6687817258883249, + "grad_norm": 8.375, + "learning_rate": 3.884731027862199e-06, + "loss": 2.4418, + "step": 15780 + }, + { + "epoch": 1.668887478849408, + "grad_norm": 8.5625, + "learning_rate": 3.8845884386941865e-06, + "loss": 2.4951, + "step": 15781 + }, + { + "epoch": 1.6689932318104908, + "grad_norm": 7.625, + "learning_rate": 3.884445843028866e-06, + "loss": 2.4358, + "step": 15782 + }, + { + "epoch": 1.6690989847715736, + "grad_norm": 8.5, + "learning_rate": 3.884303240866905e-06, + "loss": 2.6726, + "step": 15783 + }, + { + "epoch": 1.6692047377326564, + "grad_norm": 8.0, + "learning_rate": 3.884160632208978e-06, + "loss": 2.7439, + "step": 15784 + }, + { + "epoch": 1.6693104906937393, + "grad_norm": 8.0625, + "learning_rate": 3.88401801705575e-06, + "loss": 2.4171, + "step": 15785 + }, + { + "epoch": 1.6694162436548223, + "grad_norm": 9.0, + "learning_rate": 3.883875395407892e-06, + "loss": 2.7273, + "step": 15786 + }, + { + "epoch": 1.6695219966159054, + "grad_norm": 8.1875, + "learning_rate": 3.883732767266072e-06, + "loss": 2.6066, + "step": 15787 + }, + { + "epoch": 1.6696277495769882, + "grad_norm": 7.9375, + "learning_rate": 3.88359013263096e-06, + "loss": 2.3494, + "step": 15788 + }, + { + "epoch": 1.669733502538071, + "grad_norm": 7.78125, + "learning_rate": 3.883447491503226e-06, + "loss": 2.5198, + "step": 15789 + }, + { + "epoch": 1.669839255499154, + "grad_norm": 8.5625, + "learning_rate": 3.883304843883539e-06, + "loss": 2.5634, + "step": 15790 + }, + { + "epoch": 1.6699450084602367, + "grad_norm": 8.125, + "learning_rate": 3.883162189772567e-06, + "loss": 2.4974, + "step": 15791 + }, + { + "epoch": 1.6700507614213198, + "grad_norm": 8.4375, + "learning_rate": 3.883019529170981e-06, + "loss": 2.6642, + "step": 15792 + }, + { + "epoch": 1.6701565143824029, + "grad_norm": 8.375, + "learning_rate": 3.882876862079451e-06, + "loss": 2.6873, + "step": 15793 + }, + { + "epoch": 1.6702622673434857, + "grad_norm": 8.5625, + "learning_rate": 3.882734188498644e-06, + "loss": 2.5706, + "step": 15794 + }, + { + "epoch": 1.6703680203045685, + "grad_norm": 8.25, + "learning_rate": 3.882591508429232e-06, + "loss": 2.6915, + "step": 15795 + }, + { + "epoch": 1.6704737732656514, + "grad_norm": 8.1875, + "learning_rate": 3.8824488218718835e-06, + "loss": 2.4742, + "step": 15796 + }, + { + "epoch": 1.6705795262267342, + "grad_norm": 8.625, + "learning_rate": 3.882306128827267e-06, + "loss": 2.5424, + "step": 15797 + }, + { + "epoch": 1.6706852791878173, + "grad_norm": 8.1875, + "learning_rate": 3.882163429296053e-06, + "loss": 2.6119, + "step": 15798 + }, + { + "epoch": 1.6707910321489001, + "grad_norm": 7.8125, + "learning_rate": 3.882020723278913e-06, + "loss": 2.5311, + "step": 15799 + }, + { + "epoch": 1.6708967851099832, + "grad_norm": 8.125, + "learning_rate": 3.881878010776514e-06, + "loss": 2.4674, + "step": 15800 + }, + { + "epoch": 1.671002538071066, + "grad_norm": 8.375, + "learning_rate": 3.881735291789527e-06, + "loss": 2.4781, + "step": 15801 + }, + { + "epoch": 1.6711082910321489, + "grad_norm": 8.375, + "learning_rate": 3.881592566318621e-06, + "loss": 2.4633, + "step": 15802 + }, + { + "epoch": 1.6712140439932317, + "grad_norm": 8.0625, + "learning_rate": 3.881449834364467e-06, + "loss": 2.495, + "step": 15803 + }, + { + "epoch": 1.6713197969543148, + "grad_norm": 9.0625, + "learning_rate": 3.881307095927734e-06, + "loss": 2.5248, + "step": 15804 + }, + { + "epoch": 1.6714255499153976, + "grad_norm": 8.0, + "learning_rate": 3.8811643510090905e-06, + "loss": 2.5492, + "step": 15805 + }, + { + "epoch": 1.6715313028764807, + "grad_norm": 7.9375, + "learning_rate": 3.881021599609207e-06, + "loss": 2.5738, + "step": 15806 + }, + { + "epoch": 1.6716370558375635, + "grad_norm": 8.4375, + "learning_rate": 3.880878841728755e-06, + "loss": 2.3632, + "step": 15807 + }, + { + "epoch": 1.6717428087986463, + "grad_norm": 8.5625, + "learning_rate": 3.880736077368404e-06, + "loss": 2.5265, + "step": 15808 + }, + { + "epoch": 1.6718485617597292, + "grad_norm": 8.3125, + "learning_rate": 3.880593306528823e-06, + "loss": 2.5686, + "step": 15809 + }, + { + "epoch": 1.671954314720812, + "grad_norm": 7.875, + "learning_rate": 3.880450529210681e-06, + "loss": 2.438, + "step": 15810 + }, + { + "epoch": 1.672060067681895, + "grad_norm": 8.375, + "learning_rate": 3.88030774541465e-06, + "loss": 2.4564, + "step": 15811 + }, + { + "epoch": 1.6721658206429781, + "grad_norm": 7.8125, + "learning_rate": 3.8801649551414e-06, + "loss": 2.309, + "step": 15812 + }, + { + "epoch": 1.672271573604061, + "grad_norm": 8.375, + "learning_rate": 3.8800221583916e-06, + "loss": 2.4915, + "step": 15813 + }, + { + "epoch": 1.6723773265651438, + "grad_norm": 8.375, + "learning_rate": 3.87987935516592e-06, + "loss": 2.714, + "step": 15814 + }, + { + "epoch": 1.6724830795262267, + "grad_norm": 8.25, + "learning_rate": 3.879736545465031e-06, + "loss": 2.6706, + "step": 15815 + }, + { + "epoch": 1.6725888324873095, + "grad_norm": 8.3125, + "learning_rate": 3.879593729289602e-06, + "loss": 2.286, + "step": 15816 + }, + { + "epoch": 1.6726945854483926, + "grad_norm": 8.25, + "learning_rate": 3.879450906640304e-06, + "loss": 2.7019, + "step": 15817 + }, + { + "epoch": 1.6728003384094756, + "grad_norm": 8.625, + "learning_rate": 3.879308077517807e-06, + "loss": 2.4216, + "step": 15818 + }, + { + "epoch": 1.6729060913705585, + "grad_norm": 8.25, + "learning_rate": 3.879165241922782e-06, + "loss": 2.8015, + "step": 15819 + }, + { + "epoch": 1.6730118443316413, + "grad_norm": 8.375, + "learning_rate": 3.879022399855898e-06, + "loss": 2.4206, + "step": 15820 + }, + { + "epoch": 1.6731175972927241, + "grad_norm": 8.4375, + "learning_rate": 3.8788795513178265e-06, + "loss": 2.5043, + "step": 15821 + }, + { + "epoch": 1.673223350253807, + "grad_norm": 8.375, + "learning_rate": 3.878736696309237e-06, + "loss": 2.4896, + "step": 15822 + }, + { + "epoch": 1.67332910321489, + "grad_norm": 8.625, + "learning_rate": 3.8785938348307994e-06, + "loss": 2.7319, + "step": 15823 + }, + { + "epoch": 1.6734348561759729, + "grad_norm": 8.6875, + "learning_rate": 3.878450966883186e-06, + "loss": 2.5629, + "step": 15824 + }, + { + "epoch": 1.673540609137056, + "grad_norm": 8.6875, + "learning_rate": 3.878308092467065e-06, + "loss": 2.5304, + "step": 15825 + }, + { + "epoch": 1.6736463620981388, + "grad_norm": 8.125, + "learning_rate": 3.878165211583108e-06, + "loss": 2.6012, + "step": 15826 + }, + { + "epoch": 1.6737521150592216, + "grad_norm": 8.25, + "learning_rate": 3.8780223242319854e-06, + "loss": 2.3745, + "step": 15827 + }, + { + "epoch": 1.6738578680203045, + "grad_norm": 8.0, + "learning_rate": 3.8778794304143676e-06, + "loss": 2.5523, + "step": 15828 + }, + { + "epoch": 1.6739636209813875, + "grad_norm": 8.3125, + "learning_rate": 3.877736530130925e-06, + "loss": 2.6393, + "step": 15829 + }, + { + "epoch": 1.6740693739424704, + "grad_norm": 8.0625, + "learning_rate": 3.877593623382329e-06, + "loss": 2.4596, + "step": 15830 + }, + { + "epoch": 1.6741751269035534, + "grad_norm": 7.78125, + "learning_rate": 3.877450710169249e-06, + "loss": 2.32, + "step": 15831 + }, + { + "epoch": 1.6742808798646363, + "grad_norm": 8.0, + "learning_rate": 3.877307790492356e-06, + "loss": 2.5852, + "step": 15832 + }, + { + "epoch": 1.674386632825719, + "grad_norm": 8.25, + "learning_rate": 3.877164864352321e-06, + "loss": 2.5292, + "step": 15833 + }, + { + "epoch": 1.674492385786802, + "grad_norm": 7.75, + "learning_rate": 3.877021931749814e-06, + "loss": 2.4608, + "step": 15834 + }, + { + "epoch": 1.674598138747885, + "grad_norm": 8.625, + "learning_rate": 3.876878992685508e-06, + "loss": 2.3806, + "step": 15835 + }, + { + "epoch": 1.6747038917089678, + "grad_norm": 8.875, + "learning_rate": 3.87673604716007e-06, + "loss": 2.6192, + "step": 15836 + }, + { + "epoch": 1.674809644670051, + "grad_norm": 8.5, + "learning_rate": 3.876593095174174e-06, + "loss": 2.9698, + "step": 15837 + }, + { + "epoch": 1.6749153976311337, + "grad_norm": 7.9375, + "learning_rate": 3.876450136728489e-06, + "loss": 2.6137, + "step": 15838 + }, + { + "epoch": 1.6750211505922166, + "grad_norm": 7.9375, + "learning_rate": 3.8763071718236865e-06, + "loss": 2.5753, + "step": 15839 + }, + { + "epoch": 1.6751269035532994, + "grad_norm": 8.0, + "learning_rate": 3.876164200460439e-06, + "loss": 2.3, + "step": 15840 + }, + { + "epoch": 1.6752326565143822, + "grad_norm": 8.1875, + "learning_rate": 3.876021222639413e-06, + "loss": 2.5402, + "step": 15841 + }, + { + "epoch": 1.6753384094754653, + "grad_norm": 7.6875, + "learning_rate": 3.875878238361284e-06, + "loss": 2.3514, + "step": 15842 + }, + { + "epoch": 1.6754441624365484, + "grad_norm": 8.375, + "learning_rate": 3.875735247626721e-06, + "loss": 2.3329, + "step": 15843 + }, + { + "epoch": 1.6755499153976312, + "grad_norm": 8.6875, + "learning_rate": 3.875592250436394e-06, + "loss": 2.5418, + "step": 15844 + }, + { + "epoch": 1.675655668358714, + "grad_norm": 8.4375, + "learning_rate": 3.875449246790976e-06, + "loss": 2.4476, + "step": 15845 + }, + { + "epoch": 1.6757614213197969, + "grad_norm": 8.375, + "learning_rate": 3.8753062366911375e-06, + "loss": 2.6972, + "step": 15846 + }, + { + "epoch": 1.6758671742808797, + "grad_norm": 8.25, + "learning_rate": 3.875163220137549e-06, + "loss": 2.546, + "step": 15847 + }, + { + "epoch": 1.6759729272419628, + "grad_norm": 8.125, + "learning_rate": 3.8750201971308815e-06, + "loss": 2.5489, + "step": 15848 + }, + { + "epoch": 1.6760786802030458, + "grad_norm": 8.375, + "learning_rate": 3.874877167671807e-06, + "loss": 2.5382, + "step": 15849 + }, + { + "epoch": 1.6761844331641287, + "grad_norm": 8.3125, + "learning_rate": 3.874734131760996e-06, + "loss": 2.6672, + "step": 15850 + }, + { + "epoch": 1.6762901861252115, + "grad_norm": 8.625, + "learning_rate": 3.87459108939912e-06, + "loss": 2.5752, + "step": 15851 + }, + { + "epoch": 1.6763959390862944, + "grad_norm": 8.4375, + "learning_rate": 3.874448040586851e-06, + "loss": 2.5454, + "step": 15852 + }, + { + "epoch": 1.6765016920473772, + "grad_norm": 7.71875, + "learning_rate": 3.874304985324858e-06, + "loss": 2.5701, + "step": 15853 + }, + { + "epoch": 1.6766074450084603, + "grad_norm": 8.375, + "learning_rate": 3.874161923613815e-06, + "loss": 2.2701, + "step": 15854 + }, + { + "epoch": 1.676713197969543, + "grad_norm": 8.1875, + "learning_rate": 3.8740188554543925e-06, + "loss": 2.7692, + "step": 15855 + }, + { + "epoch": 1.6768189509306262, + "grad_norm": 8.375, + "learning_rate": 3.87387578084726e-06, + "loss": 2.62, + "step": 15856 + }, + { + "epoch": 1.676924703891709, + "grad_norm": 8.8125, + "learning_rate": 3.873732699793091e-06, + "loss": 2.7091, + "step": 15857 + }, + { + "epoch": 1.6770304568527918, + "grad_norm": 8.9375, + "learning_rate": 3.873589612292557e-06, + "loss": 2.5814, + "step": 15858 + }, + { + "epoch": 1.6771362098138747, + "grad_norm": 8.5, + "learning_rate": 3.873446518346328e-06, + "loss": 2.4426, + "step": 15859 + }, + { + "epoch": 1.6772419627749577, + "grad_norm": 8.25, + "learning_rate": 3.873303417955076e-06, + "loss": 2.5076, + "step": 15860 + }, + { + "epoch": 1.6773477157360406, + "grad_norm": 8.25, + "learning_rate": 3.873160311119473e-06, + "loss": 2.5934, + "step": 15861 + }, + { + "epoch": 1.6774534686971236, + "grad_norm": 8.3125, + "learning_rate": 3.873017197840191e-06, + "loss": 2.6465, + "step": 15862 + }, + { + "epoch": 1.6775592216582065, + "grad_norm": 8.125, + "learning_rate": 3.872874078117899e-06, + "loss": 2.5026, + "step": 15863 + }, + { + "epoch": 1.6776649746192893, + "grad_norm": 8.625, + "learning_rate": 3.8727309519532716e-06, + "loss": 2.1848, + "step": 15864 + }, + { + "epoch": 1.6777707275803722, + "grad_norm": 7.71875, + "learning_rate": 3.87258781934698e-06, + "loss": 2.429, + "step": 15865 + }, + { + "epoch": 1.677876480541455, + "grad_norm": 7.84375, + "learning_rate": 3.872444680299694e-06, + "loss": 2.6289, + "step": 15866 + }, + { + "epoch": 1.677982233502538, + "grad_norm": 8.0, + "learning_rate": 3.872301534812088e-06, + "loss": 2.3088, + "step": 15867 + }, + { + "epoch": 1.6780879864636211, + "grad_norm": 8.625, + "learning_rate": 3.87215838288483e-06, + "loss": 2.8197, + "step": 15868 + }, + { + "epoch": 1.678193739424704, + "grad_norm": 8.3125, + "learning_rate": 3.872015224518595e-06, + "loss": 2.5735, + "step": 15869 + }, + { + "epoch": 1.6782994923857868, + "grad_norm": 8.3125, + "learning_rate": 3.8718720597140536e-06, + "loss": 2.6885, + "step": 15870 + }, + { + "epoch": 1.6784052453468696, + "grad_norm": 7.9375, + "learning_rate": 3.871728888471877e-06, + "loss": 2.4475, + "step": 15871 + }, + { + "epoch": 1.6785109983079525, + "grad_norm": 8.4375, + "learning_rate": 3.8715857107927385e-06, + "loss": 2.3427, + "step": 15872 + }, + { + "epoch": 1.6786167512690355, + "grad_norm": 8.625, + "learning_rate": 3.871442526677309e-06, + "loss": 2.6143, + "step": 15873 + }, + { + "epoch": 1.6787225042301186, + "grad_norm": 8.3125, + "learning_rate": 3.871299336126261e-06, + "loss": 2.5032, + "step": 15874 + }, + { + "epoch": 1.6788282571912014, + "grad_norm": 8.125, + "learning_rate": 3.871156139140266e-06, + "loss": 2.5395, + "step": 15875 + }, + { + "epoch": 1.6789340101522843, + "grad_norm": 8.1875, + "learning_rate": 3.871012935719996e-06, + "loss": 2.5885, + "step": 15876 + }, + { + "epoch": 1.6790397631133671, + "grad_norm": 8.1875, + "learning_rate": 3.870869725866122e-06, + "loss": 2.5557, + "step": 15877 + }, + { + "epoch": 1.67914551607445, + "grad_norm": 7.90625, + "learning_rate": 3.8707265095793176e-06, + "loss": 2.4792, + "step": 15878 + }, + { + "epoch": 1.679251269035533, + "grad_norm": 8.625, + "learning_rate": 3.870583286860255e-06, + "loss": 2.5057, + "step": 15879 + }, + { + "epoch": 1.6793570219966159, + "grad_norm": 8.6875, + "learning_rate": 3.870440057709605e-06, + "loss": 2.5684, + "step": 15880 + }, + { + "epoch": 1.679462774957699, + "grad_norm": 8.75, + "learning_rate": 3.870296822128041e-06, + "loss": 2.9128, + "step": 15881 + }, + { + "epoch": 1.6795685279187818, + "grad_norm": 8.375, + "learning_rate": 3.870153580116234e-06, + "loss": 2.7654, + "step": 15882 + }, + { + "epoch": 1.6796742808798646, + "grad_norm": 8.0625, + "learning_rate": 3.870010331674856e-06, + "loss": 2.3944, + "step": 15883 + }, + { + "epoch": 1.6797800338409474, + "grad_norm": 8.1875, + "learning_rate": 3.869867076804581e-06, + "loss": 1.9332, + "step": 15884 + }, + { + "epoch": 1.6798857868020305, + "grad_norm": 7.875, + "learning_rate": 3.8697238155060785e-06, + "loss": 2.401, + "step": 15885 + }, + { + "epoch": 1.6799915397631133, + "grad_norm": 9.1875, + "learning_rate": 3.869580547780023e-06, + "loss": 2.8747, + "step": 15886 + }, + { + "epoch": 1.6800972927241964, + "grad_norm": 8.25, + "learning_rate": 3.869437273627087e-06, + "loss": 2.5355, + "step": 15887 + }, + { + "epoch": 1.6802030456852792, + "grad_norm": 8.625, + "learning_rate": 3.869293993047941e-06, + "loss": 2.6555, + "step": 15888 + }, + { + "epoch": 1.680308798646362, + "grad_norm": 8.4375, + "learning_rate": 3.869150706043259e-06, + "loss": 2.0855, + "step": 15889 + }, + { + "epoch": 1.680414551607445, + "grad_norm": 8.0, + "learning_rate": 3.8690074126137126e-06, + "loss": 2.486, + "step": 15890 + }, + { + "epoch": 1.680520304568528, + "grad_norm": 8.3125, + "learning_rate": 3.868864112759974e-06, + "loss": 2.5209, + "step": 15891 + }, + { + "epoch": 1.6806260575296108, + "grad_norm": 8.0, + "learning_rate": 3.868720806482716e-06, + "loss": 2.4151, + "step": 15892 + }, + { + "epoch": 1.6807318104906939, + "grad_norm": 8.625, + "learning_rate": 3.86857749378261e-06, + "loss": 2.8067, + "step": 15893 + }, + { + "epoch": 1.6808375634517767, + "grad_norm": 8.375, + "learning_rate": 3.868434174660331e-06, + "loss": 2.6661, + "step": 15894 + }, + { + "epoch": 1.6809433164128595, + "grad_norm": 7.96875, + "learning_rate": 3.868290849116549e-06, + "loss": 2.4502, + "step": 15895 + }, + { + "epoch": 1.6810490693739424, + "grad_norm": 8.25, + "learning_rate": 3.868147517151939e-06, + "loss": 2.5095, + "step": 15896 + }, + { + "epoch": 1.6811548223350252, + "grad_norm": 8.125, + "learning_rate": 3.868004178767172e-06, + "loss": 2.495, + "step": 15897 + }, + { + "epoch": 1.6812605752961083, + "grad_norm": 8.5625, + "learning_rate": 3.86786083396292e-06, + "loss": 2.5064, + "step": 15898 + }, + { + "epoch": 1.6813663282571913, + "grad_norm": 8.0, + "learning_rate": 3.867717482739857e-06, + "loss": 2.4584, + "step": 15899 + }, + { + "epoch": 1.6814720812182742, + "grad_norm": 8.0625, + "learning_rate": 3.867574125098656e-06, + "loss": 2.6139, + "step": 15900 + }, + { + "epoch": 1.681577834179357, + "grad_norm": 8.0, + "learning_rate": 3.867430761039988e-06, + "loss": 2.586, + "step": 15901 + }, + { + "epoch": 1.6816835871404399, + "grad_norm": 7.96875, + "learning_rate": 3.867287390564527e-06, + "loss": 2.4271, + "step": 15902 + }, + { + "epoch": 1.6817893401015227, + "grad_norm": 7.96875, + "learning_rate": 3.867144013672945e-06, + "loss": 2.4984, + "step": 15903 + }, + { + "epoch": 1.6818950930626058, + "grad_norm": 8.5625, + "learning_rate": 3.867000630365916e-06, + "loss": 2.7483, + "step": 15904 + }, + { + "epoch": 1.6820008460236888, + "grad_norm": 8.0625, + "learning_rate": 3.866857240644112e-06, + "loss": 2.4472, + "step": 15905 + }, + { + "epoch": 1.6821065989847717, + "grad_norm": 8.4375, + "learning_rate": 3.866713844508205e-06, + "loss": 2.7127, + "step": 15906 + }, + { + "epoch": 1.6822123519458545, + "grad_norm": 8.75, + "learning_rate": 3.86657044195887e-06, + "loss": 2.6209, + "step": 15907 + }, + { + "epoch": 1.6823181049069373, + "grad_norm": 8.375, + "learning_rate": 3.866427032996779e-06, + "loss": 2.4478, + "step": 15908 + }, + { + "epoch": 1.6824238578680202, + "grad_norm": 8.0, + "learning_rate": 3.866283617622604e-06, + "loss": 2.6245, + "step": 15909 + }, + { + "epoch": 1.6825296108291032, + "grad_norm": 8.8125, + "learning_rate": 3.86614019583702e-06, + "loss": 2.5958, + "step": 15910 + }, + { + "epoch": 1.682635363790186, + "grad_norm": 7.8125, + "learning_rate": 3.865996767640697e-06, + "loss": 2.4902, + "step": 15911 + }, + { + "epoch": 1.6827411167512691, + "grad_norm": 7.9375, + "learning_rate": 3.8658533330343105e-06, + "loss": 2.5175, + "step": 15912 + }, + { + "epoch": 1.682846869712352, + "grad_norm": 8.1875, + "learning_rate": 3.865709892018533e-06, + "loss": 2.6464, + "step": 15913 + }, + { + "epoch": 1.6829526226734348, + "grad_norm": 8.4375, + "learning_rate": 3.8655664445940375e-06, + "loss": 2.6918, + "step": 15914 + }, + { + "epoch": 1.6830583756345177, + "grad_norm": 7.8125, + "learning_rate": 3.865422990761498e-06, + "loss": 2.4705, + "step": 15915 + }, + { + "epoch": 1.6831641285956007, + "grad_norm": 8.25, + "learning_rate": 3.865279530521586e-06, + "loss": 2.642, + "step": 15916 + }, + { + "epoch": 1.6832698815566836, + "grad_norm": 8.0625, + "learning_rate": 3.865136063874975e-06, + "loss": 2.4502, + "step": 15917 + }, + { + "epoch": 1.6833756345177666, + "grad_norm": 8.5625, + "learning_rate": 3.86499259082234e-06, + "loss": 2.5541, + "step": 15918 + }, + { + "epoch": 1.6834813874788495, + "grad_norm": 8.5625, + "learning_rate": 3.864849111364353e-06, + "loss": 2.5753, + "step": 15919 + }, + { + "epoch": 1.6835871404399323, + "grad_norm": 8.5625, + "learning_rate": 3.864705625501687e-06, + "loss": 2.6027, + "step": 15920 + }, + { + "epoch": 1.6836928934010151, + "grad_norm": 8.4375, + "learning_rate": 3.864562133235014e-06, + "loss": 2.4258, + "step": 15921 + }, + { + "epoch": 1.683798646362098, + "grad_norm": 8.6875, + "learning_rate": 3.864418634565011e-06, + "loss": 2.6316, + "step": 15922 + }, + { + "epoch": 1.683904399323181, + "grad_norm": 8.5, + "learning_rate": 3.864275129492349e-06, + "loss": 2.581, + "step": 15923 + }, + { + "epoch": 1.684010152284264, + "grad_norm": 8.625, + "learning_rate": 3.864131618017702e-06, + "loss": 2.3764, + "step": 15924 + }, + { + "epoch": 1.684115905245347, + "grad_norm": 8.25, + "learning_rate": 3.863988100141743e-06, + "loss": 2.3123, + "step": 15925 + }, + { + "epoch": 1.6842216582064298, + "grad_norm": 7.875, + "learning_rate": 3.863844575865146e-06, + "loss": 2.4695, + "step": 15926 + }, + { + "epoch": 1.6843274111675126, + "grad_norm": 7.78125, + "learning_rate": 3.863701045188584e-06, + "loss": 2.4321, + "step": 15927 + }, + { + "epoch": 1.6844331641285955, + "grad_norm": 8.25, + "learning_rate": 3.863557508112729e-06, + "loss": 2.5766, + "step": 15928 + }, + { + "epoch": 1.6845389170896785, + "grad_norm": 8.375, + "learning_rate": 3.863413964638258e-06, + "loss": 2.4731, + "step": 15929 + }, + { + "epoch": 1.6846446700507616, + "grad_norm": 8.4375, + "learning_rate": 3.863270414765843e-06, + "loss": 2.5852, + "step": 15930 + }, + { + "epoch": 1.6847504230118444, + "grad_norm": 8.4375, + "learning_rate": 3.8631268584961565e-06, + "loss": 2.4755, + "step": 15931 + }, + { + "epoch": 1.6848561759729273, + "grad_norm": 8.1875, + "learning_rate": 3.862983295829875e-06, + "loss": 2.6016, + "step": 15932 + }, + { + "epoch": 1.68496192893401, + "grad_norm": 8.8125, + "learning_rate": 3.8628397267676685e-06, + "loss": 2.6276, + "step": 15933 + }, + { + "epoch": 1.685067681895093, + "grad_norm": 7.78125, + "learning_rate": 3.8626961513102135e-06, + "loss": 2.3867, + "step": 15934 + }, + { + "epoch": 1.685173434856176, + "grad_norm": 8.625, + "learning_rate": 3.8625525694581824e-06, + "loss": 2.7893, + "step": 15935 + }, + { + "epoch": 1.6852791878172588, + "grad_norm": 8.25, + "learning_rate": 3.86240898121225e-06, + "loss": 2.4604, + "step": 15936 + }, + { + "epoch": 1.685384940778342, + "grad_norm": 8.8125, + "learning_rate": 3.862265386573089e-06, + "loss": 2.4435, + "step": 15937 + }, + { + "epoch": 1.6854906937394247, + "grad_norm": 9.6875, + "learning_rate": 3.862121785541373e-06, + "loss": 2.6365, + "step": 15938 + }, + { + "epoch": 1.6855964467005076, + "grad_norm": 8.125, + "learning_rate": 3.861978178117776e-06, + "loss": 2.5647, + "step": 15939 + }, + { + "epoch": 1.6857021996615904, + "grad_norm": 8.625, + "learning_rate": 3.861834564302975e-06, + "loss": 2.5603, + "step": 15940 + }, + { + "epoch": 1.6858079526226735, + "grad_norm": 8.25, + "learning_rate": 3.86169094409764e-06, + "loss": 2.5063, + "step": 15941 + }, + { + "epoch": 1.6859137055837563, + "grad_norm": 8.1875, + "learning_rate": 3.861547317502446e-06, + "loss": 2.5048, + "step": 15942 + }, + { + "epoch": 1.6860194585448394, + "grad_norm": 8.9375, + "learning_rate": 3.861403684518068e-06, + "loss": 2.4958, + "step": 15943 + }, + { + "epoch": 1.6861252115059222, + "grad_norm": 8.625, + "learning_rate": 3.8612600451451785e-06, + "loss": 2.6911, + "step": 15944 + }, + { + "epoch": 1.686230964467005, + "grad_norm": 8.875, + "learning_rate": 3.861116399384452e-06, + "loss": 2.625, + "step": 15945 + }, + { + "epoch": 1.6863367174280879, + "grad_norm": 7.875, + "learning_rate": 3.860972747236565e-06, + "loss": 2.3836, + "step": 15946 + }, + { + "epoch": 1.686442470389171, + "grad_norm": 7.875, + "learning_rate": 3.860829088702187e-06, + "loss": 2.2589, + "step": 15947 + }, + { + "epoch": 1.6865482233502538, + "grad_norm": 8.125, + "learning_rate": 3.8606854237819955e-06, + "loss": 2.366, + "step": 15948 + }, + { + "epoch": 1.6866539763113368, + "grad_norm": 8.6875, + "learning_rate": 3.860541752476665e-06, + "loss": 2.3816, + "step": 15949 + }, + { + "epoch": 1.6867597292724197, + "grad_norm": 8.375, + "learning_rate": 3.860398074786867e-06, + "loss": 2.3261, + "step": 15950 + }, + { + "epoch": 1.6868654822335025, + "grad_norm": 8.25, + "learning_rate": 3.860254390713278e-06, + "loss": 2.6734, + "step": 15951 + }, + { + "epoch": 1.6869712351945854, + "grad_norm": 8.6875, + "learning_rate": 3.860110700256572e-06, + "loss": 2.3164, + "step": 15952 + }, + { + "epoch": 1.6870769881556682, + "grad_norm": 8.3125, + "learning_rate": 3.859967003417421e-06, + "loss": 2.3637, + "step": 15953 + }, + { + "epoch": 1.6871827411167513, + "grad_norm": 8.0625, + "learning_rate": 3.8598233001965024e-06, + "loss": 2.4988, + "step": 15954 + }, + { + "epoch": 1.6872884940778343, + "grad_norm": 8.4375, + "learning_rate": 3.85967959059449e-06, + "loss": 2.7414, + "step": 15955 + }, + { + "epoch": 1.6873942470389172, + "grad_norm": 7.90625, + "learning_rate": 3.859535874612055e-06, + "loss": 2.4194, + "step": 15956 + }, + { + "epoch": 1.6875, + "grad_norm": 8.875, + "learning_rate": 3.859392152249876e-06, + "loss": 2.69, + "step": 15957 + }, + { + "epoch": 1.6876057529610828, + "grad_norm": 8.3125, + "learning_rate": 3.859248423508625e-06, + "loss": 2.5192, + "step": 15958 + }, + { + "epoch": 1.6877115059221657, + "grad_norm": 8.25, + "learning_rate": 3.859104688388977e-06, + "loss": 2.5502, + "step": 15959 + }, + { + "epoch": 1.6878172588832487, + "grad_norm": 8.25, + "learning_rate": 3.858960946891607e-06, + "loss": 2.2842, + "step": 15960 + }, + { + "epoch": 1.6879230118443318, + "grad_norm": 8.4375, + "learning_rate": 3.8588171990171885e-06, + "loss": 2.4924, + "step": 15961 + }, + { + "epoch": 1.6880287648054146, + "grad_norm": 8.6875, + "learning_rate": 3.858673444766397e-06, + "loss": 2.5352, + "step": 15962 + }, + { + "epoch": 1.6881345177664975, + "grad_norm": 8.5625, + "learning_rate": 3.858529684139906e-06, + "loss": 2.3562, + "step": 15963 + }, + { + "epoch": 1.6882402707275803, + "grad_norm": 8.25, + "learning_rate": 3.858385917138392e-06, + "loss": 2.5091, + "step": 15964 + }, + { + "epoch": 1.6883460236886632, + "grad_norm": 8.0625, + "learning_rate": 3.858242143762527e-06, + "loss": 2.6059, + "step": 15965 + }, + { + "epoch": 1.6884517766497462, + "grad_norm": 8.4375, + "learning_rate": 3.8580983640129875e-06, + "loss": 2.4866, + "step": 15966 + }, + { + "epoch": 1.688557529610829, + "grad_norm": 8.6875, + "learning_rate": 3.857954577890449e-06, + "loss": 2.3507, + "step": 15967 + }, + { + "epoch": 1.6886632825719121, + "grad_norm": 8.1875, + "learning_rate": 3.857810785395584e-06, + "loss": 2.532, + "step": 15968 + }, + { + "epoch": 1.688769035532995, + "grad_norm": 8.5625, + "learning_rate": 3.857666986529069e-06, + "loss": 2.6056, + "step": 15969 + }, + { + "epoch": 1.6888747884940778, + "grad_norm": 8.125, + "learning_rate": 3.8575231812915784e-06, + "loss": 2.3792, + "step": 15970 + }, + { + "epoch": 1.6889805414551606, + "grad_norm": 9.4375, + "learning_rate": 3.857379369683785e-06, + "loss": 2.6078, + "step": 15971 + }, + { + "epoch": 1.6890862944162437, + "grad_norm": 8.3125, + "learning_rate": 3.857235551706366e-06, + "loss": 2.7515, + "step": 15972 + }, + { + "epoch": 1.6891920473773265, + "grad_norm": 8.1875, + "learning_rate": 3.8570917273599975e-06, + "loss": 2.3746, + "step": 15973 + }, + { + "epoch": 1.6892978003384096, + "grad_norm": 8.375, + "learning_rate": 3.85694789664535e-06, + "loss": 2.7195, + "step": 15974 + }, + { + "epoch": 1.6894035532994924, + "grad_norm": 8.8125, + "learning_rate": 3.856804059563102e-06, + "loss": 2.4307, + "step": 15975 + }, + { + "epoch": 1.6895093062605753, + "grad_norm": 8.1875, + "learning_rate": 3.856660216113928e-06, + "loss": 2.6429, + "step": 15976 + }, + { + "epoch": 1.689615059221658, + "grad_norm": 8.125, + "learning_rate": 3.856516366298502e-06, + "loss": 2.588, + "step": 15977 + }, + { + "epoch": 1.6897208121827412, + "grad_norm": 8.5625, + "learning_rate": 3.8563725101175e-06, + "loss": 2.5081, + "step": 15978 + }, + { + "epoch": 1.689826565143824, + "grad_norm": 8.0, + "learning_rate": 3.8562286475715955e-06, + "loss": 2.7117, + "step": 15979 + }, + { + "epoch": 1.689932318104907, + "grad_norm": 8.0, + "learning_rate": 3.856084778661465e-06, + "loss": 2.6261, + "step": 15980 + }, + { + "epoch": 1.69003807106599, + "grad_norm": 8.625, + "learning_rate": 3.855940903387784e-06, + "loss": 2.7445, + "step": 15981 + }, + { + "epoch": 1.6901438240270727, + "grad_norm": 9.0625, + "learning_rate": 3.855797021751226e-06, + "loss": 2.4451, + "step": 15982 + }, + { + "epoch": 1.6902495769881556, + "grad_norm": 8.3125, + "learning_rate": 3.855653133752467e-06, + "loss": 2.5277, + "step": 15983 + }, + { + "epoch": 1.6903553299492384, + "grad_norm": 9.125, + "learning_rate": 3.855509239392183e-06, + "loss": 2.8844, + "step": 15984 + }, + { + "epoch": 1.6904610829103215, + "grad_norm": 9.0, + "learning_rate": 3.855365338671048e-06, + "loss": 2.5849, + "step": 15985 + }, + { + "epoch": 1.6905668358714045, + "grad_norm": 8.125, + "learning_rate": 3.8552214315897385e-06, + "loss": 2.5368, + "step": 15986 + }, + { + "epoch": 1.6906725888324874, + "grad_norm": 10.375, + "learning_rate": 3.8550775181489284e-06, + "loss": 2.1828, + "step": 15987 + }, + { + "epoch": 1.6907783417935702, + "grad_norm": 8.0, + "learning_rate": 3.854933598349294e-06, + "loss": 2.3321, + "step": 15988 + }, + { + "epoch": 1.690884094754653, + "grad_norm": 7.78125, + "learning_rate": 3.8547896721915105e-06, + "loss": 2.4906, + "step": 15989 + }, + { + "epoch": 1.690989847715736, + "grad_norm": 8.25, + "learning_rate": 3.854645739676253e-06, + "loss": 2.5773, + "step": 15990 + }, + { + "epoch": 1.691095600676819, + "grad_norm": 8.5625, + "learning_rate": 3.854501800804197e-06, + "loss": 2.6091, + "step": 15991 + }, + { + "epoch": 1.691201353637902, + "grad_norm": 8.625, + "learning_rate": 3.854357855576017e-06, + "loss": 2.5788, + "step": 15992 + }, + { + "epoch": 1.6913071065989849, + "grad_norm": 9.0625, + "learning_rate": 3.854213903992391e-06, + "loss": 2.4664, + "step": 15993 + }, + { + "epoch": 1.6914128595600677, + "grad_norm": 8.3125, + "learning_rate": 3.854069946053993e-06, + "loss": 2.5722, + "step": 15994 + }, + { + "epoch": 1.6915186125211505, + "grad_norm": 8.3125, + "learning_rate": 3.8539259817614985e-06, + "loss": 2.3792, + "step": 15995 + }, + { + "epoch": 1.6916243654822334, + "grad_norm": 8.4375, + "learning_rate": 3.853782011115582e-06, + "loss": 2.4847, + "step": 15996 + }, + { + "epoch": 1.6917301184433164, + "grad_norm": 8.9375, + "learning_rate": 3.853638034116921e-06, + "loss": 2.6226, + "step": 15997 + }, + { + "epoch": 1.6918358714043993, + "grad_norm": 8.3125, + "learning_rate": 3.85349405076619e-06, + "loss": 2.6493, + "step": 15998 + }, + { + "epoch": 1.6919416243654823, + "grad_norm": 8.75, + "learning_rate": 3.853350061064066e-06, + "loss": 2.7079, + "step": 15999 + }, + { + "epoch": 1.6920473773265652, + "grad_norm": 8.3125, + "learning_rate": 3.8532060650112224e-06, + "loss": 2.5488, + "step": 16000 + }, + { + "epoch": 1.692153130287648, + "grad_norm": 7.90625, + "learning_rate": 3.853062062608337e-06, + "loss": 2.5779, + "step": 16001 + }, + { + "epoch": 1.6922588832487309, + "grad_norm": 8.5, + "learning_rate": 3.852918053856084e-06, + "loss": 2.5306, + "step": 16002 + }, + { + "epoch": 1.692364636209814, + "grad_norm": 8.0625, + "learning_rate": 3.852774038755141e-06, + "loss": 2.5434, + "step": 16003 + }, + { + "epoch": 1.6924703891708968, + "grad_norm": 8.0625, + "learning_rate": 3.852630017306181e-06, + "loss": 2.7236, + "step": 16004 + }, + { + "epoch": 1.6925761421319798, + "grad_norm": 8.1875, + "learning_rate": 3.852485989509883e-06, + "loss": 2.3961, + "step": 16005 + }, + { + "epoch": 1.6926818950930627, + "grad_norm": 8.1875, + "learning_rate": 3.8523419553669205e-06, + "loss": 2.4774, + "step": 16006 + }, + { + "epoch": 1.6927876480541455, + "grad_norm": 8.5625, + "learning_rate": 3.852197914877971e-06, + "loss": 2.6973, + "step": 16007 + }, + { + "epoch": 1.6928934010152283, + "grad_norm": 8.25, + "learning_rate": 3.852053868043709e-06, + "loss": 2.4861, + "step": 16008 + }, + { + "epoch": 1.6929991539763112, + "grad_norm": 8.25, + "learning_rate": 3.851909814864812e-06, + "loss": 2.5444, + "step": 16009 + }, + { + "epoch": 1.6931049069373942, + "grad_norm": 7.8125, + "learning_rate": 3.8517657553419545e-06, + "loss": 2.5313, + "step": 16010 + }, + { + "epoch": 1.6932106598984773, + "grad_norm": 8.125, + "learning_rate": 3.851621689475813e-06, + "loss": 2.6139, + "step": 16011 + }, + { + "epoch": 1.6933164128595601, + "grad_norm": 8.1875, + "learning_rate": 3.851477617267064e-06, + "loss": 2.4822, + "step": 16012 + }, + { + "epoch": 1.693422165820643, + "grad_norm": 8.25, + "learning_rate": 3.851333538716384e-06, + "loss": 2.5791, + "step": 16013 + }, + { + "epoch": 1.6935279187817258, + "grad_norm": 8.25, + "learning_rate": 3.851189453824448e-06, + "loss": 2.6766, + "step": 16014 + }, + { + "epoch": 1.6936336717428087, + "grad_norm": 7.8125, + "learning_rate": 3.851045362591932e-06, + "loss": 2.5593, + "step": 16015 + }, + { + "epoch": 1.6937394247038917, + "grad_norm": 8.3125, + "learning_rate": 3.850901265019514e-06, + "loss": 2.4571, + "step": 16016 + }, + { + "epoch": 1.6938451776649748, + "grad_norm": 8.5625, + "learning_rate": 3.850757161107867e-06, + "loss": 2.1849, + "step": 16017 + }, + { + "epoch": 1.6939509306260576, + "grad_norm": 8.5625, + "learning_rate": 3.850613050857669e-06, + "loss": 2.5581, + "step": 16018 + }, + { + "epoch": 1.6940566835871405, + "grad_norm": 8.75, + "learning_rate": 3.850468934269598e-06, + "loss": 2.6278, + "step": 16019 + }, + { + "epoch": 1.6941624365482233, + "grad_norm": 9.0625, + "learning_rate": 3.850324811344328e-06, + "loss": 2.7351, + "step": 16020 + }, + { + "epoch": 1.6942681895093061, + "grad_norm": 8.3125, + "learning_rate": 3.850180682082535e-06, + "loss": 2.4966, + "step": 16021 + }, + { + "epoch": 1.6943739424703892, + "grad_norm": 8.3125, + "learning_rate": 3.850036546484897e-06, + "loss": 2.5488, + "step": 16022 + }, + { + "epoch": 1.694479695431472, + "grad_norm": 11.75, + "learning_rate": 3.84989240455209e-06, + "loss": 2.6697, + "step": 16023 + }, + { + "epoch": 1.694585448392555, + "grad_norm": 8.5, + "learning_rate": 3.84974825628479e-06, + "loss": 2.5106, + "step": 16024 + }, + { + "epoch": 1.694691201353638, + "grad_norm": 8.0625, + "learning_rate": 3.849604101683672e-06, + "loss": 2.48, + "step": 16025 + }, + { + "epoch": 1.6947969543147208, + "grad_norm": 8.3125, + "learning_rate": 3.849459940749416e-06, + "loss": 2.5272, + "step": 16026 + }, + { + "epoch": 1.6949027072758036, + "grad_norm": 8.0625, + "learning_rate": 3.849315773482696e-06, + "loss": 2.5947, + "step": 16027 + }, + { + "epoch": 1.6950084602368867, + "grad_norm": 8.5625, + "learning_rate": 3.8491715998841875e-06, + "loss": 2.4994, + "step": 16028 + }, + { + "epoch": 1.6951142131979695, + "grad_norm": 8.1875, + "learning_rate": 3.849027419954569e-06, + "loss": 2.3589, + "step": 16029 + }, + { + "epoch": 1.6952199661590526, + "grad_norm": 8.75, + "learning_rate": 3.848883233694518e-06, + "loss": 2.5174, + "step": 16030 + }, + { + "epoch": 1.6953257191201354, + "grad_norm": 8.75, + "learning_rate": 3.848739041104708e-06, + "loss": 2.7154, + "step": 16031 + }, + { + "epoch": 1.6954314720812182, + "grad_norm": 8.3125, + "learning_rate": 3.848594842185818e-06, + "loss": 2.5148, + "step": 16032 + }, + { + "epoch": 1.695537225042301, + "grad_norm": 7.8125, + "learning_rate": 3.848450636938524e-06, + "loss": 2.4122, + "step": 16033 + }, + { + "epoch": 1.6956429780033841, + "grad_norm": 8.375, + "learning_rate": 3.8483064253635025e-06, + "loss": 2.6017, + "step": 16034 + }, + { + "epoch": 1.695748730964467, + "grad_norm": 7.875, + "learning_rate": 3.84816220746143e-06, + "loss": 2.5567, + "step": 16035 + }, + { + "epoch": 1.69585448392555, + "grad_norm": 8.6875, + "learning_rate": 3.848017983232985e-06, + "loss": 2.5406, + "step": 16036 + }, + { + "epoch": 1.6959602368866329, + "grad_norm": 8.0, + "learning_rate": 3.8478737526788415e-06, + "loss": 2.4732, + "step": 16037 + }, + { + "epoch": 1.6960659898477157, + "grad_norm": 8.0625, + "learning_rate": 3.847729515799677e-06, + "loss": 2.1569, + "step": 16038 + }, + { + "epoch": 1.6961717428087986, + "grad_norm": 8.375, + "learning_rate": 3.847585272596172e-06, + "loss": 2.4638, + "step": 16039 + }, + { + "epoch": 1.6962774957698814, + "grad_norm": 8.5625, + "learning_rate": 3.847441023068998e-06, + "loss": 2.5591, + "step": 16040 + }, + { + "epoch": 1.6963832487309645, + "grad_norm": 8.25, + "learning_rate": 3.8472967672188354e-06, + "loss": 2.0544, + "step": 16041 + }, + { + "epoch": 1.6964890016920475, + "grad_norm": 8.5625, + "learning_rate": 3.84715250504636e-06, + "loss": 2.3844, + "step": 16042 + }, + { + "epoch": 1.6965947546531304, + "grad_norm": 9.0625, + "learning_rate": 3.847008236552248e-06, + "loss": 2.7803, + "step": 16043 + }, + { + "epoch": 1.6967005076142132, + "grad_norm": 8.5, + "learning_rate": 3.846863961737178e-06, + "loss": 2.5718, + "step": 16044 + }, + { + "epoch": 1.696806260575296, + "grad_norm": 8.5625, + "learning_rate": 3.8467196806018256e-06, + "loss": 2.6525, + "step": 16045 + }, + { + "epoch": 1.6969120135363789, + "grad_norm": 7.9375, + "learning_rate": 3.846575393146869e-06, + "loss": 2.4649, + "step": 16046 + }, + { + "epoch": 1.697017766497462, + "grad_norm": 8.0, + "learning_rate": 3.846431099372986e-06, + "loss": 2.463, + "step": 16047 + }, + { + "epoch": 1.697123519458545, + "grad_norm": 8.4375, + "learning_rate": 3.8462867992808516e-06, + "loss": 2.5462, + "step": 16048 + }, + { + "epoch": 1.6972292724196278, + "grad_norm": 8.1875, + "learning_rate": 3.846142492871144e-06, + "loss": 2.467, + "step": 16049 + }, + { + "epoch": 1.6973350253807107, + "grad_norm": 8.25, + "learning_rate": 3.845998180144539e-06, + "loss": 2.5787, + "step": 16050 + }, + { + "epoch": 1.6974407783417935, + "grad_norm": 8.375, + "learning_rate": 3.845853861101716e-06, + "loss": 2.6223, + "step": 16051 + }, + { + "epoch": 1.6975465313028764, + "grad_norm": 8.3125, + "learning_rate": 3.845709535743352e-06, + "loss": 2.6727, + "step": 16052 + }, + { + "epoch": 1.6976522842639594, + "grad_norm": 8.125, + "learning_rate": 3.845565204070122e-06, + "loss": 2.5747, + "step": 16053 + }, + { + "epoch": 1.6977580372250423, + "grad_norm": 8.1875, + "learning_rate": 3.845420866082705e-06, + "loss": 2.6248, + "step": 16054 + }, + { + "epoch": 1.6978637901861253, + "grad_norm": 8.625, + "learning_rate": 3.845276521781779e-06, + "loss": 2.7063, + "step": 16055 + }, + { + "epoch": 1.6979695431472082, + "grad_norm": 9.0625, + "learning_rate": 3.84513217116802e-06, + "loss": 2.7514, + "step": 16056 + }, + { + "epoch": 1.698075296108291, + "grad_norm": 8.75, + "learning_rate": 3.844987814242107e-06, + "loss": 2.5948, + "step": 16057 + }, + { + "epoch": 1.6981810490693738, + "grad_norm": 8.4375, + "learning_rate": 3.844843451004715e-06, + "loss": 2.5264, + "step": 16058 + }, + { + "epoch": 1.698286802030457, + "grad_norm": 9.375, + "learning_rate": 3.844699081456523e-06, + "loss": 2.5674, + "step": 16059 + }, + { + "epoch": 1.6983925549915397, + "grad_norm": 8.0625, + "learning_rate": 3.8445547055982085e-06, + "loss": 2.4683, + "step": 16060 + }, + { + "epoch": 1.6984983079526228, + "grad_norm": 8.4375, + "learning_rate": 3.844410323430449e-06, + "loss": 2.4761, + "step": 16061 + }, + { + "epoch": 1.6986040609137056, + "grad_norm": 8.5, + "learning_rate": 3.84426593495392e-06, + "loss": 2.5338, + "step": 16062 + }, + { + "epoch": 1.6987098138747885, + "grad_norm": 8.0625, + "learning_rate": 3.844121540169302e-06, + "loss": 2.4646, + "step": 16063 + }, + { + "epoch": 1.6988155668358713, + "grad_norm": 8.625, + "learning_rate": 3.8439771390772716e-06, + "loss": 2.4607, + "step": 16064 + }, + { + "epoch": 1.6989213197969542, + "grad_norm": 8.25, + "learning_rate": 3.843832731678505e-06, + "loss": 2.6912, + "step": 16065 + }, + { + "epoch": 1.6990270727580372, + "grad_norm": 8.375, + "learning_rate": 3.843688317973681e-06, + "loss": 2.513, + "step": 16066 + }, + { + "epoch": 1.6991328257191203, + "grad_norm": 8.125, + "learning_rate": 3.84354389796348e-06, + "loss": 2.6018, + "step": 16067 + }, + { + "epoch": 1.6992385786802031, + "grad_norm": 8.0625, + "learning_rate": 3.843399471648574e-06, + "loss": 2.7013, + "step": 16068 + }, + { + "epoch": 1.699344331641286, + "grad_norm": 7.90625, + "learning_rate": 3.843255039029645e-06, + "loss": 2.443, + "step": 16069 + }, + { + "epoch": 1.6994500846023688, + "grad_norm": 7.71875, + "learning_rate": 3.843110600107369e-06, + "loss": 2.4872, + "step": 16070 + }, + { + "epoch": 1.6995558375634516, + "grad_norm": 8.5, + "learning_rate": 3.842966154882425e-06, + "loss": 2.4548, + "step": 16071 + }, + { + "epoch": 1.6996615905245347, + "grad_norm": 7.9375, + "learning_rate": 3.842821703355488e-06, + "loss": 2.6614, + "step": 16072 + }, + { + "epoch": 1.6997673434856178, + "grad_norm": 7.96875, + "learning_rate": 3.84267724552724e-06, + "loss": 2.4269, + "step": 16073 + }, + { + "epoch": 1.6998730964467006, + "grad_norm": 7.8125, + "learning_rate": 3.842532781398357e-06, + "loss": 2.4858, + "step": 16074 + }, + { + "epoch": 1.6999788494077834, + "grad_norm": 7.71875, + "learning_rate": 3.842388310969516e-06, + "loss": 2.5475, + "step": 16075 + }, + { + "epoch": 1.7000846023688663, + "grad_norm": 8.125, + "learning_rate": 3.842243834241396e-06, + "loss": 2.5755, + "step": 16076 + }, + { + "epoch": 1.700190355329949, + "grad_norm": 8.75, + "learning_rate": 3.8420993512146745e-06, + "loss": 2.6389, + "step": 16077 + }, + { + "epoch": 1.7002961082910322, + "grad_norm": 9.125, + "learning_rate": 3.841954861890029e-06, + "loss": 2.2668, + "step": 16078 + }, + { + "epoch": 1.700401861252115, + "grad_norm": 7.96875, + "learning_rate": 3.841810366268138e-06, + "loss": 2.4846, + "step": 16079 + }, + { + "epoch": 1.700507614213198, + "grad_norm": 8.375, + "learning_rate": 3.8416658643496804e-06, + "loss": 2.5418, + "step": 16080 + }, + { + "epoch": 1.700613367174281, + "grad_norm": 8.0625, + "learning_rate": 3.841521356135333e-06, + "loss": 2.3516, + "step": 16081 + }, + { + "epoch": 1.7007191201353637, + "grad_norm": 7.84375, + "learning_rate": 3.841376841625776e-06, + "loss": 2.5426, + "step": 16082 + }, + { + "epoch": 1.7008248730964466, + "grad_norm": 8.0, + "learning_rate": 3.841232320821685e-06, + "loss": 2.4798, + "step": 16083 + }, + { + "epoch": 1.7009306260575296, + "grad_norm": 8.0625, + "learning_rate": 3.84108779372374e-06, + "loss": 2.6632, + "step": 16084 + }, + { + "epoch": 1.7010363790186125, + "grad_norm": 8.6875, + "learning_rate": 3.840943260332618e-06, + "loss": 2.7345, + "step": 16085 + }, + { + "epoch": 1.7011421319796955, + "grad_norm": 8.1875, + "learning_rate": 3.840798720648998e-06, + "loss": 2.5756, + "step": 16086 + }, + { + "epoch": 1.7012478849407784, + "grad_norm": 8.6875, + "learning_rate": 3.840654174673557e-06, + "loss": 2.491, + "step": 16087 + }, + { + "epoch": 1.7013536379018612, + "grad_norm": 8.375, + "learning_rate": 3.840509622406975e-06, + "loss": 2.5637, + "step": 16088 + }, + { + "epoch": 1.701459390862944, + "grad_norm": 8.75, + "learning_rate": 3.840365063849929e-06, + "loss": 2.4355, + "step": 16089 + }, + { + "epoch": 1.7015651438240271, + "grad_norm": 8.3125, + "learning_rate": 3.840220499003099e-06, + "loss": 2.346, + "step": 16090 + }, + { + "epoch": 1.70167089678511, + "grad_norm": 9.0625, + "learning_rate": 3.840075927867162e-06, + "loss": 2.7113, + "step": 16091 + }, + { + "epoch": 1.701776649746193, + "grad_norm": 8.9375, + "learning_rate": 3.839931350442796e-06, + "loss": 2.5318, + "step": 16092 + }, + { + "epoch": 1.7018824027072759, + "grad_norm": 9.0, + "learning_rate": 3.839786766730681e-06, + "loss": 2.832, + "step": 16093 + }, + { + "epoch": 1.7019881556683587, + "grad_norm": 8.625, + "learning_rate": 3.8396421767314936e-06, + "loss": 2.6439, + "step": 16094 + }, + { + "epoch": 1.7020939086294415, + "grad_norm": 8.375, + "learning_rate": 3.839497580445915e-06, + "loss": 2.6346, + "step": 16095 + }, + { + "epoch": 1.7021996615905244, + "grad_norm": 8.4375, + "learning_rate": 3.839352977874622e-06, + "loss": 2.3669, + "step": 16096 + }, + { + "epoch": 1.7023054145516074, + "grad_norm": 8.9375, + "learning_rate": 3.8392083690182915e-06, + "loss": 2.6984, + "step": 16097 + }, + { + "epoch": 1.7024111675126905, + "grad_norm": 8.375, + "learning_rate": 3.839063753877604e-06, + "loss": 2.5317, + "step": 16098 + }, + { + "epoch": 1.7025169204737733, + "grad_norm": 8.25, + "learning_rate": 3.838919132453239e-06, + "loss": 2.5486, + "step": 16099 + }, + { + "epoch": 1.7026226734348562, + "grad_norm": 8.0, + "learning_rate": 3.838774504745875e-06, + "loss": 2.535, + "step": 16100 + }, + { + "epoch": 1.702728426395939, + "grad_norm": 8.5625, + "learning_rate": 3.838629870756188e-06, + "loss": 2.707, + "step": 16101 + }, + { + "epoch": 1.7028341793570219, + "grad_norm": 8.3125, + "learning_rate": 3.83848523048486e-06, + "loss": 2.4096, + "step": 16102 + }, + { + "epoch": 1.702939932318105, + "grad_norm": 8.125, + "learning_rate": 3.838340583932567e-06, + "loss": 2.709, + "step": 16103 + }, + { + "epoch": 1.703045685279188, + "grad_norm": 8.125, + "learning_rate": 3.838195931099989e-06, + "loss": 2.5123, + "step": 16104 + }, + { + "epoch": 1.7031514382402708, + "grad_norm": 7.875, + "learning_rate": 3.838051271987806e-06, + "loss": 2.4393, + "step": 16105 + }, + { + "epoch": 1.7032571912013537, + "grad_norm": 9.875, + "learning_rate": 3.837906606596694e-06, + "loss": 2.4581, + "step": 16106 + }, + { + "epoch": 1.7033629441624365, + "grad_norm": 7.96875, + "learning_rate": 3.837761934927334e-06, + "loss": 2.4548, + "step": 16107 + }, + { + "epoch": 1.7034686971235193, + "grad_norm": 8.1875, + "learning_rate": 3.837617256980405e-06, + "loss": 2.5806, + "step": 16108 + }, + { + "epoch": 1.7035744500846024, + "grad_norm": 8.6875, + "learning_rate": 3.837472572756585e-06, + "loss": 2.4988, + "step": 16109 + }, + { + "epoch": 1.7036802030456852, + "grad_norm": 8.3125, + "learning_rate": 3.837327882256553e-06, + "loss": 2.5769, + "step": 16110 + }, + { + "epoch": 1.7037859560067683, + "grad_norm": 8.1875, + "learning_rate": 3.837183185480989e-06, + "loss": 2.5883, + "step": 16111 + }, + { + "epoch": 1.7038917089678511, + "grad_norm": 8.0625, + "learning_rate": 3.837038482430571e-06, + "loss": 2.4129, + "step": 16112 + }, + { + "epoch": 1.703997461928934, + "grad_norm": 8.6875, + "learning_rate": 3.836893773105977e-06, + "loss": 2.5456, + "step": 16113 + }, + { + "epoch": 1.7041032148900168, + "grad_norm": 8.25, + "learning_rate": 3.836749057507888e-06, + "loss": 2.4083, + "step": 16114 + }, + { + "epoch": 1.7042089678510999, + "grad_norm": 8.3125, + "learning_rate": 3.836604335636982e-06, + "loss": 2.5635, + "step": 16115 + }, + { + "epoch": 1.7043147208121827, + "grad_norm": 8.0, + "learning_rate": 3.83645960749394e-06, + "loss": 2.7308, + "step": 16116 + }, + { + "epoch": 1.7044204737732658, + "grad_norm": 9.0, + "learning_rate": 3.836314873079438e-06, + "loss": 2.7173, + "step": 16117 + }, + { + "epoch": 1.7045262267343486, + "grad_norm": 8.125, + "learning_rate": 3.836170132394158e-06, + "loss": 2.5948, + "step": 16118 + }, + { + "epoch": 1.7046319796954315, + "grad_norm": 7.90625, + "learning_rate": 3.836025385438778e-06, + "loss": 2.5423, + "step": 16119 + }, + { + "epoch": 1.7047377326565143, + "grad_norm": 8.4375, + "learning_rate": 3.835880632213977e-06, + "loss": 2.4876, + "step": 16120 + }, + { + "epoch": 1.7048434856175971, + "grad_norm": 8.6875, + "learning_rate": 3.835735872720434e-06, + "loss": 2.6338, + "step": 16121 + }, + { + "epoch": 1.7049492385786802, + "grad_norm": 8.3125, + "learning_rate": 3.8355911069588285e-06, + "loss": 2.6431, + "step": 16122 + }, + { + "epoch": 1.7050549915397633, + "grad_norm": 8.0, + "learning_rate": 3.835446334929841e-06, + "loss": 2.4526, + "step": 16123 + }, + { + "epoch": 1.705160744500846, + "grad_norm": 8.5625, + "learning_rate": 3.83530155663415e-06, + "loss": 2.576, + "step": 16124 + }, + { + "epoch": 1.705266497461929, + "grad_norm": 8.0625, + "learning_rate": 3.8351567720724345e-06, + "loss": 2.5413, + "step": 16125 + }, + { + "epoch": 1.7053722504230118, + "grad_norm": 8.4375, + "learning_rate": 3.835011981245374e-06, + "loss": 2.4173, + "step": 16126 + }, + { + "epoch": 1.7054780033840946, + "grad_norm": 7.90625, + "learning_rate": 3.83486718415365e-06, + "loss": 2.4729, + "step": 16127 + }, + { + "epoch": 1.7055837563451777, + "grad_norm": 9.4375, + "learning_rate": 3.834722380797939e-06, + "loss": 2.3908, + "step": 16128 + }, + { + "epoch": 1.7056895093062607, + "grad_norm": 10.4375, + "learning_rate": 3.834577571178922e-06, + "loss": 2.438, + "step": 16129 + }, + { + "epoch": 1.7057952622673436, + "grad_norm": 8.4375, + "learning_rate": 3.8344327552972785e-06, + "loss": 2.5709, + "step": 16130 + }, + { + "epoch": 1.7059010152284264, + "grad_norm": 8.6875, + "learning_rate": 3.834287933153687e-06, + "loss": 2.7003, + "step": 16131 + }, + { + "epoch": 1.7060067681895092, + "grad_norm": 7.96875, + "learning_rate": 3.834143104748829e-06, + "loss": 2.5062, + "step": 16132 + }, + { + "epoch": 1.706112521150592, + "grad_norm": 8.4375, + "learning_rate": 3.833998270083382e-06, + "loss": 2.5579, + "step": 16133 + }, + { + "epoch": 1.7062182741116751, + "grad_norm": 8.5625, + "learning_rate": 3.833853429158028e-06, + "loss": 2.6213, + "step": 16134 + }, + { + "epoch": 1.706324027072758, + "grad_norm": 8.25, + "learning_rate": 3.833708581973444e-06, + "loss": 2.491, + "step": 16135 + }, + { + "epoch": 1.706429780033841, + "grad_norm": 8.375, + "learning_rate": 3.833563728530312e-06, + "loss": 2.4408, + "step": 16136 + }, + { + "epoch": 1.7065355329949239, + "grad_norm": 7.96875, + "learning_rate": 3.83341886882931e-06, + "loss": 2.5934, + "step": 16137 + }, + { + "epoch": 1.7066412859560067, + "grad_norm": 8.625, + "learning_rate": 3.833274002871121e-06, + "loss": 2.376, + "step": 16138 + }, + { + "epoch": 1.7067470389170896, + "grad_norm": 8.5625, + "learning_rate": 3.8331291306564185e-06, + "loss": 2.6056, + "step": 16139 + }, + { + "epoch": 1.7068527918781726, + "grad_norm": 8.125, + "learning_rate": 3.832984252185889e-06, + "loss": 2.6554, + "step": 16140 + }, + { + "epoch": 1.7069585448392555, + "grad_norm": 8.6875, + "learning_rate": 3.832839367460208e-06, + "loss": 2.6006, + "step": 16141 + }, + { + "epoch": 1.7070642978003385, + "grad_norm": 8.625, + "learning_rate": 3.832694476480058e-06, + "loss": 2.6306, + "step": 16142 + }, + { + "epoch": 1.7071700507614214, + "grad_norm": 8.25, + "learning_rate": 3.832549579246118e-06, + "loss": 2.4936, + "step": 16143 + }, + { + "epoch": 1.7072758037225042, + "grad_norm": 8.375, + "learning_rate": 3.832404675759068e-06, + "loss": 2.6766, + "step": 16144 + }, + { + "epoch": 1.707381556683587, + "grad_norm": 8.0625, + "learning_rate": 3.832259766019587e-06, + "loss": 2.4298, + "step": 16145 + }, + { + "epoch": 1.70748730964467, + "grad_norm": 9.0, + "learning_rate": 3.8321148500283565e-06, + "loss": 2.4839, + "step": 16146 + }, + { + "epoch": 1.707593062605753, + "grad_norm": 9.0, + "learning_rate": 3.831969927786055e-06, + "loss": 2.7236, + "step": 16147 + }, + { + "epoch": 1.707698815566836, + "grad_norm": 8.375, + "learning_rate": 3.831824999293364e-06, + "loss": 2.4254, + "step": 16148 + }, + { + "epoch": 1.7078045685279188, + "grad_norm": 7.8125, + "learning_rate": 3.8316800645509625e-06, + "loss": 2.477, + "step": 16149 + }, + { + "epoch": 1.7079103214890017, + "grad_norm": 8.0625, + "learning_rate": 3.8315351235595314e-06, + "loss": 2.3765, + "step": 16150 + }, + { + "epoch": 1.7080160744500845, + "grad_norm": 8.4375, + "learning_rate": 3.83139017631975e-06, + "loss": 2.4977, + "step": 16151 + }, + { + "epoch": 1.7081218274111674, + "grad_norm": 7.84375, + "learning_rate": 3.8312452228323e-06, + "loss": 2.2475, + "step": 16152 + }, + { + "epoch": 1.7082275803722504, + "grad_norm": 8.1875, + "learning_rate": 3.83110026309786e-06, + "loss": 2.5812, + "step": 16153 + }, + { + "epoch": 1.7083333333333335, + "grad_norm": 8.75, + "learning_rate": 3.830955297117112e-06, + "loss": 2.5847, + "step": 16154 + }, + { + "epoch": 1.7084390862944163, + "grad_norm": 8.0625, + "learning_rate": 3.8308103248907335e-06, + "loss": 2.6446, + "step": 16155 + }, + { + "epoch": 1.7085448392554992, + "grad_norm": 8.1875, + "learning_rate": 3.830665346419407e-06, + "loss": 2.5043, + "step": 16156 + }, + { + "epoch": 1.708650592216582, + "grad_norm": 7.84375, + "learning_rate": 3.830520361703812e-06, + "loss": 2.3947, + "step": 16157 + }, + { + "epoch": 1.7087563451776648, + "grad_norm": 8.6875, + "learning_rate": 3.83037537074463e-06, + "loss": 2.4751, + "step": 16158 + }, + { + "epoch": 1.708862098138748, + "grad_norm": 8.5, + "learning_rate": 3.83023037354254e-06, + "loss": 2.7961, + "step": 16159 + }, + { + "epoch": 1.708967851099831, + "grad_norm": 8.5, + "learning_rate": 3.830085370098222e-06, + "loss": 2.6561, + "step": 16160 + }, + { + "epoch": 1.7090736040609138, + "grad_norm": 8.5, + "learning_rate": 3.829940360412359e-06, + "loss": 2.5654, + "step": 16161 + }, + { + "epoch": 1.7091793570219966, + "grad_norm": 8.375, + "learning_rate": 3.829795344485629e-06, + "loss": 2.5894, + "step": 16162 + }, + { + "epoch": 1.7092851099830795, + "grad_norm": 8.5625, + "learning_rate": 3.829650322318713e-06, + "loss": 2.642, + "step": 16163 + }, + { + "epoch": 1.7093908629441623, + "grad_norm": 8.1875, + "learning_rate": 3.8295052939122926e-06, + "loss": 2.5695, + "step": 16164 + }, + { + "epoch": 1.7094966159052454, + "grad_norm": 8.3125, + "learning_rate": 3.829360259267047e-06, + "loss": 2.7199, + "step": 16165 + }, + { + "epoch": 1.7096023688663282, + "grad_norm": 8.875, + "learning_rate": 3.829215218383657e-06, + "loss": 2.5457, + "step": 16166 + }, + { + "epoch": 1.7097081218274113, + "grad_norm": 8.3125, + "learning_rate": 3.8290701712628045e-06, + "loss": 2.702, + "step": 16167 + }, + { + "epoch": 1.7098138747884941, + "grad_norm": 7.90625, + "learning_rate": 3.828925117905169e-06, + "loss": 2.445, + "step": 16168 + }, + { + "epoch": 1.709919627749577, + "grad_norm": 8.25, + "learning_rate": 3.828780058311431e-06, + "loss": 2.7154, + "step": 16169 + }, + { + "epoch": 1.7100253807106598, + "grad_norm": 8.3125, + "learning_rate": 3.8286349924822726e-06, + "loss": 2.3645, + "step": 16170 + }, + { + "epoch": 1.7101311336717429, + "grad_norm": 8.125, + "learning_rate": 3.828489920418372e-06, + "loss": 2.4392, + "step": 16171 + }, + { + "epoch": 1.7102368866328257, + "grad_norm": 8.3125, + "learning_rate": 3.828344842120413e-06, + "loss": 2.4573, + "step": 16172 + }, + { + "epoch": 1.7103426395939088, + "grad_norm": 8.4375, + "learning_rate": 3.828199757589075e-06, + "loss": 2.462, + "step": 16173 + }, + { + "epoch": 1.7104483925549916, + "grad_norm": 9.125, + "learning_rate": 3.8280546668250375e-06, + "loss": 2.8142, + "step": 16174 + }, + { + "epoch": 1.7105541455160744, + "grad_norm": 8.625, + "learning_rate": 3.827909569828984e-06, + "loss": 2.6735, + "step": 16175 + }, + { + "epoch": 1.7106598984771573, + "grad_norm": 8.375, + "learning_rate": 3.827764466601594e-06, + "loss": 2.4666, + "step": 16176 + }, + { + "epoch": 1.71076565143824, + "grad_norm": 8.1875, + "learning_rate": 3.827619357143546e-06, + "loss": 2.7118, + "step": 16177 + }, + { + "epoch": 1.7108714043993232, + "grad_norm": 7.96875, + "learning_rate": 3.827474241455525e-06, + "loss": 2.5412, + "step": 16178 + }, + { + "epoch": 1.7109771573604062, + "grad_norm": 9.0, + "learning_rate": 3.8273291195382116e-06, + "loss": 2.5614, + "step": 16179 + }, + { + "epoch": 1.711082910321489, + "grad_norm": 8.3125, + "learning_rate": 3.8271839913922834e-06, + "loss": 2.639, + "step": 16180 + }, + { + "epoch": 1.711188663282572, + "grad_norm": 8.75, + "learning_rate": 3.8270388570184244e-06, + "loss": 2.6043, + "step": 16181 + }, + { + "epoch": 1.7112944162436547, + "grad_norm": 8.1875, + "learning_rate": 3.826893716417315e-06, + "loss": 2.4508, + "step": 16182 + }, + { + "epoch": 1.7114001692047376, + "grad_norm": 8.1875, + "learning_rate": 3.826748569589635e-06, + "loss": 2.3433, + "step": 16183 + }, + { + "epoch": 1.7115059221658206, + "grad_norm": 8.0625, + "learning_rate": 3.826603416536068e-06, + "loss": 2.4426, + "step": 16184 + }, + { + "epoch": 1.7116116751269037, + "grad_norm": 7.84375, + "learning_rate": 3.826458257257293e-06, + "loss": 2.4726, + "step": 16185 + }, + { + "epoch": 1.7117174280879865, + "grad_norm": 9.125, + "learning_rate": 3.826313091753992e-06, + "loss": 2.2742, + "step": 16186 + }, + { + "epoch": 1.7118231810490694, + "grad_norm": 8.1875, + "learning_rate": 3.826167920026846e-06, + "loss": 2.6211, + "step": 16187 + }, + { + "epoch": 1.7119289340101522, + "grad_norm": 8.0625, + "learning_rate": 3.826022742076537e-06, + "loss": 2.4054, + "step": 16188 + }, + { + "epoch": 1.712034686971235, + "grad_norm": 7.8125, + "learning_rate": 3.825877557903745e-06, + "loss": 2.595, + "step": 16189 + }, + { + "epoch": 1.7121404399323181, + "grad_norm": 8.125, + "learning_rate": 3.825732367509152e-06, + "loss": 2.503, + "step": 16190 + }, + { + "epoch": 1.712246192893401, + "grad_norm": 8.0, + "learning_rate": 3.825587170893439e-06, + "loss": 2.5229, + "step": 16191 + }, + { + "epoch": 1.712351945854484, + "grad_norm": 7.75, + "learning_rate": 3.825441968057287e-06, + "loss": 2.3901, + "step": 16192 + }, + { + "epoch": 1.7124576988155669, + "grad_norm": 9.375, + "learning_rate": 3.825296759001379e-06, + "loss": 2.5244, + "step": 16193 + }, + { + "epoch": 1.7125634517766497, + "grad_norm": 8.4375, + "learning_rate": 3.8251515437263955e-06, + "loss": 2.5518, + "step": 16194 + }, + { + "epoch": 1.7126692047377325, + "grad_norm": 8.1875, + "learning_rate": 3.825006322233017e-06, + "loss": 2.3833, + "step": 16195 + }, + { + "epoch": 1.7127749576988156, + "grad_norm": 8.6875, + "learning_rate": 3.824861094521926e-06, + "loss": 2.4741, + "step": 16196 + }, + { + "epoch": 1.7128807106598984, + "grad_norm": 8.8125, + "learning_rate": 3.8247158605938035e-06, + "loss": 2.5648, + "step": 16197 + }, + { + "epoch": 1.7129864636209815, + "grad_norm": 8.125, + "learning_rate": 3.824570620449332e-06, + "loss": 2.4953, + "step": 16198 + }, + { + "epoch": 1.7130922165820643, + "grad_norm": 8.5, + "learning_rate": 3.824425374089191e-06, + "loss": 2.3179, + "step": 16199 + }, + { + "epoch": 1.7131979695431472, + "grad_norm": 8.875, + "learning_rate": 3.8242801215140645e-06, + "loss": 2.596, + "step": 16200 + }, + { + "epoch": 1.71330372250423, + "grad_norm": 7.96875, + "learning_rate": 3.8241348627246324e-06, + "loss": 2.4183, + "step": 16201 + }, + { + "epoch": 1.713409475465313, + "grad_norm": 8.3125, + "learning_rate": 3.823989597721578e-06, + "loss": 2.5659, + "step": 16202 + }, + { + "epoch": 1.713515228426396, + "grad_norm": 8.0625, + "learning_rate": 3.823844326505581e-06, + "loss": 2.3126, + "step": 16203 + }, + { + "epoch": 1.713620981387479, + "grad_norm": 9.0625, + "learning_rate": 3.823699049077324e-06, + "loss": 2.761, + "step": 16204 + }, + { + "epoch": 1.7137267343485618, + "grad_norm": 7.8125, + "learning_rate": 3.823553765437489e-06, + "loss": 2.5269, + "step": 16205 + }, + { + "epoch": 1.7138324873096447, + "grad_norm": 8.375, + "learning_rate": 3.823408475586757e-06, + "loss": 2.3945, + "step": 16206 + }, + { + "epoch": 1.7139382402707275, + "grad_norm": 8.25, + "learning_rate": 3.82326317952581e-06, + "loss": 2.6447, + "step": 16207 + }, + { + "epoch": 1.7140439932318103, + "grad_norm": 8.4375, + "learning_rate": 3.8231178772553315e-06, + "loss": 2.5798, + "step": 16208 + }, + { + "epoch": 1.7141497461928934, + "grad_norm": 8.875, + "learning_rate": 3.822972568776001e-06, + "loss": 2.6374, + "step": 16209 + }, + { + "epoch": 1.7142554991539765, + "grad_norm": 8.0625, + "learning_rate": 3.822827254088501e-06, + "loss": 2.4681, + "step": 16210 + }, + { + "epoch": 1.7143612521150593, + "grad_norm": 8.375, + "learning_rate": 3.822681933193515e-06, + "loss": 2.5016, + "step": 16211 + }, + { + "epoch": 1.7144670050761421, + "grad_norm": 8.875, + "learning_rate": 3.822536606091722e-06, + "loss": 2.6027, + "step": 16212 + }, + { + "epoch": 1.714572758037225, + "grad_norm": 8.3125, + "learning_rate": 3.822391272783807e-06, + "loss": 2.5776, + "step": 16213 + }, + { + "epoch": 1.7146785109983078, + "grad_norm": 8.375, + "learning_rate": 3.82224593327045e-06, + "loss": 2.6983, + "step": 16214 + }, + { + "epoch": 1.7147842639593909, + "grad_norm": 7.96875, + "learning_rate": 3.822100587552333e-06, + "loss": 2.5688, + "step": 16215 + }, + { + "epoch": 1.714890016920474, + "grad_norm": 8.5625, + "learning_rate": 3.8219552356301395e-06, + "loss": 2.68, + "step": 16216 + }, + { + "epoch": 1.7149957698815568, + "grad_norm": 8.5, + "learning_rate": 3.821809877504551e-06, + "loss": 2.6383, + "step": 16217 + }, + { + "epoch": 1.7151015228426396, + "grad_norm": 8.5, + "learning_rate": 3.821664513176249e-06, + "loss": 2.6161, + "step": 16218 + }, + { + "epoch": 1.7152072758037225, + "grad_norm": 8.125, + "learning_rate": 3.821519142645916e-06, + "loss": 2.4892, + "step": 16219 + }, + { + "epoch": 1.7153130287648053, + "grad_norm": 8.6875, + "learning_rate": 3.821373765914234e-06, + "loss": 2.4688, + "step": 16220 + }, + { + "epoch": 1.7154187817258884, + "grad_norm": 8.4375, + "learning_rate": 3.8212283829818854e-06, + "loss": 2.2673, + "step": 16221 + }, + { + "epoch": 1.7155245346869712, + "grad_norm": 8.5, + "learning_rate": 3.821082993849552e-06, + "loss": 2.6028, + "step": 16222 + }, + { + "epoch": 1.7156302876480543, + "grad_norm": 7.75, + "learning_rate": 3.820937598517917e-06, + "loss": 2.4953, + "step": 16223 + }, + { + "epoch": 1.715736040609137, + "grad_norm": 8.3125, + "learning_rate": 3.820792196987662e-06, + "loss": 2.609, + "step": 16224 + }, + { + "epoch": 1.71584179357022, + "grad_norm": 8.5, + "learning_rate": 3.820646789259469e-06, + "loss": 2.6254, + "step": 16225 + }, + { + "epoch": 1.7159475465313028, + "grad_norm": 8.25, + "learning_rate": 3.820501375334021e-06, + "loss": 2.5182, + "step": 16226 + }, + { + "epoch": 1.7160532994923858, + "grad_norm": 8.3125, + "learning_rate": 3.820355955212001e-06, + "loss": 2.5746, + "step": 16227 + }, + { + "epoch": 1.7161590524534687, + "grad_norm": 8.3125, + "learning_rate": 3.8202105288940885e-06, + "loss": 2.242, + "step": 16228 + }, + { + "epoch": 1.7162648054145517, + "grad_norm": 8.0625, + "learning_rate": 3.82006509638097e-06, + "loss": 2.45, + "step": 16229 + }, + { + "epoch": 1.7163705583756346, + "grad_norm": 8.0, + "learning_rate": 3.819919657673324e-06, + "loss": 2.4949, + "step": 16230 + }, + { + "epoch": 1.7164763113367174, + "grad_norm": 8.1875, + "learning_rate": 3.819774212771836e-06, + "loss": 2.5584, + "step": 16231 + }, + { + "epoch": 1.7165820642978002, + "grad_norm": 7.71875, + "learning_rate": 3.819628761677188e-06, + "loss": 2.4738, + "step": 16232 + }, + { + "epoch": 1.716687817258883, + "grad_norm": 8.0625, + "learning_rate": 3.81948330439006e-06, + "loss": 2.4097, + "step": 16233 + }, + { + "epoch": 1.7167935702199661, + "grad_norm": 8.3125, + "learning_rate": 3.819337840911138e-06, + "loss": 2.4697, + "step": 16234 + }, + { + "epoch": 1.7168993231810492, + "grad_norm": 8.4375, + "learning_rate": 3.819192371241103e-06, + "loss": 2.5194, + "step": 16235 + }, + { + "epoch": 1.717005076142132, + "grad_norm": 8.1875, + "learning_rate": 3.819046895380638e-06, + "loss": 2.4571, + "step": 16236 + }, + { + "epoch": 1.7171108291032149, + "grad_norm": 7.8125, + "learning_rate": 3.818901413330424e-06, + "loss": 2.0557, + "step": 16237 + }, + { + "epoch": 1.7172165820642977, + "grad_norm": 7.84375, + "learning_rate": 3.8187559250911465e-06, + "loss": 2.466, + "step": 16238 + }, + { + "epoch": 1.7173223350253806, + "grad_norm": 8.5, + "learning_rate": 3.818610430663486e-06, + "loss": 2.6826, + "step": 16239 + }, + { + "epoch": 1.7174280879864636, + "grad_norm": 8.1875, + "learning_rate": 3.818464930048126e-06, + "loss": 2.5409, + "step": 16240 + }, + { + "epoch": 1.7175338409475467, + "grad_norm": 8.5, + "learning_rate": 3.81831942324575e-06, + "loss": 2.6649, + "step": 16241 + }, + { + "epoch": 1.7176395939086295, + "grad_norm": 8.5, + "learning_rate": 3.8181739102570395e-06, + "loss": 2.6484, + "step": 16242 + }, + { + "epoch": 1.7177453468697124, + "grad_norm": 8.5625, + "learning_rate": 3.818028391082679e-06, + "loss": 2.4297, + "step": 16243 + }, + { + "epoch": 1.7178510998307952, + "grad_norm": 7.90625, + "learning_rate": 3.8178828657233495e-06, + "loss": 2.4776, + "step": 16244 + }, + { + "epoch": 1.717956852791878, + "grad_norm": 8.0625, + "learning_rate": 3.8177373341797345e-06, + "loss": 2.6105, + "step": 16245 + }, + { + "epoch": 1.718062605752961, + "grad_norm": 8.25, + "learning_rate": 3.817591796452517e-06, + "loss": 2.4986, + "step": 16246 + }, + { + "epoch": 1.718168358714044, + "grad_norm": 7.96875, + "learning_rate": 3.81744625254238e-06, + "loss": 2.4291, + "step": 16247 + }, + { + "epoch": 1.718274111675127, + "grad_norm": 9.3125, + "learning_rate": 3.8173007024500075e-06, + "loss": 2.6366, + "step": 16248 + }, + { + "epoch": 1.7183798646362098, + "grad_norm": 8.4375, + "learning_rate": 3.8171551461760806e-06, + "loss": 2.2725, + "step": 16249 + }, + { + "epoch": 1.7184856175972927, + "grad_norm": 8.625, + "learning_rate": 3.817009583721284e-06, + "loss": 2.569, + "step": 16250 + }, + { + "epoch": 1.7185913705583755, + "grad_norm": 8.5, + "learning_rate": 3.8168640150863e-06, + "loss": 2.4776, + "step": 16251 + }, + { + "epoch": 1.7186971235194586, + "grad_norm": 8.0625, + "learning_rate": 3.816718440271811e-06, + "loss": 2.49, + "step": 16252 + }, + { + "epoch": 1.7188028764805414, + "grad_norm": 8.125, + "learning_rate": 3.816572859278502e-06, + "loss": 2.4153, + "step": 16253 + }, + { + "epoch": 1.7189086294416245, + "grad_norm": 8.0625, + "learning_rate": 3.816427272107054e-06, + "loss": 2.5683, + "step": 16254 + }, + { + "epoch": 1.7190143824027073, + "grad_norm": 8.375, + "learning_rate": 3.816281678758152e-06, + "loss": 2.6798, + "step": 16255 + }, + { + "epoch": 1.7191201353637902, + "grad_norm": 8.4375, + "learning_rate": 3.816136079232478e-06, + "loss": 2.6825, + "step": 16256 + }, + { + "epoch": 1.719225888324873, + "grad_norm": 8.625, + "learning_rate": 3.815990473530715e-06, + "loss": 2.652, + "step": 16257 + }, + { + "epoch": 1.719331641285956, + "grad_norm": 8.125, + "learning_rate": 3.8158448616535486e-06, + "loss": 2.4831, + "step": 16258 + }, + { + "epoch": 1.719437394247039, + "grad_norm": 8.375, + "learning_rate": 3.815699243601659e-06, + "loss": 2.5406, + "step": 16259 + }, + { + "epoch": 1.719543147208122, + "grad_norm": 8.0625, + "learning_rate": 3.815553619375731e-06, + "loss": 2.6043, + "step": 16260 + }, + { + "epoch": 1.7196489001692048, + "grad_norm": 8.5, + "learning_rate": 3.815407988976449e-06, + "loss": 2.5507, + "step": 16261 + }, + { + "epoch": 1.7197546531302876, + "grad_norm": 8.5625, + "learning_rate": 3.815262352404495e-06, + "loss": 2.5787, + "step": 16262 + }, + { + "epoch": 1.7198604060913705, + "grad_norm": 8.1875, + "learning_rate": 3.815116709660552e-06, + "loss": 2.6491, + "step": 16263 + }, + { + "epoch": 1.7199661590524533, + "grad_norm": 8.4375, + "learning_rate": 3.814971060745305e-06, + "loss": 2.3891, + "step": 16264 + }, + { + "epoch": 1.7200719120135364, + "grad_norm": 8.25, + "learning_rate": 3.8148254056594355e-06, + "loss": 2.5933, + "step": 16265 + }, + { + "epoch": 1.7201776649746194, + "grad_norm": 8.75, + "learning_rate": 3.814679744403629e-06, + "loss": 2.7991, + "step": 16266 + }, + { + "epoch": 1.7202834179357023, + "grad_norm": 7.8125, + "learning_rate": 3.814534076978568e-06, + "loss": 2.5414, + "step": 16267 + }, + { + "epoch": 1.7203891708967851, + "grad_norm": 8.5, + "learning_rate": 3.814388403384936e-06, + "loss": 2.5212, + "step": 16268 + }, + { + "epoch": 1.720494923857868, + "grad_norm": 7.9375, + "learning_rate": 3.814242723623417e-06, + "loss": 2.5627, + "step": 16269 + }, + { + "epoch": 1.7206006768189508, + "grad_norm": 7.875, + "learning_rate": 3.814097037694694e-06, + "loss": 2.4702, + "step": 16270 + }, + { + "epoch": 1.7207064297800339, + "grad_norm": 8.125, + "learning_rate": 3.8139513455994514e-06, + "loss": 2.4192, + "step": 16271 + }, + { + "epoch": 1.720812182741117, + "grad_norm": 8.5625, + "learning_rate": 3.8138056473383724e-06, + "loss": 2.7094, + "step": 16272 + }, + { + "epoch": 1.7209179357021998, + "grad_norm": 8.3125, + "learning_rate": 3.813659942912141e-06, + "loss": 2.6828, + "step": 16273 + }, + { + "epoch": 1.7210236886632826, + "grad_norm": 8.5, + "learning_rate": 3.8135142323214413e-06, + "loss": 2.5505, + "step": 16274 + }, + { + "epoch": 1.7211294416243654, + "grad_norm": 8.6875, + "learning_rate": 3.813368515566955e-06, + "loss": 2.4916, + "step": 16275 + }, + { + "epoch": 1.7212351945854483, + "grad_norm": 8.3125, + "learning_rate": 3.813222792649368e-06, + "loss": 2.6537, + "step": 16276 + }, + { + "epoch": 1.7213409475465313, + "grad_norm": 8.0, + "learning_rate": 3.8130770635693644e-06, + "loss": 2.5622, + "step": 16277 + }, + { + "epoch": 1.7214467005076142, + "grad_norm": 9.0625, + "learning_rate": 3.812931328327627e-06, + "loss": 2.6523, + "step": 16278 + }, + { + "epoch": 1.7215524534686972, + "grad_norm": 8.125, + "learning_rate": 3.8127855869248386e-06, + "loss": 2.565, + "step": 16279 + }, + { + "epoch": 1.72165820642978, + "grad_norm": 8.25, + "learning_rate": 3.812639839361686e-06, + "loss": 2.5585, + "step": 16280 + }, + { + "epoch": 1.721763959390863, + "grad_norm": 8.25, + "learning_rate": 3.8124940856388504e-06, + "loss": 2.6575, + "step": 16281 + }, + { + "epoch": 1.7218697123519457, + "grad_norm": 7.96875, + "learning_rate": 3.8123483257570172e-06, + "loss": 2.3851, + "step": 16282 + }, + { + "epoch": 1.7219754653130288, + "grad_norm": 8.1875, + "learning_rate": 3.8122025597168696e-06, + "loss": 2.4065, + "step": 16283 + }, + { + "epoch": 1.7220812182741116, + "grad_norm": 7.84375, + "learning_rate": 3.8120567875190922e-06, + "loss": 2.496, + "step": 16284 + }, + { + "epoch": 1.7221869712351947, + "grad_norm": 8.1875, + "learning_rate": 3.8119110091643683e-06, + "loss": 2.6833, + "step": 16285 + }, + { + "epoch": 1.7222927241962775, + "grad_norm": 8.75, + "learning_rate": 3.8117652246533844e-06, + "loss": 2.7085, + "step": 16286 + }, + { + "epoch": 1.7223984771573604, + "grad_norm": 7.75, + "learning_rate": 3.811619433986822e-06, + "loss": 2.5115, + "step": 16287 + }, + { + "epoch": 1.7225042301184432, + "grad_norm": 8.5, + "learning_rate": 3.811473637165365e-06, + "loss": 2.7067, + "step": 16288 + }, + { + "epoch": 1.722609983079526, + "grad_norm": 7.90625, + "learning_rate": 3.8113278341896996e-06, + "loss": 2.4789, + "step": 16289 + }, + { + "epoch": 1.7227157360406091, + "grad_norm": 7.59375, + "learning_rate": 3.811182025060509e-06, + "loss": 2.5067, + "step": 16290 + }, + { + "epoch": 1.7228214890016922, + "grad_norm": 8.0, + "learning_rate": 3.811036209778477e-06, + "loss": 2.2145, + "step": 16291 + }, + { + "epoch": 1.722927241962775, + "grad_norm": 8.5, + "learning_rate": 3.810890388344288e-06, + "loss": 2.6665, + "step": 16292 + }, + { + "epoch": 1.7230329949238579, + "grad_norm": 8.1875, + "learning_rate": 3.810744560758628e-06, + "loss": 2.4235, + "step": 16293 + }, + { + "epoch": 1.7231387478849407, + "grad_norm": 7.8125, + "learning_rate": 3.810598727022179e-06, + "loss": 2.492, + "step": 16294 + }, + { + "epoch": 1.7232445008460235, + "grad_norm": 8.75, + "learning_rate": 3.8104528871356257e-06, + "loss": 2.8887, + "step": 16295 + }, + { + "epoch": 1.7233502538071066, + "grad_norm": 8.3125, + "learning_rate": 3.8103070410996536e-06, + "loss": 2.5026, + "step": 16296 + }, + { + "epoch": 1.7234560067681897, + "grad_norm": 7.75, + "learning_rate": 3.8101611889149457e-06, + "loss": 2.2766, + "step": 16297 + }, + { + "epoch": 1.7235617597292725, + "grad_norm": 8.3125, + "learning_rate": 3.8100153305821876e-06, + "loss": 2.5861, + "step": 16298 + }, + { + "epoch": 1.7236675126903553, + "grad_norm": 9.125, + "learning_rate": 3.8098694661020634e-06, + "loss": 2.6713, + "step": 16299 + }, + { + "epoch": 1.7237732656514382, + "grad_norm": 8.75, + "learning_rate": 3.8097235954752578e-06, + "loss": 2.651, + "step": 16300 + }, + { + "epoch": 1.723879018612521, + "grad_norm": 8.5625, + "learning_rate": 3.809577718702454e-06, + "loss": 2.6511, + "step": 16301 + }, + { + "epoch": 1.723984771573604, + "grad_norm": 8.0625, + "learning_rate": 3.8094318357843384e-06, + "loss": 2.5331, + "step": 16302 + }, + { + "epoch": 1.724090524534687, + "grad_norm": 7.875, + "learning_rate": 3.8092859467215947e-06, + "loss": 2.4276, + "step": 16303 + }, + { + "epoch": 1.72419627749577, + "grad_norm": 7.84375, + "learning_rate": 3.8091400515149073e-06, + "loss": 2.3391, + "step": 16304 + }, + { + "epoch": 1.7243020304568528, + "grad_norm": 8.5625, + "learning_rate": 3.808994150164962e-06, + "loss": 2.4534, + "step": 16305 + }, + { + "epoch": 1.7244077834179357, + "grad_norm": 8.8125, + "learning_rate": 3.8088482426724414e-06, + "loss": 2.4283, + "step": 16306 + }, + { + "epoch": 1.7245135363790185, + "grad_norm": 9.4375, + "learning_rate": 3.8087023290380317e-06, + "loss": 2.6964, + "step": 16307 + }, + { + "epoch": 1.7246192893401016, + "grad_norm": 8.625, + "learning_rate": 3.808556409262417e-06, + "loss": 2.6274, + "step": 16308 + }, + { + "epoch": 1.7247250423011844, + "grad_norm": 8.4375, + "learning_rate": 3.8084104833462826e-06, + "loss": 2.5165, + "step": 16309 + }, + { + "epoch": 1.7248307952622675, + "grad_norm": 7.875, + "learning_rate": 3.8082645512903128e-06, + "loss": 2.616, + "step": 16310 + }, + { + "epoch": 1.7249365482233503, + "grad_norm": 8.8125, + "learning_rate": 3.8081186130951918e-06, + "loss": 2.7041, + "step": 16311 + }, + { + "epoch": 1.7250423011844331, + "grad_norm": 8.9375, + "learning_rate": 3.807972668761606e-06, + "loss": 2.5341, + "step": 16312 + }, + { + "epoch": 1.725148054145516, + "grad_norm": 9.0, + "learning_rate": 3.8078267182902396e-06, + "loss": 2.2514, + "step": 16313 + }, + { + "epoch": 1.725253807106599, + "grad_norm": 8.0625, + "learning_rate": 3.8076807616817768e-06, + "loss": 2.5368, + "step": 16314 + }, + { + "epoch": 1.7253595600676819, + "grad_norm": 8.0, + "learning_rate": 3.8075347989369025e-06, + "loss": 2.5989, + "step": 16315 + }, + { + "epoch": 1.725465313028765, + "grad_norm": 7.875, + "learning_rate": 3.8073888300563032e-06, + "loss": 2.5751, + "step": 16316 + }, + { + "epoch": 1.7255710659898478, + "grad_norm": 7.96875, + "learning_rate": 3.8072428550406625e-06, + "loss": 2.4448, + "step": 16317 + }, + { + "epoch": 1.7256768189509306, + "grad_norm": 8.5, + "learning_rate": 3.8070968738906657e-06, + "loss": 2.8526, + "step": 16318 + }, + { + "epoch": 1.7257825719120135, + "grad_norm": 8.4375, + "learning_rate": 3.806950886606997e-06, + "loss": 2.6978, + "step": 16319 + }, + { + "epoch": 1.7258883248730963, + "grad_norm": 8.25, + "learning_rate": 3.8068048931903435e-06, + "loss": 2.5561, + "step": 16320 + }, + { + "epoch": 1.7259940778341794, + "grad_norm": 8.75, + "learning_rate": 3.8066588936413884e-06, + "loss": 2.5647, + "step": 16321 + }, + { + "epoch": 1.7260998307952624, + "grad_norm": 9.3125, + "learning_rate": 3.806512887960818e-06, + "loss": 2.7676, + "step": 16322 + }, + { + "epoch": 1.7262055837563453, + "grad_norm": 8.0625, + "learning_rate": 3.8063668761493165e-06, + "loss": 2.544, + "step": 16323 + }, + { + "epoch": 1.726311336717428, + "grad_norm": 8.4375, + "learning_rate": 3.80622085820757e-06, + "loss": 2.346, + "step": 16324 + }, + { + "epoch": 1.726417089678511, + "grad_norm": 8.4375, + "learning_rate": 3.806074834136263e-06, + "loss": 2.5091, + "step": 16325 + }, + { + "epoch": 1.7265228426395938, + "grad_norm": 8.0625, + "learning_rate": 3.805928803936081e-06, + "loss": 2.465, + "step": 16326 + }, + { + "epoch": 1.7266285956006768, + "grad_norm": 8.625, + "learning_rate": 3.805782767607709e-06, + "loss": 2.7336, + "step": 16327 + }, + { + "epoch": 1.72673434856176, + "grad_norm": 8.1875, + "learning_rate": 3.805636725151833e-06, + "loss": 2.6319, + "step": 16328 + }, + { + "epoch": 1.7268401015228427, + "grad_norm": 8.0625, + "learning_rate": 3.8054906765691374e-06, + "loss": 2.5719, + "step": 16329 + }, + { + "epoch": 1.7269458544839256, + "grad_norm": 8.375, + "learning_rate": 3.805344621860308e-06, + "loss": 2.3595, + "step": 16330 + }, + { + "epoch": 1.7270516074450084, + "grad_norm": 8.0625, + "learning_rate": 3.8051985610260308e-06, + "loss": 2.735, + "step": 16331 + }, + { + "epoch": 1.7271573604060912, + "grad_norm": 7.9375, + "learning_rate": 3.8050524940669898e-06, + "loss": 2.4156, + "step": 16332 + }, + { + "epoch": 1.7272631133671743, + "grad_norm": 8.6875, + "learning_rate": 3.8049064209838717e-06, + "loss": 2.5143, + "step": 16333 + }, + { + "epoch": 1.7273688663282571, + "grad_norm": 7.96875, + "learning_rate": 3.804760341777361e-06, + "loss": 2.5033, + "step": 16334 + }, + { + "epoch": 1.7274746192893402, + "grad_norm": 8.375, + "learning_rate": 3.8046142564481444e-06, + "loss": 2.5521, + "step": 16335 + }, + { + "epoch": 1.727580372250423, + "grad_norm": 8.5, + "learning_rate": 3.804468164996905e-06, + "loss": 2.4836, + "step": 16336 + }, + { + "epoch": 1.7276861252115059, + "grad_norm": 8.25, + "learning_rate": 3.8043220674243316e-06, + "loss": 2.5576, + "step": 16337 + }, + { + "epoch": 1.7277918781725887, + "grad_norm": 8.4375, + "learning_rate": 3.8041759637311077e-06, + "loss": 2.6137, + "step": 16338 + }, + { + "epoch": 1.7278976311336718, + "grad_norm": 8.0625, + "learning_rate": 3.8040298539179187e-06, + "loss": 2.6493, + "step": 16339 + }, + { + "epoch": 1.7280033840947546, + "grad_norm": 7.84375, + "learning_rate": 3.803883737985452e-06, + "loss": 2.4805, + "step": 16340 + }, + { + "epoch": 1.7281091370558377, + "grad_norm": 8.25, + "learning_rate": 3.803737615934392e-06, + "loss": 2.5763, + "step": 16341 + }, + { + "epoch": 1.7282148900169205, + "grad_norm": 8.0625, + "learning_rate": 3.8035914877654247e-06, + "loss": 2.5696, + "step": 16342 + }, + { + "epoch": 1.7283206429780034, + "grad_norm": 7.96875, + "learning_rate": 3.8034453534792355e-06, + "loss": 2.4618, + "step": 16343 + }, + { + "epoch": 1.7284263959390862, + "grad_norm": 9.5625, + "learning_rate": 3.80329921307651e-06, + "loss": 2.668, + "step": 16344 + }, + { + "epoch": 1.728532148900169, + "grad_norm": 8.0, + "learning_rate": 3.8031530665579347e-06, + "loss": 2.5126, + "step": 16345 + }, + { + "epoch": 1.728637901861252, + "grad_norm": 7.90625, + "learning_rate": 3.8030069139241945e-06, + "loss": 2.4686, + "step": 16346 + }, + { + "epoch": 1.7287436548223352, + "grad_norm": 7.78125, + "learning_rate": 3.8028607551759767e-06, + "loss": 2.5478, + "step": 16347 + }, + { + "epoch": 1.728849407783418, + "grad_norm": 8.4375, + "learning_rate": 3.8027145903139657e-06, + "loss": 2.452, + "step": 16348 + }, + { + "epoch": 1.7289551607445008, + "grad_norm": 8.125, + "learning_rate": 3.802568419338848e-06, + "loss": 2.5305, + "step": 16349 + }, + { + "epoch": 1.7290609137055837, + "grad_norm": 8.0625, + "learning_rate": 3.802422242251309e-06, + "loss": 2.2122, + "step": 16350 + }, + { + "epoch": 1.7291666666666665, + "grad_norm": 8.875, + "learning_rate": 3.802276059052036e-06, + "loss": 2.715, + "step": 16351 + }, + { + "epoch": 1.7292724196277496, + "grad_norm": 8.6875, + "learning_rate": 3.802129869741713e-06, + "loss": 2.6059, + "step": 16352 + }, + { + "epoch": 1.7293781725888326, + "grad_norm": 8.8125, + "learning_rate": 3.801983674321028e-06, + "loss": 2.759, + "step": 16353 + }, + { + "epoch": 1.7294839255499155, + "grad_norm": 8.375, + "learning_rate": 3.8018374727906648e-06, + "loss": 2.6221, + "step": 16354 + }, + { + "epoch": 1.7295896785109983, + "grad_norm": 8.0, + "learning_rate": 3.8016912651513118e-06, + "loss": 2.5638, + "step": 16355 + }, + { + "epoch": 1.7296954314720812, + "grad_norm": 8.375, + "learning_rate": 3.8015450514036544e-06, + "loss": 2.5875, + "step": 16356 + }, + { + "epoch": 1.729801184433164, + "grad_norm": 8.25, + "learning_rate": 3.801398831548378e-06, + "loss": 2.5906, + "step": 16357 + }, + { + "epoch": 1.729906937394247, + "grad_norm": 8.375, + "learning_rate": 3.8012526055861687e-06, + "loss": 2.4675, + "step": 16358 + }, + { + "epoch": 1.73001269035533, + "grad_norm": 8.0, + "learning_rate": 3.8011063735177134e-06, + "loss": 2.5282, + "step": 16359 + }, + { + "epoch": 1.730118443316413, + "grad_norm": 7.90625, + "learning_rate": 3.8009601353436985e-06, + "loss": 2.1845, + "step": 16360 + }, + { + "epoch": 1.7302241962774958, + "grad_norm": 8.1875, + "learning_rate": 3.8008138910648092e-06, + "loss": 2.5979, + "step": 16361 + }, + { + "epoch": 1.7303299492385786, + "grad_norm": 8.875, + "learning_rate": 3.800667640681732e-06, + "loss": 2.7312, + "step": 16362 + }, + { + "epoch": 1.7304357021996615, + "grad_norm": 8.25, + "learning_rate": 3.800521384195154e-06, + "loss": 2.387, + "step": 16363 + }, + { + "epoch": 1.7305414551607445, + "grad_norm": 7.90625, + "learning_rate": 3.8003751216057615e-06, + "loss": 2.4042, + "step": 16364 + }, + { + "epoch": 1.7306472081218274, + "grad_norm": 8.0625, + "learning_rate": 3.8002288529142396e-06, + "loss": 2.463, + "step": 16365 + }, + { + "epoch": 1.7307529610829104, + "grad_norm": 8.3125, + "learning_rate": 3.8000825781212756e-06, + "loss": 2.2892, + "step": 16366 + }, + { + "epoch": 1.7308587140439933, + "grad_norm": 8.375, + "learning_rate": 3.7999362972275568e-06, + "loss": 2.5714, + "step": 16367 + }, + { + "epoch": 1.7309644670050761, + "grad_norm": 8.1875, + "learning_rate": 3.799790010233767e-06, + "loss": 2.5779, + "step": 16368 + }, + { + "epoch": 1.731070219966159, + "grad_norm": 7.84375, + "learning_rate": 3.7996437171405953e-06, + "loss": 2.5162, + "step": 16369 + }, + { + "epoch": 1.731175972927242, + "grad_norm": 8.3125, + "learning_rate": 3.7994974179487278e-06, + "loss": 2.6109, + "step": 16370 + }, + { + "epoch": 1.7312817258883249, + "grad_norm": 8.9375, + "learning_rate": 3.799351112658849e-06, + "loss": 2.513, + "step": 16371 + }, + { + "epoch": 1.731387478849408, + "grad_norm": 8.125, + "learning_rate": 3.799204801271647e-06, + "loss": 2.5265, + "step": 16372 + }, + { + "epoch": 1.7314932318104908, + "grad_norm": 8.4375, + "learning_rate": 3.7990584837878084e-06, + "loss": 2.5715, + "step": 16373 + }, + { + "epoch": 1.7315989847715736, + "grad_norm": 8.9375, + "learning_rate": 3.7989121602080205e-06, + "loss": 2.659, + "step": 16374 + }, + { + "epoch": 1.7317047377326564, + "grad_norm": 8.6875, + "learning_rate": 3.798765830532968e-06, + "loss": 2.5352, + "step": 16375 + }, + { + "epoch": 1.7318104906937393, + "grad_norm": 7.78125, + "learning_rate": 3.7986194947633394e-06, + "loss": 2.4077, + "step": 16376 + }, + { + "epoch": 1.7319162436548223, + "grad_norm": 8.0625, + "learning_rate": 3.7984731528998198e-06, + "loss": 2.573, + "step": 16377 + }, + { + "epoch": 1.7320219966159054, + "grad_norm": 8.5, + "learning_rate": 3.798326804943098e-06, + "loss": 2.6586, + "step": 16378 + }, + { + "epoch": 1.7321277495769882, + "grad_norm": 8.375, + "learning_rate": 3.7981804508938585e-06, + "loss": 2.5231, + "step": 16379 + }, + { + "epoch": 1.732233502538071, + "grad_norm": 8.25, + "learning_rate": 3.798034090752789e-06, + "loss": 2.4215, + "step": 16380 + }, + { + "epoch": 1.732339255499154, + "grad_norm": 8.6875, + "learning_rate": 3.7978877245205763e-06, + "loss": 2.6478, + "step": 16381 + }, + { + "epoch": 1.7324450084602367, + "grad_norm": 8.25, + "learning_rate": 3.7977413521979077e-06, + "loss": 2.4353, + "step": 16382 + }, + { + "epoch": 1.7325507614213198, + "grad_norm": 8.0625, + "learning_rate": 3.79759497378547e-06, + "loss": 2.5512, + "step": 16383 + }, + { + "epoch": 1.7326565143824029, + "grad_norm": 7.96875, + "learning_rate": 3.79744858928395e-06, + "loss": 2.6285, + "step": 16384 + }, + { + "epoch": 1.7327622673434857, + "grad_norm": 8.25, + "learning_rate": 3.797302198694034e-06, + "loss": 2.623, + "step": 16385 + }, + { + "epoch": 1.7328680203045685, + "grad_norm": 8.375, + "learning_rate": 3.797155802016409e-06, + "loss": 2.6061, + "step": 16386 + }, + { + "epoch": 1.7329737732656514, + "grad_norm": 8.5, + "learning_rate": 3.797009399251762e-06, + "loss": 2.5944, + "step": 16387 + }, + { + "epoch": 1.7330795262267342, + "grad_norm": 8.75, + "learning_rate": 3.7968629904007815e-06, + "loss": 2.3259, + "step": 16388 + }, + { + "epoch": 1.7331852791878173, + "grad_norm": 7.78125, + "learning_rate": 3.7967165754641534e-06, + "loss": 2.433, + "step": 16389 + }, + { + "epoch": 1.7332910321489001, + "grad_norm": 8.5625, + "learning_rate": 3.796570154442563e-06, + "loss": 2.5577, + "step": 16390 + }, + { + "epoch": 1.7333967851099832, + "grad_norm": 8.375, + "learning_rate": 3.796423727336701e-06, + "loss": 2.5552, + "step": 16391 + }, + { + "epoch": 1.733502538071066, + "grad_norm": 8.3125, + "learning_rate": 3.796277294147252e-06, + "loss": 2.484, + "step": 16392 + }, + { + "epoch": 1.7336082910321489, + "grad_norm": 8.3125, + "learning_rate": 3.796130854874903e-06, + "loss": 2.5403, + "step": 16393 + }, + { + "epoch": 1.7337140439932317, + "grad_norm": 8.9375, + "learning_rate": 3.7959844095203436e-06, + "loss": 2.4703, + "step": 16394 + }, + { + "epoch": 1.7338197969543148, + "grad_norm": 8.75, + "learning_rate": 3.7958379580842586e-06, + "loss": 2.6992, + "step": 16395 + }, + { + "epoch": 1.7339255499153976, + "grad_norm": 8.25, + "learning_rate": 3.7956915005673355e-06, + "loss": 2.5738, + "step": 16396 + }, + { + "epoch": 1.7340313028764807, + "grad_norm": 8.1875, + "learning_rate": 3.795545036970263e-06, + "loss": 2.6907, + "step": 16397 + }, + { + "epoch": 1.7341370558375635, + "grad_norm": 9.125, + "learning_rate": 3.795398567293727e-06, + "loss": 2.5831, + "step": 16398 + }, + { + "epoch": 1.7342428087986463, + "grad_norm": 8.1875, + "learning_rate": 3.795252091538416e-06, + "loss": 2.5477, + "step": 16399 + }, + { + "epoch": 1.7343485617597292, + "grad_norm": 7.875, + "learning_rate": 3.7951056097050154e-06, + "loss": 2.7031, + "step": 16400 + }, + { + "epoch": 1.734454314720812, + "grad_norm": 8.25, + "learning_rate": 3.794959121794215e-06, + "loss": 2.7934, + "step": 16401 + }, + { + "epoch": 1.734560067681895, + "grad_norm": 8.0625, + "learning_rate": 3.794812627806701e-06, + "loss": 2.6197, + "step": 16402 + }, + { + "epoch": 1.7346658206429781, + "grad_norm": 7.84375, + "learning_rate": 3.7946661277431608e-06, + "loss": 2.4749, + "step": 16403 + }, + { + "epoch": 1.734771573604061, + "grad_norm": 8.125, + "learning_rate": 3.794519621604281e-06, + "loss": 2.5581, + "step": 16404 + }, + { + "epoch": 1.7348773265651438, + "grad_norm": 8.3125, + "learning_rate": 3.7943731093907513e-06, + "loss": 2.6903, + "step": 16405 + }, + { + "epoch": 1.7349830795262267, + "grad_norm": 8.25, + "learning_rate": 3.794226591103257e-06, + "loss": 2.5071, + "step": 16406 + }, + { + "epoch": 1.7350888324873095, + "grad_norm": 8.75, + "learning_rate": 3.794080066742487e-06, + "loss": 2.4072, + "step": 16407 + }, + { + "epoch": 1.7351945854483926, + "grad_norm": 8.0, + "learning_rate": 3.793933536309129e-06, + "loss": 2.5719, + "step": 16408 + }, + { + "epoch": 1.7353003384094756, + "grad_norm": 8.4375, + "learning_rate": 3.7937869998038694e-06, + "loss": 2.6804, + "step": 16409 + }, + { + "epoch": 1.7354060913705585, + "grad_norm": 8.3125, + "learning_rate": 3.793640457227397e-06, + "loss": 2.4586, + "step": 16410 + }, + { + "epoch": 1.7355118443316413, + "grad_norm": 8.875, + "learning_rate": 3.793493908580399e-06, + "loss": 2.5932, + "step": 16411 + }, + { + "epoch": 1.7356175972927241, + "grad_norm": 8.0625, + "learning_rate": 3.793347353863563e-06, + "loss": 2.5103, + "step": 16412 + }, + { + "epoch": 1.735723350253807, + "grad_norm": 8.6875, + "learning_rate": 3.793200793077577e-06, + "loss": 2.2725, + "step": 16413 + }, + { + "epoch": 1.73582910321489, + "grad_norm": 7.96875, + "learning_rate": 3.7930542262231283e-06, + "loss": 2.4267, + "step": 16414 + }, + { + "epoch": 1.7359348561759729, + "grad_norm": 8.3125, + "learning_rate": 3.7929076533009053e-06, + "loss": 2.6116, + "step": 16415 + }, + { + "epoch": 1.736040609137056, + "grad_norm": 9.25, + "learning_rate": 3.792761074311595e-06, + "loss": 2.7367, + "step": 16416 + }, + { + "epoch": 1.7361463620981388, + "grad_norm": 8.4375, + "learning_rate": 3.792614489255886e-06, + "loss": 2.3088, + "step": 16417 + }, + { + "epoch": 1.7362521150592216, + "grad_norm": 8.5, + "learning_rate": 3.792467898134466e-06, + "loss": 2.6222, + "step": 16418 + }, + { + "epoch": 1.7363578680203045, + "grad_norm": 8.6875, + "learning_rate": 3.792321300948022e-06, + "loss": 2.5135, + "step": 16419 + }, + { + "epoch": 1.7364636209813875, + "grad_norm": 9.0, + "learning_rate": 3.792174697697244e-06, + "loss": 2.3844, + "step": 16420 + }, + { + "epoch": 1.7365693739424704, + "grad_norm": 8.3125, + "learning_rate": 3.7920280883828175e-06, + "loss": 2.5886, + "step": 16421 + }, + { + "epoch": 1.7366751269035534, + "grad_norm": 8.3125, + "learning_rate": 3.791881473005432e-06, + "loss": 2.5481, + "step": 16422 + }, + { + "epoch": 1.7367808798646363, + "grad_norm": 8.125, + "learning_rate": 3.7917348515657748e-06, + "loss": 2.2814, + "step": 16423 + }, + { + "epoch": 1.736886632825719, + "grad_norm": 8.5, + "learning_rate": 3.7915882240645353e-06, + "loss": 2.5838, + "step": 16424 + }, + { + "epoch": 1.736992385786802, + "grad_norm": 8.375, + "learning_rate": 3.7914415905023987e-06, + "loss": 2.6112, + "step": 16425 + }, + { + "epoch": 1.737098138747885, + "grad_norm": 8.5625, + "learning_rate": 3.7912949508800563e-06, + "loss": 2.3907, + "step": 16426 + }, + { + "epoch": 1.7372038917089678, + "grad_norm": 9.0, + "learning_rate": 3.7911483051981945e-06, + "loss": 2.6234, + "step": 16427 + }, + { + "epoch": 1.737309644670051, + "grad_norm": 8.6875, + "learning_rate": 3.7910016534575013e-06, + "loss": 2.7097, + "step": 16428 + }, + { + "epoch": 1.7374153976311337, + "grad_norm": 8.0625, + "learning_rate": 3.790854995658666e-06, + "loss": 2.6171, + "step": 16429 + }, + { + "epoch": 1.7375211505922166, + "grad_norm": 8.0, + "learning_rate": 3.7907083318023763e-06, + "loss": 2.6444, + "step": 16430 + }, + { + "epoch": 1.7376269035532994, + "grad_norm": 8.0625, + "learning_rate": 3.7905616618893193e-06, + "loss": 2.6247, + "step": 16431 + }, + { + "epoch": 1.7377326565143822, + "grad_norm": 8.875, + "learning_rate": 3.790414985920185e-06, + "loss": 2.7642, + "step": 16432 + }, + { + "epoch": 1.7378384094754653, + "grad_norm": 7.84375, + "learning_rate": 3.790268303895661e-06, + "loss": 2.4673, + "step": 16433 + }, + { + "epoch": 1.7379441624365484, + "grad_norm": 9.0625, + "learning_rate": 3.790121615816435e-06, + "loss": 2.6338, + "step": 16434 + }, + { + "epoch": 1.7380499153976312, + "grad_norm": 8.25, + "learning_rate": 3.789974921683196e-06, + "loss": 2.5409, + "step": 16435 + }, + { + "epoch": 1.738155668358714, + "grad_norm": 9.625, + "learning_rate": 3.7898282214966327e-06, + "loss": 2.6656, + "step": 16436 + }, + { + "epoch": 1.7382614213197969, + "grad_norm": 8.4375, + "learning_rate": 3.789681515257433e-06, + "loss": 2.4923, + "step": 16437 + }, + { + "epoch": 1.7383671742808797, + "grad_norm": 7.71875, + "learning_rate": 3.7895348029662847e-06, + "loss": 2.383, + "step": 16438 + }, + { + "epoch": 1.7384729272419628, + "grad_norm": 9.0, + "learning_rate": 3.789388084623878e-06, + "loss": 2.7237, + "step": 16439 + }, + { + "epoch": 1.7385786802030458, + "grad_norm": 8.125, + "learning_rate": 3.7892413602308997e-06, + "loss": 2.4965, + "step": 16440 + }, + { + "epoch": 1.7386844331641287, + "grad_norm": 8.625, + "learning_rate": 3.7890946297880394e-06, + "loss": 2.6085, + "step": 16441 + }, + { + "epoch": 1.7387901861252115, + "grad_norm": 8.125, + "learning_rate": 3.788947893295985e-06, + "loss": 2.5741, + "step": 16442 + }, + { + "epoch": 1.7388959390862944, + "grad_norm": 8.3125, + "learning_rate": 3.7888011507554253e-06, + "loss": 2.6207, + "step": 16443 + }, + { + "epoch": 1.7390016920473772, + "grad_norm": 8.75, + "learning_rate": 3.788654402167048e-06, + "loss": 2.6886, + "step": 16444 + }, + { + "epoch": 1.7391074450084603, + "grad_norm": 7.75, + "learning_rate": 3.7885076475315442e-06, + "loss": 2.5428, + "step": 16445 + }, + { + "epoch": 1.739213197969543, + "grad_norm": 7.8125, + "learning_rate": 3.7883608868496e-06, + "loss": 2.3501, + "step": 16446 + }, + { + "epoch": 1.7393189509306262, + "grad_norm": 8.1875, + "learning_rate": 3.788214120121906e-06, + "loss": 2.5551, + "step": 16447 + }, + { + "epoch": 1.739424703891709, + "grad_norm": 8.9375, + "learning_rate": 3.788067347349149e-06, + "loss": 2.4819, + "step": 16448 + }, + { + "epoch": 1.7395304568527918, + "grad_norm": 8.4375, + "learning_rate": 3.7879205685320195e-06, + "loss": 2.624, + "step": 16449 + }, + { + "epoch": 1.7396362098138747, + "grad_norm": 8.5625, + "learning_rate": 3.787773783671205e-06, + "loss": 2.7601, + "step": 16450 + }, + { + "epoch": 1.7397419627749577, + "grad_norm": 8.0, + "learning_rate": 3.787626992767395e-06, + "loss": 2.4917, + "step": 16451 + }, + { + "epoch": 1.7398477157360406, + "grad_norm": 8.125, + "learning_rate": 3.787480195821278e-06, + "loss": 2.4089, + "step": 16452 + }, + { + "epoch": 1.7399534686971236, + "grad_norm": 7.9375, + "learning_rate": 3.7873333928335432e-06, + "loss": 2.7567, + "step": 16453 + }, + { + "epoch": 1.7400592216582065, + "grad_norm": 8.3125, + "learning_rate": 3.7871865838048787e-06, + "loss": 2.4695, + "step": 16454 + }, + { + "epoch": 1.7401649746192893, + "grad_norm": 8.75, + "learning_rate": 3.787039768735975e-06, + "loss": 2.6023, + "step": 16455 + }, + { + "epoch": 1.7402707275803722, + "grad_norm": 7.75, + "learning_rate": 3.7868929476275197e-06, + "loss": 2.4554, + "step": 16456 + }, + { + "epoch": 1.740376480541455, + "grad_norm": 8.25, + "learning_rate": 3.7867461204802013e-06, + "loss": 2.5765, + "step": 16457 + }, + { + "epoch": 1.740482233502538, + "grad_norm": 8.625, + "learning_rate": 3.786599287294711e-06, + "loss": 2.7031, + "step": 16458 + }, + { + "epoch": 1.7405879864636211, + "grad_norm": 8.4375, + "learning_rate": 3.7864524480717356e-06, + "loss": 2.4552, + "step": 16459 + }, + { + "epoch": 1.740693739424704, + "grad_norm": 7.78125, + "learning_rate": 3.7863056028119644e-06, + "loss": 2.5254, + "step": 16460 + }, + { + "epoch": 1.7407994923857868, + "grad_norm": 8.1875, + "learning_rate": 3.786158751516088e-06, + "loss": 2.4724, + "step": 16461 + }, + { + "epoch": 1.7409052453468696, + "grad_norm": 8.6875, + "learning_rate": 3.7860118941847946e-06, + "loss": 2.766, + "step": 16462 + }, + { + "epoch": 1.7410109983079525, + "grad_norm": 8.125, + "learning_rate": 3.7858650308187726e-06, + "loss": 2.2129, + "step": 16463 + }, + { + "epoch": 1.7411167512690355, + "grad_norm": 8.4375, + "learning_rate": 3.785718161418713e-06, + "loss": 2.4404, + "step": 16464 + }, + { + "epoch": 1.7412225042301186, + "grad_norm": 7.9375, + "learning_rate": 3.785571285985303e-06, + "loss": 2.5187, + "step": 16465 + }, + { + "epoch": 1.7413282571912014, + "grad_norm": 8.625, + "learning_rate": 3.7854244045192335e-06, + "loss": 2.6083, + "step": 16466 + }, + { + "epoch": 1.7414340101522843, + "grad_norm": 8.9375, + "learning_rate": 3.7852775170211924e-06, + "loss": 2.6197, + "step": 16467 + }, + { + "epoch": 1.7415397631133671, + "grad_norm": 8.125, + "learning_rate": 3.7851306234918705e-06, + "loss": 2.4923, + "step": 16468 + }, + { + "epoch": 1.74164551607445, + "grad_norm": 8.1875, + "learning_rate": 3.784983723931955e-06, + "loss": 2.5779, + "step": 16469 + }, + { + "epoch": 1.741751269035533, + "grad_norm": 8.5, + "learning_rate": 3.784836818342136e-06, + "loss": 2.5845, + "step": 16470 + }, + { + "epoch": 1.7418570219966159, + "grad_norm": 8.0, + "learning_rate": 3.784689906723105e-06, + "loss": 2.4884, + "step": 16471 + }, + { + "epoch": 1.741962774957699, + "grad_norm": 8.25, + "learning_rate": 3.7845429890755493e-06, + "loss": 2.5386, + "step": 16472 + }, + { + "epoch": 1.7420685279187818, + "grad_norm": 8.5625, + "learning_rate": 3.7843960654001576e-06, + "loss": 2.5247, + "step": 16473 + }, + { + "epoch": 1.7421742808798646, + "grad_norm": 8.125, + "learning_rate": 3.784249135697621e-06, + "loss": 2.5402, + "step": 16474 + }, + { + "epoch": 1.7422800338409474, + "grad_norm": 8.4375, + "learning_rate": 3.784102199968629e-06, + "loss": 2.5179, + "step": 16475 + }, + { + "epoch": 1.7423857868020305, + "grad_norm": 8.9375, + "learning_rate": 3.78395525821387e-06, + "loss": 2.4764, + "step": 16476 + }, + { + "epoch": 1.7424915397631133, + "grad_norm": 8.3125, + "learning_rate": 3.7838083104340345e-06, + "loss": 2.5419, + "step": 16477 + }, + { + "epoch": 1.7425972927241964, + "grad_norm": 8.25, + "learning_rate": 3.783661356629811e-06, + "loss": 2.4737, + "step": 16478 + }, + { + "epoch": 1.7427030456852792, + "grad_norm": 8.8125, + "learning_rate": 3.78351439680189e-06, + "loss": 2.6074, + "step": 16479 + }, + { + "epoch": 1.742808798646362, + "grad_norm": 8.9375, + "learning_rate": 3.7833674309509612e-06, + "loss": 2.5616, + "step": 16480 + }, + { + "epoch": 1.742914551607445, + "grad_norm": 8.1875, + "learning_rate": 3.783220459077714e-06, + "loss": 2.4668, + "step": 16481 + }, + { + "epoch": 1.743020304568528, + "grad_norm": 8.25, + "learning_rate": 3.7830734811828372e-06, + "loss": 2.4622, + "step": 16482 + }, + { + "epoch": 1.7431260575296108, + "grad_norm": 8.3125, + "learning_rate": 3.782926497267022e-06, + "loss": 2.3026, + "step": 16483 + }, + { + "epoch": 1.7432318104906939, + "grad_norm": 8.375, + "learning_rate": 3.782779507330958e-06, + "loss": 2.3647, + "step": 16484 + }, + { + "epoch": 1.7433375634517767, + "grad_norm": 8.125, + "learning_rate": 3.7826325113753333e-06, + "loss": 2.6155, + "step": 16485 + }, + { + "epoch": 1.7434433164128595, + "grad_norm": 8.3125, + "learning_rate": 3.7824855094008393e-06, + "loss": 2.6905, + "step": 16486 + }, + { + "epoch": 1.7435490693739424, + "grad_norm": 7.9375, + "learning_rate": 3.782338501408165e-06, + "loss": 2.4459, + "step": 16487 + }, + { + "epoch": 1.7436548223350252, + "grad_norm": 8.6875, + "learning_rate": 3.782191487398001e-06, + "loss": 2.1673, + "step": 16488 + }, + { + "epoch": 1.7437605752961083, + "grad_norm": 8.3125, + "learning_rate": 3.7820444673710364e-06, + "loss": 2.5368, + "step": 16489 + }, + { + "epoch": 1.7438663282571913, + "grad_norm": 7.90625, + "learning_rate": 3.781897441327962e-06, + "loss": 2.5721, + "step": 16490 + }, + { + "epoch": 1.7439720812182742, + "grad_norm": 7.75, + "learning_rate": 3.781750409269467e-06, + "loss": 2.5016, + "step": 16491 + }, + { + "epoch": 1.744077834179357, + "grad_norm": 8.4375, + "learning_rate": 3.7816033711962406e-06, + "loss": 2.4829, + "step": 16492 + }, + { + "epoch": 1.7441835871404399, + "grad_norm": 9.625, + "learning_rate": 3.781456327108975e-06, + "loss": 2.5683, + "step": 16493 + }, + { + "epoch": 1.7442893401015227, + "grad_norm": 8.5625, + "learning_rate": 3.7813092770083583e-06, + "loss": 2.5914, + "step": 16494 + }, + { + "epoch": 1.7443950930626058, + "grad_norm": 8.25, + "learning_rate": 3.7811622208950814e-06, + "loss": 2.4408, + "step": 16495 + }, + { + "epoch": 1.7445008460236888, + "grad_norm": 8.1875, + "learning_rate": 3.781015158769834e-06, + "loss": 2.6119, + "step": 16496 + }, + { + "epoch": 1.7446065989847717, + "grad_norm": 8.25, + "learning_rate": 3.7808680906333064e-06, + "loss": 2.3668, + "step": 16497 + }, + { + "epoch": 1.7447123519458545, + "grad_norm": 7.84375, + "learning_rate": 3.7807210164861885e-06, + "loss": 2.3445, + "step": 16498 + }, + { + "epoch": 1.7448181049069373, + "grad_norm": 8.25, + "learning_rate": 3.7805739363291716e-06, + "loss": 2.5276, + "step": 16499 + }, + { + "epoch": 1.7449238578680202, + "grad_norm": 8.75, + "learning_rate": 3.7804268501629438e-06, + "loss": 2.353, + "step": 16500 + }, + { + "epoch": 1.7450296108291032, + "grad_norm": 8.5625, + "learning_rate": 3.7802797579881967e-06, + "loss": 2.3993, + "step": 16501 + }, + { + "epoch": 1.745135363790186, + "grad_norm": 8.1875, + "learning_rate": 3.780132659805621e-06, + "loss": 2.5289, + "step": 16502 + }, + { + "epoch": 1.7452411167512691, + "grad_norm": 8.125, + "learning_rate": 3.779985555615906e-06, + "loss": 2.7007, + "step": 16503 + }, + { + "epoch": 1.745346869712352, + "grad_norm": 8.3125, + "learning_rate": 3.7798384454197423e-06, + "loss": 2.5464, + "step": 16504 + }, + { + "epoch": 1.7454526226734348, + "grad_norm": 8.0625, + "learning_rate": 3.7796913292178194e-06, + "loss": 2.5623, + "step": 16505 + }, + { + "epoch": 1.7455583756345177, + "grad_norm": 7.65625, + "learning_rate": 3.7795442070108297e-06, + "loss": 2.5189, + "step": 16506 + }, + { + "epoch": 1.7456641285956007, + "grad_norm": 8.25, + "learning_rate": 3.7793970787994614e-06, + "loss": 2.3981, + "step": 16507 + }, + { + "epoch": 1.7457698815566836, + "grad_norm": 8.625, + "learning_rate": 3.7792499445844066e-06, + "loss": 2.5301, + "step": 16508 + }, + { + "epoch": 1.7458756345177666, + "grad_norm": 8.375, + "learning_rate": 3.779102804366355e-06, + "loss": 2.3551, + "step": 16509 + }, + { + "epoch": 1.7459813874788495, + "grad_norm": 7.875, + "learning_rate": 3.7789556581459963e-06, + "loss": 2.5253, + "step": 16510 + }, + { + "epoch": 1.7460871404399323, + "grad_norm": 8.1875, + "learning_rate": 3.778808505924022e-06, + "loss": 2.4887, + "step": 16511 + }, + { + "epoch": 1.7461928934010151, + "grad_norm": 8.375, + "learning_rate": 3.7786613477011235e-06, + "loss": 2.4066, + "step": 16512 + }, + { + "epoch": 1.746298646362098, + "grad_norm": 8.3125, + "learning_rate": 3.7785141834779885e-06, + "loss": 2.5617, + "step": 16513 + }, + { + "epoch": 1.746404399323181, + "grad_norm": 7.84375, + "learning_rate": 3.7783670132553103e-06, + "loss": 2.428, + "step": 16514 + }, + { + "epoch": 1.746510152284264, + "grad_norm": 9.0, + "learning_rate": 3.7782198370337788e-06, + "loss": 2.7473, + "step": 16515 + }, + { + "epoch": 1.746615905245347, + "grad_norm": 8.3125, + "learning_rate": 3.7780726548140838e-06, + "loss": 2.305, + "step": 16516 + }, + { + "epoch": 1.7467216582064298, + "grad_norm": 8.125, + "learning_rate": 3.7779254665969165e-06, + "loss": 2.6217, + "step": 16517 + }, + { + "epoch": 1.7468274111675126, + "grad_norm": 8.3125, + "learning_rate": 3.777778272382968e-06, + "loss": 2.5659, + "step": 16518 + }, + { + "epoch": 1.7469331641285955, + "grad_norm": 8.5, + "learning_rate": 3.777631072172929e-06, + "loss": 2.3324, + "step": 16519 + }, + { + "epoch": 1.7470389170896785, + "grad_norm": 8.625, + "learning_rate": 3.7774838659674896e-06, + "loss": 2.3179, + "step": 16520 + }, + { + "epoch": 1.7471446700507616, + "grad_norm": 8.25, + "learning_rate": 3.777336653767341e-06, + "loss": 2.5402, + "step": 16521 + }, + { + "epoch": 1.7472504230118444, + "grad_norm": 8.375, + "learning_rate": 3.777189435573173e-06, + "loss": 2.6089, + "step": 16522 + }, + { + "epoch": 1.7473561759729273, + "grad_norm": 8.0625, + "learning_rate": 3.777042211385678e-06, + "loss": 2.4604, + "step": 16523 + }, + { + "epoch": 1.74746192893401, + "grad_norm": 8.6875, + "learning_rate": 3.776894981205546e-06, + "loss": 2.5478, + "step": 16524 + }, + { + "epoch": 1.747567681895093, + "grad_norm": 8.25, + "learning_rate": 3.776747745033469e-06, + "loss": 2.6542, + "step": 16525 + }, + { + "epoch": 1.747673434856176, + "grad_norm": 8.5, + "learning_rate": 3.7766005028701357e-06, + "loss": 2.5282, + "step": 16526 + }, + { + "epoch": 1.7477791878172588, + "grad_norm": 9.0625, + "learning_rate": 3.776453254716239e-06, + "loss": 2.911, + "step": 16527 + }, + { + "epoch": 1.747884940778342, + "grad_norm": 7.90625, + "learning_rate": 3.7763060005724692e-06, + "loss": 2.5985, + "step": 16528 + }, + { + "epoch": 1.7479906937394247, + "grad_norm": 8.125, + "learning_rate": 3.776158740439517e-06, + "loss": 2.3537, + "step": 16529 + }, + { + "epoch": 1.7480964467005076, + "grad_norm": 9.125, + "learning_rate": 3.7760114743180743e-06, + "loss": 2.5782, + "step": 16530 + }, + { + "epoch": 1.7482021996615904, + "grad_norm": 8.1875, + "learning_rate": 3.775864202208831e-06, + "loss": 2.6869, + "step": 16531 + }, + { + "epoch": 1.7483079526226735, + "grad_norm": 8.25, + "learning_rate": 3.7757169241124793e-06, + "loss": 2.5075, + "step": 16532 + }, + { + "epoch": 1.7484137055837563, + "grad_norm": 8.0625, + "learning_rate": 3.775569640029709e-06, + "loss": 2.4704, + "step": 16533 + }, + { + "epoch": 1.7485194585448394, + "grad_norm": 8.4375, + "learning_rate": 3.775422349961214e-06, + "loss": 2.5741, + "step": 16534 + }, + { + "epoch": 1.7486252115059222, + "grad_norm": 7.90625, + "learning_rate": 3.7752750539076823e-06, + "loss": 2.4779, + "step": 16535 + }, + { + "epoch": 1.748730964467005, + "grad_norm": 7.84375, + "learning_rate": 3.775127751869806e-06, + "loss": 2.4651, + "step": 16536 + }, + { + "epoch": 1.7488367174280879, + "grad_norm": 8.0625, + "learning_rate": 3.7749804438482774e-06, + "loss": 2.6227, + "step": 16537 + }, + { + "epoch": 1.748942470389171, + "grad_norm": 8.5, + "learning_rate": 3.7748331298437867e-06, + "loss": 2.3345, + "step": 16538 + }, + { + "epoch": 1.7490482233502538, + "grad_norm": 8.6875, + "learning_rate": 3.774685809857025e-06, + "loss": 2.4918, + "step": 16539 + }, + { + "epoch": 1.7491539763113368, + "grad_norm": 8.3125, + "learning_rate": 3.7745384838886845e-06, + "loss": 2.1718, + "step": 16540 + }, + { + "epoch": 1.7492597292724197, + "grad_norm": 8.5, + "learning_rate": 3.774391151939456e-06, + "loss": 2.4216, + "step": 16541 + }, + { + "epoch": 1.7493654822335025, + "grad_norm": 8.5625, + "learning_rate": 3.774243814010032e-06, + "loss": 2.5117, + "step": 16542 + }, + { + "epoch": 1.7494712351945854, + "grad_norm": 7.96875, + "learning_rate": 3.774096470101102e-06, + "loss": 2.545, + "step": 16543 + }, + { + "epoch": 1.7495769881556682, + "grad_norm": 8.125, + "learning_rate": 3.773949120213358e-06, + "loss": 2.551, + "step": 16544 + }, + { + "epoch": 1.7496827411167513, + "grad_norm": 7.96875, + "learning_rate": 3.7738017643474934e-06, + "loss": 2.4579, + "step": 16545 + }, + { + "epoch": 1.7497884940778343, + "grad_norm": 8.0625, + "learning_rate": 3.7736544025041967e-06, + "loss": 2.6119, + "step": 16546 + }, + { + "epoch": 1.7498942470389172, + "grad_norm": 8.5, + "learning_rate": 3.773507034684161e-06, + "loss": 2.6807, + "step": 16547 + }, + { + "epoch": 1.75, + "grad_norm": 8.375, + "learning_rate": 3.7733596608880784e-06, + "loss": 2.7584, + "step": 16548 + }, + { + "epoch": 1.7501057529610828, + "grad_norm": 8.75, + "learning_rate": 3.7732122811166382e-06, + "loss": 2.7713, + "step": 16549 + }, + { + "epoch": 1.7502115059221657, + "grad_norm": 7.6875, + "learning_rate": 3.7730648953705347e-06, + "loss": 2.358, + "step": 16550 + }, + { + "epoch": 1.7503172588832487, + "grad_norm": 7.90625, + "learning_rate": 3.7729175036504578e-06, + "loss": 2.4965, + "step": 16551 + }, + { + "epoch": 1.7504230118443318, + "grad_norm": 8.625, + "learning_rate": 3.7727701059570998e-06, + "loss": 2.4362, + "step": 16552 + }, + { + "epoch": 1.7505287648054146, + "grad_norm": 8.3125, + "learning_rate": 3.772622702291152e-06, + "loss": 2.6927, + "step": 16553 + }, + { + "epoch": 1.7506345177664975, + "grad_norm": 8.75, + "learning_rate": 3.7724752926533072e-06, + "loss": 2.6489, + "step": 16554 + }, + { + "epoch": 1.7507402707275803, + "grad_norm": 8.3125, + "learning_rate": 3.7723278770442554e-06, + "loss": 2.6291, + "step": 16555 + }, + { + "epoch": 1.7508460236886632, + "grad_norm": 8.4375, + "learning_rate": 3.7721804554646903e-06, + "loss": 2.7181, + "step": 16556 + }, + { + "epoch": 1.7509517766497462, + "grad_norm": 7.96875, + "learning_rate": 3.772033027915302e-06, + "loss": 2.5447, + "step": 16557 + }, + { + "epoch": 1.751057529610829, + "grad_norm": 8.125, + "learning_rate": 3.7718855943967823e-06, + "loss": 2.4975, + "step": 16558 + }, + { + "epoch": 1.7511632825719121, + "grad_norm": 8.4375, + "learning_rate": 3.7717381549098234e-06, + "loss": 2.6868, + "step": 16559 + }, + { + "epoch": 1.751269035532995, + "grad_norm": 8.125, + "learning_rate": 3.7715907094551185e-06, + "loss": 2.4429, + "step": 16560 + }, + { + "epoch": 1.7513747884940778, + "grad_norm": 8.0625, + "learning_rate": 3.7714432580333587e-06, + "loss": 2.2417, + "step": 16561 + }, + { + "epoch": 1.7514805414551606, + "grad_norm": 8.25, + "learning_rate": 3.7712958006452348e-06, + "loss": 2.1426, + "step": 16562 + }, + { + "epoch": 1.7515862944162437, + "grad_norm": 8.5625, + "learning_rate": 3.7711483372914404e-06, + "loss": 2.7917, + "step": 16563 + }, + { + "epoch": 1.7516920473773265, + "grad_norm": 8.375, + "learning_rate": 3.771000867972667e-06, + "loss": 2.6891, + "step": 16564 + }, + { + "epoch": 1.7517978003384096, + "grad_norm": 8.6875, + "learning_rate": 3.770853392689605e-06, + "loss": 2.4584, + "step": 16565 + }, + { + "epoch": 1.7519035532994924, + "grad_norm": 9.1875, + "learning_rate": 3.7707059114429484e-06, + "loss": 2.6596, + "step": 16566 + }, + { + "epoch": 1.7520093062605753, + "grad_norm": 8.8125, + "learning_rate": 3.7705584242333892e-06, + "loss": 2.3473, + "step": 16567 + }, + { + "epoch": 1.752115059221658, + "grad_norm": 8.0625, + "learning_rate": 3.7704109310616177e-06, + "loss": 2.5919, + "step": 16568 + }, + { + "epoch": 1.7522208121827412, + "grad_norm": 8.375, + "learning_rate": 3.7702634319283288e-06, + "loss": 2.6883, + "step": 16569 + }, + { + "epoch": 1.752326565143824, + "grad_norm": 8.375, + "learning_rate": 3.7701159268342123e-06, + "loss": 2.6416, + "step": 16570 + }, + { + "epoch": 1.752432318104907, + "grad_norm": 11.0, + "learning_rate": 3.7699684157799617e-06, + "loss": 2.5135, + "step": 16571 + }, + { + "epoch": 1.75253807106599, + "grad_norm": 8.8125, + "learning_rate": 3.769820898766268e-06, + "loss": 2.6121, + "step": 16572 + }, + { + "epoch": 1.7526438240270727, + "grad_norm": 8.5, + "learning_rate": 3.7696733757938255e-06, + "loss": 2.6436, + "step": 16573 + }, + { + "epoch": 1.7527495769881556, + "grad_norm": 8.25, + "learning_rate": 3.7695258468633235e-06, + "loss": 2.5427, + "step": 16574 + }, + { + "epoch": 1.7528553299492384, + "grad_norm": 8.25, + "learning_rate": 3.7693783119754568e-06, + "loss": 2.774, + "step": 16575 + }, + { + "epoch": 1.7529610829103215, + "grad_norm": 8.0625, + "learning_rate": 3.7692307711309163e-06, + "loss": 2.6994, + "step": 16576 + }, + { + "epoch": 1.7530668358714045, + "grad_norm": 7.96875, + "learning_rate": 3.7690832243303955e-06, + "loss": 2.538, + "step": 16577 + }, + { + "epoch": 1.7531725888324874, + "grad_norm": 8.125, + "learning_rate": 3.7689356715745856e-06, + "loss": 2.5326, + "step": 16578 + }, + { + "epoch": 1.7532783417935702, + "grad_norm": 8.0625, + "learning_rate": 3.76878811286418e-06, + "loss": 2.5389, + "step": 16579 + }, + { + "epoch": 1.753384094754653, + "grad_norm": 8.5625, + "learning_rate": 3.7686405481998713e-06, + "loss": 2.6971, + "step": 16580 + }, + { + "epoch": 1.753489847715736, + "grad_norm": 8.0625, + "learning_rate": 3.76849297758235e-06, + "loss": 2.4464, + "step": 16581 + }, + { + "epoch": 1.753595600676819, + "grad_norm": 8.625, + "learning_rate": 3.768345401012311e-06, + "loss": 2.574, + "step": 16582 + }, + { + "epoch": 1.753701353637902, + "grad_norm": 8.0625, + "learning_rate": 3.7681978184904453e-06, + "loss": 2.4643, + "step": 16583 + }, + { + "epoch": 1.7538071065989849, + "grad_norm": 7.90625, + "learning_rate": 3.7680502300174456e-06, + "loss": 2.5802, + "step": 16584 + }, + { + "epoch": 1.7539128595600677, + "grad_norm": 8.4375, + "learning_rate": 3.767902635594005e-06, + "loss": 2.338, + "step": 16585 + }, + { + "epoch": 1.7540186125211505, + "grad_norm": 8.3125, + "learning_rate": 3.7677550352208166e-06, + "loss": 2.4108, + "step": 16586 + }, + { + "epoch": 1.7541243654822334, + "grad_norm": 8.1875, + "learning_rate": 3.767607428898571e-06, + "loss": 2.6179, + "step": 16587 + }, + { + "epoch": 1.7542301184433164, + "grad_norm": 8.6875, + "learning_rate": 3.767459816627963e-06, + "loss": 2.5579, + "step": 16588 + }, + { + "epoch": 1.7543358714043993, + "grad_norm": 9.4375, + "learning_rate": 3.767312198409685e-06, + "loss": 2.377, + "step": 16589 + }, + { + "epoch": 1.7544416243654823, + "grad_norm": 8.1875, + "learning_rate": 3.7671645742444284e-06, + "loss": 2.456, + "step": 16590 + }, + { + "epoch": 1.7545473773265652, + "grad_norm": 7.90625, + "learning_rate": 3.767016944132887e-06, + "loss": 2.646, + "step": 16591 + }, + { + "epoch": 1.754653130287648, + "grad_norm": 8.3125, + "learning_rate": 3.7668693080757535e-06, + "loss": 2.6478, + "step": 16592 + }, + { + "epoch": 1.7547588832487309, + "grad_norm": 8.125, + "learning_rate": 3.7667216660737193e-06, + "loss": 2.6339, + "step": 16593 + }, + { + "epoch": 1.754864636209814, + "grad_norm": 8.1875, + "learning_rate": 3.766574018127479e-06, + "loss": 2.5805, + "step": 16594 + }, + { + "epoch": 1.7549703891708968, + "grad_norm": 8.5625, + "learning_rate": 3.7664263642377253e-06, + "loss": 2.3476, + "step": 16595 + }, + { + "epoch": 1.7550761421319798, + "grad_norm": 8.75, + "learning_rate": 3.7662787044051508e-06, + "loss": 2.5031, + "step": 16596 + }, + { + "epoch": 1.7551818950930627, + "grad_norm": 8.0625, + "learning_rate": 3.7661310386304474e-06, + "loss": 2.3425, + "step": 16597 + }, + { + "epoch": 1.7552876480541455, + "grad_norm": 8.25, + "learning_rate": 3.76598336691431e-06, + "loss": 2.6001, + "step": 16598 + }, + { + "epoch": 1.7553934010152283, + "grad_norm": 8.0, + "learning_rate": 3.765835689257429e-06, + "loss": 2.4868, + "step": 16599 + }, + { + "epoch": 1.7554991539763112, + "grad_norm": 8.5625, + "learning_rate": 3.7656880056604996e-06, + "loss": 2.5866, + "step": 16600 + }, + { + "epoch": 1.7556049069373942, + "grad_norm": 8.25, + "learning_rate": 3.7655403161242143e-06, + "loss": 2.4949, + "step": 16601 + }, + { + "epoch": 1.7557106598984773, + "grad_norm": 8.625, + "learning_rate": 3.765392620649265e-06, + "loss": 2.5616, + "step": 16602 + }, + { + "epoch": 1.7558164128595601, + "grad_norm": 8.375, + "learning_rate": 3.765244919236346e-06, + "loss": 2.4827, + "step": 16603 + }, + { + "epoch": 1.755922165820643, + "grad_norm": 9.1875, + "learning_rate": 3.7650972118861505e-06, + "loss": 2.6461, + "step": 16604 + }, + { + "epoch": 1.7560279187817258, + "grad_norm": 7.71875, + "learning_rate": 3.764949498599371e-06, + "loss": 2.4005, + "step": 16605 + }, + { + "epoch": 1.7561336717428087, + "grad_norm": 8.0625, + "learning_rate": 3.764801779376701e-06, + "loss": 2.5542, + "step": 16606 + }, + { + "epoch": 1.7562394247038917, + "grad_norm": 9.25, + "learning_rate": 3.764654054218833e-06, + "loss": 2.6181, + "step": 16607 + }, + { + "epoch": 1.7563451776649748, + "grad_norm": 8.125, + "learning_rate": 3.764506323126461e-06, + "loss": 2.539, + "step": 16608 + }, + { + "epoch": 1.7564509306260576, + "grad_norm": 8.875, + "learning_rate": 3.7643585861002784e-06, + "loss": 2.4508, + "step": 16609 + }, + { + "epoch": 1.7565566835871405, + "grad_norm": 8.5, + "learning_rate": 3.764210843140978e-06, + "loss": 2.4615, + "step": 16610 + }, + { + "epoch": 1.7566624365482233, + "grad_norm": 8.125, + "learning_rate": 3.7640630942492524e-06, + "loss": 2.4415, + "step": 16611 + }, + { + "epoch": 1.7567681895093061, + "grad_norm": 7.90625, + "learning_rate": 3.7639153394257966e-06, + "loss": 2.5537, + "step": 16612 + }, + { + "epoch": 1.7568739424703892, + "grad_norm": 8.1875, + "learning_rate": 3.763767578671302e-06, + "loss": 2.2538, + "step": 16613 + }, + { + "epoch": 1.756979695431472, + "grad_norm": 8.125, + "learning_rate": 3.763619811986464e-06, + "loss": 2.4076, + "step": 16614 + }, + { + "epoch": 1.757085448392555, + "grad_norm": 8.375, + "learning_rate": 3.763472039371975e-06, + "loss": 2.6139, + "step": 16615 + }, + { + "epoch": 1.757191201353638, + "grad_norm": 8.375, + "learning_rate": 3.7633242608285273e-06, + "loss": 2.7013, + "step": 16616 + }, + { + "epoch": 1.7572969543147208, + "grad_norm": 9.4375, + "learning_rate": 3.763176476356817e-06, + "loss": 2.3791, + "step": 16617 + }, + { + "epoch": 1.7574027072758036, + "grad_norm": 8.25, + "learning_rate": 3.7630286859575356e-06, + "loss": 2.5201, + "step": 16618 + }, + { + "epoch": 1.7575084602368867, + "grad_norm": 8.5, + "learning_rate": 3.7628808896313767e-06, + "loss": 2.5572, + "step": 16619 + }, + { + "epoch": 1.7576142131979695, + "grad_norm": 8.375, + "learning_rate": 3.762733087379035e-06, + "loss": 2.6044, + "step": 16620 + }, + { + "epoch": 1.7577199661590526, + "grad_norm": 8.3125, + "learning_rate": 3.7625852792012024e-06, + "loss": 2.5475, + "step": 16621 + }, + { + "epoch": 1.7578257191201354, + "grad_norm": 9.1875, + "learning_rate": 3.762437465098574e-06, + "loss": 2.6089, + "step": 16622 + }, + { + "epoch": 1.7579314720812182, + "grad_norm": 8.625, + "learning_rate": 3.7622896450718428e-06, + "loss": 2.5443, + "step": 16623 + }, + { + "epoch": 1.758037225042301, + "grad_norm": 8.125, + "learning_rate": 3.762141819121703e-06, + "loss": 2.4335, + "step": 16624 + }, + { + "epoch": 1.7581429780033841, + "grad_norm": 8.3125, + "learning_rate": 3.761993987248847e-06, + "loss": 2.6222, + "step": 16625 + }, + { + "epoch": 1.758248730964467, + "grad_norm": 8.0625, + "learning_rate": 3.7618461494539704e-06, + "loss": 2.5573, + "step": 16626 + }, + { + "epoch": 1.75835448392555, + "grad_norm": 8.25, + "learning_rate": 3.7616983057377653e-06, + "loss": 2.533, + "step": 16627 + }, + { + "epoch": 1.7584602368866329, + "grad_norm": 7.875, + "learning_rate": 3.7615504561009256e-06, + "loss": 2.4378, + "step": 16628 + }, + { + "epoch": 1.7585659898477157, + "grad_norm": 8.375, + "learning_rate": 3.761402600544146e-06, + "loss": 2.6151, + "step": 16629 + }, + { + "epoch": 1.7586717428087986, + "grad_norm": 8.125, + "learning_rate": 3.7612547390681202e-06, + "loss": 2.5407, + "step": 16630 + }, + { + "epoch": 1.7587774957698814, + "grad_norm": 7.875, + "learning_rate": 3.7611068716735415e-06, + "loss": 2.4452, + "step": 16631 + }, + { + "epoch": 1.7588832487309645, + "grad_norm": 8.5625, + "learning_rate": 3.760958998361104e-06, + "loss": 2.2421, + "step": 16632 + }, + { + "epoch": 1.7589890016920475, + "grad_norm": 7.9375, + "learning_rate": 3.760811119131502e-06, + "loss": 2.4891, + "step": 16633 + }, + { + "epoch": 1.7590947546531304, + "grad_norm": 8.625, + "learning_rate": 3.7606632339854286e-06, + "loss": 2.4448, + "step": 16634 + }, + { + "epoch": 1.7592005076142132, + "grad_norm": 7.78125, + "learning_rate": 3.7605153429235787e-06, + "loss": 2.4973, + "step": 16635 + }, + { + "epoch": 1.759306260575296, + "grad_norm": 8.625, + "learning_rate": 3.7603674459466453e-06, + "loss": 2.506, + "step": 16636 + }, + { + "epoch": 1.7594120135363789, + "grad_norm": 8.1875, + "learning_rate": 3.760219543055323e-06, + "loss": 2.5863, + "step": 16637 + }, + { + "epoch": 1.759517766497462, + "grad_norm": 8.25, + "learning_rate": 3.760071634250306e-06, + "loss": 2.5811, + "step": 16638 + }, + { + "epoch": 1.759623519458545, + "grad_norm": 7.8125, + "learning_rate": 3.759923719532288e-06, + "loss": 2.2848, + "step": 16639 + }, + { + "epoch": 1.7597292724196278, + "grad_norm": 9.1875, + "learning_rate": 3.7597757989019635e-06, + "loss": 2.6039, + "step": 16640 + }, + { + "epoch": 1.7598350253807107, + "grad_norm": 8.125, + "learning_rate": 3.759627872360026e-06, + "loss": 2.4732, + "step": 16641 + }, + { + "epoch": 1.7599407783417935, + "grad_norm": 8.5, + "learning_rate": 3.759479939907171e-06, + "loss": 2.6984, + "step": 16642 + }, + { + "epoch": 1.7600465313028764, + "grad_norm": 8.625, + "learning_rate": 3.759332001544091e-06, + "loss": 2.6868, + "step": 16643 + }, + { + "epoch": 1.7601522842639594, + "grad_norm": 8.0, + "learning_rate": 3.759184057271481e-06, + "loss": 2.472, + "step": 16644 + }, + { + "epoch": 1.7602580372250423, + "grad_norm": 8.8125, + "learning_rate": 3.7590361070900355e-06, + "loss": 2.809, + "step": 16645 + }, + { + "epoch": 1.7603637901861253, + "grad_norm": 7.8125, + "learning_rate": 3.758888151000448e-06, + "loss": 2.456, + "step": 16646 + }, + { + "epoch": 1.7604695431472082, + "grad_norm": 8.75, + "learning_rate": 3.7587401890034136e-06, + "loss": 2.51, + "step": 16647 + }, + { + "epoch": 1.760575296108291, + "grad_norm": 8.5625, + "learning_rate": 3.7585922210996263e-06, + "loss": 2.3551, + "step": 16648 + }, + { + "epoch": 1.7606810490693738, + "grad_norm": 8.6875, + "learning_rate": 3.75844424728978e-06, + "loss": 2.4539, + "step": 16649 + }, + { + "epoch": 1.760786802030457, + "grad_norm": 8.0625, + "learning_rate": 3.7582962675745706e-06, + "loss": 2.7279, + "step": 16650 + }, + { + "epoch": 1.7608925549915397, + "grad_norm": 7.84375, + "learning_rate": 3.7581482819546906e-06, + "loss": 2.5687, + "step": 16651 + }, + { + "epoch": 1.7609983079526228, + "grad_norm": 8.0, + "learning_rate": 3.7580002904308355e-06, + "loss": 2.5553, + "step": 16652 + }, + { + "epoch": 1.7611040609137056, + "grad_norm": 8.625, + "learning_rate": 3.7578522930037e-06, + "loss": 2.6239, + "step": 16653 + }, + { + "epoch": 1.7612098138747885, + "grad_norm": 8.125, + "learning_rate": 3.7577042896739767e-06, + "loss": 2.6748, + "step": 16654 + }, + { + "epoch": 1.7613155668358713, + "grad_norm": 8.75, + "learning_rate": 3.7575562804423626e-06, + "loss": 2.6237, + "step": 16655 + }, + { + "epoch": 1.7614213197969542, + "grad_norm": 8.375, + "learning_rate": 3.757408265309551e-06, + "loss": 2.4877, + "step": 16656 + }, + { + "epoch": 1.7615270727580372, + "grad_norm": 8.125, + "learning_rate": 3.7572602442762367e-06, + "loss": 2.6766, + "step": 16657 + }, + { + "epoch": 1.7616328257191203, + "grad_norm": 7.75, + "learning_rate": 3.7571122173431146e-06, + "loss": 2.4499, + "step": 16658 + }, + { + "epoch": 1.7617385786802031, + "grad_norm": 8.75, + "learning_rate": 3.756964184510879e-06, + "loss": 2.7593, + "step": 16659 + }, + { + "epoch": 1.761844331641286, + "grad_norm": 9.5, + "learning_rate": 3.7568161457802243e-06, + "loss": 2.5915, + "step": 16660 + }, + { + "epoch": 1.7619500846023688, + "grad_norm": 8.25, + "learning_rate": 3.7566681011518457e-06, + "loss": 2.403, + "step": 16661 + }, + { + "epoch": 1.7620558375634516, + "grad_norm": 8.0, + "learning_rate": 3.7565200506264377e-06, + "loss": 2.4295, + "step": 16662 + }, + { + "epoch": 1.7621615905245347, + "grad_norm": 8.5, + "learning_rate": 3.756371994204694e-06, + "loss": 2.5431, + "step": 16663 + }, + { + "epoch": 1.7622673434856178, + "grad_norm": 8.3125, + "learning_rate": 3.7562239318873104e-06, + "loss": 2.5944, + "step": 16664 + }, + { + "epoch": 1.7623730964467006, + "grad_norm": 8.375, + "learning_rate": 3.756075863674983e-06, + "loss": 2.5862, + "step": 16665 + }, + { + "epoch": 1.7624788494077834, + "grad_norm": 7.90625, + "learning_rate": 3.755927789568405e-06, + "loss": 2.3884, + "step": 16666 + }, + { + "epoch": 1.7625846023688663, + "grad_norm": 8.8125, + "learning_rate": 3.7557797095682707e-06, + "loss": 2.6493, + "step": 16667 + }, + { + "epoch": 1.762690355329949, + "grad_norm": 8.375, + "learning_rate": 3.7556316236752764e-06, + "loss": 2.4845, + "step": 16668 + }, + { + "epoch": 1.7627961082910322, + "grad_norm": 9.25, + "learning_rate": 3.755483531890117e-06, + "loss": 2.6013, + "step": 16669 + }, + { + "epoch": 1.762901861252115, + "grad_norm": 8.5625, + "learning_rate": 3.7553354342134856e-06, + "loss": 2.4346, + "step": 16670 + }, + { + "epoch": 1.763007614213198, + "grad_norm": 8.0, + "learning_rate": 3.7551873306460795e-06, + "loss": 2.4654, + "step": 16671 + }, + { + "epoch": 1.763113367174281, + "grad_norm": 7.96875, + "learning_rate": 3.7550392211885922e-06, + "loss": 2.543, + "step": 16672 + }, + { + "epoch": 1.7632191201353637, + "grad_norm": 7.8125, + "learning_rate": 3.7548911058417183e-06, + "loss": 2.4737, + "step": 16673 + }, + { + "epoch": 1.7633248730964466, + "grad_norm": 8.5, + "learning_rate": 3.7547429846061544e-06, + "loss": 2.5839, + "step": 16674 + }, + { + "epoch": 1.7634306260575296, + "grad_norm": 7.8125, + "learning_rate": 3.7545948574825952e-06, + "loss": 2.5241, + "step": 16675 + }, + { + "epoch": 1.7635363790186125, + "grad_norm": 8.625, + "learning_rate": 3.7544467244717348e-06, + "loss": 2.7731, + "step": 16676 + }, + { + "epoch": 1.7636421319796955, + "grad_norm": 8.1875, + "learning_rate": 3.7542985855742698e-06, + "loss": 2.7101, + "step": 16677 + }, + { + "epoch": 1.7637478849407784, + "grad_norm": 9.0, + "learning_rate": 3.7541504407908935e-06, + "loss": 2.787, + "step": 16678 + }, + { + "epoch": 1.7638536379018612, + "grad_norm": 15.25, + "learning_rate": 3.7540022901223027e-06, + "loss": 2.578, + "step": 16679 + }, + { + "epoch": 1.763959390862944, + "grad_norm": 8.3125, + "learning_rate": 3.7538541335691923e-06, + "loss": 2.5767, + "step": 16680 + }, + { + "epoch": 1.7640651438240271, + "grad_norm": 8.3125, + "learning_rate": 3.7537059711322565e-06, + "loss": 2.5222, + "step": 16681 + }, + { + "epoch": 1.76417089678511, + "grad_norm": 8.25, + "learning_rate": 3.7535578028121923e-06, + "loss": 2.6297, + "step": 16682 + }, + { + "epoch": 1.764276649746193, + "grad_norm": 8.875, + "learning_rate": 3.7534096286096927e-06, + "loss": 2.671, + "step": 16683 + }, + { + "epoch": 1.7643824027072759, + "grad_norm": 8.125, + "learning_rate": 3.7532614485254552e-06, + "loss": 2.5912, + "step": 16684 + }, + { + "epoch": 1.7644881556683587, + "grad_norm": 8.1875, + "learning_rate": 3.753113262560175e-06, + "loss": 2.5849, + "step": 16685 + }, + { + "epoch": 1.7645939086294415, + "grad_norm": 7.8125, + "learning_rate": 3.7529650707145456e-06, + "loss": 2.3057, + "step": 16686 + }, + { + "epoch": 1.7646996615905244, + "grad_norm": 8.6875, + "learning_rate": 3.752816872989264e-06, + "loss": 2.6748, + "step": 16687 + }, + { + "epoch": 1.7648054145516074, + "grad_norm": 9.4375, + "learning_rate": 3.752668669385025e-06, + "loss": 2.508, + "step": 16688 + }, + { + "epoch": 1.7649111675126905, + "grad_norm": 8.25, + "learning_rate": 3.7525204599025245e-06, + "loss": 2.5961, + "step": 16689 + }, + { + "epoch": 1.7650169204737733, + "grad_norm": 8.125, + "learning_rate": 3.7523722445424572e-06, + "loss": 2.556, + "step": 16690 + }, + { + "epoch": 1.7651226734348562, + "grad_norm": 8.375, + "learning_rate": 3.7522240233055195e-06, + "loss": 2.4763, + "step": 16691 + }, + { + "epoch": 1.765228426395939, + "grad_norm": 8.5625, + "learning_rate": 3.7520757961924064e-06, + "loss": 2.5459, + "step": 16692 + }, + { + "epoch": 1.7653341793570219, + "grad_norm": 8.125, + "learning_rate": 3.7519275632038145e-06, + "loss": 2.4448, + "step": 16693 + }, + { + "epoch": 1.765439932318105, + "grad_norm": 7.9375, + "learning_rate": 3.7517793243404376e-06, + "loss": 2.5475, + "step": 16694 + }, + { + "epoch": 1.765545685279188, + "grad_norm": 8.1875, + "learning_rate": 3.751631079602973e-06, + "loss": 2.6785, + "step": 16695 + }, + { + "epoch": 1.7656514382402708, + "grad_norm": 8.0, + "learning_rate": 3.7514828289921157e-06, + "loss": 2.6491, + "step": 16696 + }, + { + "epoch": 1.7657571912013537, + "grad_norm": 8.125, + "learning_rate": 3.7513345725085607e-06, + "loss": 2.5386, + "step": 16697 + }, + { + "epoch": 1.7658629441624365, + "grad_norm": 8.5, + "learning_rate": 3.7511863101530045e-06, + "loss": 2.5017, + "step": 16698 + }, + { + "epoch": 1.7659686971235193, + "grad_norm": 7.8125, + "learning_rate": 3.751038041926143e-06, + "loss": 2.5687, + "step": 16699 + }, + { + "epoch": 1.7660744500846024, + "grad_norm": 8.375, + "learning_rate": 3.7508897678286716e-06, + "loss": 2.5754, + "step": 16700 + }, + { + "epoch": 1.7661802030456852, + "grad_norm": 8.25, + "learning_rate": 3.7507414878612857e-06, + "loss": 2.6326, + "step": 16701 + }, + { + "epoch": 1.7662859560067683, + "grad_norm": 8.25, + "learning_rate": 3.7505932020246815e-06, + "loss": 2.4882, + "step": 16702 + }, + { + "epoch": 1.7663917089678511, + "grad_norm": 8.125, + "learning_rate": 3.7504449103195552e-06, + "loss": 2.3719, + "step": 16703 + }, + { + "epoch": 1.766497461928934, + "grad_norm": 8.5625, + "learning_rate": 3.7502966127466023e-06, + "loss": 2.4828, + "step": 16704 + }, + { + "epoch": 1.7666032148900168, + "grad_norm": 8.8125, + "learning_rate": 3.7501483093065185e-06, + "loss": 2.4839, + "step": 16705 + }, + { + "epoch": 1.7667089678510999, + "grad_norm": 8.5625, + "learning_rate": 3.7500000000000005e-06, + "loss": 2.4599, + "step": 16706 + }, + { + "epoch": 1.7668147208121827, + "grad_norm": 8.125, + "learning_rate": 3.749851684827744e-06, + "loss": 2.6365, + "step": 16707 + }, + { + "epoch": 1.7669204737732658, + "grad_norm": 8.375, + "learning_rate": 3.7497033637904434e-06, + "loss": 2.4105, + "step": 16708 + }, + { + "epoch": 1.7670262267343486, + "grad_norm": 7.8125, + "learning_rate": 3.7495550368887968e-06, + "loss": 2.5047, + "step": 16709 + }, + { + "epoch": 1.7671319796954315, + "grad_norm": 8.3125, + "learning_rate": 3.7494067041234993e-06, + "loss": 2.5468, + "step": 16710 + }, + { + "epoch": 1.7672377326565143, + "grad_norm": 8.5625, + "learning_rate": 3.7492583654952473e-06, + "loss": 2.5014, + "step": 16711 + }, + { + "epoch": 1.7673434856175971, + "grad_norm": 7.875, + "learning_rate": 3.749110021004737e-06, + "loss": 2.4923, + "step": 16712 + }, + { + "epoch": 1.7674492385786802, + "grad_norm": 9.1875, + "learning_rate": 3.748961670652664e-06, + "loss": 2.5925, + "step": 16713 + }, + { + "epoch": 1.7675549915397633, + "grad_norm": 7.8125, + "learning_rate": 3.7488133144397243e-06, + "loss": 2.4304, + "step": 16714 + }, + { + "epoch": 1.767660744500846, + "grad_norm": 9.125, + "learning_rate": 3.748664952366615e-06, + "loss": 2.4325, + "step": 16715 + }, + { + "epoch": 1.767766497461929, + "grad_norm": 8.0, + "learning_rate": 3.7485165844340318e-06, + "loss": 2.4825, + "step": 16716 + }, + { + "epoch": 1.7678722504230118, + "grad_norm": 8.0625, + "learning_rate": 3.74836821064267e-06, + "loss": 2.5474, + "step": 16717 + }, + { + "epoch": 1.7679780033840946, + "grad_norm": 7.96875, + "learning_rate": 3.748219830993227e-06, + "loss": 2.3758, + "step": 16718 + }, + { + "epoch": 1.7680837563451777, + "grad_norm": 8.1875, + "learning_rate": 3.7480714454863996e-06, + "loss": 2.4961, + "step": 16719 + }, + { + "epoch": 1.7681895093062607, + "grad_norm": 8.0625, + "learning_rate": 3.7479230541228834e-06, + "loss": 2.5907, + "step": 16720 + }, + { + "epoch": 1.7682952622673436, + "grad_norm": 8.5, + "learning_rate": 3.747774656903374e-06, + "loss": 2.5172, + "step": 16721 + }, + { + "epoch": 1.7684010152284264, + "grad_norm": 8.3125, + "learning_rate": 3.747626253828569e-06, + "loss": 2.5524, + "step": 16722 + }, + { + "epoch": 1.7685067681895092, + "grad_norm": 8.3125, + "learning_rate": 3.747477844899164e-06, + "loss": 2.5218, + "step": 16723 + }, + { + "epoch": 1.768612521150592, + "grad_norm": 8.625, + "learning_rate": 3.7473294301158554e-06, + "loss": 2.6532, + "step": 16724 + }, + { + "epoch": 1.7687182741116751, + "grad_norm": 8.125, + "learning_rate": 3.7471810094793405e-06, + "loss": 2.5772, + "step": 16725 + }, + { + "epoch": 1.768824027072758, + "grad_norm": 8.25, + "learning_rate": 3.747032582990314e-06, + "loss": 2.5608, + "step": 16726 + }, + { + "epoch": 1.768929780033841, + "grad_norm": 8.0, + "learning_rate": 3.7468841506494748e-06, + "loss": 2.5958, + "step": 16727 + }, + { + "epoch": 1.7690355329949239, + "grad_norm": 9.125, + "learning_rate": 3.7467357124575177e-06, + "loss": 2.2735, + "step": 16728 + }, + { + "epoch": 1.7691412859560067, + "grad_norm": 8.8125, + "learning_rate": 3.7465872684151404e-06, + "loss": 2.6029, + "step": 16729 + }, + { + "epoch": 1.7692470389170896, + "grad_norm": 8.875, + "learning_rate": 3.7464388185230384e-06, + "loss": 2.5193, + "step": 16730 + }, + { + "epoch": 1.7693527918781726, + "grad_norm": 8.75, + "learning_rate": 3.7462903627819092e-06, + "loss": 2.4438, + "step": 16731 + }, + { + "epoch": 1.7694585448392555, + "grad_norm": 8.8125, + "learning_rate": 3.7461419011924487e-06, + "loss": 2.769, + "step": 16732 + }, + { + "epoch": 1.7695642978003385, + "grad_norm": 8.8125, + "learning_rate": 3.745993433755354e-06, + "loss": 2.7802, + "step": 16733 + }, + { + "epoch": 1.7696700507614214, + "grad_norm": 8.125, + "learning_rate": 3.7458449604713216e-06, + "loss": 2.6825, + "step": 16734 + }, + { + "epoch": 1.7697758037225042, + "grad_norm": 8.0625, + "learning_rate": 3.745696481341048e-06, + "loss": 2.3945, + "step": 16735 + }, + { + "epoch": 1.769881556683587, + "grad_norm": 8.25, + "learning_rate": 3.745547996365231e-06, + "loss": 2.5336, + "step": 16736 + }, + { + "epoch": 1.76998730964467, + "grad_norm": 8.0625, + "learning_rate": 3.7453995055445663e-06, + "loss": 2.5055, + "step": 16737 + }, + { + "epoch": 1.770093062605753, + "grad_norm": 7.96875, + "learning_rate": 3.745251008879751e-06, + "loss": 2.6067, + "step": 16738 + }, + { + "epoch": 1.770198815566836, + "grad_norm": 8.375, + "learning_rate": 3.7451025063714823e-06, + "loss": 2.6127, + "step": 16739 + }, + { + "epoch": 1.7703045685279188, + "grad_norm": 8.25, + "learning_rate": 3.744953998020456e-06, + "loss": 2.584, + "step": 16740 + }, + { + "epoch": 1.7704103214890017, + "grad_norm": 8.1875, + "learning_rate": 3.744805483827371e-06, + "loss": 2.4455, + "step": 16741 + }, + { + "epoch": 1.7705160744500845, + "grad_norm": 8.375, + "learning_rate": 3.7446569637929225e-06, + "loss": 2.67, + "step": 16742 + }, + { + "epoch": 1.7706218274111674, + "grad_norm": 8.8125, + "learning_rate": 3.744508437917807e-06, + "loss": 2.4063, + "step": 16743 + }, + { + "epoch": 1.7707275803722504, + "grad_norm": 8.5, + "learning_rate": 3.7443599062027235e-06, + "loss": 2.6554, + "step": 16744 + }, + { + "epoch": 1.7708333333333335, + "grad_norm": 8.0625, + "learning_rate": 3.744211368648368e-06, + "loss": 2.523, + "step": 16745 + }, + { + "epoch": 1.7709390862944163, + "grad_norm": 8.25, + "learning_rate": 3.7440628252554363e-06, + "loss": 2.2375, + "step": 16746 + }, + { + "epoch": 1.7710448392554992, + "grad_norm": 8.375, + "learning_rate": 3.743914276024627e-06, + "loss": 2.5569, + "step": 16747 + }, + { + "epoch": 1.771150592216582, + "grad_norm": 8.375, + "learning_rate": 3.7437657209566375e-06, + "loss": 2.5896, + "step": 16748 + }, + { + "epoch": 1.7712563451776648, + "grad_norm": 8.4375, + "learning_rate": 3.7436171600521633e-06, + "loss": 2.4132, + "step": 16749 + }, + { + "epoch": 1.771362098138748, + "grad_norm": 8.1875, + "learning_rate": 3.743468593311903e-06, + "loss": 2.595, + "step": 16750 + }, + { + "epoch": 1.771467851099831, + "grad_norm": 8.5, + "learning_rate": 3.7433200207365528e-06, + "loss": 2.5794, + "step": 16751 + }, + { + "epoch": 1.7715736040609138, + "grad_norm": 8.4375, + "learning_rate": 3.7431714423268103e-06, + "loss": 2.4909, + "step": 16752 + }, + { + "epoch": 1.7716793570219966, + "grad_norm": 8.6875, + "learning_rate": 3.7430228580833723e-06, + "loss": 2.4713, + "step": 16753 + }, + { + "epoch": 1.7717851099830795, + "grad_norm": 8.875, + "learning_rate": 3.742874268006937e-06, + "loss": 2.6865, + "step": 16754 + }, + { + "epoch": 1.7718908629441623, + "grad_norm": 8.5, + "learning_rate": 3.742725672098201e-06, + "loss": 2.7849, + "step": 16755 + }, + { + "epoch": 1.7719966159052454, + "grad_norm": 7.9375, + "learning_rate": 3.7425770703578613e-06, + "loss": 2.5718, + "step": 16756 + }, + { + "epoch": 1.7721023688663282, + "grad_norm": 8.0, + "learning_rate": 3.742428462786616e-06, + "loss": 2.5582, + "step": 16757 + }, + { + "epoch": 1.7722081218274113, + "grad_norm": 8.125, + "learning_rate": 3.7422798493851618e-06, + "loss": 2.6457, + "step": 16758 + }, + { + "epoch": 1.7723138747884941, + "grad_norm": 8.625, + "learning_rate": 3.742131230154196e-06, + "loss": 2.4151, + "step": 16759 + }, + { + "epoch": 1.772419627749577, + "grad_norm": 9.1875, + "learning_rate": 3.7419826050944173e-06, + "loss": 2.6023, + "step": 16760 + }, + { + "epoch": 1.7725253807106598, + "grad_norm": 8.5625, + "learning_rate": 3.741833974206521e-06, + "loss": 2.4335, + "step": 16761 + }, + { + "epoch": 1.7726311336717429, + "grad_norm": 8.625, + "learning_rate": 3.7416853374912064e-06, + "loss": 2.4731, + "step": 16762 + }, + { + "epoch": 1.7727368866328257, + "grad_norm": 7.96875, + "learning_rate": 3.7415366949491703e-06, + "loss": 2.6112, + "step": 16763 + }, + { + "epoch": 1.7728426395939088, + "grad_norm": 8.1875, + "learning_rate": 3.7413880465811107e-06, + "loss": 2.5179, + "step": 16764 + }, + { + "epoch": 1.7729483925549916, + "grad_norm": 7.96875, + "learning_rate": 3.7412393923877243e-06, + "loss": 2.5541, + "step": 16765 + }, + { + "epoch": 1.7730541455160744, + "grad_norm": 8.6875, + "learning_rate": 3.7410907323697095e-06, + "loss": 2.4735, + "step": 16766 + }, + { + "epoch": 1.7731598984771573, + "grad_norm": 8.1875, + "learning_rate": 3.740942066527763e-06, + "loss": 2.4769, + "step": 16767 + }, + { + "epoch": 1.77326565143824, + "grad_norm": 8.5, + "learning_rate": 3.740793394862583e-06, + "loss": 2.1856, + "step": 16768 + }, + { + "epoch": 1.7733714043993232, + "grad_norm": 8.5625, + "learning_rate": 3.7406447173748673e-06, + "loss": 2.62, + "step": 16769 + }, + { + "epoch": 1.7734771573604062, + "grad_norm": 7.90625, + "learning_rate": 3.7404960340653124e-06, + "loss": 2.6396, + "step": 16770 + }, + { + "epoch": 1.773582910321489, + "grad_norm": 8.5625, + "learning_rate": 3.7403473449346176e-06, + "loss": 2.6827, + "step": 16771 + }, + { + "epoch": 1.773688663282572, + "grad_norm": 8.875, + "learning_rate": 3.7401986499834796e-06, + "loss": 2.2609, + "step": 16772 + }, + { + "epoch": 1.7737944162436547, + "grad_norm": 8.4375, + "learning_rate": 3.7400499492125977e-06, + "loss": 2.645, + "step": 16773 + }, + { + "epoch": 1.7739001692047376, + "grad_norm": 8.125, + "learning_rate": 3.739901242622668e-06, + "loss": 2.4275, + "step": 16774 + }, + { + "epoch": 1.7740059221658206, + "grad_norm": 8.25, + "learning_rate": 3.7397525302143884e-06, + "loss": 2.5406, + "step": 16775 + }, + { + "epoch": 1.7741116751269037, + "grad_norm": 8.625, + "learning_rate": 3.739603811988458e-06, + "loss": 2.6715, + "step": 16776 + }, + { + "epoch": 1.7742174280879865, + "grad_norm": 9.0625, + "learning_rate": 3.739455087945574e-06, + "loss": 2.5829, + "step": 16777 + }, + { + "epoch": 1.7743231810490694, + "grad_norm": 8.375, + "learning_rate": 3.739306358086433e-06, + "loss": 2.67, + "step": 16778 + }, + { + "epoch": 1.7744289340101522, + "grad_norm": 8.125, + "learning_rate": 3.739157622411734e-06, + "loss": 2.5211, + "step": 16779 + }, + { + "epoch": 1.774534686971235, + "grad_norm": 8.0625, + "learning_rate": 3.7390088809221763e-06, + "loss": 2.5255, + "step": 16780 + }, + { + "epoch": 1.7746404399323181, + "grad_norm": 8.3125, + "learning_rate": 3.738860133618456e-06, + "loss": 2.5765, + "step": 16781 + }, + { + "epoch": 1.774746192893401, + "grad_norm": 7.84375, + "learning_rate": 3.738711380501272e-06, + "loss": 2.4704, + "step": 16782 + }, + { + "epoch": 1.774851945854484, + "grad_norm": 8.25, + "learning_rate": 3.7385626215713224e-06, + "loss": 2.5185, + "step": 16783 + }, + { + "epoch": 1.7749576988155669, + "grad_norm": 8.9375, + "learning_rate": 3.7384138568293045e-06, + "loss": 2.6807, + "step": 16784 + }, + { + "epoch": 1.7750634517766497, + "grad_norm": 8.0625, + "learning_rate": 3.7382650862759172e-06, + "loss": 2.6844, + "step": 16785 + }, + { + "epoch": 1.7751692047377325, + "grad_norm": 8.0625, + "learning_rate": 3.7381163099118585e-06, + "loss": 2.603, + "step": 16786 + }, + { + "epoch": 1.7752749576988156, + "grad_norm": 8.375, + "learning_rate": 3.737967527737825e-06, + "loss": 2.7154, + "step": 16787 + }, + { + "epoch": 1.7753807106598984, + "grad_norm": 8.1875, + "learning_rate": 3.7378187397545174e-06, + "loss": 2.6694, + "step": 16788 + }, + { + "epoch": 1.7754864636209815, + "grad_norm": 8.0625, + "learning_rate": 3.7376699459626326e-06, + "loss": 2.5902, + "step": 16789 + }, + { + "epoch": 1.7755922165820643, + "grad_norm": 8.4375, + "learning_rate": 3.7375211463628685e-06, + "loss": 2.3978, + "step": 16790 + }, + { + "epoch": 1.7756979695431472, + "grad_norm": 8.1875, + "learning_rate": 3.7373723409559237e-06, + "loss": 2.4559, + "step": 16791 + }, + { + "epoch": 1.77580372250423, + "grad_norm": 8.9375, + "learning_rate": 3.7372235297424973e-06, + "loss": 2.6646, + "step": 16792 + }, + { + "epoch": 1.775909475465313, + "grad_norm": 8.3125, + "learning_rate": 3.737074712723287e-06, + "loss": 2.2567, + "step": 16793 + }, + { + "epoch": 1.776015228426396, + "grad_norm": 9.25, + "learning_rate": 3.7369258898989903e-06, + "loss": 2.5103, + "step": 16794 + }, + { + "epoch": 1.776120981387479, + "grad_norm": 8.375, + "learning_rate": 3.7367770612703065e-06, + "loss": 2.7949, + "step": 16795 + }, + { + "epoch": 1.7762267343485618, + "grad_norm": 8.6875, + "learning_rate": 3.7366282268379333e-06, + "loss": 2.584, + "step": 16796 + }, + { + "epoch": 1.7763324873096447, + "grad_norm": 8.4375, + "learning_rate": 3.73647938660257e-06, + "loss": 2.6697, + "step": 16797 + }, + { + "epoch": 1.7764382402707275, + "grad_norm": 8.125, + "learning_rate": 3.7363305405649146e-06, + "loss": 2.5789, + "step": 16798 + }, + { + "epoch": 1.7765439932318103, + "grad_norm": 8.375, + "learning_rate": 3.7361816887256662e-06, + "loss": 2.5668, + "step": 16799 + }, + { + "epoch": 1.7766497461928934, + "grad_norm": 8.375, + "learning_rate": 3.736032831085522e-06, + "loss": 2.4371, + "step": 16800 + }, + { + "epoch": 1.7767554991539765, + "grad_norm": 8.875, + "learning_rate": 3.735883967645182e-06, + "loss": 2.7305, + "step": 16801 + }, + { + "epoch": 1.7768612521150593, + "grad_norm": 9.0, + "learning_rate": 3.735735098405343e-06, + "loss": 2.5586, + "step": 16802 + }, + { + "epoch": 1.7769670050761421, + "grad_norm": 7.90625, + "learning_rate": 3.735586223366705e-06, + "loss": 2.5203, + "step": 16803 + }, + { + "epoch": 1.777072758037225, + "grad_norm": 8.375, + "learning_rate": 3.735437342529967e-06, + "loss": 2.5281, + "step": 16804 + }, + { + "epoch": 1.7771785109983078, + "grad_norm": 8.8125, + "learning_rate": 3.7352884558958256e-06, + "loss": 2.5025, + "step": 16805 + }, + { + "epoch": 1.7772842639593909, + "grad_norm": 8.625, + "learning_rate": 3.7351395634649814e-06, + "loss": 2.6427, + "step": 16806 + }, + { + "epoch": 1.777390016920474, + "grad_norm": 8.5, + "learning_rate": 3.7349906652381316e-06, + "loss": 2.6771, + "step": 16807 + }, + { + "epoch": 1.7774957698815568, + "grad_norm": 8.8125, + "learning_rate": 3.7348417612159764e-06, + "loss": 2.4427, + "step": 16808 + }, + { + "epoch": 1.7776015228426396, + "grad_norm": 8.5, + "learning_rate": 3.7346928513992143e-06, + "loss": 2.7403, + "step": 16809 + }, + { + "epoch": 1.7777072758037225, + "grad_norm": 8.125, + "learning_rate": 3.7345439357885426e-06, + "loss": 2.3561, + "step": 16810 + }, + { + "epoch": 1.7778130287648053, + "grad_norm": 8.4375, + "learning_rate": 3.734395014384662e-06, + "loss": 2.594, + "step": 16811 + }, + { + "epoch": 1.7779187817258884, + "grad_norm": 8.5, + "learning_rate": 3.73424608718827e-06, + "loss": 2.6653, + "step": 16812 + }, + { + "epoch": 1.7780245346869712, + "grad_norm": 8.625, + "learning_rate": 3.734097154200066e-06, + "loss": 2.628, + "step": 16813 + }, + { + "epoch": 1.7781302876480543, + "grad_norm": 8.625, + "learning_rate": 3.733948215420748e-06, + "loss": 2.6098, + "step": 16814 + }, + { + "epoch": 1.778236040609137, + "grad_norm": 8.3125, + "learning_rate": 3.733799270851017e-06, + "loss": 2.464, + "step": 16815 + }, + { + "epoch": 1.77834179357022, + "grad_norm": 8.625, + "learning_rate": 3.7336503204915694e-06, + "loss": 2.6304, + "step": 16816 + }, + { + "epoch": 1.7784475465313028, + "grad_norm": 9.5625, + "learning_rate": 3.733501364343106e-06, + "loss": 2.4522, + "step": 16817 + }, + { + "epoch": 1.7785532994923858, + "grad_norm": 8.5, + "learning_rate": 3.7333524024063256e-06, + "loss": 2.6248, + "step": 16818 + }, + { + "epoch": 1.7786590524534687, + "grad_norm": 9.5, + "learning_rate": 3.7332034346819264e-06, + "loss": 2.7271, + "step": 16819 + }, + { + "epoch": 1.7787648054145517, + "grad_norm": 8.25, + "learning_rate": 3.733054461170608e-06, + "loss": 2.511, + "step": 16820 + }, + { + "epoch": 1.7788705583756346, + "grad_norm": 8.625, + "learning_rate": 3.732905481873069e-06, + "loss": 2.7378, + "step": 16821 + }, + { + "epoch": 1.7789763113367174, + "grad_norm": 7.90625, + "learning_rate": 3.732756496790009e-06, + "loss": 2.3625, + "step": 16822 + }, + { + "epoch": 1.7790820642978002, + "grad_norm": 8.5625, + "learning_rate": 3.7326075059221263e-06, + "loss": 2.6335, + "step": 16823 + }, + { + "epoch": 1.779187817258883, + "grad_norm": 8.0625, + "learning_rate": 3.7324585092701213e-06, + "loss": 2.6951, + "step": 16824 + }, + { + "epoch": 1.7792935702199661, + "grad_norm": 8.375, + "learning_rate": 3.732309506834693e-06, + "loss": 2.2593, + "step": 16825 + }, + { + "epoch": 1.7793993231810492, + "grad_norm": 8.0625, + "learning_rate": 3.7321604986165393e-06, + "loss": 2.724, + "step": 16826 + }, + { + "epoch": 1.779505076142132, + "grad_norm": 8.125, + "learning_rate": 3.7320114846163615e-06, + "loss": 2.5704, + "step": 16827 + }, + { + "epoch": 1.7796108291032149, + "grad_norm": 8.3125, + "learning_rate": 3.731862464834857e-06, + "loss": 2.5342, + "step": 16828 + }, + { + "epoch": 1.7797165820642977, + "grad_norm": 8.375, + "learning_rate": 3.731713439272725e-06, + "loss": 2.6432, + "step": 16829 + }, + { + "epoch": 1.7798223350253806, + "grad_norm": 9.3125, + "learning_rate": 3.7315644079306666e-06, + "loss": 2.4683, + "step": 16830 + }, + { + "epoch": 1.7799280879864636, + "grad_norm": 8.25, + "learning_rate": 3.7314153708093802e-06, + "loss": 2.6081, + "step": 16831 + }, + { + "epoch": 1.7800338409475467, + "grad_norm": 7.78125, + "learning_rate": 3.731266327909564e-06, + "loss": 2.4997, + "step": 16832 + }, + { + "epoch": 1.7801395939086295, + "grad_norm": 8.75, + "learning_rate": 3.7311172792319196e-06, + "loss": 2.6994, + "step": 16833 + }, + { + "epoch": 1.7802453468697124, + "grad_norm": 9.125, + "learning_rate": 3.730968224777145e-06, + "loss": 2.7191, + "step": 16834 + }, + { + "epoch": 1.7803510998307952, + "grad_norm": 7.96875, + "learning_rate": 3.73081916454594e-06, + "loss": 2.5088, + "step": 16835 + }, + { + "epoch": 1.780456852791878, + "grad_norm": 7.84375, + "learning_rate": 3.7306700985390038e-06, + "loss": 2.2555, + "step": 16836 + }, + { + "epoch": 1.780562605752961, + "grad_norm": 8.25, + "learning_rate": 3.7305210267570365e-06, + "loss": 2.5725, + "step": 16837 + }, + { + "epoch": 1.780668358714044, + "grad_norm": 8.6875, + "learning_rate": 3.7303719492007367e-06, + "loss": 2.5165, + "step": 16838 + }, + { + "epoch": 1.780774111675127, + "grad_norm": 8.3125, + "learning_rate": 3.730222865870805e-06, + "loss": 2.5235, + "step": 16839 + }, + { + "epoch": 1.7808798646362098, + "grad_norm": 8.125, + "learning_rate": 3.7300737767679406e-06, + "loss": 2.6999, + "step": 16840 + }, + { + "epoch": 1.7809856175972927, + "grad_norm": 8.375, + "learning_rate": 3.7299246818928423e-06, + "loss": 2.6407, + "step": 16841 + }, + { + "epoch": 1.7810913705583755, + "grad_norm": 8.5625, + "learning_rate": 3.729775581246211e-06, + "loss": 2.5751, + "step": 16842 + }, + { + "epoch": 1.7811971235194586, + "grad_norm": 7.875, + "learning_rate": 3.7296264748287463e-06, + "loss": 2.4891, + "step": 16843 + }, + { + "epoch": 1.7813028764805414, + "grad_norm": 8.8125, + "learning_rate": 3.7294773626411475e-06, + "loss": 2.5269, + "step": 16844 + }, + { + "epoch": 1.7814086294416245, + "grad_norm": 8.875, + "learning_rate": 3.7293282446841138e-06, + "loss": 2.741, + "step": 16845 + }, + { + "epoch": 1.7815143824027073, + "grad_norm": 8.25, + "learning_rate": 3.7291791209583457e-06, + "loss": 2.6071, + "step": 16846 + }, + { + "epoch": 1.7816201353637902, + "grad_norm": 8.25, + "learning_rate": 3.7290299914645428e-06, + "loss": 2.5617, + "step": 16847 + }, + { + "epoch": 1.781725888324873, + "grad_norm": 8.6875, + "learning_rate": 3.728880856203404e-06, + "loss": 2.7042, + "step": 16848 + }, + { + "epoch": 1.781831641285956, + "grad_norm": 8.4375, + "learning_rate": 3.72873171517563e-06, + "loss": 2.518, + "step": 16849 + }, + { + "epoch": 1.781937394247039, + "grad_norm": 8.0625, + "learning_rate": 3.7285825683819217e-06, + "loss": 2.6609, + "step": 16850 + }, + { + "epoch": 1.782043147208122, + "grad_norm": 8.875, + "learning_rate": 3.7284334158229767e-06, + "loss": 2.7283, + "step": 16851 + }, + { + "epoch": 1.7821489001692048, + "grad_norm": 8.625, + "learning_rate": 3.728284257499497e-06, + "loss": 2.5445, + "step": 16852 + }, + { + "epoch": 1.7822546531302876, + "grad_norm": 8.625, + "learning_rate": 3.728135093412181e-06, + "loss": 2.6462, + "step": 16853 + }, + { + "epoch": 1.7823604060913705, + "grad_norm": 8.0, + "learning_rate": 3.72798592356173e-06, + "loss": 2.5816, + "step": 16854 + }, + { + "epoch": 1.7824661590524533, + "grad_norm": 8.0625, + "learning_rate": 3.727836747948843e-06, + "loss": 2.6382, + "step": 16855 + }, + { + "epoch": 1.7825719120135364, + "grad_norm": 8.0, + "learning_rate": 3.72768756657422e-06, + "loss": 2.6164, + "step": 16856 + }, + { + "epoch": 1.7826776649746194, + "grad_norm": 7.90625, + "learning_rate": 3.727538379438561e-06, + "loss": 2.6199, + "step": 16857 + }, + { + "epoch": 1.7827834179357023, + "grad_norm": 8.4375, + "learning_rate": 3.727389186542567e-06, + "loss": 2.6282, + "step": 16858 + }, + { + "epoch": 1.7828891708967851, + "grad_norm": 7.90625, + "learning_rate": 3.727239987886938e-06, + "loss": 2.5127, + "step": 16859 + }, + { + "epoch": 1.782994923857868, + "grad_norm": 8.125, + "learning_rate": 3.727090783472373e-06, + "loss": 2.5428, + "step": 16860 + }, + { + "epoch": 1.7831006768189508, + "grad_norm": 8.1875, + "learning_rate": 3.726941573299573e-06, + "loss": 2.7536, + "step": 16861 + }, + { + "epoch": 1.7832064297800339, + "grad_norm": 9.0, + "learning_rate": 3.726792357369238e-06, + "loss": 2.8328, + "step": 16862 + }, + { + "epoch": 1.783312182741117, + "grad_norm": 8.3125, + "learning_rate": 3.7266431356820685e-06, + "loss": 2.6152, + "step": 16863 + }, + { + "epoch": 1.7834179357021998, + "grad_norm": 8.3125, + "learning_rate": 3.7264939082387635e-06, + "loss": 2.4256, + "step": 16864 + }, + { + "epoch": 1.7835236886632826, + "grad_norm": 8.125, + "learning_rate": 3.726344675040025e-06, + "loss": 2.4098, + "step": 16865 + }, + { + "epoch": 1.7836294416243654, + "grad_norm": 8.75, + "learning_rate": 3.7261954360865528e-06, + "loss": 2.6096, + "step": 16866 + }, + { + "epoch": 1.7837351945854483, + "grad_norm": 8.6875, + "learning_rate": 3.726046191379046e-06, + "loss": 2.5003, + "step": 16867 + }, + { + "epoch": 1.7838409475465313, + "grad_norm": 8.1875, + "learning_rate": 3.7258969409182063e-06, + "loss": 2.3456, + "step": 16868 + }, + { + "epoch": 1.7839467005076142, + "grad_norm": 7.875, + "learning_rate": 3.7257476847047335e-06, + "loss": 2.365, + "step": 16869 + }, + { + "epoch": 1.7840524534686972, + "grad_norm": 8.3125, + "learning_rate": 3.7255984227393284e-06, + "loss": 2.5086, + "step": 16870 + }, + { + "epoch": 1.78415820642978, + "grad_norm": 8.5625, + "learning_rate": 3.7254491550226912e-06, + "loss": 2.7247, + "step": 16871 + }, + { + "epoch": 1.784263959390863, + "grad_norm": 8.125, + "learning_rate": 3.725299881555522e-06, + "loss": 2.6946, + "step": 16872 + }, + { + "epoch": 1.7843697123519457, + "grad_norm": 8.75, + "learning_rate": 3.7251506023385215e-06, + "loss": 2.8773, + "step": 16873 + }, + { + "epoch": 1.7844754653130288, + "grad_norm": 8.375, + "learning_rate": 3.7250013173723908e-06, + "loss": 2.5086, + "step": 16874 + }, + { + "epoch": 1.7845812182741116, + "grad_norm": 8.25, + "learning_rate": 3.72485202665783e-06, + "loss": 2.4932, + "step": 16875 + }, + { + "epoch": 1.7846869712351947, + "grad_norm": 8.5, + "learning_rate": 3.7247027301955385e-06, + "loss": 2.5293, + "step": 16876 + }, + { + "epoch": 1.7847927241962775, + "grad_norm": 8.1875, + "learning_rate": 3.7245534279862183e-06, + "loss": 2.4387, + "step": 16877 + }, + { + "epoch": 1.7848984771573604, + "grad_norm": 8.125, + "learning_rate": 3.7244041200305703e-06, + "loss": 2.4531, + "step": 16878 + }, + { + "epoch": 1.7850042301184432, + "grad_norm": 8.375, + "learning_rate": 3.724254806329295e-06, + "loss": 2.7026, + "step": 16879 + }, + { + "epoch": 1.785109983079526, + "grad_norm": 8.875, + "learning_rate": 3.724105486883091e-06, + "loss": 2.6725, + "step": 16880 + }, + { + "epoch": 1.7852157360406091, + "grad_norm": 8.1875, + "learning_rate": 3.7239561616926624e-06, + "loss": 2.5616, + "step": 16881 + }, + { + "epoch": 1.7853214890016922, + "grad_norm": 8.25, + "learning_rate": 3.723806830758707e-06, + "loss": 2.7028, + "step": 16882 + }, + { + "epoch": 1.785427241962775, + "grad_norm": 8.4375, + "learning_rate": 3.7236574940819265e-06, + "loss": 2.566, + "step": 16883 + }, + { + "epoch": 1.7855329949238579, + "grad_norm": 8.625, + "learning_rate": 3.7235081516630223e-06, + "loss": 2.5825, + "step": 16884 + }, + { + "epoch": 1.7856387478849407, + "grad_norm": 7.84375, + "learning_rate": 3.7233588035026945e-06, + "loss": 2.5368, + "step": 16885 + }, + { + "epoch": 1.7857445008460235, + "grad_norm": 8.125, + "learning_rate": 3.723209449601644e-06, + "loss": 2.457, + "step": 16886 + }, + { + "epoch": 1.7858502538071066, + "grad_norm": 8.3125, + "learning_rate": 3.723060089960572e-06, + "loss": 2.5691, + "step": 16887 + }, + { + "epoch": 1.7859560067681897, + "grad_norm": 8.0625, + "learning_rate": 3.7229107245801798e-06, + "loss": 2.4904, + "step": 16888 + }, + { + "epoch": 1.7860617597292725, + "grad_norm": 8.0, + "learning_rate": 3.722761353461167e-06, + "loss": 2.2656, + "step": 16889 + }, + { + "epoch": 1.7861675126903553, + "grad_norm": 8.25, + "learning_rate": 3.7226119766042356e-06, + "loss": 2.6216, + "step": 16890 + }, + { + "epoch": 1.7862732656514382, + "grad_norm": 8.0, + "learning_rate": 3.722462594010087e-06, + "loss": 2.361, + "step": 16891 + }, + { + "epoch": 1.786379018612521, + "grad_norm": 9.125, + "learning_rate": 3.72231320567942e-06, + "loss": 2.8185, + "step": 16892 + }, + { + "epoch": 1.786484771573604, + "grad_norm": 8.5625, + "learning_rate": 3.722163811612938e-06, + "loss": 2.6651, + "step": 16893 + }, + { + "epoch": 1.786590524534687, + "grad_norm": 8.1875, + "learning_rate": 3.72201441181134e-06, + "loss": 2.641, + "step": 16894 + }, + { + "epoch": 1.78669627749577, + "grad_norm": 8.0625, + "learning_rate": 3.721865006275329e-06, + "loss": 2.5655, + "step": 16895 + }, + { + "epoch": 1.7868020304568528, + "grad_norm": 8.375, + "learning_rate": 3.721715595005605e-06, + "loss": 2.6131, + "step": 16896 + }, + { + "epoch": 1.7869077834179357, + "grad_norm": 8.0, + "learning_rate": 3.72156617800287e-06, + "loss": 2.4619, + "step": 16897 + }, + { + "epoch": 1.7870135363790185, + "grad_norm": 8.0, + "learning_rate": 3.7214167552678237e-06, + "loss": 2.5792, + "step": 16898 + }, + { + "epoch": 1.7871192893401016, + "grad_norm": 7.90625, + "learning_rate": 3.7212673268011683e-06, + "loss": 2.5333, + "step": 16899 + }, + { + "epoch": 1.7872250423011844, + "grad_norm": 7.875, + "learning_rate": 3.721117892603605e-06, + "loss": 2.4953, + "step": 16900 + }, + { + "epoch": 1.7873307952622675, + "grad_norm": 8.0625, + "learning_rate": 3.7209684526758357e-06, + "loss": 2.4501, + "step": 16901 + }, + { + "epoch": 1.7874365482233503, + "grad_norm": 8.3125, + "learning_rate": 3.720819007018559e-06, + "loss": 2.6484, + "step": 16902 + }, + { + "epoch": 1.7875423011844331, + "grad_norm": 8.0625, + "learning_rate": 3.720669555632479e-06, + "loss": 2.5285, + "step": 16903 + }, + { + "epoch": 1.787648054145516, + "grad_norm": 9.75, + "learning_rate": 3.720520098518296e-06, + "loss": 2.3527, + "step": 16904 + }, + { + "epoch": 1.787753807106599, + "grad_norm": 9.0, + "learning_rate": 3.720370635676711e-06, + "loss": 2.4962, + "step": 16905 + }, + { + "epoch": 1.7878595600676819, + "grad_norm": 8.5625, + "learning_rate": 3.720221167108426e-06, + "loss": 2.5819, + "step": 16906 + }, + { + "epoch": 1.787965313028765, + "grad_norm": 8.4375, + "learning_rate": 3.7200716928141422e-06, + "loss": 2.4407, + "step": 16907 + }, + { + "epoch": 1.7880710659898478, + "grad_norm": 8.9375, + "learning_rate": 3.7199222127945612e-06, + "loss": 2.6342, + "step": 16908 + }, + { + "epoch": 1.7881768189509306, + "grad_norm": 8.6875, + "learning_rate": 3.7197727270503834e-06, + "loss": 2.6504, + "step": 16909 + }, + { + "epoch": 1.7882825719120135, + "grad_norm": 8.5, + "learning_rate": 3.719623235582312e-06, + "loss": 2.5082, + "step": 16910 + }, + { + "epoch": 1.7883883248730963, + "grad_norm": 8.1875, + "learning_rate": 3.7194737383910467e-06, + "loss": 2.4971, + "step": 16911 + }, + { + "epoch": 1.7884940778341794, + "grad_norm": 10.0, + "learning_rate": 3.7193242354772896e-06, + "loss": 2.4185, + "step": 16912 + }, + { + "epoch": 1.7885998307952624, + "grad_norm": 7.96875, + "learning_rate": 3.719174726841744e-06, + "loss": 2.527, + "step": 16913 + }, + { + "epoch": 1.7887055837563453, + "grad_norm": 8.1875, + "learning_rate": 3.7190252124851088e-06, + "loss": 2.706, + "step": 16914 + }, + { + "epoch": 1.788811336717428, + "grad_norm": 8.0, + "learning_rate": 3.718875692408087e-06, + "loss": 2.5154, + "step": 16915 + }, + { + "epoch": 1.788917089678511, + "grad_norm": 8.375, + "learning_rate": 3.7187261666113807e-06, + "loss": 2.437, + "step": 16916 + }, + { + "epoch": 1.7890228426395938, + "grad_norm": 8.9375, + "learning_rate": 3.718576635095691e-06, + "loss": 2.9498, + "step": 16917 + }, + { + "epoch": 1.7891285956006768, + "grad_norm": 8.8125, + "learning_rate": 3.7184270978617186e-06, + "loss": 2.3548, + "step": 16918 + }, + { + "epoch": 1.78923434856176, + "grad_norm": 7.9375, + "learning_rate": 3.7182775549101667e-06, + "loss": 2.4153, + "step": 16919 + }, + { + "epoch": 1.7893401015228427, + "grad_norm": 8.75, + "learning_rate": 3.718128006241737e-06, + "loss": 2.5779, + "step": 16920 + }, + { + "epoch": 1.7894458544839256, + "grad_norm": 7.8125, + "learning_rate": 3.7179784518571293e-06, + "loss": 2.4526, + "step": 16921 + }, + { + "epoch": 1.7895516074450084, + "grad_norm": 8.6875, + "learning_rate": 3.717828891757048e-06, + "loss": 2.6733, + "step": 16922 + }, + { + "epoch": 1.7896573604060912, + "grad_norm": 8.8125, + "learning_rate": 3.717679325942194e-06, + "loss": 2.6721, + "step": 16923 + }, + { + "epoch": 1.7897631133671743, + "grad_norm": 7.9375, + "learning_rate": 3.7175297544132684e-06, + "loss": 2.3464, + "step": 16924 + }, + { + "epoch": 1.7898688663282571, + "grad_norm": 8.5, + "learning_rate": 3.7173801771709738e-06, + "loss": 2.7028, + "step": 16925 + }, + { + "epoch": 1.7899746192893402, + "grad_norm": 8.3125, + "learning_rate": 3.717230594216012e-06, + "loss": 2.5473, + "step": 16926 + }, + { + "epoch": 1.790080372250423, + "grad_norm": 9.0, + "learning_rate": 3.717081005549085e-06, + "loss": 2.6235, + "step": 16927 + }, + { + "epoch": 1.7901861252115059, + "grad_norm": 7.875, + "learning_rate": 3.7169314111708936e-06, + "loss": 2.1231, + "step": 16928 + }, + { + "epoch": 1.7902918781725887, + "grad_norm": 8.0, + "learning_rate": 3.7167818110821414e-06, + "loss": 2.4117, + "step": 16929 + }, + { + "epoch": 1.7903976311336718, + "grad_norm": 8.4375, + "learning_rate": 3.7166322052835303e-06, + "loss": 2.7027, + "step": 16930 + }, + { + "epoch": 1.7905033840947546, + "grad_norm": 8.4375, + "learning_rate": 3.716482593775761e-06, + "loss": 2.6404, + "step": 16931 + }, + { + "epoch": 1.7906091370558377, + "grad_norm": 8.75, + "learning_rate": 3.716332976559537e-06, + "loss": 2.2939, + "step": 16932 + }, + { + "epoch": 1.7907148900169205, + "grad_norm": 8.0625, + "learning_rate": 3.7161833536355603e-06, + "loss": 2.458, + "step": 16933 + }, + { + "epoch": 1.7908206429780034, + "grad_norm": 8.625, + "learning_rate": 3.7160337250045315e-06, + "loss": 2.5793, + "step": 16934 + }, + { + "epoch": 1.7909263959390862, + "grad_norm": 8.3125, + "learning_rate": 3.7158840906671544e-06, + "loss": 2.7614, + "step": 16935 + }, + { + "epoch": 1.791032148900169, + "grad_norm": 8.375, + "learning_rate": 3.7157344506241304e-06, + "loss": 2.5633, + "step": 16936 + }, + { + "epoch": 1.791137901861252, + "grad_norm": 8.0625, + "learning_rate": 3.715584804876161e-06, + "loss": 2.6026, + "step": 16937 + }, + { + "epoch": 1.7912436548223352, + "grad_norm": 8.1875, + "learning_rate": 3.71543515342395e-06, + "loss": 2.5042, + "step": 16938 + }, + { + "epoch": 1.791349407783418, + "grad_norm": 8.25, + "learning_rate": 3.7152854962681996e-06, + "loss": 2.3725, + "step": 16939 + }, + { + "epoch": 1.7914551607445008, + "grad_norm": 8.1875, + "learning_rate": 3.71513583340961e-06, + "loss": 2.6707, + "step": 16940 + }, + { + "epoch": 1.7915609137055837, + "grad_norm": 8.4375, + "learning_rate": 3.714986164848886e-06, + "loss": 2.6116, + "step": 16941 + }, + { + "epoch": 1.7916666666666665, + "grad_norm": 8.5625, + "learning_rate": 3.7148364905867283e-06, + "loss": 2.4679, + "step": 16942 + }, + { + "epoch": 1.7917724196277496, + "grad_norm": 8.5625, + "learning_rate": 3.7146868106238398e-06, + "loss": 2.4982, + "step": 16943 + }, + { + "epoch": 1.7918781725888326, + "grad_norm": 8.0, + "learning_rate": 3.7145371249609235e-06, + "loss": 2.4877, + "step": 16944 + }, + { + "epoch": 1.7919839255499155, + "grad_norm": 7.96875, + "learning_rate": 3.7143874335986808e-06, + "loss": 2.5576, + "step": 16945 + }, + { + "epoch": 1.7920896785109983, + "grad_norm": 7.9375, + "learning_rate": 3.7142377365378147e-06, + "loss": 2.423, + "step": 16946 + }, + { + "epoch": 1.7921954314720812, + "grad_norm": 8.3125, + "learning_rate": 3.714088033779027e-06, + "loss": 2.5111, + "step": 16947 + }, + { + "epoch": 1.792301184433164, + "grad_norm": 8.9375, + "learning_rate": 3.713938325323021e-06, + "loss": 2.6385, + "step": 16948 + }, + { + "epoch": 1.792406937394247, + "grad_norm": 8.3125, + "learning_rate": 3.713788611170499e-06, + "loss": 2.7119, + "step": 16949 + }, + { + "epoch": 1.79251269035533, + "grad_norm": 8.0625, + "learning_rate": 3.713638891322163e-06, + "loss": 2.4708, + "step": 16950 + }, + { + "epoch": 1.792618443316413, + "grad_norm": 8.5625, + "learning_rate": 3.713489165778717e-06, + "loss": 2.6859, + "step": 16951 + }, + { + "epoch": 1.7927241962774958, + "grad_norm": 8.375, + "learning_rate": 3.7133394345408625e-06, + "loss": 2.5147, + "step": 16952 + }, + { + "epoch": 1.7928299492385786, + "grad_norm": 7.90625, + "learning_rate": 3.7131896976093014e-06, + "loss": 2.4685, + "step": 16953 + }, + { + "epoch": 1.7929357021996615, + "grad_norm": 7.96875, + "learning_rate": 3.713039954984738e-06, + "loss": 2.398, + "step": 16954 + }, + { + "epoch": 1.7930414551607445, + "grad_norm": 8.1875, + "learning_rate": 3.712890206667874e-06, + "loss": 2.6615, + "step": 16955 + }, + { + "epoch": 1.7931472081218274, + "grad_norm": 8.625, + "learning_rate": 3.712740452659412e-06, + "loss": 2.742, + "step": 16956 + }, + { + "epoch": 1.7932529610829104, + "grad_norm": 8.625, + "learning_rate": 3.7125906929600555e-06, + "loss": 2.5124, + "step": 16957 + }, + { + "epoch": 1.7933587140439933, + "grad_norm": 8.3125, + "learning_rate": 3.7124409275705065e-06, + "loss": 2.6283, + "step": 16958 + }, + { + "epoch": 1.7934644670050761, + "grad_norm": 8.0, + "learning_rate": 3.7122911564914677e-06, + "loss": 2.5217, + "step": 16959 + }, + { + "epoch": 1.793570219966159, + "grad_norm": 8.125, + "learning_rate": 3.7121413797236426e-06, + "loss": 2.1407, + "step": 16960 + }, + { + "epoch": 1.793675972927242, + "grad_norm": 8.375, + "learning_rate": 3.7119915972677345e-06, + "loss": 2.3878, + "step": 16961 + }, + { + "epoch": 1.7937817258883249, + "grad_norm": 8.6875, + "learning_rate": 3.7118418091244445e-06, + "loss": 2.5181, + "step": 16962 + }, + { + "epoch": 1.793887478849408, + "grad_norm": 8.25, + "learning_rate": 3.7116920152944773e-06, + "loss": 2.4748, + "step": 16963 + }, + { + "epoch": 1.7939932318104908, + "grad_norm": 8.4375, + "learning_rate": 3.7115422157785345e-06, + "loss": 2.6213, + "step": 16964 + }, + { + "epoch": 1.7940989847715736, + "grad_norm": 8.375, + "learning_rate": 3.7113924105773202e-06, + "loss": 2.5796, + "step": 16965 + }, + { + "epoch": 1.7942047377326564, + "grad_norm": 8.9375, + "learning_rate": 3.7112425996915357e-06, + "loss": 2.645, + "step": 16966 + }, + { + "epoch": 1.7943104906937393, + "grad_norm": 8.375, + "learning_rate": 3.7110927831218865e-06, + "loss": 2.7005, + "step": 16967 + }, + { + "epoch": 1.7944162436548223, + "grad_norm": 10.6875, + "learning_rate": 3.7109429608690735e-06, + "loss": 2.4469, + "step": 16968 + }, + { + "epoch": 1.7945219966159054, + "grad_norm": 8.0, + "learning_rate": 3.7107931329338e-06, + "loss": 2.4118, + "step": 16969 + }, + { + "epoch": 1.7946277495769882, + "grad_norm": 8.1875, + "learning_rate": 3.71064329931677e-06, + "loss": 2.6767, + "step": 16970 + }, + { + "epoch": 1.794733502538071, + "grad_norm": 8.0, + "learning_rate": 3.7104934600186867e-06, + "loss": 2.4869, + "step": 16971 + }, + { + "epoch": 1.794839255499154, + "grad_norm": 8.3125, + "learning_rate": 3.7103436150402516e-06, + "loss": 2.5594, + "step": 16972 + }, + { + "epoch": 1.7949450084602367, + "grad_norm": 8.875, + "learning_rate": 3.710193764382169e-06, + "loss": 2.7879, + "step": 16973 + }, + { + "epoch": 1.7950507614213198, + "grad_norm": 8.8125, + "learning_rate": 3.7100439080451423e-06, + "loss": 2.444, + "step": 16974 + }, + { + "epoch": 1.7951565143824029, + "grad_norm": 8.9375, + "learning_rate": 3.709894046029875e-06, + "loss": 2.576, + "step": 16975 + }, + { + "epoch": 1.7952622673434857, + "grad_norm": 8.6875, + "learning_rate": 3.7097441783370685e-06, + "loss": 2.5365, + "step": 16976 + }, + { + "epoch": 1.7953680203045685, + "grad_norm": 8.6875, + "learning_rate": 3.7095943049674286e-06, + "loss": 2.5399, + "step": 16977 + }, + { + "epoch": 1.7954737732656514, + "grad_norm": 8.1875, + "learning_rate": 3.7094444259216567e-06, + "loss": 2.3452, + "step": 16978 + }, + { + "epoch": 1.7955795262267342, + "grad_norm": 7.90625, + "learning_rate": 3.7092945412004565e-06, + "loss": 2.61, + "step": 16979 + }, + { + "epoch": 1.7956852791878173, + "grad_norm": 7.9375, + "learning_rate": 3.7091446508045324e-06, + "loss": 2.497, + "step": 16980 + }, + { + "epoch": 1.7957910321489001, + "grad_norm": 8.875, + "learning_rate": 3.708994754734586e-06, + "loss": 2.6445, + "step": 16981 + }, + { + "epoch": 1.7958967851099832, + "grad_norm": 8.5625, + "learning_rate": 3.7088448529913225e-06, + "loss": 2.5926, + "step": 16982 + }, + { + "epoch": 1.796002538071066, + "grad_norm": 7.84375, + "learning_rate": 3.708694945575444e-06, + "loss": 2.427, + "step": 16983 + }, + { + "epoch": 1.7961082910321489, + "grad_norm": 8.625, + "learning_rate": 3.7085450324876547e-06, + "loss": 2.4981, + "step": 16984 + }, + { + "epoch": 1.7962140439932317, + "grad_norm": 8.75, + "learning_rate": 3.708395113728658e-06, + "loss": 2.7491, + "step": 16985 + }, + { + "epoch": 1.7963197969543148, + "grad_norm": 8.0625, + "learning_rate": 3.708245189299157e-06, + "loss": 2.723, + "step": 16986 + }, + { + "epoch": 1.7964255499153976, + "grad_norm": 8.625, + "learning_rate": 3.7080952591998557e-06, + "loss": 2.653, + "step": 16987 + }, + { + "epoch": 1.7965313028764807, + "grad_norm": 8.9375, + "learning_rate": 3.7079453234314573e-06, + "loss": 2.6154, + "step": 16988 + }, + { + "epoch": 1.7966370558375635, + "grad_norm": 8.125, + "learning_rate": 3.7077953819946657e-06, + "loss": 2.4008, + "step": 16989 + }, + { + "epoch": 1.7967428087986463, + "grad_norm": 9.0, + "learning_rate": 3.7076454348901848e-06, + "loss": 2.6168, + "step": 16990 + }, + { + "epoch": 1.7968485617597292, + "grad_norm": 8.125, + "learning_rate": 3.7074954821187163e-06, + "loss": 2.7008, + "step": 16991 + }, + { + "epoch": 1.796954314720812, + "grad_norm": 8.0, + "learning_rate": 3.7073455236809663e-06, + "loss": 2.6735, + "step": 16992 + }, + { + "epoch": 1.797060067681895, + "grad_norm": 7.71875, + "learning_rate": 3.707195559577638e-06, + "loss": 2.3941, + "step": 16993 + }, + { + "epoch": 1.7971658206429781, + "grad_norm": 7.84375, + "learning_rate": 3.707045589809434e-06, + "loss": 2.5122, + "step": 16994 + }, + { + "epoch": 1.797271573604061, + "grad_norm": 8.75, + "learning_rate": 3.7068956143770584e-06, + "loss": 2.4475, + "step": 16995 + }, + { + "epoch": 1.7973773265651438, + "grad_norm": 8.25, + "learning_rate": 3.7067456332812156e-06, + "loss": 2.5434, + "step": 16996 + }, + { + "epoch": 1.7974830795262267, + "grad_norm": 8.3125, + "learning_rate": 3.7065956465226095e-06, + "loss": 2.4277, + "step": 16997 + }, + { + "epoch": 1.7975888324873095, + "grad_norm": 7.6875, + "learning_rate": 3.7064456541019434e-06, + "loss": 2.4625, + "step": 16998 + }, + { + "epoch": 1.7976945854483926, + "grad_norm": 8.875, + "learning_rate": 3.7062956560199213e-06, + "loss": 2.254, + "step": 16999 + }, + { + "epoch": 1.7978003384094756, + "grad_norm": 8.3125, + "learning_rate": 3.7061456522772466e-06, + "loss": 2.447, + "step": 17000 + }, + { + "epoch": 1.7979060913705585, + "grad_norm": 8.125, + "learning_rate": 3.7059956428746236e-06, + "loss": 2.4734, + "step": 17001 + }, + { + "epoch": 1.7980118443316413, + "grad_norm": 8.0, + "learning_rate": 3.7058456278127563e-06, + "loss": 2.5054, + "step": 17002 + }, + { + "epoch": 1.7981175972927241, + "grad_norm": 8.375, + "learning_rate": 3.7056956070923492e-06, + "loss": 2.5023, + "step": 17003 + }, + { + "epoch": 1.798223350253807, + "grad_norm": 8.1875, + "learning_rate": 3.7055455807141048e-06, + "loss": 2.5011, + "step": 17004 + }, + { + "epoch": 1.79832910321489, + "grad_norm": 8.0625, + "learning_rate": 3.7053955486787297e-06, + "loss": 2.4415, + "step": 17005 + }, + { + "epoch": 1.7984348561759729, + "grad_norm": 8.125, + "learning_rate": 3.705245510986925e-06, + "loss": 2.5035, + "step": 17006 + }, + { + "epoch": 1.798540609137056, + "grad_norm": 8.5625, + "learning_rate": 3.7050954676393965e-06, + "loss": 2.4471, + "step": 17007 + }, + { + "epoch": 1.7986463620981388, + "grad_norm": 8.4375, + "learning_rate": 3.7049454186368473e-06, + "loss": 2.7529, + "step": 17008 + }, + { + "epoch": 1.7987521150592216, + "grad_norm": 8.1875, + "learning_rate": 3.7047953639799823e-06, + "loss": 2.5959, + "step": 17009 + }, + { + "epoch": 1.7988578680203045, + "grad_norm": 8.125, + "learning_rate": 3.7046453036695056e-06, + "loss": 2.3301, + "step": 17010 + }, + { + "epoch": 1.7989636209813875, + "grad_norm": 8.1875, + "learning_rate": 3.7044952377061212e-06, + "loss": 2.5628, + "step": 17011 + }, + { + "epoch": 1.7990693739424704, + "grad_norm": 9.0625, + "learning_rate": 3.7043451660905326e-06, + "loss": 2.516, + "step": 17012 + }, + { + "epoch": 1.7991751269035534, + "grad_norm": 8.25, + "learning_rate": 3.7041950888234456e-06, + "loss": 2.7516, + "step": 17013 + }, + { + "epoch": 1.7992808798646363, + "grad_norm": 9.0625, + "learning_rate": 3.7040450059055635e-06, + "loss": 2.6002, + "step": 17014 + }, + { + "epoch": 1.799386632825719, + "grad_norm": 8.75, + "learning_rate": 3.7038949173375904e-06, + "loss": 2.5815, + "step": 17015 + }, + { + "epoch": 1.799492385786802, + "grad_norm": 8.4375, + "learning_rate": 3.70374482312023e-06, + "loss": 2.3606, + "step": 17016 + }, + { + "epoch": 1.799598138747885, + "grad_norm": 9.1875, + "learning_rate": 3.7035947232541887e-06, + "loss": 2.757, + "step": 17017 + }, + { + "epoch": 1.7997038917089678, + "grad_norm": 8.125, + "learning_rate": 3.703444617740169e-06, + "loss": 2.3775, + "step": 17018 + }, + { + "epoch": 1.799809644670051, + "grad_norm": 7.65625, + "learning_rate": 3.7032945065788762e-06, + "loss": 2.387, + "step": 17019 + }, + { + "epoch": 1.7999153976311337, + "grad_norm": 8.0, + "learning_rate": 3.703144389771014e-06, + "loss": 2.5557, + "step": 17020 + }, + { + "epoch": 1.8000211505922166, + "grad_norm": 7.96875, + "learning_rate": 3.7029942673172875e-06, + "loss": 2.7371, + "step": 17021 + }, + { + "epoch": 1.8001269035532994, + "grad_norm": 8.1875, + "learning_rate": 3.702844139218401e-06, + "loss": 2.5127, + "step": 17022 + }, + { + "epoch": 1.8002326565143822, + "grad_norm": 8.125, + "learning_rate": 3.7026940054750583e-06, + "loss": 2.4318, + "step": 17023 + }, + { + "epoch": 1.8003384094754653, + "grad_norm": 8.5625, + "learning_rate": 3.7025438660879654e-06, + "loss": 2.7042, + "step": 17024 + }, + { + "epoch": 1.8004441624365484, + "grad_norm": 8.4375, + "learning_rate": 3.702393721057826e-06, + "loss": 2.6291, + "step": 17025 + }, + { + "epoch": 1.8005499153976312, + "grad_norm": 7.875, + "learning_rate": 3.702243570385344e-06, + "loss": 2.2825, + "step": 17026 + }, + { + "epoch": 1.800655668358714, + "grad_norm": 7.9375, + "learning_rate": 3.7020934140712237e-06, + "loss": 2.4924, + "step": 17027 + }, + { + "epoch": 1.8007614213197969, + "grad_norm": 8.0, + "learning_rate": 3.7019432521161725e-06, + "loss": 2.4279, + "step": 17028 + }, + { + "epoch": 1.8008671742808797, + "grad_norm": 8.375, + "learning_rate": 3.701793084520893e-06, + "loss": 2.5687, + "step": 17029 + }, + { + "epoch": 1.8009729272419628, + "grad_norm": 8.6875, + "learning_rate": 3.701642911286089e-06, + "loss": 2.5897, + "step": 17030 + }, + { + "epoch": 1.8010786802030458, + "grad_norm": 8.1875, + "learning_rate": 3.7014927324124673e-06, + "loss": 2.4784, + "step": 17031 + }, + { + "epoch": 1.8011844331641287, + "grad_norm": 8.4375, + "learning_rate": 3.7013425479007305e-06, + "loss": 2.7776, + "step": 17032 + }, + { + "epoch": 1.8012901861252115, + "grad_norm": 8.75, + "learning_rate": 3.701192357751585e-06, + "loss": 2.7846, + "step": 17033 + }, + { + "epoch": 1.8013959390862944, + "grad_norm": 8.4375, + "learning_rate": 3.7010421619657355e-06, + "loss": 2.4468, + "step": 17034 + }, + { + "epoch": 1.8015016920473772, + "grad_norm": 8.25, + "learning_rate": 3.700891960543886e-06, + "loss": 2.6033, + "step": 17035 + }, + { + "epoch": 1.8016074450084603, + "grad_norm": 9.3125, + "learning_rate": 3.700741753486741e-06, + "loss": 2.7571, + "step": 17036 + }, + { + "epoch": 1.801713197969543, + "grad_norm": 8.625, + "learning_rate": 3.7005915407950065e-06, + "loss": 2.5148, + "step": 17037 + }, + { + "epoch": 1.8018189509306262, + "grad_norm": 7.90625, + "learning_rate": 3.7004413224693875e-06, + "loss": 2.5648, + "step": 17038 + }, + { + "epoch": 1.801924703891709, + "grad_norm": 8.25, + "learning_rate": 3.7002910985105877e-06, + "loss": 2.5944, + "step": 17039 + }, + { + "epoch": 1.8020304568527918, + "grad_norm": 8.1875, + "learning_rate": 3.700140868919313e-06, + "loss": 2.6937, + "step": 17040 + }, + { + "epoch": 1.8021362098138747, + "grad_norm": 8.0625, + "learning_rate": 3.6999906336962682e-06, + "loss": 2.5395, + "step": 17041 + }, + { + "epoch": 1.8022419627749577, + "grad_norm": 8.0, + "learning_rate": 3.699840392842158e-06, + "loss": 2.578, + "step": 17042 + }, + { + "epoch": 1.8023477157360406, + "grad_norm": 8.875, + "learning_rate": 3.699690146357687e-06, + "loss": 2.6772, + "step": 17043 + }, + { + "epoch": 1.8024534686971236, + "grad_norm": 8.5625, + "learning_rate": 3.6995398942435613e-06, + "loss": 2.6064, + "step": 17044 + }, + { + "epoch": 1.8025592216582065, + "grad_norm": 8.0, + "learning_rate": 3.6993896365004857e-06, + "loss": 2.6097, + "step": 17045 + }, + { + "epoch": 1.8026649746192893, + "grad_norm": 7.9375, + "learning_rate": 3.699239373129164e-06, + "loss": 2.4705, + "step": 17046 + }, + { + "epoch": 1.8027707275803722, + "grad_norm": 8.375, + "learning_rate": 3.6990891041303035e-06, + "loss": 2.6005, + "step": 17047 + }, + { + "epoch": 1.802876480541455, + "grad_norm": 8.8125, + "learning_rate": 3.698938829504608e-06, + "loss": 2.7009, + "step": 17048 + }, + { + "epoch": 1.802982233502538, + "grad_norm": 8.8125, + "learning_rate": 3.698788549252783e-06, + "loss": 2.428, + "step": 17049 + }, + { + "epoch": 1.8030879864636211, + "grad_norm": 8.25, + "learning_rate": 3.6986382633755334e-06, + "loss": 2.614, + "step": 17050 + }, + { + "epoch": 1.803193739424704, + "grad_norm": 8.5, + "learning_rate": 3.6984879718735645e-06, + "loss": 2.6367, + "step": 17051 + }, + { + "epoch": 1.8032994923857868, + "grad_norm": 8.5, + "learning_rate": 3.6983376747475823e-06, + "loss": 2.6275, + "step": 17052 + }, + { + "epoch": 1.8034052453468696, + "grad_norm": 8.8125, + "learning_rate": 3.6981873719982906e-06, + "loss": 2.6785, + "step": 17053 + }, + { + "epoch": 1.8035109983079525, + "grad_norm": 8.125, + "learning_rate": 3.6980370636263964e-06, + "loss": 2.5585, + "step": 17054 + }, + { + "epoch": 1.8036167512690355, + "grad_norm": 8.25, + "learning_rate": 3.697886749632604e-06, + "loss": 2.2038, + "step": 17055 + }, + { + "epoch": 1.8037225042301186, + "grad_norm": 8.75, + "learning_rate": 3.6977364300176187e-06, + "loss": 2.6105, + "step": 17056 + }, + { + "epoch": 1.8038282571912014, + "grad_norm": 9.25, + "learning_rate": 3.6975861047821472e-06, + "loss": 2.5768, + "step": 17057 + }, + { + "epoch": 1.8039340101522843, + "grad_norm": 8.0625, + "learning_rate": 3.697435773926893e-06, + "loss": 2.5708, + "step": 17058 + }, + { + "epoch": 1.8040397631133671, + "grad_norm": 8.4375, + "learning_rate": 3.6972854374525622e-06, + "loss": 2.7173, + "step": 17059 + }, + { + "epoch": 1.80414551607445, + "grad_norm": 8.25, + "learning_rate": 3.697135095359861e-06, + "loss": 2.6151, + "step": 17060 + }, + { + "epoch": 1.804251269035533, + "grad_norm": 7.96875, + "learning_rate": 3.6969847476494945e-06, + "loss": 2.4966, + "step": 17061 + }, + { + "epoch": 1.8043570219966159, + "grad_norm": 8.625, + "learning_rate": 3.6968343943221672e-06, + "loss": 2.5196, + "step": 17062 + }, + { + "epoch": 1.804462774957699, + "grad_norm": 8.0, + "learning_rate": 3.696684035378587e-06, + "loss": 2.5562, + "step": 17063 + }, + { + "epoch": 1.8045685279187818, + "grad_norm": 9.0, + "learning_rate": 3.696533670819458e-06, + "loss": 2.8428, + "step": 17064 + }, + { + "epoch": 1.8046742808798646, + "grad_norm": 8.625, + "learning_rate": 3.6963833006454847e-06, + "loss": 2.6409, + "step": 17065 + }, + { + "epoch": 1.8047800338409474, + "grad_norm": 8.125, + "learning_rate": 3.6962329248573747e-06, + "loss": 2.5152, + "step": 17066 + }, + { + "epoch": 1.8048857868020305, + "grad_norm": 8.4375, + "learning_rate": 3.6960825434558324e-06, + "loss": 2.6089, + "step": 17067 + }, + { + "epoch": 1.8049915397631133, + "grad_norm": 8.5625, + "learning_rate": 3.695932156441564e-06, + "loss": 2.5259, + "step": 17068 + }, + { + "epoch": 1.8050972927241964, + "grad_norm": 8.5625, + "learning_rate": 3.695781763815275e-06, + "loss": 2.5945, + "step": 17069 + }, + { + "epoch": 1.8052030456852792, + "grad_norm": 8.125, + "learning_rate": 3.695631365577671e-06, + "loss": 2.5451, + "step": 17070 + }, + { + "epoch": 1.805308798646362, + "grad_norm": 7.90625, + "learning_rate": 3.6954809617294583e-06, + "loss": 2.5477, + "step": 17071 + }, + { + "epoch": 1.805414551607445, + "grad_norm": 8.375, + "learning_rate": 3.6953305522713426e-06, + "loss": 2.7676, + "step": 17072 + }, + { + "epoch": 1.805520304568528, + "grad_norm": 8.0625, + "learning_rate": 3.6951801372040295e-06, + "loss": 2.5592, + "step": 17073 + }, + { + "epoch": 1.8056260575296108, + "grad_norm": 8.3125, + "learning_rate": 3.6950297165282246e-06, + "loss": 2.6083, + "step": 17074 + }, + { + "epoch": 1.8057318104906939, + "grad_norm": 8.5625, + "learning_rate": 3.6948792902446346e-06, + "loss": 2.3308, + "step": 17075 + }, + { + "epoch": 1.8058375634517767, + "grad_norm": 8.6875, + "learning_rate": 3.6947288583539643e-06, + "loss": 2.7487, + "step": 17076 + }, + { + "epoch": 1.8059433164128595, + "grad_norm": 8.0, + "learning_rate": 3.6945784208569198e-06, + "loss": 2.5898, + "step": 17077 + }, + { + "epoch": 1.8060490693739424, + "grad_norm": 8.125, + "learning_rate": 3.694427977754208e-06, + "loss": 2.5749, + "step": 17078 + }, + { + "epoch": 1.8061548223350252, + "grad_norm": 8.875, + "learning_rate": 3.694277529046533e-06, + "loss": 2.2255, + "step": 17079 + }, + { + "epoch": 1.8062605752961083, + "grad_norm": 8.0625, + "learning_rate": 3.694127074734603e-06, + "loss": 2.3799, + "step": 17080 + }, + { + "epoch": 1.8063663282571913, + "grad_norm": 8.4375, + "learning_rate": 3.6939766148191226e-06, + "loss": 2.603, + "step": 17081 + }, + { + "epoch": 1.8064720812182742, + "grad_norm": 8.0625, + "learning_rate": 3.6938261493007988e-06, + "loss": 2.5564, + "step": 17082 + }, + { + "epoch": 1.806577834179357, + "grad_norm": 12.4375, + "learning_rate": 3.6936756781803375e-06, + "loss": 2.4033, + "step": 17083 + }, + { + "epoch": 1.8066835871404399, + "grad_norm": 8.4375, + "learning_rate": 3.693525201458443e-06, + "loss": 2.5364, + "step": 17084 + }, + { + "epoch": 1.8067893401015227, + "grad_norm": 8.6875, + "learning_rate": 3.693374719135824e-06, + "loss": 2.5313, + "step": 17085 + }, + { + "epoch": 1.8068950930626058, + "grad_norm": 8.375, + "learning_rate": 3.693224231213185e-06, + "loss": 2.4555, + "step": 17086 + }, + { + "epoch": 1.8070008460236888, + "grad_norm": 8.8125, + "learning_rate": 3.693073737691233e-06, + "loss": 2.6749, + "step": 17087 + }, + { + "epoch": 1.8071065989847717, + "grad_norm": 8.875, + "learning_rate": 3.6929232385706733e-06, + "loss": 2.652, + "step": 17088 + }, + { + "epoch": 1.8072123519458545, + "grad_norm": 8.125, + "learning_rate": 3.6927727338522135e-06, + "loss": 2.4181, + "step": 17089 + }, + { + "epoch": 1.8073181049069373, + "grad_norm": 8.3125, + "learning_rate": 3.6926222235365584e-06, + "loss": 2.64, + "step": 17090 + }, + { + "epoch": 1.8074238578680202, + "grad_norm": 8.125, + "learning_rate": 3.6924717076244155e-06, + "loss": 2.6261, + "step": 17091 + }, + { + "epoch": 1.8075296108291032, + "grad_norm": 8.25, + "learning_rate": 3.6923211861164905e-06, + "loss": 2.5903, + "step": 17092 + }, + { + "epoch": 1.807635363790186, + "grad_norm": 8.125, + "learning_rate": 3.69217065901349e-06, + "loss": 2.4731, + "step": 17093 + }, + { + "epoch": 1.8077411167512691, + "grad_norm": 8.3125, + "learning_rate": 3.6920201263161194e-06, + "loss": 2.4478, + "step": 17094 + }, + { + "epoch": 1.807846869712352, + "grad_norm": 8.375, + "learning_rate": 3.6918695880250865e-06, + "loss": 2.6079, + "step": 17095 + }, + { + "epoch": 1.8079526226734348, + "grad_norm": 8.6875, + "learning_rate": 3.6917190441410964e-06, + "loss": 2.559, + "step": 17096 + }, + { + "epoch": 1.8080583756345177, + "grad_norm": 8.125, + "learning_rate": 3.691568494664857e-06, + "loss": 2.4824, + "step": 17097 + }, + { + "epoch": 1.8081641285956007, + "grad_norm": 8.1875, + "learning_rate": 3.691417939597074e-06, + "loss": 2.6669, + "step": 17098 + }, + { + "epoch": 1.8082698815566836, + "grad_norm": 8.0, + "learning_rate": 3.691267378938454e-06, + "loss": 2.5042, + "step": 17099 + }, + { + "epoch": 1.8083756345177666, + "grad_norm": 8.1875, + "learning_rate": 3.691116812689702e-06, + "loss": 2.5683, + "step": 17100 + }, + { + "epoch": 1.8084813874788495, + "grad_norm": 8.75, + "learning_rate": 3.690966240851528e-06, + "loss": 2.7106, + "step": 17101 + }, + { + "epoch": 1.8085871404399323, + "grad_norm": 8.5625, + "learning_rate": 3.6908156634246355e-06, + "loss": 2.6123, + "step": 17102 + }, + { + "epoch": 1.8086928934010151, + "grad_norm": 8.1875, + "learning_rate": 3.6906650804097323e-06, + "loss": 2.7442, + "step": 17103 + }, + { + "epoch": 1.808798646362098, + "grad_norm": 8.3125, + "learning_rate": 3.690514491807524e-06, + "loss": 2.5632, + "step": 17104 + }, + { + "epoch": 1.808904399323181, + "grad_norm": 8.5625, + "learning_rate": 3.690363897618719e-06, + "loss": 2.6125, + "step": 17105 + }, + { + "epoch": 1.809010152284264, + "grad_norm": 8.625, + "learning_rate": 3.6902132978440227e-06, + "loss": 2.6293, + "step": 17106 + }, + { + "epoch": 1.809115905245347, + "grad_norm": 8.375, + "learning_rate": 3.6900626924841422e-06, + "loss": 2.3289, + "step": 17107 + }, + { + "epoch": 1.8092216582064298, + "grad_norm": 8.1875, + "learning_rate": 3.689912081539785e-06, + "loss": 2.431, + "step": 17108 + }, + { + "epoch": 1.8093274111675126, + "grad_norm": 7.875, + "learning_rate": 3.6897614650116557e-06, + "loss": 2.3547, + "step": 17109 + }, + { + "epoch": 1.8094331641285955, + "grad_norm": 7.46875, + "learning_rate": 3.6896108429004633e-06, + "loss": 2.4203, + "step": 17110 + }, + { + "epoch": 1.8095389170896785, + "grad_norm": 8.625, + "learning_rate": 3.689460215206914e-06, + "loss": 2.6789, + "step": 17111 + }, + { + "epoch": 1.8096446700507616, + "grad_norm": 8.5625, + "learning_rate": 3.6893095819317137e-06, + "loss": 2.4612, + "step": 17112 + }, + { + "epoch": 1.8097504230118444, + "grad_norm": 8.5, + "learning_rate": 3.68915894307557e-06, + "loss": 2.4348, + "step": 17113 + }, + { + "epoch": 1.8098561759729273, + "grad_norm": 8.6875, + "learning_rate": 3.6890082986391905e-06, + "loss": 2.6131, + "step": 17114 + }, + { + "epoch": 1.80996192893401, + "grad_norm": 8.8125, + "learning_rate": 3.6888576486232804e-06, + "loss": 2.5269, + "step": 17115 + }, + { + "epoch": 1.810067681895093, + "grad_norm": 8.125, + "learning_rate": 3.6887069930285478e-06, + "loss": 2.486, + "step": 17116 + }, + { + "epoch": 1.810173434856176, + "grad_norm": 8.375, + "learning_rate": 3.6885563318556996e-06, + "loss": 2.5539, + "step": 17117 + }, + { + "epoch": 1.8102791878172588, + "grad_norm": 8.25, + "learning_rate": 3.6884056651054427e-06, + "loss": 2.6636, + "step": 17118 + }, + { + "epoch": 1.810384940778342, + "grad_norm": 8.8125, + "learning_rate": 3.688254992778484e-06, + "loss": 2.7563, + "step": 17119 + }, + { + "epoch": 1.8104906937394247, + "grad_norm": 8.375, + "learning_rate": 3.6881043148755306e-06, + "loss": 2.5356, + "step": 17120 + }, + { + "epoch": 1.8105964467005076, + "grad_norm": 8.1875, + "learning_rate": 3.6879536313972895e-06, + "loss": 2.421, + "step": 17121 + }, + { + "epoch": 1.8107021996615904, + "grad_norm": 8.0625, + "learning_rate": 3.687802942344468e-06, + "loss": 2.6403, + "step": 17122 + }, + { + "epoch": 1.8108079526226735, + "grad_norm": 8.125, + "learning_rate": 3.6876522477177736e-06, + "loss": 2.4974, + "step": 17123 + }, + { + "epoch": 1.8109137055837563, + "grad_norm": 8.125, + "learning_rate": 3.6875015475179116e-06, + "loss": 2.5825, + "step": 17124 + }, + { + "epoch": 1.8110194585448394, + "grad_norm": 8.1875, + "learning_rate": 3.687350841745591e-06, + "loss": 2.5237, + "step": 17125 + }, + { + "epoch": 1.8111252115059222, + "grad_norm": 8.125, + "learning_rate": 3.6872001304015193e-06, + "loss": 2.6032, + "step": 17126 + }, + { + "epoch": 1.811230964467005, + "grad_norm": 8.4375, + "learning_rate": 3.687049413486402e-06, + "loss": 2.6128, + "step": 17127 + }, + { + "epoch": 1.8113367174280879, + "grad_norm": 8.5625, + "learning_rate": 3.6868986910009476e-06, + "loss": 2.6167, + "step": 17128 + }, + { + "epoch": 1.811442470389171, + "grad_norm": 8.4375, + "learning_rate": 3.6867479629458634e-06, + "loss": 2.4418, + "step": 17129 + }, + { + "epoch": 1.8115482233502538, + "grad_norm": 8.1875, + "learning_rate": 3.6865972293218564e-06, + "loss": 2.6388, + "step": 17130 + }, + { + "epoch": 1.8116539763113368, + "grad_norm": 7.84375, + "learning_rate": 3.686446490129633e-06, + "loss": 2.4974, + "step": 17131 + }, + { + "epoch": 1.8117597292724197, + "grad_norm": 8.3125, + "learning_rate": 3.686295745369902e-06, + "loss": 2.4731, + "step": 17132 + }, + { + "epoch": 1.8118654822335025, + "grad_norm": 8.4375, + "learning_rate": 3.6861449950433702e-06, + "loss": 2.5548, + "step": 17133 + }, + { + "epoch": 1.8119712351945854, + "grad_norm": 8.5, + "learning_rate": 3.685994239150746e-06, + "loss": 2.6679, + "step": 17134 + }, + { + "epoch": 1.8120769881556682, + "grad_norm": 9.0625, + "learning_rate": 3.6858434776927344e-06, + "loss": 2.5055, + "step": 17135 + }, + { + "epoch": 1.8121827411167513, + "grad_norm": 8.1875, + "learning_rate": 3.6856927106700456e-06, + "loss": 2.4388, + "step": 17136 + }, + { + "epoch": 1.8122884940778343, + "grad_norm": 8.6875, + "learning_rate": 3.6855419380833853e-06, + "loss": 2.4245, + "step": 17137 + }, + { + "epoch": 1.8123942470389172, + "grad_norm": 8.1875, + "learning_rate": 3.6853911599334613e-06, + "loss": 2.6621, + "step": 17138 + }, + { + "epoch": 1.8125, + "grad_norm": 8.1875, + "learning_rate": 3.6852403762209822e-06, + "loss": 2.4062, + "step": 17139 + }, + { + "epoch": 1.8126057529610828, + "grad_norm": 8.875, + "learning_rate": 3.6850895869466536e-06, + "loss": 2.7259, + "step": 17140 + }, + { + "epoch": 1.8127115059221657, + "grad_norm": 8.1875, + "learning_rate": 3.6849387921111847e-06, + "loss": 2.3832, + "step": 17141 + }, + { + "epoch": 1.8128172588832487, + "grad_norm": 8.625, + "learning_rate": 3.684787991715283e-06, + "loss": 2.7166, + "step": 17142 + }, + { + "epoch": 1.8129230118443318, + "grad_norm": 8.625, + "learning_rate": 3.684637185759655e-06, + "loss": 2.5041, + "step": 17143 + }, + { + "epoch": 1.8130287648054146, + "grad_norm": 8.0625, + "learning_rate": 3.6844863742450097e-06, + "loss": 2.484, + "step": 17144 + }, + { + "epoch": 1.8131345177664975, + "grad_norm": 8.75, + "learning_rate": 3.684335557172055e-06, + "loss": 2.6175, + "step": 17145 + }, + { + "epoch": 1.8132402707275803, + "grad_norm": 7.875, + "learning_rate": 3.6841847345414977e-06, + "loss": 2.5463, + "step": 17146 + }, + { + "epoch": 1.8133460236886632, + "grad_norm": 8.5625, + "learning_rate": 3.6840339063540447e-06, + "loss": 2.5946, + "step": 17147 + }, + { + "epoch": 1.8134517766497462, + "grad_norm": 8.375, + "learning_rate": 3.683883072610406e-06, + "loss": 2.5191, + "step": 17148 + }, + { + "epoch": 1.813557529610829, + "grad_norm": 8.0625, + "learning_rate": 3.683732233311288e-06, + "loss": 2.5686, + "step": 17149 + }, + { + "epoch": 1.8136632825719121, + "grad_norm": 10.3125, + "learning_rate": 3.683581388457398e-06, + "loss": 2.9083, + "step": 17150 + }, + { + "epoch": 1.813769035532995, + "grad_norm": 8.3125, + "learning_rate": 3.683430538049445e-06, + "loss": 2.3445, + "step": 17151 + }, + { + "epoch": 1.8138747884940778, + "grad_norm": 8.5625, + "learning_rate": 3.683279682088137e-06, + "loss": 2.498, + "step": 17152 + }, + { + "epoch": 1.8139805414551606, + "grad_norm": 8.3125, + "learning_rate": 3.683128820574181e-06, + "loss": 2.5387, + "step": 17153 + }, + { + "epoch": 1.8140862944162437, + "grad_norm": 8.8125, + "learning_rate": 3.682977953508285e-06, + "loss": 2.6443, + "step": 17154 + }, + { + "epoch": 1.8141920473773265, + "grad_norm": 8.8125, + "learning_rate": 3.6828270808911575e-06, + "loss": 2.6266, + "step": 17155 + }, + { + "epoch": 1.8142978003384096, + "grad_norm": 8.3125, + "learning_rate": 3.682676202723507e-06, + "loss": 2.5523, + "step": 17156 + }, + { + "epoch": 1.8144035532994924, + "grad_norm": 7.9375, + "learning_rate": 3.6825253190060393e-06, + "loss": 2.3984, + "step": 17157 + }, + { + "epoch": 1.8145093062605753, + "grad_norm": 9.0625, + "learning_rate": 3.6823744297394653e-06, + "loss": 2.4138, + "step": 17158 + }, + { + "epoch": 1.814615059221658, + "grad_norm": 8.6875, + "learning_rate": 3.6822235349244904e-06, + "loss": 2.7273, + "step": 17159 + }, + { + "epoch": 1.8147208121827412, + "grad_norm": 8.125, + "learning_rate": 3.6820726345618246e-06, + "loss": 2.6435, + "step": 17160 + }, + { + "epoch": 1.814826565143824, + "grad_norm": 7.6875, + "learning_rate": 3.681921728652176e-06, + "loss": 2.5424, + "step": 17161 + }, + { + "epoch": 1.814932318104907, + "grad_norm": 8.875, + "learning_rate": 3.681770817196252e-06, + "loss": 2.5088, + "step": 17162 + }, + { + "epoch": 1.81503807106599, + "grad_norm": 8.375, + "learning_rate": 3.68161990019476e-06, + "loss": 2.7297, + "step": 17163 + }, + { + "epoch": 1.8151438240270727, + "grad_norm": 8.0625, + "learning_rate": 3.68146897764841e-06, + "loss": 2.5717, + "step": 17164 + }, + { + "epoch": 1.8152495769881556, + "grad_norm": 8.0625, + "learning_rate": 3.6813180495579088e-06, + "loss": 2.4352, + "step": 17165 + }, + { + "epoch": 1.8153553299492384, + "grad_norm": 7.9375, + "learning_rate": 3.681167115923965e-06, + "loss": 2.3564, + "step": 17166 + }, + { + "epoch": 1.8154610829103215, + "grad_norm": 7.875, + "learning_rate": 3.6810161767472874e-06, + "loss": 2.3932, + "step": 17167 + }, + { + "epoch": 1.8155668358714045, + "grad_norm": 8.375, + "learning_rate": 3.680865232028583e-06, + "loss": 2.6312, + "step": 17168 + }, + { + "epoch": 1.8156725888324874, + "grad_norm": 8.3125, + "learning_rate": 3.680714281768562e-06, + "loss": 2.6098, + "step": 17169 + }, + { + "epoch": 1.8157783417935702, + "grad_norm": 8.6875, + "learning_rate": 3.6805633259679315e-06, + "loss": 2.7148, + "step": 17170 + }, + { + "epoch": 1.815884094754653, + "grad_norm": 8.3125, + "learning_rate": 3.6804123646274e-06, + "loss": 2.3208, + "step": 17171 + }, + { + "epoch": 1.815989847715736, + "grad_norm": 8.625, + "learning_rate": 3.6802613977476766e-06, + "loss": 2.6043, + "step": 17172 + }, + { + "epoch": 1.816095600676819, + "grad_norm": 8.6875, + "learning_rate": 3.6801104253294684e-06, + "loss": 2.7536, + "step": 17173 + }, + { + "epoch": 1.816201353637902, + "grad_norm": 8.5, + "learning_rate": 3.679959447373485e-06, + "loss": 2.4589, + "step": 17174 + }, + { + "epoch": 1.8163071065989849, + "grad_norm": 8.5625, + "learning_rate": 3.679808463880435e-06, + "loss": 2.5002, + "step": 17175 + }, + { + "epoch": 1.8164128595600677, + "grad_norm": 8.8125, + "learning_rate": 3.6796574748510257e-06, + "loss": 2.5891, + "step": 17176 + }, + { + "epoch": 1.8165186125211505, + "grad_norm": 8.1875, + "learning_rate": 3.679506480285966e-06, + "loss": 2.4483, + "step": 17177 + }, + { + "epoch": 1.8166243654822334, + "grad_norm": 8.5625, + "learning_rate": 3.6793554801859654e-06, + "loss": 2.3964, + "step": 17178 + }, + { + "epoch": 1.8167301184433164, + "grad_norm": 8.75, + "learning_rate": 3.679204474551731e-06, + "loss": 2.5407, + "step": 17179 + }, + { + "epoch": 1.8168358714043993, + "grad_norm": 8.375, + "learning_rate": 3.6790534633839735e-06, + "loss": 2.5558, + "step": 17180 + }, + { + "epoch": 1.8169416243654823, + "grad_norm": 8.5, + "learning_rate": 3.6789024466834e-06, + "loss": 2.4866, + "step": 17181 + }, + { + "epoch": 1.8170473773265652, + "grad_norm": 8.875, + "learning_rate": 3.6787514244507196e-06, + "loss": 2.8229, + "step": 17182 + }, + { + "epoch": 1.817153130287648, + "grad_norm": 8.625, + "learning_rate": 3.67860039668664e-06, + "loss": 2.6065, + "step": 17183 + }, + { + "epoch": 1.8172588832487309, + "grad_norm": 8.0, + "learning_rate": 3.678449363391872e-06, + "loss": 2.5804, + "step": 17184 + }, + { + "epoch": 1.817364636209814, + "grad_norm": 8.1875, + "learning_rate": 3.6782983245671223e-06, + "loss": 2.3664, + "step": 17185 + }, + { + "epoch": 1.8174703891708968, + "grad_norm": 9.0625, + "learning_rate": 3.6781472802131003e-06, + "loss": 2.5308, + "step": 17186 + }, + { + "epoch": 1.8175761421319798, + "grad_norm": 8.375, + "learning_rate": 3.6779962303305155e-06, + "loss": 2.6551, + "step": 17187 + }, + { + "epoch": 1.8176818950930627, + "grad_norm": 8.3125, + "learning_rate": 3.677845174920076e-06, + "loss": 2.5641, + "step": 17188 + }, + { + "epoch": 1.8177876480541455, + "grad_norm": 8.4375, + "learning_rate": 3.6776941139824906e-06, + "loss": 2.632, + "step": 17189 + }, + { + "epoch": 1.8178934010152283, + "grad_norm": 9.125, + "learning_rate": 3.677543047518469e-06, + "loss": 2.8606, + "step": 17190 + }, + { + "epoch": 1.8179991539763112, + "grad_norm": 7.96875, + "learning_rate": 3.677391975528719e-06, + "loss": 2.5429, + "step": 17191 + }, + { + "epoch": 1.8181049069373942, + "grad_norm": 8.25, + "learning_rate": 3.6772408980139494e-06, + "loss": 2.5398, + "step": 17192 + }, + { + "epoch": 1.8182106598984773, + "grad_norm": 10.125, + "learning_rate": 3.6770898149748714e-06, + "loss": 2.5535, + "step": 17193 + }, + { + "epoch": 1.8183164128595601, + "grad_norm": 8.25, + "learning_rate": 3.6769387264121904e-06, + "loss": 2.4715, + "step": 17194 + }, + { + "epoch": 1.818422165820643, + "grad_norm": 7.78125, + "learning_rate": 3.676787632326618e-06, + "loss": 2.3583, + "step": 17195 + }, + { + "epoch": 1.8185279187817258, + "grad_norm": 8.8125, + "learning_rate": 3.676636532718863e-06, + "loss": 2.3139, + "step": 17196 + }, + { + "epoch": 1.8186336717428087, + "grad_norm": 8.4375, + "learning_rate": 3.676485427589634e-06, + "loss": 2.6775, + "step": 17197 + }, + { + "epoch": 1.8187394247038917, + "grad_norm": 8.5625, + "learning_rate": 3.6763343169396393e-06, + "loss": 2.618, + "step": 17198 + }, + { + "epoch": 1.8188451776649748, + "grad_norm": 8.125, + "learning_rate": 3.6761832007695897e-06, + "loss": 2.4153, + "step": 17199 + }, + { + "epoch": 1.8189509306260576, + "grad_norm": 7.8125, + "learning_rate": 3.676032079080193e-06, + "loss": 2.4132, + "step": 17200 + }, + { + "epoch": 1.8190566835871405, + "grad_norm": 7.9375, + "learning_rate": 3.6758809518721583e-06, + "loss": 2.5311, + "step": 17201 + }, + { + "epoch": 1.8191624365482233, + "grad_norm": 8.5625, + "learning_rate": 3.675729819146196e-06, + "loss": 2.5613, + "step": 17202 + }, + { + "epoch": 1.8192681895093061, + "grad_norm": 8.0, + "learning_rate": 3.6755786809030135e-06, + "loss": 2.4919, + "step": 17203 + }, + { + "epoch": 1.8193739424703892, + "grad_norm": 8.4375, + "learning_rate": 3.6754275371433216e-06, + "loss": 2.6779, + "step": 17204 + }, + { + "epoch": 1.819479695431472, + "grad_norm": 8.1875, + "learning_rate": 3.6752763878678287e-06, + "loss": 2.5011, + "step": 17205 + }, + { + "epoch": 1.819585448392555, + "grad_norm": 8.4375, + "learning_rate": 3.6751252330772453e-06, + "loss": 2.4839, + "step": 17206 + }, + { + "epoch": 1.819691201353638, + "grad_norm": 8.1875, + "learning_rate": 3.6749740727722793e-06, + "loss": 2.6143, + "step": 17207 + }, + { + "epoch": 1.8197969543147208, + "grad_norm": 8.25, + "learning_rate": 3.67482290695364e-06, + "loss": 2.3459, + "step": 17208 + }, + { + "epoch": 1.8199027072758036, + "grad_norm": 8.1875, + "learning_rate": 3.674671735622038e-06, + "loss": 2.5293, + "step": 17209 + }, + { + "epoch": 1.8200084602368867, + "grad_norm": 7.84375, + "learning_rate": 3.674520558778182e-06, + "loss": 2.4735, + "step": 17210 + }, + { + "epoch": 1.8201142131979695, + "grad_norm": 7.9375, + "learning_rate": 3.6743693764227806e-06, + "loss": 2.476, + "step": 17211 + }, + { + "epoch": 1.8202199661590526, + "grad_norm": 8.6875, + "learning_rate": 3.674218188556544e-06, + "loss": 2.5924, + "step": 17212 + }, + { + "epoch": 1.8203257191201354, + "grad_norm": 9.3125, + "learning_rate": 3.674066995180183e-06, + "loss": 2.409, + "step": 17213 + }, + { + "epoch": 1.8204314720812182, + "grad_norm": 8.125, + "learning_rate": 3.673915796294404e-06, + "loss": 2.4577, + "step": 17214 + }, + { + "epoch": 1.820537225042301, + "grad_norm": 7.9375, + "learning_rate": 3.67376459189992e-06, + "loss": 2.7303, + "step": 17215 + }, + { + "epoch": 1.8206429780033841, + "grad_norm": 8.8125, + "learning_rate": 3.673613381997438e-06, + "loss": 2.6156, + "step": 17216 + }, + { + "epoch": 1.820748730964467, + "grad_norm": 8.75, + "learning_rate": 3.673462166587668e-06, + "loss": 2.3705, + "step": 17217 + }, + { + "epoch": 1.82085448392555, + "grad_norm": 8.5, + "learning_rate": 3.6733109456713205e-06, + "loss": 2.5545, + "step": 17218 + }, + { + "epoch": 1.8209602368866329, + "grad_norm": 8.5, + "learning_rate": 3.6731597192491047e-06, + "loss": 2.5899, + "step": 17219 + }, + { + "epoch": 1.8210659898477157, + "grad_norm": 8.5, + "learning_rate": 3.6730084873217296e-06, + "loss": 2.6198, + "step": 17220 + }, + { + "epoch": 1.8211717428087986, + "grad_norm": 8.375, + "learning_rate": 3.6728572498899058e-06, + "loss": 2.5572, + "step": 17221 + }, + { + "epoch": 1.8212774957698814, + "grad_norm": 8.0625, + "learning_rate": 3.6727060069543423e-06, + "loss": 2.461, + "step": 17222 + }, + { + "epoch": 1.8213832487309645, + "grad_norm": 8.9375, + "learning_rate": 3.6725547585157497e-06, + "loss": 2.5716, + "step": 17223 + }, + { + "epoch": 1.8214890016920475, + "grad_norm": 8.25, + "learning_rate": 3.672403504574836e-06, + "loss": 2.4696, + "step": 17224 + }, + { + "epoch": 1.8215947546531304, + "grad_norm": 8.5, + "learning_rate": 3.672252245132314e-06, + "loss": 2.396, + "step": 17225 + }, + { + "epoch": 1.8217005076142132, + "grad_norm": 8.9375, + "learning_rate": 3.6721009801888906e-06, + "loss": 2.7047, + "step": 17226 + }, + { + "epoch": 1.821806260575296, + "grad_norm": 8.5625, + "learning_rate": 3.6719497097452765e-06, + "loss": 2.4382, + "step": 17227 + }, + { + "epoch": 1.8219120135363789, + "grad_norm": 9.0, + "learning_rate": 3.6717984338021824e-06, + "loss": 2.5482, + "step": 17228 + }, + { + "epoch": 1.822017766497462, + "grad_norm": 8.0, + "learning_rate": 3.671647152360317e-06, + "loss": 2.775, + "step": 17229 + }, + { + "epoch": 1.822123519458545, + "grad_norm": 8.3125, + "learning_rate": 3.67149586542039e-06, + "loss": 2.3857, + "step": 17230 + }, + { + "epoch": 1.8222292724196278, + "grad_norm": 8.125, + "learning_rate": 3.6713445729831133e-06, + "loss": 2.5145, + "step": 17231 + }, + { + "epoch": 1.8223350253807107, + "grad_norm": 8.5625, + "learning_rate": 3.671193275049195e-06, + "loss": 2.63, + "step": 17232 + }, + { + "epoch": 1.8224407783417935, + "grad_norm": 8.125, + "learning_rate": 3.671041971619346e-06, + "loss": 2.6179, + "step": 17233 + }, + { + "epoch": 1.8225465313028764, + "grad_norm": 9.625, + "learning_rate": 3.6708906626942753e-06, + "loss": 2.4906, + "step": 17234 + }, + { + "epoch": 1.8226522842639594, + "grad_norm": 8.4375, + "learning_rate": 3.6707393482746943e-06, + "loss": 2.1822, + "step": 17235 + }, + { + "epoch": 1.8227580372250423, + "grad_norm": 8.375, + "learning_rate": 3.6705880283613115e-06, + "loss": 2.5526, + "step": 17236 + }, + { + "epoch": 1.8228637901861253, + "grad_norm": 8.3125, + "learning_rate": 3.6704367029548386e-06, + "loss": 2.4038, + "step": 17237 + }, + { + "epoch": 1.8229695431472082, + "grad_norm": 8.625, + "learning_rate": 3.670285372055985e-06, + "loss": 2.7333, + "step": 17238 + }, + { + "epoch": 1.823075296108291, + "grad_norm": 8.25, + "learning_rate": 3.67013403566546e-06, + "loss": 2.3437, + "step": 17239 + }, + { + "epoch": 1.8231810490693738, + "grad_norm": 7.78125, + "learning_rate": 3.669982693783975e-06, + "loss": 2.4963, + "step": 17240 + }, + { + "epoch": 1.823286802030457, + "grad_norm": 8.5625, + "learning_rate": 3.6698313464122403e-06, + "loss": 2.7618, + "step": 17241 + }, + { + "epoch": 1.8233925549915397, + "grad_norm": 8.4375, + "learning_rate": 3.6696799935509652e-06, + "loss": 2.7278, + "step": 17242 + }, + { + "epoch": 1.8234983079526228, + "grad_norm": 8.375, + "learning_rate": 3.66952863520086e-06, + "loss": 2.5151, + "step": 17243 + }, + { + "epoch": 1.8236040609137056, + "grad_norm": 8.5625, + "learning_rate": 3.6693772713626353e-06, + "loss": 2.5577, + "step": 17244 + }, + { + "epoch": 1.8237098138747885, + "grad_norm": 8.625, + "learning_rate": 3.6692259020370015e-06, + "loss": 2.8699, + "step": 17245 + }, + { + "epoch": 1.8238155668358713, + "grad_norm": 8.5625, + "learning_rate": 3.669074527224668e-06, + "loss": 2.5754, + "step": 17246 + }, + { + "epoch": 1.8239213197969542, + "grad_norm": 7.96875, + "learning_rate": 3.6689231469263466e-06, + "loss": 2.5228, + "step": 17247 + }, + { + "epoch": 1.8240270727580372, + "grad_norm": 8.4375, + "learning_rate": 3.6687717611427465e-06, + "loss": 2.4971, + "step": 17248 + }, + { + "epoch": 1.8241328257191203, + "grad_norm": 8.75, + "learning_rate": 3.668620369874579e-06, + "loss": 2.6781, + "step": 17249 + }, + { + "epoch": 1.8242385786802031, + "grad_norm": 8.25, + "learning_rate": 3.6684689731225535e-06, + "loss": 2.5086, + "step": 17250 + }, + { + "epoch": 1.824344331641286, + "grad_norm": 8.375, + "learning_rate": 3.668317570887382e-06, + "loss": 2.6941, + "step": 17251 + }, + { + "epoch": 1.8244500846023688, + "grad_norm": 8.25, + "learning_rate": 3.668166163169773e-06, + "loss": 2.6317, + "step": 17252 + }, + { + "epoch": 1.8245558375634516, + "grad_norm": 8.125, + "learning_rate": 3.6680147499704388e-06, + "loss": 2.3602, + "step": 17253 + }, + { + "epoch": 1.8246615905245347, + "grad_norm": 9.0, + "learning_rate": 3.6678633312900886e-06, + "loss": 2.5681, + "step": 17254 + }, + { + "epoch": 1.8247673434856178, + "grad_norm": 8.75, + "learning_rate": 3.667711907129433e-06, + "loss": 2.5926, + "step": 17255 + }, + { + "epoch": 1.8248730964467006, + "grad_norm": 8.8125, + "learning_rate": 3.667560477489183e-06, + "loss": 2.4063, + "step": 17256 + }, + { + "epoch": 1.8249788494077834, + "grad_norm": 7.84375, + "learning_rate": 3.66740904237005e-06, + "loss": 2.495, + "step": 17257 + }, + { + "epoch": 1.8250846023688663, + "grad_norm": 7.96875, + "learning_rate": 3.6672576017727434e-06, + "loss": 2.4489, + "step": 17258 + }, + { + "epoch": 1.825190355329949, + "grad_norm": 8.5625, + "learning_rate": 3.667106155697974e-06, + "loss": 2.4891, + "step": 17259 + }, + { + "epoch": 1.8252961082910322, + "grad_norm": 8.4375, + "learning_rate": 3.666954704146454e-06, + "loss": 2.4696, + "step": 17260 + }, + { + "epoch": 1.825401861252115, + "grad_norm": 8.3125, + "learning_rate": 3.666803247118892e-06, + "loss": 2.5966, + "step": 17261 + }, + { + "epoch": 1.825507614213198, + "grad_norm": 8.3125, + "learning_rate": 3.6666517846159988e-06, + "loss": 2.372, + "step": 17262 + }, + { + "epoch": 1.825613367174281, + "grad_norm": 8.0625, + "learning_rate": 3.6665003166384865e-06, + "loss": 2.2375, + "step": 17263 + }, + { + "epoch": 1.8257191201353637, + "grad_norm": 8.3125, + "learning_rate": 3.6663488431870657e-06, + "loss": 2.4422, + "step": 17264 + }, + { + "epoch": 1.8258248730964466, + "grad_norm": 8.3125, + "learning_rate": 3.666197364262446e-06, + "loss": 2.5884, + "step": 17265 + }, + { + "epoch": 1.8259306260575296, + "grad_norm": 8.9375, + "learning_rate": 3.6660458798653397e-06, + "loss": 2.4218, + "step": 17266 + }, + { + "epoch": 1.8260363790186125, + "grad_norm": 8.75, + "learning_rate": 3.6658943899964565e-06, + "loss": 2.3453, + "step": 17267 + }, + { + "epoch": 1.8261421319796955, + "grad_norm": 8.75, + "learning_rate": 3.6657428946565076e-06, + "loss": 2.6839, + "step": 17268 + }, + { + "epoch": 1.8262478849407784, + "grad_norm": 8.375, + "learning_rate": 3.665591393846205e-06, + "loss": 2.5215, + "step": 17269 + }, + { + "epoch": 1.8263536379018612, + "grad_norm": 8.0, + "learning_rate": 3.6654398875662585e-06, + "loss": 2.3315, + "step": 17270 + }, + { + "epoch": 1.826459390862944, + "grad_norm": 8.5, + "learning_rate": 3.6652883758173785e-06, + "loss": 2.5223, + "step": 17271 + }, + { + "epoch": 1.8265651438240271, + "grad_norm": 8.25, + "learning_rate": 3.665136858600278e-06, + "loss": 2.6425, + "step": 17272 + }, + { + "epoch": 1.82667089678511, + "grad_norm": 8.25, + "learning_rate": 3.664985335915666e-06, + "loss": 2.5074, + "step": 17273 + }, + { + "epoch": 1.826776649746193, + "grad_norm": 8.75, + "learning_rate": 3.6648338077642537e-06, + "loss": 2.6439, + "step": 17274 + }, + { + "epoch": 1.8268824027072759, + "grad_norm": 8.0, + "learning_rate": 3.6646822741467532e-06, + "loss": 2.5004, + "step": 17275 + }, + { + "epoch": 1.8269881556683587, + "grad_norm": 8.1875, + "learning_rate": 3.664530735063875e-06, + "loss": 2.7209, + "step": 17276 + }, + { + "epoch": 1.8270939086294415, + "grad_norm": 8.1875, + "learning_rate": 3.6643791905163307e-06, + "loss": 2.5653, + "step": 17277 + }, + { + "epoch": 1.8271996615905244, + "grad_norm": 8.3125, + "learning_rate": 3.6642276405048307e-06, + "loss": 2.4159, + "step": 17278 + }, + { + "epoch": 1.8273054145516074, + "grad_norm": 8.6875, + "learning_rate": 3.664076085030087e-06, + "loss": 2.4339, + "step": 17279 + }, + { + "epoch": 1.8274111675126905, + "grad_norm": 8.0625, + "learning_rate": 3.66392452409281e-06, + "loss": 2.5483, + "step": 17280 + }, + { + "epoch": 1.8275169204737733, + "grad_norm": 8.1875, + "learning_rate": 3.6637729576937115e-06, + "loss": 2.6556, + "step": 17281 + }, + { + "epoch": 1.8276226734348562, + "grad_norm": 8.125, + "learning_rate": 3.6636213858335024e-06, + "loss": 2.475, + "step": 17282 + }, + { + "epoch": 1.827728426395939, + "grad_norm": 8.0625, + "learning_rate": 3.663469808512894e-06, + "loss": 2.4952, + "step": 17283 + }, + { + "epoch": 1.8278341793570219, + "grad_norm": 8.75, + "learning_rate": 3.6633182257325968e-06, + "loss": 2.5536, + "step": 17284 + }, + { + "epoch": 1.827939932318105, + "grad_norm": 8.0625, + "learning_rate": 3.6631666374933238e-06, + "loss": 2.4162, + "step": 17285 + }, + { + "epoch": 1.828045685279188, + "grad_norm": 8.125, + "learning_rate": 3.6630150437957857e-06, + "loss": 2.4223, + "step": 17286 + }, + { + "epoch": 1.8281514382402708, + "grad_norm": 8.4375, + "learning_rate": 3.662863444640693e-06, + "loss": 2.6518, + "step": 17287 + }, + { + "epoch": 1.8282571912013537, + "grad_norm": 8.6875, + "learning_rate": 3.6627118400287587e-06, + "loss": 2.6053, + "step": 17288 + }, + { + "epoch": 1.8283629441624365, + "grad_norm": 8.375, + "learning_rate": 3.662560229960693e-06, + "loss": 2.6507, + "step": 17289 + }, + { + "epoch": 1.8284686971235193, + "grad_norm": 8.4375, + "learning_rate": 3.6624086144372066e-06, + "loss": 2.4804, + "step": 17290 + }, + { + "epoch": 1.8285744500846024, + "grad_norm": 8.3125, + "learning_rate": 3.662256993459013e-06, + "loss": 2.5079, + "step": 17291 + }, + { + "epoch": 1.8286802030456852, + "grad_norm": 8.4375, + "learning_rate": 3.662105367026822e-06, + "loss": 2.4329, + "step": 17292 + }, + { + "epoch": 1.8287859560067683, + "grad_norm": 10.25, + "learning_rate": 3.661953735141346e-06, + "loss": 2.6225, + "step": 17293 + }, + { + "epoch": 1.8288917089678511, + "grad_norm": 9.0, + "learning_rate": 3.6618020978032965e-06, + "loss": 2.6306, + "step": 17294 + }, + { + "epoch": 1.828997461928934, + "grad_norm": 8.75, + "learning_rate": 3.661650455013385e-06, + "loss": 2.5676, + "step": 17295 + }, + { + "epoch": 1.8291032148900168, + "grad_norm": 8.1875, + "learning_rate": 3.6614988067723235e-06, + "loss": 2.4738, + "step": 17296 + }, + { + "epoch": 1.8292089678510999, + "grad_norm": 7.875, + "learning_rate": 3.661347153080822e-06, + "loss": 2.5065, + "step": 17297 + }, + { + "epoch": 1.8293147208121827, + "grad_norm": 8.1875, + "learning_rate": 3.661195493939594e-06, + "loss": 2.6502, + "step": 17298 + }, + { + "epoch": 1.8294204737732658, + "grad_norm": 8.1875, + "learning_rate": 3.66104382934935e-06, + "loss": 2.5534, + "step": 17299 + }, + { + "epoch": 1.8295262267343486, + "grad_norm": 8.1875, + "learning_rate": 3.6608921593108027e-06, + "loss": 2.5795, + "step": 17300 + }, + { + "epoch": 1.8296319796954315, + "grad_norm": 8.3125, + "learning_rate": 3.6607404838246623e-06, + "loss": 2.4025, + "step": 17301 + }, + { + "epoch": 1.8297377326565143, + "grad_norm": 9.25, + "learning_rate": 3.6605888028916423e-06, + "loss": 2.8103, + "step": 17302 + }, + { + "epoch": 1.8298434856175971, + "grad_norm": 9.5625, + "learning_rate": 3.6604371165124538e-06, + "loss": 2.7475, + "step": 17303 + }, + { + "epoch": 1.8299492385786802, + "grad_norm": 8.5625, + "learning_rate": 3.6602854246878085e-06, + "loss": 2.4747, + "step": 17304 + }, + { + "epoch": 1.8300549915397633, + "grad_norm": 8.875, + "learning_rate": 3.6601337274184186e-06, + "loss": 2.4907, + "step": 17305 + }, + { + "epoch": 1.830160744500846, + "grad_norm": 8.8125, + "learning_rate": 3.6599820247049945e-06, + "loss": 2.7355, + "step": 17306 + }, + { + "epoch": 1.830266497461929, + "grad_norm": 8.25, + "learning_rate": 3.65983031654825e-06, + "loss": 2.5006, + "step": 17307 + }, + { + "epoch": 1.8303722504230118, + "grad_norm": 7.9375, + "learning_rate": 3.6596786029488963e-06, + "loss": 2.4416, + "step": 17308 + }, + { + "epoch": 1.8304780033840946, + "grad_norm": 8.1875, + "learning_rate": 3.6595268839076446e-06, + "loss": 2.5248, + "step": 17309 + }, + { + "epoch": 1.8305837563451777, + "grad_norm": 7.875, + "learning_rate": 3.6593751594252074e-06, + "loss": 2.5705, + "step": 17310 + }, + { + "epoch": 1.8306895093062607, + "grad_norm": 8.0625, + "learning_rate": 3.6592234295022975e-06, + "loss": 2.5117, + "step": 17311 + }, + { + "epoch": 1.8307952622673436, + "grad_norm": 8.25, + "learning_rate": 3.659071694139626e-06, + "loss": 2.3258, + "step": 17312 + }, + { + "epoch": 1.8309010152284264, + "grad_norm": 8.25, + "learning_rate": 3.6589199533379044e-06, + "loss": 2.4615, + "step": 17313 + }, + { + "epoch": 1.8310067681895092, + "grad_norm": 8.125, + "learning_rate": 3.658768207097846e-06, + "loss": 2.6078, + "step": 17314 + }, + { + "epoch": 1.831112521150592, + "grad_norm": 8.25, + "learning_rate": 3.6586164554201627e-06, + "loss": 2.6131, + "step": 17315 + }, + { + "epoch": 1.8312182741116751, + "grad_norm": 8.6875, + "learning_rate": 3.658464698305565e-06, + "loss": 2.4334, + "step": 17316 + }, + { + "epoch": 1.831324027072758, + "grad_norm": 8.125, + "learning_rate": 3.6583129357547676e-06, + "loss": 2.5747, + "step": 17317 + }, + { + "epoch": 1.831429780033841, + "grad_norm": 8.3125, + "learning_rate": 3.65816116776848e-06, + "loss": 2.6743, + "step": 17318 + }, + { + "epoch": 1.8315355329949239, + "grad_norm": 8.1875, + "learning_rate": 3.6580093943474164e-06, + "loss": 2.1635, + "step": 17319 + }, + { + "epoch": 1.8316412859560067, + "grad_norm": 8.25, + "learning_rate": 3.6578576154922886e-06, + "loss": 2.4218, + "step": 17320 + }, + { + "epoch": 1.8317470389170896, + "grad_norm": 8.25, + "learning_rate": 3.6577058312038082e-06, + "loss": 2.6372, + "step": 17321 + }, + { + "epoch": 1.8318527918781726, + "grad_norm": 8.3125, + "learning_rate": 3.657554041482688e-06, + "loss": 2.5927, + "step": 17322 + }, + { + "epoch": 1.8319585448392555, + "grad_norm": 8.5625, + "learning_rate": 3.6574022463296405e-06, + "loss": 2.5309, + "step": 17323 + }, + { + "epoch": 1.8320642978003385, + "grad_norm": 8.3125, + "learning_rate": 3.6572504457453774e-06, + "loss": 2.6531, + "step": 17324 + }, + { + "epoch": 1.8321700507614214, + "grad_norm": 8.6875, + "learning_rate": 3.6570986397306107e-06, + "loss": 2.5318, + "step": 17325 + }, + { + "epoch": 1.8322758037225042, + "grad_norm": 8.25, + "learning_rate": 3.6569468282860537e-06, + "loss": 2.2905, + "step": 17326 + }, + { + "epoch": 1.832381556683587, + "grad_norm": 8.0, + "learning_rate": 3.656795011412418e-06, + "loss": 2.4173, + "step": 17327 + }, + { + "epoch": 1.83248730964467, + "grad_norm": 8.25, + "learning_rate": 3.6566431891104172e-06, + "loss": 2.5179, + "step": 17328 + }, + { + "epoch": 1.832593062605753, + "grad_norm": 8.1875, + "learning_rate": 3.6564913613807616e-06, + "loss": 2.5186, + "step": 17329 + }, + { + "epoch": 1.832698815566836, + "grad_norm": 7.625, + "learning_rate": 3.6563395282241663e-06, + "loss": 2.3737, + "step": 17330 + }, + { + "epoch": 1.8328045685279188, + "grad_norm": 8.3125, + "learning_rate": 3.656187689641343e-06, + "loss": 2.5028, + "step": 17331 + }, + { + "epoch": 1.8329103214890017, + "grad_norm": 8.5625, + "learning_rate": 3.6560358456330027e-06, + "loss": 2.5608, + "step": 17332 + }, + { + "epoch": 1.8330160744500845, + "grad_norm": 8.375, + "learning_rate": 3.6558839961998595e-06, + "loss": 2.785, + "step": 17333 + }, + { + "epoch": 1.8331218274111674, + "grad_norm": 8.5625, + "learning_rate": 3.655732141342625e-06, + "loss": 2.3362, + "step": 17334 + }, + { + "epoch": 1.8332275803722504, + "grad_norm": 9.4375, + "learning_rate": 3.6555802810620123e-06, + "loss": 2.8777, + "step": 17335 + }, + { + "epoch": 1.8333333333333335, + "grad_norm": 8.375, + "learning_rate": 3.6554284153587333e-06, + "loss": 2.5553, + "step": 17336 + }, + { + "epoch": 1.8334390862944163, + "grad_norm": 8.3125, + "learning_rate": 3.6552765442335024e-06, + "loss": 2.5622, + "step": 17337 + }, + { + "epoch": 1.8335448392554992, + "grad_norm": 7.84375, + "learning_rate": 3.6551246676870307e-06, + "loss": 2.4386, + "step": 17338 + }, + { + "epoch": 1.833650592216582, + "grad_norm": 8.0, + "learning_rate": 3.6549727857200314e-06, + "loss": 2.5928, + "step": 17339 + }, + { + "epoch": 1.8337563451776648, + "grad_norm": 8.5, + "learning_rate": 3.654820898333218e-06, + "loss": 2.5487, + "step": 17340 + }, + { + "epoch": 1.833862098138748, + "grad_norm": 8.0625, + "learning_rate": 3.654669005527301e-06, + "loss": 2.4768, + "step": 17341 + }, + { + "epoch": 1.833967851099831, + "grad_norm": 8.0625, + "learning_rate": 3.6545171073029958e-06, + "loss": 2.6142, + "step": 17342 + }, + { + "epoch": 1.8340736040609138, + "grad_norm": 8.8125, + "learning_rate": 3.6543652036610133e-06, + "loss": 2.5622, + "step": 17343 + }, + { + "epoch": 1.8341793570219966, + "grad_norm": 8.3125, + "learning_rate": 3.654213294602067e-06, + "loss": 2.5549, + "step": 17344 + }, + { + "epoch": 1.8342851099830795, + "grad_norm": 9.125, + "learning_rate": 3.65406138012687e-06, + "loss": 2.6794, + "step": 17345 + }, + { + "epoch": 1.8343908629441623, + "grad_norm": 8.5625, + "learning_rate": 3.653909460236135e-06, + "loss": 2.2515, + "step": 17346 + }, + { + "epoch": 1.8344966159052454, + "grad_norm": 9.1875, + "learning_rate": 3.653757534930575e-06, + "loss": 2.4659, + "step": 17347 + }, + { + "epoch": 1.8346023688663282, + "grad_norm": 9.0625, + "learning_rate": 3.653605604210902e-06, + "loss": 2.5169, + "step": 17348 + }, + { + "epoch": 1.8347081218274113, + "grad_norm": 8.9375, + "learning_rate": 3.6534536680778304e-06, + "loss": 2.5352, + "step": 17349 + }, + { + "epoch": 1.8348138747884941, + "grad_norm": 8.6875, + "learning_rate": 3.6533017265320726e-06, + "loss": 2.5165, + "step": 17350 + }, + { + "epoch": 1.834919627749577, + "grad_norm": 8.0, + "learning_rate": 3.653149779574341e-06, + "loss": 2.5138, + "step": 17351 + }, + { + "epoch": 1.8350253807106598, + "grad_norm": 8.5, + "learning_rate": 3.6529978272053494e-06, + "loss": 2.6027, + "step": 17352 + }, + { + "epoch": 1.8351311336717429, + "grad_norm": 8.875, + "learning_rate": 3.6528458694258105e-06, + "loss": 2.7715, + "step": 17353 + }, + { + "epoch": 1.8352368866328257, + "grad_norm": 8.0625, + "learning_rate": 3.652693906236437e-06, + "loss": 2.6303, + "step": 17354 + }, + { + "epoch": 1.8353426395939088, + "grad_norm": 7.84375, + "learning_rate": 3.652541937637943e-06, + "loss": 2.6174, + "step": 17355 + }, + { + "epoch": 1.8354483925549916, + "grad_norm": 8.75, + "learning_rate": 3.6523899636310413e-06, + "loss": 2.404, + "step": 17356 + }, + { + "epoch": 1.8355541455160744, + "grad_norm": 8.5, + "learning_rate": 3.6522379842164437e-06, + "loss": 2.7386, + "step": 17357 + }, + { + "epoch": 1.8356598984771573, + "grad_norm": 8.5, + "learning_rate": 3.6520859993948655e-06, + "loss": 2.4916, + "step": 17358 + }, + { + "epoch": 1.83576565143824, + "grad_norm": 8.25, + "learning_rate": 3.6519340091670185e-06, + "loss": 2.6586, + "step": 17359 + }, + { + "epoch": 1.8358714043993232, + "grad_norm": 8.4375, + "learning_rate": 3.651782013533616e-06, + "loss": 2.5907, + "step": 17360 + }, + { + "epoch": 1.8359771573604062, + "grad_norm": 8.1875, + "learning_rate": 3.6516300124953723e-06, + "loss": 2.5101, + "step": 17361 + }, + { + "epoch": 1.836082910321489, + "grad_norm": 8.4375, + "learning_rate": 3.651478006053e-06, + "loss": 2.5358, + "step": 17362 + }, + { + "epoch": 1.836188663282572, + "grad_norm": 8.5, + "learning_rate": 3.6513259942072115e-06, + "loss": 2.6797, + "step": 17363 + }, + { + "epoch": 1.8362944162436547, + "grad_norm": 8.75, + "learning_rate": 3.6511739769587214e-06, + "loss": 2.6728, + "step": 17364 + }, + { + "epoch": 1.8364001692047376, + "grad_norm": 8.4375, + "learning_rate": 3.6510219543082426e-06, + "loss": 2.664, + "step": 17365 + }, + { + "epoch": 1.8365059221658206, + "grad_norm": 8.25, + "learning_rate": 3.650869926256489e-06, + "loss": 2.5091, + "step": 17366 + }, + { + "epoch": 1.8366116751269037, + "grad_norm": 8.4375, + "learning_rate": 3.6507178928041732e-06, + "loss": 2.7898, + "step": 17367 + }, + { + "epoch": 1.8367174280879865, + "grad_norm": 8.3125, + "learning_rate": 3.650565853952009e-06, + "loss": 2.6936, + "step": 17368 + }, + { + "epoch": 1.8368231810490694, + "grad_norm": 8.375, + "learning_rate": 3.6504138097007095e-06, + "loss": 2.6523, + "step": 17369 + }, + { + "epoch": 1.8369289340101522, + "grad_norm": 8.6875, + "learning_rate": 3.6502617600509886e-06, + "loss": 2.6246, + "step": 17370 + }, + { + "epoch": 1.837034686971235, + "grad_norm": 8.375, + "learning_rate": 3.6501097050035596e-06, + "loss": 2.4488, + "step": 17371 + }, + { + "epoch": 1.8371404399323181, + "grad_norm": 8.75, + "learning_rate": 3.649957644559137e-06, + "loss": 2.4407, + "step": 17372 + }, + { + "epoch": 1.837246192893401, + "grad_norm": 7.8125, + "learning_rate": 3.6498055787184316e-06, + "loss": 2.426, + "step": 17373 + }, + { + "epoch": 1.837351945854484, + "grad_norm": 8.0, + "learning_rate": 3.6496535074821604e-06, + "loss": 2.5366, + "step": 17374 + }, + { + "epoch": 1.8374576988155669, + "grad_norm": 7.9375, + "learning_rate": 3.649501430851036e-06, + "loss": 2.3625, + "step": 17375 + }, + { + "epoch": 1.8375634517766497, + "grad_norm": 8.3125, + "learning_rate": 3.64934934882577e-06, + "loss": 2.5311, + "step": 17376 + }, + { + "epoch": 1.8376692047377325, + "grad_norm": 8.5, + "learning_rate": 3.649197261407079e-06, + "loss": 2.5584, + "step": 17377 + }, + { + "epoch": 1.8377749576988156, + "grad_norm": 8.625, + "learning_rate": 3.649045168595674e-06, + "loss": 2.6497, + "step": 17378 + }, + { + "epoch": 1.8378807106598984, + "grad_norm": 8.25, + "learning_rate": 3.6488930703922702e-06, + "loss": 2.4631, + "step": 17379 + }, + { + "epoch": 1.8379864636209815, + "grad_norm": 8.125, + "learning_rate": 3.6487409667975815e-06, + "loss": 2.5511, + "step": 17380 + }, + { + "epoch": 1.8380922165820643, + "grad_norm": 8.25, + "learning_rate": 3.6485888578123213e-06, + "loss": 2.4105, + "step": 17381 + }, + { + "epoch": 1.8381979695431472, + "grad_norm": 8.3125, + "learning_rate": 3.6484367434372037e-06, + "loss": 2.4441, + "step": 17382 + }, + { + "epoch": 1.83830372250423, + "grad_norm": 8.375, + "learning_rate": 3.6482846236729407e-06, + "loss": 2.4407, + "step": 17383 + }, + { + "epoch": 1.838409475465313, + "grad_norm": 8.75, + "learning_rate": 3.648132498520249e-06, + "loss": 2.5706, + "step": 17384 + }, + { + "epoch": 1.838515228426396, + "grad_norm": 8.1875, + "learning_rate": 3.647980367979841e-06, + "loss": 2.5863, + "step": 17385 + }, + { + "epoch": 1.838620981387479, + "grad_norm": 8.1875, + "learning_rate": 3.64782823205243e-06, + "loss": 2.5546, + "step": 17386 + }, + { + "epoch": 1.8387267343485618, + "grad_norm": 8.25, + "learning_rate": 3.647676090738731e-06, + "loss": 2.4528, + "step": 17387 + }, + { + "epoch": 1.8388324873096447, + "grad_norm": 8.0625, + "learning_rate": 3.647523944039457e-06, + "loss": 2.4837, + "step": 17388 + }, + { + "epoch": 1.8389382402707275, + "grad_norm": 8.625, + "learning_rate": 3.6473717919553226e-06, + "loss": 2.6088, + "step": 17389 + }, + { + "epoch": 1.8390439932318103, + "grad_norm": 8.4375, + "learning_rate": 3.647219634487042e-06, + "loss": 2.5596, + "step": 17390 + }, + { + "epoch": 1.8391497461928934, + "grad_norm": 8.5625, + "learning_rate": 3.647067471635329e-06, + "loss": 2.5423, + "step": 17391 + }, + { + "epoch": 1.8392554991539765, + "grad_norm": 8.125, + "learning_rate": 3.6469153034008964e-06, + "loss": 2.5902, + "step": 17392 + }, + { + "epoch": 1.8393612521150593, + "grad_norm": 8.8125, + "learning_rate": 3.64676312978446e-06, + "loss": 2.7435, + "step": 17393 + }, + { + "epoch": 1.8394670050761421, + "grad_norm": 8.4375, + "learning_rate": 3.646610950786734e-06, + "loss": 2.5447, + "step": 17394 + }, + { + "epoch": 1.839572758037225, + "grad_norm": 8.75, + "learning_rate": 3.6464587664084307e-06, + "loss": 2.3114, + "step": 17395 + }, + { + "epoch": 1.8396785109983078, + "grad_norm": 8.5625, + "learning_rate": 3.6463065766502658e-06, + "loss": 2.6843, + "step": 17396 + }, + { + "epoch": 1.8397842639593909, + "grad_norm": 8.5, + "learning_rate": 3.646154381512953e-06, + "loss": 2.5814, + "step": 17397 + }, + { + "epoch": 1.839890016920474, + "grad_norm": 8.25, + "learning_rate": 3.646002180997206e-06, + "loss": 2.4822, + "step": 17398 + }, + { + "epoch": 1.8399957698815568, + "grad_norm": 8.5625, + "learning_rate": 3.6458499751037395e-06, + "loss": 2.6173, + "step": 17399 + }, + { + "epoch": 1.8401015228426396, + "grad_norm": 8.5625, + "learning_rate": 3.645697763833268e-06, + "loss": 2.5288, + "step": 17400 + }, + { + "epoch": 1.8402072758037225, + "grad_norm": 8.125, + "learning_rate": 3.645545547186506e-06, + "loss": 2.3734, + "step": 17401 + }, + { + "epoch": 1.8403130287648053, + "grad_norm": 8.0, + "learning_rate": 3.645393325164166e-06, + "loss": 2.5902, + "step": 17402 + }, + { + "epoch": 1.8404187817258884, + "grad_norm": 7.6875, + "learning_rate": 3.645241097766965e-06, + "loss": 2.3724, + "step": 17403 + }, + { + "epoch": 1.8405245346869712, + "grad_norm": 8.3125, + "learning_rate": 3.645088864995615e-06, + "loss": 2.6034, + "step": 17404 + }, + { + "epoch": 1.8406302876480543, + "grad_norm": 7.90625, + "learning_rate": 3.644936626850831e-06, + "loss": 2.4149, + "step": 17405 + }, + { + "epoch": 1.840736040609137, + "grad_norm": 8.8125, + "learning_rate": 3.6447843833333287e-06, + "loss": 2.5927, + "step": 17406 + }, + { + "epoch": 1.84084179357022, + "grad_norm": 8.25, + "learning_rate": 3.644632134443821e-06, + "loss": 2.4258, + "step": 17407 + }, + { + "epoch": 1.8409475465313028, + "grad_norm": 7.90625, + "learning_rate": 3.644479880183022e-06, + "loss": 2.5636, + "step": 17408 + }, + { + "epoch": 1.8410532994923858, + "grad_norm": 8.6875, + "learning_rate": 3.6443276205516482e-06, + "loss": 2.6743, + "step": 17409 + }, + { + "epoch": 1.8411590524534687, + "grad_norm": 8.3125, + "learning_rate": 3.6441753555504124e-06, + "loss": 2.6052, + "step": 17410 + }, + { + "epoch": 1.8412648054145517, + "grad_norm": 8.125, + "learning_rate": 3.6440230851800296e-06, + "loss": 2.4817, + "step": 17411 + }, + { + "epoch": 1.8413705583756346, + "grad_norm": 8.0, + "learning_rate": 3.6438708094412146e-06, + "loss": 2.5856, + "step": 17412 + }, + { + "epoch": 1.8414763113367174, + "grad_norm": 8.625, + "learning_rate": 3.6437185283346816e-06, + "loss": 2.4759, + "step": 17413 + }, + { + "epoch": 1.8415820642978002, + "grad_norm": 8.4375, + "learning_rate": 3.643566241861145e-06, + "loss": 2.6049, + "step": 17414 + }, + { + "epoch": 1.841687817258883, + "grad_norm": 8.375, + "learning_rate": 3.64341395002132e-06, + "loss": 2.5165, + "step": 17415 + }, + { + "epoch": 1.8417935702199661, + "grad_norm": 7.46875, + "learning_rate": 3.6432616528159203e-06, + "loss": 2.397, + "step": 17416 + }, + { + "epoch": 1.8418993231810492, + "grad_norm": 8.0625, + "learning_rate": 3.6431093502456626e-06, + "loss": 2.6253, + "step": 17417 + }, + { + "epoch": 1.842005076142132, + "grad_norm": 8.125, + "learning_rate": 3.642957042311259e-06, + "loss": 2.4292, + "step": 17418 + }, + { + "epoch": 1.8421108291032149, + "grad_norm": 8.5, + "learning_rate": 3.642804729013426e-06, + "loss": 2.5497, + "step": 17419 + }, + { + "epoch": 1.8422165820642977, + "grad_norm": 8.5, + "learning_rate": 3.6426524103528784e-06, + "loss": 2.528, + "step": 17420 + }, + { + "epoch": 1.8423223350253806, + "grad_norm": 8.625, + "learning_rate": 3.6425000863303297e-06, + "loss": 2.7098, + "step": 17421 + }, + { + "epoch": 1.8424280879864636, + "grad_norm": 8.625, + "learning_rate": 3.642347756946495e-06, + "loss": 2.5957, + "step": 17422 + }, + { + "epoch": 1.8425338409475467, + "grad_norm": 8.375, + "learning_rate": 3.6421954222020904e-06, + "loss": 2.632, + "step": 17423 + }, + { + "epoch": 1.8426395939086295, + "grad_norm": 8.1875, + "learning_rate": 3.642043082097828e-06, + "loss": 2.4255, + "step": 17424 + }, + { + "epoch": 1.8427453468697124, + "grad_norm": 8.375, + "learning_rate": 3.6418907366344265e-06, + "loss": 2.6708, + "step": 17425 + }, + { + "epoch": 1.8428510998307952, + "grad_norm": 8.375, + "learning_rate": 3.641738385812598e-06, + "loss": 2.4935, + "step": 17426 + }, + { + "epoch": 1.842956852791878, + "grad_norm": 8.125, + "learning_rate": 3.641586029633058e-06, + "loss": 2.3236, + "step": 17427 + }, + { + "epoch": 1.843062605752961, + "grad_norm": 8.125, + "learning_rate": 3.6414336680965225e-06, + "loss": 2.5466, + "step": 17428 + }, + { + "epoch": 1.843168358714044, + "grad_norm": 8.3125, + "learning_rate": 3.641281301203705e-06, + "loss": 2.4265, + "step": 17429 + }, + { + "epoch": 1.843274111675127, + "grad_norm": 8.1875, + "learning_rate": 3.6411289289553207e-06, + "loss": 2.1969, + "step": 17430 + }, + { + "epoch": 1.8433798646362098, + "grad_norm": 8.25, + "learning_rate": 3.6409765513520863e-06, + "loss": 2.6525, + "step": 17431 + }, + { + "epoch": 1.8434856175972927, + "grad_norm": 8.0, + "learning_rate": 3.6408241683947148e-06, + "loss": 2.551, + "step": 17432 + }, + { + "epoch": 1.8435913705583755, + "grad_norm": 9.5, + "learning_rate": 3.640671780083922e-06, + "loss": 2.8884, + "step": 17433 + }, + { + "epoch": 1.8436971235194586, + "grad_norm": 8.0625, + "learning_rate": 3.6405193864204226e-06, + "loss": 2.5867, + "step": 17434 + }, + { + "epoch": 1.8438028764805414, + "grad_norm": 8.5625, + "learning_rate": 3.640366987404933e-06, + "loss": 2.5346, + "step": 17435 + }, + { + "epoch": 1.8439086294416245, + "grad_norm": 8.875, + "learning_rate": 3.6402145830381673e-06, + "loss": 2.5162, + "step": 17436 + }, + { + "epoch": 1.8440143824027073, + "grad_norm": 8.1875, + "learning_rate": 3.640062173320841e-06, + "loss": 2.5765, + "step": 17437 + }, + { + "epoch": 1.8441201353637902, + "grad_norm": 8.8125, + "learning_rate": 3.639909758253669e-06, + "loss": 2.5358, + "step": 17438 + }, + { + "epoch": 1.844225888324873, + "grad_norm": 7.46875, + "learning_rate": 3.639757337837367e-06, + "loss": 2.2896, + "step": 17439 + }, + { + "epoch": 1.844331641285956, + "grad_norm": 7.9375, + "learning_rate": 3.639604912072649e-06, + "loss": 2.4746, + "step": 17440 + }, + { + "epoch": 1.844437394247039, + "grad_norm": 8.375, + "learning_rate": 3.639452480960232e-06, + "loss": 2.5266, + "step": 17441 + }, + { + "epoch": 1.844543147208122, + "grad_norm": 8.3125, + "learning_rate": 3.6393000445008304e-06, + "loss": 2.5636, + "step": 17442 + }, + { + "epoch": 1.8446489001692048, + "grad_norm": 8.4375, + "learning_rate": 3.6391476026951596e-06, + "loss": 2.6162, + "step": 17443 + }, + { + "epoch": 1.8447546531302876, + "grad_norm": 9.0, + "learning_rate": 3.638995155543935e-06, + "loss": 2.5027, + "step": 17444 + }, + { + "epoch": 1.8448604060913705, + "grad_norm": 8.125, + "learning_rate": 3.6388427030478724e-06, + "loss": 2.5666, + "step": 17445 + }, + { + "epoch": 1.8449661590524533, + "grad_norm": 8.75, + "learning_rate": 3.638690245207686e-06, + "loss": 2.4985, + "step": 17446 + }, + { + "epoch": 1.8450719120135364, + "grad_norm": 8.1875, + "learning_rate": 3.638537782024093e-06, + "loss": 2.5851, + "step": 17447 + }, + { + "epoch": 1.8451776649746194, + "grad_norm": 7.9375, + "learning_rate": 3.638385313497807e-06, + "loss": 2.6266, + "step": 17448 + }, + { + "epoch": 1.8452834179357023, + "grad_norm": 8.25, + "learning_rate": 3.638232839629544e-06, + "loss": 2.6951, + "step": 17449 + }, + { + "epoch": 1.8453891708967851, + "grad_norm": 8.25, + "learning_rate": 3.638080360420021e-06, + "loss": 2.538, + "step": 17450 + }, + { + "epoch": 1.845494923857868, + "grad_norm": 8.6875, + "learning_rate": 3.6379278758699515e-06, + "loss": 2.7567, + "step": 17451 + }, + { + "epoch": 1.8456006768189508, + "grad_norm": 8.375, + "learning_rate": 3.6377753859800524e-06, + "loss": 2.3602, + "step": 17452 + }, + { + "epoch": 1.8457064297800339, + "grad_norm": 8.3125, + "learning_rate": 3.6376228907510382e-06, + "loss": 2.6295, + "step": 17453 + }, + { + "epoch": 1.845812182741117, + "grad_norm": 8.25, + "learning_rate": 3.637470390183625e-06, + "loss": 2.7274, + "step": 17454 + }, + { + "epoch": 1.8459179357021998, + "grad_norm": 8.125, + "learning_rate": 3.637317884278529e-06, + "loss": 2.5419, + "step": 17455 + }, + { + "epoch": 1.8460236886632826, + "grad_norm": 8.375, + "learning_rate": 3.637165373036465e-06, + "loss": 2.6175, + "step": 17456 + }, + { + "epoch": 1.8461294416243654, + "grad_norm": 8.6875, + "learning_rate": 3.6370128564581492e-06, + "loss": 2.3838, + "step": 17457 + }, + { + "epoch": 1.8462351945854483, + "grad_norm": 9.0, + "learning_rate": 3.636860334544297e-06, + "loss": 2.3721, + "step": 17458 + }, + { + "epoch": 1.8463409475465313, + "grad_norm": 8.0, + "learning_rate": 3.636707807295624e-06, + "loss": 2.3222, + "step": 17459 + }, + { + "epoch": 1.8464467005076142, + "grad_norm": 8.1875, + "learning_rate": 3.6365552747128458e-06, + "loss": 2.5509, + "step": 17460 + }, + { + "epoch": 1.8465524534686972, + "grad_norm": 8.875, + "learning_rate": 3.636402736796679e-06, + "loss": 2.6778, + "step": 17461 + }, + { + "epoch": 1.84665820642978, + "grad_norm": 8.5, + "learning_rate": 3.6362501935478388e-06, + "loss": 2.6734, + "step": 17462 + }, + { + "epoch": 1.846763959390863, + "grad_norm": 8.75, + "learning_rate": 3.6360976449670416e-06, + "loss": 2.8111, + "step": 17463 + }, + { + "epoch": 1.8468697123519457, + "grad_norm": 8.3125, + "learning_rate": 3.6359450910550025e-06, + "loss": 2.5849, + "step": 17464 + }, + { + "epoch": 1.8469754653130288, + "grad_norm": 8.25, + "learning_rate": 3.6357925318124377e-06, + "loss": 2.6469, + "step": 17465 + }, + { + "epoch": 1.8470812182741116, + "grad_norm": 7.96875, + "learning_rate": 3.6356399672400633e-06, + "loss": 2.3944, + "step": 17466 + }, + { + "epoch": 1.8471869712351947, + "grad_norm": 7.875, + "learning_rate": 3.6354873973385947e-06, + "loss": 2.4963, + "step": 17467 + }, + { + "epoch": 1.8472927241962775, + "grad_norm": 8.8125, + "learning_rate": 3.635334822108748e-06, + "loss": 2.6994, + "step": 17468 + }, + { + "epoch": 1.8473984771573604, + "grad_norm": 8.8125, + "learning_rate": 3.6351822415512395e-06, + "loss": 2.3492, + "step": 17469 + }, + { + "epoch": 1.8475042301184432, + "grad_norm": 7.90625, + "learning_rate": 3.635029655666785e-06, + "loss": 2.625, + "step": 17470 + }, + { + "epoch": 1.847609983079526, + "grad_norm": 8.5625, + "learning_rate": 3.6348770644561005e-06, + "loss": 2.5871, + "step": 17471 + }, + { + "epoch": 1.8477157360406091, + "grad_norm": 8.375, + "learning_rate": 3.6347244679199015e-06, + "loss": 2.7014, + "step": 17472 + }, + { + "epoch": 1.8478214890016922, + "grad_norm": 8.5, + "learning_rate": 3.6345718660589058e-06, + "loss": 2.5913, + "step": 17473 + }, + { + "epoch": 1.847927241962775, + "grad_norm": 9.5625, + "learning_rate": 3.634419258873828e-06, + "loss": 2.2622, + "step": 17474 + }, + { + "epoch": 1.8480329949238579, + "grad_norm": 8.6875, + "learning_rate": 3.634266646365384e-06, + "loss": 2.6161, + "step": 17475 + }, + { + "epoch": 1.8481387478849407, + "grad_norm": 8.6875, + "learning_rate": 3.634114028534291e-06, + "loss": 2.4223, + "step": 17476 + }, + { + "epoch": 1.8482445008460235, + "grad_norm": 7.71875, + "learning_rate": 3.633961405381265e-06, + "loss": 2.3315, + "step": 17477 + }, + { + "epoch": 1.8483502538071066, + "grad_norm": 11.375, + "learning_rate": 3.6338087769070214e-06, + "loss": 2.5065, + "step": 17478 + }, + { + "epoch": 1.8484560067681897, + "grad_norm": 7.84375, + "learning_rate": 3.633656143112277e-06, + "loss": 2.5451, + "step": 17479 + }, + { + "epoch": 1.8485617597292725, + "grad_norm": 8.1875, + "learning_rate": 3.6335035039977486e-06, + "loss": 2.5265, + "step": 17480 + }, + { + "epoch": 1.8486675126903553, + "grad_norm": 9.0625, + "learning_rate": 3.633350859564151e-06, + "loss": 2.8574, + "step": 17481 + }, + { + "epoch": 1.8487732656514382, + "grad_norm": 8.1875, + "learning_rate": 3.6331982098122016e-06, + "loss": 2.4617, + "step": 17482 + }, + { + "epoch": 1.848879018612521, + "grad_norm": 7.96875, + "learning_rate": 3.6330455547426167e-06, + "loss": 2.4147, + "step": 17483 + }, + { + "epoch": 1.848984771573604, + "grad_norm": 8.25, + "learning_rate": 3.6328928943561115e-06, + "loss": 2.7364, + "step": 17484 + }, + { + "epoch": 1.849090524534687, + "grad_norm": 8.3125, + "learning_rate": 3.632740228653405e-06, + "loss": 2.6246, + "step": 17485 + }, + { + "epoch": 1.84919627749577, + "grad_norm": 8.5, + "learning_rate": 3.6325875576352097e-06, + "loss": 2.7919, + "step": 17486 + }, + { + "epoch": 1.8493020304568528, + "grad_norm": 8.25, + "learning_rate": 3.632434881302246e-06, + "loss": 2.5273, + "step": 17487 + }, + { + "epoch": 1.8494077834179357, + "grad_norm": 8.625, + "learning_rate": 3.6322821996552277e-06, + "loss": 2.4635, + "step": 17488 + }, + { + "epoch": 1.8495135363790185, + "grad_norm": 7.90625, + "learning_rate": 3.6321295126948726e-06, + "loss": 2.4317, + "step": 17489 + }, + { + "epoch": 1.8496192893401016, + "grad_norm": 8.0625, + "learning_rate": 3.6319768204218964e-06, + "loss": 2.4291, + "step": 17490 + }, + { + "epoch": 1.8497250423011844, + "grad_norm": 9.5625, + "learning_rate": 3.631824122837016e-06, + "loss": 2.5075, + "step": 17491 + }, + { + "epoch": 1.8498307952622675, + "grad_norm": 8.1875, + "learning_rate": 3.6316714199409482e-06, + "loss": 2.6222, + "step": 17492 + }, + { + "epoch": 1.8499365482233503, + "grad_norm": 8.125, + "learning_rate": 3.6315187117344086e-06, + "loss": 2.652, + "step": 17493 + }, + { + "epoch": 1.8500423011844331, + "grad_norm": 7.9375, + "learning_rate": 3.631365998218115e-06, + "loss": 2.5783, + "step": 17494 + }, + { + "epoch": 1.850148054145516, + "grad_norm": 7.9375, + "learning_rate": 3.6312132793927828e-06, + "loss": 2.5072, + "step": 17495 + }, + { + "epoch": 1.850253807106599, + "grad_norm": 8.3125, + "learning_rate": 3.6310605552591307e-06, + "loss": 2.611, + "step": 17496 + }, + { + "epoch": 1.8503595600676819, + "grad_norm": 8.375, + "learning_rate": 3.630907825817872e-06, + "loss": 2.4789, + "step": 17497 + }, + { + "epoch": 1.850465313028765, + "grad_norm": 8.4375, + "learning_rate": 3.630755091069727e-06, + "loss": 2.5514, + "step": 17498 + }, + { + "epoch": 1.8505710659898478, + "grad_norm": 8.25, + "learning_rate": 3.630602351015411e-06, + "loss": 2.5763, + "step": 17499 + }, + { + "epoch": 1.8506768189509306, + "grad_norm": 8.875, + "learning_rate": 3.6304496056556393e-06, + "loss": 2.7387, + "step": 17500 + }, + { + "epoch": 1.8507825719120135, + "grad_norm": 8.5625, + "learning_rate": 3.630296854991131e-06, + "loss": 2.6939, + "step": 17501 + }, + { + "epoch": 1.8508883248730963, + "grad_norm": 9.3125, + "learning_rate": 3.630144099022601e-06, + "loss": 2.6561, + "step": 17502 + }, + { + "epoch": 1.8509940778341794, + "grad_norm": 8.3125, + "learning_rate": 3.629991337750768e-06, + "loss": 2.5517, + "step": 17503 + }, + { + "epoch": 1.8510998307952624, + "grad_norm": 8.875, + "learning_rate": 3.629838571176346e-06, + "loss": 2.6871, + "step": 17504 + }, + { + "epoch": 1.8512055837563453, + "grad_norm": 8.5625, + "learning_rate": 3.6296857993000555e-06, + "loss": 2.4989, + "step": 17505 + }, + { + "epoch": 1.851311336717428, + "grad_norm": 8.9375, + "learning_rate": 3.629533022122611e-06, + "loss": 2.7933, + "step": 17506 + }, + { + "epoch": 1.851417089678511, + "grad_norm": 8.0, + "learning_rate": 3.6293802396447296e-06, + "loss": 2.563, + "step": 17507 + }, + { + "epoch": 1.8515228426395938, + "grad_norm": 8.8125, + "learning_rate": 3.629227451867129e-06, + "loss": 2.7461, + "step": 17508 + }, + { + "epoch": 1.8516285956006768, + "grad_norm": 8.4375, + "learning_rate": 3.6290746587905264e-06, + "loss": 2.5692, + "step": 17509 + }, + { + "epoch": 1.85173434856176, + "grad_norm": 8.125, + "learning_rate": 3.628921860415637e-06, + "loss": 2.6496, + "step": 17510 + }, + { + "epoch": 1.8518401015228427, + "grad_norm": 9.1875, + "learning_rate": 3.62876905674318e-06, + "loss": 2.9279, + "step": 17511 + }, + { + "epoch": 1.8519458544839256, + "grad_norm": 8.875, + "learning_rate": 3.6286162477738714e-06, + "loss": 2.8106, + "step": 17512 + }, + { + "epoch": 1.8520516074450084, + "grad_norm": 8.1875, + "learning_rate": 3.6284634335084267e-06, + "loss": 2.706, + "step": 17513 + }, + { + "epoch": 1.8521573604060912, + "grad_norm": 8.625, + "learning_rate": 3.6283106139475664e-06, + "loss": 2.7121, + "step": 17514 + }, + { + "epoch": 1.8522631133671743, + "grad_norm": 8.1875, + "learning_rate": 3.6281577890920055e-06, + "loss": 2.5125, + "step": 17515 + }, + { + "epoch": 1.8523688663282571, + "grad_norm": 8.1875, + "learning_rate": 3.6280049589424608e-06, + "loss": 2.5602, + "step": 17516 + }, + { + "epoch": 1.8524746192893402, + "grad_norm": 8.25, + "learning_rate": 3.6278521234996513e-06, + "loss": 2.6222, + "step": 17517 + }, + { + "epoch": 1.852580372250423, + "grad_norm": 8.6875, + "learning_rate": 3.627699282764293e-06, + "loss": 2.2887, + "step": 17518 + }, + { + "epoch": 1.8526861252115059, + "grad_norm": 7.75, + "learning_rate": 3.6275464367371015e-06, + "loss": 2.5141, + "step": 17519 + }, + { + "epoch": 1.8527918781725887, + "grad_norm": 8.4375, + "learning_rate": 3.6273935854187973e-06, + "loss": 2.6631, + "step": 17520 + }, + { + "epoch": 1.8528976311336718, + "grad_norm": 8.25, + "learning_rate": 3.6272407288100965e-06, + "loss": 2.4836, + "step": 17521 + }, + { + "epoch": 1.8530033840947546, + "grad_norm": 8.8125, + "learning_rate": 3.6270878669117143e-06, + "loss": 2.5372, + "step": 17522 + }, + { + "epoch": 1.8531091370558377, + "grad_norm": 8.25, + "learning_rate": 3.6269349997243703e-06, + "loss": 2.4482, + "step": 17523 + }, + { + "epoch": 1.8532148900169205, + "grad_norm": 8.4375, + "learning_rate": 3.626782127248782e-06, + "loss": 2.4152, + "step": 17524 + }, + { + "epoch": 1.8533206429780034, + "grad_norm": 8.1875, + "learning_rate": 3.6266292494856657e-06, + "loss": 2.6944, + "step": 17525 + }, + { + "epoch": 1.8534263959390862, + "grad_norm": 8.125, + "learning_rate": 3.6264763664357383e-06, + "loss": 2.5592, + "step": 17526 + }, + { + "epoch": 1.853532148900169, + "grad_norm": 7.78125, + "learning_rate": 3.6263234780997192e-06, + "loss": 2.563, + "step": 17527 + }, + { + "epoch": 1.853637901861252, + "grad_norm": 8.8125, + "learning_rate": 3.626170584478324e-06, + "loss": 2.4561, + "step": 17528 + }, + { + "epoch": 1.8537436548223352, + "grad_norm": 8.0625, + "learning_rate": 3.6260176855722705e-06, + "loss": 2.3136, + "step": 17529 + }, + { + "epoch": 1.853849407783418, + "grad_norm": 8.6875, + "learning_rate": 3.6258647813822777e-06, + "loss": 2.7512, + "step": 17530 + }, + { + "epoch": 1.8539551607445008, + "grad_norm": 8.1875, + "learning_rate": 3.625711871909061e-06, + "loss": 2.4293, + "step": 17531 + }, + { + "epoch": 1.8540609137055837, + "grad_norm": 8.3125, + "learning_rate": 3.625558957153339e-06, + "loss": 2.4578, + "step": 17532 + }, + { + "epoch": 1.8541666666666665, + "grad_norm": 8.5625, + "learning_rate": 3.625406037115829e-06, + "loss": 2.4632, + "step": 17533 + }, + { + "epoch": 1.8542724196277496, + "grad_norm": 8.0625, + "learning_rate": 3.6252531117972495e-06, + "loss": 2.5874, + "step": 17534 + }, + { + "epoch": 1.8543781725888326, + "grad_norm": 8.1875, + "learning_rate": 3.6251001811983166e-06, + "loss": 2.5689, + "step": 17535 + }, + { + "epoch": 1.8544839255499155, + "grad_norm": 8.125, + "learning_rate": 3.6249472453197486e-06, + "loss": 2.4948, + "step": 17536 + }, + { + "epoch": 1.8545896785109983, + "grad_norm": 8.375, + "learning_rate": 3.624794304162264e-06, + "loss": 2.2177, + "step": 17537 + }, + { + "epoch": 1.8546954314720812, + "grad_norm": 8.1875, + "learning_rate": 3.6246413577265793e-06, + "loss": 2.5429, + "step": 17538 + }, + { + "epoch": 1.854801184433164, + "grad_norm": 8.5, + "learning_rate": 3.624488406013412e-06, + "loss": 2.759, + "step": 17539 + }, + { + "epoch": 1.854906937394247, + "grad_norm": 8.375, + "learning_rate": 3.624335449023482e-06, + "loss": 2.6329, + "step": 17540 + }, + { + "epoch": 1.85501269035533, + "grad_norm": 8.0625, + "learning_rate": 3.6241824867575047e-06, + "loss": 2.2128, + "step": 17541 + }, + { + "epoch": 1.855118443316413, + "grad_norm": 8.5625, + "learning_rate": 3.624029519216199e-06, + "loss": 2.6132, + "step": 17542 + }, + { + "epoch": 1.8552241962774958, + "grad_norm": 8.25, + "learning_rate": 3.6238765464002823e-06, + "loss": 2.4537, + "step": 17543 + }, + { + "epoch": 1.8553299492385786, + "grad_norm": 8.5625, + "learning_rate": 3.6237235683104732e-06, + "loss": 2.6642, + "step": 17544 + }, + { + "epoch": 1.8554357021996615, + "grad_norm": 8.9375, + "learning_rate": 3.6235705849474884e-06, + "loss": 2.6359, + "step": 17545 + }, + { + "epoch": 1.8555414551607445, + "grad_norm": 7.8125, + "learning_rate": 3.6234175963120467e-06, + "loss": 2.5285, + "step": 17546 + }, + { + "epoch": 1.8556472081218274, + "grad_norm": 8.125, + "learning_rate": 3.6232646024048655e-06, + "loss": 2.4621, + "step": 17547 + }, + { + "epoch": 1.8557529610829104, + "grad_norm": 8.5, + "learning_rate": 3.623111603226662e-06, + "loss": 2.5354, + "step": 17548 + }, + { + "epoch": 1.8558587140439933, + "grad_norm": 8.0, + "learning_rate": 3.6229585987781563e-06, + "loss": 2.5423, + "step": 17549 + }, + { + "epoch": 1.8559644670050761, + "grad_norm": 8.25, + "learning_rate": 3.622805589060065e-06, + "loss": 2.4752, + "step": 17550 + }, + { + "epoch": 1.856070219966159, + "grad_norm": 7.9375, + "learning_rate": 3.6226525740731055e-06, + "loss": 2.5114, + "step": 17551 + }, + { + "epoch": 1.856175972927242, + "grad_norm": 8.375, + "learning_rate": 3.6224995538179975e-06, + "loss": 2.6277, + "step": 17552 + }, + { + "epoch": 1.8562817258883249, + "grad_norm": 8.1875, + "learning_rate": 3.6223465282954585e-06, + "loss": 2.6221, + "step": 17553 + }, + { + "epoch": 1.856387478849408, + "grad_norm": 8.5625, + "learning_rate": 3.622193497506205e-06, + "loss": 2.6937, + "step": 17554 + }, + { + "epoch": 1.8564932318104908, + "grad_norm": 8.9375, + "learning_rate": 3.6220404614509573e-06, + "loss": 2.596, + "step": 17555 + }, + { + "epoch": 1.8565989847715736, + "grad_norm": 8.8125, + "learning_rate": 3.6218874201304326e-06, + "loss": 2.504, + "step": 17556 + }, + { + "epoch": 1.8567047377326564, + "grad_norm": 8.1875, + "learning_rate": 3.6217343735453487e-06, + "loss": 2.5307, + "step": 17557 + }, + { + "epoch": 1.8568104906937393, + "grad_norm": 8.4375, + "learning_rate": 3.6215813216964242e-06, + "loss": 2.6211, + "step": 17558 + }, + { + "epoch": 1.8569162436548223, + "grad_norm": 8.0, + "learning_rate": 3.6214282645843776e-06, + "loss": 2.528, + "step": 17559 + }, + { + "epoch": 1.8570219966159054, + "grad_norm": 8.9375, + "learning_rate": 3.621275202209926e-06, + "loss": 2.6026, + "step": 17560 + }, + { + "epoch": 1.8571277495769882, + "grad_norm": 8.25, + "learning_rate": 3.621122134573789e-06, + "loss": 2.5221, + "step": 17561 + }, + { + "epoch": 1.857233502538071, + "grad_norm": 8.25, + "learning_rate": 3.6209690616766845e-06, + "loss": 2.5852, + "step": 17562 + }, + { + "epoch": 1.857339255499154, + "grad_norm": 8.1875, + "learning_rate": 3.6208159835193313e-06, + "loss": 2.4888, + "step": 17563 + }, + { + "epoch": 1.8574450084602367, + "grad_norm": 8.4375, + "learning_rate": 3.6206629001024457e-06, + "loss": 2.6561, + "step": 17564 + }, + { + "epoch": 1.8575507614213198, + "grad_norm": 8.5, + "learning_rate": 3.6205098114267483e-06, + "loss": 2.5444, + "step": 17565 + }, + { + "epoch": 1.8576565143824029, + "grad_norm": 8.375, + "learning_rate": 3.620356717492956e-06, + "loss": 2.4607, + "step": 17566 + }, + { + "epoch": 1.8577622673434857, + "grad_norm": 8.625, + "learning_rate": 3.620203618301788e-06, + "loss": 2.5423, + "step": 17567 + }, + { + "epoch": 1.8578680203045685, + "grad_norm": 8.4375, + "learning_rate": 3.620050513853963e-06, + "loss": 2.4933, + "step": 17568 + }, + { + "epoch": 1.8579737732656514, + "grad_norm": 8.4375, + "learning_rate": 3.619897404150199e-06, + "loss": 2.4793, + "step": 17569 + }, + { + "epoch": 1.8580795262267342, + "grad_norm": 8.5625, + "learning_rate": 3.6197442891912138e-06, + "loss": 2.6625, + "step": 17570 + }, + { + "epoch": 1.8581852791878173, + "grad_norm": 7.90625, + "learning_rate": 3.6195911689777274e-06, + "loss": 2.3758, + "step": 17571 + }, + { + "epoch": 1.8582910321489001, + "grad_norm": 8.0625, + "learning_rate": 3.6194380435104577e-06, + "loss": 2.3593, + "step": 17572 + }, + { + "epoch": 1.8583967851099832, + "grad_norm": 7.9375, + "learning_rate": 3.6192849127901226e-06, + "loss": 2.5077, + "step": 17573 + }, + { + "epoch": 1.858502538071066, + "grad_norm": 8.5, + "learning_rate": 3.6191317768174413e-06, + "loss": 2.3663, + "step": 17574 + }, + { + "epoch": 1.8586082910321489, + "grad_norm": 8.25, + "learning_rate": 3.6189786355931315e-06, + "loss": 2.6733, + "step": 17575 + }, + { + "epoch": 1.8587140439932317, + "grad_norm": 8.75, + "learning_rate": 3.6188254891179138e-06, + "loss": 2.8043, + "step": 17576 + }, + { + "epoch": 1.8588197969543148, + "grad_norm": 8.625, + "learning_rate": 3.6186723373925046e-06, + "loss": 2.783, + "step": 17577 + }, + { + "epoch": 1.8589255499153976, + "grad_norm": 8.1875, + "learning_rate": 3.6185191804176244e-06, + "loss": 2.3625, + "step": 17578 + }, + { + "epoch": 1.8590313028764807, + "grad_norm": 7.875, + "learning_rate": 3.618366018193992e-06, + "loss": 2.5938, + "step": 17579 + }, + { + "epoch": 1.8591370558375635, + "grad_norm": 8.8125, + "learning_rate": 3.6182128507223236e-06, + "loss": 2.5105, + "step": 17580 + }, + { + "epoch": 1.8592428087986463, + "grad_norm": 8.625, + "learning_rate": 3.6180596780033405e-06, + "loss": 2.4844, + "step": 17581 + }, + { + "epoch": 1.8593485617597292, + "grad_norm": 7.9375, + "learning_rate": 3.6179065000377607e-06, + "loss": 2.3497, + "step": 17582 + }, + { + "epoch": 1.859454314720812, + "grad_norm": 8.375, + "learning_rate": 3.617753316826302e-06, + "loss": 2.493, + "step": 17583 + }, + { + "epoch": 1.859560067681895, + "grad_norm": 8.4375, + "learning_rate": 3.617600128369685e-06, + "loss": 2.6849, + "step": 17584 + }, + { + "epoch": 1.8596658206429781, + "grad_norm": 8.5, + "learning_rate": 3.617446934668627e-06, + "loss": 2.2238, + "step": 17585 + }, + { + "epoch": 1.859771573604061, + "grad_norm": 8.3125, + "learning_rate": 3.617293735723848e-06, + "loss": 2.55, + "step": 17586 + }, + { + "epoch": 1.8598773265651438, + "grad_norm": 8.0625, + "learning_rate": 3.617140531536066e-06, + "loss": 2.4668, + "step": 17587 + }, + { + "epoch": 1.8599830795262267, + "grad_norm": 9.5625, + "learning_rate": 3.6169873221060013e-06, + "loss": 2.7728, + "step": 17588 + }, + { + "epoch": 1.8600888324873095, + "grad_norm": 8.6875, + "learning_rate": 3.6168341074343716e-06, + "loss": 2.623, + "step": 17589 + }, + { + "epoch": 1.8601945854483926, + "grad_norm": 8.0, + "learning_rate": 3.6166808875218962e-06, + "loss": 2.674, + "step": 17590 + }, + { + "epoch": 1.8603003384094756, + "grad_norm": 8.25, + "learning_rate": 3.6165276623692934e-06, + "loss": 2.5037, + "step": 17591 + }, + { + "epoch": 1.8604060913705585, + "grad_norm": 8.25, + "learning_rate": 3.6163744319772836e-06, + "loss": 2.568, + "step": 17592 + }, + { + "epoch": 1.8605118443316413, + "grad_norm": 8.5625, + "learning_rate": 3.6162211963465847e-06, + "loss": 2.6678, + "step": 17593 + }, + { + "epoch": 1.8606175972927241, + "grad_norm": 8.4375, + "learning_rate": 3.6160679554779166e-06, + "loss": 2.61, + "step": 17594 + }, + { + "epoch": 1.860723350253807, + "grad_norm": 8.875, + "learning_rate": 3.615914709371998e-06, + "loss": 2.5445, + "step": 17595 + }, + { + "epoch": 1.86082910321489, + "grad_norm": 8.5625, + "learning_rate": 3.6157614580295482e-06, + "loss": 2.5494, + "step": 17596 + }, + { + "epoch": 1.8609348561759729, + "grad_norm": 8.1875, + "learning_rate": 3.615608201451286e-06, + "loss": 2.5136, + "step": 17597 + }, + { + "epoch": 1.861040609137056, + "grad_norm": 9.8125, + "learning_rate": 3.615454939637931e-06, + "loss": 2.5424, + "step": 17598 + }, + { + "epoch": 1.8611463620981388, + "grad_norm": 8.6875, + "learning_rate": 3.6153016725902016e-06, + "loss": 2.4748, + "step": 17599 + }, + { + "epoch": 1.8612521150592216, + "grad_norm": 8.3125, + "learning_rate": 3.615148400308819e-06, + "loss": 2.7896, + "step": 17600 + }, + { + "epoch": 1.8613578680203045, + "grad_norm": 8.25, + "learning_rate": 3.6149951227944995e-06, + "loss": 2.6546, + "step": 17601 + }, + { + "epoch": 1.8614636209813875, + "grad_norm": 8.0625, + "learning_rate": 3.614841840047964e-06, + "loss": 2.423, + "step": 17602 + }, + { + "epoch": 1.8615693739424704, + "grad_norm": 8.4375, + "learning_rate": 3.6146885520699326e-06, + "loss": 2.4008, + "step": 17603 + }, + { + "epoch": 1.8616751269035534, + "grad_norm": 7.875, + "learning_rate": 3.6145352588611234e-06, + "loss": 2.4372, + "step": 17604 + }, + { + "epoch": 1.8617808798646363, + "grad_norm": 8.9375, + "learning_rate": 3.614381960422256e-06, + "loss": 2.4272, + "step": 17605 + }, + { + "epoch": 1.861886632825719, + "grad_norm": 8.5, + "learning_rate": 3.6142286567540497e-06, + "loss": 2.5654, + "step": 17606 + }, + { + "epoch": 1.861992385786802, + "grad_norm": 8.4375, + "learning_rate": 3.6140753478572243e-06, + "loss": 2.6326, + "step": 17607 + }, + { + "epoch": 1.862098138747885, + "grad_norm": 8.5, + "learning_rate": 3.613922033732498e-06, + "loss": 2.6274, + "step": 17608 + }, + { + "epoch": 1.8622038917089678, + "grad_norm": 8.5, + "learning_rate": 3.613768714380592e-06, + "loss": 2.551, + "step": 17609 + }, + { + "epoch": 1.862309644670051, + "grad_norm": 8.3125, + "learning_rate": 3.613615389802225e-06, + "loss": 2.5661, + "step": 17610 + }, + { + "epoch": 1.8624153976311337, + "grad_norm": 7.875, + "learning_rate": 3.6134620599981164e-06, + "loss": 2.32, + "step": 17611 + }, + { + "epoch": 1.8625211505922166, + "grad_norm": 7.96875, + "learning_rate": 3.613308724968985e-06, + "loss": 2.4578, + "step": 17612 + }, + { + "epoch": 1.8626269035532994, + "grad_norm": 8.3125, + "learning_rate": 3.613155384715552e-06, + "loss": 2.5877, + "step": 17613 + }, + { + "epoch": 1.8627326565143822, + "grad_norm": 8.125, + "learning_rate": 3.613002039238536e-06, + "loss": 2.3467, + "step": 17614 + }, + { + "epoch": 1.8628384094754653, + "grad_norm": 8.5, + "learning_rate": 3.612848688538656e-06, + "loss": 2.4744, + "step": 17615 + }, + { + "epoch": 1.8629441624365484, + "grad_norm": 8.0, + "learning_rate": 3.612695332616633e-06, + "loss": 2.6383, + "step": 17616 + }, + { + "epoch": 1.8630499153976312, + "grad_norm": 8.125, + "learning_rate": 3.6125419714731857e-06, + "loss": 2.6602, + "step": 17617 + }, + { + "epoch": 1.863155668358714, + "grad_norm": 8.5625, + "learning_rate": 3.6123886051090335e-06, + "loss": 2.3781, + "step": 17618 + }, + { + "epoch": 1.8632614213197969, + "grad_norm": 8.4375, + "learning_rate": 3.6122352335248967e-06, + "loss": 2.5759, + "step": 17619 + }, + { + "epoch": 1.8633671742808797, + "grad_norm": 8.375, + "learning_rate": 3.612081856721495e-06, + "loss": 2.6623, + "step": 17620 + }, + { + "epoch": 1.8634729272419628, + "grad_norm": 8.4375, + "learning_rate": 3.6119284746995474e-06, + "loss": 2.6827, + "step": 17621 + }, + { + "epoch": 1.8635786802030458, + "grad_norm": 8.25, + "learning_rate": 3.6117750874597744e-06, + "loss": 2.7533, + "step": 17622 + }, + { + "epoch": 1.8636844331641287, + "grad_norm": 8.25, + "learning_rate": 3.6116216950028963e-06, + "loss": 2.5881, + "step": 17623 + }, + { + "epoch": 1.8637901861252115, + "grad_norm": 7.9375, + "learning_rate": 3.6114682973296315e-06, + "loss": 2.4232, + "step": 17624 + }, + { + "epoch": 1.8638959390862944, + "grad_norm": 7.6875, + "learning_rate": 3.6113148944407016e-06, + "loss": 2.4039, + "step": 17625 + }, + { + "epoch": 1.8640016920473772, + "grad_norm": 8.3125, + "learning_rate": 3.611161486336825e-06, + "loss": 2.5839, + "step": 17626 + }, + { + "epoch": 1.8641074450084603, + "grad_norm": 8.25, + "learning_rate": 3.6110080730187213e-06, + "loss": 2.1643, + "step": 17627 + }, + { + "epoch": 1.864213197969543, + "grad_norm": 8.0625, + "learning_rate": 3.610854654487111e-06, + "loss": 2.4507, + "step": 17628 + }, + { + "epoch": 1.8643189509306262, + "grad_norm": 8.0625, + "learning_rate": 3.610701230742715e-06, + "loss": 2.4265, + "step": 17629 + }, + { + "epoch": 1.864424703891709, + "grad_norm": 8.375, + "learning_rate": 3.610547801786252e-06, + "loss": 2.4371, + "step": 17630 + }, + { + "epoch": 1.8645304568527918, + "grad_norm": 8.0625, + "learning_rate": 3.6103943676184422e-06, + "loss": 2.4799, + "step": 17631 + }, + { + "epoch": 1.8646362098138747, + "grad_norm": 8.25, + "learning_rate": 3.6102409282400064e-06, + "loss": 2.512, + "step": 17632 + }, + { + "epoch": 1.8647419627749577, + "grad_norm": 8.1875, + "learning_rate": 3.6100874836516635e-06, + "loss": 2.5328, + "step": 17633 + }, + { + "epoch": 1.8648477157360406, + "grad_norm": 7.9375, + "learning_rate": 3.6099340338541337e-06, + "loss": 2.4092, + "step": 17634 + }, + { + "epoch": 1.8649534686971236, + "grad_norm": 8.3125, + "learning_rate": 3.609780578848138e-06, + "loss": 2.3826, + "step": 17635 + }, + { + "epoch": 1.8650592216582065, + "grad_norm": 8.25, + "learning_rate": 3.6096271186343964e-06, + "loss": 2.5255, + "step": 17636 + }, + { + "epoch": 1.8651649746192893, + "grad_norm": 8.625, + "learning_rate": 3.6094736532136263e-06, + "loss": 2.4887, + "step": 17637 + }, + { + "epoch": 1.8652707275803722, + "grad_norm": 8.3125, + "learning_rate": 3.6093201825865526e-06, + "loss": 2.5893, + "step": 17638 + }, + { + "epoch": 1.865376480541455, + "grad_norm": 9.125, + "learning_rate": 3.609166706753892e-06, + "loss": 2.7835, + "step": 17639 + }, + { + "epoch": 1.865482233502538, + "grad_norm": 8.6875, + "learning_rate": 3.6090132257163655e-06, + "loss": 2.4006, + "step": 17640 + }, + { + "epoch": 1.8655879864636211, + "grad_norm": 8.0, + "learning_rate": 3.608859739474694e-06, + "loss": 2.5865, + "step": 17641 + }, + { + "epoch": 1.865693739424704, + "grad_norm": 7.96875, + "learning_rate": 3.6087062480295975e-06, + "loss": 2.4059, + "step": 17642 + }, + { + "epoch": 1.8657994923857868, + "grad_norm": 8.3125, + "learning_rate": 3.6085527513817953e-06, + "loss": 2.6288, + "step": 17643 + }, + { + "epoch": 1.8659052453468696, + "grad_norm": 8.0625, + "learning_rate": 3.6083992495320084e-06, + "loss": 2.3312, + "step": 17644 + }, + { + "epoch": 1.8660109983079525, + "grad_norm": 8.5625, + "learning_rate": 3.608245742480958e-06, + "loss": 2.4579, + "step": 17645 + }, + { + "epoch": 1.8661167512690355, + "grad_norm": 8.8125, + "learning_rate": 3.6080922302293623e-06, + "loss": 2.4796, + "step": 17646 + }, + { + "epoch": 1.8662225042301186, + "grad_norm": 8.625, + "learning_rate": 3.6079387127779435e-06, + "loss": 2.3753, + "step": 17647 + }, + { + "epoch": 1.8663282571912014, + "grad_norm": 8.25, + "learning_rate": 3.607785190127422e-06, + "loss": 2.5357, + "step": 17648 + }, + { + "epoch": 1.8664340101522843, + "grad_norm": 8.0625, + "learning_rate": 3.6076316622785174e-06, + "loss": 2.587, + "step": 17649 + }, + { + "epoch": 1.8665397631133671, + "grad_norm": 8.1875, + "learning_rate": 3.6074781292319498e-06, + "loss": 2.4574, + "step": 17650 + }, + { + "epoch": 1.86664551607445, + "grad_norm": 9.9375, + "learning_rate": 3.6073245909884415e-06, + "loss": 2.8057, + "step": 17651 + }, + { + "epoch": 1.866751269035533, + "grad_norm": 8.125, + "learning_rate": 3.607171047548711e-06, + "loss": 2.5459, + "step": 17652 + }, + { + "epoch": 1.8668570219966159, + "grad_norm": 8.6875, + "learning_rate": 3.6070174989134794e-06, + "loss": 2.7471, + "step": 17653 + }, + { + "epoch": 1.866962774957699, + "grad_norm": 8.3125, + "learning_rate": 3.6068639450834676e-06, + "loss": 2.5849, + "step": 17654 + }, + { + "epoch": 1.8670685279187818, + "grad_norm": 8.4375, + "learning_rate": 3.606710386059396e-06, + "loss": 2.4384, + "step": 17655 + }, + { + "epoch": 1.8671742808798646, + "grad_norm": 8.625, + "learning_rate": 3.606556821841985e-06, + "loss": 2.7169, + "step": 17656 + }, + { + "epoch": 1.8672800338409474, + "grad_norm": 8.125, + "learning_rate": 3.6064032524319557e-06, + "loss": 2.512, + "step": 17657 + }, + { + "epoch": 1.8673857868020305, + "grad_norm": 8.1875, + "learning_rate": 3.6062496778300287e-06, + "loss": 2.5688, + "step": 17658 + }, + { + "epoch": 1.8674915397631133, + "grad_norm": 8.5625, + "learning_rate": 3.6060960980369236e-06, + "loss": 2.7232, + "step": 17659 + }, + { + "epoch": 1.8675972927241964, + "grad_norm": 8.25, + "learning_rate": 3.6059425130533627e-06, + "loss": 2.3786, + "step": 17660 + }, + { + "epoch": 1.8677030456852792, + "grad_norm": 8.5625, + "learning_rate": 3.6057889228800656e-06, + "loss": 2.5073, + "step": 17661 + }, + { + "epoch": 1.867808798646362, + "grad_norm": 8.5, + "learning_rate": 3.6056353275177537e-06, + "loss": 2.6101, + "step": 17662 + }, + { + "epoch": 1.867914551607445, + "grad_norm": 8.875, + "learning_rate": 3.605481726967146e-06, + "loss": 2.5491, + "step": 17663 + }, + { + "epoch": 1.868020304568528, + "grad_norm": 8.6875, + "learning_rate": 3.605328121228966e-06, + "loss": 2.5474, + "step": 17664 + }, + { + "epoch": 1.8681260575296108, + "grad_norm": 8.3125, + "learning_rate": 3.605174510303933e-06, + "loss": 2.4178, + "step": 17665 + }, + { + "epoch": 1.8682318104906939, + "grad_norm": 8.6875, + "learning_rate": 3.6050208941927677e-06, + "loss": 2.7917, + "step": 17666 + }, + { + "epoch": 1.8683375634517767, + "grad_norm": 8.6875, + "learning_rate": 3.604867272896192e-06, + "loss": 2.6574, + "step": 17667 + }, + { + "epoch": 1.8684433164128595, + "grad_norm": 10.1875, + "learning_rate": 3.6047136464149256e-06, + "loss": 2.8722, + "step": 17668 + }, + { + "epoch": 1.8685490693739424, + "grad_norm": 8.5, + "learning_rate": 3.6045600147496896e-06, + "loss": 2.5178, + "step": 17669 + }, + { + "epoch": 1.8686548223350252, + "grad_norm": 8.5, + "learning_rate": 3.6044063779012057e-06, + "loss": 2.494, + "step": 17670 + }, + { + "epoch": 1.8687605752961083, + "grad_norm": 8.3125, + "learning_rate": 3.6042527358701943e-06, + "loss": 2.6488, + "step": 17671 + }, + { + "epoch": 1.8688663282571913, + "grad_norm": 7.875, + "learning_rate": 3.6040990886573755e-06, + "loss": 2.4734, + "step": 17672 + }, + { + "epoch": 1.8689720812182742, + "grad_norm": 8.1875, + "learning_rate": 3.6039454362634717e-06, + "loss": 2.5296, + "step": 17673 + }, + { + "epoch": 1.869077834179357, + "grad_norm": 8.4375, + "learning_rate": 3.603791778689204e-06, + "loss": 2.4776, + "step": 17674 + }, + { + "epoch": 1.8691835871404399, + "grad_norm": 8.8125, + "learning_rate": 3.603638115935293e-06, + "loss": 2.7605, + "step": 17675 + }, + { + "epoch": 1.8692893401015227, + "grad_norm": 7.8125, + "learning_rate": 3.6034844480024596e-06, + "loss": 2.4541, + "step": 17676 + }, + { + "epoch": 1.8693950930626058, + "grad_norm": 8.125, + "learning_rate": 3.603330774891425e-06, + "loss": 2.5681, + "step": 17677 + }, + { + "epoch": 1.8695008460236888, + "grad_norm": 8.625, + "learning_rate": 3.6031770966029096e-06, + "loss": 2.6116, + "step": 17678 + }, + { + "epoch": 1.8696065989847717, + "grad_norm": 9.125, + "learning_rate": 3.603023413137636e-06, + "loss": 2.6445, + "step": 17679 + }, + { + "epoch": 1.8697123519458545, + "grad_norm": 7.8125, + "learning_rate": 3.602869724496324e-06, + "loss": 2.5266, + "step": 17680 + }, + { + "epoch": 1.8698181049069373, + "grad_norm": 8.3125, + "learning_rate": 3.602716030679696e-06, + "loss": 2.6293, + "step": 17681 + }, + { + "epoch": 1.8699238578680202, + "grad_norm": 8.5625, + "learning_rate": 3.6025623316884716e-06, + "loss": 2.6443, + "step": 17682 + }, + { + "epoch": 1.8700296108291032, + "grad_norm": 8.5, + "learning_rate": 3.602408627523375e-06, + "loss": 2.6206, + "step": 17683 + }, + { + "epoch": 1.870135363790186, + "grad_norm": 8.125, + "learning_rate": 3.602254918185124e-06, + "loss": 2.4158, + "step": 17684 + }, + { + "epoch": 1.8702411167512691, + "grad_norm": 8.75, + "learning_rate": 3.602101203674442e-06, + "loss": 2.607, + "step": 17685 + }, + { + "epoch": 1.870346869712352, + "grad_norm": 8.375, + "learning_rate": 3.60194748399205e-06, + "loss": 2.6018, + "step": 17686 + }, + { + "epoch": 1.8704526226734348, + "grad_norm": 8.25, + "learning_rate": 3.6017937591386693e-06, + "loss": 2.6758, + "step": 17687 + }, + { + "epoch": 1.8705583756345177, + "grad_norm": 8.125, + "learning_rate": 3.6016400291150205e-06, + "loss": 2.5169, + "step": 17688 + }, + { + "epoch": 1.8706641285956007, + "grad_norm": 8.375, + "learning_rate": 3.6014862939218263e-06, + "loss": 2.5791, + "step": 17689 + }, + { + "epoch": 1.8707698815566836, + "grad_norm": 8.1875, + "learning_rate": 3.6013325535598066e-06, + "loss": 2.5243, + "step": 17690 + }, + { + "epoch": 1.8708756345177666, + "grad_norm": 7.90625, + "learning_rate": 3.601178808029684e-06, + "loss": 2.4919, + "step": 17691 + }, + { + "epoch": 1.8709813874788495, + "grad_norm": 8.8125, + "learning_rate": 3.6010250573321797e-06, + "loss": 2.659, + "step": 17692 + }, + { + "epoch": 1.8710871404399323, + "grad_norm": 8.5, + "learning_rate": 3.600871301468015e-06, + "loss": 2.1546, + "step": 17693 + }, + { + "epoch": 1.8711928934010151, + "grad_norm": 8.4375, + "learning_rate": 3.600717540437911e-06, + "loss": 2.6003, + "step": 17694 + }, + { + "epoch": 1.871298646362098, + "grad_norm": 8.3125, + "learning_rate": 3.6005637742425915e-06, + "loss": 2.347, + "step": 17695 + }, + { + "epoch": 1.871404399323181, + "grad_norm": 8.1875, + "learning_rate": 3.6004100028827747e-06, + "loss": 2.7905, + "step": 17696 + }, + { + "epoch": 1.871510152284264, + "grad_norm": 9.0625, + "learning_rate": 3.6002562263591846e-06, + "loss": 2.7922, + "step": 17697 + }, + { + "epoch": 1.871615905245347, + "grad_norm": 9.9375, + "learning_rate": 3.600102444672542e-06, + "loss": 2.3657, + "step": 17698 + }, + { + "epoch": 1.8717216582064298, + "grad_norm": 8.5, + "learning_rate": 3.5999486578235674e-06, + "loss": 2.6252, + "step": 17699 + }, + { + "epoch": 1.8718274111675126, + "grad_norm": 8.8125, + "learning_rate": 3.599794865812985e-06, + "loss": 2.626, + "step": 17700 + }, + { + "epoch": 1.8719331641285955, + "grad_norm": 8.0625, + "learning_rate": 3.599641068641514e-06, + "loss": 2.6133, + "step": 17701 + }, + { + "epoch": 1.8720389170896785, + "grad_norm": 8.1875, + "learning_rate": 3.599487266309878e-06, + "loss": 2.4592, + "step": 17702 + }, + { + "epoch": 1.8721446700507616, + "grad_norm": 8.3125, + "learning_rate": 3.599333458818798e-06, + "loss": 2.589, + "step": 17703 + }, + { + "epoch": 1.8722504230118444, + "grad_norm": 8.25, + "learning_rate": 3.599179646168995e-06, + "loss": 2.498, + "step": 17704 + }, + { + "epoch": 1.8723561759729273, + "grad_norm": 9.0625, + "learning_rate": 3.5990258283611925e-06, + "loss": 2.8472, + "step": 17705 + }, + { + "epoch": 1.87246192893401, + "grad_norm": 8.625, + "learning_rate": 3.5988720053961107e-06, + "loss": 2.6264, + "step": 17706 + }, + { + "epoch": 1.872567681895093, + "grad_norm": 7.71875, + "learning_rate": 3.598718177274471e-06, + "loss": 2.3697, + "step": 17707 + }, + { + "epoch": 1.872673434856176, + "grad_norm": 8.4375, + "learning_rate": 3.5985643439969977e-06, + "loss": 2.6486, + "step": 17708 + }, + { + "epoch": 1.8727791878172588, + "grad_norm": 8.9375, + "learning_rate": 3.5984105055644107e-06, + "loss": 2.8324, + "step": 17709 + }, + { + "epoch": 1.872884940778342, + "grad_norm": 8.0625, + "learning_rate": 3.598256661977432e-06, + "loss": 2.4739, + "step": 17710 + }, + { + "epoch": 1.8729906937394247, + "grad_norm": 8.375, + "learning_rate": 3.5981028132367845e-06, + "loss": 2.5168, + "step": 17711 + }, + { + "epoch": 1.8730964467005076, + "grad_norm": 7.875, + "learning_rate": 3.5979489593431904e-06, + "loss": 2.6028, + "step": 17712 + }, + { + "epoch": 1.8732021996615904, + "grad_norm": 8.375, + "learning_rate": 3.5977951002973692e-06, + "loss": 2.643, + "step": 17713 + }, + { + "epoch": 1.8733079526226735, + "grad_norm": 8.4375, + "learning_rate": 3.5976412361000458e-06, + "loss": 2.6501, + "step": 17714 + }, + { + "epoch": 1.8734137055837563, + "grad_norm": 8.4375, + "learning_rate": 3.5974873667519404e-06, + "loss": 2.6361, + "step": 17715 + }, + { + "epoch": 1.8735194585448394, + "grad_norm": 8.0, + "learning_rate": 3.597333492253776e-06, + "loss": 2.1257, + "step": 17716 + }, + { + "epoch": 1.8736252115059222, + "grad_norm": 8.25, + "learning_rate": 3.597179612606274e-06, + "loss": 2.5572, + "step": 17717 + }, + { + "epoch": 1.873730964467005, + "grad_norm": 9.3125, + "learning_rate": 3.5970257278101572e-06, + "loss": 2.6409, + "step": 17718 + }, + { + "epoch": 1.8738367174280879, + "grad_norm": 8.25, + "learning_rate": 3.596871837866147e-06, + "loss": 2.4766, + "step": 17719 + }, + { + "epoch": 1.873942470389171, + "grad_norm": 8.9375, + "learning_rate": 3.5967179427749656e-06, + "loss": 2.4987, + "step": 17720 + }, + { + "epoch": 1.8740482233502538, + "grad_norm": 7.96875, + "learning_rate": 3.596564042537336e-06, + "loss": 2.2582, + "step": 17721 + }, + { + "epoch": 1.8741539763113368, + "grad_norm": 8.0625, + "learning_rate": 3.59641013715398e-06, + "loss": 2.4927, + "step": 17722 + }, + { + "epoch": 1.8742597292724197, + "grad_norm": 8.0, + "learning_rate": 3.596256226625619e-06, + "loss": 2.4143, + "step": 17723 + }, + { + "epoch": 1.8743654822335025, + "grad_norm": 8.3125, + "learning_rate": 3.5961023109529763e-06, + "loss": 2.4477, + "step": 17724 + }, + { + "epoch": 1.8744712351945854, + "grad_norm": 8.25, + "learning_rate": 3.595948390136773e-06, + "loss": 2.5882, + "step": 17725 + }, + { + "epoch": 1.8745769881556682, + "grad_norm": 8.0, + "learning_rate": 3.5957944641777327e-06, + "loss": 2.4705, + "step": 17726 + }, + { + "epoch": 1.8746827411167513, + "grad_norm": 8.25, + "learning_rate": 3.5956405330765776e-06, + "loss": 2.6132, + "step": 17727 + }, + { + "epoch": 1.8747884940778343, + "grad_norm": 8.3125, + "learning_rate": 3.595486596834029e-06, + "loss": 2.5761, + "step": 17728 + }, + { + "epoch": 1.8748942470389172, + "grad_norm": 8.8125, + "learning_rate": 3.59533265545081e-06, + "loss": 2.5082, + "step": 17729 + }, + { + "epoch": 1.875, + "grad_norm": 8.25, + "learning_rate": 3.595178708927643e-06, + "loss": 2.5628, + "step": 17730 + }, + { + "epoch": 1.8751057529610828, + "grad_norm": 8.4375, + "learning_rate": 3.5950247572652503e-06, + "loss": 2.4459, + "step": 17731 + }, + { + "epoch": 1.8752115059221657, + "grad_norm": 7.96875, + "learning_rate": 3.5948708004643532e-06, + "loss": 2.6034, + "step": 17732 + }, + { + "epoch": 1.8753172588832487, + "grad_norm": 8.25, + "learning_rate": 3.5947168385256762e-06, + "loss": 2.5699, + "step": 17733 + }, + { + "epoch": 1.8754230118443318, + "grad_norm": 8.0, + "learning_rate": 3.5945628714499405e-06, + "loss": 2.6144, + "step": 17734 + }, + { + "epoch": 1.8755287648054146, + "grad_norm": 7.96875, + "learning_rate": 3.5944088992378693e-06, + "loss": 2.6046, + "step": 17735 + }, + { + "epoch": 1.8756345177664975, + "grad_norm": 8.375, + "learning_rate": 3.5942549218901845e-06, + "loss": 2.4964, + "step": 17736 + }, + { + "epoch": 1.8757402707275803, + "grad_norm": 8.8125, + "learning_rate": 3.594100939407609e-06, + "loss": 2.7224, + "step": 17737 + }, + { + "epoch": 1.8758460236886632, + "grad_norm": 8.25, + "learning_rate": 3.5939469517908655e-06, + "loss": 2.4184, + "step": 17738 + }, + { + "epoch": 1.8759517766497462, + "grad_norm": 8.125, + "learning_rate": 3.593792959040675e-06, + "loss": 2.6878, + "step": 17739 + }, + { + "epoch": 1.876057529610829, + "grad_norm": 8.3125, + "learning_rate": 3.593638961157763e-06, + "loss": 2.5797, + "step": 17740 + }, + { + "epoch": 1.8761632825719121, + "grad_norm": 8.5, + "learning_rate": 3.5934849581428507e-06, + "loss": 2.3778, + "step": 17741 + }, + { + "epoch": 1.876269035532995, + "grad_norm": 8.125, + "learning_rate": 3.5933309499966595e-06, + "loss": 2.5804, + "step": 17742 + }, + { + "epoch": 1.8763747884940778, + "grad_norm": 8.1875, + "learning_rate": 3.5931769367199136e-06, + "loss": 2.6198, + "step": 17743 + }, + { + "epoch": 1.8764805414551606, + "grad_norm": 8.3125, + "learning_rate": 3.5930229183133358e-06, + "loss": 2.6524, + "step": 17744 + }, + { + "epoch": 1.8765862944162437, + "grad_norm": 8.75, + "learning_rate": 3.592868894777648e-06, + "loss": 2.5696, + "step": 17745 + }, + { + "epoch": 1.8766920473773265, + "grad_norm": 8.3125, + "learning_rate": 3.5927148661135747e-06, + "loss": 2.3221, + "step": 17746 + }, + { + "epoch": 1.8767978003384096, + "grad_norm": 9.375, + "learning_rate": 3.5925608323218364e-06, + "loss": 2.6229, + "step": 17747 + }, + { + "epoch": 1.8769035532994924, + "grad_norm": 8.875, + "learning_rate": 3.592406793403157e-06, + "loss": 2.9087, + "step": 17748 + }, + { + "epoch": 1.8770093062605753, + "grad_norm": 8.4375, + "learning_rate": 3.5922527493582597e-06, + "loss": 2.6613, + "step": 17749 + }, + { + "epoch": 1.877115059221658, + "grad_norm": 9.0625, + "learning_rate": 3.592098700187867e-06, + "loss": 2.4949, + "step": 17750 + }, + { + "epoch": 1.8772208121827412, + "grad_norm": 8.875, + "learning_rate": 3.591944645892701e-06, + "loss": 2.6476, + "step": 17751 + }, + { + "epoch": 1.877326565143824, + "grad_norm": 8.875, + "learning_rate": 3.591790586473486e-06, + "loss": 2.7738, + "step": 17752 + }, + { + "epoch": 1.877432318104907, + "grad_norm": 9.75, + "learning_rate": 3.591636521930944e-06, + "loss": 2.5474, + "step": 17753 + }, + { + "epoch": 1.87753807106599, + "grad_norm": 8.3125, + "learning_rate": 3.5914824522657986e-06, + "loss": 2.6621, + "step": 17754 + }, + { + "epoch": 1.8776438240270727, + "grad_norm": 8.4375, + "learning_rate": 3.591328377478772e-06, + "loss": 2.4153, + "step": 17755 + }, + { + "epoch": 1.8777495769881556, + "grad_norm": 8.25, + "learning_rate": 3.5911742975705884e-06, + "loss": 2.721, + "step": 17756 + }, + { + "epoch": 1.8778553299492384, + "grad_norm": 8.1875, + "learning_rate": 3.5910202125419694e-06, + "loss": 2.7379, + "step": 17757 + }, + { + "epoch": 1.8779610829103215, + "grad_norm": 7.875, + "learning_rate": 3.590866122393639e-06, + "loss": 2.5313, + "step": 17758 + }, + { + "epoch": 1.8780668358714045, + "grad_norm": 8.5, + "learning_rate": 3.59071202712632e-06, + "loss": 2.3786, + "step": 17759 + }, + { + "epoch": 1.8781725888324874, + "grad_norm": 8.625, + "learning_rate": 3.590557926740736e-06, + "loss": 2.6307, + "step": 17760 + }, + { + "epoch": 1.8782783417935702, + "grad_norm": 8.3125, + "learning_rate": 3.5904038212376086e-06, + "loss": 2.4651, + "step": 17761 + }, + { + "epoch": 1.878384094754653, + "grad_norm": 8.0625, + "learning_rate": 3.5902497106176624e-06, + "loss": 2.5039, + "step": 17762 + }, + { + "epoch": 1.878489847715736, + "grad_norm": 8.1875, + "learning_rate": 3.5900955948816207e-06, + "loss": 2.5427, + "step": 17763 + }, + { + "epoch": 1.878595600676819, + "grad_norm": 8.5, + "learning_rate": 3.5899414740302056e-06, + "loss": 2.5155, + "step": 17764 + }, + { + "epoch": 1.878701353637902, + "grad_norm": 8.1875, + "learning_rate": 3.589787348064142e-06, + "loss": 2.3819, + "step": 17765 + }, + { + "epoch": 1.8788071065989849, + "grad_norm": 8.125, + "learning_rate": 3.589633216984151e-06, + "loss": 2.4489, + "step": 17766 + }, + { + "epoch": 1.8789128595600677, + "grad_norm": 8.1875, + "learning_rate": 3.5894790807909573e-06, + "loss": 2.5897, + "step": 17767 + }, + { + "epoch": 1.8790186125211505, + "grad_norm": 8.5625, + "learning_rate": 3.5893249394852843e-06, + "loss": 2.4696, + "step": 17768 + }, + { + "epoch": 1.8791243654822334, + "grad_norm": 8.4375, + "learning_rate": 3.5891707930678542e-06, + "loss": 2.7194, + "step": 17769 + }, + { + "epoch": 1.8792301184433164, + "grad_norm": 8.125, + "learning_rate": 3.5890166415393914e-06, + "loss": 2.7064, + "step": 17770 + }, + { + "epoch": 1.8793358714043993, + "grad_norm": 7.875, + "learning_rate": 3.5888624849006187e-06, + "loss": 2.4834, + "step": 17771 + }, + { + "epoch": 1.8794416243654823, + "grad_norm": 8.625, + "learning_rate": 3.58870832315226e-06, + "loss": 2.6747, + "step": 17772 + }, + { + "epoch": 1.8795473773265652, + "grad_norm": 8.0, + "learning_rate": 3.588554156295038e-06, + "loss": 2.5572, + "step": 17773 + }, + { + "epoch": 1.879653130287648, + "grad_norm": 9.125, + "learning_rate": 3.588399984329677e-06, + "loss": 2.43, + "step": 17774 + }, + { + "epoch": 1.8797588832487309, + "grad_norm": 7.875, + "learning_rate": 3.5882458072569e-06, + "loss": 2.4294, + "step": 17775 + }, + { + "epoch": 1.879864636209814, + "grad_norm": 8.25, + "learning_rate": 3.58809162507743e-06, + "loss": 2.5226, + "step": 17776 + }, + { + "epoch": 1.8799703891708968, + "grad_norm": 8.5, + "learning_rate": 3.587937437791991e-06, + "loss": 2.5431, + "step": 17777 + }, + { + "epoch": 1.8800761421319798, + "grad_norm": 8.1875, + "learning_rate": 3.5877832454013067e-06, + "loss": 2.5677, + "step": 17778 + }, + { + "epoch": 1.8801818950930627, + "grad_norm": 8.1875, + "learning_rate": 3.587629047906101e-06, + "loss": 2.466, + "step": 17779 + }, + { + "epoch": 1.8802876480541455, + "grad_norm": 8.4375, + "learning_rate": 3.587474845307096e-06, + "loss": 2.6406, + "step": 17780 + }, + { + "epoch": 1.8803934010152283, + "grad_norm": 8.25, + "learning_rate": 3.587320637605017e-06, + "loss": 2.6647, + "step": 17781 + }, + { + "epoch": 1.8804991539763112, + "grad_norm": 8.5, + "learning_rate": 3.587166424800588e-06, + "loss": 2.2855, + "step": 17782 + }, + { + "epoch": 1.8806049069373942, + "grad_norm": 8.4375, + "learning_rate": 3.58701220689453e-06, + "loss": 2.6796, + "step": 17783 + }, + { + "epoch": 1.8807106598984773, + "grad_norm": 8.0625, + "learning_rate": 3.586857983887569e-06, + "loss": 2.4921, + "step": 17784 + }, + { + "epoch": 1.8808164128595601, + "grad_norm": 8.0, + "learning_rate": 3.586703755780428e-06, + "loss": 2.3934, + "step": 17785 + }, + { + "epoch": 1.880922165820643, + "grad_norm": 7.90625, + "learning_rate": 3.58654952257383e-06, + "loss": 2.5796, + "step": 17786 + }, + { + "epoch": 1.8810279187817258, + "grad_norm": 8.125, + "learning_rate": 3.5863952842685e-06, + "loss": 2.3213, + "step": 17787 + }, + { + "epoch": 1.8811336717428087, + "grad_norm": 8.3125, + "learning_rate": 3.5862410408651614e-06, + "loss": 2.6394, + "step": 17788 + }, + { + "epoch": 1.8812394247038917, + "grad_norm": 8.125, + "learning_rate": 3.5860867923645382e-06, + "loss": 2.509, + "step": 17789 + }, + { + "epoch": 1.8813451776649748, + "grad_norm": 8.1875, + "learning_rate": 3.585932538767353e-06, + "loss": 2.2578, + "step": 17790 + }, + { + "epoch": 1.8814509306260576, + "grad_norm": 8.1875, + "learning_rate": 3.5857782800743306e-06, + "loss": 2.4628, + "step": 17791 + }, + { + "epoch": 1.8815566835871405, + "grad_norm": 8.8125, + "learning_rate": 3.5856240162861954e-06, + "loss": 2.5415, + "step": 17792 + }, + { + "epoch": 1.8816624365482233, + "grad_norm": 8.3125, + "learning_rate": 3.5854697474036704e-06, + "loss": 2.6361, + "step": 17793 + }, + { + "epoch": 1.8817681895093061, + "grad_norm": 8.75, + "learning_rate": 3.5853154734274797e-06, + "loss": 2.5352, + "step": 17794 + }, + { + "epoch": 1.8818739424703892, + "grad_norm": 8.375, + "learning_rate": 3.5851611943583475e-06, + "loss": 2.5168, + "step": 17795 + }, + { + "epoch": 1.881979695431472, + "grad_norm": 9.125, + "learning_rate": 3.585006910196997e-06, + "loss": 2.6097, + "step": 17796 + }, + { + "epoch": 1.882085448392555, + "grad_norm": 8.4375, + "learning_rate": 3.5848526209441536e-06, + "loss": 2.531, + "step": 17797 + }, + { + "epoch": 1.882191201353638, + "grad_norm": 8.75, + "learning_rate": 3.5846983266005407e-06, + "loss": 2.6718, + "step": 17798 + }, + { + "epoch": 1.8822969543147208, + "grad_norm": 8.375, + "learning_rate": 3.584544027166882e-06, + "loss": 2.2917, + "step": 17799 + }, + { + "epoch": 1.8824027072758036, + "grad_norm": 8.25, + "learning_rate": 3.584389722643902e-06, + "loss": 2.4656, + "step": 17800 + }, + { + "epoch": 1.8825084602368867, + "grad_norm": 9.75, + "learning_rate": 3.5842354130323236e-06, + "loss": 2.7872, + "step": 17801 + }, + { + "epoch": 1.8826142131979695, + "grad_norm": 8.875, + "learning_rate": 3.584081098332872e-06, + "loss": 2.4523, + "step": 17802 + }, + { + "epoch": 1.8827199661590526, + "grad_norm": 8.75, + "learning_rate": 3.5839267785462723e-06, + "loss": 2.7287, + "step": 17803 + }, + { + "epoch": 1.8828257191201354, + "grad_norm": 8.5, + "learning_rate": 3.5837724536732466e-06, + "loss": 2.4275, + "step": 17804 + }, + { + "epoch": 1.8829314720812182, + "grad_norm": 7.90625, + "learning_rate": 3.5836181237145202e-06, + "loss": 2.4287, + "step": 17805 + }, + { + "epoch": 1.883037225042301, + "grad_norm": 8.1875, + "learning_rate": 3.5834637886708166e-06, + "loss": 2.5025, + "step": 17806 + }, + { + "epoch": 1.8831429780033841, + "grad_norm": 8.0, + "learning_rate": 3.5833094485428618e-06, + "loss": 2.4766, + "step": 17807 + }, + { + "epoch": 1.883248730964467, + "grad_norm": 8.5, + "learning_rate": 3.5831551033313785e-06, + "loss": 2.4478, + "step": 17808 + }, + { + "epoch": 1.88335448392555, + "grad_norm": 8.1875, + "learning_rate": 3.5830007530370902e-06, + "loss": 2.6188, + "step": 17809 + }, + { + "epoch": 1.8834602368866329, + "grad_norm": 9.0625, + "learning_rate": 3.5828463976607233e-06, + "loss": 2.7036, + "step": 17810 + }, + { + "epoch": 1.8835659898477157, + "grad_norm": 8.5625, + "learning_rate": 3.5826920372030006e-06, + "loss": 2.6383, + "step": 17811 + }, + { + "epoch": 1.8836717428087986, + "grad_norm": 8.125, + "learning_rate": 3.5825376716646472e-06, + "loss": 2.328, + "step": 17812 + }, + { + "epoch": 1.8837774957698814, + "grad_norm": 8.4375, + "learning_rate": 3.582383301046388e-06, + "loss": 2.5846, + "step": 17813 + }, + { + "epoch": 1.8838832487309645, + "grad_norm": 8.5, + "learning_rate": 3.582228925348945e-06, + "loss": 2.4756, + "step": 17814 + }, + { + "epoch": 1.8839890016920475, + "grad_norm": 9.5625, + "learning_rate": 3.582074544573045e-06, + "loss": 2.506, + "step": 17815 + }, + { + "epoch": 1.8840947546531304, + "grad_norm": 8.25, + "learning_rate": 3.581920158719412e-06, + "loss": 2.6274, + "step": 17816 + }, + { + "epoch": 1.8842005076142132, + "grad_norm": 8.375, + "learning_rate": 3.58176576778877e-06, + "loss": 2.5511, + "step": 17817 + }, + { + "epoch": 1.884306260575296, + "grad_norm": 8.375, + "learning_rate": 3.581611371781844e-06, + "loss": 2.5114, + "step": 17818 + }, + { + "epoch": 1.8844120135363789, + "grad_norm": 8.375, + "learning_rate": 3.5814569706993577e-06, + "loss": 2.5186, + "step": 17819 + }, + { + "epoch": 1.884517766497462, + "grad_norm": 8.3125, + "learning_rate": 3.581302564542037e-06, + "loss": 2.419, + "step": 17820 + }, + { + "epoch": 1.884623519458545, + "grad_norm": 8.5625, + "learning_rate": 3.5811481533106044e-06, + "loss": 2.3132, + "step": 17821 + }, + { + "epoch": 1.8847292724196278, + "grad_norm": 8.5, + "learning_rate": 3.5809937370057862e-06, + "loss": 2.6607, + "step": 17822 + }, + { + "epoch": 1.8848350253807107, + "grad_norm": 7.90625, + "learning_rate": 3.5808393156283067e-06, + "loss": 2.4395, + "step": 17823 + }, + { + "epoch": 1.8849407783417935, + "grad_norm": 8.6875, + "learning_rate": 3.5806848891788904e-06, + "loss": 2.5103, + "step": 17824 + }, + { + "epoch": 1.8850465313028764, + "grad_norm": 8.25, + "learning_rate": 3.5805304576582616e-06, + "loss": 2.4566, + "step": 17825 + }, + { + "epoch": 1.8851522842639594, + "grad_norm": 7.9375, + "learning_rate": 3.5803760210671458e-06, + "loss": 2.5717, + "step": 17826 + }, + { + "epoch": 1.8852580372250423, + "grad_norm": 7.9375, + "learning_rate": 3.5802215794062667e-06, + "loss": 2.374, + "step": 17827 + }, + { + "epoch": 1.8853637901861253, + "grad_norm": 8.375, + "learning_rate": 3.580067132676349e-06, + "loss": 2.5108, + "step": 17828 + }, + { + "epoch": 1.8854695431472082, + "grad_norm": 8.5625, + "learning_rate": 3.579912680878119e-06, + "loss": 2.4456, + "step": 17829 + }, + { + "epoch": 1.885575296108291, + "grad_norm": 9.125, + "learning_rate": 3.5797582240123007e-06, + "loss": 2.591, + "step": 17830 + }, + { + "epoch": 1.8856810490693738, + "grad_norm": 8.5, + "learning_rate": 3.579603762079617e-06, + "loss": 2.3529, + "step": 17831 + }, + { + "epoch": 1.885786802030457, + "grad_norm": 8.25, + "learning_rate": 3.5794492950807957e-06, + "loss": 2.4383, + "step": 17832 + }, + { + "epoch": 1.8858925549915397, + "grad_norm": 8.5625, + "learning_rate": 3.5792948230165605e-06, + "loss": 2.5486, + "step": 17833 + }, + { + "epoch": 1.8859983079526228, + "grad_norm": 8.1875, + "learning_rate": 3.579140345887635e-06, + "loss": 2.7029, + "step": 17834 + }, + { + "epoch": 1.8861040609137056, + "grad_norm": 8.25, + "learning_rate": 3.578985863694746e-06, + "loss": 2.5214, + "step": 17835 + }, + { + "epoch": 1.8862098138747885, + "grad_norm": 8.9375, + "learning_rate": 3.578831376438618e-06, + "loss": 2.8572, + "step": 17836 + }, + { + "epoch": 1.8863155668358713, + "grad_norm": 9.0625, + "learning_rate": 3.5786768841199747e-06, + "loss": 2.7678, + "step": 17837 + }, + { + "epoch": 1.8864213197969542, + "grad_norm": 8.25, + "learning_rate": 3.5785223867395426e-06, + "loss": 2.679, + "step": 17838 + }, + { + "epoch": 1.8865270727580372, + "grad_norm": 8.75, + "learning_rate": 3.578367884298046e-06, + "loss": 2.7295, + "step": 17839 + }, + { + "epoch": 1.8866328257191203, + "grad_norm": 9.0625, + "learning_rate": 3.5782133767962094e-06, + "loss": 2.505, + "step": 17840 + }, + { + "epoch": 1.8867385786802031, + "grad_norm": 7.8125, + "learning_rate": 3.5780588642347585e-06, + "loss": 2.4728, + "step": 17841 + }, + { + "epoch": 1.886844331641286, + "grad_norm": 8.6875, + "learning_rate": 3.5779043466144186e-06, + "loss": 2.7402, + "step": 17842 + }, + { + "epoch": 1.8869500846023688, + "grad_norm": 8.4375, + "learning_rate": 3.5777498239359143e-06, + "loss": 2.3998, + "step": 17843 + }, + { + "epoch": 1.8870558375634516, + "grad_norm": 7.875, + "learning_rate": 3.577595296199971e-06, + "loss": 2.5624, + "step": 17844 + }, + { + "epoch": 1.8871615905245347, + "grad_norm": 8.9375, + "learning_rate": 3.577440763407314e-06, + "loss": 2.6327, + "step": 17845 + }, + { + "epoch": 1.8872673434856178, + "grad_norm": 8.25, + "learning_rate": 3.5772862255586677e-06, + "loss": 2.4956, + "step": 17846 + }, + { + "epoch": 1.8873730964467006, + "grad_norm": 8.6875, + "learning_rate": 3.5771316826547577e-06, + "loss": 2.6207, + "step": 17847 + }, + { + "epoch": 1.8874788494077834, + "grad_norm": 8.0625, + "learning_rate": 3.5769771346963094e-06, + "loss": 2.4366, + "step": 17848 + }, + { + "epoch": 1.8875846023688663, + "grad_norm": 8.25, + "learning_rate": 3.5768225816840476e-06, + "loss": 2.6333, + "step": 17849 + }, + { + "epoch": 1.887690355329949, + "grad_norm": 10.0, + "learning_rate": 3.576668023618698e-06, + "loss": 2.572, + "step": 17850 + }, + { + "epoch": 1.8877961082910322, + "grad_norm": 7.75, + "learning_rate": 3.5765134605009854e-06, + "loss": 2.4879, + "step": 17851 + }, + { + "epoch": 1.887901861252115, + "grad_norm": 8.5, + "learning_rate": 3.576358892331636e-06, + "loss": 2.4145, + "step": 17852 + }, + { + "epoch": 1.888007614213198, + "grad_norm": 7.90625, + "learning_rate": 3.5762043191113737e-06, + "loss": 2.5269, + "step": 17853 + }, + { + "epoch": 1.888113367174281, + "grad_norm": 8.25, + "learning_rate": 3.5760497408409256e-06, + "loss": 2.6068, + "step": 17854 + }, + { + "epoch": 1.8882191201353637, + "grad_norm": 8.6875, + "learning_rate": 3.5758951575210164e-06, + "loss": 2.71, + "step": 17855 + }, + { + "epoch": 1.8883248730964466, + "grad_norm": 8.125, + "learning_rate": 3.5757405691523707e-06, + "loss": 2.408, + "step": 17856 + }, + { + "epoch": 1.8884306260575296, + "grad_norm": 8.0625, + "learning_rate": 3.5755859757357144e-06, + "loss": 2.6376, + "step": 17857 + }, + { + "epoch": 1.8885363790186125, + "grad_norm": 8.0, + "learning_rate": 3.5754313772717725e-06, + "loss": 2.6169, + "step": 17858 + }, + { + "epoch": 1.8886421319796955, + "grad_norm": 8.25, + "learning_rate": 3.575276773761272e-06, + "loss": 2.3613, + "step": 17859 + }, + { + "epoch": 1.8887478849407784, + "grad_norm": 9.125, + "learning_rate": 3.5751221652049364e-06, + "loss": 2.6242, + "step": 17860 + }, + { + "epoch": 1.8888536379018612, + "grad_norm": 8.3125, + "learning_rate": 3.574967551603493e-06, + "loss": 2.5746, + "step": 17861 + }, + { + "epoch": 1.888959390862944, + "grad_norm": 8.125, + "learning_rate": 3.5748129329576664e-06, + "loss": 2.6705, + "step": 17862 + }, + { + "epoch": 1.8890651438240271, + "grad_norm": 7.84375, + "learning_rate": 3.5746583092681818e-06, + "loss": 2.509, + "step": 17863 + }, + { + "epoch": 1.88917089678511, + "grad_norm": 8.1875, + "learning_rate": 3.5745036805357658e-06, + "loss": 2.6761, + "step": 17864 + }, + { + "epoch": 1.889276649746193, + "grad_norm": 8.0625, + "learning_rate": 3.5743490467611434e-06, + "loss": 2.6187, + "step": 17865 + }, + { + "epoch": 1.8893824027072759, + "grad_norm": 8.375, + "learning_rate": 3.5741944079450403e-06, + "loss": 2.5853, + "step": 17866 + }, + { + "epoch": 1.8894881556683587, + "grad_norm": 8.0625, + "learning_rate": 3.574039764088182e-06, + "loss": 2.5798, + "step": 17867 + }, + { + "epoch": 1.8895939086294415, + "grad_norm": 8.6875, + "learning_rate": 3.5738851151912944e-06, + "loss": 2.703, + "step": 17868 + }, + { + "epoch": 1.8896996615905244, + "grad_norm": 8.5625, + "learning_rate": 3.5737304612551027e-06, + "loss": 2.5106, + "step": 17869 + }, + { + "epoch": 1.8898054145516074, + "grad_norm": 9.25, + "learning_rate": 3.5735758022803337e-06, + "loss": 2.5713, + "step": 17870 + }, + { + "epoch": 1.8899111675126905, + "grad_norm": 7.9375, + "learning_rate": 3.573421138267713e-06, + "loss": 2.3572, + "step": 17871 + }, + { + "epoch": 1.8900169204737733, + "grad_norm": 8.625, + "learning_rate": 3.573266469217965e-06, + "loss": 2.4347, + "step": 17872 + }, + { + "epoch": 1.8901226734348562, + "grad_norm": 8.0625, + "learning_rate": 3.573111795131817e-06, + "loss": 2.44, + "step": 17873 + }, + { + "epoch": 1.890228426395939, + "grad_norm": 8.125, + "learning_rate": 3.572957116009994e-06, + "loss": 2.7338, + "step": 17874 + }, + { + "epoch": 1.8903341793570219, + "grad_norm": 8.0625, + "learning_rate": 3.5728024318532217e-06, + "loss": 2.6499, + "step": 17875 + }, + { + "epoch": 1.890439932318105, + "grad_norm": 8.125, + "learning_rate": 3.5726477426622265e-06, + "loss": 2.5227, + "step": 17876 + }, + { + "epoch": 1.890545685279188, + "grad_norm": 8.125, + "learning_rate": 3.5724930484377346e-06, + "loss": 2.3867, + "step": 17877 + }, + { + "epoch": 1.8906514382402708, + "grad_norm": 8.4375, + "learning_rate": 3.5723383491804716e-06, + "loss": 2.6713, + "step": 17878 + }, + { + "epoch": 1.8907571912013537, + "grad_norm": 8.375, + "learning_rate": 3.5721836448911623e-06, + "loss": 2.5646, + "step": 17879 + }, + { + "epoch": 1.8908629441624365, + "grad_norm": 8.4375, + "learning_rate": 3.5720289355705346e-06, + "loss": 2.5692, + "step": 17880 + }, + { + "epoch": 1.8909686971235193, + "grad_norm": 8.5, + "learning_rate": 3.571874221219313e-06, + "loss": 2.6333, + "step": 17881 + }, + { + "epoch": 1.8910744500846024, + "grad_norm": 7.9375, + "learning_rate": 3.5717195018382246e-06, + "loss": 2.6278, + "step": 17882 + }, + { + "epoch": 1.8911802030456852, + "grad_norm": 8.0625, + "learning_rate": 3.571564777427994e-06, + "loss": 2.6242, + "step": 17883 + }, + { + "epoch": 1.8912859560067683, + "grad_norm": 8.4375, + "learning_rate": 3.571410047989349e-06, + "loss": 2.6329, + "step": 17884 + }, + { + "epoch": 1.8913917089678511, + "grad_norm": 8.5625, + "learning_rate": 3.5712553135230145e-06, + "loss": 2.4509, + "step": 17885 + }, + { + "epoch": 1.891497461928934, + "grad_norm": 7.96875, + "learning_rate": 3.5711005740297166e-06, + "loss": 2.5362, + "step": 17886 + }, + { + "epoch": 1.8916032148900168, + "grad_norm": 8.3125, + "learning_rate": 3.570945829510183e-06, + "loss": 2.6164, + "step": 17887 + }, + { + "epoch": 1.8917089678510999, + "grad_norm": 8.0625, + "learning_rate": 3.570791079965137e-06, + "loss": 2.5292, + "step": 17888 + }, + { + "epoch": 1.8918147208121827, + "grad_norm": 7.96875, + "learning_rate": 3.5706363253953074e-06, + "loss": 2.4748, + "step": 17889 + }, + { + "epoch": 1.8919204737732658, + "grad_norm": 8.0625, + "learning_rate": 3.5704815658014193e-06, + "loss": 2.4311, + "step": 17890 + }, + { + "epoch": 1.8920262267343486, + "grad_norm": 8.0, + "learning_rate": 3.5703268011841984e-06, + "loss": 2.3844, + "step": 17891 + }, + { + "epoch": 1.8921319796954315, + "grad_norm": 7.90625, + "learning_rate": 3.570172031544373e-06, + "loss": 2.4851, + "step": 17892 + }, + { + "epoch": 1.8922377326565143, + "grad_norm": 8.0625, + "learning_rate": 3.5700172568826664e-06, + "loss": 2.4502, + "step": 17893 + }, + { + "epoch": 1.8923434856175971, + "grad_norm": 8.4375, + "learning_rate": 3.5698624771998075e-06, + "loss": 2.5027, + "step": 17894 + }, + { + "epoch": 1.8924492385786802, + "grad_norm": 8.4375, + "learning_rate": 3.569707692496521e-06, + "loss": 2.6281, + "step": 17895 + }, + { + "epoch": 1.8925549915397633, + "grad_norm": 8.5625, + "learning_rate": 3.569552902773534e-06, + "loss": 2.6108, + "step": 17896 + }, + { + "epoch": 1.892660744500846, + "grad_norm": 8.4375, + "learning_rate": 3.5693981080315727e-06, + "loss": 2.6856, + "step": 17897 + }, + { + "epoch": 1.892766497461929, + "grad_norm": 8.4375, + "learning_rate": 3.5692433082713627e-06, + "loss": 2.4765, + "step": 17898 + }, + { + "epoch": 1.8928722504230118, + "grad_norm": 8.75, + "learning_rate": 3.5690885034936323e-06, + "loss": 2.7737, + "step": 17899 + }, + { + "epoch": 1.8929780033840946, + "grad_norm": 8.375, + "learning_rate": 3.5689336936991067e-06, + "loss": 2.7457, + "step": 17900 + }, + { + "epoch": 1.8930837563451777, + "grad_norm": 8.5625, + "learning_rate": 3.5687788788885116e-06, + "loss": 2.5177, + "step": 17901 + }, + { + "epoch": 1.8931895093062607, + "grad_norm": 8.375, + "learning_rate": 3.5686240590625743e-06, + "loss": 2.6147, + "step": 17902 + }, + { + "epoch": 1.8932952622673436, + "grad_norm": 8.5625, + "learning_rate": 3.5684692342220215e-06, + "loss": 2.8635, + "step": 17903 + }, + { + "epoch": 1.8934010152284264, + "grad_norm": 7.875, + "learning_rate": 3.56831440436758e-06, + "loss": 2.4727, + "step": 17904 + }, + { + "epoch": 1.8935067681895092, + "grad_norm": 8.3125, + "learning_rate": 3.568159569499976e-06, + "loss": 2.504, + "step": 17905 + }, + { + "epoch": 1.893612521150592, + "grad_norm": 8.3125, + "learning_rate": 3.568004729619936e-06, + "loss": 2.3783, + "step": 17906 + }, + { + "epoch": 1.8937182741116751, + "grad_norm": 8.8125, + "learning_rate": 3.5678498847281862e-06, + "loss": 2.5919, + "step": 17907 + }, + { + "epoch": 1.893824027072758, + "grad_norm": 7.96875, + "learning_rate": 3.567695034825454e-06, + "loss": 2.4599, + "step": 17908 + }, + { + "epoch": 1.893929780033841, + "grad_norm": 8.3125, + "learning_rate": 3.567540179912466e-06, + "loss": 2.5309, + "step": 17909 + }, + { + "epoch": 1.8940355329949239, + "grad_norm": 8.5, + "learning_rate": 3.567385319989948e-06, + "loss": 2.263, + "step": 17910 + }, + { + "epoch": 1.8941412859560067, + "grad_norm": 8.6875, + "learning_rate": 3.567230455058627e-06, + "loss": 2.4788, + "step": 17911 + }, + { + "epoch": 1.8942470389170896, + "grad_norm": 8.3125, + "learning_rate": 3.56707558511923e-06, + "loss": 2.6998, + "step": 17912 + }, + { + "epoch": 1.8943527918781726, + "grad_norm": 8.25, + "learning_rate": 3.5669207101724843e-06, + "loss": 2.3416, + "step": 17913 + }, + { + "epoch": 1.8944585448392555, + "grad_norm": 8.0625, + "learning_rate": 3.5667658302191154e-06, + "loss": 2.4327, + "step": 17914 + }, + { + "epoch": 1.8945642978003385, + "grad_norm": 8.875, + "learning_rate": 3.566610945259851e-06, + "loss": 2.528, + "step": 17915 + }, + { + "epoch": 1.8946700507614214, + "grad_norm": 8.875, + "learning_rate": 3.566456055295418e-06, + "loss": 2.6114, + "step": 17916 + }, + { + "epoch": 1.8947758037225042, + "grad_norm": 8.125, + "learning_rate": 3.5663011603265422e-06, + "loss": 2.5696, + "step": 17917 + }, + { + "epoch": 1.894881556683587, + "grad_norm": 8.3125, + "learning_rate": 3.566146260353952e-06, + "loss": 2.7162, + "step": 17918 + }, + { + "epoch": 1.89498730964467, + "grad_norm": 9.3125, + "learning_rate": 3.5659913553783733e-06, + "loss": 2.6316, + "step": 17919 + }, + { + "epoch": 1.895093062605753, + "grad_norm": 8.875, + "learning_rate": 3.5658364454005324e-06, + "loss": 2.6247, + "step": 17920 + }, + { + "epoch": 1.895198815566836, + "grad_norm": 8.625, + "learning_rate": 3.565681530421157e-06, + "loss": 2.2985, + "step": 17921 + }, + { + "epoch": 1.8953045685279188, + "grad_norm": 8.0, + "learning_rate": 3.565526610440975e-06, + "loss": 2.5315, + "step": 17922 + }, + { + "epoch": 1.8954103214890017, + "grad_norm": 8.25, + "learning_rate": 3.5653716854607117e-06, + "loss": 2.7023, + "step": 17923 + }, + { + "epoch": 1.8955160744500845, + "grad_norm": 8.5, + "learning_rate": 3.5652167554810953e-06, + "loss": 2.593, + "step": 17924 + }, + { + "epoch": 1.8956218274111674, + "grad_norm": 8.3125, + "learning_rate": 3.565061820502852e-06, + "loss": 2.4928, + "step": 17925 + }, + { + "epoch": 1.8957275803722504, + "grad_norm": 8.375, + "learning_rate": 3.564906880526709e-06, + "loss": 2.6491, + "step": 17926 + }, + { + "epoch": 1.8958333333333335, + "grad_norm": 8.4375, + "learning_rate": 3.5647519355533937e-06, + "loss": 2.4905, + "step": 17927 + }, + { + "epoch": 1.8959390862944163, + "grad_norm": 8.0625, + "learning_rate": 3.5645969855836333e-06, + "loss": 2.5924, + "step": 17928 + }, + { + "epoch": 1.8960448392554992, + "grad_norm": 8.5625, + "learning_rate": 3.5644420306181544e-06, + "loss": 2.4464, + "step": 17929 + }, + { + "epoch": 1.896150592216582, + "grad_norm": 8.1875, + "learning_rate": 3.5642870706576836e-06, + "loss": 2.3812, + "step": 17930 + }, + { + "epoch": 1.8962563451776648, + "grad_norm": 8.25, + "learning_rate": 3.56413210570295e-06, + "loss": 2.6014, + "step": 17931 + }, + { + "epoch": 1.896362098138748, + "grad_norm": 8.125, + "learning_rate": 3.56397713575468e-06, + "loss": 2.6002, + "step": 17932 + }, + { + "epoch": 1.896467851099831, + "grad_norm": 8.625, + "learning_rate": 3.5638221608135993e-06, + "loss": 2.3706, + "step": 17933 + }, + { + "epoch": 1.8965736040609138, + "grad_norm": 7.9375, + "learning_rate": 3.5636671808804368e-06, + "loss": 2.5691, + "step": 17934 + }, + { + "epoch": 1.8966793570219966, + "grad_norm": 8.375, + "learning_rate": 3.563512195955919e-06, + "loss": 2.6469, + "step": 17935 + }, + { + "epoch": 1.8967851099830795, + "grad_norm": 8.75, + "learning_rate": 3.563357206040774e-06, + "loss": 2.7528, + "step": 17936 + }, + { + "epoch": 1.8968908629441623, + "grad_norm": 8.1875, + "learning_rate": 3.563202211135728e-06, + "loss": 2.4635, + "step": 17937 + }, + { + "epoch": 1.8969966159052454, + "grad_norm": 7.875, + "learning_rate": 3.563047211241509e-06, + "loss": 2.5421, + "step": 17938 + }, + { + "epoch": 1.8971023688663282, + "grad_norm": 8.3125, + "learning_rate": 3.562892206358844e-06, + "loss": 2.6045, + "step": 17939 + }, + { + "epoch": 1.8972081218274113, + "grad_norm": 8.5625, + "learning_rate": 3.5627371964884606e-06, + "loss": 2.547, + "step": 17940 + }, + { + "epoch": 1.8973138747884941, + "grad_norm": 8.8125, + "learning_rate": 3.562582181631087e-06, + "loss": 2.6173, + "step": 17941 + }, + { + "epoch": 1.897419627749577, + "grad_norm": 8.25, + "learning_rate": 3.5624271617874486e-06, + "loss": 2.5896, + "step": 17942 + }, + { + "epoch": 1.8975253807106598, + "grad_norm": 8.6875, + "learning_rate": 3.5622721369582745e-06, + "loss": 2.6005, + "step": 17943 + }, + { + "epoch": 1.8976311336717429, + "grad_norm": 8.3125, + "learning_rate": 3.562117107144292e-06, + "loss": 2.5145, + "step": 17944 + }, + { + "epoch": 1.8977368866328257, + "grad_norm": 8.25, + "learning_rate": 3.5619620723462277e-06, + "loss": 2.5396, + "step": 17945 + }, + { + "epoch": 1.8978426395939088, + "grad_norm": 8.375, + "learning_rate": 3.56180703256481e-06, + "loss": 2.5651, + "step": 17946 + }, + { + "epoch": 1.8979483925549916, + "grad_norm": 9.6875, + "learning_rate": 3.5616519878007665e-06, + "loss": 2.6748, + "step": 17947 + }, + { + "epoch": 1.8980541455160744, + "grad_norm": 8.0625, + "learning_rate": 3.561496938054825e-06, + "loss": 2.4237, + "step": 17948 + }, + { + "epoch": 1.8981598984771573, + "grad_norm": 8.25, + "learning_rate": 3.5613418833277112e-06, + "loss": 2.5279, + "step": 17949 + }, + { + "epoch": 1.89826565143824, + "grad_norm": 8.75, + "learning_rate": 3.561186823620155e-06, + "loss": 2.7539, + "step": 17950 + }, + { + "epoch": 1.8983714043993232, + "grad_norm": 8.5625, + "learning_rate": 3.5610317589328824e-06, + "loss": 2.3986, + "step": 17951 + }, + { + "epoch": 1.8984771573604062, + "grad_norm": 8.75, + "learning_rate": 3.560876689266622e-06, + "loss": 2.805, + "step": 17952 + }, + { + "epoch": 1.898582910321489, + "grad_norm": 8.25, + "learning_rate": 3.560721614622101e-06, + "loss": 2.6734, + "step": 17953 + }, + { + "epoch": 1.898688663282572, + "grad_norm": 7.9375, + "learning_rate": 3.560566535000048e-06, + "loss": 2.6133, + "step": 17954 + }, + { + "epoch": 1.8987944162436547, + "grad_norm": 9.0, + "learning_rate": 3.5604114504011888e-06, + "loss": 2.4448, + "step": 17955 + }, + { + "epoch": 1.8989001692047376, + "grad_norm": 8.5, + "learning_rate": 3.5602563608262532e-06, + "loss": 2.6678, + "step": 17956 + }, + { + "epoch": 1.8990059221658206, + "grad_norm": 8.5, + "learning_rate": 3.5601012662759677e-06, + "loss": 2.417, + "step": 17957 + }, + { + "epoch": 1.8991116751269037, + "grad_norm": 8.125, + "learning_rate": 3.5599461667510603e-06, + "loss": 2.408, + "step": 17958 + }, + { + "epoch": 1.8992174280879865, + "grad_norm": 8.125, + "learning_rate": 3.5597910622522597e-06, + "loss": 2.6528, + "step": 17959 + }, + { + "epoch": 1.8993231810490694, + "grad_norm": 8.0625, + "learning_rate": 3.5596359527802933e-06, + "loss": 2.3632, + "step": 17960 + }, + { + "epoch": 1.8994289340101522, + "grad_norm": 9.375, + "learning_rate": 3.5594808383358877e-06, + "loss": 2.5966, + "step": 17961 + }, + { + "epoch": 1.899534686971235, + "grad_norm": 8.375, + "learning_rate": 3.5593257189197726e-06, + "loss": 2.4043, + "step": 17962 + }, + { + "epoch": 1.8996404399323181, + "grad_norm": 8.25, + "learning_rate": 3.559170594532675e-06, + "loss": 2.4586, + "step": 17963 + }, + { + "epoch": 1.899746192893401, + "grad_norm": 9.9375, + "learning_rate": 3.559015465175323e-06, + "loss": 2.6848, + "step": 17964 + }, + { + "epoch": 1.899851945854484, + "grad_norm": 8.3125, + "learning_rate": 3.558860330848444e-06, + "loss": 2.5978, + "step": 17965 + }, + { + "epoch": 1.8999576988155669, + "grad_norm": 8.0, + "learning_rate": 3.558705191552767e-06, + "loss": 2.6396, + "step": 17966 + }, + { + "epoch": 1.9000634517766497, + "grad_norm": 8.375, + "learning_rate": 3.5585500472890198e-06, + "loss": 2.4815, + "step": 17967 + }, + { + "epoch": 1.9001692047377325, + "grad_norm": 8.125, + "learning_rate": 3.55839489805793e-06, + "loss": 2.5397, + "step": 17968 + }, + { + "epoch": 1.9002749576988156, + "grad_norm": 8.125, + "learning_rate": 3.558239743860226e-06, + "loss": 2.558, + "step": 17969 + }, + { + "epoch": 1.9003807106598984, + "grad_norm": 8.5625, + "learning_rate": 3.5580845846966357e-06, + "loss": 2.5995, + "step": 17970 + }, + { + "epoch": 1.9004864636209815, + "grad_norm": 8.75, + "learning_rate": 3.557929420567886e-06, + "loss": 2.6417, + "step": 17971 + }, + { + "epoch": 1.9005922165820643, + "grad_norm": 8.75, + "learning_rate": 3.5577742514747077e-06, + "loss": 2.805, + "step": 17972 + }, + { + "epoch": 1.9006979695431472, + "grad_norm": 8.5, + "learning_rate": 3.5576190774178266e-06, + "loss": 2.5717, + "step": 17973 + }, + { + "epoch": 1.90080372250423, + "grad_norm": 8.25, + "learning_rate": 3.557463898397972e-06, + "loss": 2.4659, + "step": 17974 + }, + { + "epoch": 1.900909475465313, + "grad_norm": 8.0, + "learning_rate": 3.5573087144158714e-06, + "loss": 2.3669, + "step": 17975 + }, + { + "epoch": 1.901015228426396, + "grad_norm": 7.8125, + "learning_rate": 3.5571535254722546e-06, + "loss": 2.434, + "step": 17976 + }, + { + "epoch": 1.901120981387479, + "grad_norm": 8.375, + "learning_rate": 3.556998331567847e-06, + "loss": 2.4987, + "step": 17977 + }, + { + "epoch": 1.9012267343485618, + "grad_norm": 8.125, + "learning_rate": 3.5568431327033797e-06, + "loss": 2.4743, + "step": 17978 + }, + { + "epoch": 1.9013324873096447, + "grad_norm": 8.125, + "learning_rate": 3.5566879288795798e-06, + "loss": 2.696, + "step": 17979 + }, + { + "epoch": 1.9014382402707275, + "grad_norm": 8.6875, + "learning_rate": 3.556532720097175e-06, + "loss": 2.5682, + "step": 17980 + }, + { + "epoch": 1.9015439932318103, + "grad_norm": 8.125, + "learning_rate": 3.556377506356895e-06, + "loss": 2.5623, + "step": 17981 + }, + { + "epoch": 1.9016497461928934, + "grad_norm": 8.375, + "learning_rate": 3.5562222876594664e-06, + "loss": 2.3898, + "step": 17982 + }, + { + "epoch": 1.9017554991539765, + "grad_norm": 8.4375, + "learning_rate": 3.5560670640056193e-06, + "loss": 2.5849, + "step": 17983 + }, + { + "epoch": 1.9018612521150593, + "grad_norm": 8.1875, + "learning_rate": 3.5559118353960815e-06, + "loss": 2.5647, + "step": 17984 + }, + { + "epoch": 1.9019670050761421, + "grad_norm": 8.4375, + "learning_rate": 3.5557566018315813e-06, + "loss": 2.4968, + "step": 17985 + }, + { + "epoch": 1.902072758037225, + "grad_norm": 8.5, + "learning_rate": 3.5556013633128472e-06, + "loss": 2.7098, + "step": 17986 + }, + { + "epoch": 1.9021785109983078, + "grad_norm": 8.5, + "learning_rate": 3.555446119840607e-06, + "loss": 2.2508, + "step": 17987 + }, + { + "epoch": 1.9022842639593909, + "grad_norm": 8.125, + "learning_rate": 3.5552908714155905e-06, + "loss": 2.3564, + "step": 17988 + }, + { + "epoch": 1.902390016920474, + "grad_norm": 8.25, + "learning_rate": 3.5551356180385255e-06, + "loss": 2.6659, + "step": 17989 + }, + { + "epoch": 1.9024957698815568, + "grad_norm": 8.125, + "learning_rate": 3.5549803597101396e-06, + "loss": 2.4799, + "step": 17990 + }, + { + "epoch": 1.9026015228426396, + "grad_norm": 8.1875, + "learning_rate": 3.5548250964311624e-06, + "loss": 2.5085, + "step": 17991 + }, + { + "epoch": 1.9027072758037225, + "grad_norm": 8.3125, + "learning_rate": 3.554669828202324e-06, + "loss": 2.6985, + "step": 17992 + }, + { + "epoch": 1.9028130287648053, + "grad_norm": 8.1875, + "learning_rate": 3.5545145550243505e-06, + "loss": 2.5508, + "step": 17993 + }, + { + "epoch": 1.9029187817258884, + "grad_norm": 8.375, + "learning_rate": 3.5543592768979713e-06, + "loss": 2.5639, + "step": 17994 + }, + { + "epoch": 1.9030245346869712, + "grad_norm": 8.375, + "learning_rate": 3.554203993823916e-06, + "loss": 2.5924, + "step": 17995 + }, + { + "epoch": 1.9031302876480543, + "grad_norm": 8.375, + "learning_rate": 3.5540487058029116e-06, + "loss": 2.5356, + "step": 17996 + }, + { + "epoch": 1.903236040609137, + "grad_norm": 8.6875, + "learning_rate": 3.5538934128356876e-06, + "loss": 2.5459, + "step": 17997 + }, + { + "epoch": 1.90334179357022, + "grad_norm": 8.625, + "learning_rate": 3.5537381149229736e-06, + "loss": 2.4758, + "step": 17998 + }, + { + "epoch": 1.9034475465313028, + "grad_norm": 8.125, + "learning_rate": 3.553582812065497e-06, + "loss": 2.6485, + "step": 17999 + }, + { + "epoch": 1.9035532994923858, + "grad_norm": 8.375, + "learning_rate": 3.5534275042639875e-06, + "loss": 2.5071, + "step": 18000 + }, + { + "epoch": 1.9036590524534687, + "grad_norm": 10.5, + "learning_rate": 3.5532721915191736e-06, + "loss": 2.524, + "step": 18001 + }, + { + "epoch": 1.9037648054145517, + "grad_norm": 8.1875, + "learning_rate": 3.5531168738317847e-06, + "loss": 2.488, + "step": 18002 + }, + { + "epoch": 1.9038705583756346, + "grad_norm": 8.5, + "learning_rate": 3.5529615512025473e-06, + "loss": 2.3787, + "step": 18003 + }, + { + "epoch": 1.9039763113367174, + "grad_norm": 8.5, + "learning_rate": 3.5528062236321935e-06, + "loss": 2.3971, + "step": 18004 + }, + { + "epoch": 1.9040820642978002, + "grad_norm": 8.8125, + "learning_rate": 3.5526508911214503e-06, + "loss": 2.613, + "step": 18005 + }, + { + "epoch": 1.904187817258883, + "grad_norm": 9.1875, + "learning_rate": 3.5524955536710463e-06, + "loss": 2.447, + "step": 18006 + }, + { + "epoch": 1.9042935702199661, + "grad_norm": 8.5625, + "learning_rate": 3.5523402112817123e-06, + "loss": 2.4154, + "step": 18007 + }, + { + "epoch": 1.9043993231810492, + "grad_norm": 7.84375, + "learning_rate": 3.552184863954175e-06, + "loss": 2.4441, + "step": 18008 + }, + { + "epoch": 1.904505076142132, + "grad_norm": 8.1875, + "learning_rate": 3.552029511689165e-06, + "loss": 2.5075, + "step": 18009 + }, + { + "epoch": 1.9046108291032149, + "grad_norm": 11.6875, + "learning_rate": 3.5518741544874106e-06, + "loss": 2.4879, + "step": 18010 + }, + { + "epoch": 1.9047165820642977, + "grad_norm": 9.0, + "learning_rate": 3.5517187923496414e-06, + "loss": 2.5275, + "step": 18011 + }, + { + "epoch": 1.9048223350253806, + "grad_norm": 8.3125, + "learning_rate": 3.551563425276586e-06, + "loss": 2.5994, + "step": 18012 + }, + { + "epoch": 1.9049280879864636, + "grad_norm": 9.0625, + "learning_rate": 3.551408053268973e-06, + "loss": 2.5595, + "step": 18013 + }, + { + "epoch": 1.9050338409475467, + "grad_norm": 8.0625, + "learning_rate": 3.5512526763275325e-06, + "loss": 2.3821, + "step": 18014 + }, + { + "epoch": 1.9051395939086295, + "grad_norm": 8.125, + "learning_rate": 3.5510972944529927e-06, + "loss": 2.4582, + "step": 18015 + }, + { + "epoch": 1.9052453468697124, + "grad_norm": 8.875, + "learning_rate": 3.550941907646083e-06, + "loss": 2.571, + "step": 18016 + }, + { + "epoch": 1.9053510998307952, + "grad_norm": 8.6875, + "learning_rate": 3.5507865159075334e-06, + "loss": 2.5133, + "step": 18017 + }, + { + "epoch": 1.905456852791878, + "grad_norm": 8.4375, + "learning_rate": 3.550631119238072e-06, + "loss": 2.5323, + "step": 18018 + }, + { + "epoch": 1.905562605752961, + "grad_norm": 8.25, + "learning_rate": 3.5504757176384287e-06, + "loss": 2.4042, + "step": 18019 + }, + { + "epoch": 1.905668358714044, + "grad_norm": 8.3125, + "learning_rate": 3.5503203111093327e-06, + "loss": 2.3878, + "step": 18020 + }, + { + "epoch": 1.905774111675127, + "grad_norm": 8.9375, + "learning_rate": 3.5501648996515125e-06, + "loss": 2.5919, + "step": 18021 + }, + { + "epoch": 1.9058798646362098, + "grad_norm": 8.625, + "learning_rate": 3.5500094832656982e-06, + "loss": 2.7928, + "step": 18022 + }, + { + "epoch": 1.9059856175972927, + "grad_norm": 9.125, + "learning_rate": 3.5498540619526188e-06, + "loss": 2.6612, + "step": 18023 + }, + { + "epoch": 1.9060913705583755, + "grad_norm": 8.4375, + "learning_rate": 3.5496986357130043e-06, + "loss": 2.539, + "step": 18024 + }, + { + "epoch": 1.9061971235194586, + "grad_norm": 8.125, + "learning_rate": 3.5495432045475823e-06, + "loss": 2.4303, + "step": 18025 + }, + { + "epoch": 1.9063028764805414, + "grad_norm": 8.3125, + "learning_rate": 3.549387768457084e-06, + "loss": 2.5164, + "step": 18026 + }, + { + "epoch": 1.9064086294416245, + "grad_norm": 8.0, + "learning_rate": 3.549232327442238e-06, + "loss": 2.5573, + "step": 18027 + }, + { + "epoch": 1.9065143824027073, + "grad_norm": 8.4375, + "learning_rate": 3.5490768815037736e-06, + "loss": 2.6099, + "step": 18028 + }, + { + "epoch": 1.9066201353637902, + "grad_norm": 7.9375, + "learning_rate": 3.5489214306424203e-06, + "loss": 2.5194, + "step": 18029 + }, + { + "epoch": 1.906725888324873, + "grad_norm": 8.4375, + "learning_rate": 3.548765974858908e-06, + "loss": 2.6427, + "step": 18030 + }, + { + "epoch": 1.906831641285956, + "grad_norm": 8.3125, + "learning_rate": 3.5486105141539656e-06, + "loss": 2.6209, + "step": 18031 + }, + { + "epoch": 1.906937394247039, + "grad_norm": 9.0, + "learning_rate": 3.5484550485283237e-06, + "loss": 2.2603, + "step": 18032 + }, + { + "epoch": 1.907043147208122, + "grad_norm": 9.125, + "learning_rate": 3.5482995779827107e-06, + "loss": 2.6908, + "step": 18033 + }, + { + "epoch": 1.9071489001692048, + "grad_norm": 8.25, + "learning_rate": 3.548144102517856e-06, + "loss": 2.4608, + "step": 18034 + }, + { + "epoch": 1.9072546531302876, + "grad_norm": 9.5625, + "learning_rate": 3.5479886221344895e-06, + "loss": 2.1616, + "step": 18035 + }, + { + "epoch": 1.9073604060913705, + "grad_norm": 8.125, + "learning_rate": 3.547833136833342e-06, + "loss": 2.5027, + "step": 18036 + }, + { + "epoch": 1.9074661590524533, + "grad_norm": 8.625, + "learning_rate": 3.547677646615142e-06, + "loss": 2.6357, + "step": 18037 + }, + { + "epoch": 1.9075719120135364, + "grad_norm": 8.4375, + "learning_rate": 3.547522151480619e-06, + "loss": 2.4009, + "step": 18038 + }, + { + "epoch": 1.9076776649746194, + "grad_norm": 8.5, + "learning_rate": 3.5473666514305028e-06, + "loss": 2.7597, + "step": 18039 + }, + { + "epoch": 1.9077834179357023, + "grad_norm": 8.1875, + "learning_rate": 3.547211146465524e-06, + "loss": 2.2664, + "step": 18040 + }, + { + "epoch": 1.9078891708967851, + "grad_norm": 8.9375, + "learning_rate": 3.5470556365864105e-06, + "loss": 2.2742, + "step": 18041 + }, + { + "epoch": 1.907994923857868, + "grad_norm": 7.875, + "learning_rate": 3.5469001217938937e-06, + "loss": 2.6534, + "step": 18042 + }, + { + "epoch": 1.9081006768189508, + "grad_norm": 8.25, + "learning_rate": 3.546744602088703e-06, + "loss": 2.4729, + "step": 18043 + }, + { + "epoch": 1.9082064297800339, + "grad_norm": 8.0, + "learning_rate": 3.546589077471568e-06, + "loss": 2.5526, + "step": 18044 + }, + { + "epoch": 1.908312182741117, + "grad_norm": 8.0625, + "learning_rate": 3.5464335479432176e-06, + "loss": 2.3943, + "step": 18045 + }, + { + "epoch": 1.9084179357021998, + "grad_norm": 8.3125, + "learning_rate": 3.5462780135043836e-06, + "loss": 2.4821, + "step": 18046 + }, + { + "epoch": 1.9085236886632826, + "grad_norm": 8.0625, + "learning_rate": 3.546122474155794e-06, + "loss": 2.4323, + "step": 18047 + }, + { + "epoch": 1.9086294416243654, + "grad_norm": 8.5625, + "learning_rate": 3.5459669298981804e-06, + "loss": 2.514, + "step": 18048 + }, + { + "epoch": 1.9087351945854483, + "grad_norm": 8.0625, + "learning_rate": 3.545811380732271e-06, + "loss": 2.5807, + "step": 18049 + }, + { + "epoch": 1.9088409475465313, + "grad_norm": 8.3125, + "learning_rate": 3.5456558266587967e-06, + "loss": 2.6507, + "step": 18050 + }, + { + "epoch": 1.9089467005076142, + "grad_norm": 8.875, + "learning_rate": 3.5455002676784874e-06, + "loss": 2.7444, + "step": 18051 + }, + { + "epoch": 1.9090524534686972, + "grad_norm": 8.375, + "learning_rate": 3.5453447037920726e-06, + "loss": 2.7707, + "step": 18052 + }, + { + "epoch": 1.90915820642978, + "grad_norm": 8.125, + "learning_rate": 3.5451891350002834e-06, + "loss": 2.4865, + "step": 18053 + }, + { + "epoch": 1.909263959390863, + "grad_norm": 8.125, + "learning_rate": 3.5450335613038477e-06, + "loss": 2.5552, + "step": 18054 + }, + { + "epoch": 1.9093697123519457, + "grad_norm": 8.5625, + "learning_rate": 3.544877982703498e-06, + "loss": 2.5086, + "step": 18055 + }, + { + "epoch": 1.9094754653130288, + "grad_norm": 8.25, + "learning_rate": 3.5447223991999635e-06, + "loss": 2.4912, + "step": 18056 + }, + { + "epoch": 1.9095812182741116, + "grad_norm": 8.1875, + "learning_rate": 3.544566810793973e-06, + "loss": 2.4096, + "step": 18057 + }, + { + "epoch": 1.9096869712351947, + "grad_norm": 7.96875, + "learning_rate": 3.544411217486259e-06, + "loss": 2.5095, + "step": 18058 + }, + { + "epoch": 1.9097927241962775, + "grad_norm": 8.4375, + "learning_rate": 3.544255619277549e-06, + "loss": 2.7759, + "step": 18059 + }, + { + "epoch": 1.9098984771573604, + "grad_norm": 8.25, + "learning_rate": 3.544100016168575e-06, + "loss": 2.55, + "step": 18060 + }, + { + "epoch": 1.9100042301184432, + "grad_norm": 8.4375, + "learning_rate": 3.5439444081600666e-06, + "loss": 2.6096, + "step": 18061 + }, + { + "epoch": 1.910109983079526, + "grad_norm": 8.375, + "learning_rate": 3.5437887952527538e-06, + "loss": 2.6944, + "step": 18062 + }, + { + "epoch": 1.9102157360406091, + "grad_norm": 8.1875, + "learning_rate": 3.5436331774473683e-06, + "loss": 2.4284, + "step": 18063 + }, + { + "epoch": 1.9103214890016922, + "grad_norm": 8.0, + "learning_rate": 3.543477554744638e-06, + "loss": 2.5207, + "step": 18064 + }, + { + "epoch": 1.910427241962775, + "grad_norm": 8.0625, + "learning_rate": 3.5433219271452946e-06, + "loss": 2.5897, + "step": 18065 + }, + { + "epoch": 1.9105329949238579, + "grad_norm": 8.9375, + "learning_rate": 3.543166294650068e-06, + "loss": 2.5084, + "step": 18066 + }, + { + "epoch": 1.9106387478849407, + "grad_norm": 8.3125, + "learning_rate": 3.5430106572596882e-06, + "loss": 2.6656, + "step": 18067 + }, + { + "epoch": 1.9107445008460235, + "grad_norm": 8.9375, + "learning_rate": 3.542855014974887e-06, + "loss": 2.6718, + "step": 18068 + }, + { + "epoch": 1.9108502538071066, + "grad_norm": 8.375, + "learning_rate": 3.542699367796393e-06, + "loss": 2.2128, + "step": 18069 + }, + { + "epoch": 1.9109560067681897, + "grad_norm": 8.5625, + "learning_rate": 3.5425437157249364e-06, + "loss": 2.5569, + "step": 18070 + }, + { + "epoch": 1.9110617597292725, + "grad_norm": 8.5, + "learning_rate": 3.5423880587612504e-06, + "loss": 2.4711, + "step": 18071 + }, + { + "epoch": 1.9111675126903553, + "grad_norm": 8.125, + "learning_rate": 3.542232396906062e-06, + "loss": 2.3709, + "step": 18072 + }, + { + "epoch": 1.9112732656514382, + "grad_norm": 8.3125, + "learning_rate": 3.5420767301601045e-06, + "loss": 2.2286, + "step": 18073 + }, + { + "epoch": 1.911379018612521, + "grad_norm": 8.5625, + "learning_rate": 3.5419210585241066e-06, + "loss": 2.6602, + "step": 18074 + }, + { + "epoch": 1.911484771573604, + "grad_norm": 8.375, + "learning_rate": 3.5417653819987995e-06, + "loss": 2.5005, + "step": 18075 + }, + { + "epoch": 1.911590524534687, + "grad_norm": 8.1875, + "learning_rate": 3.5416097005849132e-06, + "loss": 2.3767, + "step": 18076 + }, + { + "epoch": 1.91169627749577, + "grad_norm": 8.5625, + "learning_rate": 3.541454014283179e-06, + "loss": 2.6279, + "step": 18077 + }, + { + "epoch": 1.9118020304568528, + "grad_norm": 8.375, + "learning_rate": 3.5412983230943267e-06, + "loss": 2.6529, + "step": 18078 + }, + { + "epoch": 1.9119077834179357, + "grad_norm": 7.90625, + "learning_rate": 3.541142627019087e-06, + "loss": 2.5444, + "step": 18079 + }, + { + "epoch": 1.9120135363790185, + "grad_norm": 8.125, + "learning_rate": 3.540986926058191e-06, + "loss": 2.616, + "step": 18080 + }, + { + "epoch": 1.9121192893401016, + "grad_norm": 8.25, + "learning_rate": 3.540831220212369e-06, + "loss": 2.572, + "step": 18081 + }, + { + "epoch": 1.9122250423011844, + "grad_norm": 8.125, + "learning_rate": 3.540675509482352e-06, + "loss": 2.5318, + "step": 18082 + }, + { + "epoch": 1.9123307952622675, + "grad_norm": 8.0625, + "learning_rate": 3.5405197938688702e-06, + "loss": 2.5631, + "step": 18083 + }, + { + "epoch": 1.9124365482233503, + "grad_norm": 8.5, + "learning_rate": 3.540364073372655e-06, + "loss": 2.7913, + "step": 18084 + }, + { + "epoch": 1.9125423011844331, + "grad_norm": 8.25, + "learning_rate": 3.5402083479944365e-06, + "loss": 2.6108, + "step": 18085 + }, + { + "epoch": 1.912648054145516, + "grad_norm": 8.25, + "learning_rate": 3.5400526177349452e-06, + "loss": 2.5542, + "step": 18086 + }, + { + "epoch": 1.912753807106599, + "grad_norm": 8.0, + "learning_rate": 3.5398968825949133e-06, + "loss": 2.5184, + "step": 18087 + }, + { + "epoch": 1.9128595600676819, + "grad_norm": 8.1875, + "learning_rate": 3.5397411425750694e-06, + "loss": 2.5173, + "step": 18088 + }, + { + "epoch": 1.912965313028765, + "grad_norm": 8.125, + "learning_rate": 3.5395853976761464e-06, + "loss": 2.4315, + "step": 18089 + }, + { + "epoch": 1.9130710659898478, + "grad_norm": 8.5, + "learning_rate": 3.539429647898874e-06, + "loss": 2.5964, + "step": 18090 + }, + { + "epoch": 1.9131768189509306, + "grad_norm": 8.5, + "learning_rate": 3.5392738932439834e-06, + "loss": 2.8112, + "step": 18091 + }, + { + "epoch": 1.9132825719120135, + "grad_norm": 8.0625, + "learning_rate": 3.5391181337122054e-06, + "loss": 2.5508, + "step": 18092 + }, + { + "epoch": 1.9133883248730963, + "grad_norm": 7.6875, + "learning_rate": 3.5389623693042713e-06, + "loss": 2.5386, + "step": 18093 + }, + { + "epoch": 1.9134940778341794, + "grad_norm": 7.90625, + "learning_rate": 3.538806600020912e-06, + "loss": 2.686, + "step": 18094 + }, + { + "epoch": 1.9135998307952624, + "grad_norm": 8.5, + "learning_rate": 3.538650825862857e-06, + "loss": 2.5673, + "step": 18095 + }, + { + "epoch": 1.9137055837563453, + "grad_norm": 8.125, + "learning_rate": 3.53849504683084e-06, + "loss": 2.6662, + "step": 18096 + }, + { + "epoch": 1.913811336717428, + "grad_norm": 9.375, + "learning_rate": 3.5383392629255893e-06, + "loss": 2.6623, + "step": 18097 + }, + { + "epoch": 1.913917089678511, + "grad_norm": 8.25, + "learning_rate": 3.5381834741478376e-06, + "loss": 2.573, + "step": 18098 + }, + { + "epoch": 1.9140228426395938, + "grad_norm": 8.75, + "learning_rate": 3.538027680498315e-06, + "loss": 2.505, + "step": 18099 + }, + { + "epoch": 1.9141285956006768, + "grad_norm": 9.0, + "learning_rate": 3.5378718819777534e-06, + "loss": 2.3657, + "step": 18100 + }, + { + "epoch": 1.91423434856176, + "grad_norm": 9.0, + "learning_rate": 3.5377160785868843e-06, + "loss": 2.7546, + "step": 18101 + }, + { + "epoch": 1.9143401015228427, + "grad_norm": 8.25, + "learning_rate": 3.5375602703264367e-06, + "loss": 2.6644, + "step": 18102 + }, + { + "epoch": 1.9144458544839256, + "grad_norm": 8.625, + "learning_rate": 3.5374044571971437e-06, + "loss": 2.5795, + "step": 18103 + }, + { + "epoch": 1.9145516074450084, + "grad_norm": 8.1875, + "learning_rate": 3.5372486391997356e-06, + "loss": 2.6292, + "step": 18104 + }, + { + "epoch": 1.9146573604060912, + "grad_norm": 8.375, + "learning_rate": 3.5370928163349437e-06, + "loss": 2.5593, + "step": 18105 + }, + { + "epoch": 1.9147631133671743, + "grad_norm": 8.0625, + "learning_rate": 3.5369369886035e-06, + "loss": 2.7269, + "step": 18106 + }, + { + "epoch": 1.9148688663282571, + "grad_norm": 8.125, + "learning_rate": 3.536781156006135e-06, + "loss": 2.5774, + "step": 18107 + }, + { + "epoch": 1.9149746192893402, + "grad_norm": 8.5, + "learning_rate": 3.53662531854358e-06, + "loss": 2.547, + "step": 18108 + }, + { + "epoch": 1.915080372250423, + "grad_norm": 8.375, + "learning_rate": 3.536469476216566e-06, + "loss": 2.5948, + "step": 18109 + }, + { + "epoch": 1.9151861252115059, + "grad_norm": 8.1875, + "learning_rate": 3.5363136290258253e-06, + "loss": 2.4966, + "step": 18110 + }, + { + "epoch": 1.9152918781725887, + "grad_norm": 8.5625, + "learning_rate": 3.536157776972088e-06, + "loss": 2.5587, + "step": 18111 + }, + { + "epoch": 1.9153976311336718, + "grad_norm": 8.5625, + "learning_rate": 3.536001920056086e-06, + "loss": 2.512, + "step": 18112 + }, + { + "epoch": 1.9155033840947546, + "grad_norm": 8.875, + "learning_rate": 3.5358460582785515e-06, + "loss": 2.753, + "step": 18113 + }, + { + "epoch": 1.9156091370558377, + "grad_norm": 8.1875, + "learning_rate": 3.535690191640214e-06, + "loss": 2.5363, + "step": 18114 + }, + { + "epoch": 1.9157148900169205, + "grad_norm": 7.96875, + "learning_rate": 3.5355343201418067e-06, + "loss": 2.5246, + "step": 18115 + }, + { + "epoch": 1.9158206429780034, + "grad_norm": 8.0, + "learning_rate": 3.5353784437840606e-06, + "loss": 2.5211, + "step": 18116 + }, + { + "epoch": 1.9159263959390862, + "grad_norm": 8.5, + "learning_rate": 3.535222562567706e-06, + "loss": 2.3278, + "step": 18117 + }, + { + "epoch": 1.916032148900169, + "grad_norm": 8.75, + "learning_rate": 3.535066676493476e-06, + "loss": 2.5881, + "step": 18118 + }, + { + "epoch": 1.916137901861252, + "grad_norm": 8.8125, + "learning_rate": 3.534910785562101e-06, + "loss": 2.4981, + "step": 18119 + }, + { + "epoch": 1.9162436548223352, + "grad_norm": 8.1875, + "learning_rate": 3.534754889774314e-06, + "loss": 2.7193, + "step": 18120 + }, + { + "epoch": 1.916349407783418, + "grad_norm": 8.5, + "learning_rate": 3.534598989130844e-06, + "loss": 2.6709, + "step": 18121 + }, + { + "epoch": 1.9164551607445008, + "grad_norm": 8.125, + "learning_rate": 3.534443083632425e-06, + "loss": 2.553, + "step": 18122 + }, + { + "epoch": 1.9165609137055837, + "grad_norm": 8.0625, + "learning_rate": 3.5342871732797874e-06, + "loss": 2.4789, + "step": 18123 + }, + { + "epoch": 1.9166666666666665, + "grad_norm": 8.625, + "learning_rate": 3.534131258073663e-06, + "loss": 2.7675, + "step": 18124 + }, + { + "epoch": 1.9167724196277496, + "grad_norm": 8.0, + "learning_rate": 3.5339753380147833e-06, + "loss": 2.4839, + "step": 18125 + }, + { + "epoch": 1.9168781725888326, + "grad_norm": 8.6875, + "learning_rate": 3.533819413103881e-06, + "loss": 2.6303, + "step": 18126 + }, + { + "epoch": 1.9169839255499155, + "grad_norm": 7.65625, + "learning_rate": 3.5336634833416867e-06, + "loss": 2.4974, + "step": 18127 + }, + { + "epoch": 1.9170896785109983, + "grad_norm": 8.0, + "learning_rate": 3.5335075487289324e-06, + "loss": 2.4299, + "step": 18128 + }, + { + "epoch": 1.9171954314720812, + "grad_norm": 8.6875, + "learning_rate": 3.53335160926635e-06, + "loss": 2.6235, + "step": 18129 + }, + { + "epoch": 1.917301184433164, + "grad_norm": 8.3125, + "learning_rate": 3.5331956649546708e-06, + "loss": 2.6828, + "step": 18130 + }, + { + "epoch": 1.917406937394247, + "grad_norm": 8.5625, + "learning_rate": 3.533039715794627e-06, + "loss": 2.6548, + "step": 18131 + }, + { + "epoch": 1.91751269035533, + "grad_norm": 9.1875, + "learning_rate": 3.5328837617869503e-06, + "loss": 2.5683, + "step": 18132 + }, + { + "epoch": 1.917618443316413, + "grad_norm": 8.4375, + "learning_rate": 3.532727802932373e-06, + "loss": 2.5159, + "step": 18133 + }, + { + "epoch": 1.9177241962774958, + "grad_norm": 8.375, + "learning_rate": 3.5325718392316257e-06, + "loss": 2.5256, + "step": 18134 + }, + { + "epoch": 1.9178299492385786, + "grad_norm": 8.125, + "learning_rate": 3.5324158706854415e-06, + "loss": 2.5299, + "step": 18135 + }, + { + "epoch": 1.9179357021996615, + "grad_norm": 8.75, + "learning_rate": 3.532259897294552e-06, + "loss": 2.6477, + "step": 18136 + }, + { + "epoch": 1.9180414551607445, + "grad_norm": 8.0625, + "learning_rate": 3.5321039190596882e-06, + "loss": 2.5371, + "step": 18137 + }, + { + "epoch": 1.9181472081218274, + "grad_norm": 7.9375, + "learning_rate": 3.531947935981584e-06, + "loss": 2.5638, + "step": 18138 + }, + { + "epoch": 1.9182529610829104, + "grad_norm": 8.1875, + "learning_rate": 3.5317919480609697e-06, + "loss": 2.5902, + "step": 18139 + }, + { + "epoch": 1.9183587140439933, + "grad_norm": 8.125, + "learning_rate": 3.5316359552985773e-06, + "loss": 2.3953, + "step": 18140 + }, + { + "epoch": 1.9184644670050761, + "grad_norm": 8.0, + "learning_rate": 3.5314799576951396e-06, + "loss": 2.5712, + "step": 18141 + }, + { + "epoch": 1.918570219966159, + "grad_norm": 9.0625, + "learning_rate": 3.531323955251389e-06, + "loss": 2.4685, + "step": 18142 + }, + { + "epoch": 1.918675972927242, + "grad_norm": 8.25, + "learning_rate": 3.5311679479680554e-06, + "loss": 2.5331, + "step": 18143 + }, + { + "epoch": 1.9187817258883249, + "grad_norm": 7.8125, + "learning_rate": 3.5310119358458737e-06, + "loss": 2.4026, + "step": 18144 + }, + { + "epoch": 1.918887478849408, + "grad_norm": 8.25, + "learning_rate": 3.5308559188855746e-06, + "loss": 2.561, + "step": 18145 + }, + { + "epoch": 1.9189932318104908, + "grad_norm": 8.25, + "learning_rate": 3.5306998970878893e-06, + "loss": 2.5596, + "step": 18146 + }, + { + "epoch": 1.9190989847715736, + "grad_norm": 8.75, + "learning_rate": 3.5305438704535514e-06, + "loss": 2.7339, + "step": 18147 + }, + { + "epoch": 1.9192047377326564, + "grad_norm": 9.0, + "learning_rate": 3.5303878389832937e-06, + "loss": 2.3153, + "step": 18148 + }, + { + "epoch": 1.9193104906937393, + "grad_norm": 8.75, + "learning_rate": 3.5302318026778454e-06, + "loss": 2.6172, + "step": 18149 + }, + { + "epoch": 1.9194162436548223, + "grad_norm": 7.8125, + "learning_rate": 3.5300757615379412e-06, + "loss": 2.4881, + "step": 18150 + }, + { + "epoch": 1.9195219966159054, + "grad_norm": 8.1875, + "learning_rate": 3.529919715564313e-06, + "loss": 2.4734, + "step": 18151 + }, + { + "epoch": 1.9196277495769882, + "grad_norm": 8.875, + "learning_rate": 3.5297636647576934e-06, + "loss": 2.7115, + "step": 18152 + }, + { + "epoch": 1.919733502538071, + "grad_norm": 8.1875, + "learning_rate": 3.529607609118813e-06, + "loss": 2.6851, + "step": 18153 + }, + { + "epoch": 1.919839255499154, + "grad_norm": 8.25, + "learning_rate": 3.529451548648406e-06, + "loss": 2.6051, + "step": 18154 + }, + { + "epoch": 1.9199450084602367, + "grad_norm": 10.9375, + "learning_rate": 3.5292954833472035e-06, + "loss": 2.4275, + "step": 18155 + }, + { + "epoch": 1.9200507614213198, + "grad_norm": 7.96875, + "learning_rate": 3.5291394132159383e-06, + "loss": 2.5642, + "step": 18156 + }, + { + "epoch": 1.9201565143824029, + "grad_norm": 7.96875, + "learning_rate": 3.528983338255343e-06, + "loss": 2.5182, + "step": 18157 + }, + { + "epoch": 1.9202622673434857, + "grad_norm": 7.75, + "learning_rate": 3.5288272584661497e-06, + "loss": 2.4624, + "step": 18158 + }, + { + "epoch": 1.9203680203045685, + "grad_norm": 8.0625, + "learning_rate": 3.5286711738490897e-06, + "loss": 2.6542, + "step": 18159 + }, + { + "epoch": 1.9204737732656514, + "grad_norm": 8.375, + "learning_rate": 3.5285150844048977e-06, + "loss": 2.2428, + "step": 18160 + }, + { + "epoch": 1.9205795262267342, + "grad_norm": 8.75, + "learning_rate": 3.528358990134305e-06, + "loss": 2.4821, + "step": 18161 + }, + { + "epoch": 1.9206852791878173, + "grad_norm": 7.96875, + "learning_rate": 3.528202891038044e-06, + "loss": 2.5446, + "step": 18162 + }, + { + "epoch": 1.9207910321489001, + "grad_norm": 8.25, + "learning_rate": 3.5280467871168477e-06, + "loss": 2.559, + "step": 18163 + }, + { + "epoch": 1.9208967851099832, + "grad_norm": 8.75, + "learning_rate": 3.5278906783714485e-06, + "loss": 2.6329, + "step": 18164 + }, + { + "epoch": 1.921002538071066, + "grad_norm": 9.6875, + "learning_rate": 3.527734564802578e-06, + "loss": 2.7879, + "step": 18165 + }, + { + "epoch": 1.9211082910321489, + "grad_norm": 7.90625, + "learning_rate": 3.5275784464109697e-06, + "loss": 2.4586, + "step": 18166 + }, + { + "epoch": 1.9212140439932317, + "grad_norm": 8.75, + "learning_rate": 3.527422323197356e-06, + "loss": 2.3025, + "step": 18167 + }, + { + "epoch": 1.9213197969543148, + "grad_norm": 7.78125, + "learning_rate": 3.5272661951624698e-06, + "loss": 2.3583, + "step": 18168 + }, + { + "epoch": 1.9214255499153976, + "grad_norm": 8.0625, + "learning_rate": 3.527110062307043e-06, + "loss": 2.5186, + "step": 18169 + }, + { + "epoch": 1.9215313028764807, + "grad_norm": 7.6875, + "learning_rate": 3.526953924631809e-06, + "loss": 2.4869, + "step": 18170 + }, + { + "epoch": 1.9216370558375635, + "grad_norm": 8.1875, + "learning_rate": 3.5267977821375006e-06, + "loss": 2.4025, + "step": 18171 + }, + { + "epoch": 1.9217428087986463, + "grad_norm": 8.25, + "learning_rate": 3.5266416348248494e-06, + "loss": 2.3808, + "step": 18172 + }, + { + "epoch": 1.9218485617597292, + "grad_norm": 8.0, + "learning_rate": 3.52648548269459e-06, + "loss": 2.5723, + "step": 18173 + }, + { + "epoch": 1.921954314720812, + "grad_norm": 8.25, + "learning_rate": 3.5263293257474536e-06, + "loss": 2.6795, + "step": 18174 + }, + { + "epoch": 1.922060067681895, + "grad_norm": 8.375, + "learning_rate": 3.526173163984172e-06, + "loss": 2.609, + "step": 18175 + }, + { + "epoch": 1.9221658206429781, + "grad_norm": 9.8125, + "learning_rate": 3.526016997405481e-06, + "loss": 2.5048, + "step": 18176 + }, + { + "epoch": 1.922271573604061, + "grad_norm": 8.5, + "learning_rate": 3.5258608260121117e-06, + "loss": 2.3473, + "step": 18177 + }, + { + "epoch": 1.9223773265651438, + "grad_norm": 8.75, + "learning_rate": 3.5257046498047963e-06, + "loss": 2.6292, + "step": 18178 + }, + { + "epoch": 1.9224830795262267, + "grad_norm": 8.0, + "learning_rate": 3.5255484687842695e-06, + "loss": 2.501, + "step": 18179 + }, + { + "epoch": 1.9225888324873095, + "grad_norm": 8.375, + "learning_rate": 3.5253922829512632e-06, + "loss": 2.4452, + "step": 18180 + }, + { + "epoch": 1.9226945854483926, + "grad_norm": 7.90625, + "learning_rate": 3.5252360923065092e-06, + "loss": 2.4227, + "step": 18181 + }, + { + "epoch": 1.9228003384094756, + "grad_norm": 9.375, + "learning_rate": 3.525079896850743e-06, + "loss": 2.4838, + "step": 18182 + }, + { + "epoch": 1.9229060913705585, + "grad_norm": 8.625, + "learning_rate": 3.524923696584695e-06, + "loss": 2.6063, + "step": 18183 + }, + { + "epoch": 1.9230118443316413, + "grad_norm": 8.375, + "learning_rate": 3.524767491509099e-06, + "loss": 2.6195, + "step": 18184 + }, + { + "epoch": 1.9231175972927241, + "grad_norm": 8.1875, + "learning_rate": 3.5246112816246888e-06, + "loss": 2.4602, + "step": 18185 + }, + { + "epoch": 1.923223350253807, + "grad_norm": 7.9375, + "learning_rate": 3.524455066932197e-06, + "loss": 2.4903, + "step": 18186 + }, + { + "epoch": 1.92332910321489, + "grad_norm": 8.3125, + "learning_rate": 3.524298847432357e-06, + "loss": 2.5045, + "step": 18187 + }, + { + "epoch": 1.9234348561759729, + "grad_norm": 8.125, + "learning_rate": 3.5241426231259e-06, + "loss": 2.1193, + "step": 18188 + }, + { + "epoch": 1.923540609137056, + "grad_norm": 8.3125, + "learning_rate": 3.5239863940135623e-06, + "loss": 2.3205, + "step": 18189 + }, + { + "epoch": 1.9236463620981388, + "grad_norm": 8.1875, + "learning_rate": 3.5238301600960745e-06, + "loss": 2.5529, + "step": 18190 + }, + { + "epoch": 1.9237521150592216, + "grad_norm": 8.3125, + "learning_rate": 3.52367392137417e-06, + "loss": 2.4537, + "step": 18191 + }, + { + "epoch": 1.9238578680203045, + "grad_norm": 8.5625, + "learning_rate": 3.523517677848583e-06, + "loss": 2.6718, + "step": 18192 + }, + { + "epoch": 1.9239636209813875, + "grad_norm": 9.6875, + "learning_rate": 3.5233614295200462e-06, + "loss": 2.5548, + "step": 18193 + }, + { + "epoch": 1.9240693739424704, + "grad_norm": 7.8125, + "learning_rate": 3.5232051763892914e-06, + "loss": 2.5547, + "step": 18194 + }, + { + "epoch": 1.9241751269035534, + "grad_norm": 7.96875, + "learning_rate": 3.523048918457055e-06, + "loss": 2.6469, + "step": 18195 + }, + { + "epoch": 1.9242808798646363, + "grad_norm": 8.0, + "learning_rate": 3.522892655724068e-06, + "loss": 2.5192, + "step": 18196 + }, + { + "epoch": 1.924386632825719, + "grad_norm": 8.1875, + "learning_rate": 3.522736388191063e-06, + "loss": 2.4149, + "step": 18197 + }, + { + "epoch": 1.924492385786802, + "grad_norm": 8.1875, + "learning_rate": 3.5225801158587757e-06, + "loss": 2.8003, + "step": 18198 + }, + { + "epoch": 1.924598138747885, + "grad_norm": 8.625, + "learning_rate": 3.5224238387279382e-06, + "loss": 2.6153, + "step": 18199 + }, + { + "epoch": 1.9247038917089678, + "grad_norm": 8.4375, + "learning_rate": 3.5222675567992825e-06, + "loss": 2.6609, + "step": 18200 + }, + { + "epoch": 1.924809644670051, + "grad_norm": 8.1875, + "learning_rate": 3.522111270073544e-06, + "loss": 2.4248, + "step": 18201 + }, + { + "epoch": 1.9249153976311337, + "grad_norm": 8.3125, + "learning_rate": 3.521954978551456e-06, + "loss": 2.6362, + "step": 18202 + }, + { + "epoch": 1.9250211505922166, + "grad_norm": 8.3125, + "learning_rate": 3.52179868223375e-06, + "loss": 2.2343, + "step": 18203 + }, + { + "epoch": 1.9251269035532994, + "grad_norm": 8.375, + "learning_rate": 3.5216423811211615e-06, + "loss": 2.475, + "step": 18204 + }, + { + "epoch": 1.9252326565143822, + "grad_norm": 8.625, + "learning_rate": 3.521486075214423e-06, + "loss": 2.7288, + "step": 18205 + }, + { + "epoch": 1.9253384094754653, + "grad_norm": 8.1875, + "learning_rate": 3.5213297645142685e-06, + "loss": 2.6569, + "step": 18206 + }, + { + "epoch": 1.9254441624365484, + "grad_norm": 8.625, + "learning_rate": 3.5211734490214303e-06, + "loss": 2.5379, + "step": 18207 + }, + { + "epoch": 1.9255499153976312, + "grad_norm": 8.375, + "learning_rate": 3.5210171287366435e-06, + "loss": 2.7886, + "step": 18208 + }, + { + "epoch": 1.925655668358714, + "grad_norm": 8.3125, + "learning_rate": 3.5208608036606406e-06, + "loss": 2.5597, + "step": 18209 + }, + { + "epoch": 1.9257614213197969, + "grad_norm": 8.5, + "learning_rate": 3.5207044737941553e-06, + "loss": 2.4753, + "step": 18210 + }, + { + "epoch": 1.9258671742808797, + "grad_norm": 8.125, + "learning_rate": 3.5205481391379214e-06, + "loss": 2.4706, + "step": 18211 + }, + { + "epoch": 1.9259729272419628, + "grad_norm": 8.4375, + "learning_rate": 3.5203917996926724e-06, + "loss": 2.6236, + "step": 18212 + }, + { + "epoch": 1.9260786802030458, + "grad_norm": 8.0625, + "learning_rate": 3.5202354554591424e-06, + "loss": 2.6385, + "step": 18213 + }, + { + "epoch": 1.9261844331641287, + "grad_norm": 8.125, + "learning_rate": 3.520079106438064e-06, + "loss": 2.5008, + "step": 18214 + }, + { + "epoch": 1.9262901861252115, + "grad_norm": 8.3125, + "learning_rate": 3.5199227526301725e-06, + "loss": 2.5677, + "step": 18215 + }, + { + "epoch": 1.9263959390862944, + "grad_norm": 9.0, + "learning_rate": 3.5197663940361997e-06, + "loss": 2.5679, + "step": 18216 + }, + { + "epoch": 1.9265016920473772, + "grad_norm": 8.625, + "learning_rate": 3.5196100306568807e-06, + "loss": 2.2593, + "step": 18217 + }, + { + "epoch": 1.9266074450084603, + "grad_norm": 8.5, + "learning_rate": 3.5194536624929492e-06, + "loss": 2.6862, + "step": 18218 + }, + { + "epoch": 1.926713197969543, + "grad_norm": 10.625, + "learning_rate": 3.5192972895451383e-06, + "loss": 2.7965, + "step": 18219 + }, + { + "epoch": 1.9268189509306262, + "grad_norm": 8.0, + "learning_rate": 3.519140911814182e-06, + "loss": 2.5314, + "step": 18220 + }, + { + "epoch": 1.926924703891709, + "grad_norm": 9.9375, + "learning_rate": 3.5189845293008142e-06, + "loss": 2.4956, + "step": 18221 + }, + { + "epoch": 1.9270304568527918, + "grad_norm": 8.4375, + "learning_rate": 3.518828142005769e-06, + "loss": 2.6061, + "step": 18222 + }, + { + "epoch": 1.9271362098138747, + "grad_norm": 8.375, + "learning_rate": 3.5186717499297794e-06, + "loss": 2.5149, + "step": 18223 + }, + { + "epoch": 1.9272419627749577, + "grad_norm": 8.25, + "learning_rate": 3.5185153530735804e-06, + "loss": 2.4201, + "step": 18224 + }, + { + "epoch": 1.9273477157360406, + "grad_norm": 8.3125, + "learning_rate": 3.5183589514379058e-06, + "loss": 2.4903, + "step": 18225 + }, + { + "epoch": 1.9274534686971236, + "grad_norm": 8.0, + "learning_rate": 3.5182025450234885e-06, + "loss": 2.4686, + "step": 18226 + }, + { + "epoch": 1.9275592216582065, + "grad_norm": 7.96875, + "learning_rate": 3.5180461338310634e-06, + "loss": 2.58, + "step": 18227 + }, + { + "epoch": 1.9276649746192893, + "grad_norm": 8.3125, + "learning_rate": 3.517889717861364e-06, + "loss": 2.181, + "step": 18228 + }, + { + "epoch": 1.9277707275803722, + "grad_norm": 7.75, + "learning_rate": 3.517733297115124e-06, + "loss": 2.4787, + "step": 18229 + }, + { + "epoch": 1.927876480541455, + "grad_norm": 7.75, + "learning_rate": 3.5175768715930786e-06, + "loss": 2.4485, + "step": 18230 + }, + { + "epoch": 1.927982233502538, + "grad_norm": 8.625, + "learning_rate": 3.5174204412959613e-06, + "loss": 2.6021, + "step": 18231 + }, + { + "epoch": 1.9280879864636211, + "grad_norm": 8.1875, + "learning_rate": 3.517264006224505e-06, + "loss": 2.6721, + "step": 18232 + }, + { + "epoch": 1.928193739424704, + "grad_norm": 10.3125, + "learning_rate": 3.5171075663794452e-06, + "loss": 2.4301, + "step": 18233 + }, + { + "epoch": 1.9282994923857868, + "grad_norm": 8.0625, + "learning_rate": 3.516951121761516e-06, + "loss": 2.4938, + "step": 18234 + }, + { + "epoch": 1.9284052453468696, + "grad_norm": 8.25, + "learning_rate": 3.5167946723714507e-06, + "loss": 2.3967, + "step": 18235 + }, + { + "epoch": 1.9285109983079525, + "grad_norm": 9.3125, + "learning_rate": 3.516638218209984e-06, + "loss": 2.6728, + "step": 18236 + }, + { + "epoch": 1.9286167512690355, + "grad_norm": 8.9375, + "learning_rate": 3.51648175927785e-06, + "loss": 2.7181, + "step": 18237 + }, + { + "epoch": 1.9287225042301186, + "grad_norm": 8.5625, + "learning_rate": 3.5163252955757826e-06, + "loss": 2.6729, + "step": 18238 + }, + { + "epoch": 1.9288282571912014, + "grad_norm": 8.6875, + "learning_rate": 3.5161688271045162e-06, + "loss": 2.3864, + "step": 18239 + }, + { + "epoch": 1.9289340101522843, + "grad_norm": 8.3125, + "learning_rate": 3.5160123538647856e-06, + "loss": 2.7226, + "step": 18240 + }, + { + "epoch": 1.9290397631133671, + "grad_norm": 8.9375, + "learning_rate": 3.515855875857324e-06, + "loss": 2.6772, + "step": 18241 + }, + { + "epoch": 1.92914551607445, + "grad_norm": 7.84375, + "learning_rate": 3.5156993930828664e-06, + "loss": 2.6264, + "step": 18242 + }, + { + "epoch": 1.929251269035533, + "grad_norm": 8.25, + "learning_rate": 3.5155429055421474e-06, + "loss": 2.3757, + "step": 18243 + }, + { + "epoch": 1.9293570219966159, + "grad_norm": 7.75, + "learning_rate": 3.5153864132359008e-06, + "loss": 2.5409, + "step": 18244 + }, + { + "epoch": 1.929462774957699, + "grad_norm": 8.25, + "learning_rate": 3.5152299161648605e-06, + "loss": 2.6573, + "step": 18245 + }, + { + "epoch": 1.9295685279187818, + "grad_norm": 8.375, + "learning_rate": 3.5150734143297617e-06, + "loss": 2.5594, + "step": 18246 + }, + { + "epoch": 1.9296742808798646, + "grad_norm": 8.6875, + "learning_rate": 3.5149169077313384e-06, + "loss": 2.543, + "step": 18247 + }, + { + "epoch": 1.9297800338409474, + "grad_norm": 8.125, + "learning_rate": 3.5147603963703246e-06, + "loss": 2.4326, + "step": 18248 + }, + { + "epoch": 1.9298857868020305, + "grad_norm": 8.1875, + "learning_rate": 3.5146038802474563e-06, + "loss": 2.1117, + "step": 18249 + }, + { + "epoch": 1.9299915397631133, + "grad_norm": 8.4375, + "learning_rate": 3.5144473593634666e-06, + "loss": 2.3926, + "step": 18250 + }, + { + "epoch": 1.9300972927241964, + "grad_norm": 7.96875, + "learning_rate": 3.51429083371909e-06, + "loss": 2.5751, + "step": 18251 + }, + { + "epoch": 1.9302030456852792, + "grad_norm": 8.5625, + "learning_rate": 3.514134303315062e-06, + "loss": 2.5489, + "step": 18252 + }, + { + "epoch": 1.930308798646362, + "grad_norm": 8.75, + "learning_rate": 3.5139777681521165e-06, + "loss": 2.5802, + "step": 18253 + }, + { + "epoch": 1.930414551607445, + "grad_norm": 8.4375, + "learning_rate": 3.513821228230987e-06, + "loss": 2.7001, + "step": 18254 + }, + { + "epoch": 1.930520304568528, + "grad_norm": 8.5, + "learning_rate": 3.51366468355241e-06, + "loss": 2.4952, + "step": 18255 + }, + { + "epoch": 1.9306260575296108, + "grad_norm": 9.0625, + "learning_rate": 3.513508134117119e-06, + "loss": 2.5455, + "step": 18256 + }, + { + "epoch": 1.9307318104906939, + "grad_norm": 8.1875, + "learning_rate": 3.513351579925849e-06, + "loss": 2.556, + "step": 18257 + }, + { + "epoch": 1.9308375634517767, + "grad_norm": 8.75, + "learning_rate": 3.5131950209793342e-06, + "loss": 2.8375, + "step": 18258 + }, + { + "epoch": 1.9309433164128595, + "grad_norm": 8.375, + "learning_rate": 3.5130384572783093e-06, + "loss": 2.4246, + "step": 18259 + }, + { + "epoch": 1.9310490693739424, + "grad_norm": 8.5625, + "learning_rate": 3.51288188882351e-06, + "loss": 2.4159, + "step": 18260 + }, + { + "epoch": 1.9311548223350252, + "grad_norm": 8.375, + "learning_rate": 3.51272531561567e-06, + "loss": 2.4309, + "step": 18261 + }, + { + "epoch": 1.9312605752961083, + "grad_norm": 7.90625, + "learning_rate": 3.512568737655524e-06, + "loss": 2.5229, + "step": 18262 + }, + { + "epoch": 1.9313663282571913, + "grad_norm": 8.5, + "learning_rate": 3.512412154943807e-06, + "loss": 2.6732, + "step": 18263 + }, + { + "epoch": 1.9314720812182742, + "grad_norm": 8.4375, + "learning_rate": 3.512255567481254e-06, + "loss": 2.764, + "step": 18264 + }, + { + "epoch": 1.931577834179357, + "grad_norm": 8.5625, + "learning_rate": 3.5120989752685997e-06, + "loss": 2.3575, + "step": 18265 + }, + { + "epoch": 1.9316835871404399, + "grad_norm": 8.0625, + "learning_rate": 3.511942378306579e-06, + "loss": 2.5711, + "step": 18266 + }, + { + "epoch": 1.9317893401015227, + "grad_norm": 7.96875, + "learning_rate": 3.511785776595926e-06, + "loss": 2.4399, + "step": 18267 + }, + { + "epoch": 1.9318950930626058, + "grad_norm": 8.6875, + "learning_rate": 3.5116291701373766e-06, + "loss": 2.7362, + "step": 18268 + }, + { + "epoch": 1.9320008460236888, + "grad_norm": 8.4375, + "learning_rate": 3.5114725589316655e-06, + "loss": 2.6746, + "step": 18269 + }, + { + "epoch": 1.9321065989847717, + "grad_norm": 9.1875, + "learning_rate": 3.5113159429795264e-06, + "loss": 2.7871, + "step": 18270 + }, + { + "epoch": 1.9322123519458545, + "grad_norm": 8.25, + "learning_rate": 3.5111593222816965e-06, + "loss": 2.6106, + "step": 18271 + }, + { + "epoch": 1.9323181049069373, + "grad_norm": 8.1875, + "learning_rate": 3.511002696838909e-06, + "loss": 2.5778, + "step": 18272 + }, + { + "epoch": 1.9324238578680202, + "grad_norm": 8.125, + "learning_rate": 3.5108460666518985e-06, + "loss": 2.6101, + "step": 18273 + }, + { + "epoch": 1.9325296108291032, + "grad_norm": 8.625, + "learning_rate": 3.5106894317214014e-06, + "loss": 2.554, + "step": 18274 + }, + { + "epoch": 1.932635363790186, + "grad_norm": 7.71875, + "learning_rate": 3.5105327920481518e-06, + "loss": 2.4816, + "step": 18275 + }, + { + "epoch": 1.9327411167512691, + "grad_norm": 7.9375, + "learning_rate": 3.5103761476328855e-06, + "loss": 2.4771, + "step": 18276 + }, + { + "epoch": 1.932846869712352, + "grad_norm": 8.3125, + "learning_rate": 3.510219498476337e-06, + "loss": 2.7045, + "step": 18277 + }, + { + "epoch": 1.9329526226734348, + "grad_norm": 8.9375, + "learning_rate": 3.510062844579242e-06, + "loss": 2.8333, + "step": 18278 + }, + { + "epoch": 1.9330583756345177, + "grad_norm": 8.6875, + "learning_rate": 3.509906185942335e-06, + "loss": 2.3576, + "step": 18279 + }, + { + "epoch": 1.9331641285956007, + "grad_norm": 8.25, + "learning_rate": 3.5097495225663515e-06, + "loss": 2.3859, + "step": 18280 + }, + { + "epoch": 1.9332698815566836, + "grad_norm": 8.3125, + "learning_rate": 3.509592854452026e-06, + "loss": 2.5124, + "step": 18281 + }, + { + "epoch": 1.9333756345177666, + "grad_norm": 8.75, + "learning_rate": 3.5094361816000937e-06, + "loss": 2.6207, + "step": 18282 + }, + { + "epoch": 1.9334813874788495, + "grad_norm": 8.125, + "learning_rate": 3.5092795040112907e-06, + "loss": 2.5718, + "step": 18283 + }, + { + "epoch": 1.9335871404399323, + "grad_norm": 8.3125, + "learning_rate": 3.509122821686352e-06, + "loss": 2.6381, + "step": 18284 + }, + { + "epoch": 1.9336928934010151, + "grad_norm": 8.375, + "learning_rate": 3.5089661346260128e-06, + "loss": 2.6764, + "step": 18285 + }, + { + "epoch": 1.933798646362098, + "grad_norm": 7.875, + "learning_rate": 3.508809442831008e-06, + "loss": 2.5422, + "step": 18286 + }, + { + "epoch": 1.933904399323181, + "grad_norm": 8.4375, + "learning_rate": 3.508652746302073e-06, + "loss": 2.53, + "step": 18287 + }, + { + "epoch": 1.934010152284264, + "grad_norm": 8.4375, + "learning_rate": 3.508496045039943e-06, + "loss": 2.2066, + "step": 18288 + }, + { + "epoch": 1.934115905245347, + "grad_norm": 8.0, + "learning_rate": 3.5083393390453536e-06, + "loss": 2.5232, + "step": 18289 + }, + { + "epoch": 1.9342216582064298, + "grad_norm": 7.9375, + "learning_rate": 3.5081826283190403e-06, + "loss": 2.5957, + "step": 18290 + }, + { + "epoch": 1.9343274111675126, + "grad_norm": 7.6875, + "learning_rate": 3.508025912861738e-06, + "loss": 2.5286, + "step": 18291 + }, + { + "epoch": 1.9344331641285955, + "grad_norm": 8.4375, + "learning_rate": 3.5078691926741827e-06, + "loss": 2.6325, + "step": 18292 + }, + { + "epoch": 1.9345389170896785, + "grad_norm": 8.375, + "learning_rate": 3.507712467757109e-06, + "loss": 2.5535, + "step": 18293 + }, + { + "epoch": 1.9346446700507616, + "grad_norm": 8.8125, + "learning_rate": 3.5075557381112534e-06, + "loss": 2.476, + "step": 18294 + }, + { + "epoch": 1.9347504230118444, + "grad_norm": 8.4375, + "learning_rate": 3.5073990037373506e-06, + "loss": 2.5552, + "step": 18295 + }, + { + "epoch": 1.9348561759729273, + "grad_norm": 8.125, + "learning_rate": 3.5072422646361358e-06, + "loss": 2.5139, + "step": 18296 + }, + { + "epoch": 1.93496192893401, + "grad_norm": 8.6875, + "learning_rate": 3.5070855208083454e-06, + "loss": 2.6877, + "step": 18297 + }, + { + "epoch": 1.935067681895093, + "grad_norm": 8.25, + "learning_rate": 3.506928772254715e-06, + "loss": 2.5355, + "step": 18298 + }, + { + "epoch": 1.935173434856176, + "grad_norm": 7.90625, + "learning_rate": 3.5067720189759793e-06, + "loss": 2.2679, + "step": 18299 + }, + { + "epoch": 1.9352791878172588, + "grad_norm": 8.3125, + "learning_rate": 3.506615260972873e-06, + "loss": 2.7502, + "step": 18300 + }, + { + "epoch": 1.935384940778342, + "grad_norm": 8.0625, + "learning_rate": 3.506458498246135e-06, + "loss": 2.4824, + "step": 18301 + }, + { + "epoch": 1.9354906937394247, + "grad_norm": 8.5, + "learning_rate": 3.5063017307964976e-06, + "loss": 2.6587, + "step": 18302 + }, + { + "epoch": 1.9355964467005076, + "grad_norm": 8.5625, + "learning_rate": 3.5061449586246982e-06, + "loss": 2.7143, + "step": 18303 + }, + { + "epoch": 1.9357021996615904, + "grad_norm": 8.4375, + "learning_rate": 3.5059881817314724e-06, + "loss": 2.7299, + "step": 18304 + }, + { + "epoch": 1.9358079526226735, + "grad_norm": 8.5625, + "learning_rate": 3.5058314001175546e-06, + "loss": 2.8268, + "step": 18305 + }, + { + "epoch": 1.9359137055837563, + "grad_norm": 8.1875, + "learning_rate": 3.5056746137836824e-06, + "loss": 2.4995, + "step": 18306 + }, + { + "epoch": 1.9360194585448394, + "grad_norm": 9.4375, + "learning_rate": 3.50551782273059e-06, + "loss": 3.0541, + "step": 18307 + }, + { + "epoch": 1.9361252115059222, + "grad_norm": 8.25, + "learning_rate": 3.5053610269590135e-06, + "loss": 2.5016, + "step": 18308 + }, + { + "epoch": 1.936230964467005, + "grad_norm": 8.625, + "learning_rate": 3.505204226469689e-06, + "loss": 2.6234, + "step": 18309 + }, + { + "epoch": 1.9363367174280879, + "grad_norm": 8.25, + "learning_rate": 3.505047421263352e-06, + "loss": 2.462, + "step": 18310 + }, + { + "epoch": 1.936442470389171, + "grad_norm": 8.5, + "learning_rate": 3.5048906113407394e-06, + "loss": 2.5996, + "step": 18311 + }, + { + "epoch": 1.9365482233502538, + "grad_norm": 9.0625, + "learning_rate": 3.5047337967025853e-06, + "loss": 2.6578, + "step": 18312 + }, + { + "epoch": 1.9366539763113368, + "grad_norm": 8.625, + "learning_rate": 3.504576977349627e-06, + "loss": 2.6134, + "step": 18313 + }, + { + "epoch": 1.9367597292724197, + "grad_norm": 8.4375, + "learning_rate": 3.5044201532825994e-06, + "loss": 2.2069, + "step": 18314 + }, + { + "epoch": 1.9368654822335025, + "grad_norm": 8.125, + "learning_rate": 3.5042633245022384e-06, + "loss": 2.5866, + "step": 18315 + }, + { + "epoch": 1.9369712351945854, + "grad_norm": 8.4375, + "learning_rate": 3.5041064910092813e-06, + "loss": 2.6236, + "step": 18316 + }, + { + "epoch": 1.9370769881556682, + "grad_norm": 8.375, + "learning_rate": 3.5039496528044625e-06, + "loss": 2.5175, + "step": 18317 + }, + { + "epoch": 1.9371827411167513, + "grad_norm": 7.75, + "learning_rate": 3.5037928098885184e-06, + "loss": 2.6115, + "step": 18318 + }, + { + "epoch": 1.9372884940778343, + "grad_norm": 8.5625, + "learning_rate": 3.5036359622621855e-06, + "loss": 2.2558, + "step": 18319 + }, + { + "epoch": 1.9373942470389172, + "grad_norm": 8.875, + "learning_rate": 3.5034791099262e-06, + "loss": 2.7901, + "step": 18320 + }, + { + "epoch": 1.9375, + "grad_norm": 8.375, + "learning_rate": 3.5033222528812965e-06, + "loss": 2.5593, + "step": 18321 + }, + { + "epoch": 1.9376057529610828, + "grad_norm": 11.4375, + "learning_rate": 3.5031653911282125e-06, + "loss": 2.7266, + "step": 18322 + }, + { + "epoch": 1.9377115059221657, + "grad_norm": 8.5625, + "learning_rate": 3.5030085246676836e-06, + "loss": 2.7784, + "step": 18323 + }, + { + "epoch": 1.9378172588832487, + "grad_norm": 9.4375, + "learning_rate": 3.502851653500446e-06, + "loss": 2.5007, + "step": 18324 + }, + { + "epoch": 1.9379230118443318, + "grad_norm": 8.0625, + "learning_rate": 3.502694777627236e-06, + "loss": 2.5842, + "step": 18325 + }, + { + "epoch": 1.9380287648054146, + "grad_norm": 7.75, + "learning_rate": 3.5025378970487887e-06, + "loss": 2.2797, + "step": 18326 + }, + { + "epoch": 1.9381345177664975, + "grad_norm": 8.4375, + "learning_rate": 3.5023810117658414e-06, + "loss": 2.5956, + "step": 18327 + }, + { + "epoch": 1.9382402707275803, + "grad_norm": 8.25, + "learning_rate": 3.5022241217791295e-06, + "loss": 2.6819, + "step": 18328 + }, + { + "epoch": 1.9383460236886632, + "grad_norm": 8.8125, + "learning_rate": 3.502067227089391e-06, + "loss": 2.4405, + "step": 18329 + }, + { + "epoch": 1.9384517766497462, + "grad_norm": 8.375, + "learning_rate": 3.50191032769736e-06, + "loss": 2.5082, + "step": 18330 + }, + { + "epoch": 1.938557529610829, + "grad_norm": 8.0, + "learning_rate": 3.501753423603773e-06, + "loss": 2.5277, + "step": 18331 + }, + { + "epoch": 1.9386632825719121, + "grad_norm": 8.125, + "learning_rate": 3.501596514809368e-06, + "loss": 2.6421, + "step": 18332 + }, + { + "epoch": 1.938769035532995, + "grad_norm": 8.3125, + "learning_rate": 3.50143960131488e-06, + "loss": 2.5884, + "step": 18333 + }, + { + "epoch": 1.9388747884940778, + "grad_norm": 8.6875, + "learning_rate": 3.501282683121045e-06, + "loss": 2.7061, + "step": 18334 + }, + { + "epoch": 1.9389805414551606, + "grad_norm": 8.25, + "learning_rate": 3.5011257602286e-06, + "loss": 2.5133, + "step": 18335 + }, + { + "epoch": 1.9390862944162437, + "grad_norm": 8.5625, + "learning_rate": 3.5009688326382817e-06, + "loss": 2.3412, + "step": 18336 + }, + { + "epoch": 1.9391920473773265, + "grad_norm": 7.71875, + "learning_rate": 3.5008119003508258e-06, + "loss": 2.5778, + "step": 18337 + }, + { + "epoch": 1.9392978003384096, + "grad_norm": 7.9375, + "learning_rate": 3.5006549633669686e-06, + "loss": 2.5807, + "step": 18338 + }, + { + "epoch": 1.9394035532994924, + "grad_norm": 9.4375, + "learning_rate": 3.5004980216874474e-06, + "loss": 2.644, + "step": 18339 + }, + { + "epoch": 1.9395093062605753, + "grad_norm": 8.75, + "learning_rate": 3.500341075312998e-06, + "loss": 2.6515, + "step": 18340 + }, + { + "epoch": 1.939615059221658, + "grad_norm": 8.125, + "learning_rate": 3.5001841242443567e-06, + "loss": 2.4057, + "step": 18341 + }, + { + "epoch": 1.9397208121827412, + "grad_norm": 8.5, + "learning_rate": 3.500027168482262e-06, + "loss": 2.3245, + "step": 18342 + }, + { + "epoch": 1.939826565143824, + "grad_norm": 8.5625, + "learning_rate": 3.4998702080274465e-06, + "loss": 2.7062, + "step": 18343 + }, + { + "epoch": 1.939932318104907, + "grad_norm": 8.5, + "learning_rate": 3.49971324288065e-06, + "loss": 2.5414, + "step": 18344 + }, + { + "epoch": 1.94003807106599, + "grad_norm": 8.75, + "learning_rate": 3.499556273042608e-06, + "loss": 2.7247, + "step": 18345 + }, + { + "epoch": 1.9401438240270727, + "grad_norm": 8.6875, + "learning_rate": 3.499399298514058e-06, + "loss": 2.4944, + "step": 18346 + }, + { + "epoch": 1.9402495769881556, + "grad_norm": 8.625, + "learning_rate": 3.499242319295735e-06, + "loss": 2.6906, + "step": 18347 + }, + { + "epoch": 1.9403553299492384, + "grad_norm": 8.5625, + "learning_rate": 3.499085335388377e-06, + "loss": 2.2476, + "step": 18348 + }, + { + "epoch": 1.9404610829103215, + "grad_norm": 7.9375, + "learning_rate": 3.4989283467927197e-06, + "loss": 2.5196, + "step": 18349 + }, + { + "epoch": 1.9405668358714045, + "grad_norm": 7.78125, + "learning_rate": 3.4987713535095e-06, + "loss": 2.6292, + "step": 18350 + }, + { + "epoch": 1.9406725888324874, + "grad_norm": 7.90625, + "learning_rate": 3.4986143555394557e-06, + "loss": 2.4404, + "step": 18351 + }, + { + "epoch": 1.9407783417935702, + "grad_norm": 8.375, + "learning_rate": 3.4984573528833223e-06, + "loss": 2.4639, + "step": 18352 + }, + { + "epoch": 1.940884094754653, + "grad_norm": 7.8125, + "learning_rate": 3.4983003455418363e-06, + "loss": 2.34, + "step": 18353 + }, + { + "epoch": 1.940989847715736, + "grad_norm": 8.25, + "learning_rate": 3.4981433335157355e-06, + "loss": 2.5847, + "step": 18354 + }, + { + "epoch": 1.941095600676819, + "grad_norm": 8.1875, + "learning_rate": 3.4979863168057563e-06, + "loss": 2.3536, + "step": 18355 + }, + { + "epoch": 1.941201353637902, + "grad_norm": 8.5625, + "learning_rate": 3.4978292954126353e-06, + "loss": 2.5395, + "step": 18356 + }, + { + "epoch": 1.9413071065989849, + "grad_norm": 8.5, + "learning_rate": 3.4976722693371103e-06, + "loss": 2.5871, + "step": 18357 + }, + { + "epoch": 1.9414128595600677, + "grad_norm": 8.375, + "learning_rate": 3.4975152385799164e-06, + "loss": 2.4483, + "step": 18358 + }, + { + "epoch": 1.9415186125211505, + "grad_norm": 8.625, + "learning_rate": 3.497358203141792e-06, + "loss": 2.586, + "step": 18359 + }, + { + "epoch": 1.9416243654822334, + "grad_norm": 7.9375, + "learning_rate": 3.4972011630234732e-06, + "loss": 2.3179, + "step": 18360 + }, + { + "epoch": 1.9417301184433164, + "grad_norm": 8.3125, + "learning_rate": 3.4970441182256977e-06, + "loss": 2.6726, + "step": 18361 + }, + { + "epoch": 1.9418358714043993, + "grad_norm": 8.375, + "learning_rate": 3.4968870687492007e-06, + "loss": 2.4458, + "step": 18362 + }, + { + "epoch": 1.9419416243654823, + "grad_norm": 8.6875, + "learning_rate": 3.496730014594721e-06, + "loss": 2.535, + "step": 18363 + }, + { + "epoch": 1.9420473773265652, + "grad_norm": 8.125, + "learning_rate": 3.496572955762996e-06, + "loss": 2.5301, + "step": 18364 + }, + { + "epoch": 1.942153130287648, + "grad_norm": 8.1875, + "learning_rate": 3.49641589225476e-06, + "loss": 2.5114, + "step": 18365 + }, + { + "epoch": 1.9422588832487309, + "grad_norm": 7.96875, + "learning_rate": 3.496258824070753e-06, + "loss": 2.5782, + "step": 18366 + }, + { + "epoch": 1.942364636209814, + "grad_norm": 8.5, + "learning_rate": 3.49610175121171e-06, + "loss": 2.5933, + "step": 18367 + }, + { + "epoch": 1.9424703891708968, + "grad_norm": 8.1875, + "learning_rate": 3.4959446736783696e-06, + "loss": 2.6939, + "step": 18368 + }, + { + "epoch": 1.9425761421319798, + "grad_norm": 8.4375, + "learning_rate": 3.495787591471468e-06, + "loss": 2.58, + "step": 18369 + }, + { + "epoch": 1.9426818950930627, + "grad_norm": 8.3125, + "learning_rate": 3.495630504591743e-06, + "loss": 2.5502, + "step": 18370 + }, + { + "epoch": 1.9427876480541455, + "grad_norm": 9.0625, + "learning_rate": 3.49547341303993e-06, + "loss": 2.8058, + "step": 18371 + }, + { + "epoch": 1.9428934010152283, + "grad_norm": 8.375, + "learning_rate": 3.4953163168167677e-06, + "loss": 2.3929, + "step": 18372 + }, + { + "epoch": 1.9429991539763112, + "grad_norm": 8.0, + "learning_rate": 3.495159215922993e-06, + "loss": 2.3786, + "step": 18373 + }, + { + "epoch": 1.9431049069373942, + "grad_norm": 8.4375, + "learning_rate": 3.4950021103593434e-06, + "loss": 2.3415, + "step": 18374 + }, + { + "epoch": 1.9432106598984773, + "grad_norm": 9.3125, + "learning_rate": 3.494845000126556e-06, + "loss": 2.6672, + "step": 18375 + }, + { + "epoch": 1.9433164128595601, + "grad_norm": 8.5, + "learning_rate": 3.494687885225368e-06, + "loss": 2.5697, + "step": 18376 + }, + { + "epoch": 1.943422165820643, + "grad_norm": 9.5, + "learning_rate": 3.4945307656565165e-06, + "loss": 2.5935, + "step": 18377 + }, + { + "epoch": 1.9435279187817258, + "grad_norm": 9.3125, + "learning_rate": 3.494373641420738e-06, + "loss": 2.6029, + "step": 18378 + }, + { + "epoch": 1.9436336717428087, + "grad_norm": 8.6875, + "learning_rate": 3.494216512518772e-06, + "loss": 2.6686, + "step": 18379 + }, + { + "epoch": 1.9437394247038917, + "grad_norm": 9.1875, + "learning_rate": 3.4940593789513535e-06, + "loss": 2.173, + "step": 18380 + }, + { + "epoch": 1.9438451776649748, + "grad_norm": 8.4375, + "learning_rate": 3.4939022407192214e-06, + "loss": 2.4769, + "step": 18381 + }, + { + "epoch": 1.9439509306260576, + "grad_norm": 8.4375, + "learning_rate": 3.493745097823112e-06, + "loss": 2.5601, + "step": 18382 + }, + { + "epoch": 1.9440566835871405, + "grad_norm": 7.90625, + "learning_rate": 3.4935879502637635e-06, + "loss": 2.4972, + "step": 18383 + }, + { + "epoch": 1.9441624365482233, + "grad_norm": 7.78125, + "learning_rate": 3.493430798041914e-06, + "loss": 2.537, + "step": 18384 + }, + { + "epoch": 1.9442681895093061, + "grad_norm": 8.1875, + "learning_rate": 3.493273641158299e-06, + "loss": 2.6023, + "step": 18385 + }, + { + "epoch": 1.9443739424703892, + "grad_norm": 8.375, + "learning_rate": 3.4931164796136574e-06, + "loss": 2.5872, + "step": 18386 + }, + { + "epoch": 1.944479695431472, + "grad_norm": 8.1875, + "learning_rate": 3.492959313408727e-06, + "loss": 2.5951, + "step": 18387 + }, + { + "epoch": 1.944585448392555, + "grad_norm": 8.5, + "learning_rate": 3.492802142544244e-06, + "loss": 2.4708, + "step": 18388 + }, + { + "epoch": 1.944691201353638, + "grad_norm": 8.25, + "learning_rate": 3.4926449670209463e-06, + "loss": 2.6647, + "step": 18389 + }, + { + "epoch": 1.9447969543147208, + "grad_norm": 8.875, + "learning_rate": 3.492487786839572e-06, + "loss": 2.4875, + "step": 18390 + }, + { + "epoch": 1.9449027072758036, + "grad_norm": 8.25, + "learning_rate": 3.492330602000858e-06, + "loss": 2.5938, + "step": 18391 + }, + { + "epoch": 1.9450084602368867, + "grad_norm": 8.0, + "learning_rate": 3.492173412505543e-06, + "loss": 2.5085, + "step": 18392 + }, + { + "epoch": 1.9451142131979695, + "grad_norm": 8.5625, + "learning_rate": 3.492016218354364e-06, + "loss": 2.3976, + "step": 18393 + }, + { + "epoch": 1.9452199661590526, + "grad_norm": 8.3125, + "learning_rate": 3.4918590195480583e-06, + "loss": 2.5753, + "step": 18394 + }, + { + "epoch": 1.9453257191201354, + "grad_norm": 8.125, + "learning_rate": 3.491701816087364e-06, + "loss": 2.5677, + "step": 18395 + }, + { + "epoch": 1.9454314720812182, + "grad_norm": 8.3125, + "learning_rate": 3.4915446079730186e-06, + "loss": 2.6733, + "step": 18396 + }, + { + "epoch": 1.945537225042301, + "grad_norm": 8.8125, + "learning_rate": 3.49138739520576e-06, + "loss": 2.6298, + "step": 18397 + }, + { + "epoch": 1.9456429780033841, + "grad_norm": 8.0, + "learning_rate": 3.491230177786325e-06, + "loss": 2.5463, + "step": 18398 + }, + { + "epoch": 1.945748730964467, + "grad_norm": 8.75, + "learning_rate": 3.4910729557154533e-06, + "loss": 2.7379, + "step": 18399 + }, + { + "epoch": 1.94585448392555, + "grad_norm": 8.875, + "learning_rate": 3.490915728993881e-06, + "loss": 2.7655, + "step": 18400 + }, + { + "epoch": 1.9459602368866329, + "grad_norm": 8.625, + "learning_rate": 3.4907584976223467e-06, + "loss": 2.6475, + "step": 18401 + }, + { + "epoch": 1.9460659898477157, + "grad_norm": 8.875, + "learning_rate": 3.490601261601588e-06, + "loss": 2.4063, + "step": 18402 + }, + { + "epoch": 1.9461717428087986, + "grad_norm": 7.8125, + "learning_rate": 3.4904440209323432e-06, + "loss": 2.2861, + "step": 18403 + }, + { + "epoch": 1.9462774957698814, + "grad_norm": 8.125, + "learning_rate": 3.4902867756153485e-06, + "loss": 2.7068, + "step": 18404 + }, + { + "epoch": 1.9463832487309645, + "grad_norm": 8.25, + "learning_rate": 3.490129525651344e-06, + "loss": 2.5436, + "step": 18405 + }, + { + "epoch": 1.9464890016920475, + "grad_norm": 8.0, + "learning_rate": 3.4899722710410654e-06, + "loss": 2.3147, + "step": 18406 + }, + { + "epoch": 1.9465947546531304, + "grad_norm": 8.1875, + "learning_rate": 3.4898150117852524e-06, + "loss": 2.3259, + "step": 18407 + }, + { + "epoch": 1.9467005076142132, + "grad_norm": 8.375, + "learning_rate": 3.489657747884643e-06, + "loss": 2.5254, + "step": 18408 + }, + { + "epoch": 1.946806260575296, + "grad_norm": 8.3125, + "learning_rate": 3.4895004793399744e-06, + "loss": 2.6588, + "step": 18409 + }, + { + "epoch": 1.9469120135363789, + "grad_norm": 8.125, + "learning_rate": 3.489343206151984e-06, + "loss": 2.4916, + "step": 18410 + }, + { + "epoch": 1.947017766497462, + "grad_norm": 8.0625, + "learning_rate": 3.4891859283214115e-06, + "loss": 2.1569, + "step": 18411 + }, + { + "epoch": 1.947123519458545, + "grad_norm": 8.375, + "learning_rate": 3.489028645848993e-06, + "loss": 2.4973, + "step": 18412 + }, + { + "epoch": 1.9472292724196278, + "grad_norm": 8.875, + "learning_rate": 3.4888713587354684e-06, + "loss": 2.7186, + "step": 18413 + }, + { + "epoch": 1.9473350253807107, + "grad_norm": 8.0625, + "learning_rate": 3.4887140669815745e-06, + "loss": 2.544, + "step": 18414 + }, + { + "epoch": 1.9474407783417935, + "grad_norm": 7.6875, + "learning_rate": 3.4885567705880497e-06, + "loss": 2.3901, + "step": 18415 + }, + { + "epoch": 1.9475465313028764, + "grad_norm": 7.6875, + "learning_rate": 3.488399469555633e-06, + "loss": 2.569, + "step": 18416 + }, + { + "epoch": 1.9476522842639594, + "grad_norm": 8.6875, + "learning_rate": 3.4882421638850606e-06, + "loss": 2.7273, + "step": 18417 + }, + { + "epoch": 1.9477580372250423, + "grad_norm": 8.5625, + "learning_rate": 3.4880848535770727e-06, + "loss": 2.5201, + "step": 18418 + }, + { + "epoch": 1.9478637901861253, + "grad_norm": 8.125, + "learning_rate": 3.4879275386324073e-06, + "loss": 2.665, + "step": 18419 + }, + { + "epoch": 1.9479695431472082, + "grad_norm": 7.9375, + "learning_rate": 3.487770219051801e-06, + "loss": 2.4882, + "step": 18420 + }, + { + "epoch": 1.948075296108291, + "grad_norm": 7.8125, + "learning_rate": 3.487612894835994e-06, + "loss": 2.6459, + "step": 18421 + }, + { + "epoch": 1.9481810490693738, + "grad_norm": 7.84375, + "learning_rate": 3.487455565985723e-06, + "loss": 2.5654, + "step": 18422 + }, + { + "epoch": 1.948286802030457, + "grad_norm": 8.3125, + "learning_rate": 3.487298232501727e-06, + "loss": 2.67, + "step": 18423 + }, + { + "epoch": 1.9483925549915397, + "grad_norm": 8.125, + "learning_rate": 3.4871408943847435e-06, + "loss": 2.6153, + "step": 18424 + }, + { + "epoch": 1.9484983079526228, + "grad_norm": 8.875, + "learning_rate": 3.4869835516355123e-06, + "loss": 2.8934, + "step": 18425 + }, + { + "epoch": 1.9486040609137056, + "grad_norm": 8.3125, + "learning_rate": 3.48682620425477e-06, + "loss": 2.5596, + "step": 18426 + }, + { + "epoch": 1.9487098138747885, + "grad_norm": 8.25, + "learning_rate": 3.4866688522432574e-06, + "loss": 2.5886, + "step": 18427 + }, + { + "epoch": 1.9488155668358713, + "grad_norm": 7.8125, + "learning_rate": 3.486511495601711e-06, + "loss": 2.4111, + "step": 18428 + }, + { + "epoch": 1.9489213197969542, + "grad_norm": 8.125, + "learning_rate": 3.486354134330869e-06, + "loss": 2.273, + "step": 18429 + }, + { + "epoch": 1.9490270727580372, + "grad_norm": 8.5, + "learning_rate": 3.486196768431471e-06, + "loss": 2.472, + "step": 18430 + }, + { + "epoch": 1.9491328257191203, + "grad_norm": 8.125, + "learning_rate": 3.4860393979042545e-06, + "loss": 2.6776, + "step": 18431 + }, + { + "epoch": 1.9492385786802031, + "grad_norm": 7.96875, + "learning_rate": 3.4858820227499585e-06, + "loss": 2.5012, + "step": 18432 + }, + { + "epoch": 1.949344331641286, + "grad_norm": 8.3125, + "learning_rate": 3.485724642969321e-06, + "loss": 2.5842, + "step": 18433 + }, + { + "epoch": 1.9494500846023688, + "grad_norm": 8.3125, + "learning_rate": 3.485567258563082e-06, + "loss": 2.5673, + "step": 18434 + }, + { + "epoch": 1.9495558375634516, + "grad_norm": 8.3125, + "learning_rate": 3.4854098695319784e-06, + "loss": 2.5726, + "step": 18435 + }, + { + "epoch": 1.9496615905245347, + "grad_norm": 8.25, + "learning_rate": 3.4852524758767485e-06, + "loss": 2.6236, + "step": 18436 + }, + { + "epoch": 1.9497673434856178, + "grad_norm": 8.6875, + "learning_rate": 3.485095077598133e-06, + "loss": 2.4827, + "step": 18437 + }, + { + "epoch": 1.9498730964467006, + "grad_norm": 7.84375, + "learning_rate": 3.4849376746968684e-06, + "loss": 2.573, + "step": 18438 + }, + { + "epoch": 1.9499788494077834, + "grad_norm": 8.25, + "learning_rate": 3.484780267173694e-06, + "loss": 2.6631, + "step": 18439 + }, + { + "epoch": 1.9500846023688663, + "grad_norm": 8.125, + "learning_rate": 3.4846228550293493e-06, + "loss": 2.3662, + "step": 18440 + }, + { + "epoch": 1.950190355329949, + "grad_norm": 8.0625, + "learning_rate": 3.4844654382645716e-06, + "loss": 2.4835, + "step": 18441 + }, + { + "epoch": 1.9502961082910322, + "grad_norm": 8.0, + "learning_rate": 3.4843080168800995e-06, + "loss": 2.4258, + "step": 18442 + }, + { + "epoch": 1.950401861252115, + "grad_norm": 8.75, + "learning_rate": 3.4841505908766738e-06, + "loss": 2.5677, + "step": 18443 + }, + { + "epoch": 1.950507614213198, + "grad_norm": 8.875, + "learning_rate": 3.4839931602550312e-06, + "loss": 2.7019, + "step": 18444 + }, + { + "epoch": 1.950613367174281, + "grad_norm": 8.25, + "learning_rate": 3.4838357250159114e-06, + "loss": 2.6172, + "step": 18445 + }, + { + "epoch": 1.9507191201353637, + "grad_norm": 8.5, + "learning_rate": 3.4836782851600533e-06, + "loss": 2.7118, + "step": 18446 + }, + { + "epoch": 1.9508248730964466, + "grad_norm": 8.125, + "learning_rate": 3.4835208406881948e-06, + "loss": 2.5071, + "step": 18447 + }, + { + "epoch": 1.9509306260575296, + "grad_norm": 8.5, + "learning_rate": 3.483363391601075e-06, + "loss": 2.6187, + "step": 18448 + }, + { + "epoch": 1.9510363790186125, + "grad_norm": 8.25, + "learning_rate": 3.4832059378994336e-06, + "loss": 2.5343, + "step": 18449 + }, + { + "epoch": 1.9511421319796955, + "grad_norm": 8.375, + "learning_rate": 3.4830484795840086e-06, + "loss": 2.6931, + "step": 18450 + }, + { + "epoch": 1.9512478849407784, + "grad_norm": 8.125, + "learning_rate": 3.482891016655539e-06, + "loss": 2.6388, + "step": 18451 + }, + { + "epoch": 1.9513536379018612, + "grad_norm": 8.8125, + "learning_rate": 3.482733549114764e-06, + "loss": 2.4774, + "step": 18452 + }, + { + "epoch": 1.951459390862944, + "grad_norm": 9.3125, + "learning_rate": 3.482576076962422e-06, + "loss": 2.8611, + "step": 18453 + }, + { + "epoch": 1.9515651438240271, + "grad_norm": 8.125, + "learning_rate": 3.482418600199253e-06, + "loss": 2.5191, + "step": 18454 + }, + { + "epoch": 1.95167089678511, + "grad_norm": 8.375, + "learning_rate": 3.4822611188259946e-06, + "loss": 2.6333, + "step": 18455 + }, + { + "epoch": 1.951776649746193, + "grad_norm": 8.0625, + "learning_rate": 3.482103632843388e-06, + "loss": 2.5297, + "step": 18456 + }, + { + "epoch": 1.9518824027072759, + "grad_norm": 8.4375, + "learning_rate": 3.4819461422521693e-06, + "loss": 2.4222, + "step": 18457 + }, + { + "epoch": 1.9519881556683587, + "grad_norm": 8.375, + "learning_rate": 3.4817886470530788e-06, + "loss": 2.6849, + "step": 18458 + }, + { + "epoch": 1.9520939086294415, + "grad_norm": 8.5625, + "learning_rate": 3.481631147246856e-06, + "loss": 2.5865, + "step": 18459 + }, + { + "epoch": 1.9521996615905244, + "grad_norm": 8.0625, + "learning_rate": 3.48147364283424e-06, + "loss": 2.4768, + "step": 18460 + }, + { + "epoch": 1.9523054145516074, + "grad_norm": 8.5, + "learning_rate": 3.4813161338159697e-06, + "loss": 2.6698, + "step": 18461 + }, + { + "epoch": 1.9524111675126905, + "grad_norm": 8.125, + "learning_rate": 3.481158620192784e-06, + "loss": 2.5347, + "step": 18462 + }, + { + "epoch": 1.9525169204737733, + "grad_norm": 8.875, + "learning_rate": 3.481001101965422e-06, + "loss": 2.3368, + "step": 18463 + }, + { + "epoch": 1.9526226734348562, + "grad_norm": 8.625, + "learning_rate": 3.4808435791346228e-06, + "loss": 2.5036, + "step": 18464 + }, + { + "epoch": 1.952728426395939, + "grad_norm": 8.125, + "learning_rate": 3.4806860517011263e-06, + "loss": 2.7009, + "step": 18465 + }, + { + "epoch": 1.9528341793570219, + "grad_norm": 8.125, + "learning_rate": 3.480528519665671e-06, + "loss": 2.6962, + "step": 18466 + }, + { + "epoch": 1.952939932318105, + "grad_norm": 9.375, + "learning_rate": 3.4803709830289965e-06, + "loss": 2.5934, + "step": 18467 + }, + { + "epoch": 1.953045685279188, + "grad_norm": 8.375, + "learning_rate": 3.4802134417918417e-06, + "loss": 2.5944, + "step": 18468 + }, + { + "epoch": 1.9531514382402708, + "grad_norm": 9.125, + "learning_rate": 3.4800558959549468e-06, + "loss": 2.458, + "step": 18469 + }, + { + "epoch": 1.9532571912013537, + "grad_norm": 8.0625, + "learning_rate": 3.47989834551905e-06, + "loss": 2.3681, + "step": 18470 + }, + { + "epoch": 1.9533629441624365, + "grad_norm": 8.0625, + "learning_rate": 3.4797407904848906e-06, + "loss": 2.4714, + "step": 18471 + }, + { + "epoch": 1.9534686971235193, + "grad_norm": 8.6875, + "learning_rate": 3.479583230853209e-06, + "loss": 2.6263, + "step": 18472 + }, + { + "epoch": 1.9535744500846024, + "grad_norm": 8.375, + "learning_rate": 3.4794256666247438e-06, + "loss": 2.375, + "step": 18473 + }, + { + "epoch": 1.9536802030456852, + "grad_norm": 8.1875, + "learning_rate": 3.479268097800234e-06, + "loss": 2.5581, + "step": 18474 + }, + { + "epoch": 1.9537859560067683, + "grad_norm": 8.1875, + "learning_rate": 3.47911052438042e-06, + "loss": 2.466, + "step": 18475 + }, + { + "epoch": 1.9538917089678511, + "grad_norm": 8.125, + "learning_rate": 3.478952946366041e-06, + "loss": 2.4152, + "step": 18476 + }, + { + "epoch": 1.953997461928934, + "grad_norm": 9.6875, + "learning_rate": 3.478795363757835e-06, + "loss": 2.6455, + "step": 18477 + }, + { + "epoch": 1.9541032148900168, + "grad_norm": 8.125, + "learning_rate": 3.4786377765565437e-06, + "loss": 2.4136, + "step": 18478 + }, + { + "epoch": 1.9542089678510999, + "grad_norm": 9.5625, + "learning_rate": 3.4784801847629056e-06, + "loss": 2.8022, + "step": 18479 + }, + { + "epoch": 1.9543147208121827, + "grad_norm": 7.9375, + "learning_rate": 3.4783225883776602e-06, + "loss": 2.5505, + "step": 18480 + }, + { + "epoch": 1.9544204737732658, + "grad_norm": 8.875, + "learning_rate": 3.4781649874015467e-06, + "loss": 2.5353, + "step": 18481 + }, + { + "epoch": 1.9545262267343486, + "grad_norm": 8.625, + "learning_rate": 3.4780073818353053e-06, + "loss": 2.5342, + "step": 18482 + }, + { + "epoch": 1.9546319796954315, + "grad_norm": 8.125, + "learning_rate": 3.4778497716796743e-06, + "loss": 2.6151, + "step": 18483 + }, + { + "epoch": 1.9547377326565143, + "grad_norm": 8.75, + "learning_rate": 3.4776921569353955e-06, + "loss": 2.3173, + "step": 18484 + }, + { + "epoch": 1.9548434856175971, + "grad_norm": 8.0625, + "learning_rate": 3.4775345376032073e-06, + "loss": 2.5251, + "step": 18485 + }, + { + "epoch": 1.9549492385786802, + "grad_norm": 8.0625, + "learning_rate": 3.477376913683848e-06, + "loss": 2.4346, + "step": 18486 + }, + { + "epoch": 1.9550549915397633, + "grad_norm": 8.4375, + "learning_rate": 3.4772192851780594e-06, + "loss": 2.5565, + "step": 18487 + }, + { + "epoch": 1.955160744500846, + "grad_norm": 8.5, + "learning_rate": 3.4770616520865803e-06, + "loss": 2.5331, + "step": 18488 + }, + { + "epoch": 1.955266497461929, + "grad_norm": 8.3125, + "learning_rate": 3.4769040144101508e-06, + "loss": 2.7663, + "step": 18489 + }, + { + "epoch": 1.9553722504230118, + "grad_norm": 8.25, + "learning_rate": 3.47674637214951e-06, + "loss": 2.5352, + "step": 18490 + }, + { + "epoch": 1.9554780033840946, + "grad_norm": 8.375, + "learning_rate": 3.4765887253053977e-06, + "loss": 2.67, + "step": 18491 + }, + { + "epoch": 1.9555837563451777, + "grad_norm": 8.3125, + "learning_rate": 3.4764310738785547e-06, + "loss": 2.4681, + "step": 18492 + }, + { + "epoch": 1.9556895093062607, + "grad_norm": 8.125, + "learning_rate": 3.4762734178697187e-06, + "loss": 2.5109, + "step": 18493 + }, + { + "epoch": 1.9557952622673436, + "grad_norm": 9.125, + "learning_rate": 3.4761157572796318e-06, + "loss": 2.7089, + "step": 18494 + }, + { + "epoch": 1.9559010152284264, + "grad_norm": 8.125, + "learning_rate": 3.4759580921090325e-06, + "loss": 2.495, + "step": 18495 + }, + { + "epoch": 1.9560067681895092, + "grad_norm": 8.0, + "learning_rate": 3.4758004223586607e-06, + "loss": 2.5149, + "step": 18496 + }, + { + "epoch": 1.956112521150592, + "grad_norm": 8.125, + "learning_rate": 3.4756427480292577e-06, + "loss": 2.4965, + "step": 18497 + }, + { + "epoch": 1.9562182741116751, + "grad_norm": 8.4375, + "learning_rate": 3.4754850691215615e-06, + "loss": 2.7778, + "step": 18498 + }, + { + "epoch": 1.956324027072758, + "grad_norm": 8.1875, + "learning_rate": 3.475327385636313e-06, + "loss": 2.7188, + "step": 18499 + }, + { + "epoch": 1.956429780033841, + "grad_norm": 9.5625, + "learning_rate": 3.475169697574252e-06, + "loss": 2.4374, + "step": 18500 + }, + { + "epoch": 1.9565355329949239, + "grad_norm": 8.4375, + "learning_rate": 3.4750120049361185e-06, + "loss": 2.6009, + "step": 18501 + }, + { + "epoch": 1.9566412859560067, + "grad_norm": 8.25, + "learning_rate": 3.474854307722652e-06, + "loss": 2.2966, + "step": 18502 + }, + { + "epoch": 1.9567470389170896, + "grad_norm": 8.8125, + "learning_rate": 3.474696605934593e-06, + "loss": 2.5813, + "step": 18503 + }, + { + "epoch": 1.9568527918781726, + "grad_norm": 8.4375, + "learning_rate": 3.4745388995726825e-06, + "loss": 2.4069, + "step": 18504 + }, + { + "epoch": 1.9569585448392555, + "grad_norm": 8.0625, + "learning_rate": 3.4743811886376588e-06, + "loss": 2.5393, + "step": 18505 + }, + { + "epoch": 1.9570642978003385, + "grad_norm": 8.25, + "learning_rate": 3.4742234731302623e-06, + "loss": 2.4908, + "step": 18506 + }, + { + "epoch": 1.9571700507614214, + "grad_norm": 8.375, + "learning_rate": 3.474065753051234e-06, + "loss": 2.4692, + "step": 18507 + }, + { + "epoch": 1.9572758037225042, + "grad_norm": 8.0, + "learning_rate": 3.473908028401314e-06, + "loss": 2.436, + "step": 18508 + }, + { + "epoch": 1.957381556683587, + "grad_norm": 8.125, + "learning_rate": 3.4737502991812404e-06, + "loss": 2.3708, + "step": 18509 + }, + { + "epoch": 1.95748730964467, + "grad_norm": 8.125, + "learning_rate": 3.4735925653917562e-06, + "loss": 2.5399, + "step": 18510 + }, + { + "epoch": 1.957593062605753, + "grad_norm": 8.875, + "learning_rate": 3.4734348270336e-06, + "loss": 2.9338, + "step": 18511 + }, + { + "epoch": 1.957698815566836, + "grad_norm": 9.1875, + "learning_rate": 3.4732770841075115e-06, + "loss": 2.8713, + "step": 18512 + }, + { + "epoch": 1.9578045685279188, + "grad_norm": 8.0625, + "learning_rate": 3.4731193366142324e-06, + "loss": 2.4297, + "step": 18513 + }, + { + "epoch": 1.9579103214890017, + "grad_norm": 8.0625, + "learning_rate": 3.472961584554503e-06, + "loss": 2.6074, + "step": 18514 + }, + { + "epoch": 1.9580160744500845, + "grad_norm": 8.0, + "learning_rate": 3.472803827929061e-06, + "loss": 2.498, + "step": 18515 + }, + { + "epoch": 1.9581218274111674, + "grad_norm": 8.25, + "learning_rate": 3.4726460667386496e-06, + "loss": 2.4437, + "step": 18516 + }, + { + "epoch": 1.9582275803722504, + "grad_norm": 8.375, + "learning_rate": 3.472488300984008e-06, + "loss": 2.5589, + "step": 18517 + }, + { + "epoch": 1.9583333333333335, + "grad_norm": 9.3125, + "learning_rate": 3.4723305306658765e-06, + "loss": 2.1302, + "step": 18518 + }, + { + "epoch": 1.9584390862944163, + "grad_norm": 8.625, + "learning_rate": 3.472172755784996e-06, + "loss": 2.2763, + "step": 18519 + }, + { + "epoch": 1.9585448392554992, + "grad_norm": 8.1875, + "learning_rate": 3.4720149763421057e-06, + "loss": 2.5613, + "step": 18520 + }, + { + "epoch": 1.958650592216582, + "grad_norm": 9.25, + "learning_rate": 3.471857192337946e-06, + "loss": 2.8603, + "step": 18521 + }, + { + "epoch": 1.9587563451776648, + "grad_norm": 8.3125, + "learning_rate": 3.4716994037732587e-06, + "loss": 2.4613, + "step": 18522 + }, + { + "epoch": 1.958862098138748, + "grad_norm": 8.625, + "learning_rate": 3.471541610648784e-06, + "loss": 2.4197, + "step": 18523 + }, + { + "epoch": 1.958967851099831, + "grad_norm": 8.5, + "learning_rate": 3.4713838129652614e-06, + "loss": 2.6737, + "step": 18524 + }, + { + "epoch": 1.9590736040609138, + "grad_norm": 8.6875, + "learning_rate": 3.471226010723432e-06, + "loss": 2.5519, + "step": 18525 + }, + { + "epoch": 1.9591793570219966, + "grad_norm": 8.25, + "learning_rate": 3.4710682039240364e-06, + "loss": 2.4651, + "step": 18526 + }, + { + "epoch": 1.9592851099830795, + "grad_norm": 8.25, + "learning_rate": 3.4709103925678146e-06, + "loss": 2.5205, + "step": 18527 + }, + { + "epoch": 1.9593908629441623, + "grad_norm": 8.4375, + "learning_rate": 3.4707525766555067e-06, + "loss": 2.5796, + "step": 18528 + }, + { + "epoch": 1.9594966159052454, + "grad_norm": 8.375, + "learning_rate": 3.470594756187855e-06, + "loss": 2.6108, + "step": 18529 + }, + { + "epoch": 1.9596023688663282, + "grad_norm": 9.75, + "learning_rate": 3.4704369311655984e-06, + "loss": 2.3799, + "step": 18530 + }, + { + "epoch": 1.9597081218274113, + "grad_norm": 8.125, + "learning_rate": 3.4702791015894777e-06, + "loss": 2.5346, + "step": 18531 + }, + { + "epoch": 1.9598138747884941, + "grad_norm": 8.1875, + "learning_rate": 3.4701212674602347e-06, + "loss": 2.5573, + "step": 18532 + }, + { + "epoch": 1.959919627749577, + "grad_norm": 8.0625, + "learning_rate": 3.4699634287786098e-06, + "loss": 2.5661, + "step": 18533 + }, + { + "epoch": 1.9600253807106598, + "grad_norm": 8.1875, + "learning_rate": 3.4698055855453427e-06, + "loss": 2.4433, + "step": 18534 + }, + { + "epoch": 1.9601311336717429, + "grad_norm": 9.0625, + "learning_rate": 3.4696477377611748e-06, + "loss": 2.5051, + "step": 18535 + }, + { + "epoch": 1.9602368866328257, + "grad_norm": 9.0625, + "learning_rate": 3.4694898854268466e-06, + "loss": 2.733, + "step": 18536 + }, + { + "epoch": 1.9603426395939088, + "grad_norm": 8.4375, + "learning_rate": 3.4693320285430986e-06, + "loss": 2.595, + "step": 18537 + }, + { + "epoch": 1.9604483925549916, + "grad_norm": 9.0625, + "learning_rate": 3.4691741671106726e-06, + "loss": 2.8366, + "step": 18538 + }, + { + "epoch": 1.9605541455160744, + "grad_norm": 8.0625, + "learning_rate": 3.469016301130308e-06, + "loss": 2.6158, + "step": 18539 + }, + { + "epoch": 1.9606598984771573, + "grad_norm": 8.6875, + "learning_rate": 3.4688584306027462e-06, + "loss": 2.6088, + "step": 18540 + }, + { + "epoch": 1.96076565143824, + "grad_norm": 8.8125, + "learning_rate": 3.468700555528728e-06, + "loss": 2.4599, + "step": 18541 + }, + { + "epoch": 1.9608714043993232, + "grad_norm": 8.0625, + "learning_rate": 3.4685426759089945e-06, + "loss": 2.4825, + "step": 18542 + }, + { + "epoch": 1.9609771573604062, + "grad_norm": 9.25, + "learning_rate": 3.4683847917442863e-06, + "loss": 2.6115, + "step": 18543 + }, + { + "epoch": 1.961082910321489, + "grad_norm": 8.5625, + "learning_rate": 3.4682269030353445e-06, + "loss": 2.7769, + "step": 18544 + }, + { + "epoch": 1.961188663282572, + "grad_norm": 7.96875, + "learning_rate": 3.4680690097829096e-06, + "loss": 2.5198, + "step": 18545 + }, + { + "epoch": 1.9612944162436547, + "grad_norm": 7.90625, + "learning_rate": 3.467911111987723e-06, + "loss": 2.2615, + "step": 18546 + }, + { + "epoch": 1.9614001692047376, + "grad_norm": 7.875, + "learning_rate": 3.467753209650525e-06, + "loss": 2.5656, + "step": 18547 + }, + { + "epoch": 1.9615059221658206, + "grad_norm": 8.0, + "learning_rate": 3.4675953027720567e-06, + "loss": 2.5272, + "step": 18548 + }, + { + "epoch": 1.9616116751269037, + "grad_norm": 8.4375, + "learning_rate": 3.46743739135306e-06, + "loss": 2.419, + "step": 18549 + }, + { + "epoch": 1.9617174280879865, + "grad_norm": 7.9375, + "learning_rate": 3.4672794753942752e-06, + "loss": 2.5039, + "step": 18550 + }, + { + "epoch": 1.9618231810490694, + "grad_norm": 8.1875, + "learning_rate": 3.467121554896443e-06, + "loss": 2.409, + "step": 18551 + }, + { + "epoch": 1.9619289340101522, + "grad_norm": 8.3125, + "learning_rate": 3.466963629860306e-06, + "loss": 2.7168, + "step": 18552 + }, + { + "epoch": 1.962034686971235, + "grad_norm": 8.4375, + "learning_rate": 3.466805700286603e-06, + "loss": 2.4737, + "step": 18553 + }, + { + "epoch": 1.9621404399323181, + "grad_norm": 8.0625, + "learning_rate": 3.466647766176077e-06, + "loss": 2.5908, + "step": 18554 + }, + { + "epoch": 1.962246192893401, + "grad_norm": 8.625, + "learning_rate": 3.466489827529468e-06, + "loss": 2.8114, + "step": 18555 + }, + { + "epoch": 1.962351945854484, + "grad_norm": 9.125, + "learning_rate": 3.4663318843475175e-06, + "loss": 2.5253, + "step": 18556 + }, + { + "epoch": 1.9624576988155669, + "grad_norm": 8.6875, + "learning_rate": 3.4661739366309665e-06, + "loss": 2.7242, + "step": 18557 + }, + { + "epoch": 1.9625634517766497, + "grad_norm": 8.625, + "learning_rate": 3.466015984380557e-06, + "loss": 2.6266, + "step": 18558 + }, + { + "epoch": 1.9626692047377325, + "grad_norm": 8.25, + "learning_rate": 3.4658580275970287e-06, + "loss": 2.6466, + "step": 18559 + }, + { + "epoch": 1.9627749576988156, + "grad_norm": 7.9375, + "learning_rate": 3.4657000662811245e-06, + "loss": 2.6211, + "step": 18560 + }, + { + "epoch": 1.9628807106598984, + "grad_norm": 7.96875, + "learning_rate": 3.4655421004335842e-06, + "loss": 2.4827, + "step": 18561 + }, + { + "epoch": 1.9629864636209815, + "grad_norm": 8.75, + "learning_rate": 3.4653841300551504e-06, + "loss": 2.4494, + "step": 18562 + }, + { + "epoch": 1.9630922165820643, + "grad_norm": 8.375, + "learning_rate": 3.4652261551465634e-06, + "loss": 2.4497, + "step": 18563 + }, + { + "epoch": 1.9631979695431472, + "grad_norm": 8.9375, + "learning_rate": 3.4650681757085646e-06, + "loss": 2.6432, + "step": 18564 + }, + { + "epoch": 1.96330372250423, + "grad_norm": 8.1875, + "learning_rate": 3.464910191741896e-06, + "loss": 2.5329, + "step": 18565 + }, + { + "epoch": 1.963409475465313, + "grad_norm": 8.625, + "learning_rate": 3.464752203247298e-06, + "loss": 2.5878, + "step": 18566 + }, + { + "epoch": 1.963515228426396, + "grad_norm": 8.25, + "learning_rate": 3.4645942102255126e-06, + "loss": 2.4657, + "step": 18567 + }, + { + "epoch": 1.963620981387479, + "grad_norm": 8.625, + "learning_rate": 3.464436212677281e-06, + "loss": 2.5283, + "step": 18568 + }, + { + "epoch": 1.9637267343485618, + "grad_norm": 8.375, + "learning_rate": 3.464278210603345e-06, + "loss": 2.434, + "step": 18569 + }, + { + "epoch": 1.9638324873096447, + "grad_norm": 7.6875, + "learning_rate": 3.4641202040044457e-06, + "loss": 2.2532, + "step": 18570 + }, + { + "epoch": 1.9639382402707275, + "grad_norm": 8.8125, + "learning_rate": 3.463962192881325e-06, + "loss": 2.6015, + "step": 18571 + }, + { + "epoch": 1.9640439932318103, + "grad_norm": 8.5, + "learning_rate": 3.463804177234723e-06, + "loss": 2.5829, + "step": 18572 + }, + { + "epoch": 1.9641497461928934, + "grad_norm": 8.125, + "learning_rate": 3.4636461570653835e-06, + "loss": 2.5814, + "step": 18573 + }, + { + "epoch": 1.9642554991539765, + "grad_norm": 9.0625, + "learning_rate": 3.4634881323740456e-06, + "loss": 2.7585, + "step": 18574 + }, + { + "epoch": 1.9643612521150593, + "grad_norm": 8.125, + "learning_rate": 3.463330103161452e-06, + "loss": 2.5104, + "step": 18575 + }, + { + "epoch": 1.9644670050761421, + "grad_norm": 9.3125, + "learning_rate": 3.4631720694283445e-06, + "loss": 2.4257, + "step": 18576 + }, + { + "epoch": 1.964572758037225, + "grad_norm": 8.1875, + "learning_rate": 3.4630140311754645e-06, + "loss": 2.4213, + "step": 18577 + }, + { + "epoch": 1.9646785109983078, + "grad_norm": 8.0625, + "learning_rate": 3.4628559884035535e-06, + "loss": 2.4186, + "step": 18578 + }, + { + "epoch": 1.9647842639593909, + "grad_norm": 8.75, + "learning_rate": 3.462697941113353e-06, + "loss": 2.2932, + "step": 18579 + }, + { + "epoch": 1.964890016920474, + "grad_norm": 8.5, + "learning_rate": 3.4625398893056045e-06, + "loss": 2.5737, + "step": 18580 + }, + { + "epoch": 1.9649957698815568, + "grad_norm": 7.90625, + "learning_rate": 3.4623818329810503e-06, + "loss": 2.3138, + "step": 18581 + }, + { + "epoch": 1.9651015228426396, + "grad_norm": 8.5, + "learning_rate": 3.4622237721404316e-06, + "loss": 2.6736, + "step": 18582 + }, + { + "epoch": 1.9652072758037225, + "grad_norm": 8.25, + "learning_rate": 3.46206570678449e-06, + "loss": 2.4908, + "step": 18583 + }, + { + "epoch": 1.9653130287648053, + "grad_norm": 8.9375, + "learning_rate": 3.461907636913968e-06, + "loss": 2.8787, + "step": 18584 + }, + { + "epoch": 1.9654187817258884, + "grad_norm": 8.4375, + "learning_rate": 3.4617495625296068e-06, + "loss": 2.6221, + "step": 18585 + }, + { + "epoch": 1.9655245346869712, + "grad_norm": 8.375, + "learning_rate": 3.4615914836321485e-06, + "loss": 2.5397, + "step": 18586 + }, + { + "epoch": 1.9656302876480543, + "grad_norm": 8.375, + "learning_rate": 3.4614334002223343e-06, + "loss": 2.537, + "step": 18587 + }, + { + "epoch": 1.965736040609137, + "grad_norm": 8.25, + "learning_rate": 3.4612753123009064e-06, + "loss": 2.6406, + "step": 18588 + }, + { + "epoch": 1.96584179357022, + "grad_norm": 7.75, + "learning_rate": 3.4611172198686065e-06, + "loss": 2.4315, + "step": 18589 + }, + { + "epoch": 1.9659475465313028, + "grad_norm": 9.0, + "learning_rate": 3.4609591229261773e-06, + "loss": 2.5685, + "step": 18590 + }, + { + "epoch": 1.9660532994923858, + "grad_norm": 8.9375, + "learning_rate": 3.4608010214743585e-06, + "loss": 2.4931, + "step": 18591 + }, + { + "epoch": 1.9661590524534687, + "grad_norm": 8.1875, + "learning_rate": 3.4606429155138943e-06, + "loss": 2.4811, + "step": 18592 + }, + { + "epoch": 1.9662648054145517, + "grad_norm": 8.3125, + "learning_rate": 3.4604848050455263e-06, + "loss": 2.6743, + "step": 18593 + }, + { + "epoch": 1.9663705583756346, + "grad_norm": 8.625, + "learning_rate": 3.4603266900699956e-06, + "loss": 2.7214, + "step": 18594 + }, + { + "epoch": 1.9664763113367174, + "grad_norm": 8.0, + "learning_rate": 3.4601685705880438e-06, + "loss": 2.5957, + "step": 18595 + }, + { + "epoch": 1.9665820642978002, + "grad_norm": 8.6875, + "learning_rate": 3.4600104466004145e-06, + "loss": 2.637, + "step": 18596 + }, + { + "epoch": 1.966687817258883, + "grad_norm": 7.53125, + "learning_rate": 3.4598523181078486e-06, + "loss": 2.4362, + "step": 18597 + }, + { + "epoch": 1.9667935702199661, + "grad_norm": 8.3125, + "learning_rate": 3.4596941851110876e-06, + "loss": 2.5498, + "step": 18598 + }, + { + "epoch": 1.9668993231810492, + "grad_norm": 8.5625, + "learning_rate": 3.459536047610875e-06, + "loss": 2.5569, + "step": 18599 + }, + { + "epoch": 1.967005076142132, + "grad_norm": 8.1875, + "learning_rate": 3.4593779056079525e-06, + "loss": 2.5284, + "step": 18600 + }, + { + "epoch": 1.9671108291032149, + "grad_norm": 8.5, + "learning_rate": 3.4592197591030607e-06, + "loss": 2.5478, + "step": 18601 + }, + { + "epoch": 1.9672165820642977, + "grad_norm": 8.125, + "learning_rate": 3.4590616080969435e-06, + "loss": 2.3206, + "step": 18602 + }, + { + "epoch": 1.9673223350253806, + "grad_norm": 8.375, + "learning_rate": 3.4589034525903433e-06, + "loss": 2.3774, + "step": 18603 + }, + { + "epoch": 1.9674280879864636, + "grad_norm": 8.3125, + "learning_rate": 3.458745292584e-06, + "loss": 2.5602, + "step": 18604 + }, + { + "epoch": 1.9675338409475467, + "grad_norm": 8.125, + "learning_rate": 3.4585871280786577e-06, + "loss": 2.5015, + "step": 18605 + }, + { + "epoch": 1.9676395939086295, + "grad_norm": 8.1875, + "learning_rate": 3.4584289590750585e-06, + "loss": 2.5235, + "step": 18606 + }, + { + "epoch": 1.9677453468697124, + "grad_norm": 8.4375, + "learning_rate": 3.4582707855739434e-06, + "loss": 2.5533, + "step": 18607 + }, + { + "epoch": 1.9678510998307952, + "grad_norm": 8.6875, + "learning_rate": 3.458112607576056e-06, + "loss": 2.698, + "step": 18608 + }, + { + "epoch": 1.967956852791878, + "grad_norm": 8.1875, + "learning_rate": 3.457954425082138e-06, + "loss": 2.316, + "step": 18609 + }, + { + "epoch": 1.968062605752961, + "grad_norm": 8.0, + "learning_rate": 3.457796238092931e-06, + "loss": 2.5833, + "step": 18610 + }, + { + "epoch": 1.968168358714044, + "grad_norm": 8.1875, + "learning_rate": 3.4576380466091784e-06, + "loss": 2.5754, + "step": 18611 + }, + { + "epoch": 1.968274111675127, + "grad_norm": 8.4375, + "learning_rate": 3.457479850631622e-06, + "loss": 2.4382, + "step": 18612 + }, + { + "epoch": 1.9683798646362098, + "grad_norm": 8.3125, + "learning_rate": 3.4573216501610048e-06, + "loss": 2.638, + "step": 18613 + }, + { + "epoch": 1.9684856175972927, + "grad_norm": 8.25, + "learning_rate": 3.457163445198068e-06, + "loss": 2.621, + "step": 18614 + }, + { + "epoch": 1.9685913705583755, + "grad_norm": 8.625, + "learning_rate": 3.4570052357435545e-06, + "loss": 2.5947, + "step": 18615 + }, + { + "epoch": 1.9686971235194586, + "grad_norm": 8.125, + "learning_rate": 3.4568470217982077e-06, + "loss": 2.4495, + "step": 18616 + }, + { + "epoch": 1.9688028764805414, + "grad_norm": 9.125, + "learning_rate": 3.456688803362768e-06, + "loss": 2.4996, + "step": 18617 + }, + { + "epoch": 1.9689086294416245, + "grad_norm": 8.625, + "learning_rate": 3.4565305804379802e-06, + "loss": 2.488, + "step": 18618 + }, + { + "epoch": 1.9690143824027073, + "grad_norm": 8.0625, + "learning_rate": 3.456372353024585e-06, + "loss": 2.5, + "step": 18619 + }, + { + "epoch": 1.9691201353637902, + "grad_norm": 8.5625, + "learning_rate": 3.4562141211233254e-06, + "loss": 2.7058, + "step": 18620 + }, + { + "epoch": 1.969225888324873, + "grad_norm": 8.0625, + "learning_rate": 3.456055884734944e-06, + "loss": 2.4611, + "step": 18621 + }, + { + "epoch": 1.969331641285956, + "grad_norm": 7.78125, + "learning_rate": 3.455897643860184e-06, + "loss": 2.7044, + "step": 18622 + }, + { + "epoch": 1.969437394247039, + "grad_norm": 8.5625, + "learning_rate": 3.455739398499786e-06, + "loss": 2.3097, + "step": 18623 + }, + { + "epoch": 1.969543147208122, + "grad_norm": 8.375, + "learning_rate": 3.455581148654495e-06, + "loss": 2.7006, + "step": 18624 + }, + { + "epoch": 1.9696489001692048, + "grad_norm": 7.9375, + "learning_rate": 3.455422894325052e-06, + "loss": 2.4562, + "step": 18625 + }, + { + "epoch": 1.9697546531302876, + "grad_norm": 8.1875, + "learning_rate": 3.4552646355122003e-06, + "loss": 2.7496, + "step": 18626 + }, + { + "epoch": 1.9698604060913705, + "grad_norm": 8.4375, + "learning_rate": 3.455106372216682e-06, + "loss": 2.448, + "step": 18627 + }, + { + "epoch": 1.9699661590524533, + "grad_norm": 7.90625, + "learning_rate": 3.454948104439241e-06, + "loss": 2.5306, + "step": 18628 + }, + { + "epoch": 1.9700719120135364, + "grad_norm": 9.25, + "learning_rate": 3.4547898321806182e-06, + "loss": 2.5383, + "step": 18629 + }, + { + "epoch": 1.9701776649746194, + "grad_norm": 8.375, + "learning_rate": 3.4546315554415576e-06, + "loss": 2.74, + "step": 18630 + }, + { + "epoch": 1.9702834179357023, + "grad_norm": 7.90625, + "learning_rate": 3.4544732742228014e-06, + "loss": 2.6252, + "step": 18631 + }, + { + "epoch": 1.9703891708967851, + "grad_norm": 8.8125, + "learning_rate": 3.4543149885250934e-06, + "loss": 2.1451, + "step": 18632 + }, + { + "epoch": 1.970494923857868, + "grad_norm": 8.5625, + "learning_rate": 3.4541566983491742e-06, + "loss": 2.5964, + "step": 18633 + }, + { + "epoch": 1.9706006768189508, + "grad_norm": 8.25, + "learning_rate": 3.4539984036957886e-06, + "loss": 2.5707, + "step": 18634 + }, + { + "epoch": 1.9707064297800339, + "grad_norm": 8.25, + "learning_rate": 3.4538401045656783e-06, + "loss": 2.6736, + "step": 18635 + }, + { + "epoch": 1.970812182741117, + "grad_norm": 8.8125, + "learning_rate": 3.4536818009595855e-06, + "loss": 2.7347, + "step": 18636 + }, + { + "epoch": 1.9709179357021998, + "grad_norm": 8.0625, + "learning_rate": 3.4535234928782556e-06, + "loss": 2.5273, + "step": 18637 + }, + { + "epoch": 1.9710236886632826, + "grad_norm": 8.0625, + "learning_rate": 3.45336518032243e-06, + "loss": 2.4286, + "step": 18638 + }, + { + "epoch": 1.9711294416243654, + "grad_norm": 8.625, + "learning_rate": 3.4532068632928505e-06, + "loss": 2.754, + "step": 18639 + }, + { + "epoch": 1.9712351945854483, + "grad_norm": 8.3125, + "learning_rate": 3.4530485417902614e-06, + "loss": 2.5732, + "step": 18640 + }, + { + "epoch": 1.9713409475465313, + "grad_norm": 7.75, + "learning_rate": 3.4528902158154058e-06, + "loss": 2.6327, + "step": 18641 + }, + { + "epoch": 1.9714467005076142, + "grad_norm": 8.125, + "learning_rate": 3.4527318853690255e-06, + "loss": 2.4425, + "step": 18642 + }, + { + "epoch": 1.9715524534686972, + "grad_norm": 7.90625, + "learning_rate": 3.4525735504518643e-06, + "loss": 2.4744, + "step": 18643 + }, + { + "epoch": 1.97165820642978, + "grad_norm": 8.375, + "learning_rate": 3.452415211064666e-06, + "loss": 2.5972, + "step": 18644 + }, + { + "epoch": 1.971763959390863, + "grad_norm": 8.5625, + "learning_rate": 3.452256867208171e-06, + "loss": 2.5507, + "step": 18645 + }, + { + "epoch": 1.9718697123519457, + "grad_norm": 7.9375, + "learning_rate": 3.4520985188831245e-06, + "loss": 2.3707, + "step": 18646 + }, + { + "epoch": 1.9719754653130288, + "grad_norm": 8.1875, + "learning_rate": 3.451940166090269e-06, + "loss": 2.5053, + "step": 18647 + }, + { + "epoch": 1.9720812182741116, + "grad_norm": 8.375, + "learning_rate": 3.4517818088303483e-06, + "loss": 2.5554, + "step": 18648 + }, + { + "epoch": 1.9721869712351947, + "grad_norm": 8.5, + "learning_rate": 3.4516234471041043e-06, + "loss": 2.6009, + "step": 18649 + }, + { + "epoch": 1.9722927241962775, + "grad_norm": 8.1875, + "learning_rate": 3.4514650809122806e-06, + "loss": 2.5534, + "step": 18650 + }, + { + "epoch": 1.9723984771573604, + "grad_norm": 8.0, + "learning_rate": 3.451306710255621e-06, + "loss": 2.6055, + "step": 18651 + }, + { + "epoch": 1.9725042301184432, + "grad_norm": 8.125, + "learning_rate": 3.4511483351348675e-06, + "loss": 2.4521, + "step": 18652 + }, + { + "epoch": 1.972609983079526, + "grad_norm": 8.875, + "learning_rate": 3.4509899555507643e-06, + "loss": 2.5303, + "step": 18653 + }, + { + "epoch": 1.9727157360406091, + "grad_norm": 7.6875, + "learning_rate": 3.450831571504053e-06, + "loss": 2.6459, + "step": 18654 + }, + { + "epoch": 1.9728214890016922, + "grad_norm": 8.875, + "learning_rate": 3.4506731829954786e-06, + "loss": 2.65, + "step": 18655 + }, + { + "epoch": 1.972927241962775, + "grad_norm": 9.1875, + "learning_rate": 3.4505147900257844e-06, + "loss": 2.5197, + "step": 18656 + }, + { + "epoch": 1.9730329949238579, + "grad_norm": 8.5, + "learning_rate": 3.450356392595713e-06, + "loss": 2.6496, + "step": 18657 + }, + { + "epoch": 1.9731387478849407, + "grad_norm": 8.8125, + "learning_rate": 3.450197990706007e-06, + "loss": 2.7614, + "step": 18658 + }, + { + "epoch": 1.9732445008460235, + "grad_norm": 8.4375, + "learning_rate": 3.4500395843574115e-06, + "loss": 2.4861, + "step": 18659 + }, + { + "epoch": 1.9733502538071066, + "grad_norm": 8.5, + "learning_rate": 3.4498811735506684e-06, + "loss": 2.7086, + "step": 18660 + }, + { + "epoch": 1.9734560067681897, + "grad_norm": 8.1875, + "learning_rate": 3.4497227582865207e-06, + "loss": 2.4365, + "step": 18661 + }, + { + "epoch": 1.9735617597292725, + "grad_norm": 8.0625, + "learning_rate": 3.449564338565713e-06, + "loss": 2.3457, + "step": 18662 + }, + { + "epoch": 1.9736675126903553, + "grad_norm": 8.3125, + "learning_rate": 3.449405914388988e-06, + "loss": 2.5186, + "step": 18663 + }, + { + "epoch": 1.9737732656514382, + "grad_norm": 8.25, + "learning_rate": 3.44924748575709e-06, + "loss": 2.5657, + "step": 18664 + }, + { + "epoch": 1.973879018612521, + "grad_norm": 8.4375, + "learning_rate": 3.449089052670761e-06, + "loss": 2.6104, + "step": 18665 + }, + { + "epoch": 1.973984771573604, + "grad_norm": 8.375, + "learning_rate": 3.4489306151307454e-06, + "loss": 2.5744, + "step": 18666 + }, + { + "epoch": 1.974090524534687, + "grad_norm": 8.125, + "learning_rate": 3.4487721731377878e-06, + "loss": 2.3047, + "step": 18667 + }, + { + "epoch": 1.97419627749577, + "grad_norm": 8.6875, + "learning_rate": 3.448613726692629e-06, + "loss": 2.566, + "step": 18668 + }, + { + "epoch": 1.9743020304568528, + "grad_norm": 8.3125, + "learning_rate": 3.448455275796014e-06, + "loss": 2.4571, + "step": 18669 + }, + { + "epoch": 1.9744077834179357, + "grad_norm": 8.75, + "learning_rate": 3.4482968204486873e-06, + "loss": 2.3492, + "step": 18670 + }, + { + "epoch": 1.9745135363790185, + "grad_norm": 8.5625, + "learning_rate": 3.4481383606513902e-06, + "loss": 2.5141, + "step": 18671 + }, + { + "epoch": 1.9746192893401016, + "grad_norm": 8.75, + "learning_rate": 3.4479798964048675e-06, + "loss": 2.3823, + "step": 18672 + }, + { + "epoch": 1.9747250423011844, + "grad_norm": 8.1875, + "learning_rate": 3.447821427709864e-06, + "loss": 2.7159, + "step": 18673 + }, + { + "epoch": 1.9748307952622675, + "grad_norm": 8.9375, + "learning_rate": 3.4476629545671213e-06, + "loss": 2.514, + "step": 18674 + }, + { + "epoch": 1.9749365482233503, + "grad_norm": 8.9375, + "learning_rate": 3.4475044769773847e-06, + "loss": 2.2466, + "step": 18675 + }, + { + "epoch": 1.9750423011844331, + "grad_norm": 8.5625, + "learning_rate": 3.447345994941397e-06, + "loss": 2.7651, + "step": 18676 + }, + { + "epoch": 1.975148054145516, + "grad_norm": 8.0625, + "learning_rate": 3.447187508459901e-06, + "loss": 2.492, + "step": 18677 + }, + { + "epoch": 1.975253807106599, + "grad_norm": 8.3125, + "learning_rate": 3.4470290175336425e-06, + "loss": 2.5815, + "step": 18678 + }, + { + "epoch": 1.9753595600676819, + "grad_norm": 9.6875, + "learning_rate": 3.4468705221633634e-06, + "loss": 2.661, + "step": 18679 + }, + { + "epoch": 1.975465313028765, + "grad_norm": 7.90625, + "learning_rate": 3.4467120223498087e-06, + "loss": 2.4843, + "step": 18680 + }, + { + "epoch": 1.9755710659898478, + "grad_norm": 8.3125, + "learning_rate": 3.4465535180937216e-06, + "loss": 2.647, + "step": 18681 + }, + { + "epoch": 1.9756768189509306, + "grad_norm": 8.5625, + "learning_rate": 3.4463950093958453e-06, + "loss": 2.6509, + "step": 18682 + }, + { + "epoch": 1.9757825719120135, + "grad_norm": 8.5625, + "learning_rate": 3.4462364962569254e-06, + "loss": 2.784, + "step": 18683 + }, + { + "epoch": 1.9758883248730963, + "grad_norm": 8.4375, + "learning_rate": 3.4460779786777037e-06, + "loss": 2.5125, + "step": 18684 + }, + { + "epoch": 1.9759940778341794, + "grad_norm": 7.875, + "learning_rate": 3.445919456658926e-06, + "loss": 2.3553, + "step": 18685 + }, + { + "epoch": 1.9760998307952624, + "grad_norm": 8.5625, + "learning_rate": 3.4457609302013347e-06, + "loss": 2.4139, + "step": 18686 + }, + { + "epoch": 1.9762055837563453, + "grad_norm": 8.0625, + "learning_rate": 3.445602399305673e-06, + "loss": 2.4175, + "step": 18687 + }, + { + "epoch": 1.976311336717428, + "grad_norm": 8.9375, + "learning_rate": 3.4454438639726874e-06, + "loss": 2.6297, + "step": 18688 + }, + { + "epoch": 1.976417089678511, + "grad_norm": 8.625, + "learning_rate": 3.4452853242031198e-06, + "loss": 2.4986, + "step": 18689 + }, + { + "epoch": 1.9765228426395938, + "grad_norm": 8.5, + "learning_rate": 3.4451267799977147e-06, + "loss": 2.6997, + "step": 18690 + }, + { + "epoch": 1.9766285956006768, + "grad_norm": 8.0625, + "learning_rate": 3.4449682313572164e-06, + "loss": 2.5697, + "step": 18691 + }, + { + "epoch": 1.97673434856176, + "grad_norm": 8.4375, + "learning_rate": 3.4448096782823692e-06, + "loss": 2.8866, + "step": 18692 + }, + { + "epoch": 1.9768401015228427, + "grad_norm": 8.0625, + "learning_rate": 3.4446511207739158e-06, + "loss": 2.5011, + "step": 18693 + }, + { + "epoch": 1.9769458544839256, + "grad_norm": 8.5, + "learning_rate": 3.4444925588326017e-06, + "loss": 2.4344, + "step": 18694 + }, + { + "epoch": 1.9770516074450084, + "grad_norm": 8.9375, + "learning_rate": 3.44433399245917e-06, + "loss": 2.6351, + "step": 18695 + }, + { + "epoch": 1.9771573604060912, + "grad_norm": 8.4375, + "learning_rate": 3.4441754216543644e-06, + "loss": 2.5125, + "step": 18696 + }, + { + "epoch": 1.9772631133671743, + "grad_norm": 8.25, + "learning_rate": 3.4440168464189305e-06, + "loss": 2.4791, + "step": 18697 + }, + { + "epoch": 1.9773688663282571, + "grad_norm": 8.5, + "learning_rate": 3.4438582667536113e-06, + "loss": 2.5792, + "step": 18698 + }, + { + "epoch": 1.9774746192893402, + "grad_norm": 8.5625, + "learning_rate": 3.443699682659152e-06, + "loss": 2.8209, + "step": 18699 + }, + { + "epoch": 1.977580372250423, + "grad_norm": 7.71875, + "learning_rate": 3.4435410941362947e-06, + "loss": 2.4713, + "step": 18700 + }, + { + "epoch": 1.9776861252115059, + "grad_norm": 8.0625, + "learning_rate": 3.443382501185786e-06, + "loss": 2.6294, + "step": 18701 + }, + { + "epoch": 1.9777918781725887, + "grad_norm": 8.8125, + "learning_rate": 3.443223903808369e-06, + "loss": 2.5148, + "step": 18702 + }, + { + "epoch": 1.9778976311336718, + "grad_norm": 8.0, + "learning_rate": 3.4430653020047877e-06, + "loss": 2.5858, + "step": 18703 + }, + { + "epoch": 1.9780033840947546, + "grad_norm": 8.75, + "learning_rate": 3.442906695775787e-06, + "loss": 2.5581, + "step": 18704 + }, + { + "epoch": 1.9781091370558377, + "grad_norm": 8.25, + "learning_rate": 3.4427480851221107e-06, + "loss": 2.5403, + "step": 18705 + }, + { + "epoch": 1.9782148900169205, + "grad_norm": 7.96875, + "learning_rate": 3.4425894700445023e-06, + "loss": 2.4218, + "step": 18706 + }, + { + "epoch": 1.9783206429780034, + "grad_norm": 8.0625, + "learning_rate": 3.4424308505437077e-06, + "loss": 2.4338, + "step": 18707 + }, + { + "epoch": 1.9784263959390862, + "grad_norm": 8.4375, + "learning_rate": 3.4422722266204706e-06, + "loss": 2.8189, + "step": 18708 + }, + { + "epoch": 1.978532148900169, + "grad_norm": 8.75, + "learning_rate": 3.4421135982755348e-06, + "loss": 2.8906, + "step": 18709 + }, + { + "epoch": 1.978637901861252, + "grad_norm": 8.3125, + "learning_rate": 3.4419549655096464e-06, + "loss": 2.4676, + "step": 18710 + }, + { + "epoch": 1.9787436548223352, + "grad_norm": 8.1875, + "learning_rate": 3.441796328323548e-06, + "loss": 2.6198, + "step": 18711 + }, + { + "epoch": 1.978849407783418, + "grad_norm": 7.96875, + "learning_rate": 3.441637686717984e-06, + "loss": 2.4054, + "step": 18712 + }, + { + "epoch": 1.9789551607445008, + "grad_norm": 7.6875, + "learning_rate": 3.4414790406937003e-06, + "loss": 2.3256, + "step": 18713 + }, + { + "epoch": 1.9790609137055837, + "grad_norm": 8.5625, + "learning_rate": 3.44132039025144e-06, + "loss": 2.6443, + "step": 18714 + }, + { + "epoch": 1.9791666666666665, + "grad_norm": 8.0625, + "learning_rate": 3.4411617353919484e-06, + "loss": 2.5997, + "step": 18715 + }, + { + "epoch": 1.9792724196277496, + "grad_norm": 8.375, + "learning_rate": 3.441003076115969e-06, + "loss": 2.6554, + "step": 18716 + }, + { + "epoch": 1.9793781725888326, + "grad_norm": 8.25, + "learning_rate": 3.440844412424248e-06, + "loss": 2.4864, + "step": 18717 + }, + { + "epoch": 1.9794839255499155, + "grad_norm": 8.125, + "learning_rate": 3.440685744317529e-06, + "loss": 2.5769, + "step": 18718 + }, + { + "epoch": 1.9795896785109983, + "grad_norm": 8.5625, + "learning_rate": 3.4405270717965557e-06, + "loss": 2.4717, + "step": 18719 + }, + { + "epoch": 1.9796954314720812, + "grad_norm": 8.375, + "learning_rate": 3.440368394862074e-06, + "loss": 2.5434, + "step": 18720 + }, + { + "epoch": 1.979801184433164, + "grad_norm": 8.375, + "learning_rate": 3.4402097135148284e-06, + "loss": 2.6029, + "step": 18721 + }, + { + "epoch": 1.979906937394247, + "grad_norm": 8.75, + "learning_rate": 3.440051027755562e-06, + "loss": 2.5573, + "step": 18722 + }, + { + "epoch": 1.98001269035533, + "grad_norm": 8.25, + "learning_rate": 3.4398923375850214e-06, + "loss": 2.6299, + "step": 18723 + }, + { + "epoch": 1.980118443316413, + "grad_norm": 8.875, + "learning_rate": 3.4397336430039506e-06, + "loss": 2.6958, + "step": 18724 + }, + { + "epoch": 1.9802241962774958, + "grad_norm": 8.75, + "learning_rate": 3.4395749440130932e-06, + "loss": 2.6309, + "step": 18725 + }, + { + "epoch": 1.9803299492385786, + "grad_norm": 8.0, + "learning_rate": 3.439416240613196e-06, + "loss": 2.6301, + "step": 18726 + }, + { + "epoch": 1.9804357021996615, + "grad_norm": 8.375, + "learning_rate": 3.4392575328050023e-06, + "loss": 2.5979, + "step": 18727 + }, + { + "epoch": 1.9805414551607445, + "grad_norm": 8.125, + "learning_rate": 3.439098820589256e-06, + "loss": 2.1914, + "step": 18728 + }, + { + "epoch": 1.9806472081218274, + "grad_norm": 8.1875, + "learning_rate": 3.438940103966705e-06, + "loss": 2.4197, + "step": 18729 + }, + { + "epoch": 1.9807529610829104, + "grad_norm": 8.3125, + "learning_rate": 3.438781382938091e-06, + "loss": 2.6863, + "step": 18730 + }, + { + "epoch": 1.9808587140439933, + "grad_norm": 8.125, + "learning_rate": 3.43862265750416e-06, + "loss": 2.5823, + "step": 18731 + }, + { + "epoch": 1.9809644670050761, + "grad_norm": 8.125, + "learning_rate": 3.438463927665657e-06, + "loss": 2.6577, + "step": 18732 + }, + { + "epoch": 1.981070219966159, + "grad_norm": 8.125, + "learning_rate": 3.4383051934233257e-06, + "loss": 2.6147, + "step": 18733 + }, + { + "epoch": 1.981175972927242, + "grad_norm": 8.5, + "learning_rate": 3.4381464547779124e-06, + "loss": 2.3398, + "step": 18734 + }, + { + "epoch": 1.9812817258883249, + "grad_norm": 8.4375, + "learning_rate": 3.437987711730162e-06, + "loss": 2.7679, + "step": 18735 + }, + { + "epoch": 1.981387478849408, + "grad_norm": 8.625, + "learning_rate": 3.437828964280818e-06, + "loss": 2.5923, + "step": 18736 + }, + { + "epoch": 1.9814932318104908, + "grad_norm": 8.8125, + "learning_rate": 3.437670212430627e-06, + "loss": 2.3955, + "step": 18737 + }, + { + "epoch": 1.9815989847715736, + "grad_norm": 9.625, + "learning_rate": 3.437511456180333e-06, + "loss": 2.6601, + "step": 18738 + }, + { + "epoch": 1.9817047377326564, + "grad_norm": 8.5625, + "learning_rate": 3.4373526955306813e-06, + "loss": 2.4233, + "step": 18739 + }, + { + "epoch": 1.9818104906937393, + "grad_norm": 7.875, + "learning_rate": 3.4371939304824163e-06, + "loss": 2.4241, + "step": 18740 + }, + { + "epoch": 1.9819162436548223, + "grad_norm": 8.4375, + "learning_rate": 3.437035161036284e-06, + "loss": 2.4767, + "step": 18741 + }, + { + "epoch": 1.9820219966159054, + "grad_norm": 9.4375, + "learning_rate": 3.4368763871930283e-06, + "loss": 2.4726, + "step": 18742 + }, + { + "epoch": 1.9821277495769882, + "grad_norm": 8.25, + "learning_rate": 3.4367176089533953e-06, + "loss": 2.4548, + "step": 18743 + }, + { + "epoch": 1.982233502538071, + "grad_norm": 8.25, + "learning_rate": 3.4365588263181293e-06, + "loss": 2.5774, + "step": 18744 + }, + { + "epoch": 1.982339255499154, + "grad_norm": 8.4375, + "learning_rate": 3.4364000392879765e-06, + "loss": 2.6382, + "step": 18745 + }, + { + "epoch": 1.9824450084602367, + "grad_norm": 8.25, + "learning_rate": 3.436241247863681e-06, + "loss": 2.5481, + "step": 18746 + }, + { + "epoch": 1.9825507614213198, + "grad_norm": 8.5, + "learning_rate": 3.436082452045988e-06, + "loss": 2.6869, + "step": 18747 + }, + { + "epoch": 1.9826565143824029, + "grad_norm": 7.8125, + "learning_rate": 3.4359236518356432e-06, + "loss": 2.5421, + "step": 18748 + }, + { + "epoch": 1.9827622673434857, + "grad_norm": 7.5, + "learning_rate": 3.4357648472333917e-06, + "loss": 2.4941, + "step": 18749 + }, + { + "epoch": 1.9828680203045685, + "grad_norm": 8.8125, + "learning_rate": 3.4356060382399774e-06, + "loss": 2.4403, + "step": 18750 + }, + { + "epoch": 1.9829737732656514, + "grad_norm": 8.0, + "learning_rate": 3.435447224856147e-06, + "loss": 2.6095, + "step": 18751 + }, + { + "epoch": 1.9830795262267342, + "grad_norm": 8.125, + "learning_rate": 3.435288407082646e-06, + "loss": 2.5549, + "step": 18752 + }, + { + "epoch": 1.9831852791878173, + "grad_norm": 7.46875, + "learning_rate": 3.435129584920219e-06, + "loss": 2.3454, + "step": 18753 + }, + { + "epoch": 1.9832910321489001, + "grad_norm": 8.5625, + "learning_rate": 3.43497075836961e-06, + "loss": 2.4603, + "step": 18754 + }, + { + "epoch": 1.9833967851099832, + "grad_norm": 15.0625, + "learning_rate": 3.4348119274315674e-06, + "loss": 2.6291, + "step": 18755 + }, + { + "epoch": 1.983502538071066, + "grad_norm": 8.3125, + "learning_rate": 3.4346530921068337e-06, + "loss": 2.5951, + "step": 18756 + }, + { + "epoch": 1.9836082910321489, + "grad_norm": 8.1875, + "learning_rate": 3.4344942523961554e-06, + "loss": 2.5068, + "step": 18757 + }, + { + "epoch": 1.9837140439932317, + "grad_norm": 8.125, + "learning_rate": 3.4343354083002783e-06, + "loss": 2.5797, + "step": 18758 + }, + { + "epoch": 1.9838197969543148, + "grad_norm": 8.1875, + "learning_rate": 3.4341765598199472e-06, + "loss": 2.5083, + "step": 18759 + }, + { + "epoch": 1.9839255499153976, + "grad_norm": 8.5625, + "learning_rate": 3.4340177069559065e-06, + "loss": 2.4827, + "step": 18760 + }, + { + "epoch": 1.9840313028764807, + "grad_norm": 8.8125, + "learning_rate": 3.433858849708903e-06, + "loss": 2.6927, + "step": 18761 + }, + { + "epoch": 1.9841370558375635, + "grad_norm": 8.5625, + "learning_rate": 3.4336999880796827e-06, + "loss": 2.5895, + "step": 18762 + }, + { + "epoch": 1.9842428087986463, + "grad_norm": 8.0625, + "learning_rate": 3.4335411220689897e-06, + "loss": 2.5825, + "step": 18763 + }, + { + "epoch": 1.9843485617597292, + "grad_norm": 8.5, + "learning_rate": 3.4333822516775706e-06, + "loss": 2.454, + "step": 18764 + }, + { + "epoch": 1.984454314720812, + "grad_norm": 8.0625, + "learning_rate": 3.43322337690617e-06, + "loss": 2.5836, + "step": 18765 + }, + { + "epoch": 1.984560067681895, + "grad_norm": 8.1875, + "learning_rate": 3.433064497755533e-06, + "loss": 2.6169, + "step": 18766 + }, + { + "epoch": 1.9846658206429781, + "grad_norm": 8.1875, + "learning_rate": 3.4329056142264073e-06, + "loss": 2.4733, + "step": 18767 + }, + { + "epoch": 1.984771573604061, + "grad_norm": 8.3125, + "learning_rate": 3.4327467263195365e-06, + "loss": 2.4167, + "step": 18768 + }, + { + "epoch": 1.9848773265651438, + "grad_norm": 8.5, + "learning_rate": 3.432587834035666e-06, + "loss": 2.3954, + "step": 18769 + }, + { + "epoch": 1.9849830795262267, + "grad_norm": 9.0, + "learning_rate": 3.4324289373755424e-06, + "loss": 2.5634, + "step": 18770 + }, + { + "epoch": 1.9850888324873095, + "grad_norm": 8.8125, + "learning_rate": 3.432270036339912e-06, + "loss": 2.4938, + "step": 18771 + }, + { + "epoch": 1.9851945854483926, + "grad_norm": 8.5625, + "learning_rate": 3.4321111309295196e-06, + "loss": 2.493, + "step": 18772 + }, + { + "epoch": 1.9853003384094756, + "grad_norm": 8.375, + "learning_rate": 3.4319522211451104e-06, + "loss": 2.6601, + "step": 18773 + }, + { + "epoch": 1.9854060913705585, + "grad_norm": 8.5, + "learning_rate": 3.4317933069874308e-06, + "loss": 2.5698, + "step": 18774 + }, + { + "epoch": 1.9855118443316413, + "grad_norm": 8.1875, + "learning_rate": 3.431634388457227e-06, + "loss": 2.5902, + "step": 18775 + }, + { + "epoch": 1.9856175972927241, + "grad_norm": 8.3125, + "learning_rate": 3.4314754655552423e-06, + "loss": 2.6121, + "step": 18776 + }, + { + "epoch": 1.985723350253807, + "grad_norm": 8.25, + "learning_rate": 3.4313165382822256e-06, + "loss": 2.6528, + "step": 18777 + }, + { + "epoch": 1.98582910321489, + "grad_norm": 8.0625, + "learning_rate": 3.4311576066389204e-06, + "loss": 2.4392, + "step": 18778 + }, + { + "epoch": 1.9859348561759729, + "grad_norm": 8.5, + "learning_rate": 3.4309986706260733e-06, + "loss": 2.5741, + "step": 18779 + }, + { + "epoch": 1.986040609137056, + "grad_norm": 9.25, + "learning_rate": 3.4308397302444306e-06, + "loss": 2.5854, + "step": 18780 + }, + { + "epoch": 1.9861463620981388, + "grad_norm": 8.25, + "learning_rate": 3.430680785494739e-06, + "loss": 2.7644, + "step": 18781 + }, + { + "epoch": 1.9862521150592216, + "grad_norm": 7.96875, + "learning_rate": 3.430521836377741e-06, + "loss": 2.5167, + "step": 18782 + }, + { + "epoch": 1.9863578680203045, + "grad_norm": 8.0, + "learning_rate": 3.430362882894186e-06, + "loss": 2.4235, + "step": 18783 + }, + { + "epoch": 1.9864636209813875, + "grad_norm": 8.375, + "learning_rate": 3.430203925044818e-06, + "loss": 2.8133, + "step": 18784 + }, + { + "epoch": 1.9865693739424704, + "grad_norm": 8.0625, + "learning_rate": 3.4300449628303833e-06, + "loss": 2.4405, + "step": 18785 + }, + { + "epoch": 1.9866751269035534, + "grad_norm": 8.75, + "learning_rate": 3.4298859962516285e-06, + "loss": 2.2865, + "step": 18786 + }, + { + "epoch": 1.9867808798646363, + "grad_norm": 7.9375, + "learning_rate": 3.4297270253092984e-06, + "loss": 2.4958, + "step": 18787 + }, + { + "epoch": 1.986886632825719, + "grad_norm": 8.3125, + "learning_rate": 3.42956805000414e-06, + "loss": 2.7332, + "step": 18788 + }, + { + "epoch": 1.986992385786802, + "grad_norm": 8.75, + "learning_rate": 3.4294090703368987e-06, + "loss": 2.5602, + "step": 18789 + }, + { + "epoch": 1.987098138747885, + "grad_norm": 8.6875, + "learning_rate": 3.429250086308321e-06, + "loss": 2.6202, + "step": 18790 + }, + { + "epoch": 1.9872038917089678, + "grad_norm": 8.3125, + "learning_rate": 3.429091097919153e-06, + "loss": 2.6616, + "step": 18791 + }, + { + "epoch": 1.987309644670051, + "grad_norm": 8.25, + "learning_rate": 3.428932105170139e-06, + "loss": 2.5939, + "step": 18792 + }, + { + "epoch": 1.9874153976311337, + "grad_norm": 8.8125, + "learning_rate": 3.4287731080620278e-06, + "loss": 2.8323, + "step": 18793 + }, + { + "epoch": 1.9875211505922166, + "grad_norm": 8.125, + "learning_rate": 3.4286141065955648e-06, + "loss": 2.4838, + "step": 18794 + }, + { + "epoch": 1.9876269035532994, + "grad_norm": 9.0625, + "learning_rate": 3.4284551007714938e-06, + "loss": 2.6033, + "step": 18795 + }, + { + "epoch": 1.9877326565143822, + "grad_norm": 7.9375, + "learning_rate": 3.428296090590564e-06, + "loss": 2.4254, + "step": 18796 + }, + { + "epoch": 1.9878384094754653, + "grad_norm": 8.4375, + "learning_rate": 3.42813707605352e-06, + "loss": 2.4103, + "step": 18797 + }, + { + "epoch": 1.9879441624365484, + "grad_norm": 7.75, + "learning_rate": 3.427978057161108e-06, + "loss": 2.4094, + "step": 18798 + }, + { + "epoch": 1.9880499153976312, + "grad_norm": 8.375, + "learning_rate": 3.4278190339140747e-06, + "loss": 2.582, + "step": 18799 + }, + { + "epoch": 1.988155668358714, + "grad_norm": 8.25, + "learning_rate": 3.427660006313167e-06, + "loss": 2.5152, + "step": 18800 + }, + { + "epoch": 1.9882614213197969, + "grad_norm": 8.6875, + "learning_rate": 3.4275009743591294e-06, + "loss": 2.7068, + "step": 18801 + }, + { + "epoch": 1.9883671742808797, + "grad_norm": 8.0625, + "learning_rate": 3.4273419380527097e-06, + "loss": 2.0555, + "step": 18802 + }, + { + "epoch": 1.9884729272419628, + "grad_norm": 8.3125, + "learning_rate": 3.427182897394653e-06, + "loss": 2.6645, + "step": 18803 + }, + { + "epoch": 1.9885786802030458, + "grad_norm": 8.3125, + "learning_rate": 3.427023852385706e-06, + "loss": 2.5083, + "step": 18804 + }, + { + "epoch": 1.9886844331641287, + "grad_norm": 8.25, + "learning_rate": 3.426864803026615e-06, + "loss": 2.7058, + "step": 18805 + }, + { + "epoch": 1.9887901861252115, + "grad_norm": 8.25, + "learning_rate": 3.426705749318127e-06, + "loss": 2.5472, + "step": 18806 + }, + { + "epoch": 1.9888959390862944, + "grad_norm": 8.1875, + "learning_rate": 3.4265466912609884e-06, + "loss": 2.4783, + "step": 18807 + }, + { + "epoch": 1.9890016920473772, + "grad_norm": 7.84375, + "learning_rate": 3.4263876288559445e-06, + "loss": 2.4237, + "step": 18808 + }, + { + "epoch": 1.9891074450084603, + "grad_norm": 7.8125, + "learning_rate": 3.426228562103743e-06, + "loss": 2.5719, + "step": 18809 + }, + { + "epoch": 1.989213197969543, + "grad_norm": 8.625, + "learning_rate": 3.426069491005129e-06, + "loss": 2.6232, + "step": 18810 + }, + { + "epoch": 1.9893189509306262, + "grad_norm": 8.4375, + "learning_rate": 3.4259104155608496e-06, + "loss": 2.4773, + "step": 18811 + }, + { + "epoch": 1.989424703891709, + "grad_norm": 8.5625, + "learning_rate": 3.4257513357716517e-06, + "loss": 2.5983, + "step": 18812 + }, + { + "epoch": 1.9895304568527918, + "grad_norm": 8.375, + "learning_rate": 3.425592251638281e-06, + "loss": 2.43, + "step": 18813 + }, + { + "epoch": 1.9896362098138747, + "grad_norm": 8.125, + "learning_rate": 3.425433163161484e-06, + "loss": 2.5991, + "step": 18814 + }, + { + "epoch": 1.9897419627749577, + "grad_norm": 8.25, + "learning_rate": 3.4252740703420084e-06, + "loss": 2.2948, + "step": 18815 + }, + { + "epoch": 1.9898477157360406, + "grad_norm": 8.6875, + "learning_rate": 3.4251149731806006e-06, + "loss": 2.6767, + "step": 18816 + }, + { + "epoch": 1.9899534686971236, + "grad_norm": 8.5, + "learning_rate": 3.4249558716780053e-06, + "loss": 2.503, + "step": 18817 + }, + { + "epoch": 1.9900592216582065, + "grad_norm": 8.1875, + "learning_rate": 3.424796765834971e-06, + "loss": 2.6581, + "step": 18818 + }, + { + "epoch": 1.9901649746192893, + "grad_norm": 8.9375, + "learning_rate": 3.4246376556522443e-06, + "loss": 2.4979, + "step": 18819 + }, + { + "epoch": 1.9902707275803722, + "grad_norm": 8.25, + "learning_rate": 3.4244785411305705e-06, + "loss": 2.5299, + "step": 18820 + }, + { + "epoch": 1.990376480541455, + "grad_norm": 8.1875, + "learning_rate": 3.4243194222706972e-06, + "loss": 2.602, + "step": 18821 + }, + { + "epoch": 1.990482233502538, + "grad_norm": 8.5, + "learning_rate": 3.4241602990733703e-06, + "loss": 2.6599, + "step": 18822 + }, + { + "epoch": 1.9905879864636211, + "grad_norm": 8.0625, + "learning_rate": 3.424001171539338e-06, + "loss": 2.5891, + "step": 18823 + }, + { + "epoch": 1.990693739424704, + "grad_norm": 8.0, + "learning_rate": 3.423842039669345e-06, + "loss": 2.4103, + "step": 18824 + }, + { + "epoch": 1.9907994923857868, + "grad_norm": 8.5, + "learning_rate": 3.4236829034641404e-06, + "loss": 2.5436, + "step": 18825 + }, + { + "epoch": 1.9909052453468696, + "grad_norm": 8.375, + "learning_rate": 3.4235237629244695e-06, + "loss": 2.7309, + "step": 18826 + }, + { + "epoch": 1.9910109983079525, + "grad_norm": 8.375, + "learning_rate": 3.423364618051078e-06, + "loss": 2.3501, + "step": 18827 + }, + { + "epoch": 1.9911167512690355, + "grad_norm": 8.75, + "learning_rate": 3.423205468844716e-06, + "loss": 2.6803, + "step": 18828 + }, + { + "epoch": 1.9912225042301186, + "grad_norm": 8.1875, + "learning_rate": 3.423046315306127e-06, + "loss": 2.6043, + "step": 18829 + }, + { + "epoch": 1.9913282571912014, + "grad_norm": 8.0, + "learning_rate": 3.422887157436059e-06, + "loss": 2.5659, + "step": 18830 + }, + { + "epoch": 1.9914340101522843, + "grad_norm": 8.9375, + "learning_rate": 3.4227279952352595e-06, + "loss": 2.5915, + "step": 18831 + }, + { + "epoch": 1.9915397631133671, + "grad_norm": 7.90625, + "learning_rate": 3.4225688287044752e-06, + "loss": 2.4483, + "step": 18832 + }, + { + "epoch": 1.99164551607445, + "grad_norm": 8.1875, + "learning_rate": 3.4224096578444514e-06, + "loss": 2.613, + "step": 18833 + }, + { + "epoch": 1.991751269035533, + "grad_norm": 8.25, + "learning_rate": 3.4222504826559376e-06, + "loss": 2.682, + "step": 18834 + }, + { + "epoch": 1.9918570219966159, + "grad_norm": 7.96875, + "learning_rate": 3.4220913031396797e-06, + "loss": 2.6247, + "step": 18835 + }, + { + "epoch": 1.991962774957699, + "grad_norm": 8.1875, + "learning_rate": 3.4219321192964234e-06, + "loss": 2.4674, + "step": 18836 + }, + { + "epoch": 1.9920685279187818, + "grad_norm": 7.71875, + "learning_rate": 3.4217729311269175e-06, + "loss": 2.6446, + "step": 18837 + }, + { + "epoch": 1.9921742808798646, + "grad_norm": 8.6875, + "learning_rate": 3.4216137386319082e-06, + "loss": 2.6359, + "step": 18838 + }, + { + "epoch": 1.9922800338409474, + "grad_norm": 8.1875, + "learning_rate": 3.421454541812142e-06, + "loss": 2.7898, + "step": 18839 + }, + { + "epoch": 1.9923857868020305, + "grad_norm": 8.125, + "learning_rate": 3.4212953406683664e-06, + "loss": 2.673, + "step": 18840 + }, + { + "epoch": 1.9924915397631133, + "grad_norm": 8.5, + "learning_rate": 3.4211361352013288e-06, + "loss": 2.6109, + "step": 18841 + }, + { + "epoch": 1.9925972927241964, + "grad_norm": 7.9375, + "learning_rate": 3.4209769254117766e-06, + "loss": 2.5782, + "step": 18842 + }, + { + "epoch": 1.9927030456852792, + "grad_norm": 8.125, + "learning_rate": 3.420817711300456e-06, + "loss": 2.4743, + "step": 18843 + }, + { + "epoch": 1.992808798646362, + "grad_norm": 8.375, + "learning_rate": 3.4206584928681142e-06, + "loss": 2.4838, + "step": 18844 + }, + { + "epoch": 1.992914551607445, + "grad_norm": 8.5625, + "learning_rate": 3.4204992701154995e-06, + "loss": 2.4403, + "step": 18845 + }, + { + "epoch": 1.993020304568528, + "grad_norm": 8.375, + "learning_rate": 3.420340043043357e-06, + "loss": 2.5958, + "step": 18846 + }, + { + "epoch": 1.9931260575296108, + "grad_norm": 8.625, + "learning_rate": 3.420180811652436e-06, + "loss": 2.526, + "step": 18847 + }, + { + "epoch": 1.9932318104906939, + "grad_norm": 10.25, + "learning_rate": 3.420021575943482e-06, + "loss": 2.8819, + "step": 18848 + }, + { + "epoch": 1.9933375634517767, + "grad_norm": 8.1875, + "learning_rate": 3.419862335917243e-06, + "loss": 2.4418, + "step": 18849 + }, + { + "epoch": 1.9934433164128595, + "grad_norm": 8.1875, + "learning_rate": 3.4197030915744666e-06, + "loss": 2.763, + "step": 18850 + }, + { + "epoch": 1.9935490693739424, + "grad_norm": 8.75, + "learning_rate": 3.4195438429159e-06, + "loss": 2.4088, + "step": 18851 + }, + { + "epoch": 1.9936548223350252, + "grad_norm": 8.3125, + "learning_rate": 3.41938458994229e-06, + "loss": 2.6325, + "step": 18852 + }, + { + "epoch": 1.9937605752961083, + "grad_norm": 8.125, + "learning_rate": 3.4192253326543844e-06, + "loss": 2.5359, + "step": 18853 + }, + { + "epoch": 1.9938663282571913, + "grad_norm": 8.1875, + "learning_rate": 3.4190660710529305e-06, + "loss": 2.477, + "step": 18854 + }, + { + "epoch": 1.9939720812182742, + "grad_norm": 8.3125, + "learning_rate": 3.418906805138675e-06, + "loss": 2.5233, + "step": 18855 + }, + { + "epoch": 1.994077834179357, + "grad_norm": 8.125, + "learning_rate": 3.4187475349123657e-06, + "loss": 2.6206, + "step": 18856 + }, + { + "epoch": 1.9941835871404399, + "grad_norm": 8.0, + "learning_rate": 3.4185882603747493e-06, + "loss": 2.435, + "step": 18857 + }, + { + "epoch": 1.9942893401015227, + "grad_norm": 8.1875, + "learning_rate": 3.4184289815265747e-06, + "loss": 2.3679, + "step": 18858 + }, + { + "epoch": 1.9943950930626058, + "grad_norm": 9.3125, + "learning_rate": 3.418269698368588e-06, + "loss": 2.6609, + "step": 18859 + }, + { + "epoch": 1.9945008460236888, + "grad_norm": 8.125, + "learning_rate": 3.418110410901538e-06, + "loss": 2.271, + "step": 18860 + }, + { + "epoch": 1.9946065989847717, + "grad_norm": 7.875, + "learning_rate": 3.417951119126171e-06, + "loss": 2.5228, + "step": 18861 + }, + { + "epoch": 1.9947123519458545, + "grad_norm": 8.0625, + "learning_rate": 3.4177918230432345e-06, + "loss": 2.4567, + "step": 18862 + }, + { + "epoch": 1.9948181049069373, + "grad_norm": 8.0625, + "learning_rate": 3.4176325226534767e-06, + "loss": 2.5648, + "step": 18863 + }, + { + "epoch": 1.9949238578680202, + "grad_norm": 8.625, + "learning_rate": 3.417473217957645e-06, + "loss": 2.6833, + "step": 18864 + }, + { + "epoch": 1.9950296108291032, + "grad_norm": 8.6875, + "learning_rate": 3.4173139089564856e-06, + "loss": 2.6811, + "step": 18865 + }, + { + "epoch": 1.995135363790186, + "grad_norm": 8.0, + "learning_rate": 3.417154595650748e-06, + "loss": 2.5755, + "step": 18866 + }, + { + "epoch": 1.9952411167512691, + "grad_norm": 9.0, + "learning_rate": 3.4169952780411786e-06, + "loss": 2.09, + "step": 18867 + }, + { + "epoch": 1.995346869712352, + "grad_norm": 8.375, + "learning_rate": 3.416835956128526e-06, + "loss": 2.4722, + "step": 18868 + }, + { + "epoch": 1.9954526226734348, + "grad_norm": 8.4375, + "learning_rate": 3.416676629913537e-06, + "loss": 2.5302, + "step": 18869 + }, + { + "epoch": 1.9955583756345177, + "grad_norm": 8.5, + "learning_rate": 3.416517299396959e-06, + "loss": 2.4876, + "step": 18870 + }, + { + "epoch": 1.9956641285956007, + "grad_norm": 8.4375, + "learning_rate": 3.416357964579541e-06, + "loss": 2.643, + "step": 18871 + }, + { + "epoch": 1.9957698815566836, + "grad_norm": 8.8125, + "learning_rate": 3.4161986254620294e-06, + "loss": 2.5921, + "step": 18872 + }, + { + "epoch": 1.9958756345177666, + "grad_norm": 8.0, + "learning_rate": 3.4160392820451722e-06, + "loss": 2.5534, + "step": 18873 + }, + { + "epoch": 1.9959813874788495, + "grad_norm": 8.1875, + "learning_rate": 3.4158799343297174e-06, + "loss": 2.4153, + "step": 18874 + }, + { + "epoch": 1.9960871404399323, + "grad_norm": 8.125, + "learning_rate": 3.4157205823164126e-06, + "loss": 2.4406, + "step": 18875 + }, + { + "epoch": 1.9961928934010151, + "grad_norm": 8.75, + "learning_rate": 3.4155612260060056e-06, + "loss": 2.6772, + "step": 18876 + }, + { + "epoch": 1.996298646362098, + "grad_norm": 9.0, + "learning_rate": 3.415401865399245e-06, + "loss": 2.4994, + "step": 18877 + }, + { + "epoch": 1.996404399323181, + "grad_norm": 8.4375, + "learning_rate": 3.4152425004968762e-06, + "loss": 2.327, + "step": 18878 + }, + { + "epoch": 1.996510152284264, + "grad_norm": 7.9375, + "learning_rate": 3.4150831312996502e-06, + "loss": 2.3226, + "step": 18879 + }, + { + "epoch": 1.996615905245347, + "grad_norm": 8.375, + "learning_rate": 3.4149237578083127e-06, + "loss": 2.6014, + "step": 18880 + }, + { + "epoch": 1.9967216582064298, + "grad_norm": 8.4375, + "learning_rate": 3.4147643800236118e-06, + "loss": 2.5897, + "step": 18881 + }, + { + "epoch": 1.9968274111675126, + "grad_norm": 8.25, + "learning_rate": 3.414604997946297e-06, + "loss": 2.6266, + "step": 18882 + }, + { + "epoch": 1.9969331641285955, + "grad_norm": 8.5, + "learning_rate": 3.414445611577114e-06, + "loss": 2.5655, + "step": 18883 + }, + { + "epoch": 1.9970389170896785, + "grad_norm": 8.4375, + "learning_rate": 3.414286220916812e-06, + "loss": 2.7173, + "step": 18884 + }, + { + "epoch": 1.9971446700507616, + "grad_norm": 9.4375, + "learning_rate": 3.4141268259661388e-06, + "loss": 2.6139, + "step": 18885 + }, + { + "epoch": 1.9972504230118444, + "grad_norm": 8.1875, + "learning_rate": 3.413967426725842e-06, + "loss": 2.4239, + "step": 18886 + }, + { + "epoch": 1.9973561759729273, + "grad_norm": 7.875, + "learning_rate": 3.41380802319667e-06, + "loss": 2.3424, + "step": 18887 + }, + { + "epoch": 1.99746192893401, + "grad_norm": 7.6875, + "learning_rate": 3.413648615379371e-06, + "loss": 2.428, + "step": 18888 + }, + { + "epoch": 1.997567681895093, + "grad_norm": 8.5, + "learning_rate": 3.4134892032746923e-06, + "loss": 2.4276, + "step": 18889 + }, + { + "epoch": 1.997673434856176, + "grad_norm": 8.6875, + "learning_rate": 3.413329786883382e-06, + "loss": 2.3149, + "step": 18890 + }, + { + "epoch": 1.9977791878172588, + "grad_norm": 8.4375, + "learning_rate": 3.4131703662061895e-06, + "loss": 2.6263, + "step": 18891 + }, + { + "epoch": 1.997884940778342, + "grad_norm": 8.5, + "learning_rate": 3.4130109412438616e-06, + "loss": 2.6752, + "step": 18892 + }, + { + "epoch": 1.9979906937394247, + "grad_norm": 8.25, + "learning_rate": 3.412851511997146e-06, + "loss": 2.6152, + "step": 18893 + }, + { + "epoch": 1.9980964467005076, + "grad_norm": 8.3125, + "learning_rate": 3.412692078466793e-06, + "loss": 2.4817, + "step": 18894 + }, + { + "epoch": 1.9982021996615904, + "grad_norm": 8.6875, + "learning_rate": 3.412532640653548e-06, + "loss": 2.4385, + "step": 18895 + }, + { + "epoch": 1.9983079526226735, + "grad_norm": 9.0625, + "learning_rate": 3.4123731985581614e-06, + "loss": 2.7487, + "step": 18896 + }, + { + "epoch": 1.9984137055837563, + "grad_norm": 8.5625, + "learning_rate": 3.4122137521813803e-06, + "loss": 2.6291, + "step": 18897 + }, + { + "epoch": 1.9985194585448394, + "grad_norm": 8.4375, + "learning_rate": 3.412054301523953e-06, + "loss": 2.5574, + "step": 18898 + }, + { + "epoch": 1.9986252115059222, + "grad_norm": 9.3125, + "learning_rate": 3.411894846586629e-06, + "loss": 2.8293, + "step": 18899 + }, + { + "epoch": 1.998730964467005, + "grad_norm": 9.1875, + "learning_rate": 3.411735387370154e-06, + "loss": 2.5893, + "step": 18900 + }, + { + "epoch": 1.9988367174280879, + "grad_norm": 8.1875, + "learning_rate": 3.4115759238752778e-06, + "loss": 2.61, + "step": 18901 + }, + { + "epoch": 1.998942470389171, + "grad_norm": 8.4375, + "learning_rate": 3.4114164561027485e-06, + "loss": 2.5845, + "step": 18902 + }, + { + "epoch": 1.9990482233502538, + "grad_norm": 8.8125, + "learning_rate": 3.411256984053315e-06, + "loss": 2.6676, + "step": 18903 + }, + { + "epoch": 1.9991539763113368, + "grad_norm": 8.0625, + "learning_rate": 3.411097507727725e-06, + "loss": 2.3385, + "step": 18904 + }, + { + "epoch": 1.9992597292724197, + "grad_norm": 8.1875, + "learning_rate": 3.4109380271267278e-06, + "loss": 2.2624, + "step": 18905 + }, + { + "epoch": 1.9993654822335025, + "grad_norm": 8.0625, + "learning_rate": 3.4107785422510698e-06, + "loss": 2.5863, + "step": 18906 + }, + { + "epoch": 1.9994712351945854, + "grad_norm": 8.3125, + "learning_rate": 3.4106190531015016e-06, + "loss": 2.718, + "step": 18907 + }, + { + "epoch": 1.9995769881556682, + "grad_norm": 8.5625, + "learning_rate": 3.4104595596787704e-06, + "loss": 2.4912, + "step": 18908 + }, + { + "epoch": 1.9996827411167513, + "grad_norm": 8.5, + "learning_rate": 3.4103000619836245e-06, + "loss": 2.1543, + "step": 18909 + }, + { + "epoch": 1.9997884940778343, + "grad_norm": 8.5625, + "learning_rate": 3.4101405600168127e-06, + "loss": 2.7885, + "step": 18910 + }, + { + "epoch": 1.9998942470389172, + "grad_norm": 7.8125, + "learning_rate": 3.4099810537790833e-06, + "loss": 2.3579, + "step": 18911 + }, + { + "epoch": 2.0, + "grad_norm": 8.625, + "learning_rate": 3.4098215432711855e-06, + "loss": 2.7531, + "step": 18912 + }, + { + "epoch": 2.000105752961083, + "grad_norm": 8.3125, + "learning_rate": 3.409662028493867e-06, + "loss": 2.4423, + "step": 18913 + }, + { + "epoch": 2.0002115059221657, + "grad_norm": 8.25, + "learning_rate": 3.4095025094478774e-06, + "loss": 2.3501, + "step": 18914 + }, + { + "epoch": 2.0003172588832485, + "grad_norm": 8.3125, + "learning_rate": 3.4093429861339638e-06, + "loss": 2.7335, + "step": 18915 + }, + { + "epoch": 2.000423011844332, + "grad_norm": 8.375, + "learning_rate": 3.4091834585528747e-06, + "loss": 2.6761, + "step": 18916 + }, + { + "epoch": 2.0005287648054146, + "grad_norm": 8.75, + "learning_rate": 3.409023926705361e-06, + "loss": 2.4003, + "step": 18917 + }, + { + "epoch": 2.0006345177664975, + "grad_norm": 8.1875, + "learning_rate": 3.408864390592169e-06, + "loss": 2.492, + "step": 18918 + }, + { + "epoch": 2.0007402707275803, + "grad_norm": 7.90625, + "learning_rate": 3.4087048502140487e-06, + "loss": 2.5814, + "step": 18919 + }, + { + "epoch": 2.000846023688663, + "grad_norm": 8.0, + "learning_rate": 3.4085453055717464e-06, + "loss": 2.4463, + "step": 18920 + }, + { + "epoch": 2.000951776649746, + "grad_norm": 8.125, + "learning_rate": 3.4083857566660146e-06, + "loss": 2.5776, + "step": 18921 + }, + { + "epoch": 2.0010575296108293, + "grad_norm": 7.53125, + "learning_rate": 3.408226203497599e-06, + "loss": 2.1436, + "step": 18922 + }, + { + "epoch": 2.001163282571912, + "grad_norm": 8.0625, + "learning_rate": 3.4080666460672497e-06, + "loss": 2.4826, + "step": 18923 + }, + { + "epoch": 2.001269035532995, + "grad_norm": 8.25, + "learning_rate": 3.407907084375715e-06, + "loss": 2.6023, + "step": 18924 + }, + { + "epoch": 2.001374788494078, + "grad_norm": 8.0625, + "learning_rate": 3.407747518423743e-06, + "loss": 2.472, + "step": 18925 + }, + { + "epoch": 2.0014805414551606, + "grad_norm": 8.625, + "learning_rate": 3.407587948212084e-06, + "loss": 2.5874, + "step": 18926 + }, + { + "epoch": 2.0015862944162435, + "grad_norm": 8.5, + "learning_rate": 3.407428373741486e-06, + "loss": 2.5189, + "step": 18927 + }, + { + "epoch": 2.0016920473773268, + "grad_norm": 8.8125, + "learning_rate": 3.407268795012697e-06, + "loss": 2.468, + "step": 18928 + }, + { + "epoch": 2.0017978003384096, + "grad_norm": 7.96875, + "learning_rate": 3.4071092120264663e-06, + "loss": 2.6495, + "step": 18929 + }, + { + "epoch": 2.0019035532994924, + "grad_norm": 8.375, + "learning_rate": 3.4069496247835444e-06, + "loss": 2.5045, + "step": 18930 + }, + { + "epoch": 2.0020093062605753, + "grad_norm": 8.1875, + "learning_rate": 3.4067900332846783e-06, + "loss": 2.5601, + "step": 18931 + }, + { + "epoch": 2.002115059221658, + "grad_norm": 8.625, + "learning_rate": 3.406630437530617e-06, + "loss": 2.6531, + "step": 18932 + }, + { + "epoch": 2.002220812182741, + "grad_norm": 8.5, + "learning_rate": 3.40647083752211e-06, + "loss": 2.4165, + "step": 18933 + }, + { + "epoch": 2.0023265651438242, + "grad_norm": 7.9375, + "learning_rate": 3.406311233259907e-06, + "loss": 2.5356, + "step": 18934 + }, + { + "epoch": 2.002432318104907, + "grad_norm": 8.3125, + "learning_rate": 3.4061516247447545e-06, + "loss": 2.5973, + "step": 18935 + }, + { + "epoch": 2.00253807106599, + "grad_norm": 8.25, + "learning_rate": 3.4059920119774045e-06, + "loss": 2.4277, + "step": 18936 + }, + { + "epoch": 2.0026438240270727, + "grad_norm": 8.1875, + "learning_rate": 3.4058323949586038e-06, + "loss": 2.2825, + "step": 18937 + }, + { + "epoch": 2.0027495769881556, + "grad_norm": 7.96875, + "learning_rate": 3.405672773689102e-06, + "loss": 2.6016, + "step": 18938 + }, + { + "epoch": 2.0028553299492384, + "grad_norm": 8.1875, + "learning_rate": 3.405513148169648e-06, + "loss": 2.3445, + "step": 18939 + }, + { + "epoch": 2.0029610829103217, + "grad_norm": 8.0625, + "learning_rate": 3.4053535184009924e-06, + "loss": 2.4979, + "step": 18940 + }, + { + "epoch": 2.0030668358714045, + "grad_norm": 8.4375, + "learning_rate": 3.4051938843838815e-06, + "loss": 2.6146, + "step": 18941 + }, + { + "epoch": 2.0031725888324874, + "grad_norm": 8.1875, + "learning_rate": 3.4050342461190674e-06, + "loss": 2.5147, + "step": 18942 + }, + { + "epoch": 2.0032783417935702, + "grad_norm": 8.5, + "learning_rate": 3.4048746036072965e-06, + "loss": 2.4878, + "step": 18943 + }, + { + "epoch": 2.003384094754653, + "grad_norm": 8.25, + "learning_rate": 3.4047149568493198e-06, + "loss": 2.5791, + "step": 18944 + }, + { + "epoch": 2.003489847715736, + "grad_norm": 9.0625, + "learning_rate": 3.4045553058458856e-06, + "loss": 2.5904, + "step": 18945 + }, + { + "epoch": 2.0035956006768187, + "grad_norm": 8.0, + "learning_rate": 3.404395650597743e-06, + "loss": 2.4394, + "step": 18946 + }, + { + "epoch": 2.003701353637902, + "grad_norm": 7.8125, + "learning_rate": 3.404235991105642e-06, + "loss": 2.4768, + "step": 18947 + }, + { + "epoch": 2.003807106598985, + "grad_norm": 7.78125, + "learning_rate": 3.4040763273703316e-06, + "loss": 2.4901, + "step": 18948 + }, + { + "epoch": 2.0039128595600677, + "grad_norm": 8.1875, + "learning_rate": 3.40391665939256e-06, + "loss": 2.5123, + "step": 18949 + }, + { + "epoch": 2.0040186125211505, + "grad_norm": 9.25, + "learning_rate": 3.403756987173078e-06, + "loss": 2.5201, + "step": 18950 + }, + { + "epoch": 2.0041243654822334, + "grad_norm": 8.1875, + "learning_rate": 3.403597310712633e-06, + "loss": 2.4775, + "step": 18951 + }, + { + "epoch": 2.004230118443316, + "grad_norm": 8.8125, + "learning_rate": 3.4034376300119764e-06, + "loss": 2.4184, + "step": 18952 + }, + { + "epoch": 2.0043358714043995, + "grad_norm": 8.875, + "learning_rate": 3.403277945071857e-06, + "loss": 2.5725, + "step": 18953 + }, + { + "epoch": 2.0044416243654823, + "grad_norm": 8.5625, + "learning_rate": 3.403118255893022e-06, + "loss": 2.5805, + "step": 18954 + }, + { + "epoch": 2.004547377326565, + "grad_norm": 8.25, + "learning_rate": 3.4029585624762228e-06, + "loss": 2.3673, + "step": 18955 + }, + { + "epoch": 2.004653130287648, + "grad_norm": 8.5, + "learning_rate": 3.402798864822209e-06, + "loss": 2.5699, + "step": 18956 + }, + { + "epoch": 2.004758883248731, + "grad_norm": 7.90625, + "learning_rate": 3.402639162931729e-06, + "loss": 2.3828, + "step": 18957 + }, + { + "epoch": 2.0048646362098137, + "grad_norm": 8.1875, + "learning_rate": 3.4024794568055326e-06, + "loss": 2.4644, + "step": 18958 + }, + { + "epoch": 2.004970389170897, + "grad_norm": 8.1875, + "learning_rate": 3.402319746444369e-06, + "loss": 2.5484, + "step": 18959 + }, + { + "epoch": 2.00507614213198, + "grad_norm": 7.90625, + "learning_rate": 3.4021600318489877e-06, + "loss": 2.5008, + "step": 18960 + }, + { + "epoch": 2.0051818950930627, + "grad_norm": 8.5625, + "learning_rate": 3.4020003130201394e-06, + "loss": 2.4606, + "step": 18961 + }, + { + "epoch": 2.0052876480541455, + "grad_norm": 8.75, + "learning_rate": 3.4018405899585723e-06, + "loss": 2.6531, + "step": 18962 + }, + { + "epoch": 2.0053934010152283, + "grad_norm": 8.6875, + "learning_rate": 3.401680862665035e-06, + "loss": 2.5399, + "step": 18963 + }, + { + "epoch": 2.005499153976311, + "grad_norm": 8.25, + "learning_rate": 3.4015211311402785e-06, + "loss": 2.557, + "step": 18964 + }, + { + "epoch": 2.0056049069373945, + "grad_norm": 8.875, + "learning_rate": 3.4013613953850527e-06, + "loss": 2.6853, + "step": 18965 + }, + { + "epoch": 2.0057106598984773, + "grad_norm": 8.4375, + "learning_rate": 3.4012016554001065e-06, + "loss": 2.4679, + "step": 18966 + }, + { + "epoch": 2.00581641285956, + "grad_norm": 8.0, + "learning_rate": 3.401041911186189e-06, + "loss": 2.4199, + "step": 18967 + }, + { + "epoch": 2.005922165820643, + "grad_norm": 9.0625, + "learning_rate": 3.400882162744051e-06, + "loss": 2.5544, + "step": 18968 + }, + { + "epoch": 2.006027918781726, + "grad_norm": 8.3125, + "learning_rate": 3.400722410074441e-06, + "loss": 2.5883, + "step": 18969 + }, + { + "epoch": 2.0061336717428087, + "grad_norm": 8.125, + "learning_rate": 3.400562653178109e-06, + "loss": 2.4237, + "step": 18970 + }, + { + "epoch": 2.0062394247038915, + "grad_norm": 9.0625, + "learning_rate": 3.400402892055805e-06, + "loss": 2.9111, + "step": 18971 + }, + { + "epoch": 2.0063451776649748, + "grad_norm": 9.0625, + "learning_rate": 3.400243126708279e-06, + "loss": 2.7955, + "step": 18972 + }, + { + "epoch": 2.0064509306260576, + "grad_norm": 7.84375, + "learning_rate": 3.400083357136279e-06, + "loss": 2.5987, + "step": 18973 + }, + { + "epoch": 2.0065566835871405, + "grad_norm": 8.3125, + "learning_rate": 3.3999235833405563e-06, + "loss": 2.54, + "step": 18974 + }, + { + "epoch": 2.0066624365482233, + "grad_norm": 9.1875, + "learning_rate": 3.3997638053218614e-06, + "loss": 2.7471, + "step": 18975 + }, + { + "epoch": 2.006768189509306, + "grad_norm": 8.25, + "learning_rate": 3.399604023080941e-06, + "loss": 2.4753, + "step": 18976 + }, + { + "epoch": 2.006873942470389, + "grad_norm": 8.5, + "learning_rate": 3.3994442366185487e-06, + "loss": 2.3446, + "step": 18977 + }, + { + "epoch": 2.0069796954314723, + "grad_norm": 8.3125, + "learning_rate": 3.399284445935432e-06, + "loss": 2.4052, + "step": 18978 + }, + { + "epoch": 2.007085448392555, + "grad_norm": 8.6875, + "learning_rate": 3.3991246510323405e-06, + "loss": 2.7263, + "step": 18979 + }, + { + "epoch": 2.007191201353638, + "grad_norm": 7.90625, + "learning_rate": 3.3989648519100247e-06, + "loss": 2.1737, + "step": 18980 + }, + { + "epoch": 2.0072969543147208, + "grad_norm": 8.3125, + "learning_rate": 3.3988050485692354e-06, + "loss": 2.6632, + "step": 18981 + }, + { + "epoch": 2.0074027072758036, + "grad_norm": 9.1875, + "learning_rate": 3.3986452410107214e-06, + "loss": 2.5873, + "step": 18982 + }, + { + "epoch": 2.0075084602368864, + "grad_norm": 8.0625, + "learning_rate": 3.398485429235232e-06, + "loss": 2.3563, + "step": 18983 + }, + { + "epoch": 2.0076142131979697, + "grad_norm": 8.625, + "learning_rate": 3.3983256132435188e-06, + "loss": 2.4058, + "step": 18984 + }, + { + "epoch": 2.0077199661590526, + "grad_norm": 8.625, + "learning_rate": 3.398165793036331e-06, + "loss": 2.3707, + "step": 18985 + }, + { + "epoch": 2.0078257191201354, + "grad_norm": 8.125, + "learning_rate": 3.3980059686144173e-06, + "loss": 2.4632, + "step": 18986 + }, + { + "epoch": 2.0079314720812182, + "grad_norm": 8.1875, + "learning_rate": 3.3978461399785302e-06, + "loss": 2.5217, + "step": 18987 + }, + { + "epoch": 2.008037225042301, + "grad_norm": 7.96875, + "learning_rate": 3.3976863071294182e-06, + "loss": 2.5084, + "step": 18988 + }, + { + "epoch": 2.008142978003384, + "grad_norm": 8.75, + "learning_rate": 3.3975264700678306e-06, + "loss": 2.8383, + "step": 18989 + }, + { + "epoch": 2.008248730964467, + "grad_norm": 7.90625, + "learning_rate": 3.3973666287945185e-06, + "loss": 2.4056, + "step": 18990 + }, + { + "epoch": 2.00835448392555, + "grad_norm": 7.8125, + "learning_rate": 3.397206783310232e-06, + "loss": 2.4919, + "step": 18991 + }, + { + "epoch": 2.008460236886633, + "grad_norm": 8.375, + "learning_rate": 3.397046933615722e-06, + "loss": 2.4473, + "step": 18992 + }, + { + "epoch": 2.0085659898477157, + "grad_norm": 8.3125, + "learning_rate": 3.3968870797117367e-06, + "loss": 2.4733, + "step": 18993 + }, + { + "epoch": 2.0086717428087986, + "grad_norm": 8.375, + "learning_rate": 3.396727221599027e-06, + "loss": 2.1065, + "step": 18994 + }, + { + "epoch": 2.0087774957698814, + "grad_norm": 8.1875, + "learning_rate": 3.396567359278343e-06, + "loss": 2.3591, + "step": 18995 + }, + { + "epoch": 2.0088832487309647, + "grad_norm": 8.25, + "learning_rate": 3.396407492750436e-06, + "loss": 2.6628, + "step": 18996 + }, + { + "epoch": 2.0089890016920475, + "grad_norm": 8.6875, + "learning_rate": 3.396247622016055e-06, + "loss": 2.7359, + "step": 18997 + }, + { + "epoch": 2.0090947546531304, + "grad_norm": 8.25, + "learning_rate": 3.39608774707595e-06, + "loss": 2.338, + "step": 18998 + }, + { + "epoch": 2.009200507614213, + "grad_norm": 8.0625, + "learning_rate": 3.3959278679308718e-06, + "loss": 2.6968, + "step": 18999 + }, + { + "epoch": 2.009306260575296, + "grad_norm": 8.125, + "learning_rate": 3.3957679845815704e-06, + "loss": 2.2532, + "step": 19000 + }, + { + "epoch": 2.009412013536379, + "grad_norm": 7.9375, + "learning_rate": 3.3956080970287976e-06, + "loss": 2.5378, + "step": 19001 + }, + { + "epoch": 2.0095177664974617, + "grad_norm": 7.8125, + "learning_rate": 3.3954482052733007e-06, + "loss": 2.4976, + "step": 19002 + }, + { + "epoch": 2.009623519458545, + "grad_norm": 8.6875, + "learning_rate": 3.3952883093158317e-06, + "loss": 2.7956, + "step": 19003 + }, + { + "epoch": 2.009729272419628, + "grad_norm": 9.875, + "learning_rate": 3.395128409157142e-06, + "loss": 2.7534, + "step": 19004 + }, + { + "epoch": 2.0098350253807107, + "grad_norm": 8.375, + "learning_rate": 3.3949685047979798e-06, + "loss": 2.4278, + "step": 19005 + }, + { + "epoch": 2.0099407783417935, + "grad_norm": 8.3125, + "learning_rate": 3.3948085962390966e-06, + "loss": 2.5508, + "step": 19006 + }, + { + "epoch": 2.0100465313028764, + "grad_norm": 8.375, + "learning_rate": 3.3946486834812424e-06, + "loss": 2.6169, + "step": 19007 + }, + { + "epoch": 2.010152284263959, + "grad_norm": 8.4375, + "learning_rate": 3.394488766525168e-06, + "loss": 2.5432, + "step": 19008 + }, + { + "epoch": 2.0102580372250425, + "grad_norm": 8.1875, + "learning_rate": 3.394328845371623e-06, + "loss": 2.5278, + "step": 19009 + }, + { + "epoch": 2.0103637901861253, + "grad_norm": 8.1875, + "learning_rate": 3.39416892002136e-06, + "loss": 2.54, + "step": 19010 + }, + { + "epoch": 2.010469543147208, + "grad_norm": 9.5625, + "learning_rate": 3.3940089904751273e-06, + "loss": 2.5831, + "step": 19011 + }, + { + "epoch": 2.010575296108291, + "grad_norm": 8.0, + "learning_rate": 3.393849056733676e-06, + "loss": 2.6361, + "step": 19012 + }, + { + "epoch": 2.010681049069374, + "grad_norm": 8.0625, + "learning_rate": 3.393689118797756e-06, + "loss": 2.5199, + "step": 19013 + }, + { + "epoch": 2.0107868020304567, + "grad_norm": 8.5625, + "learning_rate": 3.393529176668119e-06, + "loss": 2.6331, + "step": 19014 + }, + { + "epoch": 2.01089255499154, + "grad_norm": 8.75, + "learning_rate": 3.3933692303455153e-06, + "loss": 2.6945, + "step": 19015 + }, + { + "epoch": 2.010998307952623, + "grad_norm": 8.25, + "learning_rate": 3.3932092798306952e-06, + "loss": 2.5198, + "step": 19016 + }, + { + "epoch": 2.0111040609137056, + "grad_norm": 8.0625, + "learning_rate": 3.393049325124409e-06, + "loss": 2.6169, + "step": 19017 + }, + { + "epoch": 2.0112098138747885, + "grad_norm": 8.875, + "learning_rate": 3.392889366227407e-06, + "loss": 2.5228, + "step": 19018 + }, + { + "epoch": 2.0113155668358713, + "grad_norm": 8.0625, + "learning_rate": 3.392729403140441e-06, + "loss": 2.5329, + "step": 19019 + }, + { + "epoch": 2.011421319796954, + "grad_norm": 8.3125, + "learning_rate": 3.3925694358642613e-06, + "loss": 2.755, + "step": 19020 + }, + { + "epoch": 2.0115270727580374, + "grad_norm": 8.375, + "learning_rate": 3.3924094643996174e-06, + "loss": 2.1106, + "step": 19021 + }, + { + "epoch": 2.0116328257191203, + "grad_norm": 8.25, + "learning_rate": 3.3922494887472616e-06, + "loss": 2.3752, + "step": 19022 + }, + { + "epoch": 2.011738578680203, + "grad_norm": 8.875, + "learning_rate": 3.392089508907944e-06, + "loss": 2.5511, + "step": 19023 + }, + { + "epoch": 2.011844331641286, + "grad_norm": 8.4375, + "learning_rate": 3.391929524882415e-06, + "loss": 2.6592, + "step": 19024 + }, + { + "epoch": 2.011950084602369, + "grad_norm": 8.1875, + "learning_rate": 3.3917695366714247e-06, + "loss": 2.3136, + "step": 19025 + }, + { + "epoch": 2.0120558375634516, + "grad_norm": 8.0, + "learning_rate": 3.391609544275726e-06, + "loss": 2.413, + "step": 19026 + }, + { + "epoch": 2.0121615905245345, + "grad_norm": 8.625, + "learning_rate": 3.3914495476960672e-06, + "loss": 2.5734, + "step": 19027 + }, + { + "epoch": 2.0122673434856178, + "grad_norm": 8.3125, + "learning_rate": 3.3912895469332008e-06, + "loss": 2.475, + "step": 19028 + }, + { + "epoch": 2.0123730964467006, + "grad_norm": 8.4375, + "learning_rate": 3.3911295419878775e-06, + "loss": 2.5426, + "step": 19029 + }, + { + "epoch": 2.0124788494077834, + "grad_norm": 11.25, + "learning_rate": 3.390969532860847e-06, + "loss": 2.485, + "step": 19030 + }, + { + "epoch": 2.0125846023688663, + "grad_norm": 8.0625, + "learning_rate": 3.390809519552861e-06, + "loss": 2.3704, + "step": 19031 + }, + { + "epoch": 2.012690355329949, + "grad_norm": 8.3125, + "learning_rate": 3.3906495020646714e-06, + "loss": 2.4523, + "step": 19032 + }, + { + "epoch": 2.012796108291032, + "grad_norm": 8.6875, + "learning_rate": 3.3904894803970263e-06, + "loss": 2.6175, + "step": 19033 + }, + { + "epoch": 2.0129018612521152, + "grad_norm": 8.5, + "learning_rate": 3.390329454550679e-06, + "loss": 2.5627, + "step": 19034 + }, + { + "epoch": 2.013007614213198, + "grad_norm": 7.96875, + "learning_rate": 3.3901694245263795e-06, + "loss": 2.4344, + "step": 19035 + }, + { + "epoch": 2.013113367174281, + "grad_norm": 8.75, + "learning_rate": 3.390009390324879e-06, + "loss": 2.523, + "step": 19036 + }, + { + "epoch": 2.0132191201353637, + "grad_norm": 8.125, + "learning_rate": 3.3898493519469283e-06, + "loss": 2.6141, + "step": 19037 + }, + { + "epoch": 2.0133248730964466, + "grad_norm": 8.75, + "learning_rate": 3.389689309393279e-06, + "loss": 2.4507, + "step": 19038 + }, + { + "epoch": 2.0134306260575294, + "grad_norm": 8.5, + "learning_rate": 3.389529262664682e-06, + "loss": 2.5086, + "step": 19039 + }, + { + "epoch": 2.0135363790186127, + "grad_norm": 7.96875, + "learning_rate": 3.389369211761887e-06, + "loss": 2.4305, + "step": 19040 + }, + { + "epoch": 2.0136421319796955, + "grad_norm": 8.375, + "learning_rate": 3.3892091566856468e-06, + "loss": 2.69, + "step": 19041 + }, + { + "epoch": 2.0137478849407784, + "grad_norm": 8.25, + "learning_rate": 3.3890490974367113e-06, + "loss": 2.4053, + "step": 19042 + }, + { + "epoch": 2.0138536379018612, + "grad_norm": 8.5625, + "learning_rate": 3.388889034015832e-06, + "loss": 2.621, + "step": 19043 + }, + { + "epoch": 2.013959390862944, + "grad_norm": 8.4375, + "learning_rate": 3.3887289664237592e-06, + "loss": 2.5818, + "step": 19044 + }, + { + "epoch": 2.014065143824027, + "grad_norm": 8.25, + "learning_rate": 3.388568894661246e-06, + "loss": 2.5549, + "step": 19045 + }, + { + "epoch": 2.01417089678511, + "grad_norm": 8.125, + "learning_rate": 3.388408818729042e-06, + "loss": 2.5422, + "step": 19046 + }, + { + "epoch": 2.014276649746193, + "grad_norm": 9.625, + "learning_rate": 3.3882487386278987e-06, + "loss": 2.5668, + "step": 19047 + }, + { + "epoch": 2.014382402707276, + "grad_norm": 8.0625, + "learning_rate": 3.3880886543585676e-06, + "loss": 2.4449, + "step": 19048 + }, + { + "epoch": 2.0144881556683587, + "grad_norm": 8.8125, + "learning_rate": 3.3879285659217997e-06, + "loss": 2.6481, + "step": 19049 + }, + { + "epoch": 2.0145939086294415, + "grad_norm": 8.25, + "learning_rate": 3.3877684733183458e-06, + "loss": 2.6404, + "step": 19050 + }, + { + "epoch": 2.0146996615905244, + "grad_norm": 8.8125, + "learning_rate": 3.387608376548958e-06, + "loss": 2.4901, + "step": 19051 + }, + { + "epoch": 2.0148054145516077, + "grad_norm": 8.5, + "learning_rate": 3.3874482756143867e-06, + "loss": 2.5358, + "step": 19052 + }, + { + "epoch": 2.0149111675126905, + "grad_norm": 7.875, + "learning_rate": 3.3872881705153838e-06, + "loss": 2.5406, + "step": 19053 + }, + { + "epoch": 2.0150169204737733, + "grad_norm": 8.1875, + "learning_rate": 3.3871280612527e-06, + "loss": 2.6225, + "step": 19054 + }, + { + "epoch": 2.015122673434856, + "grad_norm": 8.0625, + "learning_rate": 3.3869679478270877e-06, + "loss": 2.4089, + "step": 19055 + }, + { + "epoch": 2.015228426395939, + "grad_norm": 8.75, + "learning_rate": 3.386807830239297e-06, + "loss": 2.5333, + "step": 19056 + }, + { + "epoch": 2.015334179357022, + "grad_norm": 8.3125, + "learning_rate": 3.3866477084900807e-06, + "loss": 2.6954, + "step": 19057 + }, + { + "epoch": 2.0154399323181047, + "grad_norm": 8.75, + "learning_rate": 3.3864875825801885e-06, + "loss": 2.6035, + "step": 19058 + }, + { + "epoch": 2.015545685279188, + "grad_norm": 8.5625, + "learning_rate": 3.386327452510373e-06, + "loss": 2.6326, + "step": 19059 + }, + { + "epoch": 2.015651438240271, + "grad_norm": 8.3125, + "learning_rate": 3.386167318281385e-06, + "loss": 2.4295, + "step": 19060 + }, + { + "epoch": 2.0157571912013537, + "grad_norm": 8.1875, + "learning_rate": 3.3860071798939765e-06, + "loss": 2.5005, + "step": 19061 + }, + { + "epoch": 2.0158629441624365, + "grad_norm": 8.375, + "learning_rate": 3.3858470373488985e-06, + "loss": 2.3487, + "step": 19062 + }, + { + "epoch": 2.0159686971235193, + "grad_norm": 7.71875, + "learning_rate": 3.385686890646902e-06, + "loss": 2.4123, + "step": 19063 + }, + { + "epoch": 2.016074450084602, + "grad_norm": 9.375, + "learning_rate": 3.3855267397887403e-06, + "loss": 2.4744, + "step": 19064 + }, + { + "epoch": 2.0161802030456855, + "grad_norm": 8.5, + "learning_rate": 3.3853665847751636e-06, + "loss": 2.5589, + "step": 19065 + }, + { + "epoch": 2.0162859560067683, + "grad_norm": 8.125, + "learning_rate": 3.385206425606923e-06, + "loss": 2.4693, + "step": 19066 + }, + { + "epoch": 2.016391708967851, + "grad_norm": 8.0625, + "learning_rate": 3.3850462622847713e-06, + "loss": 2.5128, + "step": 19067 + }, + { + "epoch": 2.016497461928934, + "grad_norm": 8.375, + "learning_rate": 3.3848860948094587e-06, + "loss": 2.383, + "step": 19068 + }, + { + "epoch": 2.016603214890017, + "grad_norm": 8.8125, + "learning_rate": 3.384725923181738e-06, + "loss": 2.758, + "step": 19069 + }, + { + "epoch": 2.0167089678510997, + "grad_norm": 8.25, + "learning_rate": 3.3845657474023602e-06, + "loss": 2.4197, + "step": 19070 + }, + { + "epoch": 2.016814720812183, + "grad_norm": 8.1875, + "learning_rate": 3.384405567472078e-06, + "loss": 2.5177, + "step": 19071 + }, + { + "epoch": 2.0169204737732658, + "grad_norm": 7.875, + "learning_rate": 3.384245383391641e-06, + "loss": 2.2014, + "step": 19072 + }, + { + "epoch": 2.0170262267343486, + "grad_norm": 8.0, + "learning_rate": 3.3840851951618025e-06, + "loss": 2.5156, + "step": 19073 + }, + { + "epoch": 2.0171319796954315, + "grad_norm": 8.125, + "learning_rate": 3.3839250027833142e-06, + "loss": 2.6039, + "step": 19074 + }, + { + "epoch": 2.0172377326565143, + "grad_norm": 8.25, + "learning_rate": 3.383764806256927e-06, + "loss": 2.5467, + "step": 19075 + }, + { + "epoch": 2.017343485617597, + "grad_norm": 9.0625, + "learning_rate": 3.383604605583393e-06, + "loss": 2.6103, + "step": 19076 + }, + { + "epoch": 2.0174492385786804, + "grad_norm": 8.375, + "learning_rate": 3.3834444007634645e-06, + "loss": 2.4846, + "step": 19077 + }, + { + "epoch": 2.0175549915397633, + "grad_norm": 8.4375, + "learning_rate": 3.3832841917978913e-06, + "loss": 2.5871, + "step": 19078 + }, + { + "epoch": 2.017660744500846, + "grad_norm": 8.625, + "learning_rate": 3.3831239786874276e-06, + "loss": 2.7704, + "step": 19079 + }, + { + "epoch": 2.017766497461929, + "grad_norm": 8.3125, + "learning_rate": 3.3829637614328243e-06, + "loss": 2.4669, + "step": 19080 + }, + { + "epoch": 2.0178722504230118, + "grad_norm": 8.0625, + "learning_rate": 3.3828035400348336e-06, + "loss": 2.461, + "step": 19081 + }, + { + "epoch": 2.0179780033840946, + "grad_norm": 9.0625, + "learning_rate": 3.3826433144942063e-06, + "loss": 2.4902, + "step": 19082 + }, + { + "epoch": 2.0180837563451774, + "grad_norm": 7.59375, + "learning_rate": 3.382483084811695e-06, + "loss": 2.2851, + "step": 19083 + }, + { + "epoch": 2.0181895093062607, + "grad_norm": 7.84375, + "learning_rate": 3.382322850988052e-06, + "loss": 2.397, + "step": 19084 + }, + { + "epoch": 2.0182952622673436, + "grad_norm": 8.4375, + "learning_rate": 3.3821626130240276e-06, + "loss": 2.6643, + "step": 19085 + }, + { + "epoch": 2.0184010152284264, + "grad_norm": 8.625, + "learning_rate": 3.3820023709203757e-06, + "loss": 2.6856, + "step": 19086 + }, + { + "epoch": 2.0185067681895092, + "grad_norm": 8.3125, + "learning_rate": 3.3818421246778473e-06, + "loss": 2.5696, + "step": 19087 + }, + { + "epoch": 2.018612521150592, + "grad_norm": 8.25, + "learning_rate": 3.381681874297194e-06, + "loss": 2.5978, + "step": 19088 + }, + { + "epoch": 2.018718274111675, + "grad_norm": 8.375, + "learning_rate": 3.381521619779169e-06, + "loss": 2.4584, + "step": 19089 + }, + { + "epoch": 2.018824027072758, + "grad_norm": 8.4375, + "learning_rate": 3.3813613611245234e-06, + "loss": 2.1322, + "step": 19090 + }, + { + "epoch": 2.018929780033841, + "grad_norm": 8.5625, + "learning_rate": 3.3812010983340087e-06, + "loss": 2.5344, + "step": 19091 + }, + { + "epoch": 2.019035532994924, + "grad_norm": 8.75, + "learning_rate": 3.3810408314083786e-06, + "loss": 2.6744, + "step": 19092 + }, + { + "epoch": 2.0191412859560067, + "grad_norm": 8.375, + "learning_rate": 3.380880560348384e-06, + "loss": 2.5563, + "step": 19093 + }, + { + "epoch": 2.0192470389170896, + "grad_norm": 8.75, + "learning_rate": 3.3807202851547765e-06, + "loss": 2.5634, + "step": 19094 + }, + { + "epoch": 2.0193527918781724, + "grad_norm": 8.25, + "learning_rate": 3.3805600058283097e-06, + "loss": 2.2776, + "step": 19095 + }, + { + "epoch": 2.0194585448392557, + "grad_norm": 8.125, + "learning_rate": 3.3803997223697344e-06, + "loss": 2.6078, + "step": 19096 + }, + { + "epoch": 2.0195642978003385, + "grad_norm": 8.5, + "learning_rate": 3.380239434779804e-06, + "loss": 2.5782, + "step": 19097 + }, + { + "epoch": 2.0196700507614214, + "grad_norm": 8.5625, + "learning_rate": 3.380079143059269e-06, + "loss": 2.6106, + "step": 19098 + }, + { + "epoch": 2.019775803722504, + "grad_norm": 8.875, + "learning_rate": 3.379918847208883e-06, + "loss": 2.4132, + "step": 19099 + }, + { + "epoch": 2.019881556683587, + "grad_norm": 8.4375, + "learning_rate": 3.379758547229398e-06, + "loss": 2.5291, + "step": 19100 + }, + { + "epoch": 2.01998730964467, + "grad_norm": 8.3125, + "learning_rate": 3.379598243121565e-06, + "loss": 2.3884, + "step": 19101 + }, + { + "epoch": 2.020093062605753, + "grad_norm": 9.0625, + "learning_rate": 3.3794379348861383e-06, + "loss": 2.6451, + "step": 19102 + }, + { + "epoch": 2.020198815566836, + "grad_norm": 8.5625, + "learning_rate": 3.3792776225238686e-06, + "loss": 2.358, + "step": 19103 + }, + { + "epoch": 2.020304568527919, + "grad_norm": 8.3125, + "learning_rate": 3.3791173060355087e-06, + "loss": 2.543, + "step": 19104 + }, + { + "epoch": 2.0204103214890017, + "grad_norm": 8.25, + "learning_rate": 3.37895698542181e-06, + "loss": 2.5739, + "step": 19105 + }, + { + "epoch": 2.0205160744500845, + "grad_norm": 8.8125, + "learning_rate": 3.3787966606835267e-06, + "loss": 2.5313, + "step": 19106 + }, + { + "epoch": 2.0206218274111674, + "grad_norm": 8.1875, + "learning_rate": 3.3786363318214093e-06, + "loss": 2.4243, + "step": 19107 + }, + { + "epoch": 2.0207275803722506, + "grad_norm": 9.3125, + "learning_rate": 3.3784759988362116e-06, + "loss": 2.8225, + "step": 19108 + }, + { + "epoch": 2.0208333333333335, + "grad_norm": 7.875, + "learning_rate": 3.3783156617286854e-06, + "loss": 2.2481, + "step": 19109 + }, + { + "epoch": 2.0209390862944163, + "grad_norm": 8.3125, + "learning_rate": 3.378155320499583e-06, + "loss": 2.5345, + "step": 19110 + }, + { + "epoch": 2.021044839255499, + "grad_norm": 7.90625, + "learning_rate": 3.3779949751496565e-06, + "loss": 2.4179, + "step": 19111 + }, + { + "epoch": 2.021150592216582, + "grad_norm": 8.5625, + "learning_rate": 3.377834625679659e-06, + "loss": 1.956, + "step": 19112 + }, + { + "epoch": 2.021256345177665, + "grad_norm": 9.5, + "learning_rate": 3.377674272090342e-06, + "loss": 2.6189, + "step": 19113 + }, + { + "epoch": 2.0213620981387477, + "grad_norm": 8.4375, + "learning_rate": 3.377513914382459e-06, + "loss": 2.4106, + "step": 19114 + }, + { + "epoch": 2.021467851099831, + "grad_norm": 8.625, + "learning_rate": 3.377353552556762e-06, + "loss": 2.5933, + "step": 19115 + }, + { + "epoch": 2.021573604060914, + "grad_norm": 8.75, + "learning_rate": 3.377193186614004e-06, + "loss": 2.8173, + "step": 19116 + }, + { + "epoch": 2.0216793570219966, + "grad_norm": 8.5, + "learning_rate": 3.377032816554936e-06, + "loss": 2.5922, + "step": 19117 + }, + { + "epoch": 2.0217851099830795, + "grad_norm": 8.25, + "learning_rate": 3.3768724423803127e-06, + "loss": 2.5683, + "step": 19118 + }, + { + "epoch": 2.0218908629441623, + "grad_norm": 8.5625, + "learning_rate": 3.3767120640908857e-06, + "loss": 2.6845, + "step": 19119 + }, + { + "epoch": 2.021996615905245, + "grad_norm": 8.6875, + "learning_rate": 3.3765516816874074e-06, + "loss": 2.4875, + "step": 19120 + }, + { + "epoch": 2.0221023688663284, + "grad_norm": 8.1875, + "learning_rate": 3.37639129517063e-06, + "loss": 2.5231, + "step": 19121 + }, + { + "epoch": 2.0222081218274113, + "grad_norm": 8.0, + "learning_rate": 3.3762309045413077e-06, + "loss": 2.483, + "step": 19122 + }, + { + "epoch": 2.022313874788494, + "grad_norm": 8.6875, + "learning_rate": 3.3760705098001906e-06, + "loss": 2.4963, + "step": 19123 + }, + { + "epoch": 2.022419627749577, + "grad_norm": 7.875, + "learning_rate": 3.375910110948034e-06, + "loss": 2.3326, + "step": 19124 + }, + { + "epoch": 2.02252538071066, + "grad_norm": 8.3125, + "learning_rate": 3.37574970798559e-06, + "loss": 2.5615, + "step": 19125 + }, + { + "epoch": 2.0226311336717426, + "grad_norm": 8.375, + "learning_rate": 3.37558930091361e-06, + "loss": 2.6065, + "step": 19126 + }, + { + "epoch": 2.022736886632826, + "grad_norm": 8.125, + "learning_rate": 3.3754288897328474e-06, + "loss": 2.5156, + "step": 19127 + }, + { + "epoch": 2.0228426395939088, + "grad_norm": 8.3125, + "learning_rate": 3.3752684744440556e-06, + "loss": 2.5431, + "step": 19128 + }, + { + "epoch": 2.0229483925549916, + "grad_norm": 8.0625, + "learning_rate": 3.3751080550479863e-06, + "loss": 2.4429, + "step": 19129 + }, + { + "epoch": 2.0230541455160744, + "grad_norm": 8.0625, + "learning_rate": 3.3749476315453933e-06, + "loss": 2.4009, + "step": 19130 + }, + { + "epoch": 2.0231598984771573, + "grad_norm": 9.0625, + "learning_rate": 3.374787203937028e-06, + "loss": 2.5115, + "step": 19131 + }, + { + "epoch": 2.02326565143824, + "grad_norm": 8.125, + "learning_rate": 3.3746267722236455e-06, + "loss": 2.4662, + "step": 19132 + }, + { + "epoch": 2.0233714043993234, + "grad_norm": 8.25, + "learning_rate": 3.374466336405996e-06, + "loss": 2.589, + "step": 19133 + }, + { + "epoch": 2.0234771573604062, + "grad_norm": 8.1875, + "learning_rate": 3.3743058964848343e-06, + "loss": 2.4398, + "step": 19134 + }, + { + "epoch": 2.023582910321489, + "grad_norm": 8.5625, + "learning_rate": 3.374145452460913e-06, + "loss": 2.4793, + "step": 19135 + }, + { + "epoch": 2.023688663282572, + "grad_norm": 8.5, + "learning_rate": 3.373985004334983e-06, + "loss": 2.512, + "step": 19136 + }, + { + "epoch": 2.0237944162436547, + "grad_norm": 8.6875, + "learning_rate": 3.3738245521078005e-06, + "loss": 2.4917, + "step": 19137 + }, + { + "epoch": 2.0239001692047376, + "grad_norm": 9.25, + "learning_rate": 3.3736640957801164e-06, + "loss": 2.5115, + "step": 19138 + }, + { + "epoch": 2.0240059221658204, + "grad_norm": 8.3125, + "learning_rate": 3.3735036353526835e-06, + "loss": 2.4532, + "step": 19139 + }, + { + "epoch": 2.0241116751269037, + "grad_norm": 8.0625, + "learning_rate": 3.3733431708262553e-06, + "loss": 2.4561, + "step": 19140 + }, + { + "epoch": 2.0242174280879865, + "grad_norm": 9.0625, + "learning_rate": 3.3731827022015853e-06, + "loss": 2.6206, + "step": 19141 + }, + { + "epoch": 2.0243231810490694, + "grad_norm": 8.4375, + "learning_rate": 3.373022229479425e-06, + "loss": 2.6162, + "step": 19142 + }, + { + "epoch": 2.0244289340101522, + "grad_norm": 8.375, + "learning_rate": 3.37286175266053e-06, + "loss": 2.3649, + "step": 19143 + }, + { + "epoch": 2.024534686971235, + "grad_norm": 8.625, + "learning_rate": 3.3727012717456507e-06, + "loss": 2.6454, + "step": 19144 + }, + { + "epoch": 2.024640439932318, + "grad_norm": 8.1875, + "learning_rate": 3.3725407867355412e-06, + "loss": 2.5226, + "step": 19145 + }, + { + "epoch": 2.024746192893401, + "grad_norm": 8.1875, + "learning_rate": 3.3723802976309554e-06, + "loss": 2.6521, + "step": 19146 + }, + { + "epoch": 2.024851945854484, + "grad_norm": 8.0, + "learning_rate": 3.3722198044326453e-06, + "loss": 2.52, + "step": 19147 + }, + { + "epoch": 2.024957698815567, + "grad_norm": 8.5, + "learning_rate": 3.3720593071413645e-06, + "loss": 2.5092, + "step": 19148 + }, + { + "epoch": 2.0250634517766497, + "grad_norm": 8.3125, + "learning_rate": 3.371898805757865e-06, + "loss": 2.7095, + "step": 19149 + }, + { + "epoch": 2.0251692047377325, + "grad_norm": 8.875, + "learning_rate": 3.3717383002829017e-06, + "loss": 2.6209, + "step": 19150 + }, + { + "epoch": 2.0252749576988154, + "grad_norm": 7.75, + "learning_rate": 3.3715777907172274e-06, + "loss": 2.4151, + "step": 19151 + }, + { + "epoch": 2.0253807106598987, + "grad_norm": 8.4375, + "learning_rate": 3.371417277061595e-06, + "loss": 2.4482, + "step": 19152 + }, + { + "epoch": 2.0254864636209815, + "grad_norm": 8.4375, + "learning_rate": 3.371256759316758e-06, + "loss": 2.6207, + "step": 19153 + }, + { + "epoch": 2.0255922165820643, + "grad_norm": 8.625, + "learning_rate": 3.3710962374834686e-06, + "loss": 2.6905, + "step": 19154 + }, + { + "epoch": 2.025697969543147, + "grad_norm": 8.1875, + "learning_rate": 3.370935711562481e-06, + "loss": 2.3331, + "step": 19155 + }, + { + "epoch": 2.02580372250423, + "grad_norm": 8.6875, + "learning_rate": 3.3707751815545487e-06, + "loss": 2.3196, + "step": 19156 + }, + { + "epoch": 2.025909475465313, + "grad_norm": 7.875, + "learning_rate": 3.3706146474604245e-06, + "loss": 2.572, + "step": 19157 + }, + { + "epoch": 2.026015228426396, + "grad_norm": 9.5, + "learning_rate": 3.3704541092808605e-06, + "loss": 2.6575, + "step": 19158 + }, + { + "epoch": 2.026120981387479, + "grad_norm": 8.125, + "learning_rate": 3.370293567016613e-06, + "loss": 2.2846, + "step": 19159 + }, + { + "epoch": 2.026226734348562, + "grad_norm": 8.125, + "learning_rate": 3.3701330206684336e-06, + "loss": 2.2076, + "step": 19160 + }, + { + "epoch": 2.0263324873096447, + "grad_norm": 8.3125, + "learning_rate": 3.369972470237075e-06, + "loss": 2.7249, + "step": 19161 + }, + { + "epoch": 2.0264382402707275, + "grad_norm": 8.375, + "learning_rate": 3.369811915723292e-06, + "loss": 2.62, + "step": 19162 + }, + { + "epoch": 2.0265439932318103, + "grad_norm": 8.375, + "learning_rate": 3.3696513571278378e-06, + "loss": 2.6222, + "step": 19163 + }, + { + "epoch": 2.0266497461928936, + "grad_norm": 8.4375, + "learning_rate": 3.3694907944514643e-06, + "loss": 2.2278, + "step": 19164 + }, + { + "epoch": 2.0267554991539765, + "grad_norm": 8.6875, + "learning_rate": 3.3693302276949267e-06, + "loss": 2.7959, + "step": 19165 + }, + { + "epoch": 2.0268612521150593, + "grad_norm": 7.96875, + "learning_rate": 3.369169656858978e-06, + "loss": 2.5409, + "step": 19166 + }, + { + "epoch": 2.026967005076142, + "grad_norm": 8.25, + "learning_rate": 3.3690090819443715e-06, + "loss": 2.6431, + "step": 19167 + }, + { + "epoch": 2.027072758037225, + "grad_norm": 8.375, + "learning_rate": 3.3688485029518603e-06, + "loss": 2.3555, + "step": 19168 + }, + { + "epoch": 2.027178510998308, + "grad_norm": 8.125, + "learning_rate": 3.368687919882199e-06, + "loss": 2.4217, + "step": 19169 + }, + { + "epoch": 2.0272842639593907, + "grad_norm": 8.625, + "learning_rate": 3.368527332736141e-06, + "loss": 2.5742, + "step": 19170 + }, + { + "epoch": 2.027390016920474, + "grad_norm": 8.3125, + "learning_rate": 3.368366741514439e-06, + "loss": 2.4706, + "step": 19171 + }, + { + "epoch": 2.0274957698815568, + "grad_norm": 8.375, + "learning_rate": 3.3682061462178474e-06, + "loss": 2.5653, + "step": 19172 + }, + { + "epoch": 2.0276015228426396, + "grad_norm": 8.25, + "learning_rate": 3.3680455468471185e-06, + "loss": 2.2993, + "step": 19173 + }, + { + "epoch": 2.0277072758037225, + "grad_norm": 8.4375, + "learning_rate": 3.367884943403007e-06, + "loss": 2.6526, + "step": 19174 + }, + { + "epoch": 2.0278130287648053, + "grad_norm": 8.5, + "learning_rate": 3.367724335886267e-06, + "loss": 2.5763, + "step": 19175 + }, + { + "epoch": 2.027918781725888, + "grad_norm": 8.0625, + "learning_rate": 3.367563724297651e-06, + "loss": 2.2318, + "step": 19176 + }, + { + "epoch": 2.0280245346869714, + "grad_norm": 8.125, + "learning_rate": 3.367403108637914e-06, + "loss": 2.5103, + "step": 19177 + }, + { + "epoch": 2.0281302876480543, + "grad_norm": 8.3125, + "learning_rate": 3.3672424889078085e-06, + "loss": 2.5213, + "step": 19178 + }, + { + "epoch": 2.028236040609137, + "grad_norm": 8.5625, + "learning_rate": 3.3670818651080887e-06, + "loss": 2.4893, + "step": 19179 + }, + { + "epoch": 2.02834179357022, + "grad_norm": 8.5625, + "learning_rate": 3.366921237239508e-06, + "loss": 2.116, + "step": 19180 + }, + { + "epoch": 2.0284475465313028, + "grad_norm": 8.4375, + "learning_rate": 3.3667606053028213e-06, + "loss": 2.5945, + "step": 19181 + }, + { + "epoch": 2.0285532994923856, + "grad_norm": 8.625, + "learning_rate": 3.3665999692987814e-06, + "loss": 2.6588, + "step": 19182 + }, + { + "epoch": 2.028659052453469, + "grad_norm": 8.9375, + "learning_rate": 3.3664393292281415e-06, + "loss": 2.8263, + "step": 19183 + }, + { + "epoch": 2.0287648054145517, + "grad_norm": 8.5, + "learning_rate": 3.3662786850916574e-06, + "loss": 2.545, + "step": 19184 + }, + { + "epoch": 2.0288705583756346, + "grad_norm": 8.5, + "learning_rate": 3.366118036890081e-06, + "loss": 2.5488, + "step": 19185 + }, + { + "epoch": 2.0289763113367174, + "grad_norm": 8.625, + "learning_rate": 3.3659573846241668e-06, + "loss": 2.633, + "step": 19186 + }, + { + "epoch": 2.0290820642978002, + "grad_norm": 8.1875, + "learning_rate": 3.3657967282946692e-06, + "loss": 2.4225, + "step": 19187 + }, + { + "epoch": 2.029187817258883, + "grad_norm": 8.375, + "learning_rate": 3.3656360679023413e-06, + "loss": 2.6943, + "step": 19188 + }, + { + "epoch": 2.0292935702199664, + "grad_norm": 8.25, + "learning_rate": 3.3654754034479377e-06, + "loss": 2.6062, + "step": 19189 + }, + { + "epoch": 2.029399323181049, + "grad_norm": 8.375, + "learning_rate": 3.365314734932211e-06, + "loss": 2.4437, + "step": 19190 + }, + { + "epoch": 2.029505076142132, + "grad_norm": 8.6875, + "learning_rate": 3.3651540623559177e-06, + "loss": 2.4888, + "step": 19191 + }, + { + "epoch": 2.029610829103215, + "grad_norm": 8.0, + "learning_rate": 3.364993385719809e-06, + "loss": 2.3896, + "step": 19192 + }, + { + "epoch": 2.0297165820642977, + "grad_norm": 8.1875, + "learning_rate": 3.36483270502464e-06, + "loss": 2.4083, + "step": 19193 + }, + { + "epoch": 2.0298223350253806, + "grad_norm": 8.625, + "learning_rate": 3.364672020271165e-06, + "loss": 2.4576, + "step": 19194 + }, + { + "epoch": 2.0299280879864634, + "grad_norm": 8.375, + "learning_rate": 3.3645113314601386e-06, + "loss": 2.5673, + "step": 19195 + }, + { + "epoch": 2.0300338409475467, + "grad_norm": 8.3125, + "learning_rate": 3.3643506385923134e-06, + "loss": 2.6832, + "step": 19196 + }, + { + "epoch": 2.0301395939086295, + "grad_norm": 8.4375, + "learning_rate": 3.3641899416684447e-06, + "loss": 2.5431, + "step": 19197 + }, + { + "epoch": 2.0302453468697124, + "grad_norm": 8.3125, + "learning_rate": 3.364029240689286e-06, + "loss": 2.4339, + "step": 19198 + }, + { + "epoch": 2.030351099830795, + "grad_norm": 8.1875, + "learning_rate": 3.363868535655591e-06, + "loss": 2.5293, + "step": 19199 + }, + { + "epoch": 2.030456852791878, + "grad_norm": 8.5, + "learning_rate": 3.3637078265681146e-06, + "loss": 2.484, + "step": 19200 + }, + { + "epoch": 2.030562605752961, + "grad_norm": 8.375, + "learning_rate": 3.363547113427611e-06, + "loss": 2.478, + "step": 19201 + }, + { + "epoch": 2.030668358714044, + "grad_norm": 8.5625, + "learning_rate": 3.3633863962348327e-06, + "loss": 2.6518, + "step": 19202 + }, + { + "epoch": 2.030774111675127, + "grad_norm": 8.4375, + "learning_rate": 3.3632256749905356e-06, + "loss": 2.5878, + "step": 19203 + }, + { + "epoch": 2.03087986463621, + "grad_norm": 7.6875, + "learning_rate": 3.3630649496954745e-06, + "loss": 2.3402, + "step": 19204 + }, + { + "epoch": 2.0309856175972927, + "grad_norm": 8.6875, + "learning_rate": 3.362904220350402e-06, + "loss": 2.8251, + "step": 19205 + }, + { + "epoch": 2.0310913705583755, + "grad_norm": 8.375, + "learning_rate": 3.3627434869560728e-06, + "loss": 2.5472, + "step": 19206 + }, + { + "epoch": 2.0311971235194584, + "grad_norm": 8.375, + "learning_rate": 3.362582749513241e-06, + "loss": 2.1703, + "step": 19207 + }, + { + "epoch": 2.0313028764805416, + "grad_norm": 8.3125, + "learning_rate": 3.3624220080226615e-06, + "loss": 2.6874, + "step": 19208 + }, + { + "epoch": 2.0314086294416245, + "grad_norm": 8.375, + "learning_rate": 3.362261262485088e-06, + "loss": 2.136, + "step": 19209 + }, + { + "epoch": 2.0315143824027073, + "grad_norm": 8.6875, + "learning_rate": 3.362100512901275e-06, + "loss": 2.6816, + "step": 19210 + }, + { + "epoch": 2.03162013536379, + "grad_norm": 8.0625, + "learning_rate": 3.361939759271977e-06, + "loss": 2.4543, + "step": 19211 + }, + { + "epoch": 2.031725888324873, + "grad_norm": 8.5, + "learning_rate": 3.3617790015979478e-06, + "loss": 2.1676, + "step": 19212 + }, + { + "epoch": 2.031831641285956, + "grad_norm": 8.875, + "learning_rate": 3.3616182398799435e-06, + "loss": 2.5058, + "step": 19213 + }, + { + "epoch": 2.031937394247039, + "grad_norm": 8.5625, + "learning_rate": 3.3614574741187164e-06, + "loss": 2.6985, + "step": 19214 + }, + { + "epoch": 2.032043147208122, + "grad_norm": 8.5, + "learning_rate": 3.361296704315021e-06, + "loss": 2.4364, + "step": 19215 + }, + { + "epoch": 2.032148900169205, + "grad_norm": 8.75, + "learning_rate": 3.3611359304696133e-06, + "loss": 2.5308, + "step": 19216 + }, + { + "epoch": 2.0322546531302876, + "grad_norm": 8.4375, + "learning_rate": 3.360975152583247e-06, + "loss": 2.5319, + "step": 19217 + }, + { + "epoch": 2.0323604060913705, + "grad_norm": 8.5625, + "learning_rate": 3.3608143706566757e-06, + "loss": 2.3231, + "step": 19218 + }, + { + "epoch": 2.0324661590524533, + "grad_norm": 8.3125, + "learning_rate": 3.360653584690656e-06, + "loss": 2.4372, + "step": 19219 + }, + { + "epoch": 2.0325719120135366, + "grad_norm": 8.0, + "learning_rate": 3.360492794685939e-06, + "loss": 2.6272, + "step": 19220 + }, + { + "epoch": 2.0326776649746194, + "grad_norm": 8.0, + "learning_rate": 3.360332000643283e-06, + "loss": 2.3687, + "step": 19221 + }, + { + "epoch": 2.0327834179357023, + "grad_norm": 7.9375, + "learning_rate": 3.360171202563439e-06, + "loss": 2.4015, + "step": 19222 + }, + { + "epoch": 2.032889170896785, + "grad_norm": 8.4375, + "learning_rate": 3.3600104004471653e-06, + "loss": 2.671, + "step": 19223 + }, + { + "epoch": 2.032994923857868, + "grad_norm": 7.59375, + "learning_rate": 3.3598495942952136e-06, + "loss": 2.4967, + "step": 19224 + }, + { + "epoch": 2.033100676818951, + "grad_norm": 8.0, + "learning_rate": 3.359688784108339e-06, + "loss": 2.5143, + "step": 19225 + }, + { + "epoch": 2.0332064297800336, + "grad_norm": 8.5625, + "learning_rate": 3.3595279698872974e-06, + "loss": 2.6271, + "step": 19226 + }, + { + "epoch": 2.033312182741117, + "grad_norm": 8.0625, + "learning_rate": 3.3593671516328424e-06, + "loss": 2.4029, + "step": 19227 + }, + { + "epoch": 2.0334179357021998, + "grad_norm": 8.5, + "learning_rate": 3.3592063293457283e-06, + "loss": 2.4534, + "step": 19228 + }, + { + "epoch": 2.0335236886632826, + "grad_norm": 8.3125, + "learning_rate": 3.3590455030267107e-06, + "loss": 2.5662, + "step": 19229 + }, + { + "epoch": 2.0336294416243654, + "grad_norm": 8.375, + "learning_rate": 3.3588846726765435e-06, + "loss": 2.5533, + "step": 19230 + }, + { + "epoch": 2.0337351945854483, + "grad_norm": 8.125, + "learning_rate": 3.358723838295982e-06, + "loss": 2.4408, + "step": 19231 + }, + { + "epoch": 2.033840947546531, + "grad_norm": 8.4375, + "learning_rate": 3.358562999885781e-06, + "loss": 2.5677, + "step": 19232 + }, + { + "epoch": 2.0339467005076144, + "grad_norm": 8.75, + "learning_rate": 3.358402157446695e-06, + "loss": 2.5249, + "step": 19233 + }, + { + "epoch": 2.0340524534686972, + "grad_norm": 8.5625, + "learning_rate": 3.358241310979478e-06, + "loss": 2.5269, + "step": 19234 + }, + { + "epoch": 2.03415820642978, + "grad_norm": 8.25, + "learning_rate": 3.3580804604848865e-06, + "loss": 2.6297, + "step": 19235 + }, + { + "epoch": 2.034263959390863, + "grad_norm": 8.625, + "learning_rate": 3.3579196059636744e-06, + "loss": 2.5621, + "step": 19236 + }, + { + "epoch": 2.0343697123519457, + "grad_norm": 8.1875, + "learning_rate": 3.3577587474165956e-06, + "loss": 2.3988, + "step": 19237 + }, + { + "epoch": 2.0344754653130286, + "grad_norm": 8.25, + "learning_rate": 3.3575978848444057e-06, + "loss": 2.4428, + "step": 19238 + }, + { + "epoch": 2.034581218274112, + "grad_norm": 8.0, + "learning_rate": 3.3574370182478605e-06, + "loss": 2.539, + "step": 19239 + }, + { + "epoch": 2.0346869712351947, + "grad_norm": 8.5, + "learning_rate": 3.3572761476277137e-06, + "loss": 2.6038, + "step": 19240 + }, + { + "epoch": 2.0347927241962775, + "grad_norm": 9.1875, + "learning_rate": 3.3571152729847206e-06, + "loss": 2.5205, + "step": 19241 + }, + { + "epoch": 2.0348984771573604, + "grad_norm": 8.375, + "learning_rate": 3.3569543943196357e-06, + "loss": 2.7028, + "step": 19242 + }, + { + "epoch": 2.0350042301184432, + "grad_norm": 7.625, + "learning_rate": 3.356793511633215e-06, + "loss": 2.3324, + "step": 19243 + }, + { + "epoch": 2.035109983079526, + "grad_norm": 8.1875, + "learning_rate": 3.3566326249262127e-06, + "loss": 2.4236, + "step": 19244 + }, + { + "epoch": 2.0352157360406093, + "grad_norm": 9.375, + "learning_rate": 3.3564717341993836e-06, + "loss": 2.7126, + "step": 19245 + }, + { + "epoch": 2.035321489001692, + "grad_norm": 8.5, + "learning_rate": 3.356310839453483e-06, + "loss": 2.6268, + "step": 19246 + }, + { + "epoch": 2.035427241962775, + "grad_norm": 8.8125, + "learning_rate": 3.356149940689265e-06, + "loss": 2.625, + "step": 19247 + }, + { + "epoch": 2.035532994923858, + "grad_norm": 8.25, + "learning_rate": 3.3559890379074866e-06, + "loss": 2.5209, + "step": 19248 + }, + { + "epoch": 2.0356387478849407, + "grad_norm": 8.625, + "learning_rate": 3.3558281311089015e-06, + "loss": 2.4668, + "step": 19249 + }, + { + "epoch": 2.0357445008460235, + "grad_norm": 7.875, + "learning_rate": 3.3556672202942648e-06, + "loss": 2.5688, + "step": 19250 + }, + { + "epoch": 2.035850253807107, + "grad_norm": 8.125, + "learning_rate": 3.3555063054643322e-06, + "loss": 2.419, + "step": 19251 + }, + { + "epoch": 2.0359560067681897, + "grad_norm": 8.3125, + "learning_rate": 3.3553453866198583e-06, + "loss": 2.4523, + "step": 19252 + }, + { + "epoch": 2.0360617597292725, + "grad_norm": 8.375, + "learning_rate": 3.355184463761597e-06, + "loss": 2.4487, + "step": 19253 + }, + { + "epoch": 2.0361675126903553, + "grad_norm": 8.4375, + "learning_rate": 3.3550235368903065e-06, + "loss": 2.6182, + "step": 19254 + }, + { + "epoch": 2.036273265651438, + "grad_norm": 8.625, + "learning_rate": 3.354862606006739e-06, + "loss": 2.676, + "step": 19255 + }, + { + "epoch": 2.036379018612521, + "grad_norm": 8.4375, + "learning_rate": 3.354701671111652e-06, + "loss": 2.5429, + "step": 19256 + }, + { + "epoch": 2.036484771573604, + "grad_norm": 8.3125, + "learning_rate": 3.3545407322057987e-06, + "loss": 2.6072, + "step": 19257 + }, + { + "epoch": 2.036590524534687, + "grad_norm": 8.5625, + "learning_rate": 3.354379789289936e-06, + "loss": 2.4094, + "step": 19258 + }, + { + "epoch": 2.03669627749577, + "grad_norm": 7.90625, + "learning_rate": 3.354218842364818e-06, + "loss": 2.5724, + "step": 19259 + }, + { + "epoch": 2.036802030456853, + "grad_norm": 9.0625, + "learning_rate": 3.3540578914312004e-06, + "loss": 2.8316, + "step": 19260 + }, + { + "epoch": 2.0369077834179357, + "grad_norm": 8.5, + "learning_rate": 3.353896936489839e-06, + "loss": 2.4682, + "step": 19261 + }, + { + "epoch": 2.0370135363790185, + "grad_norm": 8.25, + "learning_rate": 3.353735977541488e-06, + "loss": 2.7073, + "step": 19262 + }, + { + "epoch": 2.0371192893401013, + "grad_norm": 8.3125, + "learning_rate": 3.353575014586903e-06, + "loss": 2.6953, + "step": 19263 + }, + { + "epoch": 2.0372250423011846, + "grad_norm": 8.0625, + "learning_rate": 3.353414047626839e-06, + "loss": 2.554, + "step": 19264 + }, + { + "epoch": 2.0373307952622675, + "grad_norm": 9.0, + "learning_rate": 3.353253076662053e-06, + "loss": 2.5453, + "step": 19265 + }, + { + "epoch": 2.0374365482233503, + "grad_norm": 9.1875, + "learning_rate": 3.3530921016932983e-06, + "loss": 2.7953, + "step": 19266 + }, + { + "epoch": 2.037542301184433, + "grad_norm": 8.5625, + "learning_rate": 3.3529311227213326e-06, + "loss": 2.6497, + "step": 19267 + }, + { + "epoch": 2.037648054145516, + "grad_norm": 8.3125, + "learning_rate": 3.352770139746909e-06, + "loss": 2.5348, + "step": 19268 + }, + { + "epoch": 2.037753807106599, + "grad_norm": 8.3125, + "learning_rate": 3.352609152770784e-06, + "loss": 2.5191, + "step": 19269 + }, + { + "epoch": 2.037859560067682, + "grad_norm": 8.0625, + "learning_rate": 3.352448161793713e-06, + "loss": 2.6607, + "step": 19270 + }, + { + "epoch": 2.037965313028765, + "grad_norm": 8.5625, + "learning_rate": 3.352287166816452e-06, + "loss": 2.3944, + "step": 19271 + }, + { + "epoch": 2.0380710659898478, + "grad_norm": 8.5625, + "learning_rate": 3.352126167839755e-06, + "loss": 2.5726, + "step": 19272 + }, + { + "epoch": 2.0381768189509306, + "grad_norm": 8.4375, + "learning_rate": 3.351965164864378e-06, + "loss": 2.4362, + "step": 19273 + }, + { + "epoch": 2.0382825719120135, + "grad_norm": 8.75, + "learning_rate": 3.3518041578910782e-06, + "loss": 2.0599, + "step": 19274 + }, + { + "epoch": 2.0383883248730963, + "grad_norm": 8.5, + "learning_rate": 3.3516431469206094e-06, + "loss": 2.4752, + "step": 19275 + }, + { + "epoch": 2.0384940778341796, + "grad_norm": 8.5625, + "learning_rate": 3.351482131953728e-06, + "loss": 2.2874, + "step": 19276 + }, + { + "epoch": 2.0385998307952624, + "grad_norm": 8.0625, + "learning_rate": 3.351321112991189e-06, + "loss": 2.3561, + "step": 19277 + }, + { + "epoch": 2.0387055837563453, + "grad_norm": 8.6875, + "learning_rate": 3.3511600900337484e-06, + "loss": 2.4926, + "step": 19278 + }, + { + "epoch": 2.038811336717428, + "grad_norm": 8.375, + "learning_rate": 3.3509990630821604e-06, + "loss": 2.5975, + "step": 19279 + }, + { + "epoch": 2.038917089678511, + "grad_norm": 8.25, + "learning_rate": 3.3508380321371826e-06, + "loss": 2.4583, + "step": 19280 + }, + { + "epoch": 2.0390228426395938, + "grad_norm": 8.5625, + "learning_rate": 3.3506769971995705e-06, + "loss": 2.5568, + "step": 19281 + }, + { + "epoch": 2.0391285956006766, + "grad_norm": 8.5, + "learning_rate": 3.3505159582700773e-06, + "loss": 2.7587, + "step": 19282 + }, + { + "epoch": 2.03923434856176, + "grad_norm": 8.4375, + "learning_rate": 3.3503549153494623e-06, + "loss": 2.6016, + "step": 19283 + }, + { + "epoch": 2.0393401015228427, + "grad_norm": 8.5625, + "learning_rate": 3.3501938684384792e-06, + "loss": 2.5898, + "step": 19284 + }, + { + "epoch": 2.0394458544839256, + "grad_norm": 8.5, + "learning_rate": 3.3500328175378838e-06, + "loss": 2.6649, + "step": 19285 + }, + { + "epoch": 2.0395516074450084, + "grad_norm": 8.375, + "learning_rate": 3.3498717626484323e-06, + "loss": 2.4338, + "step": 19286 + }, + { + "epoch": 2.0396573604060912, + "grad_norm": 9.1875, + "learning_rate": 3.3497107037708803e-06, + "loss": 2.5746, + "step": 19287 + }, + { + "epoch": 2.039763113367174, + "grad_norm": 8.25, + "learning_rate": 3.3495496409059826e-06, + "loss": 2.5545, + "step": 19288 + }, + { + "epoch": 2.0398688663282574, + "grad_norm": 9.6875, + "learning_rate": 3.3493885740544967e-06, + "loss": 2.7568, + "step": 19289 + }, + { + "epoch": 2.03997461928934, + "grad_norm": 8.75, + "learning_rate": 3.349227503217178e-06, + "loss": 2.4548, + "step": 19290 + }, + { + "epoch": 2.040080372250423, + "grad_norm": 8.25, + "learning_rate": 3.3490664283947805e-06, + "loss": 2.4748, + "step": 19291 + }, + { + "epoch": 2.040186125211506, + "grad_norm": 8.625, + "learning_rate": 3.348905349588062e-06, + "loss": 2.4538, + "step": 19292 + }, + { + "epoch": 2.0402918781725887, + "grad_norm": 8.4375, + "learning_rate": 3.348744266797779e-06, + "loss": 2.677, + "step": 19293 + }, + { + "epoch": 2.0403976311336716, + "grad_norm": 9.25, + "learning_rate": 3.3485831800246855e-06, + "loss": 2.5243, + "step": 19294 + }, + { + "epoch": 2.040503384094755, + "grad_norm": 8.75, + "learning_rate": 3.3484220892695373e-06, + "loss": 2.768, + "step": 19295 + }, + { + "epoch": 2.0406091370558377, + "grad_norm": 8.5625, + "learning_rate": 3.3482609945330924e-06, + "loss": 2.6319, + "step": 19296 + }, + { + "epoch": 2.0407148900169205, + "grad_norm": 8.375, + "learning_rate": 3.348099895816106e-06, + "loss": 2.5565, + "step": 19297 + }, + { + "epoch": 2.0408206429780034, + "grad_norm": 8.5, + "learning_rate": 3.3479387931193323e-06, + "loss": 2.8267, + "step": 19298 + }, + { + "epoch": 2.040926395939086, + "grad_norm": 10.1875, + "learning_rate": 3.3477776864435294e-06, + "loss": 2.5749, + "step": 19299 + }, + { + "epoch": 2.041032148900169, + "grad_norm": 8.25, + "learning_rate": 3.3476165757894525e-06, + "loss": 2.2778, + "step": 19300 + }, + { + "epoch": 2.0411379018612523, + "grad_norm": 8.3125, + "learning_rate": 3.347455461157857e-06, + "loss": 2.5811, + "step": 19301 + }, + { + "epoch": 2.041243654822335, + "grad_norm": 8.5625, + "learning_rate": 3.347294342549501e-06, + "loss": 2.5903, + "step": 19302 + }, + { + "epoch": 2.041349407783418, + "grad_norm": 8.1875, + "learning_rate": 3.3471332199651386e-06, + "loss": 2.6268, + "step": 19303 + }, + { + "epoch": 2.041455160744501, + "grad_norm": 8.8125, + "learning_rate": 3.3469720934055257e-06, + "loss": 2.6079, + "step": 19304 + }, + { + "epoch": 2.0415609137055837, + "grad_norm": 8.3125, + "learning_rate": 3.3468109628714195e-06, + "loss": 2.5133, + "step": 19305 + }, + { + "epoch": 2.0416666666666665, + "grad_norm": 9.5625, + "learning_rate": 3.3466498283635764e-06, + "loss": 2.6413, + "step": 19306 + }, + { + "epoch": 2.0417724196277494, + "grad_norm": 8.1875, + "learning_rate": 3.3464886898827515e-06, + "loss": 2.4614, + "step": 19307 + }, + { + "epoch": 2.0418781725888326, + "grad_norm": 8.9375, + "learning_rate": 3.346327547429701e-06, + "loss": 2.7322, + "step": 19308 + }, + { + "epoch": 2.0419839255499155, + "grad_norm": 8.25, + "learning_rate": 3.346166401005182e-06, + "loss": 2.3916, + "step": 19309 + }, + { + "epoch": 2.0420896785109983, + "grad_norm": 8.5, + "learning_rate": 3.3460052506099506e-06, + "loss": 2.5286, + "step": 19310 + }, + { + "epoch": 2.042195431472081, + "grad_norm": 9.375, + "learning_rate": 3.3458440962447613e-06, + "loss": 2.7219, + "step": 19311 + }, + { + "epoch": 2.042301184433164, + "grad_norm": 8.5625, + "learning_rate": 3.3456829379103726e-06, + "loss": 2.4074, + "step": 19312 + }, + { + "epoch": 2.042406937394247, + "grad_norm": 8.5625, + "learning_rate": 3.3455217756075398e-06, + "loss": 2.549, + "step": 19313 + }, + { + "epoch": 2.04251269035533, + "grad_norm": 8.875, + "learning_rate": 3.3453606093370182e-06, + "loss": 2.7544, + "step": 19314 + }, + { + "epoch": 2.042618443316413, + "grad_norm": 8.375, + "learning_rate": 3.3451994390995657e-06, + "loss": 2.6477, + "step": 19315 + }, + { + "epoch": 2.042724196277496, + "grad_norm": 8.375, + "learning_rate": 3.3450382648959377e-06, + "loss": 2.5147, + "step": 19316 + }, + { + "epoch": 2.0428299492385786, + "grad_norm": 8.4375, + "learning_rate": 3.3448770867268903e-06, + "loss": 2.4168, + "step": 19317 + }, + { + "epoch": 2.0429357021996615, + "grad_norm": 8.1875, + "learning_rate": 3.344715904593181e-06, + "loss": 2.4333, + "step": 19318 + }, + { + "epoch": 2.0430414551607443, + "grad_norm": 8.5, + "learning_rate": 3.344554718495565e-06, + "loss": 2.5921, + "step": 19319 + }, + { + "epoch": 2.0431472081218276, + "grad_norm": 11.3125, + "learning_rate": 3.3443935284347993e-06, + "loss": 2.4931, + "step": 19320 + }, + { + "epoch": 2.0432529610829104, + "grad_norm": 8.0625, + "learning_rate": 3.3442323344116405e-06, + "loss": 2.4809, + "step": 19321 + }, + { + "epoch": 2.0433587140439933, + "grad_norm": 8.6875, + "learning_rate": 3.3440711364268443e-06, + "loss": 2.5255, + "step": 19322 + }, + { + "epoch": 2.043464467005076, + "grad_norm": 8.5625, + "learning_rate": 3.343909934481167e-06, + "loss": 2.3919, + "step": 19323 + }, + { + "epoch": 2.043570219966159, + "grad_norm": 8.375, + "learning_rate": 3.3437487285753666e-06, + "loss": 2.6579, + "step": 19324 + }, + { + "epoch": 2.043675972927242, + "grad_norm": 8.75, + "learning_rate": 3.343587518710198e-06, + "loss": 2.4909, + "step": 19325 + }, + { + "epoch": 2.043781725888325, + "grad_norm": 8.5625, + "learning_rate": 3.3434263048864176e-06, + "loss": 2.5529, + "step": 19326 + }, + { + "epoch": 2.043887478849408, + "grad_norm": 8.625, + "learning_rate": 3.343265087104783e-06, + "loss": 2.6476, + "step": 19327 + }, + { + "epoch": 2.0439932318104908, + "grad_norm": 8.5, + "learning_rate": 3.3431038653660497e-06, + "loss": 2.6913, + "step": 19328 + }, + { + "epoch": 2.0440989847715736, + "grad_norm": 8.875, + "learning_rate": 3.342942639670976e-06, + "loss": 2.5823, + "step": 19329 + }, + { + "epoch": 2.0442047377326564, + "grad_norm": 8.9375, + "learning_rate": 3.3427814100203158e-06, + "loss": 2.7286, + "step": 19330 + }, + { + "epoch": 2.0443104906937393, + "grad_norm": 8.375, + "learning_rate": 3.342620176414828e-06, + "loss": 2.4374, + "step": 19331 + }, + { + "epoch": 2.0444162436548226, + "grad_norm": 8.625, + "learning_rate": 3.3424589388552687e-06, + "loss": 2.4246, + "step": 19332 + }, + { + "epoch": 2.0445219966159054, + "grad_norm": 8.25, + "learning_rate": 3.3422976973423933e-06, + "loss": 2.529, + "step": 19333 + }, + { + "epoch": 2.0446277495769882, + "grad_norm": 8.5, + "learning_rate": 3.3421364518769594e-06, + "loss": 2.5759, + "step": 19334 + }, + { + "epoch": 2.044733502538071, + "grad_norm": 9.0, + "learning_rate": 3.341975202459724e-06, + "loss": 2.4235, + "step": 19335 + }, + { + "epoch": 2.044839255499154, + "grad_norm": 8.5, + "learning_rate": 3.3418139490914427e-06, + "loss": 2.289, + "step": 19336 + }, + { + "epoch": 2.0449450084602367, + "grad_norm": 8.0625, + "learning_rate": 3.3416526917728732e-06, + "loss": 2.3647, + "step": 19337 + }, + { + "epoch": 2.0450507614213196, + "grad_norm": 8.375, + "learning_rate": 3.341491430504772e-06, + "loss": 2.3247, + "step": 19338 + }, + { + "epoch": 2.045156514382403, + "grad_norm": 8.375, + "learning_rate": 3.3413301652878956e-06, + "loss": 2.498, + "step": 19339 + }, + { + "epoch": 2.0452622673434857, + "grad_norm": 8.1875, + "learning_rate": 3.3411688961230003e-06, + "loss": 2.4824, + "step": 19340 + }, + { + "epoch": 2.0453680203045685, + "grad_norm": 9.125, + "learning_rate": 3.3410076230108447e-06, + "loss": 2.4656, + "step": 19341 + }, + { + "epoch": 2.0454737732656514, + "grad_norm": 8.3125, + "learning_rate": 3.340846345952183e-06, + "loss": 2.5349, + "step": 19342 + }, + { + "epoch": 2.0455795262267342, + "grad_norm": 7.9375, + "learning_rate": 3.340685064947774e-06, + "loss": 2.5082, + "step": 19343 + }, + { + "epoch": 2.045685279187817, + "grad_norm": 8.75, + "learning_rate": 3.340523779998373e-06, + "loss": 2.4439, + "step": 19344 + }, + { + "epoch": 2.0457910321489003, + "grad_norm": 8.6875, + "learning_rate": 3.3403624911047384e-06, + "loss": 2.4714, + "step": 19345 + }, + { + "epoch": 2.045896785109983, + "grad_norm": 8.0, + "learning_rate": 3.3402011982676257e-06, + "loss": 2.5048, + "step": 19346 + }, + { + "epoch": 2.046002538071066, + "grad_norm": 9.4375, + "learning_rate": 3.340039901487793e-06, + "loss": 2.7746, + "step": 19347 + }, + { + "epoch": 2.046108291032149, + "grad_norm": 8.3125, + "learning_rate": 3.3398786007659967e-06, + "loss": 2.4869, + "step": 19348 + }, + { + "epoch": 2.0462140439932317, + "grad_norm": 8.375, + "learning_rate": 3.3397172961029933e-06, + "loss": 2.4418, + "step": 19349 + }, + { + "epoch": 2.0463197969543145, + "grad_norm": 7.875, + "learning_rate": 3.3395559874995405e-06, + "loss": 2.5441, + "step": 19350 + }, + { + "epoch": 2.046425549915398, + "grad_norm": 9.0625, + "learning_rate": 3.3393946749563943e-06, + "loss": 2.6115, + "step": 19351 + }, + { + "epoch": 2.0465313028764807, + "grad_norm": 8.0625, + "learning_rate": 3.3392333584743124e-06, + "loss": 2.1763, + "step": 19352 + }, + { + "epoch": 2.0466370558375635, + "grad_norm": 8.25, + "learning_rate": 3.339072038054051e-06, + "loss": 2.5817, + "step": 19353 + }, + { + "epoch": 2.0467428087986463, + "grad_norm": 8.375, + "learning_rate": 3.3389107136963684e-06, + "loss": 2.5539, + "step": 19354 + }, + { + "epoch": 2.046848561759729, + "grad_norm": 8.375, + "learning_rate": 3.3387493854020207e-06, + "loss": 2.4857, + "step": 19355 + }, + { + "epoch": 2.046954314720812, + "grad_norm": 8.6875, + "learning_rate": 3.338588053171765e-06, + "loss": 2.4498, + "step": 19356 + }, + { + "epoch": 2.0470600676818953, + "grad_norm": 8.3125, + "learning_rate": 3.3384267170063594e-06, + "loss": 2.3119, + "step": 19357 + }, + { + "epoch": 2.047165820642978, + "grad_norm": 8.1875, + "learning_rate": 3.338265376906559e-06, + "loss": 2.5596, + "step": 19358 + }, + { + "epoch": 2.047271573604061, + "grad_norm": 8.125, + "learning_rate": 3.338104032873123e-06, + "loss": 2.4701, + "step": 19359 + }, + { + "epoch": 2.047377326565144, + "grad_norm": 8.3125, + "learning_rate": 3.337942684906807e-06, + "loss": 2.4159, + "step": 19360 + }, + { + "epoch": 2.0474830795262267, + "grad_norm": 8.0625, + "learning_rate": 3.3377813330083687e-06, + "loss": 2.4025, + "step": 19361 + }, + { + "epoch": 2.0475888324873095, + "grad_norm": 8.5, + "learning_rate": 3.337619977178565e-06, + "loss": 2.5746, + "step": 19362 + }, + { + "epoch": 2.047694585448393, + "grad_norm": 8.25, + "learning_rate": 3.3374586174181534e-06, + "loss": 2.279, + "step": 19363 + }, + { + "epoch": 2.0478003384094756, + "grad_norm": 8.3125, + "learning_rate": 3.3372972537278913e-06, + "loss": 2.7057, + "step": 19364 + }, + { + "epoch": 2.0479060913705585, + "grad_norm": 7.96875, + "learning_rate": 3.3371358861085356e-06, + "loss": 2.2945, + "step": 19365 + }, + { + "epoch": 2.0480118443316413, + "grad_norm": 8.125, + "learning_rate": 3.336974514560844e-06, + "loss": 2.5307, + "step": 19366 + }, + { + "epoch": 2.048117597292724, + "grad_norm": 8.375, + "learning_rate": 3.336813139085573e-06, + "loss": 2.513, + "step": 19367 + }, + { + "epoch": 2.048223350253807, + "grad_norm": 7.96875, + "learning_rate": 3.3366517596834796e-06, + "loss": 2.2812, + "step": 19368 + }, + { + "epoch": 2.04832910321489, + "grad_norm": 7.90625, + "learning_rate": 3.3364903763553227e-06, + "loss": 2.6606, + "step": 19369 + }, + { + "epoch": 2.048434856175973, + "grad_norm": 8.375, + "learning_rate": 3.3363289891018567e-06, + "loss": 2.5829, + "step": 19370 + }, + { + "epoch": 2.048540609137056, + "grad_norm": 8.625, + "learning_rate": 3.3361675979238423e-06, + "loss": 2.4426, + "step": 19371 + }, + { + "epoch": 2.0486463620981388, + "grad_norm": 7.65625, + "learning_rate": 3.336006202822035e-06, + "loss": 2.4112, + "step": 19372 + }, + { + "epoch": 2.0487521150592216, + "grad_norm": 8.5625, + "learning_rate": 3.335844803797193e-06, + "loss": 2.659, + "step": 19373 + }, + { + "epoch": 2.0488578680203045, + "grad_norm": 8.5625, + "learning_rate": 3.3356834008500725e-06, + "loss": 2.5978, + "step": 19374 + }, + { + "epoch": 2.0489636209813873, + "grad_norm": 8.375, + "learning_rate": 3.3355219939814322e-06, + "loss": 2.5907, + "step": 19375 + }, + { + "epoch": 2.0490693739424706, + "grad_norm": 8.3125, + "learning_rate": 3.335360583192029e-06, + "loss": 2.3088, + "step": 19376 + }, + { + "epoch": 2.0491751269035534, + "grad_norm": 8.0625, + "learning_rate": 3.335199168482619e-06, + "loss": 2.4058, + "step": 19377 + }, + { + "epoch": 2.0492808798646363, + "grad_norm": 8.0625, + "learning_rate": 3.335037749853962e-06, + "loss": 2.5309, + "step": 19378 + }, + { + "epoch": 2.049386632825719, + "grad_norm": 8.75, + "learning_rate": 3.334876327306814e-06, + "loss": 2.4294, + "step": 19379 + }, + { + "epoch": 2.049492385786802, + "grad_norm": 8.375, + "learning_rate": 3.3347149008419332e-06, + "loss": 2.6176, + "step": 19380 + }, + { + "epoch": 2.0495981387478848, + "grad_norm": 8.1875, + "learning_rate": 3.334553470460076e-06, + "loss": 2.5968, + "step": 19381 + }, + { + "epoch": 2.049703891708968, + "grad_norm": 8.3125, + "learning_rate": 3.3343920361620015e-06, + "loss": 2.6031, + "step": 19382 + }, + { + "epoch": 2.049809644670051, + "grad_norm": 8.25, + "learning_rate": 3.334230597948466e-06, + "loss": 2.6251, + "step": 19383 + }, + { + "epoch": 2.0499153976311337, + "grad_norm": 9.0, + "learning_rate": 3.3340691558202277e-06, + "loss": 2.8325, + "step": 19384 + }, + { + "epoch": 2.0500211505922166, + "grad_norm": 8.375, + "learning_rate": 3.3339077097780437e-06, + "loss": 2.6036, + "step": 19385 + }, + { + "epoch": 2.0501269035532994, + "grad_norm": 9.5, + "learning_rate": 3.333746259822672e-06, + "loss": 2.7137, + "step": 19386 + }, + { + "epoch": 2.0502326565143822, + "grad_norm": 8.625, + "learning_rate": 3.3335848059548704e-06, + "loss": 2.4124, + "step": 19387 + }, + { + "epoch": 2.0503384094754655, + "grad_norm": 8.25, + "learning_rate": 3.3334233481753952e-06, + "loss": 2.6731, + "step": 19388 + }, + { + "epoch": 2.0504441624365484, + "grad_norm": 8.1875, + "learning_rate": 3.3332618864850063e-06, + "loss": 2.6302, + "step": 19389 + }, + { + "epoch": 2.050549915397631, + "grad_norm": 8.375, + "learning_rate": 3.3331004208844593e-06, + "loss": 2.7156, + "step": 19390 + }, + { + "epoch": 2.050655668358714, + "grad_norm": 8.8125, + "learning_rate": 3.3329389513745138e-06, + "loss": 2.4854, + "step": 19391 + }, + { + "epoch": 2.050761421319797, + "grad_norm": 8.6875, + "learning_rate": 3.332777477955926e-06, + "loss": 2.6161, + "step": 19392 + }, + { + "epoch": 2.0508671742808797, + "grad_norm": 8.0625, + "learning_rate": 3.332616000629453e-06, + "loss": 2.3623, + "step": 19393 + }, + { + "epoch": 2.0509729272419626, + "grad_norm": 8.6875, + "learning_rate": 3.3324545193958546e-06, + "loss": 2.3746, + "step": 19394 + }, + { + "epoch": 2.051078680203046, + "grad_norm": 8.1875, + "learning_rate": 3.332293034255888e-06, + "loss": 2.5074, + "step": 19395 + }, + { + "epoch": 2.0511844331641287, + "grad_norm": 8.3125, + "learning_rate": 3.3321315452103096e-06, + "loss": 2.597, + "step": 19396 + }, + { + "epoch": 2.0512901861252115, + "grad_norm": 8.375, + "learning_rate": 3.3319700522598784e-06, + "loss": 2.6949, + "step": 19397 + }, + { + "epoch": 2.0513959390862944, + "grad_norm": 8.4375, + "learning_rate": 3.3318085554053525e-06, + "loss": 2.4206, + "step": 19398 + }, + { + "epoch": 2.051501692047377, + "grad_norm": 8.125, + "learning_rate": 3.3316470546474896e-06, + "loss": 2.3404, + "step": 19399 + }, + { + "epoch": 2.05160744500846, + "grad_norm": 8.6875, + "learning_rate": 3.3314855499870464e-06, + "loss": 2.8236, + "step": 19400 + }, + { + "epoch": 2.0517131979695433, + "grad_norm": 8.375, + "learning_rate": 3.331324041424782e-06, + "loss": 2.5225, + "step": 19401 + }, + { + "epoch": 2.051818950930626, + "grad_norm": 8.375, + "learning_rate": 3.331162528961454e-06, + "loss": 2.7294, + "step": 19402 + }, + { + "epoch": 2.051924703891709, + "grad_norm": 8.1875, + "learning_rate": 3.33100101259782e-06, + "loss": 2.4496, + "step": 19403 + }, + { + "epoch": 2.052030456852792, + "grad_norm": 8.5625, + "learning_rate": 3.3308394923346378e-06, + "loss": 2.571, + "step": 19404 + }, + { + "epoch": 2.0521362098138747, + "grad_norm": 8.5, + "learning_rate": 3.330677968172666e-06, + "loss": 2.422, + "step": 19405 + }, + { + "epoch": 2.0522419627749575, + "grad_norm": 8.625, + "learning_rate": 3.330516440112661e-06, + "loss": 2.6826, + "step": 19406 + }, + { + "epoch": 2.052347715736041, + "grad_norm": 8.25, + "learning_rate": 3.3303549081553838e-06, + "loss": 2.3054, + "step": 19407 + }, + { + "epoch": 2.0524534686971236, + "grad_norm": 8.0625, + "learning_rate": 3.33019337230159e-06, + "loss": 2.3881, + "step": 19408 + }, + { + "epoch": 2.0525592216582065, + "grad_norm": 8.25, + "learning_rate": 3.330031832552038e-06, + "loss": 2.6298, + "step": 19409 + }, + { + "epoch": 2.0526649746192893, + "grad_norm": 8.125, + "learning_rate": 3.3298702889074864e-06, + "loss": 2.4724, + "step": 19410 + }, + { + "epoch": 2.052770727580372, + "grad_norm": 8.125, + "learning_rate": 3.329708741368693e-06, + "loss": 2.5843, + "step": 19411 + }, + { + "epoch": 2.052876480541455, + "grad_norm": 9.125, + "learning_rate": 3.3295471899364153e-06, + "loss": 2.8665, + "step": 19412 + }, + { + "epoch": 2.0529822335025383, + "grad_norm": 8.5625, + "learning_rate": 3.329385634611413e-06, + "loss": 2.6674, + "step": 19413 + }, + { + "epoch": 2.053087986463621, + "grad_norm": 8.625, + "learning_rate": 3.329224075394442e-06, + "loss": 2.4889, + "step": 19414 + }, + { + "epoch": 2.053193739424704, + "grad_norm": 8.1875, + "learning_rate": 3.329062512286262e-06, + "loss": 2.5571, + "step": 19415 + }, + { + "epoch": 2.053299492385787, + "grad_norm": 8.375, + "learning_rate": 3.32890094528763e-06, + "loss": 2.5151, + "step": 19416 + }, + { + "epoch": 2.0534052453468696, + "grad_norm": 8.125, + "learning_rate": 3.3287393743993056e-06, + "loss": 2.2618, + "step": 19417 + }, + { + "epoch": 2.0535109983079525, + "grad_norm": 8.25, + "learning_rate": 3.3285777996220464e-06, + "loss": 2.5027, + "step": 19418 + }, + { + "epoch": 2.0536167512690353, + "grad_norm": 8.4375, + "learning_rate": 3.32841622095661e-06, + "loss": 2.6817, + "step": 19419 + }, + { + "epoch": 2.0537225042301186, + "grad_norm": 8.0625, + "learning_rate": 3.3282546384037556e-06, + "loss": 2.4577, + "step": 19420 + }, + { + "epoch": 2.0538282571912014, + "grad_norm": 7.78125, + "learning_rate": 3.3280930519642405e-06, + "loss": 2.5189, + "step": 19421 + }, + { + "epoch": 2.0539340101522843, + "grad_norm": 8.1875, + "learning_rate": 3.327931461638824e-06, + "loss": 2.4396, + "step": 19422 + }, + { + "epoch": 2.054039763113367, + "grad_norm": 8.375, + "learning_rate": 3.327769867428262e-06, + "loss": 2.6112, + "step": 19423 + }, + { + "epoch": 2.05414551607445, + "grad_norm": 8.75, + "learning_rate": 3.3276082693333167e-06, + "loss": 2.5956, + "step": 19424 + }, + { + "epoch": 2.054251269035533, + "grad_norm": 9.0, + "learning_rate": 3.327446667354743e-06, + "loss": 2.672, + "step": 19425 + }, + { + "epoch": 2.054357021996616, + "grad_norm": 8.75, + "learning_rate": 3.327285061493301e-06, + "loss": 2.6889, + "step": 19426 + }, + { + "epoch": 2.054462774957699, + "grad_norm": 8.3125, + "learning_rate": 3.3271234517497486e-06, + "loss": 2.5782, + "step": 19427 + }, + { + "epoch": 2.0545685279187818, + "grad_norm": 8.1875, + "learning_rate": 3.3269618381248437e-06, + "loss": 2.5339, + "step": 19428 + }, + { + "epoch": 2.0546742808798646, + "grad_norm": 8.0625, + "learning_rate": 3.3268002206193454e-06, + "loss": 2.3586, + "step": 19429 + }, + { + "epoch": 2.0547800338409474, + "grad_norm": 8.75, + "learning_rate": 3.3266385992340123e-06, + "loss": 2.6319, + "step": 19430 + }, + { + "epoch": 2.0548857868020303, + "grad_norm": 8.625, + "learning_rate": 3.3264769739696016e-06, + "loss": 2.6658, + "step": 19431 + }, + { + "epoch": 2.0549915397631136, + "grad_norm": 7.8125, + "learning_rate": 3.3263153448268727e-06, + "loss": 2.3973, + "step": 19432 + }, + { + "epoch": 2.0550972927241964, + "grad_norm": 8.4375, + "learning_rate": 3.326153711806584e-06, + "loss": 2.4284, + "step": 19433 + }, + { + "epoch": 2.0552030456852792, + "grad_norm": 8.5625, + "learning_rate": 3.325992074909494e-06, + "loss": 2.5071, + "step": 19434 + }, + { + "epoch": 2.055308798646362, + "grad_norm": 8.5625, + "learning_rate": 3.32583043413636e-06, + "loss": 2.738, + "step": 19435 + }, + { + "epoch": 2.055414551607445, + "grad_norm": 8.5625, + "learning_rate": 3.3256687894879426e-06, + "loss": 2.4644, + "step": 19436 + }, + { + "epoch": 2.0555203045685277, + "grad_norm": 8.125, + "learning_rate": 3.325507140964999e-06, + "loss": 2.4447, + "step": 19437 + }, + { + "epoch": 2.055626057529611, + "grad_norm": 8.6875, + "learning_rate": 3.325345488568288e-06, + "loss": 2.5653, + "step": 19438 + }, + { + "epoch": 2.055731810490694, + "grad_norm": 8.5625, + "learning_rate": 3.325183832298568e-06, + "loss": 2.5987, + "step": 19439 + }, + { + "epoch": 2.0558375634517767, + "grad_norm": 8.375, + "learning_rate": 3.3250221721565976e-06, + "loss": 2.6637, + "step": 19440 + }, + { + "epoch": 2.0559433164128595, + "grad_norm": 8.5, + "learning_rate": 3.324860508143135e-06, + "loss": 2.5594, + "step": 19441 + }, + { + "epoch": 2.0560490693739424, + "grad_norm": 8.0, + "learning_rate": 3.3246988402589407e-06, + "loss": 2.4861, + "step": 19442 + }, + { + "epoch": 2.0561548223350252, + "grad_norm": 8.3125, + "learning_rate": 3.3245371685047717e-06, + "loss": 2.5803, + "step": 19443 + }, + { + "epoch": 2.0562605752961085, + "grad_norm": 8.6875, + "learning_rate": 3.3243754928813864e-06, + "loss": 2.4031, + "step": 19444 + }, + { + "epoch": 2.0563663282571913, + "grad_norm": 10.75, + "learning_rate": 3.3242138133895453e-06, + "loss": 2.4573, + "step": 19445 + }, + { + "epoch": 2.056472081218274, + "grad_norm": 8.4375, + "learning_rate": 3.3240521300300047e-06, + "loss": 2.6286, + "step": 19446 + }, + { + "epoch": 2.056577834179357, + "grad_norm": 8.3125, + "learning_rate": 3.3238904428035246e-06, + "loss": 2.4376, + "step": 19447 + }, + { + "epoch": 2.05668358714044, + "grad_norm": 8.8125, + "learning_rate": 3.323728751710864e-06, + "loss": 2.5131, + "step": 19448 + }, + { + "epoch": 2.0567893401015227, + "grad_norm": 8.625, + "learning_rate": 3.3235670567527812e-06, + "loss": 2.5895, + "step": 19449 + }, + { + "epoch": 2.0568950930626055, + "grad_norm": 8.4375, + "learning_rate": 3.323405357930035e-06, + "loss": 2.6112, + "step": 19450 + }, + { + "epoch": 2.057000846023689, + "grad_norm": 8.375, + "learning_rate": 3.3232436552433838e-06, + "loss": 2.5763, + "step": 19451 + }, + { + "epoch": 2.0571065989847717, + "grad_norm": 8.125, + "learning_rate": 3.323081948693587e-06, + "loss": 2.6362, + "step": 19452 + }, + { + "epoch": 2.0572123519458545, + "grad_norm": 8.875, + "learning_rate": 3.322920238281404e-06, + "loss": 2.4248, + "step": 19453 + }, + { + "epoch": 2.0573181049069373, + "grad_norm": 8.3125, + "learning_rate": 3.322758524007592e-06, + "loss": 2.3146, + "step": 19454 + }, + { + "epoch": 2.05742385786802, + "grad_norm": 8.6875, + "learning_rate": 3.3225968058729114e-06, + "loss": 2.5071, + "step": 19455 + }, + { + "epoch": 2.057529610829103, + "grad_norm": 8.9375, + "learning_rate": 3.32243508387812e-06, + "loss": 2.7438, + "step": 19456 + }, + { + "epoch": 2.0576353637901863, + "grad_norm": 8.0625, + "learning_rate": 3.322273358023977e-06, + "loss": 2.5712, + "step": 19457 + }, + { + "epoch": 2.057741116751269, + "grad_norm": 8.8125, + "learning_rate": 3.322111628311242e-06, + "loss": 2.6386, + "step": 19458 + }, + { + "epoch": 2.057846869712352, + "grad_norm": 8.3125, + "learning_rate": 3.3219498947406726e-06, + "loss": 2.2338, + "step": 19459 + }, + { + "epoch": 2.057952622673435, + "grad_norm": 8.375, + "learning_rate": 3.3217881573130285e-06, + "loss": 2.4356, + "step": 19460 + }, + { + "epoch": 2.0580583756345177, + "grad_norm": 8.5, + "learning_rate": 3.321626416029069e-06, + "loss": 2.4679, + "step": 19461 + }, + { + "epoch": 2.0581641285956005, + "grad_norm": 9.8125, + "learning_rate": 3.321464670889553e-06, + "loss": 2.636, + "step": 19462 + }, + { + "epoch": 2.058269881556684, + "grad_norm": 9.0, + "learning_rate": 3.3213029218952386e-06, + "loss": 2.5601, + "step": 19463 + }, + { + "epoch": 2.0583756345177666, + "grad_norm": 8.8125, + "learning_rate": 3.3211411690468862e-06, + "loss": 2.5254, + "step": 19464 + }, + { + "epoch": 2.0584813874788495, + "grad_norm": 8.1875, + "learning_rate": 3.3209794123452543e-06, + "loss": 2.5848, + "step": 19465 + }, + { + "epoch": 2.0585871404399323, + "grad_norm": 7.90625, + "learning_rate": 3.320817651791101e-06, + "loss": 2.5128, + "step": 19466 + }, + { + "epoch": 2.058692893401015, + "grad_norm": 8.0, + "learning_rate": 3.320655887385187e-06, + "loss": 2.5082, + "step": 19467 + }, + { + "epoch": 2.058798646362098, + "grad_norm": 8.6875, + "learning_rate": 3.3204941191282698e-06, + "loss": 2.6368, + "step": 19468 + }, + { + "epoch": 2.0589043993231813, + "grad_norm": 8.375, + "learning_rate": 3.32033234702111e-06, + "loss": 2.5403, + "step": 19469 + }, + { + "epoch": 2.059010152284264, + "grad_norm": 8.875, + "learning_rate": 3.3201705710644648e-06, + "loss": 2.5275, + "step": 19470 + }, + { + "epoch": 2.059115905245347, + "grad_norm": 8.375, + "learning_rate": 3.320008791259096e-06, + "loss": 2.7389, + "step": 19471 + }, + { + "epoch": 2.0592216582064298, + "grad_norm": 8.1875, + "learning_rate": 3.319847007605761e-06, + "loss": 2.5049, + "step": 19472 + }, + { + "epoch": 2.0593274111675126, + "grad_norm": 7.75, + "learning_rate": 3.3196852201052183e-06, + "loss": 2.4449, + "step": 19473 + }, + { + "epoch": 2.0594331641285955, + "grad_norm": 8.9375, + "learning_rate": 3.319523428758229e-06, + "loss": 2.3891, + "step": 19474 + }, + { + "epoch": 2.0595389170896787, + "grad_norm": 8.1875, + "learning_rate": 3.3193616335655516e-06, + "loss": 2.3598, + "step": 19475 + }, + { + "epoch": 2.0596446700507616, + "grad_norm": 7.875, + "learning_rate": 3.3191998345279437e-06, + "loss": 2.493, + "step": 19476 + }, + { + "epoch": 2.0597504230118444, + "grad_norm": 8.625, + "learning_rate": 3.3190380316461675e-06, + "loss": 2.6444, + "step": 19477 + }, + { + "epoch": 2.0598561759729273, + "grad_norm": 8.5625, + "learning_rate": 3.3188762249209805e-06, + "loss": 2.4928, + "step": 19478 + }, + { + "epoch": 2.05996192893401, + "grad_norm": 8.5625, + "learning_rate": 3.3187144143531412e-06, + "loss": 2.4931, + "step": 19479 + }, + { + "epoch": 2.060067681895093, + "grad_norm": 9.0625, + "learning_rate": 3.3185525999434113e-06, + "loss": 2.5006, + "step": 19480 + }, + { + "epoch": 2.0601734348561758, + "grad_norm": 8.3125, + "learning_rate": 3.3183907816925487e-06, + "loss": 2.5052, + "step": 19481 + }, + { + "epoch": 2.060279187817259, + "grad_norm": 8.5625, + "learning_rate": 3.318228959601313e-06, + "loss": 2.5358, + "step": 19482 + }, + { + "epoch": 2.060384940778342, + "grad_norm": 8.625, + "learning_rate": 3.3180671336704627e-06, + "loss": 2.489, + "step": 19483 + }, + { + "epoch": 2.0604906937394247, + "grad_norm": 7.75, + "learning_rate": 3.3179053039007584e-06, + "loss": 2.3597, + "step": 19484 + }, + { + "epoch": 2.0605964467005076, + "grad_norm": 8.0, + "learning_rate": 3.317743470292959e-06, + "loss": 2.5315, + "step": 19485 + }, + { + "epoch": 2.0607021996615904, + "grad_norm": 9.3125, + "learning_rate": 3.3175816328478233e-06, + "loss": 2.604, + "step": 19486 + }, + { + "epoch": 2.0608079526226732, + "grad_norm": 8.4375, + "learning_rate": 3.317419791566112e-06, + "loss": 2.5574, + "step": 19487 + }, + { + "epoch": 2.0609137055837565, + "grad_norm": 8.5625, + "learning_rate": 3.3172579464485845e-06, + "loss": 2.404, + "step": 19488 + }, + { + "epoch": 2.0610194585448394, + "grad_norm": 8.625, + "learning_rate": 3.317096097495999e-06, + "loss": 2.5009, + "step": 19489 + }, + { + "epoch": 2.061125211505922, + "grad_norm": 7.9375, + "learning_rate": 3.3169342447091157e-06, + "loss": 2.543, + "step": 19490 + }, + { + "epoch": 2.061230964467005, + "grad_norm": 8.0, + "learning_rate": 3.3167723880886945e-06, + "loss": 2.4101, + "step": 19491 + }, + { + "epoch": 2.061336717428088, + "grad_norm": 8.4375, + "learning_rate": 3.316610527635493e-06, + "loss": 2.5799, + "step": 19492 + }, + { + "epoch": 2.0614424703891707, + "grad_norm": 8.3125, + "learning_rate": 3.316448663350274e-06, + "loss": 2.5199, + "step": 19493 + }, + { + "epoch": 2.061548223350254, + "grad_norm": 8.6875, + "learning_rate": 3.3162867952337946e-06, + "loss": 2.5663, + "step": 19494 + }, + { + "epoch": 2.061653976311337, + "grad_norm": 8.125, + "learning_rate": 3.3161249232868153e-06, + "loss": 2.5082, + "step": 19495 + }, + { + "epoch": 2.0617597292724197, + "grad_norm": 8.875, + "learning_rate": 3.3159630475100955e-06, + "loss": 2.7199, + "step": 19496 + }, + { + "epoch": 2.0618654822335025, + "grad_norm": 8.5, + "learning_rate": 3.315801167904395e-06, + "loss": 2.5362, + "step": 19497 + }, + { + "epoch": 2.0619712351945854, + "grad_norm": 8.25, + "learning_rate": 3.315639284470473e-06, + "loss": 2.531, + "step": 19498 + }, + { + "epoch": 2.062076988155668, + "grad_norm": 8.0625, + "learning_rate": 3.31547739720909e-06, + "loss": 2.5102, + "step": 19499 + }, + { + "epoch": 2.0621827411167515, + "grad_norm": 8.3125, + "learning_rate": 3.3153155061210046e-06, + "loss": 2.4619, + "step": 19500 + }, + { + "epoch": 2.0622884940778343, + "grad_norm": 8.5, + "learning_rate": 3.3151536112069767e-06, + "loss": 2.5083, + "step": 19501 + }, + { + "epoch": 2.062394247038917, + "grad_norm": 8.1875, + "learning_rate": 3.3149917124677668e-06, + "loss": 2.5282, + "step": 19502 + }, + { + "epoch": 2.0625, + "grad_norm": 8.125, + "learning_rate": 3.3148298099041336e-06, + "loss": 2.6332, + "step": 19503 + }, + { + "epoch": 2.062605752961083, + "grad_norm": 8.3125, + "learning_rate": 3.314667903516838e-06, + "loss": 2.6879, + "step": 19504 + }, + { + "epoch": 2.0627115059221657, + "grad_norm": 8.6875, + "learning_rate": 3.314505993306638e-06, + "loss": 2.5227, + "step": 19505 + }, + { + "epoch": 2.0628172588832485, + "grad_norm": 9.25, + "learning_rate": 3.314344079274296e-06, + "loss": 2.418, + "step": 19506 + }, + { + "epoch": 2.062923011844332, + "grad_norm": 8.625, + "learning_rate": 3.31418216142057e-06, + "loss": 2.5327, + "step": 19507 + }, + { + "epoch": 2.0630287648054146, + "grad_norm": 8.5, + "learning_rate": 3.314020239746219e-06, + "loss": 2.5602, + "step": 19508 + }, + { + "epoch": 2.0631345177664975, + "grad_norm": 8.1875, + "learning_rate": 3.3138583142520054e-06, + "loss": 2.5047, + "step": 19509 + }, + { + "epoch": 2.0632402707275803, + "grad_norm": 8.5, + "learning_rate": 3.3136963849386867e-06, + "loss": 2.5767, + "step": 19510 + }, + { + "epoch": 2.063346023688663, + "grad_norm": 8.1875, + "learning_rate": 3.313534451807023e-06, + "loss": 2.6467, + "step": 19511 + }, + { + "epoch": 2.063451776649746, + "grad_norm": 8.375, + "learning_rate": 3.3133725148577754e-06, + "loss": 2.6255, + "step": 19512 + }, + { + "epoch": 2.0635575296108293, + "grad_norm": 8.8125, + "learning_rate": 3.3132105740917044e-06, + "loss": 2.4908, + "step": 19513 + }, + { + "epoch": 2.063663282571912, + "grad_norm": 8.5, + "learning_rate": 3.313048629509567e-06, + "loss": 2.5738, + "step": 19514 + }, + { + "epoch": 2.063769035532995, + "grad_norm": 9.0625, + "learning_rate": 3.3128866811121263e-06, + "loss": 2.5725, + "step": 19515 + }, + { + "epoch": 2.063874788494078, + "grad_norm": 8.25, + "learning_rate": 3.3127247289001406e-06, + "loss": 2.6173, + "step": 19516 + }, + { + "epoch": 2.0639805414551606, + "grad_norm": 8.6875, + "learning_rate": 3.312562772874369e-06, + "loss": 2.7049, + "step": 19517 + }, + { + "epoch": 2.0640862944162435, + "grad_norm": 8.375, + "learning_rate": 3.312400813035574e-06, + "loss": 2.4643, + "step": 19518 + }, + { + "epoch": 2.0641920473773268, + "grad_norm": 8.8125, + "learning_rate": 3.3122388493845136e-06, + "loss": 2.6197, + "step": 19519 + }, + { + "epoch": 2.0642978003384096, + "grad_norm": 8.6875, + "learning_rate": 3.3120768819219483e-06, + "loss": 2.4519, + "step": 19520 + }, + { + "epoch": 2.0644035532994924, + "grad_norm": 8.8125, + "learning_rate": 3.311914910648639e-06, + "loss": 2.6275, + "step": 19521 + }, + { + "epoch": 2.0645093062605753, + "grad_norm": 8.25, + "learning_rate": 3.3117529355653444e-06, + "loss": 2.3968, + "step": 19522 + }, + { + "epoch": 2.064615059221658, + "grad_norm": 8.25, + "learning_rate": 3.3115909566728262e-06, + "loss": 2.4587, + "step": 19523 + }, + { + "epoch": 2.064720812182741, + "grad_norm": 8.1875, + "learning_rate": 3.3114289739718425e-06, + "loss": 2.3761, + "step": 19524 + }, + { + "epoch": 2.0648265651438242, + "grad_norm": 9.1875, + "learning_rate": 3.3112669874631552e-06, + "loss": 2.5164, + "step": 19525 + }, + { + "epoch": 2.064932318104907, + "grad_norm": 7.8125, + "learning_rate": 3.311104997147524e-06, + "loss": 2.5019, + "step": 19526 + }, + { + "epoch": 2.06503807106599, + "grad_norm": 8.0625, + "learning_rate": 3.310943003025707e-06, + "loss": 2.5271, + "step": 19527 + }, + { + "epoch": 2.0651438240270727, + "grad_norm": 8.375, + "learning_rate": 3.3107810050984677e-06, + "loss": 2.5447, + "step": 19528 + }, + { + "epoch": 2.0652495769881556, + "grad_norm": 8.4375, + "learning_rate": 3.310619003366565e-06, + "loss": 2.2679, + "step": 19529 + }, + { + "epoch": 2.0653553299492384, + "grad_norm": 8.375, + "learning_rate": 3.3104569978307577e-06, + "loss": 2.6183, + "step": 19530 + }, + { + "epoch": 2.0654610829103213, + "grad_norm": 8.375, + "learning_rate": 3.310294988491808e-06, + "loss": 2.6708, + "step": 19531 + }, + { + "epoch": 2.0655668358714045, + "grad_norm": 8.0625, + "learning_rate": 3.3101329753504753e-06, + "loss": 2.6181, + "step": 19532 + }, + { + "epoch": 2.0656725888324874, + "grad_norm": 7.65625, + "learning_rate": 3.309970958407519e-06, + "loss": 2.4326, + "step": 19533 + }, + { + "epoch": 2.0657783417935702, + "grad_norm": 8.375, + "learning_rate": 3.309808937663701e-06, + "loss": 2.412, + "step": 19534 + }, + { + "epoch": 2.065884094754653, + "grad_norm": 8.1875, + "learning_rate": 3.3096469131197813e-06, + "loss": 2.2561, + "step": 19535 + }, + { + "epoch": 2.065989847715736, + "grad_norm": 9.0, + "learning_rate": 3.309484884776519e-06, + "loss": 2.515, + "step": 19536 + }, + { + "epoch": 2.0660956006768187, + "grad_norm": 8.25, + "learning_rate": 3.3093228526346756e-06, + "loss": 2.4915, + "step": 19537 + }, + { + "epoch": 2.066201353637902, + "grad_norm": 8.625, + "learning_rate": 3.3091608166950106e-06, + "loss": 2.6603, + "step": 19538 + }, + { + "epoch": 2.066307106598985, + "grad_norm": 8.6875, + "learning_rate": 3.3089987769582854e-06, + "loss": 2.7762, + "step": 19539 + }, + { + "epoch": 2.0664128595600677, + "grad_norm": 9.0, + "learning_rate": 3.308836733425259e-06, + "loss": 2.4492, + "step": 19540 + }, + { + "epoch": 2.0665186125211505, + "grad_norm": 8.125, + "learning_rate": 3.3086746860966935e-06, + "loss": 2.538, + "step": 19541 + }, + { + "epoch": 2.0666243654822334, + "grad_norm": 8.5625, + "learning_rate": 3.308512634973348e-06, + "loss": 2.6853, + "step": 19542 + }, + { + "epoch": 2.066730118443316, + "grad_norm": 8.4375, + "learning_rate": 3.3083505800559836e-06, + "loss": 2.5129, + "step": 19543 + }, + { + "epoch": 2.0668358714043995, + "grad_norm": 8.375, + "learning_rate": 3.308188521345361e-06, + "loss": 1.859, + "step": 19544 + }, + { + "epoch": 2.0669416243654823, + "grad_norm": 8.25, + "learning_rate": 3.3080264588422396e-06, + "loss": 2.6576, + "step": 19545 + }, + { + "epoch": 2.067047377326565, + "grad_norm": 8.6875, + "learning_rate": 3.3078643925473795e-06, + "loss": 2.812, + "step": 19546 + }, + { + "epoch": 2.067153130287648, + "grad_norm": 7.71875, + "learning_rate": 3.3077023224615434e-06, + "loss": 2.5544, + "step": 19547 + }, + { + "epoch": 2.067258883248731, + "grad_norm": 7.9375, + "learning_rate": 3.3075402485854908e-06, + "loss": 2.4389, + "step": 19548 + }, + { + "epoch": 2.0673646362098137, + "grad_norm": 8.4375, + "learning_rate": 3.3073781709199816e-06, + "loss": 2.4148, + "step": 19549 + }, + { + "epoch": 2.067470389170897, + "grad_norm": 8.0625, + "learning_rate": 3.3072160894657767e-06, + "loss": 2.65, + "step": 19550 + }, + { + "epoch": 2.06757614213198, + "grad_norm": 8.4375, + "learning_rate": 3.3070540042236375e-06, + "loss": 2.4697, + "step": 19551 + }, + { + "epoch": 2.0676818950930627, + "grad_norm": 7.96875, + "learning_rate": 3.3068919151943234e-06, + "loss": 2.4117, + "step": 19552 + }, + { + "epoch": 2.0677876480541455, + "grad_norm": 8.5625, + "learning_rate": 3.3067298223785958e-06, + "loss": 2.7337, + "step": 19553 + }, + { + "epoch": 2.0678934010152283, + "grad_norm": 8.1875, + "learning_rate": 3.306567725777215e-06, + "loss": 2.5095, + "step": 19554 + }, + { + "epoch": 2.067999153976311, + "grad_norm": 8.0625, + "learning_rate": 3.3064056253909415e-06, + "loss": 2.6163, + "step": 19555 + }, + { + "epoch": 2.0681049069373945, + "grad_norm": 11.1875, + "learning_rate": 3.3062435212205365e-06, + "loss": 2.6388, + "step": 19556 + }, + { + "epoch": 2.0682106598984773, + "grad_norm": 8.3125, + "learning_rate": 3.30608141326676e-06, + "loss": 2.4989, + "step": 19557 + }, + { + "epoch": 2.06831641285956, + "grad_norm": 8.5625, + "learning_rate": 3.3059193015303735e-06, + "loss": 2.5755, + "step": 19558 + }, + { + "epoch": 2.068422165820643, + "grad_norm": 8.5, + "learning_rate": 3.3057571860121372e-06, + "loss": 2.3308, + "step": 19559 + }, + { + "epoch": 2.068527918781726, + "grad_norm": 8.625, + "learning_rate": 3.3055950667128124e-06, + "loss": 2.6388, + "step": 19560 + }, + { + "epoch": 2.0686336717428087, + "grad_norm": 7.90625, + "learning_rate": 3.3054329436331587e-06, + "loss": 2.5294, + "step": 19561 + }, + { + "epoch": 2.0687394247038915, + "grad_norm": 8.75, + "learning_rate": 3.305270816773938e-06, + "loss": 2.6549, + "step": 19562 + }, + { + "epoch": 2.0688451776649748, + "grad_norm": 8.375, + "learning_rate": 3.305108686135911e-06, + "loss": 2.5153, + "step": 19563 + }, + { + "epoch": 2.0689509306260576, + "grad_norm": 8.3125, + "learning_rate": 3.304946551719838e-06, + "loss": 2.4966, + "step": 19564 + }, + { + "epoch": 2.0690566835871405, + "grad_norm": 8.25, + "learning_rate": 3.3047844135264795e-06, + "loss": 2.4578, + "step": 19565 + }, + { + "epoch": 2.0691624365482233, + "grad_norm": 8.375, + "learning_rate": 3.3046222715565973e-06, + "loss": 2.6369, + "step": 19566 + }, + { + "epoch": 2.069268189509306, + "grad_norm": 8.25, + "learning_rate": 3.3044601258109523e-06, + "loss": 2.5428, + "step": 19567 + }, + { + "epoch": 2.069373942470389, + "grad_norm": 8.4375, + "learning_rate": 3.3042979762903037e-06, + "loss": 2.618, + "step": 19568 + }, + { + "epoch": 2.0694796954314723, + "grad_norm": 8.0, + "learning_rate": 3.3041358229954145e-06, + "loss": 2.6793, + "step": 19569 + }, + { + "epoch": 2.069585448392555, + "grad_norm": 8.3125, + "learning_rate": 3.3039736659270455e-06, + "loss": 2.5289, + "step": 19570 + }, + { + "epoch": 2.069691201353638, + "grad_norm": 8.75, + "learning_rate": 3.3038115050859554e-06, + "loss": 2.5585, + "step": 19571 + }, + { + "epoch": 2.0697969543147208, + "grad_norm": 8.5, + "learning_rate": 3.3036493404729077e-06, + "loss": 2.6663, + "step": 19572 + }, + { + "epoch": 2.0699027072758036, + "grad_norm": 8.25, + "learning_rate": 3.3034871720886626e-06, + "loss": 2.5011, + "step": 19573 + }, + { + "epoch": 2.0700084602368864, + "grad_norm": 8.6875, + "learning_rate": 3.30332499993398e-06, + "loss": 2.5942, + "step": 19574 + }, + { + "epoch": 2.0701142131979697, + "grad_norm": 8.4375, + "learning_rate": 3.3031628240096213e-06, + "loss": 2.2507, + "step": 19575 + }, + { + "epoch": 2.0702199661590526, + "grad_norm": 7.8125, + "learning_rate": 3.3030006443163488e-06, + "loss": 2.4766, + "step": 19576 + }, + { + "epoch": 2.0703257191201354, + "grad_norm": 8.4375, + "learning_rate": 3.3028384608549223e-06, + "loss": 2.4909, + "step": 19577 + }, + { + "epoch": 2.0704314720812182, + "grad_norm": 8.6875, + "learning_rate": 3.302676273626104e-06, + "loss": 2.3978, + "step": 19578 + }, + { + "epoch": 2.070537225042301, + "grad_norm": 8.625, + "learning_rate": 3.3025140826306535e-06, + "loss": 2.6387, + "step": 19579 + }, + { + "epoch": 2.070642978003384, + "grad_norm": 8.125, + "learning_rate": 3.3023518878693328e-06, + "loss": 2.4179, + "step": 19580 + }, + { + "epoch": 2.070748730964467, + "grad_norm": 8.0, + "learning_rate": 3.3021896893429027e-06, + "loss": 2.5348, + "step": 19581 + }, + { + "epoch": 2.07085448392555, + "grad_norm": 8.6875, + "learning_rate": 3.3020274870521253e-06, + "loss": 2.556, + "step": 19582 + }, + { + "epoch": 2.070960236886633, + "grad_norm": 8.5625, + "learning_rate": 3.301865280997759e-06, + "loss": 2.4809, + "step": 19583 + }, + { + "epoch": 2.0710659898477157, + "grad_norm": 9.25, + "learning_rate": 3.3017030711805683e-06, + "loss": 2.4445, + "step": 19584 + }, + { + "epoch": 2.0711717428087986, + "grad_norm": 7.96875, + "learning_rate": 3.301540857601313e-06, + "loss": 2.3626, + "step": 19585 + }, + { + "epoch": 2.0712774957698814, + "grad_norm": 8.0625, + "learning_rate": 3.301378640260754e-06, + "loss": 2.359, + "step": 19586 + }, + { + "epoch": 2.0713832487309647, + "grad_norm": 7.90625, + "learning_rate": 3.301216419159653e-06, + "loss": 2.5085, + "step": 19587 + }, + { + "epoch": 2.0714890016920475, + "grad_norm": 8.5625, + "learning_rate": 3.3010541942987707e-06, + "loss": 2.5542, + "step": 19588 + }, + { + "epoch": 2.0715947546531304, + "grad_norm": 8.1875, + "learning_rate": 3.30089196567887e-06, + "loss": 2.6001, + "step": 19589 + }, + { + "epoch": 2.071700507614213, + "grad_norm": 8.3125, + "learning_rate": 3.3007297333007087e-06, + "loss": 2.4548, + "step": 19590 + }, + { + "epoch": 2.071806260575296, + "grad_norm": 9.125, + "learning_rate": 3.3005674971650507e-06, + "loss": 2.6145, + "step": 19591 + }, + { + "epoch": 2.071912013536379, + "grad_norm": 7.96875, + "learning_rate": 3.300405257272658e-06, + "loss": 2.4775, + "step": 19592 + }, + { + "epoch": 2.0720177664974617, + "grad_norm": 8.375, + "learning_rate": 3.3002430136242902e-06, + "loss": 2.4472, + "step": 19593 + }, + { + "epoch": 2.072123519458545, + "grad_norm": 8.375, + "learning_rate": 3.300080766220709e-06, + "loss": 2.6126, + "step": 19594 + }, + { + "epoch": 2.072229272419628, + "grad_norm": 8.125, + "learning_rate": 3.2999185150626765e-06, + "loss": 2.6463, + "step": 19595 + }, + { + "epoch": 2.0723350253807107, + "grad_norm": 8.5, + "learning_rate": 3.2997562601509534e-06, + "loss": 2.3809, + "step": 19596 + }, + { + "epoch": 2.0724407783417935, + "grad_norm": 8.875, + "learning_rate": 3.299594001486301e-06, + "loss": 2.6366, + "step": 19597 + }, + { + "epoch": 2.0725465313028764, + "grad_norm": 8.5625, + "learning_rate": 3.2994317390694814e-06, + "loss": 2.5748, + "step": 19598 + }, + { + "epoch": 2.072652284263959, + "grad_norm": 8.5625, + "learning_rate": 3.2992694729012555e-06, + "loss": 2.5751, + "step": 19599 + }, + { + "epoch": 2.0727580372250425, + "grad_norm": 8.3125, + "learning_rate": 3.299107202982384e-06, + "loss": 2.5417, + "step": 19600 + }, + { + "epoch": 2.0728637901861253, + "grad_norm": 8.3125, + "learning_rate": 3.29894492931363e-06, + "loss": 2.4863, + "step": 19601 + }, + { + "epoch": 2.072969543147208, + "grad_norm": 9.0, + "learning_rate": 3.2987826518957542e-06, + "loss": 2.5576, + "step": 19602 + }, + { + "epoch": 2.073075296108291, + "grad_norm": 8.375, + "learning_rate": 3.2986203707295178e-06, + "loss": 2.523, + "step": 19603 + }, + { + "epoch": 2.073181049069374, + "grad_norm": 7.875, + "learning_rate": 3.2984580858156836e-06, + "loss": 2.48, + "step": 19604 + }, + { + "epoch": 2.0732868020304567, + "grad_norm": 7.84375, + "learning_rate": 3.298295797155011e-06, + "loss": 2.531, + "step": 19605 + }, + { + "epoch": 2.07339255499154, + "grad_norm": 8.3125, + "learning_rate": 3.298133504748263e-06, + "loss": 2.4015, + "step": 19606 + }, + { + "epoch": 2.073498307952623, + "grad_norm": 8.875, + "learning_rate": 3.2979712085962013e-06, + "loss": 2.5802, + "step": 19607 + }, + { + "epoch": 2.0736040609137056, + "grad_norm": 8.375, + "learning_rate": 3.2978089086995867e-06, + "loss": 2.6154, + "step": 19608 + }, + { + "epoch": 2.0737098138747885, + "grad_norm": 8.625, + "learning_rate": 3.2976466050591817e-06, + "loss": 2.5459, + "step": 19609 + }, + { + "epoch": 2.0738155668358713, + "grad_norm": 7.9375, + "learning_rate": 3.297484297675746e-06, + "loss": 2.5559, + "step": 19610 + }, + { + "epoch": 2.073921319796954, + "grad_norm": 8.0, + "learning_rate": 3.297321986550044e-06, + "loss": 2.5061, + "step": 19611 + }, + { + "epoch": 2.0740270727580374, + "grad_norm": 8.5, + "learning_rate": 3.297159671682836e-06, + "loss": 2.5512, + "step": 19612 + }, + { + "epoch": 2.0741328257191203, + "grad_norm": 8.4375, + "learning_rate": 3.2969973530748828e-06, + "loss": 2.7191, + "step": 19613 + }, + { + "epoch": 2.074238578680203, + "grad_norm": 8.1875, + "learning_rate": 3.296835030726948e-06, + "loss": 2.6786, + "step": 19614 + }, + { + "epoch": 2.074344331641286, + "grad_norm": 8.5625, + "learning_rate": 3.2966727046397917e-06, + "loss": 2.6586, + "step": 19615 + }, + { + "epoch": 2.074450084602369, + "grad_norm": 8.0625, + "learning_rate": 3.2965103748141764e-06, + "loss": 2.4059, + "step": 19616 + }, + { + "epoch": 2.0745558375634516, + "grad_norm": 8.5625, + "learning_rate": 3.2963480412508634e-06, + "loss": 2.7972, + "step": 19617 + }, + { + "epoch": 2.074661590524535, + "grad_norm": 8.75, + "learning_rate": 3.296185703950615e-06, + "loss": 2.524, + "step": 19618 + }, + { + "epoch": 2.0747673434856178, + "grad_norm": 7.90625, + "learning_rate": 3.2960233629141926e-06, + "loss": 2.6513, + "step": 19619 + }, + { + "epoch": 2.0748730964467006, + "grad_norm": 8.1875, + "learning_rate": 3.2958610181423578e-06, + "loss": 2.6574, + "step": 19620 + }, + { + "epoch": 2.0749788494077834, + "grad_norm": 8.4375, + "learning_rate": 3.2956986696358733e-06, + "loss": 2.5967, + "step": 19621 + }, + { + "epoch": 2.0750846023688663, + "grad_norm": 8.3125, + "learning_rate": 3.2955363173955e-06, + "loss": 2.5486, + "step": 19622 + }, + { + "epoch": 2.075190355329949, + "grad_norm": 8.0, + "learning_rate": 3.2953739614220004e-06, + "loss": 2.4763, + "step": 19623 + }, + { + "epoch": 2.075296108291032, + "grad_norm": 8.4375, + "learning_rate": 3.295211601716136e-06, + "loss": 2.523, + "step": 19624 + }, + { + "epoch": 2.0754018612521152, + "grad_norm": 8.9375, + "learning_rate": 3.295049238278669e-06, + "loss": 2.6598, + "step": 19625 + }, + { + "epoch": 2.075507614213198, + "grad_norm": 8.375, + "learning_rate": 3.294886871110361e-06, + "loss": 2.6291, + "step": 19626 + }, + { + "epoch": 2.075613367174281, + "grad_norm": 7.8125, + "learning_rate": 3.2947245002119737e-06, + "loss": 2.6969, + "step": 19627 + }, + { + "epoch": 2.0757191201353637, + "grad_norm": 8.375, + "learning_rate": 3.2945621255842697e-06, + "loss": 2.3938, + "step": 19628 + }, + { + "epoch": 2.0758248730964466, + "grad_norm": 8.0625, + "learning_rate": 3.2943997472280105e-06, + "loss": 2.5329, + "step": 19629 + }, + { + "epoch": 2.0759306260575294, + "grad_norm": 8.5625, + "learning_rate": 3.294237365143958e-06, + "loss": 2.5577, + "step": 19630 + }, + { + "epoch": 2.0760363790186127, + "grad_norm": 8.625, + "learning_rate": 3.2940749793328745e-06, + "loss": 2.5325, + "step": 19631 + }, + { + "epoch": 2.0761421319796955, + "grad_norm": 8.125, + "learning_rate": 3.293912589795522e-06, + "loss": 2.4414, + "step": 19632 + }, + { + "epoch": 2.0762478849407784, + "grad_norm": 8.5625, + "learning_rate": 3.2937501965326625e-06, + "loss": 2.5478, + "step": 19633 + }, + { + "epoch": 2.0763536379018612, + "grad_norm": 8.375, + "learning_rate": 3.293587799545058e-06, + "loss": 2.52, + "step": 19634 + }, + { + "epoch": 2.076459390862944, + "grad_norm": 8.75, + "learning_rate": 3.2934253988334695e-06, + "loss": 2.6725, + "step": 19635 + }, + { + "epoch": 2.076565143824027, + "grad_norm": 8.4375, + "learning_rate": 3.2932629943986612e-06, + "loss": 2.4723, + "step": 19636 + }, + { + "epoch": 2.07667089678511, + "grad_norm": 8.875, + "learning_rate": 3.293100586241394e-06, + "loss": 2.7818, + "step": 19637 + }, + { + "epoch": 2.076776649746193, + "grad_norm": 8.0625, + "learning_rate": 3.2929381743624293e-06, + "loss": 2.4309, + "step": 19638 + }, + { + "epoch": 2.076882402707276, + "grad_norm": 8.4375, + "learning_rate": 3.292775758762531e-06, + "loss": 2.5789, + "step": 19639 + }, + { + "epoch": 2.0769881556683587, + "grad_norm": 8.375, + "learning_rate": 3.29261333944246e-06, + "loss": 2.6561, + "step": 19640 + }, + { + "epoch": 2.0770939086294415, + "grad_norm": 8.125, + "learning_rate": 3.292450916402979e-06, + "loss": 2.6949, + "step": 19641 + }, + { + "epoch": 2.0771996615905244, + "grad_norm": 8.3125, + "learning_rate": 3.29228848964485e-06, + "loss": 2.7046, + "step": 19642 + }, + { + "epoch": 2.077305414551607, + "grad_norm": 8.5, + "learning_rate": 3.292126059168835e-06, + "loss": 2.5541, + "step": 19643 + }, + { + "epoch": 2.0774111675126905, + "grad_norm": 8.625, + "learning_rate": 3.2919636249756957e-06, + "loss": 2.5655, + "step": 19644 + }, + { + "epoch": 2.0775169204737733, + "grad_norm": 8.25, + "learning_rate": 3.2918011870661954e-06, + "loss": 2.461, + "step": 19645 + }, + { + "epoch": 2.077622673434856, + "grad_norm": 8.6875, + "learning_rate": 3.291638745441096e-06, + "loss": 2.5157, + "step": 19646 + }, + { + "epoch": 2.077728426395939, + "grad_norm": 8.1875, + "learning_rate": 3.29147630010116e-06, + "loss": 2.4481, + "step": 19647 + }, + { + "epoch": 2.077834179357022, + "grad_norm": 8.25, + "learning_rate": 3.2913138510471493e-06, + "loss": 2.5137, + "step": 19648 + }, + { + "epoch": 2.0779399323181047, + "grad_norm": 8.875, + "learning_rate": 3.291151398279826e-06, + "loss": 2.6054, + "step": 19649 + }, + { + "epoch": 2.078045685279188, + "grad_norm": 8.4375, + "learning_rate": 3.2909889417999535e-06, + "loss": 2.5512, + "step": 19650 + }, + { + "epoch": 2.078151438240271, + "grad_norm": 7.9375, + "learning_rate": 3.2908264816082926e-06, + "loss": 2.6603, + "step": 19651 + }, + { + "epoch": 2.0782571912013537, + "grad_norm": 8.125, + "learning_rate": 3.290664017705607e-06, + "loss": 2.4698, + "step": 19652 + }, + { + "epoch": 2.0783629441624365, + "grad_norm": 9.25, + "learning_rate": 3.2905015500926586e-06, + "loss": 2.4189, + "step": 19653 + }, + { + "epoch": 2.0784686971235193, + "grad_norm": 8.6875, + "learning_rate": 3.290339078770209e-06, + "loss": 2.77, + "step": 19654 + }, + { + "epoch": 2.078574450084602, + "grad_norm": 8.5, + "learning_rate": 3.2901766037390216e-06, + "loss": 2.5554, + "step": 19655 + }, + { + "epoch": 2.0786802030456855, + "grad_norm": 9.0625, + "learning_rate": 3.290014124999859e-06, + "loss": 2.4367, + "step": 19656 + }, + { + "epoch": 2.0787859560067683, + "grad_norm": 8.125, + "learning_rate": 3.289851642553483e-06, + "loss": 2.5278, + "step": 19657 + }, + { + "epoch": 2.078891708967851, + "grad_norm": 8.6875, + "learning_rate": 3.289689156400656e-06, + "loss": 2.4437, + "step": 19658 + }, + { + "epoch": 2.078997461928934, + "grad_norm": 8.0625, + "learning_rate": 3.2895266665421417e-06, + "loss": 2.4471, + "step": 19659 + }, + { + "epoch": 2.079103214890017, + "grad_norm": 8.3125, + "learning_rate": 3.2893641729787007e-06, + "loss": 2.3186, + "step": 19660 + }, + { + "epoch": 2.0792089678510997, + "grad_norm": 9.0, + "learning_rate": 3.289201675711097e-06, + "loss": 2.3883, + "step": 19661 + }, + { + "epoch": 2.079314720812183, + "grad_norm": 8.375, + "learning_rate": 3.289039174740092e-06, + "loss": 2.5497, + "step": 19662 + }, + { + "epoch": 2.0794204737732658, + "grad_norm": 8.3125, + "learning_rate": 3.288876670066449e-06, + "loss": 2.4775, + "step": 19663 + }, + { + "epoch": 2.0795262267343486, + "grad_norm": 8.4375, + "learning_rate": 3.2887141616909307e-06, + "loss": 2.5386, + "step": 19664 + }, + { + "epoch": 2.0796319796954315, + "grad_norm": 8.8125, + "learning_rate": 3.2885516496143e-06, + "loss": 2.5806, + "step": 19665 + }, + { + "epoch": 2.0797377326565143, + "grad_norm": 8.3125, + "learning_rate": 3.288389133837318e-06, + "loss": 2.5384, + "step": 19666 + }, + { + "epoch": 2.079843485617597, + "grad_norm": 8.125, + "learning_rate": 3.2882266143607483e-06, + "loss": 2.4899, + "step": 19667 + }, + { + "epoch": 2.0799492385786804, + "grad_norm": 8.5625, + "learning_rate": 3.288064091185354e-06, + "loss": 2.6075, + "step": 19668 + }, + { + "epoch": 2.0800549915397633, + "grad_norm": 8.0625, + "learning_rate": 3.287901564311897e-06, + "loss": 2.509, + "step": 19669 + }, + { + "epoch": 2.080160744500846, + "grad_norm": 9.5625, + "learning_rate": 3.2877390337411398e-06, + "loss": 2.8383, + "step": 19670 + }, + { + "epoch": 2.080266497461929, + "grad_norm": 8.5, + "learning_rate": 3.2875764994738455e-06, + "loss": 2.5707, + "step": 19671 + }, + { + "epoch": 2.0803722504230118, + "grad_norm": 8.125, + "learning_rate": 3.2874139615107776e-06, + "loss": 2.5697, + "step": 19672 + }, + { + "epoch": 2.0804780033840946, + "grad_norm": 8.1875, + "learning_rate": 3.2872514198526974e-06, + "loss": 2.4795, + "step": 19673 + }, + { + "epoch": 2.0805837563451774, + "grad_norm": 8.3125, + "learning_rate": 3.287088874500368e-06, + "loss": 2.6843, + "step": 19674 + }, + { + "epoch": 2.0806895093062607, + "grad_norm": 7.875, + "learning_rate": 3.286926325454554e-06, + "loss": 2.3072, + "step": 19675 + }, + { + "epoch": 2.0807952622673436, + "grad_norm": 9.0, + "learning_rate": 3.2867637727160146e-06, + "loss": 2.7494, + "step": 19676 + }, + { + "epoch": 2.0809010152284264, + "grad_norm": 7.8125, + "learning_rate": 3.2866012162855155e-06, + "loss": 2.5181, + "step": 19677 + }, + { + "epoch": 2.0810067681895092, + "grad_norm": 8.4375, + "learning_rate": 3.286438656163819e-06, + "loss": 2.4876, + "step": 19678 + }, + { + "epoch": 2.081112521150592, + "grad_norm": 8.3125, + "learning_rate": 3.2862760923516867e-06, + "loss": 2.5509, + "step": 19679 + }, + { + "epoch": 2.081218274111675, + "grad_norm": 8.5625, + "learning_rate": 3.286113524849882e-06, + "loss": 2.4805, + "step": 19680 + }, + { + "epoch": 2.081324027072758, + "grad_norm": 8.5, + "learning_rate": 3.285950953659169e-06, + "loss": 2.4356, + "step": 19681 + }, + { + "epoch": 2.081429780033841, + "grad_norm": 8.1875, + "learning_rate": 3.285788378780309e-06, + "loss": 2.4555, + "step": 19682 + }, + { + "epoch": 2.081535532994924, + "grad_norm": 7.90625, + "learning_rate": 3.2856258002140663e-06, + "loss": 2.4914, + "step": 19683 + }, + { + "epoch": 2.0816412859560067, + "grad_norm": 8.25, + "learning_rate": 3.2854632179612024e-06, + "loss": 2.6235, + "step": 19684 + }, + { + "epoch": 2.0817470389170896, + "grad_norm": 8.9375, + "learning_rate": 3.285300632022481e-06, + "loss": 2.5148, + "step": 19685 + }, + { + "epoch": 2.0818527918781724, + "grad_norm": 8.25, + "learning_rate": 3.2851380423986644e-06, + "loss": 2.5333, + "step": 19686 + }, + { + "epoch": 2.0819585448392557, + "grad_norm": 8.5, + "learning_rate": 3.2849754490905174e-06, + "loss": 2.6587, + "step": 19687 + }, + { + "epoch": 2.0820642978003385, + "grad_norm": 8.6875, + "learning_rate": 3.284812852098801e-06, + "loss": 2.7312, + "step": 19688 + }, + { + "epoch": 2.0821700507614214, + "grad_norm": 9.125, + "learning_rate": 3.2846502514242772e-06, + "loss": 2.6051, + "step": 19689 + }, + { + "epoch": 2.082275803722504, + "grad_norm": 8.8125, + "learning_rate": 3.284487647067713e-06, + "loss": 2.4255, + "step": 19690 + }, + { + "epoch": 2.082381556683587, + "grad_norm": 8.0, + "learning_rate": 3.284325039029868e-06, + "loss": 2.3264, + "step": 19691 + }, + { + "epoch": 2.08248730964467, + "grad_norm": 8.625, + "learning_rate": 3.2841624273115065e-06, + "loss": 2.623, + "step": 19692 + }, + { + "epoch": 2.082593062605753, + "grad_norm": 8.4375, + "learning_rate": 3.283999811913392e-06, + "loss": 2.5497, + "step": 19693 + }, + { + "epoch": 2.082698815566836, + "grad_norm": 8.4375, + "learning_rate": 3.2838371928362867e-06, + "loss": 2.4734, + "step": 19694 + }, + { + "epoch": 2.082804568527919, + "grad_norm": 8.1875, + "learning_rate": 3.2836745700809538e-06, + "loss": 2.4424, + "step": 19695 + }, + { + "epoch": 2.0829103214890017, + "grad_norm": 8.0625, + "learning_rate": 3.2835119436481566e-06, + "loss": 2.5424, + "step": 19696 + }, + { + "epoch": 2.0830160744500845, + "grad_norm": 8.375, + "learning_rate": 3.283349313538659e-06, + "loss": 2.6124, + "step": 19697 + }, + { + "epoch": 2.0831218274111674, + "grad_norm": 7.90625, + "learning_rate": 3.2831866797532226e-06, + "loss": 2.5966, + "step": 19698 + }, + { + "epoch": 2.0832275803722506, + "grad_norm": 8.0625, + "learning_rate": 3.2830240422926113e-06, + "loss": 2.3717, + "step": 19699 + }, + { + "epoch": 2.0833333333333335, + "grad_norm": 8.75, + "learning_rate": 3.2828614011575892e-06, + "loss": 2.6997, + "step": 19700 + }, + { + "epoch": 2.0834390862944163, + "grad_norm": 7.6875, + "learning_rate": 3.2826987563489188e-06, + "loss": 2.4966, + "step": 19701 + }, + { + "epoch": 2.083544839255499, + "grad_norm": 7.90625, + "learning_rate": 3.2825361078673623e-06, + "loss": 2.4281, + "step": 19702 + }, + { + "epoch": 2.083650592216582, + "grad_norm": 8.6875, + "learning_rate": 3.282373455713685e-06, + "loss": 2.5407, + "step": 19703 + }, + { + "epoch": 2.083756345177665, + "grad_norm": 9.25, + "learning_rate": 3.282210799888649e-06, + "loss": 2.7926, + "step": 19704 + }, + { + "epoch": 2.0838620981387477, + "grad_norm": 8.8125, + "learning_rate": 3.2820481403930167e-06, + "loss": 2.4994, + "step": 19705 + }, + { + "epoch": 2.083967851099831, + "grad_norm": 8.25, + "learning_rate": 3.281885477227552e-06, + "loss": 2.4349, + "step": 19706 + }, + { + "epoch": 2.084073604060914, + "grad_norm": 8.5, + "learning_rate": 3.2817228103930197e-06, + "loss": 2.5134, + "step": 19707 + }, + { + "epoch": 2.0841793570219966, + "grad_norm": 8.5, + "learning_rate": 3.281560139890181e-06, + "loss": 2.588, + "step": 19708 + }, + { + "epoch": 2.0842851099830795, + "grad_norm": 8.375, + "learning_rate": 3.281397465719801e-06, + "loss": 2.5023, + "step": 19709 + }, + { + "epoch": 2.0843908629441623, + "grad_norm": 9.0625, + "learning_rate": 3.2812347878826423e-06, + "loss": 2.6946, + "step": 19710 + }, + { + "epoch": 2.084496615905245, + "grad_norm": 8.6875, + "learning_rate": 3.2810721063794674e-06, + "loss": 2.4308, + "step": 19711 + }, + { + "epoch": 2.0846023688663284, + "grad_norm": 8.25, + "learning_rate": 3.2809094212110416e-06, + "loss": 2.7751, + "step": 19712 + }, + { + "epoch": 2.0847081218274113, + "grad_norm": 8.625, + "learning_rate": 3.2807467323781273e-06, + "loss": 2.5784, + "step": 19713 + }, + { + "epoch": 2.084813874788494, + "grad_norm": 8.3125, + "learning_rate": 3.280584039881487e-06, + "loss": 2.26, + "step": 19714 + }, + { + "epoch": 2.084919627749577, + "grad_norm": 8.3125, + "learning_rate": 3.280421343721885e-06, + "loss": 2.6692, + "step": 19715 + }, + { + "epoch": 2.08502538071066, + "grad_norm": 9.125, + "learning_rate": 3.2802586439000853e-06, + "loss": 2.6494, + "step": 19716 + }, + { + "epoch": 2.0851311336717426, + "grad_norm": 8.25, + "learning_rate": 3.2800959404168514e-06, + "loss": 2.4868, + "step": 19717 + }, + { + "epoch": 2.085236886632826, + "grad_norm": 8.0625, + "learning_rate": 3.2799332332729457e-06, + "loss": 2.4007, + "step": 19718 + }, + { + "epoch": 2.0853426395939088, + "grad_norm": 8.25, + "learning_rate": 3.279770522469132e-06, + "loss": 2.6501, + "step": 19719 + }, + { + "epoch": 2.0854483925549916, + "grad_norm": 8.1875, + "learning_rate": 3.2796078080061756e-06, + "loss": 2.5191, + "step": 19720 + }, + { + "epoch": 2.0855541455160744, + "grad_norm": 9.1875, + "learning_rate": 3.2794450898848367e-06, + "loss": 2.492, + "step": 19721 + }, + { + "epoch": 2.0856598984771573, + "grad_norm": 9.25, + "learning_rate": 3.279282368105882e-06, + "loss": 2.5541, + "step": 19722 + }, + { + "epoch": 2.08576565143824, + "grad_norm": 8.4375, + "learning_rate": 3.2791196426700738e-06, + "loss": 2.6398, + "step": 19723 + }, + { + "epoch": 2.0858714043993234, + "grad_norm": 8.0, + "learning_rate": 3.2789569135781747e-06, + "loss": 2.4599, + "step": 19724 + }, + { + "epoch": 2.0859771573604062, + "grad_norm": 8.6875, + "learning_rate": 3.2787941808309502e-06, + "loss": 2.2125, + "step": 19725 + }, + { + "epoch": 2.086082910321489, + "grad_norm": 8.6875, + "learning_rate": 3.278631444429163e-06, + "loss": 2.5688, + "step": 19726 + }, + { + "epoch": 2.086188663282572, + "grad_norm": 8.5, + "learning_rate": 3.2784687043735767e-06, + "loss": 2.4317, + "step": 19727 + }, + { + "epoch": 2.0862944162436547, + "grad_norm": 7.75, + "learning_rate": 3.278305960664956e-06, + "loss": 2.4634, + "step": 19728 + }, + { + "epoch": 2.0864001692047376, + "grad_norm": 9.5625, + "learning_rate": 3.278143213304063e-06, + "loss": 2.5031, + "step": 19729 + }, + { + "epoch": 2.086505922165821, + "grad_norm": 8.75, + "learning_rate": 3.277980462291662e-06, + "loss": 2.6795, + "step": 19730 + }, + { + "epoch": 2.0866116751269037, + "grad_norm": 8.25, + "learning_rate": 3.2778177076285177e-06, + "loss": 2.3586, + "step": 19731 + }, + { + "epoch": 2.0867174280879865, + "grad_norm": 8.0, + "learning_rate": 3.2776549493153926e-06, + "loss": 2.4519, + "step": 19732 + }, + { + "epoch": 2.0868231810490694, + "grad_norm": 8.6875, + "learning_rate": 3.2774921873530506e-06, + "loss": 2.6323, + "step": 19733 + }, + { + "epoch": 2.0869289340101522, + "grad_norm": 8.4375, + "learning_rate": 3.2773294217422557e-06, + "loss": 2.3356, + "step": 19734 + }, + { + "epoch": 2.087034686971235, + "grad_norm": 8.0625, + "learning_rate": 3.277166652483772e-06, + "loss": 2.3787, + "step": 19735 + }, + { + "epoch": 2.087140439932318, + "grad_norm": 8.1875, + "learning_rate": 3.2770038795783633e-06, + "loss": 2.6284, + "step": 19736 + }, + { + "epoch": 2.087246192893401, + "grad_norm": 8.0625, + "learning_rate": 3.2768411030267923e-06, + "loss": 2.413, + "step": 19737 + }, + { + "epoch": 2.087351945854484, + "grad_norm": 8.0, + "learning_rate": 3.276678322829825e-06, + "loss": 2.6238, + "step": 19738 + }, + { + "epoch": 2.087457698815567, + "grad_norm": 8.375, + "learning_rate": 3.276515538988223e-06, + "loss": 2.5152, + "step": 19739 + }, + { + "epoch": 2.0875634517766497, + "grad_norm": 7.96875, + "learning_rate": 3.2763527515027516e-06, + "loss": 2.3642, + "step": 19740 + }, + { + "epoch": 2.0876692047377325, + "grad_norm": 7.96875, + "learning_rate": 3.2761899603741743e-06, + "loss": 2.297, + "step": 19741 + }, + { + "epoch": 2.0877749576988154, + "grad_norm": 8.1875, + "learning_rate": 3.2760271656032537e-06, + "loss": 2.4856, + "step": 19742 + }, + { + "epoch": 2.0878807106598987, + "grad_norm": 8.1875, + "learning_rate": 3.2758643671907563e-06, + "loss": 2.5365, + "step": 19743 + }, + { + "epoch": 2.0879864636209815, + "grad_norm": 8.0625, + "learning_rate": 3.2757015651374445e-06, + "loss": 2.3535, + "step": 19744 + }, + { + "epoch": 2.0880922165820643, + "grad_norm": 8.625, + "learning_rate": 3.2755387594440828e-06, + "loss": 2.649, + "step": 19745 + }, + { + "epoch": 2.088197969543147, + "grad_norm": 8.0, + "learning_rate": 3.275375950111434e-06, + "loss": 2.5547, + "step": 19746 + }, + { + "epoch": 2.08830372250423, + "grad_norm": 8.75, + "learning_rate": 3.275213137140264e-06, + "loss": 2.689, + "step": 19747 + }, + { + "epoch": 2.088409475465313, + "grad_norm": 8.3125, + "learning_rate": 3.2750503205313357e-06, + "loss": 2.7005, + "step": 19748 + }, + { + "epoch": 2.088515228426396, + "grad_norm": 8.5, + "learning_rate": 3.2748875002854126e-06, + "loss": 2.3551, + "step": 19749 + }, + { + "epoch": 2.088620981387479, + "grad_norm": 8.3125, + "learning_rate": 3.2747246764032596e-06, + "loss": 2.7, + "step": 19750 + }, + { + "epoch": 2.088726734348562, + "grad_norm": 8.0625, + "learning_rate": 3.2745618488856408e-06, + "loss": 2.1608, + "step": 19751 + }, + { + "epoch": 2.0888324873096447, + "grad_norm": 8.5625, + "learning_rate": 3.2743990177333197e-06, + "loss": 2.6214, + "step": 19752 + }, + { + "epoch": 2.0889382402707275, + "grad_norm": 8.3125, + "learning_rate": 3.2742361829470604e-06, + "loss": 2.3872, + "step": 19753 + }, + { + "epoch": 2.0890439932318103, + "grad_norm": 7.8125, + "learning_rate": 3.2740733445276287e-06, + "loss": 2.493, + "step": 19754 + }, + { + "epoch": 2.089149746192893, + "grad_norm": 8.625, + "learning_rate": 3.2739105024757868e-06, + "loss": 2.4852, + "step": 19755 + }, + { + "epoch": 2.0892554991539765, + "grad_norm": 8.8125, + "learning_rate": 3.2737476567922983e-06, + "loss": 2.5793, + "step": 19756 + }, + { + "epoch": 2.0893612521150593, + "grad_norm": 8.125, + "learning_rate": 3.27358480747793e-06, + "loss": 2.4993, + "step": 19757 + }, + { + "epoch": 2.089467005076142, + "grad_norm": 8.3125, + "learning_rate": 3.273421954533444e-06, + "loss": 2.4863, + "step": 19758 + }, + { + "epoch": 2.089572758037225, + "grad_norm": 8.6875, + "learning_rate": 3.273259097959604e-06, + "loss": 2.7439, + "step": 19759 + }, + { + "epoch": 2.089678510998308, + "grad_norm": 8.0625, + "learning_rate": 3.273096237757176e-06, + "loss": 2.4071, + "step": 19760 + }, + { + "epoch": 2.0897842639593907, + "grad_norm": 8.3125, + "learning_rate": 3.272933373926924e-06, + "loss": 2.67, + "step": 19761 + }, + { + "epoch": 2.089890016920474, + "grad_norm": 8.75, + "learning_rate": 3.2727705064696118e-06, + "loss": 2.5486, + "step": 19762 + }, + { + "epoch": 2.0899957698815568, + "grad_norm": 8.0, + "learning_rate": 3.2726076353860036e-06, + "loss": 2.5125, + "step": 19763 + }, + { + "epoch": 2.0901015228426396, + "grad_norm": 7.9375, + "learning_rate": 3.272444760676864e-06, + "loss": 2.3089, + "step": 19764 + }, + { + "epoch": 2.0902072758037225, + "grad_norm": 8.3125, + "learning_rate": 3.2722818823429566e-06, + "loss": 2.3778, + "step": 19765 + }, + { + "epoch": 2.0903130287648053, + "grad_norm": 7.71875, + "learning_rate": 3.2721190003850466e-06, + "loss": 2.5391, + "step": 19766 + }, + { + "epoch": 2.090418781725888, + "grad_norm": 8.4375, + "learning_rate": 3.2719561148038974e-06, + "loss": 2.5212, + "step": 19767 + }, + { + "epoch": 2.0905245346869714, + "grad_norm": 7.9375, + "learning_rate": 3.2717932256002737e-06, + "loss": 2.4244, + "step": 19768 + }, + { + "epoch": 2.0906302876480543, + "grad_norm": 8.375, + "learning_rate": 3.2716303327749403e-06, + "loss": 2.4684, + "step": 19769 + }, + { + "epoch": 2.090736040609137, + "grad_norm": 9.3125, + "learning_rate": 3.271467436328662e-06, + "loss": 2.5278, + "step": 19770 + }, + { + "epoch": 2.09084179357022, + "grad_norm": 8.25, + "learning_rate": 3.271304536262202e-06, + "loss": 2.4131, + "step": 19771 + }, + { + "epoch": 2.0909475465313028, + "grad_norm": 8.625, + "learning_rate": 3.271141632576325e-06, + "loss": 2.4264, + "step": 19772 + }, + { + "epoch": 2.0910532994923856, + "grad_norm": 8.1875, + "learning_rate": 3.2709787252717963e-06, + "loss": 2.6101, + "step": 19773 + }, + { + "epoch": 2.091159052453469, + "grad_norm": 8.1875, + "learning_rate": 3.2708158143493796e-06, + "loss": 2.429, + "step": 19774 + }, + { + "epoch": 2.0912648054145517, + "grad_norm": 8.125, + "learning_rate": 3.270652899809839e-06, + "loss": 2.4833, + "step": 19775 + }, + { + "epoch": 2.0913705583756346, + "grad_norm": 8.4375, + "learning_rate": 3.27048998165394e-06, + "loss": 2.6261, + "step": 19776 + }, + { + "epoch": 2.0914763113367174, + "grad_norm": 8.1875, + "learning_rate": 3.2703270598824464e-06, + "loss": 2.5218, + "step": 19777 + }, + { + "epoch": 2.0915820642978002, + "grad_norm": 7.96875, + "learning_rate": 3.2701641344961226e-06, + "loss": 2.4902, + "step": 19778 + }, + { + "epoch": 2.091687817258883, + "grad_norm": 8.25, + "learning_rate": 3.270001205495734e-06, + "loss": 2.4765, + "step": 19779 + }, + { + "epoch": 2.0917935702199664, + "grad_norm": 9.1875, + "learning_rate": 3.2698382728820454e-06, + "loss": 2.6775, + "step": 19780 + }, + { + "epoch": 2.091899323181049, + "grad_norm": 8.6875, + "learning_rate": 3.2696753366558196e-06, + "loss": 2.5153, + "step": 19781 + }, + { + "epoch": 2.092005076142132, + "grad_norm": 8.375, + "learning_rate": 3.269512396817823e-06, + "loss": 2.4429, + "step": 19782 + }, + { + "epoch": 2.092110829103215, + "grad_norm": 8.8125, + "learning_rate": 3.2693494533688192e-06, + "loss": 2.5837, + "step": 19783 + }, + { + "epoch": 2.0922165820642977, + "grad_norm": 8.75, + "learning_rate": 3.2691865063095723e-06, + "loss": 2.6302, + "step": 19784 + }, + { + "epoch": 2.0923223350253806, + "grad_norm": 8.125, + "learning_rate": 3.2690235556408486e-06, + "loss": 2.5725, + "step": 19785 + }, + { + "epoch": 2.0924280879864634, + "grad_norm": 8.5625, + "learning_rate": 3.2688606013634113e-06, + "loss": 2.5806, + "step": 19786 + }, + { + "epoch": 2.0925338409475467, + "grad_norm": 8.3125, + "learning_rate": 3.268697643478026e-06, + "loss": 2.6114, + "step": 19787 + }, + { + "epoch": 2.0926395939086295, + "grad_norm": 8.6875, + "learning_rate": 3.2685346819854564e-06, + "loss": 2.6786, + "step": 19788 + }, + { + "epoch": 2.0927453468697124, + "grad_norm": 8.5, + "learning_rate": 3.268371716886469e-06, + "loss": 2.4913, + "step": 19789 + }, + { + "epoch": 2.092851099830795, + "grad_norm": 8.6875, + "learning_rate": 3.268208748181827e-06, + "loss": 2.5457, + "step": 19790 + }, + { + "epoch": 2.092956852791878, + "grad_norm": 8.5, + "learning_rate": 3.268045775872295e-06, + "loss": 2.5688, + "step": 19791 + }, + { + "epoch": 2.093062605752961, + "grad_norm": 8.4375, + "learning_rate": 3.267882799958638e-06, + "loss": 2.6181, + "step": 19792 + }, + { + "epoch": 2.093168358714044, + "grad_norm": 8.6875, + "learning_rate": 3.2677198204416226e-06, + "loss": 2.5418, + "step": 19793 + }, + { + "epoch": 2.093274111675127, + "grad_norm": 8.75, + "learning_rate": 3.267556837322011e-06, + "loss": 2.5494, + "step": 19794 + }, + { + "epoch": 2.09337986463621, + "grad_norm": 8.5, + "learning_rate": 3.2673938506005686e-06, + "loss": 2.6116, + "step": 19795 + }, + { + "epoch": 2.0934856175972927, + "grad_norm": 8.5, + "learning_rate": 3.267230860278062e-06, + "loss": 2.5284, + "step": 19796 + }, + { + "epoch": 2.0935913705583755, + "grad_norm": 8.75, + "learning_rate": 3.2670678663552534e-06, + "loss": 2.5224, + "step": 19797 + }, + { + "epoch": 2.0936971235194584, + "grad_norm": 8.875, + "learning_rate": 3.26690486883291e-06, + "loss": 2.7039, + "step": 19798 + }, + { + "epoch": 2.0938028764805416, + "grad_norm": 8.0625, + "learning_rate": 3.2667418677117956e-06, + "loss": 2.4606, + "step": 19799 + }, + { + "epoch": 2.0939086294416245, + "grad_norm": 8.625, + "learning_rate": 3.2665788629926747e-06, + "loss": 2.5521, + "step": 19800 + }, + { + "epoch": 2.0940143824027073, + "grad_norm": 8.1875, + "learning_rate": 3.266415854676313e-06, + "loss": 2.5589, + "step": 19801 + }, + { + "epoch": 2.09412013536379, + "grad_norm": 7.6875, + "learning_rate": 3.2662528427634755e-06, + "loss": 2.4973, + "step": 19802 + }, + { + "epoch": 2.094225888324873, + "grad_norm": 8.0625, + "learning_rate": 3.2660898272549264e-06, + "loss": 2.5242, + "step": 19803 + }, + { + "epoch": 2.094331641285956, + "grad_norm": 8.1875, + "learning_rate": 3.2659268081514306e-06, + "loss": 2.5586, + "step": 19804 + }, + { + "epoch": 2.094437394247039, + "grad_norm": 8.5625, + "learning_rate": 3.2657637854537542e-06, + "loss": 2.4294, + "step": 19805 + }, + { + "epoch": 2.094543147208122, + "grad_norm": 8.3125, + "learning_rate": 3.265600759162662e-06, + "loss": 2.6352, + "step": 19806 + }, + { + "epoch": 2.094648900169205, + "grad_norm": 8.375, + "learning_rate": 3.2654377292789174e-06, + "loss": 2.4859, + "step": 19807 + }, + { + "epoch": 2.0947546531302876, + "grad_norm": 8.125, + "learning_rate": 3.2652746958032876e-06, + "loss": 2.5124, + "step": 19808 + }, + { + "epoch": 2.0948604060913705, + "grad_norm": 8.0625, + "learning_rate": 3.2651116587365363e-06, + "loss": 2.4389, + "step": 19809 + }, + { + "epoch": 2.0949661590524533, + "grad_norm": 8.75, + "learning_rate": 3.2649486180794286e-06, + "loss": 2.5859, + "step": 19810 + }, + { + "epoch": 2.0950719120135366, + "grad_norm": 8.5625, + "learning_rate": 3.2647855738327305e-06, + "loss": 2.7954, + "step": 19811 + }, + { + "epoch": 2.0951776649746194, + "grad_norm": 8.625, + "learning_rate": 3.2646225259972063e-06, + "loss": 2.4565, + "step": 19812 + }, + { + "epoch": 2.0952834179357023, + "grad_norm": 8.0, + "learning_rate": 3.26445947457362e-06, + "loss": 2.3876, + "step": 19813 + }, + { + "epoch": 2.095389170896785, + "grad_norm": 8.5625, + "learning_rate": 3.264296419562739e-06, + "loss": 2.6325, + "step": 19814 + }, + { + "epoch": 2.095494923857868, + "grad_norm": 8.6875, + "learning_rate": 3.264133360965328e-06, + "loss": 2.4849, + "step": 19815 + }, + { + "epoch": 2.095600676818951, + "grad_norm": 8.8125, + "learning_rate": 3.2639702987821515e-06, + "loss": 2.2235, + "step": 19816 + }, + { + "epoch": 2.0957064297800336, + "grad_norm": 8.1875, + "learning_rate": 3.2638072330139746e-06, + "loss": 2.5369, + "step": 19817 + }, + { + "epoch": 2.095812182741117, + "grad_norm": 8.4375, + "learning_rate": 3.263644163661563e-06, + "loss": 2.4394, + "step": 19818 + }, + { + "epoch": 2.0959179357021998, + "grad_norm": 8.25, + "learning_rate": 3.263481090725681e-06, + "loss": 2.5155, + "step": 19819 + }, + { + "epoch": 2.0960236886632826, + "grad_norm": 9.125, + "learning_rate": 3.2633180142070954e-06, + "loss": 2.6439, + "step": 19820 + }, + { + "epoch": 2.0961294416243654, + "grad_norm": 8.4375, + "learning_rate": 3.263154934106569e-06, + "loss": 2.6113, + "step": 19821 + }, + { + "epoch": 2.0962351945854483, + "grad_norm": 8.4375, + "learning_rate": 3.2629918504248703e-06, + "loss": 2.5307, + "step": 19822 + }, + { + "epoch": 2.096340947546531, + "grad_norm": 7.90625, + "learning_rate": 3.262828763162762e-06, + "loss": 2.4421, + "step": 19823 + }, + { + "epoch": 2.0964467005076144, + "grad_norm": 7.90625, + "learning_rate": 3.2626656723210108e-06, + "loss": 2.3439, + "step": 19824 + }, + { + "epoch": 2.0965524534686972, + "grad_norm": 8.375, + "learning_rate": 3.2625025779003817e-06, + "loss": 2.4959, + "step": 19825 + }, + { + "epoch": 2.09665820642978, + "grad_norm": 8.25, + "learning_rate": 3.2623394799016393e-06, + "loss": 2.5694, + "step": 19826 + }, + { + "epoch": 2.096763959390863, + "grad_norm": 8.4375, + "learning_rate": 3.26217637832555e-06, + "loss": 2.6289, + "step": 19827 + }, + { + "epoch": 2.0968697123519457, + "grad_norm": 8.4375, + "learning_rate": 3.262013273172878e-06, + "loss": 2.2228, + "step": 19828 + }, + { + "epoch": 2.0969754653130286, + "grad_norm": 8.5625, + "learning_rate": 3.26185016444439e-06, + "loss": 2.4951, + "step": 19829 + }, + { + "epoch": 2.097081218274112, + "grad_norm": 8.0, + "learning_rate": 3.2616870521408502e-06, + "loss": 2.6309, + "step": 19830 + }, + { + "epoch": 2.0971869712351947, + "grad_norm": 8.25, + "learning_rate": 3.2615239362630245e-06, + "loss": 2.5102, + "step": 19831 + }, + { + "epoch": 2.0972927241962775, + "grad_norm": 9.125, + "learning_rate": 3.261360816811679e-06, + "loss": 2.8645, + "step": 19832 + }, + { + "epoch": 2.0973984771573604, + "grad_norm": 8.4375, + "learning_rate": 3.2611976937875782e-06, + "loss": 2.7128, + "step": 19833 + }, + { + "epoch": 2.0975042301184432, + "grad_norm": 8.625, + "learning_rate": 3.2610345671914882e-06, + "loss": 2.4163, + "step": 19834 + }, + { + "epoch": 2.097609983079526, + "grad_norm": 8.3125, + "learning_rate": 3.260871437024174e-06, + "loss": 2.6357, + "step": 19835 + }, + { + "epoch": 2.0977157360406093, + "grad_norm": 8.5, + "learning_rate": 3.260708303286402e-06, + "loss": 2.4196, + "step": 19836 + }, + { + "epoch": 2.097821489001692, + "grad_norm": 8.9375, + "learning_rate": 3.2605451659789365e-06, + "loss": 2.4869, + "step": 19837 + }, + { + "epoch": 2.097927241962775, + "grad_norm": 8.5625, + "learning_rate": 3.260382025102543e-06, + "loss": 2.8021, + "step": 19838 + }, + { + "epoch": 2.098032994923858, + "grad_norm": 8.0, + "learning_rate": 3.2602188806579883e-06, + "loss": 2.4443, + "step": 19839 + }, + { + "epoch": 2.0981387478849407, + "grad_norm": 8.75, + "learning_rate": 3.2600557326460374e-06, + "loss": 2.3161, + "step": 19840 + }, + { + "epoch": 2.0982445008460235, + "grad_norm": 9.125, + "learning_rate": 3.259892581067456e-06, + "loss": 2.7753, + "step": 19841 + }, + { + "epoch": 2.098350253807107, + "grad_norm": 8.3125, + "learning_rate": 3.2597294259230082e-06, + "loss": 2.3587, + "step": 19842 + }, + { + "epoch": 2.0984560067681897, + "grad_norm": 8.75, + "learning_rate": 3.2595662672134627e-06, + "loss": 2.5744, + "step": 19843 + }, + { + "epoch": 2.0985617597292725, + "grad_norm": 7.71875, + "learning_rate": 3.2594031049395826e-06, + "loss": 2.3032, + "step": 19844 + }, + { + "epoch": 2.0986675126903553, + "grad_norm": 8.5625, + "learning_rate": 3.259239939102134e-06, + "loss": 2.5207, + "step": 19845 + }, + { + "epoch": 2.098773265651438, + "grad_norm": 8.125, + "learning_rate": 3.2590767697018834e-06, + "loss": 2.5956, + "step": 19846 + }, + { + "epoch": 2.098879018612521, + "grad_norm": 8.375, + "learning_rate": 3.2589135967395957e-06, + "loss": 2.6539, + "step": 19847 + }, + { + "epoch": 2.098984771573604, + "grad_norm": 8.125, + "learning_rate": 3.2587504202160364e-06, + "loss": 2.6135, + "step": 19848 + }, + { + "epoch": 2.099090524534687, + "grad_norm": 8.25, + "learning_rate": 3.258587240131972e-06, + "loss": 2.4581, + "step": 19849 + }, + { + "epoch": 2.09919627749577, + "grad_norm": 8.3125, + "learning_rate": 3.2584240564881684e-06, + "loss": 2.5119, + "step": 19850 + }, + { + "epoch": 2.099302030456853, + "grad_norm": 8.25, + "learning_rate": 3.25826086928539e-06, + "loss": 2.4996, + "step": 19851 + }, + { + "epoch": 2.0994077834179357, + "grad_norm": 8.5, + "learning_rate": 3.2580976785244045e-06, + "loss": 2.6599, + "step": 19852 + }, + { + "epoch": 2.0995135363790185, + "grad_norm": 9.875, + "learning_rate": 3.2579344842059763e-06, + "loss": 2.7277, + "step": 19853 + }, + { + "epoch": 2.0996192893401013, + "grad_norm": 8.375, + "learning_rate": 3.257771286330871e-06, + "loss": 2.3715, + "step": 19854 + }, + { + "epoch": 2.0997250423011846, + "grad_norm": 7.875, + "learning_rate": 3.257608084899856e-06, + "loss": 2.3643, + "step": 19855 + }, + { + "epoch": 2.0998307952622675, + "grad_norm": 8.375, + "learning_rate": 3.2574448799136955e-06, + "loss": 2.5785, + "step": 19856 + }, + { + "epoch": 2.0999365482233503, + "grad_norm": 8.125, + "learning_rate": 3.257281671373156e-06, + "loss": 2.4605, + "step": 19857 + }, + { + "epoch": 2.100042301184433, + "grad_norm": 8.75, + "learning_rate": 3.2571184592790028e-06, + "loss": 2.6545, + "step": 19858 + }, + { + "epoch": 2.100148054145516, + "grad_norm": 8.625, + "learning_rate": 3.256955243632003e-06, + "loss": 2.4336, + "step": 19859 + }, + { + "epoch": 2.100253807106599, + "grad_norm": 9.0, + "learning_rate": 3.256792024432922e-06, + "loss": 2.76, + "step": 19860 + }, + { + "epoch": 2.100359560067682, + "grad_norm": 8.5625, + "learning_rate": 3.256628801682525e-06, + "loss": 2.5013, + "step": 19861 + }, + { + "epoch": 2.100465313028765, + "grad_norm": 8.6875, + "learning_rate": 3.256465575381579e-06, + "loss": 2.7379, + "step": 19862 + }, + { + "epoch": 2.1005710659898478, + "grad_norm": 8.5625, + "learning_rate": 3.256302345530849e-06, + "loss": 2.394, + "step": 19863 + }, + { + "epoch": 2.1006768189509306, + "grad_norm": 8.25, + "learning_rate": 3.2561391121311013e-06, + "loss": 2.5976, + "step": 19864 + }, + { + "epoch": 2.1007825719120135, + "grad_norm": 8.375, + "learning_rate": 3.2559758751831028e-06, + "loss": 2.4042, + "step": 19865 + }, + { + "epoch": 2.1008883248730963, + "grad_norm": 8.25, + "learning_rate": 3.2558126346876172e-06, + "loss": 2.5291, + "step": 19866 + }, + { + "epoch": 2.100994077834179, + "grad_norm": 8.25, + "learning_rate": 3.255649390645413e-06, + "loss": 2.5641, + "step": 19867 + }, + { + "epoch": 2.1010998307952624, + "grad_norm": 8.6875, + "learning_rate": 3.2554861430572554e-06, + "loss": 2.5497, + "step": 19868 + }, + { + "epoch": 2.1012055837563453, + "grad_norm": 8.375, + "learning_rate": 3.2553228919239106e-06, + "loss": 2.3737, + "step": 19869 + }, + { + "epoch": 2.101311336717428, + "grad_norm": 9.0, + "learning_rate": 3.2551596372461437e-06, + "loss": 2.8186, + "step": 19870 + }, + { + "epoch": 2.101417089678511, + "grad_norm": 8.3125, + "learning_rate": 3.2549963790247218e-06, + "loss": 2.6336, + "step": 19871 + }, + { + "epoch": 2.1015228426395938, + "grad_norm": 8.3125, + "learning_rate": 3.254833117260411e-06, + "loss": 2.595, + "step": 19872 + }, + { + "epoch": 2.1016285956006766, + "grad_norm": 8.1875, + "learning_rate": 3.254669851953976e-06, + "loss": 2.6235, + "step": 19873 + }, + { + "epoch": 2.10173434856176, + "grad_norm": 8.5, + "learning_rate": 3.2545065831061847e-06, + "loss": 2.6778, + "step": 19874 + }, + { + "epoch": 2.1018401015228427, + "grad_norm": 8.4375, + "learning_rate": 3.2543433107178025e-06, + "loss": 2.4836, + "step": 19875 + }, + { + "epoch": 2.1019458544839256, + "grad_norm": 8.125, + "learning_rate": 3.254180034789596e-06, + "loss": 2.3565, + "step": 19876 + }, + { + "epoch": 2.1020516074450084, + "grad_norm": 8.5625, + "learning_rate": 3.2540167553223308e-06, + "loss": 2.4793, + "step": 19877 + }, + { + "epoch": 2.1021573604060912, + "grad_norm": 9.3125, + "learning_rate": 3.253853472316773e-06, + "loss": 2.5697, + "step": 19878 + }, + { + "epoch": 2.102263113367174, + "grad_norm": 8.0625, + "learning_rate": 3.2536901857736902e-06, + "loss": 2.4674, + "step": 19879 + }, + { + "epoch": 2.1023688663282574, + "grad_norm": 8.6875, + "learning_rate": 3.2535268956938463e-06, + "loss": 2.5604, + "step": 19880 + }, + { + "epoch": 2.10247461928934, + "grad_norm": 8.3125, + "learning_rate": 3.2533636020780096e-06, + "loss": 2.441, + "step": 19881 + }, + { + "epoch": 2.102580372250423, + "grad_norm": 8.625, + "learning_rate": 3.2532003049269457e-06, + "loss": 2.4561, + "step": 19882 + }, + { + "epoch": 2.102686125211506, + "grad_norm": 8.8125, + "learning_rate": 3.2530370042414205e-06, + "loss": 2.3881, + "step": 19883 + }, + { + "epoch": 2.1027918781725887, + "grad_norm": 8.3125, + "learning_rate": 3.2528737000222004e-06, + "loss": 2.709, + "step": 19884 + }, + { + "epoch": 2.1028976311336716, + "grad_norm": 8.125, + "learning_rate": 3.2527103922700525e-06, + "loss": 2.6055, + "step": 19885 + }, + { + "epoch": 2.103003384094755, + "grad_norm": 8.25, + "learning_rate": 3.2525470809857425e-06, + "loss": 2.6511, + "step": 19886 + }, + { + "epoch": 2.1031091370558377, + "grad_norm": 8.125, + "learning_rate": 3.2523837661700365e-06, + "loss": 2.5604, + "step": 19887 + }, + { + "epoch": 2.1032148900169205, + "grad_norm": 8.375, + "learning_rate": 3.252220447823702e-06, + "loss": 2.7011, + "step": 19888 + }, + { + "epoch": 2.1033206429780034, + "grad_norm": 8.625, + "learning_rate": 3.2520571259475033e-06, + "loss": 2.6711, + "step": 19889 + }, + { + "epoch": 2.103426395939086, + "grad_norm": 8.3125, + "learning_rate": 3.251893800542209e-06, + "loss": 2.6274, + "step": 19890 + }, + { + "epoch": 2.103532148900169, + "grad_norm": 7.6875, + "learning_rate": 3.2517304716085845e-06, + "loss": 2.4012, + "step": 19891 + }, + { + "epoch": 2.1036379018612523, + "grad_norm": 9.125, + "learning_rate": 3.2515671391473963e-06, + "loss": 2.6764, + "step": 19892 + }, + { + "epoch": 2.103743654822335, + "grad_norm": 8.4375, + "learning_rate": 3.25140380315941e-06, + "loss": 2.6432, + "step": 19893 + }, + { + "epoch": 2.103849407783418, + "grad_norm": 8.3125, + "learning_rate": 3.2512404636453945e-06, + "loss": 2.6459, + "step": 19894 + }, + { + "epoch": 2.103955160744501, + "grad_norm": 8.625, + "learning_rate": 3.2510771206061144e-06, + "loss": 2.6394, + "step": 19895 + }, + { + "epoch": 2.1040609137055837, + "grad_norm": 8.75, + "learning_rate": 3.250913774042336e-06, + "loss": 2.473, + "step": 19896 + }, + { + "epoch": 2.1041666666666665, + "grad_norm": 7.96875, + "learning_rate": 3.2507504239548265e-06, + "loss": 2.6493, + "step": 19897 + }, + { + "epoch": 2.1042724196277494, + "grad_norm": 8.125, + "learning_rate": 3.2505870703443525e-06, + "loss": 2.5605, + "step": 19898 + }, + { + "epoch": 2.1043781725888326, + "grad_norm": 8.0, + "learning_rate": 3.25042371321168e-06, + "loss": 2.2592, + "step": 19899 + }, + { + "epoch": 2.1044839255499155, + "grad_norm": 8.125, + "learning_rate": 3.250260352557576e-06, + "loss": 2.5548, + "step": 19900 + }, + { + "epoch": 2.1045896785109983, + "grad_norm": 8.1875, + "learning_rate": 3.250096988382807e-06, + "loss": 1.9792, + "step": 19901 + }, + { + "epoch": 2.104695431472081, + "grad_norm": 7.90625, + "learning_rate": 3.2499336206881406e-06, + "loss": 2.5427, + "step": 19902 + }, + { + "epoch": 2.104801184433164, + "grad_norm": 8.625, + "learning_rate": 3.249770249474341e-06, + "loss": 2.5198, + "step": 19903 + }, + { + "epoch": 2.104906937394247, + "grad_norm": 8.3125, + "learning_rate": 3.249606874742177e-06, + "loss": 2.5697, + "step": 19904 + }, + { + "epoch": 2.10501269035533, + "grad_norm": 8.0, + "learning_rate": 3.2494434964924137e-06, + "loss": 2.5492, + "step": 19905 + }, + { + "epoch": 2.105118443316413, + "grad_norm": 8.375, + "learning_rate": 3.2492801147258196e-06, + "loss": 2.6388, + "step": 19906 + }, + { + "epoch": 2.105224196277496, + "grad_norm": 8.1875, + "learning_rate": 3.2491167294431604e-06, + "loss": 2.4692, + "step": 19907 + }, + { + "epoch": 2.1053299492385786, + "grad_norm": 7.90625, + "learning_rate": 3.2489533406452022e-06, + "loss": 2.4386, + "step": 19908 + }, + { + "epoch": 2.1054357021996615, + "grad_norm": 8.3125, + "learning_rate": 3.248789948332713e-06, + "loss": 2.531, + "step": 19909 + }, + { + "epoch": 2.1055414551607443, + "grad_norm": 8.9375, + "learning_rate": 3.2486265525064574e-06, + "loss": 2.5295, + "step": 19910 + }, + { + "epoch": 2.1056472081218276, + "grad_norm": 8.875, + "learning_rate": 3.248463153167205e-06, + "loss": 2.6159, + "step": 19911 + }, + { + "epoch": 2.1057529610829104, + "grad_norm": 8.8125, + "learning_rate": 3.2482997503157206e-06, + "loss": 2.6202, + "step": 19912 + }, + { + "epoch": 2.1058587140439933, + "grad_norm": 8.0625, + "learning_rate": 3.248136343952771e-06, + "loss": 2.551, + "step": 19913 + }, + { + "epoch": 2.105964467005076, + "grad_norm": 9.0, + "learning_rate": 3.2479729340791242e-06, + "loss": 2.6367, + "step": 19914 + }, + { + "epoch": 2.106070219966159, + "grad_norm": 8.6875, + "learning_rate": 3.247809520695546e-06, + "loss": 2.5161, + "step": 19915 + }, + { + "epoch": 2.106175972927242, + "grad_norm": 8.5625, + "learning_rate": 3.247646103802804e-06, + "loss": 2.5752, + "step": 19916 + }, + { + "epoch": 2.106281725888325, + "grad_norm": 8.4375, + "learning_rate": 3.247482683401665e-06, + "loss": 2.5547, + "step": 19917 + }, + { + "epoch": 2.106387478849408, + "grad_norm": 8.4375, + "learning_rate": 3.2473192594928943e-06, + "loss": 2.4568, + "step": 19918 + }, + { + "epoch": 2.1064932318104908, + "grad_norm": 8.3125, + "learning_rate": 3.24715583207726e-06, + "loss": 2.5497, + "step": 19919 + }, + { + "epoch": 2.1065989847715736, + "grad_norm": 8.375, + "learning_rate": 3.24699240115553e-06, + "loss": 2.5236, + "step": 19920 + }, + { + "epoch": 2.1067047377326564, + "grad_norm": 8.375, + "learning_rate": 3.2468289667284697e-06, + "loss": 2.6066, + "step": 19921 + }, + { + "epoch": 2.1068104906937393, + "grad_norm": 8.6875, + "learning_rate": 3.2466655287968463e-06, + "loss": 2.7919, + "step": 19922 + }, + { + "epoch": 2.1069162436548226, + "grad_norm": 9.3125, + "learning_rate": 3.246502087361427e-06, + "loss": 2.6526, + "step": 19923 + }, + { + "epoch": 2.1070219966159054, + "grad_norm": 10.1875, + "learning_rate": 3.2463386424229785e-06, + "loss": 2.5787, + "step": 19924 + }, + { + "epoch": 2.1071277495769882, + "grad_norm": 8.1875, + "learning_rate": 3.246175193982268e-06, + "loss": 2.6157, + "step": 19925 + }, + { + "epoch": 2.107233502538071, + "grad_norm": 8.5, + "learning_rate": 3.2460117420400638e-06, + "loss": 2.3447, + "step": 19926 + }, + { + "epoch": 2.107339255499154, + "grad_norm": 8.3125, + "learning_rate": 3.24584828659713e-06, + "loss": 2.5953, + "step": 19927 + }, + { + "epoch": 2.1074450084602367, + "grad_norm": 8.9375, + "learning_rate": 3.2456848276542354e-06, + "loss": 2.6588, + "step": 19928 + }, + { + "epoch": 2.1075507614213196, + "grad_norm": 7.65625, + "learning_rate": 3.2455213652121474e-06, + "loss": 2.4443, + "step": 19929 + }, + { + "epoch": 2.107656514382403, + "grad_norm": 8.3125, + "learning_rate": 3.2453578992716326e-06, + "loss": 2.5159, + "step": 19930 + }, + { + "epoch": 2.1077622673434857, + "grad_norm": 8.125, + "learning_rate": 3.2451944298334574e-06, + "loss": 2.3006, + "step": 19931 + }, + { + "epoch": 2.1078680203045685, + "grad_norm": 8.6875, + "learning_rate": 3.2450309568983906e-06, + "loss": 2.6308, + "step": 19932 + }, + { + "epoch": 2.1079737732656514, + "grad_norm": 8.9375, + "learning_rate": 3.2448674804671974e-06, + "loss": 2.5605, + "step": 19933 + }, + { + "epoch": 2.1080795262267342, + "grad_norm": 8.375, + "learning_rate": 3.244704000540646e-06, + "loss": 2.473, + "step": 19934 + }, + { + "epoch": 2.108185279187817, + "grad_norm": 8.3125, + "learning_rate": 3.244540517119503e-06, + "loss": 2.7165, + "step": 19935 + }, + { + "epoch": 2.1082910321489003, + "grad_norm": 7.84375, + "learning_rate": 3.244377030204536e-06, + "loss": 2.3487, + "step": 19936 + }, + { + "epoch": 2.108396785109983, + "grad_norm": 8.125, + "learning_rate": 3.2442135397965124e-06, + "loss": 2.3111, + "step": 19937 + }, + { + "epoch": 2.108502538071066, + "grad_norm": 9.1875, + "learning_rate": 3.244050045896199e-06, + "loss": 2.8818, + "step": 19938 + }, + { + "epoch": 2.108608291032149, + "grad_norm": 8.625, + "learning_rate": 3.2438865485043623e-06, + "loss": 2.509, + "step": 19939 + }, + { + "epoch": 2.1087140439932317, + "grad_norm": 8.1875, + "learning_rate": 3.2437230476217707e-06, + "loss": 2.5468, + "step": 19940 + }, + { + "epoch": 2.1088197969543145, + "grad_norm": 8.125, + "learning_rate": 3.2435595432491912e-06, + "loss": 2.5855, + "step": 19941 + }, + { + "epoch": 2.108925549915398, + "grad_norm": 8.25, + "learning_rate": 3.2433960353873912e-06, + "loss": 2.5567, + "step": 19942 + }, + { + "epoch": 2.1090313028764807, + "grad_norm": 9.0625, + "learning_rate": 3.2432325240371364e-06, + "loss": 2.7038, + "step": 19943 + }, + { + "epoch": 2.1091370558375635, + "grad_norm": 8.5625, + "learning_rate": 3.2430690091991968e-06, + "loss": 2.6001, + "step": 19944 + }, + { + "epoch": 2.1092428087986463, + "grad_norm": 8.75, + "learning_rate": 3.242905490874337e-06, + "loss": 2.4567, + "step": 19945 + }, + { + "epoch": 2.109348561759729, + "grad_norm": 8.9375, + "learning_rate": 3.2427419690633267e-06, + "loss": 2.4653, + "step": 19946 + }, + { + "epoch": 2.109454314720812, + "grad_norm": 8.5625, + "learning_rate": 3.242578443766931e-06, + "loss": 2.5452, + "step": 19947 + }, + { + "epoch": 2.1095600676818953, + "grad_norm": 8.0625, + "learning_rate": 3.242414914985919e-06, + "loss": 2.5302, + "step": 19948 + }, + { + "epoch": 2.109665820642978, + "grad_norm": 8.3125, + "learning_rate": 3.242251382721058e-06, + "loss": 2.6174, + "step": 19949 + }, + { + "epoch": 2.109771573604061, + "grad_norm": 8.5625, + "learning_rate": 3.242087846973114e-06, + "loss": 2.4629, + "step": 19950 + }, + { + "epoch": 2.109877326565144, + "grad_norm": 8.0625, + "learning_rate": 3.241924307742855e-06, + "loss": 2.6057, + "step": 19951 + }, + { + "epoch": 2.1099830795262267, + "grad_norm": 8.375, + "learning_rate": 3.2417607650310496e-06, + "loss": 2.7028, + "step": 19952 + }, + { + "epoch": 2.1100888324873095, + "grad_norm": 10.0625, + "learning_rate": 3.241597218838464e-06, + "loss": 2.7958, + "step": 19953 + }, + { + "epoch": 2.110194585448393, + "grad_norm": 8.0625, + "learning_rate": 3.2414336691658648e-06, + "loss": 2.4598, + "step": 19954 + }, + { + "epoch": 2.1103003384094756, + "grad_norm": 8.25, + "learning_rate": 3.241270116014022e-06, + "loss": 2.6016, + "step": 19955 + }, + { + "epoch": 2.1104060913705585, + "grad_norm": 8.0, + "learning_rate": 3.241106559383701e-06, + "loss": 2.4646, + "step": 19956 + }, + { + "epoch": 2.1105118443316413, + "grad_norm": 8.9375, + "learning_rate": 3.2409429992756702e-06, + "loss": 2.4976, + "step": 19957 + }, + { + "epoch": 2.110617597292724, + "grad_norm": 8.625, + "learning_rate": 3.2407794356906973e-06, + "loss": 2.588, + "step": 19958 + }, + { + "epoch": 2.110723350253807, + "grad_norm": 8.3125, + "learning_rate": 3.2406158686295487e-06, + "loss": 2.5218, + "step": 19959 + }, + { + "epoch": 2.11082910321489, + "grad_norm": 8.1875, + "learning_rate": 3.2404522980929937e-06, + "loss": 2.4658, + "step": 19960 + }, + { + "epoch": 2.110934856175973, + "grad_norm": 8.5, + "learning_rate": 3.2402887240817987e-06, + "loss": 2.6516, + "step": 19961 + }, + { + "epoch": 2.111040609137056, + "grad_norm": 7.6875, + "learning_rate": 3.240125146596731e-06, + "loss": 2.4536, + "step": 19962 + }, + { + "epoch": 2.1111463620981388, + "grad_norm": 8.875, + "learning_rate": 3.2399615656385584e-06, + "loss": 2.5134, + "step": 19963 + }, + { + "epoch": 2.1112521150592216, + "grad_norm": 9.25, + "learning_rate": 3.239797981208049e-06, + "loss": 2.6205, + "step": 19964 + }, + { + "epoch": 2.1113578680203045, + "grad_norm": 8.3125, + "learning_rate": 3.2396343933059716e-06, + "loss": 2.5068, + "step": 19965 + }, + { + "epoch": 2.1114636209813873, + "grad_norm": 8.875, + "learning_rate": 3.239470801933091e-06, + "loss": 2.5297, + "step": 19966 + }, + { + "epoch": 2.1115693739424706, + "grad_norm": 8.1875, + "learning_rate": 3.2393072070901773e-06, + "loss": 2.3756, + "step": 19967 + }, + { + "epoch": 2.1116751269035534, + "grad_norm": 8.25, + "learning_rate": 3.2391436087779966e-06, + "loss": 2.4278, + "step": 19968 + }, + { + "epoch": 2.1117808798646363, + "grad_norm": 8.125, + "learning_rate": 3.2389800069973173e-06, + "loss": 2.5964, + "step": 19969 + }, + { + "epoch": 2.111886632825719, + "grad_norm": 8.1875, + "learning_rate": 3.238816401748907e-06, + "loss": 2.4391, + "step": 19970 + }, + { + "epoch": 2.111992385786802, + "grad_norm": 8.0625, + "learning_rate": 3.2386527930335342e-06, + "loss": 2.4317, + "step": 19971 + }, + { + "epoch": 2.1120981387478848, + "grad_norm": 8.3125, + "learning_rate": 3.238489180851965e-06, + "loss": 2.5818, + "step": 19972 + }, + { + "epoch": 2.112203891708968, + "grad_norm": 8.8125, + "learning_rate": 3.2383255652049683e-06, + "loss": 2.5617, + "step": 19973 + }, + { + "epoch": 2.112309644670051, + "grad_norm": 8.1875, + "learning_rate": 3.238161946093312e-06, + "loss": 2.2775, + "step": 19974 + }, + { + "epoch": 2.1124153976311337, + "grad_norm": 8.25, + "learning_rate": 3.237998323517764e-06, + "loss": 2.5235, + "step": 19975 + }, + { + "epoch": 2.1125211505922166, + "grad_norm": 8.4375, + "learning_rate": 3.237834697479091e-06, + "loss": 2.4575, + "step": 19976 + }, + { + "epoch": 2.1126269035532994, + "grad_norm": 8.9375, + "learning_rate": 3.2376710679780614e-06, + "loss": 2.5202, + "step": 19977 + }, + { + "epoch": 2.1127326565143822, + "grad_norm": 8.0625, + "learning_rate": 3.2375074350154434e-06, + "loss": 2.5039, + "step": 19978 + }, + { + "epoch": 2.112838409475465, + "grad_norm": 7.9375, + "learning_rate": 3.2373437985920046e-06, + "loss": 2.4106, + "step": 19979 + }, + { + "epoch": 2.1129441624365484, + "grad_norm": 9.0, + "learning_rate": 3.2371801587085133e-06, + "loss": 2.4057, + "step": 19980 + }, + { + "epoch": 2.113049915397631, + "grad_norm": 8.8125, + "learning_rate": 3.237016515365736e-06, + "loss": 2.547, + "step": 19981 + }, + { + "epoch": 2.113155668358714, + "grad_norm": 8.625, + "learning_rate": 3.236852868564443e-06, + "loss": 2.6047, + "step": 19982 + }, + { + "epoch": 2.113261421319797, + "grad_norm": 8.6875, + "learning_rate": 3.2366892183053998e-06, + "loss": 2.5878, + "step": 19983 + }, + { + "epoch": 2.1133671742808797, + "grad_norm": 8.5625, + "learning_rate": 3.2365255645893764e-06, + "loss": 2.4773, + "step": 19984 + }, + { + "epoch": 2.1134729272419626, + "grad_norm": 8.875, + "learning_rate": 3.2363619074171388e-06, + "loss": 2.6773, + "step": 19985 + }, + { + "epoch": 2.113578680203046, + "grad_norm": 8.1875, + "learning_rate": 3.2361982467894565e-06, + "loss": 2.384, + "step": 19986 + }, + { + "epoch": 2.1136844331641287, + "grad_norm": 8.1875, + "learning_rate": 3.236034582707097e-06, + "loss": 2.4671, + "step": 19987 + }, + { + "epoch": 2.1137901861252115, + "grad_norm": 7.90625, + "learning_rate": 3.2358709151708278e-06, + "loss": 2.3158, + "step": 19988 + }, + { + "epoch": 2.1138959390862944, + "grad_norm": 8.9375, + "learning_rate": 3.235707244181417e-06, + "loss": 2.5563, + "step": 19989 + }, + { + "epoch": 2.114001692047377, + "grad_norm": 8.9375, + "learning_rate": 3.2355435697396336e-06, + "loss": 2.9208, + "step": 19990 + }, + { + "epoch": 2.11410744500846, + "grad_norm": 8.625, + "learning_rate": 3.235379891846245e-06, + "loss": 2.3751, + "step": 19991 + }, + { + "epoch": 2.1142131979695433, + "grad_norm": 9.0, + "learning_rate": 3.235216210502019e-06, + "loss": 2.3007, + "step": 19992 + }, + { + "epoch": 2.114318950930626, + "grad_norm": 8.375, + "learning_rate": 3.2350525257077246e-06, + "loss": 2.5684, + "step": 19993 + }, + { + "epoch": 2.114424703891709, + "grad_norm": 8.4375, + "learning_rate": 3.23488883746413e-06, + "loss": 2.5995, + "step": 19994 + }, + { + "epoch": 2.114530456852792, + "grad_norm": 8.625, + "learning_rate": 3.234725145772001e-06, + "loss": 2.5467, + "step": 19995 + }, + { + "epoch": 2.1146362098138747, + "grad_norm": 8.8125, + "learning_rate": 3.2345614506321077e-06, + "loss": 2.4266, + "step": 19996 + }, + { + "epoch": 2.1147419627749575, + "grad_norm": 8.3125, + "learning_rate": 3.2343977520452186e-06, + "loss": 2.8297, + "step": 19997 + }, + { + "epoch": 2.114847715736041, + "grad_norm": 8.3125, + "learning_rate": 3.2342340500121007e-06, + "loss": 2.4171, + "step": 19998 + }, + { + "epoch": 2.1149534686971236, + "grad_norm": 8.0, + "learning_rate": 3.2340703445335224e-06, + "loss": 2.5229, + "step": 19999 + }, + { + "epoch": 2.1150592216582065, + "grad_norm": 8.5, + "learning_rate": 3.2339066356102535e-06, + "loss": 2.253, + "step": 20000 + }, + { + "epoch": 2.1151649746192893, + "grad_norm": 8.1875, + "learning_rate": 3.2337429232430595e-06, + "loss": 2.6127, + "step": 20001 + }, + { + "epoch": 2.115270727580372, + "grad_norm": 8.875, + "learning_rate": 3.233579207432711e-06, + "loss": 2.7121, + "step": 20002 + }, + { + "epoch": 2.115376480541455, + "grad_norm": 8.375, + "learning_rate": 3.233415488179975e-06, + "loss": 2.6334, + "step": 20003 + }, + { + "epoch": 2.1154822335025383, + "grad_norm": 8.375, + "learning_rate": 3.23325176548562e-06, + "loss": 2.5467, + "step": 20004 + }, + { + "epoch": 2.115587986463621, + "grad_norm": 8.375, + "learning_rate": 3.2330880393504145e-06, + "loss": 2.532, + "step": 20005 + }, + { + "epoch": 2.115693739424704, + "grad_norm": 8.8125, + "learning_rate": 3.2329243097751266e-06, + "loss": 2.6607, + "step": 20006 + }, + { + "epoch": 2.115799492385787, + "grad_norm": 9.375, + "learning_rate": 3.232760576760524e-06, + "loss": 2.7481, + "step": 20007 + }, + { + "epoch": 2.1159052453468696, + "grad_norm": 8.375, + "learning_rate": 3.2325968403073766e-06, + "loss": 2.6383, + "step": 20008 + }, + { + "epoch": 2.1160109983079525, + "grad_norm": 8.1875, + "learning_rate": 3.2324331004164518e-06, + "loss": 2.5712, + "step": 20009 + }, + { + "epoch": 2.1161167512690353, + "grad_norm": 8.375, + "learning_rate": 3.2322693570885176e-06, + "loss": 2.4439, + "step": 20010 + }, + { + "epoch": 2.1162225042301186, + "grad_norm": 9.5625, + "learning_rate": 3.232105610324343e-06, + "loss": 2.2604, + "step": 20011 + }, + { + "epoch": 2.1163282571912014, + "grad_norm": 8.375, + "learning_rate": 3.231941860124696e-06, + "loss": 2.6243, + "step": 20012 + }, + { + "epoch": 2.1164340101522843, + "grad_norm": 8.0625, + "learning_rate": 3.2317781064903454e-06, + "loss": 2.6808, + "step": 20013 + }, + { + "epoch": 2.116539763113367, + "grad_norm": 8.5625, + "learning_rate": 3.23161434942206e-06, + "loss": 2.5677, + "step": 20014 + }, + { + "epoch": 2.11664551607445, + "grad_norm": 8.875, + "learning_rate": 3.2314505889206065e-06, + "loss": 2.4493, + "step": 20015 + }, + { + "epoch": 2.116751269035533, + "grad_norm": 8.75, + "learning_rate": 3.231286824986755e-06, + "loss": 2.6072, + "step": 20016 + }, + { + "epoch": 2.116857021996616, + "grad_norm": 8.75, + "learning_rate": 3.2311230576212736e-06, + "loss": 2.712, + "step": 20017 + }, + { + "epoch": 2.116962774957699, + "grad_norm": 8.25, + "learning_rate": 3.2309592868249304e-06, + "loss": 2.4102, + "step": 20018 + }, + { + "epoch": 2.1170685279187818, + "grad_norm": 8.5625, + "learning_rate": 3.230795512598495e-06, + "loss": 2.5488, + "step": 20019 + }, + { + "epoch": 2.1171742808798646, + "grad_norm": 8.5625, + "learning_rate": 3.2306317349427347e-06, + "loss": 2.4546, + "step": 20020 + }, + { + "epoch": 2.1172800338409474, + "grad_norm": 8.3125, + "learning_rate": 3.2304679538584183e-06, + "loss": 2.5884, + "step": 20021 + }, + { + "epoch": 2.1173857868020303, + "grad_norm": 8.5, + "learning_rate": 3.2303041693463153e-06, + "loss": 2.5045, + "step": 20022 + }, + { + "epoch": 2.1174915397631136, + "grad_norm": 8.125, + "learning_rate": 3.2301403814071925e-06, + "loss": 2.5779, + "step": 20023 + }, + { + "epoch": 2.1175972927241964, + "grad_norm": 8.3125, + "learning_rate": 3.22997659004182e-06, + "loss": 2.2154, + "step": 20024 + }, + { + "epoch": 2.1177030456852792, + "grad_norm": 8.3125, + "learning_rate": 3.229812795250966e-06, + "loss": 2.2321, + "step": 20025 + }, + { + "epoch": 2.117808798646362, + "grad_norm": 9.0, + "learning_rate": 3.229648997035399e-06, + "loss": 2.2826, + "step": 20026 + }, + { + "epoch": 2.117914551607445, + "grad_norm": 7.75, + "learning_rate": 3.229485195395887e-06, + "loss": 2.3448, + "step": 20027 + }, + { + "epoch": 2.1180203045685277, + "grad_norm": 8.25, + "learning_rate": 3.2293213903332003e-06, + "loss": 2.5617, + "step": 20028 + }, + { + "epoch": 2.118126057529611, + "grad_norm": 8.4375, + "learning_rate": 3.2291575818481062e-06, + "loss": 2.5208, + "step": 20029 + }, + { + "epoch": 2.118231810490694, + "grad_norm": 8.125, + "learning_rate": 3.228993769941374e-06, + "loss": 2.5198, + "step": 20030 + }, + { + "epoch": 2.1183375634517767, + "grad_norm": 8.8125, + "learning_rate": 3.228829954613772e-06, + "loss": 2.4664, + "step": 20031 + }, + { + "epoch": 2.1184433164128595, + "grad_norm": 8.0625, + "learning_rate": 3.2286661358660685e-06, + "loss": 2.2974, + "step": 20032 + }, + { + "epoch": 2.1185490693739424, + "grad_norm": 8.375, + "learning_rate": 3.228502313699034e-06, + "loss": 2.6535, + "step": 20033 + }, + { + "epoch": 2.1186548223350252, + "grad_norm": 8.3125, + "learning_rate": 3.228338488113435e-06, + "loss": 2.4784, + "step": 20034 + }, + { + "epoch": 2.1187605752961085, + "grad_norm": 8.4375, + "learning_rate": 3.2281746591100422e-06, + "loss": 2.7395, + "step": 20035 + }, + { + "epoch": 2.1188663282571913, + "grad_norm": 8.375, + "learning_rate": 3.228010826689623e-06, + "loss": 2.3457, + "step": 20036 + }, + { + "epoch": 2.118972081218274, + "grad_norm": 8.75, + "learning_rate": 3.2278469908529467e-06, + "loss": 2.5758, + "step": 20037 + }, + { + "epoch": 2.119077834179357, + "grad_norm": 8.1875, + "learning_rate": 3.2276831516007827e-06, + "loss": 2.5514, + "step": 20038 + }, + { + "epoch": 2.11918358714044, + "grad_norm": 8.0, + "learning_rate": 3.227519308933899e-06, + "loss": 2.3592, + "step": 20039 + }, + { + "epoch": 2.1192893401015227, + "grad_norm": 7.9375, + "learning_rate": 3.2273554628530645e-06, + "loss": 2.5126, + "step": 20040 + }, + { + "epoch": 2.1193950930626055, + "grad_norm": 8.25, + "learning_rate": 3.2271916133590494e-06, + "loss": 2.6806, + "step": 20041 + }, + { + "epoch": 2.119500846023689, + "grad_norm": 8.625, + "learning_rate": 3.22702776045262e-06, + "loss": 2.5817, + "step": 20042 + }, + { + "epoch": 2.1196065989847717, + "grad_norm": 8.9375, + "learning_rate": 3.226863904134547e-06, + "loss": 2.3807, + "step": 20043 + }, + { + "epoch": 2.1197123519458545, + "grad_norm": 8.25, + "learning_rate": 3.2267000444055997e-06, + "loss": 2.5812, + "step": 20044 + }, + { + "epoch": 2.1198181049069373, + "grad_norm": 8.8125, + "learning_rate": 3.226536181266546e-06, + "loss": 2.4294, + "step": 20045 + }, + { + "epoch": 2.11992385786802, + "grad_norm": 8.75, + "learning_rate": 3.2263723147181542e-06, + "loss": 2.359, + "step": 20046 + }, + { + "epoch": 2.120029610829103, + "grad_norm": 8.1875, + "learning_rate": 3.2262084447611953e-06, + "loss": 2.6904, + "step": 20047 + }, + { + "epoch": 2.1201353637901863, + "grad_norm": 8.625, + "learning_rate": 3.2260445713964374e-06, + "loss": 2.5084, + "step": 20048 + }, + { + "epoch": 2.120241116751269, + "grad_norm": 8.5, + "learning_rate": 3.225880694624648e-06, + "loss": 2.603, + "step": 20049 + }, + { + "epoch": 2.120346869712352, + "grad_norm": 8.9375, + "learning_rate": 3.2257168144465982e-06, + "loss": 2.6195, + "step": 20050 + }, + { + "epoch": 2.120452622673435, + "grad_norm": 8.5, + "learning_rate": 3.2255529308630556e-06, + "loss": 2.4422, + "step": 20051 + }, + { + "epoch": 2.1205583756345177, + "grad_norm": 8.3125, + "learning_rate": 3.2253890438747905e-06, + "loss": 2.6132, + "step": 20052 + }, + { + "epoch": 2.1206641285956005, + "grad_norm": 8.0, + "learning_rate": 3.2252251534825713e-06, + "loss": 2.4323, + "step": 20053 + }, + { + "epoch": 2.120769881556684, + "grad_norm": 8.125, + "learning_rate": 3.2250612596871665e-06, + "loss": 2.5568, + "step": 20054 + }, + { + "epoch": 2.1208756345177666, + "grad_norm": 8.4375, + "learning_rate": 3.2248973624893464e-06, + "loss": 2.5462, + "step": 20055 + }, + { + "epoch": 2.1209813874788495, + "grad_norm": 8.3125, + "learning_rate": 3.2247334618898792e-06, + "loss": 2.6118, + "step": 20056 + }, + { + "epoch": 2.1210871404399323, + "grad_norm": 8.5625, + "learning_rate": 3.224569557889534e-06, + "loss": 2.4921, + "step": 20057 + }, + { + "epoch": 2.121192893401015, + "grad_norm": 9.0, + "learning_rate": 3.224405650489081e-06, + "loss": 2.6903, + "step": 20058 + }, + { + "epoch": 2.121298646362098, + "grad_norm": 8.4375, + "learning_rate": 3.2242417396892878e-06, + "loss": 2.6053, + "step": 20059 + }, + { + "epoch": 2.1214043993231813, + "grad_norm": 9.25, + "learning_rate": 3.2240778254909243e-06, + "loss": 2.6935, + "step": 20060 + }, + { + "epoch": 2.121510152284264, + "grad_norm": 8.5, + "learning_rate": 3.22391390789476e-06, + "loss": 2.4852, + "step": 20061 + }, + { + "epoch": 2.121615905245347, + "grad_norm": 9.1875, + "learning_rate": 3.2237499869015632e-06, + "loss": 2.81, + "step": 20062 + }, + { + "epoch": 2.1217216582064298, + "grad_norm": 8.4375, + "learning_rate": 3.2235860625121047e-06, + "loss": 2.7272, + "step": 20063 + }, + { + "epoch": 2.1218274111675126, + "grad_norm": 8.9375, + "learning_rate": 3.2234221347271523e-06, + "loss": 2.5035, + "step": 20064 + }, + { + "epoch": 2.1219331641285955, + "grad_norm": 7.96875, + "learning_rate": 3.2232582035474757e-06, + "loss": 2.4183, + "step": 20065 + }, + { + "epoch": 2.1220389170896787, + "grad_norm": 7.9375, + "learning_rate": 3.2230942689738444e-06, + "loss": 2.3344, + "step": 20066 + }, + { + "epoch": 2.1221446700507616, + "grad_norm": 8.4375, + "learning_rate": 3.222930331007027e-06, + "loss": 2.5037, + "step": 20067 + }, + { + "epoch": 2.1222504230118444, + "grad_norm": 9.5625, + "learning_rate": 3.2227663896477933e-06, + "loss": 2.6538, + "step": 20068 + }, + { + "epoch": 2.1223561759729273, + "grad_norm": 7.8125, + "learning_rate": 3.2226024448969116e-06, + "loss": 2.43, + "step": 20069 + }, + { + "epoch": 2.12246192893401, + "grad_norm": 7.96875, + "learning_rate": 3.2224384967551537e-06, + "loss": 2.5322, + "step": 20070 + }, + { + "epoch": 2.122567681895093, + "grad_norm": 9.25, + "learning_rate": 3.2222745452232867e-06, + "loss": 2.7063, + "step": 20071 + }, + { + "epoch": 2.1226734348561758, + "grad_norm": 8.1875, + "learning_rate": 3.222110590302081e-06, + "loss": 2.6914, + "step": 20072 + }, + { + "epoch": 2.122779187817259, + "grad_norm": 8.125, + "learning_rate": 3.2219466319923057e-06, + "loss": 2.5347, + "step": 20073 + }, + { + "epoch": 2.122884940778342, + "grad_norm": 8.25, + "learning_rate": 3.2217826702947296e-06, + "loss": 2.4152, + "step": 20074 + }, + { + "epoch": 2.1229906937394247, + "grad_norm": 8.5625, + "learning_rate": 3.2216187052101232e-06, + "loss": 2.4103, + "step": 20075 + }, + { + "epoch": 2.1230964467005076, + "grad_norm": 9.1875, + "learning_rate": 3.2214547367392553e-06, + "loss": 2.5146, + "step": 20076 + }, + { + "epoch": 2.1232021996615904, + "grad_norm": 8.625, + "learning_rate": 3.221290764882895e-06, + "loss": 2.6878, + "step": 20077 + }, + { + "epoch": 2.1233079526226732, + "grad_norm": 8.5625, + "learning_rate": 3.2211267896418117e-06, + "loss": 2.5292, + "step": 20078 + }, + { + "epoch": 2.1234137055837565, + "grad_norm": 8.125, + "learning_rate": 3.2209628110167764e-06, + "loss": 2.4559, + "step": 20079 + }, + { + "epoch": 2.1235194585448394, + "grad_norm": 8.5625, + "learning_rate": 3.2207988290085568e-06, + "loss": 2.4268, + "step": 20080 + }, + { + "epoch": 2.123625211505922, + "grad_norm": 8.4375, + "learning_rate": 3.220634843617923e-06, + "loss": 2.5497, + "step": 20081 + }, + { + "epoch": 2.123730964467005, + "grad_norm": 8.25, + "learning_rate": 3.2204708548456454e-06, + "loss": 2.4772, + "step": 20082 + }, + { + "epoch": 2.123836717428088, + "grad_norm": 8.25, + "learning_rate": 3.2203068626924922e-06, + "loss": 2.5515, + "step": 20083 + }, + { + "epoch": 2.1239424703891707, + "grad_norm": 8.375, + "learning_rate": 3.2201428671592334e-06, + "loss": 2.3485, + "step": 20084 + }, + { + "epoch": 2.124048223350254, + "grad_norm": 8.75, + "learning_rate": 3.2199788682466387e-06, + "loss": 2.5129, + "step": 20085 + }, + { + "epoch": 2.124153976311337, + "grad_norm": 10.1875, + "learning_rate": 3.2198148659554774e-06, + "loss": 2.3216, + "step": 20086 + }, + { + "epoch": 2.1242597292724197, + "grad_norm": 8.4375, + "learning_rate": 3.2196508602865194e-06, + "loss": 2.5971, + "step": 20087 + }, + { + "epoch": 2.1243654822335025, + "grad_norm": 8.5, + "learning_rate": 3.2194868512405346e-06, + "loss": 2.5996, + "step": 20088 + }, + { + "epoch": 2.1244712351945854, + "grad_norm": 8.125, + "learning_rate": 3.219322838818292e-06, + "loss": 2.665, + "step": 20089 + }, + { + "epoch": 2.124576988155668, + "grad_norm": 8.8125, + "learning_rate": 3.2191588230205616e-06, + "loss": 2.8322, + "step": 20090 + }, + { + "epoch": 2.1246827411167515, + "grad_norm": 8.5, + "learning_rate": 3.2189948038481132e-06, + "loss": 2.591, + "step": 20091 + }, + { + "epoch": 2.1247884940778343, + "grad_norm": 7.9375, + "learning_rate": 3.2188307813017167e-06, + "loss": 2.479, + "step": 20092 + }, + { + "epoch": 2.124894247038917, + "grad_norm": 8.75, + "learning_rate": 3.2186667553821406e-06, + "loss": 2.4173, + "step": 20093 + }, + { + "epoch": 2.125, + "grad_norm": 9.3125, + "learning_rate": 3.2185027260901553e-06, + "loss": 2.7844, + "step": 20094 + }, + { + "epoch": 2.125105752961083, + "grad_norm": 8.5625, + "learning_rate": 3.218338693426531e-06, + "loss": 2.4069, + "step": 20095 + }, + { + "epoch": 2.1252115059221657, + "grad_norm": 8.625, + "learning_rate": 3.2181746573920363e-06, + "loss": 2.64, + "step": 20096 + }, + { + "epoch": 2.125317258883249, + "grad_norm": 9.125, + "learning_rate": 3.2180106179874416e-06, + "loss": 2.5303, + "step": 20097 + }, + { + "epoch": 2.125423011844332, + "grad_norm": 8.1875, + "learning_rate": 3.2178465752135175e-06, + "loss": 2.5992, + "step": 20098 + }, + { + "epoch": 2.1255287648054146, + "grad_norm": 8.0, + "learning_rate": 3.2176825290710328e-06, + "loss": 2.5257, + "step": 20099 + }, + { + "epoch": 2.1256345177664975, + "grad_norm": 8.25, + "learning_rate": 3.2175184795607573e-06, + "loss": 2.3775, + "step": 20100 + }, + { + "epoch": 2.1257402707275803, + "grad_norm": 9.0, + "learning_rate": 3.217354426683461e-06, + "loss": 2.6924, + "step": 20101 + }, + { + "epoch": 2.125846023688663, + "grad_norm": 7.84375, + "learning_rate": 3.2171903704399145e-06, + "loss": 2.6862, + "step": 20102 + }, + { + "epoch": 2.125951776649746, + "grad_norm": 8.6875, + "learning_rate": 3.2170263108308858e-06, + "loss": 2.5953, + "step": 20103 + }, + { + "epoch": 2.1260575296108293, + "grad_norm": 8.125, + "learning_rate": 3.2168622478571465e-06, + "loss": 2.6937, + "step": 20104 + }, + { + "epoch": 2.126163282571912, + "grad_norm": 8.125, + "learning_rate": 3.216698181519466e-06, + "loss": 2.561, + "step": 20105 + }, + { + "epoch": 2.126269035532995, + "grad_norm": 8.0, + "learning_rate": 3.2165341118186135e-06, + "loss": 2.3617, + "step": 20106 + }, + { + "epoch": 2.126374788494078, + "grad_norm": 8.875, + "learning_rate": 3.2163700387553596e-06, + "loss": 2.8817, + "step": 20107 + }, + { + "epoch": 2.1264805414551606, + "grad_norm": 8.5625, + "learning_rate": 3.2162059623304746e-06, + "loss": 2.6047, + "step": 20108 + }, + { + "epoch": 2.1265862944162435, + "grad_norm": 8.3125, + "learning_rate": 3.216041882544727e-06, + "loss": 2.1476, + "step": 20109 + }, + { + "epoch": 2.1266920473773268, + "grad_norm": 8.9375, + "learning_rate": 3.2158777993988886e-06, + "loss": 2.5764, + "step": 20110 + }, + { + "epoch": 2.1267978003384096, + "grad_norm": 8.25, + "learning_rate": 3.2157137128937288e-06, + "loss": 2.5983, + "step": 20111 + }, + { + "epoch": 2.1269035532994924, + "grad_norm": 8.125, + "learning_rate": 3.215549623030016e-06, + "loss": 2.4894, + "step": 20112 + }, + { + "epoch": 2.1270093062605753, + "grad_norm": 8.0625, + "learning_rate": 3.215385529808522e-06, + "loss": 2.6199, + "step": 20113 + }, + { + "epoch": 2.127115059221658, + "grad_norm": 8.625, + "learning_rate": 3.2152214332300163e-06, + "loss": 2.7543, + "step": 20114 + }, + { + "epoch": 2.127220812182741, + "grad_norm": 8.5625, + "learning_rate": 3.2150573332952694e-06, + "loss": 2.4323, + "step": 20115 + }, + { + "epoch": 2.1273265651438242, + "grad_norm": 8.0625, + "learning_rate": 3.21489323000505e-06, + "loss": 2.4518, + "step": 20116 + }, + { + "epoch": 2.127432318104907, + "grad_norm": 8.375, + "learning_rate": 3.21472912336013e-06, + "loss": 2.2622, + "step": 20117 + }, + { + "epoch": 2.12753807106599, + "grad_norm": 8.5, + "learning_rate": 3.2145650133612783e-06, + "loss": 2.5573, + "step": 20118 + }, + { + "epoch": 2.1276438240270727, + "grad_norm": 8.0, + "learning_rate": 3.214400900009265e-06, + "loss": 2.5647, + "step": 20119 + }, + { + "epoch": 2.1277495769881556, + "grad_norm": 8.25, + "learning_rate": 3.2142367833048605e-06, + "loss": 2.6259, + "step": 20120 + }, + { + "epoch": 2.1278553299492384, + "grad_norm": 8.125, + "learning_rate": 3.214072663248835e-06, + "loss": 2.0363, + "step": 20121 + }, + { + "epoch": 2.1279610829103213, + "grad_norm": 8.0, + "learning_rate": 3.213908539841958e-06, + "loss": 2.3965, + "step": 20122 + }, + { + "epoch": 2.1280668358714045, + "grad_norm": 8.4375, + "learning_rate": 3.2137444130850005e-06, + "loss": 2.2552, + "step": 20123 + }, + { + "epoch": 2.1281725888324874, + "grad_norm": 8.0625, + "learning_rate": 3.213580282978733e-06, + "loss": 2.6043, + "step": 20124 + }, + { + "epoch": 2.1282783417935702, + "grad_norm": 8.1875, + "learning_rate": 3.2134161495239242e-06, + "loss": 2.4782, + "step": 20125 + }, + { + "epoch": 2.128384094754653, + "grad_norm": 8.1875, + "learning_rate": 3.2132520127213457e-06, + "loss": 2.524, + "step": 20126 + }, + { + "epoch": 2.128489847715736, + "grad_norm": 8.1875, + "learning_rate": 3.2130878725717674e-06, + "loss": 2.43, + "step": 20127 + }, + { + "epoch": 2.1285956006768187, + "grad_norm": 8.5625, + "learning_rate": 3.2129237290759587e-06, + "loss": 2.6023, + "step": 20128 + }, + { + "epoch": 2.128701353637902, + "grad_norm": 8.25, + "learning_rate": 3.212759582234691e-06, + "loss": 2.6261, + "step": 20129 + }, + { + "epoch": 2.128807106598985, + "grad_norm": 8.1875, + "learning_rate": 3.212595432048734e-06, + "loss": 2.4088, + "step": 20130 + }, + { + "epoch": 2.1289128595600677, + "grad_norm": 9.125, + "learning_rate": 3.2124312785188578e-06, + "loss": 2.4932, + "step": 20131 + }, + { + "epoch": 2.1290186125211505, + "grad_norm": 8.4375, + "learning_rate": 3.2122671216458328e-06, + "loss": 2.5978, + "step": 20132 + }, + { + "epoch": 2.1291243654822334, + "grad_norm": 8.8125, + "learning_rate": 3.21210296143043e-06, + "loss": 2.5518, + "step": 20133 + }, + { + "epoch": 2.129230118443316, + "grad_norm": 7.96875, + "learning_rate": 3.2119387978734194e-06, + "loss": 2.5769, + "step": 20134 + }, + { + "epoch": 2.1293358714043995, + "grad_norm": 8.25, + "learning_rate": 3.2117746309755706e-06, + "loss": 2.4035, + "step": 20135 + }, + { + "epoch": 2.1294416243654823, + "grad_norm": 8.875, + "learning_rate": 3.2116104607376553e-06, + "loss": 2.6367, + "step": 20136 + }, + { + "epoch": 2.129547377326565, + "grad_norm": 8.0625, + "learning_rate": 3.211446287160443e-06, + "loss": 2.6313, + "step": 20137 + }, + { + "epoch": 2.129653130287648, + "grad_norm": 8.0, + "learning_rate": 3.2112821102447033e-06, + "loss": 2.6432, + "step": 20138 + }, + { + "epoch": 2.129758883248731, + "grad_norm": 8.25, + "learning_rate": 3.2111179299912087e-06, + "loss": 2.6805, + "step": 20139 + }, + { + "epoch": 2.1298646362098137, + "grad_norm": 8.625, + "learning_rate": 3.210953746400728e-06, + "loss": 2.6363, + "step": 20140 + }, + { + "epoch": 2.129970389170897, + "grad_norm": 8.4375, + "learning_rate": 3.210789559474032e-06, + "loss": 2.5651, + "step": 20141 + }, + { + "epoch": 2.13007614213198, + "grad_norm": 8.5, + "learning_rate": 3.2106253692118915e-06, + "loss": 2.4359, + "step": 20142 + }, + { + "epoch": 2.1301818950930627, + "grad_norm": 8.9375, + "learning_rate": 3.2104611756150767e-06, + "loss": 2.8649, + "step": 20143 + }, + { + "epoch": 2.1302876480541455, + "grad_norm": 8.6875, + "learning_rate": 3.210296978684358e-06, + "loss": 2.5309, + "step": 20144 + }, + { + "epoch": 2.1303934010152283, + "grad_norm": 8.25, + "learning_rate": 3.2101327784205066e-06, + "loss": 2.5112, + "step": 20145 + }, + { + "epoch": 2.130499153976311, + "grad_norm": 8.125, + "learning_rate": 3.2099685748242926e-06, + "loss": 2.4283, + "step": 20146 + }, + { + "epoch": 2.1306049069373945, + "grad_norm": 8.3125, + "learning_rate": 3.2098043678964857e-06, + "loss": 2.4415, + "step": 20147 + }, + { + "epoch": 2.1307106598984773, + "grad_norm": 8.4375, + "learning_rate": 3.2096401576378578e-06, + "loss": 2.6391, + "step": 20148 + }, + { + "epoch": 2.13081641285956, + "grad_norm": 8.4375, + "learning_rate": 3.2094759440491786e-06, + "loss": 2.7009, + "step": 20149 + }, + { + "epoch": 2.130922165820643, + "grad_norm": 8.3125, + "learning_rate": 3.209311727131219e-06, + "loss": 2.7046, + "step": 20150 + }, + { + "epoch": 2.131027918781726, + "grad_norm": 8.5625, + "learning_rate": 3.209147506884749e-06, + "loss": 2.6923, + "step": 20151 + }, + { + "epoch": 2.1311336717428087, + "grad_norm": 8.75, + "learning_rate": 3.2089832833105406e-06, + "loss": 2.5952, + "step": 20152 + }, + { + "epoch": 2.1312394247038915, + "grad_norm": 8.5625, + "learning_rate": 3.2088190564093636e-06, + "loss": 2.347, + "step": 20153 + }, + { + "epoch": 2.1313451776649748, + "grad_norm": 8.125, + "learning_rate": 3.208654826181988e-06, + "loss": 2.5741, + "step": 20154 + }, + { + "epoch": 2.1314509306260576, + "grad_norm": 8.625, + "learning_rate": 3.208490592629186e-06, + "loss": 2.5746, + "step": 20155 + }, + { + "epoch": 2.1315566835871405, + "grad_norm": 8.6875, + "learning_rate": 3.2083263557517276e-06, + "loss": 2.6548, + "step": 20156 + }, + { + "epoch": 2.1316624365482233, + "grad_norm": 8.6875, + "learning_rate": 3.2081621155503816e-06, + "loss": 2.6133, + "step": 20157 + }, + { + "epoch": 2.131768189509306, + "grad_norm": 8.9375, + "learning_rate": 3.2079978720259213e-06, + "loss": 2.493, + "step": 20158 + }, + { + "epoch": 2.131873942470389, + "grad_norm": 8.4375, + "learning_rate": 3.207833625179117e-06, + "loss": 2.5384, + "step": 20159 + }, + { + "epoch": 2.1319796954314723, + "grad_norm": 9.3125, + "learning_rate": 3.2076693750107385e-06, + "loss": 2.4881, + "step": 20160 + }, + { + "epoch": 2.132085448392555, + "grad_norm": 9.125, + "learning_rate": 3.2075051215215576e-06, + "loss": 2.5246, + "step": 20161 + }, + { + "epoch": 2.132191201353638, + "grad_norm": 8.375, + "learning_rate": 3.2073408647123443e-06, + "loss": 2.3705, + "step": 20162 + }, + { + "epoch": 2.1322969543147208, + "grad_norm": 8.3125, + "learning_rate": 3.2071766045838694e-06, + "loss": 2.278, + "step": 20163 + }, + { + "epoch": 2.1324027072758036, + "grad_norm": 8.5625, + "learning_rate": 3.2070123411369044e-06, + "loss": 2.501, + "step": 20164 + }, + { + "epoch": 2.1325084602368864, + "grad_norm": 8.375, + "learning_rate": 3.20684807437222e-06, + "loss": 2.2455, + "step": 20165 + }, + { + "epoch": 2.1326142131979697, + "grad_norm": 8.4375, + "learning_rate": 3.206683804290585e-06, + "loss": 2.5542, + "step": 20166 + }, + { + "epoch": 2.1327199661590526, + "grad_norm": 8.25, + "learning_rate": 3.2065195308927728e-06, + "loss": 2.5006, + "step": 20167 + }, + { + "epoch": 2.1328257191201354, + "grad_norm": 8.625, + "learning_rate": 3.2063552541795532e-06, + "loss": 2.758, + "step": 20168 + }, + { + "epoch": 2.1329314720812182, + "grad_norm": 8.5625, + "learning_rate": 3.2061909741516985e-06, + "loss": 2.5928, + "step": 20169 + }, + { + "epoch": 2.133037225042301, + "grad_norm": 8.5625, + "learning_rate": 3.206026690809977e-06, + "loss": 2.6878, + "step": 20170 + }, + { + "epoch": 2.133142978003384, + "grad_norm": 9.25, + "learning_rate": 3.2058624041551613e-06, + "loss": 2.5397, + "step": 20171 + }, + { + "epoch": 2.133248730964467, + "grad_norm": 8.375, + "learning_rate": 3.205698114188023e-06, + "loss": 2.5606, + "step": 20172 + }, + { + "epoch": 2.13335448392555, + "grad_norm": 8.4375, + "learning_rate": 3.2055338209093304e-06, + "loss": 2.5562, + "step": 20173 + }, + { + "epoch": 2.133460236886633, + "grad_norm": 8.875, + "learning_rate": 3.2053695243198573e-06, + "loss": 2.6348, + "step": 20174 + }, + { + "epoch": 2.1335659898477157, + "grad_norm": 8.375, + "learning_rate": 3.205205224420373e-06, + "loss": 2.6142, + "step": 20175 + }, + { + "epoch": 2.1336717428087986, + "grad_norm": 7.9375, + "learning_rate": 3.2050409212116485e-06, + "loss": 2.4504, + "step": 20176 + }, + { + "epoch": 2.1337774957698814, + "grad_norm": 8.5625, + "learning_rate": 3.2048766146944566e-06, + "loss": 2.5905, + "step": 20177 + }, + { + "epoch": 2.1338832487309647, + "grad_norm": 8.625, + "learning_rate": 3.2047123048695667e-06, + "loss": 2.5951, + "step": 20178 + }, + { + "epoch": 2.1339890016920475, + "grad_norm": 8.625, + "learning_rate": 3.2045479917377493e-06, + "loss": 2.5414, + "step": 20179 + }, + { + "epoch": 2.1340947546531304, + "grad_norm": 8.75, + "learning_rate": 3.204383675299777e-06, + "loss": 2.6078, + "step": 20180 + }, + { + "epoch": 2.134200507614213, + "grad_norm": 8.75, + "learning_rate": 3.2042193555564204e-06, + "loss": 2.4654, + "step": 20181 + }, + { + "epoch": 2.134306260575296, + "grad_norm": 8.5625, + "learning_rate": 3.20405503250845e-06, + "loss": 2.4841, + "step": 20182 + }, + { + "epoch": 2.134412013536379, + "grad_norm": 7.84375, + "learning_rate": 3.2038907061566375e-06, + "loss": 2.5483, + "step": 20183 + }, + { + "epoch": 2.1345177664974617, + "grad_norm": 8.6875, + "learning_rate": 3.2037263765017534e-06, + "loss": 2.5992, + "step": 20184 + }, + { + "epoch": 2.134623519458545, + "grad_norm": 8.5625, + "learning_rate": 3.2035620435445695e-06, + "loss": 2.4562, + "step": 20185 + }, + { + "epoch": 2.134729272419628, + "grad_norm": 8.75, + "learning_rate": 3.203397707285857e-06, + "loss": 2.6475, + "step": 20186 + }, + { + "epoch": 2.1348350253807107, + "grad_norm": 8.4375, + "learning_rate": 3.203233367726386e-06, + "loss": 2.6418, + "step": 20187 + }, + { + "epoch": 2.1349407783417935, + "grad_norm": 8.4375, + "learning_rate": 3.2030690248669295e-06, + "loss": 2.6606, + "step": 20188 + }, + { + "epoch": 2.1350465313028764, + "grad_norm": 8.4375, + "learning_rate": 3.2029046787082564e-06, + "loss": 2.4516, + "step": 20189 + }, + { + "epoch": 2.135152284263959, + "grad_norm": 8.3125, + "learning_rate": 3.20274032925114e-06, + "loss": 2.6551, + "step": 20190 + }, + { + "epoch": 2.1352580372250425, + "grad_norm": 7.78125, + "learning_rate": 3.2025759764963504e-06, + "loss": 2.5115, + "step": 20191 + }, + { + "epoch": 2.1353637901861253, + "grad_norm": 8.125, + "learning_rate": 3.2024116204446587e-06, + "loss": 2.5239, + "step": 20192 + }, + { + "epoch": 2.135469543147208, + "grad_norm": 8.3125, + "learning_rate": 3.202247261096837e-06, + "loss": 2.5512, + "step": 20193 + }, + { + "epoch": 2.135575296108291, + "grad_norm": 9.0625, + "learning_rate": 3.2020828984536554e-06, + "loss": 2.6266, + "step": 20194 + }, + { + "epoch": 2.135681049069374, + "grad_norm": 8.6875, + "learning_rate": 3.2019185325158864e-06, + "loss": 2.6873, + "step": 20195 + }, + { + "epoch": 2.1357868020304567, + "grad_norm": 8.875, + "learning_rate": 3.2017541632843014e-06, + "loss": 2.6363, + "step": 20196 + }, + { + "epoch": 2.13589255499154, + "grad_norm": 8.4375, + "learning_rate": 3.2015897907596704e-06, + "loss": 2.5226, + "step": 20197 + }, + { + "epoch": 2.135998307952623, + "grad_norm": 8.5, + "learning_rate": 3.201425414942766e-06, + "loss": 2.6164, + "step": 20198 + }, + { + "epoch": 2.1361040609137056, + "grad_norm": 8.5625, + "learning_rate": 3.2012610358343587e-06, + "loss": 2.3975, + "step": 20199 + }, + { + "epoch": 2.1362098138747885, + "grad_norm": 8.125, + "learning_rate": 3.20109665343522e-06, + "loss": 2.6432, + "step": 20200 + }, + { + "epoch": 2.1363155668358713, + "grad_norm": 8.125, + "learning_rate": 3.2009322677461212e-06, + "loss": 2.5201, + "step": 20201 + }, + { + "epoch": 2.136421319796954, + "grad_norm": 8.1875, + "learning_rate": 3.200767878767834e-06, + "loss": 2.5769, + "step": 20202 + }, + { + "epoch": 2.136527072758037, + "grad_norm": 8.375, + "learning_rate": 3.2006034865011306e-06, + "loss": 2.4649, + "step": 20203 + }, + { + "epoch": 2.1366328257191203, + "grad_norm": 7.8125, + "learning_rate": 3.200439090946781e-06, + "loss": 2.5268, + "step": 20204 + }, + { + "epoch": 2.136738578680203, + "grad_norm": 8.625, + "learning_rate": 3.200274692105557e-06, + "loss": 2.6084, + "step": 20205 + }, + { + "epoch": 2.136844331641286, + "grad_norm": 8.375, + "learning_rate": 3.2001102899782303e-06, + "loss": 2.6512, + "step": 20206 + }, + { + "epoch": 2.136950084602369, + "grad_norm": 8.75, + "learning_rate": 3.199945884565573e-06, + "loss": 2.3183, + "step": 20207 + }, + { + "epoch": 2.1370558375634516, + "grad_norm": 8.25, + "learning_rate": 3.1997814758683554e-06, + "loss": 2.5431, + "step": 20208 + }, + { + "epoch": 2.137161590524535, + "grad_norm": 8.375, + "learning_rate": 3.1996170638873496e-06, + "loss": 2.4273, + "step": 20209 + }, + { + "epoch": 2.1372673434856178, + "grad_norm": 8.4375, + "learning_rate": 3.1994526486233275e-06, + "loss": 2.4988, + "step": 20210 + }, + { + "epoch": 2.1373730964467006, + "grad_norm": 8.9375, + "learning_rate": 3.1992882300770587e-06, + "loss": 2.5057, + "step": 20211 + }, + { + "epoch": 2.1374788494077834, + "grad_norm": 8.75, + "learning_rate": 3.1991238082493178e-06, + "loss": 2.7463, + "step": 20212 + }, + { + "epoch": 2.1375846023688663, + "grad_norm": 8.0625, + "learning_rate": 3.1989593831408742e-06, + "loss": 2.2091, + "step": 20213 + }, + { + "epoch": 2.137690355329949, + "grad_norm": 8.3125, + "learning_rate": 3.1987949547524997e-06, + "loss": 2.5642, + "step": 20214 + }, + { + "epoch": 2.137796108291032, + "grad_norm": 8.625, + "learning_rate": 3.198630523084967e-06, + "loss": 2.2771, + "step": 20215 + }, + { + "epoch": 2.1379018612521152, + "grad_norm": 8.125, + "learning_rate": 3.1984660881390474e-06, + "loss": 2.6009, + "step": 20216 + }, + { + "epoch": 2.138007614213198, + "grad_norm": 8.125, + "learning_rate": 3.1983016499155108e-06, + "loss": 2.6063, + "step": 20217 + }, + { + "epoch": 2.138113367174281, + "grad_norm": 7.8125, + "learning_rate": 3.1981372084151315e-06, + "loss": 2.5256, + "step": 20218 + }, + { + "epoch": 2.1382191201353637, + "grad_norm": 8.6875, + "learning_rate": 3.197972763638678e-06, + "loss": 2.6908, + "step": 20219 + }, + { + "epoch": 2.1383248730964466, + "grad_norm": 8.9375, + "learning_rate": 3.1978083155869255e-06, + "loss": 2.4916, + "step": 20220 + }, + { + "epoch": 2.1384306260575294, + "grad_norm": 8.375, + "learning_rate": 3.1976438642606434e-06, + "loss": 2.8039, + "step": 20221 + }, + { + "epoch": 2.1385363790186127, + "grad_norm": 8.5, + "learning_rate": 3.1974794096606036e-06, + "loss": 2.6662, + "step": 20222 + }, + { + "epoch": 2.1386421319796955, + "grad_norm": 8.75, + "learning_rate": 3.197314951787579e-06, + "loss": 2.593, + "step": 20223 + }, + { + "epoch": 2.1387478849407784, + "grad_norm": 8.6875, + "learning_rate": 3.1971504906423396e-06, + "loss": 2.7022, + "step": 20224 + }, + { + "epoch": 2.1388536379018612, + "grad_norm": 8.3125, + "learning_rate": 3.1969860262256584e-06, + "loss": 2.3934, + "step": 20225 + }, + { + "epoch": 2.138959390862944, + "grad_norm": 8.1875, + "learning_rate": 3.1968215585383078e-06, + "loss": 2.6198, + "step": 20226 + }, + { + "epoch": 2.139065143824027, + "grad_norm": 8.1875, + "learning_rate": 3.196657087581057e-06, + "loss": 2.4711, + "step": 20227 + }, + { + "epoch": 2.13917089678511, + "grad_norm": 8.3125, + "learning_rate": 3.1964926133546802e-06, + "loss": 2.5502, + "step": 20228 + }, + { + "epoch": 2.139276649746193, + "grad_norm": 8.25, + "learning_rate": 3.1963281358599484e-06, + "loss": 2.5239, + "step": 20229 + }, + { + "epoch": 2.139382402707276, + "grad_norm": 8.25, + "learning_rate": 3.1961636550976332e-06, + "loss": 2.4828, + "step": 20230 + }, + { + "epoch": 2.1394881556683587, + "grad_norm": 8.375, + "learning_rate": 3.195999171068507e-06, + "loss": 2.6281, + "step": 20231 + }, + { + "epoch": 2.1395939086294415, + "grad_norm": 8.6875, + "learning_rate": 3.1958346837733416e-06, + "loss": 2.5337, + "step": 20232 + }, + { + "epoch": 2.1396996615905244, + "grad_norm": 8.25, + "learning_rate": 3.195670193212908e-06, + "loss": 2.4962, + "step": 20233 + }, + { + "epoch": 2.139805414551607, + "grad_norm": 8.3125, + "learning_rate": 3.195505699387979e-06, + "loss": 2.6202, + "step": 20234 + }, + { + "epoch": 2.1399111675126905, + "grad_norm": 8.0625, + "learning_rate": 3.195341202299327e-06, + "loss": 2.3178, + "step": 20235 + }, + { + "epoch": 2.1400169204737733, + "grad_norm": 8.1875, + "learning_rate": 3.1951767019477214e-06, + "loss": 2.5997, + "step": 20236 + }, + { + "epoch": 2.140122673434856, + "grad_norm": 8.25, + "learning_rate": 3.1950121983339367e-06, + "loss": 2.5417, + "step": 20237 + }, + { + "epoch": 2.140228426395939, + "grad_norm": 8.75, + "learning_rate": 3.1948476914587443e-06, + "loss": 2.4699, + "step": 20238 + }, + { + "epoch": 2.140334179357022, + "grad_norm": 9.1875, + "learning_rate": 3.1946831813229156e-06, + "loss": 2.7572, + "step": 20239 + }, + { + "epoch": 2.1404399323181047, + "grad_norm": 8.25, + "learning_rate": 3.194518667927223e-06, + "loss": 2.6847, + "step": 20240 + }, + { + "epoch": 2.140545685279188, + "grad_norm": 7.875, + "learning_rate": 3.194354151272438e-06, + "loss": 2.3537, + "step": 20241 + }, + { + "epoch": 2.140651438240271, + "grad_norm": 8.375, + "learning_rate": 3.1941896313593334e-06, + "loss": 2.2963, + "step": 20242 + }, + { + "epoch": 2.1407571912013537, + "grad_norm": 8.6875, + "learning_rate": 3.1940251081886803e-06, + "loss": 2.6283, + "step": 20243 + }, + { + "epoch": 2.1408629441624365, + "grad_norm": 7.96875, + "learning_rate": 3.193860581761252e-06, + "loss": 2.3459, + "step": 20244 + }, + { + "epoch": 2.1409686971235193, + "grad_norm": 8.1875, + "learning_rate": 3.193696052077819e-06, + "loss": 2.3328, + "step": 20245 + }, + { + "epoch": 2.141074450084602, + "grad_norm": 8.5, + "learning_rate": 3.1935315191391536e-06, + "loss": 2.5792, + "step": 20246 + }, + { + "epoch": 2.1411802030456855, + "grad_norm": 9.0, + "learning_rate": 3.1933669829460296e-06, + "loss": 2.5804, + "step": 20247 + }, + { + "epoch": 2.1412859560067683, + "grad_norm": 9.25, + "learning_rate": 3.1932024434992178e-06, + "loss": 2.8243, + "step": 20248 + }, + { + "epoch": 2.141391708967851, + "grad_norm": 8.4375, + "learning_rate": 3.19303790079949e-06, + "loss": 2.5735, + "step": 20249 + }, + { + "epoch": 2.141497461928934, + "grad_norm": 8.3125, + "learning_rate": 3.1928733548476187e-06, + "loss": 2.6561, + "step": 20250 + }, + { + "epoch": 2.141603214890017, + "grad_norm": 7.96875, + "learning_rate": 3.1927088056443765e-06, + "loss": 2.537, + "step": 20251 + }, + { + "epoch": 2.1417089678510997, + "grad_norm": 8.125, + "learning_rate": 3.192544253190535e-06, + "loss": 2.7142, + "step": 20252 + }, + { + "epoch": 2.141814720812183, + "grad_norm": 8.9375, + "learning_rate": 3.192379697486867e-06, + "loss": 2.4916, + "step": 20253 + }, + { + "epoch": 2.1419204737732658, + "grad_norm": 8.0, + "learning_rate": 3.192215138534144e-06, + "loss": 2.4583, + "step": 20254 + }, + { + "epoch": 2.1420262267343486, + "grad_norm": 9.25, + "learning_rate": 3.192050576333138e-06, + "loss": 2.9364, + "step": 20255 + }, + { + "epoch": 2.1421319796954315, + "grad_norm": 8.875, + "learning_rate": 3.191886010884622e-06, + "loss": 2.6595, + "step": 20256 + }, + { + "epoch": 2.1422377326565143, + "grad_norm": 8.1875, + "learning_rate": 3.1917214421893677e-06, + "loss": 2.2855, + "step": 20257 + }, + { + "epoch": 2.142343485617597, + "grad_norm": 8.625, + "learning_rate": 3.191556870248148e-06, + "loss": 2.6891, + "step": 20258 + }, + { + "epoch": 2.1424492385786804, + "grad_norm": 8.4375, + "learning_rate": 3.1913922950617345e-06, + "loss": 2.5541, + "step": 20259 + }, + { + "epoch": 2.1425549915397633, + "grad_norm": 8.625, + "learning_rate": 3.1912277166308996e-06, + "loss": 2.4757, + "step": 20260 + }, + { + "epoch": 2.142660744500846, + "grad_norm": 8.125, + "learning_rate": 3.1910631349564164e-06, + "loss": 2.5746, + "step": 20261 + }, + { + "epoch": 2.142766497461929, + "grad_norm": 8.25, + "learning_rate": 3.1908985500390553e-06, + "loss": 2.1318, + "step": 20262 + }, + { + "epoch": 2.1428722504230118, + "grad_norm": 8.5625, + "learning_rate": 3.1907339618795913e-06, + "loss": 2.5845, + "step": 20263 + }, + { + "epoch": 2.1429780033840946, + "grad_norm": 8.5625, + "learning_rate": 3.1905693704787943e-06, + "loss": 2.5196, + "step": 20264 + }, + { + "epoch": 2.1430837563451774, + "grad_norm": 8.8125, + "learning_rate": 3.190404775837438e-06, + "loss": 2.6517, + "step": 20265 + }, + { + "epoch": 2.1431895093062607, + "grad_norm": 8.375, + "learning_rate": 3.190240177956294e-06, + "loss": 2.539, + "step": 20266 + }, + { + "epoch": 2.1432952622673436, + "grad_norm": 7.9375, + "learning_rate": 3.1900755768361364e-06, + "loss": 2.4394, + "step": 20267 + }, + { + "epoch": 2.1434010152284264, + "grad_norm": 8.1875, + "learning_rate": 3.1899109724777348e-06, + "loss": 2.2608, + "step": 20268 + }, + { + "epoch": 2.1435067681895092, + "grad_norm": 8.625, + "learning_rate": 3.1897463648818644e-06, + "loss": 2.72, + "step": 20269 + }, + { + "epoch": 2.143612521150592, + "grad_norm": 8.125, + "learning_rate": 3.189581754049296e-06, + "loss": 2.4907, + "step": 20270 + }, + { + "epoch": 2.143718274111675, + "grad_norm": 8.4375, + "learning_rate": 3.1894171399808022e-06, + "loss": 2.4044, + "step": 20271 + }, + { + "epoch": 2.143824027072758, + "grad_norm": 8.9375, + "learning_rate": 3.1892525226771555e-06, + "loss": 2.7409, + "step": 20272 + }, + { + "epoch": 2.143929780033841, + "grad_norm": 8.125, + "learning_rate": 3.18908790213913e-06, + "loss": 2.5937, + "step": 20273 + }, + { + "epoch": 2.144035532994924, + "grad_norm": 8.8125, + "learning_rate": 3.1889232783674954e-06, + "loss": 2.6161, + "step": 20274 + }, + { + "epoch": 2.1441412859560067, + "grad_norm": 9.0, + "learning_rate": 3.188758651363026e-06, + "loss": 2.572, + "step": 20275 + }, + { + "epoch": 2.1442470389170896, + "grad_norm": 8.3125, + "learning_rate": 3.1885940211264943e-06, + "loss": 2.4102, + "step": 20276 + }, + { + "epoch": 2.1443527918781724, + "grad_norm": 8.3125, + "learning_rate": 3.188429387658672e-06, + "loss": 2.4122, + "step": 20277 + }, + { + "epoch": 2.1444585448392557, + "grad_norm": 8.375, + "learning_rate": 3.188264750960332e-06, + "loss": 2.3832, + "step": 20278 + }, + { + "epoch": 2.1445642978003385, + "grad_norm": 9.0, + "learning_rate": 3.1881001110322475e-06, + "loss": 2.6065, + "step": 20279 + }, + { + "epoch": 2.1446700507614214, + "grad_norm": 8.375, + "learning_rate": 3.1879354678751905e-06, + "loss": 2.3741, + "step": 20280 + }, + { + "epoch": 2.144775803722504, + "grad_norm": 8.125, + "learning_rate": 3.187770821489933e-06, + "loss": 2.5456, + "step": 20281 + }, + { + "epoch": 2.144881556683587, + "grad_norm": 8.125, + "learning_rate": 3.187606171877249e-06, + "loss": 2.5336, + "step": 20282 + }, + { + "epoch": 2.14498730964467, + "grad_norm": 8.0625, + "learning_rate": 3.187441519037911e-06, + "loss": 2.3522, + "step": 20283 + }, + { + "epoch": 2.145093062605753, + "grad_norm": 8.3125, + "learning_rate": 3.1872768629726904e-06, + "loss": 2.4652, + "step": 20284 + }, + { + "epoch": 2.145198815566836, + "grad_norm": 8.0625, + "learning_rate": 3.1871122036823603e-06, + "loss": 2.4979, + "step": 20285 + }, + { + "epoch": 2.145304568527919, + "grad_norm": 8.1875, + "learning_rate": 3.1869475411676942e-06, + "loss": 2.4512, + "step": 20286 + }, + { + "epoch": 2.1454103214890017, + "grad_norm": 8.5625, + "learning_rate": 3.186782875429464e-06, + "loss": 2.872, + "step": 20287 + }, + { + "epoch": 2.1455160744500845, + "grad_norm": 8.1875, + "learning_rate": 3.186618206468443e-06, + "loss": 2.5787, + "step": 20288 + }, + { + "epoch": 2.1456218274111674, + "grad_norm": 7.78125, + "learning_rate": 3.1864535342854036e-06, + "loss": 2.4738, + "step": 20289 + }, + { + "epoch": 2.1457275803722506, + "grad_norm": 8.125, + "learning_rate": 3.186288858881118e-06, + "loss": 2.423, + "step": 20290 + }, + { + "epoch": 2.1458333333333335, + "grad_norm": 8.5, + "learning_rate": 3.186124180256359e-06, + "loss": 2.5846, + "step": 20291 + }, + { + "epoch": 2.1459390862944163, + "grad_norm": 8.75, + "learning_rate": 3.1859594984119006e-06, + "loss": 2.4242, + "step": 20292 + }, + { + "epoch": 2.146044839255499, + "grad_norm": 8.9375, + "learning_rate": 3.185794813348515e-06, + "loss": 2.7899, + "step": 20293 + }, + { + "epoch": 2.146150592216582, + "grad_norm": 8.4375, + "learning_rate": 3.185630125066975e-06, + "loss": 2.6247, + "step": 20294 + }, + { + "epoch": 2.146256345177665, + "grad_norm": 8.3125, + "learning_rate": 3.185465433568052e-06, + "loss": 2.5558, + "step": 20295 + }, + { + "epoch": 2.1463620981387477, + "grad_norm": 8.25, + "learning_rate": 3.1853007388525215e-06, + "loss": 2.5802, + "step": 20296 + }, + { + "epoch": 2.146467851099831, + "grad_norm": 8.5, + "learning_rate": 3.185136040921154e-06, + "loss": 2.7128, + "step": 20297 + }, + { + "epoch": 2.146573604060914, + "grad_norm": 8.0625, + "learning_rate": 3.1849713397747238e-06, + "loss": 2.336, + "step": 20298 + }, + { + "epoch": 2.1466793570219966, + "grad_norm": 8.3125, + "learning_rate": 3.1848066354140026e-06, + "loss": 2.3404, + "step": 20299 + }, + { + "epoch": 2.1467851099830795, + "grad_norm": 8.4375, + "learning_rate": 3.1846419278397643e-06, + "loss": 2.4792, + "step": 20300 + }, + { + "epoch": 2.1468908629441623, + "grad_norm": 8.375, + "learning_rate": 3.1844772170527816e-06, + "loss": 2.6486, + "step": 20301 + }, + { + "epoch": 2.146996615905245, + "grad_norm": 9.0625, + "learning_rate": 3.1843125030538276e-06, + "loss": 2.6225, + "step": 20302 + }, + { + "epoch": 2.1471023688663284, + "grad_norm": 7.90625, + "learning_rate": 3.1841477858436738e-06, + "loss": 2.545, + "step": 20303 + }, + { + "epoch": 2.1472081218274113, + "grad_norm": 7.90625, + "learning_rate": 3.1839830654230953e-06, + "loss": 2.4266, + "step": 20304 + }, + { + "epoch": 2.147313874788494, + "grad_norm": 8.5, + "learning_rate": 3.1838183417928638e-06, + "loss": 2.432, + "step": 20305 + }, + { + "epoch": 2.147419627749577, + "grad_norm": 9.1875, + "learning_rate": 3.1836536149537517e-06, + "loss": 2.7819, + "step": 20306 + }, + { + "epoch": 2.14752538071066, + "grad_norm": 8.1875, + "learning_rate": 3.1834888849065336e-06, + "loss": 2.5768, + "step": 20307 + }, + { + "epoch": 2.1476311336717426, + "grad_norm": 9.0625, + "learning_rate": 3.1833241516519813e-06, + "loss": 2.748, + "step": 20308 + }, + { + "epoch": 2.147736886632826, + "grad_norm": 8.8125, + "learning_rate": 3.1831594151908684e-06, + "loss": 2.2647, + "step": 20309 + }, + { + "epoch": 2.1478426395939088, + "grad_norm": 8.0625, + "learning_rate": 3.1829946755239675e-06, + "loss": 2.6003, + "step": 20310 + }, + { + "epoch": 2.1479483925549916, + "grad_norm": 8.125, + "learning_rate": 3.1828299326520523e-06, + "loss": 2.5931, + "step": 20311 + }, + { + "epoch": 2.1480541455160744, + "grad_norm": 9.0625, + "learning_rate": 3.1826651865758958e-06, + "loss": 2.6474, + "step": 20312 + }, + { + "epoch": 2.1481598984771573, + "grad_norm": 8.125, + "learning_rate": 3.18250043729627e-06, + "loss": 2.5682, + "step": 20313 + }, + { + "epoch": 2.14826565143824, + "grad_norm": 8.5, + "learning_rate": 3.18233568481395e-06, + "loss": 2.5004, + "step": 20314 + }, + { + "epoch": 2.148371404399323, + "grad_norm": 7.90625, + "learning_rate": 3.1821709291297067e-06, + "loss": 2.5046, + "step": 20315 + }, + { + "epoch": 2.1484771573604062, + "grad_norm": 8.1875, + "learning_rate": 3.182006170244314e-06, + "loss": 2.6742, + "step": 20316 + }, + { + "epoch": 2.148582910321489, + "grad_norm": 8.75, + "learning_rate": 3.181841408158545e-06, + "loss": 2.4048, + "step": 20317 + }, + { + "epoch": 2.148688663282572, + "grad_norm": 7.96875, + "learning_rate": 3.1816766428731742e-06, + "loss": 2.678, + "step": 20318 + }, + { + "epoch": 2.1487944162436547, + "grad_norm": 8.25, + "learning_rate": 3.181511874388973e-06, + "loss": 2.6057, + "step": 20319 + }, + { + "epoch": 2.1489001692047376, + "grad_norm": 8.5625, + "learning_rate": 3.1813471027067155e-06, + "loss": 2.4073, + "step": 20320 + }, + { + "epoch": 2.149005922165821, + "grad_norm": 7.96875, + "learning_rate": 3.181182327827175e-06, + "loss": 2.4237, + "step": 20321 + }, + { + "epoch": 2.1491116751269037, + "grad_norm": 8.4375, + "learning_rate": 3.1810175497511244e-06, + "loss": 2.6907, + "step": 20322 + }, + { + "epoch": 2.1492174280879865, + "grad_norm": 8.4375, + "learning_rate": 3.1808527684793365e-06, + "loss": 2.7748, + "step": 20323 + }, + { + "epoch": 2.1493231810490694, + "grad_norm": 8.125, + "learning_rate": 3.180687984012586e-06, + "loss": 2.611, + "step": 20324 + }, + { + "epoch": 2.1494289340101522, + "grad_norm": 8.9375, + "learning_rate": 3.180523196351644e-06, + "loss": 2.7447, + "step": 20325 + }, + { + "epoch": 2.149534686971235, + "grad_norm": 8.4375, + "learning_rate": 3.1803584054972853e-06, + "loss": 2.6991, + "step": 20326 + }, + { + "epoch": 2.149640439932318, + "grad_norm": 8.1875, + "learning_rate": 3.180193611450283e-06, + "loss": 2.5582, + "step": 20327 + }, + { + "epoch": 2.149746192893401, + "grad_norm": 9.1875, + "learning_rate": 3.1800288142114107e-06, + "loss": 2.652, + "step": 20328 + }, + { + "epoch": 2.149851945854484, + "grad_norm": 8.0625, + "learning_rate": 3.179864013781441e-06, + "loss": 2.569, + "step": 20329 + }, + { + "epoch": 2.149957698815567, + "grad_norm": 8.1875, + "learning_rate": 3.1796992101611473e-06, + "loss": 2.4635, + "step": 20330 + }, + { + "epoch": 2.1500634517766497, + "grad_norm": 8.3125, + "learning_rate": 3.1795344033513033e-06, + "loss": 2.5086, + "step": 20331 + }, + { + "epoch": 2.1501692047377325, + "grad_norm": 8.5, + "learning_rate": 3.1793695933526818e-06, + "loss": 2.4851, + "step": 20332 + }, + { + "epoch": 2.1502749576988154, + "grad_norm": 8.625, + "learning_rate": 3.1792047801660574e-06, + "loss": 2.5561, + "step": 20333 + }, + { + "epoch": 2.1503807106598987, + "grad_norm": 8.0, + "learning_rate": 3.1790399637922026e-06, + "loss": 2.4414, + "step": 20334 + }, + { + "epoch": 2.1504864636209815, + "grad_norm": 8.0625, + "learning_rate": 3.1788751442318903e-06, + "loss": 2.5265, + "step": 20335 + }, + { + "epoch": 2.1505922165820643, + "grad_norm": 9.3125, + "learning_rate": 3.1787103214858956e-06, + "loss": 2.3102, + "step": 20336 + }, + { + "epoch": 2.150697969543147, + "grad_norm": 8.375, + "learning_rate": 3.1785454955549905e-06, + "loss": 2.4332, + "step": 20337 + }, + { + "epoch": 2.15080372250423, + "grad_norm": 8.4375, + "learning_rate": 3.1783806664399484e-06, + "loss": 2.4813, + "step": 20338 + }, + { + "epoch": 2.150909475465313, + "grad_norm": 8.5, + "learning_rate": 3.1782158341415443e-06, + "loss": 2.4783, + "step": 20339 + }, + { + "epoch": 2.151015228426396, + "grad_norm": 8.0625, + "learning_rate": 3.17805099866055e-06, + "loss": 2.5668, + "step": 20340 + }, + { + "epoch": 2.151120981387479, + "grad_norm": 9.125, + "learning_rate": 3.1778861599977396e-06, + "loss": 2.6084, + "step": 20341 + }, + { + "epoch": 2.151226734348562, + "grad_norm": 8.5625, + "learning_rate": 3.1777213181538874e-06, + "loss": 2.7262, + "step": 20342 + }, + { + "epoch": 2.1513324873096447, + "grad_norm": 10.125, + "learning_rate": 3.1775564731297653e-06, + "loss": 2.601, + "step": 20343 + }, + { + "epoch": 2.1514382402707275, + "grad_norm": 8.4375, + "learning_rate": 3.1773916249261483e-06, + "loss": 2.5293, + "step": 20344 + }, + { + "epoch": 2.1515439932318103, + "grad_norm": 8.75, + "learning_rate": 3.1772267735438094e-06, + "loss": 2.48, + "step": 20345 + }, + { + "epoch": 2.151649746192893, + "grad_norm": 8.375, + "learning_rate": 3.1770619189835227e-06, + "loss": 2.451, + "step": 20346 + }, + { + "epoch": 2.1517554991539765, + "grad_norm": 8.1875, + "learning_rate": 3.176897061246061e-06, + "loss": 2.5866, + "step": 20347 + }, + { + "epoch": 2.1518612521150593, + "grad_norm": 8.3125, + "learning_rate": 3.1767322003321978e-06, + "loss": 2.4915, + "step": 20348 + }, + { + "epoch": 2.151967005076142, + "grad_norm": 9.3125, + "learning_rate": 3.1765673362427074e-06, + "loss": 2.7895, + "step": 20349 + }, + { + "epoch": 2.152072758037225, + "grad_norm": 8.6875, + "learning_rate": 3.1764024689783635e-06, + "loss": 2.6384, + "step": 20350 + }, + { + "epoch": 2.152178510998308, + "grad_norm": 8.5625, + "learning_rate": 3.1762375985399386e-06, + "loss": 2.5872, + "step": 20351 + }, + { + "epoch": 2.152284263959391, + "grad_norm": 8.375, + "learning_rate": 3.1760727249282076e-06, + "loss": 2.616, + "step": 20352 + }, + { + "epoch": 2.152390016920474, + "grad_norm": 8.625, + "learning_rate": 3.1759078481439444e-06, + "loss": 2.5714, + "step": 20353 + }, + { + "epoch": 2.1524957698815568, + "grad_norm": 8.3125, + "learning_rate": 3.1757429681879214e-06, + "loss": 2.4695, + "step": 20354 + }, + { + "epoch": 2.1526015228426396, + "grad_norm": 8.375, + "learning_rate": 3.1755780850609137e-06, + "loss": 2.5272, + "step": 20355 + }, + { + "epoch": 2.1527072758037225, + "grad_norm": 9.0625, + "learning_rate": 3.1754131987636937e-06, + "loss": 2.3897, + "step": 20356 + }, + { + "epoch": 2.1528130287648053, + "grad_norm": 8.25, + "learning_rate": 3.175248309297036e-06, + "loss": 2.5258, + "step": 20357 + }, + { + "epoch": 2.152918781725888, + "grad_norm": 8.3125, + "learning_rate": 3.1750834166617143e-06, + "loss": 2.6525, + "step": 20358 + }, + { + "epoch": 2.1530245346869714, + "grad_norm": 8.1875, + "learning_rate": 3.1749185208585028e-06, + "loss": 2.4757, + "step": 20359 + }, + { + "epoch": 2.1531302876480543, + "grad_norm": 8.625, + "learning_rate": 3.1747536218881736e-06, + "loss": 2.5444, + "step": 20360 + }, + { + "epoch": 2.153236040609137, + "grad_norm": 8.1875, + "learning_rate": 3.174588719751502e-06, + "loss": 2.3573, + "step": 20361 + }, + { + "epoch": 2.15334179357022, + "grad_norm": 9.0, + "learning_rate": 3.174423814449261e-06, + "loss": 2.5914, + "step": 20362 + }, + { + "epoch": 2.1534475465313028, + "grad_norm": 8.375, + "learning_rate": 3.174258905982226e-06, + "loss": 2.5502, + "step": 20363 + }, + { + "epoch": 2.1535532994923856, + "grad_norm": 8.5625, + "learning_rate": 3.174093994351169e-06, + "loss": 2.4864, + "step": 20364 + }, + { + "epoch": 2.153659052453469, + "grad_norm": 8.3125, + "learning_rate": 3.1739290795568652e-06, + "loss": 2.4989, + "step": 20365 + }, + { + "epoch": 2.1537648054145517, + "grad_norm": 8.1875, + "learning_rate": 3.1737641616000868e-06, + "loss": 2.6582, + "step": 20366 + }, + { + "epoch": 2.1538705583756346, + "grad_norm": 8.125, + "learning_rate": 3.1735992404816096e-06, + "loss": 2.5652, + "step": 20367 + }, + { + "epoch": 2.1539763113367174, + "grad_norm": 8.625, + "learning_rate": 3.173434316202206e-06, + "loss": 2.5894, + "step": 20368 + }, + { + "epoch": 2.1540820642978002, + "grad_norm": 8.5, + "learning_rate": 3.1732693887626516e-06, + "loss": 2.586, + "step": 20369 + }, + { + "epoch": 2.154187817258883, + "grad_norm": 9.125, + "learning_rate": 3.1731044581637175e-06, + "loss": 2.7055, + "step": 20370 + }, + { + "epoch": 2.1542935702199664, + "grad_norm": 8.25, + "learning_rate": 3.172939524406181e-06, + "loss": 2.4759, + "step": 20371 + }, + { + "epoch": 2.154399323181049, + "grad_norm": 8.375, + "learning_rate": 3.1727745874908143e-06, + "loss": 2.7138, + "step": 20372 + }, + { + "epoch": 2.154505076142132, + "grad_norm": 8.6875, + "learning_rate": 3.172609647418391e-06, + "loss": 2.6379, + "step": 20373 + }, + { + "epoch": 2.154610829103215, + "grad_norm": 8.625, + "learning_rate": 3.1724447041896865e-06, + "loss": 2.5473, + "step": 20374 + }, + { + "epoch": 2.1547165820642977, + "grad_norm": 9.125, + "learning_rate": 3.1722797578054744e-06, + "loss": 2.7954, + "step": 20375 + }, + { + "epoch": 2.1548223350253806, + "grad_norm": 8.5, + "learning_rate": 3.172114808266527e-06, + "loss": 2.5149, + "step": 20376 + }, + { + "epoch": 2.1549280879864634, + "grad_norm": 8.9375, + "learning_rate": 3.1719498555736206e-06, + "loss": 2.4858, + "step": 20377 + }, + { + "epoch": 2.1550338409475467, + "grad_norm": 8.0, + "learning_rate": 3.171784899727528e-06, + "loss": 2.2166, + "step": 20378 + }, + { + "epoch": 2.1551395939086295, + "grad_norm": 8.5, + "learning_rate": 3.1716199407290236e-06, + "loss": 2.6286, + "step": 20379 + }, + { + "epoch": 2.1552453468697124, + "grad_norm": 8.875, + "learning_rate": 3.171454978578882e-06, + "loss": 2.7551, + "step": 20380 + }, + { + "epoch": 2.155351099830795, + "grad_norm": 8.125, + "learning_rate": 3.171290013277876e-06, + "loss": 2.4599, + "step": 20381 + }, + { + "epoch": 2.155456852791878, + "grad_norm": 8.5625, + "learning_rate": 3.171125044826781e-06, + "loss": 2.5136, + "step": 20382 + }, + { + "epoch": 2.155562605752961, + "grad_norm": 8.375, + "learning_rate": 3.17096007322637e-06, + "loss": 2.4817, + "step": 20383 + }, + { + "epoch": 2.155668358714044, + "grad_norm": 8.5, + "learning_rate": 3.1707950984774187e-06, + "loss": 2.6006, + "step": 20384 + }, + { + "epoch": 2.155774111675127, + "grad_norm": 8.25, + "learning_rate": 3.1706301205807006e-06, + "loss": 2.5101, + "step": 20385 + }, + { + "epoch": 2.15587986463621, + "grad_norm": 8.4375, + "learning_rate": 3.170465139536989e-06, + "loss": 2.2959, + "step": 20386 + }, + { + "epoch": 2.1559856175972927, + "grad_norm": 8.0625, + "learning_rate": 3.170300155347058e-06, + "loss": 2.5855, + "step": 20387 + }, + { + "epoch": 2.1560913705583755, + "grad_norm": 8.125, + "learning_rate": 3.1701351680116836e-06, + "loss": 2.424, + "step": 20388 + }, + { + "epoch": 2.1561971235194584, + "grad_norm": 8.0625, + "learning_rate": 3.1699701775316384e-06, + "loss": 2.2936, + "step": 20389 + }, + { + "epoch": 2.1563028764805416, + "grad_norm": 8.1875, + "learning_rate": 3.169805183907697e-06, + "loss": 2.2581, + "step": 20390 + }, + { + "epoch": 2.1564086294416245, + "grad_norm": 8.3125, + "learning_rate": 3.1696401871406345e-06, + "loss": 2.3339, + "step": 20391 + }, + { + "epoch": 2.1565143824027073, + "grad_norm": 8.4375, + "learning_rate": 3.169475187231224e-06, + "loss": 2.6097, + "step": 20392 + }, + { + "epoch": 2.15662013536379, + "grad_norm": 8.25, + "learning_rate": 3.169310184180241e-06, + "loss": 2.5637, + "step": 20393 + }, + { + "epoch": 2.156725888324873, + "grad_norm": 8.0625, + "learning_rate": 3.1691451779884584e-06, + "loss": 2.6631, + "step": 20394 + }, + { + "epoch": 2.156831641285956, + "grad_norm": 8.3125, + "learning_rate": 3.1689801686566506e-06, + "loss": 2.5192, + "step": 20395 + }, + { + "epoch": 2.156937394247039, + "grad_norm": 8.1875, + "learning_rate": 3.168815156185593e-06, + "loss": 2.5625, + "step": 20396 + }, + { + "epoch": 2.157043147208122, + "grad_norm": 8.25, + "learning_rate": 3.1686501405760593e-06, + "loss": 2.4966, + "step": 20397 + }, + { + "epoch": 2.157148900169205, + "grad_norm": 8.25, + "learning_rate": 3.1684851218288247e-06, + "loss": 2.4172, + "step": 20398 + }, + { + "epoch": 2.1572546531302876, + "grad_norm": 8.4375, + "learning_rate": 3.1683200999446615e-06, + "loss": 2.4888, + "step": 20399 + }, + { + "epoch": 2.1573604060913705, + "grad_norm": 8.0625, + "learning_rate": 3.1681550749243466e-06, + "loss": 2.5772, + "step": 20400 + }, + { + "epoch": 2.1574661590524533, + "grad_norm": 8.5625, + "learning_rate": 3.1679900467686533e-06, + "loss": 2.8747, + "step": 20401 + }, + { + "epoch": 2.1575719120135366, + "grad_norm": 9.0625, + "learning_rate": 3.1678250154783553e-06, + "loss": 2.6053, + "step": 20402 + }, + { + "epoch": 2.1576776649746194, + "grad_norm": 8.125, + "learning_rate": 3.1676599810542276e-06, + "loss": 2.4587, + "step": 20403 + }, + { + "epoch": 2.1577834179357023, + "grad_norm": 8.375, + "learning_rate": 3.1674949434970446e-06, + "loss": 2.8424, + "step": 20404 + }, + { + "epoch": 2.157889170896785, + "grad_norm": 8.75, + "learning_rate": 3.1673299028075803e-06, + "loss": 2.5042, + "step": 20405 + }, + { + "epoch": 2.157994923857868, + "grad_norm": 8.3125, + "learning_rate": 3.167164858986611e-06, + "loss": 2.71, + "step": 20406 + }, + { + "epoch": 2.158100676818951, + "grad_norm": 8.1875, + "learning_rate": 3.1669998120349093e-06, + "loss": 2.4193, + "step": 20407 + }, + { + "epoch": 2.1582064297800336, + "grad_norm": 8.4375, + "learning_rate": 3.16683476195325e-06, + "loss": 2.6721, + "step": 20408 + }, + { + "epoch": 2.158312182741117, + "grad_norm": 8.5, + "learning_rate": 3.166669708742408e-06, + "loss": 2.3421, + "step": 20409 + }, + { + "epoch": 2.1584179357021998, + "grad_norm": 8.4375, + "learning_rate": 3.1665046524031585e-06, + "loss": 2.6075, + "step": 20410 + }, + { + "epoch": 2.1585236886632826, + "grad_norm": 8.4375, + "learning_rate": 3.166339592936274e-06, + "loss": 2.5135, + "step": 20411 + }, + { + "epoch": 2.1586294416243654, + "grad_norm": 8.25, + "learning_rate": 3.1661745303425312e-06, + "loss": 2.5519, + "step": 20412 + }, + { + "epoch": 2.1587351945854483, + "grad_norm": 8.5625, + "learning_rate": 3.166009464622704e-06, + "loss": 2.5066, + "step": 20413 + }, + { + "epoch": 2.158840947546531, + "grad_norm": 7.84375, + "learning_rate": 3.1658443957775657e-06, + "loss": 2.486, + "step": 20414 + }, + { + "epoch": 2.1589467005076144, + "grad_norm": 8.125, + "learning_rate": 3.1656793238078925e-06, + "loss": 2.397, + "step": 20415 + }, + { + "epoch": 2.1590524534686972, + "grad_norm": 8.4375, + "learning_rate": 3.1655142487144588e-06, + "loss": 2.6839, + "step": 20416 + }, + { + "epoch": 2.15915820642978, + "grad_norm": 8.625, + "learning_rate": 3.1653491704980387e-06, + "loss": 2.759, + "step": 20417 + }, + { + "epoch": 2.159263959390863, + "grad_norm": 7.8125, + "learning_rate": 3.1651840891594065e-06, + "loss": 2.4776, + "step": 20418 + }, + { + "epoch": 2.1593697123519457, + "grad_norm": 8.1875, + "learning_rate": 3.1650190046993384e-06, + "loss": 2.4029, + "step": 20419 + }, + { + "epoch": 2.1594754653130286, + "grad_norm": 8.25, + "learning_rate": 3.1648539171186075e-06, + "loss": 2.3411, + "step": 20420 + }, + { + "epoch": 2.159581218274112, + "grad_norm": 8.375, + "learning_rate": 3.1646888264179895e-06, + "loss": 2.6557, + "step": 20421 + }, + { + "epoch": 2.1596869712351947, + "grad_norm": 8.125, + "learning_rate": 3.164523732598258e-06, + "loss": 2.5528, + "step": 20422 + }, + { + "epoch": 2.1597927241962775, + "grad_norm": 8.625, + "learning_rate": 3.1643586356601886e-06, + "loss": 2.29, + "step": 20423 + }, + { + "epoch": 2.1598984771573604, + "grad_norm": 8.625, + "learning_rate": 3.1641935356045563e-06, + "loss": 2.5659, + "step": 20424 + }, + { + "epoch": 2.1600042301184432, + "grad_norm": 8.0625, + "learning_rate": 3.1640284324321354e-06, + "loss": 2.4422, + "step": 20425 + }, + { + "epoch": 2.160109983079526, + "grad_norm": 8.6875, + "learning_rate": 3.1638633261437e-06, + "loss": 2.4906, + "step": 20426 + }, + { + "epoch": 2.160215736040609, + "grad_norm": 8.125, + "learning_rate": 3.1636982167400258e-06, + "loss": 2.5945, + "step": 20427 + }, + { + "epoch": 2.160321489001692, + "grad_norm": 8.625, + "learning_rate": 3.1635331042218876e-06, + "loss": 2.9028, + "step": 20428 + }, + { + "epoch": 2.160427241962775, + "grad_norm": 8.0, + "learning_rate": 3.1633679885900597e-06, + "loss": 1.8613, + "step": 20429 + }, + { + "epoch": 2.160532994923858, + "grad_norm": 8.4375, + "learning_rate": 3.163202869845317e-06, + "loss": 2.3729, + "step": 20430 + }, + { + "epoch": 2.1606387478849407, + "grad_norm": 8.5625, + "learning_rate": 3.1630377479884348e-06, + "loss": 2.5649, + "step": 20431 + }, + { + "epoch": 2.1607445008460235, + "grad_norm": 8.625, + "learning_rate": 3.1628726230201875e-06, + "loss": 2.4163, + "step": 20432 + }, + { + "epoch": 2.160850253807107, + "grad_norm": 8.0625, + "learning_rate": 3.16270749494135e-06, + "loss": 2.5185, + "step": 20433 + }, + { + "epoch": 2.1609560067681897, + "grad_norm": 8.5625, + "learning_rate": 3.162542363752697e-06, + "loss": 2.786, + "step": 20434 + }, + { + "epoch": 2.1610617597292725, + "grad_norm": 8.125, + "learning_rate": 3.1623772294550044e-06, + "loss": 2.534, + "step": 20435 + }, + { + "epoch": 2.1611675126903553, + "grad_norm": 7.90625, + "learning_rate": 3.1622120920490464e-06, + "loss": 2.5637, + "step": 20436 + }, + { + "epoch": 2.161273265651438, + "grad_norm": 8.25, + "learning_rate": 3.162046951535597e-06, + "loss": 2.6091, + "step": 20437 + }, + { + "epoch": 2.161379018612521, + "grad_norm": 8.625, + "learning_rate": 3.1618818079154325e-06, + "loss": 2.5263, + "step": 20438 + }, + { + "epoch": 2.161484771573604, + "grad_norm": 8.75, + "learning_rate": 3.1617166611893275e-06, + "loss": 2.6121, + "step": 20439 + }, + { + "epoch": 2.161590524534687, + "grad_norm": 8.375, + "learning_rate": 3.1615515113580564e-06, + "loss": 2.6195, + "step": 20440 + }, + { + "epoch": 2.16169627749577, + "grad_norm": 7.84375, + "learning_rate": 3.1613863584223944e-06, + "loss": 2.6148, + "step": 20441 + }, + { + "epoch": 2.161802030456853, + "grad_norm": 8.25, + "learning_rate": 3.161221202383117e-06, + "loss": 2.6404, + "step": 20442 + }, + { + "epoch": 2.1619077834179357, + "grad_norm": 8.6875, + "learning_rate": 3.161056043240999e-06, + "loss": 2.4972, + "step": 20443 + }, + { + "epoch": 2.1620135363790185, + "grad_norm": 8.875, + "learning_rate": 3.160890880996816e-06, + "loss": 2.5229, + "step": 20444 + }, + { + "epoch": 2.1621192893401013, + "grad_norm": 7.9375, + "learning_rate": 3.1607257156513415e-06, + "loss": 2.4335, + "step": 20445 + }, + { + "epoch": 2.1622250423011846, + "grad_norm": 8.125, + "learning_rate": 3.1605605472053517e-06, + "loss": 2.006, + "step": 20446 + }, + { + "epoch": 2.1623307952622675, + "grad_norm": 8.625, + "learning_rate": 3.160395375659622e-06, + "loss": 2.6153, + "step": 20447 + }, + { + "epoch": 2.1624365482233503, + "grad_norm": 8.1875, + "learning_rate": 3.1602302010149264e-06, + "loss": 2.6149, + "step": 20448 + }, + { + "epoch": 2.162542301184433, + "grad_norm": 8.1875, + "learning_rate": 3.16006502327204e-06, + "loss": 2.5248, + "step": 20449 + }, + { + "epoch": 2.162648054145516, + "grad_norm": 8.3125, + "learning_rate": 3.159899842431739e-06, + "loss": 2.4876, + "step": 20450 + }, + { + "epoch": 2.162753807106599, + "grad_norm": 8.3125, + "learning_rate": 3.159734658494798e-06, + "loss": 2.6096, + "step": 20451 + }, + { + "epoch": 2.162859560067682, + "grad_norm": 8.4375, + "learning_rate": 3.159569471461992e-06, + "loss": 2.5228, + "step": 20452 + }, + { + "epoch": 2.162965313028765, + "grad_norm": 8.5625, + "learning_rate": 3.159404281334096e-06, + "loss": 2.5781, + "step": 20453 + }, + { + "epoch": 2.1630710659898478, + "grad_norm": 8.375, + "learning_rate": 3.1592390881118855e-06, + "loss": 2.5003, + "step": 20454 + }, + { + "epoch": 2.1631768189509306, + "grad_norm": 8.375, + "learning_rate": 3.1590738917961362e-06, + "loss": 2.4944, + "step": 20455 + }, + { + "epoch": 2.1632825719120135, + "grad_norm": 8.375, + "learning_rate": 3.1589086923876215e-06, + "loss": 2.3388, + "step": 20456 + }, + { + "epoch": 2.1633883248730963, + "grad_norm": 8.4375, + "learning_rate": 3.158743489887119e-06, + "loss": 2.6692, + "step": 20457 + }, + { + "epoch": 2.163494077834179, + "grad_norm": 7.96875, + "learning_rate": 3.1585782842954016e-06, + "loss": 2.3259, + "step": 20458 + }, + { + "epoch": 2.1635998307952624, + "grad_norm": 8.375, + "learning_rate": 3.1584130756132457e-06, + "loss": 2.5283, + "step": 20459 + }, + { + "epoch": 2.1637055837563453, + "grad_norm": 8.4375, + "learning_rate": 3.158247863841427e-06, + "loss": 2.4832, + "step": 20460 + }, + { + "epoch": 2.163811336717428, + "grad_norm": 8.125, + "learning_rate": 3.15808264898072e-06, + "loss": 2.4805, + "step": 20461 + }, + { + "epoch": 2.163917089678511, + "grad_norm": 8.0625, + "learning_rate": 3.157917431031901e-06, + "loss": 2.6008, + "step": 20462 + }, + { + "epoch": 2.1640228426395938, + "grad_norm": 8.125, + "learning_rate": 3.1577522099957442e-06, + "loss": 2.4499, + "step": 20463 + }, + { + "epoch": 2.164128595600677, + "grad_norm": 8.625, + "learning_rate": 3.1575869858730256e-06, + "loss": 2.5793, + "step": 20464 + }, + { + "epoch": 2.16423434856176, + "grad_norm": 8.625, + "learning_rate": 3.1574217586645194e-06, + "loss": 2.5001, + "step": 20465 + }, + { + "epoch": 2.1643401015228427, + "grad_norm": 8.5, + "learning_rate": 3.1572565283710022e-06, + "loss": 2.4832, + "step": 20466 + }, + { + "epoch": 2.1644458544839256, + "grad_norm": 8.5, + "learning_rate": 3.157091294993249e-06, + "loss": 2.3824, + "step": 20467 + }, + { + "epoch": 2.1645516074450084, + "grad_norm": 8.9375, + "learning_rate": 3.156926058532035e-06, + "loss": 2.6357, + "step": 20468 + }, + { + "epoch": 2.1646573604060912, + "grad_norm": 8.3125, + "learning_rate": 3.1567608189881355e-06, + "loss": 2.5996, + "step": 20469 + }, + { + "epoch": 2.164763113367174, + "grad_norm": 8.25, + "learning_rate": 3.1565955763623264e-06, + "loss": 2.6432, + "step": 20470 + }, + { + "epoch": 2.1648688663282574, + "grad_norm": 8.4375, + "learning_rate": 3.156430330655383e-06, + "loss": 2.6759, + "step": 20471 + }, + { + "epoch": 2.16497461928934, + "grad_norm": 8.4375, + "learning_rate": 3.1562650818680796e-06, + "loss": 2.5436, + "step": 20472 + }, + { + "epoch": 2.165080372250423, + "grad_norm": 8.75, + "learning_rate": 3.1560998300011936e-06, + "loss": 2.5642, + "step": 20473 + }, + { + "epoch": 2.165186125211506, + "grad_norm": 9.4375, + "learning_rate": 3.155934575055499e-06, + "loss": 2.6631, + "step": 20474 + }, + { + "epoch": 2.1652918781725887, + "grad_norm": 8.4375, + "learning_rate": 3.1557693170317717e-06, + "loss": 2.5059, + "step": 20475 + }, + { + "epoch": 2.1653976311336716, + "grad_norm": 7.78125, + "learning_rate": 3.1556040559307866e-06, + "loss": 2.4404, + "step": 20476 + }, + { + "epoch": 2.165503384094755, + "grad_norm": 8.125, + "learning_rate": 3.1554387917533207e-06, + "loss": 2.5, + "step": 20477 + }, + { + "epoch": 2.1656091370558377, + "grad_norm": 8.375, + "learning_rate": 3.155273524500148e-06, + "loss": 2.5381, + "step": 20478 + }, + { + "epoch": 2.1657148900169205, + "grad_norm": 8.125, + "learning_rate": 3.1551082541720453e-06, + "loss": 2.4367, + "step": 20479 + }, + { + "epoch": 2.1658206429780034, + "grad_norm": 8.4375, + "learning_rate": 3.1549429807697875e-06, + "loss": 2.5677, + "step": 20480 + }, + { + "epoch": 2.165926395939086, + "grad_norm": 8.875, + "learning_rate": 3.154777704294149e-06, + "loss": 2.6702, + "step": 20481 + }, + { + "epoch": 2.166032148900169, + "grad_norm": 8.0625, + "learning_rate": 3.1546124247459074e-06, + "loss": 2.3666, + "step": 20482 + }, + { + "epoch": 2.1661379018612523, + "grad_norm": 8.1875, + "learning_rate": 3.154447142125838e-06, + "loss": 2.3875, + "step": 20483 + }, + { + "epoch": 2.166243654822335, + "grad_norm": 8.375, + "learning_rate": 3.154281856434715e-06, + "loss": 2.5643, + "step": 20484 + }, + { + "epoch": 2.166349407783418, + "grad_norm": 8.125, + "learning_rate": 3.154116567673314e-06, + "loss": 2.5473, + "step": 20485 + }, + { + "epoch": 2.166455160744501, + "grad_norm": 8.6875, + "learning_rate": 3.1539512758424134e-06, + "loss": 2.4943, + "step": 20486 + }, + { + "epoch": 2.1665609137055837, + "grad_norm": 9.5, + "learning_rate": 3.1537859809427864e-06, + "loss": 2.6018, + "step": 20487 + }, + { + "epoch": 2.1666666666666665, + "grad_norm": 8.3125, + "learning_rate": 3.153620682975208e-06, + "loss": 2.5321, + "step": 20488 + }, + { + "epoch": 2.1667724196277494, + "grad_norm": 8.3125, + "learning_rate": 3.1534553819404562e-06, + "loss": 2.5528, + "step": 20489 + }, + { + "epoch": 2.1668781725888326, + "grad_norm": 9.375, + "learning_rate": 3.1532900778393053e-06, + "loss": 2.9105, + "step": 20490 + }, + { + "epoch": 2.1669839255499155, + "grad_norm": 8.625, + "learning_rate": 3.153124770672531e-06, + "loss": 2.5278, + "step": 20491 + }, + { + "epoch": 2.1670896785109983, + "grad_norm": 7.96875, + "learning_rate": 3.15295946044091e-06, + "loss": 2.6633, + "step": 20492 + }, + { + "epoch": 2.167195431472081, + "grad_norm": 8.1875, + "learning_rate": 3.1527941471452168e-06, + "loss": 2.6117, + "step": 20493 + }, + { + "epoch": 2.167301184433164, + "grad_norm": 9.5, + "learning_rate": 3.1526288307862264e-06, + "loss": 2.6843, + "step": 20494 + }, + { + "epoch": 2.167406937394247, + "grad_norm": 8.5625, + "learning_rate": 3.1524635113647173e-06, + "loss": 2.359, + "step": 20495 + }, + { + "epoch": 2.16751269035533, + "grad_norm": 8.875, + "learning_rate": 3.1522981888814636e-06, + "loss": 2.997, + "step": 20496 + }, + { + "epoch": 2.167618443316413, + "grad_norm": 8.25, + "learning_rate": 3.1521328633372404e-06, + "loss": 2.4878, + "step": 20497 + }, + { + "epoch": 2.167724196277496, + "grad_norm": 9.25, + "learning_rate": 3.151967534732826e-06, + "loss": 2.2334, + "step": 20498 + }, + { + "epoch": 2.1678299492385786, + "grad_norm": 10.75, + "learning_rate": 3.151802203068994e-06, + "loss": 2.3674, + "step": 20499 + }, + { + "epoch": 2.1679357021996615, + "grad_norm": 9.0625, + "learning_rate": 3.1516368683465203e-06, + "loss": 2.5705, + "step": 20500 + }, + { + "epoch": 2.1680414551607443, + "grad_norm": 8.0625, + "learning_rate": 3.1514715305661815e-06, + "loss": 2.5814, + "step": 20501 + }, + { + "epoch": 2.1681472081218276, + "grad_norm": 8.125, + "learning_rate": 3.1513061897287534e-06, + "loss": 2.1671, + "step": 20502 + }, + { + "epoch": 2.1682529610829104, + "grad_norm": 8.375, + "learning_rate": 3.151140845835011e-06, + "loss": 2.3465, + "step": 20503 + }, + { + "epoch": 2.1683587140439933, + "grad_norm": 8.375, + "learning_rate": 3.1509754988857316e-06, + "loss": 2.4868, + "step": 20504 + }, + { + "epoch": 2.168464467005076, + "grad_norm": 9.3125, + "learning_rate": 3.150810148881691e-06, + "loss": 2.6525, + "step": 20505 + }, + { + "epoch": 2.168570219966159, + "grad_norm": 8.5, + "learning_rate": 3.150644795823664e-06, + "loss": 2.606, + "step": 20506 + }, + { + "epoch": 2.168675972927242, + "grad_norm": 8.375, + "learning_rate": 3.1504794397124265e-06, + "loss": 2.4643, + "step": 20507 + }, + { + "epoch": 2.168781725888325, + "grad_norm": 8.3125, + "learning_rate": 3.1503140805487558e-06, + "loss": 2.5561, + "step": 20508 + }, + { + "epoch": 2.168887478849408, + "grad_norm": 8.5, + "learning_rate": 3.1501487183334274e-06, + "loss": 2.5056, + "step": 20509 + }, + { + "epoch": 2.1689932318104908, + "grad_norm": 8.1875, + "learning_rate": 3.149983353067216e-06, + "loss": 2.2736, + "step": 20510 + }, + { + "epoch": 2.1690989847715736, + "grad_norm": 8.3125, + "learning_rate": 3.1498179847508984e-06, + "loss": 2.5136, + "step": 20511 + }, + { + "epoch": 2.1692047377326564, + "grad_norm": 8.5625, + "learning_rate": 3.149652613385252e-06, + "loss": 2.6451, + "step": 20512 + }, + { + "epoch": 2.1693104906937393, + "grad_norm": 8.25, + "learning_rate": 3.14948723897105e-06, + "loss": 2.5157, + "step": 20513 + }, + { + "epoch": 2.1694162436548226, + "grad_norm": 8.375, + "learning_rate": 3.1493218615090715e-06, + "loss": 2.5428, + "step": 20514 + }, + { + "epoch": 2.1695219966159054, + "grad_norm": 8.5625, + "learning_rate": 3.149156481000091e-06, + "loss": 2.5801, + "step": 20515 + }, + { + "epoch": 2.1696277495769882, + "grad_norm": 8.0, + "learning_rate": 3.1489910974448833e-06, + "loss": 2.5619, + "step": 20516 + }, + { + "epoch": 2.169733502538071, + "grad_norm": 8.5625, + "learning_rate": 3.1488257108442264e-06, + "loss": 2.4637, + "step": 20517 + }, + { + "epoch": 2.169839255499154, + "grad_norm": 8.4375, + "learning_rate": 3.1486603211988964e-06, + "loss": 2.6029, + "step": 20518 + }, + { + "epoch": 2.1699450084602367, + "grad_norm": 8.75, + "learning_rate": 3.1484949285096684e-06, + "loss": 2.6953, + "step": 20519 + }, + { + "epoch": 2.1700507614213196, + "grad_norm": 8.3125, + "learning_rate": 3.1483295327773187e-06, + "loss": 2.4301, + "step": 20520 + }, + { + "epoch": 2.170156514382403, + "grad_norm": 9.0625, + "learning_rate": 3.148164134002624e-06, + "loss": 2.5804, + "step": 20521 + }, + { + "epoch": 2.1702622673434857, + "grad_norm": 8.3125, + "learning_rate": 3.14799873218636e-06, + "loss": 2.3575, + "step": 20522 + }, + { + "epoch": 2.1703680203045685, + "grad_norm": 8.5625, + "learning_rate": 3.1478333273293027e-06, + "loss": 2.521, + "step": 20523 + }, + { + "epoch": 2.1704737732656514, + "grad_norm": 9.3125, + "learning_rate": 3.147667919432229e-06, + "loss": 2.4929, + "step": 20524 + }, + { + "epoch": 2.1705795262267342, + "grad_norm": 9.1875, + "learning_rate": 3.1475025084959155e-06, + "loss": 2.6752, + "step": 20525 + }, + { + "epoch": 2.170685279187817, + "grad_norm": 8.375, + "learning_rate": 3.1473370945211358e-06, + "loss": 2.5237, + "step": 20526 + }, + { + "epoch": 2.1707910321489003, + "grad_norm": 9.3125, + "learning_rate": 3.1471716775086693e-06, + "loss": 2.4189, + "step": 20527 + }, + { + "epoch": 2.170896785109983, + "grad_norm": 8.375, + "learning_rate": 3.1470062574592895e-06, + "loss": 2.4641, + "step": 20528 + }, + { + "epoch": 2.171002538071066, + "grad_norm": 9.3125, + "learning_rate": 3.146840834373774e-06, + "loss": 2.6272, + "step": 20529 + }, + { + "epoch": 2.171108291032149, + "grad_norm": 8.375, + "learning_rate": 3.1466754082529e-06, + "loss": 2.599, + "step": 20530 + }, + { + "epoch": 2.1712140439932317, + "grad_norm": 8.1875, + "learning_rate": 3.1465099790974424e-06, + "loss": 2.5872, + "step": 20531 + }, + { + "epoch": 2.1713197969543145, + "grad_norm": 8.3125, + "learning_rate": 3.1463445469081776e-06, + "loss": 2.4963, + "step": 20532 + }, + { + "epoch": 2.171425549915398, + "grad_norm": 8.8125, + "learning_rate": 3.1461791116858826e-06, + "loss": 2.7428, + "step": 20533 + }, + { + "epoch": 2.1715313028764807, + "grad_norm": 8.5625, + "learning_rate": 3.1460136734313333e-06, + "loss": 2.3311, + "step": 20534 + }, + { + "epoch": 2.1716370558375635, + "grad_norm": 8.6875, + "learning_rate": 3.1458482321453056e-06, + "loss": 2.6557, + "step": 20535 + }, + { + "epoch": 2.1717428087986463, + "grad_norm": 8.8125, + "learning_rate": 3.1456827878285767e-06, + "loss": 2.6117, + "step": 20536 + }, + { + "epoch": 2.171848561759729, + "grad_norm": 8.375, + "learning_rate": 3.1455173404819228e-06, + "loss": 2.6312, + "step": 20537 + }, + { + "epoch": 2.171954314720812, + "grad_norm": 8.6875, + "learning_rate": 3.145351890106119e-06, + "loss": 2.5414, + "step": 20538 + }, + { + "epoch": 2.1720600676818953, + "grad_norm": 8.1875, + "learning_rate": 3.1451864367019437e-06, + "loss": 2.1436, + "step": 20539 + }, + { + "epoch": 2.172165820642978, + "grad_norm": 8.1875, + "learning_rate": 3.145020980270172e-06, + "loss": 2.3586, + "step": 20540 + }, + { + "epoch": 2.172271573604061, + "grad_norm": 8.25, + "learning_rate": 3.1448555208115806e-06, + "loss": 2.4655, + "step": 20541 + }, + { + "epoch": 2.172377326565144, + "grad_norm": 7.9375, + "learning_rate": 3.1446900583269457e-06, + "loss": 2.515, + "step": 20542 + }, + { + "epoch": 2.1724830795262267, + "grad_norm": 7.78125, + "learning_rate": 3.144524592817045e-06, + "loss": 2.4402, + "step": 20543 + }, + { + "epoch": 2.1725888324873095, + "grad_norm": 8.125, + "learning_rate": 3.144359124282653e-06, + "loss": 2.6761, + "step": 20544 + }, + { + "epoch": 2.172694585448393, + "grad_norm": 8.125, + "learning_rate": 3.1441936527245475e-06, + "loss": 2.5312, + "step": 20545 + }, + { + "epoch": 2.1728003384094756, + "grad_norm": 8.6875, + "learning_rate": 3.1440281781435046e-06, + "loss": 2.8107, + "step": 20546 + }, + { + "epoch": 2.1729060913705585, + "grad_norm": 7.6875, + "learning_rate": 3.1438627005403006e-06, + "loss": 2.2473, + "step": 20547 + }, + { + "epoch": 2.1730118443316413, + "grad_norm": 8.1875, + "learning_rate": 3.1436972199157123e-06, + "loss": 2.4776, + "step": 20548 + }, + { + "epoch": 2.173117597292724, + "grad_norm": 8.25, + "learning_rate": 3.1435317362705165e-06, + "loss": 2.5312, + "step": 20549 + }, + { + "epoch": 2.173223350253807, + "grad_norm": 8.0625, + "learning_rate": 3.14336624960549e-06, + "loss": 2.5311, + "step": 20550 + }, + { + "epoch": 2.17332910321489, + "grad_norm": 8.875, + "learning_rate": 3.1432007599214077e-06, + "loss": 2.5071, + "step": 20551 + }, + { + "epoch": 2.173434856175973, + "grad_norm": 8.5625, + "learning_rate": 3.1430352672190486e-06, + "loss": 2.3392, + "step": 20552 + }, + { + "epoch": 2.173540609137056, + "grad_norm": 9.0625, + "learning_rate": 3.142869771499187e-06, + "loss": 2.6862, + "step": 20553 + }, + { + "epoch": 2.1736463620981388, + "grad_norm": 8.375, + "learning_rate": 3.1427042727626005e-06, + "loss": 2.7517, + "step": 20554 + }, + { + "epoch": 2.1737521150592216, + "grad_norm": 8.25, + "learning_rate": 3.1425387710100663e-06, + "loss": 2.5852, + "step": 20555 + }, + { + "epoch": 2.1738578680203045, + "grad_norm": 8.125, + "learning_rate": 3.1423732662423594e-06, + "loss": 2.5304, + "step": 20556 + }, + { + "epoch": 2.1739636209813873, + "grad_norm": 8.75, + "learning_rate": 3.1422077584602584e-06, + "loss": 2.451, + "step": 20557 + }, + { + "epoch": 2.1740693739424706, + "grad_norm": 8.3125, + "learning_rate": 3.1420422476645386e-06, + "loss": 2.4588, + "step": 20558 + }, + { + "epoch": 2.1741751269035534, + "grad_norm": 8.875, + "learning_rate": 3.1418767338559773e-06, + "loss": 2.6326, + "step": 20559 + }, + { + "epoch": 2.1742808798646363, + "grad_norm": 8.6875, + "learning_rate": 3.1417112170353516e-06, + "loss": 2.6072, + "step": 20560 + }, + { + "epoch": 2.174386632825719, + "grad_norm": 8.3125, + "learning_rate": 3.1415456972034365e-06, + "loss": 2.5507, + "step": 20561 + }, + { + "epoch": 2.174492385786802, + "grad_norm": 8.3125, + "learning_rate": 3.1413801743610105e-06, + "loss": 2.4067, + "step": 20562 + }, + { + "epoch": 2.1745981387478848, + "grad_norm": 8.0, + "learning_rate": 3.1412146485088503e-06, + "loss": 2.4523, + "step": 20563 + }, + { + "epoch": 2.174703891708968, + "grad_norm": 8.625, + "learning_rate": 3.14104911964773e-06, + "loss": 2.4284, + "step": 20564 + }, + { + "epoch": 2.174809644670051, + "grad_norm": 8.875, + "learning_rate": 3.1408835877784298e-06, + "loss": 2.4767, + "step": 20565 + }, + { + "epoch": 2.1749153976311337, + "grad_norm": 8.375, + "learning_rate": 3.140718052901725e-06, + "loss": 2.452, + "step": 20566 + }, + { + "epoch": 2.1750211505922166, + "grad_norm": 8.375, + "learning_rate": 3.140552515018392e-06, + "loss": 2.6747, + "step": 20567 + }, + { + "epoch": 2.1751269035532994, + "grad_norm": 8.3125, + "learning_rate": 3.140386974129208e-06, + "loss": 2.5723, + "step": 20568 + }, + { + "epoch": 2.1752326565143822, + "grad_norm": 8.375, + "learning_rate": 3.140221430234951e-06, + "loss": 2.5441, + "step": 20569 + }, + { + "epoch": 2.175338409475465, + "grad_norm": 7.75, + "learning_rate": 3.1400558833363953e-06, + "loss": 2.419, + "step": 20570 + }, + { + "epoch": 2.1754441624365484, + "grad_norm": 8.5625, + "learning_rate": 3.13989033343432e-06, + "loss": 2.5461, + "step": 20571 + }, + { + "epoch": 2.175549915397631, + "grad_norm": 8.3125, + "learning_rate": 3.1397247805295005e-06, + "loss": 2.4812, + "step": 20572 + }, + { + "epoch": 2.175655668358714, + "grad_norm": 8.4375, + "learning_rate": 3.139559224622714e-06, + "loss": 2.534, + "step": 20573 + }, + { + "epoch": 2.175761421319797, + "grad_norm": 8.0, + "learning_rate": 3.139393665714738e-06, + "loss": 2.4905, + "step": 20574 + }, + { + "epoch": 2.1758671742808797, + "grad_norm": 8.25, + "learning_rate": 3.1392281038063486e-06, + "loss": 2.36, + "step": 20575 + }, + { + "epoch": 2.175972927241963, + "grad_norm": 8.625, + "learning_rate": 3.139062538898324e-06, + "loss": 2.7758, + "step": 20576 + }, + { + "epoch": 2.176078680203046, + "grad_norm": 8.3125, + "learning_rate": 3.1388969709914397e-06, + "loss": 2.6209, + "step": 20577 + }, + { + "epoch": 2.1761844331641287, + "grad_norm": 8.25, + "learning_rate": 3.1387314000864737e-06, + "loss": 2.4854, + "step": 20578 + }, + { + "epoch": 2.1762901861252115, + "grad_norm": 8.0625, + "learning_rate": 3.1385658261842024e-06, + "loss": 2.385, + "step": 20579 + }, + { + "epoch": 2.1763959390862944, + "grad_norm": 8.1875, + "learning_rate": 3.1384002492854025e-06, + "loss": 2.4886, + "step": 20580 + }, + { + "epoch": 2.176501692047377, + "grad_norm": 8.375, + "learning_rate": 3.1382346693908517e-06, + "loss": 2.5187, + "step": 20581 + }, + { + "epoch": 2.17660744500846, + "grad_norm": 8.0625, + "learning_rate": 3.1380690865013257e-06, + "loss": 2.3004, + "step": 20582 + }, + { + "epoch": 2.1767131979695433, + "grad_norm": 8.25, + "learning_rate": 3.1379035006176027e-06, + "loss": 2.512, + "step": 20583 + }, + { + "epoch": 2.176818950930626, + "grad_norm": 8.75, + "learning_rate": 3.1377379117404606e-06, + "loss": 2.7714, + "step": 20584 + }, + { + "epoch": 2.176924703891709, + "grad_norm": 8.3125, + "learning_rate": 3.1375723198706744e-06, + "loss": 2.682, + "step": 20585 + }, + { + "epoch": 2.177030456852792, + "grad_norm": 8.4375, + "learning_rate": 3.137406725009022e-06, + "loss": 2.5562, + "step": 20586 + }, + { + "epoch": 2.1771362098138747, + "grad_norm": 8.6875, + "learning_rate": 3.137241127156281e-06, + "loss": 2.3758, + "step": 20587 + }, + { + "epoch": 2.1772419627749575, + "grad_norm": 8.875, + "learning_rate": 3.137075526313228e-06, + "loss": 2.6965, + "step": 20588 + }, + { + "epoch": 2.177347715736041, + "grad_norm": 8.4375, + "learning_rate": 3.1369099224806397e-06, + "loss": 2.4301, + "step": 20589 + }, + { + "epoch": 2.1774534686971236, + "grad_norm": 8.4375, + "learning_rate": 3.1367443156592935e-06, + "loss": 2.6926, + "step": 20590 + }, + { + "epoch": 2.1775592216582065, + "grad_norm": 8.0625, + "learning_rate": 3.136578705849967e-06, + "loss": 2.3725, + "step": 20591 + }, + { + "epoch": 2.1776649746192893, + "grad_norm": 8.125, + "learning_rate": 3.136413093053437e-06, + "loss": 2.5461, + "step": 20592 + }, + { + "epoch": 2.177770727580372, + "grad_norm": 8.3125, + "learning_rate": 3.1362474772704804e-06, + "loss": 2.4759, + "step": 20593 + }, + { + "epoch": 2.177876480541455, + "grad_norm": 8.1875, + "learning_rate": 3.136081858501875e-06, + "loss": 2.4805, + "step": 20594 + }, + { + "epoch": 2.1779822335025383, + "grad_norm": 8.9375, + "learning_rate": 3.1359162367483976e-06, + "loss": 2.5634, + "step": 20595 + }, + { + "epoch": 2.178087986463621, + "grad_norm": 8.3125, + "learning_rate": 3.135750612010825e-06, + "loss": 2.502, + "step": 20596 + }, + { + "epoch": 2.178193739424704, + "grad_norm": 8.4375, + "learning_rate": 3.135584984289935e-06, + "loss": 2.7291, + "step": 20597 + }, + { + "epoch": 2.178299492385787, + "grad_norm": 7.96875, + "learning_rate": 3.1354193535865045e-06, + "loss": 2.3984, + "step": 20598 + }, + { + "epoch": 2.1784052453468696, + "grad_norm": 8.125, + "learning_rate": 3.135253719901311e-06, + "loss": 2.0908, + "step": 20599 + }, + { + "epoch": 2.1785109983079525, + "grad_norm": 7.96875, + "learning_rate": 3.135088083235131e-06, + "loss": 2.3589, + "step": 20600 + }, + { + "epoch": 2.1786167512690353, + "grad_norm": 8.3125, + "learning_rate": 3.134922443588743e-06, + "loss": 2.663, + "step": 20601 + }, + { + "epoch": 2.1787225042301186, + "grad_norm": 8.5, + "learning_rate": 3.134756800962923e-06, + "loss": 2.6312, + "step": 20602 + }, + { + "epoch": 2.1788282571912014, + "grad_norm": 8.3125, + "learning_rate": 3.1345911553584498e-06, + "loss": 2.5851, + "step": 20603 + }, + { + "epoch": 2.1789340101522843, + "grad_norm": 8.1875, + "learning_rate": 3.1344255067761e-06, + "loss": 2.4048, + "step": 20604 + }, + { + "epoch": 2.179039763113367, + "grad_norm": 8.5, + "learning_rate": 3.13425985521665e-06, + "loss": 2.3968, + "step": 20605 + }, + { + "epoch": 2.17914551607445, + "grad_norm": 8.3125, + "learning_rate": 3.1340942006808785e-06, + "loss": 2.4702, + "step": 20606 + }, + { + "epoch": 2.179251269035533, + "grad_norm": 8.375, + "learning_rate": 3.133928543169562e-06, + "loss": 2.5019, + "step": 20607 + }, + { + "epoch": 2.179357021996616, + "grad_norm": 8.5625, + "learning_rate": 3.1337628826834774e-06, + "loss": 2.5797, + "step": 20608 + }, + { + "epoch": 2.179462774957699, + "grad_norm": 8.6875, + "learning_rate": 3.1335972192234034e-06, + "loss": 2.6687, + "step": 20609 + }, + { + "epoch": 2.1795685279187818, + "grad_norm": 8.1875, + "learning_rate": 3.133431552790117e-06, + "loss": 2.467, + "step": 20610 + }, + { + "epoch": 2.1796742808798646, + "grad_norm": 8.125, + "learning_rate": 3.1332658833843956e-06, + "loss": 2.5558, + "step": 20611 + }, + { + "epoch": 2.1797800338409474, + "grad_norm": 8.25, + "learning_rate": 3.1331002110070153e-06, + "loss": 2.44, + "step": 20612 + }, + { + "epoch": 2.1798857868020303, + "grad_norm": 8.5625, + "learning_rate": 3.132934535658756e-06, + "loss": 2.6189, + "step": 20613 + }, + { + "epoch": 2.1799915397631136, + "grad_norm": 8.4375, + "learning_rate": 3.132768857340393e-06, + "loss": 2.7419, + "step": 20614 + }, + { + "epoch": 2.1800972927241964, + "grad_norm": 8.375, + "learning_rate": 3.132603176052704e-06, + "loss": 2.9245, + "step": 20615 + }, + { + "epoch": 2.1802030456852792, + "grad_norm": 8.3125, + "learning_rate": 3.1324374917964685e-06, + "loss": 2.6305, + "step": 20616 + }, + { + "epoch": 2.180308798646362, + "grad_norm": 8.25, + "learning_rate": 3.132271804572461e-06, + "loss": 2.4077, + "step": 20617 + }, + { + "epoch": 2.180414551607445, + "grad_norm": 8.5625, + "learning_rate": 3.1321061143814613e-06, + "loss": 2.6508, + "step": 20618 + }, + { + "epoch": 2.1805203045685277, + "grad_norm": 8.8125, + "learning_rate": 3.131940421224246e-06, + "loss": 2.8308, + "step": 20619 + }, + { + "epoch": 2.180626057529611, + "grad_norm": 9.1875, + "learning_rate": 3.131774725101593e-06, + "loss": 2.7575, + "step": 20620 + }, + { + "epoch": 2.180731810490694, + "grad_norm": 8.125, + "learning_rate": 3.1316090260142794e-06, + "loss": 2.5802, + "step": 20621 + }, + { + "epoch": 2.1808375634517767, + "grad_norm": 8.625, + "learning_rate": 3.131443323963083e-06, + "loss": 2.3971, + "step": 20622 + }, + { + "epoch": 2.1809433164128595, + "grad_norm": 8.875, + "learning_rate": 3.1312776189487814e-06, + "loss": 2.8, + "step": 20623 + }, + { + "epoch": 2.1810490693739424, + "grad_norm": 8.6875, + "learning_rate": 3.1311119109721518e-06, + "loss": 2.5816, + "step": 20624 + }, + { + "epoch": 2.1811548223350252, + "grad_norm": 8.1875, + "learning_rate": 3.1309462000339726e-06, + "loss": 2.3912, + "step": 20625 + }, + { + "epoch": 2.1812605752961085, + "grad_norm": 8.5625, + "learning_rate": 3.1307804861350203e-06, + "loss": 2.484, + "step": 20626 + }, + { + "epoch": 2.1813663282571913, + "grad_norm": 8.125, + "learning_rate": 3.1306147692760737e-06, + "loss": 2.5463, + "step": 20627 + }, + { + "epoch": 2.181472081218274, + "grad_norm": 8.0, + "learning_rate": 3.1304490494579095e-06, + "loss": 2.4105, + "step": 20628 + }, + { + "epoch": 2.181577834179357, + "grad_norm": 8.375, + "learning_rate": 3.130283326681306e-06, + "loss": 2.4235, + "step": 20629 + }, + { + "epoch": 2.18168358714044, + "grad_norm": 8.4375, + "learning_rate": 3.1301176009470415e-06, + "loss": 2.6643, + "step": 20630 + }, + { + "epoch": 2.1817893401015227, + "grad_norm": 8.4375, + "learning_rate": 3.1299518722558908e-06, + "loss": 2.5806, + "step": 20631 + }, + { + "epoch": 2.1818950930626055, + "grad_norm": 8.0625, + "learning_rate": 3.129786140608635e-06, + "loss": 2.4356, + "step": 20632 + }, + { + "epoch": 2.182000846023689, + "grad_norm": 8.125, + "learning_rate": 3.129620406006051e-06, + "loss": 2.4041, + "step": 20633 + }, + { + "epoch": 2.1821065989847717, + "grad_norm": 9.0625, + "learning_rate": 3.1294546684489145e-06, + "loss": 2.7309, + "step": 20634 + }, + { + "epoch": 2.1822123519458545, + "grad_norm": 8.375, + "learning_rate": 3.1292889279380047e-06, + "loss": 2.4759, + "step": 20635 + }, + { + "epoch": 2.1823181049069373, + "grad_norm": 8.125, + "learning_rate": 3.1291231844741e-06, + "loss": 2.6439, + "step": 20636 + }, + { + "epoch": 2.18242385786802, + "grad_norm": 8.75, + "learning_rate": 3.1289574380579773e-06, + "loss": 2.6305, + "step": 20637 + }, + { + "epoch": 2.182529610829103, + "grad_norm": 8.5625, + "learning_rate": 3.128791688690415e-06, + "loss": 2.6946, + "step": 20638 + }, + { + "epoch": 2.1826353637901863, + "grad_norm": 8.375, + "learning_rate": 3.1286259363721906e-06, + "loss": 2.5848, + "step": 20639 + }, + { + "epoch": 2.182741116751269, + "grad_norm": 7.875, + "learning_rate": 3.1284601811040805e-06, + "loss": 2.3558, + "step": 20640 + }, + { + "epoch": 2.182846869712352, + "grad_norm": 8.0625, + "learning_rate": 3.128294422886865e-06, + "loss": 2.5878, + "step": 20641 + }, + { + "epoch": 2.182952622673435, + "grad_norm": 8.3125, + "learning_rate": 3.1281286617213204e-06, + "loss": 2.4139, + "step": 20642 + }, + { + "epoch": 2.1830583756345177, + "grad_norm": 8.375, + "learning_rate": 3.1279628976082242e-06, + "loss": 2.6348, + "step": 20643 + }, + { + "epoch": 2.1831641285956005, + "grad_norm": 8.9375, + "learning_rate": 3.1277971305483555e-06, + "loss": 2.4531, + "step": 20644 + }, + { + "epoch": 2.183269881556684, + "grad_norm": 8.875, + "learning_rate": 3.1276313605424926e-06, + "loss": 2.4229, + "step": 20645 + }, + { + "epoch": 2.1833756345177666, + "grad_norm": 8.4375, + "learning_rate": 3.127465587591411e-06, + "loss": 2.7345, + "step": 20646 + }, + { + "epoch": 2.1834813874788495, + "grad_norm": 9.125, + "learning_rate": 3.127299811695891e-06, + "loss": 2.2626, + "step": 20647 + }, + { + "epoch": 2.1835871404399323, + "grad_norm": 8.625, + "learning_rate": 3.127134032856709e-06, + "loss": 2.6325, + "step": 20648 + }, + { + "epoch": 2.183692893401015, + "grad_norm": 8.625, + "learning_rate": 3.1269682510746435e-06, + "loss": 2.6655, + "step": 20649 + }, + { + "epoch": 2.183798646362098, + "grad_norm": 8.375, + "learning_rate": 3.1268024663504727e-06, + "loss": 2.6045, + "step": 20650 + }, + { + "epoch": 2.1839043993231813, + "grad_norm": 8.75, + "learning_rate": 3.126636678684974e-06, + "loss": 2.7386, + "step": 20651 + }, + { + "epoch": 2.184010152284264, + "grad_norm": 8.875, + "learning_rate": 3.1264708880789256e-06, + "loss": 2.7113, + "step": 20652 + }, + { + "epoch": 2.184115905245347, + "grad_norm": 8.0, + "learning_rate": 3.1263050945331053e-06, + "loss": 2.5232, + "step": 20653 + }, + { + "epoch": 2.1842216582064298, + "grad_norm": 7.9375, + "learning_rate": 3.126139298048292e-06, + "loss": 2.5912, + "step": 20654 + }, + { + "epoch": 2.1843274111675126, + "grad_norm": 8.625, + "learning_rate": 3.125973498625263e-06, + "loss": 2.4154, + "step": 20655 + }, + { + "epoch": 2.1844331641285955, + "grad_norm": 8.3125, + "learning_rate": 3.1258076962647954e-06, + "loss": 2.6087, + "step": 20656 + }, + { + "epoch": 2.1845389170896787, + "grad_norm": 8.5, + "learning_rate": 3.125641890967669e-06, + "loss": 2.502, + "step": 20657 + }, + { + "epoch": 2.1846446700507616, + "grad_norm": 8.5, + "learning_rate": 3.1254760827346616e-06, + "loss": 2.6505, + "step": 20658 + }, + { + "epoch": 2.1847504230118444, + "grad_norm": 8.5, + "learning_rate": 3.1253102715665495e-06, + "loss": 2.6511, + "step": 20659 + }, + { + "epoch": 2.1848561759729273, + "grad_norm": 8.25, + "learning_rate": 3.1251444574641132e-06, + "loss": 2.5541, + "step": 20660 + }, + { + "epoch": 2.18496192893401, + "grad_norm": 7.96875, + "learning_rate": 3.1249786404281296e-06, + "loss": 2.5976, + "step": 20661 + }, + { + "epoch": 2.185067681895093, + "grad_norm": 8.5, + "learning_rate": 3.1248128204593757e-06, + "loss": 2.5398, + "step": 20662 + }, + { + "epoch": 2.1851734348561758, + "grad_norm": 8.3125, + "learning_rate": 3.124646997558631e-06, + "loss": 2.4247, + "step": 20663 + }, + { + "epoch": 2.185279187817259, + "grad_norm": 8.8125, + "learning_rate": 3.124481171726675e-06, + "loss": 2.6458, + "step": 20664 + }, + { + "epoch": 2.185384940778342, + "grad_norm": 8.1875, + "learning_rate": 3.124315342964283e-06, + "loss": 2.4356, + "step": 20665 + }, + { + "epoch": 2.1854906937394247, + "grad_norm": 8.8125, + "learning_rate": 3.1241495112722343e-06, + "loss": 2.7644, + "step": 20666 + }, + { + "epoch": 2.1855964467005076, + "grad_norm": 8.0625, + "learning_rate": 3.1239836766513075e-06, + "loss": 2.5601, + "step": 20667 + }, + { + "epoch": 2.1857021996615904, + "grad_norm": 9.0625, + "learning_rate": 3.123817839102281e-06, + "loss": 2.5384, + "step": 20668 + }, + { + "epoch": 2.1858079526226732, + "grad_norm": 8.0625, + "learning_rate": 3.1236519986259315e-06, + "loss": 2.5731, + "step": 20669 + }, + { + "epoch": 2.1859137055837565, + "grad_norm": 7.875, + "learning_rate": 3.1234861552230396e-06, + "loss": 2.3171, + "step": 20670 + }, + { + "epoch": 2.1860194585448394, + "grad_norm": 8.75, + "learning_rate": 3.123320308894381e-06, + "loss": 2.6783, + "step": 20671 + }, + { + "epoch": 2.186125211505922, + "grad_norm": 8.25, + "learning_rate": 3.1231544596407348e-06, + "loss": 2.4533, + "step": 20672 + }, + { + "epoch": 2.186230964467005, + "grad_norm": 9.3125, + "learning_rate": 3.1229886074628805e-06, + "loss": 2.7269, + "step": 20673 + }, + { + "epoch": 2.186336717428088, + "grad_norm": 8.625, + "learning_rate": 3.1228227523615958e-06, + "loss": 2.4595, + "step": 20674 + }, + { + "epoch": 2.1864424703891707, + "grad_norm": 8.3125, + "learning_rate": 3.1226568943376573e-06, + "loss": 2.4855, + "step": 20675 + }, + { + "epoch": 2.186548223350254, + "grad_norm": 7.90625, + "learning_rate": 3.1224910333918456e-06, + "loss": 2.4308, + "step": 20676 + }, + { + "epoch": 2.186653976311337, + "grad_norm": 8.0625, + "learning_rate": 3.1223251695249384e-06, + "loss": 2.3456, + "step": 20677 + }, + { + "epoch": 2.1867597292724197, + "grad_norm": 8.5, + "learning_rate": 3.1221593027377127e-06, + "loss": 2.5711, + "step": 20678 + }, + { + "epoch": 2.1868654822335025, + "grad_norm": 8.3125, + "learning_rate": 3.121993433030948e-06, + "loss": 2.5327, + "step": 20679 + }, + { + "epoch": 2.1869712351945854, + "grad_norm": 8.5, + "learning_rate": 3.1218275604054234e-06, + "loss": 2.508, + "step": 20680 + }, + { + "epoch": 2.187076988155668, + "grad_norm": 8.0625, + "learning_rate": 3.1216616848619158e-06, + "loss": 2.2797, + "step": 20681 + }, + { + "epoch": 2.187182741116751, + "grad_norm": 8.25, + "learning_rate": 3.1214958064012034e-06, + "loss": 2.4055, + "step": 20682 + }, + { + "epoch": 2.1872884940778343, + "grad_norm": 8.875, + "learning_rate": 3.121329925024067e-06, + "loss": 2.6455, + "step": 20683 + }, + { + "epoch": 2.187394247038917, + "grad_norm": 8.25, + "learning_rate": 3.121164040731283e-06, + "loss": 2.6727, + "step": 20684 + }, + { + "epoch": 2.1875, + "grad_norm": 8.25, + "learning_rate": 3.120998153523629e-06, + "loss": 2.5229, + "step": 20685 + }, + { + "epoch": 2.187605752961083, + "grad_norm": 8.3125, + "learning_rate": 3.1208322634018855e-06, + "loss": 2.5096, + "step": 20686 + }, + { + "epoch": 2.1877115059221657, + "grad_norm": 8.0, + "learning_rate": 3.1206663703668307e-06, + "loss": 2.3861, + "step": 20687 + }, + { + "epoch": 2.187817258883249, + "grad_norm": 8.5, + "learning_rate": 3.12050047441924e-06, + "loss": 2.5495, + "step": 20688 + }, + { + "epoch": 2.187923011844332, + "grad_norm": 7.9375, + "learning_rate": 3.120334575559897e-06, + "loss": 2.4717, + "step": 20689 + }, + { + "epoch": 2.1880287648054146, + "grad_norm": 8.5625, + "learning_rate": 3.1201686737895766e-06, + "loss": 2.5387, + "step": 20690 + }, + { + "epoch": 2.1881345177664975, + "grad_norm": 8.3125, + "learning_rate": 3.120002769109058e-06, + "loss": 2.5632, + "step": 20691 + }, + { + "epoch": 2.1882402707275803, + "grad_norm": 8.625, + "learning_rate": 3.1198368615191206e-06, + "loss": 2.786, + "step": 20692 + }, + { + "epoch": 2.188346023688663, + "grad_norm": 8.5625, + "learning_rate": 3.119670951020542e-06, + "loss": 2.6931, + "step": 20693 + }, + { + "epoch": 2.188451776649746, + "grad_norm": 9.9375, + "learning_rate": 3.1195050376141005e-06, + "loss": 2.8361, + "step": 20694 + }, + { + "epoch": 2.1885575296108293, + "grad_norm": 8.3125, + "learning_rate": 3.1193391213005763e-06, + "loss": 2.3566, + "step": 20695 + }, + { + "epoch": 2.188663282571912, + "grad_norm": 9.1875, + "learning_rate": 3.1191732020807465e-06, + "loss": 2.5321, + "step": 20696 + }, + { + "epoch": 2.188769035532995, + "grad_norm": 8.375, + "learning_rate": 3.1190072799553896e-06, + "loss": 2.6066, + "step": 20697 + }, + { + "epoch": 2.188874788494078, + "grad_norm": 7.8125, + "learning_rate": 3.118841354925285e-06, + "loss": 2.4502, + "step": 20698 + }, + { + "epoch": 2.1889805414551606, + "grad_norm": 8.8125, + "learning_rate": 3.118675426991211e-06, + "loss": 2.5803, + "step": 20699 + }, + { + "epoch": 2.1890862944162435, + "grad_norm": 8.5, + "learning_rate": 3.118509496153947e-06, + "loss": 2.5798, + "step": 20700 + }, + { + "epoch": 2.1891920473773268, + "grad_norm": 8.0625, + "learning_rate": 3.1183435624142695e-06, + "loss": 2.3929, + "step": 20701 + }, + { + "epoch": 2.1892978003384096, + "grad_norm": 12.625, + "learning_rate": 3.1181776257729596e-06, + "loss": 2.4791, + "step": 20702 + }, + { + "epoch": 2.1894035532994924, + "grad_norm": 8.5625, + "learning_rate": 3.1180116862307948e-06, + "loss": 2.4082, + "step": 20703 + }, + { + "epoch": 2.1895093062605753, + "grad_norm": 8.1875, + "learning_rate": 3.1178457437885534e-06, + "loss": 2.6605, + "step": 20704 + }, + { + "epoch": 2.189615059221658, + "grad_norm": 8.9375, + "learning_rate": 3.1176797984470153e-06, + "loss": 2.6309, + "step": 20705 + }, + { + "epoch": 2.189720812182741, + "grad_norm": 8.0625, + "learning_rate": 3.117513850206958e-06, + "loss": 2.4446, + "step": 20706 + }, + { + "epoch": 2.1898265651438242, + "grad_norm": 7.90625, + "learning_rate": 3.1173478990691602e-06, + "loss": 2.5536, + "step": 20707 + }, + { + "epoch": 2.189932318104907, + "grad_norm": 8.125, + "learning_rate": 3.1171819450344023e-06, + "loss": 2.4621, + "step": 20708 + }, + { + "epoch": 2.19003807106599, + "grad_norm": 8.6875, + "learning_rate": 3.1170159881034616e-06, + "loss": 2.5817, + "step": 20709 + }, + { + "epoch": 2.1901438240270727, + "grad_norm": 8.0, + "learning_rate": 3.1168500282771176e-06, + "loss": 2.5169, + "step": 20710 + }, + { + "epoch": 2.1902495769881556, + "grad_norm": 8.5625, + "learning_rate": 3.116684065556148e-06, + "loss": 2.7456, + "step": 20711 + }, + { + "epoch": 2.1903553299492384, + "grad_norm": 8.5625, + "learning_rate": 3.116518099941333e-06, + "loss": 2.5469, + "step": 20712 + }, + { + "epoch": 2.1904610829103213, + "grad_norm": 9.0, + "learning_rate": 3.11635213143345e-06, + "loss": 2.5274, + "step": 20713 + }, + { + "epoch": 2.1905668358714045, + "grad_norm": 8.3125, + "learning_rate": 3.1161861600332788e-06, + "loss": 2.637, + "step": 20714 + }, + { + "epoch": 2.1906725888324874, + "grad_norm": 8.8125, + "learning_rate": 3.116020185741598e-06, + "loss": 2.519, + "step": 20715 + }, + { + "epoch": 2.1907783417935702, + "grad_norm": 8.125, + "learning_rate": 3.1158542085591866e-06, + "loss": 2.4296, + "step": 20716 + }, + { + "epoch": 2.190884094754653, + "grad_norm": 8.375, + "learning_rate": 3.115688228486823e-06, + "loss": 2.1477, + "step": 20717 + }, + { + "epoch": 2.190989847715736, + "grad_norm": 8.0625, + "learning_rate": 3.1155222455252864e-06, + "loss": 2.333, + "step": 20718 + }, + { + "epoch": 2.1910956006768187, + "grad_norm": 8.5625, + "learning_rate": 3.1153562596753556e-06, + "loss": 2.6753, + "step": 20719 + }, + { + "epoch": 2.191201353637902, + "grad_norm": 7.96875, + "learning_rate": 3.1151902709378094e-06, + "loss": 2.475, + "step": 20720 + }, + { + "epoch": 2.191307106598985, + "grad_norm": 8.4375, + "learning_rate": 3.115024279313427e-06, + "loss": 2.7535, + "step": 20721 + }, + { + "epoch": 2.1914128595600677, + "grad_norm": 8.3125, + "learning_rate": 3.114858284802988e-06, + "loss": 2.5243, + "step": 20722 + }, + { + "epoch": 2.1915186125211505, + "grad_norm": 8.6875, + "learning_rate": 3.114692287407269e-06, + "loss": 2.4771, + "step": 20723 + }, + { + "epoch": 2.1916243654822334, + "grad_norm": 8.3125, + "learning_rate": 3.1145262871270517e-06, + "loss": 2.6452, + "step": 20724 + }, + { + "epoch": 2.191730118443316, + "grad_norm": 8.875, + "learning_rate": 3.114360283963113e-06, + "loss": 2.0402, + "step": 20725 + }, + { + "epoch": 2.1918358714043995, + "grad_norm": 8.5625, + "learning_rate": 3.114194277916233e-06, + "loss": 2.6678, + "step": 20726 + }, + { + "epoch": 2.1919416243654823, + "grad_norm": 8.9375, + "learning_rate": 3.114028268987191e-06, + "loss": 2.3471, + "step": 20727 + }, + { + "epoch": 2.192047377326565, + "grad_norm": 8.375, + "learning_rate": 3.1138622571767653e-06, + "loss": 2.7264, + "step": 20728 + }, + { + "epoch": 2.192153130287648, + "grad_norm": 9.0625, + "learning_rate": 3.113696242485735e-06, + "loss": 2.6749, + "step": 20729 + }, + { + "epoch": 2.192258883248731, + "grad_norm": 8.5, + "learning_rate": 3.1135302249148793e-06, + "loss": 2.5841, + "step": 20730 + }, + { + "epoch": 2.1923646362098137, + "grad_norm": 9.0625, + "learning_rate": 3.113364204464977e-06, + "loss": 2.5354, + "step": 20731 + }, + { + "epoch": 2.192470389170897, + "grad_norm": 8.875, + "learning_rate": 3.1131981811368074e-06, + "loss": 2.5494, + "step": 20732 + }, + { + "epoch": 2.19257614213198, + "grad_norm": 8.625, + "learning_rate": 3.113032154931149e-06, + "loss": 2.5412, + "step": 20733 + }, + { + "epoch": 2.1926818950930627, + "grad_norm": 8.0, + "learning_rate": 3.1128661258487818e-06, + "loss": 2.4411, + "step": 20734 + }, + { + "epoch": 2.1927876480541455, + "grad_norm": 8.75, + "learning_rate": 3.112700093890485e-06, + "loss": 2.5947, + "step": 20735 + }, + { + "epoch": 2.1928934010152283, + "grad_norm": 9.125, + "learning_rate": 3.112534059057037e-06, + "loss": 2.7141, + "step": 20736 + }, + { + "epoch": 2.192999153976311, + "grad_norm": 7.84375, + "learning_rate": 3.112368021349217e-06, + "loss": 2.4801, + "step": 20737 + }, + { + "epoch": 2.1931049069373945, + "grad_norm": 8.25, + "learning_rate": 3.1122019807678045e-06, + "loss": 2.4842, + "step": 20738 + }, + { + "epoch": 2.1932106598984773, + "grad_norm": 8.25, + "learning_rate": 3.112035937313578e-06, + "loss": 2.6322, + "step": 20739 + }, + { + "epoch": 2.19331641285956, + "grad_norm": 8.4375, + "learning_rate": 3.1118698909873178e-06, + "loss": 2.708, + "step": 20740 + }, + { + "epoch": 2.193422165820643, + "grad_norm": 9.0, + "learning_rate": 3.111703841789802e-06, + "loss": 2.6177, + "step": 20741 + }, + { + "epoch": 2.193527918781726, + "grad_norm": 8.5, + "learning_rate": 3.11153778972181e-06, + "loss": 2.4038, + "step": 20742 + }, + { + "epoch": 2.1936336717428087, + "grad_norm": 8.125, + "learning_rate": 3.1113717347841223e-06, + "loss": 2.4097, + "step": 20743 + }, + { + "epoch": 2.1937394247038915, + "grad_norm": 8.25, + "learning_rate": 3.111205676977517e-06, + "loss": 2.4963, + "step": 20744 + }, + { + "epoch": 2.1938451776649748, + "grad_norm": 9.5625, + "learning_rate": 3.1110396163027724e-06, + "loss": 2.8581, + "step": 20745 + }, + { + "epoch": 2.1939509306260576, + "grad_norm": 8.9375, + "learning_rate": 3.1108735527606694e-06, + "loss": 2.5017, + "step": 20746 + }, + { + "epoch": 2.1940566835871405, + "grad_norm": 8.375, + "learning_rate": 3.1107074863519866e-06, + "loss": 2.4695, + "step": 20747 + }, + { + "epoch": 2.1941624365482233, + "grad_norm": 8.6875, + "learning_rate": 3.110541417077503e-06, + "loss": 2.2657, + "step": 20748 + }, + { + "epoch": 2.194268189509306, + "grad_norm": 7.90625, + "learning_rate": 3.1103753449379986e-06, + "loss": 2.3714, + "step": 20749 + }, + { + "epoch": 2.194373942470389, + "grad_norm": 8.5625, + "learning_rate": 3.1102092699342523e-06, + "loss": 2.4718, + "step": 20750 + }, + { + "epoch": 2.1944796954314723, + "grad_norm": 8.5, + "learning_rate": 3.1100431920670438e-06, + "loss": 2.6074, + "step": 20751 + }, + { + "epoch": 2.194585448392555, + "grad_norm": 8.3125, + "learning_rate": 3.109877111337151e-06, + "loss": 2.6269, + "step": 20752 + }, + { + "epoch": 2.194691201353638, + "grad_norm": 8.1875, + "learning_rate": 3.109711027745356e-06, + "loss": 2.5112, + "step": 20753 + }, + { + "epoch": 2.1947969543147208, + "grad_norm": 8.5625, + "learning_rate": 3.109544941292435e-06, + "loss": 2.6452, + "step": 20754 + }, + { + "epoch": 2.1949027072758036, + "grad_norm": 8.25, + "learning_rate": 3.1093788519791695e-06, + "loss": 2.6158, + "step": 20755 + }, + { + "epoch": 2.1950084602368864, + "grad_norm": 7.96875, + "learning_rate": 3.109212759806338e-06, + "loss": 2.5156, + "step": 20756 + }, + { + "epoch": 2.1951142131979697, + "grad_norm": 8.75, + "learning_rate": 3.109046664774721e-06, + "loss": 2.5135, + "step": 20757 + }, + { + "epoch": 2.1952199661590526, + "grad_norm": 9.3125, + "learning_rate": 3.1088805668850965e-06, + "loss": 2.5869, + "step": 20758 + }, + { + "epoch": 2.1953257191201354, + "grad_norm": 7.90625, + "learning_rate": 3.108714466138244e-06, + "loss": 2.4948, + "step": 20759 + }, + { + "epoch": 2.1954314720812182, + "grad_norm": 8.875, + "learning_rate": 3.1085483625349444e-06, + "loss": 2.5109, + "step": 20760 + }, + { + "epoch": 2.195537225042301, + "grad_norm": 8.25, + "learning_rate": 3.1083822560759752e-06, + "loss": 2.5682, + "step": 20761 + }, + { + "epoch": 2.195642978003384, + "grad_norm": 8.4375, + "learning_rate": 3.108216146762118e-06, + "loss": 2.4946, + "step": 20762 + }, + { + "epoch": 2.195748730964467, + "grad_norm": 8.375, + "learning_rate": 3.108050034594151e-06, + "loss": 2.5623, + "step": 20763 + }, + { + "epoch": 2.19585448392555, + "grad_norm": 7.9375, + "learning_rate": 3.107883919572853e-06, + "loss": 2.5182, + "step": 20764 + }, + { + "epoch": 2.195960236886633, + "grad_norm": 8.5625, + "learning_rate": 3.107717801699005e-06, + "loss": 2.6713, + "step": 20765 + }, + { + "epoch": 2.1960659898477157, + "grad_norm": 9.5, + "learning_rate": 3.1075516809733863e-06, + "loss": 2.4204, + "step": 20766 + }, + { + "epoch": 2.1961717428087986, + "grad_norm": 8.375, + "learning_rate": 3.1073855573967753e-06, + "loss": 2.525, + "step": 20767 + }, + { + "epoch": 2.1962774957698814, + "grad_norm": 8.125, + "learning_rate": 3.1072194309699523e-06, + "loss": 2.5294, + "step": 20768 + }, + { + "epoch": 2.1963832487309647, + "grad_norm": 8.3125, + "learning_rate": 3.1070533016936977e-06, + "loss": 2.6803, + "step": 20769 + }, + { + "epoch": 2.1964890016920475, + "grad_norm": 8.9375, + "learning_rate": 3.1068871695687896e-06, + "loss": 2.4183, + "step": 20770 + }, + { + "epoch": 2.1965947546531304, + "grad_norm": 8.9375, + "learning_rate": 3.106721034596008e-06, + "loss": 2.5898, + "step": 20771 + }, + { + "epoch": 2.196700507614213, + "grad_norm": 9.5625, + "learning_rate": 3.1065548967761332e-06, + "loss": 2.4758, + "step": 20772 + }, + { + "epoch": 2.196806260575296, + "grad_norm": 7.84375, + "learning_rate": 3.1063887561099443e-06, + "loss": 2.5151, + "step": 20773 + }, + { + "epoch": 2.196912013536379, + "grad_norm": 9.1875, + "learning_rate": 3.1062226125982207e-06, + "loss": 2.92, + "step": 20774 + }, + { + "epoch": 2.1970177664974617, + "grad_norm": 8.25, + "learning_rate": 3.1060564662417424e-06, + "loss": 2.5644, + "step": 20775 + }, + { + "epoch": 2.197123519458545, + "grad_norm": 8.375, + "learning_rate": 3.1058903170412896e-06, + "loss": 2.7121, + "step": 20776 + }, + { + "epoch": 2.197229272419628, + "grad_norm": 8.375, + "learning_rate": 3.10572416499764e-06, + "loss": 2.6506, + "step": 20777 + }, + { + "epoch": 2.1973350253807107, + "grad_norm": 8.375, + "learning_rate": 3.1055580101115757e-06, + "loss": 2.7149, + "step": 20778 + }, + { + "epoch": 2.1974407783417935, + "grad_norm": 8.375, + "learning_rate": 3.105391852383875e-06, + "loss": 2.6458, + "step": 20779 + }, + { + "epoch": 2.1975465313028764, + "grad_norm": 8.5625, + "learning_rate": 3.105225691815318e-06, + "loss": 2.7905, + "step": 20780 + }, + { + "epoch": 2.197652284263959, + "grad_norm": 8.5, + "learning_rate": 3.105059528406684e-06, + "loss": 2.5051, + "step": 20781 + }, + { + "epoch": 2.1977580372250425, + "grad_norm": 8.375, + "learning_rate": 3.104893362158754e-06, + "loss": 2.6422, + "step": 20782 + }, + { + "epoch": 2.1978637901861253, + "grad_norm": 8.3125, + "learning_rate": 3.104727193072306e-06, + "loss": 2.511, + "step": 20783 + }, + { + "epoch": 2.197969543147208, + "grad_norm": 8.0625, + "learning_rate": 3.104561021148121e-06, + "loss": 2.6832, + "step": 20784 + }, + { + "epoch": 2.198075296108291, + "grad_norm": 8.75, + "learning_rate": 3.1043948463869776e-06, + "loss": 2.4673, + "step": 20785 + }, + { + "epoch": 2.198181049069374, + "grad_norm": 8.0, + "learning_rate": 3.104228668789657e-06, + "loss": 2.2449, + "step": 20786 + }, + { + "epoch": 2.1982868020304567, + "grad_norm": 8.25, + "learning_rate": 3.1040624883569388e-06, + "loss": 2.4673, + "step": 20787 + }, + { + "epoch": 2.19839255499154, + "grad_norm": 8.125, + "learning_rate": 3.103896305089602e-06, + "loss": 2.4249, + "step": 20788 + }, + { + "epoch": 2.198498307952623, + "grad_norm": 8.1875, + "learning_rate": 3.103730118988427e-06, + "loss": 2.6971, + "step": 20789 + }, + { + "epoch": 2.1986040609137056, + "grad_norm": 8.5625, + "learning_rate": 3.1035639300541932e-06, + "loss": 2.7312, + "step": 20790 + }, + { + "epoch": 2.1987098138747885, + "grad_norm": 8.6875, + "learning_rate": 3.1033977382876806e-06, + "loss": 2.6095, + "step": 20791 + }, + { + "epoch": 2.1988155668358713, + "grad_norm": 8.0625, + "learning_rate": 3.10323154368967e-06, + "loss": 2.4421, + "step": 20792 + }, + { + "epoch": 2.198921319796954, + "grad_norm": 8.75, + "learning_rate": 3.1030653462609394e-06, + "loss": 2.8023, + "step": 20793 + }, + { + "epoch": 2.199027072758037, + "grad_norm": 8.25, + "learning_rate": 3.10289914600227e-06, + "loss": 2.6123, + "step": 20794 + }, + { + "epoch": 2.1991328257191203, + "grad_norm": 8.75, + "learning_rate": 3.102732942914442e-06, + "loss": 2.5594, + "step": 20795 + }, + { + "epoch": 2.199238578680203, + "grad_norm": 9.0, + "learning_rate": 3.1025667369982345e-06, + "loss": 2.5126, + "step": 20796 + }, + { + "epoch": 2.199344331641286, + "grad_norm": 8.1875, + "learning_rate": 3.102400528254428e-06, + "loss": 2.4001, + "step": 20797 + }, + { + "epoch": 2.199450084602369, + "grad_norm": 8.5, + "learning_rate": 3.102234316683802e-06, + "loss": 2.6349, + "step": 20798 + }, + { + "epoch": 2.1995558375634516, + "grad_norm": 8.5625, + "learning_rate": 3.102068102287136e-06, + "loss": 2.607, + "step": 20799 + }, + { + "epoch": 2.199661590524535, + "grad_norm": 8.4375, + "learning_rate": 3.1019018850652116e-06, + "loss": 2.4443, + "step": 20800 + }, + { + "epoch": 2.1997673434856178, + "grad_norm": 8.8125, + "learning_rate": 3.101735665018808e-06, + "loss": 2.6373, + "step": 20801 + }, + { + "epoch": 2.1998730964467006, + "grad_norm": 8.75, + "learning_rate": 3.1015694421487042e-06, + "loss": 2.6412, + "step": 20802 + }, + { + "epoch": 2.1999788494077834, + "grad_norm": 10.0625, + "learning_rate": 3.101403216455681e-06, + "loss": 2.814, + "step": 20803 + }, + { + "epoch": 2.2000846023688663, + "grad_norm": 8.25, + "learning_rate": 3.101236987940519e-06, + "loss": 2.4651, + "step": 20804 + }, + { + "epoch": 2.200190355329949, + "grad_norm": 8.25, + "learning_rate": 3.1010707566039977e-06, + "loss": 2.167, + "step": 20805 + }, + { + "epoch": 2.200296108291032, + "grad_norm": 8.625, + "learning_rate": 3.100904522446897e-06, + "loss": 2.1527, + "step": 20806 + }, + { + "epoch": 2.2004018612521152, + "grad_norm": 8.125, + "learning_rate": 3.1007382854699975e-06, + "loss": 2.3421, + "step": 20807 + }, + { + "epoch": 2.200507614213198, + "grad_norm": 8.4375, + "learning_rate": 3.1005720456740783e-06, + "loss": 2.5091, + "step": 20808 + }, + { + "epoch": 2.200613367174281, + "grad_norm": 8.375, + "learning_rate": 3.1004058030599204e-06, + "loss": 2.5336, + "step": 20809 + }, + { + "epoch": 2.2007191201353637, + "grad_norm": 8.875, + "learning_rate": 3.100239557628304e-06, + "loss": 2.6267, + "step": 20810 + }, + { + "epoch": 2.2008248730964466, + "grad_norm": 7.75, + "learning_rate": 3.1000733093800084e-06, + "loss": 2.4721, + "step": 20811 + }, + { + "epoch": 2.2009306260575294, + "grad_norm": 7.8125, + "learning_rate": 3.0999070583158143e-06, + "loss": 2.1496, + "step": 20812 + }, + { + "epoch": 2.2010363790186127, + "grad_norm": 8.3125, + "learning_rate": 3.0997408044365014e-06, + "loss": 2.6817, + "step": 20813 + }, + { + "epoch": 2.2011421319796955, + "grad_norm": 8.4375, + "learning_rate": 3.099574547742851e-06, + "loss": 2.5455, + "step": 20814 + }, + { + "epoch": 2.2012478849407784, + "grad_norm": 8.25, + "learning_rate": 3.0994082882356415e-06, + "loss": 2.455, + "step": 20815 + }, + { + "epoch": 2.2013536379018612, + "grad_norm": 8.6875, + "learning_rate": 3.0992420259156543e-06, + "loss": 2.7368, + "step": 20816 + }, + { + "epoch": 2.201459390862944, + "grad_norm": 7.5625, + "learning_rate": 3.0990757607836696e-06, + "loss": 2.3201, + "step": 20817 + }, + { + "epoch": 2.201565143824027, + "grad_norm": 7.90625, + "learning_rate": 3.0989094928404674e-06, + "loss": 2.4592, + "step": 20818 + }, + { + "epoch": 2.20167089678511, + "grad_norm": 8.0625, + "learning_rate": 3.098743222086828e-06, + "loss": 2.4468, + "step": 20819 + }, + { + "epoch": 2.201776649746193, + "grad_norm": 8.4375, + "learning_rate": 3.0985769485235317e-06, + "loss": 2.7959, + "step": 20820 + }, + { + "epoch": 2.201882402707276, + "grad_norm": 8.3125, + "learning_rate": 3.098410672151357e-06, + "loss": 2.5945, + "step": 20821 + }, + { + "epoch": 2.2019881556683587, + "grad_norm": 8.25, + "learning_rate": 3.098244392971087e-06, + "loss": 2.5091, + "step": 20822 + }, + { + "epoch": 2.2020939086294415, + "grad_norm": 8.5625, + "learning_rate": 3.098078110983501e-06, + "loss": 2.6125, + "step": 20823 + }, + { + "epoch": 2.2021996615905244, + "grad_norm": 8.625, + "learning_rate": 3.097911826189379e-06, + "loss": 2.6219, + "step": 20824 + }, + { + "epoch": 2.202305414551607, + "grad_norm": 8.1875, + "learning_rate": 3.0977455385895013e-06, + "loss": 2.5172, + "step": 20825 + }, + { + "epoch": 2.2024111675126905, + "grad_norm": 8.4375, + "learning_rate": 3.097579248184648e-06, + "loss": 2.5883, + "step": 20826 + }, + { + "epoch": 2.2025169204737733, + "grad_norm": 8.875, + "learning_rate": 3.0974129549756e-06, + "loss": 2.5894, + "step": 20827 + }, + { + "epoch": 2.202622673434856, + "grad_norm": 8.3125, + "learning_rate": 3.097246658963137e-06, + "loss": 2.5022, + "step": 20828 + }, + { + "epoch": 2.202728426395939, + "grad_norm": 8.4375, + "learning_rate": 3.0970803601480396e-06, + "loss": 2.7268, + "step": 20829 + }, + { + "epoch": 2.202834179357022, + "grad_norm": 8.5, + "learning_rate": 3.0969140585310887e-06, + "loss": 2.4017, + "step": 20830 + }, + { + "epoch": 2.2029399323181047, + "grad_norm": 8.875, + "learning_rate": 3.096747754113063e-06, + "loss": 2.6002, + "step": 20831 + }, + { + "epoch": 2.203045685279188, + "grad_norm": 8.3125, + "learning_rate": 3.096581446894746e-06, + "loss": 2.5069, + "step": 20832 + }, + { + "epoch": 2.203151438240271, + "grad_norm": 8.5, + "learning_rate": 3.0964151368769154e-06, + "loss": 2.5915, + "step": 20833 + }, + { + "epoch": 2.2032571912013537, + "grad_norm": 8.375, + "learning_rate": 3.096248824060352e-06, + "loss": 2.3784, + "step": 20834 + }, + { + "epoch": 2.2033629441624365, + "grad_norm": 8.875, + "learning_rate": 3.0960825084458374e-06, + "loss": 2.7749, + "step": 20835 + }, + { + "epoch": 2.2034686971235193, + "grad_norm": 9.0625, + "learning_rate": 3.0959161900341515e-06, + "loss": 2.5267, + "step": 20836 + }, + { + "epoch": 2.203574450084602, + "grad_norm": 8.4375, + "learning_rate": 3.095749868826074e-06, + "loss": 2.5595, + "step": 20837 + }, + { + "epoch": 2.2036802030456855, + "grad_norm": 8.3125, + "learning_rate": 3.095583544822387e-06, + "loss": 2.447, + "step": 20838 + }, + { + "epoch": 2.2037859560067683, + "grad_norm": 8.5625, + "learning_rate": 3.095417218023869e-06, + "loss": 2.452, + "step": 20839 + }, + { + "epoch": 2.203891708967851, + "grad_norm": 8.8125, + "learning_rate": 3.095250888431302e-06, + "loss": 2.7294, + "step": 20840 + }, + { + "epoch": 2.203997461928934, + "grad_norm": 8.8125, + "learning_rate": 3.0950845560454655e-06, + "loss": 2.5381, + "step": 20841 + }, + { + "epoch": 2.204103214890017, + "grad_norm": 8.8125, + "learning_rate": 3.0949182208671413e-06, + "loss": 2.3976, + "step": 20842 + }, + { + "epoch": 2.2042089678510997, + "grad_norm": 8.6875, + "learning_rate": 3.094751882897109e-06, + "loss": 2.6498, + "step": 20843 + }, + { + "epoch": 2.204314720812183, + "grad_norm": 8.1875, + "learning_rate": 3.0945855421361485e-06, + "loss": 2.4849, + "step": 20844 + }, + { + "epoch": 2.2044204737732658, + "grad_norm": 8.375, + "learning_rate": 3.0944191985850426e-06, + "loss": 2.4958, + "step": 20845 + }, + { + "epoch": 2.2045262267343486, + "grad_norm": 8.25, + "learning_rate": 3.0942528522445704e-06, + "loss": 2.5536, + "step": 20846 + }, + { + "epoch": 2.2046319796954315, + "grad_norm": 8.6875, + "learning_rate": 3.0940865031155116e-06, + "loss": 2.6438, + "step": 20847 + }, + { + "epoch": 2.2047377326565143, + "grad_norm": 9.0625, + "learning_rate": 3.093920151198648e-06, + "loss": 2.4871, + "step": 20848 + }, + { + "epoch": 2.204843485617597, + "grad_norm": 8.25, + "learning_rate": 3.0937537964947604e-06, + "loss": 2.4974, + "step": 20849 + }, + { + "epoch": 2.2049492385786804, + "grad_norm": 8.125, + "learning_rate": 3.093587439004629e-06, + "loss": 2.5749, + "step": 20850 + }, + { + "epoch": 2.2050549915397633, + "grad_norm": 8.375, + "learning_rate": 3.0934210787290344e-06, + "loss": 2.5202, + "step": 20851 + }, + { + "epoch": 2.205160744500846, + "grad_norm": 9.6875, + "learning_rate": 3.0932547156687578e-06, + "loss": 2.566, + "step": 20852 + }, + { + "epoch": 2.205266497461929, + "grad_norm": 8.6875, + "learning_rate": 3.093088349824579e-06, + "loss": 2.8362, + "step": 20853 + }, + { + "epoch": 2.2053722504230118, + "grad_norm": 9.4375, + "learning_rate": 3.0929219811972796e-06, + "loss": 2.4837, + "step": 20854 + }, + { + "epoch": 2.2054780033840946, + "grad_norm": 8.3125, + "learning_rate": 3.0927556097876395e-06, + "loss": 2.5425, + "step": 20855 + }, + { + "epoch": 2.2055837563451774, + "grad_norm": 8.6875, + "learning_rate": 3.0925892355964394e-06, + "loss": 2.5916, + "step": 20856 + }, + { + "epoch": 2.2056895093062607, + "grad_norm": 8.0625, + "learning_rate": 3.09242285862446e-06, + "loss": 2.43, + "step": 20857 + }, + { + "epoch": 2.2057952622673436, + "grad_norm": 8.3125, + "learning_rate": 3.092256478872484e-06, + "loss": 2.498, + "step": 20858 + }, + { + "epoch": 2.2059010152284264, + "grad_norm": 8.3125, + "learning_rate": 3.0920900963412896e-06, + "loss": 2.5069, + "step": 20859 + }, + { + "epoch": 2.2060067681895092, + "grad_norm": 8.0625, + "learning_rate": 3.0919237110316584e-06, + "loss": 2.6906, + "step": 20860 + }, + { + "epoch": 2.206112521150592, + "grad_norm": 8.3125, + "learning_rate": 3.091757322944372e-06, + "loss": 2.4534, + "step": 20861 + }, + { + "epoch": 2.206218274111675, + "grad_norm": 8.5, + "learning_rate": 3.09159093208021e-06, + "loss": 2.5947, + "step": 20862 + }, + { + "epoch": 2.206324027072758, + "grad_norm": 8.5625, + "learning_rate": 3.091424538439954e-06, + "loss": 2.6037, + "step": 20863 + }, + { + "epoch": 2.206429780033841, + "grad_norm": 7.8125, + "learning_rate": 3.0912581420243843e-06, + "loss": 2.2052, + "step": 20864 + }, + { + "epoch": 2.206535532994924, + "grad_norm": 8.8125, + "learning_rate": 3.091091742834282e-06, + "loss": 2.4356, + "step": 20865 + }, + { + "epoch": 2.2066412859560067, + "grad_norm": 8.625, + "learning_rate": 3.0909253408704275e-06, + "loss": 2.4678, + "step": 20866 + }, + { + "epoch": 2.2067470389170896, + "grad_norm": 8.3125, + "learning_rate": 3.0907589361336027e-06, + "loss": 2.6527, + "step": 20867 + }, + { + "epoch": 2.2068527918781724, + "grad_norm": 8.1875, + "learning_rate": 3.0905925286245874e-06, + "loss": 2.3627, + "step": 20868 + }, + { + "epoch": 2.2069585448392557, + "grad_norm": 8.0625, + "learning_rate": 3.0904261183441632e-06, + "loss": 2.505, + "step": 20869 + }, + { + "epoch": 2.2070642978003385, + "grad_norm": 8.875, + "learning_rate": 3.0902597052931104e-06, + "loss": 2.7543, + "step": 20870 + }, + { + "epoch": 2.2071700507614214, + "grad_norm": 8.0, + "learning_rate": 3.0900932894722107e-06, + "loss": 2.2984, + "step": 20871 + }, + { + "epoch": 2.207275803722504, + "grad_norm": 8.3125, + "learning_rate": 3.089926870882244e-06, + "loss": 2.5615, + "step": 20872 + }, + { + "epoch": 2.207381556683587, + "grad_norm": 8.375, + "learning_rate": 3.089760449523992e-06, + "loss": 2.6791, + "step": 20873 + }, + { + "epoch": 2.20748730964467, + "grad_norm": 8.625, + "learning_rate": 3.0895940253982354e-06, + "loss": 2.58, + "step": 20874 + }, + { + "epoch": 2.207593062605753, + "grad_norm": 8.625, + "learning_rate": 3.0894275985057553e-06, + "loss": 2.6802, + "step": 20875 + }, + { + "epoch": 2.207698815566836, + "grad_norm": 8.5625, + "learning_rate": 3.089261168847332e-06, + "loss": 2.683, + "step": 20876 + }, + { + "epoch": 2.207804568527919, + "grad_norm": 8.9375, + "learning_rate": 3.089094736423748e-06, + "loss": 2.5284, + "step": 20877 + }, + { + "epoch": 2.2079103214890017, + "grad_norm": 8.1875, + "learning_rate": 3.088928301235783e-06, + "loss": 2.419, + "step": 20878 + }, + { + "epoch": 2.2080160744500845, + "grad_norm": 8.9375, + "learning_rate": 3.0887618632842175e-06, + "loss": 2.5776, + "step": 20879 + }, + { + "epoch": 2.2081218274111674, + "grad_norm": 8.0625, + "learning_rate": 3.0885954225698343e-06, + "loss": 2.6251, + "step": 20880 + }, + { + "epoch": 2.2082275803722506, + "grad_norm": 8.1875, + "learning_rate": 3.0884289790934135e-06, + "loss": 2.2732, + "step": 20881 + }, + { + "epoch": 2.2083333333333335, + "grad_norm": 8.125, + "learning_rate": 3.088262532855735e-06, + "loss": 2.3931, + "step": 20882 + }, + { + "epoch": 2.2084390862944163, + "grad_norm": 8.0, + "learning_rate": 3.088096083857582e-06, + "loss": 2.4604, + "step": 20883 + }, + { + "epoch": 2.208544839255499, + "grad_norm": 8.3125, + "learning_rate": 3.0879296320997348e-06, + "loss": 2.4887, + "step": 20884 + }, + { + "epoch": 2.208650592216582, + "grad_norm": 8.5, + "learning_rate": 3.0877631775829743e-06, + "loss": 2.587, + "step": 20885 + }, + { + "epoch": 2.208756345177665, + "grad_norm": 8.75, + "learning_rate": 3.087596720308081e-06, + "loss": 2.7523, + "step": 20886 + }, + { + "epoch": 2.2088620981387477, + "grad_norm": 8.0, + "learning_rate": 3.087430260275837e-06, + "loss": 2.4892, + "step": 20887 + }, + { + "epoch": 2.208967851099831, + "grad_norm": 8.5625, + "learning_rate": 3.0872637974870223e-06, + "loss": 2.6711, + "step": 20888 + }, + { + "epoch": 2.209073604060914, + "grad_norm": 8.25, + "learning_rate": 3.0870973319424197e-06, + "loss": 2.6797, + "step": 20889 + }, + { + "epoch": 2.2091793570219966, + "grad_norm": 8.625, + "learning_rate": 3.0869308636428098e-06, + "loss": 2.7632, + "step": 20890 + }, + { + "epoch": 2.2092851099830795, + "grad_norm": 8.3125, + "learning_rate": 3.0867643925889723e-06, + "loss": 2.6321, + "step": 20891 + }, + { + "epoch": 2.2093908629441623, + "grad_norm": 8.5, + "learning_rate": 3.0865979187816895e-06, + "loss": 2.5046, + "step": 20892 + }, + { + "epoch": 2.209496615905245, + "grad_norm": 8.25, + "learning_rate": 3.086431442221743e-06, + "loss": 2.6374, + "step": 20893 + }, + { + "epoch": 2.2096023688663284, + "grad_norm": 8.5625, + "learning_rate": 3.086264962909914e-06, + "loss": 2.6743, + "step": 20894 + }, + { + "epoch": 2.2097081218274113, + "grad_norm": 8.3125, + "learning_rate": 3.086098480846983e-06, + "loss": 2.4131, + "step": 20895 + }, + { + "epoch": 2.209813874788494, + "grad_norm": 8.25, + "learning_rate": 3.0859319960337314e-06, + "loss": 2.4218, + "step": 20896 + }, + { + "epoch": 2.209919627749577, + "grad_norm": 8.375, + "learning_rate": 3.085765508470941e-06, + "loss": 2.3317, + "step": 20897 + }, + { + "epoch": 2.21002538071066, + "grad_norm": 8.1875, + "learning_rate": 3.085599018159392e-06, + "loss": 2.6074, + "step": 20898 + }, + { + "epoch": 2.2101311336717426, + "grad_norm": 8.0625, + "learning_rate": 3.0854325250998663e-06, + "loss": 2.5409, + "step": 20899 + }, + { + "epoch": 2.210236886632826, + "grad_norm": 9.25, + "learning_rate": 3.0852660292931452e-06, + "loss": 2.7258, + "step": 20900 + }, + { + "epoch": 2.2103426395939088, + "grad_norm": 8.0, + "learning_rate": 3.0850995307400104e-06, + "loss": 2.4073, + "step": 20901 + }, + { + "epoch": 2.2104483925549916, + "grad_norm": 8.6875, + "learning_rate": 3.084933029441243e-06, + "loss": 2.5668, + "step": 20902 + }, + { + "epoch": 2.2105541455160744, + "grad_norm": 8.3125, + "learning_rate": 3.084766525397623e-06, + "loss": 2.463, + "step": 20903 + }, + { + "epoch": 2.2106598984771573, + "grad_norm": 8.25, + "learning_rate": 3.084600018609934e-06, + "loss": 2.5521, + "step": 20904 + }, + { + "epoch": 2.21076565143824, + "grad_norm": 8.625, + "learning_rate": 3.084433509078956e-06, + "loss": 2.7098, + "step": 20905 + }, + { + "epoch": 2.210871404399323, + "grad_norm": 8.125, + "learning_rate": 3.0842669968054707e-06, + "loss": 2.5151, + "step": 20906 + }, + { + "epoch": 2.2109771573604062, + "grad_norm": 8.6875, + "learning_rate": 3.0841004817902586e-06, + "loss": 2.5443, + "step": 20907 + }, + { + "epoch": 2.211082910321489, + "grad_norm": 8.3125, + "learning_rate": 3.0839339640341026e-06, + "loss": 2.6213, + "step": 20908 + }, + { + "epoch": 2.211188663282572, + "grad_norm": 8.5, + "learning_rate": 3.083767443537783e-06, + "loss": 2.4672, + "step": 20909 + }, + { + "epoch": 2.2112944162436547, + "grad_norm": 8.375, + "learning_rate": 3.083600920302081e-06, + "loss": 2.6266, + "step": 20910 + }, + { + "epoch": 2.2114001692047376, + "grad_norm": 8.25, + "learning_rate": 3.083434394327779e-06, + "loss": 2.4504, + "step": 20911 + }, + { + "epoch": 2.211505922165821, + "grad_norm": 7.96875, + "learning_rate": 3.0832678656156583e-06, + "loss": 2.5208, + "step": 20912 + }, + { + "epoch": 2.2116116751269037, + "grad_norm": 8.0, + "learning_rate": 3.0831013341665006e-06, + "loss": 2.4395, + "step": 20913 + }, + { + "epoch": 2.2117174280879865, + "grad_norm": 8.3125, + "learning_rate": 3.0829347999810855e-06, + "loss": 2.6711, + "step": 20914 + }, + { + "epoch": 2.2118231810490694, + "grad_norm": 8.3125, + "learning_rate": 3.0827682630601966e-06, + "loss": 2.4012, + "step": 20915 + }, + { + "epoch": 2.2119289340101522, + "grad_norm": 8.4375, + "learning_rate": 3.0826017234046144e-06, + "loss": 2.5234, + "step": 20916 + }, + { + "epoch": 2.212034686971235, + "grad_norm": 8.5, + "learning_rate": 3.0824351810151203e-06, + "loss": 2.6372, + "step": 20917 + }, + { + "epoch": 2.212140439932318, + "grad_norm": 8.6875, + "learning_rate": 3.0822686358924965e-06, + "loss": 2.4173, + "step": 20918 + }, + { + "epoch": 2.212246192893401, + "grad_norm": 8.6875, + "learning_rate": 3.082102088037524e-06, + "loss": 2.516, + "step": 20919 + }, + { + "epoch": 2.212351945854484, + "grad_norm": 8.4375, + "learning_rate": 3.0819355374509846e-06, + "loss": 2.3588, + "step": 20920 + }, + { + "epoch": 2.212457698815567, + "grad_norm": 8.8125, + "learning_rate": 3.0817689841336597e-06, + "loss": 2.7019, + "step": 20921 + }, + { + "epoch": 2.2125634517766497, + "grad_norm": 8.125, + "learning_rate": 3.0816024280863306e-06, + "loss": 2.3743, + "step": 20922 + }, + { + "epoch": 2.2126692047377325, + "grad_norm": 8.6875, + "learning_rate": 3.081435869309779e-06, + "loss": 2.6226, + "step": 20923 + }, + { + "epoch": 2.2127749576988154, + "grad_norm": 8.125, + "learning_rate": 3.0812693078047874e-06, + "loss": 2.5789, + "step": 20924 + }, + { + "epoch": 2.2128807106598987, + "grad_norm": 8.3125, + "learning_rate": 3.0811027435721364e-06, + "loss": 2.3938, + "step": 20925 + }, + { + "epoch": 2.2129864636209815, + "grad_norm": 8.1875, + "learning_rate": 3.0809361766126072e-06, + "loss": 2.5035, + "step": 20926 + }, + { + "epoch": 2.2130922165820643, + "grad_norm": 8.125, + "learning_rate": 3.0807696069269826e-06, + "loss": 2.5626, + "step": 20927 + }, + { + "epoch": 2.213197969543147, + "grad_norm": 8.6875, + "learning_rate": 3.0806030345160442e-06, + "loss": 2.409, + "step": 20928 + }, + { + "epoch": 2.21330372250423, + "grad_norm": 8.375, + "learning_rate": 3.080436459380573e-06, + "loss": 2.6814, + "step": 20929 + }, + { + "epoch": 2.213409475465313, + "grad_norm": 8.3125, + "learning_rate": 3.0802698815213503e-06, + "loss": 2.5094, + "step": 20930 + }, + { + "epoch": 2.213515228426396, + "grad_norm": 8.25, + "learning_rate": 3.080103300939159e-06, + "loss": 2.526, + "step": 20931 + }, + { + "epoch": 2.213620981387479, + "grad_norm": 8.4375, + "learning_rate": 3.0799367176347805e-06, + "loss": 2.7179, + "step": 20932 + }, + { + "epoch": 2.213726734348562, + "grad_norm": 8.8125, + "learning_rate": 3.079770131608995e-06, + "loss": 2.4132, + "step": 20933 + }, + { + "epoch": 2.2138324873096447, + "grad_norm": 8.1875, + "learning_rate": 3.0796035428625866e-06, + "loss": 2.5901, + "step": 20934 + }, + { + "epoch": 2.2139382402707275, + "grad_norm": 8.625, + "learning_rate": 3.079436951396335e-06, + "loss": 2.7441, + "step": 20935 + }, + { + "epoch": 2.2140439932318103, + "grad_norm": 8.3125, + "learning_rate": 3.0792703572110233e-06, + "loss": 2.4686, + "step": 20936 + }, + { + "epoch": 2.214149746192893, + "grad_norm": 8.875, + "learning_rate": 3.079103760307432e-06, + "loss": 2.3688, + "step": 20937 + }, + { + "epoch": 2.2142554991539765, + "grad_norm": 8.8125, + "learning_rate": 3.0789371606863444e-06, + "loss": 2.5679, + "step": 20938 + }, + { + "epoch": 2.2143612521150593, + "grad_norm": 8.8125, + "learning_rate": 3.078770558348541e-06, + "loss": 2.7046, + "step": 20939 + }, + { + "epoch": 2.214467005076142, + "grad_norm": 8.75, + "learning_rate": 3.078603953294804e-06, + "loss": 2.5601, + "step": 20940 + }, + { + "epoch": 2.214572758037225, + "grad_norm": 8.3125, + "learning_rate": 3.078437345525916e-06, + "loss": 2.4665, + "step": 20941 + }, + { + "epoch": 2.214678510998308, + "grad_norm": 7.78125, + "learning_rate": 3.078270735042657e-06, + "loss": 2.3827, + "step": 20942 + }, + { + "epoch": 2.214784263959391, + "grad_norm": 8.125, + "learning_rate": 3.078104121845811e-06, + "loss": 2.4877, + "step": 20943 + }, + { + "epoch": 2.214890016920474, + "grad_norm": 8.5, + "learning_rate": 3.077937505936159e-06, + "loss": 2.6206, + "step": 20944 + }, + { + "epoch": 2.2149957698815568, + "grad_norm": 8.625, + "learning_rate": 3.0777708873144813e-06, + "loss": 2.5995, + "step": 20945 + }, + { + "epoch": 2.2151015228426396, + "grad_norm": 7.96875, + "learning_rate": 3.0776042659815614e-06, + "loss": 2.3995, + "step": 20946 + }, + { + "epoch": 2.2152072758037225, + "grad_norm": 8.625, + "learning_rate": 3.0774376419381818e-06, + "loss": 2.6752, + "step": 20947 + }, + { + "epoch": 2.2153130287648053, + "grad_norm": 7.96875, + "learning_rate": 3.077271015185123e-06, + "loss": 2.3963, + "step": 20948 + }, + { + "epoch": 2.215418781725888, + "grad_norm": 11.375, + "learning_rate": 3.0771043857231674e-06, + "loss": 2.478, + "step": 20949 + }, + { + "epoch": 2.2155245346869714, + "grad_norm": 8.3125, + "learning_rate": 3.076937753553097e-06, + "loss": 2.57, + "step": 20950 + }, + { + "epoch": 2.2156302876480543, + "grad_norm": 8.625, + "learning_rate": 3.0767711186756942e-06, + "loss": 2.6072, + "step": 20951 + }, + { + "epoch": 2.215736040609137, + "grad_norm": 8.625, + "learning_rate": 3.0766044810917396e-06, + "loss": 2.7531, + "step": 20952 + }, + { + "epoch": 2.21584179357022, + "grad_norm": 7.71875, + "learning_rate": 3.076437840802017e-06, + "loss": 2.0894, + "step": 20953 + }, + { + "epoch": 2.2159475465313028, + "grad_norm": 7.9375, + "learning_rate": 3.0762711978073055e-06, + "loss": 2.4587, + "step": 20954 + }, + { + "epoch": 2.2160532994923856, + "grad_norm": 8.25, + "learning_rate": 3.076104552108391e-06, + "loss": 2.4064, + "step": 20955 + }, + { + "epoch": 2.216159052453469, + "grad_norm": 8.875, + "learning_rate": 3.075937903706052e-06, + "loss": 2.6202, + "step": 20956 + }, + { + "epoch": 2.2162648054145517, + "grad_norm": 8.3125, + "learning_rate": 3.075771252601073e-06, + "loss": 2.3796, + "step": 20957 + }, + { + "epoch": 2.2163705583756346, + "grad_norm": 8.25, + "learning_rate": 3.0756045987942346e-06, + "loss": 2.5473, + "step": 20958 + }, + { + "epoch": 2.2164763113367174, + "grad_norm": 8.5, + "learning_rate": 3.075437942286319e-06, + "loss": 2.6553, + "step": 20959 + }, + { + "epoch": 2.2165820642978002, + "grad_norm": 8.5, + "learning_rate": 3.075271283078109e-06, + "loss": 2.7549, + "step": 20960 + }, + { + "epoch": 2.216687817258883, + "grad_norm": 8.1875, + "learning_rate": 3.0751046211703855e-06, + "loss": 2.6966, + "step": 20961 + }, + { + "epoch": 2.2167935702199664, + "grad_norm": 8.125, + "learning_rate": 3.0749379565639313e-06, + "loss": 2.5831, + "step": 20962 + }, + { + "epoch": 2.216899323181049, + "grad_norm": 8.5, + "learning_rate": 3.074771289259529e-06, + "loss": 2.5446, + "step": 20963 + }, + { + "epoch": 2.217005076142132, + "grad_norm": 8.3125, + "learning_rate": 3.0746046192579597e-06, + "loss": 2.7217, + "step": 20964 + }, + { + "epoch": 2.217110829103215, + "grad_norm": 9.1875, + "learning_rate": 3.074437946560006e-06, + "loss": 2.5892, + "step": 20965 + }, + { + "epoch": 2.2172165820642977, + "grad_norm": 9.25, + "learning_rate": 3.07427127116645e-06, + "loss": 2.8357, + "step": 20966 + }, + { + "epoch": 2.2173223350253806, + "grad_norm": 8.0625, + "learning_rate": 3.074104593078074e-06, + "loss": 2.626, + "step": 20967 + }, + { + "epoch": 2.2174280879864634, + "grad_norm": 8.375, + "learning_rate": 3.073937912295659e-06, + "loss": 2.5187, + "step": 20968 + }, + { + "epoch": 2.2175338409475467, + "grad_norm": 8.25, + "learning_rate": 3.0737712288199897e-06, + "loss": 2.4453, + "step": 20969 + }, + { + "epoch": 2.2176395939086295, + "grad_norm": 8.8125, + "learning_rate": 3.073604542651846e-06, + "loss": 2.3919, + "step": 20970 + }, + { + "epoch": 2.2177453468697124, + "grad_norm": 8.6875, + "learning_rate": 3.07343785379201e-06, + "loss": 2.7165, + "step": 20971 + }, + { + "epoch": 2.217851099830795, + "grad_norm": 9.25, + "learning_rate": 3.0732711622412647e-06, + "loss": 2.6541, + "step": 20972 + }, + { + "epoch": 2.217956852791878, + "grad_norm": 8.4375, + "learning_rate": 3.0731044680003934e-06, + "loss": 2.4988, + "step": 20973 + }, + { + "epoch": 2.218062605752961, + "grad_norm": 8.4375, + "learning_rate": 3.072937771070177e-06, + "loss": 2.5196, + "step": 20974 + }, + { + "epoch": 2.218168358714044, + "grad_norm": 8.1875, + "learning_rate": 3.072771071451397e-06, + "loss": 2.5592, + "step": 20975 + }, + { + "epoch": 2.218274111675127, + "grad_norm": 8.1875, + "learning_rate": 3.0726043691448376e-06, + "loss": 2.6184, + "step": 20976 + }, + { + "epoch": 2.21837986463621, + "grad_norm": 8.125, + "learning_rate": 3.07243766415128e-06, + "loss": 2.4097, + "step": 20977 + }, + { + "epoch": 2.2184856175972927, + "grad_norm": 8.3125, + "learning_rate": 3.072270956471506e-06, + "loss": 2.4678, + "step": 20978 + }, + { + "epoch": 2.2185913705583755, + "grad_norm": 8.5625, + "learning_rate": 3.072104246106299e-06, + "loss": 2.4326, + "step": 20979 + }, + { + "epoch": 2.2186971235194584, + "grad_norm": 8.0625, + "learning_rate": 3.0719375330564406e-06, + "loss": 2.5965, + "step": 20980 + }, + { + "epoch": 2.2188028764805416, + "grad_norm": 8.3125, + "learning_rate": 3.0717708173227122e-06, + "loss": 2.4835, + "step": 20981 + }, + { + "epoch": 2.2189086294416245, + "grad_norm": 8.0625, + "learning_rate": 3.0716040989058986e-06, + "loss": 2.5944, + "step": 20982 + }, + { + "epoch": 2.2190143824027073, + "grad_norm": 8.875, + "learning_rate": 3.0714373778067806e-06, + "loss": 2.7661, + "step": 20983 + }, + { + "epoch": 2.21912013536379, + "grad_norm": 8.5625, + "learning_rate": 3.0712706540261405e-06, + "loss": 2.7426, + "step": 20984 + }, + { + "epoch": 2.219225888324873, + "grad_norm": 8.5625, + "learning_rate": 3.071103927564761e-06, + "loss": 2.8467, + "step": 20985 + }, + { + "epoch": 2.219331641285956, + "grad_norm": 8.5, + "learning_rate": 3.0709371984234243e-06, + "loss": 2.6317, + "step": 20986 + }, + { + "epoch": 2.219437394247039, + "grad_norm": 8.5, + "learning_rate": 3.070770466602912e-06, + "loss": 2.4054, + "step": 20987 + }, + { + "epoch": 2.219543147208122, + "grad_norm": 8.4375, + "learning_rate": 3.0706037321040083e-06, + "loss": 2.6284, + "step": 20988 + }, + { + "epoch": 2.219648900169205, + "grad_norm": 8.5, + "learning_rate": 3.0704369949274943e-06, + "loss": 2.6901, + "step": 20989 + }, + { + "epoch": 2.2197546531302876, + "grad_norm": 8.1875, + "learning_rate": 3.070270255074153e-06, + "loss": 2.5009, + "step": 20990 + }, + { + "epoch": 2.2198604060913705, + "grad_norm": 8.125, + "learning_rate": 3.070103512544766e-06, + "loss": 2.379, + "step": 20991 + }, + { + "epoch": 2.2199661590524533, + "grad_norm": 8.375, + "learning_rate": 3.069936767340117e-06, + "loss": 2.5165, + "step": 20992 + }, + { + "epoch": 2.2200719120135366, + "grad_norm": 8.3125, + "learning_rate": 3.069770019460988e-06, + "loss": 2.5663, + "step": 20993 + }, + { + "epoch": 2.2201776649746194, + "grad_norm": 8.5, + "learning_rate": 3.0696032689081606e-06, + "loss": 2.654, + "step": 20994 + }, + { + "epoch": 2.2202834179357023, + "grad_norm": 8.25, + "learning_rate": 3.069436515682419e-06, + "loss": 2.6114, + "step": 20995 + }, + { + "epoch": 2.220389170896785, + "grad_norm": 8.3125, + "learning_rate": 3.069269759784543e-06, + "loss": 2.538, + "step": 20996 + }, + { + "epoch": 2.220494923857868, + "grad_norm": 8.5, + "learning_rate": 3.069103001215319e-06, + "loss": 2.5005, + "step": 20997 + }, + { + "epoch": 2.220600676818951, + "grad_norm": 8.625, + "learning_rate": 3.0689362399755255e-06, + "loss": 2.6865, + "step": 20998 + }, + { + "epoch": 2.2207064297800336, + "grad_norm": 8.3125, + "learning_rate": 3.0687694760659487e-06, + "loss": 2.3375, + "step": 20999 + }, + { + "epoch": 2.220812182741117, + "grad_norm": 8.75, + "learning_rate": 3.0686027094873676e-06, + "loss": 2.743, + "step": 21000 + }, + { + "epoch": 2.2209179357021998, + "grad_norm": 8.375, + "learning_rate": 3.0684359402405676e-06, + "loss": 2.6988, + "step": 21001 + }, + { + "epoch": 2.2210236886632826, + "grad_norm": 8.125, + "learning_rate": 3.0682691683263303e-06, + "loss": 2.5091, + "step": 21002 + }, + { + "epoch": 2.2211294416243654, + "grad_norm": 8.3125, + "learning_rate": 3.068102393745438e-06, + "loss": 2.4715, + "step": 21003 + }, + { + "epoch": 2.2212351945854483, + "grad_norm": 8.5625, + "learning_rate": 3.0679356164986736e-06, + "loss": 2.5753, + "step": 21004 + }, + { + "epoch": 2.221340947546531, + "grad_norm": 7.875, + "learning_rate": 3.06776883658682e-06, + "loss": 2.3659, + "step": 21005 + }, + { + "epoch": 2.2214467005076144, + "grad_norm": 8.25, + "learning_rate": 3.0676020540106583e-06, + "loss": 2.5009, + "step": 21006 + }, + { + "epoch": 2.2215524534686972, + "grad_norm": 8.25, + "learning_rate": 3.0674352687709724e-06, + "loss": 2.5656, + "step": 21007 + }, + { + "epoch": 2.22165820642978, + "grad_norm": 8.375, + "learning_rate": 3.0672684808685453e-06, + "loss": 2.7434, + "step": 21008 + }, + { + "epoch": 2.221763959390863, + "grad_norm": 8.8125, + "learning_rate": 3.06710169030416e-06, + "loss": 2.6942, + "step": 21009 + }, + { + "epoch": 2.2218697123519457, + "grad_norm": 8.625, + "learning_rate": 3.066934897078597e-06, + "loss": 2.4608, + "step": 21010 + }, + { + "epoch": 2.2219754653130286, + "grad_norm": 8.625, + "learning_rate": 3.0667681011926414e-06, + "loss": 2.6518, + "step": 21011 + }, + { + "epoch": 2.222081218274112, + "grad_norm": 8.125, + "learning_rate": 3.0666013026470746e-06, + "loss": 2.4723, + "step": 21012 + }, + { + "epoch": 2.2221869712351947, + "grad_norm": 8.0625, + "learning_rate": 3.0664345014426794e-06, + "loss": 2.5281, + "step": 21013 + }, + { + "epoch": 2.2222927241962775, + "grad_norm": 8.4375, + "learning_rate": 3.06626769758024e-06, + "loss": 2.6774, + "step": 21014 + }, + { + "epoch": 2.2223984771573604, + "grad_norm": 8.3125, + "learning_rate": 3.0661008910605363e-06, + "loss": 2.6043, + "step": 21015 + }, + { + "epoch": 2.2225042301184432, + "grad_norm": 8.75, + "learning_rate": 3.0659340818843524e-06, + "loss": 2.6407, + "step": 21016 + }, + { + "epoch": 2.222609983079526, + "grad_norm": 8.1875, + "learning_rate": 3.065767270052473e-06, + "loss": 2.5042, + "step": 21017 + }, + { + "epoch": 2.222715736040609, + "grad_norm": 8.3125, + "learning_rate": 3.0656004555656782e-06, + "loss": 2.5184, + "step": 21018 + }, + { + "epoch": 2.222821489001692, + "grad_norm": 9.0, + "learning_rate": 3.065433638424751e-06, + "loss": 2.5145, + "step": 21019 + }, + { + "epoch": 2.222927241962775, + "grad_norm": 8.1875, + "learning_rate": 3.065266818630476e-06, + "loss": 2.4678, + "step": 21020 + }, + { + "epoch": 2.223032994923858, + "grad_norm": 8.125, + "learning_rate": 3.065099996183636e-06, + "loss": 2.5168, + "step": 21021 + }, + { + "epoch": 2.2231387478849407, + "grad_norm": 8.75, + "learning_rate": 3.064933171085011e-06, + "loss": 2.2102, + "step": 21022 + }, + { + "epoch": 2.2232445008460235, + "grad_norm": 7.5625, + "learning_rate": 3.0647663433353868e-06, + "loss": 2.3135, + "step": 21023 + }, + { + "epoch": 2.223350253807107, + "grad_norm": 8.375, + "learning_rate": 3.0645995129355443e-06, + "loss": 2.3515, + "step": 21024 + }, + { + "epoch": 2.2234560067681897, + "grad_norm": 8.6875, + "learning_rate": 3.064432679886268e-06, + "loss": 2.633, + "step": 21025 + }, + { + "epoch": 2.2235617597292725, + "grad_norm": 8.9375, + "learning_rate": 3.064265844188339e-06, + "loss": 2.594, + "step": 21026 + }, + { + "epoch": 2.2236675126903553, + "grad_norm": 8.5, + "learning_rate": 3.0640990058425425e-06, + "loss": 2.4155, + "step": 21027 + }, + { + "epoch": 2.223773265651438, + "grad_norm": 8.5, + "learning_rate": 3.063932164849659e-06, + "loss": 2.7334, + "step": 21028 + }, + { + "epoch": 2.223879018612521, + "grad_norm": 8.5625, + "learning_rate": 3.063765321210473e-06, + "loss": 2.4875, + "step": 21029 + }, + { + "epoch": 2.223984771573604, + "grad_norm": 8.5, + "learning_rate": 3.0635984749257667e-06, + "loss": 2.5302, + "step": 21030 + }, + { + "epoch": 2.224090524534687, + "grad_norm": 8.625, + "learning_rate": 3.063431625996324e-06, + "loss": 2.5178, + "step": 21031 + }, + { + "epoch": 2.22419627749577, + "grad_norm": 8.5, + "learning_rate": 3.0632647744229255e-06, + "loss": 2.7302, + "step": 21032 + }, + { + "epoch": 2.224302030456853, + "grad_norm": 8.625, + "learning_rate": 3.0630979202063567e-06, + "loss": 2.4973, + "step": 21033 + }, + { + "epoch": 2.2244077834179357, + "grad_norm": 7.9375, + "learning_rate": 3.0629310633473996e-06, + "loss": 2.3867, + "step": 21034 + }, + { + "epoch": 2.2245135363790185, + "grad_norm": 8.5625, + "learning_rate": 3.062764203846837e-06, + "loss": 2.7291, + "step": 21035 + }, + { + "epoch": 2.2246192893401013, + "grad_norm": 8.0625, + "learning_rate": 3.0625973417054528e-06, + "loss": 2.5693, + "step": 21036 + }, + { + "epoch": 2.2247250423011846, + "grad_norm": 8.625, + "learning_rate": 3.062430476924029e-06, + "loss": 2.3811, + "step": 21037 + }, + { + "epoch": 2.2248307952622675, + "grad_norm": 9.375, + "learning_rate": 3.0622636095033492e-06, + "loss": 2.5922, + "step": 21038 + }, + { + "epoch": 2.2249365482233503, + "grad_norm": 8.375, + "learning_rate": 3.0620967394441964e-06, + "loss": 2.7217, + "step": 21039 + }, + { + "epoch": 2.225042301184433, + "grad_norm": 8.9375, + "learning_rate": 3.061929866747353e-06, + "loss": 2.2103, + "step": 21040 + }, + { + "epoch": 2.225148054145516, + "grad_norm": 8.5625, + "learning_rate": 3.0617629914136025e-06, + "loss": 2.3588, + "step": 21041 + }, + { + "epoch": 2.225253807106599, + "grad_norm": 8.1875, + "learning_rate": 3.061596113443728e-06, + "loss": 2.6052, + "step": 21042 + }, + { + "epoch": 2.225359560067682, + "grad_norm": 8.125, + "learning_rate": 3.0614292328385133e-06, + "loss": 2.4454, + "step": 21043 + }, + { + "epoch": 2.225465313028765, + "grad_norm": 8.5, + "learning_rate": 3.0612623495987404e-06, + "loss": 2.5741, + "step": 21044 + }, + { + "epoch": 2.2255710659898478, + "grad_norm": 8.125, + "learning_rate": 3.0610954637251927e-06, + "loss": 2.5309, + "step": 21045 + }, + { + "epoch": 2.2256768189509306, + "grad_norm": 8.875, + "learning_rate": 3.060928575218654e-06, + "loss": 2.4708, + "step": 21046 + }, + { + "epoch": 2.2257825719120135, + "grad_norm": 8.0625, + "learning_rate": 3.0607616840799065e-06, + "loss": 2.5498, + "step": 21047 + }, + { + "epoch": 2.2258883248730963, + "grad_norm": 8.375, + "learning_rate": 3.060594790309734e-06, + "loss": 2.4271, + "step": 21048 + }, + { + "epoch": 2.225994077834179, + "grad_norm": 8.6875, + "learning_rate": 3.0604278939089195e-06, + "loss": 2.7432, + "step": 21049 + }, + { + "epoch": 2.2260998307952624, + "grad_norm": 8.5625, + "learning_rate": 3.060260994878245e-06, + "loss": 2.6696, + "step": 21050 + }, + { + "epoch": 2.2262055837563453, + "grad_norm": 8.8125, + "learning_rate": 3.0600940932184953e-06, + "loss": 2.6357, + "step": 21051 + }, + { + "epoch": 2.226311336717428, + "grad_norm": 8.4375, + "learning_rate": 3.059927188930454e-06, + "loss": 2.7785, + "step": 21052 + }, + { + "epoch": 2.226417089678511, + "grad_norm": 8.5625, + "learning_rate": 3.059760282014903e-06, + "loss": 2.5912, + "step": 21053 + }, + { + "epoch": 2.2265228426395938, + "grad_norm": 9.0625, + "learning_rate": 3.0595933724726253e-06, + "loss": 2.8377, + "step": 21054 + }, + { + "epoch": 2.226628595600677, + "grad_norm": 8.375, + "learning_rate": 3.0594264603044056e-06, + "loss": 2.3977, + "step": 21055 + }, + { + "epoch": 2.22673434856176, + "grad_norm": 8.4375, + "learning_rate": 3.0592595455110258e-06, + "loss": 2.5875, + "step": 21056 + }, + { + "epoch": 2.2268401015228427, + "grad_norm": 8.875, + "learning_rate": 3.0590926280932698e-06, + "loss": 2.6242, + "step": 21057 + }, + { + "epoch": 2.2269458544839256, + "grad_norm": 8.3125, + "learning_rate": 3.0589257080519207e-06, + "loss": 2.518, + "step": 21058 + }, + { + "epoch": 2.2270516074450084, + "grad_norm": 8.375, + "learning_rate": 3.058758785387762e-06, + "loss": 2.5612, + "step": 21059 + }, + { + "epoch": 2.2271573604060912, + "grad_norm": 8.5, + "learning_rate": 3.058591860101577e-06, + "loss": 2.5175, + "step": 21060 + }, + { + "epoch": 2.227263113367174, + "grad_norm": 8.375, + "learning_rate": 3.0584249321941476e-06, + "loss": 2.6097, + "step": 21061 + }, + { + "epoch": 2.2273688663282574, + "grad_norm": 8.125, + "learning_rate": 3.05825800166626e-06, + "loss": 2.5448, + "step": 21062 + }, + { + "epoch": 2.22747461928934, + "grad_norm": 9.4375, + "learning_rate": 3.0580910685186952e-06, + "loss": 2.4914, + "step": 21063 + }, + { + "epoch": 2.227580372250423, + "grad_norm": 7.875, + "learning_rate": 3.0579241327522375e-06, + "loss": 2.4726, + "step": 21064 + }, + { + "epoch": 2.227686125211506, + "grad_norm": 8.5625, + "learning_rate": 3.05775719436767e-06, + "loss": 2.682, + "step": 21065 + }, + { + "epoch": 2.2277918781725887, + "grad_norm": 9.25, + "learning_rate": 3.0575902533657762e-06, + "loss": 2.7624, + "step": 21066 + }, + { + "epoch": 2.2278976311336716, + "grad_norm": 8.625, + "learning_rate": 3.0574233097473388e-06, + "loss": 2.7028, + "step": 21067 + }, + { + "epoch": 2.228003384094755, + "grad_norm": 8.625, + "learning_rate": 3.0572563635131424e-06, + "loss": 2.5931, + "step": 21068 + }, + { + "epoch": 2.2281091370558377, + "grad_norm": 8.6875, + "learning_rate": 3.0570894146639695e-06, + "loss": 2.5563, + "step": 21069 + }, + { + "epoch": 2.2282148900169205, + "grad_norm": 8.1875, + "learning_rate": 3.0569224632006034e-06, + "loss": 2.6282, + "step": 21070 + }, + { + "epoch": 2.2283206429780034, + "grad_norm": 8.5625, + "learning_rate": 3.056755509123829e-06, + "loss": 2.7727, + "step": 21071 + }, + { + "epoch": 2.228426395939086, + "grad_norm": 7.96875, + "learning_rate": 3.0565885524344284e-06, + "loss": 2.546, + "step": 21072 + }, + { + "epoch": 2.228532148900169, + "grad_norm": 8.0625, + "learning_rate": 3.056421593133185e-06, + "loss": 2.6343, + "step": 21073 + }, + { + "epoch": 2.2286379018612523, + "grad_norm": 8.375, + "learning_rate": 3.056254631220882e-06, + "loss": 2.6332, + "step": 21074 + }, + { + "epoch": 2.228743654822335, + "grad_norm": 8.625, + "learning_rate": 3.056087666698305e-06, + "loss": 2.5585, + "step": 21075 + }, + { + "epoch": 2.228849407783418, + "grad_norm": 8.25, + "learning_rate": 3.055920699566235e-06, + "loss": 2.4719, + "step": 21076 + }, + { + "epoch": 2.228955160744501, + "grad_norm": 8.4375, + "learning_rate": 3.0557537298254563e-06, + "loss": 2.4779, + "step": 21077 + }, + { + "epoch": 2.2290609137055837, + "grad_norm": 8.125, + "learning_rate": 3.055586757476754e-06, + "loss": 2.4049, + "step": 21078 + }, + { + "epoch": 2.2291666666666665, + "grad_norm": 8.5, + "learning_rate": 3.055419782520909e-06, + "loss": 2.5322, + "step": 21079 + }, + { + "epoch": 2.2292724196277494, + "grad_norm": 8.375, + "learning_rate": 3.0552528049587067e-06, + "loss": 2.625, + "step": 21080 + }, + { + "epoch": 2.2293781725888326, + "grad_norm": 8.25, + "learning_rate": 3.0550858247909297e-06, + "loss": 2.7002, + "step": 21081 + }, + { + "epoch": 2.2294839255499155, + "grad_norm": 8.25, + "learning_rate": 3.054918842018363e-06, + "loss": 2.411, + "step": 21082 + }, + { + "epoch": 2.2295896785109983, + "grad_norm": 8.5, + "learning_rate": 3.0547518566417878e-06, + "loss": 2.5506, + "step": 21083 + }, + { + "epoch": 2.229695431472081, + "grad_norm": 8.875, + "learning_rate": 3.0545848686619894e-06, + "loss": 2.4993, + "step": 21084 + }, + { + "epoch": 2.229801184433164, + "grad_norm": 8.5, + "learning_rate": 3.0544178780797517e-06, + "loss": 2.547, + "step": 21085 + }, + { + "epoch": 2.229906937394247, + "grad_norm": 8.4375, + "learning_rate": 3.054250884895856e-06, + "loss": 2.436, + "step": 21086 + }, + { + "epoch": 2.23001269035533, + "grad_norm": 8.375, + "learning_rate": 3.0540838891110892e-06, + "loss": 2.4088, + "step": 21087 + }, + { + "epoch": 2.230118443316413, + "grad_norm": 8.625, + "learning_rate": 3.0539168907262327e-06, + "loss": 2.4864, + "step": 21088 + }, + { + "epoch": 2.230224196277496, + "grad_norm": 7.9375, + "learning_rate": 3.0537498897420704e-06, + "loss": 2.4885, + "step": 21089 + }, + { + "epoch": 2.2303299492385786, + "grad_norm": 8.8125, + "learning_rate": 3.0535828861593873e-06, + "loss": 2.376, + "step": 21090 + }, + { + "epoch": 2.2304357021996615, + "grad_norm": 10.4375, + "learning_rate": 3.053415879978966e-06, + "loss": 2.7079, + "step": 21091 + }, + { + "epoch": 2.2305414551607443, + "grad_norm": 8.5625, + "learning_rate": 3.0532488712015894e-06, + "loss": 2.5925, + "step": 21092 + }, + { + "epoch": 2.2306472081218276, + "grad_norm": 7.90625, + "learning_rate": 3.0530818598280425e-06, + "loss": 2.3317, + "step": 21093 + }, + { + "epoch": 2.2307529610829104, + "grad_norm": 8.4375, + "learning_rate": 3.0529148458591095e-06, + "loss": 2.5856, + "step": 21094 + }, + { + "epoch": 2.2308587140439933, + "grad_norm": 8.875, + "learning_rate": 3.052747829295572e-06, + "loss": 2.5087, + "step": 21095 + }, + { + "epoch": 2.230964467005076, + "grad_norm": 8.375, + "learning_rate": 3.0525808101382153e-06, + "loss": 2.562, + "step": 21096 + }, + { + "epoch": 2.231070219966159, + "grad_norm": 8.3125, + "learning_rate": 3.0524137883878236e-06, + "loss": 2.5988, + "step": 21097 + }, + { + "epoch": 2.231175972927242, + "grad_norm": 8.125, + "learning_rate": 3.052246764045179e-06, + "loss": 2.5684, + "step": 21098 + }, + { + "epoch": 2.231281725888325, + "grad_norm": 9.0625, + "learning_rate": 3.052079737111067e-06, + "loss": 2.5766, + "step": 21099 + }, + { + "epoch": 2.231387478849408, + "grad_norm": 8.125, + "learning_rate": 3.0519127075862705e-06, + "loss": 2.5694, + "step": 21100 + }, + { + "epoch": 2.2314932318104908, + "grad_norm": 8.0625, + "learning_rate": 3.051745675471573e-06, + "loss": 2.5317, + "step": 21101 + }, + { + "epoch": 2.2315989847715736, + "grad_norm": 12.0625, + "learning_rate": 3.0515786407677584e-06, + "loss": 2.5449, + "step": 21102 + }, + { + "epoch": 2.2317047377326564, + "grad_norm": 8.6875, + "learning_rate": 3.0514116034756115e-06, + "loss": 2.0398, + "step": 21103 + }, + { + "epoch": 2.2318104906937393, + "grad_norm": 8.9375, + "learning_rate": 3.051244563595915e-06, + "loss": 2.547, + "step": 21104 + }, + { + "epoch": 2.2319162436548226, + "grad_norm": 7.8125, + "learning_rate": 3.051077521129453e-06, + "loss": 2.4032, + "step": 21105 + }, + { + "epoch": 2.2320219966159054, + "grad_norm": 8.9375, + "learning_rate": 3.0509104760770103e-06, + "loss": 2.7505, + "step": 21106 + }, + { + "epoch": 2.2321277495769882, + "grad_norm": 8.25, + "learning_rate": 3.05074342843937e-06, + "loss": 2.6471, + "step": 21107 + }, + { + "epoch": 2.232233502538071, + "grad_norm": 8.25, + "learning_rate": 3.0505763782173153e-06, + "loss": 2.436, + "step": 21108 + }, + { + "epoch": 2.232339255499154, + "grad_norm": 7.875, + "learning_rate": 3.0504093254116316e-06, + "loss": 2.5535, + "step": 21109 + }, + { + "epoch": 2.2324450084602367, + "grad_norm": 8.4375, + "learning_rate": 3.0502422700231022e-06, + "loss": 2.4912, + "step": 21110 + }, + { + "epoch": 2.2325507614213196, + "grad_norm": 8.125, + "learning_rate": 3.0500752120525102e-06, + "loss": 2.5966, + "step": 21111 + }, + { + "epoch": 2.232656514382403, + "grad_norm": 8.3125, + "learning_rate": 3.0499081515006408e-06, + "loss": 2.6182, + "step": 21112 + }, + { + "epoch": 2.2327622673434857, + "grad_norm": 8.9375, + "learning_rate": 3.0497410883682766e-06, + "loss": 2.7145, + "step": 21113 + }, + { + "epoch": 2.2328680203045685, + "grad_norm": 9.0, + "learning_rate": 3.0495740226562028e-06, + "loss": 2.8096, + "step": 21114 + }, + { + "epoch": 2.2329737732656514, + "grad_norm": 8.375, + "learning_rate": 3.049406954365202e-06, + "loss": 2.223, + "step": 21115 + }, + { + "epoch": 2.2330795262267342, + "grad_norm": 8.625, + "learning_rate": 3.04923988349606e-06, + "loss": 2.4938, + "step": 21116 + }, + { + "epoch": 2.233185279187817, + "grad_norm": 7.96875, + "learning_rate": 3.0490728100495603e-06, + "loss": 2.5787, + "step": 21117 + }, + { + "epoch": 2.2332910321489003, + "grad_norm": 8.125, + "learning_rate": 3.0489057340264854e-06, + "loss": 2.0968, + "step": 21118 + }, + { + "epoch": 2.233396785109983, + "grad_norm": 8.6875, + "learning_rate": 3.048738655427621e-06, + "loss": 2.6973, + "step": 21119 + }, + { + "epoch": 2.233502538071066, + "grad_norm": 8.25, + "learning_rate": 3.0485715742537507e-06, + "loss": 2.3839, + "step": 21120 + }, + { + "epoch": 2.233608291032149, + "grad_norm": 8.25, + "learning_rate": 3.0484044905056574e-06, + "loss": 2.5604, + "step": 21121 + }, + { + "epoch": 2.2337140439932317, + "grad_norm": 9.0625, + "learning_rate": 3.048237404184127e-06, + "loss": 2.7314, + "step": 21122 + }, + { + "epoch": 2.2338197969543145, + "grad_norm": 8.1875, + "learning_rate": 3.048070315289943e-06, + "loss": 2.5182, + "step": 21123 + }, + { + "epoch": 2.233925549915398, + "grad_norm": 8.125, + "learning_rate": 3.0479032238238878e-06, + "loss": 2.5226, + "step": 21124 + }, + { + "epoch": 2.2340313028764807, + "grad_norm": 8.875, + "learning_rate": 3.047736129786748e-06, + "loss": 2.496, + "step": 21125 + }, + { + "epoch": 2.2341370558375635, + "grad_norm": 8.5625, + "learning_rate": 3.0475690331793063e-06, + "loss": 2.6557, + "step": 21126 + }, + { + "epoch": 2.2342428087986463, + "grad_norm": 8.25, + "learning_rate": 3.0474019340023466e-06, + "loss": 2.6623, + "step": 21127 + }, + { + "epoch": 2.234348561759729, + "grad_norm": 9.0625, + "learning_rate": 3.0472348322566543e-06, + "loss": 2.6204, + "step": 21128 + }, + { + "epoch": 2.234454314720812, + "grad_norm": 7.96875, + "learning_rate": 3.0470677279430123e-06, + "loss": 2.4824, + "step": 21129 + }, + { + "epoch": 2.2345600676818953, + "grad_norm": 8.5, + "learning_rate": 3.046900621062205e-06, + "loss": 2.663, + "step": 21130 + }, + { + "epoch": 2.234665820642978, + "grad_norm": 7.875, + "learning_rate": 3.0467335116150163e-06, + "loss": 2.4035, + "step": 21131 + }, + { + "epoch": 2.234771573604061, + "grad_norm": 9.0, + "learning_rate": 3.046566399602232e-06, + "loss": 2.5848, + "step": 21132 + }, + { + "epoch": 2.234877326565144, + "grad_norm": 8.4375, + "learning_rate": 3.0463992850246347e-06, + "loss": 2.4165, + "step": 21133 + }, + { + "epoch": 2.2349830795262267, + "grad_norm": 8.3125, + "learning_rate": 3.0462321678830086e-06, + "loss": 2.7399, + "step": 21134 + }, + { + "epoch": 2.2350888324873095, + "grad_norm": 8.5, + "learning_rate": 3.046065048178139e-06, + "loss": 2.8873, + "step": 21135 + }, + { + "epoch": 2.235194585448393, + "grad_norm": 8.0, + "learning_rate": 3.0458979259108094e-06, + "loss": 2.4569, + "step": 21136 + }, + { + "epoch": 2.2353003384094756, + "grad_norm": 9.0, + "learning_rate": 3.0457308010818036e-06, + "loss": 2.4745, + "step": 21137 + }, + { + "epoch": 2.2354060913705585, + "grad_norm": 8.4375, + "learning_rate": 3.045563673691907e-06, + "loss": 2.3966, + "step": 21138 + }, + { + "epoch": 2.2355118443316413, + "grad_norm": 8.8125, + "learning_rate": 3.0453965437419026e-06, + "loss": 2.5557, + "step": 21139 + }, + { + "epoch": 2.235617597292724, + "grad_norm": 8.5625, + "learning_rate": 3.0452294112325746e-06, + "loss": 2.5022, + "step": 21140 + }, + { + "epoch": 2.235723350253807, + "grad_norm": 8.6875, + "learning_rate": 3.0450622761647096e-06, + "loss": 2.4194, + "step": 21141 + }, + { + "epoch": 2.23582910321489, + "grad_norm": 8.1875, + "learning_rate": 3.0448951385390897e-06, + "loss": 2.4937, + "step": 21142 + }, + { + "epoch": 2.235934856175973, + "grad_norm": 8.25, + "learning_rate": 3.044727998356499e-06, + "loss": 2.4951, + "step": 21143 + }, + { + "epoch": 2.236040609137056, + "grad_norm": 8.625, + "learning_rate": 3.044560855617724e-06, + "loss": 2.4852, + "step": 21144 + }, + { + "epoch": 2.2361463620981388, + "grad_norm": 8.375, + "learning_rate": 3.0443937103235465e-06, + "loss": 2.4541, + "step": 21145 + }, + { + "epoch": 2.2362521150592216, + "grad_norm": 9.375, + "learning_rate": 3.044226562474752e-06, + "loss": 2.5852, + "step": 21146 + }, + { + "epoch": 2.2363578680203045, + "grad_norm": 8.0, + "learning_rate": 3.0440594120721256e-06, + "loss": 2.3578, + "step": 21147 + }, + { + "epoch": 2.2364636209813873, + "grad_norm": 8.25, + "learning_rate": 3.04389225911645e-06, + "loss": 2.5396, + "step": 21148 + }, + { + "epoch": 2.2365693739424706, + "grad_norm": 7.96875, + "learning_rate": 3.043725103608511e-06, + "loss": 2.4825, + "step": 21149 + }, + { + "epoch": 2.2366751269035534, + "grad_norm": 9.0625, + "learning_rate": 3.0435579455490923e-06, + "loss": 2.6657, + "step": 21150 + }, + { + "epoch": 2.2367808798646363, + "grad_norm": 8.0625, + "learning_rate": 3.0433907849389783e-06, + "loss": 2.5582, + "step": 21151 + }, + { + "epoch": 2.236886632825719, + "grad_norm": 9.0625, + "learning_rate": 3.043223621778954e-06, + "loss": 2.6437, + "step": 21152 + }, + { + "epoch": 2.236992385786802, + "grad_norm": 8.5625, + "learning_rate": 3.043056456069803e-06, + "loss": 2.2805, + "step": 21153 + }, + { + "epoch": 2.2370981387478848, + "grad_norm": 8.375, + "learning_rate": 3.0428892878123104e-06, + "loss": 2.515, + "step": 21154 + }, + { + "epoch": 2.237203891708968, + "grad_norm": 8.4375, + "learning_rate": 3.0427221170072605e-06, + "loss": 2.521, + "step": 21155 + }, + { + "epoch": 2.237309644670051, + "grad_norm": 8.3125, + "learning_rate": 3.0425549436554375e-06, + "loss": 2.5666, + "step": 21156 + }, + { + "epoch": 2.2374153976311337, + "grad_norm": 8.3125, + "learning_rate": 3.0423877677576257e-06, + "loss": 2.5824, + "step": 21157 + }, + { + "epoch": 2.2375211505922166, + "grad_norm": 8.625, + "learning_rate": 3.042220589314611e-06, + "loss": 2.6555, + "step": 21158 + }, + { + "epoch": 2.2376269035532994, + "grad_norm": 8.875, + "learning_rate": 3.0420534083271754e-06, + "loss": 2.5919, + "step": 21159 + }, + { + "epoch": 2.2377326565143822, + "grad_norm": 8.375, + "learning_rate": 3.0418862247961056e-06, + "loss": 2.5353, + "step": 21160 + }, + { + "epoch": 2.237838409475465, + "grad_norm": 8.4375, + "learning_rate": 3.0417190387221856e-06, + "loss": 2.8069, + "step": 21161 + }, + { + "epoch": 2.2379441624365484, + "grad_norm": 8.4375, + "learning_rate": 3.041551850106199e-06, + "loss": 2.5024, + "step": 21162 + }, + { + "epoch": 2.238049915397631, + "grad_norm": 8.125, + "learning_rate": 3.041384658948932e-06, + "loss": 2.6095, + "step": 21163 + }, + { + "epoch": 2.238155668358714, + "grad_norm": 8.5, + "learning_rate": 3.0412174652511676e-06, + "loss": 2.3248, + "step": 21164 + }, + { + "epoch": 2.238261421319797, + "grad_norm": 8.1875, + "learning_rate": 3.0410502690136913e-06, + "loss": 2.566, + "step": 21165 + }, + { + "epoch": 2.2383671742808797, + "grad_norm": 8.4375, + "learning_rate": 3.0408830702372867e-06, + "loss": 2.1959, + "step": 21166 + }, + { + "epoch": 2.238472927241963, + "grad_norm": 7.96875, + "learning_rate": 3.0407158689227394e-06, + "loss": 2.4192, + "step": 21167 + }, + { + "epoch": 2.238578680203046, + "grad_norm": 9.4375, + "learning_rate": 3.040548665070834e-06, + "loss": 2.6306, + "step": 21168 + }, + { + "epoch": 2.2386844331641287, + "grad_norm": 8.6875, + "learning_rate": 3.0403814586823545e-06, + "loss": 2.5687, + "step": 21169 + }, + { + "epoch": 2.2387901861252115, + "grad_norm": 8.5625, + "learning_rate": 3.040214249758086e-06, + "loss": 2.5256, + "step": 21170 + }, + { + "epoch": 2.2388959390862944, + "grad_norm": 8.875, + "learning_rate": 3.0400470382988136e-06, + "loss": 2.7962, + "step": 21171 + }, + { + "epoch": 2.239001692047377, + "grad_norm": 8.375, + "learning_rate": 3.03987982430532e-06, + "loss": 2.49, + "step": 21172 + }, + { + "epoch": 2.23910744500846, + "grad_norm": 8.0625, + "learning_rate": 3.0397126077783923e-06, + "loss": 2.5442, + "step": 21173 + }, + { + "epoch": 2.2392131979695433, + "grad_norm": 7.9375, + "learning_rate": 3.0395453887188135e-06, + "loss": 2.5649, + "step": 21174 + }, + { + "epoch": 2.239318950930626, + "grad_norm": 7.90625, + "learning_rate": 3.039378167127368e-06, + "loss": 2.5324, + "step": 21175 + }, + { + "epoch": 2.239424703891709, + "grad_norm": 8.3125, + "learning_rate": 3.039210943004843e-06, + "loss": 2.4663, + "step": 21176 + }, + { + "epoch": 2.239530456852792, + "grad_norm": 8.0625, + "learning_rate": 3.0390437163520208e-06, + "loss": 2.4589, + "step": 21177 + }, + { + "epoch": 2.2396362098138747, + "grad_norm": 8.1875, + "learning_rate": 3.0388764871696867e-06, + "loss": 2.3828, + "step": 21178 + }, + { + "epoch": 2.2397419627749575, + "grad_norm": 8.375, + "learning_rate": 3.038709255458626e-06, + "loss": 2.6777, + "step": 21179 + }, + { + "epoch": 2.239847715736041, + "grad_norm": 8.25, + "learning_rate": 3.038542021219623e-06, + "loss": 2.5949, + "step": 21180 + }, + { + "epoch": 2.2399534686971236, + "grad_norm": 8.6875, + "learning_rate": 3.038374784453463e-06, + "loss": 2.7971, + "step": 21181 + }, + { + "epoch": 2.2400592216582065, + "grad_norm": 9.25, + "learning_rate": 3.0382075451609296e-06, + "loss": 2.6167, + "step": 21182 + }, + { + "epoch": 2.2401649746192893, + "grad_norm": 8.375, + "learning_rate": 3.0380403033428092e-06, + "loss": 2.5295, + "step": 21183 + }, + { + "epoch": 2.240270727580372, + "grad_norm": 8.625, + "learning_rate": 3.0378730589998843e-06, + "loss": 2.4879, + "step": 21184 + }, + { + "epoch": 2.240376480541455, + "grad_norm": 9.0, + "learning_rate": 3.037705812132942e-06, + "loss": 2.7622, + "step": 21185 + }, + { + "epoch": 2.2404822335025383, + "grad_norm": 8.0625, + "learning_rate": 3.037538562742766e-06, + "loss": 2.4063, + "step": 21186 + }, + { + "epoch": 2.240587986463621, + "grad_norm": 9.0625, + "learning_rate": 3.0373713108301417e-06, + "loss": 2.4695, + "step": 21187 + }, + { + "epoch": 2.240693739424704, + "grad_norm": 8.3125, + "learning_rate": 3.0372040563958534e-06, + "loss": 2.5925, + "step": 21188 + }, + { + "epoch": 2.240799492385787, + "grad_norm": 8.375, + "learning_rate": 3.0370367994406863e-06, + "loss": 2.537, + "step": 21189 + }, + { + "epoch": 2.2409052453468696, + "grad_norm": 8.1875, + "learning_rate": 3.036869539965425e-06, + "loss": 2.5069, + "step": 21190 + }, + { + "epoch": 2.2410109983079525, + "grad_norm": 8.5625, + "learning_rate": 3.0367022779708545e-06, + "loss": 2.5663, + "step": 21191 + }, + { + "epoch": 2.2411167512690353, + "grad_norm": 8.6875, + "learning_rate": 3.0365350134577594e-06, + "loss": 2.4969, + "step": 21192 + }, + { + "epoch": 2.2412225042301186, + "grad_norm": 8.25, + "learning_rate": 3.0363677464269253e-06, + "loss": 2.5017, + "step": 21193 + }, + { + "epoch": 2.2413282571912014, + "grad_norm": 8.3125, + "learning_rate": 3.0362004768791366e-06, + "loss": 2.371, + "step": 21194 + }, + { + "epoch": 2.2414340101522843, + "grad_norm": 8.375, + "learning_rate": 3.0360332048151786e-06, + "loss": 2.4699, + "step": 21195 + }, + { + "epoch": 2.241539763113367, + "grad_norm": 7.96875, + "learning_rate": 3.0358659302358363e-06, + "loss": 2.541, + "step": 21196 + }, + { + "epoch": 2.24164551607445, + "grad_norm": 9.1875, + "learning_rate": 3.0356986531418938e-06, + "loss": 2.848, + "step": 21197 + }, + { + "epoch": 2.241751269035533, + "grad_norm": 9.125, + "learning_rate": 3.035531373534137e-06, + "loss": 2.649, + "step": 21198 + }, + { + "epoch": 2.241857021996616, + "grad_norm": 9.25, + "learning_rate": 3.0353640914133504e-06, + "loss": 2.5192, + "step": 21199 + }, + { + "epoch": 2.241962774957699, + "grad_norm": 8.5625, + "learning_rate": 3.0351968067803183e-06, + "loss": 2.6153, + "step": 21200 + }, + { + "epoch": 2.2420685279187818, + "grad_norm": 8.0625, + "learning_rate": 3.035029519635827e-06, + "loss": 2.3336, + "step": 21201 + }, + { + "epoch": 2.2421742808798646, + "grad_norm": 8.4375, + "learning_rate": 3.0348622299806613e-06, + "loss": 2.5559, + "step": 21202 + }, + { + "epoch": 2.2422800338409474, + "grad_norm": 8.625, + "learning_rate": 3.034694937815606e-06, + "loss": 2.8263, + "step": 21203 + }, + { + "epoch": 2.2423857868020303, + "grad_norm": 8.0625, + "learning_rate": 3.034527643141445e-06, + "loss": 2.5569, + "step": 21204 + }, + { + "epoch": 2.2424915397631136, + "grad_norm": 8.1875, + "learning_rate": 3.0343603459589655e-06, + "loss": 2.5118, + "step": 21205 + }, + { + "epoch": 2.2425972927241964, + "grad_norm": 9.0625, + "learning_rate": 3.0341930462689512e-06, + "loss": 2.7013, + "step": 21206 + }, + { + "epoch": 2.2427030456852792, + "grad_norm": 8.9375, + "learning_rate": 3.034025744072187e-06, + "loss": 2.5293, + "step": 21207 + }, + { + "epoch": 2.242808798646362, + "grad_norm": 8.25, + "learning_rate": 3.0338584393694592e-06, + "loss": 2.5995, + "step": 21208 + }, + { + "epoch": 2.242914551607445, + "grad_norm": 8.5, + "learning_rate": 3.0336911321615515e-06, + "loss": 2.5619, + "step": 21209 + }, + { + "epoch": 2.2430203045685277, + "grad_norm": 8.1875, + "learning_rate": 3.0335238224492486e-06, + "loss": 2.5577, + "step": 21210 + }, + { + "epoch": 2.243126057529611, + "grad_norm": 8.5625, + "learning_rate": 3.0333565102333377e-06, + "loss": 2.608, + "step": 21211 + }, + { + "epoch": 2.243231810490694, + "grad_norm": 8.375, + "learning_rate": 3.033189195514603e-06, + "loss": 2.5986, + "step": 21212 + }, + { + "epoch": 2.2433375634517767, + "grad_norm": 8.3125, + "learning_rate": 3.0330218782938293e-06, + "loss": 2.6101, + "step": 21213 + }, + { + "epoch": 2.2434433164128595, + "grad_norm": 8.75, + "learning_rate": 3.032854558571802e-06, + "loss": 2.6631, + "step": 21214 + }, + { + "epoch": 2.2435490693739424, + "grad_norm": 8.0625, + "learning_rate": 3.0326872363493067e-06, + "loss": 2.4686, + "step": 21215 + }, + { + "epoch": 2.2436548223350252, + "grad_norm": 8.375, + "learning_rate": 3.0325199116271264e-06, + "loss": 2.3121, + "step": 21216 + }, + { + "epoch": 2.2437605752961085, + "grad_norm": 8.3125, + "learning_rate": 3.0323525844060497e-06, + "loss": 2.5029, + "step": 21217 + }, + { + "epoch": 2.2438663282571913, + "grad_norm": 8.625, + "learning_rate": 3.032185254686859e-06, + "loss": 2.7303, + "step": 21218 + }, + { + "epoch": 2.243972081218274, + "grad_norm": 8.6875, + "learning_rate": 3.0320179224703412e-06, + "loss": 2.5511, + "step": 21219 + }, + { + "epoch": 2.244077834179357, + "grad_norm": 8.875, + "learning_rate": 3.0318505877572802e-06, + "loss": 2.5389, + "step": 21220 + }, + { + "epoch": 2.24418358714044, + "grad_norm": 8.625, + "learning_rate": 3.0316832505484624e-06, + "loss": 2.5961, + "step": 21221 + }, + { + "epoch": 2.2442893401015227, + "grad_norm": 8.375, + "learning_rate": 3.0315159108446723e-06, + "loss": 2.5961, + "step": 21222 + }, + { + "epoch": 2.2443950930626055, + "grad_norm": 8.75, + "learning_rate": 3.0313485686466954e-06, + "loss": 2.5774, + "step": 21223 + }, + { + "epoch": 2.244500846023689, + "grad_norm": 8.25, + "learning_rate": 3.0311812239553173e-06, + "loss": 2.5857, + "step": 21224 + }, + { + "epoch": 2.2446065989847717, + "grad_norm": 7.96875, + "learning_rate": 3.031013876771323e-06, + "loss": 2.4814, + "step": 21225 + }, + { + "epoch": 2.2447123519458545, + "grad_norm": 9.0625, + "learning_rate": 3.030846527095497e-06, + "loss": 2.3942, + "step": 21226 + }, + { + "epoch": 2.2448181049069373, + "grad_norm": 8.3125, + "learning_rate": 3.0306791749286265e-06, + "loss": 2.4962, + "step": 21227 + }, + { + "epoch": 2.24492385786802, + "grad_norm": 8.25, + "learning_rate": 3.0305118202714945e-06, + "loss": 2.5927, + "step": 21228 + }, + { + "epoch": 2.245029610829103, + "grad_norm": 8.5, + "learning_rate": 3.0303444631248874e-06, + "loss": 2.5664, + "step": 21229 + }, + { + "epoch": 2.2451353637901863, + "grad_norm": 8.25, + "learning_rate": 3.0301771034895917e-06, + "loss": 2.6133, + "step": 21230 + }, + { + "epoch": 2.245241116751269, + "grad_norm": 8.5, + "learning_rate": 3.0300097413663916e-06, + "loss": 2.66, + "step": 21231 + }, + { + "epoch": 2.245346869712352, + "grad_norm": 8.5625, + "learning_rate": 3.029842376756072e-06, + "loss": 2.6006, + "step": 21232 + }, + { + "epoch": 2.245452622673435, + "grad_norm": 8.5, + "learning_rate": 3.0296750096594192e-06, + "loss": 2.8322, + "step": 21233 + }, + { + "epoch": 2.2455583756345177, + "grad_norm": 8.0625, + "learning_rate": 3.0295076400772182e-06, + "loss": 2.5278, + "step": 21234 + }, + { + "epoch": 2.2456641285956005, + "grad_norm": 8.625, + "learning_rate": 3.029340268010254e-06, + "loss": 2.3552, + "step": 21235 + }, + { + "epoch": 2.245769881556684, + "grad_norm": 7.9375, + "learning_rate": 3.029172893459313e-06, + "loss": 2.4864, + "step": 21236 + }, + { + "epoch": 2.2458756345177666, + "grad_norm": 8.0625, + "learning_rate": 3.029005516425179e-06, + "loss": 2.4297, + "step": 21237 + }, + { + "epoch": 2.2459813874788495, + "grad_norm": 8.5625, + "learning_rate": 3.0288381369086393e-06, + "loss": 2.6007, + "step": 21238 + }, + { + "epoch": 2.2460871404399323, + "grad_norm": 8.625, + "learning_rate": 3.028670754910478e-06, + "loss": 2.2909, + "step": 21239 + }, + { + "epoch": 2.246192893401015, + "grad_norm": 8.4375, + "learning_rate": 3.028503370431482e-06, + "loss": 2.578, + "step": 21240 + }, + { + "epoch": 2.246298646362098, + "grad_norm": 8.375, + "learning_rate": 3.0283359834724357e-06, + "loss": 2.451, + "step": 21241 + }, + { + "epoch": 2.2464043993231813, + "grad_norm": 8.3125, + "learning_rate": 3.028168594034124e-06, + "loss": 2.0638, + "step": 21242 + }, + { + "epoch": 2.246510152284264, + "grad_norm": 8.625, + "learning_rate": 3.028001202117333e-06, + "loss": 2.7882, + "step": 21243 + }, + { + "epoch": 2.246615905245347, + "grad_norm": 8.625, + "learning_rate": 3.027833807722849e-06, + "loss": 2.6152, + "step": 21244 + }, + { + "epoch": 2.2467216582064298, + "grad_norm": 8.375, + "learning_rate": 3.027666410851456e-06, + "loss": 1.8536, + "step": 21245 + }, + { + "epoch": 2.2468274111675126, + "grad_norm": 7.96875, + "learning_rate": 3.027499011503941e-06, + "loss": 2.4779, + "step": 21246 + }, + { + "epoch": 2.2469331641285955, + "grad_norm": 8.375, + "learning_rate": 3.027331609681089e-06, + "loss": 2.4945, + "step": 21247 + }, + { + "epoch": 2.2470389170896787, + "grad_norm": 8.4375, + "learning_rate": 3.027164205383685e-06, + "loss": 2.206, + "step": 21248 + }, + { + "epoch": 2.2471446700507616, + "grad_norm": 7.96875, + "learning_rate": 3.0269967986125154e-06, + "loss": 2.6309, + "step": 21249 + }, + { + "epoch": 2.2472504230118444, + "grad_norm": 8.8125, + "learning_rate": 3.0268293893683655e-06, + "loss": 2.5614, + "step": 21250 + }, + { + "epoch": 2.2473561759729273, + "grad_norm": 8.4375, + "learning_rate": 3.02666197765202e-06, + "loss": 2.4978, + "step": 21251 + }, + { + "epoch": 2.24746192893401, + "grad_norm": 8.125, + "learning_rate": 3.0264945634642656e-06, + "loss": 2.6478, + "step": 21252 + }, + { + "epoch": 2.247567681895093, + "grad_norm": 8.625, + "learning_rate": 3.0263271468058876e-06, + "loss": 2.6009, + "step": 21253 + }, + { + "epoch": 2.2476734348561758, + "grad_norm": 10.1875, + "learning_rate": 3.026159727677671e-06, + "loss": 2.568, + "step": 21254 + }, + { + "epoch": 2.247779187817259, + "grad_norm": 8.25, + "learning_rate": 3.0259923060804027e-06, + "loss": 2.6237, + "step": 21255 + }, + { + "epoch": 2.247884940778342, + "grad_norm": 8.6875, + "learning_rate": 3.0258248820148673e-06, + "loss": 2.476, + "step": 21256 + }, + { + "epoch": 2.2479906937394247, + "grad_norm": 8.5625, + "learning_rate": 3.0256574554818512e-06, + "loss": 2.5689, + "step": 21257 + }, + { + "epoch": 2.2480964467005076, + "grad_norm": 9.0, + "learning_rate": 3.025490026482139e-06, + "loss": 2.6576, + "step": 21258 + }, + { + "epoch": 2.2482021996615904, + "grad_norm": 8.25, + "learning_rate": 3.0253225950165175e-06, + "loss": 2.3766, + "step": 21259 + }, + { + "epoch": 2.2483079526226732, + "grad_norm": 8.5625, + "learning_rate": 3.025155161085772e-06, + "loss": 2.5556, + "step": 21260 + }, + { + "epoch": 2.2484137055837565, + "grad_norm": 8.375, + "learning_rate": 3.024987724690688e-06, + "loss": 2.4715, + "step": 21261 + }, + { + "epoch": 2.2485194585448394, + "grad_norm": 8.625, + "learning_rate": 3.0248202858320515e-06, + "loss": 2.4778, + "step": 21262 + }, + { + "epoch": 2.248625211505922, + "grad_norm": 8.625, + "learning_rate": 3.024652844510647e-06, + "loss": 2.675, + "step": 21263 + }, + { + "epoch": 2.248730964467005, + "grad_norm": 8.4375, + "learning_rate": 3.0244854007272616e-06, + "loss": 2.5232, + "step": 21264 + }, + { + "epoch": 2.248836717428088, + "grad_norm": 8.25, + "learning_rate": 3.024317954482682e-06, + "loss": 2.4815, + "step": 21265 + }, + { + "epoch": 2.2489424703891707, + "grad_norm": 8.5625, + "learning_rate": 3.0241505057776916e-06, + "loss": 2.4544, + "step": 21266 + }, + { + "epoch": 2.249048223350254, + "grad_norm": 8.0, + "learning_rate": 3.0239830546130773e-06, + "loss": 2.5317, + "step": 21267 + }, + { + "epoch": 2.249153976311337, + "grad_norm": 8.375, + "learning_rate": 3.023815600989625e-06, + "loss": 2.5395, + "step": 21268 + }, + { + "epoch": 2.2492597292724197, + "grad_norm": 8.75, + "learning_rate": 3.023648144908121e-06, + "loss": 2.3531, + "step": 21269 + }, + { + "epoch": 2.2493654822335025, + "grad_norm": 8.5625, + "learning_rate": 3.0234806863693487e-06, + "loss": 2.5557, + "step": 21270 + }, + { + "epoch": 2.2494712351945854, + "grad_norm": 8.5, + "learning_rate": 3.023313225374097e-06, + "loss": 2.8069, + "step": 21271 + }, + { + "epoch": 2.249576988155668, + "grad_norm": 8.1875, + "learning_rate": 3.02314576192315e-06, + "loss": 2.5692, + "step": 21272 + }, + { + "epoch": 2.249682741116751, + "grad_norm": 9.125, + "learning_rate": 3.0229782960172944e-06, + "loss": 2.8308, + "step": 21273 + }, + { + "epoch": 2.2497884940778343, + "grad_norm": 8.5625, + "learning_rate": 3.022810827657315e-06, + "loss": 2.6735, + "step": 21274 + }, + { + "epoch": 2.249894247038917, + "grad_norm": 9.875, + "learning_rate": 3.0226433568439983e-06, + "loss": 2.7847, + "step": 21275 + }, + { + "epoch": 2.25, + "grad_norm": 8.4375, + "learning_rate": 3.0224758835781298e-06, + "loss": 2.4969, + "step": 21276 + }, + { + "epoch": 2.250105752961083, + "grad_norm": 8.0, + "learning_rate": 3.0223084078604956e-06, + "loss": 2.4484, + "step": 21277 + }, + { + "epoch": 2.2502115059221657, + "grad_norm": 9.0625, + "learning_rate": 3.0221409296918823e-06, + "loss": 2.3382, + "step": 21278 + }, + { + "epoch": 2.250317258883249, + "grad_norm": 8.0, + "learning_rate": 3.021973449073076e-06, + "loss": 2.4678, + "step": 21279 + }, + { + "epoch": 2.250423011844332, + "grad_norm": 8.3125, + "learning_rate": 3.0218059660048597e-06, + "loss": 2.6153, + "step": 21280 + }, + { + "epoch": 2.2505287648054146, + "grad_norm": 8.1875, + "learning_rate": 3.021638480488022e-06, + "loss": 2.6787, + "step": 21281 + }, + { + "epoch": 2.2506345177664975, + "grad_norm": 8.375, + "learning_rate": 3.0214709925233486e-06, + "loss": 2.4109, + "step": 21282 + }, + { + "epoch": 2.2507402707275803, + "grad_norm": 8.375, + "learning_rate": 3.021303502111625e-06, + "loss": 2.5436, + "step": 21283 + }, + { + "epoch": 2.250846023688663, + "grad_norm": 8.8125, + "learning_rate": 3.021136009253638e-06, + "loss": 2.6201, + "step": 21284 + }, + { + "epoch": 2.250951776649746, + "grad_norm": 8.0625, + "learning_rate": 3.020968513950172e-06, + "loss": 2.4406, + "step": 21285 + }, + { + "epoch": 2.2510575296108293, + "grad_norm": 8.8125, + "learning_rate": 3.0208010162020142e-06, + "loss": 2.6298, + "step": 21286 + }, + { + "epoch": 2.251163282571912, + "grad_norm": 8.1875, + "learning_rate": 3.0206335160099496e-06, + "loss": 2.6089, + "step": 21287 + }, + { + "epoch": 2.251269035532995, + "grad_norm": 8.0625, + "learning_rate": 3.020466013374766e-06, + "loss": 2.5618, + "step": 21288 + }, + { + "epoch": 2.251374788494078, + "grad_norm": 7.9375, + "learning_rate": 3.0202985082972474e-06, + "loss": 2.5677, + "step": 21289 + }, + { + "epoch": 2.2514805414551606, + "grad_norm": 8.3125, + "learning_rate": 3.0201310007781805e-06, + "loss": 2.4697, + "step": 21290 + }, + { + "epoch": 2.2515862944162435, + "grad_norm": 8.5, + "learning_rate": 3.019963490818352e-06, + "loss": 2.4453, + "step": 21291 + }, + { + "epoch": 2.2516920473773268, + "grad_norm": 8.5, + "learning_rate": 3.0197959784185477e-06, + "loss": 2.7971, + "step": 21292 + }, + { + "epoch": 2.2517978003384096, + "grad_norm": 8.1875, + "learning_rate": 3.019628463579553e-06, + "loss": 2.4898, + "step": 21293 + }, + { + "epoch": 2.2519035532994924, + "grad_norm": 8.3125, + "learning_rate": 3.0194609463021545e-06, + "loss": 2.8178, + "step": 21294 + }, + { + "epoch": 2.2520093062605753, + "grad_norm": 8.25, + "learning_rate": 3.019293426587139e-06, + "loss": 2.5535, + "step": 21295 + }, + { + "epoch": 2.252115059221658, + "grad_norm": 8.4375, + "learning_rate": 3.0191259044352906e-06, + "loss": 2.5573, + "step": 21296 + }, + { + "epoch": 2.252220812182741, + "grad_norm": 8.3125, + "learning_rate": 3.0189583798473977e-06, + "loss": 2.583, + "step": 21297 + }, + { + "epoch": 2.2523265651438242, + "grad_norm": 8.375, + "learning_rate": 3.0187908528242444e-06, + "loss": 2.7054, + "step": 21298 + }, + { + "epoch": 2.252432318104907, + "grad_norm": 8.25, + "learning_rate": 3.0186233233666185e-06, + "loss": 2.535, + "step": 21299 + }, + { + "epoch": 2.25253807106599, + "grad_norm": 8.5625, + "learning_rate": 3.0184557914753056e-06, + "loss": 2.5351, + "step": 21300 + }, + { + "epoch": 2.2526438240270727, + "grad_norm": 8.5, + "learning_rate": 3.0182882571510912e-06, + "loss": 2.5227, + "step": 21301 + }, + { + "epoch": 2.2527495769881556, + "grad_norm": 8.8125, + "learning_rate": 3.0181207203947616e-06, + "loss": 2.6187, + "step": 21302 + }, + { + "epoch": 2.2528553299492384, + "grad_norm": 8.125, + "learning_rate": 3.0179531812071043e-06, + "loss": 2.5307, + "step": 21303 + }, + { + "epoch": 2.2529610829103213, + "grad_norm": 7.90625, + "learning_rate": 3.0177856395889046e-06, + "loss": 2.596, + "step": 21304 + }, + { + "epoch": 2.2530668358714045, + "grad_norm": 8.5625, + "learning_rate": 3.0176180955409477e-06, + "loss": 2.2251, + "step": 21305 + }, + { + "epoch": 2.2531725888324874, + "grad_norm": 9.3125, + "learning_rate": 3.0174505490640217e-06, + "loss": 2.5849, + "step": 21306 + }, + { + "epoch": 2.2532783417935702, + "grad_norm": 8.75, + "learning_rate": 3.017283000158911e-06, + "loss": 2.6096, + "step": 21307 + }, + { + "epoch": 2.253384094754653, + "grad_norm": 8.9375, + "learning_rate": 3.017115448826404e-06, + "loss": 2.572, + "step": 21308 + }, + { + "epoch": 2.253489847715736, + "grad_norm": 8.375, + "learning_rate": 3.016947895067284e-06, + "loss": 2.7268, + "step": 21309 + }, + { + "epoch": 2.253595600676819, + "grad_norm": 8.375, + "learning_rate": 3.0167803388823404e-06, + "loss": 2.2963, + "step": 21310 + }, + { + "epoch": 2.253701353637902, + "grad_norm": 8.5, + "learning_rate": 3.0166127802723578e-06, + "loss": 2.3093, + "step": 21311 + }, + { + "epoch": 2.253807106598985, + "grad_norm": 8.25, + "learning_rate": 3.016445219238122e-06, + "loss": 2.5972, + "step": 21312 + }, + { + "epoch": 2.2539128595600677, + "grad_norm": 8.8125, + "learning_rate": 3.0162776557804208e-06, + "loss": 2.6208, + "step": 21313 + }, + { + "epoch": 2.2540186125211505, + "grad_norm": 8.375, + "learning_rate": 3.0161100899000393e-06, + "loss": 2.5619, + "step": 21314 + }, + { + "epoch": 2.2541243654822334, + "grad_norm": 8.375, + "learning_rate": 3.0159425215977645e-06, + "loss": 2.6077, + "step": 21315 + }, + { + "epoch": 2.254230118443316, + "grad_norm": 8.5625, + "learning_rate": 3.0157749508743815e-06, + "loss": 2.0281, + "step": 21316 + }, + { + "epoch": 2.2543358714043995, + "grad_norm": 8.1875, + "learning_rate": 3.015607377730679e-06, + "loss": 2.5964, + "step": 21317 + }, + { + "epoch": 2.2544416243654823, + "grad_norm": 8.375, + "learning_rate": 3.0154398021674408e-06, + "loss": 2.491, + "step": 21318 + }, + { + "epoch": 2.254547377326565, + "grad_norm": 8.5, + "learning_rate": 3.0152722241854547e-06, + "loss": 2.538, + "step": 21319 + }, + { + "epoch": 2.254653130287648, + "grad_norm": 8.4375, + "learning_rate": 3.0151046437855076e-06, + "loss": 2.6423, + "step": 21320 + }, + { + "epoch": 2.254758883248731, + "grad_norm": 7.90625, + "learning_rate": 3.0149370609683843e-06, + "loss": 2.4726, + "step": 21321 + }, + { + "epoch": 2.2548646362098137, + "grad_norm": 8.4375, + "learning_rate": 3.014769475734872e-06, + "loss": 2.6847, + "step": 21322 + }, + { + "epoch": 2.2549703891708965, + "grad_norm": 8.3125, + "learning_rate": 3.0146018880857582e-06, + "loss": 2.5852, + "step": 21323 + }, + { + "epoch": 2.25507614213198, + "grad_norm": 8.5625, + "learning_rate": 3.0144342980218268e-06, + "loss": 2.6618, + "step": 21324 + }, + { + "epoch": 2.2551818950930627, + "grad_norm": 8.5, + "learning_rate": 3.014266705543866e-06, + "loss": 2.467, + "step": 21325 + }, + { + "epoch": 2.2552876480541455, + "grad_norm": 8.25, + "learning_rate": 3.0140991106526618e-06, + "loss": 2.7029, + "step": 21326 + }, + { + "epoch": 2.2553934010152283, + "grad_norm": 8.625, + "learning_rate": 3.013931513349001e-06, + "loss": 2.6703, + "step": 21327 + }, + { + "epoch": 2.255499153976311, + "grad_norm": 8.6875, + "learning_rate": 3.0137639136336693e-06, + "loss": 2.8107, + "step": 21328 + }, + { + "epoch": 2.2556049069373945, + "grad_norm": 8.0625, + "learning_rate": 3.0135963115074546e-06, + "loss": 2.5114, + "step": 21329 + }, + { + "epoch": 2.2557106598984773, + "grad_norm": 9.25, + "learning_rate": 3.0134287069711423e-06, + "loss": 2.6266, + "step": 21330 + }, + { + "epoch": 2.25581641285956, + "grad_norm": 8.4375, + "learning_rate": 3.013261100025518e-06, + "loss": 2.7457, + "step": 21331 + }, + { + "epoch": 2.255922165820643, + "grad_norm": 8.4375, + "learning_rate": 3.0130934906713705e-06, + "loss": 2.5792, + "step": 21332 + }, + { + "epoch": 2.256027918781726, + "grad_norm": 8.5625, + "learning_rate": 3.0129258789094843e-06, + "loss": 2.5517, + "step": 21333 + }, + { + "epoch": 2.2561336717428087, + "grad_norm": 8.5625, + "learning_rate": 3.012758264740647e-06, + "loss": 2.6447, + "step": 21334 + }, + { + "epoch": 2.2562394247038915, + "grad_norm": 8.6875, + "learning_rate": 3.0125906481656446e-06, + "loss": 2.5156, + "step": 21335 + }, + { + "epoch": 2.2563451776649748, + "grad_norm": 9.4375, + "learning_rate": 3.0124230291852645e-06, + "loss": 2.6577, + "step": 21336 + }, + { + "epoch": 2.2564509306260576, + "grad_norm": 8.1875, + "learning_rate": 3.012255407800292e-06, + "loss": 2.4962, + "step": 21337 + }, + { + "epoch": 2.2565566835871405, + "grad_norm": 8.6875, + "learning_rate": 3.0120877840115146e-06, + "loss": 2.6229, + "step": 21338 + }, + { + "epoch": 2.2566624365482233, + "grad_norm": 8.125, + "learning_rate": 3.011920157819719e-06, + "loss": 2.5202, + "step": 21339 + }, + { + "epoch": 2.256768189509306, + "grad_norm": 8.125, + "learning_rate": 3.0117525292256917e-06, + "loss": 2.3014, + "step": 21340 + }, + { + "epoch": 2.2568739424703894, + "grad_norm": 8.125, + "learning_rate": 3.0115848982302186e-06, + "loss": 2.4624, + "step": 21341 + }, + { + "epoch": 2.2569796954314723, + "grad_norm": 8.8125, + "learning_rate": 3.0114172648340873e-06, + "loss": 2.5573, + "step": 21342 + }, + { + "epoch": 2.257085448392555, + "grad_norm": 10.875, + "learning_rate": 3.0112496290380834e-06, + "loss": 2.3893, + "step": 21343 + }, + { + "epoch": 2.257191201353638, + "grad_norm": 8.6875, + "learning_rate": 3.011081990842994e-06, + "loss": 2.6547, + "step": 21344 + }, + { + "epoch": 2.2572969543147208, + "grad_norm": 8.25, + "learning_rate": 3.010914350249606e-06, + "loss": 2.4615, + "step": 21345 + }, + { + "epoch": 2.2574027072758036, + "grad_norm": 8.125, + "learning_rate": 3.010746707258706e-06, + "loss": 2.4564, + "step": 21346 + }, + { + "epoch": 2.2575084602368864, + "grad_norm": 8.5625, + "learning_rate": 3.010579061871081e-06, + "loss": 2.6545, + "step": 21347 + }, + { + "epoch": 2.2576142131979697, + "grad_norm": 8.6875, + "learning_rate": 3.0104114140875164e-06, + "loss": 2.6948, + "step": 21348 + }, + { + "epoch": 2.2577199661590526, + "grad_norm": 8.5625, + "learning_rate": 3.0102437639088007e-06, + "loss": 2.5727, + "step": 21349 + }, + { + "epoch": 2.2578257191201354, + "grad_norm": 8.5625, + "learning_rate": 3.010076111335719e-06, + "loss": 2.5442, + "step": 21350 + }, + { + "epoch": 2.2579314720812182, + "grad_norm": 8.5, + "learning_rate": 3.009908456369059e-06, + "loss": 2.6214, + "step": 21351 + }, + { + "epoch": 2.258037225042301, + "grad_norm": 8.6875, + "learning_rate": 3.0097407990096067e-06, + "loss": 2.5749, + "step": 21352 + }, + { + "epoch": 2.258142978003384, + "grad_norm": 8.5625, + "learning_rate": 3.0095731392581494e-06, + "loss": 2.4936, + "step": 21353 + }, + { + "epoch": 2.2582487309644668, + "grad_norm": 9.125, + "learning_rate": 3.0094054771154745e-06, + "loss": 2.6446, + "step": 21354 + }, + { + "epoch": 2.25835448392555, + "grad_norm": 8.4375, + "learning_rate": 3.009237812582368e-06, + "loss": 2.6572, + "step": 21355 + }, + { + "epoch": 2.258460236886633, + "grad_norm": 9.4375, + "learning_rate": 3.009070145659616e-06, + "loss": 2.6451, + "step": 21356 + }, + { + "epoch": 2.2585659898477157, + "grad_norm": 8.125, + "learning_rate": 3.0089024763480062e-06, + "loss": 2.4387, + "step": 21357 + }, + { + "epoch": 2.2586717428087986, + "grad_norm": 8.25, + "learning_rate": 3.0087348046483255e-06, + "loss": 2.6546, + "step": 21358 + }, + { + "epoch": 2.2587774957698814, + "grad_norm": 8.75, + "learning_rate": 3.00856713056136e-06, + "loss": 2.4379, + "step": 21359 + }, + { + "epoch": 2.2588832487309647, + "grad_norm": 7.90625, + "learning_rate": 3.008399454087897e-06, + "loss": 2.3112, + "step": 21360 + }, + { + "epoch": 2.2589890016920475, + "grad_norm": 8.25, + "learning_rate": 3.008231775228724e-06, + "loss": 2.4455, + "step": 21361 + }, + { + "epoch": 2.2590947546531304, + "grad_norm": 8.5, + "learning_rate": 3.008064093984627e-06, + "loss": 2.5459, + "step": 21362 + }, + { + "epoch": 2.259200507614213, + "grad_norm": 8.3125, + "learning_rate": 3.0078964103563924e-06, + "loss": 2.7013, + "step": 21363 + }, + { + "epoch": 2.259306260575296, + "grad_norm": 8.4375, + "learning_rate": 3.0077287243448077e-06, + "loss": 2.5003, + "step": 21364 + }, + { + "epoch": 2.259412013536379, + "grad_norm": 8.375, + "learning_rate": 3.0075610359506602e-06, + "loss": 2.5574, + "step": 21365 + }, + { + "epoch": 2.2595177664974617, + "grad_norm": 8.5625, + "learning_rate": 3.0073933451747362e-06, + "loss": 2.6156, + "step": 21366 + }, + { + "epoch": 2.259623519458545, + "grad_norm": 9.375, + "learning_rate": 3.007225652017823e-06, + "loss": 2.7675, + "step": 21367 + }, + { + "epoch": 2.259729272419628, + "grad_norm": 8.25, + "learning_rate": 3.007057956480707e-06, + "loss": 2.4144, + "step": 21368 + }, + { + "epoch": 2.2598350253807107, + "grad_norm": 8.375, + "learning_rate": 3.0068902585641747e-06, + "loss": 2.4617, + "step": 21369 + }, + { + "epoch": 2.2599407783417935, + "grad_norm": 8.375, + "learning_rate": 3.006722558269014e-06, + "loss": 2.6927, + "step": 21370 + }, + { + "epoch": 2.2600465313028764, + "grad_norm": 8.9375, + "learning_rate": 3.006554855596012e-06, + "loss": 2.4975, + "step": 21371 + }, + { + "epoch": 2.260152284263959, + "grad_norm": 9.125, + "learning_rate": 3.0063871505459554e-06, + "loss": 2.7003, + "step": 21372 + }, + { + "epoch": 2.2602580372250425, + "grad_norm": 8.4375, + "learning_rate": 3.006219443119631e-06, + "loss": 2.483, + "step": 21373 + }, + { + "epoch": 2.2603637901861253, + "grad_norm": 8.375, + "learning_rate": 3.006051733317826e-06, + "loss": 2.4268, + "step": 21374 + }, + { + "epoch": 2.260469543147208, + "grad_norm": 9.0625, + "learning_rate": 3.0058840211413264e-06, + "loss": 2.3854, + "step": 21375 + }, + { + "epoch": 2.260575296108291, + "grad_norm": 8.1875, + "learning_rate": 3.0057163065909206e-06, + "loss": 2.5997, + "step": 21376 + }, + { + "epoch": 2.260681049069374, + "grad_norm": 8.375, + "learning_rate": 3.005548589667395e-06, + "loss": 2.6148, + "step": 21377 + }, + { + "epoch": 2.2607868020304567, + "grad_norm": 8.5, + "learning_rate": 3.005380870371536e-06, + "loss": 2.4588, + "step": 21378 + }, + { + "epoch": 2.26089255499154, + "grad_norm": 8.0625, + "learning_rate": 3.0052131487041316e-06, + "loss": 2.5989, + "step": 21379 + }, + { + "epoch": 2.260998307952623, + "grad_norm": 8.625, + "learning_rate": 3.0050454246659683e-06, + "loss": 2.6241, + "step": 21380 + }, + { + "epoch": 2.2611040609137056, + "grad_norm": 8.75, + "learning_rate": 3.004877698257834e-06, + "loss": 2.7723, + "step": 21381 + }, + { + "epoch": 2.2612098138747885, + "grad_norm": 8.375, + "learning_rate": 3.0047099694805144e-06, + "loss": 2.5786, + "step": 21382 + }, + { + "epoch": 2.2613155668358713, + "grad_norm": 9.0625, + "learning_rate": 3.0045422383347977e-06, + "loss": 2.6466, + "step": 21383 + }, + { + "epoch": 2.261421319796954, + "grad_norm": 8.9375, + "learning_rate": 3.004374504821471e-06, + "loss": 2.3781, + "step": 21384 + }, + { + "epoch": 2.261527072758037, + "grad_norm": 8.375, + "learning_rate": 3.004206768941321e-06, + "loss": 2.4257, + "step": 21385 + }, + { + "epoch": 2.2616328257191203, + "grad_norm": 8.375, + "learning_rate": 3.0040390306951344e-06, + "loss": 2.6157, + "step": 21386 + }, + { + "epoch": 2.261738578680203, + "grad_norm": 7.90625, + "learning_rate": 3.0038712900836987e-06, + "loss": 2.5257, + "step": 21387 + }, + { + "epoch": 2.261844331641286, + "grad_norm": 8.1875, + "learning_rate": 3.0037035471078006e-06, + "loss": 2.4861, + "step": 21388 + }, + { + "epoch": 2.261950084602369, + "grad_norm": 8.3125, + "learning_rate": 3.0035358017682287e-06, + "loss": 2.6246, + "step": 21389 + }, + { + "epoch": 2.2620558375634516, + "grad_norm": 8.5, + "learning_rate": 3.0033680540657694e-06, + "loss": 2.7477, + "step": 21390 + }, + { + "epoch": 2.262161590524535, + "grad_norm": 8.8125, + "learning_rate": 3.0032003040012093e-06, + "loss": 2.5849, + "step": 21391 + }, + { + "epoch": 2.2622673434856178, + "grad_norm": 8.0625, + "learning_rate": 3.003032551575336e-06, + "loss": 2.4506, + "step": 21392 + }, + { + "epoch": 2.2623730964467006, + "grad_norm": 8.1875, + "learning_rate": 3.002864796788937e-06, + "loss": 2.5684, + "step": 21393 + }, + { + "epoch": 2.2624788494077834, + "grad_norm": 8.1875, + "learning_rate": 3.0026970396427985e-06, + "loss": 2.4831, + "step": 21394 + }, + { + "epoch": 2.2625846023688663, + "grad_norm": 8.4375, + "learning_rate": 3.002529280137709e-06, + "loss": 2.532, + "step": 21395 + }, + { + "epoch": 2.262690355329949, + "grad_norm": 8.3125, + "learning_rate": 3.002361518274454e-06, + "loss": 2.3242, + "step": 21396 + }, + { + "epoch": 2.262796108291032, + "grad_norm": 8.3125, + "learning_rate": 3.0021937540538233e-06, + "loss": 2.5021, + "step": 21397 + }, + { + "epoch": 2.2629018612521152, + "grad_norm": 7.90625, + "learning_rate": 3.002025987476602e-06, + "loss": 2.4204, + "step": 21398 + }, + { + "epoch": 2.263007614213198, + "grad_norm": 8.3125, + "learning_rate": 3.001858218543578e-06, + "loss": 2.6111, + "step": 21399 + }, + { + "epoch": 2.263113367174281, + "grad_norm": 8.5, + "learning_rate": 3.001690447255539e-06, + "loss": 2.5493, + "step": 21400 + }, + { + "epoch": 2.2632191201353637, + "grad_norm": 8.875, + "learning_rate": 3.0015226736132713e-06, + "loss": 2.5868, + "step": 21401 + }, + { + "epoch": 2.2633248730964466, + "grad_norm": 8.875, + "learning_rate": 3.001354897617563e-06, + "loss": 2.7861, + "step": 21402 + }, + { + "epoch": 2.2634306260575294, + "grad_norm": 8.1875, + "learning_rate": 3.001187119269202e-06, + "loss": 2.3943, + "step": 21403 + }, + { + "epoch": 2.2635363790186127, + "grad_norm": 9.25, + "learning_rate": 3.001019338568974e-06, + "loss": 2.5281, + "step": 21404 + }, + { + "epoch": 2.2636421319796955, + "grad_norm": 8.5, + "learning_rate": 3.0008515555176672e-06, + "loss": 2.4794, + "step": 21405 + }, + { + "epoch": 2.2637478849407784, + "grad_norm": 8.0625, + "learning_rate": 3.000683770116069e-06, + "loss": 2.6025, + "step": 21406 + }, + { + "epoch": 2.2638536379018612, + "grad_norm": 8.375, + "learning_rate": 3.0005159823649667e-06, + "loss": 2.2646, + "step": 21407 + }, + { + "epoch": 2.263959390862944, + "grad_norm": 8.3125, + "learning_rate": 3.0003481922651478e-06, + "loss": 2.3974, + "step": 21408 + }, + { + "epoch": 2.264065143824027, + "grad_norm": 8.5625, + "learning_rate": 3.0001803998173993e-06, + "loss": 2.4119, + "step": 21409 + }, + { + "epoch": 2.26417089678511, + "grad_norm": 8.4375, + "learning_rate": 3.0000126050225086e-06, + "loss": 2.5415, + "step": 21410 + }, + { + "epoch": 2.264276649746193, + "grad_norm": 8.375, + "learning_rate": 2.9998448078812637e-06, + "loss": 2.4881, + "step": 21411 + }, + { + "epoch": 2.264382402707276, + "grad_norm": 8.4375, + "learning_rate": 2.9996770083944517e-06, + "loss": 2.4788, + "step": 21412 + }, + { + "epoch": 2.2644881556683587, + "grad_norm": 8.4375, + "learning_rate": 2.9995092065628585e-06, + "loss": 2.1977, + "step": 21413 + }, + { + "epoch": 2.2645939086294415, + "grad_norm": 8.875, + "learning_rate": 2.9993414023872735e-06, + "loss": 2.898, + "step": 21414 + }, + { + "epoch": 2.2646996615905244, + "grad_norm": 8.3125, + "learning_rate": 2.999173595868484e-06, + "loss": 2.6592, + "step": 21415 + }, + { + "epoch": 2.264805414551607, + "grad_norm": 8.25, + "learning_rate": 2.9990057870072768e-06, + "loss": 2.4826, + "step": 21416 + }, + { + "epoch": 2.2649111675126905, + "grad_norm": 9.1875, + "learning_rate": 2.9988379758044397e-06, + "loss": 2.5618, + "step": 21417 + }, + { + "epoch": 2.2650169204737733, + "grad_norm": 9.0625, + "learning_rate": 2.99867016226076e-06, + "loss": 2.5951, + "step": 21418 + }, + { + "epoch": 2.265122673434856, + "grad_norm": 8.5625, + "learning_rate": 2.9985023463770254e-06, + "loss": 2.5275, + "step": 21419 + }, + { + "epoch": 2.265228426395939, + "grad_norm": 8.3125, + "learning_rate": 2.998334528154022e-06, + "loss": 2.4911, + "step": 21420 + }, + { + "epoch": 2.265334179357022, + "grad_norm": 8.125, + "learning_rate": 2.9981667075925393e-06, + "loss": 2.382, + "step": 21421 + }, + { + "epoch": 2.265439932318105, + "grad_norm": 8.4375, + "learning_rate": 2.997998884693364e-06, + "loss": 2.6989, + "step": 21422 + }, + { + "epoch": 2.265545685279188, + "grad_norm": 8.0625, + "learning_rate": 2.997831059457283e-06, + "loss": 2.365, + "step": 21423 + }, + { + "epoch": 2.265651438240271, + "grad_norm": 8.625, + "learning_rate": 2.9976632318850853e-06, + "loss": 2.5365, + "step": 21424 + }, + { + "epoch": 2.2657571912013537, + "grad_norm": 8.3125, + "learning_rate": 2.997495401977557e-06, + "loss": 2.4608, + "step": 21425 + }, + { + "epoch": 2.2658629441624365, + "grad_norm": 7.84375, + "learning_rate": 2.9973275697354864e-06, + "loss": 2.4743, + "step": 21426 + }, + { + "epoch": 2.2659686971235193, + "grad_norm": 8.4375, + "learning_rate": 2.9971597351596615e-06, + "loss": 2.4721, + "step": 21427 + }, + { + "epoch": 2.266074450084602, + "grad_norm": 8.625, + "learning_rate": 2.996991898250869e-06, + "loss": 2.7404, + "step": 21428 + }, + { + "epoch": 2.2661802030456855, + "grad_norm": 8.75, + "learning_rate": 2.9968240590098963e-06, + "loss": 2.6474, + "step": 21429 + }, + { + "epoch": 2.2662859560067683, + "grad_norm": 8.25, + "learning_rate": 2.996656217437532e-06, + "loss": 2.3554, + "step": 21430 + }, + { + "epoch": 2.266391708967851, + "grad_norm": 8.0625, + "learning_rate": 2.996488373534562e-06, + "loss": 2.4979, + "step": 21431 + }, + { + "epoch": 2.266497461928934, + "grad_norm": 8.625, + "learning_rate": 2.9963205273017763e-06, + "loss": 2.4118, + "step": 21432 + }, + { + "epoch": 2.266603214890017, + "grad_norm": 8.3125, + "learning_rate": 2.9961526787399606e-06, + "loss": 2.6036, + "step": 21433 + }, + { + "epoch": 2.2667089678510997, + "grad_norm": 8.4375, + "learning_rate": 2.9959848278499047e-06, + "loss": 2.6066, + "step": 21434 + }, + { + "epoch": 2.2668147208121825, + "grad_norm": 9.125, + "learning_rate": 2.995816974632394e-06, + "loss": 2.5582, + "step": 21435 + }, + { + "epoch": 2.2669204737732658, + "grad_norm": 8.625, + "learning_rate": 2.995649119088217e-06, + "loss": 2.531, + "step": 21436 + }, + { + "epoch": 2.2670262267343486, + "grad_norm": 8.125, + "learning_rate": 2.9954812612181614e-06, + "loss": 2.5542, + "step": 21437 + }, + { + "epoch": 2.2671319796954315, + "grad_norm": 8.75, + "learning_rate": 2.995313401023015e-06, + "loss": 2.7908, + "step": 21438 + }, + { + "epoch": 2.2672377326565143, + "grad_norm": 8.3125, + "learning_rate": 2.995145538503565e-06, + "loss": 2.5001, + "step": 21439 + }, + { + "epoch": 2.267343485617597, + "grad_norm": 8.5, + "learning_rate": 2.9949776736605996e-06, + "loss": 2.5562, + "step": 21440 + }, + { + "epoch": 2.2674492385786804, + "grad_norm": 8.875, + "learning_rate": 2.9948098064949064e-06, + "loss": 2.6378, + "step": 21441 + }, + { + "epoch": 2.2675549915397633, + "grad_norm": 8.3125, + "learning_rate": 2.994641937007273e-06, + "loss": 2.6028, + "step": 21442 + }, + { + "epoch": 2.267660744500846, + "grad_norm": 8.625, + "learning_rate": 2.9944740651984883e-06, + "loss": 2.4438, + "step": 21443 + }, + { + "epoch": 2.267766497461929, + "grad_norm": 8.6875, + "learning_rate": 2.994306191069338e-06, + "loss": 2.3864, + "step": 21444 + }, + { + "epoch": 2.2678722504230118, + "grad_norm": 8.75, + "learning_rate": 2.9941383146206114e-06, + "loss": 2.6206, + "step": 21445 + }, + { + "epoch": 2.2679780033840946, + "grad_norm": 10.1875, + "learning_rate": 2.9939704358530956e-06, + "loss": 2.4107, + "step": 21446 + }, + { + "epoch": 2.2680837563451774, + "grad_norm": 8.3125, + "learning_rate": 2.9938025547675793e-06, + "loss": 2.5387, + "step": 21447 + }, + { + "epoch": 2.2681895093062607, + "grad_norm": 9.5625, + "learning_rate": 2.9936346713648485e-06, + "loss": 2.7717, + "step": 21448 + }, + { + "epoch": 2.2682952622673436, + "grad_norm": 8.5, + "learning_rate": 2.9934667856456923e-06, + "loss": 2.5585, + "step": 21449 + }, + { + "epoch": 2.2684010152284264, + "grad_norm": 8.5, + "learning_rate": 2.9932988976108983e-06, + "loss": 2.6249, + "step": 21450 + }, + { + "epoch": 2.2685067681895092, + "grad_norm": 8.5625, + "learning_rate": 2.993131007261255e-06, + "loss": 2.5658, + "step": 21451 + }, + { + "epoch": 2.268612521150592, + "grad_norm": 7.9375, + "learning_rate": 2.992963114597549e-06, + "loss": 2.439, + "step": 21452 + }, + { + "epoch": 2.2687182741116754, + "grad_norm": 8.0625, + "learning_rate": 2.992795219620569e-06, + "loss": 2.5086, + "step": 21453 + }, + { + "epoch": 2.268824027072758, + "grad_norm": 8.375, + "learning_rate": 2.9926273223311024e-06, + "loss": 2.6416, + "step": 21454 + }, + { + "epoch": 2.268929780033841, + "grad_norm": 8.25, + "learning_rate": 2.992459422729937e-06, + "loss": 2.5917, + "step": 21455 + }, + { + "epoch": 2.269035532994924, + "grad_norm": 8.3125, + "learning_rate": 2.9922915208178614e-06, + "loss": 2.5659, + "step": 21456 + }, + { + "epoch": 2.2691412859560067, + "grad_norm": 8.5625, + "learning_rate": 2.992123616595663e-06, + "loss": 2.5525, + "step": 21457 + }, + { + "epoch": 2.2692470389170896, + "grad_norm": 8.3125, + "learning_rate": 2.9919557100641284e-06, + "loss": 2.4416, + "step": 21458 + }, + { + "epoch": 2.2693527918781724, + "grad_norm": 8.375, + "learning_rate": 2.9917878012240487e-06, + "loss": 2.5783, + "step": 21459 + }, + { + "epoch": 2.2694585448392557, + "grad_norm": 8.0, + "learning_rate": 2.9916198900762095e-06, + "loss": 2.5931, + "step": 21460 + }, + { + "epoch": 2.2695642978003385, + "grad_norm": 8.1875, + "learning_rate": 2.9914519766213985e-06, + "loss": 2.565, + "step": 21461 + }, + { + "epoch": 2.2696700507614214, + "grad_norm": 8.375, + "learning_rate": 2.991284060860405e-06, + "loss": 2.598, + "step": 21462 + }, + { + "epoch": 2.269775803722504, + "grad_norm": 8.625, + "learning_rate": 2.9911161427940162e-06, + "loss": 2.6247, + "step": 21463 + }, + { + "epoch": 2.269881556683587, + "grad_norm": 8.1875, + "learning_rate": 2.9909482224230204e-06, + "loss": 2.49, + "step": 21464 + }, + { + "epoch": 2.26998730964467, + "grad_norm": 8.0, + "learning_rate": 2.990780299748205e-06, + "loss": 2.324, + "step": 21465 + }, + { + "epoch": 2.2700930626057527, + "grad_norm": 8.4375, + "learning_rate": 2.9906123747703587e-06, + "loss": 2.5474, + "step": 21466 + }, + { + "epoch": 2.270198815566836, + "grad_norm": 8.25, + "learning_rate": 2.9904444474902688e-06, + "loss": 2.585, + "step": 21467 + }, + { + "epoch": 2.270304568527919, + "grad_norm": 8.1875, + "learning_rate": 2.9902765179087234e-06, + "loss": 2.4776, + "step": 21468 + }, + { + "epoch": 2.2704103214890017, + "grad_norm": 8.5625, + "learning_rate": 2.9901085860265117e-06, + "loss": 2.6378, + "step": 21469 + }, + { + "epoch": 2.2705160744500845, + "grad_norm": 8.625, + "learning_rate": 2.989940651844421e-06, + "loss": 2.7176, + "step": 21470 + }, + { + "epoch": 2.2706218274111674, + "grad_norm": 8.5, + "learning_rate": 2.989772715363238e-06, + "loss": 2.3034, + "step": 21471 + }, + { + "epoch": 2.2707275803722506, + "grad_norm": 8.4375, + "learning_rate": 2.989604776583753e-06, + "loss": 2.6603, + "step": 21472 + }, + { + "epoch": 2.2708333333333335, + "grad_norm": 9.625, + "learning_rate": 2.989436835506753e-06, + "loss": 2.8974, + "step": 21473 + }, + { + "epoch": 2.2709390862944163, + "grad_norm": 8.25, + "learning_rate": 2.9892688921330254e-06, + "loss": 2.5662, + "step": 21474 + }, + { + "epoch": 2.271044839255499, + "grad_norm": 8.5, + "learning_rate": 2.989100946463359e-06, + "loss": 2.6433, + "step": 21475 + }, + { + "epoch": 2.271150592216582, + "grad_norm": 8.3125, + "learning_rate": 2.9889329984985428e-06, + "loss": 2.4627, + "step": 21476 + }, + { + "epoch": 2.271256345177665, + "grad_norm": 8.375, + "learning_rate": 2.9887650482393634e-06, + "loss": 2.6806, + "step": 21477 + }, + { + "epoch": 2.2713620981387477, + "grad_norm": 8.375, + "learning_rate": 2.9885970956866094e-06, + "loss": 2.6114, + "step": 21478 + }, + { + "epoch": 2.271467851099831, + "grad_norm": 8.5625, + "learning_rate": 2.9884291408410693e-06, + "loss": 2.4902, + "step": 21479 + }, + { + "epoch": 2.271573604060914, + "grad_norm": 7.6875, + "learning_rate": 2.988261183703531e-06, + "loss": 2.5009, + "step": 21480 + }, + { + "epoch": 2.2716793570219966, + "grad_norm": 8.25, + "learning_rate": 2.988093224274783e-06, + "loss": 2.6373, + "step": 21481 + }, + { + "epoch": 2.2717851099830795, + "grad_norm": 8.0625, + "learning_rate": 2.9879252625556127e-06, + "loss": 2.5392, + "step": 21482 + }, + { + "epoch": 2.2718908629441623, + "grad_norm": 8.4375, + "learning_rate": 2.9877572985468084e-06, + "loss": 2.5854, + "step": 21483 + }, + { + "epoch": 2.271996615905245, + "grad_norm": 8.25, + "learning_rate": 2.9875893322491584e-06, + "loss": 2.4809, + "step": 21484 + }, + { + "epoch": 2.2721023688663284, + "grad_norm": 9.8125, + "learning_rate": 2.9874213636634518e-06, + "loss": 2.5652, + "step": 21485 + }, + { + "epoch": 2.2722081218274113, + "grad_norm": 8.375, + "learning_rate": 2.9872533927904762e-06, + "loss": 2.6299, + "step": 21486 + }, + { + "epoch": 2.272313874788494, + "grad_norm": 7.96875, + "learning_rate": 2.9870854196310195e-06, + "loss": 2.2936, + "step": 21487 + }, + { + "epoch": 2.272419627749577, + "grad_norm": 8.875, + "learning_rate": 2.98691744418587e-06, + "loss": 2.8347, + "step": 21488 + }, + { + "epoch": 2.27252538071066, + "grad_norm": 8.125, + "learning_rate": 2.986749466455816e-06, + "loss": 2.6128, + "step": 21489 + }, + { + "epoch": 2.2726311336717426, + "grad_norm": 8.8125, + "learning_rate": 2.986581486441646e-06, + "loss": 2.6931, + "step": 21490 + }, + { + "epoch": 2.272736886632826, + "grad_norm": 8.3125, + "learning_rate": 2.9864135041441482e-06, + "loss": 2.3963, + "step": 21491 + }, + { + "epoch": 2.2728426395939088, + "grad_norm": 8.0625, + "learning_rate": 2.986245519564111e-06, + "loss": 2.3727, + "step": 21492 + }, + { + "epoch": 2.2729483925549916, + "grad_norm": 8.0625, + "learning_rate": 2.986077532702321e-06, + "loss": 2.4856, + "step": 21493 + }, + { + "epoch": 2.2730541455160744, + "grad_norm": 8.1875, + "learning_rate": 2.985909543559569e-06, + "loss": 2.66, + "step": 21494 + }, + { + "epoch": 2.2731598984771573, + "grad_norm": 8.5625, + "learning_rate": 2.985741552136643e-06, + "loss": 2.5904, + "step": 21495 + }, + { + "epoch": 2.27326565143824, + "grad_norm": 8.625, + "learning_rate": 2.985573558434329e-06, + "loss": 2.5933, + "step": 21496 + }, + { + "epoch": 2.273371404399323, + "grad_norm": 8.5, + "learning_rate": 2.985405562453418e-06, + "loss": 2.3581, + "step": 21497 + }, + { + "epoch": 2.2734771573604062, + "grad_norm": 8.75, + "learning_rate": 2.9852375641946974e-06, + "loss": 2.6015, + "step": 21498 + }, + { + "epoch": 2.273582910321489, + "grad_norm": 7.75, + "learning_rate": 2.985069563658954e-06, + "loss": 2.434, + "step": 21499 + }, + { + "epoch": 2.273688663282572, + "grad_norm": 8.125, + "learning_rate": 2.984901560846979e-06, + "loss": 2.558, + "step": 21500 + }, + { + "epoch": 2.2737944162436547, + "grad_norm": 8.0, + "learning_rate": 2.9847335557595584e-06, + "loss": 2.3868, + "step": 21501 + }, + { + "epoch": 2.2739001692047376, + "grad_norm": 9.5, + "learning_rate": 2.984565548397482e-06, + "loss": 2.5101, + "step": 21502 + }, + { + "epoch": 2.274005922165821, + "grad_norm": 8.3125, + "learning_rate": 2.984397538761537e-06, + "loss": 2.6503, + "step": 21503 + }, + { + "epoch": 2.2741116751269037, + "grad_norm": 8.125, + "learning_rate": 2.9842295268525127e-06, + "loss": 2.3899, + "step": 21504 + }, + { + "epoch": 2.2742174280879865, + "grad_norm": 8.3125, + "learning_rate": 2.984061512671198e-06, + "loss": 2.5119, + "step": 21505 + }, + { + "epoch": 2.2743231810490694, + "grad_norm": 8.1875, + "learning_rate": 2.9838934962183795e-06, + "loss": 2.6538, + "step": 21506 + }, + { + "epoch": 2.2744289340101522, + "grad_norm": 8.875, + "learning_rate": 2.9837254774948476e-06, + "loss": 2.4865, + "step": 21507 + }, + { + "epoch": 2.274534686971235, + "grad_norm": 8.8125, + "learning_rate": 2.9835574565013897e-06, + "loss": 2.0705, + "step": 21508 + }, + { + "epoch": 2.274640439932318, + "grad_norm": 8.8125, + "learning_rate": 2.983389433238794e-06, + "loss": 2.7594, + "step": 21509 + }, + { + "epoch": 2.274746192893401, + "grad_norm": 8.3125, + "learning_rate": 2.9832214077078497e-06, + "loss": 2.3616, + "step": 21510 + }, + { + "epoch": 2.274851945854484, + "grad_norm": 8.75, + "learning_rate": 2.9830533799093448e-06, + "loss": 2.7048, + "step": 21511 + }, + { + "epoch": 2.274957698815567, + "grad_norm": 8.625, + "learning_rate": 2.9828853498440674e-06, + "loss": 2.5071, + "step": 21512 + }, + { + "epoch": 2.2750634517766497, + "grad_norm": 7.9375, + "learning_rate": 2.9827173175128076e-06, + "loss": 2.418, + "step": 21513 + }, + { + "epoch": 2.2751692047377325, + "grad_norm": 8.5, + "learning_rate": 2.9825492829163525e-06, + "loss": 2.7558, + "step": 21514 + }, + { + "epoch": 2.2752749576988154, + "grad_norm": 8.5, + "learning_rate": 2.982381246055491e-06, + "loss": 2.6953, + "step": 21515 + }, + { + "epoch": 2.2753807106598987, + "grad_norm": 8.125, + "learning_rate": 2.982213206931012e-06, + "loss": 2.4514, + "step": 21516 + }, + { + "epoch": 2.2754864636209815, + "grad_norm": 7.90625, + "learning_rate": 2.982045165543703e-06, + "loss": 2.5434, + "step": 21517 + }, + { + "epoch": 2.2755922165820643, + "grad_norm": 8.5625, + "learning_rate": 2.9818771218943534e-06, + "loss": 2.6252, + "step": 21518 + }, + { + "epoch": 2.275697969543147, + "grad_norm": 8.625, + "learning_rate": 2.9817090759837514e-06, + "loss": 2.5813, + "step": 21519 + }, + { + "epoch": 2.27580372250423, + "grad_norm": 9.1875, + "learning_rate": 2.9815410278126854e-06, + "loss": 2.676, + "step": 21520 + }, + { + "epoch": 2.275909475465313, + "grad_norm": 8.5, + "learning_rate": 2.981372977381945e-06, + "loss": 2.5186, + "step": 21521 + }, + { + "epoch": 2.276015228426396, + "grad_norm": 8.375, + "learning_rate": 2.9812049246923173e-06, + "loss": 2.6477, + "step": 21522 + }, + { + "epoch": 2.276120981387479, + "grad_norm": 8.125, + "learning_rate": 2.9810368697445925e-06, + "loss": 2.5085, + "step": 21523 + }, + { + "epoch": 2.276226734348562, + "grad_norm": 8.5, + "learning_rate": 2.9808688125395584e-06, + "loss": 2.3514, + "step": 21524 + }, + { + "epoch": 2.2763324873096447, + "grad_norm": 8.375, + "learning_rate": 2.980700753078003e-06, + "loss": 2.5287, + "step": 21525 + }, + { + "epoch": 2.2764382402707275, + "grad_norm": 7.9375, + "learning_rate": 2.9805326913607164e-06, + "loss": 2.6505, + "step": 21526 + }, + { + "epoch": 2.2765439932318103, + "grad_norm": 8.9375, + "learning_rate": 2.9803646273884857e-06, + "loss": 2.5824, + "step": 21527 + }, + { + "epoch": 2.276649746192893, + "grad_norm": 8.375, + "learning_rate": 2.9801965611620996e-06, + "loss": 2.5569, + "step": 21528 + }, + { + "epoch": 2.2767554991539765, + "grad_norm": 8.25, + "learning_rate": 2.9800284926823484e-06, + "loss": 2.4576, + "step": 21529 + }, + { + "epoch": 2.2768612521150593, + "grad_norm": 8.375, + "learning_rate": 2.9798604219500207e-06, + "loss": 2.6215, + "step": 21530 + }, + { + "epoch": 2.276967005076142, + "grad_norm": 8.5, + "learning_rate": 2.9796923489659024e-06, + "loss": 2.6601, + "step": 21531 + }, + { + "epoch": 2.277072758037225, + "grad_norm": 8.5625, + "learning_rate": 2.9795242737307856e-06, + "loss": 2.7219, + "step": 21532 + }, + { + "epoch": 2.277178510998308, + "grad_norm": 9.125, + "learning_rate": 2.979356196245457e-06, + "loss": 2.5768, + "step": 21533 + }, + { + "epoch": 2.277284263959391, + "grad_norm": 8.0625, + "learning_rate": 2.9791881165107057e-06, + "loss": 2.4506, + "step": 21534 + }, + { + "epoch": 2.277390016920474, + "grad_norm": 8.375, + "learning_rate": 2.9790200345273206e-06, + "loss": 2.5776, + "step": 21535 + }, + { + "epoch": 2.2774957698815568, + "grad_norm": 8.6875, + "learning_rate": 2.9788519502960905e-06, + "loss": 2.3286, + "step": 21536 + }, + { + "epoch": 2.2776015228426396, + "grad_norm": 7.96875, + "learning_rate": 2.978683863817804e-06, + "loss": 2.5574, + "step": 21537 + }, + { + "epoch": 2.2777072758037225, + "grad_norm": 8.5, + "learning_rate": 2.978515775093249e-06, + "loss": 2.5094, + "step": 21538 + }, + { + "epoch": 2.2778130287648053, + "grad_norm": 9.125, + "learning_rate": 2.9783476841232163e-06, + "loss": 2.7637, + "step": 21539 + }, + { + "epoch": 2.277918781725888, + "grad_norm": 8.3125, + "learning_rate": 2.9781795909084937e-06, + "loss": 2.6965, + "step": 21540 + }, + { + "epoch": 2.2780245346869714, + "grad_norm": 8.0, + "learning_rate": 2.978011495449869e-06, + "loss": 2.4375, + "step": 21541 + }, + { + "epoch": 2.2781302876480543, + "grad_norm": 8.375, + "learning_rate": 2.9778433977481324e-06, + "loss": 2.4959, + "step": 21542 + }, + { + "epoch": 2.278236040609137, + "grad_norm": 8.75, + "learning_rate": 2.9776752978040722e-06, + "loss": 2.671, + "step": 21543 + }, + { + "epoch": 2.27834179357022, + "grad_norm": 8.0625, + "learning_rate": 2.977507195618477e-06, + "loss": 2.5016, + "step": 21544 + }, + { + "epoch": 2.2784475465313028, + "grad_norm": 8.375, + "learning_rate": 2.9773390911921358e-06, + "loss": 2.6614, + "step": 21545 + }, + { + "epoch": 2.2785532994923856, + "grad_norm": 9.0625, + "learning_rate": 2.9771709845258375e-06, + "loss": 2.7441, + "step": 21546 + }, + { + "epoch": 2.2786590524534684, + "grad_norm": 8.4375, + "learning_rate": 2.9770028756203707e-06, + "loss": 2.7334, + "step": 21547 + }, + { + "epoch": 2.2787648054145517, + "grad_norm": 7.90625, + "learning_rate": 2.9768347644765246e-06, + "loss": 2.2376, + "step": 21548 + }, + { + "epoch": 2.2788705583756346, + "grad_norm": 8.4375, + "learning_rate": 2.976666651095088e-06, + "loss": 2.4826, + "step": 21549 + }, + { + "epoch": 2.2789763113367174, + "grad_norm": 8.25, + "learning_rate": 2.97649853547685e-06, + "loss": 2.6516, + "step": 21550 + }, + { + "epoch": 2.2790820642978002, + "grad_norm": 8.3125, + "learning_rate": 2.976330417622599e-06, + "loss": 2.614, + "step": 21551 + }, + { + "epoch": 2.279187817258883, + "grad_norm": 8.8125, + "learning_rate": 2.976162297533125e-06, + "loss": 2.7831, + "step": 21552 + }, + { + "epoch": 2.2792935702199664, + "grad_norm": 8.1875, + "learning_rate": 2.9759941752092147e-06, + "loss": 2.2502, + "step": 21553 + }, + { + "epoch": 2.279399323181049, + "grad_norm": 8.25, + "learning_rate": 2.9758260506516595e-06, + "loss": 2.4584, + "step": 21554 + }, + { + "epoch": 2.279505076142132, + "grad_norm": 8.625, + "learning_rate": 2.9756579238612467e-06, + "loss": 2.5976, + "step": 21555 + }, + { + "epoch": 2.279610829103215, + "grad_norm": 8.1875, + "learning_rate": 2.9754897948387657e-06, + "loss": 2.4803, + "step": 21556 + }, + { + "epoch": 2.2797165820642977, + "grad_norm": 7.65625, + "learning_rate": 2.975321663585006e-06, + "loss": 2.4484, + "step": 21557 + }, + { + "epoch": 2.2798223350253806, + "grad_norm": 8.5625, + "learning_rate": 2.9751535301007555e-06, + "loss": 2.6453, + "step": 21558 + }, + { + "epoch": 2.2799280879864634, + "grad_norm": 10.375, + "learning_rate": 2.9749853943868044e-06, + "loss": 2.6521, + "step": 21559 + }, + { + "epoch": 2.2800338409475467, + "grad_norm": 8.6875, + "learning_rate": 2.9748172564439408e-06, + "loss": 2.4158, + "step": 21560 + }, + { + "epoch": 2.2801395939086295, + "grad_norm": 8.4375, + "learning_rate": 2.974649116272954e-06, + "loss": 2.6579, + "step": 21561 + }, + { + "epoch": 2.2802453468697124, + "grad_norm": 7.71875, + "learning_rate": 2.974480973874634e-06, + "loss": 2.352, + "step": 21562 + }, + { + "epoch": 2.280351099830795, + "grad_norm": 8.4375, + "learning_rate": 2.9743128292497675e-06, + "loss": 2.6014, + "step": 21563 + }, + { + "epoch": 2.280456852791878, + "grad_norm": 8.625, + "learning_rate": 2.974144682399145e-06, + "loss": 2.5238, + "step": 21564 + }, + { + "epoch": 2.2805626057529613, + "grad_norm": 8.375, + "learning_rate": 2.973976533323556e-06, + "loss": 2.6766, + "step": 21565 + }, + { + "epoch": 2.280668358714044, + "grad_norm": 8.3125, + "learning_rate": 2.9738083820237884e-06, + "loss": 2.4579, + "step": 21566 + }, + { + "epoch": 2.280774111675127, + "grad_norm": 8.0, + "learning_rate": 2.973640228500633e-06, + "loss": 2.4602, + "step": 21567 + }, + { + "epoch": 2.28087986463621, + "grad_norm": 8.4375, + "learning_rate": 2.9734720727548765e-06, + "loss": 2.6135, + "step": 21568 + }, + { + "epoch": 2.2809856175972927, + "grad_norm": 8.0625, + "learning_rate": 2.9733039147873095e-06, + "loss": 2.6293, + "step": 21569 + }, + { + "epoch": 2.2810913705583755, + "grad_norm": 8.6875, + "learning_rate": 2.973135754598721e-06, + "loss": 2.6886, + "step": 21570 + }, + { + "epoch": 2.2811971235194584, + "grad_norm": 7.96875, + "learning_rate": 2.9729675921899005e-06, + "loss": 2.4075, + "step": 21571 + }, + { + "epoch": 2.2813028764805416, + "grad_norm": 8.1875, + "learning_rate": 2.9727994275616357e-06, + "loss": 2.578, + "step": 21572 + }, + { + "epoch": 2.2814086294416245, + "grad_norm": 7.90625, + "learning_rate": 2.9726312607147163e-06, + "loss": 2.4942, + "step": 21573 + }, + { + "epoch": 2.2815143824027073, + "grad_norm": 8.1875, + "learning_rate": 2.972463091649932e-06, + "loss": 2.1412, + "step": 21574 + }, + { + "epoch": 2.28162013536379, + "grad_norm": 7.875, + "learning_rate": 2.9722949203680723e-06, + "loss": 2.3074, + "step": 21575 + }, + { + "epoch": 2.281725888324873, + "grad_norm": 8.5625, + "learning_rate": 2.972126746869925e-06, + "loss": 2.6944, + "step": 21576 + }, + { + "epoch": 2.281831641285956, + "grad_norm": 7.8125, + "learning_rate": 2.9719585711562804e-06, + "loss": 2.5012, + "step": 21577 + }, + { + "epoch": 2.2819373942470387, + "grad_norm": 9.0, + "learning_rate": 2.9717903932279267e-06, + "loss": 2.4149, + "step": 21578 + }, + { + "epoch": 2.282043147208122, + "grad_norm": 8.375, + "learning_rate": 2.9716222130856537e-06, + "loss": 2.397, + "step": 21579 + }, + { + "epoch": 2.282148900169205, + "grad_norm": 8.4375, + "learning_rate": 2.9714540307302514e-06, + "loss": 2.5179, + "step": 21580 + }, + { + "epoch": 2.2822546531302876, + "grad_norm": 8.0, + "learning_rate": 2.971285846162508e-06, + "loss": 2.5768, + "step": 21581 + }, + { + "epoch": 2.2823604060913705, + "grad_norm": 8.0, + "learning_rate": 2.9711176593832113e-06, + "loss": 2.5157, + "step": 21582 + }, + { + "epoch": 2.2824661590524533, + "grad_norm": 8.375, + "learning_rate": 2.9709494703931528e-06, + "loss": 2.5363, + "step": 21583 + }, + { + "epoch": 2.2825719120135366, + "grad_norm": 8.625, + "learning_rate": 2.970781279193122e-06, + "loss": 2.6918, + "step": 21584 + }, + { + "epoch": 2.2826776649746194, + "grad_norm": 9.5625, + "learning_rate": 2.9706130857839065e-06, + "loss": 2.7068, + "step": 21585 + }, + { + "epoch": 2.2827834179357023, + "grad_norm": 8.375, + "learning_rate": 2.9704448901662963e-06, + "loss": 2.5577, + "step": 21586 + }, + { + "epoch": 2.282889170896785, + "grad_norm": 8.375, + "learning_rate": 2.9702766923410814e-06, + "loss": 2.684, + "step": 21587 + }, + { + "epoch": 2.282994923857868, + "grad_norm": 8.625, + "learning_rate": 2.9701084923090493e-06, + "loss": 2.5086, + "step": 21588 + }, + { + "epoch": 2.283100676818951, + "grad_norm": 8.5, + "learning_rate": 2.9699402900709906e-06, + "loss": 2.4195, + "step": 21589 + }, + { + "epoch": 2.2832064297800336, + "grad_norm": 7.90625, + "learning_rate": 2.9697720856276944e-06, + "loss": 2.5643, + "step": 21590 + }, + { + "epoch": 2.283312182741117, + "grad_norm": 9.5625, + "learning_rate": 2.9696038789799504e-06, + "loss": 2.5243, + "step": 21591 + }, + { + "epoch": 2.2834179357021998, + "grad_norm": 8.5625, + "learning_rate": 2.9694356701285465e-06, + "loss": 2.6668, + "step": 21592 + }, + { + "epoch": 2.2835236886632826, + "grad_norm": 8.0625, + "learning_rate": 2.9692674590742737e-06, + "loss": 2.6028, + "step": 21593 + }, + { + "epoch": 2.2836294416243654, + "grad_norm": 8.875, + "learning_rate": 2.9690992458179206e-06, + "loss": 2.763, + "step": 21594 + }, + { + "epoch": 2.2837351945854483, + "grad_norm": 8.3125, + "learning_rate": 2.968931030360276e-06, + "loss": 2.4971, + "step": 21595 + }, + { + "epoch": 2.2838409475465316, + "grad_norm": 8.75, + "learning_rate": 2.9687628127021306e-06, + "loss": 2.5803, + "step": 21596 + }, + { + "epoch": 2.2839467005076144, + "grad_norm": 8.3125, + "learning_rate": 2.9685945928442726e-06, + "loss": 2.4568, + "step": 21597 + }, + { + "epoch": 2.2840524534686972, + "grad_norm": 8.3125, + "learning_rate": 2.968426370787492e-06, + "loss": 2.492, + "step": 21598 + }, + { + "epoch": 2.28415820642978, + "grad_norm": 8.3125, + "learning_rate": 2.9682581465325775e-06, + "loss": 2.5323, + "step": 21599 + }, + { + "epoch": 2.284263959390863, + "grad_norm": 8.25, + "learning_rate": 2.9680899200803198e-06, + "loss": 2.5744, + "step": 21600 + }, + { + "epoch": 2.2843697123519457, + "grad_norm": 7.90625, + "learning_rate": 2.967921691431507e-06, + "loss": 2.4355, + "step": 21601 + }, + { + "epoch": 2.2844754653130286, + "grad_norm": 8.25, + "learning_rate": 2.9677534605869294e-06, + "loss": 2.6002, + "step": 21602 + }, + { + "epoch": 2.284581218274112, + "grad_norm": 8.25, + "learning_rate": 2.9675852275473765e-06, + "loss": 2.2134, + "step": 21603 + }, + { + "epoch": 2.2846869712351947, + "grad_norm": 8.8125, + "learning_rate": 2.9674169923136364e-06, + "loss": 2.4632, + "step": 21604 + }, + { + "epoch": 2.2847927241962775, + "grad_norm": 8.0625, + "learning_rate": 2.9672487548865002e-06, + "loss": 2.4972, + "step": 21605 + }, + { + "epoch": 2.2848984771573604, + "grad_norm": 8.8125, + "learning_rate": 2.967080515266757e-06, + "loss": 2.7657, + "step": 21606 + }, + { + "epoch": 2.2850042301184432, + "grad_norm": 8.6875, + "learning_rate": 2.9669122734551954e-06, + "loss": 2.5409, + "step": 21607 + }, + { + "epoch": 2.285109983079526, + "grad_norm": 8.4375, + "learning_rate": 2.966744029452605e-06, + "loss": 2.5153, + "step": 21608 + }, + { + "epoch": 2.285215736040609, + "grad_norm": 8.0625, + "learning_rate": 2.966575783259776e-06, + "loss": 2.6059, + "step": 21609 + }, + { + "epoch": 2.285321489001692, + "grad_norm": 8.8125, + "learning_rate": 2.9664075348774983e-06, + "loss": 2.3432, + "step": 21610 + }, + { + "epoch": 2.285427241962775, + "grad_norm": 8.0, + "learning_rate": 2.9662392843065606e-06, + "loss": 2.4489, + "step": 21611 + }, + { + "epoch": 2.285532994923858, + "grad_norm": 8.1875, + "learning_rate": 2.9660710315477524e-06, + "loss": 2.6692, + "step": 21612 + }, + { + "epoch": 2.2856387478849407, + "grad_norm": 8.875, + "learning_rate": 2.9659027766018637e-06, + "loss": 2.5677, + "step": 21613 + }, + { + "epoch": 2.2857445008460235, + "grad_norm": 8.625, + "learning_rate": 2.9657345194696834e-06, + "loss": 2.5707, + "step": 21614 + }, + { + "epoch": 2.285850253807107, + "grad_norm": 8.25, + "learning_rate": 2.9655662601520026e-06, + "loss": 2.4948, + "step": 21615 + }, + { + "epoch": 2.2859560067681897, + "grad_norm": 7.9375, + "learning_rate": 2.965397998649609e-06, + "loss": 2.5364, + "step": 21616 + }, + { + "epoch": 2.2860617597292725, + "grad_norm": 8.8125, + "learning_rate": 2.965229734963292e-06, + "loss": 2.5892, + "step": 21617 + }, + { + "epoch": 2.2861675126903553, + "grad_norm": 8.3125, + "learning_rate": 2.965061469093843e-06, + "loss": 2.5112, + "step": 21618 + }, + { + "epoch": 2.286273265651438, + "grad_norm": 8.0625, + "learning_rate": 2.9648932010420505e-06, + "loss": 2.5057, + "step": 21619 + }, + { + "epoch": 2.286379018612521, + "grad_norm": 8.4375, + "learning_rate": 2.9647249308087045e-06, + "loss": 2.7088, + "step": 21620 + }, + { + "epoch": 2.286484771573604, + "grad_norm": 8.5, + "learning_rate": 2.964556658394595e-06, + "loss": 2.4864, + "step": 21621 + }, + { + "epoch": 2.286590524534687, + "grad_norm": 8.8125, + "learning_rate": 2.9643883838005103e-06, + "loss": 2.4475, + "step": 21622 + }, + { + "epoch": 2.28669627749577, + "grad_norm": 8.8125, + "learning_rate": 2.9642201070272416e-06, + "loss": 2.5142, + "step": 21623 + }, + { + "epoch": 2.286802030456853, + "grad_norm": 7.75, + "learning_rate": 2.964051828075577e-06, + "loss": 2.3911, + "step": 21624 + }, + { + "epoch": 2.2869077834179357, + "grad_norm": 8.625, + "learning_rate": 2.9638835469463077e-06, + "loss": 2.659, + "step": 21625 + }, + { + "epoch": 2.2870135363790185, + "grad_norm": 8.4375, + "learning_rate": 2.9637152636402223e-06, + "loss": 2.46, + "step": 21626 + }, + { + "epoch": 2.2871192893401013, + "grad_norm": 8.125, + "learning_rate": 2.96354697815811e-06, + "loss": 2.4149, + "step": 21627 + }, + { + "epoch": 2.2872250423011846, + "grad_norm": 8.6875, + "learning_rate": 2.9633786905007623e-06, + "loss": 2.4339, + "step": 21628 + }, + { + "epoch": 2.2873307952622675, + "grad_norm": 9.125, + "learning_rate": 2.963210400668968e-06, + "loss": 2.5863, + "step": 21629 + }, + { + "epoch": 2.2874365482233503, + "grad_norm": 8.25, + "learning_rate": 2.963042108663516e-06, + "loss": 2.4852, + "step": 21630 + }, + { + "epoch": 2.287542301184433, + "grad_norm": 8.625, + "learning_rate": 2.9628738144851975e-06, + "loss": 2.4315, + "step": 21631 + }, + { + "epoch": 2.287648054145516, + "grad_norm": 8.6875, + "learning_rate": 2.9627055181348014e-06, + "loss": 2.5316, + "step": 21632 + }, + { + "epoch": 2.287753807106599, + "grad_norm": 8.3125, + "learning_rate": 2.9625372196131168e-06, + "loss": 2.3949, + "step": 21633 + }, + { + "epoch": 2.287859560067682, + "grad_norm": 8.3125, + "learning_rate": 2.962368918920935e-06, + "loss": 2.6861, + "step": 21634 + }, + { + "epoch": 2.287965313028765, + "grad_norm": 8.875, + "learning_rate": 2.9622006160590444e-06, + "loss": 2.6124, + "step": 21635 + }, + { + "epoch": 2.2880710659898478, + "grad_norm": 8.25, + "learning_rate": 2.9620323110282356e-06, + "loss": 2.4594, + "step": 21636 + }, + { + "epoch": 2.2881768189509306, + "grad_norm": 8.5, + "learning_rate": 2.9618640038292986e-06, + "loss": 2.5048, + "step": 21637 + }, + { + "epoch": 2.2882825719120135, + "grad_norm": 8.5625, + "learning_rate": 2.961695694463022e-06, + "loss": 2.1154, + "step": 21638 + }, + { + "epoch": 2.2883883248730963, + "grad_norm": 8.625, + "learning_rate": 2.9615273829301967e-06, + "loss": 2.5812, + "step": 21639 + }, + { + "epoch": 2.288494077834179, + "grad_norm": 8.3125, + "learning_rate": 2.9613590692316125e-06, + "loss": 2.6239, + "step": 21640 + }, + { + "epoch": 2.2885998307952624, + "grad_norm": 8.1875, + "learning_rate": 2.961190753368058e-06, + "loss": 2.5768, + "step": 21641 + }, + { + "epoch": 2.2887055837563453, + "grad_norm": 8.3125, + "learning_rate": 2.9610224353403246e-06, + "loss": 2.5811, + "step": 21642 + }, + { + "epoch": 2.288811336717428, + "grad_norm": 8.1875, + "learning_rate": 2.960854115149201e-06, + "loss": 2.6333, + "step": 21643 + }, + { + "epoch": 2.288917089678511, + "grad_norm": 8.6875, + "learning_rate": 2.960685792795478e-06, + "loss": 2.6052, + "step": 21644 + }, + { + "epoch": 2.2890228426395938, + "grad_norm": 7.9375, + "learning_rate": 2.960517468279945e-06, + "loss": 2.387, + "step": 21645 + }, + { + "epoch": 2.289128595600677, + "grad_norm": 8.125, + "learning_rate": 2.9603491416033912e-06, + "loss": 2.4435, + "step": 21646 + }, + { + "epoch": 2.28923434856176, + "grad_norm": 9.0, + "learning_rate": 2.9601808127666075e-06, + "loss": 2.4899, + "step": 21647 + }, + { + "epoch": 2.2893401015228427, + "grad_norm": 7.8125, + "learning_rate": 2.960012481770384e-06, + "loss": 2.4865, + "step": 21648 + }, + { + "epoch": 2.2894458544839256, + "grad_norm": 8.625, + "learning_rate": 2.9598441486155093e-06, + "loss": 2.551, + "step": 21649 + }, + { + "epoch": 2.2895516074450084, + "grad_norm": 8.375, + "learning_rate": 2.959675813302775e-06, + "loss": 2.7347, + "step": 21650 + }, + { + "epoch": 2.2896573604060912, + "grad_norm": 8.1875, + "learning_rate": 2.959507475832969e-06, + "loss": 2.5334, + "step": 21651 + }, + { + "epoch": 2.289763113367174, + "grad_norm": 8.1875, + "learning_rate": 2.9593391362068826e-06, + "loss": 2.527, + "step": 21652 + }, + { + "epoch": 2.2898688663282574, + "grad_norm": 9.125, + "learning_rate": 2.9591707944253055e-06, + "loss": 2.5965, + "step": 21653 + }, + { + "epoch": 2.28997461928934, + "grad_norm": 8.4375, + "learning_rate": 2.959002450489028e-06, + "loss": 2.566, + "step": 21654 + }, + { + "epoch": 2.290080372250423, + "grad_norm": 7.78125, + "learning_rate": 2.9588341043988394e-06, + "loss": 2.4613, + "step": 21655 + }, + { + "epoch": 2.290186125211506, + "grad_norm": 8.25, + "learning_rate": 2.9586657561555305e-06, + "loss": 2.4835, + "step": 21656 + }, + { + "epoch": 2.2902918781725887, + "grad_norm": 7.5625, + "learning_rate": 2.958497405759891e-06, + "loss": 2.4662, + "step": 21657 + }, + { + "epoch": 2.2903976311336716, + "grad_norm": 9.3125, + "learning_rate": 2.9583290532127094e-06, + "loss": 2.5555, + "step": 21658 + }, + { + "epoch": 2.290503384094755, + "grad_norm": 8.4375, + "learning_rate": 2.958160698514778e-06, + "loss": 2.4036, + "step": 21659 + }, + { + "epoch": 2.2906091370558377, + "grad_norm": 8.125, + "learning_rate": 2.957992341666886e-06, + "loss": 2.4318, + "step": 21660 + }, + { + "epoch": 2.2907148900169205, + "grad_norm": 8.75, + "learning_rate": 2.9578239826698225e-06, + "loss": 2.5263, + "step": 21661 + }, + { + "epoch": 2.2908206429780034, + "grad_norm": 8.0625, + "learning_rate": 2.9576556215243785e-06, + "loss": 2.7473, + "step": 21662 + }, + { + "epoch": 2.290926395939086, + "grad_norm": 8.0625, + "learning_rate": 2.957487258231344e-06, + "loss": 2.5236, + "step": 21663 + }, + { + "epoch": 2.291032148900169, + "grad_norm": 8.1875, + "learning_rate": 2.957318892791509e-06, + "loss": 2.4299, + "step": 21664 + }, + { + "epoch": 2.2911379018612523, + "grad_norm": 8.8125, + "learning_rate": 2.9571505252056632e-06, + "loss": 2.6379, + "step": 21665 + }, + { + "epoch": 2.291243654822335, + "grad_norm": 8.5, + "learning_rate": 2.9569821554745976e-06, + "loss": 2.5662, + "step": 21666 + }, + { + "epoch": 2.291349407783418, + "grad_norm": 8.625, + "learning_rate": 2.9568137835991012e-06, + "loss": 2.4928, + "step": 21667 + }, + { + "epoch": 2.291455160744501, + "grad_norm": 8.1875, + "learning_rate": 2.956645409579965e-06, + "loss": 2.3881, + "step": 21668 + }, + { + "epoch": 2.2915609137055837, + "grad_norm": 8.125, + "learning_rate": 2.9564770334179777e-06, + "loss": 2.589, + "step": 21669 + }, + { + "epoch": 2.2916666666666665, + "grad_norm": 8.375, + "learning_rate": 2.9563086551139307e-06, + "loss": 2.5897, + "step": 21670 + }, + { + "epoch": 2.2917724196277494, + "grad_norm": 8.0625, + "learning_rate": 2.9561402746686137e-06, + "loss": 2.4711, + "step": 21671 + }, + { + "epoch": 2.2918781725888326, + "grad_norm": 8.375, + "learning_rate": 2.9559718920828172e-06, + "loss": 2.6336, + "step": 21672 + }, + { + "epoch": 2.2919839255499155, + "grad_norm": 8.25, + "learning_rate": 2.9558035073573317e-06, + "loss": 2.6034, + "step": 21673 + }, + { + "epoch": 2.2920896785109983, + "grad_norm": 8.5625, + "learning_rate": 2.9556351204929457e-06, + "loss": 2.5193, + "step": 21674 + }, + { + "epoch": 2.292195431472081, + "grad_norm": 8.125, + "learning_rate": 2.955466731490451e-06, + "loss": 2.3476, + "step": 21675 + }, + { + "epoch": 2.292301184433164, + "grad_norm": 8.125, + "learning_rate": 2.9552983403506376e-06, + "loss": 2.3875, + "step": 21676 + }, + { + "epoch": 2.2924069373942473, + "grad_norm": 8.8125, + "learning_rate": 2.9551299470742943e-06, + "loss": 2.4855, + "step": 21677 + }, + { + "epoch": 2.29251269035533, + "grad_norm": 8.5, + "learning_rate": 2.9549615516622134e-06, + "loss": 2.5454, + "step": 21678 + }, + { + "epoch": 2.292618443316413, + "grad_norm": 8.0, + "learning_rate": 2.9547931541151825e-06, + "loss": 2.4682, + "step": 21679 + }, + { + "epoch": 2.292724196277496, + "grad_norm": 8.375, + "learning_rate": 2.9546247544339945e-06, + "loss": 2.3396, + "step": 21680 + }, + { + "epoch": 2.2928299492385786, + "grad_norm": 8.375, + "learning_rate": 2.954456352619438e-06, + "loss": 2.4906, + "step": 21681 + }, + { + "epoch": 2.2929357021996615, + "grad_norm": 8.125, + "learning_rate": 2.954287948672304e-06, + "loss": 2.5098, + "step": 21682 + }, + { + "epoch": 2.2930414551607443, + "grad_norm": 8.25, + "learning_rate": 2.9541195425933828e-06, + "loss": 2.5886, + "step": 21683 + }, + { + "epoch": 2.2931472081218276, + "grad_norm": 8.375, + "learning_rate": 2.9539511343834637e-06, + "loss": 2.3074, + "step": 21684 + }, + { + "epoch": 2.2932529610829104, + "grad_norm": 8.9375, + "learning_rate": 2.9537827240433376e-06, + "loss": 2.6465, + "step": 21685 + }, + { + "epoch": 2.2933587140439933, + "grad_norm": 8.1875, + "learning_rate": 2.9536143115737953e-06, + "loss": 2.6166, + "step": 21686 + }, + { + "epoch": 2.293464467005076, + "grad_norm": 8.1875, + "learning_rate": 2.9534458969756257e-06, + "loss": 2.6299, + "step": 21687 + }, + { + "epoch": 2.293570219966159, + "grad_norm": 8.8125, + "learning_rate": 2.95327748024962e-06, + "loss": 2.6489, + "step": 21688 + }, + { + "epoch": 2.293675972927242, + "grad_norm": 8.4375, + "learning_rate": 2.9531090613965697e-06, + "loss": 2.604, + "step": 21689 + }, + { + "epoch": 2.2937817258883246, + "grad_norm": 8.625, + "learning_rate": 2.952940640417262e-06, + "loss": 2.5766, + "step": 21690 + }, + { + "epoch": 2.293887478849408, + "grad_norm": 8.875, + "learning_rate": 2.952772217312491e-06, + "loss": 2.2815, + "step": 21691 + }, + { + "epoch": 2.2939932318104908, + "grad_norm": 8.5625, + "learning_rate": 2.952603792083045e-06, + "loss": 2.2988, + "step": 21692 + }, + { + "epoch": 2.2940989847715736, + "grad_norm": 8.4375, + "learning_rate": 2.9524353647297132e-06, + "loss": 2.4538, + "step": 21693 + }, + { + "epoch": 2.2942047377326564, + "grad_norm": 8.6875, + "learning_rate": 2.9522669352532884e-06, + "loss": 2.4801, + "step": 21694 + }, + { + "epoch": 2.2943104906937393, + "grad_norm": 8.625, + "learning_rate": 2.95209850365456e-06, + "loss": 2.5001, + "step": 21695 + }, + { + "epoch": 2.2944162436548226, + "grad_norm": 8.3125, + "learning_rate": 2.951930069934317e-06, + "loss": 2.5476, + "step": 21696 + }, + { + "epoch": 2.2945219966159054, + "grad_norm": 8.75, + "learning_rate": 2.951761634093351e-06, + "loss": 2.5037, + "step": 21697 + }, + { + "epoch": 2.2946277495769882, + "grad_norm": 8.1875, + "learning_rate": 2.9515931961324536e-06, + "loss": 2.5447, + "step": 21698 + }, + { + "epoch": 2.294733502538071, + "grad_norm": 8.5, + "learning_rate": 2.9514247560524134e-06, + "loss": 2.624, + "step": 21699 + }, + { + "epoch": 2.294839255499154, + "grad_norm": 8.25, + "learning_rate": 2.9512563138540212e-06, + "loss": 2.7767, + "step": 21700 + }, + { + "epoch": 2.2949450084602367, + "grad_norm": 8.375, + "learning_rate": 2.9510878695380683e-06, + "loss": 2.4994, + "step": 21701 + }, + { + "epoch": 2.2950507614213196, + "grad_norm": 7.8125, + "learning_rate": 2.9509194231053445e-06, + "loss": 2.5047, + "step": 21702 + }, + { + "epoch": 2.295156514382403, + "grad_norm": 8.25, + "learning_rate": 2.95075097455664e-06, + "loss": 2.5024, + "step": 21703 + }, + { + "epoch": 2.2952622673434857, + "grad_norm": 8.9375, + "learning_rate": 2.9505825238927456e-06, + "loss": 2.2855, + "step": 21704 + }, + { + "epoch": 2.2953680203045685, + "grad_norm": 8.1875, + "learning_rate": 2.950414071114451e-06, + "loss": 2.4896, + "step": 21705 + }, + { + "epoch": 2.2954737732656514, + "grad_norm": 8.75, + "learning_rate": 2.9502456162225478e-06, + "loss": 2.5549, + "step": 21706 + }, + { + "epoch": 2.2955795262267342, + "grad_norm": 9.25, + "learning_rate": 2.950077159217827e-06, + "loss": 2.6975, + "step": 21707 + }, + { + "epoch": 2.2956852791878175, + "grad_norm": 9.0625, + "learning_rate": 2.9499087001010774e-06, + "loss": 2.5845, + "step": 21708 + }, + { + "epoch": 2.2957910321489003, + "grad_norm": 8.5625, + "learning_rate": 2.94974023887309e-06, + "loss": 2.3904, + "step": 21709 + }, + { + "epoch": 2.295896785109983, + "grad_norm": 8.1875, + "learning_rate": 2.949571775534656e-06, + "loss": 2.5396, + "step": 21710 + }, + { + "epoch": 2.296002538071066, + "grad_norm": 8.125, + "learning_rate": 2.9494033100865655e-06, + "loss": 2.4927, + "step": 21711 + }, + { + "epoch": 2.296108291032149, + "grad_norm": 8.8125, + "learning_rate": 2.949234842529609e-06, + "loss": 2.4714, + "step": 21712 + }, + { + "epoch": 2.2962140439932317, + "grad_norm": 8.3125, + "learning_rate": 2.9490663728645774e-06, + "loss": 2.8002, + "step": 21713 + }, + { + "epoch": 2.2963197969543145, + "grad_norm": 8.0, + "learning_rate": 2.9488979010922603e-06, + "loss": 2.3971, + "step": 21714 + }, + { + "epoch": 2.296425549915398, + "grad_norm": 8.625, + "learning_rate": 2.9487294272134493e-06, + "loss": 2.3644, + "step": 21715 + }, + { + "epoch": 2.2965313028764807, + "grad_norm": 8.125, + "learning_rate": 2.9485609512289343e-06, + "loss": 2.4808, + "step": 21716 + }, + { + "epoch": 2.2966370558375635, + "grad_norm": 8.125, + "learning_rate": 2.948392473139507e-06, + "loss": 2.5375, + "step": 21717 + }, + { + "epoch": 2.2967428087986463, + "grad_norm": 8.375, + "learning_rate": 2.9482239929459573e-06, + "loss": 2.5534, + "step": 21718 + }, + { + "epoch": 2.296848561759729, + "grad_norm": 8.9375, + "learning_rate": 2.9480555106490745e-06, + "loss": 2.6342, + "step": 21719 + }, + { + "epoch": 2.296954314720812, + "grad_norm": 8.6875, + "learning_rate": 2.947887026249652e-06, + "loss": 2.7012, + "step": 21720 + }, + { + "epoch": 2.297060067681895, + "grad_norm": 8.375, + "learning_rate": 2.9477185397484775e-06, + "loss": 2.3487, + "step": 21721 + }, + { + "epoch": 2.297165820642978, + "grad_norm": 8.375, + "learning_rate": 2.9475500511463433e-06, + "loss": 2.4901, + "step": 21722 + }, + { + "epoch": 2.297271573604061, + "grad_norm": 8.6875, + "learning_rate": 2.9473815604440396e-06, + "loss": 2.6366, + "step": 21723 + }, + { + "epoch": 2.297377326565144, + "grad_norm": 8.4375, + "learning_rate": 2.9472130676423576e-06, + "loss": 2.5859, + "step": 21724 + }, + { + "epoch": 2.2974830795262267, + "grad_norm": 8.5, + "learning_rate": 2.9470445727420875e-06, + "loss": 2.7545, + "step": 21725 + }, + { + "epoch": 2.2975888324873095, + "grad_norm": 8.0, + "learning_rate": 2.94687607574402e-06, + "loss": 2.6175, + "step": 21726 + }, + { + "epoch": 2.297694585448393, + "grad_norm": 8.25, + "learning_rate": 2.9467075766489467e-06, + "loss": 2.5445, + "step": 21727 + }, + { + "epoch": 2.2978003384094756, + "grad_norm": 8.25, + "learning_rate": 2.946539075457656e-06, + "loss": 2.7158, + "step": 21728 + }, + { + "epoch": 2.2979060913705585, + "grad_norm": 8.9375, + "learning_rate": 2.9463705721709413e-06, + "loss": 2.7735, + "step": 21729 + }, + { + "epoch": 2.2980118443316413, + "grad_norm": 8.1875, + "learning_rate": 2.9462020667895913e-06, + "loss": 2.2163, + "step": 21730 + }, + { + "epoch": 2.298117597292724, + "grad_norm": 8.8125, + "learning_rate": 2.9460335593143976e-06, + "loss": 2.74, + "step": 21731 + }, + { + "epoch": 2.298223350253807, + "grad_norm": 8.3125, + "learning_rate": 2.9458650497461507e-06, + "loss": 2.6348, + "step": 21732 + }, + { + "epoch": 2.29832910321489, + "grad_norm": 8.3125, + "learning_rate": 2.9456965380856418e-06, + "loss": 2.4625, + "step": 21733 + }, + { + "epoch": 2.298434856175973, + "grad_norm": 7.90625, + "learning_rate": 2.9455280243336615e-06, + "loss": 2.2728, + "step": 21734 + }, + { + "epoch": 2.298540609137056, + "grad_norm": 8.375, + "learning_rate": 2.945359508491e-06, + "loss": 2.4977, + "step": 21735 + }, + { + "epoch": 2.2986463620981388, + "grad_norm": 8.5625, + "learning_rate": 2.9451909905584487e-06, + "loss": 2.5027, + "step": 21736 + }, + { + "epoch": 2.2987521150592216, + "grad_norm": 8.125, + "learning_rate": 2.945022470536798e-06, + "loss": 2.3361, + "step": 21737 + }, + { + "epoch": 2.2988578680203045, + "grad_norm": 8.75, + "learning_rate": 2.9448539484268383e-06, + "loss": 2.6233, + "step": 21738 + }, + { + "epoch": 2.2989636209813873, + "grad_norm": 8.0, + "learning_rate": 2.944685424229362e-06, + "loss": 2.5364, + "step": 21739 + }, + { + "epoch": 2.2990693739424706, + "grad_norm": 8.1875, + "learning_rate": 2.944516897945159e-06, + "loss": 2.4582, + "step": 21740 + }, + { + "epoch": 2.2991751269035534, + "grad_norm": 8.375, + "learning_rate": 2.9443483695750184e-06, + "loss": 2.3816, + "step": 21741 + }, + { + "epoch": 2.2992808798646363, + "grad_norm": 8.375, + "learning_rate": 2.944179839119734e-06, + "loss": 2.5903, + "step": 21742 + }, + { + "epoch": 2.299386632825719, + "grad_norm": 8.3125, + "learning_rate": 2.944011306580095e-06, + "loss": 2.4135, + "step": 21743 + }, + { + "epoch": 2.299492385786802, + "grad_norm": 8.4375, + "learning_rate": 2.943842771956892e-06, + "loss": 2.621, + "step": 21744 + }, + { + "epoch": 2.2995981387478848, + "grad_norm": 8.125, + "learning_rate": 2.9436742352509175e-06, + "loss": 2.4489, + "step": 21745 + }, + { + "epoch": 2.299703891708968, + "grad_norm": 9.125, + "learning_rate": 2.9435056964629604e-06, + "loss": 2.5068, + "step": 21746 + }, + { + "epoch": 2.299809644670051, + "grad_norm": 9.0, + "learning_rate": 2.943337155593813e-06, + "loss": 2.3031, + "step": 21747 + }, + { + "epoch": 2.2999153976311337, + "grad_norm": 8.0625, + "learning_rate": 2.9431686126442657e-06, + "loss": 2.4297, + "step": 21748 + }, + { + "epoch": 2.3000211505922166, + "grad_norm": 8.3125, + "learning_rate": 2.943000067615109e-06, + "loss": 2.2248, + "step": 21749 + }, + { + "epoch": 2.3001269035532994, + "grad_norm": 8.375, + "learning_rate": 2.9428315205071335e-06, + "loss": 2.3128, + "step": 21750 + }, + { + "epoch": 2.3002326565143822, + "grad_norm": 9.125, + "learning_rate": 2.942662971321131e-06, + "loss": 2.5971, + "step": 21751 + }, + { + "epoch": 2.300338409475465, + "grad_norm": 8.8125, + "learning_rate": 2.9424944200578933e-06, + "loss": 2.4921, + "step": 21752 + }, + { + "epoch": 2.3004441624365484, + "grad_norm": 8.5625, + "learning_rate": 2.9423258667182096e-06, + "loss": 2.6658, + "step": 21753 + }, + { + "epoch": 2.300549915397631, + "grad_norm": 9.125, + "learning_rate": 2.9421573113028713e-06, + "loss": 2.7111, + "step": 21754 + }, + { + "epoch": 2.300655668358714, + "grad_norm": 8.1875, + "learning_rate": 2.9419887538126702e-06, + "loss": 2.6129, + "step": 21755 + }, + { + "epoch": 2.300761421319797, + "grad_norm": 8.125, + "learning_rate": 2.9418201942483964e-06, + "loss": 2.6023, + "step": 21756 + }, + { + "epoch": 2.3008671742808797, + "grad_norm": 8.4375, + "learning_rate": 2.941651632610841e-06, + "loss": 2.5467, + "step": 21757 + }, + { + "epoch": 2.300972927241963, + "grad_norm": 8.75, + "learning_rate": 2.9414830689007947e-06, + "loss": 2.5788, + "step": 21758 + }, + { + "epoch": 2.301078680203046, + "grad_norm": 8.125, + "learning_rate": 2.9413145031190495e-06, + "loss": 2.3983, + "step": 21759 + }, + { + "epoch": 2.3011844331641287, + "grad_norm": 8.0, + "learning_rate": 2.9411459352663952e-06, + "loss": 2.5255, + "step": 21760 + }, + { + "epoch": 2.3012901861252115, + "grad_norm": 8.6875, + "learning_rate": 2.940977365343624e-06, + "loss": 2.5052, + "step": 21761 + }, + { + "epoch": 2.3013959390862944, + "grad_norm": 8.375, + "learning_rate": 2.940808793351526e-06, + "loss": 2.4266, + "step": 21762 + }, + { + "epoch": 2.301501692047377, + "grad_norm": 7.625, + "learning_rate": 2.940640219290893e-06, + "loss": 2.428, + "step": 21763 + }, + { + "epoch": 2.30160744500846, + "grad_norm": 8.9375, + "learning_rate": 2.9404716431625158e-06, + "loss": 2.5969, + "step": 21764 + }, + { + "epoch": 2.3017131979695433, + "grad_norm": 8.125, + "learning_rate": 2.9403030649671856e-06, + "loss": 2.2618, + "step": 21765 + }, + { + "epoch": 2.301818950930626, + "grad_norm": 8.625, + "learning_rate": 2.940134484705692e-06, + "loss": 2.4575, + "step": 21766 + }, + { + "epoch": 2.301924703891709, + "grad_norm": 8.6875, + "learning_rate": 2.9399659023788274e-06, + "loss": 2.4854, + "step": 21767 + }, + { + "epoch": 2.302030456852792, + "grad_norm": 8.625, + "learning_rate": 2.939797317987384e-06, + "loss": 2.5525, + "step": 21768 + }, + { + "epoch": 2.3021362098138747, + "grad_norm": 8.75, + "learning_rate": 2.939628731532151e-06, + "loss": 2.5799, + "step": 21769 + }, + { + "epoch": 2.3022419627749575, + "grad_norm": 8.5625, + "learning_rate": 2.9394601430139196e-06, + "loss": 2.5038, + "step": 21770 + }, + { + "epoch": 2.302347715736041, + "grad_norm": 8.1875, + "learning_rate": 2.9392915524334825e-06, + "loss": 2.4685, + "step": 21771 + }, + { + "epoch": 2.3024534686971236, + "grad_norm": 8.9375, + "learning_rate": 2.9391229597916293e-06, + "loss": 2.6334, + "step": 21772 + }, + { + "epoch": 2.3025592216582065, + "grad_norm": 8.25, + "learning_rate": 2.9389543650891516e-06, + "loss": 2.5188, + "step": 21773 + }, + { + "epoch": 2.3026649746192893, + "grad_norm": 8.375, + "learning_rate": 2.9387857683268405e-06, + "loss": 2.4826, + "step": 21774 + }, + { + "epoch": 2.302770727580372, + "grad_norm": 8.0625, + "learning_rate": 2.9386171695054877e-06, + "loss": 2.3626, + "step": 21775 + }, + { + "epoch": 2.302876480541455, + "grad_norm": 8.9375, + "learning_rate": 2.938448568625883e-06, + "loss": 2.5509, + "step": 21776 + }, + { + "epoch": 2.3029822335025383, + "grad_norm": 8.625, + "learning_rate": 2.9382799656888194e-06, + "loss": 2.5407, + "step": 21777 + }, + { + "epoch": 2.303087986463621, + "grad_norm": 7.96875, + "learning_rate": 2.938111360695087e-06, + "loss": 2.4449, + "step": 21778 + }, + { + "epoch": 2.303193739424704, + "grad_norm": 8.0, + "learning_rate": 2.937942753645477e-06, + "loss": 2.6507, + "step": 21779 + }, + { + "epoch": 2.303299492385787, + "grad_norm": 7.9375, + "learning_rate": 2.937774144540781e-06, + "loss": 2.4149, + "step": 21780 + }, + { + "epoch": 2.3034052453468696, + "grad_norm": 8.125, + "learning_rate": 2.93760553338179e-06, + "loss": 2.3495, + "step": 21781 + }, + { + "epoch": 2.3035109983079525, + "grad_norm": 8.5625, + "learning_rate": 2.937436920169294e-06, + "loss": 2.4781, + "step": 21782 + }, + { + "epoch": 2.3036167512690353, + "grad_norm": 8.125, + "learning_rate": 2.9372683049040874e-06, + "loss": 2.4662, + "step": 21783 + }, + { + "epoch": 2.3037225042301186, + "grad_norm": 8.75, + "learning_rate": 2.9370996875869583e-06, + "loss": 2.2835, + "step": 21784 + }, + { + "epoch": 2.3038282571912014, + "grad_norm": 8.1875, + "learning_rate": 2.936931068218699e-06, + "loss": 2.3654, + "step": 21785 + }, + { + "epoch": 2.3039340101522843, + "grad_norm": 8.375, + "learning_rate": 2.936762446800101e-06, + "loss": 2.5805, + "step": 21786 + }, + { + "epoch": 2.304039763113367, + "grad_norm": 8.125, + "learning_rate": 2.936593823331956e-06, + "loss": 2.6173, + "step": 21787 + }, + { + "epoch": 2.30414551607445, + "grad_norm": 8.6875, + "learning_rate": 2.9364251978150547e-06, + "loss": 2.5527, + "step": 21788 + }, + { + "epoch": 2.3042512690355332, + "grad_norm": 8.25, + "learning_rate": 2.9362565702501877e-06, + "loss": 2.4389, + "step": 21789 + }, + { + "epoch": 2.304357021996616, + "grad_norm": 8.0625, + "learning_rate": 2.9360879406381483e-06, + "loss": 2.5843, + "step": 21790 + }, + { + "epoch": 2.304462774957699, + "grad_norm": 8.25, + "learning_rate": 2.9359193089797256e-06, + "loss": 2.6589, + "step": 21791 + }, + { + "epoch": 2.3045685279187818, + "grad_norm": 8.5, + "learning_rate": 2.935750675275712e-06, + "loss": 2.6313, + "step": 21792 + }, + { + "epoch": 2.3046742808798646, + "grad_norm": 8.0, + "learning_rate": 2.9355820395268986e-06, + "loss": 2.4389, + "step": 21793 + }, + { + "epoch": 2.3047800338409474, + "grad_norm": 8.4375, + "learning_rate": 2.935413401734077e-06, + "loss": 2.429, + "step": 21794 + }, + { + "epoch": 2.3048857868020303, + "grad_norm": 8.375, + "learning_rate": 2.935244761898038e-06, + "loss": 2.6727, + "step": 21795 + }, + { + "epoch": 2.3049915397631136, + "grad_norm": 9.25, + "learning_rate": 2.935076120019574e-06, + "loss": 2.7077, + "step": 21796 + }, + { + "epoch": 2.3050972927241964, + "grad_norm": 8.0, + "learning_rate": 2.9349074760994756e-06, + "loss": 2.5265, + "step": 21797 + }, + { + "epoch": 2.3052030456852792, + "grad_norm": 8.125, + "learning_rate": 2.9347388301385336e-06, + "loss": 2.2051, + "step": 21798 + }, + { + "epoch": 2.305308798646362, + "grad_norm": 8.375, + "learning_rate": 2.934570182137541e-06, + "loss": 2.448, + "step": 21799 + }, + { + "epoch": 2.305414551607445, + "grad_norm": 8.875, + "learning_rate": 2.934401532097288e-06, + "loss": 2.3012, + "step": 21800 + }, + { + "epoch": 2.3055203045685277, + "grad_norm": 8.25, + "learning_rate": 2.934232880018566e-06, + "loss": 2.4753, + "step": 21801 + }, + { + "epoch": 2.3056260575296106, + "grad_norm": 8.625, + "learning_rate": 2.934064225902167e-06, + "loss": 2.5751, + "step": 21802 + }, + { + "epoch": 2.305731810490694, + "grad_norm": 8.9375, + "learning_rate": 2.933895569748881e-06, + "loss": 2.5823, + "step": 21803 + }, + { + "epoch": 2.3058375634517767, + "grad_norm": 8.3125, + "learning_rate": 2.9337269115595024e-06, + "loss": 2.1944, + "step": 21804 + }, + { + "epoch": 2.3059433164128595, + "grad_norm": 8.5625, + "learning_rate": 2.9335582513348194e-06, + "loss": 2.6399, + "step": 21805 + }, + { + "epoch": 2.3060490693739424, + "grad_norm": 8.3125, + "learning_rate": 2.933389589075625e-06, + "loss": 2.2054, + "step": 21806 + }, + { + "epoch": 2.3061548223350252, + "grad_norm": 8.125, + "learning_rate": 2.9332209247827116e-06, + "loss": 2.6032, + "step": 21807 + }, + { + "epoch": 2.3062605752961085, + "grad_norm": 8.25, + "learning_rate": 2.933052258456868e-06, + "loss": 2.5454, + "step": 21808 + }, + { + "epoch": 2.3063663282571913, + "grad_norm": 8.0625, + "learning_rate": 2.932883590098889e-06, + "loss": 2.3157, + "step": 21809 + }, + { + "epoch": 2.306472081218274, + "grad_norm": 8.3125, + "learning_rate": 2.9327149197095632e-06, + "loss": 2.1832, + "step": 21810 + }, + { + "epoch": 2.306577834179357, + "grad_norm": 8.5, + "learning_rate": 2.932546247289683e-06, + "loss": 2.6167, + "step": 21811 + }, + { + "epoch": 2.30668358714044, + "grad_norm": 8.3125, + "learning_rate": 2.9323775728400407e-06, + "loss": 2.4235, + "step": 21812 + }, + { + "epoch": 2.3067893401015227, + "grad_norm": 8.4375, + "learning_rate": 2.9322088963614274e-06, + "loss": 2.4274, + "step": 21813 + }, + { + "epoch": 2.3068950930626055, + "grad_norm": 8.75, + "learning_rate": 2.9320402178546344e-06, + "loss": 2.6929, + "step": 21814 + }, + { + "epoch": 2.307000846023689, + "grad_norm": 7.96875, + "learning_rate": 2.931871537320453e-06, + "loss": 2.495, + "step": 21815 + }, + { + "epoch": 2.3071065989847717, + "grad_norm": 8.5, + "learning_rate": 2.931702854759676e-06, + "loss": 2.6692, + "step": 21816 + }, + { + "epoch": 2.3072123519458545, + "grad_norm": 8.6875, + "learning_rate": 2.931534170173093e-06, + "loss": 2.685, + "step": 21817 + }, + { + "epoch": 2.3073181049069373, + "grad_norm": 8.25, + "learning_rate": 2.9313654835614976e-06, + "loss": 2.3976, + "step": 21818 + }, + { + "epoch": 2.30742385786802, + "grad_norm": 8.125, + "learning_rate": 2.93119679492568e-06, + "loss": 2.2295, + "step": 21819 + }, + { + "epoch": 2.3075296108291035, + "grad_norm": 8.0625, + "learning_rate": 2.931028104266432e-06, + "loss": 2.4089, + "step": 21820 + }, + { + "epoch": 2.3076353637901863, + "grad_norm": 8.25, + "learning_rate": 2.9308594115845446e-06, + "loss": 2.5791, + "step": 21821 + }, + { + "epoch": 2.307741116751269, + "grad_norm": 8.3125, + "learning_rate": 2.9306907168808113e-06, + "loss": 2.3526, + "step": 21822 + }, + { + "epoch": 2.307846869712352, + "grad_norm": 7.84375, + "learning_rate": 2.930522020156023e-06, + "loss": 2.5028, + "step": 21823 + }, + { + "epoch": 2.307952622673435, + "grad_norm": 8.125, + "learning_rate": 2.9303533214109697e-06, + "loss": 2.5123, + "step": 21824 + }, + { + "epoch": 2.3080583756345177, + "grad_norm": 8.1875, + "learning_rate": 2.930184620646445e-06, + "loss": 2.5144, + "step": 21825 + }, + { + "epoch": 2.3081641285956005, + "grad_norm": 8.6875, + "learning_rate": 2.9300159178632404e-06, + "loss": 2.5393, + "step": 21826 + }, + { + "epoch": 2.308269881556684, + "grad_norm": 8.4375, + "learning_rate": 2.929847213062146e-06, + "loss": 2.4434, + "step": 21827 + }, + { + "epoch": 2.3083756345177666, + "grad_norm": 8.375, + "learning_rate": 2.9296785062439546e-06, + "loss": 2.7482, + "step": 21828 + }, + { + "epoch": 2.3084813874788495, + "grad_norm": 8.3125, + "learning_rate": 2.9295097974094578e-06, + "loss": 2.6358, + "step": 21829 + }, + { + "epoch": 2.3085871404399323, + "grad_norm": 9.25, + "learning_rate": 2.9293410865594473e-06, + "loss": 2.5982, + "step": 21830 + }, + { + "epoch": 2.308692893401015, + "grad_norm": 8.5, + "learning_rate": 2.9291723736947143e-06, + "loss": 2.5105, + "step": 21831 + }, + { + "epoch": 2.308798646362098, + "grad_norm": 8.5, + "learning_rate": 2.9290036588160516e-06, + "loss": 2.6912, + "step": 21832 + }, + { + "epoch": 2.308904399323181, + "grad_norm": 9.0, + "learning_rate": 2.928834941924249e-06, + "loss": 2.5887, + "step": 21833 + }, + { + "epoch": 2.309010152284264, + "grad_norm": 9.0, + "learning_rate": 2.928666223020101e-06, + "loss": 2.5522, + "step": 21834 + }, + { + "epoch": 2.309115905245347, + "grad_norm": 8.4375, + "learning_rate": 2.9284975021043964e-06, + "loss": 2.7198, + "step": 21835 + }, + { + "epoch": 2.3092216582064298, + "grad_norm": 8.8125, + "learning_rate": 2.9283287791779285e-06, + "loss": 2.7084, + "step": 21836 + }, + { + "epoch": 2.3093274111675126, + "grad_norm": 8.8125, + "learning_rate": 2.9281600542414894e-06, + "loss": 2.8108, + "step": 21837 + }, + { + "epoch": 2.3094331641285955, + "grad_norm": 8.5, + "learning_rate": 2.92799132729587e-06, + "loss": 2.6053, + "step": 21838 + }, + { + "epoch": 2.3095389170896787, + "grad_norm": 7.90625, + "learning_rate": 2.9278225983418623e-06, + "loss": 2.4215, + "step": 21839 + }, + { + "epoch": 2.3096446700507616, + "grad_norm": 10.8125, + "learning_rate": 2.927653867380258e-06, + "loss": 2.2331, + "step": 21840 + }, + { + "epoch": 2.3097504230118444, + "grad_norm": 7.96875, + "learning_rate": 2.9274851344118494e-06, + "loss": 2.4192, + "step": 21841 + }, + { + "epoch": 2.3098561759729273, + "grad_norm": 8.5625, + "learning_rate": 2.9273163994374275e-06, + "loss": 2.561, + "step": 21842 + }, + { + "epoch": 2.30996192893401, + "grad_norm": 8.1875, + "learning_rate": 2.9271476624577842e-06, + "loss": 2.577, + "step": 21843 + }, + { + "epoch": 2.310067681895093, + "grad_norm": 8.5625, + "learning_rate": 2.926978923473713e-06, + "loss": 2.6179, + "step": 21844 + }, + { + "epoch": 2.3101734348561758, + "grad_norm": 8.4375, + "learning_rate": 2.926810182486003e-06, + "loss": 2.5441, + "step": 21845 + }, + { + "epoch": 2.310279187817259, + "grad_norm": 8.8125, + "learning_rate": 2.926641439495448e-06, + "loss": 2.7359, + "step": 21846 + }, + { + "epoch": 2.310384940778342, + "grad_norm": 8.6875, + "learning_rate": 2.9264726945028383e-06, + "loss": 2.5699, + "step": 21847 + }, + { + "epoch": 2.3104906937394247, + "grad_norm": 8.25, + "learning_rate": 2.9263039475089676e-06, + "loss": 2.6556, + "step": 21848 + }, + { + "epoch": 2.3105964467005076, + "grad_norm": 9.0, + "learning_rate": 2.9261351985146265e-06, + "loss": 2.4717, + "step": 21849 + }, + { + "epoch": 2.3107021996615904, + "grad_norm": 8.375, + "learning_rate": 2.925966447520607e-06, + "loss": 2.4917, + "step": 21850 + }, + { + "epoch": 2.3108079526226732, + "grad_norm": 8.4375, + "learning_rate": 2.925797694527702e-06, + "loss": 2.6415, + "step": 21851 + }, + { + "epoch": 2.3109137055837565, + "grad_norm": 8.4375, + "learning_rate": 2.9256289395367017e-06, + "loss": 2.4616, + "step": 21852 + }, + { + "epoch": 2.3110194585448394, + "grad_norm": 8.3125, + "learning_rate": 2.925460182548399e-06, + "loss": 2.4663, + "step": 21853 + }, + { + "epoch": 2.311125211505922, + "grad_norm": 8.25, + "learning_rate": 2.9252914235635866e-06, + "loss": 2.3561, + "step": 21854 + }, + { + "epoch": 2.311230964467005, + "grad_norm": 8.625, + "learning_rate": 2.9251226625830543e-06, + "loss": 2.5, + "step": 21855 + }, + { + "epoch": 2.311336717428088, + "grad_norm": 8.75, + "learning_rate": 2.9249538996075955e-06, + "loss": 2.5871, + "step": 21856 + }, + { + "epoch": 2.3114424703891707, + "grad_norm": 8.0625, + "learning_rate": 2.9247851346380017e-06, + "loss": 2.3681, + "step": 21857 + }, + { + "epoch": 2.311548223350254, + "grad_norm": 9.1875, + "learning_rate": 2.9246163676750655e-06, + "loss": 2.7097, + "step": 21858 + }, + { + "epoch": 2.311653976311337, + "grad_norm": 8.25, + "learning_rate": 2.924447598719578e-06, + "loss": 2.554, + "step": 21859 + }, + { + "epoch": 2.3117597292724197, + "grad_norm": 8.375, + "learning_rate": 2.9242788277723315e-06, + "loss": 2.4533, + "step": 21860 + }, + { + "epoch": 2.3118654822335025, + "grad_norm": 9.1875, + "learning_rate": 2.924110054834119e-06, + "loss": 2.7677, + "step": 21861 + }, + { + "epoch": 2.3119712351945854, + "grad_norm": 8.8125, + "learning_rate": 2.9239412799057298e-06, + "loss": 2.7352, + "step": 21862 + }, + { + "epoch": 2.312076988155668, + "grad_norm": 8.6875, + "learning_rate": 2.9237725029879587e-06, + "loss": 2.6712, + "step": 21863 + }, + { + "epoch": 2.312182741116751, + "grad_norm": 9.0, + "learning_rate": 2.9236037240815966e-06, + "loss": 2.3465, + "step": 21864 + }, + { + "epoch": 2.3122884940778343, + "grad_norm": 8.6875, + "learning_rate": 2.923434943187435e-06, + "loss": 2.5086, + "step": 21865 + }, + { + "epoch": 2.312394247038917, + "grad_norm": 8.4375, + "learning_rate": 2.9232661603062656e-06, + "loss": 2.7613, + "step": 21866 + }, + { + "epoch": 2.3125, + "grad_norm": 8.3125, + "learning_rate": 2.9230973754388824e-06, + "loss": 2.3873, + "step": 21867 + }, + { + "epoch": 2.312605752961083, + "grad_norm": 8.1875, + "learning_rate": 2.922928588586076e-06, + "loss": 2.0432, + "step": 21868 + }, + { + "epoch": 2.3127115059221657, + "grad_norm": 8.5625, + "learning_rate": 2.922759799748639e-06, + "loss": 2.609, + "step": 21869 + }, + { + "epoch": 2.312817258883249, + "grad_norm": 8.3125, + "learning_rate": 2.9225910089273628e-06, + "loss": 2.6309, + "step": 21870 + }, + { + "epoch": 2.312923011844332, + "grad_norm": 8.3125, + "learning_rate": 2.92242221612304e-06, + "loss": 2.4965, + "step": 21871 + }, + { + "epoch": 2.3130287648054146, + "grad_norm": 8.8125, + "learning_rate": 2.922253421336462e-06, + "loss": 2.5706, + "step": 21872 + }, + { + "epoch": 2.3131345177664975, + "grad_norm": 8.4375, + "learning_rate": 2.9220846245684214e-06, + "loss": 2.641, + "step": 21873 + }, + { + "epoch": 2.3132402707275803, + "grad_norm": 8.5, + "learning_rate": 2.921915825819711e-06, + "loss": 2.4637, + "step": 21874 + }, + { + "epoch": 2.313346023688663, + "grad_norm": 8.625, + "learning_rate": 2.921747025091122e-06, + "loss": 2.5087, + "step": 21875 + }, + { + "epoch": 2.313451776649746, + "grad_norm": 8.4375, + "learning_rate": 2.921578222383446e-06, + "loss": 2.3262, + "step": 21876 + }, + { + "epoch": 2.3135575296108293, + "grad_norm": 8.5, + "learning_rate": 2.921409417697477e-06, + "loss": 2.5851, + "step": 21877 + }, + { + "epoch": 2.313663282571912, + "grad_norm": 8.5625, + "learning_rate": 2.9212406110340047e-06, + "loss": 2.6964, + "step": 21878 + }, + { + "epoch": 2.313769035532995, + "grad_norm": 8.9375, + "learning_rate": 2.921071802393823e-06, + "loss": 2.6555, + "step": 21879 + }, + { + "epoch": 2.313874788494078, + "grad_norm": 8.0, + "learning_rate": 2.9209029917777243e-06, + "loss": 2.7516, + "step": 21880 + }, + { + "epoch": 2.3139805414551606, + "grad_norm": 8.25, + "learning_rate": 2.920734179186499e-06, + "loss": 2.3873, + "step": 21881 + }, + { + "epoch": 2.3140862944162435, + "grad_norm": 8.5625, + "learning_rate": 2.92056536462094e-06, + "loss": 2.74, + "step": 21882 + }, + { + "epoch": 2.3141920473773268, + "grad_norm": 8.3125, + "learning_rate": 2.920396548081841e-06, + "loss": 2.6157, + "step": 21883 + }, + { + "epoch": 2.3142978003384096, + "grad_norm": 9.125, + "learning_rate": 2.9202277295699923e-06, + "loss": 2.4164, + "step": 21884 + }, + { + "epoch": 2.3144035532994924, + "grad_norm": 8.75, + "learning_rate": 2.9200589090861866e-06, + "loss": 2.6775, + "step": 21885 + }, + { + "epoch": 2.3145093062605753, + "grad_norm": 8.1875, + "learning_rate": 2.919890086631216e-06, + "loss": 2.5688, + "step": 21886 + }, + { + "epoch": 2.314615059221658, + "grad_norm": 8.0625, + "learning_rate": 2.9197212622058735e-06, + "loss": 2.5439, + "step": 21887 + }, + { + "epoch": 2.314720812182741, + "grad_norm": 8.5, + "learning_rate": 2.9195524358109505e-06, + "loss": 2.7352, + "step": 21888 + }, + { + "epoch": 2.3148265651438242, + "grad_norm": 8.625, + "learning_rate": 2.9193836074472405e-06, + "loss": 2.6748, + "step": 21889 + }, + { + "epoch": 2.314932318104907, + "grad_norm": 8.25, + "learning_rate": 2.9192147771155332e-06, + "loss": 2.4708, + "step": 21890 + }, + { + "epoch": 2.31503807106599, + "grad_norm": 8.125, + "learning_rate": 2.9190459448166225e-06, + "loss": 2.5086, + "step": 21891 + }, + { + "epoch": 2.3151438240270727, + "grad_norm": 8.4375, + "learning_rate": 2.9188771105513014e-06, + "loss": 2.3405, + "step": 21892 + }, + { + "epoch": 2.3152495769881556, + "grad_norm": 8.5625, + "learning_rate": 2.9187082743203617e-06, + "loss": 2.5471, + "step": 21893 + }, + { + "epoch": 2.3153553299492384, + "grad_norm": 8.75, + "learning_rate": 2.918539436124594e-06, + "loss": 2.5028, + "step": 21894 + }, + { + "epoch": 2.3154610829103213, + "grad_norm": 8.5625, + "learning_rate": 2.918370595964793e-06, + "loss": 2.5457, + "step": 21895 + }, + { + "epoch": 2.3155668358714045, + "grad_norm": 8.5, + "learning_rate": 2.9182017538417493e-06, + "loss": 2.4732, + "step": 21896 + }, + { + "epoch": 2.3156725888324874, + "grad_norm": 8.0, + "learning_rate": 2.918032909756256e-06, + "loss": 2.6124, + "step": 21897 + }, + { + "epoch": 2.3157783417935702, + "grad_norm": 9.1875, + "learning_rate": 2.9178640637091053e-06, + "loss": 2.7031, + "step": 21898 + }, + { + "epoch": 2.315884094754653, + "grad_norm": 8.8125, + "learning_rate": 2.9176952157010897e-06, + "loss": 2.6458, + "step": 21899 + }, + { + "epoch": 2.315989847715736, + "grad_norm": 8.125, + "learning_rate": 2.9175263657330006e-06, + "loss": 2.5384, + "step": 21900 + }, + { + "epoch": 2.316095600676819, + "grad_norm": 8.3125, + "learning_rate": 2.917357513805631e-06, + "loss": 2.6549, + "step": 21901 + }, + { + "epoch": 2.316201353637902, + "grad_norm": 8.375, + "learning_rate": 2.9171886599197743e-06, + "loss": 2.7199, + "step": 21902 + }, + { + "epoch": 2.316307106598985, + "grad_norm": 8.6875, + "learning_rate": 2.917019804076221e-06, + "loss": 2.6276, + "step": 21903 + }, + { + "epoch": 2.3164128595600677, + "grad_norm": 9.125, + "learning_rate": 2.9168509462757644e-06, + "loss": 2.5818, + "step": 21904 + }, + { + "epoch": 2.3165186125211505, + "grad_norm": 8.125, + "learning_rate": 2.916682086519197e-06, + "loss": 2.2668, + "step": 21905 + }, + { + "epoch": 2.3166243654822334, + "grad_norm": 8.5625, + "learning_rate": 2.91651322480731e-06, + "loss": 2.6033, + "step": 21906 + }, + { + "epoch": 2.316730118443316, + "grad_norm": 8.375, + "learning_rate": 2.9163443611408982e-06, + "loss": 2.4883, + "step": 21907 + }, + { + "epoch": 2.3168358714043995, + "grad_norm": 8.3125, + "learning_rate": 2.9161754955207528e-06, + "loss": 2.656, + "step": 21908 + }, + { + "epoch": 2.3169416243654823, + "grad_norm": 8.375, + "learning_rate": 2.9160066279476644e-06, + "loss": 2.3668, + "step": 21909 + }, + { + "epoch": 2.317047377326565, + "grad_norm": 8.25, + "learning_rate": 2.915837758422428e-06, + "loss": 2.4546, + "step": 21910 + }, + { + "epoch": 2.317153130287648, + "grad_norm": 8.6875, + "learning_rate": 2.915668886945835e-06, + "loss": 2.5833, + "step": 21911 + }, + { + "epoch": 2.317258883248731, + "grad_norm": 8.375, + "learning_rate": 2.9155000135186783e-06, + "loss": 2.4315, + "step": 21912 + }, + { + "epoch": 2.3173646362098137, + "grad_norm": 8.25, + "learning_rate": 2.915331138141749e-06, + "loss": 2.6, + "step": 21913 + }, + { + "epoch": 2.3174703891708965, + "grad_norm": 8.75, + "learning_rate": 2.9151622608158413e-06, + "loss": 2.5932, + "step": 21914 + }, + { + "epoch": 2.31757614213198, + "grad_norm": 8.125, + "learning_rate": 2.914993381541747e-06, + "loss": 2.5308, + "step": 21915 + }, + { + "epoch": 2.3176818950930627, + "grad_norm": 8.375, + "learning_rate": 2.914824500320258e-06, + "loss": 2.4899, + "step": 21916 + }, + { + "epoch": 2.3177876480541455, + "grad_norm": 7.9375, + "learning_rate": 2.914655617152168e-06, + "loss": 2.406, + "step": 21917 + }, + { + "epoch": 2.3178934010152283, + "grad_norm": 8.4375, + "learning_rate": 2.9144867320382676e-06, + "loss": 2.6183, + "step": 21918 + }, + { + "epoch": 2.317999153976311, + "grad_norm": 8.625, + "learning_rate": 2.9143178449793515e-06, + "loss": 2.3587, + "step": 21919 + }, + { + "epoch": 2.3181049069373945, + "grad_norm": 9.25, + "learning_rate": 2.9141489559762106e-06, + "loss": 2.5105, + "step": 21920 + }, + { + "epoch": 2.3182106598984773, + "grad_norm": 8.3125, + "learning_rate": 2.9139800650296385e-06, + "loss": 2.5664, + "step": 21921 + }, + { + "epoch": 2.31831641285956, + "grad_norm": 8.375, + "learning_rate": 2.9138111721404264e-06, + "loss": 2.4953, + "step": 21922 + }, + { + "epoch": 2.318422165820643, + "grad_norm": 8.1875, + "learning_rate": 2.9136422773093685e-06, + "loss": 2.2982, + "step": 21923 + }, + { + "epoch": 2.318527918781726, + "grad_norm": 8.0625, + "learning_rate": 2.9134733805372566e-06, + "loss": 2.2394, + "step": 21924 + }, + { + "epoch": 2.3186336717428087, + "grad_norm": 8.1875, + "learning_rate": 2.913304481824883e-06, + "loss": 2.5768, + "step": 21925 + }, + { + "epoch": 2.3187394247038915, + "grad_norm": 8.125, + "learning_rate": 2.91313558117304e-06, + "loss": 2.6729, + "step": 21926 + }, + { + "epoch": 2.3188451776649748, + "grad_norm": 8.8125, + "learning_rate": 2.912966678582522e-06, + "loss": 2.7651, + "step": 21927 + }, + { + "epoch": 2.3189509306260576, + "grad_norm": 8.6875, + "learning_rate": 2.912797774054119e-06, + "loss": 2.5112, + "step": 21928 + }, + { + "epoch": 2.3190566835871405, + "grad_norm": 8.75, + "learning_rate": 2.9126288675886257e-06, + "loss": 2.5033, + "step": 21929 + }, + { + "epoch": 2.3191624365482233, + "grad_norm": 8.875, + "learning_rate": 2.9124599591868336e-06, + "loss": 2.4827, + "step": 21930 + }, + { + "epoch": 2.319268189509306, + "grad_norm": 8.5625, + "learning_rate": 2.912291048849536e-06, + "loss": 2.5798, + "step": 21931 + }, + { + "epoch": 2.3193739424703894, + "grad_norm": 9.25, + "learning_rate": 2.9121221365775243e-06, + "loss": 2.3595, + "step": 21932 + }, + { + "epoch": 2.3194796954314723, + "grad_norm": 8.25, + "learning_rate": 2.9119532223715923e-06, + "loss": 2.5772, + "step": 21933 + }, + { + "epoch": 2.319585448392555, + "grad_norm": 8.6875, + "learning_rate": 2.911784306232533e-06, + "loss": 2.4789, + "step": 21934 + }, + { + "epoch": 2.319691201353638, + "grad_norm": 8.0, + "learning_rate": 2.911615388161137e-06, + "loss": 2.3223, + "step": 21935 + }, + { + "epoch": 2.3197969543147208, + "grad_norm": 8.5625, + "learning_rate": 2.911446468158199e-06, + "loss": 2.5713, + "step": 21936 + }, + { + "epoch": 2.3199027072758036, + "grad_norm": 8.1875, + "learning_rate": 2.9112775462245112e-06, + "loss": 2.5102, + "step": 21937 + }, + { + "epoch": 2.3200084602368864, + "grad_norm": 8.1875, + "learning_rate": 2.9111086223608666e-06, + "loss": 2.4886, + "step": 21938 + }, + { + "epoch": 2.3201142131979697, + "grad_norm": 8.375, + "learning_rate": 2.910939696568056e-06, + "loss": 2.5438, + "step": 21939 + }, + { + "epoch": 2.3202199661590526, + "grad_norm": 8.375, + "learning_rate": 2.9107707688468745e-06, + "loss": 2.6355, + "step": 21940 + }, + { + "epoch": 2.3203257191201354, + "grad_norm": 8.375, + "learning_rate": 2.9106018391981127e-06, + "loss": 2.6346, + "step": 21941 + }, + { + "epoch": 2.3204314720812182, + "grad_norm": 8.5, + "learning_rate": 2.9104329076225656e-06, + "loss": 2.6337, + "step": 21942 + }, + { + "epoch": 2.320537225042301, + "grad_norm": 9.5, + "learning_rate": 2.9102639741210248e-06, + "loss": 2.7157, + "step": 21943 + }, + { + "epoch": 2.320642978003384, + "grad_norm": 8.75, + "learning_rate": 2.9100950386942823e-06, + "loss": 2.6014, + "step": 21944 + }, + { + "epoch": 2.3207487309644668, + "grad_norm": 9.5, + "learning_rate": 2.9099261013431314e-06, + "loss": 2.548, + "step": 21945 + }, + { + "epoch": 2.32085448392555, + "grad_norm": 8.0625, + "learning_rate": 2.909757162068365e-06, + "loss": 2.541, + "step": 21946 + }, + { + "epoch": 2.320960236886633, + "grad_norm": 8.5625, + "learning_rate": 2.9095882208707767e-06, + "loss": 2.6331, + "step": 21947 + }, + { + "epoch": 2.3210659898477157, + "grad_norm": 8.1875, + "learning_rate": 2.9094192777511573e-06, + "loss": 2.5384, + "step": 21948 + }, + { + "epoch": 2.3211717428087986, + "grad_norm": 8.8125, + "learning_rate": 2.909250332710301e-06, + "loss": 2.7856, + "step": 21949 + }, + { + "epoch": 2.3212774957698814, + "grad_norm": 8.875, + "learning_rate": 2.9090813857490007e-06, + "loss": 2.4866, + "step": 21950 + }, + { + "epoch": 2.3213832487309647, + "grad_norm": 8.5, + "learning_rate": 2.9089124368680483e-06, + "loss": 2.4382, + "step": 21951 + }, + { + "epoch": 2.3214890016920475, + "grad_norm": 8.9375, + "learning_rate": 2.908743486068237e-06, + "loss": 2.3719, + "step": 21952 + }, + { + "epoch": 2.3215947546531304, + "grad_norm": 7.78125, + "learning_rate": 2.90857453335036e-06, + "loss": 2.3445, + "step": 21953 + }, + { + "epoch": 2.321700507614213, + "grad_norm": 8.25, + "learning_rate": 2.9084055787152097e-06, + "loss": 2.5472, + "step": 21954 + }, + { + "epoch": 2.321806260575296, + "grad_norm": 8.3125, + "learning_rate": 2.908236622163579e-06, + "loss": 2.3817, + "step": 21955 + }, + { + "epoch": 2.321912013536379, + "grad_norm": 8.5625, + "learning_rate": 2.9080676636962614e-06, + "loss": 2.4982, + "step": 21956 + }, + { + "epoch": 2.3220177664974617, + "grad_norm": 8.5, + "learning_rate": 2.9078987033140487e-06, + "loss": 2.6148, + "step": 21957 + }, + { + "epoch": 2.322123519458545, + "grad_norm": 8.25, + "learning_rate": 2.9077297410177334e-06, + "loss": 2.6739, + "step": 21958 + }, + { + "epoch": 2.322229272419628, + "grad_norm": 8.5, + "learning_rate": 2.9075607768081104e-06, + "loss": 2.4449, + "step": 21959 + }, + { + "epoch": 2.3223350253807107, + "grad_norm": 8.1875, + "learning_rate": 2.9073918106859705e-06, + "loss": 2.3842, + "step": 21960 + }, + { + "epoch": 2.3224407783417935, + "grad_norm": 8.4375, + "learning_rate": 2.907222842652108e-06, + "loss": 2.5889, + "step": 21961 + }, + { + "epoch": 2.3225465313028764, + "grad_norm": 7.96875, + "learning_rate": 2.907053872707315e-06, + "loss": 2.5505, + "step": 21962 + }, + { + "epoch": 2.322652284263959, + "grad_norm": 8.3125, + "learning_rate": 2.906884900852385e-06, + "loss": 2.6947, + "step": 21963 + }, + { + "epoch": 2.3227580372250425, + "grad_norm": 8.1875, + "learning_rate": 2.90671592708811e-06, + "loss": 2.4323, + "step": 21964 + }, + { + "epoch": 2.3228637901861253, + "grad_norm": 8.1875, + "learning_rate": 2.906546951415284e-06, + "loss": 2.3779, + "step": 21965 + }, + { + "epoch": 2.322969543147208, + "grad_norm": 7.84375, + "learning_rate": 2.906377973834699e-06, + "loss": 2.4124, + "step": 21966 + }, + { + "epoch": 2.323075296108291, + "grad_norm": 8.25, + "learning_rate": 2.906208994347149e-06, + "loss": 2.4461, + "step": 21967 + }, + { + "epoch": 2.323181049069374, + "grad_norm": 9.0625, + "learning_rate": 2.906040012953426e-06, + "loss": 2.4429, + "step": 21968 + }, + { + "epoch": 2.3232868020304567, + "grad_norm": 8.5, + "learning_rate": 2.9058710296543235e-06, + "loss": 2.6502, + "step": 21969 + }, + { + "epoch": 2.32339255499154, + "grad_norm": 9.25, + "learning_rate": 2.9057020444506336e-06, + "loss": 2.6425, + "step": 21970 + }, + { + "epoch": 2.323498307952623, + "grad_norm": 8.4375, + "learning_rate": 2.9055330573431495e-06, + "loss": 2.5693, + "step": 21971 + }, + { + "epoch": 2.3236040609137056, + "grad_norm": 8.5, + "learning_rate": 2.905364068332666e-06, + "loss": 2.7965, + "step": 21972 + }, + { + "epoch": 2.3237098138747885, + "grad_norm": 8.0, + "learning_rate": 2.9051950774199746e-06, + "loss": 2.5091, + "step": 21973 + }, + { + "epoch": 2.3238155668358713, + "grad_norm": 8.75, + "learning_rate": 2.9050260846058677e-06, + "loss": 2.671, + "step": 21974 + }, + { + "epoch": 2.323921319796954, + "grad_norm": 7.625, + "learning_rate": 2.9048570898911392e-06, + "loss": 2.3812, + "step": 21975 + }, + { + "epoch": 2.324027072758037, + "grad_norm": 8.0625, + "learning_rate": 2.904688093276582e-06, + "loss": 2.4542, + "step": 21976 + }, + { + "epoch": 2.3241328257191203, + "grad_norm": 8.25, + "learning_rate": 2.9045190947629894e-06, + "loss": 2.4448, + "step": 21977 + }, + { + "epoch": 2.324238578680203, + "grad_norm": 8.5, + "learning_rate": 2.9043500943511537e-06, + "loss": 2.38, + "step": 21978 + }, + { + "epoch": 2.324344331641286, + "grad_norm": 8.3125, + "learning_rate": 2.9041810920418685e-06, + "loss": 2.5336, + "step": 21979 + }, + { + "epoch": 2.324450084602369, + "grad_norm": 8.4375, + "learning_rate": 2.904012087835926e-06, + "loss": 2.5488, + "step": 21980 + }, + { + "epoch": 2.3245558375634516, + "grad_norm": 8.1875, + "learning_rate": 2.9038430817341213e-06, + "loss": 2.4431, + "step": 21981 + }, + { + "epoch": 2.324661590524535, + "grad_norm": 8.6875, + "learning_rate": 2.903674073737246e-06, + "loss": 2.7192, + "step": 21982 + }, + { + "epoch": 2.3247673434856178, + "grad_norm": 8.1875, + "learning_rate": 2.903505063846093e-06, + "loss": 2.4685, + "step": 21983 + }, + { + "epoch": 2.3248730964467006, + "grad_norm": 9.3125, + "learning_rate": 2.9033360520614558e-06, + "loss": 2.5846, + "step": 21984 + }, + { + "epoch": 2.3249788494077834, + "grad_norm": 8.0, + "learning_rate": 2.9031670383841274e-06, + "loss": 2.3508, + "step": 21985 + }, + { + "epoch": 2.3250846023688663, + "grad_norm": 8.5, + "learning_rate": 2.9029980228149006e-06, + "loss": 2.656, + "step": 21986 + }, + { + "epoch": 2.325190355329949, + "grad_norm": 8.5, + "learning_rate": 2.9028290053545697e-06, + "loss": 2.27, + "step": 21987 + }, + { + "epoch": 2.325296108291032, + "grad_norm": 8.8125, + "learning_rate": 2.902659986003926e-06, + "loss": 2.6469, + "step": 21988 + }, + { + "epoch": 2.3254018612521152, + "grad_norm": 8.5, + "learning_rate": 2.902490964763765e-06, + "loss": 2.6779, + "step": 21989 + }, + { + "epoch": 2.325507614213198, + "grad_norm": 8.375, + "learning_rate": 2.9023219416348775e-06, + "loss": 2.5716, + "step": 21990 + }, + { + "epoch": 2.325613367174281, + "grad_norm": 9.4375, + "learning_rate": 2.9021529166180586e-06, + "loss": 2.5058, + "step": 21991 + }, + { + "epoch": 2.3257191201353637, + "grad_norm": 8.125, + "learning_rate": 2.9019838897141003e-06, + "loss": 2.3649, + "step": 21992 + }, + { + "epoch": 2.3258248730964466, + "grad_norm": 8.125, + "learning_rate": 2.901814860923795e-06, + "loss": 2.3146, + "step": 21993 + }, + { + "epoch": 2.3259306260575294, + "grad_norm": 8.5, + "learning_rate": 2.901645830247938e-06, + "loss": 2.4028, + "step": 21994 + }, + { + "epoch": 2.3260363790186127, + "grad_norm": 8.625, + "learning_rate": 2.9014767976873207e-06, + "loss": 2.5469, + "step": 21995 + }, + { + "epoch": 2.3261421319796955, + "grad_norm": 8.5, + "learning_rate": 2.901307763242737e-06, + "loss": 2.6938, + "step": 21996 + }, + { + "epoch": 2.3262478849407784, + "grad_norm": 8.125, + "learning_rate": 2.9011387269149804e-06, + "loss": 2.5994, + "step": 21997 + }, + { + "epoch": 2.3263536379018612, + "grad_norm": 7.71875, + "learning_rate": 2.900969688704844e-06, + "loss": 2.4534, + "step": 21998 + }, + { + "epoch": 2.326459390862944, + "grad_norm": 8.4375, + "learning_rate": 2.90080064861312e-06, + "loss": 2.5682, + "step": 21999 + }, + { + "epoch": 2.326565143824027, + "grad_norm": 8.5625, + "learning_rate": 2.9006316066406037e-06, + "loss": 2.3914, + "step": 22000 + }, + { + "epoch": 2.32667089678511, + "grad_norm": 8.75, + "learning_rate": 2.9004625627880866e-06, + "loss": 2.6873, + "step": 22001 + }, + { + "epoch": 2.326776649746193, + "grad_norm": 8.875, + "learning_rate": 2.9002935170563617e-06, + "loss": 2.7099, + "step": 22002 + }, + { + "epoch": 2.326882402707276, + "grad_norm": 9.0625, + "learning_rate": 2.9001244694462243e-06, + "loss": 2.7716, + "step": 22003 + }, + { + "epoch": 2.3269881556683587, + "grad_norm": 8.375, + "learning_rate": 2.899955419958466e-06, + "loss": 2.5253, + "step": 22004 + }, + { + "epoch": 2.3270939086294415, + "grad_norm": 7.84375, + "learning_rate": 2.89978636859388e-06, + "loss": 2.4457, + "step": 22005 + }, + { + "epoch": 2.3271996615905244, + "grad_norm": 8.3125, + "learning_rate": 2.89961731535326e-06, + "loss": 2.6039, + "step": 22006 + }, + { + "epoch": 2.327305414551607, + "grad_norm": 8.4375, + "learning_rate": 2.8994482602374003e-06, + "loss": 2.556, + "step": 22007 + }, + { + "epoch": 2.3274111675126905, + "grad_norm": 7.875, + "learning_rate": 2.899279203247093e-06, + "loss": 2.4755, + "step": 22008 + }, + { + "epoch": 2.3275169204737733, + "grad_norm": 8.1875, + "learning_rate": 2.899110144383131e-06, + "loss": 2.5329, + "step": 22009 + }, + { + "epoch": 2.327622673434856, + "grad_norm": 8.6875, + "learning_rate": 2.898941083646309e-06, + "loss": 2.5845, + "step": 22010 + }, + { + "epoch": 2.327728426395939, + "grad_norm": 8.9375, + "learning_rate": 2.89877202103742e-06, + "loss": 2.5237, + "step": 22011 + }, + { + "epoch": 2.327834179357022, + "grad_norm": 8.3125, + "learning_rate": 2.898602956557256e-06, + "loss": 2.6806, + "step": 22012 + }, + { + "epoch": 2.327939932318105, + "grad_norm": 8.4375, + "learning_rate": 2.8984338902066116e-06, + "loss": 2.5563, + "step": 22013 + }, + { + "epoch": 2.328045685279188, + "grad_norm": 8.0625, + "learning_rate": 2.8982648219862797e-06, + "loss": 2.4496, + "step": 22014 + }, + { + "epoch": 2.328151438240271, + "grad_norm": 8.6875, + "learning_rate": 2.8980957518970542e-06, + "loss": 2.8433, + "step": 22015 + }, + { + "epoch": 2.3282571912013537, + "grad_norm": 8.0, + "learning_rate": 2.897926679939728e-06, + "loss": 2.3307, + "step": 22016 + }, + { + "epoch": 2.3283629441624365, + "grad_norm": 8.0625, + "learning_rate": 2.8977576061150954e-06, + "loss": 2.6029, + "step": 22017 + }, + { + "epoch": 2.3284686971235193, + "grad_norm": 8.75, + "learning_rate": 2.897588530423948e-06, + "loss": 2.8191, + "step": 22018 + }, + { + "epoch": 2.328574450084602, + "grad_norm": 8.125, + "learning_rate": 2.897419452867081e-06, + "loss": 2.4852, + "step": 22019 + }, + { + "epoch": 2.3286802030456855, + "grad_norm": 8.875, + "learning_rate": 2.8972503734452874e-06, + "loss": 2.6147, + "step": 22020 + }, + { + "epoch": 2.3287859560067683, + "grad_norm": 8.5625, + "learning_rate": 2.8970812921593596e-06, + "loss": 2.5318, + "step": 22021 + }, + { + "epoch": 2.328891708967851, + "grad_norm": 8.25, + "learning_rate": 2.8969122090100914e-06, + "loss": 2.3413, + "step": 22022 + }, + { + "epoch": 2.328997461928934, + "grad_norm": 8.6875, + "learning_rate": 2.8967431239982774e-06, + "loss": 2.7434, + "step": 22023 + }, + { + "epoch": 2.329103214890017, + "grad_norm": 8.0, + "learning_rate": 2.896574037124709e-06, + "loss": 2.5184, + "step": 22024 + }, + { + "epoch": 2.3292089678510997, + "grad_norm": 8.4375, + "learning_rate": 2.896404948390181e-06, + "loss": 2.4876, + "step": 22025 + }, + { + "epoch": 2.3293147208121825, + "grad_norm": 8.875, + "learning_rate": 2.8962358577954875e-06, + "loss": 2.6537, + "step": 22026 + }, + { + "epoch": 2.3294204737732658, + "grad_norm": 8.25, + "learning_rate": 2.8960667653414213e-06, + "loss": 2.4478, + "step": 22027 + }, + { + "epoch": 2.3295262267343486, + "grad_norm": 7.875, + "learning_rate": 2.895897671028775e-06, + "loss": 2.5821, + "step": 22028 + }, + { + "epoch": 2.3296319796954315, + "grad_norm": 8.4375, + "learning_rate": 2.8957285748583436e-06, + "loss": 2.5683, + "step": 22029 + }, + { + "epoch": 2.3297377326565143, + "grad_norm": 8.75, + "learning_rate": 2.8955594768309192e-06, + "loss": 2.7937, + "step": 22030 + }, + { + "epoch": 2.329843485617597, + "grad_norm": 8.375, + "learning_rate": 2.8953903769472963e-06, + "loss": 2.6389, + "step": 22031 + }, + { + "epoch": 2.3299492385786804, + "grad_norm": 8.9375, + "learning_rate": 2.895221275208268e-06, + "loss": 2.5824, + "step": 22032 + }, + { + "epoch": 2.3300549915397633, + "grad_norm": 8.125, + "learning_rate": 2.8950521716146273e-06, + "loss": 2.4695, + "step": 22033 + }, + { + "epoch": 2.330160744500846, + "grad_norm": 8.5625, + "learning_rate": 2.894883066167169e-06, + "loss": 2.3819, + "step": 22034 + }, + { + "epoch": 2.330266497461929, + "grad_norm": 8.9375, + "learning_rate": 2.8947139588666854e-06, + "loss": 2.6724, + "step": 22035 + }, + { + "epoch": 2.3303722504230118, + "grad_norm": 8.0, + "learning_rate": 2.8945448497139716e-06, + "loss": 2.5347, + "step": 22036 + }, + { + "epoch": 2.3304780033840946, + "grad_norm": 8.25, + "learning_rate": 2.894375738709819e-06, + "loss": 2.5069, + "step": 22037 + }, + { + "epoch": 2.3305837563451774, + "grad_norm": 8.375, + "learning_rate": 2.8942066258550234e-06, + "loss": 2.6197, + "step": 22038 + }, + { + "epoch": 2.3306895093062607, + "grad_norm": 9.3125, + "learning_rate": 2.894037511150377e-06, + "loss": 2.7444, + "step": 22039 + }, + { + "epoch": 2.3307952622673436, + "grad_norm": 8.625, + "learning_rate": 2.893868394596674e-06, + "loss": 2.6246, + "step": 22040 + }, + { + "epoch": 2.3309010152284264, + "grad_norm": 8.1875, + "learning_rate": 2.893699276194707e-06, + "loss": 2.5019, + "step": 22041 + }, + { + "epoch": 2.3310067681895092, + "grad_norm": 8.6875, + "learning_rate": 2.8935301559452707e-06, + "loss": 2.6489, + "step": 22042 + }, + { + "epoch": 2.331112521150592, + "grad_norm": 8.625, + "learning_rate": 2.8933610338491587e-06, + "loss": 2.4752, + "step": 22043 + }, + { + "epoch": 2.3312182741116754, + "grad_norm": 8.375, + "learning_rate": 2.8931919099071633e-06, + "loss": 2.4445, + "step": 22044 + }, + { + "epoch": 2.331324027072758, + "grad_norm": 8.1875, + "learning_rate": 2.8930227841200798e-06, + "loss": 2.4568, + "step": 22045 + }, + { + "epoch": 2.331429780033841, + "grad_norm": 8.375, + "learning_rate": 2.892853656488701e-06, + "loss": 2.5901, + "step": 22046 + }, + { + "epoch": 2.331535532994924, + "grad_norm": 8.6875, + "learning_rate": 2.8926845270138203e-06, + "loss": 2.4037, + "step": 22047 + }, + { + "epoch": 2.3316412859560067, + "grad_norm": 8.25, + "learning_rate": 2.892515395696232e-06, + "loss": 2.4722, + "step": 22048 + }, + { + "epoch": 2.3317470389170896, + "grad_norm": 8.75, + "learning_rate": 2.8923462625367298e-06, + "loss": 2.7064, + "step": 22049 + }, + { + "epoch": 2.3318527918781724, + "grad_norm": 8.5625, + "learning_rate": 2.892177127536106e-06, + "loss": 2.6766, + "step": 22050 + }, + { + "epoch": 2.3319585448392557, + "grad_norm": 8.4375, + "learning_rate": 2.8920079906951565e-06, + "loss": 2.7083, + "step": 22051 + }, + { + "epoch": 2.3320642978003385, + "grad_norm": 8.4375, + "learning_rate": 2.8918388520146735e-06, + "loss": 2.2132, + "step": 22052 + }, + { + "epoch": 2.3321700507614214, + "grad_norm": 8.5625, + "learning_rate": 2.8916697114954504e-06, + "loss": 2.5591, + "step": 22053 + }, + { + "epoch": 2.332275803722504, + "grad_norm": 8.0, + "learning_rate": 2.8915005691382824e-06, + "loss": 2.535, + "step": 22054 + }, + { + "epoch": 2.332381556683587, + "grad_norm": 8.0, + "learning_rate": 2.891331424943962e-06, + "loss": 2.306, + "step": 22055 + }, + { + "epoch": 2.33248730964467, + "grad_norm": 8.6875, + "learning_rate": 2.8911622789132833e-06, + "loss": 2.51, + "step": 22056 + }, + { + "epoch": 2.3325930626057527, + "grad_norm": 8.25, + "learning_rate": 2.89099313104704e-06, + "loss": 2.5084, + "step": 22057 + }, + { + "epoch": 2.332698815566836, + "grad_norm": 8.375, + "learning_rate": 2.8908239813460265e-06, + "loss": 2.5756, + "step": 22058 + }, + { + "epoch": 2.332804568527919, + "grad_norm": 8.1875, + "learning_rate": 2.890654829811035e-06, + "loss": 2.3449, + "step": 22059 + }, + { + "epoch": 2.3329103214890017, + "grad_norm": 8.3125, + "learning_rate": 2.8904856764428596e-06, + "loss": 2.6011, + "step": 22060 + }, + { + "epoch": 2.3330160744500845, + "grad_norm": 8.5, + "learning_rate": 2.8903165212422957e-06, + "loss": 2.5095, + "step": 22061 + }, + { + "epoch": 2.3331218274111674, + "grad_norm": 8.375, + "learning_rate": 2.890147364210136e-06, + "loss": 2.4055, + "step": 22062 + }, + { + "epoch": 2.3332275803722506, + "grad_norm": 8.375, + "learning_rate": 2.8899782053471737e-06, + "loss": 2.3341, + "step": 22063 + }, + { + "epoch": 2.3333333333333335, + "grad_norm": 8.125, + "learning_rate": 2.889809044654204e-06, + "loss": 2.6755, + "step": 22064 + }, + { + "epoch": 2.3334390862944163, + "grad_norm": 10.1875, + "learning_rate": 2.8896398821320193e-06, + "loss": 2.5837, + "step": 22065 + }, + { + "epoch": 2.333544839255499, + "grad_norm": 8.25, + "learning_rate": 2.8894707177814137e-06, + "loss": 2.555, + "step": 22066 + }, + { + "epoch": 2.333650592216582, + "grad_norm": 8.5625, + "learning_rate": 2.8893015516031817e-06, + "loss": 2.6367, + "step": 22067 + }, + { + "epoch": 2.333756345177665, + "grad_norm": 10.375, + "learning_rate": 2.8891323835981165e-06, + "loss": 2.7415, + "step": 22068 + }, + { + "epoch": 2.3338620981387477, + "grad_norm": 8.5, + "learning_rate": 2.888963213767012e-06, + "loss": 2.6148, + "step": 22069 + }, + { + "epoch": 2.333967851099831, + "grad_norm": 8.5625, + "learning_rate": 2.888794042110663e-06, + "loss": 2.4133, + "step": 22070 + }, + { + "epoch": 2.334073604060914, + "grad_norm": 8.5625, + "learning_rate": 2.8886248686298623e-06, + "loss": 2.452, + "step": 22071 + }, + { + "epoch": 2.3341793570219966, + "grad_norm": 8.1875, + "learning_rate": 2.888455693325404e-06, + "loss": 2.4477, + "step": 22072 + }, + { + "epoch": 2.3342851099830795, + "grad_norm": 8.5625, + "learning_rate": 2.8882865161980818e-06, + "loss": 2.4016, + "step": 22073 + }, + { + "epoch": 2.3343908629441623, + "grad_norm": 8.25, + "learning_rate": 2.8881173372486898e-06, + "loss": 2.6578, + "step": 22074 + }, + { + "epoch": 2.334496615905245, + "grad_norm": 8.125, + "learning_rate": 2.8879481564780215e-06, + "loss": 2.5301, + "step": 22075 + }, + { + "epoch": 2.3346023688663284, + "grad_norm": 8.8125, + "learning_rate": 2.8877789738868723e-06, + "loss": 2.6199, + "step": 22076 + }, + { + "epoch": 2.3347081218274113, + "grad_norm": 8.9375, + "learning_rate": 2.8876097894760337e-06, + "loss": 2.7675, + "step": 22077 + }, + { + "epoch": 2.334813874788494, + "grad_norm": 8.25, + "learning_rate": 2.8874406032463016e-06, + "loss": 2.3843, + "step": 22078 + }, + { + "epoch": 2.334919627749577, + "grad_norm": 9.4375, + "learning_rate": 2.8872714151984688e-06, + "loss": 2.5789, + "step": 22079 + }, + { + "epoch": 2.33502538071066, + "grad_norm": 8.3125, + "learning_rate": 2.88710222533333e-06, + "loss": 2.4944, + "step": 22080 + }, + { + "epoch": 2.3351311336717426, + "grad_norm": 8.75, + "learning_rate": 2.8869330336516793e-06, + "loss": 2.6575, + "step": 22081 + }, + { + "epoch": 2.335236886632826, + "grad_norm": 8.8125, + "learning_rate": 2.8867638401543092e-06, + "loss": 2.5654, + "step": 22082 + }, + { + "epoch": 2.3353426395939088, + "grad_norm": 9.0, + "learning_rate": 2.886594644842015e-06, + "loss": 2.5576, + "step": 22083 + }, + { + "epoch": 2.3354483925549916, + "grad_norm": 7.90625, + "learning_rate": 2.88642544771559e-06, + "loss": 2.5327, + "step": 22084 + }, + { + "epoch": 2.3355541455160744, + "grad_norm": 9.3125, + "learning_rate": 2.8862562487758283e-06, + "loss": 2.6154, + "step": 22085 + }, + { + "epoch": 2.3356598984771573, + "grad_norm": 8.4375, + "learning_rate": 2.8860870480235245e-06, + "loss": 2.6468, + "step": 22086 + }, + { + "epoch": 2.33576565143824, + "grad_norm": 8.5625, + "learning_rate": 2.8859178454594718e-06, + "loss": 2.2605, + "step": 22087 + }, + { + "epoch": 2.335871404399323, + "grad_norm": 8.125, + "learning_rate": 2.8857486410844643e-06, + "loss": 2.5092, + "step": 22088 + }, + { + "epoch": 2.3359771573604062, + "grad_norm": 8.5, + "learning_rate": 2.8855794348992967e-06, + "loss": 2.6063, + "step": 22089 + }, + { + "epoch": 2.336082910321489, + "grad_norm": 8.3125, + "learning_rate": 2.8854102269047625e-06, + "loss": 2.6027, + "step": 22090 + }, + { + "epoch": 2.336188663282572, + "grad_norm": 8.375, + "learning_rate": 2.885241017101655e-06, + "loss": 2.6459, + "step": 22091 + }, + { + "epoch": 2.3362944162436547, + "grad_norm": 8.9375, + "learning_rate": 2.8850718054907694e-06, + "loss": 2.6346, + "step": 22092 + }, + { + "epoch": 2.3364001692047376, + "grad_norm": 9.1875, + "learning_rate": 2.8849025920728997e-06, + "loss": 2.8081, + "step": 22093 + }, + { + "epoch": 2.336505922165821, + "grad_norm": 8.4375, + "learning_rate": 2.884733376848839e-06, + "loss": 2.6155, + "step": 22094 + }, + { + "epoch": 2.3366116751269037, + "grad_norm": 8.375, + "learning_rate": 2.884564159819381e-06, + "loss": 2.5961, + "step": 22095 + }, + { + "epoch": 2.3367174280879865, + "grad_norm": 8.25, + "learning_rate": 2.8843949409853222e-06, + "loss": 2.5109, + "step": 22096 + }, + { + "epoch": 2.3368231810490694, + "grad_norm": 8.75, + "learning_rate": 2.8842257203474554e-06, + "loss": 2.4876, + "step": 22097 + }, + { + "epoch": 2.3369289340101522, + "grad_norm": 7.75, + "learning_rate": 2.884056497906573e-06, + "loss": 2.4714, + "step": 22098 + }, + { + "epoch": 2.337034686971235, + "grad_norm": 8.3125, + "learning_rate": 2.8838872736634715e-06, + "loss": 2.4364, + "step": 22099 + }, + { + "epoch": 2.337140439932318, + "grad_norm": 8.6875, + "learning_rate": 2.8837180476189443e-06, + "loss": 2.3988, + "step": 22100 + }, + { + "epoch": 2.337246192893401, + "grad_norm": 9.125, + "learning_rate": 2.8835488197737843e-06, + "loss": 2.8496, + "step": 22101 + }, + { + "epoch": 2.337351945854484, + "grad_norm": 8.625, + "learning_rate": 2.8833795901287876e-06, + "loss": 2.6847, + "step": 22102 + }, + { + "epoch": 2.337457698815567, + "grad_norm": 8.1875, + "learning_rate": 2.883210358684746e-06, + "loss": 2.5127, + "step": 22103 + }, + { + "epoch": 2.3375634517766497, + "grad_norm": 8.3125, + "learning_rate": 2.8830411254424555e-06, + "loss": 2.5009, + "step": 22104 + }, + { + "epoch": 2.3376692047377325, + "grad_norm": 8.1875, + "learning_rate": 2.88287189040271e-06, + "loss": 2.5645, + "step": 22105 + }, + { + "epoch": 2.3377749576988154, + "grad_norm": 8.875, + "learning_rate": 2.8827026535663035e-06, + "loss": 2.5729, + "step": 22106 + }, + { + "epoch": 2.3378807106598987, + "grad_norm": 8.5625, + "learning_rate": 2.882533414934029e-06, + "loss": 2.6556, + "step": 22107 + }, + { + "epoch": 2.3379864636209815, + "grad_norm": 8.3125, + "learning_rate": 2.882364174506682e-06, + "loss": 2.5083, + "step": 22108 + }, + { + "epoch": 2.3380922165820643, + "grad_norm": 8.5, + "learning_rate": 2.8821949322850575e-06, + "loss": 2.5153, + "step": 22109 + }, + { + "epoch": 2.338197969543147, + "grad_norm": 8.0625, + "learning_rate": 2.882025688269947e-06, + "loss": 2.4655, + "step": 22110 + }, + { + "epoch": 2.33830372250423, + "grad_norm": 8.625, + "learning_rate": 2.8818564424621463e-06, + "loss": 2.2681, + "step": 22111 + }, + { + "epoch": 2.338409475465313, + "grad_norm": 7.875, + "learning_rate": 2.8816871948624496e-06, + "loss": 2.4442, + "step": 22112 + }, + { + "epoch": 2.338515228426396, + "grad_norm": 8.3125, + "learning_rate": 2.8815179454716517e-06, + "loss": 2.3022, + "step": 22113 + }, + { + "epoch": 2.338620981387479, + "grad_norm": 8.0625, + "learning_rate": 2.8813486942905455e-06, + "loss": 2.4995, + "step": 22114 + }, + { + "epoch": 2.338726734348562, + "grad_norm": 8.5625, + "learning_rate": 2.8811794413199264e-06, + "loss": 2.5934, + "step": 22115 + }, + { + "epoch": 2.3388324873096447, + "grad_norm": 8.1875, + "learning_rate": 2.881010186560588e-06, + "loss": 2.4901, + "step": 22116 + }, + { + "epoch": 2.3389382402707275, + "grad_norm": 8.25, + "learning_rate": 2.880840930013324e-06, + "loss": 2.3977, + "step": 22117 + }, + { + "epoch": 2.3390439932318103, + "grad_norm": 8.9375, + "learning_rate": 2.8806716716789296e-06, + "loss": 2.7481, + "step": 22118 + }, + { + "epoch": 2.339149746192893, + "grad_norm": 8.125, + "learning_rate": 2.880502411558199e-06, + "loss": 2.5217, + "step": 22119 + }, + { + "epoch": 2.3392554991539765, + "grad_norm": 9.0625, + "learning_rate": 2.8803331496519253e-06, + "loss": 2.6478, + "step": 22120 + }, + { + "epoch": 2.3393612521150593, + "grad_norm": 8.125, + "learning_rate": 2.8801638859609044e-06, + "loss": 2.5517, + "step": 22121 + }, + { + "epoch": 2.339467005076142, + "grad_norm": 8.0, + "learning_rate": 2.8799946204859296e-06, + "loss": 2.406, + "step": 22122 + }, + { + "epoch": 2.339572758037225, + "grad_norm": 8.25, + "learning_rate": 2.8798253532277954e-06, + "loss": 2.5682, + "step": 22123 + }, + { + "epoch": 2.339678510998308, + "grad_norm": 8.9375, + "learning_rate": 2.879656084187297e-06, + "loss": 2.1296, + "step": 22124 + }, + { + "epoch": 2.339784263959391, + "grad_norm": 8.4375, + "learning_rate": 2.879486813365227e-06, + "loss": 2.6144, + "step": 22125 + }, + { + "epoch": 2.339890016920474, + "grad_norm": 8.25, + "learning_rate": 2.8793175407623807e-06, + "loss": 2.5983, + "step": 22126 + }, + { + "epoch": 2.3399957698815568, + "grad_norm": 8.0625, + "learning_rate": 2.879148266379553e-06, + "loss": 2.4512, + "step": 22127 + }, + { + "epoch": 2.3401015228426396, + "grad_norm": 8.5625, + "learning_rate": 2.878978990217537e-06, + "loss": 2.5831, + "step": 22128 + }, + { + "epoch": 2.3402072758037225, + "grad_norm": 9.0, + "learning_rate": 2.8788097122771275e-06, + "loss": 2.8587, + "step": 22129 + }, + { + "epoch": 2.3403130287648053, + "grad_norm": 8.25, + "learning_rate": 2.878640432559119e-06, + "loss": 2.5846, + "step": 22130 + }, + { + "epoch": 2.340418781725888, + "grad_norm": 8.4375, + "learning_rate": 2.878471151064306e-06, + "loss": 2.4595, + "step": 22131 + }, + { + "epoch": 2.3405245346869714, + "grad_norm": 8.1875, + "learning_rate": 2.878301867793483e-06, + "loss": 2.4753, + "step": 22132 + }, + { + "epoch": 2.3406302876480543, + "grad_norm": 8.5, + "learning_rate": 2.8781325827474425e-06, + "loss": 2.5728, + "step": 22133 + }, + { + "epoch": 2.340736040609137, + "grad_norm": 8.5625, + "learning_rate": 2.877963295926982e-06, + "loss": 2.6744, + "step": 22134 + }, + { + "epoch": 2.34084179357022, + "grad_norm": 8.3125, + "learning_rate": 2.8777940073328943e-06, + "loss": 2.4837, + "step": 22135 + }, + { + "epoch": 2.3409475465313028, + "grad_norm": 8.0, + "learning_rate": 2.8776247169659734e-06, + "loss": 2.5946, + "step": 22136 + }, + { + "epoch": 2.3410532994923856, + "grad_norm": 9.3125, + "learning_rate": 2.8774554248270142e-06, + "loss": 2.7625, + "step": 22137 + }, + { + "epoch": 2.3411590524534684, + "grad_norm": 8.5, + "learning_rate": 2.877286130916811e-06, + "loss": 2.5942, + "step": 22138 + }, + { + "epoch": 2.3412648054145517, + "grad_norm": 8.625, + "learning_rate": 2.877116835236158e-06, + "loss": 2.7253, + "step": 22139 + }, + { + "epoch": 2.3413705583756346, + "grad_norm": 8.875, + "learning_rate": 2.8769475377858504e-06, + "loss": 2.683, + "step": 22140 + }, + { + "epoch": 2.3414763113367174, + "grad_norm": 8.25, + "learning_rate": 2.876778238566682e-06, + "loss": 2.3783, + "step": 22141 + }, + { + "epoch": 2.3415820642978002, + "grad_norm": 8.375, + "learning_rate": 2.876608937579447e-06, + "loss": 2.5181, + "step": 22142 + }, + { + "epoch": 2.341687817258883, + "grad_norm": 8.6875, + "learning_rate": 2.8764396348249407e-06, + "loss": 2.6919, + "step": 22143 + }, + { + "epoch": 2.3417935702199664, + "grad_norm": 8.25, + "learning_rate": 2.8762703303039574e-06, + "loss": 2.4995, + "step": 22144 + }, + { + "epoch": 2.341899323181049, + "grad_norm": 8.9375, + "learning_rate": 2.876101024017291e-06, + "loss": 2.5006, + "step": 22145 + }, + { + "epoch": 2.342005076142132, + "grad_norm": 8.4375, + "learning_rate": 2.875931715965737e-06, + "loss": 2.6052, + "step": 22146 + }, + { + "epoch": 2.342110829103215, + "grad_norm": 8.5, + "learning_rate": 2.875762406150089e-06, + "loss": 2.7614, + "step": 22147 + }, + { + "epoch": 2.3422165820642977, + "grad_norm": 8.5625, + "learning_rate": 2.875593094571141e-06, + "loss": 2.4628, + "step": 22148 + }, + { + "epoch": 2.3423223350253806, + "grad_norm": 8.1875, + "learning_rate": 2.8754237812296876e-06, + "loss": 2.3938, + "step": 22149 + }, + { + "epoch": 2.3424280879864634, + "grad_norm": 9.0, + "learning_rate": 2.875254466126525e-06, + "loss": 2.5574, + "step": 22150 + }, + { + "epoch": 2.3425338409475467, + "grad_norm": 8.75, + "learning_rate": 2.8750851492624473e-06, + "loss": 2.2732, + "step": 22151 + }, + { + "epoch": 2.3426395939086295, + "grad_norm": 8.4375, + "learning_rate": 2.8749158306382476e-06, + "loss": 2.5865, + "step": 22152 + }, + { + "epoch": 2.3427453468697124, + "grad_norm": 8.3125, + "learning_rate": 2.874746510254721e-06, + "loss": 2.4402, + "step": 22153 + }, + { + "epoch": 2.342851099830795, + "grad_norm": 8.25, + "learning_rate": 2.874577188112663e-06, + "loss": 2.6845, + "step": 22154 + }, + { + "epoch": 2.342956852791878, + "grad_norm": 8.4375, + "learning_rate": 2.8744078642128663e-06, + "loss": 2.6222, + "step": 22155 + }, + { + "epoch": 2.3430626057529613, + "grad_norm": 8.75, + "learning_rate": 2.8742385385561273e-06, + "loss": 2.685, + "step": 22156 + }, + { + "epoch": 2.343168358714044, + "grad_norm": 8.3125, + "learning_rate": 2.87406921114324e-06, + "loss": 2.4799, + "step": 22157 + }, + { + "epoch": 2.343274111675127, + "grad_norm": 8.1875, + "learning_rate": 2.8738998819749985e-06, + "loss": 2.5159, + "step": 22158 + }, + { + "epoch": 2.34337986463621, + "grad_norm": 7.9375, + "learning_rate": 2.8737305510521985e-06, + "loss": 2.4985, + "step": 22159 + }, + { + "epoch": 2.3434856175972927, + "grad_norm": 8.4375, + "learning_rate": 2.873561218375634e-06, + "loss": 2.6956, + "step": 22160 + }, + { + "epoch": 2.3435913705583755, + "grad_norm": 8.0625, + "learning_rate": 2.8733918839460983e-06, + "loss": 2.4584, + "step": 22161 + }, + { + "epoch": 2.3436971235194584, + "grad_norm": 8.6875, + "learning_rate": 2.8732225477643884e-06, + "loss": 2.688, + "step": 22162 + }, + { + "epoch": 2.3438028764805416, + "grad_norm": 8.6875, + "learning_rate": 2.8730532098312973e-06, + "loss": 2.5313, + "step": 22163 + }, + { + "epoch": 2.3439086294416245, + "grad_norm": 8.5625, + "learning_rate": 2.87288387014762e-06, + "loss": 2.5303, + "step": 22164 + }, + { + "epoch": 2.3440143824027073, + "grad_norm": 8.5625, + "learning_rate": 2.8727145287141506e-06, + "loss": 2.4958, + "step": 22165 + }, + { + "epoch": 2.34412013536379, + "grad_norm": 8.1875, + "learning_rate": 2.8725451855316855e-06, + "loss": 2.482, + "step": 22166 + }, + { + "epoch": 2.344225888324873, + "grad_norm": 8.0625, + "learning_rate": 2.8723758406010176e-06, + "loss": 2.5306, + "step": 22167 + }, + { + "epoch": 2.344331641285956, + "grad_norm": 8.375, + "learning_rate": 2.8722064939229414e-06, + "loss": 2.3631, + "step": 22168 + }, + { + "epoch": 2.3444373942470387, + "grad_norm": 8.6875, + "learning_rate": 2.8720371454982537e-06, + "loss": 2.517, + "step": 22169 + }, + { + "epoch": 2.344543147208122, + "grad_norm": 8.375, + "learning_rate": 2.8718677953277473e-06, + "loss": 2.4071, + "step": 22170 + }, + { + "epoch": 2.344648900169205, + "grad_norm": 8.6875, + "learning_rate": 2.8716984434122174e-06, + "loss": 2.7379, + "step": 22171 + }, + { + "epoch": 2.3447546531302876, + "grad_norm": 8.4375, + "learning_rate": 2.871529089752459e-06, + "loss": 2.3718, + "step": 22172 + }, + { + "epoch": 2.3448604060913705, + "grad_norm": 8.375, + "learning_rate": 2.871359734349267e-06, + "loss": 2.3467, + "step": 22173 + }, + { + "epoch": 2.3449661590524533, + "grad_norm": 8.875, + "learning_rate": 2.8711903772034335e-06, + "loss": 2.5456, + "step": 22174 + }, + { + "epoch": 2.3450719120135366, + "grad_norm": 8.5625, + "learning_rate": 2.8710210183157573e-06, + "loss": 2.4712, + "step": 22175 + }, + { + "epoch": 2.3451776649746194, + "grad_norm": 8.875, + "learning_rate": 2.8708516576870303e-06, + "loss": 2.6086, + "step": 22176 + }, + { + "epoch": 2.3452834179357023, + "grad_norm": 9.1875, + "learning_rate": 2.8706822953180487e-06, + "loss": 2.7195, + "step": 22177 + }, + { + "epoch": 2.345389170896785, + "grad_norm": 7.96875, + "learning_rate": 2.870512931209607e-06, + "loss": 2.2765, + "step": 22178 + }, + { + "epoch": 2.345494923857868, + "grad_norm": 8.25, + "learning_rate": 2.870343565362499e-06, + "loss": 2.4438, + "step": 22179 + }, + { + "epoch": 2.345600676818951, + "grad_norm": 8.1875, + "learning_rate": 2.87017419777752e-06, + "loss": 2.4567, + "step": 22180 + }, + { + "epoch": 2.3457064297800336, + "grad_norm": 8.625, + "learning_rate": 2.8700048284554655e-06, + "loss": 2.7245, + "step": 22181 + }, + { + "epoch": 2.345812182741117, + "grad_norm": 8.875, + "learning_rate": 2.86983545739713e-06, + "loss": 2.7254, + "step": 22182 + }, + { + "epoch": 2.3459179357021998, + "grad_norm": 8.5, + "learning_rate": 2.8696660846033064e-06, + "loss": 2.7233, + "step": 22183 + }, + { + "epoch": 2.3460236886632826, + "grad_norm": 8.9375, + "learning_rate": 2.8694967100747917e-06, + "loss": 2.6538, + "step": 22184 + }, + { + "epoch": 2.3461294416243654, + "grad_norm": 8.9375, + "learning_rate": 2.8693273338123806e-06, + "loss": 2.6307, + "step": 22185 + }, + { + "epoch": 2.3462351945854483, + "grad_norm": 8.375, + "learning_rate": 2.8691579558168676e-06, + "loss": 2.4568, + "step": 22186 + }, + { + "epoch": 2.3463409475465316, + "grad_norm": 7.84375, + "learning_rate": 2.868988576089046e-06, + "loss": 2.4363, + "step": 22187 + }, + { + "epoch": 2.3464467005076144, + "grad_norm": 8.75, + "learning_rate": 2.8688191946297134e-06, + "loss": 2.3429, + "step": 22188 + }, + { + "epoch": 2.3465524534686972, + "grad_norm": 8.4375, + "learning_rate": 2.8686498114396625e-06, + "loss": 2.3543, + "step": 22189 + }, + { + "epoch": 2.34665820642978, + "grad_norm": 8.625, + "learning_rate": 2.8684804265196887e-06, + "loss": 2.6701, + "step": 22190 + }, + { + "epoch": 2.346763959390863, + "grad_norm": 9.125, + "learning_rate": 2.8683110398705873e-06, + "loss": 2.9252, + "step": 22191 + }, + { + "epoch": 2.3468697123519457, + "grad_norm": 8.75, + "learning_rate": 2.868141651493152e-06, + "loss": 2.6794, + "step": 22192 + }, + { + "epoch": 2.3469754653130286, + "grad_norm": 8.375, + "learning_rate": 2.8679722613881787e-06, + "loss": 2.4451, + "step": 22193 + }, + { + "epoch": 2.347081218274112, + "grad_norm": 9.0625, + "learning_rate": 2.8678028695564626e-06, + "loss": 2.4537, + "step": 22194 + }, + { + "epoch": 2.3471869712351947, + "grad_norm": 8.5, + "learning_rate": 2.8676334759987982e-06, + "loss": 2.214, + "step": 22195 + }, + { + "epoch": 2.3472927241962775, + "grad_norm": 8.25, + "learning_rate": 2.86746408071598e-06, + "loss": 2.4419, + "step": 22196 + }, + { + "epoch": 2.3473984771573604, + "grad_norm": 8.125, + "learning_rate": 2.8672946837088026e-06, + "loss": 2.432, + "step": 22197 + }, + { + "epoch": 2.3475042301184432, + "grad_norm": 7.75, + "learning_rate": 2.8671252849780624e-06, + "loss": 2.4186, + "step": 22198 + }, + { + "epoch": 2.347609983079526, + "grad_norm": 9.3125, + "learning_rate": 2.8669558845245526e-06, + "loss": 2.3473, + "step": 22199 + }, + { + "epoch": 2.347715736040609, + "grad_norm": 8.75, + "learning_rate": 2.86678648234907e-06, + "loss": 2.5397, + "step": 22200 + }, + { + "epoch": 2.347821489001692, + "grad_norm": 8.25, + "learning_rate": 2.866617078452407e-06, + "loss": 2.5167, + "step": 22201 + }, + { + "epoch": 2.347927241962775, + "grad_norm": 8.125, + "learning_rate": 2.8664476728353607e-06, + "loss": 2.5042, + "step": 22202 + }, + { + "epoch": 2.348032994923858, + "grad_norm": 8.0625, + "learning_rate": 2.866278265498725e-06, + "loss": 2.5102, + "step": 22203 + }, + { + "epoch": 2.3481387478849407, + "grad_norm": 7.90625, + "learning_rate": 2.866108856443296e-06, + "loss": 2.323, + "step": 22204 + }, + { + "epoch": 2.3482445008460235, + "grad_norm": 8.4375, + "learning_rate": 2.8659394456698673e-06, + "loss": 2.6078, + "step": 22205 + }, + { + "epoch": 2.348350253807107, + "grad_norm": 8.875, + "learning_rate": 2.865770033179235e-06, + "loss": 2.4619, + "step": 22206 + }, + { + "epoch": 2.3484560067681897, + "grad_norm": 8.0625, + "learning_rate": 2.865600618972193e-06, + "loss": 2.446, + "step": 22207 + }, + { + "epoch": 2.3485617597292725, + "grad_norm": 8.1875, + "learning_rate": 2.8654312030495374e-06, + "loss": 2.4508, + "step": 22208 + }, + { + "epoch": 2.3486675126903553, + "grad_norm": 8.3125, + "learning_rate": 2.8652617854120617e-06, + "loss": 2.5035, + "step": 22209 + }, + { + "epoch": 2.348773265651438, + "grad_norm": 8.5, + "learning_rate": 2.8650923660605615e-06, + "loss": 2.4024, + "step": 22210 + }, + { + "epoch": 2.348879018612521, + "grad_norm": 8.8125, + "learning_rate": 2.8649229449958334e-06, + "loss": 2.8085, + "step": 22211 + }, + { + "epoch": 2.348984771573604, + "grad_norm": 8.8125, + "learning_rate": 2.8647535222186707e-06, + "loss": 2.5659, + "step": 22212 + }, + { + "epoch": 2.349090524534687, + "grad_norm": 8.375, + "learning_rate": 2.864584097729869e-06, + "loss": 2.5614, + "step": 22213 + }, + { + "epoch": 2.34919627749577, + "grad_norm": 8.1875, + "learning_rate": 2.8644146715302236e-06, + "loss": 2.3287, + "step": 22214 + }, + { + "epoch": 2.349302030456853, + "grad_norm": 8.9375, + "learning_rate": 2.8642452436205282e-06, + "loss": 2.5485, + "step": 22215 + }, + { + "epoch": 2.3494077834179357, + "grad_norm": 8.1875, + "learning_rate": 2.8640758140015797e-06, + "loss": 2.4879, + "step": 22216 + }, + { + "epoch": 2.3495135363790185, + "grad_norm": 9.0625, + "learning_rate": 2.863906382674172e-06, + "loss": 2.6809, + "step": 22217 + }, + { + "epoch": 2.3496192893401013, + "grad_norm": 7.90625, + "learning_rate": 2.8637369496391e-06, + "loss": 2.4606, + "step": 22218 + }, + { + "epoch": 2.3497250423011846, + "grad_norm": 8.5625, + "learning_rate": 2.863567514897159e-06, + "loss": 2.5887, + "step": 22219 + }, + { + "epoch": 2.3498307952622675, + "grad_norm": 8.3125, + "learning_rate": 2.8633980784491456e-06, + "loss": 2.5686, + "step": 22220 + }, + { + "epoch": 2.3499365482233503, + "grad_norm": 8.5625, + "learning_rate": 2.863228640295853e-06, + "loss": 2.6083, + "step": 22221 + }, + { + "epoch": 2.350042301184433, + "grad_norm": 7.84375, + "learning_rate": 2.863059200438077e-06, + "loss": 2.509, + "step": 22222 + }, + { + "epoch": 2.350148054145516, + "grad_norm": 8.375, + "learning_rate": 2.862889758876613e-06, + "loss": 2.5934, + "step": 22223 + }, + { + "epoch": 2.350253807106599, + "grad_norm": 8.3125, + "learning_rate": 2.8627203156122556e-06, + "loss": 2.5013, + "step": 22224 + }, + { + "epoch": 2.350359560067682, + "grad_norm": 8.5625, + "learning_rate": 2.8625508706457993e-06, + "loss": 2.5274, + "step": 22225 + }, + { + "epoch": 2.350465313028765, + "grad_norm": 8.5625, + "learning_rate": 2.8623814239780414e-06, + "loss": 2.4336, + "step": 22226 + }, + { + "epoch": 2.3505710659898478, + "grad_norm": 8.5625, + "learning_rate": 2.862211975609774e-06, + "loss": 2.5506, + "step": 22227 + }, + { + "epoch": 2.3506768189509306, + "grad_norm": 8.25, + "learning_rate": 2.8620425255417947e-06, + "loss": 2.5132, + "step": 22228 + }, + { + "epoch": 2.3507825719120135, + "grad_norm": 8.8125, + "learning_rate": 2.861873073774899e-06, + "loss": 2.582, + "step": 22229 + }, + { + "epoch": 2.3508883248730963, + "grad_norm": 8.6875, + "learning_rate": 2.8617036203098803e-06, + "loss": 2.609, + "step": 22230 + }, + { + "epoch": 2.350994077834179, + "grad_norm": 8.875, + "learning_rate": 2.861534165147534e-06, + "loss": 2.578, + "step": 22231 + }, + { + "epoch": 2.3510998307952624, + "grad_norm": 8.375, + "learning_rate": 2.861364708288656e-06, + "loss": 2.2435, + "step": 22232 + }, + { + "epoch": 2.3512055837563453, + "grad_norm": 8.125, + "learning_rate": 2.861195249734041e-06, + "loss": 2.5902, + "step": 22233 + }, + { + "epoch": 2.351311336717428, + "grad_norm": 8.375, + "learning_rate": 2.8610257894844847e-06, + "loss": 2.5118, + "step": 22234 + }, + { + "epoch": 2.351417089678511, + "grad_norm": 8.1875, + "learning_rate": 2.860856327540782e-06, + "loss": 2.6467, + "step": 22235 + }, + { + "epoch": 2.3515228426395938, + "grad_norm": 9.25, + "learning_rate": 2.8606868639037277e-06, + "loss": 2.6346, + "step": 22236 + }, + { + "epoch": 2.351628595600677, + "grad_norm": 9.125, + "learning_rate": 2.860517398574118e-06, + "loss": 2.7984, + "step": 22237 + }, + { + "epoch": 2.35173434856176, + "grad_norm": 7.875, + "learning_rate": 2.8603479315527473e-06, + "loss": 2.3445, + "step": 22238 + }, + { + "epoch": 2.3518401015228427, + "grad_norm": 8.4375, + "learning_rate": 2.860178462840412e-06, + "loss": 2.6802, + "step": 22239 + }, + { + "epoch": 2.3519458544839256, + "grad_norm": 8.3125, + "learning_rate": 2.8600089924379056e-06, + "loss": 2.5639, + "step": 22240 + }, + { + "epoch": 2.3520516074450084, + "grad_norm": 8.1875, + "learning_rate": 2.859839520346024e-06, + "loss": 2.4948, + "step": 22241 + }, + { + "epoch": 2.3521573604060912, + "grad_norm": 8.375, + "learning_rate": 2.8596700465655638e-06, + "loss": 2.7173, + "step": 22242 + }, + { + "epoch": 2.352263113367174, + "grad_norm": 8.6875, + "learning_rate": 2.8595005710973185e-06, + "loss": 2.7007, + "step": 22243 + }, + { + "epoch": 2.3523688663282574, + "grad_norm": 8.5, + "learning_rate": 2.859331093942083e-06, + "loss": 2.5733, + "step": 22244 + }, + { + "epoch": 2.35247461928934, + "grad_norm": 8.375, + "learning_rate": 2.8591616151006545e-06, + "loss": 2.2968, + "step": 22245 + }, + { + "epoch": 2.352580372250423, + "grad_norm": 8.0625, + "learning_rate": 2.858992134573828e-06, + "loss": 2.6379, + "step": 22246 + }, + { + "epoch": 2.352686125211506, + "grad_norm": 8.75, + "learning_rate": 2.8588226523623976e-06, + "loss": 2.4707, + "step": 22247 + }, + { + "epoch": 2.3527918781725887, + "grad_norm": 8.875, + "learning_rate": 2.858653168467159e-06, + "loss": 2.5476, + "step": 22248 + }, + { + "epoch": 2.3528976311336716, + "grad_norm": 8.3125, + "learning_rate": 2.858483682888909e-06, + "loss": 2.511, + "step": 22249 + }, + { + "epoch": 2.353003384094755, + "grad_norm": 8.5, + "learning_rate": 2.8583141956284406e-06, + "loss": 2.6616, + "step": 22250 + }, + { + "epoch": 2.3531091370558377, + "grad_norm": 8.5, + "learning_rate": 2.8581447066865507e-06, + "loss": 2.602, + "step": 22251 + }, + { + "epoch": 2.3532148900169205, + "grad_norm": 8.0625, + "learning_rate": 2.857975216064034e-06, + "loss": 2.5068, + "step": 22252 + }, + { + "epoch": 2.3533206429780034, + "grad_norm": 8.3125, + "learning_rate": 2.8578057237616856e-06, + "loss": 2.4725, + "step": 22253 + }, + { + "epoch": 2.353426395939086, + "grad_norm": 7.90625, + "learning_rate": 2.8576362297803016e-06, + "loss": 2.4483, + "step": 22254 + }, + { + "epoch": 2.353532148900169, + "grad_norm": 8.1875, + "learning_rate": 2.8574667341206776e-06, + "loss": 2.466, + "step": 22255 + }, + { + "epoch": 2.3536379018612523, + "grad_norm": 8.875, + "learning_rate": 2.8572972367836083e-06, + "loss": 2.361, + "step": 22256 + }, + { + "epoch": 2.353743654822335, + "grad_norm": 9.0, + "learning_rate": 2.857127737769888e-06, + "loss": 2.9267, + "step": 22257 + }, + { + "epoch": 2.353849407783418, + "grad_norm": 8.3125, + "learning_rate": 2.8569582370803143e-06, + "loss": 2.5885, + "step": 22258 + }, + { + "epoch": 2.353955160744501, + "grad_norm": 8.9375, + "learning_rate": 2.8567887347156818e-06, + "loss": 2.5372, + "step": 22259 + }, + { + "epoch": 2.3540609137055837, + "grad_norm": 8.75, + "learning_rate": 2.8566192306767852e-06, + "loss": 2.7281, + "step": 22260 + }, + { + "epoch": 2.3541666666666665, + "grad_norm": 8.125, + "learning_rate": 2.8564497249644205e-06, + "loss": 2.4622, + "step": 22261 + }, + { + "epoch": 2.3542724196277494, + "grad_norm": 8.4375, + "learning_rate": 2.8562802175793835e-06, + "loss": 2.5196, + "step": 22262 + }, + { + "epoch": 2.3543781725888326, + "grad_norm": 8.5, + "learning_rate": 2.8561107085224677e-06, + "loss": 2.5761, + "step": 22263 + }, + { + "epoch": 2.3544839255499155, + "grad_norm": 9.5625, + "learning_rate": 2.855941197794471e-06, + "loss": 2.9505, + "step": 22264 + }, + { + "epoch": 2.3545896785109983, + "grad_norm": 8.0, + "learning_rate": 2.855771685396188e-06, + "loss": 2.325, + "step": 22265 + }, + { + "epoch": 2.354695431472081, + "grad_norm": 8.1875, + "learning_rate": 2.855602171328414e-06, + "loss": 2.5006, + "step": 22266 + }, + { + "epoch": 2.354801184433164, + "grad_norm": 8.1875, + "learning_rate": 2.855432655591944e-06, + "loss": 2.6304, + "step": 22267 + }, + { + "epoch": 2.3549069373942473, + "grad_norm": 8.5, + "learning_rate": 2.8552631381875745e-06, + "loss": 2.5504, + "step": 22268 + }, + { + "epoch": 2.35501269035533, + "grad_norm": 8.1875, + "learning_rate": 2.8550936191161e-06, + "loss": 2.4648, + "step": 22269 + }, + { + "epoch": 2.355118443316413, + "grad_norm": 8.25, + "learning_rate": 2.8549240983783166e-06, + "loss": 2.4326, + "step": 22270 + }, + { + "epoch": 2.355224196277496, + "grad_norm": 9.125, + "learning_rate": 2.854754575975019e-06, + "loss": 2.7009, + "step": 22271 + }, + { + "epoch": 2.3553299492385786, + "grad_norm": 8.375, + "learning_rate": 2.854585051907004e-06, + "loss": 2.3076, + "step": 22272 + }, + { + "epoch": 2.3554357021996615, + "grad_norm": 9.1875, + "learning_rate": 2.8544155261750655e-06, + "loss": 2.5094, + "step": 22273 + }, + { + "epoch": 2.3555414551607443, + "grad_norm": 7.65625, + "learning_rate": 2.854245998780001e-06, + "loss": 2.4707, + "step": 22274 + }, + { + "epoch": 2.3556472081218276, + "grad_norm": 8.5625, + "learning_rate": 2.8540764697226042e-06, + "loss": 2.2279, + "step": 22275 + }, + { + "epoch": 2.3557529610829104, + "grad_norm": 8.25, + "learning_rate": 2.8539069390036712e-06, + "loss": 2.5208, + "step": 22276 + }, + { + "epoch": 2.3558587140439933, + "grad_norm": 8.375, + "learning_rate": 2.853737406623998e-06, + "loss": 2.4866, + "step": 22277 + }, + { + "epoch": 2.355964467005076, + "grad_norm": 8.375, + "learning_rate": 2.8535678725843803e-06, + "loss": 2.5783, + "step": 22278 + }, + { + "epoch": 2.356070219966159, + "grad_norm": 8.1875, + "learning_rate": 2.8533983368856123e-06, + "loss": 2.583, + "step": 22279 + }, + { + "epoch": 2.356175972927242, + "grad_norm": 8.5625, + "learning_rate": 2.85322879952849e-06, + "loss": 2.7342, + "step": 22280 + }, + { + "epoch": 2.3562817258883246, + "grad_norm": 8.3125, + "learning_rate": 2.853059260513811e-06, + "loss": 2.4764, + "step": 22281 + }, + { + "epoch": 2.356387478849408, + "grad_norm": 8.25, + "learning_rate": 2.852889719842368e-06, + "loss": 2.5339, + "step": 22282 + }, + { + "epoch": 2.3564932318104908, + "grad_norm": 8.875, + "learning_rate": 2.852720177514959e-06, + "loss": 2.5133, + "step": 22283 + }, + { + "epoch": 2.3565989847715736, + "grad_norm": 8.5, + "learning_rate": 2.8525506335323775e-06, + "loss": 2.4697, + "step": 22284 + }, + { + "epoch": 2.3567047377326564, + "grad_norm": 8.5, + "learning_rate": 2.85238108789542e-06, + "loss": 2.569, + "step": 22285 + }, + { + "epoch": 2.3568104906937393, + "grad_norm": 8.8125, + "learning_rate": 2.8522115406048827e-06, + "loss": 2.5566, + "step": 22286 + }, + { + "epoch": 2.3569162436548226, + "grad_norm": 8.75, + "learning_rate": 2.852041991661561e-06, + "loss": 2.2193, + "step": 22287 + }, + { + "epoch": 2.3570219966159054, + "grad_norm": 8.6875, + "learning_rate": 2.8518724410662495e-06, + "loss": 2.4294, + "step": 22288 + }, + { + "epoch": 2.3571277495769882, + "grad_norm": 8.375, + "learning_rate": 2.8517028888197446e-06, + "loss": 2.4369, + "step": 22289 + }, + { + "epoch": 2.357233502538071, + "grad_norm": 8.9375, + "learning_rate": 2.8515333349228417e-06, + "loss": 2.499, + "step": 22290 + }, + { + "epoch": 2.357339255499154, + "grad_norm": 8.0625, + "learning_rate": 2.8513637793763375e-06, + "loss": 2.4643, + "step": 22291 + }, + { + "epoch": 2.3574450084602367, + "grad_norm": 9.25, + "learning_rate": 2.851194222181026e-06, + "loss": 2.5174, + "step": 22292 + }, + { + "epoch": 2.3575507614213196, + "grad_norm": 8.0, + "learning_rate": 2.851024663337704e-06, + "loss": 2.5447, + "step": 22293 + }, + { + "epoch": 2.357656514382403, + "grad_norm": 8.5625, + "learning_rate": 2.8508551028471677e-06, + "loss": 2.5862, + "step": 22294 + }, + { + "epoch": 2.3577622673434857, + "grad_norm": 8.0, + "learning_rate": 2.85068554071021e-06, + "loss": 2.4792, + "step": 22295 + }, + { + "epoch": 2.3578680203045685, + "grad_norm": 9.0, + "learning_rate": 2.8505159769276302e-06, + "loss": 2.3764, + "step": 22296 + }, + { + "epoch": 2.3579737732656514, + "grad_norm": 8.6875, + "learning_rate": 2.8503464115002216e-06, + "loss": 2.5501, + "step": 22297 + }, + { + "epoch": 2.3580795262267342, + "grad_norm": 8.3125, + "learning_rate": 2.8501768444287796e-06, + "loss": 2.3498, + "step": 22298 + }, + { + "epoch": 2.3581852791878175, + "grad_norm": 8.75, + "learning_rate": 2.8500072757141018e-06, + "loss": 2.7068, + "step": 22299 + }, + { + "epoch": 2.3582910321489003, + "grad_norm": 8.4375, + "learning_rate": 2.849837705356984e-06, + "loss": 2.6736, + "step": 22300 + }, + { + "epoch": 2.358396785109983, + "grad_norm": 8.1875, + "learning_rate": 2.8496681333582186e-06, + "loss": 2.4797, + "step": 22301 + }, + { + "epoch": 2.358502538071066, + "grad_norm": 8.125, + "learning_rate": 2.8494985597186053e-06, + "loss": 2.2586, + "step": 22302 + }, + { + "epoch": 2.358608291032149, + "grad_norm": 8.375, + "learning_rate": 2.849328984438938e-06, + "loss": 2.4025, + "step": 22303 + }, + { + "epoch": 2.3587140439932317, + "grad_norm": 9.0625, + "learning_rate": 2.8491594075200126e-06, + "loss": 2.7961, + "step": 22304 + }, + { + "epoch": 2.3588197969543145, + "grad_norm": 8.4375, + "learning_rate": 2.848989828962625e-06, + "loss": 2.5316, + "step": 22305 + }, + { + "epoch": 2.358925549915398, + "grad_norm": 8.875, + "learning_rate": 2.848820248767571e-06, + "loss": 2.8336, + "step": 22306 + }, + { + "epoch": 2.3590313028764807, + "grad_norm": 8.125, + "learning_rate": 2.8486506669356456e-06, + "loss": 2.586, + "step": 22307 + }, + { + "epoch": 2.3591370558375635, + "grad_norm": 8.6875, + "learning_rate": 2.848481083467645e-06, + "loss": 2.5474, + "step": 22308 + }, + { + "epoch": 2.3592428087986463, + "grad_norm": 8.75, + "learning_rate": 2.8483114983643664e-06, + "loss": 2.7239, + "step": 22309 + }, + { + "epoch": 2.359348561759729, + "grad_norm": 7.875, + "learning_rate": 2.848141911626604e-06, + "loss": 2.3889, + "step": 22310 + }, + { + "epoch": 2.359454314720812, + "grad_norm": 7.875, + "learning_rate": 2.847972323255153e-06, + "loss": 2.4162, + "step": 22311 + }, + { + "epoch": 2.359560067681895, + "grad_norm": 8.75, + "learning_rate": 2.8478027332508113e-06, + "loss": 2.6677, + "step": 22312 + }, + { + "epoch": 2.359665820642978, + "grad_norm": 8.5625, + "learning_rate": 2.8476331416143732e-06, + "loss": 2.6355, + "step": 22313 + }, + { + "epoch": 2.359771573604061, + "grad_norm": 8.1875, + "learning_rate": 2.8474635483466353e-06, + "loss": 2.6505, + "step": 22314 + }, + { + "epoch": 2.359877326565144, + "grad_norm": 8.75, + "learning_rate": 2.847293953448393e-06, + "loss": 2.6467, + "step": 22315 + }, + { + "epoch": 2.3599830795262267, + "grad_norm": 9.0, + "learning_rate": 2.8471243569204416e-06, + "loss": 2.7143, + "step": 22316 + }, + { + "epoch": 2.3600888324873095, + "grad_norm": 8.625, + "learning_rate": 2.846954758763577e-06, + "loss": 2.6747, + "step": 22317 + }, + { + "epoch": 2.360194585448393, + "grad_norm": 9.125, + "learning_rate": 2.8467851589785972e-06, + "loss": 2.5734, + "step": 22318 + }, + { + "epoch": 2.3603003384094756, + "grad_norm": 8.5625, + "learning_rate": 2.846615557566296e-06, + "loss": 2.6257, + "step": 22319 + }, + { + "epoch": 2.3604060913705585, + "grad_norm": 8.875, + "learning_rate": 2.846445954527469e-06, + "loss": 2.6103, + "step": 22320 + }, + { + "epoch": 2.3605118443316413, + "grad_norm": 8.4375, + "learning_rate": 2.8462763498629142e-06, + "loss": 2.4835, + "step": 22321 + }, + { + "epoch": 2.360617597292724, + "grad_norm": 8.5625, + "learning_rate": 2.8461067435734254e-06, + "loss": 2.3047, + "step": 22322 + }, + { + "epoch": 2.360723350253807, + "grad_norm": 8.4375, + "learning_rate": 2.8459371356597987e-06, + "loss": 2.6467, + "step": 22323 + }, + { + "epoch": 2.36082910321489, + "grad_norm": 8.375, + "learning_rate": 2.8457675261228313e-06, + "loss": 2.634, + "step": 22324 + }, + { + "epoch": 2.360934856175973, + "grad_norm": 8.5, + "learning_rate": 2.845597914963317e-06, + "loss": 2.6021, + "step": 22325 + }, + { + "epoch": 2.361040609137056, + "grad_norm": 18.375, + "learning_rate": 2.845428302182055e-06, + "loss": 2.6532, + "step": 22326 + }, + { + "epoch": 2.3611463620981388, + "grad_norm": 8.25, + "learning_rate": 2.8452586877798377e-06, + "loss": 2.5027, + "step": 22327 + }, + { + "epoch": 2.3612521150592216, + "grad_norm": 8.3125, + "learning_rate": 2.845089071757463e-06, + "loss": 2.5598, + "step": 22328 + }, + { + "epoch": 2.3613578680203045, + "grad_norm": 8.1875, + "learning_rate": 2.8449194541157273e-06, + "loss": 2.4726, + "step": 22329 + }, + { + "epoch": 2.3614636209813873, + "grad_norm": 8.1875, + "learning_rate": 2.8447498348554243e-06, + "loss": 2.4527, + "step": 22330 + }, + { + "epoch": 2.3615693739424706, + "grad_norm": 8.25, + "learning_rate": 2.844580213977352e-06, + "loss": 2.7196, + "step": 22331 + }, + { + "epoch": 2.3616751269035534, + "grad_norm": 8.6875, + "learning_rate": 2.8444105914823063e-06, + "loss": 2.5884, + "step": 22332 + }, + { + "epoch": 2.3617808798646363, + "grad_norm": 8.0625, + "learning_rate": 2.8442409673710808e-06, + "loss": 2.514, + "step": 22333 + }, + { + "epoch": 2.361886632825719, + "grad_norm": 9.125, + "learning_rate": 2.8440713416444745e-06, + "loss": 2.6929, + "step": 22334 + }, + { + "epoch": 2.361992385786802, + "grad_norm": 8.875, + "learning_rate": 2.843901714303282e-06, + "loss": 2.6079, + "step": 22335 + }, + { + "epoch": 2.3620981387478848, + "grad_norm": 8.8125, + "learning_rate": 2.8437320853482993e-06, + "loss": 2.5744, + "step": 22336 + }, + { + "epoch": 2.362203891708968, + "grad_norm": 8.625, + "learning_rate": 2.8435624547803225e-06, + "loss": 2.6628, + "step": 22337 + }, + { + "epoch": 2.362309644670051, + "grad_norm": 9.0625, + "learning_rate": 2.843392822600148e-06, + "loss": 2.6754, + "step": 22338 + }, + { + "epoch": 2.3624153976311337, + "grad_norm": 8.125, + "learning_rate": 2.843223188808571e-06, + "loss": 2.2742, + "step": 22339 + }, + { + "epoch": 2.3625211505922166, + "grad_norm": 7.96875, + "learning_rate": 2.843053553406388e-06, + "loss": 2.432, + "step": 22340 + }, + { + "epoch": 2.3626269035532994, + "grad_norm": 8.9375, + "learning_rate": 2.8428839163943955e-06, + "loss": 2.4821, + "step": 22341 + }, + { + "epoch": 2.3627326565143822, + "grad_norm": 8.3125, + "learning_rate": 2.842714277773388e-06, + "loss": 2.5403, + "step": 22342 + }, + { + "epoch": 2.362838409475465, + "grad_norm": 8.625, + "learning_rate": 2.8425446375441627e-06, + "loss": 2.5794, + "step": 22343 + }, + { + "epoch": 2.3629441624365484, + "grad_norm": 8.75, + "learning_rate": 2.8423749957075157e-06, + "loss": 2.6112, + "step": 22344 + }, + { + "epoch": 2.363049915397631, + "grad_norm": 8.375, + "learning_rate": 2.8422053522642434e-06, + "loss": 2.3365, + "step": 22345 + }, + { + "epoch": 2.363155668358714, + "grad_norm": 8.5, + "learning_rate": 2.842035707215141e-06, + "loss": 2.6369, + "step": 22346 + }, + { + "epoch": 2.363261421319797, + "grad_norm": 8.3125, + "learning_rate": 2.8418660605610047e-06, + "loss": 2.4095, + "step": 22347 + }, + { + "epoch": 2.3633671742808797, + "grad_norm": 9.125, + "learning_rate": 2.8416964123026313e-06, + "loss": 2.6851, + "step": 22348 + }, + { + "epoch": 2.363472927241963, + "grad_norm": 8.125, + "learning_rate": 2.841526762440816e-06, + "loss": 2.3424, + "step": 22349 + }, + { + "epoch": 2.363578680203046, + "grad_norm": 8.625, + "learning_rate": 2.8413571109763555e-06, + "loss": 2.5149, + "step": 22350 + }, + { + "epoch": 2.3636844331641287, + "grad_norm": 8.4375, + "learning_rate": 2.841187457910045e-06, + "loss": 2.5663, + "step": 22351 + }, + { + "epoch": 2.3637901861252115, + "grad_norm": 8.125, + "learning_rate": 2.841017803242682e-06, + "loss": 2.4405, + "step": 22352 + }, + { + "epoch": 2.3638959390862944, + "grad_norm": 8.1875, + "learning_rate": 2.840848146975062e-06, + "loss": 2.6124, + "step": 22353 + }, + { + "epoch": 2.364001692047377, + "grad_norm": 8.1875, + "learning_rate": 2.840678489107981e-06, + "loss": 2.4999, + "step": 22354 + }, + { + "epoch": 2.36410744500846, + "grad_norm": 8.1875, + "learning_rate": 2.8405088296422347e-06, + "loss": 2.4748, + "step": 22355 + }, + { + "epoch": 2.3642131979695433, + "grad_norm": 7.84375, + "learning_rate": 2.8403391685786204e-06, + "loss": 2.4182, + "step": 22356 + }, + { + "epoch": 2.364318950930626, + "grad_norm": 9.1875, + "learning_rate": 2.8401695059179336e-06, + "loss": 2.7236, + "step": 22357 + }, + { + "epoch": 2.364424703891709, + "grad_norm": 7.9375, + "learning_rate": 2.83999984166097e-06, + "loss": 2.4107, + "step": 22358 + }, + { + "epoch": 2.364530456852792, + "grad_norm": 8.3125, + "learning_rate": 2.839830175808526e-06, + "loss": 2.3243, + "step": 22359 + }, + { + "epoch": 2.3646362098138747, + "grad_norm": 8.125, + "learning_rate": 2.8396605083613976e-06, + "loss": 2.4907, + "step": 22360 + }, + { + "epoch": 2.3647419627749575, + "grad_norm": 8.125, + "learning_rate": 2.8394908393203826e-06, + "loss": 2.5068, + "step": 22361 + }, + { + "epoch": 2.364847715736041, + "grad_norm": 8.5625, + "learning_rate": 2.8393211686862748e-06, + "loss": 2.6929, + "step": 22362 + }, + { + "epoch": 2.3649534686971236, + "grad_norm": 8.25, + "learning_rate": 2.839151496459872e-06, + "loss": 2.4411, + "step": 22363 + }, + { + "epoch": 2.3650592216582065, + "grad_norm": 8.375, + "learning_rate": 2.8389818226419707e-06, + "loss": 2.6744, + "step": 22364 + }, + { + "epoch": 2.3651649746192893, + "grad_norm": 8.1875, + "learning_rate": 2.838812147233365e-06, + "loss": 2.6712, + "step": 22365 + }, + { + "epoch": 2.365270727580372, + "grad_norm": 8.875, + "learning_rate": 2.8386424702348538e-06, + "loss": 2.4084, + "step": 22366 + }, + { + "epoch": 2.365376480541455, + "grad_norm": 8.5, + "learning_rate": 2.838472791647231e-06, + "loss": 2.6677, + "step": 22367 + }, + { + "epoch": 2.3654822335025383, + "grad_norm": 8.4375, + "learning_rate": 2.8383031114712937e-06, + "loss": 2.5241, + "step": 22368 + }, + { + "epoch": 2.365587986463621, + "grad_norm": 8.1875, + "learning_rate": 2.8381334297078383e-06, + "loss": 2.3753, + "step": 22369 + }, + { + "epoch": 2.365693739424704, + "grad_norm": 8.1875, + "learning_rate": 2.8379637463576615e-06, + "loss": 2.1568, + "step": 22370 + }, + { + "epoch": 2.365799492385787, + "grad_norm": 8.9375, + "learning_rate": 2.8377940614215587e-06, + "loss": 2.608, + "step": 22371 + }, + { + "epoch": 2.3659052453468696, + "grad_norm": 8.5, + "learning_rate": 2.8376243749003273e-06, + "loss": 2.6703, + "step": 22372 + }, + { + "epoch": 2.3660109983079525, + "grad_norm": 8.0625, + "learning_rate": 2.8374546867947622e-06, + "loss": 2.4017, + "step": 22373 + }, + { + "epoch": 2.3661167512690353, + "grad_norm": 7.96875, + "learning_rate": 2.8372849971056598e-06, + "loss": 2.5832, + "step": 22374 + }, + { + "epoch": 2.3662225042301186, + "grad_norm": 9.0, + "learning_rate": 2.837115305833818e-06, + "loss": 2.5754, + "step": 22375 + }, + { + "epoch": 2.3663282571912014, + "grad_norm": 8.0625, + "learning_rate": 2.8369456129800317e-06, + "loss": 2.512, + "step": 22376 + }, + { + "epoch": 2.3664340101522843, + "grad_norm": 9.0, + "learning_rate": 2.8367759185450964e-06, + "loss": 2.6306, + "step": 22377 + }, + { + "epoch": 2.366539763113367, + "grad_norm": 8.6875, + "learning_rate": 2.8366062225298098e-06, + "loss": 2.5616, + "step": 22378 + }, + { + "epoch": 2.36664551607445, + "grad_norm": 8.75, + "learning_rate": 2.8364365249349685e-06, + "loss": 2.4692, + "step": 22379 + }, + { + "epoch": 2.3667512690355332, + "grad_norm": 8.3125, + "learning_rate": 2.836266825761368e-06, + "loss": 2.578, + "step": 22380 + }, + { + "epoch": 2.366857021996616, + "grad_norm": 8.1875, + "learning_rate": 2.836097125009804e-06, + "loss": 2.4641, + "step": 22381 + }, + { + "epoch": 2.366962774957699, + "grad_norm": 7.8125, + "learning_rate": 2.8359274226810746e-06, + "loss": 2.4884, + "step": 22382 + }, + { + "epoch": 2.3670685279187818, + "grad_norm": 8.8125, + "learning_rate": 2.8357577187759755e-06, + "loss": 2.5617, + "step": 22383 + }, + { + "epoch": 2.3671742808798646, + "grad_norm": 8.3125, + "learning_rate": 2.8355880132953016e-06, + "loss": 2.5638, + "step": 22384 + }, + { + "epoch": 2.3672800338409474, + "grad_norm": 8.5, + "learning_rate": 2.835418306239851e-06, + "loss": 2.6272, + "step": 22385 + }, + { + "epoch": 2.3673857868020303, + "grad_norm": 8.125, + "learning_rate": 2.8352485976104193e-06, + "loss": 2.292, + "step": 22386 + }, + { + "epoch": 2.3674915397631136, + "grad_norm": 8.0625, + "learning_rate": 2.8350788874078027e-06, + "loss": 2.5422, + "step": 22387 + }, + { + "epoch": 2.3675972927241964, + "grad_norm": 8.625, + "learning_rate": 2.834909175632799e-06, + "loss": 2.5233, + "step": 22388 + }, + { + "epoch": 2.3677030456852792, + "grad_norm": 8.8125, + "learning_rate": 2.834739462286203e-06, + "loss": 2.4079, + "step": 22389 + }, + { + "epoch": 2.367808798646362, + "grad_norm": 7.90625, + "learning_rate": 2.8345697473688113e-06, + "loss": 2.5001, + "step": 22390 + }, + { + "epoch": 2.367914551607445, + "grad_norm": 8.3125, + "learning_rate": 2.8344000308814206e-06, + "loss": 2.5676, + "step": 22391 + }, + { + "epoch": 2.3680203045685277, + "grad_norm": 8.375, + "learning_rate": 2.834230312824828e-06, + "loss": 2.5311, + "step": 22392 + }, + { + "epoch": 2.3681260575296106, + "grad_norm": 8.3125, + "learning_rate": 2.834060593199829e-06, + "loss": 2.5355, + "step": 22393 + }, + { + "epoch": 2.368231810490694, + "grad_norm": 8.125, + "learning_rate": 2.83389087200722e-06, + "loss": 2.6328, + "step": 22394 + }, + { + "epoch": 2.3683375634517767, + "grad_norm": 8.25, + "learning_rate": 2.8337211492477974e-06, + "loss": 2.6696, + "step": 22395 + }, + { + "epoch": 2.3684433164128595, + "grad_norm": 9.3125, + "learning_rate": 2.833551424922358e-06, + "loss": 2.7902, + "step": 22396 + }, + { + "epoch": 2.3685490693739424, + "grad_norm": 7.9375, + "learning_rate": 2.8333816990316986e-06, + "loss": 2.4881, + "step": 22397 + }, + { + "epoch": 2.3686548223350252, + "grad_norm": 9.0, + "learning_rate": 2.833211971576615e-06, + "loss": 2.5746, + "step": 22398 + }, + { + "epoch": 2.3687605752961085, + "grad_norm": 8.125, + "learning_rate": 2.8330422425579045e-06, + "loss": 2.5652, + "step": 22399 + }, + { + "epoch": 2.3688663282571913, + "grad_norm": 8.625, + "learning_rate": 2.832872511976362e-06, + "loss": 2.6681, + "step": 22400 + }, + { + "epoch": 2.368972081218274, + "grad_norm": 8.625, + "learning_rate": 2.8327027798327855e-06, + "loss": 2.519, + "step": 22401 + }, + { + "epoch": 2.369077834179357, + "grad_norm": 8.4375, + "learning_rate": 2.832533046127971e-06, + "loss": 2.5969, + "step": 22402 + }, + { + "epoch": 2.36918358714044, + "grad_norm": 8.1875, + "learning_rate": 2.8323633108627148e-06, + "loss": 2.7916, + "step": 22403 + }, + { + "epoch": 2.3692893401015227, + "grad_norm": 8.0625, + "learning_rate": 2.832193574037813e-06, + "loss": 2.6269, + "step": 22404 + }, + { + "epoch": 2.3693950930626055, + "grad_norm": 8.375, + "learning_rate": 2.832023835654063e-06, + "loss": 2.4877, + "step": 22405 + }, + { + "epoch": 2.369500846023689, + "grad_norm": 9.3125, + "learning_rate": 2.831854095712261e-06, + "loss": 2.6578, + "step": 22406 + }, + { + "epoch": 2.3696065989847717, + "grad_norm": 8.625, + "learning_rate": 2.8316843542132034e-06, + "loss": 2.5086, + "step": 22407 + }, + { + "epoch": 2.3697123519458545, + "grad_norm": 8.5, + "learning_rate": 2.8315146111576875e-06, + "loss": 2.6325, + "step": 22408 + }, + { + "epoch": 2.3698181049069373, + "grad_norm": 8.5625, + "learning_rate": 2.8313448665465076e-06, + "loss": 2.5963, + "step": 22409 + }, + { + "epoch": 2.36992385786802, + "grad_norm": 9.375, + "learning_rate": 2.831175120380463e-06, + "loss": 2.6228, + "step": 22410 + }, + { + "epoch": 2.3700296108291035, + "grad_norm": 7.5625, + "learning_rate": 2.8310053726603487e-06, + "loss": 2.4783, + "step": 22411 + }, + { + "epoch": 2.3701353637901863, + "grad_norm": 8.6875, + "learning_rate": 2.8308356233869617e-06, + "loss": 2.4116, + "step": 22412 + }, + { + "epoch": 2.370241116751269, + "grad_norm": 8.75, + "learning_rate": 2.8306658725610976e-06, + "loss": 2.5583, + "step": 22413 + }, + { + "epoch": 2.370346869712352, + "grad_norm": 9.0, + "learning_rate": 2.8304961201835547e-06, + "loss": 2.6429, + "step": 22414 + }, + { + "epoch": 2.370452622673435, + "grad_norm": 9.0625, + "learning_rate": 2.8303263662551284e-06, + "loss": 2.514, + "step": 22415 + }, + { + "epoch": 2.3705583756345177, + "grad_norm": 8.25, + "learning_rate": 2.8301566107766153e-06, + "loss": 2.3306, + "step": 22416 + }, + { + "epoch": 2.3706641285956005, + "grad_norm": 9.375, + "learning_rate": 2.8299868537488128e-06, + "loss": 2.4051, + "step": 22417 + }, + { + "epoch": 2.370769881556684, + "grad_norm": 8.1875, + "learning_rate": 2.8298170951725167e-06, + "loss": 2.554, + "step": 22418 + }, + { + "epoch": 2.3708756345177666, + "grad_norm": 8.375, + "learning_rate": 2.829647335048524e-06, + "loss": 2.2699, + "step": 22419 + }, + { + "epoch": 2.3709813874788495, + "grad_norm": 8.75, + "learning_rate": 2.829477573377631e-06, + "loss": 2.677, + "step": 22420 + }, + { + "epoch": 2.3710871404399323, + "grad_norm": 8.25, + "learning_rate": 2.829307810160634e-06, + "loss": 2.6016, + "step": 22421 + }, + { + "epoch": 2.371192893401015, + "grad_norm": 8.875, + "learning_rate": 2.8291380453983304e-06, + "loss": 2.5653, + "step": 22422 + }, + { + "epoch": 2.371298646362098, + "grad_norm": 7.90625, + "learning_rate": 2.828968279091517e-06, + "loss": 2.3867, + "step": 22423 + }, + { + "epoch": 2.371404399323181, + "grad_norm": 9.125, + "learning_rate": 2.82879851124099e-06, + "loss": 2.5385, + "step": 22424 + }, + { + "epoch": 2.371510152284264, + "grad_norm": 8.25, + "learning_rate": 2.8286287418475457e-06, + "loss": 2.4916, + "step": 22425 + }, + { + "epoch": 2.371615905245347, + "grad_norm": 8.6875, + "learning_rate": 2.8284589709119815e-06, + "loss": 2.5073, + "step": 22426 + }, + { + "epoch": 2.3717216582064298, + "grad_norm": 8.1875, + "learning_rate": 2.8282891984350934e-06, + "loss": 2.474, + "step": 22427 + }, + { + "epoch": 2.3718274111675126, + "grad_norm": 8.6875, + "learning_rate": 2.828119424417678e-06, + "loss": 2.4755, + "step": 22428 + }, + { + "epoch": 2.3719331641285955, + "grad_norm": 8.5, + "learning_rate": 2.827949648860533e-06, + "loss": 2.556, + "step": 22429 + }, + { + "epoch": 2.3720389170896787, + "grad_norm": 8.625, + "learning_rate": 2.8277798717644546e-06, + "loss": 2.4121, + "step": 22430 + }, + { + "epoch": 2.3721446700507616, + "grad_norm": 8.375, + "learning_rate": 2.8276100931302384e-06, + "loss": 2.6928, + "step": 22431 + }, + { + "epoch": 2.3722504230118444, + "grad_norm": 8.25, + "learning_rate": 2.827440312958682e-06, + "loss": 2.3713, + "step": 22432 + }, + { + "epoch": 2.3723561759729273, + "grad_norm": 8.3125, + "learning_rate": 2.8272705312505832e-06, + "loss": 2.394, + "step": 22433 + }, + { + "epoch": 2.37246192893401, + "grad_norm": 8.8125, + "learning_rate": 2.827100748006737e-06, + "loss": 2.6375, + "step": 22434 + }, + { + "epoch": 2.372567681895093, + "grad_norm": 8.5, + "learning_rate": 2.8269309632279406e-06, + "loss": 2.7578, + "step": 22435 + }, + { + "epoch": 2.3726734348561758, + "grad_norm": 8.5, + "learning_rate": 2.8267611769149916e-06, + "loss": 2.7689, + "step": 22436 + }, + { + "epoch": 2.372779187817259, + "grad_norm": 8.5, + "learning_rate": 2.8265913890686857e-06, + "loss": 2.6047, + "step": 22437 + }, + { + "epoch": 2.372884940778342, + "grad_norm": 8.4375, + "learning_rate": 2.8264215996898197e-06, + "loss": 2.6354, + "step": 22438 + }, + { + "epoch": 2.3729906937394247, + "grad_norm": 8.3125, + "learning_rate": 2.8262518087791903e-06, + "loss": 2.3081, + "step": 22439 + }, + { + "epoch": 2.3730964467005076, + "grad_norm": 8.8125, + "learning_rate": 2.826082016337595e-06, + "loss": 2.7082, + "step": 22440 + }, + { + "epoch": 2.3732021996615904, + "grad_norm": 8.375, + "learning_rate": 2.82591222236583e-06, + "loss": 2.5325, + "step": 22441 + }, + { + "epoch": 2.3733079526226732, + "grad_norm": 8.875, + "learning_rate": 2.8257424268646934e-06, + "loss": 2.6041, + "step": 22442 + }, + { + "epoch": 2.3734137055837565, + "grad_norm": 8.125, + "learning_rate": 2.8255726298349795e-06, + "loss": 2.5856, + "step": 22443 + }, + { + "epoch": 2.3735194585448394, + "grad_norm": 8.1875, + "learning_rate": 2.8254028312774866e-06, + "loss": 2.4626, + "step": 22444 + }, + { + "epoch": 2.373625211505922, + "grad_norm": 9.25, + "learning_rate": 2.8252330311930123e-06, + "loss": 2.5489, + "step": 22445 + }, + { + "epoch": 2.373730964467005, + "grad_norm": 8.25, + "learning_rate": 2.8250632295823514e-06, + "loss": 2.4836, + "step": 22446 + }, + { + "epoch": 2.373836717428088, + "grad_norm": 8.5625, + "learning_rate": 2.824893426446302e-06, + "loss": 2.5176, + "step": 22447 + }, + { + "epoch": 2.3739424703891707, + "grad_norm": 8.625, + "learning_rate": 2.8247236217856598e-06, + "loss": 2.215, + "step": 22448 + }, + { + "epoch": 2.374048223350254, + "grad_norm": 8.3125, + "learning_rate": 2.8245538156012234e-06, + "loss": 2.4255, + "step": 22449 + }, + { + "epoch": 2.374153976311337, + "grad_norm": 8.375, + "learning_rate": 2.824384007893789e-06, + "loss": 2.5991, + "step": 22450 + }, + { + "epoch": 2.3742597292724197, + "grad_norm": 8.6875, + "learning_rate": 2.824214198664152e-06, + "loss": 2.4483, + "step": 22451 + }, + { + "epoch": 2.3743654822335025, + "grad_norm": 8.1875, + "learning_rate": 2.8240443879131114e-06, + "loss": 2.5091, + "step": 22452 + }, + { + "epoch": 2.3744712351945854, + "grad_norm": 8.375, + "learning_rate": 2.8238745756414627e-06, + "loss": 2.6927, + "step": 22453 + }, + { + "epoch": 2.374576988155668, + "grad_norm": 8.6875, + "learning_rate": 2.8237047618500026e-06, + "loss": 2.3756, + "step": 22454 + }, + { + "epoch": 2.374682741116751, + "grad_norm": 8.5, + "learning_rate": 2.823534946539529e-06, + "loss": 2.5966, + "step": 22455 + }, + { + "epoch": 2.3747884940778343, + "grad_norm": 8.1875, + "learning_rate": 2.8233651297108385e-06, + "loss": 2.4771, + "step": 22456 + }, + { + "epoch": 2.374894247038917, + "grad_norm": 8.5625, + "learning_rate": 2.823195311364726e-06, + "loss": 2.4137, + "step": 22457 + }, + { + "epoch": 2.375, + "grad_norm": 8.8125, + "learning_rate": 2.8230254915019917e-06, + "loss": 2.5267, + "step": 22458 + }, + { + "epoch": 2.375105752961083, + "grad_norm": 8.0625, + "learning_rate": 2.8228556701234305e-06, + "loss": 2.502, + "step": 22459 + }, + { + "epoch": 2.3752115059221657, + "grad_norm": 8.5625, + "learning_rate": 2.8226858472298395e-06, + "loss": 2.5849, + "step": 22460 + }, + { + "epoch": 2.375317258883249, + "grad_norm": 8.25, + "learning_rate": 2.8225160228220162e-06, + "loss": 2.4673, + "step": 22461 + }, + { + "epoch": 2.375423011844332, + "grad_norm": 8.125, + "learning_rate": 2.8223461969007572e-06, + "loss": 2.4621, + "step": 22462 + }, + { + "epoch": 2.3755287648054146, + "grad_norm": 8.375, + "learning_rate": 2.8221763694668587e-06, + "loss": 2.3251, + "step": 22463 + }, + { + "epoch": 2.3756345177664975, + "grad_norm": 8.9375, + "learning_rate": 2.822006540521119e-06, + "loss": 2.4819, + "step": 22464 + }, + { + "epoch": 2.3757402707275803, + "grad_norm": 8.375, + "learning_rate": 2.821836710064334e-06, + "loss": 2.402, + "step": 22465 + }, + { + "epoch": 2.375846023688663, + "grad_norm": 8.6875, + "learning_rate": 2.8216668780973007e-06, + "loss": 2.5952, + "step": 22466 + }, + { + "epoch": 2.375951776649746, + "grad_norm": 8.875, + "learning_rate": 2.8214970446208157e-06, + "loss": 2.4698, + "step": 22467 + }, + { + "epoch": 2.3760575296108293, + "grad_norm": 8.4375, + "learning_rate": 2.8213272096356774e-06, + "loss": 2.5822, + "step": 22468 + }, + { + "epoch": 2.376163282571912, + "grad_norm": 7.875, + "learning_rate": 2.8211573731426823e-06, + "loss": 2.3729, + "step": 22469 + }, + { + "epoch": 2.376269035532995, + "grad_norm": 8.5625, + "learning_rate": 2.8209875351426263e-06, + "loss": 2.3149, + "step": 22470 + }, + { + "epoch": 2.376374788494078, + "grad_norm": 8.6875, + "learning_rate": 2.8208176956363078e-06, + "loss": 2.6873, + "step": 22471 + }, + { + "epoch": 2.3764805414551606, + "grad_norm": 9.0625, + "learning_rate": 2.820647854624522e-06, + "loss": 2.5726, + "step": 22472 + }, + { + "epoch": 2.3765862944162435, + "grad_norm": 8.4375, + "learning_rate": 2.8204780121080676e-06, + "loss": 2.516, + "step": 22473 + }, + { + "epoch": 2.3766920473773268, + "grad_norm": 7.65625, + "learning_rate": 2.820308168087741e-06, + "loss": 2.4871, + "step": 22474 + }, + { + "epoch": 2.3767978003384096, + "grad_norm": 8.5, + "learning_rate": 2.8201383225643385e-06, + "loss": 2.5199, + "step": 22475 + }, + { + "epoch": 2.3769035532994924, + "grad_norm": 8.6875, + "learning_rate": 2.819968475538658e-06, + "loss": 2.5456, + "step": 22476 + }, + { + "epoch": 2.3770093062605753, + "grad_norm": 9.0625, + "learning_rate": 2.819798627011497e-06, + "loss": 2.338, + "step": 22477 + }, + { + "epoch": 2.377115059221658, + "grad_norm": 8.6875, + "learning_rate": 2.8196287769836516e-06, + "loss": 2.5529, + "step": 22478 + }, + { + "epoch": 2.377220812182741, + "grad_norm": 8.4375, + "learning_rate": 2.8194589254559183e-06, + "loss": 2.6635, + "step": 22479 + }, + { + "epoch": 2.3773265651438242, + "grad_norm": 8.1875, + "learning_rate": 2.8192890724290963e-06, + "loss": 2.3261, + "step": 22480 + }, + { + "epoch": 2.377432318104907, + "grad_norm": 8.125, + "learning_rate": 2.8191192179039806e-06, + "loss": 2.6165, + "step": 22481 + }, + { + "epoch": 2.37753807106599, + "grad_norm": 8.6875, + "learning_rate": 2.8189493618813683e-06, + "loss": 2.6063, + "step": 22482 + }, + { + "epoch": 2.3776438240270727, + "grad_norm": 8.375, + "learning_rate": 2.818779504362058e-06, + "loss": 2.6369, + "step": 22483 + }, + { + "epoch": 2.3777495769881556, + "grad_norm": 8.25, + "learning_rate": 2.8186096453468453e-06, + "loss": 2.5054, + "step": 22484 + }, + { + "epoch": 2.3778553299492384, + "grad_norm": 8.375, + "learning_rate": 2.818439784836528e-06, + "loss": 2.4277, + "step": 22485 + }, + { + "epoch": 2.3779610829103213, + "grad_norm": 8.375, + "learning_rate": 2.8182699228319025e-06, + "loss": 2.6458, + "step": 22486 + }, + { + "epoch": 2.3780668358714045, + "grad_norm": 8.3125, + "learning_rate": 2.8181000593337675e-06, + "loss": 2.4033, + "step": 22487 + }, + { + "epoch": 2.3781725888324874, + "grad_norm": 8.8125, + "learning_rate": 2.817930194342919e-06, + "loss": 2.6214, + "step": 22488 + }, + { + "epoch": 2.3782783417935702, + "grad_norm": 8.625, + "learning_rate": 2.8177603278601528e-06, + "loss": 2.5582, + "step": 22489 + }, + { + "epoch": 2.378384094754653, + "grad_norm": 9.0, + "learning_rate": 2.8175904598862685e-06, + "loss": 2.8828, + "step": 22490 + }, + { + "epoch": 2.378489847715736, + "grad_norm": 8.9375, + "learning_rate": 2.817420590422062e-06, + "loss": 2.2273, + "step": 22491 + }, + { + "epoch": 2.378595600676819, + "grad_norm": 8.25, + "learning_rate": 2.81725071946833e-06, + "loss": 2.5678, + "step": 22492 + }, + { + "epoch": 2.378701353637902, + "grad_norm": 9.8125, + "learning_rate": 2.81708084702587e-06, + "loss": 2.6146, + "step": 22493 + }, + { + "epoch": 2.378807106598985, + "grad_norm": 8.5625, + "learning_rate": 2.8169109730954796e-06, + "loss": 2.5733, + "step": 22494 + }, + { + "epoch": 2.3789128595600677, + "grad_norm": 8.3125, + "learning_rate": 2.816741097677956e-06, + "loss": 2.5047, + "step": 22495 + }, + { + "epoch": 2.3790186125211505, + "grad_norm": 8.8125, + "learning_rate": 2.8165712207740954e-06, + "loss": 2.7183, + "step": 22496 + }, + { + "epoch": 2.3791243654822334, + "grad_norm": 8.0625, + "learning_rate": 2.816401342384696e-06, + "loss": 2.4179, + "step": 22497 + }, + { + "epoch": 2.379230118443316, + "grad_norm": 8.25, + "learning_rate": 2.8162314625105536e-06, + "loss": 2.62, + "step": 22498 + }, + { + "epoch": 2.3793358714043995, + "grad_norm": 8.375, + "learning_rate": 2.816061581152467e-06, + "loss": 2.4994, + "step": 22499 + }, + { + "epoch": 2.3794416243654823, + "grad_norm": 8.375, + "learning_rate": 2.815891698311233e-06, + "loss": 2.5921, + "step": 22500 + }, + { + "epoch": 2.379547377326565, + "grad_norm": 9.125, + "learning_rate": 2.815721813987648e-06, + "loss": 2.4273, + "step": 22501 + }, + { + "epoch": 2.379653130287648, + "grad_norm": 8.375, + "learning_rate": 2.8155519281825096e-06, + "loss": 2.5388, + "step": 22502 + }, + { + "epoch": 2.379758883248731, + "grad_norm": 8.5, + "learning_rate": 2.8153820408966147e-06, + "loss": 2.4768, + "step": 22503 + }, + { + "epoch": 2.3798646362098137, + "grad_norm": 8.5625, + "learning_rate": 2.8152121521307618e-06, + "loss": 2.5772, + "step": 22504 + }, + { + "epoch": 2.3799703891708965, + "grad_norm": 8.5, + "learning_rate": 2.8150422618857466e-06, + "loss": 2.5821, + "step": 22505 + }, + { + "epoch": 2.38007614213198, + "grad_norm": 8.375, + "learning_rate": 2.814872370162367e-06, + "loss": 2.5089, + "step": 22506 + }, + { + "epoch": 2.3801818950930627, + "grad_norm": 8.875, + "learning_rate": 2.8147024769614205e-06, + "loss": 2.4451, + "step": 22507 + }, + { + "epoch": 2.3802876480541455, + "grad_norm": 8.5, + "learning_rate": 2.8145325822837032e-06, + "loss": 2.5913, + "step": 22508 + }, + { + "epoch": 2.3803934010152283, + "grad_norm": 8.4375, + "learning_rate": 2.8143626861300137e-06, + "loss": 2.599, + "step": 22509 + }, + { + "epoch": 2.380499153976311, + "grad_norm": 8.625, + "learning_rate": 2.8141927885011485e-06, + "loss": 2.6571, + "step": 22510 + }, + { + "epoch": 2.3806049069373945, + "grad_norm": 8.25, + "learning_rate": 2.8140228893979047e-06, + "loss": 2.6419, + "step": 22511 + }, + { + "epoch": 2.3807106598984773, + "grad_norm": 8.375, + "learning_rate": 2.8138529888210807e-06, + "loss": 2.499, + "step": 22512 + }, + { + "epoch": 2.38081641285956, + "grad_norm": 8.1875, + "learning_rate": 2.8136830867714726e-06, + "loss": 2.6412, + "step": 22513 + }, + { + "epoch": 2.380922165820643, + "grad_norm": 8.9375, + "learning_rate": 2.813513183249878e-06, + "loss": 2.6834, + "step": 22514 + }, + { + "epoch": 2.381027918781726, + "grad_norm": 9.25, + "learning_rate": 2.8133432782570947e-06, + "loss": 2.7329, + "step": 22515 + }, + { + "epoch": 2.3811336717428087, + "grad_norm": 8.625, + "learning_rate": 2.813173371793919e-06, + "loss": 2.5819, + "step": 22516 + }, + { + "epoch": 2.3812394247038915, + "grad_norm": 7.9375, + "learning_rate": 2.8130034638611487e-06, + "loss": 2.446, + "step": 22517 + }, + { + "epoch": 2.3813451776649748, + "grad_norm": 8.625, + "learning_rate": 2.8128335544595814e-06, + "loss": 2.5426, + "step": 22518 + }, + { + "epoch": 2.3814509306260576, + "grad_norm": 8.875, + "learning_rate": 2.812663643590014e-06, + "loss": 2.4139, + "step": 22519 + }, + { + "epoch": 2.3815566835871405, + "grad_norm": 8.4375, + "learning_rate": 2.812493731253244e-06, + "loss": 2.5358, + "step": 22520 + }, + { + "epoch": 2.3816624365482233, + "grad_norm": 8.6875, + "learning_rate": 2.812323817450069e-06, + "loss": 2.6356, + "step": 22521 + }, + { + "epoch": 2.381768189509306, + "grad_norm": 8.3125, + "learning_rate": 2.812153902181286e-06, + "loss": 2.4092, + "step": 22522 + }, + { + "epoch": 2.3818739424703894, + "grad_norm": 8.3125, + "learning_rate": 2.8119839854476927e-06, + "loss": 2.3039, + "step": 22523 + }, + { + "epoch": 2.3819796954314723, + "grad_norm": 8.3125, + "learning_rate": 2.8118140672500854e-06, + "loss": 2.4828, + "step": 22524 + }, + { + "epoch": 2.382085448392555, + "grad_norm": 9.5, + "learning_rate": 2.811644147589263e-06, + "loss": 2.5769, + "step": 22525 + }, + { + "epoch": 2.382191201353638, + "grad_norm": 8.0625, + "learning_rate": 2.811474226466022e-06, + "loss": 2.4045, + "step": 22526 + }, + { + "epoch": 2.3822969543147208, + "grad_norm": 8.75, + "learning_rate": 2.8113043038811588e-06, + "loss": 2.6231, + "step": 22527 + }, + { + "epoch": 2.3824027072758036, + "grad_norm": 7.875, + "learning_rate": 2.811134379835472e-06, + "loss": 2.4595, + "step": 22528 + }, + { + "epoch": 2.3825084602368864, + "grad_norm": 8.8125, + "learning_rate": 2.81096445432976e-06, + "loss": 2.8691, + "step": 22529 + }, + { + "epoch": 2.3826142131979697, + "grad_norm": 8.3125, + "learning_rate": 2.810794527364818e-06, + "loss": 2.6351, + "step": 22530 + }, + { + "epoch": 2.3827199661590526, + "grad_norm": 8.75, + "learning_rate": 2.810624598941445e-06, + "loss": 2.7976, + "step": 22531 + }, + { + "epoch": 2.3828257191201354, + "grad_norm": 9.375, + "learning_rate": 2.8104546690604377e-06, + "loss": 2.6683, + "step": 22532 + }, + { + "epoch": 2.3829314720812182, + "grad_norm": 8.6875, + "learning_rate": 2.810284737722593e-06, + "loss": 2.5819, + "step": 22533 + }, + { + "epoch": 2.383037225042301, + "grad_norm": 8.4375, + "learning_rate": 2.8101148049287097e-06, + "loss": 2.4253, + "step": 22534 + }, + { + "epoch": 2.383142978003384, + "grad_norm": 8.875, + "learning_rate": 2.809944870679584e-06, + "loss": 2.5008, + "step": 22535 + }, + { + "epoch": 2.3832487309644668, + "grad_norm": 7.96875, + "learning_rate": 2.809774934976014e-06, + "loss": 2.5336, + "step": 22536 + }, + { + "epoch": 2.38335448392555, + "grad_norm": 8.0, + "learning_rate": 2.8096049978187962e-06, + "loss": 2.4152, + "step": 22537 + }, + { + "epoch": 2.383460236886633, + "grad_norm": 8.3125, + "learning_rate": 2.80943505920873e-06, + "loss": 2.4074, + "step": 22538 + }, + { + "epoch": 2.3835659898477157, + "grad_norm": 8.25, + "learning_rate": 2.809265119146611e-06, + "loss": 2.6222, + "step": 22539 + }, + { + "epoch": 2.3836717428087986, + "grad_norm": 7.96875, + "learning_rate": 2.809095177633237e-06, + "loss": 2.5467, + "step": 22540 + }, + { + "epoch": 2.3837774957698814, + "grad_norm": 9.625, + "learning_rate": 2.808925234669406e-06, + "loss": 2.6964, + "step": 22541 + }, + { + "epoch": 2.3838832487309647, + "grad_norm": 8.1875, + "learning_rate": 2.8087552902559157e-06, + "loss": 2.5742, + "step": 22542 + }, + { + "epoch": 2.3839890016920475, + "grad_norm": 8.0, + "learning_rate": 2.8085853443935623e-06, + "loss": 2.6784, + "step": 22543 + }, + { + "epoch": 2.3840947546531304, + "grad_norm": 8.5625, + "learning_rate": 2.8084153970831447e-06, + "loss": 2.5567, + "step": 22544 + }, + { + "epoch": 2.384200507614213, + "grad_norm": 8.375, + "learning_rate": 2.8082454483254595e-06, + "loss": 2.5534, + "step": 22545 + }, + { + "epoch": 2.384306260575296, + "grad_norm": 8.5625, + "learning_rate": 2.8080754981213037e-06, + "loss": 2.6511, + "step": 22546 + }, + { + "epoch": 2.384412013536379, + "grad_norm": 8.625, + "learning_rate": 2.8079055464714766e-06, + "loss": 2.498, + "step": 22547 + }, + { + "epoch": 2.3845177664974617, + "grad_norm": 8.0625, + "learning_rate": 2.8077355933767747e-06, + "loss": 2.384, + "step": 22548 + }, + { + "epoch": 2.384623519458545, + "grad_norm": 7.84375, + "learning_rate": 2.807565638837995e-06, + "loss": 2.5442, + "step": 22549 + }, + { + "epoch": 2.384729272419628, + "grad_norm": 8.0, + "learning_rate": 2.8073956828559363e-06, + "loss": 2.4272, + "step": 22550 + }, + { + "epoch": 2.3848350253807107, + "grad_norm": 8.6875, + "learning_rate": 2.807225725431395e-06, + "loss": 2.4956, + "step": 22551 + }, + { + "epoch": 2.3849407783417935, + "grad_norm": 8.3125, + "learning_rate": 2.807055766565169e-06, + "loss": 2.5026, + "step": 22552 + }, + { + "epoch": 2.3850465313028764, + "grad_norm": 7.9375, + "learning_rate": 2.8068858062580564e-06, + "loss": 2.4618, + "step": 22553 + }, + { + "epoch": 2.385152284263959, + "grad_norm": 8.5, + "learning_rate": 2.806715844510853e-06, + "loss": 2.3607, + "step": 22554 + }, + { + "epoch": 2.3852580372250425, + "grad_norm": 8.375, + "learning_rate": 2.8065458813243586e-06, + "loss": 2.3979, + "step": 22555 + }, + { + "epoch": 2.3853637901861253, + "grad_norm": 8.8125, + "learning_rate": 2.8063759166993687e-06, + "loss": 2.5226, + "step": 22556 + }, + { + "epoch": 2.385469543147208, + "grad_norm": 8.5625, + "learning_rate": 2.806205950636683e-06, + "loss": 2.4544, + "step": 22557 + }, + { + "epoch": 2.385575296108291, + "grad_norm": 8.4375, + "learning_rate": 2.8060359831370975e-06, + "loss": 2.5301, + "step": 22558 + }, + { + "epoch": 2.385681049069374, + "grad_norm": 8.75, + "learning_rate": 2.8058660142014103e-06, + "loss": 2.5975, + "step": 22559 + }, + { + "epoch": 2.3857868020304567, + "grad_norm": 8.4375, + "learning_rate": 2.805696043830419e-06, + "loss": 2.5464, + "step": 22560 + }, + { + "epoch": 2.38589255499154, + "grad_norm": 8.25, + "learning_rate": 2.805526072024922e-06, + "loss": 2.5264, + "step": 22561 + }, + { + "epoch": 2.385998307952623, + "grad_norm": 8.375, + "learning_rate": 2.8053560987857146e-06, + "loss": 2.4875, + "step": 22562 + }, + { + "epoch": 2.3861040609137056, + "grad_norm": 8.125, + "learning_rate": 2.8051861241135968e-06, + "loss": 2.5432, + "step": 22563 + }, + { + "epoch": 2.3862098138747885, + "grad_norm": 9.125, + "learning_rate": 2.8050161480093656e-06, + "loss": 2.6383, + "step": 22564 + }, + { + "epoch": 2.3863155668358713, + "grad_norm": 8.4375, + "learning_rate": 2.8048461704738177e-06, + "loss": 2.6058, + "step": 22565 + }, + { + "epoch": 2.386421319796954, + "grad_norm": 10.8125, + "learning_rate": 2.804676191507752e-06, + "loss": 2.5162, + "step": 22566 + }, + { + "epoch": 2.386527072758037, + "grad_norm": 8.75, + "learning_rate": 2.8045062111119652e-06, + "loss": 2.6387, + "step": 22567 + }, + { + "epoch": 2.3866328257191203, + "grad_norm": 8.3125, + "learning_rate": 2.804336229287255e-06, + "loss": 2.5792, + "step": 22568 + }, + { + "epoch": 2.386738578680203, + "grad_norm": 8.3125, + "learning_rate": 2.80416624603442e-06, + "loss": 2.6142, + "step": 22569 + }, + { + "epoch": 2.386844331641286, + "grad_norm": 8.1875, + "learning_rate": 2.803996261354257e-06, + "loss": 2.6362, + "step": 22570 + }, + { + "epoch": 2.386950084602369, + "grad_norm": 8.375, + "learning_rate": 2.803826275247564e-06, + "loss": 2.3301, + "step": 22571 + }, + { + "epoch": 2.3870558375634516, + "grad_norm": 8.125, + "learning_rate": 2.8036562877151375e-06, + "loss": 2.4156, + "step": 22572 + }, + { + "epoch": 2.387161590524535, + "grad_norm": 8.625, + "learning_rate": 2.8034862987577777e-06, + "loss": 2.5825, + "step": 22573 + }, + { + "epoch": 2.3872673434856178, + "grad_norm": 8.875, + "learning_rate": 2.8033163083762806e-06, + "loss": 2.649, + "step": 22574 + }, + { + "epoch": 2.3873730964467006, + "grad_norm": 8.625, + "learning_rate": 2.803146316571443e-06, + "loss": 2.4528, + "step": 22575 + }, + { + "epoch": 2.3874788494077834, + "grad_norm": 8.0625, + "learning_rate": 2.8029763233440642e-06, + "loss": 2.4628, + "step": 22576 + }, + { + "epoch": 2.3875846023688663, + "grad_norm": 8.1875, + "learning_rate": 2.8028063286949424e-06, + "loss": 2.2976, + "step": 22577 + }, + { + "epoch": 2.387690355329949, + "grad_norm": 8.25, + "learning_rate": 2.802636332624873e-06, + "loss": 2.4838, + "step": 22578 + }, + { + "epoch": 2.387796108291032, + "grad_norm": 8.25, + "learning_rate": 2.802466335134656e-06, + "loss": 2.5575, + "step": 22579 + }, + { + "epoch": 2.3879018612521152, + "grad_norm": 8.4375, + "learning_rate": 2.802296336225088e-06, + "loss": 2.6276, + "step": 22580 + }, + { + "epoch": 2.388007614213198, + "grad_norm": 8.5625, + "learning_rate": 2.8021263358969657e-06, + "loss": 2.6937, + "step": 22581 + }, + { + "epoch": 2.388113367174281, + "grad_norm": 9.5, + "learning_rate": 2.8019563341510897e-06, + "loss": 2.9105, + "step": 22582 + }, + { + "epoch": 2.3882191201353637, + "grad_norm": 9.875, + "learning_rate": 2.801786330988256e-06, + "loss": 2.5375, + "step": 22583 + }, + { + "epoch": 2.3883248730964466, + "grad_norm": 8.5625, + "learning_rate": 2.8016163264092615e-06, + "loss": 2.7171, + "step": 22584 + }, + { + "epoch": 2.3884306260575294, + "grad_norm": 8.5625, + "learning_rate": 2.8014463204149054e-06, + "loss": 2.7985, + "step": 22585 + }, + { + "epoch": 2.3885363790186127, + "grad_norm": 9.0, + "learning_rate": 2.8012763130059857e-06, + "loss": 2.4486, + "step": 22586 + }, + { + "epoch": 2.3886421319796955, + "grad_norm": 9.0625, + "learning_rate": 2.8011063041832985e-06, + "loss": 2.5858, + "step": 22587 + }, + { + "epoch": 2.3887478849407784, + "grad_norm": 8.625, + "learning_rate": 2.800936293947643e-06, + "loss": 2.8786, + "step": 22588 + }, + { + "epoch": 2.3888536379018612, + "grad_norm": 8.4375, + "learning_rate": 2.8007662822998164e-06, + "loss": 2.5211, + "step": 22589 + }, + { + "epoch": 2.388959390862944, + "grad_norm": 8.0, + "learning_rate": 2.800596269240617e-06, + "loss": 2.551, + "step": 22590 + }, + { + "epoch": 2.389065143824027, + "grad_norm": 8.3125, + "learning_rate": 2.8004262547708415e-06, + "loss": 2.7283, + "step": 22591 + }, + { + "epoch": 2.38917089678511, + "grad_norm": 8.5625, + "learning_rate": 2.800256238891289e-06, + "loss": 2.5869, + "step": 22592 + }, + { + "epoch": 2.389276649746193, + "grad_norm": 8.5, + "learning_rate": 2.800086221602757e-06, + "loss": 2.5579, + "step": 22593 + }, + { + "epoch": 2.389382402707276, + "grad_norm": 8.9375, + "learning_rate": 2.799916202906042e-06, + "loss": 2.7725, + "step": 22594 + }, + { + "epoch": 2.3894881556683587, + "grad_norm": 8.125, + "learning_rate": 2.7997461828019446e-06, + "loss": 2.471, + "step": 22595 + }, + { + "epoch": 2.3895939086294415, + "grad_norm": 8.375, + "learning_rate": 2.7995761612912604e-06, + "loss": 2.5169, + "step": 22596 + }, + { + "epoch": 2.3896996615905244, + "grad_norm": 8.75, + "learning_rate": 2.7994061383747865e-06, + "loss": 2.6517, + "step": 22597 + }, + { + "epoch": 2.389805414551607, + "grad_norm": 8.875, + "learning_rate": 2.7992361140533234e-06, + "loss": 2.7282, + "step": 22598 + }, + { + "epoch": 2.3899111675126905, + "grad_norm": 8.5625, + "learning_rate": 2.7990660883276673e-06, + "loss": 2.5344, + "step": 22599 + }, + { + "epoch": 2.3900169204737733, + "grad_norm": 8.75, + "learning_rate": 2.798896061198616e-06, + "loss": 2.5092, + "step": 22600 + }, + { + "epoch": 2.390122673434856, + "grad_norm": 8.25, + "learning_rate": 2.798726032666968e-06, + "loss": 2.3356, + "step": 22601 + }, + { + "epoch": 2.390228426395939, + "grad_norm": 8.1875, + "learning_rate": 2.7985560027335213e-06, + "loss": 2.4599, + "step": 22602 + }, + { + "epoch": 2.390334179357022, + "grad_norm": 9.1875, + "learning_rate": 2.7983859713990726e-06, + "loss": 2.6356, + "step": 22603 + }, + { + "epoch": 2.390439932318105, + "grad_norm": 8.625, + "learning_rate": 2.7982159386644213e-06, + "loss": 2.3926, + "step": 22604 + }, + { + "epoch": 2.390545685279188, + "grad_norm": 8.8125, + "learning_rate": 2.7980459045303644e-06, + "loss": 2.4694, + "step": 22605 + }, + { + "epoch": 2.390651438240271, + "grad_norm": 8.4375, + "learning_rate": 2.7978758689976994e-06, + "loss": 2.552, + "step": 22606 + }, + { + "epoch": 2.3907571912013537, + "grad_norm": 8.125, + "learning_rate": 2.7977058320672255e-06, + "loss": 2.4682, + "step": 22607 + }, + { + "epoch": 2.3908629441624365, + "grad_norm": 8.0625, + "learning_rate": 2.797535793739739e-06, + "loss": 2.4503, + "step": 22608 + }, + { + "epoch": 2.3909686971235193, + "grad_norm": 8.625, + "learning_rate": 2.79736575401604e-06, + "loss": 2.5569, + "step": 22609 + }, + { + "epoch": 2.391074450084602, + "grad_norm": 8.4375, + "learning_rate": 2.7971957128969234e-06, + "loss": 2.711, + "step": 22610 + }, + { + "epoch": 2.3911802030456855, + "grad_norm": 9.0, + "learning_rate": 2.7970256703831903e-06, + "loss": 2.5737, + "step": 22611 + }, + { + "epoch": 2.3912859560067683, + "grad_norm": 8.125, + "learning_rate": 2.796855626475637e-06, + "loss": 2.5577, + "step": 22612 + }, + { + "epoch": 2.391391708967851, + "grad_norm": 8.25, + "learning_rate": 2.7966855811750613e-06, + "loss": 2.416, + "step": 22613 + }, + { + "epoch": 2.391497461928934, + "grad_norm": 8.3125, + "learning_rate": 2.796515534482262e-06, + "loss": 2.6367, + "step": 22614 + }, + { + "epoch": 2.391603214890017, + "grad_norm": 8.4375, + "learning_rate": 2.796345486398036e-06, + "loss": 2.7345, + "step": 22615 + }, + { + "epoch": 2.3917089678510997, + "grad_norm": 8.8125, + "learning_rate": 2.7961754369231815e-06, + "loss": 2.713, + "step": 22616 + }, + { + "epoch": 2.3918147208121825, + "grad_norm": 8.1875, + "learning_rate": 2.796005386058497e-06, + "loss": 2.553, + "step": 22617 + }, + { + "epoch": 2.3919204737732658, + "grad_norm": 8.4375, + "learning_rate": 2.795835333804781e-06, + "loss": 2.586, + "step": 22618 + }, + { + "epoch": 2.3920262267343486, + "grad_norm": 8.875, + "learning_rate": 2.7956652801628302e-06, + "loss": 2.6477, + "step": 22619 + }, + { + "epoch": 2.3921319796954315, + "grad_norm": 8.125, + "learning_rate": 2.795495225133444e-06, + "loss": 2.6134, + "step": 22620 + }, + { + "epoch": 2.3922377326565143, + "grad_norm": 8.625, + "learning_rate": 2.7953251687174187e-06, + "loss": 2.6957, + "step": 22621 + }, + { + "epoch": 2.392343485617597, + "grad_norm": 8.5625, + "learning_rate": 2.795155110915553e-06, + "loss": 2.7078, + "step": 22622 + }, + { + "epoch": 2.3924492385786804, + "grad_norm": 8.25, + "learning_rate": 2.7949850517286457e-06, + "loss": 2.5847, + "step": 22623 + }, + { + "epoch": 2.3925549915397633, + "grad_norm": 8.6875, + "learning_rate": 2.7948149911574945e-06, + "loss": 2.5443, + "step": 22624 + }, + { + "epoch": 2.392660744500846, + "grad_norm": 8.125, + "learning_rate": 2.7946449292028967e-06, + "loss": 2.7578, + "step": 22625 + }, + { + "epoch": 2.392766497461929, + "grad_norm": 8.5625, + "learning_rate": 2.79447486586565e-06, + "loss": 2.4639, + "step": 22626 + }, + { + "epoch": 2.3928722504230118, + "grad_norm": 8.6875, + "learning_rate": 2.794304801146554e-06, + "loss": 2.5118, + "step": 22627 + }, + { + "epoch": 2.3929780033840946, + "grad_norm": 8.6875, + "learning_rate": 2.794134735046406e-06, + "loss": 2.5805, + "step": 22628 + }, + { + "epoch": 2.3930837563451774, + "grad_norm": 8.6875, + "learning_rate": 2.7939646675660036e-06, + "loss": 2.5811, + "step": 22629 + }, + { + "epoch": 2.3931895093062607, + "grad_norm": 9.1875, + "learning_rate": 2.7937945987061455e-06, + "loss": 2.5984, + "step": 22630 + }, + { + "epoch": 2.3932952622673436, + "grad_norm": 8.3125, + "learning_rate": 2.7936245284676294e-06, + "loss": 2.6302, + "step": 22631 + }, + { + "epoch": 2.3934010152284264, + "grad_norm": 8.25, + "learning_rate": 2.7934544568512533e-06, + "loss": 2.3187, + "step": 22632 + }, + { + "epoch": 2.3935067681895092, + "grad_norm": 8.4375, + "learning_rate": 2.793284383857816e-06, + "loss": 2.3551, + "step": 22633 + }, + { + "epoch": 2.393612521150592, + "grad_norm": 8.3125, + "learning_rate": 2.7931143094881143e-06, + "loss": 2.6654, + "step": 22634 + }, + { + "epoch": 2.3937182741116754, + "grad_norm": 9.1875, + "learning_rate": 2.792944233742947e-06, + "loss": 2.4707, + "step": 22635 + }, + { + "epoch": 2.393824027072758, + "grad_norm": 9.75, + "learning_rate": 2.792774156623113e-06, + "loss": 2.6842, + "step": 22636 + }, + { + "epoch": 2.393929780033841, + "grad_norm": 8.625, + "learning_rate": 2.7926040781294096e-06, + "loss": 2.5131, + "step": 22637 + }, + { + "epoch": 2.394035532994924, + "grad_norm": 8.0, + "learning_rate": 2.792433998262634e-06, + "loss": 2.6162, + "step": 22638 + }, + { + "epoch": 2.3941412859560067, + "grad_norm": 8.4375, + "learning_rate": 2.792263917023586e-06, + "loss": 2.368, + "step": 22639 + }, + { + "epoch": 2.3942470389170896, + "grad_norm": 8.875, + "learning_rate": 2.7920938344130634e-06, + "loss": 2.6144, + "step": 22640 + }, + { + "epoch": 2.3943527918781724, + "grad_norm": 8.75, + "learning_rate": 2.7919237504318626e-06, + "loss": 2.2903, + "step": 22641 + }, + { + "epoch": 2.3944585448392557, + "grad_norm": 8.0625, + "learning_rate": 2.791753665080784e-06, + "loss": 2.553, + "step": 22642 + }, + { + "epoch": 2.3945642978003385, + "grad_norm": 9.1875, + "learning_rate": 2.7915835783606243e-06, + "loss": 2.4779, + "step": 22643 + }, + { + "epoch": 2.3946700507614214, + "grad_norm": 8.375, + "learning_rate": 2.7914134902721823e-06, + "loss": 2.2682, + "step": 22644 + }, + { + "epoch": 2.394775803722504, + "grad_norm": 8.4375, + "learning_rate": 2.791243400816256e-06, + "loss": 2.5466, + "step": 22645 + }, + { + "epoch": 2.394881556683587, + "grad_norm": 8.3125, + "learning_rate": 2.7910733099936437e-06, + "loss": 2.6073, + "step": 22646 + }, + { + "epoch": 2.39498730964467, + "grad_norm": 8.4375, + "learning_rate": 2.7909032178051433e-06, + "loss": 2.6029, + "step": 22647 + }, + { + "epoch": 2.3950930626057527, + "grad_norm": 8.5625, + "learning_rate": 2.790733124251553e-06, + "loss": 2.4187, + "step": 22648 + }, + { + "epoch": 2.395198815566836, + "grad_norm": 8.1875, + "learning_rate": 2.790563029333671e-06, + "loss": 2.4261, + "step": 22649 + }, + { + "epoch": 2.395304568527919, + "grad_norm": 8.0625, + "learning_rate": 2.790392933052296e-06, + "loss": 2.602, + "step": 22650 + }, + { + "epoch": 2.3954103214890017, + "grad_norm": 8.5, + "learning_rate": 2.790222835408225e-06, + "loss": 2.5782, + "step": 22651 + }, + { + "epoch": 2.3955160744500845, + "grad_norm": 8.4375, + "learning_rate": 2.7900527364022566e-06, + "loss": 2.5861, + "step": 22652 + }, + { + "epoch": 2.3956218274111674, + "grad_norm": 8.375, + "learning_rate": 2.7898826360351907e-06, + "loss": 2.2777, + "step": 22653 + }, + { + "epoch": 2.3957275803722506, + "grad_norm": 8.4375, + "learning_rate": 2.789712534307823e-06, + "loss": 2.5494, + "step": 22654 + }, + { + "epoch": 2.3958333333333335, + "grad_norm": 8.3125, + "learning_rate": 2.7895424312209536e-06, + "loss": 2.5156, + "step": 22655 + }, + { + "epoch": 2.3959390862944163, + "grad_norm": 8.125, + "learning_rate": 2.7893723267753796e-06, + "loss": 2.4079, + "step": 22656 + }, + { + "epoch": 2.396044839255499, + "grad_norm": 8.625, + "learning_rate": 2.7892022209718995e-06, + "loss": 2.6031, + "step": 22657 + }, + { + "epoch": 2.396150592216582, + "grad_norm": 8.25, + "learning_rate": 2.7890321138113117e-06, + "loss": 2.5541, + "step": 22658 + }, + { + "epoch": 2.396256345177665, + "grad_norm": 8.4375, + "learning_rate": 2.788862005294415e-06, + "loss": 2.4758, + "step": 22659 + }, + { + "epoch": 2.3963620981387477, + "grad_norm": 8.0, + "learning_rate": 2.7886918954220066e-06, + "loss": 2.5469, + "step": 22660 + }, + { + "epoch": 2.396467851099831, + "grad_norm": 8.0, + "learning_rate": 2.788521784194885e-06, + "loss": 2.509, + "step": 22661 + }, + { + "epoch": 2.396573604060914, + "grad_norm": 8.4375, + "learning_rate": 2.7883516716138486e-06, + "loss": 2.6543, + "step": 22662 + }, + { + "epoch": 2.3966793570219966, + "grad_norm": 7.90625, + "learning_rate": 2.7881815576796967e-06, + "loss": 2.453, + "step": 22663 + }, + { + "epoch": 2.3967851099830795, + "grad_norm": 8.0625, + "learning_rate": 2.7880114423932253e-06, + "loss": 2.5536, + "step": 22664 + }, + { + "epoch": 2.3968908629441623, + "grad_norm": 8.5625, + "learning_rate": 2.787841325755235e-06, + "loss": 2.5346, + "step": 22665 + }, + { + "epoch": 2.396996615905245, + "grad_norm": 8.375, + "learning_rate": 2.787671207766523e-06, + "loss": 2.5952, + "step": 22666 + }, + { + "epoch": 2.3971023688663284, + "grad_norm": 8.125, + "learning_rate": 2.787501088427887e-06, + "loss": 2.4452, + "step": 22667 + }, + { + "epoch": 2.3972081218274113, + "grad_norm": 8.125, + "learning_rate": 2.7873309677401266e-06, + "loss": 2.5598, + "step": 22668 + }, + { + "epoch": 2.397313874788494, + "grad_norm": 8.125, + "learning_rate": 2.787160845704039e-06, + "loss": 2.6135, + "step": 22669 + }, + { + "epoch": 2.397419627749577, + "grad_norm": 8.3125, + "learning_rate": 2.786990722320423e-06, + "loss": 2.6139, + "step": 22670 + }, + { + "epoch": 2.39752538071066, + "grad_norm": 8.4375, + "learning_rate": 2.786820597590077e-06, + "loss": 2.5579, + "step": 22671 + }, + { + "epoch": 2.3976311336717426, + "grad_norm": 8.5, + "learning_rate": 2.7866504715138e-06, + "loss": 2.4508, + "step": 22672 + }, + { + "epoch": 2.397736886632826, + "grad_norm": 8.1875, + "learning_rate": 2.786480344092388e-06, + "loss": 2.5317, + "step": 22673 + }, + { + "epoch": 2.3978426395939088, + "grad_norm": 8.3125, + "learning_rate": 2.7863102153266424e-06, + "loss": 2.4298, + "step": 22674 + }, + { + "epoch": 2.3979483925549916, + "grad_norm": 8.1875, + "learning_rate": 2.7861400852173596e-06, + "loss": 2.5645, + "step": 22675 + }, + { + "epoch": 2.3980541455160744, + "grad_norm": 8.25, + "learning_rate": 2.7859699537653386e-06, + "loss": 2.5304, + "step": 22676 + }, + { + "epoch": 2.3981598984771573, + "grad_norm": 8.5625, + "learning_rate": 2.7857998209713772e-06, + "loss": 2.5049, + "step": 22677 + }, + { + "epoch": 2.39826565143824, + "grad_norm": 8.5625, + "learning_rate": 2.7856296868362743e-06, + "loss": 2.5304, + "step": 22678 + }, + { + "epoch": 2.398371404399323, + "grad_norm": 8.9375, + "learning_rate": 2.7854595513608277e-06, + "loss": 2.7589, + "step": 22679 + }, + { + "epoch": 2.3984771573604062, + "grad_norm": 8.75, + "learning_rate": 2.7852894145458363e-06, + "loss": 2.6454, + "step": 22680 + }, + { + "epoch": 2.398582910321489, + "grad_norm": 9.25, + "learning_rate": 2.785119276392099e-06, + "loss": 2.3217, + "step": 22681 + }, + { + "epoch": 2.398688663282572, + "grad_norm": 8.3125, + "learning_rate": 2.7849491369004135e-06, + "loss": 2.4594, + "step": 22682 + }, + { + "epoch": 2.3987944162436547, + "grad_norm": 8.625, + "learning_rate": 2.7847789960715777e-06, + "loss": 2.7483, + "step": 22683 + }, + { + "epoch": 2.3989001692047376, + "grad_norm": 9.0, + "learning_rate": 2.7846088539063903e-06, + "loss": 2.5888, + "step": 22684 + }, + { + "epoch": 2.399005922165821, + "grad_norm": 10.125, + "learning_rate": 2.784438710405651e-06, + "loss": 2.8178, + "step": 22685 + }, + { + "epoch": 2.3991116751269037, + "grad_norm": 14.3125, + "learning_rate": 2.784268565570156e-06, + "loss": 2.5401, + "step": 22686 + }, + { + "epoch": 2.3992174280879865, + "grad_norm": 8.1875, + "learning_rate": 2.7840984194007055e-06, + "loss": 2.6077, + "step": 22687 + }, + { + "epoch": 2.3993231810490694, + "grad_norm": 8.25, + "learning_rate": 2.783928271898097e-06, + "loss": 2.6397, + "step": 22688 + }, + { + "epoch": 2.3994289340101522, + "grad_norm": 8.9375, + "learning_rate": 2.7837581230631295e-06, + "loss": 2.8614, + "step": 22689 + }, + { + "epoch": 2.399534686971235, + "grad_norm": 8.875, + "learning_rate": 2.783587972896601e-06, + "loss": 2.6514, + "step": 22690 + }, + { + "epoch": 2.399640439932318, + "grad_norm": 8.5, + "learning_rate": 2.783417821399311e-06, + "loss": 2.4821, + "step": 22691 + }, + { + "epoch": 2.399746192893401, + "grad_norm": 8.375, + "learning_rate": 2.7832476685720556e-06, + "loss": 2.5179, + "step": 22692 + }, + { + "epoch": 2.399851945854484, + "grad_norm": 11.0, + "learning_rate": 2.7830775144156357e-06, + "loss": 2.6448, + "step": 22693 + }, + { + "epoch": 2.399957698815567, + "grad_norm": 8.1875, + "learning_rate": 2.7829073589308484e-06, + "loss": 2.4421, + "step": 22694 + }, + { + "epoch": 2.4000634517766497, + "grad_norm": 8.1875, + "learning_rate": 2.782737202118493e-06, + "loss": 2.348, + "step": 22695 + }, + { + "epoch": 2.4001692047377325, + "grad_norm": 8.5625, + "learning_rate": 2.782567043979367e-06, + "loss": 2.6234, + "step": 22696 + }, + { + "epoch": 2.4002749576988154, + "grad_norm": 8.3125, + "learning_rate": 2.7823968845142697e-06, + "loss": 2.5824, + "step": 22697 + }, + { + "epoch": 2.4003807106598987, + "grad_norm": 9.25, + "learning_rate": 2.7822267237239996e-06, + "loss": 2.6204, + "step": 22698 + }, + { + "epoch": 2.4004864636209815, + "grad_norm": 8.0625, + "learning_rate": 2.782056561609354e-06, + "loss": 2.6446, + "step": 22699 + }, + { + "epoch": 2.4005922165820643, + "grad_norm": 8.3125, + "learning_rate": 2.7818863981711337e-06, + "loss": 2.43, + "step": 22700 + }, + { + "epoch": 2.400697969543147, + "grad_norm": 8.6875, + "learning_rate": 2.781716233410135e-06, + "loss": 2.6333, + "step": 22701 + }, + { + "epoch": 2.40080372250423, + "grad_norm": 8.25, + "learning_rate": 2.781546067327157e-06, + "loss": 2.5869, + "step": 22702 + }, + { + "epoch": 2.400909475465313, + "grad_norm": 8.1875, + "learning_rate": 2.7813758999229988e-06, + "loss": 2.5794, + "step": 22703 + }, + { + "epoch": 2.401015228426396, + "grad_norm": 8.4375, + "learning_rate": 2.7812057311984593e-06, + "loss": 2.4965, + "step": 22704 + }, + { + "epoch": 2.401120981387479, + "grad_norm": 8.625, + "learning_rate": 2.7810355611543343e-06, + "loss": 2.3902, + "step": 22705 + }, + { + "epoch": 2.401226734348562, + "grad_norm": 8.6875, + "learning_rate": 2.780865389791426e-06, + "loss": 2.4649, + "step": 22706 + }, + { + "epoch": 2.4013324873096447, + "grad_norm": 8.625, + "learning_rate": 2.7806952171105313e-06, + "loss": 2.8235, + "step": 22707 + }, + { + "epoch": 2.4014382402707275, + "grad_norm": 8.0, + "learning_rate": 2.780525043112448e-06, + "loss": 2.4385, + "step": 22708 + }, + { + "epoch": 2.4015439932318103, + "grad_norm": 8.5625, + "learning_rate": 2.7803548677979757e-06, + "loss": 2.5548, + "step": 22709 + }, + { + "epoch": 2.401649746192893, + "grad_norm": 8.9375, + "learning_rate": 2.780184691167913e-06, + "loss": 2.5686, + "step": 22710 + }, + { + "epoch": 2.4017554991539765, + "grad_norm": 8.875, + "learning_rate": 2.7800145132230577e-06, + "loss": 2.5394, + "step": 22711 + }, + { + "epoch": 2.4018612521150593, + "grad_norm": 8.75, + "learning_rate": 2.7798443339642088e-06, + "loss": 2.5757, + "step": 22712 + }, + { + "epoch": 2.401967005076142, + "grad_norm": 8.125, + "learning_rate": 2.7796741533921657e-06, + "loss": 2.7297, + "step": 22713 + }, + { + "epoch": 2.402072758037225, + "grad_norm": 8.375, + "learning_rate": 2.779503971507725e-06, + "loss": 2.4136, + "step": 22714 + }, + { + "epoch": 2.402178510998308, + "grad_norm": 8.5625, + "learning_rate": 2.7793337883116865e-06, + "loss": 2.4481, + "step": 22715 + }, + { + "epoch": 2.402284263959391, + "grad_norm": 8.8125, + "learning_rate": 2.7791636038048497e-06, + "loss": 2.3348, + "step": 22716 + }, + { + "epoch": 2.402390016920474, + "grad_norm": 8.6875, + "learning_rate": 2.7789934179880124e-06, + "loss": 2.7453, + "step": 22717 + }, + { + "epoch": 2.4024957698815568, + "grad_norm": 8.25, + "learning_rate": 2.7788232308619722e-06, + "loss": 2.5327, + "step": 22718 + }, + { + "epoch": 2.4026015228426396, + "grad_norm": 8.75, + "learning_rate": 2.778653042427529e-06, + "loss": 2.7586, + "step": 22719 + }, + { + "epoch": 2.4027072758037225, + "grad_norm": 8.8125, + "learning_rate": 2.778482852685481e-06, + "loss": 2.5693, + "step": 22720 + }, + { + "epoch": 2.4028130287648053, + "grad_norm": 8.3125, + "learning_rate": 2.778312661636627e-06, + "loss": 2.4665, + "step": 22721 + }, + { + "epoch": 2.402918781725888, + "grad_norm": 8.4375, + "learning_rate": 2.7781424692817654e-06, + "loss": 2.4121, + "step": 22722 + }, + { + "epoch": 2.4030245346869714, + "grad_norm": 8.4375, + "learning_rate": 2.777972275621695e-06, + "loss": 2.4778, + "step": 22723 + }, + { + "epoch": 2.4031302876480543, + "grad_norm": 8.25, + "learning_rate": 2.7778020806572135e-06, + "loss": 2.6002, + "step": 22724 + }, + { + "epoch": 2.403236040609137, + "grad_norm": 8.625, + "learning_rate": 2.777631884389122e-06, + "loss": 2.467, + "step": 22725 + }, + { + "epoch": 2.40334179357022, + "grad_norm": 8.125, + "learning_rate": 2.777461686818217e-06, + "loss": 2.6192, + "step": 22726 + }, + { + "epoch": 2.4034475465313028, + "grad_norm": 8.375, + "learning_rate": 2.777291487945297e-06, + "loss": 2.4434, + "step": 22727 + }, + { + "epoch": 2.4035532994923856, + "grad_norm": 8.5, + "learning_rate": 2.777121287771163e-06, + "loss": 2.6641, + "step": 22728 + }, + { + "epoch": 2.4036590524534684, + "grad_norm": 8.1875, + "learning_rate": 2.7769510862966115e-06, + "loss": 2.5819, + "step": 22729 + }, + { + "epoch": 2.4037648054145517, + "grad_norm": 8.8125, + "learning_rate": 2.7767808835224406e-06, + "loss": 2.6776, + "step": 22730 + }, + { + "epoch": 2.4038705583756346, + "grad_norm": 7.96875, + "learning_rate": 2.7766106794494515e-06, + "loss": 2.4955, + "step": 22731 + }, + { + "epoch": 2.4039763113367174, + "grad_norm": 8.4375, + "learning_rate": 2.776440474078441e-06, + "loss": 2.2551, + "step": 22732 + }, + { + "epoch": 2.4040820642978002, + "grad_norm": 8.125, + "learning_rate": 2.7762702674102094e-06, + "loss": 2.5934, + "step": 22733 + }, + { + "epoch": 2.404187817258883, + "grad_norm": 8.1875, + "learning_rate": 2.776100059445553e-06, + "loss": 2.4139, + "step": 22734 + }, + { + "epoch": 2.4042935702199664, + "grad_norm": 8.625, + "learning_rate": 2.7759298501852733e-06, + "loss": 2.6258, + "step": 22735 + }, + { + "epoch": 2.404399323181049, + "grad_norm": 8.375, + "learning_rate": 2.775759639630167e-06, + "loss": 2.4614, + "step": 22736 + }, + { + "epoch": 2.404505076142132, + "grad_norm": 8.5, + "learning_rate": 2.7755894277810334e-06, + "loss": 2.4206, + "step": 22737 + }, + { + "epoch": 2.404610829103215, + "grad_norm": 8.0625, + "learning_rate": 2.7754192146386715e-06, + "loss": 2.5228, + "step": 22738 + }, + { + "epoch": 2.4047165820642977, + "grad_norm": 8.4375, + "learning_rate": 2.775249000203881e-06, + "loss": 2.6057, + "step": 22739 + }, + { + "epoch": 2.4048223350253806, + "grad_norm": 8.25, + "learning_rate": 2.775078784477457e-06, + "loss": 2.6447, + "step": 22740 + }, + { + "epoch": 2.4049280879864634, + "grad_norm": 8.4375, + "learning_rate": 2.774908567460202e-06, + "loss": 2.4671, + "step": 22741 + }, + { + "epoch": 2.4050338409475467, + "grad_norm": 8.5625, + "learning_rate": 2.7747383491529144e-06, + "loss": 2.5923, + "step": 22742 + }, + { + "epoch": 2.4051395939086295, + "grad_norm": 8.8125, + "learning_rate": 2.774568129556391e-06, + "loss": 2.6614, + "step": 22743 + }, + { + "epoch": 2.4052453468697124, + "grad_norm": 8.3125, + "learning_rate": 2.7743979086714325e-06, + "loss": 2.4459, + "step": 22744 + }, + { + "epoch": 2.405351099830795, + "grad_norm": 8.75, + "learning_rate": 2.774227686498836e-06, + "loss": 2.4089, + "step": 22745 + }, + { + "epoch": 2.405456852791878, + "grad_norm": 8.5, + "learning_rate": 2.7740574630394012e-06, + "loss": 2.5813, + "step": 22746 + }, + { + "epoch": 2.4055626057529613, + "grad_norm": 8.75, + "learning_rate": 2.7738872382939275e-06, + "loss": 2.4644, + "step": 22747 + }, + { + "epoch": 2.405668358714044, + "grad_norm": 8.3125, + "learning_rate": 2.773717012263213e-06, + "loss": 2.427, + "step": 22748 + }, + { + "epoch": 2.405774111675127, + "grad_norm": 7.875, + "learning_rate": 2.7735467849480557e-06, + "loss": 2.5704, + "step": 22749 + }, + { + "epoch": 2.40587986463621, + "grad_norm": 8.4375, + "learning_rate": 2.7733765563492555e-06, + "loss": 2.42, + "step": 22750 + }, + { + "epoch": 2.4059856175972927, + "grad_norm": 7.875, + "learning_rate": 2.773206326467611e-06, + "loss": 2.4211, + "step": 22751 + }, + { + "epoch": 2.4060913705583755, + "grad_norm": 7.96875, + "learning_rate": 2.7730360953039215e-06, + "loss": 2.5382, + "step": 22752 + }, + { + "epoch": 2.4061971235194584, + "grad_norm": 8.4375, + "learning_rate": 2.7728658628589844e-06, + "loss": 2.6056, + "step": 22753 + }, + { + "epoch": 2.4063028764805416, + "grad_norm": 8.6875, + "learning_rate": 2.7726956291336e-06, + "loss": 2.3621, + "step": 22754 + }, + { + "epoch": 2.4064086294416245, + "grad_norm": 8.125, + "learning_rate": 2.772525394128566e-06, + "loss": 2.4721, + "step": 22755 + }, + { + "epoch": 2.4065143824027073, + "grad_norm": 8.3125, + "learning_rate": 2.772355157844682e-06, + "loss": 2.5615, + "step": 22756 + }, + { + "epoch": 2.40662013536379, + "grad_norm": 8.9375, + "learning_rate": 2.772184920282747e-06, + "loss": 2.6525, + "step": 22757 + }, + { + "epoch": 2.406725888324873, + "grad_norm": 9.25, + "learning_rate": 2.772014681443559e-06, + "loss": 2.5687, + "step": 22758 + }, + { + "epoch": 2.406831641285956, + "grad_norm": 8.3125, + "learning_rate": 2.771844441327917e-06, + "loss": 2.5656, + "step": 22759 + }, + { + "epoch": 2.4069373942470387, + "grad_norm": 8.5, + "learning_rate": 2.7716741999366213e-06, + "loss": 2.4614, + "step": 22760 + }, + { + "epoch": 2.407043147208122, + "grad_norm": 9.1875, + "learning_rate": 2.7715039572704694e-06, + "loss": 2.5922, + "step": 22761 + }, + { + "epoch": 2.407148900169205, + "grad_norm": 9.5, + "learning_rate": 2.7713337133302597e-06, + "loss": 2.5846, + "step": 22762 + }, + { + "epoch": 2.4072546531302876, + "grad_norm": 9.1875, + "learning_rate": 2.771163468116793e-06, + "loss": 2.629, + "step": 22763 + }, + { + "epoch": 2.4073604060913705, + "grad_norm": 8.4375, + "learning_rate": 2.7709932216308664e-06, + "loss": 2.2388, + "step": 22764 + }, + { + "epoch": 2.4074661590524533, + "grad_norm": 8.125, + "learning_rate": 2.7708229738732796e-06, + "loss": 2.5253, + "step": 22765 + }, + { + "epoch": 2.4075719120135366, + "grad_norm": 8.25, + "learning_rate": 2.770652724844831e-06, + "loss": 2.6325, + "step": 22766 + }, + { + "epoch": 2.4076776649746194, + "grad_norm": 8.25, + "learning_rate": 2.77048247454632e-06, + "loss": 2.5059, + "step": 22767 + }, + { + "epoch": 2.4077834179357023, + "grad_norm": 9.1875, + "learning_rate": 2.7703122229785454e-06, + "loss": 2.8089, + "step": 22768 + }, + { + "epoch": 2.407889170896785, + "grad_norm": 8.0, + "learning_rate": 2.770141970142306e-06, + "loss": 2.5904, + "step": 22769 + }, + { + "epoch": 2.407994923857868, + "grad_norm": 8.25, + "learning_rate": 2.7699717160384016e-06, + "loss": 2.3419, + "step": 22770 + }, + { + "epoch": 2.408100676818951, + "grad_norm": 8.375, + "learning_rate": 2.76980146066763e-06, + "loss": 2.4387, + "step": 22771 + }, + { + "epoch": 2.4082064297800336, + "grad_norm": 8.75, + "learning_rate": 2.76963120403079e-06, + "loss": 2.6821, + "step": 22772 + }, + { + "epoch": 2.408312182741117, + "grad_norm": 7.96875, + "learning_rate": 2.7694609461286807e-06, + "loss": 2.4621, + "step": 22773 + }, + { + "epoch": 2.4084179357021998, + "grad_norm": 8.4375, + "learning_rate": 2.769290686962103e-06, + "loss": 2.558, + "step": 22774 + }, + { + "epoch": 2.4085236886632826, + "grad_norm": 9.5625, + "learning_rate": 2.769120426531853e-06, + "loss": 2.6407, + "step": 22775 + }, + { + "epoch": 2.4086294416243654, + "grad_norm": 8.1875, + "learning_rate": 2.76895016483873e-06, + "loss": 2.257, + "step": 22776 + }, + { + "epoch": 2.4087351945854483, + "grad_norm": 8.25, + "learning_rate": 2.7687799018835353e-06, + "loss": 2.6232, + "step": 22777 + }, + { + "epoch": 2.4088409475465316, + "grad_norm": 8.625, + "learning_rate": 2.7686096376670657e-06, + "loss": 2.5915, + "step": 22778 + }, + { + "epoch": 2.4089467005076144, + "grad_norm": 8.875, + "learning_rate": 2.768439372190122e-06, + "loss": 2.6213, + "step": 22779 + }, + { + "epoch": 2.4090524534686972, + "grad_norm": 8.4375, + "learning_rate": 2.7682691054535017e-06, + "loss": 2.4598, + "step": 22780 + }, + { + "epoch": 2.40915820642978, + "grad_norm": 8.375, + "learning_rate": 2.7680988374580036e-06, + "loss": 2.5097, + "step": 22781 + }, + { + "epoch": 2.409263959390863, + "grad_norm": 8.3125, + "learning_rate": 2.7679285682044277e-06, + "loss": 2.4266, + "step": 22782 + }, + { + "epoch": 2.4093697123519457, + "grad_norm": 8.625, + "learning_rate": 2.767758297693573e-06, + "loss": 2.5531, + "step": 22783 + }, + { + "epoch": 2.4094754653130286, + "grad_norm": 8.5625, + "learning_rate": 2.7675880259262374e-06, + "loss": 2.4927, + "step": 22784 + }, + { + "epoch": 2.409581218274112, + "grad_norm": 8.3125, + "learning_rate": 2.7674177529032204e-06, + "loss": 2.461, + "step": 22785 + }, + { + "epoch": 2.4096869712351947, + "grad_norm": 8.4375, + "learning_rate": 2.767247478625322e-06, + "loss": 2.6367, + "step": 22786 + }, + { + "epoch": 2.4097927241962775, + "grad_norm": 8.125, + "learning_rate": 2.7670772030933406e-06, + "loss": 2.6337, + "step": 22787 + }, + { + "epoch": 2.4098984771573604, + "grad_norm": 9.125, + "learning_rate": 2.7669069263080744e-06, + "loss": 2.7035, + "step": 22788 + }, + { + "epoch": 2.4100042301184432, + "grad_norm": 8.3125, + "learning_rate": 2.766736648270324e-06, + "loss": 2.575, + "step": 22789 + }, + { + "epoch": 2.410109983079526, + "grad_norm": 8.75, + "learning_rate": 2.766566368980887e-06, + "loss": 2.5773, + "step": 22790 + }, + { + "epoch": 2.410215736040609, + "grad_norm": 8.6875, + "learning_rate": 2.7663960884405627e-06, + "loss": 2.4854, + "step": 22791 + }, + { + "epoch": 2.410321489001692, + "grad_norm": 8.625, + "learning_rate": 2.7662258066501513e-06, + "loss": 2.7864, + "step": 22792 + }, + { + "epoch": 2.410427241962775, + "grad_norm": 8.375, + "learning_rate": 2.766055523610451e-06, + "loss": 2.322, + "step": 22793 + }, + { + "epoch": 2.410532994923858, + "grad_norm": 8.9375, + "learning_rate": 2.765885239322261e-06, + "loss": 2.5059, + "step": 22794 + }, + { + "epoch": 2.4106387478849407, + "grad_norm": 8.625, + "learning_rate": 2.765714953786379e-06, + "loss": 2.7271, + "step": 22795 + }, + { + "epoch": 2.4107445008460235, + "grad_norm": 8.875, + "learning_rate": 2.7655446670036066e-06, + "loss": 2.6871, + "step": 22796 + }, + { + "epoch": 2.410850253807107, + "grad_norm": 8.375, + "learning_rate": 2.7653743789747416e-06, + "loss": 2.6167, + "step": 22797 + }, + { + "epoch": 2.4109560067681897, + "grad_norm": 8.1875, + "learning_rate": 2.765204089700583e-06, + "loss": 2.4851, + "step": 22798 + }, + { + "epoch": 2.4110617597292725, + "grad_norm": 9.25, + "learning_rate": 2.7650337991819305e-06, + "loss": 2.4832, + "step": 22799 + }, + { + "epoch": 2.4111675126903553, + "grad_norm": 8.5625, + "learning_rate": 2.7648635074195823e-06, + "loss": 2.4298, + "step": 22800 + }, + { + "epoch": 2.411273265651438, + "grad_norm": 8.625, + "learning_rate": 2.7646932144143386e-06, + "loss": 2.5566, + "step": 22801 + }, + { + "epoch": 2.411379018612521, + "grad_norm": 8.1875, + "learning_rate": 2.7645229201669964e-06, + "loss": 2.5907, + "step": 22802 + }, + { + "epoch": 2.411484771573604, + "grad_norm": 8.4375, + "learning_rate": 2.764352624678358e-06, + "loss": 2.7459, + "step": 22803 + }, + { + "epoch": 2.411590524534687, + "grad_norm": 7.96875, + "learning_rate": 2.7641823279492195e-06, + "loss": 2.5413, + "step": 22804 + }, + { + "epoch": 2.41169627749577, + "grad_norm": 9.125, + "learning_rate": 2.7640120299803822e-06, + "loss": 2.2691, + "step": 22805 + }, + { + "epoch": 2.411802030456853, + "grad_norm": 8.3125, + "learning_rate": 2.763841730772645e-06, + "loss": 2.5092, + "step": 22806 + }, + { + "epoch": 2.4119077834179357, + "grad_norm": 8.5, + "learning_rate": 2.7636714303268055e-06, + "loss": 2.4642, + "step": 22807 + }, + { + "epoch": 2.4120135363790185, + "grad_norm": 8.125, + "learning_rate": 2.7635011286436643e-06, + "loss": 2.4741, + "step": 22808 + }, + { + "epoch": 2.4121192893401013, + "grad_norm": 8.625, + "learning_rate": 2.76333082572402e-06, + "loss": 2.5662, + "step": 22809 + }, + { + "epoch": 2.4122250423011846, + "grad_norm": 8.25, + "learning_rate": 2.763160521568671e-06, + "loss": 2.4893, + "step": 22810 + }, + { + "epoch": 2.4123307952622675, + "grad_norm": 8.75, + "learning_rate": 2.762990216178418e-06, + "loss": 2.4969, + "step": 22811 + }, + { + "epoch": 2.4124365482233503, + "grad_norm": 8.25, + "learning_rate": 2.7628199095540597e-06, + "loss": 2.6276, + "step": 22812 + }, + { + "epoch": 2.412542301184433, + "grad_norm": 8.4375, + "learning_rate": 2.7626496016963943e-06, + "loss": 2.5373, + "step": 22813 + }, + { + "epoch": 2.412648054145516, + "grad_norm": 8.125, + "learning_rate": 2.762479292606223e-06, + "loss": 2.518, + "step": 22814 + }, + { + "epoch": 2.412753807106599, + "grad_norm": 8.4375, + "learning_rate": 2.762308982284343e-06, + "loss": 2.8431, + "step": 22815 + }, + { + "epoch": 2.412859560067682, + "grad_norm": 8.8125, + "learning_rate": 2.7621386707315532e-06, + "loss": 2.5824, + "step": 22816 + }, + { + "epoch": 2.412965313028765, + "grad_norm": 8.375, + "learning_rate": 2.7619683579486556e-06, + "loss": 2.4749, + "step": 22817 + }, + { + "epoch": 2.4130710659898478, + "grad_norm": 8.5, + "learning_rate": 2.7617980439364468e-06, + "loss": 2.5087, + "step": 22818 + }, + { + "epoch": 2.4131768189509306, + "grad_norm": 8.125, + "learning_rate": 2.7616277286957268e-06, + "loss": 2.4116, + "step": 22819 + }, + { + "epoch": 2.4132825719120135, + "grad_norm": 8.5, + "learning_rate": 2.7614574122272948e-06, + "loss": 2.6729, + "step": 22820 + }, + { + "epoch": 2.4133883248730963, + "grad_norm": 8.8125, + "learning_rate": 2.7612870945319504e-06, + "loss": 2.6039, + "step": 22821 + }, + { + "epoch": 2.413494077834179, + "grad_norm": 9.125, + "learning_rate": 2.7611167756104927e-06, + "loss": 2.5496, + "step": 22822 + }, + { + "epoch": 2.4135998307952624, + "grad_norm": 8.0, + "learning_rate": 2.76094645546372e-06, + "loss": 2.4134, + "step": 22823 + }, + { + "epoch": 2.4137055837563453, + "grad_norm": 7.9375, + "learning_rate": 2.7607761340924332e-06, + "loss": 2.4999, + "step": 22824 + }, + { + "epoch": 2.413811336717428, + "grad_norm": 8.8125, + "learning_rate": 2.7606058114974306e-06, + "loss": 2.3834, + "step": 22825 + }, + { + "epoch": 2.413917089678511, + "grad_norm": 7.875, + "learning_rate": 2.76043548767951e-06, + "loss": 2.5117, + "step": 22826 + }, + { + "epoch": 2.4140228426395938, + "grad_norm": 8.125, + "learning_rate": 2.760265162639474e-06, + "loss": 2.7414, + "step": 22827 + }, + { + "epoch": 2.414128595600677, + "grad_norm": 8.875, + "learning_rate": 2.76009483637812e-06, + "loss": 2.8451, + "step": 22828 + }, + { + "epoch": 2.41423434856176, + "grad_norm": 8.25, + "learning_rate": 2.759924508896246e-06, + "loss": 2.5964, + "step": 22829 + }, + { + "epoch": 2.4143401015228427, + "grad_norm": 8.375, + "learning_rate": 2.7597541801946533e-06, + "loss": 2.3672, + "step": 22830 + }, + { + "epoch": 2.4144458544839256, + "grad_norm": 8.625, + "learning_rate": 2.75958385027414e-06, + "loss": 2.5321, + "step": 22831 + }, + { + "epoch": 2.4145516074450084, + "grad_norm": 8.5, + "learning_rate": 2.7594135191355064e-06, + "loss": 2.467, + "step": 22832 + }, + { + "epoch": 2.4146573604060912, + "grad_norm": 7.96875, + "learning_rate": 2.759243186779551e-06, + "loss": 2.5436, + "step": 22833 + }, + { + "epoch": 2.414763113367174, + "grad_norm": 8.0625, + "learning_rate": 2.759072853207074e-06, + "loss": 2.694, + "step": 22834 + }, + { + "epoch": 2.4148688663282574, + "grad_norm": 8.5, + "learning_rate": 2.7589025184188735e-06, + "loss": 2.4643, + "step": 22835 + }, + { + "epoch": 2.41497461928934, + "grad_norm": 8.4375, + "learning_rate": 2.758732182415749e-06, + "loss": 2.5666, + "step": 22836 + }, + { + "epoch": 2.415080372250423, + "grad_norm": 8.375, + "learning_rate": 2.7585618451985015e-06, + "loss": 2.5593, + "step": 22837 + }, + { + "epoch": 2.415186125211506, + "grad_norm": 8.5625, + "learning_rate": 2.7583915067679273e-06, + "loss": 2.4762, + "step": 22838 + }, + { + "epoch": 2.4152918781725887, + "grad_norm": 8.1875, + "learning_rate": 2.7582211671248284e-06, + "loss": 2.4412, + "step": 22839 + }, + { + "epoch": 2.4153976311336716, + "grad_norm": 8.1875, + "learning_rate": 2.758050826270003e-06, + "loss": 2.4374, + "step": 22840 + }, + { + "epoch": 2.415503384094755, + "grad_norm": 8.5625, + "learning_rate": 2.7578804842042504e-06, + "loss": 2.5795, + "step": 22841 + }, + { + "epoch": 2.4156091370558377, + "grad_norm": 7.71875, + "learning_rate": 2.7577101409283704e-06, + "loss": 2.561, + "step": 22842 + }, + { + "epoch": 2.4157148900169205, + "grad_norm": 8.3125, + "learning_rate": 2.7575397964431624e-06, + "loss": 2.5215, + "step": 22843 + }, + { + "epoch": 2.4158206429780034, + "grad_norm": 9.0625, + "learning_rate": 2.757369450749425e-06, + "loss": 2.8481, + "step": 22844 + }, + { + "epoch": 2.415926395939086, + "grad_norm": 8.625, + "learning_rate": 2.7571991038479586e-06, + "loss": 2.7668, + "step": 22845 + }, + { + "epoch": 2.416032148900169, + "grad_norm": 7.875, + "learning_rate": 2.757028755739561e-06, + "loss": 2.5642, + "step": 22846 + }, + { + "epoch": 2.4161379018612523, + "grad_norm": 8.5, + "learning_rate": 2.756858406425033e-06, + "loss": 2.1733, + "step": 22847 + }, + { + "epoch": 2.416243654822335, + "grad_norm": 8.25, + "learning_rate": 2.756688055905174e-06, + "loss": 2.3716, + "step": 22848 + }, + { + "epoch": 2.416349407783418, + "grad_norm": 8.125, + "learning_rate": 2.7565177041807823e-06, + "loss": 2.4391, + "step": 22849 + }, + { + "epoch": 2.416455160744501, + "grad_norm": 8.0625, + "learning_rate": 2.7563473512526586e-06, + "loss": 2.4659, + "step": 22850 + }, + { + "epoch": 2.4165609137055837, + "grad_norm": 8.0625, + "learning_rate": 2.756176997121601e-06, + "loss": 2.5595, + "step": 22851 + }, + { + "epoch": 2.4166666666666665, + "grad_norm": 8.1875, + "learning_rate": 2.75600664178841e-06, + "loss": 2.5722, + "step": 22852 + }, + { + "epoch": 2.4167724196277494, + "grad_norm": 7.8125, + "learning_rate": 2.7558362852538843e-06, + "loss": 2.4628, + "step": 22853 + }, + { + "epoch": 2.4168781725888326, + "grad_norm": 8.625, + "learning_rate": 2.7556659275188237e-06, + "loss": 2.5623, + "step": 22854 + }, + { + "epoch": 2.4169839255499155, + "grad_norm": 9.25, + "learning_rate": 2.755495568584027e-06, + "loss": 2.4212, + "step": 22855 + }, + { + "epoch": 2.4170896785109983, + "grad_norm": 8.375, + "learning_rate": 2.7553252084502947e-06, + "loss": 2.6946, + "step": 22856 + }, + { + "epoch": 2.417195431472081, + "grad_norm": 8.875, + "learning_rate": 2.7551548471184255e-06, + "loss": 2.6624, + "step": 22857 + }, + { + "epoch": 2.417301184433164, + "grad_norm": 8.375, + "learning_rate": 2.754984484589218e-06, + "loss": 2.4061, + "step": 22858 + }, + { + "epoch": 2.4174069373942473, + "grad_norm": 8.5625, + "learning_rate": 2.754814120863474e-06, + "loss": 2.5141, + "step": 22859 + }, + { + "epoch": 2.41751269035533, + "grad_norm": 8.375, + "learning_rate": 2.7546437559419916e-06, + "loss": 2.5273, + "step": 22860 + }, + { + "epoch": 2.417618443316413, + "grad_norm": 8.125, + "learning_rate": 2.7544733898255687e-06, + "loss": 2.6213, + "step": 22861 + }, + { + "epoch": 2.417724196277496, + "grad_norm": 8.3125, + "learning_rate": 2.7543030225150077e-06, + "loss": 2.5335, + "step": 22862 + }, + { + "epoch": 2.4178299492385786, + "grad_norm": 8.125, + "learning_rate": 2.7541326540111064e-06, + "loss": 2.4353, + "step": 22863 + }, + { + "epoch": 2.4179357021996615, + "grad_norm": 8.0625, + "learning_rate": 2.7539622843146636e-06, + "loss": 2.5016, + "step": 22864 + }, + { + "epoch": 2.4180414551607443, + "grad_norm": 9.0625, + "learning_rate": 2.7537919134264802e-06, + "loss": 2.5089, + "step": 22865 + }, + { + "epoch": 2.4181472081218276, + "grad_norm": 8.625, + "learning_rate": 2.753621541347355e-06, + "loss": 2.6576, + "step": 22866 + }, + { + "epoch": 2.4182529610829104, + "grad_norm": 8.625, + "learning_rate": 2.753451168078088e-06, + "loss": 2.5236, + "step": 22867 + }, + { + "epoch": 2.4183587140439933, + "grad_norm": 8.5625, + "learning_rate": 2.7532807936194776e-06, + "loss": 2.2935, + "step": 22868 + }, + { + "epoch": 2.418464467005076, + "grad_norm": 9.25, + "learning_rate": 2.753110417972325e-06, + "loss": 2.2571, + "step": 22869 + }, + { + "epoch": 2.418570219966159, + "grad_norm": 8.5625, + "learning_rate": 2.752940041137428e-06, + "loss": 2.6706, + "step": 22870 + }, + { + "epoch": 2.418675972927242, + "grad_norm": 8.3125, + "learning_rate": 2.7527696631155877e-06, + "loss": 2.5457, + "step": 22871 + }, + { + "epoch": 2.4187817258883246, + "grad_norm": 10.5625, + "learning_rate": 2.7525992839076023e-06, + "loss": 2.2697, + "step": 22872 + }, + { + "epoch": 2.418887478849408, + "grad_norm": 8.875, + "learning_rate": 2.752428903514271e-06, + "loss": 2.5977, + "step": 22873 + }, + { + "epoch": 2.4189932318104908, + "grad_norm": 8.5625, + "learning_rate": 2.7522585219363946e-06, + "loss": 2.7163, + "step": 22874 + }, + { + "epoch": 2.4190989847715736, + "grad_norm": 7.9375, + "learning_rate": 2.7520881391747723e-06, + "loss": 2.6076, + "step": 22875 + }, + { + "epoch": 2.4192047377326564, + "grad_norm": 8.375, + "learning_rate": 2.7519177552302036e-06, + "loss": 2.5091, + "step": 22876 + }, + { + "epoch": 2.4193104906937393, + "grad_norm": 8.3125, + "learning_rate": 2.7517473701034875e-06, + "loss": 2.3675, + "step": 22877 + }, + { + "epoch": 2.4194162436548226, + "grad_norm": 8.125, + "learning_rate": 2.751576983795424e-06, + "loss": 2.2804, + "step": 22878 + }, + { + "epoch": 2.4195219966159054, + "grad_norm": 8.1875, + "learning_rate": 2.7514065963068134e-06, + "loss": 2.3944, + "step": 22879 + }, + { + "epoch": 2.4196277495769882, + "grad_norm": 7.875, + "learning_rate": 2.7512362076384534e-06, + "loss": 2.4537, + "step": 22880 + }, + { + "epoch": 2.419733502538071, + "grad_norm": 8.625, + "learning_rate": 2.7510658177911455e-06, + "loss": 2.7169, + "step": 22881 + }, + { + "epoch": 2.419839255499154, + "grad_norm": 8.875, + "learning_rate": 2.7508954267656872e-06, + "loss": 2.7231, + "step": 22882 + }, + { + "epoch": 2.4199450084602367, + "grad_norm": 8.3125, + "learning_rate": 2.7507250345628807e-06, + "loss": 2.6156, + "step": 22883 + }, + { + "epoch": 2.4200507614213196, + "grad_norm": 7.875, + "learning_rate": 2.7505546411835233e-06, + "loss": 2.3764, + "step": 22884 + }, + { + "epoch": 2.420156514382403, + "grad_norm": 8.3125, + "learning_rate": 2.750384246628416e-06, + "loss": 2.4919, + "step": 22885 + }, + { + "epoch": 2.4202622673434857, + "grad_norm": 8.375, + "learning_rate": 2.750213850898357e-06, + "loss": 2.6895, + "step": 22886 + }, + { + "epoch": 2.4203680203045685, + "grad_norm": 8.25, + "learning_rate": 2.7500434539941472e-06, + "loss": 2.6886, + "step": 22887 + }, + { + "epoch": 2.4204737732656514, + "grad_norm": 8.3125, + "learning_rate": 2.749873055916586e-06, + "loss": 2.6206, + "step": 22888 + }, + { + "epoch": 2.4205795262267342, + "grad_norm": 9.125, + "learning_rate": 2.749702656666472e-06, + "loss": 2.8886, + "step": 22889 + }, + { + "epoch": 2.4206852791878175, + "grad_norm": 8.0625, + "learning_rate": 2.7495322562446064e-06, + "loss": 2.5658, + "step": 22890 + }, + { + "epoch": 2.4207910321489003, + "grad_norm": 8.25, + "learning_rate": 2.7493618546517874e-06, + "loss": 2.4741, + "step": 22891 + }, + { + "epoch": 2.420896785109983, + "grad_norm": 8.4375, + "learning_rate": 2.7491914518888157e-06, + "loss": 2.6474, + "step": 22892 + }, + { + "epoch": 2.421002538071066, + "grad_norm": 8.625, + "learning_rate": 2.7490210479564896e-06, + "loss": 2.5714, + "step": 22893 + }, + { + "epoch": 2.421108291032149, + "grad_norm": 7.90625, + "learning_rate": 2.7488506428556104e-06, + "loss": 2.4706, + "step": 22894 + }, + { + "epoch": 2.4212140439932317, + "grad_norm": 8.75, + "learning_rate": 2.748680236586977e-06, + "loss": 2.692, + "step": 22895 + }, + { + "epoch": 2.4213197969543145, + "grad_norm": 8.1875, + "learning_rate": 2.7485098291513877e-06, + "loss": 2.3719, + "step": 22896 + }, + { + "epoch": 2.421425549915398, + "grad_norm": 8.125, + "learning_rate": 2.748339420549645e-06, + "loss": 2.6249, + "step": 22897 + }, + { + "epoch": 2.4215313028764807, + "grad_norm": 8.1875, + "learning_rate": 2.7481690107825465e-06, + "loss": 2.6147, + "step": 22898 + }, + { + "epoch": 2.4216370558375635, + "grad_norm": 8.8125, + "learning_rate": 2.747998599850892e-06, + "loss": 2.6179, + "step": 22899 + }, + { + "epoch": 2.4217428087986463, + "grad_norm": 9.0625, + "learning_rate": 2.7478281877554806e-06, + "loss": 2.7517, + "step": 22900 + }, + { + "epoch": 2.421848561759729, + "grad_norm": 8.625, + "learning_rate": 2.7476577744971146e-06, + "loss": 2.5817, + "step": 22901 + }, + { + "epoch": 2.421954314720812, + "grad_norm": 8.5, + "learning_rate": 2.747487360076591e-06, + "loss": 2.5349, + "step": 22902 + }, + { + "epoch": 2.422060067681895, + "grad_norm": 8.5, + "learning_rate": 2.747316944494711e-06, + "loss": 2.6115, + "step": 22903 + }, + { + "epoch": 2.422165820642978, + "grad_norm": 8.0, + "learning_rate": 2.7471465277522735e-06, + "loss": 2.4341, + "step": 22904 + }, + { + "epoch": 2.422271573604061, + "grad_norm": 8.25, + "learning_rate": 2.746976109850078e-06, + "loss": 2.4553, + "step": 22905 + }, + { + "epoch": 2.422377326565144, + "grad_norm": 8.625, + "learning_rate": 2.746805690788925e-06, + "loss": 2.5877, + "step": 22906 + }, + { + "epoch": 2.4224830795262267, + "grad_norm": 8.3125, + "learning_rate": 2.7466352705696135e-06, + "loss": 2.505, + "step": 22907 + }, + { + "epoch": 2.4225888324873095, + "grad_norm": 8.0625, + "learning_rate": 2.746464849192944e-06, + "loss": 2.5605, + "step": 22908 + }, + { + "epoch": 2.422694585448393, + "grad_norm": 8.5625, + "learning_rate": 2.746294426659715e-06, + "loss": 2.777, + "step": 22909 + }, + { + "epoch": 2.4228003384094756, + "grad_norm": 8.125, + "learning_rate": 2.746124002970728e-06, + "loss": 2.6495, + "step": 22910 + }, + { + "epoch": 2.4229060913705585, + "grad_norm": 8.125, + "learning_rate": 2.7459535781267814e-06, + "loss": 2.5804, + "step": 22911 + }, + { + "epoch": 2.4230118443316413, + "grad_norm": 8.25, + "learning_rate": 2.745783152128675e-06, + "loss": 2.6406, + "step": 22912 + }, + { + "epoch": 2.423117597292724, + "grad_norm": 8.9375, + "learning_rate": 2.7456127249772092e-06, + "loss": 2.4306, + "step": 22913 + }, + { + "epoch": 2.423223350253807, + "grad_norm": 8.6875, + "learning_rate": 2.745442296673183e-06, + "loss": 2.6582, + "step": 22914 + }, + { + "epoch": 2.42332910321489, + "grad_norm": 8.125, + "learning_rate": 2.745271867217396e-06, + "loss": 2.5876, + "step": 22915 + }, + { + "epoch": 2.423434856175973, + "grad_norm": 7.875, + "learning_rate": 2.7451014366106492e-06, + "loss": 2.5236, + "step": 22916 + }, + { + "epoch": 2.423540609137056, + "grad_norm": 8.3125, + "learning_rate": 2.7449310048537414e-06, + "loss": 2.4788, + "step": 22917 + }, + { + "epoch": 2.4236463620981388, + "grad_norm": 8.875, + "learning_rate": 2.744760571947473e-06, + "loss": 2.2428, + "step": 22918 + }, + { + "epoch": 2.4237521150592216, + "grad_norm": 8.4375, + "learning_rate": 2.744590137892642e-06, + "loss": 2.5093, + "step": 22919 + }, + { + "epoch": 2.4238578680203045, + "grad_norm": 8.0, + "learning_rate": 2.744419702690051e-06, + "loss": 2.5487, + "step": 22920 + }, + { + "epoch": 2.4239636209813873, + "grad_norm": 8.125, + "learning_rate": 2.744249266340498e-06, + "loss": 2.4614, + "step": 22921 + }, + { + "epoch": 2.4240693739424706, + "grad_norm": 8.0625, + "learning_rate": 2.7440788288447824e-06, + "loss": 2.3479, + "step": 22922 + }, + { + "epoch": 2.4241751269035534, + "grad_norm": 8.3125, + "learning_rate": 2.743908390203705e-06, + "loss": 2.5548, + "step": 22923 + }, + { + "epoch": 2.4242808798646363, + "grad_norm": 8.375, + "learning_rate": 2.743737950418065e-06, + "loss": 2.3632, + "step": 22924 + }, + { + "epoch": 2.424386632825719, + "grad_norm": 8.0625, + "learning_rate": 2.743567509488663e-06, + "loss": 2.3701, + "step": 22925 + }, + { + "epoch": 2.424492385786802, + "grad_norm": 9.6875, + "learning_rate": 2.743397067416298e-06, + "loss": 2.5745, + "step": 22926 + }, + { + "epoch": 2.4245981387478848, + "grad_norm": 8.875, + "learning_rate": 2.7432266242017707e-06, + "loss": 2.5435, + "step": 22927 + }, + { + "epoch": 2.424703891708968, + "grad_norm": 8.0625, + "learning_rate": 2.7430561798458793e-06, + "loss": 2.2423, + "step": 22928 + }, + { + "epoch": 2.424809644670051, + "grad_norm": 8.625, + "learning_rate": 2.7428857343494254e-06, + "loss": 2.6767, + "step": 22929 + }, + { + "epoch": 2.4249153976311337, + "grad_norm": 8.6875, + "learning_rate": 2.7427152877132086e-06, + "loss": 2.4906, + "step": 22930 + }, + { + "epoch": 2.4250211505922166, + "grad_norm": 8.125, + "learning_rate": 2.742544839938027e-06, + "loss": 2.3507, + "step": 22931 + }, + { + "epoch": 2.4251269035532994, + "grad_norm": 8.25, + "learning_rate": 2.7423743910246827e-06, + "loss": 2.5656, + "step": 22932 + }, + { + "epoch": 2.4252326565143822, + "grad_norm": 8.6875, + "learning_rate": 2.7422039409739736e-06, + "loss": 2.6949, + "step": 22933 + }, + { + "epoch": 2.425338409475465, + "grad_norm": 8.1875, + "learning_rate": 2.742033489786701e-06, + "loss": 2.3317, + "step": 22934 + }, + { + "epoch": 2.4254441624365484, + "grad_norm": 8.4375, + "learning_rate": 2.7418630374636633e-06, + "loss": 2.6049, + "step": 22935 + }, + { + "epoch": 2.425549915397631, + "grad_norm": 8.3125, + "learning_rate": 2.7416925840056626e-06, + "loss": 2.7051, + "step": 22936 + }, + { + "epoch": 2.425655668358714, + "grad_norm": 8.5625, + "learning_rate": 2.741522129413497e-06, + "loss": 2.424, + "step": 22937 + }, + { + "epoch": 2.425761421319797, + "grad_norm": 8.5, + "learning_rate": 2.7413516736879666e-06, + "loss": 2.4174, + "step": 22938 + }, + { + "epoch": 2.4258671742808797, + "grad_norm": 8.4375, + "learning_rate": 2.741181216829872e-06, + "loss": 2.67, + "step": 22939 + }, + { + "epoch": 2.425972927241963, + "grad_norm": 8.6875, + "learning_rate": 2.741010758840012e-06, + "loss": 2.2406, + "step": 22940 + }, + { + "epoch": 2.426078680203046, + "grad_norm": 7.625, + "learning_rate": 2.7408402997191876e-06, + "loss": 2.4776, + "step": 22941 + }, + { + "epoch": 2.4261844331641287, + "grad_norm": 8.5, + "learning_rate": 2.740669839468198e-06, + "loss": 2.6283, + "step": 22942 + }, + { + "epoch": 2.4262901861252115, + "grad_norm": 8.4375, + "learning_rate": 2.7404993780878423e-06, + "loss": 2.6469, + "step": 22943 + }, + { + "epoch": 2.4263959390862944, + "grad_norm": 8.375, + "learning_rate": 2.740328915578922e-06, + "loss": 2.7008, + "step": 22944 + }, + { + "epoch": 2.426501692047377, + "grad_norm": 8.625, + "learning_rate": 2.740158451942237e-06, + "loss": 2.5934, + "step": 22945 + }, + { + "epoch": 2.42660744500846, + "grad_norm": 8.3125, + "learning_rate": 2.739987987178587e-06, + "loss": 2.5736, + "step": 22946 + }, + { + "epoch": 2.4267131979695433, + "grad_norm": 8.0, + "learning_rate": 2.73981752128877e-06, + "loss": 2.3323, + "step": 22947 + }, + { + "epoch": 2.426818950930626, + "grad_norm": 8.5625, + "learning_rate": 2.7396470542735885e-06, + "loss": 2.6508, + "step": 22948 + }, + { + "epoch": 2.426924703891709, + "grad_norm": 8.4375, + "learning_rate": 2.739476586133841e-06, + "loss": 2.5035, + "step": 22949 + }, + { + "epoch": 2.427030456852792, + "grad_norm": 9.3125, + "learning_rate": 2.739306116870328e-06, + "loss": 2.7583, + "step": 22950 + }, + { + "epoch": 2.4271362098138747, + "grad_norm": 9.0, + "learning_rate": 2.739135646483849e-06, + "loss": 2.4355, + "step": 22951 + }, + { + "epoch": 2.4272419627749575, + "grad_norm": 8.6875, + "learning_rate": 2.7389651749752043e-06, + "loss": 2.6184, + "step": 22952 + }, + { + "epoch": 2.427347715736041, + "grad_norm": 7.96875, + "learning_rate": 2.7387947023451943e-06, + "loss": 2.4784, + "step": 22953 + }, + { + "epoch": 2.4274534686971236, + "grad_norm": 8.4375, + "learning_rate": 2.7386242285946173e-06, + "loss": 2.6483, + "step": 22954 + }, + { + "epoch": 2.4275592216582065, + "grad_norm": 8.375, + "learning_rate": 2.7384537537242754e-06, + "loss": 2.3805, + "step": 22955 + }, + { + "epoch": 2.4276649746192893, + "grad_norm": 8.1875, + "learning_rate": 2.738283277734968e-06, + "loss": 2.5276, + "step": 22956 + }, + { + "epoch": 2.427770727580372, + "grad_norm": 8.75, + "learning_rate": 2.7381128006274934e-06, + "loss": 2.5027, + "step": 22957 + }, + { + "epoch": 2.427876480541455, + "grad_norm": 7.9375, + "learning_rate": 2.7379423224026537e-06, + "loss": 2.5599, + "step": 22958 + }, + { + "epoch": 2.4279822335025383, + "grad_norm": 7.96875, + "learning_rate": 2.737771843061247e-06, + "loss": 2.501, + "step": 22959 + }, + { + "epoch": 2.428087986463621, + "grad_norm": 8.3125, + "learning_rate": 2.7376013626040754e-06, + "loss": 2.5809, + "step": 22960 + }, + { + "epoch": 2.428193739424704, + "grad_norm": 7.84375, + "learning_rate": 2.7374308810319365e-06, + "loss": 2.431, + "step": 22961 + }, + { + "epoch": 2.428299492385787, + "grad_norm": 8.3125, + "learning_rate": 2.7372603983456324e-06, + "loss": 2.4373, + "step": 22962 + }, + { + "epoch": 2.4284052453468696, + "grad_norm": 8.5625, + "learning_rate": 2.7370899145459624e-06, + "loss": 2.573, + "step": 22963 + }, + { + "epoch": 2.4285109983079525, + "grad_norm": 9.0625, + "learning_rate": 2.7369194296337264e-06, + "loss": 2.4784, + "step": 22964 + }, + { + "epoch": 2.4286167512690353, + "grad_norm": 7.9375, + "learning_rate": 2.7367489436097246e-06, + "loss": 2.5479, + "step": 22965 + }, + { + "epoch": 2.4287225042301186, + "grad_norm": 8.125, + "learning_rate": 2.736578456474756e-06, + "loss": 2.7006, + "step": 22966 + }, + { + "epoch": 2.4288282571912014, + "grad_norm": 8.6875, + "learning_rate": 2.736407968229622e-06, + "loss": 2.6914, + "step": 22967 + }, + { + "epoch": 2.4289340101522843, + "grad_norm": 8.375, + "learning_rate": 2.736237478875122e-06, + "loss": 2.6462, + "step": 22968 + }, + { + "epoch": 2.429039763113367, + "grad_norm": 8.625, + "learning_rate": 2.736066988412056e-06, + "loss": 2.7241, + "step": 22969 + }, + { + "epoch": 2.42914551607445, + "grad_norm": 8.875, + "learning_rate": 2.7358964968412243e-06, + "loss": 2.458, + "step": 22970 + }, + { + "epoch": 2.4292512690355332, + "grad_norm": 8.25, + "learning_rate": 2.735726004163427e-06, + "loss": 2.705, + "step": 22971 + }, + { + "epoch": 2.429357021996616, + "grad_norm": 8.3125, + "learning_rate": 2.735555510379464e-06, + "loss": 2.6161, + "step": 22972 + }, + { + "epoch": 2.429462774957699, + "grad_norm": 8.3125, + "learning_rate": 2.7353850154901344e-06, + "loss": 2.6608, + "step": 22973 + }, + { + "epoch": 2.4295685279187818, + "grad_norm": 8.5625, + "learning_rate": 2.7352145194962405e-06, + "loss": 2.5762, + "step": 22974 + }, + { + "epoch": 2.4296742808798646, + "grad_norm": 8.8125, + "learning_rate": 2.735044022398581e-06, + "loss": 2.6385, + "step": 22975 + }, + { + "epoch": 2.4297800338409474, + "grad_norm": 7.9375, + "learning_rate": 2.7348735241979554e-06, + "loss": 2.631, + "step": 22976 + }, + { + "epoch": 2.4298857868020303, + "grad_norm": 7.6875, + "learning_rate": 2.7347030248951645e-06, + "loss": 2.4161, + "step": 22977 + }, + { + "epoch": 2.4299915397631136, + "grad_norm": 8.5625, + "learning_rate": 2.734532524491008e-06, + "loss": 2.4785, + "step": 22978 + }, + { + "epoch": 2.4300972927241964, + "grad_norm": 9.5625, + "learning_rate": 2.734362022986286e-06, + "loss": 2.6827, + "step": 22979 + }, + { + "epoch": 2.4302030456852792, + "grad_norm": 8.4375, + "learning_rate": 2.7341915203818e-06, + "loss": 2.5426, + "step": 22980 + }, + { + "epoch": 2.430308798646362, + "grad_norm": 8.1875, + "learning_rate": 2.734021016678349e-06, + "loss": 2.5177, + "step": 22981 + }, + { + "epoch": 2.430414551607445, + "grad_norm": 8.3125, + "learning_rate": 2.7338505118767315e-06, + "loss": 2.1678, + "step": 22982 + }, + { + "epoch": 2.4305203045685277, + "grad_norm": 8.125, + "learning_rate": 2.733680005977751e-06, + "loss": 2.5811, + "step": 22983 + }, + { + "epoch": 2.4306260575296106, + "grad_norm": 8.125, + "learning_rate": 2.733509498982205e-06, + "loss": 2.4903, + "step": 22984 + }, + { + "epoch": 2.430731810490694, + "grad_norm": 8.125, + "learning_rate": 2.7333389908908944e-06, + "loss": 2.594, + "step": 22985 + }, + { + "epoch": 2.4308375634517767, + "grad_norm": 8.5, + "learning_rate": 2.7331684817046194e-06, + "loss": 2.7547, + "step": 22986 + }, + { + "epoch": 2.4309433164128595, + "grad_norm": 8.9375, + "learning_rate": 2.7329979714241804e-06, + "loss": 2.8153, + "step": 22987 + }, + { + "epoch": 2.4310490693739424, + "grad_norm": 8.0, + "learning_rate": 2.732827460050377e-06, + "loss": 2.4867, + "step": 22988 + }, + { + "epoch": 2.4311548223350252, + "grad_norm": 8.5, + "learning_rate": 2.7326569475840088e-06, + "loss": 2.5282, + "step": 22989 + }, + { + "epoch": 2.4312605752961085, + "grad_norm": 10.5, + "learning_rate": 2.732486434025878e-06, + "loss": 2.6083, + "step": 22990 + }, + { + "epoch": 2.4313663282571913, + "grad_norm": 8.1875, + "learning_rate": 2.732315919376782e-06, + "loss": 2.6259, + "step": 22991 + }, + { + "epoch": 2.431472081218274, + "grad_norm": 8.75, + "learning_rate": 2.7321454036375233e-06, + "loss": 2.4288, + "step": 22992 + }, + { + "epoch": 2.431577834179357, + "grad_norm": 8.25, + "learning_rate": 2.7319748868089012e-06, + "loss": 2.4858, + "step": 22993 + }, + { + "epoch": 2.43168358714044, + "grad_norm": 8.6875, + "learning_rate": 2.731804368891716e-06, + "loss": 2.5961, + "step": 22994 + }, + { + "epoch": 2.4317893401015227, + "grad_norm": 8.6875, + "learning_rate": 2.7316338498867674e-06, + "loss": 2.6054, + "step": 22995 + }, + { + "epoch": 2.4318950930626055, + "grad_norm": 8.4375, + "learning_rate": 2.731463329794856e-06, + "loss": 2.5069, + "step": 22996 + }, + { + "epoch": 2.432000846023689, + "grad_norm": 8.3125, + "learning_rate": 2.7312928086167816e-06, + "loss": 2.7448, + "step": 22997 + }, + { + "epoch": 2.4321065989847717, + "grad_norm": 8.125, + "learning_rate": 2.731122286353344e-06, + "loss": 2.6005, + "step": 22998 + }, + { + "epoch": 2.4322123519458545, + "grad_norm": 8.625, + "learning_rate": 2.7309517630053454e-06, + "loss": 2.5143, + "step": 22999 + }, + { + "epoch": 2.4323181049069373, + "grad_norm": 7.625, + "learning_rate": 2.7307812385735845e-06, + "loss": 2.4485, + "step": 23000 + }, + { + "epoch": 2.43242385786802, + "grad_norm": 8.3125, + "learning_rate": 2.7306107130588607e-06, + "loss": 2.4833, + "step": 23001 + }, + { + "epoch": 2.4325296108291035, + "grad_norm": 8.5625, + "learning_rate": 2.7304401864619756e-06, + "loss": 2.5644, + "step": 23002 + }, + { + "epoch": 2.4326353637901863, + "grad_norm": 8.6875, + "learning_rate": 2.7302696587837297e-06, + "loss": 2.4476, + "step": 23003 + }, + { + "epoch": 2.432741116751269, + "grad_norm": 8.6875, + "learning_rate": 2.7300991300249213e-06, + "loss": 2.5291, + "step": 23004 + }, + { + "epoch": 2.432846869712352, + "grad_norm": 8.5625, + "learning_rate": 2.7299286001863522e-06, + "loss": 2.2986, + "step": 23005 + }, + { + "epoch": 2.432952622673435, + "grad_norm": 8.3125, + "learning_rate": 2.729758069268822e-06, + "loss": 2.4721, + "step": 23006 + }, + { + "epoch": 2.4330583756345177, + "grad_norm": 8.3125, + "learning_rate": 2.729587537273132e-06, + "loss": 2.6235, + "step": 23007 + }, + { + "epoch": 2.4331641285956005, + "grad_norm": 8.3125, + "learning_rate": 2.7294170042000805e-06, + "loss": 2.4736, + "step": 23008 + }, + { + "epoch": 2.433269881556684, + "grad_norm": 9.75, + "learning_rate": 2.7292464700504695e-06, + "loss": 2.593, + "step": 23009 + }, + { + "epoch": 2.4333756345177666, + "grad_norm": 8.375, + "learning_rate": 2.729075934825098e-06, + "loss": 2.4626, + "step": 23010 + }, + { + "epoch": 2.4334813874788495, + "grad_norm": 8.125, + "learning_rate": 2.7289053985247675e-06, + "loss": 2.4906, + "step": 23011 + }, + { + "epoch": 2.4335871404399323, + "grad_norm": 7.90625, + "learning_rate": 2.728734861150277e-06, + "loss": 2.2953, + "step": 23012 + }, + { + "epoch": 2.433692893401015, + "grad_norm": 8.0, + "learning_rate": 2.7285643227024283e-06, + "loss": 2.5143, + "step": 23013 + }, + { + "epoch": 2.433798646362098, + "grad_norm": 8.1875, + "learning_rate": 2.7283937831820193e-06, + "loss": 2.5308, + "step": 23014 + }, + { + "epoch": 2.433904399323181, + "grad_norm": 8.1875, + "learning_rate": 2.7282232425898524e-06, + "loss": 2.5787, + "step": 23015 + }, + { + "epoch": 2.434010152284264, + "grad_norm": 8.4375, + "learning_rate": 2.728052700926728e-06, + "loss": 2.4339, + "step": 23016 + }, + { + "epoch": 2.434115905245347, + "grad_norm": 7.875, + "learning_rate": 2.7278821581934437e-06, + "loss": 2.5078, + "step": 23017 + }, + { + "epoch": 2.4342216582064298, + "grad_norm": 8.3125, + "learning_rate": 2.7277116143908034e-06, + "loss": 2.3571, + "step": 23018 + }, + { + "epoch": 2.4343274111675126, + "grad_norm": 9.125, + "learning_rate": 2.727541069519605e-06, + "loss": 2.6392, + "step": 23019 + }, + { + "epoch": 2.4344331641285955, + "grad_norm": 8.8125, + "learning_rate": 2.727370523580649e-06, + "loss": 2.5863, + "step": 23020 + }, + { + "epoch": 2.4345389170896787, + "grad_norm": 8.25, + "learning_rate": 2.727199976574737e-06, + "loss": 2.4233, + "step": 23021 + }, + { + "epoch": 2.4346446700507616, + "grad_norm": 8.375, + "learning_rate": 2.727029428502668e-06, + "loss": 2.5844, + "step": 23022 + }, + { + "epoch": 2.4347504230118444, + "grad_norm": 8.3125, + "learning_rate": 2.7268588793652424e-06, + "loss": 2.478, + "step": 23023 + }, + { + "epoch": 2.4348561759729273, + "grad_norm": 8.4375, + "learning_rate": 2.7266883291632605e-06, + "loss": 2.4099, + "step": 23024 + }, + { + "epoch": 2.43496192893401, + "grad_norm": 9.4375, + "learning_rate": 2.726517777897524e-06, + "loss": 2.6889, + "step": 23025 + }, + { + "epoch": 2.435067681895093, + "grad_norm": 8.3125, + "learning_rate": 2.7263472255688315e-06, + "loss": 2.488, + "step": 23026 + }, + { + "epoch": 2.4351734348561758, + "grad_norm": 8.5625, + "learning_rate": 2.7261766721779844e-06, + "loss": 2.5717, + "step": 23027 + }, + { + "epoch": 2.435279187817259, + "grad_norm": 7.84375, + "learning_rate": 2.726006117725783e-06, + "loss": 2.3178, + "step": 23028 + }, + { + "epoch": 2.435384940778342, + "grad_norm": 8.8125, + "learning_rate": 2.7258355622130274e-06, + "loss": 2.5117, + "step": 23029 + }, + { + "epoch": 2.4354906937394247, + "grad_norm": 8.5, + "learning_rate": 2.7256650056405166e-06, + "loss": 2.7573, + "step": 23030 + }, + { + "epoch": 2.4355964467005076, + "grad_norm": 8.375, + "learning_rate": 2.7254944480090533e-06, + "loss": 2.1748, + "step": 23031 + }, + { + "epoch": 2.4357021996615904, + "grad_norm": 8.0, + "learning_rate": 2.725323889319436e-06, + "loss": 2.3829, + "step": 23032 + }, + { + "epoch": 2.4358079526226732, + "grad_norm": 9.0, + "learning_rate": 2.7251533295724666e-06, + "loss": 2.6625, + "step": 23033 + }, + { + "epoch": 2.4359137055837565, + "grad_norm": 8.75, + "learning_rate": 2.7249827687689445e-06, + "loss": 2.2673, + "step": 23034 + }, + { + "epoch": 2.4360194585448394, + "grad_norm": 7.875, + "learning_rate": 2.7248122069096704e-06, + "loss": 2.407, + "step": 23035 + }, + { + "epoch": 2.436125211505922, + "grad_norm": 7.9375, + "learning_rate": 2.7246416439954442e-06, + "loss": 2.3597, + "step": 23036 + }, + { + "epoch": 2.436230964467005, + "grad_norm": 8.6875, + "learning_rate": 2.724471080027067e-06, + "loss": 2.6219, + "step": 23037 + }, + { + "epoch": 2.436336717428088, + "grad_norm": 9.0, + "learning_rate": 2.7243005150053386e-06, + "loss": 2.6728, + "step": 23038 + }, + { + "epoch": 2.4364424703891707, + "grad_norm": 8.375, + "learning_rate": 2.7241299489310596e-06, + "loss": 2.6052, + "step": 23039 + }, + { + "epoch": 2.436548223350254, + "grad_norm": 9.0, + "learning_rate": 2.7239593818050307e-06, + "loss": 2.519, + "step": 23040 + }, + { + "epoch": 2.436653976311337, + "grad_norm": 8.9375, + "learning_rate": 2.7237888136280517e-06, + "loss": 2.4189, + "step": 23041 + }, + { + "epoch": 2.4367597292724197, + "grad_norm": 8.4375, + "learning_rate": 2.7236182444009233e-06, + "loss": 2.6979, + "step": 23042 + }, + { + "epoch": 2.4368654822335025, + "grad_norm": 8.3125, + "learning_rate": 2.723447674124446e-06, + "loss": 2.612, + "step": 23043 + }, + { + "epoch": 2.4369712351945854, + "grad_norm": 8.6875, + "learning_rate": 2.7232771027994204e-06, + "loss": 2.6294, + "step": 23044 + }, + { + "epoch": 2.437076988155668, + "grad_norm": 9.0, + "learning_rate": 2.7231065304266467e-06, + "loss": 2.3563, + "step": 23045 + }, + { + "epoch": 2.437182741116751, + "grad_norm": 8.3125, + "learning_rate": 2.7229359570069246e-06, + "loss": 2.3056, + "step": 23046 + }, + { + "epoch": 2.4372884940778343, + "grad_norm": 8.0625, + "learning_rate": 2.722765382541056e-06, + "loss": 2.4115, + "step": 23047 + }, + { + "epoch": 2.437394247038917, + "grad_norm": 8.4375, + "learning_rate": 2.722594807029841e-06, + "loss": 2.456, + "step": 23048 + }, + { + "epoch": 2.4375, + "grad_norm": 8.3125, + "learning_rate": 2.7224242304740784e-06, + "loss": 2.4911, + "step": 23049 + }, + { + "epoch": 2.437605752961083, + "grad_norm": 8.25, + "learning_rate": 2.72225365287457e-06, + "loss": 2.4389, + "step": 23050 + }, + { + "epoch": 2.4377115059221657, + "grad_norm": 7.9375, + "learning_rate": 2.722083074232117e-06, + "loss": 2.4711, + "step": 23051 + }, + { + "epoch": 2.437817258883249, + "grad_norm": 8.125, + "learning_rate": 2.721912494547518e-06, + "loss": 2.2445, + "step": 23052 + }, + { + "epoch": 2.437923011844332, + "grad_norm": 8.5625, + "learning_rate": 2.721741913821575e-06, + "loss": 2.6574, + "step": 23053 + }, + { + "epoch": 2.4380287648054146, + "grad_norm": 8.125, + "learning_rate": 2.7215713320550877e-06, + "loss": 2.4329, + "step": 23054 + }, + { + "epoch": 2.4381345177664975, + "grad_norm": 8.625, + "learning_rate": 2.7214007492488566e-06, + "loss": 2.4139, + "step": 23055 + }, + { + "epoch": 2.4382402707275803, + "grad_norm": 8.375, + "learning_rate": 2.7212301654036827e-06, + "loss": 2.7068, + "step": 23056 + }, + { + "epoch": 2.438346023688663, + "grad_norm": 8.25, + "learning_rate": 2.7210595805203666e-06, + "loss": 2.3859, + "step": 23057 + }, + { + "epoch": 2.438451776649746, + "grad_norm": 8.5, + "learning_rate": 2.7208889945997067e-06, + "loss": 2.6029, + "step": 23058 + }, + { + "epoch": 2.4385575296108293, + "grad_norm": 9.4375, + "learning_rate": 2.7207184076425063e-06, + "loss": 2.7841, + "step": 23059 + }, + { + "epoch": 2.438663282571912, + "grad_norm": 8.3125, + "learning_rate": 2.7205478196495645e-06, + "loss": 2.4299, + "step": 23060 + }, + { + "epoch": 2.438769035532995, + "grad_norm": 8.25, + "learning_rate": 2.720377230621682e-06, + "loss": 2.4276, + "step": 23061 + }, + { + "epoch": 2.438874788494078, + "grad_norm": 8.25, + "learning_rate": 2.720206640559659e-06, + "loss": 2.6698, + "step": 23062 + }, + { + "epoch": 2.4389805414551606, + "grad_norm": 8.75, + "learning_rate": 2.720036049464297e-06, + "loss": 2.7101, + "step": 23063 + }, + { + "epoch": 2.4390862944162435, + "grad_norm": 8.4375, + "learning_rate": 2.7198654573363953e-06, + "loss": 2.4718, + "step": 23064 + }, + { + "epoch": 2.4391920473773268, + "grad_norm": 8.125, + "learning_rate": 2.7196948641767546e-06, + "loss": 2.5853, + "step": 23065 + }, + { + "epoch": 2.4392978003384096, + "grad_norm": 7.8125, + "learning_rate": 2.719524269986176e-06, + "loss": 2.5626, + "step": 23066 + }, + { + "epoch": 2.4394035532994924, + "grad_norm": 8.4375, + "learning_rate": 2.71935367476546e-06, + "loss": 2.5284, + "step": 23067 + }, + { + "epoch": 2.4395093062605753, + "grad_norm": 8.5625, + "learning_rate": 2.719183078515406e-06, + "loss": 2.364, + "step": 23068 + }, + { + "epoch": 2.439615059221658, + "grad_norm": 9.0, + "learning_rate": 2.719012481236817e-06, + "loss": 2.2492, + "step": 23069 + }, + { + "epoch": 2.439720812182741, + "grad_norm": 8.375, + "learning_rate": 2.7188418829304913e-06, + "loss": 2.5089, + "step": 23070 + }, + { + "epoch": 2.4398265651438242, + "grad_norm": 8.625, + "learning_rate": 2.7186712835972294e-06, + "loss": 1.9164, + "step": 23071 + }, + { + "epoch": 2.439932318104907, + "grad_norm": 7.9375, + "learning_rate": 2.7185006832378337e-06, + "loss": 2.2889, + "step": 23072 + }, + { + "epoch": 2.44003807106599, + "grad_norm": 8.3125, + "learning_rate": 2.7183300818531033e-06, + "loss": 2.5025, + "step": 23073 + }, + { + "epoch": 2.4401438240270727, + "grad_norm": 8.4375, + "learning_rate": 2.7181594794438395e-06, + "loss": 2.3842, + "step": 23074 + }, + { + "epoch": 2.4402495769881556, + "grad_norm": 8.25, + "learning_rate": 2.717988876010842e-06, + "loss": 2.2271, + "step": 23075 + }, + { + "epoch": 2.4403553299492384, + "grad_norm": 8.625, + "learning_rate": 2.717818271554911e-06, + "loss": 2.5667, + "step": 23076 + }, + { + "epoch": 2.4404610829103213, + "grad_norm": 10.0, + "learning_rate": 2.7176476660768495e-06, + "loss": 2.7429, + "step": 23077 + }, + { + "epoch": 2.4405668358714045, + "grad_norm": 8.25, + "learning_rate": 2.717477059577456e-06, + "loss": 2.5707, + "step": 23078 + }, + { + "epoch": 2.4406725888324874, + "grad_norm": 8.1875, + "learning_rate": 2.717306452057531e-06, + "loss": 2.5506, + "step": 23079 + }, + { + "epoch": 2.4407783417935702, + "grad_norm": 8.75, + "learning_rate": 2.7171358435178767e-06, + "loss": 2.6214, + "step": 23080 + }, + { + "epoch": 2.440884094754653, + "grad_norm": 9.625, + "learning_rate": 2.716965233959292e-06, + "loss": 2.5943, + "step": 23081 + }, + { + "epoch": 2.440989847715736, + "grad_norm": 7.75, + "learning_rate": 2.716794623382578e-06, + "loss": 2.5373, + "step": 23082 + }, + { + "epoch": 2.441095600676819, + "grad_norm": 7.71875, + "learning_rate": 2.7166240117885363e-06, + "loss": 2.4172, + "step": 23083 + }, + { + "epoch": 2.441201353637902, + "grad_norm": 8.625, + "learning_rate": 2.716453399177966e-06, + "loss": 2.6184, + "step": 23084 + }, + { + "epoch": 2.441307106598985, + "grad_norm": 8.5, + "learning_rate": 2.716282785551668e-06, + "loss": 2.548, + "step": 23085 + }, + { + "epoch": 2.4414128595600677, + "grad_norm": 8.75, + "learning_rate": 2.716112170910445e-06, + "loss": 2.8459, + "step": 23086 + }, + { + "epoch": 2.4415186125211505, + "grad_norm": 7.875, + "learning_rate": 2.7159415552550943e-06, + "loss": 2.5801, + "step": 23087 + }, + { + "epoch": 2.4416243654822334, + "grad_norm": 9.0, + "learning_rate": 2.715770938586419e-06, + "loss": 2.6882, + "step": 23088 + }, + { + "epoch": 2.441730118443316, + "grad_norm": 8.4375, + "learning_rate": 2.7156003209052197e-06, + "loss": 2.5484, + "step": 23089 + }, + { + "epoch": 2.4418358714043995, + "grad_norm": 8.25, + "learning_rate": 2.7154297022122942e-06, + "loss": 2.511, + "step": 23090 + }, + { + "epoch": 2.4419416243654823, + "grad_norm": 8.4375, + "learning_rate": 2.715259082508447e-06, + "loss": 2.6939, + "step": 23091 + }, + { + "epoch": 2.442047377326565, + "grad_norm": 9.5, + "learning_rate": 2.7150884617944767e-06, + "loss": 2.3575, + "step": 23092 + }, + { + "epoch": 2.442153130287648, + "grad_norm": 8.3125, + "learning_rate": 2.7149178400711833e-06, + "loss": 2.4471, + "step": 23093 + }, + { + "epoch": 2.442258883248731, + "grad_norm": 8.25, + "learning_rate": 2.714747217339369e-06, + "loss": 2.5678, + "step": 23094 + }, + { + "epoch": 2.4423646362098137, + "grad_norm": 8.5, + "learning_rate": 2.7145765935998344e-06, + "loss": 2.5537, + "step": 23095 + }, + { + "epoch": 2.4424703891708965, + "grad_norm": 8.3125, + "learning_rate": 2.7144059688533788e-06, + "loss": 2.5944, + "step": 23096 + }, + { + "epoch": 2.44257614213198, + "grad_norm": 8.5, + "learning_rate": 2.7142353431008035e-06, + "loss": 2.4506, + "step": 23097 + }, + { + "epoch": 2.4426818950930627, + "grad_norm": 9.1875, + "learning_rate": 2.71406471634291e-06, + "loss": 2.4715, + "step": 23098 + }, + { + "epoch": 2.4427876480541455, + "grad_norm": 8.3125, + "learning_rate": 2.713894088580498e-06, + "loss": 2.48, + "step": 23099 + }, + { + "epoch": 2.4428934010152283, + "grad_norm": 8.25, + "learning_rate": 2.7137234598143684e-06, + "loss": 2.6228, + "step": 23100 + }, + { + "epoch": 2.442999153976311, + "grad_norm": 8.625, + "learning_rate": 2.713552830045323e-06, + "loss": 2.401, + "step": 23101 + }, + { + "epoch": 2.4431049069373945, + "grad_norm": 8.1875, + "learning_rate": 2.7133821992741605e-06, + "loss": 2.6048, + "step": 23102 + }, + { + "epoch": 2.4432106598984773, + "grad_norm": 8.4375, + "learning_rate": 2.7132115675016818e-06, + "loss": 2.5954, + "step": 23103 + }, + { + "epoch": 2.44331641285956, + "grad_norm": 8.0625, + "learning_rate": 2.7130409347286903e-06, + "loss": 2.4086, + "step": 23104 + }, + { + "epoch": 2.443422165820643, + "grad_norm": 8.3125, + "learning_rate": 2.7128703009559836e-06, + "loss": 2.6029, + "step": 23105 + }, + { + "epoch": 2.443527918781726, + "grad_norm": 8.1875, + "learning_rate": 2.712699666184364e-06, + "loss": 2.4768, + "step": 23106 + }, + { + "epoch": 2.4436336717428087, + "grad_norm": 8.0, + "learning_rate": 2.712529030414632e-06, + "loss": 2.4994, + "step": 23107 + }, + { + "epoch": 2.4437394247038915, + "grad_norm": 9.1875, + "learning_rate": 2.7123583936475883e-06, + "loss": 2.4724, + "step": 23108 + }, + { + "epoch": 2.4438451776649748, + "grad_norm": 8.0625, + "learning_rate": 2.7121877558840333e-06, + "loss": 2.4336, + "step": 23109 + }, + { + "epoch": 2.4439509306260576, + "grad_norm": 8.8125, + "learning_rate": 2.7120171171247677e-06, + "loss": 2.4788, + "step": 23110 + }, + { + "epoch": 2.4440566835871405, + "grad_norm": 8.25, + "learning_rate": 2.7118464773705934e-06, + "loss": 2.6068, + "step": 23111 + }, + { + "epoch": 2.4441624365482233, + "grad_norm": 8.3125, + "learning_rate": 2.711675836622309e-06, + "loss": 2.5602, + "step": 23112 + }, + { + "epoch": 2.444268189509306, + "grad_norm": 8.3125, + "learning_rate": 2.7115051948807163e-06, + "loss": 2.6009, + "step": 23113 + }, + { + "epoch": 2.4443739424703894, + "grad_norm": 8.375, + "learning_rate": 2.7113345521466173e-06, + "loss": 2.4428, + "step": 23114 + }, + { + "epoch": 2.4444796954314723, + "grad_norm": 8.125, + "learning_rate": 2.7111639084208114e-06, + "loss": 2.5116, + "step": 23115 + }, + { + "epoch": 2.444585448392555, + "grad_norm": 8.125, + "learning_rate": 2.7109932637040993e-06, + "loss": 2.5448, + "step": 23116 + }, + { + "epoch": 2.444691201353638, + "grad_norm": 8.4375, + "learning_rate": 2.7108226179972823e-06, + "loss": 2.5763, + "step": 23117 + }, + { + "epoch": 2.4447969543147208, + "grad_norm": 8.125, + "learning_rate": 2.7106519713011614e-06, + "loss": 2.4052, + "step": 23118 + }, + { + "epoch": 2.4449027072758036, + "grad_norm": 8.4375, + "learning_rate": 2.710481323616536e-06, + "loss": 2.5353, + "step": 23119 + }, + { + "epoch": 2.4450084602368864, + "grad_norm": 8.125, + "learning_rate": 2.710310674944209e-06, + "loss": 2.4346, + "step": 23120 + }, + { + "epoch": 2.4451142131979697, + "grad_norm": 8.0625, + "learning_rate": 2.710140025284979e-06, + "loss": 2.5697, + "step": 23121 + }, + { + "epoch": 2.4452199661590526, + "grad_norm": 8.5, + "learning_rate": 2.709969374639648e-06, + "loss": 2.5346, + "step": 23122 + }, + { + "epoch": 2.4453257191201354, + "grad_norm": 8.375, + "learning_rate": 2.7097987230090167e-06, + "loss": 2.5733, + "step": 23123 + }, + { + "epoch": 2.4454314720812182, + "grad_norm": 8.375, + "learning_rate": 2.7096280703938867e-06, + "loss": 2.1864, + "step": 23124 + }, + { + "epoch": 2.445537225042301, + "grad_norm": 8.625, + "learning_rate": 2.7094574167950566e-06, + "loss": 2.3802, + "step": 23125 + }, + { + "epoch": 2.445642978003384, + "grad_norm": 8.0625, + "learning_rate": 2.709286762213329e-06, + "loss": 2.4988, + "step": 23126 + }, + { + "epoch": 2.4457487309644668, + "grad_norm": 8.3125, + "learning_rate": 2.7091161066495047e-06, + "loss": 2.5679, + "step": 23127 + }, + { + "epoch": 2.44585448392555, + "grad_norm": 8.9375, + "learning_rate": 2.708945450104383e-06, + "loss": 2.7389, + "step": 23128 + }, + { + "epoch": 2.445960236886633, + "grad_norm": 8.25, + "learning_rate": 2.708774792578766e-06, + "loss": 2.5449, + "step": 23129 + }, + { + "epoch": 2.4460659898477157, + "grad_norm": 8.375, + "learning_rate": 2.708604134073455e-06, + "loss": 2.4012, + "step": 23130 + }, + { + "epoch": 2.4461717428087986, + "grad_norm": 9.4375, + "learning_rate": 2.7084334745892498e-06, + "loss": 2.7087, + "step": 23131 + }, + { + "epoch": 2.4462774957698814, + "grad_norm": 8.1875, + "learning_rate": 2.708262814126951e-06, + "loss": 2.4387, + "step": 23132 + }, + { + "epoch": 2.4463832487309647, + "grad_norm": 8.9375, + "learning_rate": 2.708092152687361e-06, + "loss": 2.7007, + "step": 23133 + }, + { + "epoch": 2.4464890016920475, + "grad_norm": 8.1875, + "learning_rate": 2.707921490271279e-06, + "loss": 2.4759, + "step": 23134 + }, + { + "epoch": 2.4465947546531304, + "grad_norm": 8.25, + "learning_rate": 2.7077508268795068e-06, + "loss": 2.5213, + "step": 23135 + }, + { + "epoch": 2.446700507614213, + "grad_norm": 9.125, + "learning_rate": 2.707580162512845e-06, + "loss": 2.7057, + "step": 23136 + }, + { + "epoch": 2.446806260575296, + "grad_norm": 8.75, + "learning_rate": 2.7074094971720944e-06, + "loss": 2.4705, + "step": 23137 + }, + { + "epoch": 2.446912013536379, + "grad_norm": 8.75, + "learning_rate": 2.7072388308580543e-06, + "loss": 2.621, + "step": 23138 + }, + { + "epoch": 2.4470177664974617, + "grad_norm": 8.75, + "learning_rate": 2.707068163571529e-06, + "loss": 2.5687, + "step": 23139 + }, + { + "epoch": 2.447123519458545, + "grad_norm": 9.0625, + "learning_rate": 2.706897495313317e-06, + "loss": 2.4511, + "step": 23140 + }, + { + "epoch": 2.447229272419628, + "grad_norm": 8.1875, + "learning_rate": 2.7067268260842196e-06, + "loss": 2.7291, + "step": 23141 + }, + { + "epoch": 2.4473350253807107, + "grad_norm": 8.625, + "learning_rate": 2.706556155885037e-06, + "loss": 2.6997, + "step": 23142 + }, + { + "epoch": 2.4474407783417935, + "grad_norm": 8.3125, + "learning_rate": 2.706385484716572e-06, + "loss": 2.5812, + "step": 23143 + }, + { + "epoch": 2.4475465313028764, + "grad_norm": 8.8125, + "learning_rate": 2.7062148125796238e-06, + "loss": 2.6092, + "step": 23144 + }, + { + "epoch": 2.447652284263959, + "grad_norm": 8.875, + "learning_rate": 2.706044139474994e-06, + "loss": 2.6491, + "step": 23145 + }, + { + "epoch": 2.4477580372250425, + "grad_norm": 8.3125, + "learning_rate": 2.7058734654034837e-06, + "loss": 2.4364, + "step": 23146 + }, + { + "epoch": 2.4478637901861253, + "grad_norm": 9.8125, + "learning_rate": 2.7057027903658928e-06, + "loss": 2.5758, + "step": 23147 + }, + { + "epoch": 2.447969543147208, + "grad_norm": 8.125, + "learning_rate": 2.7055321143630227e-06, + "loss": 2.3482, + "step": 23148 + }, + { + "epoch": 2.448075296108291, + "grad_norm": 8.4375, + "learning_rate": 2.705361437395675e-06, + "loss": 2.5432, + "step": 23149 + }, + { + "epoch": 2.448181049069374, + "grad_norm": 8.1875, + "learning_rate": 2.7051907594646497e-06, + "loss": 2.5952, + "step": 23150 + }, + { + "epoch": 2.4482868020304567, + "grad_norm": 8.5625, + "learning_rate": 2.705020080570748e-06, + "loss": 2.7565, + "step": 23151 + }, + { + "epoch": 2.44839255499154, + "grad_norm": 8.4375, + "learning_rate": 2.7048494007147713e-06, + "loss": 2.6817, + "step": 23152 + }, + { + "epoch": 2.448498307952623, + "grad_norm": 8.0625, + "learning_rate": 2.7046787198975205e-06, + "loss": 2.4408, + "step": 23153 + }, + { + "epoch": 2.4486040609137056, + "grad_norm": 8.5625, + "learning_rate": 2.7045080381197953e-06, + "loss": 2.4869, + "step": 23154 + }, + { + "epoch": 2.4487098138747885, + "grad_norm": 8.75, + "learning_rate": 2.7043373553823974e-06, + "loss": 2.2875, + "step": 23155 + }, + { + "epoch": 2.4488155668358713, + "grad_norm": 8.1875, + "learning_rate": 2.7041666716861287e-06, + "loss": 2.5303, + "step": 23156 + }, + { + "epoch": 2.448921319796954, + "grad_norm": 8.3125, + "learning_rate": 2.7039959870317885e-06, + "loss": 2.4973, + "step": 23157 + }, + { + "epoch": 2.449027072758037, + "grad_norm": 8.5, + "learning_rate": 2.703825301420179e-06, + "loss": 2.5866, + "step": 23158 + }, + { + "epoch": 2.4491328257191203, + "grad_norm": 8.0, + "learning_rate": 2.7036546148521007e-06, + "loss": 2.3944, + "step": 23159 + }, + { + "epoch": 2.449238578680203, + "grad_norm": 7.875, + "learning_rate": 2.7034839273283547e-06, + "loss": 2.5233, + "step": 23160 + }, + { + "epoch": 2.449344331641286, + "grad_norm": 9.0625, + "learning_rate": 2.7033132388497417e-06, + "loss": 2.6031, + "step": 23161 + }, + { + "epoch": 2.449450084602369, + "grad_norm": 8.5625, + "learning_rate": 2.703142549417063e-06, + "loss": 2.4042, + "step": 23162 + }, + { + "epoch": 2.4495558375634516, + "grad_norm": 8.625, + "learning_rate": 2.7029718590311194e-06, + "loss": 2.3786, + "step": 23163 + }, + { + "epoch": 2.449661590524535, + "grad_norm": 9.1875, + "learning_rate": 2.702801167692712e-06, + "loss": 2.5854, + "step": 23164 + }, + { + "epoch": 2.4497673434856178, + "grad_norm": 8.0, + "learning_rate": 2.7026304754026412e-06, + "loss": 2.4153, + "step": 23165 + }, + { + "epoch": 2.4498730964467006, + "grad_norm": 8.25, + "learning_rate": 2.7024597821617087e-06, + "loss": 2.4312, + "step": 23166 + }, + { + "epoch": 2.4499788494077834, + "grad_norm": 8.125, + "learning_rate": 2.702289087970715e-06, + "loss": 2.4505, + "step": 23167 + }, + { + "epoch": 2.4500846023688663, + "grad_norm": 8.1875, + "learning_rate": 2.7021183928304617e-06, + "loss": 2.532, + "step": 23168 + }, + { + "epoch": 2.450190355329949, + "grad_norm": 8.25, + "learning_rate": 2.701947696741749e-06, + "loss": 2.562, + "step": 23169 + }, + { + "epoch": 2.450296108291032, + "grad_norm": 8.9375, + "learning_rate": 2.7017769997053788e-06, + "loss": 2.5343, + "step": 23170 + }, + { + "epoch": 2.4504018612521152, + "grad_norm": 7.90625, + "learning_rate": 2.7016063017221515e-06, + "loss": 2.4653, + "step": 23171 + }, + { + "epoch": 2.450507614213198, + "grad_norm": 8.4375, + "learning_rate": 2.7014356027928684e-06, + "loss": 2.4661, + "step": 23172 + }, + { + "epoch": 2.450613367174281, + "grad_norm": 8.1875, + "learning_rate": 2.70126490291833e-06, + "loss": 2.5842, + "step": 23173 + }, + { + "epoch": 2.4507191201353637, + "grad_norm": 8.5, + "learning_rate": 2.701094202099338e-06, + "loss": 2.4628, + "step": 23174 + }, + { + "epoch": 2.4508248730964466, + "grad_norm": 8.3125, + "learning_rate": 2.7009235003366934e-06, + "loss": 2.4897, + "step": 23175 + }, + { + "epoch": 2.4509306260575294, + "grad_norm": 9.375, + "learning_rate": 2.7007527976311965e-06, + "loss": 2.6117, + "step": 23176 + }, + { + "epoch": 2.4510363790186127, + "grad_norm": 9.0, + "learning_rate": 2.700582093983649e-06, + "loss": 2.8422, + "step": 23177 + }, + { + "epoch": 2.4511421319796955, + "grad_norm": 8.4375, + "learning_rate": 2.7004113893948526e-06, + "loss": 2.6455, + "step": 23178 + }, + { + "epoch": 2.4512478849407784, + "grad_norm": 8.625, + "learning_rate": 2.7002406838656057e-06, + "loss": 2.1361, + "step": 23179 + }, + { + "epoch": 2.4513536379018612, + "grad_norm": 8.1875, + "learning_rate": 2.7000699773967126e-06, + "loss": 2.697, + "step": 23180 + }, + { + "epoch": 2.451459390862944, + "grad_norm": 8.5, + "learning_rate": 2.699899269988973e-06, + "loss": 2.5863, + "step": 23181 + }, + { + "epoch": 2.451565143824027, + "grad_norm": 8.1875, + "learning_rate": 2.699728561643187e-06, + "loss": 2.4743, + "step": 23182 + }, + { + "epoch": 2.45167089678511, + "grad_norm": 8.25, + "learning_rate": 2.6995578523601567e-06, + "loss": 2.663, + "step": 23183 + }, + { + "epoch": 2.451776649746193, + "grad_norm": 8.5, + "learning_rate": 2.699387142140684e-06, + "loss": 2.4119, + "step": 23184 + }, + { + "epoch": 2.451882402707276, + "grad_norm": 8.4375, + "learning_rate": 2.6992164309855686e-06, + "loss": 2.5411, + "step": 23185 + }, + { + "epoch": 2.4519881556683587, + "grad_norm": 8.5625, + "learning_rate": 2.699045718895611e-06, + "loss": 2.5258, + "step": 23186 + }, + { + "epoch": 2.4520939086294415, + "grad_norm": 8.25, + "learning_rate": 2.698875005871614e-06, + "loss": 2.5199, + "step": 23187 + }, + { + "epoch": 2.4521996615905244, + "grad_norm": 8.4375, + "learning_rate": 2.698704291914378e-06, + "loss": 2.3553, + "step": 23188 + }, + { + "epoch": 2.452305414551607, + "grad_norm": 8.1875, + "learning_rate": 2.6985335770247035e-06, + "loss": 2.5731, + "step": 23189 + }, + { + "epoch": 2.4524111675126905, + "grad_norm": 8.9375, + "learning_rate": 2.698362861203393e-06, + "loss": 2.6679, + "step": 23190 + }, + { + "epoch": 2.4525169204737733, + "grad_norm": 8.6875, + "learning_rate": 2.698192144451246e-06, + "loss": 2.6441, + "step": 23191 + }, + { + "epoch": 2.452622673434856, + "grad_norm": 8.0625, + "learning_rate": 2.6980214267690643e-06, + "loss": 2.3873, + "step": 23192 + }, + { + "epoch": 2.452728426395939, + "grad_norm": 8.75, + "learning_rate": 2.6978507081576493e-06, + "loss": 2.648, + "step": 23193 + }, + { + "epoch": 2.452834179357022, + "grad_norm": 9.0, + "learning_rate": 2.697679988617802e-06, + "loss": 2.3361, + "step": 23194 + }, + { + "epoch": 2.452939932318105, + "grad_norm": 8.125, + "learning_rate": 2.6975092681503227e-06, + "loss": 2.4007, + "step": 23195 + }, + { + "epoch": 2.453045685279188, + "grad_norm": 8.4375, + "learning_rate": 2.6973385467560136e-06, + "loss": 2.5149, + "step": 23196 + }, + { + "epoch": 2.453151438240271, + "grad_norm": 8.3125, + "learning_rate": 2.697167824435676e-06, + "loss": 2.6383, + "step": 23197 + }, + { + "epoch": 2.4532571912013537, + "grad_norm": 9.8125, + "learning_rate": 2.69699710119011e-06, + "loss": 2.6602, + "step": 23198 + }, + { + "epoch": 2.4533629441624365, + "grad_norm": 8.375, + "learning_rate": 2.696826377020117e-06, + "loss": 2.3804, + "step": 23199 + }, + { + "epoch": 2.4534686971235193, + "grad_norm": 8.6875, + "learning_rate": 2.696655651926498e-06, + "loss": 2.5047, + "step": 23200 + }, + { + "epoch": 2.453574450084602, + "grad_norm": 8.875, + "learning_rate": 2.696484925910055e-06, + "loss": 2.628, + "step": 23201 + }, + { + "epoch": 2.4536802030456855, + "grad_norm": 8.6875, + "learning_rate": 2.696314198971588e-06, + "loss": 2.6034, + "step": 23202 + }, + { + "epoch": 2.4537859560067683, + "grad_norm": 8.3125, + "learning_rate": 2.6961434711118996e-06, + "loss": 2.3671, + "step": 23203 + }, + { + "epoch": 2.453891708967851, + "grad_norm": 8.25, + "learning_rate": 2.69597274233179e-06, + "loss": 2.3615, + "step": 23204 + }, + { + "epoch": 2.453997461928934, + "grad_norm": 8.0, + "learning_rate": 2.69580201263206e-06, + "loss": 2.6154, + "step": 23205 + }, + { + "epoch": 2.454103214890017, + "grad_norm": 8.125, + "learning_rate": 2.6956312820135113e-06, + "loss": 2.5149, + "step": 23206 + }, + { + "epoch": 2.4542089678510997, + "grad_norm": 8.0625, + "learning_rate": 2.6954605504769453e-06, + "loss": 2.3492, + "step": 23207 + }, + { + "epoch": 2.4543147208121825, + "grad_norm": 8.375, + "learning_rate": 2.6952898180231625e-06, + "loss": 2.3118, + "step": 23208 + }, + { + "epoch": 2.4544204737732658, + "grad_norm": 8.4375, + "learning_rate": 2.695119084652964e-06, + "loss": 2.7927, + "step": 23209 + }, + { + "epoch": 2.4545262267343486, + "grad_norm": 8.1875, + "learning_rate": 2.6949483503671524e-06, + "loss": 2.3993, + "step": 23210 + }, + { + "epoch": 2.4546319796954315, + "grad_norm": 8.375, + "learning_rate": 2.6947776151665273e-06, + "loss": 2.6174, + "step": 23211 + }, + { + "epoch": 2.4547377326565143, + "grad_norm": 8.3125, + "learning_rate": 2.6946068790518914e-06, + "loss": 2.7567, + "step": 23212 + }, + { + "epoch": 2.454843485617597, + "grad_norm": 8.0625, + "learning_rate": 2.6944361420240443e-06, + "loss": 2.5284, + "step": 23213 + }, + { + "epoch": 2.4549492385786804, + "grad_norm": 8.5625, + "learning_rate": 2.6942654040837877e-06, + "loss": 2.5163, + "step": 23214 + }, + { + "epoch": 2.4550549915397633, + "grad_norm": 8.0625, + "learning_rate": 2.6940946652319237e-06, + "loss": 2.3998, + "step": 23215 + }, + { + "epoch": 2.455160744500846, + "grad_norm": 8.4375, + "learning_rate": 2.693923925469253e-06, + "loss": 2.4963, + "step": 23216 + }, + { + "epoch": 2.455266497461929, + "grad_norm": 8.375, + "learning_rate": 2.693753184796575e-06, + "loss": 2.4949, + "step": 23217 + }, + { + "epoch": 2.4553722504230118, + "grad_norm": 8.3125, + "learning_rate": 2.693582443214694e-06, + "loss": 2.3936, + "step": 23218 + }, + { + "epoch": 2.4554780033840946, + "grad_norm": 8.9375, + "learning_rate": 2.6934117007244093e-06, + "loss": 2.6298, + "step": 23219 + }, + { + "epoch": 2.4555837563451774, + "grad_norm": 9.1875, + "learning_rate": 2.6932409573265233e-06, + "loss": 2.4623, + "step": 23220 + }, + { + "epoch": 2.4556895093062607, + "grad_norm": 9.0, + "learning_rate": 2.6930702130218356e-06, + "loss": 2.6859, + "step": 23221 + }, + { + "epoch": 2.4557952622673436, + "grad_norm": 8.0625, + "learning_rate": 2.6928994678111488e-06, + "loss": 2.4397, + "step": 23222 + }, + { + "epoch": 2.4559010152284264, + "grad_norm": 8.4375, + "learning_rate": 2.6927287216952637e-06, + "loss": 2.6224, + "step": 23223 + }, + { + "epoch": 2.4560067681895092, + "grad_norm": 7.625, + "learning_rate": 2.6925579746749807e-06, + "loss": 2.3431, + "step": 23224 + }, + { + "epoch": 2.456112521150592, + "grad_norm": 7.8125, + "learning_rate": 2.692387226751103e-06, + "loss": 2.5187, + "step": 23225 + }, + { + "epoch": 2.4562182741116754, + "grad_norm": 8.375, + "learning_rate": 2.6922164779244308e-06, + "loss": 2.5125, + "step": 23226 + }, + { + "epoch": 2.456324027072758, + "grad_norm": 8.5, + "learning_rate": 2.6920457281957642e-06, + "loss": 2.6884, + "step": 23227 + }, + { + "epoch": 2.456429780033841, + "grad_norm": 8.3125, + "learning_rate": 2.691874977565906e-06, + "loss": 2.443, + "step": 23228 + }, + { + "epoch": 2.456535532994924, + "grad_norm": 8.125, + "learning_rate": 2.6917042260356574e-06, + "loss": 2.3901, + "step": 23229 + }, + { + "epoch": 2.4566412859560067, + "grad_norm": 8.5625, + "learning_rate": 2.6915334736058185e-06, + "loss": 2.5922, + "step": 23230 + }, + { + "epoch": 2.4567470389170896, + "grad_norm": 8.5625, + "learning_rate": 2.691362720277192e-06, + "loss": 2.4591, + "step": 23231 + }, + { + "epoch": 2.4568527918781724, + "grad_norm": 8.0625, + "learning_rate": 2.6911919660505785e-06, + "loss": 2.4691, + "step": 23232 + }, + { + "epoch": 2.4569585448392557, + "grad_norm": 8.625, + "learning_rate": 2.691021210926779e-06, + "loss": 2.5919, + "step": 23233 + }, + { + "epoch": 2.4570642978003385, + "grad_norm": 8.625, + "learning_rate": 2.6908504549065954e-06, + "loss": 2.5897, + "step": 23234 + }, + { + "epoch": 2.4571700507614214, + "grad_norm": 8.4375, + "learning_rate": 2.690679697990829e-06, + "loss": 2.4405, + "step": 23235 + }, + { + "epoch": 2.457275803722504, + "grad_norm": 8.375, + "learning_rate": 2.6905089401802797e-06, + "loss": 2.5143, + "step": 23236 + }, + { + "epoch": 2.457381556683587, + "grad_norm": 8.3125, + "learning_rate": 2.6903381814757503e-06, + "loss": 2.2, + "step": 23237 + }, + { + "epoch": 2.45748730964467, + "grad_norm": 8.5, + "learning_rate": 2.690167421878042e-06, + "loss": 2.575, + "step": 23238 + }, + { + "epoch": 2.4575930626057527, + "grad_norm": 8.125, + "learning_rate": 2.6899966613879556e-06, + "loss": 2.3308, + "step": 23239 + }, + { + "epoch": 2.457698815566836, + "grad_norm": 8.5, + "learning_rate": 2.6898259000062926e-06, + "loss": 2.3156, + "step": 23240 + }, + { + "epoch": 2.457804568527919, + "grad_norm": 8.0, + "learning_rate": 2.689655137733854e-06, + "loss": 2.4838, + "step": 23241 + }, + { + "epoch": 2.4579103214890017, + "grad_norm": 8.6875, + "learning_rate": 2.689484374571442e-06, + "loss": 2.3196, + "step": 23242 + }, + { + "epoch": 2.4580160744500845, + "grad_norm": 8.125, + "learning_rate": 2.689313610519857e-06, + "loss": 2.2954, + "step": 23243 + }, + { + "epoch": 2.4581218274111674, + "grad_norm": 8.125, + "learning_rate": 2.6891428455799e-06, + "loss": 2.5533, + "step": 23244 + }, + { + "epoch": 2.4582275803722506, + "grad_norm": 8.5, + "learning_rate": 2.6889720797523737e-06, + "loss": 2.5144, + "step": 23245 + }, + { + "epoch": 2.4583333333333335, + "grad_norm": 8.5, + "learning_rate": 2.6888013130380784e-06, + "loss": 2.203, + "step": 23246 + }, + { + "epoch": 2.4584390862944163, + "grad_norm": 8.375, + "learning_rate": 2.6886305454378166e-06, + "loss": 2.5347, + "step": 23247 + }, + { + "epoch": 2.458544839255499, + "grad_norm": 8.375, + "learning_rate": 2.6884597769523877e-06, + "loss": 2.6074, + "step": 23248 + }, + { + "epoch": 2.458650592216582, + "grad_norm": 8.0625, + "learning_rate": 2.6882890075825945e-06, + "loss": 2.4543, + "step": 23249 + }, + { + "epoch": 2.458756345177665, + "grad_norm": 8.1875, + "learning_rate": 2.6881182373292384e-06, + "loss": 2.3893, + "step": 23250 + }, + { + "epoch": 2.4588620981387477, + "grad_norm": 8.3125, + "learning_rate": 2.6879474661931203e-06, + "loss": 2.4674, + "step": 23251 + }, + { + "epoch": 2.458967851099831, + "grad_norm": 8.25, + "learning_rate": 2.687776694175041e-06, + "loss": 2.4673, + "step": 23252 + }, + { + "epoch": 2.459073604060914, + "grad_norm": 8.1875, + "learning_rate": 2.6876059212758025e-06, + "loss": 2.2189, + "step": 23253 + }, + { + "epoch": 2.4591793570219966, + "grad_norm": 8.5, + "learning_rate": 2.6874351474962067e-06, + "loss": 2.4689, + "step": 23254 + }, + { + "epoch": 2.4592851099830795, + "grad_norm": 8.875, + "learning_rate": 2.6872643728370545e-06, + "loss": 2.2468, + "step": 23255 + }, + { + "epoch": 2.4593908629441623, + "grad_norm": 8.375, + "learning_rate": 2.6870935972991462e-06, + "loss": 2.6157, + "step": 23256 + }, + { + "epoch": 2.459496615905245, + "grad_norm": 8.5, + "learning_rate": 2.6869228208832854e-06, + "loss": 2.6242, + "step": 23257 + }, + { + "epoch": 2.4596023688663284, + "grad_norm": 8.125, + "learning_rate": 2.6867520435902715e-06, + "loss": 2.3911, + "step": 23258 + }, + { + "epoch": 2.4597081218274113, + "grad_norm": 8.3125, + "learning_rate": 2.6865812654209067e-06, + "loss": 2.6368, + "step": 23259 + }, + { + "epoch": 2.459813874788494, + "grad_norm": 9.0, + "learning_rate": 2.6864104863759928e-06, + "loss": 2.5622, + "step": 23260 + }, + { + "epoch": 2.459919627749577, + "grad_norm": 8.125, + "learning_rate": 2.6862397064563306e-06, + "loss": 2.4616, + "step": 23261 + }, + { + "epoch": 2.46002538071066, + "grad_norm": 8.5, + "learning_rate": 2.6860689256627205e-06, + "loss": 2.6933, + "step": 23262 + }, + { + "epoch": 2.4601311336717426, + "grad_norm": 8.8125, + "learning_rate": 2.6858981439959665e-06, + "loss": 2.6562, + "step": 23263 + }, + { + "epoch": 2.460236886632826, + "grad_norm": 8.75, + "learning_rate": 2.685727361456868e-06, + "loss": 2.6556, + "step": 23264 + }, + { + "epoch": 2.4603426395939088, + "grad_norm": 8.5625, + "learning_rate": 2.685556578046227e-06, + "loss": 2.8297, + "step": 23265 + }, + { + "epoch": 2.4604483925549916, + "grad_norm": 8.5, + "learning_rate": 2.6853857937648447e-06, + "loss": 2.6971, + "step": 23266 + }, + { + "epoch": 2.4605541455160744, + "grad_norm": 8.3125, + "learning_rate": 2.6852150086135233e-06, + "loss": 2.5398, + "step": 23267 + }, + { + "epoch": 2.4606598984771573, + "grad_norm": 7.96875, + "learning_rate": 2.6850442225930622e-06, + "loss": 2.4836, + "step": 23268 + }, + { + "epoch": 2.46076565143824, + "grad_norm": 8.5, + "learning_rate": 2.6848734357042653e-06, + "loss": 2.6018, + "step": 23269 + }, + { + "epoch": 2.460871404399323, + "grad_norm": 8.5, + "learning_rate": 2.684702647947933e-06, + "loss": 2.7861, + "step": 23270 + }, + { + "epoch": 2.4609771573604062, + "grad_norm": 8.125, + "learning_rate": 2.6845318593248663e-06, + "loss": 2.5917, + "step": 23271 + }, + { + "epoch": 2.461082910321489, + "grad_norm": 8.375, + "learning_rate": 2.6843610698358667e-06, + "loss": 2.3262, + "step": 23272 + }, + { + "epoch": 2.461188663282572, + "grad_norm": 8.625, + "learning_rate": 2.684190279481737e-06, + "loss": 2.5701, + "step": 23273 + }, + { + "epoch": 2.4612944162436547, + "grad_norm": 7.90625, + "learning_rate": 2.6840194882632773e-06, + "loss": 2.5575, + "step": 23274 + }, + { + "epoch": 2.4614001692047376, + "grad_norm": 9.125, + "learning_rate": 2.6838486961812893e-06, + "loss": 2.7192, + "step": 23275 + }, + { + "epoch": 2.461505922165821, + "grad_norm": 8.5, + "learning_rate": 2.6836779032365753e-06, + "loss": 2.4687, + "step": 23276 + }, + { + "epoch": 2.4616116751269037, + "grad_norm": 8.375, + "learning_rate": 2.6835071094299354e-06, + "loss": 2.4283, + "step": 23277 + }, + { + "epoch": 2.4617174280879865, + "grad_norm": 8.125, + "learning_rate": 2.683336314762171e-06, + "loss": 2.5595, + "step": 23278 + }, + { + "epoch": 2.4618231810490694, + "grad_norm": 8.5, + "learning_rate": 2.6831655192340856e-06, + "loss": 2.5837, + "step": 23279 + }, + { + "epoch": 2.4619289340101522, + "grad_norm": 8.25, + "learning_rate": 2.6829947228464782e-06, + "loss": 2.4941, + "step": 23280 + }, + { + "epoch": 2.462034686971235, + "grad_norm": 8.875, + "learning_rate": 2.682823925600151e-06, + "loss": 2.6866, + "step": 23281 + }, + { + "epoch": 2.462140439932318, + "grad_norm": 8.5625, + "learning_rate": 2.6826531274959074e-06, + "loss": 2.6334, + "step": 23282 + }, + { + "epoch": 2.462246192893401, + "grad_norm": 8.5625, + "learning_rate": 2.682482328534547e-06, + "loss": 2.4066, + "step": 23283 + }, + { + "epoch": 2.462351945854484, + "grad_norm": 9.5, + "learning_rate": 2.6823115287168707e-06, + "loss": 2.9435, + "step": 23284 + }, + { + "epoch": 2.462457698815567, + "grad_norm": 8.4375, + "learning_rate": 2.6821407280436815e-06, + "loss": 2.5321, + "step": 23285 + }, + { + "epoch": 2.4625634517766497, + "grad_norm": 8.25, + "learning_rate": 2.681969926515781e-06, + "loss": 2.4309, + "step": 23286 + }, + { + "epoch": 2.4626692047377325, + "grad_norm": 13.25, + "learning_rate": 2.681799124133969e-06, + "loss": 2.5244, + "step": 23287 + }, + { + "epoch": 2.4627749576988154, + "grad_norm": 8.3125, + "learning_rate": 2.681628320899049e-06, + "loss": 2.5445, + "step": 23288 + }, + { + "epoch": 2.4628807106598987, + "grad_norm": 8.375, + "learning_rate": 2.6814575168118206e-06, + "loss": 2.5277, + "step": 23289 + }, + { + "epoch": 2.4629864636209815, + "grad_norm": 8.4375, + "learning_rate": 2.6812867118730875e-06, + "loss": 2.4947, + "step": 23290 + }, + { + "epoch": 2.4630922165820643, + "grad_norm": 7.9375, + "learning_rate": 2.681115906083649e-06, + "loss": 2.4402, + "step": 23291 + }, + { + "epoch": 2.463197969543147, + "grad_norm": 8.625, + "learning_rate": 2.680945099444308e-06, + "loss": 2.6756, + "step": 23292 + }, + { + "epoch": 2.46330372250423, + "grad_norm": 8.875, + "learning_rate": 2.680774291955866e-06, + "loss": 2.4893, + "step": 23293 + }, + { + "epoch": 2.463409475465313, + "grad_norm": 8.4375, + "learning_rate": 2.680603483619123e-06, + "loss": 2.4722, + "step": 23294 + }, + { + "epoch": 2.463515228426396, + "grad_norm": 7.875, + "learning_rate": 2.680432674434883e-06, + "loss": 2.5306, + "step": 23295 + }, + { + "epoch": 2.463620981387479, + "grad_norm": 8.5625, + "learning_rate": 2.6802618644039463e-06, + "loss": 2.579, + "step": 23296 + }, + { + "epoch": 2.463726734348562, + "grad_norm": 8.0, + "learning_rate": 2.680091053527113e-06, + "loss": 2.4473, + "step": 23297 + }, + { + "epoch": 2.4638324873096447, + "grad_norm": 9.1875, + "learning_rate": 2.6799202418051866e-06, + "loss": 2.7159, + "step": 23298 + }, + { + "epoch": 2.4639382402707275, + "grad_norm": 9.5625, + "learning_rate": 2.679749429238969e-06, + "loss": 2.5501, + "step": 23299 + }, + { + "epoch": 2.4640439932318103, + "grad_norm": 8.25, + "learning_rate": 2.67957861582926e-06, + "loss": 2.4195, + "step": 23300 + }, + { + "epoch": 2.464149746192893, + "grad_norm": 9.4375, + "learning_rate": 2.679407801576862e-06, + "loss": 2.453, + "step": 23301 + }, + { + "epoch": 2.4642554991539765, + "grad_norm": 8.5, + "learning_rate": 2.6792369864825774e-06, + "loss": 2.6806, + "step": 23302 + }, + { + "epoch": 2.4643612521150593, + "grad_norm": 8.3125, + "learning_rate": 2.679066170547206e-06, + "loss": 2.4469, + "step": 23303 + }, + { + "epoch": 2.464467005076142, + "grad_norm": 9.375, + "learning_rate": 2.6788953537715507e-06, + "loss": 2.8136, + "step": 23304 + }, + { + "epoch": 2.464572758037225, + "grad_norm": 8.5, + "learning_rate": 2.678724536156413e-06, + "loss": 2.8283, + "step": 23305 + }, + { + "epoch": 2.464678510998308, + "grad_norm": 8.3125, + "learning_rate": 2.6785537177025933e-06, + "loss": 2.5731, + "step": 23306 + }, + { + "epoch": 2.464784263959391, + "grad_norm": 8.625, + "learning_rate": 2.678382898410894e-06, + "loss": 2.6645, + "step": 23307 + }, + { + "epoch": 2.464890016920474, + "grad_norm": 8.6875, + "learning_rate": 2.6782120782821175e-06, + "loss": 2.3988, + "step": 23308 + }, + { + "epoch": 2.4649957698815568, + "grad_norm": 8.125, + "learning_rate": 2.6780412573170644e-06, + "loss": 2.4783, + "step": 23309 + }, + { + "epoch": 2.4651015228426396, + "grad_norm": 8.1875, + "learning_rate": 2.677870435516536e-06, + "loss": 2.6756, + "step": 23310 + }, + { + "epoch": 2.4652072758037225, + "grad_norm": 8.625, + "learning_rate": 2.6776996128813353e-06, + "loss": 2.7224, + "step": 23311 + }, + { + "epoch": 2.4653130287648053, + "grad_norm": 8.375, + "learning_rate": 2.6775287894122627e-06, + "loss": 2.4617, + "step": 23312 + }, + { + "epoch": 2.465418781725888, + "grad_norm": 8.75, + "learning_rate": 2.6773579651101196e-06, + "loss": 2.662, + "step": 23313 + }, + { + "epoch": 2.4655245346869714, + "grad_norm": 7.8125, + "learning_rate": 2.6771871399757087e-06, + "loss": 2.3796, + "step": 23314 + }, + { + "epoch": 2.4656302876480543, + "grad_norm": 8.25, + "learning_rate": 2.6770163140098303e-06, + "loss": 2.494, + "step": 23315 + }, + { + "epoch": 2.465736040609137, + "grad_norm": 8.1875, + "learning_rate": 2.6768454872132866e-06, + "loss": 2.5232, + "step": 23316 + }, + { + "epoch": 2.46584179357022, + "grad_norm": 8.4375, + "learning_rate": 2.67667465958688e-06, + "loss": 2.5422, + "step": 23317 + }, + { + "epoch": 2.4659475465313028, + "grad_norm": 8.125, + "learning_rate": 2.676503831131412e-06, + "loss": 2.3285, + "step": 23318 + }, + { + "epoch": 2.4660532994923856, + "grad_norm": 8.3125, + "learning_rate": 2.6763330018476825e-06, + "loss": 2.4685, + "step": 23319 + }, + { + "epoch": 2.4661590524534684, + "grad_norm": 8.0, + "learning_rate": 2.6761621717364956e-06, + "loss": 2.6232, + "step": 23320 + }, + { + "epoch": 2.4662648054145517, + "grad_norm": 8.4375, + "learning_rate": 2.675991340798651e-06, + "loss": 2.272, + "step": 23321 + }, + { + "epoch": 2.4663705583756346, + "grad_norm": 8.5, + "learning_rate": 2.6758205090349503e-06, + "loss": 2.5578, + "step": 23322 + }, + { + "epoch": 2.4664763113367174, + "grad_norm": 8.625, + "learning_rate": 2.675649676446197e-06, + "loss": 2.6024, + "step": 23323 + }, + { + "epoch": 2.4665820642978002, + "grad_norm": 8.8125, + "learning_rate": 2.6754788430331906e-06, + "loss": 2.4837, + "step": 23324 + }, + { + "epoch": 2.466687817258883, + "grad_norm": 8.4375, + "learning_rate": 2.6753080087967347e-06, + "loss": 2.6366, + "step": 23325 + }, + { + "epoch": 2.4667935702199664, + "grad_norm": 8.1875, + "learning_rate": 2.67513717373763e-06, + "loss": 2.4727, + "step": 23326 + }, + { + "epoch": 2.466899323181049, + "grad_norm": 9.0, + "learning_rate": 2.6749663378566777e-06, + "loss": 2.7899, + "step": 23327 + }, + { + "epoch": 2.467005076142132, + "grad_norm": 9.5, + "learning_rate": 2.67479550115468e-06, + "loss": 2.7931, + "step": 23328 + }, + { + "epoch": 2.467110829103215, + "grad_norm": 8.625, + "learning_rate": 2.6746246636324384e-06, + "loss": 2.6531, + "step": 23329 + }, + { + "epoch": 2.4672165820642977, + "grad_norm": 8.0625, + "learning_rate": 2.6744538252907547e-06, + "loss": 2.4201, + "step": 23330 + }, + { + "epoch": 2.4673223350253806, + "grad_norm": 9.0, + "learning_rate": 2.6742829861304315e-06, + "loss": 2.7013, + "step": 23331 + }, + { + "epoch": 2.4674280879864634, + "grad_norm": 8.75, + "learning_rate": 2.674112146152268e-06, + "loss": 2.6025, + "step": 23332 + }, + { + "epoch": 2.4675338409475467, + "grad_norm": 8.875, + "learning_rate": 2.6739413053570677e-06, + "loss": 2.5371, + "step": 23333 + }, + { + "epoch": 2.4676395939086295, + "grad_norm": 8.6875, + "learning_rate": 2.673770463745633e-06, + "loss": 2.6328, + "step": 23334 + }, + { + "epoch": 2.4677453468697124, + "grad_norm": 8.3125, + "learning_rate": 2.673599621318764e-06, + "loss": 2.483, + "step": 23335 + }, + { + "epoch": 2.467851099830795, + "grad_norm": 8.6875, + "learning_rate": 2.6734287780772626e-06, + "loss": 2.477, + "step": 23336 + }, + { + "epoch": 2.467956852791878, + "grad_norm": 8.8125, + "learning_rate": 2.6732579340219318e-06, + "loss": 2.5204, + "step": 23337 + }, + { + "epoch": 2.4680626057529613, + "grad_norm": 8.6875, + "learning_rate": 2.6730870891535717e-06, + "loss": 2.7349, + "step": 23338 + }, + { + "epoch": 2.468168358714044, + "grad_norm": 8.3125, + "learning_rate": 2.6729162434729845e-06, + "loss": 2.5731, + "step": 23339 + }, + { + "epoch": 2.468274111675127, + "grad_norm": 8.3125, + "learning_rate": 2.6727453969809723e-06, + "loss": 2.4445, + "step": 23340 + }, + { + "epoch": 2.46837986463621, + "grad_norm": 8.0625, + "learning_rate": 2.6725745496783365e-06, + "loss": 2.5803, + "step": 23341 + }, + { + "epoch": 2.4684856175972927, + "grad_norm": 8.3125, + "learning_rate": 2.672403701565879e-06, + "loss": 2.5839, + "step": 23342 + }, + { + "epoch": 2.4685913705583755, + "grad_norm": 8.1875, + "learning_rate": 2.672232852644402e-06, + "loss": 2.3683, + "step": 23343 + }, + { + "epoch": 2.4686971235194584, + "grad_norm": 8.5625, + "learning_rate": 2.6720620029147065e-06, + "loss": 2.3873, + "step": 23344 + }, + { + "epoch": 2.4688028764805416, + "grad_norm": 8.5, + "learning_rate": 2.6718911523775936e-06, + "loss": 2.5827, + "step": 23345 + }, + { + "epoch": 2.4689086294416245, + "grad_norm": 7.75, + "learning_rate": 2.6717203010338665e-06, + "loss": 2.375, + "step": 23346 + }, + { + "epoch": 2.4690143824027073, + "grad_norm": 8.0625, + "learning_rate": 2.6715494488843265e-06, + "loss": 2.2172, + "step": 23347 + }, + { + "epoch": 2.46912013536379, + "grad_norm": 8.75, + "learning_rate": 2.6713785959297744e-06, + "loss": 2.4586, + "step": 23348 + }, + { + "epoch": 2.469225888324873, + "grad_norm": 8.625, + "learning_rate": 2.6712077421710133e-06, + "loss": 2.5145, + "step": 23349 + }, + { + "epoch": 2.469331641285956, + "grad_norm": 8.125, + "learning_rate": 2.671036887608844e-06, + "loss": 2.6381, + "step": 23350 + }, + { + "epoch": 2.4694373942470387, + "grad_norm": 8.1875, + "learning_rate": 2.670866032244067e-06, + "loss": 2.5713, + "step": 23351 + }, + { + "epoch": 2.469543147208122, + "grad_norm": 8.75, + "learning_rate": 2.6706951760774874e-06, + "loss": 2.7037, + "step": 23352 + }, + { + "epoch": 2.469648900169205, + "grad_norm": 8.5625, + "learning_rate": 2.670524319109905e-06, + "loss": 2.5598, + "step": 23353 + }, + { + "epoch": 2.4697546531302876, + "grad_norm": 8.75, + "learning_rate": 2.670353461342121e-06, + "loss": 2.5631, + "step": 23354 + }, + { + "epoch": 2.4698604060913705, + "grad_norm": 7.9375, + "learning_rate": 2.670182602774939e-06, + "loss": 2.4864, + "step": 23355 + }, + { + "epoch": 2.4699661590524533, + "grad_norm": 8.5625, + "learning_rate": 2.6700117434091595e-06, + "loss": 2.464, + "step": 23356 + }, + { + "epoch": 2.4700719120135366, + "grad_norm": 8.5625, + "learning_rate": 2.6698408832455837e-06, + "loss": 2.6909, + "step": 23357 + }, + { + "epoch": 2.4701776649746194, + "grad_norm": 9.0625, + "learning_rate": 2.6696700222850142e-06, + "loss": 2.7221, + "step": 23358 + }, + { + "epoch": 2.4702834179357023, + "grad_norm": 8.25, + "learning_rate": 2.669499160528253e-06, + "loss": 2.4761, + "step": 23359 + }, + { + "epoch": 2.470389170896785, + "grad_norm": 8.25, + "learning_rate": 2.669328297976101e-06, + "loss": 2.703, + "step": 23360 + }, + { + "epoch": 2.470494923857868, + "grad_norm": 8.1875, + "learning_rate": 2.669157434629361e-06, + "loss": 2.3716, + "step": 23361 + }, + { + "epoch": 2.470600676818951, + "grad_norm": 8.5, + "learning_rate": 2.668986570488834e-06, + "loss": 2.3527, + "step": 23362 + }, + { + "epoch": 2.4707064297800336, + "grad_norm": 8.625, + "learning_rate": 2.668815705555323e-06, + "loss": 2.4588, + "step": 23363 + }, + { + "epoch": 2.470812182741117, + "grad_norm": 8.6875, + "learning_rate": 2.6686448398296276e-06, + "loss": 2.5591, + "step": 23364 + }, + { + "epoch": 2.4709179357021998, + "grad_norm": 8.0, + "learning_rate": 2.6684739733125525e-06, + "loss": 2.5584, + "step": 23365 + }, + { + "epoch": 2.4710236886632826, + "grad_norm": 8.3125, + "learning_rate": 2.668303106004897e-06, + "loss": 2.4534, + "step": 23366 + }, + { + "epoch": 2.4711294416243654, + "grad_norm": 8.1875, + "learning_rate": 2.6681322379074635e-06, + "loss": 2.4223, + "step": 23367 + }, + { + "epoch": 2.4712351945854483, + "grad_norm": 8.25, + "learning_rate": 2.6679613690210548e-06, + "loss": 2.364, + "step": 23368 + }, + { + "epoch": 2.4713409475465316, + "grad_norm": 8.375, + "learning_rate": 2.667790499346472e-06, + "loss": 2.4971, + "step": 23369 + }, + { + "epoch": 2.4714467005076144, + "grad_norm": 9.4375, + "learning_rate": 2.667619628884517e-06, + "loss": 2.7954, + "step": 23370 + }, + { + "epoch": 2.4715524534686972, + "grad_norm": 8.8125, + "learning_rate": 2.6674487576359915e-06, + "loss": 2.5736, + "step": 23371 + }, + { + "epoch": 2.47165820642978, + "grad_norm": 8.5, + "learning_rate": 2.667277885601698e-06, + "loss": 2.566, + "step": 23372 + }, + { + "epoch": 2.471763959390863, + "grad_norm": 8.375, + "learning_rate": 2.667107012782437e-06, + "loss": 2.5433, + "step": 23373 + }, + { + "epoch": 2.4718697123519457, + "grad_norm": 8.3125, + "learning_rate": 2.6669361391790117e-06, + "loss": 2.3389, + "step": 23374 + }, + { + "epoch": 2.4719754653130286, + "grad_norm": 8.1875, + "learning_rate": 2.6667652647922234e-06, + "loss": 2.4246, + "step": 23375 + }, + { + "epoch": 2.472081218274112, + "grad_norm": 8.5, + "learning_rate": 2.666594389622873e-06, + "loss": 2.4955, + "step": 23376 + }, + { + "epoch": 2.4721869712351947, + "grad_norm": 8.25, + "learning_rate": 2.6664235136717637e-06, + "loss": 2.5928, + "step": 23377 + }, + { + "epoch": 2.4722927241962775, + "grad_norm": 8.8125, + "learning_rate": 2.6662526369396974e-06, + "loss": 2.3426, + "step": 23378 + }, + { + "epoch": 2.4723984771573604, + "grad_norm": 9.125, + "learning_rate": 2.6660817594274757e-06, + "loss": 2.6155, + "step": 23379 + }, + { + "epoch": 2.4725042301184432, + "grad_norm": 8.875, + "learning_rate": 2.665910881135899e-06, + "loss": 2.375, + "step": 23380 + }, + { + "epoch": 2.472609983079526, + "grad_norm": 8.5, + "learning_rate": 2.665740002065772e-06, + "loss": 2.4292, + "step": 23381 + }, + { + "epoch": 2.472715736040609, + "grad_norm": 8.4375, + "learning_rate": 2.665569122217894e-06, + "loss": 2.6764, + "step": 23382 + }, + { + "epoch": 2.472821489001692, + "grad_norm": 8.375, + "learning_rate": 2.665398241593068e-06, + "loss": 2.6167, + "step": 23383 + }, + { + "epoch": 2.472927241962775, + "grad_norm": 8.375, + "learning_rate": 2.6652273601920957e-06, + "loss": 2.6193, + "step": 23384 + }, + { + "epoch": 2.473032994923858, + "grad_norm": 8.25, + "learning_rate": 2.665056478015779e-06, + "loss": 2.3309, + "step": 23385 + }, + { + "epoch": 2.4731387478849407, + "grad_norm": 8.1875, + "learning_rate": 2.6648855950649188e-06, + "loss": 2.5315, + "step": 23386 + }, + { + "epoch": 2.4732445008460235, + "grad_norm": 8.5, + "learning_rate": 2.6647147113403194e-06, + "loss": 2.4535, + "step": 23387 + }, + { + "epoch": 2.473350253807107, + "grad_norm": 7.96875, + "learning_rate": 2.6645438268427804e-06, + "loss": 2.4616, + "step": 23388 + }, + { + "epoch": 2.4734560067681897, + "grad_norm": 8.3125, + "learning_rate": 2.6643729415731045e-06, + "loss": 2.5295, + "step": 23389 + }, + { + "epoch": 2.4735617597292725, + "grad_norm": 8.375, + "learning_rate": 2.664202055532094e-06, + "loss": 2.7359, + "step": 23390 + }, + { + "epoch": 2.4736675126903553, + "grad_norm": 8.1875, + "learning_rate": 2.6640311687205504e-06, + "loss": 2.6917, + "step": 23391 + }, + { + "epoch": 2.473773265651438, + "grad_norm": 8.1875, + "learning_rate": 2.6638602811392753e-06, + "loss": 2.5387, + "step": 23392 + }, + { + "epoch": 2.473879018612521, + "grad_norm": 8.8125, + "learning_rate": 2.663689392789071e-06, + "loss": 2.2903, + "step": 23393 + }, + { + "epoch": 2.473984771573604, + "grad_norm": 8.375, + "learning_rate": 2.66351850367074e-06, + "loss": 2.6071, + "step": 23394 + }, + { + "epoch": 2.474090524534687, + "grad_norm": 8.8125, + "learning_rate": 2.663347613785083e-06, + "loss": 2.6131, + "step": 23395 + }, + { + "epoch": 2.47419627749577, + "grad_norm": 8.0625, + "learning_rate": 2.663176723132902e-06, + "loss": 2.4292, + "step": 23396 + }, + { + "epoch": 2.474302030456853, + "grad_norm": 8.1875, + "learning_rate": 2.663005831715e-06, + "loss": 2.1245, + "step": 23397 + }, + { + "epoch": 2.4744077834179357, + "grad_norm": 8.625, + "learning_rate": 2.662834939532178e-06, + "loss": 2.5695, + "step": 23398 + }, + { + "epoch": 2.4745135363790185, + "grad_norm": 8.375, + "learning_rate": 2.662664046585238e-06, + "loss": 2.3602, + "step": 23399 + }, + { + "epoch": 2.4746192893401013, + "grad_norm": 8.75, + "learning_rate": 2.6624931528749826e-06, + "loss": 2.6735, + "step": 23400 + }, + { + "epoch": 2.4747250423011846, + "grad_norm": 8.5, + "learning_rate": 2.6623222584022136e-06, + "loss": 2.6941, + "step": 23401 + }, + { + "epoch": 2.4748307952622675, + "grad_norm": 8.125, + "learning_rate": 2.662151363167732e-06, + "loss": 2.56, + "step": 23402 + }, + { + "epoch": 2.4749365482233503, + "grad_norm": 8.0625, + "learning_rate": 2.6619804671723404e-06, + "loss": 2.6376, + "step": 23403 + }, + { + "epoch": 2.475042301184433, + "grad_norm": 8.3125, + "learning_rate": 2.6618095704168404e-06, + "loss": 2.666, + "step": 23404 + }, + { + "epoch": 2.475148054145516, + "grad_norm": 8.375, + "learning_rate": 2.6616386729020348e-06, + "loss": 2.4101, + "step": 23405 + }, + { + "epoch": 2.475253807106599, + "grad_norm": 8.3125, + "learning_rate": 2.6614677746287255e-06, + "loss": 2.4934, + "step": 23406 + }, + { + "epoch": 2.475359560067682, + "grad_norm": 8.25, + "learning_rate": 2.661296875597713e-06, + "loss": 2.5923, + "step": 23407 + }, + { + "epoch": 2.475465313028765, + "grad_norm": 8.0625, + "learning_rate": 2.661125975809801e-06, + "loss": 2.4239, + "step": 23408 + }, + { + "epoch": 2.4755710659898478, + "grad_norm": 8.125, + "learning_rate": 2.6609550752657903e-06, + "loss": 2.5747, + "step": 23409 + }, + { + "epoch": 2.4756768189509306, + "grad_norm": 8.5, + "learning_rate": 2.6607841739664837e-06, + "loss": 2.5236, + "step": 23410 + }, + { + "epoch": 2.4757825719120135, + "grad_norm": 8.5625, + "learning_rate": 2.660613271912682e-06, + "loss": 2.5122, + "step": 23411 + }, + { + "epoch": 2.4758883248730963, + "grad_norm": 8.625, + "learning_rate": 2.6604423691051878e-06, + "loss": 2.4431, + "step": 23412 + }, + { + "epoch": 2.475994077834179, + "grad_norm": 8.5, + "learning_rate": 2.6602714655448034e-06, + "loss": 2.2448, + "step": 23413 + }, + { + "epoch": 2.4760998307952624, + "grad_norm": 8.5625, + "learning_rate": 2.660100561232332e-06, + "loss": 2.7135, + "step": 23414 + }, + { + "epoch": 2.4762055837563453, + "grad_norm": 7.875, + "learning_rate": 2.6599296561685724e-06, + "loss": 2.6133, + "step": 23415 + }, + { + "epoch": 2.476311336717428, + "grad_norm": 8.125, + "learning_rate": 2.6597587503543287e-06, + "loss": 2.5978, + "step": 23416 + }, + { + "epoch": 2.476417089678511, + "grad_norm": 8.25, + "learning_rate": 2.6595878437904033e-06, + "loss": 2.521, + "step": 23417 + }, + { + "epoch": 2.4765228426395938, + "grad_norm": 7.78125, + "learning_rate": 2.659416936477596e-06, + "loss": 2.4347, + "step": 23418 + }, + { + "epoch": 2.476628595600677, + "grad_norm": 7.8125, + "learning_rate": 2.6592460284167114e-06, + "loss": 2.5355, + "step": 23419 + }, + { + "epoch": 2.47673434856176, + "grad_norm": 8.9375, + "learning_rate": 2.65907511960855e-06, + "loss": 2.6027, + "step": 23420 + }, + { + "epoch": 2.4768401015228427, + "grad_norm": 9.3125, + "learning_rate": 2.658904210053913e-06, + "loss": 2.5455, + "step": 23421 + }, + { + "epoch": 2.4769458544839256, + "grad_norm": 7.71875, + "learning_rate": 2.6587332997536055e-06, + "loss": 2.4972, + "step": 23422 + }, + { + "epoch": 2.4770516074450084, + "grad_norm": 8.5, + "learning_rate": 2.658562388708427e-06, + "loss": 2.4797, + "step": 23423 + }, + { + "epoch": 2.4771573604060912, + "grad_norm": 8.375, + "learning_rate": 2.658391476919179e-06, + "loss": 2.7703, + "step": 23424 + }, + { + "epoch": 2.477263113367174, + "grad_norm": 8.375, + "learning_rate": 2.6582205643866653e-06, + "loss": 2.6748, + "step": 23425 + }, + { + "epoch": 2.4773688663282574, + "grad_norm": 8.3125, + "learning_rate": 2.6580496511116877e-06, + "loss": 2.4763, + "step": 23426 + }, + { + "epoch": 2.47747461928934, + "grad_norm": 8.4375, + "learning_rate": 2.6578787370950467e-06, + "loss": 2.4119, + "step": 23427 + }, + { + "epoch": 2.477580372250423, + "grad_norm": 8.25, + "learning_rate": 2.657707822337546e-06, + "loss": 2.6094, + "step": 23428 + }, + { + "epoch": 2.477686125211506, + "grad_norm": 8.1875, + "learning_rate": 2.657536906839987e-06, + "loss": 2.6913, + "step": 23429 + }, + { + "epoch": 2.4777918781725887, + "grad_norm": 8.25, + "learning_rate": 2.657365990603171e-06, + "loss": 2.6542, + "step": 23430 + }, + { + "epoch": 2.4778976311336716, + "grad_norm": 8.625, + "learning_rate": 2.6571950736279017e-06, + "loss": 2.7235, + "step": 23431 + }, + { + "epoch": 2.478003384094755, + "grad_norm": 8.8125, + "learning_rate": 2.65702415591498e-06, + "loss": 2.6445, + "step": 23432 + }, + { + "epoch": 2.4781091370558377, + "grad_norm": 8.625, + "learning_rate": 2.6568532374652083e-06, + "loss": 2.6611, + "step": 23433 + }, + { + "epoch": 2.4782148900169205, + "grad_norm": 8.8125, + "learning_rate": 2.656682318279388e-06, + "loss": 2.6029, + "step": 23434 + }, + { + "epoch": 2.4783206429780034, + "grad_norm": 8.5625, + "learning_rate": 2.656511398358322e-06, + "loss": 2.4302, + "step": 23435 + }, + { + "epoch": 2.478426395939086, + "grad_norm": 8.1875, + "learning_rate": 2.6563404777028117e-06, + "loss": 2.4848, + "step": 23436 + }, + { + "epoch": 2.478532148900169, + "grad_norm": 8.5625, + "learning_rate": 2.65616955631366e-06, + "loss": 2.4496, + "step": 23437 + }, + { + "epoch": 2.4786379018612523, + "grad_norm": 8.6875, + "learning_rate": 2.6559986341916683e-06, + "loss": 2.6682, + "step": 23438 + }, + { + "epoch": 2.478743654822335, + "grad_norm": 9.6875, + "learning_rate": 2.655827711337638e-06, + "loss": 2.6522, + "step": 23439 + }, + { + "epoch": 2.478849407783418, + "grad_norm": 8.9375, + "learning_rate": 2.6556567877523727e-06, + "loss": 2.7916, + "step": 23440 + }, + { + "epoch": 2.478955160744501, + "grad_norm": 8.625, + "learning_rate": 2.6554858634366737e-06, + "loss": 2.6111, + "step": 23441 + }, + { + "epoch": 2.4790609137055837, + "grad_norm": 8.5625, + "learning_rate": 2.6553149383913428e-06, + "loss": 2.5919, + "step": 23442 + }, + { + "epoch": 2.4791666666666665, + "grad_norm": 9.0, + "learning_rate": 2.6551440126171827e-06, + "loss": 2.6349, + "step": 23443 + }, + { + "epoch": 2.4792724196277494, + "grad_norm": 8.125, + "learning_rate": 2.6549730861149953e-06, + "loss": 2.5695, + "step": 23444 + }, + { + "epoch": 2.4793781725888326, + "grad_norm": 8.4375, + "learning_rate": 2.654802158885583e-06, + "loss": 2.3279, + "step": 23445 + }, + { + "epoch": 2.4794839255499155, + "grad_norm": 8.4375, + "learning_rate": 2.6546312309297465e-06, + "loss": 2.3638, + "step": 23446 + }, + { + "epoch": 2.4795896785109983, + "grad_norm": 8.3125, + "learning_rate": 2.6544603022482894e-06, + "loss": 2.2786, + "step": 23447 + }, + { + "epoch": 2.479695431472081, + "grad_norm": 8.125, + "learning_rate": 2.654289372842013e-06, + "loss": 2.496, + "step": 23448 + }, + { + "epoch": 2.479801184433164, + "grad_norm": 8.25, + "learning_rate": 2.6541184427117206e-06, + "loss": 2.6182, + "step": 23449 + }, + { + "epoch": 2.4799069373942473, + "grad_norm": 8.5, + "learning_rate": 2.653947511858212e-06, + "loss": 2.4551, + "step": 23450 + }, + { + "epoch": 2.48001269035533, + "grad_norm": 8.5, + "learning_rate": 2.6537765802822916e-06, + "loss": 2.63, + "step": 23451 + }, + { + "epoch": 2.480118443316413, + "grad_norm": 8.5625, + "learning_rate": 2.653605647984761e-06, + "loss": 2.4546, + "step": 23452 + }, + { + "epoch": 2.480224196277496, + "grad_norm": 8.3125, + "learning_rate": 2.6534347149664208e-06, + "loss": 2.5541, + "step": 23453 + }, + { + "epoch": 2.4803299492385786, + "grad_norm": 8.0, + "learning_rate": 2.653263781228075e-06, + "loss": 2.5078, + "step": 23454 + }, + { + "epoch": 2.4804357021996615, + "grad_norm": 8.625, + "learning_rate": 2.653092846770525e-06, + "loss": 2.7053, + "step": 23455 + }, + { + "epoch": 2.4805414551607443, + "grad_norm": 8.3125, + "learning_rate": 2.652921911594572e-06, + "loss": 2.4263, + "step": 23456 + }, + { + "epoch": 2.4806472081218276, + "grad_norm": 8.9375, + "learning_rate": 2.6527509757010193e-06, + "loss": 2.5515, + "step": 23457 + }, + { + "epoch": 2.4807529610829104, + "grad_norm": 9.0, + "learning_rate": 2.6525800390906693e-06, + "loss": 2.5971, + "step": 23458 + }, + { + "epoch": 2.4808587140439933, + "grad_norm": 8.625, + "learning_rate": 2.6524091017643232e-06, + "loss": 2.39, + "step": 23459 + }, + { + "epoch": 2.480964467005076, + "grad_norm": 8.1875, + "learning_rate": 2.652238163722784e-06, + "loss": 2.4187, + "step": 23460 + }, + { + "epoch": 2.481070219966159, + "grad_norm": 8.625, + "learning_rate": 2.6520672249668526e-06, + "loss": 2.5787, + "step": 23461 + }, + { + "epoch": 2.481175972927242, + "grad_norm": 8.125, + "learning_rate": 2.6518962854973324e-06, + "loss": 2.4149, + "step": 23462 + }, + { + "epoch": 2.4812817258883246, + "grad_norm": 8.25, + "learning_rate": 2.651725345315025e-06, + "loss": 2.6258, + "step": 23463 + }, + { + "epoch": 2.481387478849408, + "grad_norm": 8.4375, + "learning_rate": 2.651554404420733e-06, + "loss": 2.2165, + "step": 23464 + }, + { + "epoch": 2.4814932318104908, + "grad_norm": 8.9375, + "learning_rate": 2.6513834628152574e-06, + "loss": 2.516, + "step": 23465 + }, + { + "epoch": 2.4815989847715736, + "grad_norm": 8.6875, + "learning_rate": 2.651212520499401e-06, + "loss": 2.5556, + "step": 23466 + }, + { + "epoch": 2.4817047377326564, + "grad_norm": 8.625, + "learning_rate": 2.6510415774739667e-06, + "loss": 2.6444, + "step": 23467 + }, + { + "epoch": 2.4818104906937393, + "grad_norm": 8.75, + "learning_rate": 2.650870633739756e-06, + "loss": 2.3903, + "step": 23468 + }, + { + "epoch": 2.4819162436548226, + "grad_norm": 8.0, + "learning_rate": 2.650699689297571e-06, + "loss": 2.6114, + "step": 23469 + }, + { + "epoch": 2.4820219966159054, + "grad_norm": 8.4375, + "learning_rate": 2.650528744148214e-06, + "loss": 2.7024, + "step": 23470 + }, + { + "epoch": 2.4821277495769882, + "grad_norm": 8.125, + "learning_rate": 2.6503577982924878e-06, + "loss": 2.5322, + "step": 23471 + }, + { + "epoch": 2.482233502538071, + "grad_norm": 8.4375, + "learning_rate": 2.650186851731193e-06, + "loss": 2.6084, + "step": 23472 + }, + { + "epoch": 2.482339255499154, + "grad_norm": 8.375, + "learning_rate": 2.650015904465133e-06, + "loss": 2.5885, + "step": 23473 + }, + { + "epoch": 2.4824450084602367, + "grad_norm": 8.625, + "learning_rate": 2.6498449564951094e-06, + "loss": 2.7245, + "step": 23474 + }, + { + "epoch": 2.4825507614213196, + "grad_norm": 8.25, + "learning_rate": 2.649674007821925e-06, + "loss": 2.5488, + "step": 23475 + }, + { + "epoch": 2.482656514382403, + "grad_norm": 8.0625, + "learning_rate": 2.6495030584463814e-06, + "loss": 2.5375, + "step": 23476 + }, + { + "epoch": 2.4827622673434857, + "grad_norm": 8.625, + "learning_rate": 2.649332108369282e-06, + "loss": 2.6229, + "step": 23477 + }, + { + "epoch": 2.4828680203045685, + "grad_norm": 8.6875, + "learning_rate": 2.649161157591427e-06, + "loss": 2.5604, + "step": 23478 + }, + { + "epoch": 2.4829737732656514, + "grad_norm": 8.3125, + "learning_rate": 2.648990206113621e-06, + "loss": 2.602, + "step": 23479 + }, + { + "epoch": 2.4830795262267342, + "grad_norm": 8.8125, + "learning_rate": 2.648819253936664e-06, + "loss": 2.7026, + "step": 23480 + }, + { + "epoch": 2.4831852791878175, + "grad_norm": 8.75, + "learning_rate": 2.6486483010613588e-06, + "loss": 2.5214, + "step": 23481 + }, + { + "epoch": 2.4832910321489003, + "grad_norm": 8.0, + "learning_rate": 2.648477347488509e-06, + "loss": 2.4488, + "step": 23482 + }, + { + "epoch": 2.483396785109983, + "grad_norm": 8.6875, + "learning_rate": 2.6483063932189147e-06, + "loss": 2.4206, + "step": 23483 + }, + { + "epoch": 2.483502538071066, + "grad_norm": 8.4375, + "learning_rate": 2.64813543825338e-06, + "loss": 2.4459, + "step": 23484 + }, + { + "epoch": 2.483608291032149, + "grad_norm": 8.5625, + "learning_rate": 2.6479644825927054e-06, + "loss": 2.4992, + "step": 23485 + }, + { + "epoch": 2.4837140439932317, + "grad_norm": 8.125, + "learning_rate": 2.6477935262376943e-06, + "loss": 2.5688, + "step": 23486 + }, + { + "epoch": 2.4838197969543145, + "grad_norm": 8.8125, + "learning_rate": 2.6476225691891487e-06, + "loss": 2.6557, + "step": 23487 + }, + { + "epoch": 2.483925549915398, + "grad_norm": 8.5625, + "learning_rate": 2.647451611447871e-06, + "loss": 2.6299, + "step": 23488 + }, + { + "epoch": 2.4840313028764807, + "grad_norm": 8.5, + "learning_rate": 2.6472806530146635e-06, + "loss": 2.5799, + "step": 23489 + }, + { + "epoch": 2.4841370558375635, + "grad_norm": 8.5, + "learning_rate": 2.6471096938903273e-06, + "loss": 2.5316, + "step": 23490 + }, + { + "epoch": 2.4842428087986463, + "grad_norm": 11.125, + "learning_rate": 2.646938734075666e-06, + "loss": 2.7236, + "step": 23491 + }, + { + "epoch": 2.484348561759729, + "grad_norm": 8.625, + "learning_rate": 2.64676777357148e-06, + "loss": 2.3576, + "step": 23492 + }, + { + "epoch": 2.484454314720812, + "grad_norm": 8.3125, + "learning_rate": 2.6465968123785747e-06, + "loss": 2.519, + "step": 23493 + }, + { + "epoch": 2.484560067681895, + "grad_norm": 8.75, + "learning_rate": 2.6464258504977493e-06, + "loss": 2.5906, + "step": 23494 + }, + { + "epoch": 2.484665820642978, + "grad_norm": 8.8125, + "learning_rate": 2.6462548879298083e-06, + "loss": 2.6898, + "step": 23495 + }, + { + "epoch": 2.484771573604061, + "grad_norm": 8.5625, + "learning_rate": 2.6460839246755525e-06, + "loss": 2.5365, + "step": 23496 + }, + { + "epoch": 2.484877326565144, + "grad_norm": 8.8125, + "learning_rate": 2.645912960735784e-06, + "loss": 2.493, + "step": 23497 + }, + { + "epoch": 2.4849830795262267, + "grad_norm": 8.8125, + "learning_rate": 2.6457419961113063e-06, + "loss": 2.5959, + "step": 23498 + }, + { + "epoch": 2.4850888324873095, + "grad_norm": 8.75, + "learning_rate": 2.6455710308029215e-06, + "loss": 2.5917, + "step": 23499 + }, + { + "epoch": 2.485194585448393, + "grad_norm": 9.6875, + "learning_rate": 2.6454000648114304e-06, + "loss": 2.5109, + "step": 23500 + }, + { + "epoch": 2.4853003384094756, + "grad_norm": 9.25, + "learning_rate": 2.645229098137636e-06, + "loss": 2.6904, + "step": 23501 + }, + { + "epoch": 2.4854060913705585, + "grad_norm": 9.0625, + "learning_rate": 2.6450581307823415e-06, + "loss": 2.7105, + "step": 23502 + }, + { + "epoch": 2.4855118443316413, + "grad_norm": 8.1875, + "learning_rate": 2.644887162746349e-06, + "loss": 2.4505, + "step": 23503 + }, + { + "epoch": 2.485617597292724, + "grad_norm": 8.3125, + "learning_rate": 2.6447161940304592e-06, + "loss": 2.4438, + "step": 23504 + }, + { + "epoch": 2.485723350253807, + "grad_norm": 8.6875, + "learning_rate": 2.6445452246354763e-06, + "loss": 2.2382, + "step": 23505 + }, + { + "epoch": 2.48582910321489, + "grad_norm": 8.5, + "learning_rate": 2.644374254562202e-06, + "loss": 2.6132, + "step": 23506 + }, + { + "epoch": 2.485934856175973, + "grad_norm": 8.6875, + "learning_rate": 2.6442032838114374e-06, + "loss": 2.5438, + "step": 23507 + }, + { + "epoch": 2.486040609137056, + "grad_norm": 8.375, + "learning_rate": 2.644032312383986e-06, + "loss": 2.6193, + "step": 23508 + }, + { + "epoch": 2.4861463620981388, + "grad_norm": 8.25, + "learning_rate": 2.6438613402806506e-06, + "loss": 2.6239, + "step": 23509 + }, + { + "epoch": 2.4862521150592216, + "grad_norm": 8.4375, + "learning_rate": 2.6436903675022318e-06, + "loss": 2.5752, + "step": 23510 + }, + { + "epoch": 2.4863578680203045, + "grad_norm": 9.1875, + "learning_rate": 2.6435193940495337e-06, + "loss": 2.9327, + "step": 23511 + }, + { + "epoch": 2.4864636209813873, + "grad_norm": 8.5, + "learning_rate": 2.643348419923357e-06, + "loss": 2.6073, + "step": 23512 + }, + { + "epoch": 2.4865693739424706, + "grad_norm": 8.75, + "learning_rate": 2.643177445124505e-06, + "loss": 2.5527, + "step": 23513 + }, + { + "epoch": 2.4866751269035534, + "grad_norm": 8.0625, + "learning_rate": 2.6430064696537807e-06, + "loss": 2.5276, + "step": 23514 + }, + { + "epoch": 2.4867808798646363, + "grad_norm": 8.75, + "learning_rate": 2.642835493511985e-06, + "loss": 2.6763, + "step": 23515 + }, + { + "epoch": 2.486886632825719, + "grad_norm": 8.0625, + "learning_rate": 2.6426645166999204e-06, + "loss": 2.5198, + "step": 23516 + }, + { + "epoch": 2.486992385786802, + "grad_norm": 8.625, + "learning_rate": 2.6424935392183897e-06, + "loss": 2.5176, + "step": 23517 + }, + { + "epoch": 2.4870981387478848, + "grad_norm": 8.375, + "learning_rate": 2.6423225610681956e-06, + "loss": 2.5624, + "step": 23518 + }, + { + "epoch": 2.487203891708968, + "grad_norm": 8.3125, + "learning_rate": 2.642151582250139e-06, + "loss": 2.5249, + "step": 23519 + }, + { + "epoch": 2.487309644670051, + "grad_norm": 8.9375, + "learning_rate": 2.6419806027650237e-06, + "loss": 2.6307, + "step": 23520 + }, + { + "epoch": 2.4874153976311337, + "grad_norm": 8.625, + "learning_rate": 2.6418096226136515e-06, + "loss": 2.6714, + "step": 23521 + }, + { + "epoch": 2.4875211505922166, + "grad_norm": 12.5, + "learning_rate": 2.6416386417968248e-06, + "loss": 2.6589, + "step": 23522 + }, + { + "epoch": 2.4876269035532994, + "grad_norm": 8.5625, + "learning_rate": 2.6414676603153453e-06, + "loss": 2.6421, + "step": 23523 + }, + { + "epoch": 2.4877326565143822, + "grad_norm": 7.96875, + "learning_rate": 2.6412966781700173e-06, + "loss": 2.5557, + "step": 23524 + }, + { + "epoch": 2.487838409475465, + "grad_norm": 8.625, + "learning_rate": 2.641125695361641e-06, + "loss": 2.5477, + "step": 23525 + }, + { + "epoch": 2.4879441624365484, + "grad_norm": 8.625, + "learning_rate": 2.640954711891019e-06, + "loss": 2.4165, + "step": 23526 + }, + { + "epoch": 2.488049915397631, + "grad_norm": 8.5625, + "learning_rate": 2.6407837277589544e-06, + "loss": 2.4668, + "step": 23527 + }, + { + "epoch": 2.488155668358714, + "grad_norm": 8.625, + "learning_rate": 2.64061274296625e-06, + "loss": 2.2708, + "step": 23528 + }, + { + "epoch": 2.488261421319797, + "grad_norm": 8.3125, + "learning_rate": 2.6404417575137063e-06, + "loss": 2.625, + "step": 23529 + }, + { + "epoch": 2.4883671742808797, + "grad_norm": 8.875, + "learning_rate": 2.640270771402128e-06, + "loss": 2.726, + "step": 23530 + }, + { + "epoch": 2.488472927241963, + "grad_norm": 8.4375, + "learning_rate": 2.640099784632316e-06, + "loss": 2.477, + "step": 23531 + }, + { + "epoch": 2.488578680203046, + "grad_norm": 8.1875, + "learning_rate": 2.639928797205073e-06, + "loss": 2.5844, + "step": 23532 + }, + { + "epoch": 2.4886844331641287, + "grad_norm": 8.9375, + "learning_rate": 2.639757809121202e-06, + "loss": 2.5605, + "step": 23533 + }, + { + "epoch": 2.4887901861252115, + "grad_norm": 8.0, + "learning_rate": 2.639586820381504e-06, + "loss": 2.3614, + "step": 23534 + }, + { + "epoch": 2.4888959390862944, + "grad_norm": 8.3125, + "learning_rate": 2.6394158309867817e-06, + "loss": 2.2213, + "step": 23535 + }, + { + "epoch": 2.489001692047377, + "grad_norm": 8.125, + "learning_rate": 2.6392448409378384e-06, + "loss": 2.2986, + "step": 23536 + }, + { + "epoch": 2.48910744500846, + "grad_norm": 8.375, + "learning_rate": 2.639073850235476e-06, + "loss": 2.4013, + "step": 23537 + }, + { + "epoch": 2.4892131979695433, + "grad_norm": 8.25, + "learning_rate": 2.6389028588804978e-06, + "loss": 2.5143, + "step": 23538 + }, + { + "epoch": 2.489318950930626, + "grad_norm": 8.5, + "learning_rate": 2.638731866873704e-06, + "loss": 2.3381, + "step": 23539 + }, + { + "epoch": 2.489424703891709, + "grad_norm": 8.4375, + "learning_rate": 2.638560874215899e-06, + "loss": 2.4932, + "step": 23540 + }, + { + "epoch": 2.489530456852792, + "grad_norm": 8.3125, + "learning_rate": 2.6383898809078844e-06, + "loss": 2.5505, + "step": 23541 + }, + { + "epoch": 2.4896362098138747, + "grad_norm": 8.125, + "learning_rate": 2.638218886950462e-06, + "loss": 2.4054, + "step": 23542 + }, + { + "epoch": 2.4897419627749575, + "grad_norm": 8.3125, + "learning_rate": 2.638047892344436e-06, + "loss": 2.4821, + "step": 23543 + }, + { + "epoch": 2.489847715736041, + "grad_norm": 9.9375, + "learning_rate": 2.6378768970906072e-06, + "loss": 2.5624, + "step": 23544 + }, + { + "epoch": 2.4899534686971236, + "grad_norm": 7.9375, + "learning_rate": 2.6377059011897776e-06, + "loss": 2.4352, + "step": 23545 + }, + { + "epoch": 2.4900592216582065, + "grad_norm": 7.9375, + "learning_rate": 2.6375349046427516e-06, + "loss": 2.6401, + "step": 23546 + }, + { + "epoch": 2.4901649746192893, + "grad_norm": 8.8125, + "learning_rate": 2.637363907450331e-06, + "loss": 2.6259, + "step": 23547 + }, + { + "epoch": 2.490270727580372, + "grad_norm": 8.5, + "learning_rate": 2.6371929096133166e-06, + "loss": 2.4371, + "step": 23548 + }, + { + "epoch": 2.490376480541455, + "grad_norm": 7.9375, + "learning_rate": 2.637021911132513e-06, + "loss": 2.55, + "step": 23549 + }, + { + "epoch": 2.4904822335025383, + "grad_norm": 8.5, + "learning_rate": 2.636850912008721e-06, + "loss": 2.3077, + "step": 23550 + }, + { + "epoch": 2.490587986463621, + "grad_norm": 8.1875, + "learning_rate": 2.636679912242743e-06, + "loss": 2.4922, + "step": 23551 + }, + { + "epoch": 2.490693739424704, + "grad_norm": 8.5625, + "learning_rate": 2.6365089118353836e-06, + "loss": 2.3628, + "step": 23552 + }, + { + "epoch": 2.490799492385787, + "grad_norm": 9.25, + "learning_rate": 2.636337910787443e-06, + "loss": 2.5845, + "step": 23553 + }, + { + "epoch": 2.4909052453468696, + "grad_norm": 8.125, + "learning_rate": 2.6361669090997237e-06, + "loss": 2.2417, + "step": 23554 + }, + { + "epoch": 2.4910109983079525, + "grad_norm": 7.75, + "learning_rate": 2.6359959067730294e-06, + "loss": 2.6551, + "step": 23555 + }, + { + "epoch": 2.4911167512690353, + "grad_norm": 7.84375, + "learning_rate": 2.635824903808162e-06, + "loss": 2.4451, + "step": 23556 + }, + { + "epoch": 2.4912225042301186, + "grad_norm": 8.6875, + "learning_rate": 2.635653900205924e-06, + "loss": 2.8968, + "step": 23557 + }, + { + "epoch": 2.4913282571912014, + "grad_norm": 8.75, + "learning_rate": 2.635482895967117e-06, + "loss": 2.0691, + "step": 23558 + }, + { + "epoch": 2.4914340101522843, + "grad_norm": 8.9375, + "learning_rate": 2.635311891092545e-06, + "loss": 2.7984, + "step": 23559 + }, + { + "epoch": 2.491539763113367, + "grad_norm": 8.0625, + "learning_rate": 2.6351408855830098e-06, + "loss": 2.5788, + "step": 23560 + }, + { + "epoch": 2.49164551607445, + "grad_norm": 8.375, + "learning_rate": 2.6349698794393127e-06, + "loss": 2.6231, + "step": 23561 + }, + { + "epoch": 2.4917512690355332, + "grad_norm": 8.4375, + "learning_rate": 2.634798872662258e-06, + "loss": 2.4403, + "step": 23562 + }, + { + "epoch": 2.491857021996616, + "grad_norm": 8.875, + "learning_rate": 2.6346278652526465e-06, + "loss": 2.5489, + "step": 23563 + }, + { + "epoch": 2.491962774957699, + "grad_norm": 8.625, + "learning_rate": 2.634456857211282e-06, + "loss": 2.5286, + "step": 23564 + }, + { + "epoch": 2.4920685279187818, + "grad_norm": 8.4375, + "learning_rate": 2.634285848538966e-06, + "loss": 2.4291, + "step": 23565 + }, + { + "epoch": 2.4921742808798646, + "grad_norm": 8.6875, + "learning_rate": 2.6341148392365022e-06, + "loss": 2.5534, + "step": 23566 + }, + { + "epoch": 2.4922800338409474, + "grad_norm": 8.625, + "learning_rate": 2.633943829304692e-06, + "loss": 2.636, + "step": 23567 + }, + { + "epoch": 2.4923857868020303, + "grad_norm": 8.4375, + "learning_rate": 2.6337728187443386e-06, + "loss": 2.4928, + "step": 23568 + }, + { + "epoch": 2.4924915397631136, + "grad_norm": 8.25, + "learning_rate": 2.6336018075562435e-06, + "loss": 2.3835, + "step": 23569 + }, + { + "epoch": 2.4925972927241964, + "grad_norm": 8.4375, + "learning_rate": 2.63343079574121e-06, + "loss": 2.3208, + "step": 23570 + }, + { + "epoch": 2.4927030456852792, + "grad_norm": 9.0, + "learning_rate": 2.63325978330004e-06, + "loss": 2.4587, + "step": 23571 + }, + { + "epoch": 2.492808798646362, + "grad_norm": 8.125, + "learning_rate": 2.6330887702335363e-06, + "loss": 2.632, + "step": 23572 + }, + { + "epoch": 2.492914551607445, + "grad_norm": 8.4375, + "learning_rate": 2.632917756542502e-06, + "loss": 2.568, + "step": 23573 + }, + { + "epoch": 2.4930203045685277, + "grad_norm": 8.125, + "learning_rate": 2.632746742227738e-06, + "loss": 2.3309, + "step": 23574 + }, + { + "epoch": 2.4931260575296106, + "grad_norm": 8.6875, + "learning_rate": 2.6325757272900492e-06, + "loss": 2.4299, + "step": 23575 + }, + { + "epoch": 2.493231810490694, + "grad_norm": 8.25, + "learning_rate": 2.6324047117302367e-06, + "loss": 2.6131, + "step": 23576 + }, + { + "epoch": 2.4933375634517767, + "grad_norm": 8.3125, + "learning_rate": 2.6322336955491017e-06, + "loss": 2.5637, + "step": 23577 + }, + { + "epoch": 2.4934433164128595, + "grad_norm": 8.5, + "learning_rate": 2.632062678747449e-06, + "loss": 2.6372, + "step": 23578 + }, + { + "epoch": 2.4935490693739424, + "grad_norm": 8.3125, + "learning_rate": 2.6318916613260805e-06, + "loss": 2.4969, + "step": 23579 + }, + { + "epoch": 2.4936548223350252, + "grad_norm": 8.3125, + "learning_rate": 2.6317206432857967e-06, + "loss": 2.5557, + "step": 23580 + }, + { + "epoch": 2.4937605752961085, + "grad_norm": 8.1875, + "learning_rate": 2.631549624627403e-06, + "loss": 2.5791, + "step": 23581 + }, + { + "epoch": 2.4938663282571913, + "grad_norm": 8.875, + "learning_rate": 2.631378605351701e-06, + "loss": 2.5116, + "step": 23582 + }, + { + "epoch": 2.493972081218274, + "grad_norm": 8.1875, + "learning_rate": 2.631207585459492e-06, + "loss": 2.633, + "step": 23583 + }, + { + "epoch": 2.494077834179357, + "grad_norm": 8.375, + "learning_rate": 2.63103656495158e-06, + "loss": 2.4157, + "step": 23584 + }, + { + "epoch": 2.49418358714044, + "grad_norm": 9.0625, + "learning_rate": 2.6308655438287677e-06, + "loss": 2.6358, + "step": 23585 + }, + { + "epoch": 2.4942893401015227, + "grad_norm": 7.9375, + "learning_rate": 2.6306945220918555e-06, + "loss": 2.5574, + "step": 23586 + }, + { + "epoch": 2.4943950930626055, + "grad_norm": 8.1875, + "learning_rate": 2.6305234997416484e-06, + "loss": 2.5527, + "step": 23587 + }, + { + "epoch": 2.494500846023689, + "grad_norm": 8.0625, + "learning_rate": 2.630352476778948e-06, + "loss": 2.5134, + "step": 23588 + }, + { + "epoch": 2.4946065989847717, + "grad_norm": 7.8125, + "learning_rate": 2.630181453204556e-06, + "loss": 2.4997, + "step": 23589 + }, + { + "epoch": 2.4947123519458545, + "grad_norm": 8.25, + "learning_rate": 2.6300104290192758e-06, + "loss": 2.5142, + "step": 23590 + }, + { + "epoch": 2.4948181049069373, + "grad_norm": 8.0625, + "learning_rate": 2.6298394042239106e-06, + "loss": 2.1814, + "step": 23591 + }, + { + "epoch": 2.49492385786802, + "grad_norm": 8.125, + "learning_rate": 2.6296683788192616e-06, + "loss": 2.4673, + "step": 23592 + }, + { + "epoch": 2.4950296108291035, + "grad_norm": 9.125, + "learning_rate": 2.629497352806132e-06, + "loss": 2.6716, + "step": 23593 + }, + { + "epoch": 2.4951353637901863, + "grad_norm": 8.5625, + "learning_rate": 2.629326326185324e-06, + "loss": 2.5519, + "step": 23594 + }, + { + "epoch": 2.495241116751269, + "grad_norm": 8.5, + "learning_rate": 2.6291552989576415e-06, + "loss": 2.5868, + "step": 23595 + }, + { + "epoch": 2.495346869712352, + "grad_norm": 8.3125, + "learning_rate": 2.6289842711238845e-06, + "loss": 2.4437, + "step": 23596 + }, + { + "epoch": 2.495452622673435, + "grad_norm": 8.3125, + "learning_rate": 2.628813242684858e-06, + "loss": 2.602, + "step": 23597 + }, + { + "epoch": 2.4955583756345177, + "grad_norm": 8.5, + "learning_rate": 2.628642213641363e-06, + "loss": 2.4745, + "step": 23598 + }, + { + "epoch": 2.4956641285956005, + "grad_norm": 8.5, + "learning_rate": 2.628471183994203e-06, + "loss": 2.5855, + "step": 23599 + }, + { + "epoch": 2.495769881556684, + "grad_norm": 9.0, + "learning_rate": 2.6283001537441806e-06, + "loss": 2.8152, + "step": 23600 + }, + { + "epoch": 2.4958756345177666, + "grad_norm": 8.125, + "learning_rate": 2.628129122892098e-06, + "loss": 2.4871, + "step": 23601 + }, + { + "epoch": 2.4959813874788495, + "grad_norm": 8.8125, + "learning_rate": 2.627958091438757e-06, + "loss": 2.6225, + "step": 23602 + }, + { + "epoch": 2.4960871404399323, + "grad_norm": 8.5, + "learning_rate": 2.627787059384962e-06, + "loss": 2.282, + "step": 23603 + }, + { + "epoch": 2.496192893401015, + "grad_norm": 8.4375, + "learning_rate": 2.627616026731515e-06, + "loss": 2.5884, + "step": 23604 + }, + { + "epoch": 2.496298646362098, + "grad_norm": 8.3125, + "learning_rate": 2.6274449934792166e-06, + "loss": 2.5721, + "step": 23605 + }, + { + "epoch": 2.496404399323181, + "grad_norm": 8.5625, + "learning_rate": 2.6272739596288723e-06, + "loss": 2.6443, + "step": 23606 + }, + { + "epoch": 2.496510152284264, + "grad_norm": 8.3125, + "learning_rate": 2.6271029251812826e-06, + "loss": 2.5369, + "step": 23607 + }, + { + "epoch": 2.496615905245347, + "grad_norm": 8.25, + "learning_rate": 2.6269318901372516e-06, + "loss": 2.4577, + "step": 23608 + }, + { + "epoch": 2.4967216582064298, + "grad_norm": 8.125, + "learning_rate": 2.62676085449758e-06, + "loss": 2.7129, + "step": 23609 + }, + { + "epoch": 2.4968274111675126, + "grad_norm": 8.375, + "learning_rate": 2.6265898182630724e-06, + "loss": 2.3964, + "step": 23610 + }, + { + "epoch": 2.4969331641285955, + "grad_norm": 8.375, + "learning_rate": 2.6264187814345302e-06, + "loss": 2.4192, + "step": 23611 + }, + { + "epoch": 2.4970389170896787, + "grad_norm": 9.3125, + "learning_rate": 2.6262477440127567e-06, + "loss": 2.6816, + "step": 23612 + }, + { + "epoch": 2.4971446700507616, + "grad_norm": 8.4375, + "learning_rate": 2.6260767059985543e-06, + "loss": 2.6326, + "step": 23613 + }, + { + "epoch": 2.4972504230118444, + "grad_norm": 8.4375, + "learning_rate": 2.625905667392725e-06, + "loss": 2.5927, + "step": 23614 + }, + { + "epoch": 2.4973561759729273, + "grad_norm": 8.3125, + "learning_rate": 2.625734628196072e-06, + "loss": 2.4639, + "step": 23615 + }, + { + "epoch": 2.49746192893401, + "grad_norm": 8.3125, + "learning_rate": 2.6255635884093977e-06, + "loss": 2.3675, + "step": 23616 + }, + { + "epoch": 2.497567681895093, + "grad_norm": 8.875, + "learning_rate": 2.625392548033505e-06, + "loss": 2.5158, + "step": 23617 + }, + { + "epoch": 2.4976734348561758, + "grad_norm": 8.25, + "learning_rate": 2.6252215070691967e-06, + "loss": 2.6141, + "step": 23618 + }, + { + "epoch": 2.497779187817259, + "grad_norm": 8.25, + "learning_rate": 2.625050465517275e-06, + "loss": 2.6463, + "step": 23619 + }, + { + "epoch": 2.497884940778342, + "grad_norm": 8.8125, + "learning_rate": 2.6248794233785425e-06, + "loss": 2.6842, + "step": 23620 + }, + { + "epoch": 2.4979906937394247, + "grad_norm": 8.125, + "learning_rate": 2.6247083806538014e-06, + "loss": 2.4456, + "step": 23621 + }, + { + "epoch": 2.4980964467005076, + "grad_norm": 8.8125, + "learning_rate": 2.6245373373438554e-06, + "loss": 2.3918, + "step": 23622 + }, + { + "epoch": 2.4982021996615904, + "grad_norm": 8.4375, + "learning_rate": 2.6243662934495067e-06, + "loss": 2.6479, + "step": 23623 + }, + { + "epoch": 2.4983079526226732, + "grad_norm": 8.375, + "learning_rate": 2.6241952489715577e-06, + "loss": 2.6285, + "step": 23624 + }, + { + "epoch": 2.4984137055837565, + "grad_norm": 8.5, + "learning_rate": 2.624024203910811e-06, + "loss": 2.6148, + "step": 23625 + }, + { + "epoch": 2.4985194585448394, + "grad_norm": 8.3125, + "learning_rate": 2.62385315826807e-06, + "loss": 2.5266, + "step": 23626 + }, + { + "epoch": 2.498625211505922, + "grad_norm": 8.75, + "learning_rate": 2.6236821120441362e-06, + "loss": 2.732, + "step": 23627 + }, + { + "epoch": 2.498730964467005, + "grad_norm": 8.0625, + "learning_rate": 2.6235110652398126e-06, + "loss": 2.3424, + "step": 23628 + }, + { + "epoch": 2.498836717428088, + "grad_norm": 7.90625, + "learning_rate": 2.623340017855903e-06, + "loss": 2.4093, + "step": 23629 + }, + { + "epoch": 2.4989424703891707, + "grad_norm": 9.0625, + "learning_rate": 2.623168969893209e-06, + "loss": 2.5239, + "step": 23630 + }, + { + "epoch": 2.499048223350254, + "grad_norm": 8.875, + "learning_rate": 2.6229979213525327e-06, + "loss": 2.589, + "step": 23631 + }, + { + "epoch": 2.499153976311337, + "grad_norm": 9.5625, + "learning_rate": 2.622826872234678e-06, + "loss": 2.6291, + "step": 23632 + }, + { + "epoch": 2.4992597292724197, + "grad_norm": 8.5625, + "learning_rate": 2.6226558225404468e-06, + "loss": 2.5657, + "step": 23633 + }, + { + "epoch": 2.4993654822335025, + "grad_norm": 8.3125, + "learning_rate": 2.6224847722706407e-06, + "loss": 2.4871, + "step": 23634 + }, + { + "epoch": 2.4994712351945854, + "grad_norm": 8.5, + "learning_rate": 2.6223137214260656e-06, + "loss": 2.6654, + "step": 23635 + }, + { + "epoch": 2.499576988155668, + "grad_norm": 8.8125, + "learning_rate": 2.622142670007522e-06, + "loss": 2.545, + "step": 23636 + }, + { + "epoch": 2.499682741116751, + "grad_norm": 8.5625, + "learning_rate": 2.6219716180158118e-06, + "loss": 2.5153, + "step": 23637 + }, + { + "epoch": 2.4997884940778343, + "grad_norm": 8.3125, + "learning_rate": 2.62180056545174e-06, + "loss": 2.6132, + "step": 23638 + }, + { + "epoch": 2.499894247038917, + "grad_norm": 7.71875, + "learning_rate": 2.621629512316107e-06, + "loss": 2.4044, + "step": 23639 + }, + { + "epoch": 2.5, + "grad_norm": 8.0625, + "learning_rate": 2.6214584586097164e-06, + "loss": 2.3695, + "step": 23640 + }, + { + "epoch": 2.500105752961083, + "grad_norm": 8.625, + "learning_rate": 2.6212874043333713e-06, + "loss": 2.7474, + "step": 23641 + }, + { + "epoch": 2.5002115059221657, + "grad_norm": 8.625, + "learning_rate": 2.6211163494878737e-06, + "loss": 2.6133, + "step": 23642 + }, + { + "epoch": 2.500317258883249, + "grad_norm": 9.0625, + "learning_rate": 2.6209452940740272e-06, + "loss": 2.5095, + "step": 23643 + }, + { + "epoch": 2.500423011844332, + "grad_norm": 8.6875, + "learning_rate": 2.620774238092633e-06, + "loss": 2.5836, + "step": 23644 + }, + { + "epoch": 2.5005287648054146, + "grad_norm": 8.0625, + "learning_rate": 2.6206031815444956e-06, + "loss": 2.3809, + "step": 23645 + }, + { + "epoch": 2.5006345177664975, + "grad_norm": 8.4375, + "learning_rate": 2.6204321244304164e-06, + "loss": 2.4807, + "step": 23646 + }, + { + "epoch": 2.5007402707275803, + "grad_norm": 8.0625, + "learning_rate": 2.6202610667511985e-06, + "loss": 2.549, + "step": 23647 + }, + { + "epoch": 2.500846023688663, + "grad_norm": 9.0625, + "learning_rate": 2.620090008507645e-06, + "loss": 2.3009, + "step": 23648 + }, + { + "epoch": 2.500951776649746, + "grad_norm": 8.5625, + "learning_rate": 2.6199189497005578e-06, + "loss": 2.6444, + "step": 23649 + }, + { + "epoch": 2.5010575296108293, + "grad_norm": 8.5625, + "learning_rate": 2.61974789033074e-06, + "loss": 2.738, + "step": 23650 + }, + { + "epoch": 2.501163282571912, + "grad_norm": 9.0625, + "learning_rate": 2.619576830398994e-06, + "loss": 2.7532, + "step": 23651 + }, + { + "epoch": 2.501269035532995, + "grad_norm": 8.8125, + "learning_rate": 2.619405769906123e-06, + "loss": 2.1859, + "step": 23652 + }, + { + "epoch": 2.501374788494078, + "grad_norm": 8.125, + "learning_rate": 2.6192347088529295e-06, + "loss": 2.5062, + "step": 23653 + }, + { + "epoch": 2.5014805414551606, + "grad_norm": 8.4375, + "learning_rate": 2.619063647240217e-06, + "loss": 2.5428, + "step": 23654 + }, + { + "epoch": 2.501586294416244, + "grad_norm": 8.875, + "learning_rate": 2.6188925850687873e-06, + "loss": 2.7041, + "step": 23655 + }, + { + "epoch": 2.5016920473773263, + "grad_norm": 8.8125, + "learning_rate": 2.618721522339443e-06, + "loss": 2.5199, + "step": 23656 + }, + { + "epoch": 2.5017978003384096, + "grad_norm": 8.4375, + "learning_rate": 2.6185504590529876e-06, + "loss": 2.6037, + "step": 23657 + }, + { + "epoch": 2.5019035532994924, + "grad_norm": 8.9375, + "learning_rate": 2.6183793952102227e-06, + "loss": 2.4896, + "step": 23658 + }, + { + "epoch": 2.5020093062605753, + "grad_norm": 8.5, + "learning_rate": 2.618208330811952e-06, + "loss": 2.6951, + "step": 23659 + }, + { + "epoch": 2.502115059221658, + "grad_norm": 8.8125, + "learning_rate": 2.6180372658589775e-06, + "loss": 2.7382, + "step": 23660 + }, + { + "epoch": 2.502220812182741, + "grad_norm": 7.96875, + "learning_rate": 2.6178662003521033e-06, + "loss": 2.6418, + "step": 23661 + }, + { + "epoch": 2.5023265651438242, + "grad_norm": 8.6875, + "learning_rate": 2.617695134292131e-06, + "loss": 2.6603, + "step": 23662 + }, + { + "epoch": 2.502432318104907, + "grad_norm": 8.5625, + "learning_rate": 2.617524067679863e-06, + "loss": 2.5692, + "step": 23663 + }, + { + "epoch": 2.50253807106599, + "grad_norm": 7.875, + "learning_rate": 2.617353000516103e-06, + "loss": 2.4545, + "step": 23664 + }, + { + "epoch": 2.5026438240270727, + "grad_norm": 8.25, + "learning_rate": 2.6171819328016534e-06, + "loss": 2.4368, + "step": 23665 + }, + { + "epoch": 2.5027495769881556, + "grad_norm": 8.1875, + "learning_rate": 2.6170108645373167e-06, + "loss": 2.5528, + "step": 23666 + }, + { + "epoch": 2.5028553299492384, + "grad_norm": 8.4375, + "learning_rate": 2.616839795723896e-06, + "loss": 2.624, + "step": 23667 + }, + { + "epoch": 2.5029610829103213, + "grad_norm": 8.25, + "learning_rate": 2.6166687263621944e-06, + "loss": 2.3956, + "step": 23668 + }, + { + "epoch": 2.5030668358714045, + "grad_norm": 8.1875, + "learning_rate": 2.616497656453013e-06, + "loss": 2.6174, + "step": 23669 + }, + { + "epoch": 2.5031725888324874, + "grad_norm": 8.5, + "learning_rate": 2.6163265859971564e-06, + "loss": 2.4498, + "step": 23670 + }, + { + "epoch": 2.5032783417935702, + "grad_norm": 8.375, + "learning_rate": 2.616155514995427e-06, + "loss": 2.4703, + "step": 23671 + }, + { + "epoch": 2.503384094754653, + "grad_norm": 8.3125, + "learning_rate": 2.6159844434486272e-06, + "loss": 2.6702, + "step": 23672 + }, + { + "epoch": 2.503489847715736, + "grad_norm": 8.125, + "learning_rate": 2.6158133713575596e-06, + "loss": 2.426, + "step": 23673 + }, + { + "epoch": 2.503595600676819, + "grad_norm": 8.25, + "learning_rate": 2.6156422987230272e-06, + "loss": 2.4571, + "step": 23674 + }, + { + "epoch": 2.503701353637902, + "grad_norm": 8.6875, + "learning_rate": 2.615471225545833e-06, + "loss": 2.661, + "step": 23675 + }, + { + "epoch": 2.503807106598985, + "grad_norm": 8.375, + "learning_rate": 2.6153001518267795e-06, + "loss": 2.4515, + "step": 23676 + }, + { + "epoch": 2.5039128595600677, + "grad_norm": 8.5, + "learning_rate": 2.6151290775666698e-06, + "loss": 2.7254, + "step": 23677 + }, + { + "epoch": 2.5040186125211505, + "grad_norm": 7.96875, + "learning_rate": 2.614958002766306e-06, + "loss": 2.3946, + "step": 23678 + }, + { + "epoch": 2.5041243654822334, + "grad_norm": 8.5625, + "learning_rate": 2.614786927426491e-06, + "loss": 2.4781, + "step": 23679 + }, + { + "epoch": 2.504230118443316, + "grad_norm": 8.6875, + "learning_rate": 2.6146158515480286e-06, + "loss": 2.4837, + "step": 23680 + }, + { + "epoch": 2.5043358714043995, + "grad_norm": 8.75, + "learning_rate": 2.614444775131721e-06, + "loss": 2.5178, + "step": 23681 + }, + { + "epoch": 2.5044416243654823, + "grad_norm": 8.4375, + "learning_rate": 2.61427369817837e-06, + "loss": 2.5669, + "step": 23682 + }, + { + "epoch": 2.504547377326565, + "grad_norm": 8.5625, + "learning_rate": 2.61410262068878e-06, + "loss": 2.6899, + "step": 23683 + }, + { + "epoch": 2.504653130287648, + "grad_norm": 8.5625, + "learning_rate": 2.6139315426637527e-06, + "loss": 2.5355, + "step": 23684 + }, + { + "epoch": 2.504758883248731, + "grad_norm": 8.6875, + "learning_rate": 2.6137604641040915e-06, + "loss": 2.3704, + "step": 23685 + }, + { + "epoch": 2.504864636209814, + "grad_norm": 8.125, + "learning_rate": 2.6135893850105987e-06, + "loss": 2.5788, + "step": 23686 + }, + { + "epoch": 2.5049703891708965, + "grad_norm": 8.25, + "learning_rate": 2.6134183053840774e-06, + "loss": 2.4666, + "step": 23687 + }, + { + "epoch": 2.50507614213198, + "grad_norm": 8.25, + "learning_rate": 2.6132472252253306e-06, + "loss": 2.4596, + "step": 23688 + }, + { + "epoch": 2.5051818950930627, + "grad_norm": 8.1875, + "learning_rate": 2.6130761445351604e-06, + "loss": 2.4421, + "step": 23689 + }, + { + "epoch": 2.5052876480541455, + "grad_norm": 8.375, + "learning_rate": 2.612905063314371e-06, + "loss": 2.5251, + "step": 23690 + }, + { + "epoch": 2.5053934010152283, + "grad_norm": 8.125, + "learning_rate": 2.612733981563764e-06, + "loss": 2.5165, + "step": 23691 + }, + { + "epoch": 2.505499153976311, + "grad_norm": 8.1875, + "learning_rate": 2.6125628992841423e-06, + "loss": 2.4801, + "step": 23692 + }, + { + "epoch": 2.5056049069373945, + "grad_norm": 8.8125, + "learning_rate": 2.612391816476309e-06, + "loss": 2.6161, + "step": 23693 + }, + { + "epoch": 2.5057106598984773, + "grad_norm": 8.6875, + "learning_rate": 2.6122207331410664e-06, + "loss": 2.1644, + "step": 23694 + }, + { + "epoch": 2.50581641285956, + "grad_norm": 8.375, + "learning_rate": 2.612049649279218e-06, + "loss": 2.5458, + "step": 23695 + }, + { + "epoch": 2.505922165820643, + "grad_norm": 8.375, + "learning_rate": 2.6118785648915666e-06, + "loss": 2.5769, + "step": 23696 + }, + { + "epoch": 2.506027918781726, + "grad_norm": 8.0625, + "learning_rate": 2.6117074799789156e-06, + "loss": 2.4312, + "step": 23697 + }, + { + "epoch": 2.5061336717428087, + "grad_norm": 8.1875, + "learning_rate": 2.6115363945420664e-06, + "loss": 2.4492, + "step": 23698 + }, + { + "epoch": 2.5062394247038915, + "grad_norm": 8.625, + "learning_rate": 2.6113653085818223e-06, + "loss": 2.4983, + "step": 23699 + }, + { + "epoch": 2.5063451776649748, + "grad_norm": 9.3125, + "learning_rate": 2.6111942220989873e-06, + "loss": 2.5516, + "step": 23700 + }, + { + "epoch": 2.5064509306260576, + "grad_norm": 8.625, + "learning_rate": 2.6110231350943623e-06, + "loss": 2.5929, + "step": 23701 + }, + { + "epoch": 2.5065566835871405, + "grad_norm": 8.0625, + "learning_rate": 2.6108520475687514e-06, + "loss": 2.5543, + "step": 23702 + }, + { + "epoch": 2.5066624365482233, + "grad_norm": 8.6875, + "learning_rate": 2.610680959522957e-06, + "loss": 2.5134, + "step": 23703 + }, + { + "epoch": 2.506768189509306, + "grad_norm": 8.5, + "learning_rate": 2.610509870957782e-06, + "loss": 2.5203, + "step": 23704 + }, + { + "epoch": 2.5068739424703894, + "grad_norm": 8.125, + "learning_rate": 2.6103387818740294e-06, + "loss": 2.5953, + "step": 23705 + }, + { + "epoch": 2.506979695431472, + "grad_norm": 8.5625, + "learning_rate": 2.610167692272503e-06, + "loss": 2.6759, + "step": 23706 + }, + { + "epoch": 2.507085448392555, + "grad_norm": 8.1875, + "learning_rate": 2.6099966021540035e-06, + "loss": 2.3937, + "step": 23707 + }, + { + "epoch": 2.507191201353638, + "grad_norm": 8.3125, + "learning_rate": 2.6098255115193357e-06, + "loss": 2.7038, + "step": 23708 + }, + { + "epoch": 2.5072969543147208, + "grad_norm": 9.0, + "learning_rate": 2.6096544203693013e-06, + "loss": 2.5359, + "step": 23709 + }, + { + "epoch": 2.5074027072758036, + "grad_norm": 7.90625, + "learning_rate": 2.6094833287047037e-06, + "loss": 2.455, + "step": 23710 + }, + { + "epoch": 2.5075084602368864, + "grad_norm": 8.125, + "learning_rate": 2.6093122365263455e-06, + "loss": 2.4115, + "step": 23711 + }, + { + "epoch": 2.5076142131979697, + "grad_norm": 8.25, + "learning_rate": 2.60914114383503e-06, + "loss": 2.5415, + "step": 23712 + }, + { + "epoch": 2.5077199661590526, + "grad_norm": 9.3125, + "learning_rate": 2.6089700506315595e-06, + "loss": 2.5297, + "step": 23713 + }, + { + "epoch": 2.5078257191201354, + "grad_norm": 8.1875, + "learning_rate": 2.6087989569167365e-06, + "loss": 2.3661, + "step": 23714 + }, + { + "epoch": 2.5079314720812182, + "grad_norm": 8.4375, + "learning_rate": 2.6086278626913656e-06, + "loss": 2.6101, + "step": 23715 + }, + { + "epoch": 2.508037225042301, + "grad_norm": 8.8125, + "learning_rate": 2.608456767956248e-06, + "loss": 2.623, + "step": 23716 + }, + { + "epoch": 2.5081429780033844, + "grad_norm": 8.0625, + "learning_rate": 2.6082856727121876e-06, + "loss": 2.5982, + "step": 23717 + }, + { + "epoch": 2.5082487309644668, + "grad_norm": 8.0625, + "learning_rate": 2.6081145769599865e-06, + "loss": 2.608, + "step": 23718 + }, + { + "epoch": 2.50835448392555, + "grad_norm": 8.1875, + "learning_rate": 2.6079434807004478e-06, + "loss": 2.255, + "step": 23719 + }, + { + "epoch": 2.508460236886633, + "grad_norm": 8.6875, + "learning_rate": 2.6077723839343745e-06, + "loss": 2.3834, + "step": 23720 + }, + { + "epoch": 2.5085659898477157, + "grad_norm": 8.4375, + "learning_rate": 2.60760128666257e-06, + "loss": 2.7219, + "step": 23721 + }, + { + "epoch": 2.5086717428087986, + "grad_norm": 8.5625, + "learning_rate": 2.6074301888858356e-06, + "loss": 2.6562, + "step": 23722 + }, + { + "epoch": 2.5087774957698814, + "grad_norm": 8.3125, + "learning_rate": 2.6072590906049755e-06, + "loss": 2.4788, + "step": 23723 + }, + { + "epoch": 2.5088832487309647, + "grad_norm": 8.8125, + "learning_rate": 2.6070879918207933e-06, + "loss": 2.4717, + "step": 23724 + }, + { + "epoch": 2.5089890016920475, + "grad_norm": 8.625, + "learning_rate": 2.6069168925340905e-06, + "loss": 2.7332, + "step": 23725 + }, + { + "epoch": 2.5090947546531304, + "grad_norm": 8.375, + "learning_rate": 2.60674579274567e-06, + "loss": 2.7049, + "step": 23726 + }, + { + "epoch": 2.509200507614213, + "grad_norm": 8.4375, + "learning_rate": 2.6065746924563355e-06, + "loss": 2.5802, + "step": 23727 + }, + { + "epoch": 2.509306260575296, + "grad_norm": 8.9375, + "learning_rate": 2.60640359166689e-06, + "loss": 2.5598, + "step": 23728 + }, + { + "epoch": 2.509412013536379, + "grad_norm": 9.0625, + "learning_rate": 2.606232490378135e-06, + "loss": 2.7256, + "step": 23729 + }, + { + "epoch": 2.5095177664974617, + "grad_norm": 9.0, + "learning_rate": 2.6060613885908754e-06, + "loss": 2.6295, + "step": 23730 + }, + { + "epoch": 2.509623519458545, + "grad_norm": 7.96875, + "learning_rate": 2.6058902863059123e-06, + "loss": 2.6074, + "step": 23731 + }, + { + "epoch": 2.509729272419628, + "grad_norm": 8.75, + "learning_rate": 2.6057191835240497e-06, + "loss": 2.3372, + "step": 23732 + }, + { + "epoch": 2.5098350253807107, + "grad_norm": 8.0625, + "learning_rate": 2.6055480802460895e-06, + "loss": 2.5019, + "step": 23733 + }, + { + "epoch": 2.5099407783417935, + "grad_norm": 8.0625, + "learning_rate": 2.6053769764728367e-06, + "loss": 2.669, + "step": 23734 + }, + { + "epoch": 2.5100465313028764, + "grad_norm": 8.3125, + "learning_rate": 2.6052058722050923e-06, + "loss": 2.0955, + "step": 23735 + }, + { + "epoch": 2.5101522842639596, + "grad_norm": 8.6875, + "learning_rate": 2.6050347674436594e-06, + "loss": 2.659, + "step": 23736 + }, + { + "epoch": 2.510258037225042, + "grad_norm": 8.125, + "learning_rate": 2.6048636621893414e-06, + "loss": 2.6755, + "step": 23737 + }, + { + "epoch": 2.5103637901861253, + "grad_norm": 8.6875, + "learning_rate": 2.6046925564429414e-06, + "loss": 2.7019, + "step": 23738 + }, + { + "epoch": 2.510469543147208, + "grad_norm": 8.1875, + "learning_rate": 2.604521450205262e-06, + "loss": 2.6235, + "step": 23739 + }, + { + "epoch": 2.510575296108291, + "grad_norm": 8.5625, + "learning_rate": 2.6043503434771057e-06, + "loss": 2.5478, + "step": 23740 + }, + { + "epoch": 2.510681049069374, + "grad_norm": 8.125, + "learning_rate": 2.604179236259276e-06, + "loss": 2.5338, + "step": 23741 + }, + { + "epoch": 2.5107868020304567, + "grad_norm": 8.3125, + "learning_rate": 2.6040081285525757e-06, + "loss": 2.6228, + "step": 23742 + }, + { + "epoch": 2.51089255499154, + "grad_norm": 8.5625, + "learning_rate": 2.6038370203578083e-06, + "loss": 2.3719, + "step": 23743 + }, + { + "epoch": 2.510998307952623, + "grad_norm": 8.125, + "learning_rate": 2.6036659116757767e-06, + "loss": 2.2284, + "step": 23744 + }, + { + "epoch": 2.5111040609137056, + "grad_norm": 8.375, + "learning_rate": 2.6034948025072822e-06, + "loss": 2.3928, + "step": 23745 + }, + { + "epoch": 2.5112098138747885, + "grad_norm": 8.375, + "learning_rate": 2.6033236928531296e-06, + "loss": 2.4361, + "step": 23746 + }, + { + "epoch": 2.5113155668358713, + "grad_norm": 9.0, + "learning_rate": 2.6031525827141213e-06, + "loss": 2.6155, + "step": 23747 + }, + { + "epoch": 2.511421319796954, + "grad_norm": 9.3125, + "learning_rate": 2.6029814720910595e-06, + "loss": 2.7151, + "step": 23748 + }, + { + "epoch": 2.511527072758037, + "grad_norm": 8.5, + "learning_rate": 2.6028103609847476e-06, + "loss": 2.4567, + "step": 23749 + }, + { + "epoch": 2.5116328257191203, + "grad_norm": 8.625, + "learning_rate": 2.6026392493959894e-06, + "loss": 2.7448, + "step": 23750 + }, + { + "epoch": 2.511738578680203, + "grad_norm": 8.625, + "learning_rate": 2.602468137325587e-06, + "loss": 2.6647, + "step": 23751 + }, + { + "epoch": 2.511844331641286, + "grad_norm": 8.125, + "learning_rate": 2.6022970247743434e-06, + "loss": 2.4557, + "step": 23752 + }, + { + "epoch": 2.511950084602369, + "grad_norm": 8.1875, + "learning_rate": 2.6021259117430623e-06, + "loss": 2.5963, + "step": 23753 + }, + { + "epoch": 2.5120558375634516, + "grad_norm": 8.625, + "learning_rate": 2.601954798232546e-06, + "loss": 2.657, + "step": 23754 + }, + { + "epoch": 2.512161590524535, + "grad_norm": 9.375, + "learning_rate": 2.601783684243597e-06, + "loss": 2.616, + "step": 23755 + }, + { + "epoch": 2.5122673434856178, + "grad_norm": 8.6875, + "learning_rate": 2.601612569777019e-06, + "loss": 2.5611, + "step": 23756 + }, + { + "epoch": 2.5123730964467006, + "grad_norm": 8.5, + "learning_rate": 2.6014414548336143e-06, + "loss": 2.6326, + "step": 23757 + }, + { + "epoch": 2.5124788494077834, + "grad_norm": 9.125, + "learning_rate": 2.601270339414187e-06, + "loss": 2.4882, + "step": 23758 + }, + { + "epoch": 2.5125846023688663, + "grad_norm": 9.125, + "learning_rate": 2.601099223519539e-06, + "loss": 2.3482, + "step": 23759 + }, + { + "epoch": 2.512690355329949, + "grad_norm": 8.8125, + "learning_rate": 2.6009281071504737e-06, + "loss": 2.545, + "step": 23760 + }, + { + "epoch": 2.512796108291032, + "grad_norm": 8.125, + "learning_rate": 2.6007569903077938e-06, + "loss": 2.5281, + "step": 23761 + }, + { + "epoch": 2.5129018612521152, + "grad_norm": 8.375, + "learning_rate": 2.6005858729923035e-06, + "loss": 2.3679, + "step": 23762 + }, + { + "epoch": 2.513007614213198, + "grad_norm": 8.25, + "learning_rate": 2.600414755204804e-06, + "loss": 2.4808, + "step": 23763 + }, + { + "epoch": 2.513113367174281, + "grad_norm": 8.1875, + "learning_rate": 2.6002436369460994e-06, + "loss": 2.5595, + "step": 23764 + }, + { + "epoch": 2.5132191201353637, + "grad_norm": 8.5, + "learning_rate": 2.600072518216993e-06, + "loss": 2.4935, + "step": 23765 + }, + { + "epoch": 2.5133248730964466, + "grad_norm": 8.3125, + "learning_rate": 2.5999013990182857e-06, + "loss": 2.7196, + "step": 23766 + }, + { + "epoch": 2.51343062605753, + "grad_norm": 8.375, + "learning_rate": 2.5997302793507836e-06, + "loss": 2.6776, + "step": 23767 + }, + { + "epoch": 2.5135363790186123, + "grad_norm": 8.375, + "learning_rate": 2.599559159215287e-06, + "loss": 2.5202, + "step": 23768 + }, + { + "epoch": 2.5136421319796955, + "grad_norm": 8.1875, + "learning_rate": 2.599388038612601e-06, + "loss": 2.5867, + "step": 23769 + }, + { + "epoch": 2.5137478849407784, + "grad_norm": 8.4375, + "learning_rate": 2.5992169175435265e-06, + "loss": 2.4341, + "step": 23770 + }, + { + "epoch": 2.5138536379018612, + "grad_norm": 8.3125, + "learning_rate": 2.599045796008868e-06, + "loss": 2.5716, + "step": 23771 + }, + { + "epoch": 2.513959390862944, + "grad_norm": 8.0, + "learning_rate": 2.5988746740094284e-06, + "loss": 2.4498, + "step": 23772 + }, + { + "epoch": 2.514065143824027, + "grad_norm": 8.0625, + "learning_rate": 2.5987035515460106e-06, + "loss": 2.6385, + "step": 23773 + }, + { + "epoch": 2.51417089678511, + "grad_norm": 8.625, + "learning_rate": 2.5985324286194162e-06, + "loss": 2.7368, + "step": 23774 + }, + { + "epoch": 2.514276649746193, + "grad_norm": 9.4375, + "learning_rate": 2.5983613052304502e-06, + "loss": 2.8066, + "step": 23775 + }, + { + "epoch": 2.514382402707276, + "grad_norm": 8.25, + "learning_rate": 2.598190181379915e-06, + "loss": 2.2995, + "step": 23776 + }, + { + "epoch": 2.5144881556683587, + "grad_norm": 8.375, + "learning_rate": 2.598019057068613e-06, + "loss": 2.5105, + "step": 23777 + }, + { + "epoch": 2.5145939086294415, + "grad_norm": 8.6875, + "learning_rate": 2.597847932297348e-06, + "loss": 2.526, + "step": 23778 + }, + { + "epoch": 2.5146996615905244, + "grad_norm": 8.625, + "learning_rate": 2.5976768070669228e-06, + "loss": 2.6362, + "step": 23779 + }, + { + "epoch": 2.514805414551607, + "grad_norm": 8.1875, + "learning_rate": 2.5975056813781395e-06, + "loss": 2.3665, + "step": 23780 + }, + { + "epoch": 2.5149111675126905, + "grad_norm": 8.3125, + "learning_rate": 2.597334555231803e-06, + "loss": 2.528, + "step": 23781 + }, + { + "epoch": 2.5150169204737733, + "grad_norm": 8.1875, + "learning_rate": 2.597163428628715e-06, + "loss": 2.2432, + "step": 23782 + }, + { + "epoch": 2.515122673434856, + "grad_norm": 8.0, + "learning_rate": 2.596992301569678e-06, + "loss": 2.5586, + "step": 23783 + }, + { + "epoch": 2.515228426395939, + "grad_norm": 8.1875, + "learning_rate": 2.596821174055496e-06, + "loss": 2.5096, + "step": 23784 + }, + { + "epoch": 2.515334179357022, + "grad_norm": 8.4375, + "learning_rate": 2.596650046086972e-06, + "loss": 2.5294, + "step": 23785 + }, + { + "epoch": 2.515439932318105, + "grad_norm": 8.875, + "learning_rate": 2.5964789176649095e-06, + "loss": 2.5247, + "step": 23786 + }, + { + "epoch": 2.515545685279188, + "grad_norm": 8.75, + "learning_rate": 2.59630778879011e-06, + "loss": 2.6485, + "step": 23787 + }, + { + "epoch": 2.515651438240271, + "grad_norm": 8.4375, + "learning_rate": 2.596136659463378e-06, + "loss": 2.7714, + "step": 23788 + }, + { + "epoch": 2.5157571912013537, + "grad_norm": 8.625, + "learning_rate": 2.5959655296855156e-06, + "loss": 2.6838, + "step": 23789 + }, + { + "epoch": 2.5158629441624365, + "grad_norm": 8.125, + "learning_rate": 2.5957943994573265e-06, + "loss": 2.6149, + "step": 23790 + }, + { + "epoch": 2.5159686971235193, + "grad_norm": 8.3125, + "learning_rate": 2.5956232687796132e-06, + "loss": 2.4963, + "step": 23791 + }, + { + "epoch": 2.516074450084602, + "grad_norm": 8.6875, + "learning_rate": 2.59545213765318e-06, + "loss": 2.5865, + "step": 23792 + }, + { + "epoch": 2.5161802030456855, + "grad_norm": 8.1875, + "learning_rate": 2.5952810060788273e-06, + "loss": 2.3993, + "step": 23793 + }, + { + "epoch": 2.5162859560067683, + "grad_norm": 8.0625, + "learning_rate": 2.5951098740573606e-06, + "loss": 2.3461, + "step": 23794 + }, + { + "epoch": 2.516391708967851, + "grad_norm": 8.75, + "learning_rate": 2.594938741589582e-06, + "loss": 2.5686, + "step": 23795 + }, + { + "epoch": 2.516497461928934, + "grad_norm": 7.90625, + "learning_rate": 2.5947676086762948e-06, + "loss": 2.4654, + "step": 23796 + }, + { + "epoch": 2.516603214890017, + "grad_norm": 8.25, + "learning_rate": 2.5945964753183024e-06, + "loss": 2.4231, + "step": 23797 + }, + { + "epoch": 2.5167089678511, + "grad_norm": 8.5, + "learning_rate": 2.594425341516407e-06, + "loss": 2.508, + "step": 23798 + }, + { + "epoch": 2.5168147208121825, + "grad_norm": 8.0, + "learning_rate": 2.594254207271412e-06, + "loss": 2.5509, + "step": 23799 + }, + { + "epoch": 2.5169204737732658, + "grad_norm": 7.8125, + "learning_rate": 2.594083072584121e-06, + "loss": 2.4999, + "step": 23800 + }, + { + "epoch": 2.5170262267343486, + "grad_norm": 8.25, + "learning_rate": 2.593911937455337e-06, + "loss": 2.3387, + "step": 23801 + }, + { + "epoch": 2.5171319796954315, + "grad_norm": 8.25, + "learning_rate": 2.5937408018858613e-06, + "loss": 2.4101, + "step": 23802 + }, + { + "epoch": 2.5172377326565143, + "grad_norm": 8.8125, + "learning_rate": 2.5935696658764987e-06, + "loss": 2.6841, + "step": 23803 + }, + { + "epoch": 2.517343485617597, + "grad_norm": 8.25, + "learning_rate": 2.593398529428053e-06, + "loss": 2.3922, + "step": 23804 + }, + { + "epoch": 2.5174492385786804, + "grad_norm": 8.0, + "learning_rate": 2.5932273925413255e-06, + "loss": 2.5244, + "step": 23805 + }, + { + "epoch": 2.5175549915397633, + "grad_norm": 7.9375, + "learning_rate": 2.5930562552171196e-06, + "loss": 2.4095, + "step": 23806 + }, + { + "epoch": 2.517660744500846, + "grad_norm": 8.5, + "learning_rate": 2.5928851174562394e-06, + "loss": 2.7622, + "step": 23807 + }, + { + "epoch": 2.517766497461929, + "grad_norm": 9.0, + "learning_rate": 2.5927139792594874e-06, + "loss": 2.5194, + "step": 23808 + }, + { + "epoch": 2.5178722504230118, + "grad_norm": 8.5, + "learning_rate": 2.592542840627666e-06, + "loss": 2.5387, + "step": 23809 + }, + { + "epoch": 2.5179780033840946, + "grad_norm": 8.6875, + "learning_rate": 2.592371701561579e-06, + "loss": 2.4746, + "step": 23810 + }, + { + "epoch": 2.5180837563451774, + "grad_norm": 8.5625, + "learning_rate": 2.5922005620620296e-06, + "loss": 2.727, + "step": 23811 + }, + { + "epoch": 2.5181895093062607, + "grad_norm": 8.6875, + "learning_rate": 2.592029422129821e-06, + "loss": 2.4762, + "step": 23812 + }, + { + "epoch": 2.5182952622673436, + "grad_norm": 9.125, + "learning_rate": 2.591858281765755e-06, + "loss": 2.5223, + "step": 23813 + }, + { + "epoch": 2.5184010152284264, + "grad_norm": 8.1875, + "learning_rate": 2.5916871409706367e-06, + "loss": 2.4939, + "step": 23814 + }, + { + "epoch": 2.5185067681895092, + "grad_norm": 8.6875, + "learning_rate": 2.5915159997452672e-06, + "loss": 2.5732, + "step": 23815 + }, + { + "epoch": 2.518612521150592, + "grad_norm": 8.6875, + "learning_rate": 2.5913448580904516e-06, + "loss": 2.5765, + "step": 23816 + }, + { + "epoch": 2.5187182741116754, + "grad_norm": 9.1875, + "learning_rate": 2.5911737160069916e-06, + "loss": 2.6378, + "step": 23817 + }, + { + "epoch": 2.5188240270727578, + "grad_norm": 8.1875, + "learning_rate": 2.59100257349569e-06, + "loss": 2.5003, + "step": 23818 + }, + { + "epoch": 2.518929780033841, + "grad_norm": 8.625, + "learning_rate": 2.5908314305573508e-06, + "loss": 2.5738, + "step": 23819 + }, + { + "epoch": 2.519035532994924, + "grad_norm": 8.4375, + "learning_rate": 2.590660287192777e-06, + "loss": 2.2835, + "step": 23820 + }, + { + "epoch": 2.5191412859560067, + "grad_norm": 8.75, + "learning_rate": 2.590489143402772e-06, + "loss": 2.7448, + "step": 23821 + }, + { + "epoch": 2.5192470389170896, + "grad_norm": 8.375, + "learning_rate": 2.5903179991881376e-06, + "loss": 2.7158, + "step": 23822 + }, + { + "epoch": 2.5193527918781724, + "grad_norm": 8.25, + "learning_rate": 2.590146854549678e-06, + "loss": 2.5298, + "step": 23823 + }, + { + "epoch": 2.5194585448392557, + "grad_norm": 8.3125, + "learning_rate": 2.589975709488197e-06, + "loss": 2.5102, + "step": 23824 + }, + { + "epoch": 2.5195642978003385, + "grad_norm": 8.9375, + "learning_rate": 2.5898045640044954e-06, + "loss": 2.6115, + "step": 23825 + }, + { + "epoch": 2.5196700507614214, + "grad_norm": 8.125, + "learning_rate": 2.5896334180993786e-06, + "loss": 2.2953, + "step": 23826 + }, + { + "epoch": 2.519775803722504, + "grad_norm": 8.875, + "learning_rate": 2.589462271773649e-06, + "loss": 2.58, + "step": 23827 + }, + { + "epoch": 2.519881556683587, + "grad_norm": 8.0, + "learning_rate": 2.589291125028109e-06, + "loss": 2.5612, + "step": 23828 + }, + { + "epoch": 2.5199873096446703, + "grad_norm": 7.9375, + "learning_rate": 2.5891199778635622e-06, + "loss": 2.4315, + "step": 23829 + }, + { + "epoch": 2.5200930626057527, + "grad_norm": 8.4375, + "learning_rate": 2.588948830280812e-06, + "loss": 2.6485, + "step": 23830 + }, + { + "epoch": 2.520198815566836, + "grad_norm": 7.875, + "learning_rate": 2.5887776822806616e-06, + "loss": 2.501, + "step": 23831 + }, + { + "epoch": 2.520304568527919, + "grad_norm": 8.875, + "learning_rate": 2.5886065338639134e-06, + "loss": 2.7445, + "step": 23832 + }, + { + "epoch": 2.5204103214890017, + "grad_norm": 9.0625, + "learning_rate": 2.5884353850313716e-06, + "loss": 2.6483, + "step": 23833 + }, + { + "epoch": 2.5205160744500845, + "grad_norm": 9.0625, + "learning_rate": 2.5882642357838378e-06, + "loss": 2.4407, + "step": 23834 + }, + { + "epoch": 2.5206218274111674, + "grad_norm": 8.875, + "learning_rate": 2.5880930861221167e-06, + "loss": 2.7652, + "step": 23835 + }, + { + "epoch": 2.5207275803722506, + "grad_norm": 8.3125, + "learning_rate": 2.587921936047011e-06, + "loss": 2.4982, + "step": 23836 + }, + { + "epoch": 2.5208333333333335, + "grad_norm": 8.25, + "learning_rate": 2.5877507855593227e-06, + "loss": 2.5266, + "step": 23837 + }, + { + "epoch": 2.5209390862944163, + "grad_norm": 8.5, + "learning_rate": 2.5875796346598556e-06, + "loss": 2.7713, + "step": 23838 + }, + { + "epoch": 2.521044839255499, + "grad_norm": 7.84375, + "learning_rate": 2.5874084833494146e-06, + "loss": 2.4723, + "step": 23839 + }, + { + "epoch": 2.521150592216582, + "grad_norm": 8.5, + "learning_rate": 2.587237331628801e-06, + "loss": 2.649, + "step": 23840 + }, + { + "epoch": 2.521256345177665, + "grad_norm": 8.875, + "learning_rate": 2.5870661794988173e-06, + "loss": 2.8285, + "step": 23841 + }, + { + "epoch": 2.5213620981387477, + "grad_norm": 9.125, + "learning_rate": 2.5868950269602682e-06, + "loss": 2.4403, + "step": 23842 + }, + { + "epoch": 2.521467851099831, + "grad_norm": 8.125, + "learning_rate": 2.5867238740139567e-06, + "loss": 2.4357, + "step": 23843 + }, + { + "epoch": 2.521573604060914, + "grad_norm": 8.6875, + "learning_rate": 2.5865527206606844e-06, + "loss": 2.6356, + "step": 23844 + }, + { + "epoch": 2.5216793570219966, + "grad_norm": 9.0, + "learning_rate": 2.5863815669012564e-06, + "loss": 2.5247, + "step": 23845 + }, + { + "epoch": 2.5217851099830795, + "grad_norm": 8.375, + "learning_rate": 2.5862104127364744e-06, + "loss": 2.5296, + "step": 23846 + }, + { + "epoch": 2.5218908629441623, + "grad_norm": 9.1875, + "learning_rate": 2.5860392581671435e-06, + "loss": 2.5195, + "step": 23847 + }, + { + "epoch": 2.5219966159052456, + "grad_norm": 8.0, + "learning_rate": 2.5858681031940643e-06, + "loss": 2.5244, + "step": 23848 + }, + { + "epoch": 2.522102368866328, + "grad_norm": 8.3125, + "learning_rate": 2.5856969478180417e-06, + "loss": 2.438, + "step": 23849 + }, + { + "epoch": 2.5222081218274113, + "grad_norm": 8.375, + "learning_rate": 2.585525792039878e-06, + "loss": 2.4594, + "step": 23850 + }, + { + "epoch": 2.522313874788494, + "grad_norm": 8.375, + "learning_rate": 2.5853546358603777e-06, + "loss": 2.5611, + "step": 23851 + }, + { + "epoch": 2.522419627749577, + "grad_norm": 8.0625, + "learning_rate": 2.5851834792803422e-06, + "loss": 2.5451, + "step": 23852 + }, + { + "epoch": 2.52252538071066, + "grad_norm": 8.875, + "learning_rate": 2.5850123223005754e-06, + "loss": 2.4811, + "step": 23853 + }, + { + "epoch": 2.5226311336717426, + "grad_norm": 8.75, + "learning_rate": 2.5848411649218815e-06, + "loss": 2.441, + "step": 23854 + }, + { + "epoch": 2.522736886632826, + "grad_norm": 8.625, + "learning_rate": 2.5846700071450613e-06, + "loss": 2.4688, + "step": 23855 + }, + { + "epoch": 2.5228426395939088, + "grad_norm": 8.25, + "learning_rate": 2.5844988489709205e-06, + "loss": 2.2774, + "step": 23856 + }, + { + "epoch": 2.5229483925549916, + "grad_norm": 8.875, + "learning_rate": 2.58432769040026e-06, + "loss": 2.6519, + "step": 23857 + }, + { + "epoch": 2.5230541455160744, + "grad_norm": 8.75, + "learning_rate": 2.584156531433885e-06, + "loss": 2.7159, + "step": 23858 + }, + { + "epoch": 2.5231598984771573, + "grad_norm": 9.1875, + "learning_rate": 2.5839853720725984e-06, + "loss": 2.7151, + "step": 23859 + }, + { + "epoch": 2.52326565143824, + "grad_norm": 8.25, + "learning_rate": 2.5838142123172016e-06, + "loss": 2.5726, + "step": 23860 + }, + { + "epoch": 2.523371404399323, + "grad_norm": 8.0625, + "learning_rate": 2.5836430521684994e-06, + "loss": 2.4782, + "step": 23861 + }, + { + "epoch": 2.5234771573604062, + "grad_norm": 8.375, + "learning_rate": 2.583471891627295e-06, + "loss": 2.5187, + "step": 23862 + }, + { + "epoch": 2.523582910321489, + "grad_norm": 9.4375, + "learning_rate": 2.5833007306943906e-06, + "loss": 2.5599, + "step": 23863 + }, + { + "epoch": 2.523688663282572, + "grad_norm": 8.75, + "learning_rate": 2.58312956937059e-06, + "loss": 2.7394, + "step": 23864 + }, + { + "epoch": 2.5237944162436547, + "grad_norm": 8.625, + "learning_rate": 2.5829584076566967e-06, + "loss": 2.6301, + "step": 23865 + }, + { + "epoch": 2.5239001692047376, + "grad_norm": 7.96875, + "learning_rate": 2.5827872455535134e-06, + "loss": 2.3341, + "step": 23866 + }, + { + "epoch": 2.524005922165821, + "grad_norm": 8.125, + "learning_rate": 2.5826160830618428e-06, + "loss": 2.2688, + "step": 23867 + }, + { + "epoch": 2.5241116751269037, + "grad_norm": 8.875, + "learning_rate": 2.58244492018249e-06, + "loss": 2.6334, + "step": 23868 + }, + { + "epoch": 2.5242174280879865, + "grad_norm": 9.125, + "learning_rate": 2.5822737569162554e-06, + "loss": 2.5447, + "step": 23869 + }, + { + "epoch": 2.5243231810490694, + "grad_norm": 8.9375, + "learning_rate": 2.582102593263944e-06, + "loss": 2.6131, + "step": 23870 + }, + { + "epoch": 2.5244289340101522, + "grad_norm": 8.0625, + "learning_rate": 2.5819314292263596e-06, + "loss": 2.5123, + "step": 23871 + }, + { + "epoch": 2.524534686971235, + "grad_norm": 8.625, + "learning_rate": 2.581760264804304e-06, + "loss": 2.4364, + "step": 23872 + }, + { + "epoch": 2.524640439932318, + "grad_norm": 7.90625, + "learning_rate": 2.5815890999985806e-06, + "loss": 2.5311, + "step": 23873 + }, + { + "epoch": 2.524746192893401, + "grad_norm": 7.96875, + "learning_rate": 2.581417934809994e-06, + "loss": 2.5001, + "step": 23874 + }, + { + "epoch": 2.524851945854484, + "grad_norm": 8.0625, + "learning_rate": 2.5812467692393458e-06, + "loss": 2.6212, + "step": 23875 + }, + { + "epoch": 2.524957698815567, + "grad_norm": 9.0, + "learning_rate": 2.5810756032874392e-06, + "loss": 2.5092, + "step": 23876 + }, + { + "epoch": 2.5250634517766497, + "grad_norm": 8.0625, + "learning_rate": 2.5809044369550785e-06, + "loss": 2.5567, + "step": 23877 + }, + { + "epoch": 2.5251692047377325, + "grad_norm": 8.5625, + "learning_rate": 2.580733270243066e-06, + "loss": 2.5986, + "step": 23878 + }, + { + "epoch": 2.525274957698816, + "grad_norm": 8.5, + "learning_rate": 2.5805621031522057e-06, + "loss": 2.5431, + "step": 23879 + }, + { + "epoch": 2.525380710659898, + "grad_norm": 8.3125, + "learning_rate": 2.5803909356833e-06, + "loss": 2.572, + "step": 23880 + }, + { + "epoch": 2.5254864636209815, + "grad_norm": 8.6875, + "learning_rate": 2.580219767837152e-06, + "loss": 2.613, + "step": 23881 + }, + { + "epoch": 2.5255922165820643, + "grad_norm": 8.1875, + "learning_rate": 2.5800485996145665e-06, + "loss": 2.5066, + "step": 23882 + }, + { + "epoch": 2.525697969543147, + "grad_norm": 9.0625, + "learning_rate": 2.579877431016345e-06, + "loss": 2.7612, + "step": 23883 + }, + { + "epoch": 2.52580372250423, + "grad_norm": 8.4375, + "learning_rate": 2.5797062620432915e-06, + "loss": 2.6067, + "step": 23884 + }, + { + "epoch": 2.525909475465313, + "grad_norm": 8.5625, + "learning_rate": 2.5795350926962097e-06, + "loss": 2.6339, + "step": 23885 + }, + { + "epoch": 2.526015228426396, + "grad_norm": 8.125, + "learning_rate": 2.5793639229759014e-06, + "loss": 2.556, + "step": 23886 + }, + { + "epoch": 2.526120981387479, + "grad_norm": 7.71875, + "learning_rate": 2.579192752883172e-06, + "loss": 2.6053, + "step": 23887 + }, + { + "epoch": 2.526226734348562, + "grad_norm": 7.6875, + "learning_rate": 2.579021582418822e-06, + "loss": 2.2826, + "step": 23888 + }, + { + "epoch": 2.5263324873096447, + "grad_norm": 9.0, + "learning_rate": 2.5788504115836567e-06, + "loss": 2.6403, + "step": 23889 + }, + { + "epoch": 2.5264382402707275, + "grad_norm": 9.0, + "learning_rate": 2.5786792403784777e-06, + "loss": 2.5382, + "step": 23890 + }, + { + "epoch": 2.5265439932318103, + "grad_norm": 9.0, + "learning_rate": 2.5785080688040905e-06, + "loss": 2.5355, + "step": 23891 + }, + { + "epoch": 2.526649746192893, + "grad_norm": 8.1875, + "learning_rate": 2.578336896861296e-06, + "loss": 2.649, + "step": 23892 + }, + { + "epoch": 2.5267554991539765, + "grad_norm": 8.5625, + "learning_rate": 2.5781657245508995e-06, + "loss": 2.4861, + "step": 23893 + }, + { + "epoch": 2.5268612521150593, + "grad_norm": 8.1875, + "learning_rate": 2.5779945518737026e-06, + "loss": 2.6478, + "step": 23894 + }, + { + "epoch": 2.526967005076142, + "grad_norm": 8.25, + "learning_rate": 2.577823378830509e-06, + "loss": 2.5809, + "step": 23895 + }, + { + "epoch": 2.527072758037225, + "grad_norm": 8.625, + "learning_rate": 2.5776522054221227e-06, + "loss": 2.6286, + "step": 23896 + }, + { + "epoch": 2.527178510998308, + "grad_norm": 8.0, + "learning_rate": 2.5774810316493467e-06, + "loss": 2.5783, + "step": 23897 + }, + { + "epoch": 2.527284263959391, + "grad_norm": 8.8125, + "learning_rate": 2.577309857512983e-06, + "loss": 2.5422, + "step": 23898 + }, + { + "epoch": 2.527390016920474, + "grad_norm": 8.75, + "learning_rate": 2.577138683013835e-06, + "loss": 2.5716, + "step": 23899 + }, + { + "epoch": 2.5274957698815568, + "grad_norm": 8.0625, + "learning_rate": 2.5769675081527085e-06, + "loss": 2.6221, + "step": 23900 + }, + { + "epoch": 2.5276015228426396, + "grad_norm": 8.0625, + "learning_rate": 2.576796332930404e-06, + "loss": 2.4336, + "step": 23901 + }, + { + "epoch": 2.5277072758037225, + "grad_norm": 8.1875, + "learning_rate": 2.5766251573477258e-06, + "loss": 2.5579, + "step": 23902 + }, + { + "epoch": 2.5278130287648053, + "grad_norm": 7.71875, + "learning_rate": 2.5764539814054773e-06, + "loss": 2.4391, + "step": 23903 + }, + { + "epoch": 2.527918781725888, + "grad_norm": 9.125, + "learning_rate": 2.576282805104462e-06, + "loss": 2.659, + "step": 23904 + }, + { + "epoch": 2.5280245346869714, + "grad_norm": 7.78125, + "learning_rate": 2.5761116284454818e-06, + "loss": 2.3109, + "step": 23905 + }, + { + "epoch": 2.5281302876480543, + "grad_norm": 8.5, + "learning_rate": 2.575940451429341e-06, + "loss": 2.5218, + "step": 23906 + }, + { + "epoch": 2.528236040609137, + "grad_norm": 8.625, + "learning_rate": 2.575769274056843e-06, + "loss": 2.8422, + "step": 23907 + }, + { + "epoch": 2.52834179357022, + "grad_norm": 7.9375, + "learning_rate": 2.57559809632879e-06, + "loss": 2.4775, + "step": 23908 + }, + { + "epoch": 2.5284475465313028, + "grad_norm": 8.75, + "learning_rate": 2.575426918245987e-06, + "loss": 2.6755, + "step": 23909 + }, + { + "epoch": 2.528553299492386, + "grad_norm": 8.25, + "learning_rate": 2.5752557398092364e-06, + "loss": 2.5495, + "step": 23910 + }, + { + "epoch": 2.5286590524534684, + "grad_norm": 8.125, + "learning_rate": 2.575084561019341e-06, + "loss": 2.3572, + "step": 23911 + }, + { + "epoch": 2.5287648054145517, + "grad_norm": 8.375, + "learning_rate": 2.5749133818771046e-06, + "loss": 2.49, + "step": 23912 + }, + { + "epoch": 2.5288705583756346, + "grad_norm": 8.375, + "learning_rate": 2.5747422023833304e-06, + "loss": 2.4686, + "step": 23913 + }, + { + "epoch": 2.5289763113367174, + "grad_norm": 8.4375, + "learning_rate": 2.574571022538821e-06, + "loss": 2.6735, + "step": 23914 + }, + { + "epoch": 2.5290820642978002, + "grad_norm": 9.6875, + "learning_rate": 2.574399842344381e-06, + "loss": 2.5844, + "step": 23915 + }, + { + "epoch": 2.529187817258883, + "grad_norm": 8.1875, + "learning_rate": 2.574228661800813e-06, + "loss": 2.4786, + "step": 23916 + }, + { + "epoch": 2.5292935702199664, + "grad_norm": 8.0625, + "learning_rate": 2.5740574809089197e-06, + "loss": 2.4486, + "step": 23917 + }, + { + "epoch": 2.529399323181049, + "grad_norm": 8.0625, + "learning_rate": 2.5738862996695046e-06, + "loss": 2.4399, + "step": 23918 + }, + { + "epoch": 2.529505076142132, + "grad_norm": 8.5, + "learning_rate": 2.5737151180833726e-06, + "loss": 2.698, + "step": 23919 + }, + { + "epoch": 2.529610829103215, + "grad_norm": 8.3125, + "learning_rate": 2.5735439361513247e-06, + "loss": 2.553, + "step": 23920 + }, + { + "epoch": 2.5297165820642977, + "grad_norm": 8.0625, + "learning_rate": 2.5733727538741655e-06, + "loss": 2.4825, + "step": 23921 + }, + { + "epoch": 2.5298223350253806, + "grad_norm": 8.875, + "learning_rate": 2.573201571252698e-06, + "loss": 2.4577, + "step": 23922 + }, + { + "epoch": 2.5299280879864634, + "grad_norm": 8.6875, + "learning_rate": 2.573030388287725e-06, + "loss": 2.5557, + "step": 23923 + }, + { + "epoch": 2.5300338409475467, + "grad_norm": 8.25, + "learning_rate": 2.572859204980051e-06, + "loss": 2.596, + "step": 23924 + }, + { + "epoch": 2.5301395939086295, + "grad_norm": 8.625, + "learning_rate": 2.572688021330478e-06, + "loss": 2.4428, + "step": 23925 + }, + { + "epoch": 2.5302453468697124, + "grad_norm": 8.3125, + "learning_rate": 2.57251683733981e-06, + "loss": 2.6153, + "step": 23926 + }, + { + "epoch": 2.530351099830795, + "grad_norm": 8.25, + "learning_rate": 2.5723456530088505e-06, + "loss": 2.6052, + "step": 23927 + }, + { + "epoch": 2.530456852791878, + "grad_norm": 8.5, + "learning_rate": 2.572174468338402e-06, + "loss": 2.5609, + "step": 23928 + }, + { + "epoch": 2.5305626057529613, + "grad_norm": 8.25, + "learning_rate": 2.5720032833292687e-06, + "loss": 2.5689, + "step": 23929 + }, + { + "epoch": 2.5306683587140437, + "grad_norm": 8.1875, + "learning_rate": 2.5718320979822532e-06, + "loss": 2.4849, + "step": 23930 + }, + { + "epoch": 2.530774111675127, + "grad_norm": 8.625, + "learning_rate": 2.571660912298159e-06, + "loss": 2.6101, + "step": 23931 + }, + { + "epoch": 2.53087986463621, + "grad_norm": 8.6875, + "learning_rate": 2.571489726277789e-06, + "loss": 2.401, + "step": 23932 + }, + { + "epoch": 2.5309856175972927, + "grad_norm": 9.375, + "learning_rate": 2.5713185399219474e-06, + "loss": 2.5027, + "step": 23933 + }, + { + "epoch": 2.5310913705583755, + "grad_norm": 8.375, + "learning_rate": 2.571147353231437e-06, + "loss": 2.6946, + "step": 23934 + }, + { + "epoch": 2.5311971235194584, + "grad_norm": 8.75, + "learning_rate": 2.5709761662070616e-06, + "loss": 2.4947, + "step": 23935 + }, + { + "epoch": 2.5313028764805416, + "grad_norm": 9.3125, + "learning_rate": 2.5708049788496237e-06, + "loss": 2.5168, + "step": 23936 + }, + { + "epoch": 2.5314086294416245, + "grad_norm": 8.375, + "learning_rate": 2.5706337911599265e-06, + "loss": 2.603, + "step": 23937 + }, + { + "epoch": 2.5315143824027073, + "grad_norm": 8.75, + "learning_rate": 2.5704626031387747e-06, + "loss": 2.5181, + "step": 23938 + }, + { + "epoch": 2.53162013536379, + "grad_norm": 9.5, + "learning_rate": 2.5702914147869704e-06, + "loss": 2.434, + "step": 23939 + }, + { + "epoch": 2.531725888324873, + "grad_norm": 8.5, + "learning_rate": 2.570120226105317e-06, + "loss": 2.6984, + "step": 23940 + }, + { + "epoch": 2.5318316412859563, + "grad_norm": 8.3125, + "learning_rate": 2.5699490370946183e-06, + "loss": 2.5991, + "step": 23941 + }, + { + "epoch": 2.5319373942470387, + "grad_norm": 8.125, + "learning_rate": 2.5697778477556772e-06, + "loss": 2.5165, + "step": 23942 + }, + { + "epoch": 2.532043147208122, + "grad_norm": 8.6875, + "learning_rate": 2.569606658089297e-06, + "loss": 2.547, + "step": 23943 + }, + { + "epoch": 2.532148900169205, + "grad_norm": 9.3125, + "learning_rate": 2.5694354680962812e-06, + "loss": 2.5997, + "step": 23944 + }, + { + "epoch": 2.5322546531302876, + "grad_norm": 9.25, + "learning_rate": 2.569264277777434e-06, + "loss": 2.7074, + "step": 23945 + }, + { + "epoch": 2.5323604060913705, + "grad_norm": 8.8125, + "learning_rate": 2.569093087133557e-06, + "loss": 2.5427, + "step": 23946 + }, + { + "epoch": 2.5324661590524533, + "grad_norm": 8.0625, + "learning_rate": 2.568921896165455e-06, + "loss": 2.5287, + "step": 23947 + }, + { + "epoch": 2.5325719120135366, + "grad_norm": 8.4375, + "learning_rate": 2.5687507048739307e-06, + "loss": 2.527, + "step": 23948 + }, + { + "epoch": 2.5326776649746194, + "grad_norm": 8.6875, + "learning_rate": 2.5685795132597867e-06, + "loss": 2.1607, + "step": 23949 + }, + { + "epoch": 2.5327834179357023, + "grad_norm": 8.6875, + "learning_rate": 2.568408321323828e-06, + "loss": 2.7443, + "step": 23950 + }, + { + "epoch": 2.532889170896785, + "grad_norm": 7.875, + "learning_rate": 2.568237129066857e-06, + "loss": 2.3594, + "step": 23951 + }, + { + "epoch": 2.532994923857868, + "grad_norm": 7.96875, + "learning_rate": 2.5680659364896766e-06, + "loss": 2.384, + "step": 23952 + }, + { + "epoch": 2.533100676818951, + "grad_norm": 8.5625, + "learning_rate": 2.56789474359309e-06, + "loss": 2.3022, + "step": 23953 + }, + { + "epoch": 2.5332064297800336, + "grad_norm": 8.125, + "learning_rate": 2.5677235503779017e-06, + "loss": 2.6457, + "step": 23954 + }, + { + "epoch": 2.533312182741117, + "grad_norm": 7.75, + "learning_rate": 2.567552356844915e-06, + "loss": 2.5388, + "step": 23955 + }, + { + "epoch": 2.5334179357021998, + "grad_norm": 8.625, + "learning_rate": 2.5673811629949323e-06, + "loss": 2.7397, + "step": 23956 + }, + { + "epoch": 2.5335236886632826, + "grad_norm": 8.125, + "learning_rate": 2.5672099688287577e-06, + "loss": 2.6154, + "step": 23957 + }, + { + "epoch": 2.5336294416243654, + "grad_norm": 8.5, + "learning_rate": 2.567038774347194e-06, + "loss": 2.4728, + "step": 23958 + }, + { + "epoch": 2.5337351945854483, + "grad_norm": 8.75, + "learning_rate": 2.5668675795510447e-06, + "loss": 2.6699, + "step": 23959 + }, + { + "epoch": 2.5338409475465316, + "grad_norm": 9.4375, + "learning_rate": 2.566696384441113e-06, + "loss": 2.585, + "step": 23960 + }, + { + "epoch": 2.533946700507614, + "grad_norm": 8.375, + "learning_rate": 2.566525189018202e-06, + "loss": 2.5377, + "step": 23961 + }, + { + "epoch": 2.5340524534686972, + "grad_norm": 8.625, + "learning_rate": 2.566353993283116e-06, + "loss": 2.3243, + "step": 23962 + }, + { + "epoch": 2.53415820642978, + "grad_norm": 9.125, + "learning_rate": 2.5661827972366583e-06, + "loss": 2.6083, + "step": 23963 + }, + { + "epoch": 2.534263959390863, + "grad_norm": 8.3125, + "learning_rate": 2.5660116008796317e-06, + "loss": 2.4918, + "step": 23964 + }, + { + "epoch": 2.5343697123519457, + "grad_norm": 8.1875, + "learning_rate": 2.5658404042128387e-06, + "loss": 2.617, + "step": 23965 + }, + { + "epoch": 2.5344754653130286, + "grad_norm": 9.875, + "learning_rate": 2.5656692072370846e-06, + "loss": 2.6246, + "step": 23966 + }, + { + "epoch": 2.534581218274112, + "grad_norm": 8.5625, + "learning_rate": 2.565498009953172e-06, + "loss": 2.7397, + "step": 23967 + }, + { + "epoch": 2.5346869712351947, + "grad_norm": 8.8125, + "learning_rate": 2.565326812361903e-06, + "loss": 2.5844, + "step": 23968 + }, + { + "epoch": 2.5347927241962775, + "grad_norm": 8.75, + "learning_rate": 2.5651556144640827e-06, + "loss": 2.5288, + "step": 23969 + }, + { + "epoch": 2.5348984771573604, + "grad_norm": 8.3125, + "learning_rate": 2.5649844162605133e-06, + "loss": 2.557, + "step": 23970 + }, + { + "epoch": 2.5350042301184432, + "grad_norm": 8.5625, + "learning_rate": 2.5648132177519984e-06, + "loss": 2.5618, + "step": 23971 + }, + { + "epoch": 2.535109983079526, + "grad_norm": 8.3125, + "learning_rate": 2.5646420189393423e-06, + "loss": 2.1522, + "step": 23972 + }, + { + "epoch": 2.535215736040609, + "grad_norm": 8.1875, + "learning_rate": 2.5644708198233477e-06, + "loss": 2.4971, + "step": 23973 + }, + { + "epoch": 2.535321489001692, + "grad_norm": 8.5, + "learning_rate": 2.5642996204048174e-06, + "loss": 2.3244, + "step": 23974 + }, + { + "epoch": 2.535427241962775, + "grad_norm": 8.1875, + "learning_rate": 2.564128420684555e-06, + "loss": 2.1068, + "step": 23975 + }, + { + "epoch": 2.535532994923858, + "grad_norm": 8.8125, + "learning_rate": 2.563957220663365e-06, + "loss": 2.7892, + "step": 23976 + }, + { + "epoch": 2.5356387478849407, + "grad_norm": 8.5, + "learning_rate": 2.5637860203420493e-06, + "loss": 2.5875, + "step": 23977 + }, + { + "epoch": 2.5357445008460235, + "grad_norm": 8.4375, + "learning_rate": 2.5636148197214124e-06, + "loss": 2.6854, + "step": 23978 + }, + { + "epoch": 2.535850253807107, + "grad_norm": 8.5625, + "learning_rate": 2.5634436188022566e-06, + "loss": 2.567, + "step": 23979 + }, + { + "epoch": 2.5359560067681897, + "grad_norm": 7.96875, + "learning_rate": 2.5632724175853865e-06, + "loss": 2.554, + "step": 23980 + }, + { + "epoch": 2.5360617597292725, + "grad_norm": 8.6875, + "learning_rate": 2.563101216071604e-06, + "loss": 2.5243, + "step": 23981 + }, + { + "epoch": 2.5361675126903553, + "grad_norm": 8.4375, + "learning_rate": 2.5629300142617136e-06, + "loss": 2.4649, + "step": 23982 + }, + { + "epoch": 2.536273265651438, + "grad_norm": 9.75, + "learning_rate": 2.562758812156519e-06, + "loss": 2.5884, + "step": 23983 + }, + { + "epoch": 2.536379018612521, + "grad_norm": 8.625, + "learning_rate": 2.562587609756822e-06, + "loss": 2.6959, + "step": 23984 + }, + { + "epoch": 2.536484771573604, + "grad_norm": 8.625, + "learning_rate": 2.5624164070634277e-06, + "loss": 2.5568, + "step": 23985 + }, + { + "epoch": 2.536590524534687, + "grad_norm": 8.3125, + "learning_rate": 2.5622452040771387e-06, + "loss": 2.4987, + "step": 23986 + }, + { + "epoch": 2.53669627749577, + "grad_norm": 8.5, + "learning_rate": 2.5620740007987577e-06, + "loss": 2.7581, + "step": 23987 + }, + { + "epoch": 2.536802030456853, + "grad_norm": 8.8125, + "learning_rate": 2.561902797229089e-06, + "loss": 2.4712, + "step": 23988 + }, + { + "epoch": 2.5369077834179357, + "grad_norm": 9.0, + "learning_rate": 2.5617315933689363e-06, + "loss": 2.4034, + "step": 23989 + }, + { + "epoch": 2.5370135363790185, + "grad_norm": 8.75, + "learning_rate": 2.561560389219103e-06, + "loss": 2.493, + "step": 23990 + }, + { + "epoch": 2.537119289340102, + "grad_norm": 8.375, + "learning_rate": 2.561389184780391e-06, + "loss": 2.3795, + "step": 23991 + }, + { + "epoch": 2.537225042301184, + "grad_norm": 8.75, + "learning_rate": 2.5612179800536046e-06, + "loss": 2.5782, + "step": 23992 + }, + { + "epoch": 2.5373307952622675, + "grad_norm": 8.5625, + "learning_rate": 2.561046775039548e-06, + "loss": 2.7222, + "step": 23993 + }, + { + "epoch": 2.5374365482233503, + "grad_norm": 9.0625, + "learning_rate": 2.5608755697390224e-06, + "loss": 2.4031, + "step": 23994 + }, + { + "epoch": 2.537542301184433, + "grad_norm": 10.625, + "learning_rate": 2.560704364152834e-06, + "loss": 2.6893, + "step": 23995 + }, + { + "epoch": 2.537648054145516, + "grad_norm": 8.6875, + "learning_rate": 2.5605331582817842e-06, + "loss": 2.6118, + "step": 23996 + }, + { + "epoch": 2.537753807106599, + "grad_norm": 8.4375, + "learning_rate": 2.560361952126677e-06, + "loss": 2.5189, + "step": 23997 + }, + { + "epoch": 2.537859560067682, + "grad_norm": 8.4375, + "learning_rate": 2.560190745688317e-06, + "loss": 2.4899, + "step": 23998 + }, + { + "epoch": 2.537965313028765, + "grad_norm": 8.75, + "learning_rate": 2.560019538967505e-06, + "loss": 2.496, + "step": 23999 + }, + { + "epoch": 2.5380710659898478, + "grad_norm": 8.6875, + "learning_rate": 2.5598483319650464e-06, + "loss": 2.4145, + "step": 24000 + }, + { + "epoch": 2.5381768189509306, + "grad_norm": 8.5625, + "learning_rate": 2.559677124681745e-06, + "loss": 2.4129, + "step": 24001 + }, + { + "epoch": 2.5382825719120135, + "grad_norm": 8.125, + "learning_rate": 2.559505917118402e-06, + "loss": 2.3792, + "step": 24002 + }, + { + "epoch": 2.5383883248730963, + "grad_norm": 8.4375, + "learning_rate": 2.5593347092758225e-06, + "loss": 2.1607, + "step": 24003 + }, + { + "epoch": 2.538494077834179, + "grad_norm": 8.25, + "learning_rate": 2.5591635011548093e-06, + "loss": 2.4613, + "step": 24004 + }, + { + "epoch": 2.5385998307952624, + "grad_norm": 8.75, + "learning_rate": 2.558992292756166e-06, + "loss": 2.5402, + "step": 24005 + }, + { + "epoch": 2.5387055837563453, + "grad_norm": 8.8125, + "learning_rate": 2.5588210840806966e-06, + "loss": 2.4618, + "step": 24006 + }, + { + "epoch": 2.538811336717428, + "grad_norm": 8.375, + "learning_rate": 2.5586498751292027e-06, + "loss": 2.3376, + "step": 24007 + }, + { + "epoch": 2.538917089678511, + "grad_norm": 8.8125, + "learning_rate": 2.55847866590249e-06, + "loss": 2.7589, + "step": 24008 + }, + { + "epoch": 2.5390228426395938, + "grad_norm": 8.5, + "learning_rate": 2.5583074564013603e-06, + "loss": 2.5512, + "step": 24009 + }, + { + "epoch": 2.539128595600677, + "grad_norm": 8.3125, + "learning_rate": 2.558136246626618e-06, + "loss": 2.4387, + "step": 24010 + }, + { + "epoch": 2.53923434856176, + "grad_norm": 8.1875, + "learning_rate": 2.5579650365790657e-06, + "loss": 2.1135, + "step": 24011 + }, + { + "epoch": 2.5393401015228427, + "grad_norm": 8.375, + "learning_rate": 2.5577938262595078e-06, + "loss": 2.6534, + "step": 24012 + }, + { + "epoch": 2.5394458544839256, + "grad_norm": 9.0, + "learning_rate": 2.5576226156687463e-06, + "loss": 2.7526, + "step": 24013 + }, + { + "epoch": 2.5395516074450084, + "grad_norm": 8.125, + "learning_rate": 2.5574514048075853e-06, + "loss": 2.6702, + "step": 24014 + }, + { + "epoch": 2.5396573604060912, + "grad_norm": 8.5, + "learning_rate": 2.557280193676829e-06, + "loss": 2.5257, + "step": 24015 + }, + { + "epoch": 2.539763113367174, + "grad_norm": 8.6875, + "learning_rate": 2.55710898227728e-06, + "loss": 2.6543, + "step": 24016 + }, + { + "epoch": 2.5398688663282574, + "grad_norm": 8.375, + "learning_rate": 2.556937770609742e-06, + "loss": 2.6915, + "step": 24017 + }, + { + "epoch": 2.53997461928934, + "grad_norm": 8.375, + "learning_rate": 2.5567665586750186e-06, + "loss": 2.7402, + "step": 24018 + }, + { + "epoch": 2.540080372250423, + "grad_norm": 8.0625, + "learning_rate": 2.556595346473912e-06, + "loss": 2.419, + "step": 24019 + }, + { + "epoch": 2.540186125211506, + "grad_norm": 8.75, + "learning_rate": 2.556424134007227e-06, + "loss": 2.6961, + "step": 24020 + }, + { + "epoch": 2.5402918781725887, + "grad_norm": 8.5, + "learning_rate": 2.556252921275767e-06, + "loss": 2.6544, + "step": 24021 + }, + { + "epoch": 2.540397631133672, + "grad_norm": 8.625, + "learning_rate": 2.5560817082803346e-06, + "loss": 2.4133, + "step": 24022 + }, + { + "epoch": 2.5405033840947544, + "grad_norm": 8.75, + "learning_rate": 2.555910495021734e-06, + "loss": 2.3673, + "step": 24023 + }, + { + "epoch": 2.5406091370558377, + "grad_norm": 8.4375, + "learning_rate": 2.555739281500768e-06, + "loss": 2.5339, + "step": 24024 + }, + { + "epoch": 2.5407148900169205, + "grad_norm": 8.8125, + "learning_rate": 2.5555680677182408e-06, + "loss": 2.561, + "step": 24025 + }, + { + "epoch": 2.5408206429780034, + "grad_norm": 8.5, + "learning_rate": 2.5553968536749553e-06, + "loss": 2.6244, + "step": 24026 + }, + { + "epoch": 2.540926395939086, + "grad_norm": 8.8125, + "learning_rate": 2.5552256393717145e-06, + "loss": 2.6208, + "step": 24027 + }, + { + "epoch": 2.541032148900169, + "grad_norm": 8.5625, + "learning_rate": 2.5550544248093236e-06, + "loss": 2.5501, + "step": 24028 + }, + { + "epoch": 2.5411379018612523, + "grad_norm": 8.75, + "learning_rate": 2.5548832099885834e-06, + "loss": 2.4862, + "step": 24029 + }, + { + "epoch": 2.541243654822335, + "grad_norm": 8.75, + "learning_rate": 2.5547119949102995e-06, + "loss": 2.4619, + "step": 24030 + }, + { + "epoch": 2.541349407783418, + "grad_norm": 9.0625, + "learning_rate": 2.554540779575274e-06, + "loss": 2.5412, + "step": 24031 + }, + { + "epoch": 2.541455160744501, + "grad_norm": 8.1875, + "learning_rate": 2.5543695639843106e-06, + "loss": 2.3859, + "step": 24032 + }, + { + "epoch": 2.5415609137055837, + "grad_norm": 8.8125, + "learning_rate": 2.554198348138214e-06, + "loss": 2.7054, + "step": 24033 + }, + { + "epoch": 2.5416666666666665, + "grad_norm": 8.375, + "learning_rate": 2.5540271320377868e-06, + "loss": 2.157, + "step": 24034 + }, + { + "epoch": 2.5417724196277494, + "grad_norm": 8.4375, + "learning_rate": 2.553855915683832e-06, + "loss": 2.5931, + "step": 24035 + }, + { + "epoch": 2.5418781725888326, + "grad_norm": 8.375, + "learning_rate": 2.5536846990771536e-06, + "loss": 2.6811, + "step": 24036 + }, + { + "epoch": 2.5419839255499155, + "grad_norm": 8.5, + "learning_rate": 2.5535134822185546e-06, + "loss": 2.5959, + "step": 24037 + }, + { + "epoch": 2.5420896785109983, + "grad_norm": 8.125, + "learning_rate": 2.5533422651088387e-06, + "loss": 2.4412, + "step": 24038 + }, + { + "epoch": 2.542195431472081, + "grad_norm": 9.5, + "learning_rate": 2.55317104774881e-06, + "loss": 2.4391, + "step": 24039 + }, + { + "epoch": 2.542301184433164, + "grad_norm": 9.0, + "learning_rate": 2.55299983013927e-06, + "loss": 2.6534, + "step": 24040 + }, + { + "epoch": 2.5424069373942473, + "grad_norm": 8.25, + "learning_rate": 2.5528286122810247e-06, + "loss": 2.677, + "step": 24041 + }, + { + "epoch": 2.5425126903553297, + "grad_norm": 9.375, + "learning_rate": 2.5526573941748754e-06, + "loss": 2.6857, + "step": 24042 + }, + { + "epoch": 2.542618443316413, + "grad_norm": 8.0625, + "learning_rate": 2.5524861758216275e-06, + "loss": 2.3767, + "step": 24043 + }, + { + "epoch": 2.542724196277496, + "grad_norm": 8.3125, + "learning_rate": 2.5523149572220828e-06, + "loss": 2.4863, + "step": 24044 + }, + { + "epoch": 2.5428299492385786, + "grad_norm": 8.1875, + "learning_rate": 2.5521437383770454e-06, + "loss": 2.4074, + "step": 24045 + }, + { + "epoch": 2.5429357021996615, + "grad_norm": 8.75, + "learning_rate": 2.551972519287319e-06, + "loss": 2.5741, + "step": 24046 + }, + { + "epoch": 2.5430414551607443, + "grad_norm": 8.1875, + "learning_rate": 2.5518012999537074e-06, + "loss": 2.4687, + "step": 24047 + }, + { + "epoch": 2.5431472081218276, + "grad_norm": 8.3125, + "learning_rate": 2.5516300803770124e-06, + "loss": 2.6287, + "step": 24048 + }, + { + "epoch": 2.5432529610829104, + "grad_norm": 8.875, + "learning_rate": 2.5514588605580388e-06, + "loss": 2.2037, + "step": 24049 + }, + { + "epoch": 2.5433587140439933, + "grad_norm": 8.25, + "learning_rate": 2.55128764049759e-06, + "loss": 2.6311, + "step": 24050 + }, + { + "epoch": 2.543464467005076, + "grad_norm": 8.125, + "learning_rate": 2.5511164201964687e-06, + "loss": 2.3483, + "step": 24051 + }, + { + "epoch": 2.543570219966159, + "grad_norm": 8.3125, + "learning_rate": 2.55094519965548e-06, + "loss": 2.5383, + "step": 24052 + }, + { + "epoch": 2.5436759729272422, + "grad_norm": 8.625, + "learning_rate": 2.550773978875426e-06, + "loss": 2.6476, + "step": 24053 + }, + { + "epoch": 2.5437817258883246, + "grad_norm": 9.25, + "learning_rate": 2.5506027578571096e-06, + "loss": 2.4583, + "step": 24054 + }, + { + "epoch": 2.543887478849408, + "grad_norm": 8.375, + "learning_rate": 2.5504315366013364e-06, + "loss": 2.4048, + "step": 24055 + }, + { + "epoch": 2.5439932318104908, + "grad_norm": 7.9375, + "learning_rate": 2.5502603151089082e-06, + "loss": 2.4573, + "step": 24056 + }, + { + "epoch": 2.5440989847715736, + "grad_norm": 8.625, + "learning_rate": 2.5500890933806287e-06, + "loss": 2.5161, + "step": 24057 + }, + { + "epoch": 2.5442047377326564, + "grad_norm": 8.375, + "learning_rate": 2.549917871417301e-06, + "loss": 2.4066, + "step": 24058 + }, + { + "epoch": 2.5443104906937393, + "grad_norm": 8.0625, + "learning_rate": 2.5497466492197306e-06, + "loss": 2.4654, + "step": 24059 + }, + { + "epoch": 2.5444162436548226, + "grad_norm": 8.25, + "learning_rate": 2.5495754267887187e-06, + "loss": 2.4283, + "step": 24060 + }, + { + "epoch": 2.5445219966159054, + "grad_norm": 8.125, + "learning_rate": 2.5494042041250688e-06, + "loss": 2.5397, + "step": 24061 + }, + { + "epoch": 2.5446277495769882, + "grad_norm": 8.1875, + "learning_rate": 2.5492329812295864e-06, + "loss": 2.5711, + "step": 24062 + }, + { + "epoch": 2.544733502538071, + "grad_norm": 8.25, + "learning_rate": 2.5490617581030736e-06, + "loss": 2.5666, + "step": 24063 + }, + { + "epoch": 2.544839255499154, + "grad_norm": 8.0, + "learning_rate": 2.5488905347463335e-06, + "loss": 2.5255, + "step": 24064 + }, + { + "epoch": 2.5449450084602367, + "grad_norm": 8.3125, + "learning_rate": 2.5487193111601707e-06, + "loss": 2.5094, + "step": 24065 + }, + { + "epoch": 2.5450507614213196, + "grad_norm": 7.78125, + "learning_rate": 2.5485480873453877e-06, + "loss": 2.4821, + "step": 24066 + }, + { + "epoch": 2.545156514382403, + "grad_norm": 7.75, + "learning_rate": 2.5483768633027877e-06, + "loss": 2.6081, + "step": 24067 + }, + { + "epoch": 2.5452622673434857, + "grad_norm": 8.1875, + "learning_rate": 2.548205639033176e-06, + "loss": 2.525, + "step": 24068 + }, + { + "epoch": 2.5453680203045685, + "grad_norm": 8.625, + "learning_rate": 2.548034414537355e-06, + "loss": 2.5997, + "step": 24069 + }, + { + "epoch": 2.5454737732656514, + "grad_norm": 9.6875, + "learning_rate": 2.5478631898161272e-06, + "loss": 2.6874, + "step": 24070 + }, + { + "epoch": 2.5455795262267342, + "grad_norm": 8.75, + "learning_rate": 2.547691964870298e-06, + "loss": 2.6126, + "step": 24071 + }, + { + "epoch": 2.5456852791878175, + "grad_norm": 8.3125, + "learning_rate": 2.5475207397006696e-06, + "loss": 2.6691, + "step": 24072 + }, + { + "epoch": 2.5457910321489, + "grad_norm": 8.1875, + "learning_rate": 2.547349514308045e-06, + "loss": 2.5918, + "step": 24073 + }, + { + "epoch": 2.545896785109983, + "grad_norm": 8.5625, + "learning_rate": 2.5471782886932295e-06, + "loss": 2.5797, + "step": 24074 + }, + { + "epoch": 2.546002538071066, + "grad_norm": 8.625, + "learning_rate": 2.5470070628570253e-06, + "loss": 2.6283, + "step": 24075 + }, + { + "epoch": 2.546108291032149, + "grad_norm": 8.5, + "learning_rate": 2.546835836800236e-06, + "loss": 2.683, + "step": 24076 + }, + { + "epoch": 2.5462140439932317, + "grad_norm": 8.5625, + "learning_rate": 2.5466646105236653e-06, + "loss": 2.2881, + "step": 24077 + }, + { + "epoch": 2.5463197969543145, + "grad_norm": 8.4375, + "learning_rate": 2.546493384028117e-06, + "loss": 2.8156, + "step": 24078 + }, + { + "epoch": 2.546425549915398, + "grad_norm": 8.0625, + "learning_rate": 2.5463221573143944e-06, + "loss": 2.5122, + "step": 24079 + }, + { + "epoch": 2.5465313028764807, + "grad_norm": 7.6875, + "learning_rate": 2.5461509303832997e-06, + "loss": 2.4019, + "step": 24080 + }, + { + "epoch": 2.5466370558375635, + "grad_norm": 8.3125, + "learning_rate": 2.5459797032356387e-06, + "loss": 2.526, + "step": 24081 + }, + { + "epoch": 2.5467428087986463, + "grad_norm": 8.1875, + "learning_rate": 2.5458084758722133e-06, + "loss": 2.5283, + "step": 24082 + }, + { + "epoch": 2.546848561759729, + "grad_norm": 7.96875, + "learning_rate": 2.5456372482938275e-06, + "loss": 2.3047, + "step": 24083 + }, + { + "epoch": 2.546954314720812, + "grad_norm": 8.0625, + "learning_rate": 2.5454660205012855e-06, + "loss": 2.3445, + "step": 24084 + }, + { + "epoch": 2.547060067681895, + "grad_norm": 8.375, + "learning_rate": 2.5452947924953887e-06, + "loss": 2.477, + "step": 24085 + }, + { + "epoch": 2.547165820642978, + "grad_norm": 8.125, + "learning_rate": 2.5451235642769424e-06, + "loss": 2.568, + "step": 24086 + }, + { + "epoch": 2.547271573604061, + "grad_norm": 8.5625, + "learning_rate": 2.5449523358467502e-06, + "loss": 2.5744, + "step": 24087 + }, + { + "epoch": 2.547377326565144, + "grad_norm": 8.6875, + "learning_rate": 2.544781107205615e-06, + "loss": 2.5242, + "step": 24088 + }, + { + "epoch": 2.5474830795262267, + "grad_norm": 8.625, + "learning_rate": 2.54460987835434e-06, + "loss": 2.8018, + "step": 24089 + }, + { + "epoch": 2.5475888324873095, + "grad_norm": 8.375, + "learning_rate": 2.5444386492937295e-06, + "loss": 2.4387, + "step": 24090 + }, + { + "epoch": 2.547694585448393, + "grad_norm": 8.9375, + "learning_rate": 2.5442674200245863e-06, + "loss": 2.6134, + "step": 24091 + }, + { + "epoch": 2.5478003384094756, + "grad_norm": 8.4375, + "learning_rate": 2.5440961905477142e-06, + "loss": 2.4877, + "step": 24092 + }, + { + "epoch": 2.5479060913705585, + "grad_norm": 7.78125, + "learning_rate": 2.5439249608639162e-06, + "loss": 2.4393, + "step": 24093 + }, + { + "epoch": 2.5480118443316413, + "grad_norm": 8.3125, + "learning_rate": 2.5437537309739974e-06, + "loss": 2.6163, + "step": 24094 + }, + { + "epoch": 2.548117597292724, + "grad_norm": 8.375, + "learning_rate": 2.5435825008787596e-06, + "loss": 2.6438, + "step": 24095 + }, + { + "epoch": 2.548223350253807, + "grad_norm": 8.25, + "learning_rate": 2.5434112705790072e-06, + "loss": 2.6025, + "step": 24096 + }, + { + "epoch": 2.54832910321489, + "grad_norm": 8.0, + "learning_rate": 2.543240040075543e-06, + "loss": 2.3314, + "step": 24097 + }, + { + "epoch": 2.548434856175973, + "grad_norm": 8.6875, + "learning_rate": 2.5430688093691718e-06, + "loss": 2.4082, + "step": 24098 + }, + { + "epoch": 2.548540609137056, + "grad_norm": 8.125, + "learning_rate": 2.542897578460696e-06, + "loss": 2.4496, + "step": 24099 + }, + { + "epoch": 2.5486463620981388, + "grad_norm": 8.3125, + "learning_rate": 2.5427263473509188e-06, + "loss": 2.6496, + "step": 24100 + }, + { + "epoch": 2.5487521150592216, + "grad_norm": 9.0, + "learning_rate": 2.542555116040645e-06, + "loss": 2.6212, + "step": 24101 + }, + { + "epoch": 2.5488578680203045, + "grad_norm": 7.90625, + "learning_rate": 2.5423838845306763e-06, + "loss": 2.5174, + "step": 24102 + }, + { + "epoch": 2.5489636209813877, + "grad_norm": 8.375, + "learning_rate": 2.542212652821819e-06, + "loss": 2.5768, + "step": 24103 + }, + { + "epoch": 2.54906937394247, + "grad_norm": 8.3125, + "learning_rate": 2.5420414209148742e-06, + "loss": 2.4363, + "step": 24104 + }, + { + "epoch": 2.5491751269035534, + "grad_norm": 8.25, + "learning_rate": 2.5418701888106457e-06, + "loss": 2.6286, + "step": 24105 + }, + { + "epoch": 2.5492808798646363, + "grad_norm": 8.4375, + "learning_rate": 2.541698956509938e-06, + "loss": 2.5265, + "step": 24106 + }, + { + "epoch": 2.549386632825719, + "grad_norm": 8.0625, + "learning_rate": 2.541527724013555e-06, + "loss": 2.4273, + "step": 24107 + }, + { + "epoch": 2.549492385786802, + "grad_norm": 8.8125, + "learning_rate": 2.5413564913222983e-06, + "loss": 2.6545, + "step": 24108 + }, + { + "epoch": 2.5495981387478848, + "grad_norm": 8.75, + "learning_rate": 2.541185258436973e-06, + "loss": 2.6193, + "step": 24109 + }, + { + "epoch": 2.549703891708968, + "grad_norm": 8.125, + "learning_rate": 2.541014025358382e-06, + "loss": 2.4035, + "step": 24110 + }, + { + "epoch": 2.549809644670051, + "grad_norm": 8.625, + "learning_rate": 2.5408427920873284e-06, + "loss": 2.5039, + "step": 24111 + }, + { + "epoch": 2.5499153976311337, + "grad_norm": 8.5, + "learning_rate": 2.5406715586246165e-06, + "loss": 2.5225, + "step": 24112 + }, + { + "epoch": 2.5500211505922166, + "grad_norm": 8.125, + "learning_rate": 2.5405003249710504e-06, + "loss": 2.4625, + "step": 24113 + }, + { + "epoch": 2.5501269035532994, + "grad_norm": 7.875, + "learning_rate": 2.540329091127432e-06, + "loss": 2.49, + "step": 24114 + }, + { + "epoch": 2.5502326565143822, + "grad_norm": 8.0625, + "learning_rate": 2.5401578570945655e-06, + "loss": 2.4007, + "step": 24115 + }, + { + "epoch": 2.550338409475465, + "grad_norm": 8.625, + "learning_rate": 2.5399866228732556e-06, + "loss": 2.2544, + "step": 24116 + }, + { + "epoch": 2.5504441624365484, + "grad_norm": 8.5, + "learning_rate": 2.5398153884643046e-06, + "loss": 2.5984, + "step": 24117 + }, + { + "epoch": 2.550549915397631, + "grad_norm": 8.6875, + "learning_rate": 2.5396441538685156e-06, + "loss": 2.7381, + "step": 24118 + }, + { + "epoch": 2.550655668358714, + "grad_norm": 8.625, + "learning_rate": 2.5394729190866935e-06, + "loss": 2.5746, + "step": 24119 + }, + { + "epoch": 2.550761421319797, + "grad_norm": 8.25, + "learning_rate": 2.5393016841196406e-06, + "loss": 2.6162, + "step": 24120 + }, + { + "epoch": 2.5508671742808797, + "grad_norm": 8.625, + "learning_rate": 2.5391304489681607e-06, + "loss": 2.3844, + "step": 24121 + }, + { + "epoch": 2.550972927241963, + "grad_norm": 8.125, + "learning_rate": 2.5389592136330584e-06, + "loss": 2.5018, + "step": 24122 + }, + { + "epoch": 2.551078680203046, + "grad_norm": 8.5625, + "learning_rate": 2.5387879781151363e-06, + "loss": 2.6416, + "step": 24123 + }, + { + "epoch": 2.5511844331641287, + "grad_norm": 8.625, + "learning_rate": 2.5386167424151974e-06, + "loss": 2.4323, + "step": 24124 + }, + { + "epoch": 2.5512901861252115, + "grad_norm": 8.6875, + "learning_rate": 2.538445506534047e-06, + "loss": 2.4837, + "step": 24125 + }, + { + "epoch": 2.5513959390862944, + "grad_norm": 8.25, + "learning_rate": 2.5382742704724872e-06, + "loss": 2.502, + "step": 24126 + }, + { + "epoch": 2.551501692047377, + "grad_norm": 8.625, + "learning_rate": 2.5381030342313214e-06, + "loss": 2.4819, + "step": 24127 + }, + { + "epoch": 2.55160744500846, + "grad_norm": 8.5625, + "learning_rate": 2.5379317978113537e-06, + "loss": 2.6476, + "step": 24128 + }, + { + "epoch": 2.5517131979695433, + "grad_norm": 8.1875, + "learning_rate": 2.5377605612133876e-06, + "loss": 2.562, + "step": 24129 + }, + { + "epoch": 2.551818950930626, + "grad_norm": 8.1875, + "learning_rate": 2.537589324438227e-06, + "loss": 2.5911, + "step": 24130 + }, + { + "epoch": 2.551924703891709, + "grad_norm": 8.0625, + "learning_rate": 2.5374180874866748e-06, + "loss": 2.2614, + "step": 24131 + }, + { + "epoch": 2.552030456852792, + "grad_norm": 7.875, + "learning_rate": 2.5372468503595354e-06, + "loss": 2.2885, + "step": 24132 + }, + { + "epoch": 2.5521362098138747, + "grad_norm": 8.1875, + "learning_rate": 2.537075613057611e-06, + "loss": 2.5217, + "step": 24133 + }, + { + "epoch": 2.552241962774958, + "grad_norm": 8.5, + "learning_rate": 2.5369043755817063e-06, + "loss": 2.6083, + "step": 24134 + }, + { + "epoch": 2.5523477157360404, + "grad_norm": 8.1875, + "learning_rate": 2.5367331379326244e-06, + "loss": 2.6611, + "step": 24135 + }, + { + "epoch": 2.5524534686971236, + "grad_norm": 8.75, + "learning_rate": 2.536561900111169e-06, + "loss": 2.3365, + "step": 24136 + }, + { + "epoch": 2.5525592216582065, + "grad_norm": 9.125, + "learning_rate": 2.5363906621181435e-06, + "loss": 2.7458, + "step": 24137 + }, + { + "epoch": 2.5526649746192893, + "grad_norm": 8.5625, + "learning_rate": 2.536219423954351e-06, + "loss": 2.2144, + "step": 24138 + }, + { + "epoch": 2.552770727580372, + "grad_norm": 7.84375, + "learning_rate": 2.5360481856205966e-06, + "loss": 2.4631, + "step": 24139 + }, + { + "epoch": 2.552876480541455, + "grad_norm": 8.125, + "learning_rate": 2.5358769471176813e-06, + "loss": 2.4618, + "step": 24140 + }, + { + "epoch": 2.5529822335025383, + "grad_norm": 8.4375, + "learning_rate": 2.5357057084464114e-06, + "loss": 2.4262, + "step": 24141 + }, + { + "epoch": 2.553087986463621, + "grad_norm": 8.3125, + "learning_rate": 2.5355344696075896e-06, + "loss": 2.2885, + "step": 24142 + }, + { + "epoch": 2.553193739424704, + "grad_norm": 8.4375, + "learning_rate": 2.535363230602018e-06, + "loss": 2.4851, + "step": 24143 + }, + { + "epoch": 2.553299492385787, + "grad_norm": 8.8125, + "learning_rate": 2.5351919914305017e-06, + "loss": 2.4096, + "step": 24144 + }, + { + "epoch": 2.5534052453468696, + "grad_norm": 8.6875, + "learning_rate": 2.5350207520938434e-06, + "loss": 2.3189, + "step": 24145 + }, + { + "epoch": 2.5535109983079525, + "grad_norm": 8.125, + "learning_rate": 2.5348495125928476e-06, + "loss": 2.2654, + "step": 24146 + }, + { + "epoch": 2.5536167512690353, + "grad_norm": 8.125, + "learning_rate": 2.5346782729283164e-06, + "loss": 2.5793, + "step": 24147 + }, + { + "epoch": 2.5537225042301186, + "grad_norm": 8.75, + "learning_rate": 2.5345070331010553e-06, + "loss": 2.6103, + "step": 24148 + }, + { + "epoch": 2.5538282571912014, + "grad_norm": 8.125, + "learning_rate": 2.534335793111866e-06, + "loss": 2.4754, + "step": 24149 + }, + { + "epoch": 2.5539340101522843, + "grad_norm": 8.3125, + "learning_rate": 2.5341645529615535e-06, + "loss": 2.4187, + "step": 24150 + }, + { + "epoch": 2.554039763113367, + "grad_norm": 8.1875, + "learning_rate": 2.5339933126509204e-06, + "loss": 2.4104, + "step": 24151 + }, + { + "epoch": 2.55414551607445, + "grad_norm": 8.4375, + "learning_rate": 2.5338220721807706e-06, + "loss": 2.5456, + "step": 24152 + }, + { + "epoch": 2.5542512690355332, + "grad_norm": 8.3125, + "learning_rate": 2.5336508315519076e-06, + "loss": 2.6211, + "step": 24153 + }, + { + "epoch": 2.5543570219966156, + "grad_norm": 8.6875, + "learning_rate": 2.533479590765135e-06, + "loss": 2.5581, + "step": 24154 + }, + { + "epoch": 2.554462774957699, + "grad_norm": 8.8125, + "learning_rate": 2.533308349821256e-06, + "loss": 2.7814, + "step": 24155 + }, + { + "epoch": 2.5545685279187818, + "grad_norm": 8.3125, + "learning_rate": 2.5331371087210748e-06, + "loss": 2.6874, + "step": 24156 + }, + { + "epoch": 2.5546742808798646, + "grad_norm": 7.875, + "learning_rate": 2.532965867465395e-06, + "loss": 2.4937, + "step": 24157 + }, + { + "epoch": 2.5547800338409474, + "grad_norm": 8.625, + "learning_rate": 2.53279462605502e-06, + "loss": 2.4658, + "step": 24158 + }, + { + "epoch": 2.5548857868020303, + "grad_norm": 8.8125, + "learning_rate": 2.532623384490753e-06, + "loss": 2.8454, + "step": 24159 + }, + { + "epoch": 2.5549915397631136, + "grad_norm": 8.75, + "learning_rate": 2.532452142773398e-06, + "loss": 2.6137, + "step": 24160 + }, + { + "epoch": 2.5550972927241964, + "grad_norm": 8.4375, + "learning_rate": 2.532280900903758e-06, + "loss": 2.3287, + "step": 24161 + }, + { + "epoch": 2.5552030456852792, + "grad_norm": 8.9375, + "learning_rate": 2.532109658882637e-06, + "loss": 2.6762, + "step": 24162 + }, + { + "epoch": 2.555308798646362, + "grad_norm": 8.625, + "learning_rate": 2.5319384167108392e-06, + "loss": 2.7022, + "step": 24163 + }, + { + "epoch": 2.555414551607445, + "grad_norm": 8.5, + "learning_rate": 2.5317671743891657e-06, + "loss": 2.5184, + "step": 24164 + }, + { + "epoch": 2.555520304568528, + "grad_norm": 8.5625, + "learning_rate": 2.5315959319184236e-06, + "loss": 2.5348, + "step": 24165 + }, + { + "epoch": 2.5556260575296106, + "grad_norm": 8.5625, + "learning_rate": 2.5314246892994136e-06, + "loss": 2.5101, + "step": 24166 + }, + { + "epoch": 2.555731810490694, + "grad_norm": 8.125, + "learning_rate": 2.5312534465329414e-06, + "loss": 2.4343, + "step": 24167 + }, + { + "epoch": 2.5558375634517767, + "grad_norm": 8.0625, + "learning_rate": 2.531082203619809e-06, + "loss": 2.5165, + "step": 24168 + }, + { + "epoch": 2.5559433164128595, + "grad_norm": 9.4375, + "learning_rate": 2.5309109605608207e-06, + "loss": 2.4538, + "step": 24169 + }, + { + "epoch": 2.5560490693739424, + "grad_norm": 8.3125, + "learning_rate": 2.5307397173567797e-06, + "loss": 2.6683, + "step": 24170 + }, + { + "epoch": 2.5561548223350252, + "grad_norm": 8.1875, + "learning_rate": 2.53056847400849e-06, + "loss": 2.5505, + "step": 24171 + }, + { + "epoch": 2.5562605752961085, + "grad_norm": 12.25, + "learning_rate": 2.5303972305167547e-06, + "loss": 2.6958, + "step": 24172 + }, + { + "epoch": 2.5563663282571913, + "grad_norm": 8.375, + "learning_rate": 2.5302259868823775e-06, + "loss": 2.4279, + "step": 24173 + }, + { + "epoch": 2.556472081218274, + "grad_norm": 8.125, + "learning_rate": 2.5300547431061624e-06, + "loss": 2.4952, + "step": 24174 + }, + { + "epoch": 2.556577834179357, + "grad_norm": 8.75, + "learning_rate": 2.5298834991889126e-06, + "loss": 2.4547, + "step": 24175 + }, + { + "epoch": 2.55668358714044, + "grad_norm": 9.3125, + "learning_rate": 2.5297122551314318e-06, + "loss": 2.6818, + "step": 24176 + }, + { + "epoch": 2.5567893401015227, + "grad_norm": 8.4375, + "learning_rate": 2.529541010934524e-06, + "loss": 2.4932, + "step": 24177 + }, + { + "epoch": 2.5568950930626055, + "grad_norm": 7.8125, + "learning_rate": 2.5293697665989913e-06, + "loss": 2.3579, + "step": 24178 + }, + { + "epoch": 2.557000846023689, + "grad_norm": 8.5, + "learning_rate": 2.5291985221256393e-06, + "loss": 2.4903, + "step": 24179 + }, + { + "epoch": 2.5571065989847717, + "grad_norm": 8.25, + "learning_rate": 2.52902727751527e-06, + "loss": 2.5897, + "step": 24180 + }, + { + "epoch": 2.5572123519458545, + "grad_norm": 8.1875, + "learning_rate": 2.5288560327686877e-06, + "loss": 2.4432, + "step": 24181 + }, + { + "epoch": 2.5573181049069373, + "grad_norm": 8.5625, + "learning_rate": 2.5286847878866953e-06, + "loss": 2.5196, + "step": 24182 + }, + { + "epoch": 2.55742385786802, + "grad_norm": 8.625, + "learning_rate": 2.528513542870098e-06, + "loss": 2.4803, + "step": 24183 + }, + { + "epoch": 2.5575296108291035, + "grad_norm": 8.25, + "learning_rate": 2.5283422977196977e-06, + "loss": 2.4995, + "step": 24184 + }, + { + "epoch": 2.557635363790186, + "grad_norm": 8.8125, + "learning_rate": 2.5281710524362984e-06, + "loss": 2.6278, + "step": 24185 + }, + { + "epoch": 2.557741116751269, + "grad_norm": 8.6875, + "learning_rate": 2.5279998070207045e-06, + "loss": 2.6096, + "step": 24186 + }, + { + "epoch": 2.557846869712352, + "grad_norm": 8.5625, + "learning_rate": 2.5278285614737188e-06, + "loss": 2.5093, + "step": 24187 + }, + { + "epoch": 2.557952622673435, + "grad_norm": 8.125, + "learning_rate": 2.5276573157961443e-06, + "loss": 2.5953, + "step": 24188 + }, + { + "epoch": 2.5580583756345177, + "grad_norm": 8.125, + "learning_rate": 2.527486069988786e-06, + "loss": 2.4849, + "step": 24189 + }, + { + "epoch": 2.5581641285956005, + "grad_norm": 8.4375, + "learning_rate": 2.527314824052447e-06, + "loss": 2.5535, + "step": 24190 + }, + { + "epoch": 2.558269881556684, + "grad_norm": 8.125, + "learning_rate": 2.5271435779879296e-06, + "loss": 2.5606, + "step": 24191 + }, + { + "epoch": 2.5583756345177666, + "grad_norm": 8.25, + "learning_rate": 2.5269723317960392e-06, + "loss": 2.5046, + "step": 24192 + }, + { + "epoch": 2.5584813874788495, + "grad_norm": 8.375, + "learning_rate": 2.5268010854775793e-06, + "loss": 2.689, + "step": 24193 + }, + { + "epoch": 2.5585871404399323, + "grad_norm": 8.75, + "learning_rate": 2.5266298390333515e-06, + "loss": 2.4854, + "step": 24194 + }, + { + "epoch": 2.558692893401015, + "grad_norm": 9.125, + "learning_rate": 2.526458592464162e-06, + "loss": 2.5438, + "step": 24195 + }, + { + "epoch": 2.558798646362098, + "grad_norm": 10.125, + "learning_rate": 2.526287345770813e-06, + "loss": 2.4752, + "step": 24196 + }, + { + "epoch": 2.558904399323181, + "grad_norm": 9.0, + "learning_rate": 2.5261160989541077e-06, + "loss": 2.4592, + "step": 24197 + }, + { + "epoch": 2.559010152284264, + "grad_norm": 8.0625, + "learning_rate": 2.5259448520148505e-06, + "loss": 2.3903, + "step": 24198 + }, + { + "epoch": 2.559115905245347, + "grad_norm": 8.125, + "learning_rate": 2.525773604953845e-06, + "loss": 2.4457, + "step": 24199 + }, + { + "epoch": 2.5592216582064298, + "grad_norm": 8.875, + "learning_rate": 2.525602357771894e-06, + "loss": 2.6855, + "step": 24200 + }, + { + "epoch": 2.5593274111675126, + "grad_norm": 8.5625, + "learning_rate": 2.5254311104698013e-06, + "loss": 2.4942, + "step": 24201 + }, + { + "epoch": 2.5594331641285955, + "grad_norm": 8.25, + "learning_rate": 2.5252598630483715e-06, + "loss": 2.5261, + "step": 24202 + }, + { + "epoch": 2.5595389170896787, + "grad_norm": 8.5, + "learning_rate": 2.5250886155084075e-06, + "loss": 2.5752, + "step": 24203 + }, + { + "epoch": 2.5596446700507616, + "grad_norm": 8.875, + "learning_rate": 2.5249173678507126e-06, + "loss": 2.8695, + "step": 24204 + }, + { + "epoch": 2.5597504230118444, + "grad_norm": 8.625, + "learning_rate": 2.524746120076091e-06, + "loss": 2.7499, + "step": 24205 + }, + { + "epoch": 2.5598561759729273, + "grad_norm": 8.375, + "learning_rate": 2.524574872185346e-06, + "loss": 2.5774, + "step": 24206 + }, + { + "epoch": 2.55996192893401, + "grad_norm": 8.4375, + "learning_rate": 2.524403624179281e-06, + "loss": 2.5455, + "step": 24207 + }, + { + "epoch": 2.560067681895093, + "grad_norm": 8.0625, + "learning_rate": 2.524232376058699e-06, + "loss": 2.5957, + "step": 24208 + }, + { + "epoch": 2.5601734348561758, + "grad_norm": 8.3125, + "learning_rate": 2.5240611278244055e-06, + "loss": 2.3848, + "step": 24209 + }, + { + "epoch": 2.560279187817259, + "grad_norm": 8.125, + "learning_rate": 2.523889879477202e-06, + "loss": 2.4857, + "step": 24210 + }, + { + "epoch": 2.560384940778342, + "grad_norm": 8.375, + "learning_rate": 2.5237186310178936e-06, + "loss": 2.4432, + "step": 24211 + }, + { + "epoch": 2.5604906937394247, + "grad_norm": 8.75, + "learning_rate": 2.523547382447284e-06, + "loss": 2.4695, + "step": 24212 + }, + { + "epoch": 2.5605964467005076, + "grad_norm": 8.3125, + "learning_rate": 2.523376133766175e-06, + "loss": 2.534, + "step": 24213 + }, + { + "epoch": 2.5607021996615904, + "grad_norm": 8.375, + "learning_rate": 2.523204884975372e-06, + "loss": 2.722, + "step": 24214 + }, + { + "epoch": 2.5608079526226737, + "grad_norm": 8.625, + "learning_rate": 2.523033636075678e-06, + "loss": 2.3961, + "step": 24215 + }, + { + "epoch": 2.560913705583756, + "grad_norm": 8.5625, + "learning_rate": 2.5228623870678965e-06, + "loss": 2.3582, + "step": 24216 + }, + { + "epoch": 2.5610194585448394, + "grad_norm": 8.0625, + "learning_rate": 2.5226911379528307e-06, + "loss": 2.6167, + "step": 24217 + }, + { + "epoch": 2.561125211505922, + "grad_norm": 8.625, + "learning_rate": 2.522519888731285e-06, + "loss": 2.4798, + "step": 24218 + }, + { + "epoch": 2.561230964467005, + "grad_norm": 8.0, + "learning_rate": 2.522348639404063e-06, + "loss": 2.5588, + "step": 24219 + }, + { + "epoch": 2.561336717428088, + "grad_norm": 8.5, + "learning_rate": 2.5221773899719676e-06, + "loss": 2.3001, + "step": 24220 + }, + { + "epoch": 2.5614424703891707, + "grad_norm": 8.4375, + "learning_rate": 2.522006140435803e-06, + "loss": 2.6119, + "step": 24221 + }, + { + "epoch": 2.561548223350254, + "grad_norm": 8.1875, + "learning_rate": 2.521834890796373e-06, + "loss": 2.3908, + "step": 24222 + }, + { + "epoch": 2.561653976311337, + "grad_norm": 8.875, + "learning_rate": 2.52166364105448e-06, + "loss": 2.7077, + "step": 24223 + }, + { + "epoch": 2.5617597292724197, + "grad_norm": 8.5625, + "learning_rate": 2.5214923912109285e-06, + "loss": 2.2106, + "step": 24224 + }, + { + "epoch": 2.5618654822335025, + "grad_norm": 8.5625, + "learning_rate": 2.521321141266522e-06, + "loss": 2.2996, + "step": 24225 + }, + { + "epoch": 2.5619712351945854, + "grad_norm": 8.4375, + "learning_rate": 2.521149891222064e-06, + "loss": 2.4937, + "step": 24226 + }, + { + "epoch": 2.562076988155668, + "grad_norm": 8.875, + "learning_rate": 2.5209786410783586e-06, + "loss": 2.6015, + "step": 24227 + }, + { + "epoch": 2.562182741116751, + "grad_norm": 7.9375, + "learning_rate": 2.520807390836209e-06, + "loss": 2.6053, + "step": 24228 + }, + { + "epoch": 2.5622884940778343, + "grad_norm": 8.25, + "learning_rate": 2.5206361404964188e-06, + "loss": 2.5123, + "step": 24229 + }, + { + "epoch": 2.562394247038917, + "grad_norm": 8.3125, + "learning_rate": 2.520464890059792e-06, + "loss": 2.4381, + "step": 24230 + }, + { + "epoch": 2.5625, + "grad_norm": 8.8125, + "learning_rate": 2.520293639527131e-06, + "loss": 2.5875, + "step": 24231 + }, + { + "epoch": 2.562605752961083, + "grad_norm": 8.5, + "learning_rate": 2.5201223888992404e-06, + "loss": 2.6876, + "step": 24232 + }, + { + "epoch": 2.5627115059221657, + "grad_norm": 8.125, + "learning_rate": 2.5199511381769244e-06, + "loss": 2.5288, + "step": 24233 + }, + { + "epoch": 2.562817258883249, + "grad_norm": 8.1875, + "learning_rate": 2.519779887360985e-06, + "loss": 2.5739, + "step": 24234 + }, + { + "epoch": 2.562923011844332, + "grad_norm": 8.75, + "learning_rate": 2.519608636452227e-06, + "loss": 2.586, + "step": 24235 + }, + { + "epoch": 2.5630287648054146, + "grad_norm": 8.6875, + "learning_rate": 2.5194373854514537e-06, + "loss": 2.5497, + "step": 24236 + }, + { + "epoch": 2.5631345177664975, + "grad_norm": 8.5, + "learning_rate": 2.519266134359469e-06, + "loss": 2.5145, + "step": 24237 + }, + { + "epoch": 2.5632402707275803, + "grad_norm": 8.5, + "learning_rate": 2.5190948831770755e-06, + "loss": 2.6606, + "step": 24238 + }, + { + "epoch": 2.563346023688663, + "grad_norm": 8.3125, + "learning_rate": 2.518923631905078e-06, + "loss": 2.5691, + "step": 24239 + }, + { + "epoch": 2.563451776649746, + "grad_norm": 8.375, + "learning_rate": 2.5187523805442796e-06, + "loss": 2.6402, + "step": 24240 + }, + { + "epoch": 2.5635575296108293, + "grad_norm": 8.75, + "learning_rate": 2.5185811290954843e-06, + "loss": 2.511, + "step": 24241 + }, + { + "epoch": 2.563663282571912, + "grad_norm": 7.90625, + "learning_rate": 2.5184098775594945e-06, + "loss": 2.3892, + "step": 24242 + }, + { + "epoch": 2.563769035532995, + "grad_norm": 8.875, + "learning_rate": 2.5182386259371154e-06, + "loss": 2.3553, + "step": 24243 + }, + { + "epoch": 2.563874788494078, + "grad_norm": 8.3125, + "learning_rate": 2.5180673742291496e-06, + "loss": 2.4674, + "step": 24244 + }, + { + "epoch": 2.5639805414551606, + "grad_norm": 8.125, + "learning_rate": 2.517896122436401e-06, + "loss": 2.4339, + "step": 24245 + }, + { + "epoch": 2.564086294416244, + "grad_norm": 7.71875, + "learning_rate": 2.5177248705596735e-06, + "loss": 2.4568, + "step": 24246 + }, + { + "epoch": 2.5641920473773263, + "grad_norm": 8.0625, + "learning_rate": 2.51755361859977e-06, + "loss": 2.4395, + "step": 24247 + }, + { + "epoch": 2.5642978003384096, + "grad_norm": 8.625, + "learning_rate": 2.5173823665574947e-06, + "loss": 2.6231, + "step": 24248 + }, + { + "epoch": 2.5644035532994924, + "grad_norm": 8.375, + "learning_rate": 2.5172111144336513e-06, + "loss": 2.5231, + "step": 24249 + }, + { + "epoch": 2.5645093062605753, + "grad_norm": 8.5, + "learning_rate": 2.517039862229043e-06, + "loss": 2.3006, + "step": 24250 + }, + { + "epoch": 2.564615059221658, + "grad_norm": 8.375, + "learning_rate": 2.5168686099444734e-06, + "loss": 2.6487, + "step": 24251 + }, + { + "epoch": 2.564720812182741, + "grad_norm": 8.875, + "learning_rate": 2.516697357580747e-06, + "loss": 2.3402, + "step": 24252 + }, + { + "epoch": 2.5648265651438242, + "grad_norm": 8.5625, + "learning_rate": 2.5165261051386658e-06, + "loss": 2.4522, + "step": 24253 + }, + { + "epoch": 2.564932318104907, + "grad_norm": 8.3125, + "learning_rate": 2.5163548526190353e-06, + "loss": 2.5824, + "step": 24254 + }, + { + "epoch": 2.56503807106599, + "grad_norm": 8.125, + "learning_rate": 2.516183600022657e-06, + "loss": 2.488, + "step": 24255 + }, + { + "epoch": 2.5651438240270727, + "grad_norm": 8.8125, + "learning_rate": 2.5160123473503368e-06, + "loss": 2.5762, + "step": 24256 + }, + { + "epoch": 2.5652495769881556, + "grad_norm": 8.5625, + "learning_rate": 2.5158410946028764e-06, + "loss": 2.7048, + "step": 24257 + }, + { + "epoch": 2.5653553299492384, + "grad_norm": 8.0625, + "learning_rate": 2.5156698417810805e-06, + "loss": 2.4031, + "step": 24258 + }, + { + "epoch": 2.5654610829103213, + "grad_norm": 8.625, + "learning_rate": 2.515498588885753e-06, + "loss": 2.2992, + "step": 24259 + }, + { + "epoch": 2.5655668358714045, + "grad_norm": 8.1875, + "learning_rate": 2.5153273359176967e-06, + "loss": 2.5006, + "step": 24260 + }, + { + "epoch": 2.5656725888324874, + "grad_norm": 8.8125, + "learning_rate": 2.515156082877715e-06, + "loss": 2.3271, + "step": 24261 + }, + { + "epoch": 2.5657783417935702, + "grad_norm": 8.0625, + "learning_rate": 2.514984829766612e-06, + "loss": 2.4824, + "step": 24262 + }, + { + "epoch": 2.565884094754653, + "grad_norm": 8.6875, + "learning_rate": 2.5148135765851915e-06, + "loss": 2.3981, + "step": 24263 + }, + { + "epoch": 2.565989847715736, + "grad_norm": 8.75, + "learning_rate": 2.5146423233342565e-06, + "loss": 2.2843, + "step": 24264 + }, + { + "epoch": 2.566095600676819, + "grad_norm": 8.4375, + "learning_rate": 2.514471070014612e-06, + "loss": 2.3055, + "step": 24265 + }, + { + "epoch": 2.566201353637902, + "grad_norm": 9.375, + "learning_rate": 2.5142998166270606e-06, + "loss": 2.5534, + "step": 24266 + }, + { + "epoch": 2.566307106598985, + "grad_norm": 9.5625, + "learning_rate": 2.5141285631724057e-06, + "loss": 2.681, + "step": 24267 + }, + { + "epoch": 2.5664128595600677, + "grad_norm": 8.5, + "learning_rate": 2.513957309651451e-06, + "loss": 2.6495, + "step": 24268 + }, + { + "epoch": 2.5665186125211505, + "grad_norm": 8.75, + "learning_rate": 2.5137860560650007e-06, + "loss": 2.7324, + "step": 24269 + }, + { + "epoch": 2.5666243654822334, + "grad_norm": 7.96875, + "learning_rate": 2.513614802413858e-06, + "loss": 2.7018, + "step": 24270 + }, + { + "epoch": 2.566730118443316, + "grad_norm": 8.9375, + "learning_rate": 2.5134435486988263e-06, + "loss": 2.6899, + "step": 24271 + }, + { + "epoch": 2.5668358714043995, + "grad_norm": 9.3125, + "learning_rate": 2.5132722949207094e-06, + "loss": 2.562, + "step": 24272 + }, + { + "epoch": 2.5669416243654823, + "grad_norm": 7.96875, + "learning_rate": 2.513101041080312e-06, + "loss": 2.3532, + "step": 24273 + }, + { + "epoch": 2.567047377326565, + "grad_norm": 7.71875, + "learning_rate": 2.512929787178436e-06, + "loss": 2.1066, + "step": 24274 + }, + { + "epoch": 2.567153130287648, + "grad_norm": 8.3125, + "learning_rate": 2.512758533215886e-06, + "loss": 2.5446, + "step": 24275 + }, + { + "epoch": 2.567258883248731, + "grad_norm": 8.4375, + "learning_rate": 2.5125872791934657e-06, + "loss": 2.2514, + "step": 24276 + }, + { + "epoch": 2.567364636209814, + "grad_norm": 8.4375, + "learning_rate": 2.5124160251119776e-06, + "loss": 2.4082, + "step": 24277 + }, + { + "epoch": 2.5674703891708965, + "grad_norm": 8.5, + "learning_rate": 2.512244770972227e-06, + "loss": 2.4547, + "step": 24278 + }, + { + "epoch": 2.56757614213198, + "grad_norm": 9.0, + "learning_rate": 2.5120735167750158e-06, + "loss": 2.5231, + "step": 24279 + }, + { + "epoch": 2.5676818950930627, + "grad_norm": 8.1875, + "learning_rate": 2.5119022625211493e-06, + "loss": 2.6628, + "step": 24280 + }, + { + "epoch": 2.5677876480541455, + "grad_norm": 8.5625, + "learning_rate": 2.5117310082114303e-06, + "loss": 2.4723, + "step": 24281 + }, + { + "epoch": 2.5678934010152283, + "grad_norm": 8.6875, + "learning_rate": 2.511559753846662e-06, + "loss": 2.5171, + "step": 24282 + }, + { + "epoch": 2.567999153976311, + "grad_norm": 8.375, + "learning_rate": 2.5113884994276493e-06, + "loss": 2.5523, + "step": 24283 + }, + { + "epoch": 2.5681049069373945, + "grad_norm": 8.75, + "learning_rate": 2.511217244955195e-06, + "loss": 2.6214, + "step": 24284 + }, + { + "epoch": 2.5682106598984773, + "grad_norm": 8.875, + "learning_rate": 2.511045990430102e-06, + "loss": 2.7199, + "step": 24285 + }, + { + "epoch": 2.56831641285956, + "grad_norm": 8.25, + "learning_rate": 2.5108747358531747e-06, + "loss": 2.4816, + "step": 24286 + }, + { + "epoch": 2.568422165820643, + "grad_norm": 9.125, + "learning_rate": 2.5107034812252178e-06, + "loss": 2.5195, + "step": 24287 + }, + { + "epoch": 2.568527918781726, + "grad_norm": 8.25, + "learning_rate": 2.5105322265470327e-06, + "loss": 2.5472, + "step": 24288 + }, + { + "epoch": 2.5686336717428087, + "grad_norm": 7.9375, + "learning_rate": 2.510360971819425e-06, + "loss": 2.4585, + "step": 24289 + }, + { + "epoch": 2.5687394247038915, + "grad_norm": 8.6875, + "learning_rate": 2.5101897170431965e-06, + "loss": 2.5294, + "step": 24290 + }, + { + "epoch": 2.5688451776649748, + "grad_norm": 8.0, + "learning_rate": 2.5100184622191524e-06, + "loss": 2.4872, + "step": 24291 + }, + { + "epoch": 2.5689509306260576, + "grad_norm": 8.625, + "learning_rate": 2.509847207348096e-06, + "loss": 2.6869, + "step": 24292 + }, + { + "epoch": 2.5690566835871405, + "grad_norm": 8.1875, + "learning_rate": 2.50967595243083e-06, + "loss": 2.4503, + "step": 24293 + }, + { + "epoch": 2.5691624365482233, + "grad_norm": 8.6875, + "learning_rate": 2.5095046974681596e-06, + "loss": 2.141, + "step": 24294 + }, + { + "epoch": 2.569268189509306, + "grad_norm": 8.875, + "learning_rate": 2.5093334424608874e-06, + "loss": 2.7385, + "step": 24295 + }, + { + "epoch": 2.5693739424703894, + "grad_norm": 8.0625, + "learning_rate": 2.5091621874098163e-06, + "loss": 2.5185, + "step": 24296 + }, + { + "epoch": 2.569479695431472, + "grad_norm": 9.375, + "learning_rate": 2.508990932315751e-06, + "loss": 2.6799, + "step": 24297 + }, + { + "epoch": 2.569585448392555, + "grad_norm": 8.125, + "learning_rate": 2.5088196771794957e-06, + "loss": 2.4551, + "step": 24298 + }, + { + "epoch": 2.569691201353638, + "grad_norm": 8.125, + "learning_rate": 2.5086484220018526e-06, + "loss": 2.3566, + "step": 24299 + }, + { + "epoch": 2.5697969543147208, + "grad_norm": 8.3125, + "learning_rate": 2.5084771667836268e-06, + "loss": 2.4549, + "step": 24300 + }, + { + "epoch": 2.5699027072758036, + "grad_norm": 9.0, + "learning_rate": 2.5083059115256205e-06, + "loss": 2.068, + "step": 24301 + }, + { + "epoch": 2.5700084602368864, + "grad_norm": 8.75, + "learning_rate": 2.5081346562286376e-06, + "loss": 2.4104, + "step": 24302 + }, + { + "epoch": 2.5701142131979697, + "grad_norm": 8.0625, + "learning_rate": 2.5079634008934827e-06, + "loss": 2.5483, + "step": 24303 + }, + { + "epoch": 2.5702199661590526, + "grad_norm": 8.0, + "learning_rate": 2.5077921455209593e-06, + "loss": 2.5275, + "step": 24304 + }, + { + "epoch": 2.5703257191201354, + "grad_norm": 8.9375, + "learning_rate": 2.5076208901118694e-06, + "loss": 2.3952, + "step": 24305 + }, + { + "epoch": 2.5704314720812182, + "grad_norm": 7.96875, + "learning_rate": 2.5074496346670174e-06, + "loss": 2.3755, + "step": 24306 + }, + { + "epoch": 2.570537225042301, + "grad_norm": 8.1875, + "learning_rate": 2.5072783791872084e-06, + "loss": 2.3979, + "step": 24307 + }, + { + "epoch": 2.5706429780033844, + "grad_norm": 8.5625, + "learning_rate": 2.5071071236732453e-06, + "loss": 2.6552, + "step": 24308 + }, + { + "epoch": 2.5707487309644668, + "grad_norm": 8.375, + "learning_rate": 2.5069358681259303e-06, + "loss": 2.5806, + "step": 24309 + }, + { + "epoch": 2.57085448392555, + "grad_norm": 8.0, + "learning_rate": 2.506764612546069e-06, + "loss": 2.5939, + "step": 24310 + }, + { + "epoch": 2.570960236886633, + "grad_norm": 8.5625, + "learning_rate": 2.5065933569344636e-06, + "loss": 2.4873, + "step": 24311 + }, + { + "epoch": 2.5710659898477157, + "grad_norm": 8.1875, + "learning_rate": 2.5064221012919183e-06, + "loss": 2.5612, + "step": 24312 + }, + { + "epoch": 2.5711717428087986, + "grad_norm": 11.625, + "learning_rate": 2.506250845619237e-06, + "loss": 2.5072, + "step": 24313 + }, + { + "epoch": 2.5712774957698814, + "grad_norm": 8.4375, + "learning_rate": 2.506079589917223e-06, + "loss": 2.3657, + "step": 24314 + }, + { + "epoch": 2.5713832487309647, + "grad_norm": 8.3125, + "learning_rate": 2.5059083341866792e-06, + "loss": 2.5316, + "step": 24315 + }, + { + "epoch": 2.5714890016920475, + "grad_norm": 8.25, + "learning_rate": 2.5057370784284107e-06, + "loss": 2.714, + "step": 24316 + }, + { + "epoch": 2.5715947546531304, + "grad_norm": 8.4375, + "learning_rate": 2.5055658226432204e-06, + "loss": 2.6803, + "step": 24317 + }, + { + "epoch": 2.571700507614213, + "grad_norm": 8.0, + "learning_rate": 2.505394566831912e-06, + "loss": 2.4376, + "step": 24318 + }, + { + "epoch": 2.571806260575296, + "grad_norm": 8.75, + "learning_rate": 2.505223310995289e-06, + "loss": 2.7756, + "step": 24319 + }, + { + "epoch": 2.571912013536379, + "grad_norm": 8.4375, + "learning_rate": 2.505052055134156e-06, + "loss": 2.5903, + "step": 24320 + }, + { + "epoch": 2.5720177664974617, + "grad_norm": 9.1875, + "learning_rate": 2.504880799249314e-06, + "loss": 2.6242, + "step": 24321 + }, + { + "epoch": 2.572123519458545, + "grad_norm": 8.4375, + "learning_rate": 2.5047095433415698e-06, + "loss": 2.5498, + "step": 24322 + }, + { + "epoch": 2.572229272419628, + "grad_norm": 8.625, + "learning_rate": 2.5045382874117245e-06, + "loss": 2.3038, + "step": 24323 + }, + { + "epoch": 2.5723350253807107, + "grad_norm": 9.0625, + "learning_rate": 2.504367031460584e-06, + "loss": 2.6352, + "step": 24324 + }, + { + "epoch": 2.5724407783417935, + "grad_norm": 8.8125, + "learning_rate": 2.50419577548895e-06, + "loss": 2.6155, + "step": 24325 + }, + { + "epoch": 2.5725465313028764, + "grad_norm": 8.375, + "learning_rate": 2.5040245194976274e-06, + "loss": 2.6197, + "step": 24326 + }, + { + "epoch": 2.5726522842639596, + "grad_norm": 8.375, + "learning_rate": 2.5038532634874194e-06, + "loss": 2.4381, + "step": 24327 + }, + { + "epoch": 2.572758037225042, + "grad_norm": 8.25, + "learning_rate": 2.5036820074591295e-06, + "loss": 2.5263, + "step": 24328 + }, + { + "epoch": 2.5728637901861253, + "grad_norm": 9.9375, + "learning_rate": 2.5035107514135613e-06, + "loss": 2.3806, + "step": 24329 + }, + { + "epoch": 2.572969543147208, + "grad_norm": 7.875, + "learning_rate": 2.5033394953515196e-06, + "loss": 2.4195, + "step": 24330 + }, + { + "epoch": 2.573075296108291, + "grad_norm": 8.5625, + "learning_rate": 2.5031682392738056e-06, + "loss": 2.657, + "step": 24331 + }, + { + "epoch": 2.573181049069374, + "grad_norm": 8.8125, + "learning_rate": 2.5029969831812246e-06, + "loss": 2.2797, + "step": 24332 + }, + { + "epoch": 2.5732868020304567, + "grad_norm": 9.0625, + "learning_rate": 2.5028257270745802e-06, + "loss": 2.4722, + "step": 24333 + }, + { + "epoch": 2.57339255499154, + "grad_norm": 8.125, + "learning_rate": 2.5026544709546756e-06, + "loss": 2.5284, + "step": 24334 + }, + { + "epoch": 2.573498307952623, + "grad_norm": 8.3125, + "learning_rate": 2.502483214822315e-06, + "loss": 2.5792, + "step": 24335 + }, + { + "epoch": 2.5736040609137056, + "grad_norm": 8.1875, + "learning_rate": 2.502311958678302e-06, + "loss": 2.4479, + "step": 24336 + }, + { + "epoch": 2.5737098138747885, + "grad_norm": 8.8125, + "learning_rate": 2.5021407025234395e-06, + "loss": 2.5272, + "step": 24337 + }, + { + "epoch": 2.5738155668358713, + "grad_norm": 8.5, + "learning_rate": 2.5019694463585315e-06, + "loss": 2.4601, + "step": 24338 + }, + { + "epoch": 2.573921319796954, + "grad_norm": 8.625, + "learning_rate": 2.501798190184382e-06, + "loss": 2.5538, + "step": 24339 + }, + { + "epoch": 2.574027072758037, + "grad_norm": 8.4375, + "learning_rate": 2.501626934001794e-06, + "loss": 2.3551, + "step": 24340 + }, + { + "epoch": 2.5741328257191203, + "grad_norm": 8.0625, + "learning_rate": 2.5014556778115712e-06, + "loss": 2.343, + "step": 24341 + }, + { + "epoch": 2.574238578680203, + "grad_norm": 8.25, + "learning_rate": 2.5012844216145184e-06, + "loss": 2.4408, + "step": 24342 + }, + { + "epoch": 2.574344331641286, + "grad_norm": 8.5625, + "learning_rate": 2.501113165411438e-06, + "loss": 2.5543, + "step": 24343 + }, + { + "epoch": 2.574450084602369, + "grad_norm": 8.875, + "learning_rate": 2.5009419092031333e-06, + "loss": 2.6327, + "step": 24344 + }, + { + "epoch": 2.5745558375634516, + "grad_norm": 8.0, + "learning_rate": 2.500770652990409e-06, + "loss": 2.5183, + "step": 24345 + }, + { + "epoch": 2.574661590524535, + "grad_norm": 8.125, + "learning_rate": 2.500599396774069e-06, + "loss": 2.5268, + "step": 24346 + }, + { + "epoch": 2.5747673434856178, + "grad_norm": 8.25, + "learning_rate": 2.5004281405549152e-06, + "loss": 2.4305, + "step": 24347 + }, + { + "epoch": 2.5748730964467006, + "grad_norm": 8.1875, + "learning_rate": 2.5002568843337533e-06, + "loss": 2.5244, + "step": 24348 + }, + { + "epoch": 2.5749788494077834, + "grad_norm": 7.96875, + "learning_rate": 2.5000856281113856e-06, + "loss": 2.3472, + "step": 24349 + }, + { + "epoch": 2.5750846023688663, + "grad_norm": 8.3125, + "learning_rate": 2.499914371888615e-06, + "loss": 2.6013, + "step": 24350 + }, + { + "epoch": 2.575190355329949, + "grad_norm": 8.1875, + "learning_rate": 2.499743115666248e-06, + "loss": 2.6449, + "step": 24351 + }, + { + "epoch": 2.575296108291032, + "grad_norm": 8.375, + "learning_rate": 2.4995718594450856e-06, + "loss": 2.6163, + "step": 24352 + }, + { + "epoch": 2.5754018612521152, + "grad_norm": 8.25, + "learning_rate": 2.4994006032259324e-06, + "loss": 2.5586, + "step": 24353 + }, + { + "epoch": 2.575507614213198, + "grad_norm": 8.75, + "learning_rate": 2.4992293470095914e-06, + "loss": 2.4266, + "step": 24354 + }, + { + "epoch": 2.575613367174281, + "grad_norm": 8.5, + "learning_rate": 2.499058090796867e-06, + "loss": 2.5151, + "step": 24355 + }, + { + "epoch": 2.5757191201353637, + "grad_norm": 8.0, + "learning_rate": 2.4988868345885628e-06, + "loss": 2.4733, + "step": 24356 + }, + { + "epoch": 2.5758248730964466, + "grad_norm": 8.3125, + "learning_rate": 2.498715578385482e-06, + "loss": 2.3589, + "step": 24357 + }, + { + "epoch": 2.57593062605753, + "grad_norm": 9.0625, + "learning_rate": 2.4985443221884296e-06, + "loss": 2.6823, + "step": 24358 + }, + { + "epoch": 2.5760363790186123, + "grad_norm": 8.0, + "learning_rate": 2.498373065998207e-06, + "loss": 2.5513, + "step": 24359 + }, + { + "epoch": 2.5761421319796955, + "grad_norm": 8.25, + "learning_rate": 2.498201809815619e-06, + "loss": 2.2509, + "step": 24360 + }, + { + "epoch": 2.5762478849407784, + "grad_norm": 9.1875, + "learning_rate": 2.4980305536414693e-06, + "loss": 2.4156, + "step": 24361 + }, + { + "epoch": 2.5763536379018612, + "grad_norm": 8.1875, + "learning_rate": 2.4978592974765614e-06, + "loss": 2.4684, + "step": 24362 + }, + { + "epoch": 2.576459390862944, + "grad_norm": 8.125, + "learning_rate": 2.4976880413216984e-06, + "loss": 2.6738, + "step": 24363 + }, + { + "epoch": 2.576565143824027, + "grad_norm": 8.6875, + "learning_rate": 2.497516785177685e-06, + "loss": 2.7532, + "step": 24364 + }, + { + "epoch": 2.57667089678511, + "grad_norm": 8.3125, + "learning_rate": 2.497345529045325e-06, + "loss": 2.5754, + "step": 24365 + }, + { + "epoch": 2.576776649746193, + "grad_norm": 8.9375, + "learning_rate": 2.4971742729254198e-06, + "loss": 2.2256, + "step": 24366 + }, + { + "epoch": 2.576882402707276, + "grad_norm": 8.375, + "learning_rate": 2.4970030168187763e-06, + "loss": 2.5076, + "step": 24367 + }, + { + "epoch": 2.5769881556683587, + "grad_norm": 8.5625, + "learning_rate": 2.496831760726195e-06, + "loss": 2.4868, + "step": 24368 + }, + { + "epoch": 2.5770939086294415, + "grad_norm": 8.375, + "learning_rate": 2.496660504648482e-06, + "loss": 2.4632, + "step": 24369 + }, + { + "epoch": 2.5771996615905244, + "grad_norm": 8.625, + "learning_rate": 2.496489248586439e-06, + "loss": 2.6465, + "step": 24370 + }, + { + "epoch": 2.577305414551607, + "grad_norm": 9.1875, + "learning_rate": 2.496317992540872e-06, + "loss": 2.5845, + "step": 24371 + }, + { + "epoch": 2.5774111675126905, + "grad_norm": 8.125, + "learning_rate": 2.496146736512581e-06, + "loss": 2.2546, + "step": 24372 + }, + { + "epoch": 2.5775169204737733, + "grad_norm": 8.0, + "learning_rate": 2.4959754805023726e-06, + "loss": 2.5534, + "step": 24373 + }, + { + "epoch": 2.577622673434856, + "grad_norm": 8.5, + "learning_rate": 2.4958042245110505e-06, + "loss": 2.5829, + "step": 24374 + }, + { + "epoch": 2.577728426395939, + "grad_norm": 9.0, + "learning_rate": 2.495632968539417e-06, + "loss": 2.6632, + "step": 24375 + }, + { + "epoch": 2.577834179357022, + "grad_norm": 7.90625, + "learning_rate": 2.495461712588276e-06, + "loss": 2.455, + "step": 24376 + }, + { + "epoch": 2.577939932318105, + "grad_norm": 7.9375, + "learning_rate": 2.495290456658431e-06, + "loss": 2.5147, + "step": 24377 + }, + { + "epoch": 2.578045685279188, + "grad_norm": 8.625, + "learning_rate": 2.4951192007506866e-06, + "loss": 2.588, + "step": 24378 + }, + { + "epoch": 2.578151438240271, + "grad_norm": 8.1875, + "learning_rate": 2.4949479448658454e-06, + "loss": 2.4897, + "step": 24379 + }, + { + "epoch": 2.5782571912013537, + "grad_norm": 9.5625, + "learning_rate": 2.4947766890047123e-06, + "loss": 2.5883, + "step": 24380 + }, + { + "epoch": 2.5783629441624365, + "grad_norm": 9.5, + "learning_rate": 2.494605433168089e-06, + "loss": 2.7417, + "step": 24381 + }, + { + "epoch": 2.5784686971235193, + "grad_norm": 8.1875, + "learning_rate": 2.4944341773567796e-06, + "loss": 2.5598, + "step": 24382 + }, + { + "epoch": 2.578574450084602, + "grad_norm": 7.9375, + "learning_rate": 2.4942629215715897e-06, + "loss": 2.4488, + "step": 24383 + }, + { + "epoch": 2.5786802030456855, + "grad_norm": 8.3125, + "learning_rate": 2.494091665813321e-06, + "loss": 2.5114, + "step": 24384 + }, + { + "epoch": 2.5787859560067683, + "grad_norm": 8.5, + "learning_rate": 2.493920410082778e-06, + "loss": 2.6592, + "step": 24385 + }, + { + "epoch": 2.578891708967851, + "grad_norm": 8.4375, + "learning_rate": 2.493749154380764e-06, + "loss": 2.5651, + "step": 24386 + }, + { + "epoch": 2.578997461928934, + "grad_norm": 8.5, + "learning_rate": 2.493577898708083e-06, + "loss": 2.4308, + "step": 24387 + }, + { + "epoch": 2.579103214890017, + "grad_norm": 8.875, + "learning_rate": 2.493406643065537e-06, + "loss": 2.4113, + "step": 24388 + }, + { + "epoch": 2.5792089678511, + "grad_norm": 8.5625, + "learning_rate": 2.4932353874539324e-06, + "loss": 2.3472, + "step": 24389 + }, + { + "epoch": 2.5793147208121825, + "grad_norm": 8.9375, + "learning_rate": 2.49306413187407e-06, + "loss": 2.6482, + "step": 24390 + }, + { + "epoch": 2.5794204737732658, + "grad_norm": 8.3125, + "learning_rate": 2.4928928763267555e-06, + "loss": 2.5251, + "step": 24391 + }, + { + "epoch": 2.5795262267343486, + "grad_norm": 8.5625, + "learning_rate": 2.492721620812792e-06, + "loss": 2.6882, + "step": 24392 + }, + { + "epoch": 2.5796319796954315, + "grad_norm": 8.6875, + "learning_rate": 2.4925503653329825e-06, + "loss": 2.4547, + "step": 24393 + }, + { + "epoch": 2.5797377326565143, + "grad_norm": 8.5, + "learning_rate": 2.492379109888132e-06, + "loss": 2.5426, + "step": 24394 + }, + { + "epoch": 2.579843485617597, + "grad_norm": 9.5, + "learning_rate": 2.492207854479042e-06, + "loss": 2.4876, + "step": 24395 + }, + { + "epoch": 2.5799492385786804, + "grad_norm": 8.25, + "learning_rate": 2.4920365991065185e-06, + "loss": 2.6164, + "step": 24396 + }, + { + "epoch": 2.5800549915397633, + "grad_norm": 8.9375, + "learning_rate": 2.491865343771363e-06, + "loss": 2.5985, + "step": 24397 + }, + { + "epoch": 2.580160744500846, + "grad_norm": 8.375, + "learning_rate": 2.4916940884743807e-06, + "loss": 2.5296, + "step": 24398 + }, + { + "epoch": 2.580266497461929, + "grad_norm": 8.4375, + "learning_rate": 2.4915228332163745e-06, + "loss": 2.6309, + "step": 24399 + }, + { + "epoch": 2.5803722504230118, + "grad_norm": 8.0, + "learning_rate": 2.4913515779981474e-06, + "loss": 2.3561, + "step": 24400 + }, + { + "epoch": 2.5804780033840946, + "grad_norm": 8.1875, + "learning_rate": 2.491180322820505e-06, + "loss": 2.4967, + "step": 24401 + }, + { + "epoch": 2.5805837563451774, + "grad_norm": 8.5, + "learning_rate": 2.491009067684249e-06, + "loss": 2.4209, + "step": 24402 + }, + { + "epoch": 2.5806895093062607, + "grad_norm": 8.375, + "learning_rate": 2.490837812590185e-06, + "loss": 2.5384, + "step": 24403 + }, + { + "epoch": 2.5807952622673436, + "grad_norm": 9.875, + "learning_rate": 2.4906665575391134e-06, + "loss": 2.5113, + "step": 24404 + }, + { + "epoch": 2.5809010152284264, + "grad_norm": 8.375, + "learning_rate": 2.4904953025318416e-06, + "loss": 2.6304, + "step": 24405 + }, + { + "epoch": 2.5810067681895092, + "grad_norm": 8.3125, + "learning_rate": 2.4903240475691703e-06, + "loss": 2.531, + "step": 24406 + }, + { + "epoch": 2.581112521150592, + "grad_norm": 8.625, + "learning_rate": 2.4901527926519047e-06, + "loss": 2.6824, + "step": 24407 + }, + { + "epoch": 2.5812182741116754, + "grad_norm": 8.375, + "learning_rate": 2.489981537780848e-06, + "loss": 2.4734, + "step": 24408 + }, + { + "epoch": 2.5813240270727578, + "grad_norm": 8.3125, + "learning_rate": 2.489810282956804e-06, + "loss": 2.623, + "step": 24409 + }, + { + "epoch": 2.581429780033841, + "grad_norm": 9.3125, + "learning_rate": 2.489639028180576e-06, + "loss": 2.6261, + "step": 24410 + }, + { + "epoch": 2.581535532994924, + "grad_norm": 8.8125, + "learning_rate": 2.4894677734529677e-06, + "loss": 2.6793, + "step": 24411 + }, + { + "epoch": 2.5816412859560067, + "grad_norm": 8.1875, + "learning_rate": 2.4892965187747835e-06, + "loss": 2.474, + "step": 24412 + }, + { + "epoch": 2.5817470389170896, + "grad_norm": 8.25, + "learning_rate": 2.4891252641468257e-06, + "loss": 2.5178, + "step": 24413 + }, + { + "epoch": 2.5818527918781724, + "grad_norm": 8.375, + "learning_rate": 2.4889540095698987e-06, + "loss": 2.5948, + "step": 24414 + }, + { + "epoch": 2.5819585448392557, + "grad_norm": 8.125, + "learning_rate": 2.488782755044806e-06, + "loss": 2.5574, + "step": 24415 + }, + { + "epoch": 2.5820642978003385, + "grad_norm": 8.0, + "learning_rate": 2.488611500572351e-06, + "loss": 2.5108, + "step": 24416 + }, + { + "epoch": 2.5821700507614214, + "grad_norm": 8.5625, + "learning_rate": 2.4884402461533382e-06, + "loss": 2.6294, + "step": 24417 + }, + { + "epoch": 2.582275803722504, + "grad_norm": 8.4375, + "learning_rate": 2.4882689917885697e-06, + "loss": 2.0415, + "step": 24418 + }, + { + "epoch": 2.582381556683587, + "grad_norm": 7.9375, + "learning_rate": 2.4880977374788515e-06, + "loss": 2.5559, + "step": 24419 + }, + { + "epoch": 2.5824873096446703, + "grad_norm": 8.25, + "learning_rate": 2.4879264832249838e-06, + "loss": 2.6362, + "step": 24420 + }, + { + "epoch": 2.5825930626057527, + "grad_norm": 8.5625, + "learning_rate": 2.4877552290277745e-06, + "loss": 2.8969, + "step": 24421 + }, + { + "epoch": 2.582698815566836, + "grad_norm": 8.875, + "learning_rate": 2.4875839748880228e-06, + "loss": 2.7463, + "step": 24422 + }, + { + "epoch": 2.582804568527919, + "grad_norm": 8.625, + "learning_rate": 2.4874127208065355e-06, + "loss": 2.6692, + "step": 24423 + }, + { + "epoch": 2.5829103214890017, + "grad_norm": 8.5625, + "learning_rate": 2.487241466784115e-06, + "loss": 2.7005, + "step": 24424 + }, + { + "epoch": 2.5830160744500845, + "grad_norm": 8.375, + "learning_rate": 2.487070212821564e-06, + "loss": 2.5152, + "step": 24425 + }, + { + "epoch": 2.5831218274111674, + "grad_norm": 8.625, + "learning_rate": 2.4868989589196886e-06, + "loss": 2.4557, + "step": 24426 + }, + { + "epoch": 2.5832275803722506, + "grad_norm": 8.1875, + "learning_rate": 2.48672770507929e-06, + "loss": 2.6187, + "step": 24427 + }, + { + "epoch": 2.5833333333333335, + "grad_norm": 8.5, + "learning_rate": 2.4865564513011746e-06, + "loss": 2.5227, + "step": 24428 + }, + { + "epoch": 2.5834390862944163, + "grad_norm": 8.375, + "learning_rate": 2.486385197586143e-06, + "loss": 2.6022, + "step": 24429 + }, + { + "epoch": 2.583544839255499, + "grad_norm": 9.0, + "learning_rate": 2.4862139439350006e-06, + "loss": 2.5989, + "step": 24430 + }, + { + "epoch": 2.583650592216582, + "grad_norm": 9.0, + "learning_rate": 2.4860426903485495e-06, + "loss": 2.415, + "step": 24431 + }, + { + "epoch": 2.583756345177665, + "grad_norm": 9.3125, + "learning_rate": 2.4858714368275956e-06, + "loss": 2.8335, + "step": 24432 + }, + { + "epoch": 2.5838620981387477, + "grad_norm": 8.5, + "learning_rate": 2.4857001833729402e-06, + "loss": 2.6932, + "step": 24433 + }, + { + "epoch": 2.583967851099831, + "grad_norm": 8.5, + "learning_rate": 2.485528929985388e-06, + "loss": 2.6131, + "step": 24434 + }, + { + "epoch": 2.584073604060914, + "grad_norm": 7.90625, + "learning_rate": 2.485357676665744e-06, + "loss": 2.518, + "step": 24435 + }, + { + "epoch": 2.5841793570219966, + "grad_norm": 8.5, + "learning_rate": 2.485186423414809e-06, + "loss": 2.6013, + "step": 24436 + }, + { + "epoch": 2.5842851099830795, + "grad_norm": 9.0625, + "learning_rate": 2.485015170233389e-06, + "loss": 2.7491, + "step": 24437 + }, + { + "epoch": 2.5843908629441623, + "grad_norm": 8.0625, + "learning_rate": 2.484843917122286e-06, + "loss": 2.5181, + "step": 24438 + }, + { + "epoch": 2.5844966159052456, + "grad_norm": 8.75, + "learning_rate": 2.4846726640823046e-06, + "loss": 2.5068, + "step": 24439 + }, + { + "epoch": 2.584602368866328, + "grad_norm": 9.25, + "learning_rate": 2.484501411114248e-06, + "loss": 2.8118, + "step": 24440 + }, + { + "epoch": 2.5847081218274113, + "grad_norm": 8.625, + "learning_rate": 2.4843301582189208e-06, + "loss": 2.569, + "step": 24441 + }, + { + "epoch": 2.584813874788494, + "grad_norm": 8.625, + "learning_rate": 2.484158905397124e-06, + "loss": 2.7049, + "step": 24442 + }, + { + "epoch": 2.584919627749577, + "grad_norm": 8.5, + "learning_rate": 2.483987652649663e-06, + "loss": 2.7594, + "step": 24443 + }, + { + "epoch": 2.58502538071066, + "grad_norm": 8.375, + "learning_rate": 2.4838163999773436e-06, + "loss": 2.397, + "step": 24444 + }, + { + "epoch": 2.5851311336717426, + "grad_norm": 8.1875, + "learning_rate": 2.4836451473809656e-06, + "loss": 2.6623, + "step": 24445 + }, + { + "epoch": 2.585236886632826, + "grad_norm": 8.4375, + "learning_rate": 2.483473894861335e-06, + "loss": 2.4978, + "step": 24446 + }, + { + "epoch": 2.5853426395939088, + "grad_norm": 8.25, + "learning_rate": 2.483302642419254e-06, + "loss": 2.6315, + "step": 24447 + }, + { + "epoch": 2.5854483925549916, + "grad_norm": 8.875, + "learning_rate": 2.483131390055527e-06, + "loss": 2.627, + "step": 24448 + }, + { + "epoch": 2.5855541455160744, + "grad_norm": 8.3125, + "learning_rate": 2.4829601377709574e-06, + "loss": 2.4972, + "step": 24449 + }, + { + "epoch": 2.5856598984771573, + "grad_norm": 8.5, + "learning_rate": 2.48278888556635e-06, + "loss": 2.5024, + "step": 24450 + }, + { + "epoch": 2.58576565143824, + "grad_norm": 8.5, + "learning_rate": 2.482617633442506e-06, + "loss": 2.6597, + "step": 24451 + }, + { + "epoch": 2.585871404399323, + "grad_norm": 8.125, + "learning_rate": 2.48244638140023e-06, + "loss": 2.4954, + "step": 24452 + }, + { + "epoch": 2.5859771573604062, + "grad_norm": 8.6875, + "learning_rate": 2.4822751294403274e-06, + "loss": 2.5149, + "step": 24453 + }, + { + "epoch": 2.586082910321489, + "grad_norm": 8.3125, + "learning_rate": 2.4821038775635995e-06, + "loss": 2.4857, + "step": 24454 + }, + { + "epoch": 2.586188663282572, + "grad_norm": 8.5625, + "learning_rate": 2.481932625770851e-06, + "loss": 2.5858, + "step": 24455 + }, + { + "epoch": 2.5862944162436547, + "grad_norm": 8.25, + "learning_rate": 2.481761374062885e-06, + "loss": 2.3727, + "step": 24456 + }, + { + "epoch": 2.5864001692047376, + "grad_norm": 7.96875, + "learning_rate": 2.4815901224405063e-06, + "loss": 2.4507, + "step": 24457 + }, + { + "epoch": 2.586505922165821, + "grad_norm": 8.1875, + "learning_rate": 2.4814188709045165e-06, + "loss": 2.3976, + "step": 24458 + }, + { + "epoch": 2.5866116751269037, + "grad_norm": 8.6875, + "learning_rate": 2.481247619455721e-06, + "loss": 2.7092, + "step": 24459 + }, + { + "epoch": 2.5867174280879865, + "grad_norm": 8.375, + "learning_rate": 2.481076368094923e-06, + "loss": 2.4863, + "step": 24460 + }, + { + "epoch": 2.5868231810490694, + "grad_norm": 9.0625, + "learning_rate": 2.480905116822925e-06, + "loss": 2.8229, + "step": 24461 + }, + { + "epoch": 2.5869289340101522, + "grad_norm": 8.375, + "learning_rate": 2.4807338656405324e-06, + "loss": 2.6417, + "step": 24462 + }, + { + "epoch": 2.587034686971235, + "grad_norm": 8.25, + "learning_rate": 2.4805626145485467e-06, + "loss": 2.57, + "step": 24463 + }, + { + "epoch": 2.587140439932318, + "grad_norm": 8.0625, + "learning_rate": 2.4803913635477738e-06, + "loss": 2.428, + "step": 24464 + }, + { + "epoch": 2.587246192893401, + "grad_norm": 8.3125, + "learning_rate": 2.4802201126390153e-06, + "loss": 2.5099, + "step": 24465 + }, + { + "epoch": 2.587351945854484, + "grad_norm": 8.875, + "learning_rate": 2.4800488618230772e-06, + "loss": 2.5095, + "step": 24466 + }, + { + "epoch": 2.587457698815567, + "grad_norm": 8.4375, + "learning_rate": 2.4798776111007605e-06, + "loss": 2.5974, + "step": 24467 + }, + { + "epoch": 2.5875634517766497, + "grad_norm": 8.5, + "learning_rate": 2.47970636047287e-06, + "loss": 2.5879, + "step": 24468 + }, + { + "epoch": 2.5876692047377325, + "grad_norm": 8.4375, + "learning_rate": 2.4795351099402095e-06, + "loss": 2.592, + "step": 24469 + }, + { + "epoch": 2.587774957698816, + "grad_norm": 8.5625, + "learning_rate": 2.4793638595035816e-06, + "loss": 2.4474, + "step": 24470 + }, + { + "epoch": 2.587880710659898, + "grad_norm": 8.5625, + "learning_rate": 2.4791926091637917e-06, + "loss": 2.6301, + "step": 24471 + }, + { + "epoch": 2.5879864636209815, + "grad_norm": 8.375, + "learning_rate": 2.479021358921642e-06, + "loss": 2.5919, + "step": 24472 + }, + { + "epoch": 2.5880922165820643, + "grad_norm": 8.5, + "learning_rate": 2.4788501087779366e-06, + "loss": 2.5289, + "step": 24473 + }, + { + "epoch": 2.588197969543147, + "grad_norm": 8.0, + "learning_rate": 2.4786788587334783e-06, + "loss": 2.4877, + "step": 24474 + }, + { + "epoch": 2.58830372250423, + "grad_norm": 8.4375, + "learning_rate": 2.4785076087890727e-06, + "loss": 2.673, + "step": 24475 + }, + { + "epoch": 2.588409475465313, + "grad_norm": 8.375, + "learning_rate": 2.4783363589455213e-06, + "loss": 2.7246, + "step": 24476 + }, + { + "epoch": 2.588515228426396, + "grad_norm": 8.5, + "learning_rate": 2.4781651092036286e-06, + "loss": 2.3212, + "step": 24477 + }, + { + "epoch": 2.588620981387479, + "grad_norm": 8.25, + "learning_rate": 2.4779938595641977e-06, + "loss": 2.59, + "step": 24478 + }, + { + "epoch": 2.588726734348562, + "grad_norm": 8.5, + "learning_rate": 2.4778226100280324e-06, + "loss": 2.4578, + "step": 24479 + }, + { + "epoch": 2.5888324873096447, + "grad_norm": 8.4375, + "learning_rate": 2.477651360595938e-06, + "loss": 2.3989, + "step": 24480 + }, + { + "epoch": 2.5889382402707275, + "grad_norm": 8.125, + "learning_rate": 2.477480111268715e-06, + "loss": 2.4016, + "step": 24481 + }, + { + "epoch": 2.5890439932318103, + "grad_norm": 8.5, + "learning_rate": 2.47730886204717e-06, + "loss": 2.6219, + "step": 24482 + }, + { + "epoch": 2.589149746192893, + "grad_norm": 8.8125, + "learning_rate": 2.4771376129321047e-06, + "loss": 2.6812, + "step": 24483 + }, + { + "epoch": 2.5892554991539765, + "grad_norm": 9.125, + "learning_rate": 2.476966363924323e-06, + "loss": 2.492, + "step": 24484 + }, + { + "epoch": 2.5893612521150593, + "grad_norm": 7.96875, + "learning_rate": 2.4767951150246287e-06, + "loss": 2.4414, + "step": 24485 + }, + { + "epoch": 2.589467005076142, + "grad_norm": 9.0625, + "learning_rate": 2.4766238662338257e-06, + "loss": 2.5683, + "step": 24486 + }, + { + "epoch": 2.589572758037225, + "grad_norm": 8.3125, + "learning_rate": 2.476452617552717e-06, + "loss": 2.6393, + "step": 24487 + }, + { + "epoch": 2.589678510998308, + "grad_norm": 8.375, + "learning_rate": 2.4762813689821064e-06, + "loss": 2.5397, + "step": 24488 + }, + { + "epoch": 2.589784263959391, + "grad_norm": 8.5625, + "learning_rate": 2.4761101205227987e-06, + "loss": 2.6724, + "step": 24489 + }, + { + "epoch": 2.589890016920474, + "grad_norm": 8.0625, + "learning_rate": 2.475938872175595e-06, + "loss": 2.404, + "step": 24490 + }, + { + "epoch": 2.5899957698815568, + "grad_norm": 8.3125, + "learning_rate": 2.475767623941302e-06, + "loss": 2.3509, + "step": 24491 + }, + { + "epoch": 2.5901015228426396, + "grad_norm": 8.0, + "learning_rate": 2.47559637582072e-06, + "loss": 2.5094, + "step": 24492 + }, + { + "epoch": 2.5902072758037225, + "grad_norm": 8.125, + "learning_rate": 2.4754251278146554e-06, + "loss": 2.4486, + "step": 24493 + }, + { + "epoch": 2.5903130287648053, + "grad_norm": 8.3125, + "learning_rate": 2.47525387992391e-06, + "loss": 2.3228, + "step": 24494 + }, + { + "epoch": 2.590418781725888, + "grad_norm": 8.125, + "learning_rate": 2.4750826321492886e-06, + "loss": 2.3948, + "step": 24495 + }, + { + "epoch": 2.5905245346869714, + "grad_norm": 8.0, + "learning_rate": 2.474911384491593e-06, + "loss": 2.4215, + "step": 24496 + }, + { + "epoch": 2.5906302876480543, + "grad_norm": 8.125, + "learning_rate": 2.4747401369516285e-06, + "loss": 2.6229, + "step": 24497 + }, + { + "epoch": 2.590736040609137, + "grad_norm": 8.1875, + "learning_rate": 2.474568889530199e-06, + "loss": 2.5568, + "step": 24498 + }, + { + "epoch": 2.59084179357022, + "grad_norm": 9.1875, + "learning_rate": 2.474397642228107e-06, + "loss": 2.5415, + "step": 24499 + }, + { + "epoch": 2.5909475465313028, + "grad_norm": 7.96875, + "learning_rate": 2.474226395046156e-06, + "loss": 2.5792, + "step": 24500 + }, + { + "epoch": 2.591053299492386, + "grad_norm": 8.0625, + "learning_rate": 2.4740551479851503e-06, + "loss": 2.5701, + "step": 24501 + }, + { + "epoch": 2.5911590524534684, + "grad_norm": 9.4375, + "learning_rate": 2.473883901045893e-06, + "loss": 2.6808, + "step": 24502 + }, + { + "epoch": 2.5912648054145517, + "grad_norm": 8.4375, + "learning_rate": 2.4737126542291877e-06, + "loss": 2.7027, + "step": 24503 + }, + { + "epoch": 2.5913705583756346, + "grad_norm": 8.375, + "learning_rate": 2.473541407535839e-06, + "loss": 2.2251, + "step": 24504 + }, + { + "epoch": 2.5914763113367174, + "grad_norm": 8.1875, + "learning_rate": 2.473370160966649e-06, + "loss": 2.5992, + "step": 24505 + }, + { + "epoch": 2.5915820642978002, + "grad_norm": 8.0625, + "learning_rate": 2.473198914522421e-06, + "loss": 2.6201, + "step": 24506 + }, + { + "epoch": 2.591687817258883, + "grad_norm": 8.25, + "learning_rate": 2.4730276682039616e-06, + "loss": 2.5187, + "step": 24507 + }, + { + "epoch": 2.5917935702199664, + "grad_norm": 9.0625, + "learning_rate": 2.472856422012071e-06, + "loss": 2.389, + "step": 24508 + }, + { + "epoch": 2.591899323181049, + "grad_norm": 8.8125, + "learning_rate": 2.4726851759475543e-06, + "loss": 2.691, + "step": 24509 + }, + { + "epoch": 2.592005076142132, + "grad_norm": 8.6875, + "learning_rate": 2.4725139300112147e-06, + "loss": 2.3086, + "step": 24510 + }, + { + "epoch": 2.592110829103215, + "grad_norm": 8.5625, + "learning_rate": 2.472342684203857e-06, + "loss": 2.7268, + "step": 24511 + }, + { + "epoch": 2.5922165820642977, + "grad_norm": 7.96875, + "learning_rate": 2.472171438526282e-06, + "loss": 2.5348, + "step": 24512 + }, + { + "epoch": 2.5923223350253806, + "grad_norm": 8.625, + "learning_rate": 2.4720001929792968e-06, + "loss": 2.5149, + "step": 24513 + }, + { + "epoch": 2.5924280879864634, + "grad_norm": 9.1875, + "learning_rate": 2.4718289475637024e-06, + "loss": 2.5131, + "step": 24514 + }, + { + "epoch": 2.5925338409475467, + "grad_norm": 8.25, + "learning_rate": 2.4716577022803028e-06, + "loss": 2.5478, + "step": 24515 + }, + { + "epoch": 2.5926395939086295, + "grad_norm": 8.0625, + "learning_rate": 2.471486457129903e-06, + "loss": 2.5067, + "step": 24516 + }, + { + "epoch": 2.5927453468697124, + "grad_norm": 8.0625, + "learning_rate": 2.4713152121133047e-06, + "loss": 2.5323, + "step": 24517 + }, + { + "epoch": 2.592851099830795, + "grad_norm": 8.1875, + "learning_rate": 2.471143967231313e-06, + "loss": 2.4849, + "step": 24518 + }, + { + "epoch": 2.592956852791878, + "grad_norm": 7.875, + "learning_rate": 2.4709727224847304e-06, + "loss": 2.5185, + "step": 24519 + }, + { + "epoch": 2.5930626057529613, + "grad_norm": 9.0, + "learning_rate": 2.4708014778743623e-06, + "loss": 2.4132, + "step": 24520 + }, + { + "epoch": 2.5931683587140437, + "grad_norm": 8.6875, + "learning_rate": 2.4706302334010095e-06, + "loss": 2.5271, + "step": 24521 + }, + { + "epoch": 2.593274111675127, + "grad_norm": 8.375, + "learning_rate": 2.470458989065476e-06, + "loss": 2.528, + "step": 24522 + }, + { + "epoch": 2.59337986463621, + "grad_norm": 8.4375, + "learning_rate": 2.470287744868569e-06, + "loss": 2.3799, + "step": 24523 + }, + { + "epoch": 2.5934856175972927, + "grad_norm": 8.75, + "learning_rate": 2.470116500811088e-06, + "loss": 2.6157, + "step": 24524 + }, + { + "epoch": 2.5935913705583755, + "grad_norm": 7.96875, + "learning_rate": 2.469945256893838e-06, + "loss": 2.4683, + "step": 24525 + }, + { + "epoch": 2.5936971235194584, + "grad_norm": 8.375, + "learning_rate": 2.4697740131176225e-06, + "loss": 2.6792, + "step": 24526 + }, + { + "epoch": 2.5938028764805416, + "grad_norm": 9.375, + "learning_rate": 2.4696027694832465e-06, + "loss": 2.307, + "step": 24527 + }, + { + "epoch": 2.5939086294416245, + "grad_norm": 8.5625, + "learning_rate": 2.4694315259915104e-06, + "loss": 2.6187, + "step": 24528 + }, + { + "epoch": 2.5940143824027073, + "grad_norm": 8.1875, + "learning_rate": 2.4692602826432215e-06, + "loss": 2.5106, + "step": 24529 + }, + { + "epoch": 2.59412013536379, + "grad_norm": 8.75, + "learning_rate": 2.46908903943918e-06, + "loss": 2.7567, + "step": 24530 + }, + { + "epoch": 2.594225888324873, + "grad_norm": 8.75, + "learning_rate": 2.4689177963801915e-06, + "loss": 2.5773, + "step": 24531 + }, + { + "epoch": 2.5943316412859563, + "grad_norm": 8.8125, + "learning_rate": 2.4687465534670594e-06, + "loss": 2.6501, + "step": 24532 + }, + { + "epoch": 2.5944373942470387, + "grad_norm": 9.0, + "learning_rate": 2.4685753107005864e-06, + "loss": 2.6228, + "step": 24533 + }, + { + "epoch": 2.594543147208122, + "grad_norm": 8.375, + "learning_rate": 2.468404068081577e-06, + "loss": 2.5001, + "step": 24534 + }, + { + "epoch": 2.594648900169205, + "grad_norm": 8.25, + "learning_rate": 2.468232825610834e-06, + "loss": 2.513, + "step": 24535 + }, + { + "epoch": 2.5947546531302876, + "grad_norm": 8.25, + "learning_rate": 2.4680615832891624e-06, + "loss": 2.4939, + "step": 24536 + }, + { + "epoch": 2.5948604060913705, + "grad_norm": 7.9375, + "learning_rate": 2.4678903411173637e-06, + "loss": 2.5892, + "step": 24537 + }, + { + "epoch": 2.5949661590524533, + "grad_norm": 9.0, + "learning_rate": 2.467719099096243e-06, + "loss": 2.622, + "step": 24538 + }, + { + "epoch": 2.5950719120135366, + "grad_norm": 8.6875, + "learning_rate": 2.467547857226603e-06, + "loss": 2.4713, + "step": 24539 + }, + { + "epoch": 2.5951776649746194, + "grad_norm": 8.9375, + "learning_rate": 2.4673766155092475e-06, + "loss": 2.6077, + "step": 24540 + }, + { + "epoch": 2.5952834179357023, + "grad_norm": 8.3125, + "learning_rate": 2.4672053739449806e-06, + "loss": 2.5749, + "step": 24541 + }, + { + "epoch": 2.595389170896785, + "grad_norm": 8.5, + "learning_rate": 2.4670341325346053e-06, + "loss": 2.3945, + "step": 24542 + }, + { + "epoch": 2.595494923857868, + "grad_norm": 8.1875, + "learning_rate": 2.4668628912789256e-06, + "loss": 2.6419, + "step": 24543 + }, + { + "epoch": 2.595600676818951, + "grad_norm": 9.3125, + "learning_rate": 2.466691650178744e-06, + "loss": 2.7378, + "step": 24544 + }, + { + "epoch": 2.5957064297800336, + "grad_norm": 8.0, + "learning_rate": 2.466520409234866e-06, + "loss": 2.3889, + "step": 24545 + }, + { + "epoch": 2.595812182741117, + "grad_norm": 7.96875, + "learning_rate": 2.466349168448093e-06, + "loss": 2.3991, + "step": 24546 + }, + { + "epoch": 2.5959179357021998, + "grad_norm": 8.9375, + "learning_rate": 2.4661779278192306e-06, + "loss": 2.7263, + "step": 24547 + }, + { + "epoch": 2.5960236886632826, + "grad_norm": 8.6875, + "learning_rate": 2.4660066873490805e-06, + "loss": 2.5289, + "step": 24548 + }, + { + "epoch": 2.5961294416243654, + "grad_norm": 9.0, + "learning_rate": 2.465835447038447e-06, + "loss": 2.7528, + "step": 24549 + }, + { + "epoch": 2.5962351945854483, + "grad_norm": 8.5625, + "learning_rate": 2.4656642068881347e-06, + "loss": 2.3573, + "step": 24550 + }, + { + "epoch": 2.5963409475465316, + "grad_norm": 7.96875, + "learning_rate": 2.465492966898945e-06, + "loss": 2.5015, + "step": 24551 + }, + { + "epoch": 2.596446700507614, + "grad_norm": 8.9375, + "learning_rate": 2.4653217270716845e-06, + "loss": 2.4016, + "step": 24552 + }, + { + "epoch": 2.5965524534686972, + "grad_norm": 8.125, + "learning_rate": 2.4651504874071537e-06, + "loss": 2.4963, + "step": 24553 + }, + { + "epoch": 2.59665820642978, + "grad_norm": 8.4375, + "learning_rate": 2.464979247906158e-06, + "loss": 2.6431, + "step": 24554 + }, + { + "epoch": 2.596763959390863, + "grad_norm": 8.1875, + "learning_rate": 2.464808008569499e-06, + "loss": 2.6751, + "step": 24555 + }, + { + "epoch": 2.5968697123519457, + "grad_norm": 8.4375, + "learning_rate": 2.464636769397983e-06, + "loss": 2.6955, + "step": 24556 + }, + { + "epoch": 2.5969754653130286, + "grad_norm": 7.96875, + "learning_rate": 2.4644655303924113e-06, + "loss": 2.5278, + "step": 24557 + }, + { + "epoch": 2.597081218274112, + "grad_norm": 8.3125, + "learning_rate": 2.4642942915535885e-06, + "loss": 2.4162, + "step": 24558 + }, + { + "epoch": 2.5971869712351947, + "grad_norm": 8.375, + "learning_rate": 2.464123052882319e-06, + "loss": 2.5543, + "step": 24559 + }, + { + "epoch": 2.5972927241962775, + "grad_norm": 8.5, + "learning_rate": 2.463951814379404e-06, + "loss": 2.2679, + "step": 24560 + }, + { + "epoch": 2.5973984771573604, + "grad_norm": 8.125, + "learning_rate": 2.4637805760456494e-06, + "loss": 2.3062, + "step": 24561 + }, + { + "epoch": 2.5975042301184432, + "grad_norm": 8.8125, + "learning_rate": 2.4636093378818573e-06, + "loss": 2.4948, + "step": 24562 + }, + { + "epoch": 2.597609983079526, + "grad_norm": 9.0, + "learning_rate": 2.4634380998888317e-06, + "loss": 2.7148, + "step": 24563 + }, + { + "epoch": 2.597715736040609, + "grad_norm": 8.125, + "learning_rate": 2.463266862067376e-06, + "loss": 2.5028, + "step": 24564 + }, + { + "epoch": 2.597821489001692, + "grad_norm": 8.0, + "learning_rate": 2.463095624418295e-06, + "loss": 2.472, + "step": 24565 + }, + { + "epoch": 2.597927241962775, + "grad_norm": 8.1875, + "learning_rate": 2.4629243869423893e-06, + "loss": 2.444, + "step": 24566 + }, + { + "epoch": 2.598032994923858, + "grad_norm": 8.3125, + "learning_rate": 2.462753149640465e-06, + "loss": 2.5312, + "step": 24567 + }, + { + "epoch": 2.5981387478849407, + "grad_norm": 8.8125, + "learning_rate": 2.462581912513326e-06, + "loss": 2.544, + "step": 24568 + }, + { + "epoch": 2.5982445008460235, + "grad_norm": 8.6875, + "learning_rate": 2.4624106755617736e-06, + "loss": 2.6415, + "step": 24569 + }, + { + "epoch": 2.598350253807107, + "grad_norm": 8.6875, + "learning_rate": 2.462239438786613e-06, + "loss": 2.5415, + "step": 24570 + }, + { + "epoch": 2.5984560067681897, + "grad_norm": 8.1875, + "learning_rate": 2.4620682021886467e-06, + "loss": 2.3572, + "step": 24571 + }, + { + "epoch": 2.5985617597292725, + "grad_norm": 8.5625, + "learning_rate": 2.4618969657686794e-06, + "loss": 2.3297, + "step": 24572 + }, + { + "epoch": 2.5986675126903553, + "grad_norm": 8.625, + "learning_rate": 2.4617257295275136e-06, + "loss": 2.8874, + "step": 24573 + }, + { + "epoch": 2.598773265651438, + "grad_norm": 7.90625, + "learning_rate": 2.4615544934659544e-06, + "loss": 2.4528, + "step": 24574 + }, + { + "epoch": 2.598879018612521, + "grad_norm": 8.75, + "learning_rate": 2.461383257584803e-06, + "loss": 2.6426, + "step": 24575 + }, + { + "epoch": 2.598984771573604, + "grad_norm": 8.1875, + "learning_rate": 2.461212021884864e-06, + "loss": 2.3947, + "step": 24576 + }, + { + "epoch": 2.599090524534687, + "grad_norm": 8.75, + "learning_rate": 2.4610407863669425e-06, + "loss": 2.6886, + "step": 24577 + }, + { + "epoch": 2.59919627749577, + "grad_norm": 8.25, + "learning_rate": 2.4608695510318393e-06, + "loss": 2.345, + "step": 24578 + }, + { + "epoch": 2.599302030456853, + "grad_norm": 8.75, + "learning_rate": 2.4606983158803602e-06, + "loss": 2.7781, + "step": 24579 + }, + { + "epoch": 2.5994077834179357, + "grad_norm": 8.5625, + "learning_rate": 2.4605270809133073e-06, + "loss": 2.5623, + "step": 24580 + }, + { + "epoch": 2.5995135363790185, + "grad_norm": 7.96875, + "learning_rate": 2.4603558461314857e-06, + "loss": 2.316, + "step": 24581 + }, + { + "epoch": 2.599619289340102, + "grad_norm": 8.0625, + "learning_rate": 2.460184611535696e-06, + "loss": 2.5145, + "step": 24582 + }, + { + "epoch": 2.599725042301184, + "grad_norm": 8.6875, + "learning_rate": 2.4600133771267457e-06, + "loss": 2.5247, + "step": 24583 + }, + { + "epoch": 2.5998307952622675, + "grad_norm": 9.125, + "learning_rate": 2.459842142905435e-06, + "loss": 2.4506, + "step": 24584 + }, + { + "epoch": 2.5999365482233503, + "grad_norm": 8.625, + "learning_rate": 2.4596709088725683e-06, + "loss": 2.5897, + "step": 24585 + }, + { + "epoch": 2.600042301184433, + "grad_norm": 8.0625, + "learning_rate": 2.459499675028951e-06, + "loss": 2.4162, + "step": 24586 + }, + { + "epoch": 2.600148054145516, + "grad_norm": 8.5, + "learning_rate": 2.459328441375384e-06, + "loss": 2.6476, + "step": 24587 + }, + { + "epoch": 2.600253807106599, + "grad_norm": 9.0, + "learning_rate": 2.459157207912672e-06, + "loss": 2.505, + "step": 24588 + }, + { + "epoch": 2.600359560067682, + "grad_norm": 8.875, + "learning_rate": 2.458985974641619e-06, + "loss": 2.5875, + "step": 24589 + }, + { + "epoch": 2.600465313028765, + "grad_norm": 8.375, + "learning_rate": 2.4588147415630284e-06, + "loss": 2.5672, + "step": 24590 + }, + { + "epoch": 2.6005710659898478, + "grad_norm": 8.125, + "learning_rate": 2.458643508677703e-06, + "loss": 2.4733, + "step": 24591 + }, + { + "epoch": 2.6006768189509306, + "grad_norm": 9.0, + "learning_rate": 2.4584722759864467e-06, + "loss": 2.6568, + "step": 24592 + }, + { + "epoch": 2.6007825719120135, + "grad_norm": 8.125, + "learning_rate": 2.4583010434900622e-06, + "loss": 2.6031, + "step": 24593 + }, + { + "epoch": 2.6008883248730963, + "grad_norm": 8.0625, + "learning_rate": 2.4581298111893547e-06, + "loss": 2.4348, + "step": 24594 + }, + { + "epoch": 2.600994077834179, + "grad_norm": 8.875, + "learning_rate": 2.4579585790851266e-06, + "loss": 2.6827, + "step": 24595 + }, + { + "epoch": 2.6010998307952624, + "grad_norm": 8.625, + "learning_rate": 2.4577873471781814e-06, + "loss": 2.4834, + "step": 24596 + }, + { + "epoch": 2.6012055837563453, + "grad_norm": 7.90625, + "learning_rate": 2.457616115469324e-06, + "loss": 2.6548, + "step": 24597 + }, + { + "epoch": 2.601311336717428, + "grad_norm": 8.4375, + "learning_rate": 2.4574448839593557e-06, + "loss": 2.2935, + "step": 24598 + }, + { + "epoch": 2.601417089678511, + "grad_norm": 8.6875, + "learning_rate": 2.4572736526490825e-06, + "loss": 2.7827, + "step": 24599 + }, + { + "epoch": 2.6015228426395938, + "grad_norm": 8.4375, + "learning_rate": 2.4571024215393054e-06, + "loss": 2.1035, + "step": 24600 + }, + { + "epoch": 2.601628595600677, + "grad_norm": 8.125, + "learning_rate": 2.4569311906308295e-06, + "loss": 2.4136, + "step": 24601 + }, + { + "epoch": 2.60173434856176, + "grad_norm": 9.5, + "learning_rate": 2.4567599599244573e-06, + "loss": 2.7244, + "step": 24602 + }, + { + "epoch": 2.6018401015228427, + "grad_norm": 9.0625, + "learning_rate": 2.456588729420993e-06, + "loss": 2.6516, + "step": 24603 + }, + { + "epoch": 2.6019458544839256, + "grad_norm": 9.3125, + "learning_rate": 2.456417499121241e-06, + "loss": 2.5104, + "step": 24604 + }, + { + "epoch": 2.6020516074450084, + "grad_norm": 8.9375, + "learning_rate": 2.456246269026003e-06, + "loss": 2.5835, + "step": 24605 + }, + { + "epoch": 2.6021573604060912, + "grad_norm": 8.375, + "learning_rate": 2.456075039136084e-06, + "loss": 2.4493, + "step": 24606 + }, + { + "epoch": 2.602263113367174, + "grad_norm": 8.375, + "learning_rate": 2.4559038094522866e-06, + "loss": 2.5304, + "step": 24607 + }, + { + "epoch": 2.6023688663282574, + "grad_norm": 8.6875, + "learning_rate": 2.455732579975415e-06, + "loss": 2.4121, + "step": 24608 + }, + { + "epoch": 2.60247461928934, + "grad_norm": 8.0625, + "learning_rate": 2.4555613507062713e-06, + "loss": 2.4579, + "step": 24609 + }, + { + "epoch": 2.602580372250423, + "grad_norm": 8.75, + "learning_rate": 2.455390121645661e-06, + "loss": 2.2383, + "step": 24610 + }, + { + "epoch": 2.602686125211506, + "grad_norm": 8.4375, + "learning_rate": 2.4552188927943858e-06, + "loss": 2.183, + "step": 24611 + }, + { + "epoch": 2.6027918781725887, + "grad_norm": 8.25, + "learning_rate": 2.45504766415325e-06, + "loss": 2.5546, + "step": 24612 + }, + { + "epoch": 2.602897631133672, + "grad_norm": 9.625, + "learning_rate": 2.4548764357230584e-06, + "loss": 2.7434, + "step": 24613 + }, + { + "epoch": 2.6030033840947544, + "grad_norm": 8.4375, + "learning_rate": 2.4547052075046113e-06, + "loss": 2.5901, + "step": 24614 + }, + { + "epoch": 2.6031091370558377, + "grad_norm": 8.125, + "learning_rate": 2.454533979498716e-06, + "loss": 2.4121, + "step": 24615 + }, + { + "epoch": 2.6032148900169205, + "grad_norm": 8.25, + "learning_rate": 2.454362751706173e-06, + "loss": 2.5181, + "step": 24616 + }, + { + "epoch": 2.6033206429780034, + "grad_norm": 8.125, + "learning_rate": 2.4541915241277875e-06, + "loss": 2.7611, + "step": 24617 + }, + { + "epoch": 2.603426395939086, + "grad_norm": 8.1875, + "learning_rate": 2.454020296764362e-06, + "loss": 2.4351, + "step": 24618 + }, + { + "epoch": 2.603532148900169, + "grad_norm": 8.0625, + "learning_rate": 2.4538490696167015e-06, + "loss": 2.5165, + "step": 24619 + }, + { + "epoch": 2.6036379018612523, + "grad_norm": 8.0, + "learning_rate": 2.4536778426856065e-06, + "loss": 2.4461, + "step": 24620 + }, + { + "epoch": 2.603743654822335, + "grad_norm": 8.375, + "learning_rate": 2.4535066159718833e-06, + "loss": 2.6007, + "step": 24621 + }, + { + "epoch": 2.603849407783418, + "grad_norm": 8.4375, + "learning_rate": 2.4533353894763356e-06, + "loss": 2.6256, + "step": 24622 + }, + { + "epoch": 2.603955160744501, + "grad_norm": 8.75, + "learning_rate": 2.4531641631997645e-06, + "loss": 2.4956, + "step": 24623 + }, + { + "epoch": 2.6040609137055837, + "grad_norm": 8.875, + "learning_rate": 2.4529929371429755e-06, + "loss": 2.3427, + "step": 24624 + }, + { + "epoch": 2.6041666666666665, + "grad_norm": 7.84375, + "learning_rate": 2.452821711306771e-06, + "loss": 2.4443, + "step": 24625 + }, + { + "epoch": 2.6042724196277494, + "grad_norm": 8.5, + "learning_rate": 2.4526504856919557e-06, + "loss": 2.5229, + "step": 24626 + }, + { + "epoch": 2.6043781725888326, + "grad_norm": 8.0625, + "learning_rate": 2.4524792602993312e-06, + "loss": 2.3549, + "step": 24627 + }, + { + "epoch": 2.6044839255499155, + "grad_norm": 8.375, + "learning_rate": 2.4523080351297034e-06, + "loss": 2.5976, + "step": 24628 + }, + { + "epoch": 2.6045896785109983, + "grad_norm": 8.75, + "learning_rate": 2.4521368101838736e-06, + "loss": 2.5512, + "step": 24629 + }, + { + "epoch": 2.604695431472081, + "grad_norm": 8.0625, + "learning_rate": 2.451965585462646e-06, + "loss": 2.4685, + "step": 24630 + }, + { + "epoch": 2.604801184433164, + "grad_norm": 8.25, + "learning_rate": 2.4517943609668244e-06, + "loss": 2.5754, + "step": 24631 + }, + { + "epoch": 2.6049069373942473, + "grad_norm": 7.75, + "learning_rate": 2.4516231366972123e-06, + "loss": 2.5079, + "step": 24632 + }, + { + "epoch": 2.6050126903553297, + "grad_norm": 9.6875, + "learning_rate": 2.451451912654613e-06, + "loss": 2.6838, + "step": 24633 + }, + { + "epoch": 2.605118443316413, + "grad_norm": 8.8125, + "learning_rate": 2.4512806888398298e-06, + "loss": 2.3703, + "step": 24634 + }, + { + "epoch": 2.605224196277496, + "grad_norm": 8.8125, + "learning_rate": 2.4511094652536678e-06, + "loss": 2.6092, + "step": 24635 + }, + { + "epoch": 2.6053299492385786, + "grad_norm": 8.5625, + "learning_rate": 2.4509382418969272e-06, + "loss": 2.2597, + "step": 24636 + }, + { + "epoch": 2.6054357021996615, + "grad_norm": 8.1875, + "learning_rate": 2.4507670187704136e-06, + "loss": 2.5211, + "step": 24637 + }, + { + "epoch": 2.6055414551607443, + "grad_norm": 8.6875, + "learning_rate": 2.4505957958749317e-06, + "loss": 2.4889, + "step": 24638 + }, + { + "epoch": 2.6056472081218276, + "grad_norm": 8.125, + "learning_rate": 2.450424573211282e-06, + "loss": 2.5606, + "step": 24639 + }, + { + "epoch": 2.6057529610829104, + "grad_norm": 8.3125, + "learning_rate": 2.4502533507802707e-06, + "loss": 2.4959, + "step": 24640 + }, + { + "epoch": 2.6058587140439933, + "grad_norm": 8.5625, + "learning_rate": 2.450082128582699e-06, + "loss": 2.5642, + "step": 24641 + }, + { + "epoch": 2.605964467005076, + "grad_norm": 8.5625, + "learning_rate": 2.449910906619372e-06, + "loss": 2.1304, + "step": 24642 + }, + { + "epoch": 2.606070219966159, + "grad_norm": 8.5625, + "learning_rate": 2.449739684891092e-06, + "loss": 2.5513, + "step": 24643 + }, + { + "epoch": 2.6061759729272422, + "grad_norm": 8.0625, + "learning_rate": 2.449568463398665e-06, + "loss": 2.3688, + "step": 24644 + }, + { + "epoch": 2.6062817258883246, + "grad_norm": 8.3125, + "learning_rate": 2.449397242142891e-06, + "loss": 2.5067, + "step": 24645 + }, + { + "epoch": 2.606387478849408, + "grad_norm": 9.5625, + "learning_rate": 2.449226021124575e-06, + "loss": 2.7388, + "step": 24646 + }, + { + "epoch": 2.6064932318104908, + "grad_norm": 8.625, + "learning_rate": 2.449054800344521e-06, + "loss": 2.4812, + "step": 24647 + }, + { + "epoch": 2.6065989847715736, + "grad_norm": 8.6875, + "learning_rate": 2.4488835798035312e-06, + "loss": 2.7806, + "step": 24648 + }, + { + "epoch": 2.6067047377326564, + "grad_norm": 7.90625, + "learning_rate": 2.448712359502411e-06, + "loss": 2.5817, + "step": 24649 + }, + { + "epoch": 2.6068104906937393, + "grad_norm": 8.25, + "learning_rate": 2.4485411394419617e-06, + "loss": 2.4816, + "step": 24650 + }, + { + "epoch": 2.6069162436548226, + "grad_norm": 7.75, + "learning_rate": 2.448369919622989e-06, + "loss": 2.392, + "step": 24651 + }, + { + "epoch": 2.6070219966159054, + "grad_norm": 8.1875, + "learning_rate": 2.4481987000462934e-06, + "loss": 2.5195, + "step": 24652 + }, + { + "epoch": 2.6071277495769882, + "grad_norm": 8.25, + "learning_rate": 2.448027480712682e-06, + "loss": 2.3773, + "step": 24653 + }, + { + "epoch": 2.607233502538071, + "grad_norm": 8.375, + "learning_rate": 2.447856261622955e-06, + "loss": 2.5723, + "step": 24654 + }, + { + "epoch": 2.607339255499154, + "grad_norm": 8.625, + "learning_rate": 2.4476850427779176e-06, + "loss": 2.5734, + "step": 24655 + }, + { + "epoch": 2.6074450084602367, + "grad_norm": 8.3125, + "learning_rate": 2.4475138241783733e-06, + "loss": 2.5809, + "step": 24656 + }, + { + "epoch": 2.6075507614213196, + "grad_norm": 8.5, + "learning_rate": 2.4473426058251246e-06, + "loss": 2.5534, + "step": 24657 + }, + { + "epoch": 2.607656514382403, + "grad_norm": 8.75, + "learning_rate": 2.447171387718976e-06, + "loss": 2.6133, + "step": 24658 + }, + { + "epoch": 2.6077622673434857, + "grad_norm": 8.875, + "learning_rate": 2.44700016986073e-06, + "loss": 2.6472, + "step": 24659 + }, + { + "epoch": 2.6078680203045685, + "grad_norm": 8.5625, + "learning_rate": 2.4468289522511914e-06, + "loss": 2.7622, + "step": 24660 + }, + { + "epoch": 2.6079737732656514, + "grad_norm": 8.625, + "learning_rate": 2.446657734891162e-06, + "loss": 2.5536, + "step": 24661 + }, + { + "epoch": 2.6080795262267342, + "grad_norm": 7.71875, + "learning_rate": 2.4464865177814467e-06, + "loss": 2.4314, + "step": 24662 + }, + { + "epoch": 2.6081852791878175, + "grad_norm": 9.0625, + "learning_rate": 2.4463153009228473e-06, + "loss": 2.8044, + "step": 24663 + }, + { + "epoch": 2.6082910321489, + "grad_norm": 8.375, + "learning_rate": 2.4461440843161686e-06, + "loss": 2.629, + "step": 24664 + }, + { + "epoch": 2.608396785109983, + "grad_norm": 8.8125, + "learning_rate": 2.4459728679622137e-06, + "loss": 2.6284, + "step": 24665 + }, + { + "epoch": 2.608502538071066, + "grad_norm": 10.5, + "learning_rate": 2.4458016518617863e-06, + "loss": 2.681, + "step": 24666 + }, + { + "epoch": 2.608608291032149, + "grad_norm": 8.6875, + "learning_rate": 2.44563043601569e-06, + "loss": 2.3739, + "step": 24667 + }, + { + "epoch": 2.6087140439932317, + "grad_norm": 8.3125, + "learning_rate": 2.4454592204247265e-06, + "loss": 2.4929, + "step": 24668 + }, + { + "epoch": 2.6088197969543145, + "grad_norm": 8.8125, + "learning_rate": 2.445288005089702e-06, + "loss": 2.5197, + "step": 24669 + }, + { + "epoch": 2.608925549915398, + "grad_norm": 9.0, + "learning_rate": 2.4451167900114174e-06, + "loss": 2.572, + "step": 24670 + }, + { + "epoch": 2.6090313028764807, + "grad_norm": 8.1875, + "learning_rate": 2.444945575190678e-06, + "loss": 2.4329, + "step": 24671 + }, + { + "epoch": 2.6091370558375635, + "grad_norm": 9.0, + "learning_rate": 2.444774360628286e-06, + "loss": 2.5871, + "step": 24672 + }, + { + "epoch": 2.6092428087986463, + "grad_norm": 8.5, + "learning_rate": 2.444603146325045e-06, + "loss": 2.5494, + "step": 24673 + }, + { + "epoch": 2.609348561759729, + "grad_norm": 8.75, + "learning_rate": 2.4444319322817596e-06, + "loss": 2.7424, + "step": 24674 + }, + { + "epoch": 2.609454314720812, + "grad_norm": 8.5, + "learning_rate": 2.444260718499232e-06, + "loss": 2.2817, + "step": 24675 + }, + { + "epoch": 2.609560067681895, + "grad_norm": 8.375, + "learning_rate": 2.4440895049782672e-06, + "loss": 2.5642, + "step": 24676 + }, + { + "epoch": 2.609665820642978, + "grad_norm": 8.0, + "learning_rate": 2.443918291719666e-06, + "loss": 2.6628, + "step": 24677 + }, + { + "epoch": 2.609771573604061, + "grad_norm": 9.0625, + "learning_rate": 2.443747078724234e-06, + "loss": 2.4079, + "step": 24678 + }, + { + "epoch": 2.609877326565144, + "grad_norm": 8.9375, + "learning_rate": 2.4435758659927737e-06, + "loss": 2.3849, + "step": 24679 + }, + { + "epoch": 2.6099830795262267, + "grad_norm": 8.5625, + "learning_rate": 2.443404653526089e-06, + "loss": 2.2987, + "step": 24680 + }, + { + "epoch": 2.6100888324873095, + "grad_norm": 8.375, + "learning_rate": 2.4432334413249826e-06, + "loss": 2.4263, + "step": 24681 + }, + { + "epoch": 2.610194585448393, + "grad_norm": 8.3125, + "learning_rate": 2.443062229390258e-06, + "loss": 2.6248, + "step": 24682 + }, + { + "epoch": 2.6103003384094756, + "grad_norm": 9.125, + "learning_rate": 2.442891017722721e-06, + "loss": 2.5651, + "step": 24683 + }, + { + "epoch": 2.6104060913705585, + "grad_norm": 8.375, + "learning_rate": 2.4427198063231714e-06, + "loss": 2.5142, + "step": 24684 + }, + { + "epoch": 2.6105118443316413, + "grad_norm": 8.125, + "learning_rate": 2.4425485951924155e-06, + "loss": 2.5956, + "step": 24685 + }, + { + "epoch": 2.610617597292724, + "grad_norm": 8.4375, + "learning_rate": 2.4423773843312545e-06, + "loss": 2.4054, + "step": 24686 + }, + { + "epoch": 2.610723350253807, + "grad_norm": 8.4375, + "learning_rate": 2.4422061737404935e-06, + "loss": 2.5203, + "step": 24687 + }, + { + "epoch": 2.61082910321489, + "grad_norm": 8.375, + "learning_rate": 2.4420349634209347e-06, + "loss": 2.5733, + "step": 24688 + }, + { + "epoch": 2.610934856175973, + "grad_norm": 8.625, + "learning_rate": 2.4418637533733835e-06, + "loss": 2.5797, + "step": 24689 + }, + { + "epoch": 2.611040609137056, + "grad_norm": 8.0, + "learning_rate": 2.44169254359864e-06, + "loss": 2.4204, + "step": 24690 + }, + { + "epoch": 2.6111463620981388, + "grad_norm": 8.0625, + "learning_rate": 2.4415213340975104e-06, + "loss": 2.534, + "step": 24691 + }, + { + "epoch": 2.6112521150592216, + "grad_norm": 8.3125, + "learning_rate": 2.441350124870798e-06, + "loss": 2.5187, + "step": 24692 + }, + { + "epoch": 2.6113578680203045, + "grad_norm": 8.5625, + "learning_rate": 2.4411789159193046e-06, + "loss": 2.7414, + "step": 24693 + }, + { + "epoch": 2.6114636209813877, + "grad_norm": 8.4375, + "learning_rate": 2.4410077072438346e-06, + "loss": 2.6833, + "step": 24694 + }, + { + "epoch": 2.61156937394247, + "grad_norm": 8.0, + "learning_rate": 2.440836498845191e-06, + "loss": 2.447, + "step": 24695 + }, + { + "epoch": 2.6116751269035534, + "grad_norm": 9.6875, + "learning_rate": 2.4406652907241783e-06, + "loss": 2.6647, + "step": 24696 + }, + { + "epoch": 2.6117808798646363, + "grad_norm": 8.375, + "learning_rate": 2.4404940828815983e-06, + "loss": 2.4948, + "step": 24697 + }, + { + "epoch": 2.611886632825719, + "grad_norm": 8.375, + "learning_rate": 2.4403228753182568e-06, + "loss": 2.4021, + "step": 24698 + }, + { + "epoch": 2.611992385786802, + "grad_norm": 8.875, + "learning_rate": 2.440151668034954e-06, + "loss": 2.8301, + "step": 24699 + }, + { + "epoch": 2.6120981387478848, + "grad_norm": 8.375, + "learning_rate": 2.4399804610324952e-06, + "loss": 2.7212, + "step": 24700 + }, + { + "epoch": 2.612203891708968, + "grad_norm": 9.0, + "learning_rate": 2.4398092543116842e-06, + "loss": 2.5832, + "step": 24701 + }, + { + "epoch": 2.612309644670051, + "grad_norm": 8.5, + "learning_rate": 2.439638047873323e-06, + "loss": 2.6416, + "step": 24702 + }, + { + "epoch": 2.6124153976311337, + "grad_norm": 8.375, + "learning_rate": 2.439466841718216e-06, + "loss": 2.4741, + "step": 24703 + }, + { + "epoch": 2.6125211505922166, + "grad_norm": 8.1875, + "learning_rate": 2.4392956358471664e-06, + "loss": 2.4428, + "step": 24704 + }, + { + "epoch": 2.6126269035532994, + "grad_norm": 8.6875, + "learning_rate": 2.4391244302609784e-06, + "loss": 2.7079, + "step": 24705 + }, + { + "epoch": 2.6127326565143822, + "grad_norm": 8.1875, + "learning_rate": 2.438953224960453e-06, + "loss": 2.4749, + "step": 24706 + }, + { + "epoch": 2.612838409475465, + "grad_norm": 8.0, + "learning_rate": 2.438782019946397e-06, + "loss": 2.4741, + "step": 24707 + }, + { + "epoch": 2.6129441624365484, + "grad_norm": 8.0, + "learning_rate": 2.4386108152196104e-06, + "loss": 2.4148, + "step": 24708 + }, + { + "epoch": 2.613049915397631, + "grad_norm": 8.4375, + "learning_rate": 2.4384396107808984e-06, + "loss": 2.5179, + "step": 24709 + }, + { + "epoch": 2.613155668358714, + "grad_norm": 8.375, + "learning_rate": 2.4382684066310645e-06, + "loss": 2.2993, + "step": 24710 + }, + { + "epoch": 2.613261421319797, + "grad_norm": 8.125, + "learning_rate": 2.438097202770911e-06, + "loss": 2.5391, + "step": 24711 + }, + { + "epoch": 2.6133671742808797, + "grad_norm": 8.5, + "learning_rate": 2.437925999201243e-06, + "loss": 2.4614, + "step": 24712 + }, + { + "epoch": 2.613472927241963, + "grad_norm": 8.25, + "learning_rate": 2.437754795922862e-06, + "loss": 2.5425, + "step": 24713 + }, + { + "epoch": 2.613578680203046, + "grad_norm": 8.125, + "learning_rate": 2.4375835929365736e-06, + "loss": 2.3051, + "step": 24714 + }, + { + "epoch": 2.6136844331641287, + "grad_norm": 8.1875, + "learning_rate": 2.437412390243179e-06, + "loss": 2.6976, + "step": 24715 + }, + { + "epoch": 2.6137901861252115, + "grad_norm": 8.0625, + "learning_rate": 2.4372411878434823e-06, + "loss": 2.4747, + "step": 24716 + }, + { + "epoch": 2.6138959390862944, + "grad_norm": 8.4375, + "learning_rate": 2.4370699857382873e-06, + "loss": 2.6636, + "step": 24717 + }, + { + "epoch": 2.614001692047377, + "grad_norm": 7.78125, + "learning_rate": 2.4368987839283962e-06, + "loss": 2.5248, + "step": 24718 + }, + { + "epoch": 2.61410744500846, + "grad_norm": 7.96875, + "learning_rate": 2.436727582414615e-06, + "loss": 2.5695, + "step": 24719 + }, + { + "epoch": 2.6142131979695433, + "grad_norm": 8.125, + "learning_rate": 2.436556381197744e-06, + "loss": 2.5416, + "step": 24720 + }, + { + "epoch": 2.614318950930626, + "grad_norm": 8.6875, + "learning_rate": 2.436385180278589e-06, + "loss": 2.3892, + "step": 24721 + }, + { + "epoch": 2.614424703891709, + "grad_norm": 8.625, + "learning_rate": 2.436213979657951e-06, + "loss": 2.621, + "step": 24722 + }, + { + "epoch": 2.614530456852792, + "grad_norm": 8.75, + "learning_rate": 2.4360427793366362e-06, + "loss": 2.4955, + "step": 24723 + }, + { + "epoch": 2.6146362098138747, + "grad_norm": 8.375, + "learning_rate": 2.4358715793154454e-06, + "loss": 2.5365, + "step": 24724 + }, + { + "epoch": 2.614741962774958, + "grad_norm": 7.9375, + "learning_rate": 2.435700379595184e-06, + "loss": 2.2825, + "step": 24725 + }, + { + "epoch": 2.6148477157360404, + "grad_norm": 8.8125, + "learning_rate": 2.4355291801766536e-06, + "loss": 2.81, + "step": 24726 + }, + { + "epoch": 2.6149534686971236, + "grad_norm": 9.25, + "learning_rate": 2.435357981060658e-06, + "loss": 2.6982, + "step": 24727 + }, + { + "epoch": 2.6150592216582065, + "grad_norm": 8.1875, + "learning_rate": 2.435186782248002e-06, + "loss": 2.4532, + "step": 24728 + }, + { + "epoch": 2.6151649746192893, + "grad_norm": 8.25, + "learning_rate": 2.435015583739487e-06, + "loss": 2.6572, + "step": 24729 + }, + { + "epoch": 2.615270727580372, + "grad_norm": 9.125, + "learning_rate": 2.434844385535919e-06, + "loss": 2.5762, + "step": 24730 + }, + { + "epoch": 2.615376480541455, + "grad_norm": 9.125, + "learning_rate": 2.434673187638098e-06, + "loss": 2.5419, + "step": 24731 + }, + { + "epoch": 2.6154822335025383, + "grad_norm": 8.3125, + "learning_rate": 2.4345019900468294e-06, + "loss": 2.7549, + "step": 24732 + }, + { + "epoch": 2.615587986463621, + "grad_norm": 8.6875, + "learning_rate": 2.4343307927629163e-06, + "loss": 2.7617, + "step": 24733 + }, + { + "epoch": 2.615693739424704, + "grad_norm": 8.5625, + "learning_rate": 2.434159595787162e-06, + "loss": 2.5358, + "step": 24734 + }, + { + "epoch": 2.615799492385787, + "grad_norm": 8.5, + "learning_rate": 2.433988399120369e-06, + "loss": 2.59, + "step": 24735 + }, + { + "epoch": 2.6159052453468696, + "grad_norm": 8.5, + "learning_rate": 2.433817202763342e-06, + "loss": 2.7515, + "step": 24736 + }, + { + "epoch": 2.6160109983079525, + "grad_norm": 8.25, + "learning_rate": 2.4336460067168847e-06, + "loss": 2.5531, + "step": 24737 + }, + { + "epoch": 2.6161167512690353, + "grad_norm": 8.4375, + "learning_rate": 2.4334748109817984e-06, + "loss": 2.3659, + "step": 24738 + }, + { + "epoch": 2.6162225042301186, + "grad_norm": 8.375, + "learning_rate": 2.4333036155588886e-06, + "loss": 2.4654, + "step": 24739 + }, + { + "epoch": 2.6163282571912014, + "grad_norm": 8.875, + "learning_rate": 2.4331324204489566e-06, + "loss": 2.5219, + "step": 24740 + }, + { + "epoch": 2.6164340101522843, + "grad_norm": 8.1875, + "learning_rate": 2.4329612256528075e-06, + "loss": 2.5027, + "step": 24741 + }, + { + "epoch": 2.616539763113367, + "grad_norm": 7.5, + "learning_rate": 2.432790031171243e-06, + "loss": 2.4843, + "step": 24742 + }, + { + "epoch": 2.61664551607445, + "grad_norm": 8.625, + "learning_rate": 2.432618837005069e-06, + "loss": 2.4013, + "step": 24743 + }, + { + "epoch": 2.6167512690355332, + "grad_norm": 9.625, + "learning_rate": 2.4324476431550857e-06, + "loss": 2.589, + "step": 24744 + }, + { + "epoch": 2.6168570219966156, + "grad_norm": 7.9375, + "learning_rate": 2.432276449622098e-06, + "loss": 2.5423, + "step": 24745 + }, + { + "epoch": 2.616962774957699, + "grad_norm": 8.25, + "learning_rate": 2.432105256406911e-06, + "loss": 2.467, + "step": 24746 + }, + { + "epoch": 2.6170685279187818, + "grad_norm": 8.5625, + "learning_rate": 2.4319340635103246e-06, + "loss": 2.3225, + "step": 24747 + }, + { + "epoch": 2.6171742808798646, + "grad_norm": 8.3125, + "learning_rate": 2.4317628709331444e-06, + "loss": 2.6305, + "step": 24748 + }, + { + "epoch": 2.6172800338409474, + "grad_norm": 8.4375, + "learning_rate": 2.4315916786761727e-06, + "loss": 2.2952, + "step": 24749 + }, + { + "epoch": 2.6173857868020303, + "grad_norm": 7.6875, + "learning_rate": 2.4314204867402137e-06, + "loss": 2.2206, + "step": 24750 + }, + { + "epoch": 2.6174915397631136, + "grad_norm": 8.5, + "learning_rate": 2.4312492951260697e-06, + "loss": 2.7, + "step": 24751 + }, + { + "epoch": 2.6175972927241964, + "grad_norm": 8.6875, + "learning_rate": 2.4310781038345453e-06, + "loss": 2.6637, + "step": 24752 + }, + { + "epoch": 2.6177030456852792, + "grad_norm": 8.5, + "learning_rate": 2.4309069128664435e-06, + "loss": 2.4818, + "step": 24753 + }, + { + "epoch": 2.617808798646362, + "grad_norm": 8.375, + "learning_rate": 2.4307357222225664e-06, + "loss": 2.5195, + "step": 24754 + }, + { + "epoch": 2.617914551607445, + "grad_norm": 8.6875, + "learning_rate": 2.430564531903719e-06, + "loss": 2.6338, + "step": 24755 + }, + { + "epoch": 2.618020304568528, + "grad_norm": 8.4375, + "learning_rate": 2.4303933419107035e-06, + "loss": 2.648, + "step": 24756 + }, + { + "epoch": 2.6181260575296106, + "grad_norm": 8.8125, + "learning_rate": 2.4302221522443236e-06, + "loss": 2.6672, + "step": 24757 + }, + { + "epoch": 2.618231810490694, + "grad_norm": 8.5, + "learning_rate": 2.430050962905382e-06, + "loss": 2.5893, + "step": 24758 + }, + { + "epoch": 2.6183375634517767, + "grad_norm": 8.4375, + "learning_rate": 2.4298797738946845e-06, + "loss": 2.7602, + "step": 24759 + }, + { + "epoch": 2.6184433164128595, + "grad_norm": 8.4375, + "learning_rate": 2.4297085852130304e-06, + "loss": 2.5944, + "step": 24760 + }, + { + "epoch": 2.6185490693739424, + "grad_norm": 8.3125, + "learning_rate": 2.4295373968612253e-06, + "loss": 2.2427, + "step": 24761 + }, + { + "epoch": 2.6186548223350252, + "grad_norm": 9.0, + "learning_rate": 2.429366208840074e-06, + "loss": 2.5552, + "step": 24762 + }, + { + "epoch": 2.6187605752961085, + "grad_norm": 8.375, + "learning_rate": 2.429195021150377e-06, + "loss": 2.2081, + "step": 24763 + }, + { + "epoch": 2.6188663282571913, + "grad_norm": 9.25, + "learning_rate": 2.4290238337929396e-06, + "loss": 2.6057, + "step": 24764 + }, + { + "epoch": 2.618972081218274, + "grad_norm": 8.4375, + "learning_rate": 2.4288526467685635e-06, + "loss": 2.6083, + "step": 24765 + }, + { + "epoch": 2.619077834179357, + "grad_norm": 8.3125, + "learning_rate": 2.428681460078053e-06, + "loss": 2.4671, + "step": 24766 + }, + { + "epoch": 2.61918358714044, + "grad_norm": 8.25, + "learning_rate": 2.4285102737222112e-06, + "loss": 2.5566, + "step": 24767 + }, + { + "epoch": 2.6192893401015227, + "grad_norm": 9.125, + "learning_rate": 2.4283390877018423e-06, + "loss": 2.6858, + "step": 24768 + }, + { + "epoch": 2.6193950930626055, + "grad_norm": 8.125, + "learning_rate": 2.428167902017748e-06, + "loss": 2.6441, + "step": 24769 + }, + { + "epoch": 2.619500846023689, + "grad_norm": 8.625, + "learning_rate": 2.4279967166707318e-06, + "loss": 2.6223, + "step": 24770 + }, + { + "epoch": 2.6196065989847717, + "grad_norm": 8.9375, + "learning_rate": 2.4278255316615986e-06, + "loss": 2.4487, + "step": 24771 + }, + { + "epoch": 2.6197123519458545, + "grad_norm": 8.5625, + "learning_rate": 2.42765434699115e-06, + "loss": 2.6683, + "step": 24772 + }, + { + "epoch": 2.6198181049069373, + "grad_norm": 8.5, + "learning_rate": 2.4274831626601904e-06, + "loss": 2.1376, + "step": 24773 + }, + { + "epoch": 2.61992385786802, + "grad_norm": 8.4375, + "learning_rate": 2.427311978669522e-06, + "loss": 2.5773, + "step": 24774 + }, + { + "epoch": 2.6200296108291035, + "grad_norm": 8.4375, + "learning_rate": 2.42714079501995e-06, + "loss": 2.3454, + "step": 24775 + }, + { + "epoch": 2.620135363790186, + "grad_norm": 8.125, + "learning_rate": 2.4269696117122753e-06, + "loss": 2.5648, + "step": 24776 + }, + { + "epoch": 2.620241116751269, + "grad_norm": 8.3125, + "learning_rate": 2.4267984287473033e-06, + "loss": 2.3105, + "step": 24777 + }, + { + "epoch": 2.620346869712352, + "grad_norm": 8.25, + "learning_rate": 2.4266272461258354e-06, + "loss": 2.6147, + "step": 24778 + }, + { + "epoch": 2.620452622673435, + "grad_norm": 8.5625, + "learning_rate": 2.4264560638486757e-06, + "loss": 2.5871, + "step": 24779 + }, + { + "epoch": 2.6205583756345177, + "grad_norm": 8.6875, + "learning_rate": 2.4262848819166286e-06, + "loss": 2.5208, + "step": 24780 + }, + { + "epoch": 2.6206641285956005, + "grad_norm": 9.3125, + "learning_rate": 2.4261137003304954e-06, + "loss": 2.6892, + "step": 24781 + }, + { + "epoch": 2.620769881556684, + "grad_norm": 8.25, + "learning_rate": 2.425942519091081e-06, + "loss": 2.4789, + "step": 24782 + }, + { + "epoch": 2.6208756345177666, + "grad_norm": 8.625, + "learning_rate": 2.425771338199188e-06, + "loss": 2.5357, + "step": 24783 + }, + { + "epoch": 2.6209813874788495, + "grad_norm": 8.125, + "learning_rate": 2.4256001576556206e-06, + "loss": 2.5549, + "step": 24784 + }, + { + "epoch": 2.6210871404399323, + "grad_norm": 8.875, + "learning_rate": 2.42542897746118e-06, + "loss": 2.5528, + "step": 24785 + }, + { + "epoch": 2.621192893401015, + "grad_norm": 8.375, + "learning_rate": 2.425257797616671e-06, + "loss": 2.5841, + "step": 24786 + }, + { + "epoch": 2.621298646362098, + "grad_norm": 8.1875, + "learning_rate": 2.4250866181228963e-06, + "loss": 2.4057, + "step": 24787 + }, + { + "epoch": 2.621404399323181, + "grad_norm": 9.0, + "learning_rate": 2.4249154389806593e-06, + "loss": 2.5541, + "step": 24788 + }, + { + "epoch": 2.621510152284264, + "grad_norm": 8.625, + "learning_rate": 2.4247442601907644e-06, + "loss": 2.5612, + "step": 24789 + }, + { + "epoch": 2.621615905245347, + "grad_norm": 8.25, + "learning_rate": 2.424573081754013e-06, + "loss": 2.0513, + "step": 24790 + }, + { + "epoch": 2.6217216582064298, + "grad_norm": 8.4375, + "learning_rate": 2.4244019036712108e-06, + "loss": 2.5214, + "step": 24791 + }, + { + "epoch": 2.6218274111675126, + "grad_norm": 8.25, + "learning_rate": 2.424230725943158e-06, + "loss": 2.5855, + "step": 24792 + }, + { + "epoch": 2.6219331641285955, + "grad_norm": 8.1875, + "learning_rate": 2.42405954857066e-06, + "loss": 2.4479, + "step": 24793 + }, + { + "epoch": 2.6220389170896787, + "grad_norm": 9.0625, + "learning_rate": 2.423888371554519e-06, + "loss": 2.4686, + "step": 24794 + }, + { + "epoch": 2.6221446700507616, + "grad_norm": 8.5625, + "learning_rate": 2.4237171948955397e-06, + "loss": 2.6489, + "step": 24795 + }, + { + "epoch": 2.6222504230118444, + "grad_norm": 8.5625, + "learning_rate": 2.4235460185945236e-06, + "loss": 2.5053, + "step": 24796 + }, + { + "epoch": 2.6223561759729273, + "grad_norm": 8.5625, + "learning_rate": 2.423374842652274e-06, + "loss": 2.5817, + "step": 24797 + }, + { + "epoch": 2.62246192893401, + "grad_norm": 8.125, + "learning_rate": 2.4232036670695967e-06, + "loss": 2.4863, + "step": 24798 + }, + { + "epoch": 2.622567681895093, + "grad_norm": 8.1875, + "learning_rate": 2.423032491847292e-06, + "loss": 2.4926, + "step": 24799 + }, + { + "epoch": 2.6226734348561758, + "grad_norm": 8.4375, + "learning_rate": 2.4228613169861653e-06, + "loss": 2.5447, + "step": 24800 + }, + { + "epoch": 2.622779187817259, + "grad_norm": 8.5625, + "learning_rate": 2.4226901424870183e-06, + "loss": 2.6234, + "step": 24801 + }, + { + "epoch": 2.622884940778342, + "grad_norm": 8.625, + "learning_rate": 2.4225189683506546e-06, + "loss": 2.5304, + "step": 24802 + }, + { + "epoch": 2.6229906937394247, + "grad_norm": 8.4375, + "learning_rate": 2.4223477945778777e-06, + "loss": 2.435, + "step": 24803 + }, + { + "epoch": 2.6230964467005076, + "grad_norm": 8.1875, + "learning_rate": 2.422176621169492e-06, + "loss": 2.4517, + "step": 24804 + }, + { + "epoch": 2.6232021996615904, + "grad_norm": 8.5625, + "learning_rate": 2.422005448126298e-06, + "loss": 2.7195, + "step": 24805 + }, + { + "epoch": 2.6233079526226737, + "grad_norm": 8.6875, + "learning_rate": 2.421834275449101e-06, + "loss": 2.4876, + "step": 24806 + }, + { + "epoch": 2.623413705583756, + "grad_norm": 8.3125, + "learning_rate": 2.4216631031387043e-06, + "loss": 2.6219, + "step": 24807 + }, + { + "epoch": 2.6235194585448394, + "grad_norm": 8.4375, + "learning_rate": 2.4214919311959103e-06, + "loss": 2.6218, + "step": 24808 + }, + { + "epoch": 2.623625211505922, + "grad_norm": 8.3125, + "learning_rate": 2.4213207596215227e-06, + "loss": 2.6499, + "step": 24809 + }, + { + "epoch": 2.623730964467005, + "grad_norm": 8.3125, + "learning_rate": 2.421149588416344e-06, + "loss": 2.5175, + "step": 24810 + }, + { + "epoch": 2.623836717428088, + "grad_norm": 9.0625, + "learning_rate": 2.420978417581179e-06, + "loss": 2.563, + "step": 24811 + }, + { + "epoch": 2.6239424703891707, + "grad_norm": 8.5, + "learning_rate": 2.420807247116829e-06, + "loss": 2.731, + "step": 24812 + }, + { + "epoch": 2.624048223350254, + "grad_norm": 8.3125, + "learning_rate": 2.4206360770241e-06, + "loss": 2.4036, + "step": 24813 + }, + { + "epoch": 2.624153976311337, + "grad_norm": 8.125, + "learning_rate": 2.4204649073037907e-06, + "loss": 2.4354, + "step": 24814 + }, + { + "epoch": 2.6242597292724197, + "grad_norm": 8.625, + "learning_rate": 2.420293737956708e-06, + "loss": 2.5897, + "step": 24815 + }, + { + "epoch": 2.6243654822335025, + "grad_norm": 8.3125, + "learning_rate": 2.4201225689836556e-06, + "loss": 2.608, + "step": 24816 + }, + { + "epoch": 2.6244712351945854, + "grad_norm": 7.90625, + "learning_rate": 2.419951400385434e-06, + "loss": 2.3774, + "step": 24817 + }, + { + "epoch": 2.624576988155668, + "grad_norm": 8.25, + "learning_rate": 2.4197802321628484e-06, + "loss": 2.5339, + "step": 24818 + }, + { + "epoch": 2.624682741116751, + "grad_norm": 8.25, + "learning_rate": 2.419609064316701e-06, + "loss": 2.5282, + "step": 24819 + }, + { + "epoch": 2.6247884940778343, + "grad_norm": 8.8125, + "learning_rate": 2.4194378968477956e-06, + "loss": 2.7701, + "step": 24820 + }, + { + "epoch": 2.624894247038917, + "grad_norm": 8.5625, + "learning_rate": 2.4192667297569347e-06, + "loss": 2.5526, + "step": 24821 + }, + { + "epoch": 2.625, + "grad_norm": 8.5, + "learning_rate": 2.419095563044923e-06, + "loss": 2.4773, + "step": 24822 + }, + { + "epoch": 2.625105752961083, + "grad_norm": 9.125, + "learning_rate": 2.418924396712562e-06, + "loss": 2.5844, + "step": 24823 + }, + { + "epoch": 2.6252115059221657, + "grad_norm": 8.375, + "learning_rate": 2.4187532307606555e-06, + "loss": 2.3955, + "step": 24824 + }, + { + "epoch": 2.625317258883249, + "grad_norm": 8.4375, + "learning_rate": 2.4185820651900074e-06, + "loss": 2.5134, + "step": 24825 + }, + { + "epoch": 2.625423011844332, + "grad_norm": 8.375, + "learning_rate": 2.4184109000014194e-06, + "loss": 2.3325, + "step": 24826 + }, + { + "epoch": 2.6255287648054146, + "grad_norm": 8.125, + "learning_rate": 2.4182397351956966e-06, + "loss": 2.3141, + "step": 24827 + }, + { + "epoch": 2.6256345177664975, + "grad_norm": 9.0625, + "learning_rate": 2.418068570773641e-06, + "loss": 2.5278, + "step": 24828 + }, + { + "epoch": 2.6257402707275803, + "grad_norm": 8.3125, + "learning_rate": 2.4178974067360567e-06, + "loss": 2.5214, + "step": 24829 + }, + { + "epoch": 2.625846023688663, + "grad_norm": 8.5, + "learning_rate": 2.4177262430837455e-06, + "loss": 2.3668, + "step": 24830 + }, + { + "epoch": 2.625951776649746, + "grad_norm": 8.375, + "learning_rate": 2.417555079817512e-06, + "loss": 2.3035, + "step": 24831 + }, + { + "epoch": 2.6260575296108293, + "grad_norm": 8.4375, + "learning_rate": 2.4173839169381576e-06, + "loss": 2.4758, + "step": 24832 + }, + { + "epoch": 2.626163282571912, + "grad_norm": 8.75, + "learning_rate": 2.4172127544464874e-06, + "loss": 2.465, + "step": 24833 + }, + { + "epoch": 2.626269035532995, + "grad_norm": 8.125, + "learning_rate": 2.417041592343304e-06, + "loss": 2.5709, + "step": 24834 + }, + { + "epoch": 2.626374788494078, + "grad_norm": 8.625, + "learning_rate": 2.41687043062941e-06, + "loss": 2.5531, + "step": 24835 + }, + { + "epoch": 2.6264805414551606, + "grad_norm": 8.1875, + "learning_rate": 2.41669926930561e-06, + "loss": 2.4641, + "step": 24836 + }, + { + "epoch": 2.626586294416244, + "grad_norm": 8.875, + "learning_rate": 2.4165281083727055e-06, + "loss": 2.5359, + "step": 24837 + }, + { + "epoch": 2.6266920473773263, + "grad_norm": 7.875, + "learning_rate": 2.4163569478315014e-06, + "loss": 2.3282, + "step": 24838 + }, + { + "epoch": 2.6267978003384096, + "grad_norm": 8.9375, + "learning_rate": 2.4161857876827992e-06, + "loss": 2.5188, + "step": 24839 + }, + { + "epoch": 2.6269035532994924, + "grad_norm": 8.6875, + "learning_rate": 2.416014627927403e-06, + "loss": 2.7134, + "step": 24840 + }, + { + "epoch": 2.6270093062605753, + "grad_norm": 8.25, + "learning_rate": 2.4158434685661153e-06, + "loss": 2.4232, + "step": 24841 + }, + { + "epoch": 2.627115059221658, + "grad_norm": 8.3125, + "learning_rate": 2.4156723095997402e-06, + "loss": 2.5715, + "step": 24842 + }, + { + "epoch": 2.627220812182741, + "grad_norm": 9.0, + "learning_rate": 2.4155011510290804e-06, + "loss": 2.6005, + "step": 24843 + }, + { + "epoch": 2.6273265651438242, + "grad_norm": 8.6875, + "learning_rate": 2.415329992854939e-06, + "loss": 2.6193, + "step": 24844 + }, + { + "epoch": 2.627432318104907, + "grad_norm": 8.6875, + "learning_rate": 2.41515883507812e-06, + "loss": 2.586, + "step": 24845 + }, + { + "epoch": 2.62753807106599, + "grad_norm": 8.4375, + "learning_rate": 2.414987677699425e-06, + "loss": 2.6727, + "step": 24846 + }, + { + "epoch": 2.6276438240270727, + "grad_norm": 8.3125, + "learning_rate": 2.414816520719659e-06, + "loss": 2.5998, + "step": 24847 + }, + { + "epoch": 2.6277495769881556, + "grad_norm": 8.625, + "learning_rate": 2.414645364139623e-06, + "loss": 2.2224, + "step": 24848 + }, + { + "epoch": 2.6278553299492384, + "grad_norm": 7.5625, + "learning_rate": 2.4144742079601225e-06, + "loss": 2.3494, + "step": 24849 + }, + { + "epoch": 2.6279610829103213, + "grad_norm": 8.3125, + "learning_rate": 2.4143030521819587e-06, + "loss": 2.4917, + "step": 24850 + }, + { + "epoch": 2.6280668358714045, + "grad_norm": 8.6875, + "learning_rate": 2.414131896805936e-06, + "loss": 2.4832, + "step": 24851 + }, + { + "epoch": 2.6281725888324874, + "grad_norm": 8.5, + "learning_rate": 2.4139607418328573e-06, + "loss": 2.5095, + "step": 24852 + }, + { + "epoch": 2.6282783417935702, + "grad_norm": 13.0, + "learning_rate": 2.413789587263525e-06, + "loss": 2.3771, + "step": 24853 + }, + { + "epoch": 2.628384094754653, + "grad_norm": 8.5, + "learning_rate": 2.413618433098745e-06, + "loss": 2.5434, + "step": 24854 + }, + { + "epoch": 2.628489847715736, + "grad_norm": 8.0625, + "learning_rate": 2.4134472793393165e-06, + "loss": 2.0865, + "step": 24855 + }, + { + "epoch": 2.628595600676819, + "grad_norm": 8.8125, + "learning_rate": 2.413276125986045e-06, + "loss": 2.4828, + "step": 24856 + }, + { + "epoch": 2.628701353637902, + "grad_norm": 8.5625, + "learning_rate": 2.4131049730397326e-06, + "loss": 2.6176, + "step": 24857 + }, + { + "epoch": 2.628807106598985, + "grad_norm": 8.3125, + "learning_rate": 2.4129338205011836e-06, + "loss": 2.5311, + "step": 24858 + }, + { + "epoch": 2.6289128595600677, + "grad_norm": 8.375, + "learning_rate": 2.4127626683712e-06, + "loss": 2.4284, + "step": 24859 + }, + { + "epoch": 2.6290186125211505, + "grad_norm": 8.6875, + "learning_rate": 2.412591516650586e-06, + "loss": 2.8635, + "step": 24860 + }, + { + "epoch": 2.6291243654822334, + "grad_norm": 8.0, + "learning_rate": 2.4124203653401448e-06, + "loss": 2.3484, + "step": 24861 + }, + { + "epoch": 2.629230118443316, + "grad_norm": 14.8125, + "learning_rate": 2.412249214440678e-06, + "loss": 2.4197, + "step": 24862 + }, + { + "epoch": 2.6293358714043995, + "grad_norm": 7.84375, + "learning_rate": 2.4120780639529907e-06, + "loss": 2.4163, + "step": 24863 + }, + { + "epoch": 2.6294416243654823, + "grad_norm": 7.90625, + "learning_rate": 2.411906913877884e-06, + "loss": 2.5544, + "step": 24864 + }, + { + "epoch": 2.629547377326565, + "grad_norm": 8.125, + "learning_rate": 2.411735764216163e-06, + "loss": 2.2859, + "step": 24865 + }, + { + "epoch": 2.629653130287648, + "grad_norm": 8.0, + "learning_rate": 2.4115646149686293e-06, + "loss": 2.5502, + "step": 24866 + }, + { + "epoch": 2.629758883248731, + "grad_norm": 8.25, + "learning_rate": 2.411393466136087e-06, + "loss": 2.5349, + "step": 24867 + }, + { + "epoch": 2.629864636209814, + "grad_norm": 8.125, + "learning_rate": 2.411222317719339e-06, + "loss": 2.4335, + "step": 24868 + }, + { + "epoch": 2.6299703891708965, + "grad_norm": 8.125, + "learning_rate": 2.411051169719188e-06, + "loss": 2.4967, + "step": 24869 + }, + { + "epoch": 2.63007614213198, + "grad_norm": 8.1875, + "learning_rate": 2.4108800221364386e-06, + "loss": 2.6422, + "step": 24870 + }, + { + "epoch": 2.6301818950930627, + "grad_norm": 8.5, + "learning_rate": 2.410708874971892e-06, + "loss": 2.5196, + "step": 24871 + }, + { + "epoch": 2.6302876480541455, + "grad_norm": 8.25, + "learning_rate": 2.4105377282263523e-06, + "loss": 2.6615, + "step": 24872 + }, + { + "epoch": 2.6303934010152283, + "grad_norm": 9.1875, + "learning_rate": 2.410366581900622e-06, + "loss": 2.4772, + "step": 24873 + }, + { + "epoch": 2.630499153976311, + "grad_norm": 8.6875, + "learning_rate": 2.410195435995505e-06, + "loss": 2.432, + "step": 24874 + }, + { + "epoch": 2.6306049069373945, + "grad_norm": 9.0625, + "learning_rate": 2.4100242905118036e-06, + "loss": 2.4298, + "step": 24875 + }, + { + "epoch": 2.6307106598984773, + "grad_norm": 8.6875, + "learning_rate": 2.409853145450322e-06, + "loss": 2.5266, + "step": 24876 + }, + { + "epoch": 2.63081641285956, + "grad_norm": 8.4375, + "learning_rate": 2.409682000811863e-06, + "loss": 2.5224, + "step": 24877 + }, + { + "epoch": 2.630922165820643, + "grad_norm": 9.125, + "learning_rate": 2.409510856597229e-06, + "loss": 2.4449, + "step": 24878 + }, + { + "epoch": 2.631027918781726, + "grad_norm": 8.4375, + "learning_rate": 2.4093397128072237e-06, + "loss": 2.3559, + "step": 24879 + }, + { + "epoch": 2.6311336717428087, + "grad_norm": 8.9375, + "learning_rate": 2.4091685694426496e-06, + "loss": 2.612, + "step": 24880 + }, + { + "epoch": 2.6312394247038915, + "grad_norm": 8.6875, + "learning_rate": 2.408997426504311e-06, + "loss": 2.4526, + "step": 24881 + }, + { + "epoch": 2.6313451776649748, + "grad_norm": 9.3125, + "learning_rate": 2.408826283993009e-06, + "loss": 2.2723, + "step": 24882 + }, + { + "epoch": 2.6314509306260576, + "grad_norm": 8.0, + "learning_rate": 2.40865514190955e-06, + "loss": 2.546, + "step": 24883 + }, + { + "epoch": 2.6315566835871405, + "grad_norm": 8.125, + "learning_rate": 2.408484000254733e-06, + "loss": 2.4436, + "step": 24884 + }, + { + "epoch": 2.6316624365482233, + "grad_norm": 8.9375, + "learning_rate": 2.4083128590293637e-06, + "loss": 2.6357, + "step": 24885 + }, + { + "epoch": 2.631768189509306, + "grad_norm": 8.3125, + "learning_rate": 2.4081417182342455e-06, + "loss": 2.5612, + "step": 24886 + }, + { + "epoch": 2.6318739424703894, + "grad_norm": 8.4375, + "learning_rate": 2.4079705778701798e-06, + "loss": 2.4745, + "step": 24887 + }, + { + "epoch": 2.631979695431472, + "grad_norm": 8.9375, + "learning_rate": 2.407799437937971e-06, + "loss": 2.8127, + "step": 24888 + }, + { + "epoch": 2.632085448392555, + "grad_norm": 8.25, + "learning_rate": 2.4076282984384213e-06, + "loss": 2.5544, + "step": 24889 + }, + { + "epoch": 2.632191201353638, + "grad_norm": 8.5, + "learning_rate": 2.407457159372335e-06, + "loss": 2.5879, + "step": 24890 + }, + { + "epoch": 2.6322969543147208, + "grad_norm": 8.875, + "learning_rate": 2.4072860207405134e-06, + "loss": 2.5655, + "step": 24891 + }, + { + "epoch": 2.6324027072758036, + "grad_norm": 8.5, + "learning_rate": 2.407114882543762e-06, + "loss": 2.5616, + "step": 24892 + }, + { + "epoch": 2.6325084602368864, + "grad_norm": 8.4375, + "learning_rate": 2.4069437447828813e-06, + "loss": 2.6249, + "step": 24893 + }, + { + "epoch": 2.6326142131979697, + "grad_norm": 8.5, + "learning_rate": 2.4067726074586754e-06, + "loss": 2.4787, + "step": 24894 + }, + { + "epoch": 2.6327199661590526, + "grad_norm": 9.3125, + "learning_rate": 2.406601470571948e-06, + "loss": 2.7359, + "step": 24895 + }, + { + "epoch": 2.6328257191201354, + "grad_norm": 9.0625, + "learning_rate": 2.4064303341235013e-06, + "loss": 2.4319, + "step": 24896 + }, + { + "epoch": 2.6329314720812182, + "grad_norm": 7.96875, + "learning_rate": 2.406259198114139e-06, + "loss": 2.4353, + "step": 24897 + }, + { + "epoch": 2.633037225042301, + "grad_norm": 8.6875, + "learning_rate": 2.406088062544664e-06, + "loss": 2.5343, + "step": 24898 + }, + { + "epoch": 2.6331429780033844, + "grad_norm": 9.8125, + "learning_rate": 2.4059169274158803e-06, + "loss": 2.4428, + "step": 24899 + }, + { + "epoch": 2.6332487309644668, + "grad_norm": 8.4375, + "learning_rate": 2.4057457927285883e-06, + "loss": 2.3682, + "step": 24900 + }, + { + "epoch": 2.63335448392555, + "grad_norm": 8.5, + "learning_rate": 2.405574658483594e-06, + "loss": 2.3998, + "step": 24901 + }, + { + "epoch": 2.633460236886633, + "grad_norm": 8.5, + "learning_rate": 2.4054035246816984e-06, + "loss": 2.5525, + "step": 24902 + }, + { + "epoch": 2.6335659898477157, + "grad_norm": 8.875, + "learning_rate": 2.4052323913237052e-06, + "loss": 2.6263, + "step": 24903 + }, + { + "epoch": 2.6336717428087986, + "grad_norm": 8.0625, + "learning_rate": 2.4050612584104187e-06, + "loss": 2.5647, + "step": 24904 + }, + { + "epoch": 2.6337774957698814, + "grad_norm": 8.4375, + "learning_rate": 2.40489012594264e-06, + "loss": 2.6063, + "step": 24905 + }, + { + "epoch": 2.6338832487309647, + "grad_norm": 9.0, + "learning_rate": 2.404718993921173e-06, + "loss": 2.5466, + "step": 24906 + }, + { + "epoch": 2.6339890016920475, + "grad_norm": 8.8125, + "learning_rate": 2.4045478623468212e-06, + "loss": 2.5226, + "step": 24907 + }, + { + "epoch": 2.6340947546531304, + "grad_norm": 8.25, + "learning_rate": 2.4043767312203876e-06, + "loss": 2.6303, + "step": 24908 + }, + { + "epoch": 2.634200507614213, + "grad_norm": 8.4375, + "learning_rate": 2.4042056005426743e-06, + "loss": 2.8044, + "step": 24909 + }, + { + "epoch": 2.634306260575296, + "grad_norm": 8.8125, + "learning_rate": 2.404034470314485e-06, + "loss": 2.7823, + "step": 24910 + }, + { + "epoch": 2.634412013536379, + "grad_norm": 8.625, + "learning_rate": 2.403863340536623e-06, + "loss": 2.7159, + "step": 24911 + }, + { + "epoch": 2.6345177664974617, + "grad_norm": 8.9375, + "learning_rate": 2.4036922112098903e-06, + "loss": 2.5536, + "step": 24912 + }, + { + "epoch": 2.634623519458545, + "grad_norm": 8.4375, + "learning_rate": 2.4035210823350913e-06, + "loss": 2.4438, + "step": 24913 + }, + { + "epoch": 2.634729272419628, + "grad_norm": 8.3125, + "learning_rate": 2.403349953913028e-06, + "loss": 2.3992, + "step": 24914 + }, + { + "epoch": 2.6348350253807107, + "grad_norm": 7.96875, + "learning_rate": 2.4031788259445047e-06, + "loss": 2.5091, + "step": 24915 + }, + { + "epoch": 2.6349407783417935, + "grad_norm": 8.1875, + "learning_rate": 2.4030076984303227e-06, + "loss": 2.4662, + "step": 24916 + }, + { + "epoch": 2.6350465313028764, + "grad_norm": 8.8125, + "learning_rate": 2.4028365713712866e-06, + "loss": 2.323, + "step": 24917 + }, + { + "epoch": 2.6351522842639596, + "grad_norm": 8.8125, + "learning_rate": 2.402665444768198e-06, + "loss": 2.7817, + "step": 24918 + }, + { + "epoch": 2.635258037225042, + "grad_norm": 8.5, + "learning_rate": 2.402494318621861e-06, + "loss": 2.334, + "step": 24919 + }, + { + "epoch": 2.6353637901861253, + "grad_norm": 8.375, + "learning_rate": 2.402323192933078e-06, + "loss": 2.699, + "step": 24920 + }, + { + "epoch": 2.635469543147208, + "grad_norm": 8.5, + "learning_rate": 2.4021520677026523e-06, + "loss": 2.6288, + "step": 24921 + }, + { + "epoch": 2.635575296108291, + "grad_norm": 8.5, + "learning_rate": 2.401980942931388e-06, + "loss": 2.256, + "step": 24922 + }, + { + "epoch": 2.635681049069374, + "grad_norm": 8.625, + "learning_rate": 2.401809818620085e-06, + "loss": 2.4365, + "step": 24923 + }, + { + "epoch": 2.6357868020304567, + "grad_norm": 8.4375, + "learning_rate": 2.4016386947695506e-06, + "loss": 2.5224, + "step": 24924 + }, + { + "epoch": 2.63589255499154, + "grad_norm": 8.5, + "learning_rate": 2.401467571380584e-06, + "loss": 2.429, + "step": 24925 + }, + { + "epoch": 2.635998307952623, + "grad_norm": 8.625, + "learning_rate": 2.4012964484539906e-06, + "loss": 2.5195, + "step": 24926 + }, + { + "epoch": 2.6361040609137056, + "grad_norm": 8.3125, + "learning_rate": 2.401125325990572e-06, + "loss": 2.6255, + "step": 24927 + }, + { + "epoch": 2.6362098138747885, + "grad_norm": 8.5625, + "learning_rate": 2.4009542039911327e-06, + "loss": 2.6847, + "step": 24928 + }, + { + "epoch": 2.6363155668358713, + "grad_norm": 8.75, + "learning_rate": 2.400783082456474e-06, + "loss": 2.4478, + "step": 24929 + }, + { + "epoch": 2.636421319796954, + "grad_norm": 8.9375, + "learning_rate": 2.4006119613873994e-06, + "loss": 2.7218, + "step": 24930 + }, + { + "epoch": 2.636527072758037, + "grad_norm": 8.25, + "learning_rate": 2.4004408407847136e-06, + "loss": 2.0751, + "step": 24931 + }, + { + "epoch": 2.6366328257191203, + "grad_norm": 8.5625, + "learning_rate": 2.400269720649217e-06, + "loss": 2.4241, + "step": 24932 + }, + { + "epoch": 2.636738578680203, + "grad_norm": 8.1875, + "learning_rate": 2.4000986009817147e-06, + "loss": 2.5341, + "step": 24933 + }, + { + "epoch": 2.636844331641286, + "grad_norm": 8.3125, + "learning_rate": 2.399927481783008e-06, + "loss": 2.4306, + "step": 24934 + }, + { + "epoch": 2.636950084602369, + "grad_norm": 8.25, + "learning_rate": 2.399756363053901e-06, + "loss": 2.5412, + "step": 24935 + }, + { + "epoch": 2.6370558375634516, + "grad_norm": 8.4375, + "learning_rate": 2.3995852447951963e-06, + "loss": 2.5216, + "step": 24936 + }, + { + "epoch": 2.637161590524535, + "grad_norm": 8.375, + "learning_rate": 2.3994141270076978e-06, + "loss": 2.4828, + "step": 24937 + }, + { + "epoch": 2.6372673434856178, + "grad_norm": 8.5, + "learning_rate": 2.3992430096922066e-06, + "loss": 2.6454, + "step": 24938 + }, + { + "epoch": 2.6373730964467006, + "grad_norm": 8.125, + "learning_rate": 2.3990718928495263e-06, + "loss": 2.6006, + "step": 24939 + }, + { + "epoch": 2.6374788494077834, + "grad_norm": 8.8125, + "learning_rate": 2.398900776480462e-06, + "loss": 2.6702, + "step": 24940 + }, + { + "epoch": 2.6375846023688663, + "grad_norm": 8.375, + "learning_rate": 2.398729660585814e-06, + "loss": 2.4525, + "step": 24941 + }, + { + "epoch": 2.637690355329949, + "grad_norm": 8.5, + "learning_rate": 2.3985585451663866e-06, + "loss": 2.5806, + "step": 24942 + }, + { + "epoch": 2.637796108291032, + "grad_norm": 8.4375, + "learning_rate": 2.398387430222982e-06, + "loss": 2.2087, + "step": 24943 + }, + { + "epoch": 2.6379018612521152, + "grad_norm": 10.375, + "learning_rate": 2.398216315756404e-06, + "loss": 2.7402, + "step": 24944 + }, + { + "epoch": 2.638007614213198, + "grad_norm": 8.4375, + "learning_rate": 2.3980452017674547e-06, + "loss": 2.779, + "step": 24945 + }, + { + "epoch": 2.638113367174281, + "grad_norm": 8.5625, + "learning_rate": 2.397874088256939e-06, + "loss": 2.4851, + "step": 24946 + }, + { + "epoch": 2.6382191201353637, + "grad_norm": 8.5625, + "learning_rate": 2.397702975225657e-06, + "loss": 2.3806, + "step": 24947 + }, + { + "epoch": 2.6383248730964466, + "grad_norm": 8.5, + "learning_rate": 2.3975318626744132e-06, + "loss": 2.5602, + "step": 24948 + }, + { + "epoch": 2.63843062605753, + "grad_norm": 8.6875, + "learning_rate": 2.397360750604011e-06, + "loss": 2.5266, + "step": 24949 + }, + { + "epoch": 2.6385363790186123, + "grad_norm": 8.5, + "learning_rate": 2.3971896390152524e-06, + "loss": 2.5407, + "step": 24950 + }, + { + "epoch": 2.6386421319796955, + "grad_norm": 8.6875, + "learning_rate": 2.397018527908941e-06, + "loss": 2.6055, + "step": 24951 + }, + { + "epoch": 2.6387478849407784, + "grad_norm": 8.4375, + "learning_rate": 2.3968474172858796e-06, + "loss": 2.4007, + "step": 24952 + }, + { + "epoch": 2.6388536379018612, + "grad_norm": 8.25, + "learning_rate": 2.3966763071468717e-06, + "loss": 2.6009, + "step": 24953 + }, + { + "epoch": 2.638959390862944, + "grad_norm": 8.75, + "learning_rate": 2.396505197492718e-06, + "loss": 2.5995, + "step": 24954 + }, + { + "epoch": 2.639065143824027, + "grad_norm": 8.625, + "learning_rate": 2.396334088324225e-06, + "loss": 2.614, + "step": 24955 + }, + { + "epoch": 2.63917089678511, + "grad_norm": 7.84375, + "learning_rate": 2.396162979642192e-06, + "loss": 2.4431, + "step": 24956 + }, + { + "epoch": 2.639276649746193, + "grad_norm": 9.125, + "learning_rate": 2.3959918714474243e-06, + "loss": 2.4854, + "step": 24957 + }, + { + "epoch": 2.639382402707276, + "grad_norm": 7.96875, + "learning_rate": 2.3958207637407244e-06, + "loss": 2.4813, + "step": 24958 + }, + { + "epoch": 2.6394881556683587, + "grad_norm": 8.375, + "learning_rate": 2.3956496565228947e-06, + "loss": 2.5274, + "step": 24959 + }, + { + "epoch": 2.6395939086294415, + "grad_norm": 8.1875, + "learning_rate": 2.395478549794739e-06, + "loss": 2.498, + "step": 24960 + }, + { + "epoch": 2.6396996615905244, + "grad_norm": 8.4375, + "learning_rate": 2.395307443557059e-06, + "loss": 2.2503, + "step": 24961 + }, + { + "epoch": 2.639805414551607, + "grad_norm": 8.4375, + "learning_rate": 2.39513633781066e-06, + "loss": 2.4171, + "step": 24962 + }, + { + "epoch": 2.6399111675126905, + "grad_norm": 9.25, + "learning_rate": 2.3949652325563414e-06, + "loss": 2.63, + "step": 24963 + }, + { + "epoch": 2.6400169204737733, + "grad_norm": 8.25, + "learning_rate": 2.394794127794909e-06, + "loss": 2.6057, + "step": 24964 + }, + { + "epoch": 2.640122673434856, + "grad_norm": 8.5, + "learning_rate": 2.394623023527164e-06, + "loss": 2.5293, + "step": 24965 + }, + { + "epoch": 2.640228426395939, + "grad_norm": 8.1875, + "learning_rate": 2.3944519197539104e-06, + "loss": 2.4132, + "step": 24966 + }, + { + "epoch": 2.640334179357022, + "grad_norm": 8.375, + "learning_rate": 2.3942808164759507e-06, + "loss": 2.4144, + "step": 24967 + }, + { + "epoch": 2.640439932318105, + "grad_norm": 9.1875, + "learning_rate": 2.394109713694088e-06, + "loss": 2.7591, + "step": 24968 + }, + { + "epoch": 2.640545685279188, + "grad_norm": 8.625, + "learning_rate": 2.3939386114091263e-06, + "loss": 2.4535, + "step": 24969 + }, + { + "epoch": 2.640651438240271, + "grad_norm": 8.75, + "learning_rate": 2.393767509621866e-06, + "loss": 2.5948, + "step": 24970 + }, + { + "epoch": 2.6407571912013537, + "grad_norm": 8.8125, + "learning_rate": 2.3935964083331113e-06, + "loss": 2.7867, + "step": 24971 + }, + { + "epoch": 2.6408629441624365, + "grad_norm": 8.125, + "learning_rate": 2.393425307543665e-06, + "loss": 2.4919, + "step": 24972 + }, + { + "epoch": 2.6409686971235193, + "grad_norm": 8.3125, + "learning_rate": 2.3932542072543306e-06, + "loss": 2.5251, + "step": 24973 + }, + { + "epoch": 2.641074450084602, + "grad_norm": 8.6875, + "learning_rate": 2.3930831074659103e-06, + "loss": 2.4675, + "step": 24974 + }, + { + "epoch": 2.6411802030456855, + "grad_norm": 8.5625, + "learning_rate": 2.392912008179207e-06, + "loss": 2.5172, + "step": 24975 + }, + { + "epoch": 2.6412859560067683, + "grad_norm": 7.96875, + "learning_rate": 2.3927409093950253e-06, + "loss": 2.3305, + "step": 24976 + }, + { + "epoch": 2.641391708967851, + "grad_norm": 9.1875, + "learning_rate": 2.3925698111141644e-06, + "loss": 2.5634, + "step": 24977 + }, + { + "epoch": 2.641497461928934, + "grad_norm": 8.8125, + "learning_rate": 2.3923987133374317e-06, + "loss": 2.6503, + "step": 24978 + }, + { + "epoch": 2.641603214890017, + "grad_norm": 8.625, + "learning_rate": 2.3922276160656264e-06, + "loss": 2.5169, + "step": 24979 + }, + { + "epoch": 2.6417089678511, + "grad_norm": 8.375, + "learning_rate": 2.392056519299553e-06, + "loss": 2.7906, + "step": 24980 + }, + { + "epoch": 2.6418147208121825, + "grad_norm": 8.0625, + "learning_rate": 2.3918854230400143e-06, + "loss": 2.5179, + "step": 24981 + }, + { + "epoch": 2.6419204737732658, + "grad_norm": 9.0625, + "learning_rate": 2.391714327287813e-06, + "loss": 2.4308, + "step": 24982 + }, + { + "epoch": 2.6420262267343486, + "grad_norm": 8.25, + "learning_rate": 2.3915432320437522e-06, + "loss": 2.4352, + "step": 24983 + }, + { + "epoch": 2.6421319796954315, + "grad_norm": 8.375, + "learning_rate": 2.391372137308635e-06, + "loss": 2.6137, + "step": 24984 + }, + { + "epoch": 2.6422377326565143, + "grad_norm": 8.375, + "learning_rate": 2.391201043083264e-06, + "loss": 2.6003, + "step": 24985 + }, + { + "epoch": 2.642343485617597, + "grad_norm": 8.1875, + "learning_rate": 2.3910299493684413e-06, + "loss": 2.632, + "step": 24986 + }, + { + "epoch": 2.6424492385786804, + "grad_norm": 8.375, + "learning_rate": 2.3908588561649708e-06, + "loss": 2.38, + "step": 24987 + }, + { + "epoch": 2.6425549915397633, + "grad_norm": 8.1875, + "learning_rate": 2.390687763473655e-06, + "loss": 2.5631, + "step": 24988 + }, + { + "epoch": 2.642660744500846, + "grad_norm": 8.625, + "learning_rate": 2.390516671295297e-06, + "loss": 2.6023, + "step": 24989 + }, + { + "epoch": 2.642766497461929, + "grad_norm": 9.1875, + "learning_rate": 2.390345579630699e-06, + "loss": 2.6347, + "step": 24990 + }, + { + "epoch": 2.6428722504230118, + "grad_norm": 8.4375, + "learning_rate": 2.3901744884806643e-06, + "loss": 2.4817, + "step": 24991 + }, + { + "epoch": 2.6429780033840946, + "grad_norm": 8.5, + "learning_rate": 2.3900033978459974e-06, + "loss": 2.6839, + "step": 24992 + }, + { + "epoch": 2.6430837563451774, + "grad_norm": 8.5625, + "learning_rate": 2.3898323077274975e-06, + "loss": 2.6125, + "step": 24993 + }, + { + "epoch": 2.6431895093062607, + "grad_norm": 9.125, + "learning_rate": 2.389661218125971e-06, + "loss": 2.5128, + "step": 24994 + }, + { + "epoch": 2.6432952622673436, + "grad_norm": 8.3125, + "learning_rate": 2.3894901290422185e-06, + "loss": 2.5839, + "step": 24995 + }, + { + "epoch": 2.6434010152284264, + "grad_norm": 7.96875, + "learning_rate": 2.389319040477044e-06, + "loss": 2.6465, + "step": 24996 + }, + { + "epoch": 2.6435067681895092, + "grad_norm": 8.3125, + "learning_rate": 2.3891479524312495e-06, + "loss": 2.5236, + "step": 24997 + }, + { + "epoch": 2.643612521150592, + "grad_norm": 8.4375, + "learning_rate": 2.388976864905639e-06, + "loss": 2.6246, + "step": 24998 + }, + { + "epoch": 2.6437182741116754, + "grad_norm": 8.75, + "learning_rate": 2.388805777901014e-06, + "loss": 2.5972, + "step": 24999 + }, + { + "epoch": 2.6438240270727578, + "grad_norm": 8.625, + "learning_rate": 2.3886346914181777e-06, + "loss": 2.4096, + "step": 25000 + }, + { + "epoch": 2.643929780033841, + "grad_norm": 8.25, + "learning_rate": 2.3884636054579345e-06, + "loss": 2.3979, + "step": 25001 + }, + { + "epoch": 2.644035532994924, + "grad_norm": 8.25, + "learning_rate": 2.3882925200210852e-06, + "loss": 2.5057, + "step": 25002 + }, + { + "epoch": 2.6441412859560067, + "grad_norm": 8.25, + "learning_rate": 2.3881214351084338e-06, + "loss": 2.5659, + "step": 25003 + }, + { + "epoch": 2.6442470389170896, + "grad_norm": 8.4375, + "learning_rate": 2.387950350720782e-06, + "loss": 2.6445, + "step": 25004 + }, + { + "epoch": 2.6443527918781724, + "grad_norm": 8.125, + "learning_rate": 2.3877792668589344e-06, + "loss": 2.5471, + "step": 25005 + }, + { + "epoch": 2.6444585448392557, + "grad_norm": 8.3125, + "learning_rate": 2.3876081835236916e-06, + "loss": 2.5186, + "step": 25006 + }, + { + "epoch": 2.6445642978003385, + "grad_norm": 8.8125, + "learning_rate": 2.387437100715859e-06, + "loss": 2.5873, + "step": 25007 + }, + { + "epoch": 2.6446700507614214, + "grad_norm": 8.5625, + "learning_rate": 2.3872660184362374e-06, + "loss": 2.619, + "step": 25008 + }, + { + "epoch": 2.644775803722504, + "grad_norm": 8.25, + "learning_rate": 2.387094936685629e-06, + "loss": 2.5202, + "step": 25009 + }, + { + "epoch": 2.644881556683587, + "grad_norm": 8.375, + "learning_rate": 2.38692385546484e-06, + "loss": 2.6795, + "step": 25010 + }, + { + "epoch": 2.6449873096446703, + "grad_norm": 8.0625, + "learning_rate": 2.38675277477467e-06, + "loss": 2.0971, + "step": 25011 + }, + { + "epoch": 2.6450930626057527, + "grad_norm": 8.3125, + "learning_rate": 2.3865816946159234e-06, + "loss": 2.4171, + "step": 25012 + }, + { + "epoch": 2.645198815566836, + "grad_norm": 7.96875, + "learning_rate": 2.3864106149894017e-06, + "loss": 2.4664, + "step": 25013 + }, + { + "epoch": 2.645304568527919, + "grad_norm": 8.0, + "learning_rate": 2.38623953589591e-06, + "loss": 2.4884, + "step": 25014 + }, + { + "epoch": 2.6454103214890017, + "grad_norm": 8.125, + "learning_rate": 2.3860684573362477e-06, + "loss": 2.3175, + "step": 25015 + }, + { + "epoch": 2.6455160744500845, + "grad_norm": 8.5, + "learning_rate": 2.3858973793112213e-06, + "loss": 2.6095, + "step": 25016 + }, + { + "epoch": 2.6456218274111674, + "grad_norm": 8.75, + "learning_rate": 2.3857263018216307e-06, + "loss": 2.6244, + "step": 25017 + }, + { + "epoch": 2.6457275803722506, + "grad_norm": 8.5, + "learning_rate": 2.38555522486828e-06, + "loss": 2.618, + "step": 25018 + }, + { + "epoch": 2.6458333333333335, + "grad_norm": 8.25, + "learning_rate": 2.3853841484519723e-06, + "loss": 2.6491, + "step": 25019 + }, + { + "epoch": 2.6459390862944163, + "grad_norm": 8.5, + "learning_rate": 2.385213072573509e-06, + "loss": 2.7121, + "step": 25020 + }, + { + "epoch": 2.646044839255499, + "grad_norm": 8.4375, + "learning_rate": 2.385041997233695e-06, + "loss": 2.5399, + "step": 25021 + }, + { + "epoch": 2.646150592216582, + "grad_norm": 8.375, + "learning_rate": 2.3848709224333306e-06, + "loss": 2.4681, + "step": 25022 + }, + { + "epoch": 2.646256345177665, + "grad_norm": 8.4375, + "learning_rate": 2.3846998481732217e-06, + "loss": 2.5336, + "step": 25023 + }, + { + "epoch": 2.6463620981387477, + "grad_norm": 8.375, + "learning_rate": 2.384528774454168e-06, + "loss": 2.4962, + "step": 25024 + }, + { + "epoch": 2.646467851099831, + "grad_norm": 7.75, + "learning_rate": 2.3843577012769736e-06, + "loss": 2.4338, + "step": 25025 + }, + { + "epoch": 2.646573604060914, + "grad_norm": 8.5625, + "learning_rate": 2.384186628642441e-06, + "loss": 2.5958, + "step": 25026 + }, + { + "epoch": 2.6466793570219966, + "grad_norm": 8.9375, + "learning_rate": 2.384015556551373e-06, + "loss": 2.4756, + "step": 25027 + }, + { + "epoch": 2.6467851099830795, + "grad_norm": 8.5625, + "learning_rate": 2.3838444850045734e-06, + "loss": 2.5672, + "step": 25028 + }, + { + "epoch": 2.6468908629441623, + "grad_norm": 8.375, + "learning_rate": 2.3836734140028435e-06, + "loss": 2.5383, + "step": 25029 + }, + { + "epoch": 2.6469966159052456, + "grad_norm": 8.1875, + "learning_rate": 2.383502343546988e-06, + "loss": 2.4657, + "step": 25030 + }, + { + "epoch": 2.647102368866328, + "grad_norm": 8.3125, + "learning_rate": 2.3833312736378064e-06, + "loss": 2.4424, + "step": 25031 + }, + { + "epoch": 2.6472081218274113, + "grad_norm": 8.1875, + "learning_rate": 2.383160204276105e-06, + "loss": 2.3596, + "step": 25032 + }, + { + "epoch": 2.647313874788494, + "grad_norm": 8.875, + "learning_rate": 2.382989135462684e-06, + "loss": 2.6767, + "step": 25033 + }, + { + "epoch": 2.647419627749577, + "grad_norm": 8.25, + "learning_rate": 2.382818067198348e-06, + "loss": 2.5448, + "step": 25034 + }, + { + "epoch": 2.64752538071066, + "grad_norm": 8.5625, + "learning_rate": 2.382646999483898e-06, + "loss": 2.7138, + "step": 25035 + }, + { + "epoch": 2.6476311336717426, + "grad_norm": 8.1875, + "learning_rate": 2.3824759323201374e-06, + "loss": 2.4122, + "step": 25036 + }, + { + "epoch": 2.647736886632826, + "grad_norm": 8.0625, + "learning_rate": 2.38230486570787e-06, + "loss": 2.538, + "step": 25037 + }, + { + "epoch": 2.6478426395939088, + "grad_norm": 8.0625, + "learning_rate": 2.382133799647897e-06, + "loss": 2.463, + "step": 25038 + }, + { + "epoch": 2.6479483925549916, + "grad_norm": 8.0625, + "learning_rate": 2.3819627341410233e-06, + "loss": 2.4251, + "step": 25039 + }, + { + "epoch": 2.6480541455160744, + "grad_norm": 8.3125, + "learning_rate": 2.381791669188049e-06, + "loss": 2.5978, + "step": 25040 + }, + { + "epoch": 2.6481598984771573, + "grad_norm": 8.75, + "learning_rate": 2.381620604789778e-06, + "loss": 2.705, + "step": 25041 + }, + { + "epoch": 2.64826565143824, + "grad_norm": 8.8125, + "learning_rate": 2.3814495409470132e-06, + "loss": 2.469, + "step": 25042 + }, + { + "epoch": 2.648371404399323, + "grad_norm": 8.4375, + "learning_rate": 2.381278477660558e-06, + "loss": 2.5229, + "step": 25043 + }, + { + "epoch": 2.6484771573604062, + "grad_norm": 8.8125, + "learning_rate": 2.381107414931213e-06, + "loss": 2.655, + "step": 25044 + }, + { + "epoch": 2.648582910321489, + "grad_norm": 8.4375, + "learning_rate": 2.380936352759783e-06, + "loss": 2.4215, + "step": 25045 + }, + { + "epoch": 2.648688663282572, + "grad_norm": 8.4375, + "learning_rate": 2.380765291147071e-06, + "loss": 2.3503, + "step": 25046 + }, + { + "epoch": 2.6487944162436547, + "grad_norm": 8.4375, + "learning_rate": 2.380594230093877e-06, + "loss": 2.7326, + "step": 25047 + }, + { + "epoch": 2.6489001692047376, + "grad_norm": 8.625, + "learning_rate": 2.3804231696010065e-06, + "loss": 2.6102, + "step": 25048 + }, + { + "epoch": 2.649005922165821, + "grad_norm": 8.5625, + "learning_rate": 2.3802521096692608e-06, + "loss": 2.6459, + "step": 25049 + }, + { + "epoch": 2.6491116751269037, + "grad_norm": 8.125, + "learning_rate": 2.380081050299443e-06, + "loss": 2.3843, + "step": 25050 + }, + { + "epoch": 2.6492174280879865, + "grad_norm": 8.0625, + "learning_rate": 2.3799099914923555e-06, + "loss": 2.4736, + "step": 25051 + }, + { + "epoch": 2.6493231810490694, + "grad_norm": 8.4375, + "learning_rate": 2.3797389332488023e-06, + "loss": 2.5726, + "step": 25052 + }, + { + "epoch": 2.6494289340101522, + "grad_norm": 8.6875, + "learning_rate": 2.379567875569584e-06, + "loss": 2.4866, + "step": 25053 + }, + { + "epoch": 2.649534686971235, + "grad_norm": 8.875, + "learning_rate": 2.3793968184555044e-06, + "loss": 2.632, + "step": 25054 + }, + { + "epoch": 2.649640439932318, + "grad_norm": 8.3125, + "learning_rate": 2.3792257619073673e-06, + "loss": 2.573, + "step": 25055 + }, + { + "epoch": 2.649746192893401, + "grad_norm": 9.1875, + "learning_rate": 2.3790547059259736e-06, + "loss": 2.8292, + "step": 25056 + }, + { + "epoch": 2.649851945854484, + "grad_norm": 8.4375, + "learning_rate": 2.3788836505121267e-06, + "loss": 2.5812, + "step": 25057 + }, + { + "epoch": 2.649957698815567, + "grad_norm": 8.0, + "learning_rate": 2.378712595666629e-06, + "loss": 2.4747, + "step": 25058 + }, + { + "epoch": 2.6500634517766497, + "grad_norm": 8.625, + "learning_rate": 2.3785415413902844e-06, + "loss": 2.5637, + "step": 25059 + }, + { + "epoch": 2.6501692047377325, + "grad_norm": 8.75, + "learning_rate": 2.3783704876838937e-06, + "loss": 2.6126, + "step": 25060 + }, + { + "epoch": 2.650274957698816, + "grad_norm": 8.125, + "learning_rate": 2.3781994345482618e-06, + "loss": 2.4898, + "step": 25061 + }, + { + "epoch": 2.650380710659898, + "grad_norm": 7.875, + "learning_rate": 2.3780283819841886e-06, + "loss": 2.4405, + "step": 25062 + }, + { + "epoch": 2.6504864636209815, + "grad_norm": 8.625, + "learning_rate": 2.3778573299924785e-06, + "loss": 2.7347, + "step": 25063 + }, + { + "epoch": 2.6505922165820643, + "grad_norm": 8.25, + "learning_rate": 2.3776862785739352e-06, + "loss": 2.4374, + "step": 25064 + }, + { + "epoch": 2.650697969543147, + "grad_norm": 9.0, + "learning_rate": 2.377515227729359e-06, + "loss": 2.5414, + "step": 25065 + }, + { + "epoch": 2.65080372250423, + "grad_norm": 8.25, + "learning_rate": 2.3773441774595545e-06, + "loss": 2.3456, + "step": 25066 + }, + { + "epoch": 2.650909475465313, + "grad_norm": 8.375, + "learning_rate": 2.377173127765323e-06, + "loss": 2.4505, + "step": 25067 + }, + { + "epoch": 2.651015228426396, + "grad_norm": 8.875, + "learning_rate": 2.3770020786474686e-06, + "loss": 2.5503, + "step": 25068 + }, + { + "epoch": 2.651120981387479, + "grad_norm": 9.0, + "learning_rate": 2.376831030106792e-06, + "loss": 2.6991, + "step": 25069 + }, + { + "epoch": 2.651226734348562, + "grad_norm": 8.25, + "learning_rate": 2.3766599821440985e-06, + "loss": 2.4868, + "step": 25070 + }, + { + "epoch": 2.6513324873096447, + "grad_norm": 8.0625, + "learning_rate": 2.376488934760188e-06, + "loss": 2.5182, + "step": 25071 + }, + { + "epoch": 2.6514382402707275, + "grad_norm": 8.1875, + "learning_rate": 2.3763178879558646e-06, + "loss": 2.6238, + "step": 25072 + }, + { + "epoch": 2.6515439932318103, + "grad_norm": 8.1875, + "learning_rate": 2.3761468417319313e-06, + "loss": 2.4223, + "step": 25073 + }, + { + "epoch": 2.651649746192893, + "grad_norm": 9.375, + "learning_rate": 2.3759757960891893e-06, + "loss": 2.6157, + "step": 25074 + }, + { + "epoch": 2.6517554991539765, + "grad_norm": 9.125, + "learning_rate": 2.375804751028443e-06, + "loss": 2.6284, + "step": 25075 + }, + { + "epoch": 2.6518612521150593, + "grad_norm": 8.5, + "learning_rate": 2.3756337065504937e-06, + "loss": 2.3469, + "step": 25076 + }, + { + "epoch": 2.651967005076142, + "grad_norm": 8.3125, + "learning_rate": 2.3754626626561454e-06, + "loss": 2.5166, + "step": 25077 + }, + { + "epoch": 2.652072758037225, + "grad_norm": 8.25, + "learning_rate": 2.375291619346199e-06, + "loss": 2.4174, + "step": 25078 + }, + { + "epoch": 2.652178510998308, + "grad_norm": 8.25, + "learning_rate": 2.3751205766214588e-06, + "loss": 2.3223, + "step": 25079 + }, + { + "epoch": 2.652284263959391, + "grad_norm": 8.5625, + "learning_rate": 2.374949534482726e-06, + "loss": 2.596, + "step": 25080 + }, + { + "epoch": 2.652390016920474, + "grad_norm": 8.25, + "learning_rate": 2.3747784929308037e-06, + "loss": 2.4811, + "step": 25081 + }, + { + "epoch": 2.6524957698815568, + "grad_norm": 9.125, + "learning_rate": 2.3746074519664952e-06, + "loss": 2.6203, + "step": 25082 + }, + { + "epoch": 2.6526015228426396, + "grad_norm": 8.6875, + "learning_rate": 2.3744364115906023e-06, + "loss": 2.6817, + "step": 25083 + }, + { + "epoch": 2.6527072758037225, + "grad_norm": 8.3125, + "learning_rate": 2.3742653718039293e-06, + "loss": 2.5985, + "step": 25084 + }, + { + "epoch": 2.6528130287648053, + "grad_norm": 8.5, + "learning_rate": 2.3740943326072752e-06, + "loss": 2.2628, + "step": 25085 + }, + { + "epoch": 2.652918781725888, + "grad_norm": 9.0, + "learning_rate": 2.373923294001447e-06, + "loss": 2.6306, + "step": 25086 + }, + { + "epoch": 2.6530245346869714, + "grad_norm": 8.9375, + "learning_rate": 2.3737522559872438e-06, + "loss": 2.7125, + "step": 25087 + }, + { + "epoch": 2.6531302876480543, + "grad_norm": 8.6875, + "learning_rate": 2.3735812185654706e-06, + "loss": 2.6517, + "step": 25088 + }, + { + "epoch": 2.653236040609137, + "grad_norm": 8.4375, + "learning_rate": 2.373410181736928e-06, + "loss": 2.2092, + "step": 25089 + }, + { + "epoch": 2.65334179357022, + "grad_norm": 8.9375, + "learning_rate": 2.37323914550242e-06, + "loss": 2.5547, + "step": 25090 + }, + { + "epoch": 2.6534475465313028, + "grad_norm": 8.4375, + "learning_rate": 2.3730681098627497e-06, + "loss": 2.6263, + "step": 25091 + }, + { + "epoch": 2.653553299492386, + "grad_norm": 8.5625, + "learning_rate": 2.372897074818718e-06, + "loss": 2.313, + "step": 25092 + }, + { + "epoch": 2.6536590524534684, + "grad_norm": 8.75, + "learning_rate": 2.372726040371129e-06, + "loss": 2.6802, + "step": 25093 + }, + { + "epoch": 2.6537648054145517, + "grad_norm": 8.0625, + "learning_rate": 2.372555006520784e-06, + "loss": 2.4193, + "step": 25094 + }, + { + "epoch": 2.6538705583756346, + "grad_norm": 8.875, + "learning_rate": 2.3723839732684867e-06, + "loss": 2.5773, + "step": 25095 + }, + { + "epoch": 2.6539763113367174, + "grad_norm": 8.75, + "learning_rate": 2.3722129406150385e-06, + "loss": 2.7473, + "step": 25096 + }, + { + "epoch": 2.6540820642978002, + "grad_norm": 8.125, + "learning_rate": 2.3720419085612426e-06, + "loss": 2.4217, + "step": 25097 + }, + { + "epoch": 2.654187817258883, + "grad_norm": 8.5, + "learning_rate": 2.3718708771079025e-06, + "loss": 2.4278, + "step": 25098 + }, + { + "epoch": 2.6542935702199664, + "grad_norm": 8.0625, + "learning_rate": 2.37169984625582e-06, + "loss": 2.5577, + "step": 25099 + }, + { + "epoch": 2.654399323181049, + "grad_norm": 8.25, + "learning_rate": 2.371528816005798e-06, + "loss": 2.7516, + "step": 25100 + }, + { + "epoch": 2.654505076142132, + "grad_norm": 8.375, + "learning_rate": 2.371357786358637e-06, + "loss": 2.5714, + "step": 25101 + }, + { + "epoch": 2.654610829103215, + "grad_norm": 8.9375, + "learning_rate": 2.3711867573151432e-06, + "loss": 2.5766, + "step": 25102 + }, + { + "epoch": 2.6547165820642977, + "grad_norm": 8.3125, + "learning_rate": 2.3710157288761164e-06, + "loss": 2.5576, + "step": 25103 + }, + { + "epoch": 2.6548223350253806, + "grad_norm": 8.0625, + "learning_rate": 2.37084470104236e-06, + "loss": 2.5679, + "step": 25104 + }, + { + "epoch": 2.6549280879864634, + "grad_norm": 8.3125, + "learning_rate": 2.3706736738146764e-06, + "loss": 2.5247, + "step": 25105 + }, + { + "epoch": 2.6550338409475467, + "grad_norm": 8.3125, + "learning_rate": 2.3705026471938684e-06, + "loss": 2.5934, + "step": 25106 + }, + { + "epoch": 2.6551395939086295, + "grad_norm": 8.6875, + "learning_rate": 2.3703316211807393e-06, + "loss": 2.7387, + "step": 25107 + }, + { + "epoch": 2.6552453468697124, + "grad_norm": 8.5, + "learning_rate": 2.3701605957760898e-06, + "loss": 2.5631, + "step": 25108 + }, + { + "epoch": 2.655351099830795, + "grad_norm": 8.375, + "learning_rate": 2.369989570980725e-06, + "loss": 2.5066, + "step": 25109 + }, + { + "epoch": 2.655456852791878, + "grad_norm": 8.5, + "learning_rate": 2.3698185467954447e-06, + "loss": 2.7043, + "step": 25110 + }, + { + "epoch": 2.6555626057529613, + "grad_norm": 9.375, + "learning_rate": 2.369647523221053e-06, + "loss": 2.5792, + "step": 25111 + }, + { + "epoch": 2.6556683587140437, + "grad_norm": 8.0, + "learning_rate": 2.369476500258352e-06, + "loss": 2.4823, + "step": 25112 + }, + { + "epoch": 2.655774111675127, + "grad_norm": 7.96875, + "learning_rate": 2.369305477908145e-06, + "loss": 2.5216, + "step": 25113 + }, + { + "epoch": 2.65587986463621, + "grad_norm": 8.6875, + "learning_rate": 2.369134456171233e-06, + "loss": 2.46, + "step": 25114 + }, + { + "epoch": 2.6559856175972927, + "grad_norm": 8.8125, + "learning_rate": 2.36896343504842e-06, + "loss": 2.5594, + "step": 25115 + }, + { + "epoch": 2.6560913705583755, + "grad_norm": 9.0, + "learning_rate": 2.3687924145405086e-06, + "loss": 2.4079, + "step": 25116 + }, + { + "epoch": 2.6561971235194584, + "grad_norm": 8.1875, + "learning_rate": 2.368621394648299e-06, + "loss": 2.4179, + "step": 25117 + }, + { + "epoch": 2.6563028764805416, + "grad_norm": 8.5625, + "learning_rate": 2.3684503753725975e-06, + "loss": 2.3056, + "step": 25118 + }, + { + "epoch": 2.6564086294416245, + "grad_norm": 8.0625, + "learning_rate": 2.3682793567142037e-06, + "loss": 2.5591, + "step": 25119 + }, + { + "epoch": 2.6565143824027073, + "grad_norm": 8.5625, + "learning_rate": 2.3681083386739212e-06, + "loss": 2.3593, + "step": 25120 + }, + { + "epoch": 2.65662013536379, + "grad_norm": 7.90625, + "learning_rate": 2.3679373212525518e-06, + "loss": 2.4008, + "step": 25121 + }, + { + "epoch": 2.656725888324873, + "grad_norm": 8.0625, + "learning_rate": 2.3677663044508995e-06, + "loss": 2.511, + "step": 25122 + }, + { + "epoch": 2.6568316412859563, + "grad_norm": 8.5, + "learning_rate": 2.3675952882697646e-06, + "loss": 2.6554, + "step": 25123 + }, + { + "epoch": 2.6569373942470387, + "grad_norm": 8.3125, + "learning_rate": 2.367424272709951e-06, + "loss": 2.4437, + "step": 25124 + }, + { + "epoch": 2.657043147208122, + "grad_norm": 8.75, + "learning_rate": 2.3672532577722623e-06, + "loss": 2.6343, + "step": 25125 + }, + { + "epoch": 2.657148900169205, + "grad_norm": 8.5, + "learning_rate": 2.367082243457499e-06, + "loss": 2.5307, + "step": 25126 + }, + { + "epoch": 2.6572546531302876, + "grad_norm": 9.375, + "learning_rate": 2.3669112297664646e-06, + "loss": 2.6059, + "step": 25127 + }, + { + "epoch": 2.6573604060913705, + "grad_norm": 8.125, + "learning_rate": 2.366740216699961e-06, + "loss": 2.5867, + "step": 25128 + }, + { + "epoch": 2.6574661590524533, + "grad_norm": 8.9375, + "learning_rate": 2.366569204258791e-06, + "loss": 2.6721, + "step": 25129 + }, + { + "epoch": 2.6575719120135366, + "grad_norm": 8.125, + "learning_rate": 2.3663981924437573e-06, + "loss": 2.4484, + "step": 25130 + }, + { + "epoch": 2.6576776649746194, + "grad_norm": 7.90625, + "learning_rate": 2.366227181255663e-06, + "loss": 2.4554, + "step": 25131 + }, + { + "epoch": 2.6577834179357023, + "grad_norm": 8.1875, + "learning_rate": 2.3660561706953087e-06, + "loss": 2.5237, + "step": 25132 + }, + { + "epoch": 2.657889170896785, + "grad_norm": 8.25, + "learning_rate": 2.3658851607634977e-06, + "loss": 2.4836, + "step": 25133 + }, + { + "epoch": 2.657994923857868, + "grad_norm": 8.375, + "learning_rate": 2.3657141514610344e-06, + "loss": 2.5447, + "step": 25134 + }, + { + "epoch": 2.658100676818951, + "grad_norm": 8.6875, + "learning_rate": 2.3655431427887186e-06, + "loss": 2.6429, + "step": 25135 + }, + { + "epoch": 2.6582064297800336, + "grad_norm": 8.5, + "learning_rate": 2.3653721347473543e-06, + "loss": 2.6547, + "step": 25136 + }, + { + "epoch": 2.658312182741117, + "grad_norm": 8.3125, + "learning_rate": 2.3652011273377427e-06, + "loss": 2.6015, + "step": 25137 + }, + { + "epoch": 2.6584179357021998, + "grad_norm": 8.375, + "learning_rate": 2.3650301205606886e-06, + "loss": 2.498, + "step": 25138 + }, + { + "epoch": 2.6585236886632826, + "grad_norm": 8.4375, + "learning_rate": 2.364859114416991e-06, + "loss": 2.6852, + "step": 25139 + }, + { + "epoch": 2.6586294416243654, + "grad_norm": 8.125, + "learning_rate": 2.364688108907456e-06, + "loss": 2.5144, + "step": 25140 + }, + { + "epoch": 2.6587351945854483, + "grad_norm": 8.1875, + "learning_rate": 2.3645171040328833e-06, + "loss": 2.4284, + "step": 25141 + }, + { + "epoch": 2.6588409475465316, + "grad_norm": 8.1875, + "learning_rate": 2.364346099794077e-06, + "loss": 2.6596, + "step": 25142 + }, + { + "epoch": 2.658946700507614, + "grad_norm": 8.1875, + "learning_rate": 2.3641750961918386e-06, + "loss": 2.3924, + "step": 25143 + }, + { + "epoch": 2.6590524534686972, + "grad_norm": 8.9375, + "learning_rate": 2.364004093226971e-06, + "loss": 2.6083, + "step": 25144 + }, + { + "epoch": 2.65915820642978, + "grad_norm": 8.8125, + "learning_rate": 2.3638330909002767e-06, + "loss": 2.748, + "step": 25145 + }, + { + "epoch": 2.659263959390863, + "grad_norm": 8.1875, + "learning_rate": 2.3636620892125578e-06, + "loss": 2.5004, + "step": 25146 + }, + { + "epoch": 2.6593697123519457, + "grad_norm": 8.8125, + "learning_rate": 2.3634910881646176e-06, + "loss": 2.7237, + "step": 25147 + }, + { + "epoch": 2.6594754653130286, + "grad_norm": 8.4375, + "learning_rate": 2.363320087757257e-06, + "loss": 2.4408, + "step": 25148 + }, + { + "epoch": 2.659581218274112, + "grad_norm": 8.625, + "learning_rate": 2.3631490879912803e-06, + "loss": 2.4853, + "step": 25149 + }, + { + "epoch": 2.6596869712351947, + "grad_norm": 8.5, + "learning_rate": 2.362978088867488e-06, + "loss": 2.3468, + "step": 25150 + }, + { + "epoch": 2.6597927241962775, + "grad_norm": 8.3125, + "learning_rate": 2.362807090386684e-06, + "loss": 2.373, + "step": 25151 + }, + { + "epoch": 2.6598984771573604, + "grad_norm": 7.96875, + "learning_rate": 2.36263609254967e-06, + "loss": 2.4946, + "step": 25152 + }, + { + "epoch": 2.6600042301184432, + "grad_norm": 8.3125, + "learning_rate": 2.3624650953572484e-06, + "loss": 2.4251, + "step": 25153 + }, + { + "epoch": 2.660109983079526, + "grad_norm": 8.5, + "learning_rate": 2.3622940988102232e-06, + "loss": 2.45, + "step": 25154 + }, + { + "epoch": 2.660215736040609, + "grad_norm": 9.0, + "learning_rate": 2.362123102909393e-06, + "loss": 2.6887, + "step": 25155 + }, + { + "epoch": 2.660321489001692, + "grad_norm": 8.375, + "learning_rate": 2.361952107655565e-06, + "loss": 2.6772, + "step": 25156 + }, + { + "epoch": 2.660427241962775, + "grad_norm": 8.375, + "learning_rate": 2.3617811130495383e-06, + "loss": 2.3415, + "step": 25157 + }, + { + "epoch": 2.660532994923858, + "grad_norm": 8.5625, + "learning_rate": 2.361610119092117e-06, + "loss": 2.4443, + "step": 25158 + }, + { + "epoch": 2.6606387478849407, + "grad_norm": 8.3125, + "learning_rate": 2.3614391257841017e-06, + "loss": 2.7538, + "step": 25159 + }, + { + "epoch": 2.6607445008460235, + "grad_norm": 7.96875, + "learning_rate": 2.361268133126296e-06, + "loss": 2.3682, + "step": 25160 + }, + { + "epoch": 2.660850253807107, + "grad_norm": 8.25, + "learning_rate": 2.3610971411195035e-06, + "loss": 2.421, + "step": 25161 + }, + { + "epoch": 2.6609560067681897, + "grad_norm": 8.5, + "learning_rate": 2.360926149764524e-06, + "loss": 2.4152, + "step": 25162 + }, + { + "epoch": 2.6610617597292725, + "grad_norm": 8.25, + "learning_rate": 2.3607551590621625e-06, + "loss": 2.4155, + "step": 25163 + }, + { + "epoch": 2.6611675126903553, + "grad_norm": 7.90625, + "learning_rate": 2.3605841690132188e-06, + "loss": 2.4638, + "step": 25164 + }, + { + "epoch": 2.661273265651438, + "grad_norm": 8.8125, + "learning_rate": 2.3604131796184975e-06, + "loss": 2.8041, + "step": 25165 + }, + { + "epoch": 2.661379018612521, + "grad_norm": 8.3125, + "learning_rate": 2.3602421908787995e-06, + "loss": 2.5607, + "step": 25166 + }, + { + "epoch": 2.661484771573604, + "grad_norm": 8.5, + "learning_rate": 2.3600712027949277e-06, + "loss": 2.2755, + "step": 25167 + }, + { + "epoch": 2.661590524534687, + "grad_norm": 8.375, + "learning_rate": 2.3599002153676844e-06, + "loss": 2.5278, + "step": 25168 + }, + { + "epoch": 2.66169627749577, + "grad_norm": 9.75, + "learning_rate": 2.359729228597872e-06, + "loss": 2.5835, + "step": 25169 + }, + { + "epoch": 2.661802030456853, + "grad_norm": 8.4375, + "learning_rate": 2.359558242486294e-06, + "loss": 2.6408, + "step": 25170 + }, + { + "epoch": 2.6619077834179357, + "grad_norm": 8.0, + "learning_rate": 2.3593872570337505e-06, + "loss": 2.47, + "step": 25171 + }, + { + "epoch": 2.6620135363790185, + "grad_norm": 9.25, + "learning_rate": 2.3592162722410465e-06, + "loss": 2.5832, + "step": 25172 + }, + { + "epoch": 2.662119289340102, + "grad_norm": 8.4375, + "learning_rate": 2.359045288108982e-06, + "loss": 2.5346, + "step": 25173 + }, + { + "epoch": 2.662225042301184, + "grad_norm": 8.1875, + "learning_rate": 2.3588743046383604e-06, + "loss": 2.3928, + "step": 25174 + }, + { + "epoch": 2.6623307952622675, + "grad_norm": 9.0, + "learning_rate": 2.358703321829984e-06, + "loss": 2.6815, + "step": 25175 + }, + { + "epoch": 2.6624365482233503, + "grad_norm": 8.4375, + "learning_rate": 2.3585323396846555e-06, + "loss": 2.4399, + "step": 25176 + }, + { + "epoch": 2.662542301184433, + "grad_norm": 8.0625, + "learning_rate": 2.3583613582031756e-06, + "loss": 2.6008, + "step": 25177 + }, + { + "epoch": 2.662648054145516, + "grad_norm": 8.5, + "learning_rate": 2.3581903773863485e-06, + "loss": 2.7329, + "step": 25178 + }, + { + "epoch": 2.662753807106599, + "grad_norm": 8.375, + "learning_rate": 2.358019397234977e-06, + "loss": 2.2457, + "step": 25179 + }, + { + "epoch": 2.662859560067682, + "grad_norm": 8.875, + "learning_rate": 2.3578484177498616e-06, + "loss": 2.5587, + "step": 25180 + }, + { + "epoch": 2.662965313028765, + "grad_norm": 8.1875, + "learning_rate": 2.3576774389318057e-06, + "loss": 2.5601, + "step": 25181 + }, + { + "epoch": 2.6630710659898478, + "grad_norm": 8.75, + "learning_rate": 2.3575064607816107e-06, + "loss": 2.59, + "step": 25182 + }, + { + "epoch": 2.6631768189509306, + "grad_norm": 8.5625, + "learning_rate": 2.3573354833000804e-06, + "loss": 2.494, + "step": 25183 + }, + { + "epoch": 2.6632825719120135, + "grad_norm": 8.1875, + "learning_rate": 2.3571645064880158e-06, + "loss": 2.4205, + "step": 25184 + }, + { + "epoch": 2.6633883248730963, + "grad_norm": 8.3125, + "learning_rate": 2.3569935303462206e-06, + "loss": 2.5237, + "step": 25185 + }, + { + "epoch": 2.663494077834179, + "grad_norm": 8.5, + "learning_rate": 2.3568225548754953e-06, + "loss": 2.6607, + "step": 25186 + }, + { + "epoch": 2.6635998307952624, + "grad_norm": 8.25, + "learning_rate": 2.3566515800766424e-06, + "loss": 2.5184, + "step": 25187 + }, + { + "epoch": 2.6637055837563453, + "grad_norm": 8.8125, + "learning_rate": 2.356480605950467e-06, + "loss": 2.6316, + "step": 25188 + }, + { + "epoch": 2.663811336717428, + "grad_norm": 8.4375, + "learning_rate": 2.3563096324977686e-06, + "loss": 2.5349, + "step": 25189 + }, + { + "epoch": 2.663917089678511, + "grad_norm": 8.8125, + "learning_rate": 2.3561386597193507e-06, + "loss": 2.6103, + "step": 25190 + }, + { + "epoch": 2.6640228426395938, + "grad_norm": 9.0, + "learning_rate": 2.355967687616014e-06, + "loss": 2.5399, + "step": 25191 + }, + { + "epoch": 2.664128595600677, + "grad_norm": 8.5, + "learning_rate": 2.355796716188564e-06, + "loss": 2.7556, + "step": 25192 + }, + { + "epoch": 2.66423434856176, + "grad_norm": 8.875, + "learning_rate": 2.3556257454377988e-06, + "loss": 2.4548, + "step": 25193 + }, + { + "epoch": 2.6643401015228427, + "grad_norm": 8.625, + "learning_rate": 2.355454775364525e-06, + "loss": 2.3767, + "step": 25194 + }, + { + "epoch": 2.6644458544839256, + "grad_norm": 8.1875, + "learning_rate": 2.3552838059695416e-06, + "loss": 2.4021, + "step": 25195 + }, + { + "epoch": 2.6645516074450084, + "grad_norm": 8.0, + "learning_rate": 2.355112837253652e-06, + "loss": 2.4208, + "step": 25196 + }, + { + "epoch": 2.6646573604060912, + "grad_norm": 8.875, + "learning_rate": 2.354941869217659e-06, + "loss": 2.5381, + "step": 25197 + }, + { + "epoch": 2.664763113367174, + "grad_norm": 8.75, + "learning_rate": 2.354770901862364e-06, + "loss": 2.8359, + "step": 25198 + }, + { + "epoch": 2.6648688663282574, + "grad_norm": 8.625, + "learning_rate": 2.354599935188571e-06, + "loss": 2.4863, + "step": 25199 + }, + { + "epoch": 2.66497461928934, + "grad_norm": 8.4375, + "learning_rate": 2.3544289691970794e-06, + "loss": 2.3692, + "step": 25200 + }, + { + "epoch": 2.665080372250423, + "grad_norm": 9.9375, + "learning_rate": 2.3542580038886945e-06, + "loss": 2.4934, + "step": 25201 + }, + { + "epoch": 2.665186125211506, + "grad_norm": 7.78125, + "learning_rate": 2.354087039264217e-06, + "loss": 2.4389, + "step": 25202 + }, + { + "epoch": 2.6652918781725887, + "grad_norm": 8.75, + "learning_rate": 2.3539160753244488e-06, + "loss": 2.5236, + "step": 25203 + }, + { + "epoch": 2.665397631133672, + "grad_norm": 8.1875, + "learning_rate": 2.3537451120701925e-06, + "loss": 2.4342, + "step": 25204 + }, + { + "epoch": 2.6655033840947544, + "grad_norm": 8.5, + "learning_rate": 2.3535741495022507e-06, + "loss": 2.5813, + "step": 25205 + }, + { + "epoch": 2.6656091370558377, + "grad_norm": 8.625, + "learning_rate": 2.353403187621426e-06, + "loss": 2.7908, + "step": 25206 + }, + { + "epoch": 2.6657148900169205, + "grad_norm": 8.5625, + "learning_rate": 2.35323222642852e-06, + "loss": 2.6823, + "step": 25207 + }, + { + "epoch": 2.6658206429780034, + "grad_norm": 8.5, + "learning_rate": 2.3530612659243357e-06, + "loss": 2.555, + "step": 25208 + }, + { + "epoch": 2.665926395939086, + "grad_norm": 8.625, + "learning_rate": 2.352890306109673e-06, + "loss": 2.5622, + "step": 25209 + }, + { + "epoch": 2.666032148900169, + "grad_norm": 8.125, + "learning_rate": 2.352719346985338e-06, + "loss": 2.5834, + "step": 25210 + }, + { + "epoch": 2.6661379018612523, + "grad_norm": 8.5625, + "learning_rate": 2.35254838855213e-06, + "loss": 2.4949, + "step": 25211 + }, + { + "epoch": 2.666243654822335, + "grad_norm": 8.5, + "learning_rate": 2.3523774308108517e-06, + "loss": 2.7047, + "step": 25212 + }, + { + "epoch": 2.666349407783418, + "grad_norm": 8.5, + "learning_rate": 2.352206473762306e-06, + "loss": 2.7262, + "step": 25213 + }, + { + "epoch": 2.666455160744501, + "grad_norm": 8.5625, + "learning_rate": 2.352035517407295e-06, + "loss": 2.7732, + "step": 25214 + }, + { + "epoch": 2.6665609137055837, + "grad_norm": 8.25, + "learning_rate": 2.351864561746621e-06, + "loss": 2.3014, + "step": 25215 + }, + { + "epoch": 2.6666666666666665, + "grad_norm": 8.3125, + "learning_rate": 2.3516936067810857e-06, + "loss": 2.5049, + "step": 25216 + }, + { + "epoch": 2.6667724196277494, + "grad_norm": 8.3125, + "learning_rate": 2.351522652511492e-06, + "loss": 2.61, + "step": 25217 + }, + { + "epoch": 2.6668781725888326, + "grad_norm": 8.4375, + "learning_rate": 2.3513516989386416e-06, + "loss": 2.3648, + "step": 25218 + }, + { + "epoch": 2.6669839255499155, + "grad_norm": 8.75, + "learning_rate": 2.351180746063337e-06, + "loss": 2.5265, + "step": 25219 + }, + { + "epoch": 2.6670896785109983, + "grad_norm": 7.90625, + "learning_rate": 2.35100979388638e-06, + "loss": 2.3984, + "step": 25220 + }, + { + "epoch": 2.667195431472081, + "grad_norm": 8.3125, + "learning_rate": 2.3508388424085728e-06, + "loss": 2.6235, + "step": 25221 + }, + { + "epoch": 2.667301184433164, + "grad_norm": 8.25, + "learning_rate": 2.3506678916307184e-06, + "loss": 2.5475, + "step": 25222 + }, + { + "epoch": 2.6674069373942473, + "grad_norm": 8.0, + "learning_rate": 2.350496941553618e-06, + "loss": 2.6093, + "step": 25223 + }, + { + "epoch": 2.6675126903553297, + "grad_norm": 8.4375, + "learning_rate": 2.3503259921780756e-06, + "loss": 2.5572, + "step": 25224 + }, + { + "epoch": 2.667618443316413, + "grad_norm": 9.5, + "learning_rate": 2.3501550435048905e-06, + "loss": 2.7339, + "step": 25225 + }, + { + "epoch": 2.667724196277496, + "grad_norm": 8.1875, + "learning_rate": 2.3499840955348682e-06, + "loss": 2.4747, + "step": 25226 + }, + { + "epoch": 2.6678299492385786, + "grad_norm": 8.375, + "learning_rate": 2.3498131482688082e-06, + "loss": 2.5107, + "step": 25227 + }, + { + "epoch": 2.6679357021996615, + "grad_norm": 8.1875, + "learning_rate": 2.349642201707514e-06, + "loss": 2.5252, + "step": 25228 + }, + { + "epoch": 2.6680414551607443, + "grad_norm": 8.375, + "learning_rate": 2.3494712558517866e-06, + "loss": 2.5288, + "step": 25229 + }, + { + "epoch": 2.6681472081218276, + "grad_norm": 8.375, + "learning_rate": 2.3493003107024293e-06, + "loss": 2.343, + "step": 25230 + }, + { + "epoch": 2.6682529610829104, + "grad_norm": 8.3125, + "learning_rate": 2.349129366260244e-06, + "loss": 2.3912, + "step": 25231 + }, + { + "epoch": 2.6683587140439933, + "grad_norm": 8.4375, + "learning_rate": 2.3489584225260333e-06, + "loss": 2.4582, + "step": 25232 + }, + { + "epoch": 2.668464467005076, + "grad_norm": 8.5625, + "learning_rate": 2.3487874795005997e-06, + "loss": 2.4809, + "step": 25233 + }, + { + "epoch": 2.668570219966159, + "grad_norm": 8.0, + "learning_rate": 2.3486165371847434e-06, + "loss": 2.5241, + "step": 25234 + }, + { + "epoch": 2.6686759729272422, + "grad_norm": 8.6875, + "learning_rate": 2.3484455955792686e-06, + "loss": 2.6211, + "step": 25235 + }, + { + "epoch": 2.6687817258883246, + "grad_norm": 8.1875, + "learning_rate": 2.3482746546849754e-06, + "loss": 2.4653, + "step": 25236 + }, + { + "epoch": 2.668887478849408, + "grad_norm": 8.5625, + "learning_rate": 2.3481037145026685e-06, + "loss": 2.4434, + "step": 25237 + }, + { + "epoch": 2.6689932318104908, + "grad_norm": 8.0625, + "learning_rate": 2.3479327750331478e-06, + "loss": 2.4441, + "step": 25238 + }, + { + "epoch": 2.6690989847715736, + "grad_norm": 8.25, + "learning_rate": 2.3477618362772164e-06, + "loss": 2.3446, + "step": 25239 + }, + { + "epoch": 2.6692047377326564, + "grad_norm": 8.875, + "learning_rate": 2.3475908982356776e-06, + "loss": 2.3774, + "step": 25240 + }, + { + "epoch": 2.6693104906937393, + "grad_norm": 8.375, + "learning_rate": 2.3474199609093307e-06, + "loss": 2.4255, + "step": 25241 + }, + { + "epoch": 2.6694162436548226, + "grad_norm": 8.25, + "learning_rate": 2.347249024298981e-06, + "loss": 2.533, + "step": 25242 + }, + { + "epoch": 2.6695219966159054, + "grad_norm": 8.1875, + "learning_rate": 2.3470780884054286e-06, + "loss": 2.4906, + "step": 25243 + }, + { + "epoch": 2.6696277495769882, + "grad_norm": 8.75, + "learning_rate": 2.3469071532294764e-06, + "loss": 2.7274, + "step": 25244 + }, + { + "epoch": 2.669733502538071, + "grad_norm": 8.625, + "learning_rate": 2.346736218771926e-06, + "loss": 2.6455, + "step": 25245 + }, + { + "epoch": 2.669839255499154, + "grad_norm": 8.0, + "learning_rate": 2.3465652850335805e-06, + "loss": 2.2222, + "step": 25246 + }, + { + "epoch": 2.6699450084602367, + "grad_norm": 8.0, + "learning_rate": 2.3463943520152398e-06, + "loss": 2.3617, + "step": 25247 + }, + { + "epoch": 2.6700507614213196, + "grad_norm": 8.125, + "learning_rate": 2.3462234197177084e-06, + "loss": 2.4823, + "step": 25248 + }, + { + "epoch": 2.670156514382403, + "grad_norm": 8.6875, + "learning_rate": 2.3460524881417885e-06, + "loss": 2.4943, + "step": 25249 + }, + { + "epoch": 2.6702622673434857, + "grad_norm": 9.125, + "learning_rate": 2.3458815572882802e-06, + "loss": 2.7942, + "step": 25250 + }, + { + "epoch": 2.6703680203045685, + "grad_norm": 9.0625, + "learning_rate": 2.3457106271579873e-06, + "loss": 2.3876, + "step": 25251 + }, + { + "epoch": 2.6704737732656514, + "grad_norm": 8.9375, + "learning_rate": 2.345539697751711e-06, + "loss": 2.6891, + "step": 25252 + }, + { + "epoch": 2.6705795262267342, + "grad_norm": 8.3125, + "learning_rate": 2.3453687690702543e-06, + "loss": 2.524, + "step": 25253 + }, + { + "epoch": 2.6706852791878175, + "grad_norm": 8.5625, + "learning_rate": 2.3451978411144178e-06, + "loss": 2.6212, + "step": 25254 + }, + { + "epoch": 2.6707910321489, + "grad_norm": 8.5, + "learning_rate": 2.3450269138850055e-06, + "loss": 2.7052, + "step": 25255 + }, + { + "epoch": 2.670896785109983, + "grad_norm": 7.96875, + "learning_rate": 2.344855987382818e-06, + "loss": 2.447, + "step": 25256 + }, + { + "epoch": 2.671002538071066, + "grad_norm": 8.125, + "learning_rate": 2.344685061608657e-06, + "loss": 2.3421, + "step": 25257 + }, + { + "epoch": 2.671108291032149, + "grad_norm": 7.96875, + "learning_rate": 2.344514136563327e-06, + "loss": 2.4938, + "step": 25258 + }, + { + "epoch": 2.6712140439932317, + "grad_norm": 8.5625, + "learning_rate": 2.3443432122476277e-06, + "loss": 2.3916, + "step": 25259 + }, + { + "epoch": 2.6713197969543145, + "grad_norm": 8.8125, + "learning_rate": 2.3441722886623626e-06, + "loss": 2.328, + "step": 25260 + }, + { + "epoch": 2.671425549915398, + "grad_norm": 8.75, + "learning_rate": 2.3440013658083325e-06, + "loss": 2.289, + "step": 25261 + }, + { + "epoch": 2.6715313028764807, + "grad_norm": 8.125, + "learning_rate": 2.3438304436863414e-06, + "loss": 2.5078, + "step": 25262 + }, + { + "epoch": 2.6716370558375635, + "grad_norm": 8.4375, + "learning_rate": 2.3436595222971887e-06, + "loss": 2.5091, + "step": 25263 + }, + { + "epoch": 2.6717428087986463, + "grad_norm": 8.4375, + "learning_rate": 2.3434886016416793e-06, + "loss": 2.6207, + "step": 25264 + }, + { + "epoch": 2.671848561759729, + "grad_norm": 8.875, + "learning_rate": 2.343317681720613e-06, + "loss": 2.6337, + "step": 25265 + }, + { + "epoch": 2.671954314720812, + "grad_norm": 8.375, + "learning_rate": 2.3431467625347925e-06, + "loss": 2.5272, + "step": 25266 + }, + { + "epoch": 2.672060067681895, + "grad_norm": 8.5, + "learning_rate": 2.342975844085021e-06, + "loss": 2.6252, + "step": 25267 + }, + { + "epoch": 2.672165820642978, + "grad_norm": 7.9375, + "learning_rate": 2.3428049263720987e-06, + "loss": 2.4532, + "step": 25268 + }, + { + "epoch": 2.672271573604061, + "grad_norm": 8.1875, + "learning_rate": 2.3426340093968293e-06, + "loss": 2.149, + "step": 25269 + }, + { + "epoch": 2.672377326565144, + "grad_norm": 8.3125, + "learning_rate": 2.3424630931600133e-06, + "loss": 2.4765, + "step": 25270 + }, + { + "epoch": 2.6724830795262267, + "grad_norm": 8.8125, + "learning_rate": 2.342292177662455e-06, + "loss": 2.6715, + "step": 25271 + }, + { + "epoch": 2.6725888324873095, + "grad_norm": 8.0625, + "learning_rate": 2.3421212629049537e-06, + "loss": 2.5505, + "step": 25272 + }, + { + "epoch": 2.672694585448393, + "grad_norm": 8.4375, + "learning_rate": 2.341950348888314e-06, + "loss": 2.6441, + "step": 25273 + }, + { + "epoch": 2.6728003384094756, + "grad_norm": 8.625, + "learning_rate": 2.341779435613335e-06, + "loss": 2.4353, + "step": 25274 + }, + { + "epoch": 2.6729060913705585, + "grad_norm": 8.1875, + "learning_rate": 2.341608523080821e-06, + "loss": 2.5846, + "step": 25275 + }, + { + "epoch": 2.6730118443316413, + "grad_norm": 8.375, + "learning_rate": 2.341437611291574e-06, + "loss": 2.7707, + "step": 25276 + }, + { + "epoch": 2.673117597292724, + "grad_norm": 8.25, + "learning_rate": 2.341266700246395e-06, + "loss": 2.5712, + "step": 25277 + }, + { + "epoch": 2.673223350253807, + "grad_norm": 8.625, + "learning_rate": 2.3410957899460873e-06, + "loss": 2.5226, + "step": 25278 + }, + { + "epoch": 2.67332910321489, + "grad_norm": 8.75, + "learning_rate": 2.3409248803914507e-06, + "loss": 2.7056, + "step": 25279 + }, + { + "epoch": 2.673434856175973, + "grad_norm": 7.96875, + "learning_rate": 2.34075397158329e-06, + "loss": 2.576, + "step": 25280 + }, + { + "epoch": 2.673540609137056, + "grad_norm": 8.125, + "learning_rate": 2.3405830635224048e-06, + "loss": 2.4641, + "step": 25281 + }, + { + "epoch": 2.6736463620981388, + "grad_norm": 8.5, + "learning_rate": 2.3404121562095984e-06, + "loss": 2.3157, + "step": 25282 + }, + { + "epoch": 2.6737521150592216, + "grad_norm": 8.875, + "learning_rate": 2.340241249645672e-06, + "loss": 2.5998, + "step": 25283 + }, + { + "epoch": 2.6738578680203045, + "grad_norm": 8.6875, + "learning_rate": 2.340070343831428e-06, + "loss": 2.6296, + "step": 25284 + }, + { + "epoch": 2.6739636209813877, + "grad_norm": 9.25, + "learning_rate": 2.3398994387676695e-06, + "loss": 2.47, + "step": 25285 + }, + { + "epoch": 2.67406937394247, + "grad_norm": 8.1875, + "learning_rate": 2.339728534455196e-06, + "loss": 2.049, + "step": 25286 + }, + { + "epoch": 2.6741751269035534, + "grad_norm": 8.75, + "learning_rate": 2.339557630894813e-06, + "loss": 2.5485, + "step": 25287 + }, + { + "epoch": 2.6742808798646363, + "grad_norm": 8.375, + "learning_rate": 2.339386728087319e-06, + "loss": 2.5329, + "step": 25288 + }, + { + "epoch": 2.674386632825719, + "grad_norm": 8.4375, + "learning_rate": 2.3392158260335175e-06, + "loss": 2.6717, + "step": 25289 + }, + { + "epoch": 2.674492385786802, + "grad_norm": 8.4375, + "learning_rate": 2.33904492473421e-06, + "loss": 2.5978, + "step": 25290 + }, + { + "epoch": 2.6745981387478848, + "grad_norm": 8.625, + "learning_rate": 2.3388740241902e-06, + "loss": 2.679, + "step": 25291 + }, + { + "epoch": 2.674703891708968, + "grad_norm": 7.875, + "learning_rate": 2.3387031244022873e-06, + "loss": 2.512, + "step": 25292 + }, + { + "epoch": 2.674809644670051, + "grad_norm": 8.0, + "learning_rate": 2.338532225371275e-06, + "loss": 2.6202, + "step": 25293 + }, + { + "epoch": 2.6749153976311337, + "grad_norm": 8.5625, + "learning_rate": 2.338361327097966e-06, + "loss": 2.4555, + "step": 25294 + }, + { + "epoch": 2.6750211505922166, + "grad_norm": 8.5, + "learning_rate": 2.338190429583159e-06, + "loss": 2.4812, + "step": 25295 + }, + { + "epoch": 2.6751269035532994, + "grad_norm": 8.625, + "learning_rate": 2.3380195328276605e-06, + "loss": 2.7895, + "step": 25296 + }, + { + "epoch": 2.6752326565143822, + "grad_norm": 8.0625, + "learning_rate": 2.337848636832269e-06, + "loss": 2.6143, + "step": 25297 + }, + { + "epoch": 2.675338409475465, + "grad_norm": 9.0625, + "learning_rate": 2.3376777415977877e-06, + "loss": 2.6804, + "step": 25298 + }, + { + "epoch": 2.6754441624365484, + "grad_norm": 8.0625, + "learning_rate": 2.337506847125018e-06, + "loss": 2.3585, + "step": 25299 + }, + { + "epoch": 2.675549915397631, + "grad_norm": 8.9375, + "learning_rate": 2.337335953414763e-06, + "loss": 2.6178, + "step": 25300 + }, + { + "epoch": 2.675655668358714, + "grad_norm": 8.75, + "learning_rate": 2.3371650604678224e-06, + "loss": 2.6787, + "step": 25301 + }, + { + "epoch": 2.675761421319797, + "grad_norm": 8.1875, + "learning_rate": 2.336994168285e-06, + "loss": 2.557, + "step": 25302 + }, + { + "epoch": 2.6758671742808797, + "grad_norm": 9.25, + "learning_rate": 2.336823276867099e-06, + "loss": 2.5666, + "step": 25303 + }, + { + "epoch": 2.675972927241963, + "grad_norm": 8.6875, + "learning_rate": 2.3366523862149184e-06, + "loss": 2.45, + "step": 25304 + }, + { + "epoch": 2.676078680203046, + "grad_norm": 8.5, + "learning_rate": 2.3364814963292614e-06, + "loss": 2.342, + "step": 25305 + }, + { + "epoch": 2.6761844331641287, + "grad_norm": 8.5625, + "learning_rate": 2.3363106072109294e-06, + "loss": 2.5445, + "step": 25306 + }, + { + "epoch": 2.6762901861252115, + "grad_norm": 8.5, + "learning_rate": 2.3361397188607255e-06, + "loss": 2.5694, + "step": 25307 + }, + { + "epoch": 2.6763959390862944, + "grad_norm": 8.4375, + "learning_rate": 2.33596883127945e-06, + "loss": 2.5125, + "step": 25308 + }, + { + "epoch": 2.676501692047377, + "grad_norm": 8.75, + "learning_rate": 2.3357979444679072e-06, + "loss": 2.617, + "step": 25309 + }, + { + "epoch": 2.67660744500846, + "grad_norm": 8.6875, + "learning_rate": 2.335627058426896e-06, + "loss": 2.5409, + "step": 25310 + }, + { + "epoch": 2.6767131979695433, + "grad_norm": 8.4375, + "learning_rate": 2.3354561731572195e-06, + "loss": 2.5021, + "step": 25311 + }, + { + "epoch": 2.676818950930626, + "grad_norm": 8.5, + "learning_rate": 2.3352852886596814e-06, + "loss": 2.5832, + "step": 25312 + }, + { + "epoch": 2.676924703891709, + "grad_norm": 8.5625, + "learning_rate": 2.3351144049350812e-06, + "loss": 2.6623, + "step": 25313 + }, + { + "epoch": 2.677030456852792, + "grad_norm": 8.5625, + "learning_rate": 2.3349435219842223e-06, + "loss": 2.5976, + "step": 25314 + }, + { + "epoch": 2.6771362098138747, + "grad_norm": 8.625, + "learning_rate": 2.334772639807905e-06, + "loss": 2.5846, + "step": 25315 + }, + { + "epoch": 2.677241962774958, + "grad_norm": 7.9375, + "learning_rate": 2.3346017584069336e-06, + "loss": 2.4822, + "step": 25316 + }, + { + "epoch": 2.6773477157360404, + "grad_norm": 8.0625, + "learning_rate": 2.3344308777821063e-06, + "loss": 2.4173, + "step": 25317 + }, + { + "epoch": 2.6774534686971236, + "grad_norm": 8.375, + "learning_rate": 2.3342599979342294e-06, + "loss": 2.5722, + "step": 25318 + }, + { + "epoch": 2.6775592216582065, + "grad_norm": 8.25, + "learning_rate": 2.3340891188641014e-06, + "loss": 2.574, + "step": 25319 + }, + { + "epoch": 2.6776649746192893, + "grad_norm": 8.0625, + "learning_rate": 2.333918240572525e-06, + "loss": 2.645, + "step": 25320 + }, + { + "epoch": 2.677770727580372, + "grad_norm": 8.25, + "learning_rate": 2.333747363060303e-06, + "loss": 2.6243, + "step": 25321 + }, + { + "epoch": 2.677876480541455, + "grad_norm": 8.375, + "learning_rate": 2.3335764863282363e-06, + "loss": 2.5556, + "step": 25322 + }, + { + "epoch": 2.6779822335025383, + "grad_norm": 8.75, + "learning_rate": 2.3334056103771272e-06, + "loss": 2.4444, + "step": 25323 + }, + { + "epoch": 2.678087986463621, + "grad_norm": 8.4375, + "learning_rate": 2.3332347352077775e-06, + "loss": 2.555, + "step": 25324 + }, + { + "epoch": 2.678193739424704, + "grad_norm": 8.125, + "learning_rate": 2.3330638608209896e-06, + "loss": 2.1377, + "step": 25325 + }, + { + "epoch": 2.678299492385787, + "grad_norm": 8.4375, + "learning_rate": 2.3328929872175636e-06, + "loss": 2.6851, + "step": 25326 + }, + { + "epoch": 2.6784052453468696, + "grad_norm": 9.1875, + "learning_rate": 2.332722114398302e-06, + "loss": 2.6254, + "step": 25327 + }, + { + "epoch": 2.6785109983079525, + "grad_norm": 8.625, + "learning_rate": 2.3325512423640093e-06, + "loss": 2.4652, + "step": 25328 + }, + { + "epoch": 2.6786167512690353, + "grad_norm": 8.0625, + "learning_rate": 2.3323803711154836e-06, + "loss": 2.6249, + "step": 25329 + }, + { + "epoch": 2.6787225042301186, + "grad_norm": 8.875, + "learning_rate": 2.3322095006535287e-06, + "loss": 2.3206, + "step": 25330 + }, + { + "epoch": 2.6788282571912014, + "grad_norm": 8.4375, + "learning_rate": 2.3320386309789456e-06, + "loss": 2.4826, + "step": 25331 + }, + { + "epoch": 2.6789340101522843, + "grad_norm": 8.1875, + "learning_rate": 2.3318677620925377e-06, + "loss": 2.4505, + "step": 25332 + }, + { + "epoch": 2.679039763113367, + "grad_norm": 7.875, + "learning_rate": 2.3316968939951033e-06, + "loss": 2.3465, + "step": 25333 + }, + { + "epoch": 2.67914551607445, + "grad_norm": 7.8125, + "learning_rate": 2.331526026687449e-06, + "loss": 2.3501, + "step": 25334 + }, + { + "epoch": 2.6792512690355332, + "grad_norm": 8.1875, + "learning_rate": 2.331355160170373e-06, + "loss": 2.5206, + "step": 25335 + }, + { + "epoch": 2.6793570219966156, + "grad_norm": 8.5, + "learning_rate": 2.3311842944446777e-06, + "loss": 2.441, + "step": 25336 + }, + { + "epoch": 2.679462774957699, + "grad_norm": 9.0, + "learning_rate": 2.3310134295111663e-06, + "loss": 2.7108, + "step": 25337 + }, + { + "epoch": 2.6795685279187818, + "grad_norm": 8.5625, + "learning_rate": 2.3308425653706395e-06, + "loss": 2.6222, + "step": 25338 + }, + { + "epoch": 2.6796742808798646, + "grad_norm": 7.9375, + "learning_rate": 2.3306717020238995e-06, + "loss": 2.4442, + "step": 25339 + }, + { + "epoch": 2.6797800338409474, + "grad_norm": 8.4375, + "learning_rate": 2.3305008394717475e-06, + "loss": 2.6399, + "step": 25340 + }, + { + "epoch": 2.6798857868020303, + "grad_norm": 9.25, + "learning_rate": 2.330329977714987e-06, + "loss": 2.6534, + "step": 25341 + }, + { + "epoch": 2.6799915397631136, + "grad_norm": 9.0, + "learning_rate": 2.3301591167544176e-06, + "loss": 2.4839, + "step": 25342 + }, + { + "epoch": 2.6800972927241964, + "grad_norm": 8.3125, + "learning_rate": 2.3299882565908422e-06, + "loss": 2.5219, + "step": 25343 + }, + { + "epoch": 2.6802030456852792, + "grad_norm": 8.25, + "learning_rate": 2.3298173972250618e-06, + "loss": 2.6177, + "step": 25344 + }, + { + "epoch": 2.680308798646362, + "grad_norm": 8.5, + "learning_rate": 2.3296465386578788e-06, + "loss": 2.6328, + "step": 25345 + }, + { + "epoch": 2.680414551607445, + "grad_norm": 8.4375, + "learning_rate": 2.3294756808900954e-06, + "loss": 2.6466, + "step": 25346 + }, + { + "epoch": 2.680520304568528, + "grad_norm": 9.5, + "learning_rate": 2.3293048239225126e-06, + "loss": 2.6167, + "step": 25347 + }, + { + "epoch": 2.6806260575296106, + "grad_norm": 8.6875, + "learning_rate": 2.3291339677559332e-06, + "loss": 2.6607, + "step": 25348 + }, + { + "epoch": 2.680731810490694, + "grad_norm": 8.8125, + "learning_rate": 2.328963112391157e-06, + "loss": 2.6927, + "step": 25349 + }, + { + "epoch": 2.6808375634517767, + "grad_norm": 7.9375, + "learning_rate": 2.3287922578289884e-06, + "loss": 2.5212, + "step": 25350 + }, + { + "epoch": 2.6809433164128595, + "grad_norm": 8.75, + "learning_rate": 2.3286214040702264e-06, + "loss": 2.3231, + "step": 25351 + }, + { + "epoch": 2.6810490693739424, + "grad_norm": 8.0625, + "learning_rate": 2.3284505511156747e-06, + "loss": 2.5758, + "step": 25352 + }, + { + "epoch": 2.6811548223350252, + "grad_norm": 8.6875, + "learning_rate": 2.328279698966134e-06, + "loss": 2.6182, + "step": 25353 + }, + { + "epoch": 2.6812605752961085, + "grad_norm": 8.0, + "learning_rate": 2.328108847622407e-06, + "loss": 2.5229, + "step": 25354 + }, + { + "epoch": 2.6813663282571913, + "grad_norm": 8.4375, + "learning_rate": 2.3279379970852944e-06, + "loss": 2.5365, + "step": 25355 + }, + { + "epoch": 2.681472081218274, + "grad_norm": 8.0, + "learning_rate": 2.3277671473555983e-06, + "loss": 2.4631, + "step": 25356 + }, + { + "epoch": 2.681577834179357, + "grad_norm": 7.96875, + "learning_rate": 2.3275962984341216e-06, + "loss": 2.4155, + "step": 25357 + }, + { + "epoch": 2.68168358714044, + "grad_norm": 8.375, + "learning_rate": 2.327425450321664e-06, + "loss": 2.4365, + "step": 25358 + }, + { + "epoch": 2.6817893401015227, + "grad_norm": 9.0, + "learning_rate": 2.3272546030190285e-06, + "loss": 2.613, + "step": 25359 + }, + { + "epoch": 2.6818950930626055, + "grad_norm": 7.875, + "learning_rate": 2.327083756527016e-06, + "loss": 2.4741, + "step": 25360 + }, + { + "epoch": 2.682000846023689, + "grad_norm": 8.6875, + "learning_rate": 2.3269129108464296e-06, + "loss": 2.4246, + "step": 25361 + }, + { + "epoch": 2.6821065989847717, + "grad_norm": 8.125, + "learning_rate": 2.326742065978069e-06, + "loss": 2.5946, + "step": 25362 + }, + { + "epoch": 2.6822123519458545, + "grad_norm": 8.3125, + "learning_rate": 2.3265712219227374e-06, + "loss": 2.5043, + "step": 25363 + }, + { + "epoch": 2.6823181049069373, + "grad_norm": 7.9375, + "learning_rate": 2.3264003786812374e-06, + "loss": 2.3983, + "step": 25364 + }, + { + "epoch": 2.68242385786802, + "grad_norm": 8.5625, + "learning_rate": 2.3262295362543672e-06, + "loss": 2.5946, + "step": 25365 + }, + { + "epoch": 2.6825296108291035, + "grad_norm": 8.9375, + "learning_rate": 2.3260586946429327e-06, + "loss": 2.6381, + "step": 25366 + }, + { + "epoch": 2.682635363790186, + "grad_norm": 8.5625, + "learning_rate": 2.3258878538477325e-06, + "loss": 2.4613, + "step": 25367 + }, + { + "epoch": 2.682741116751269, + "grad_norm": 8.1875, + "learning_rate": 2.32571701386957e-06, + "loss": 2.6187, + "step": 25368 + }, + { + "epoch": 2.682846869712352, + "grad_norm": 8.5, + "learning_rate": 2.3255461747092457e-06, + "loss": 2.5258, + "step": 25369 + }, + { + "epoch": 2.682952622673435, + "grad_norm": 8.3125, + "learning_rate": 2.325375336367563e-06, + "loss": 2.6094, + "step": 25370 + }, + { + "epoch": 2.6830583756345177, + "grad_norm": 8.0625, + "learning_rate": 2.3252044988453205e-06, + "loss": 2.4948, + "step": 25371 + }, + { + "epoch": 2.6831641285956005, + "grad_norm": 8.625, + "learning_rate": 2.3250336621433227e-06, + "loss": 2.6723, + "step": 25372 + }, + { + "epoch": 2.683269881556684, + "grad_norm": 8.6875, + "learning_rate": 2.3248628262623714e-06, + "loss": 2.4521, + "step": 25373 + }, + { + "epoch": 2.6833756345177666, + "grad_norm": 8.1875, + "learning_rate": 2.3246919912032657e-06, + "loss": 2.5019, + "step": 25374 + }, + { + "epoch": 2.6834813874788495, + "grad_norm": 8.125, + "learning_rate": 2.3245211569668098e-06, + "loss": 2.5776, + "step": 25375 + }, + { + "epoch": 2.6835871404399323, + "grad_norm": 8.0625, + "learning_rate": 2.3243503235538037e-06, + "loss": 2.3913, + "step": 25376 + }, + { + "epoch": 2.683692893401015, + "grad_norm": 8.125, + "learning_rate": 2.32417949096505e-06, + "loss": 2.469, + "step": 25377 + }, + { + "epoch": 2.683798646362098, + "grad_norm": 9.0, + "learning_rate": 2.3240086592013498e-06, + "loss": 2.9798, + "step": 25378 + }, + { + "epoch": 2.683904399323181, + "grad_norm": 8.375, + "learning_rate": 2.323837828263506e-06, + "loss": 2.354, + "step": 25379 + }, + { + "epoch": 2.684010152284264, + "grad_norm": 8.75, + "learning_rate": 2.323666998152318e-06, + "loss": 2.6951, + "step": 25380 + }, + { + "epoch": 2.684115905245347, + "grad_norm": 9.375, + "learning_rate": 2.323496168868588e-06, + "loss": 2.7367, + "step": 25381 + }, + { + "epoch": 2.6842216582064298, + "grad_norm": 8.25, + "learning_rate": 2.3233253404131203e-06, + "loss": 2.4303, + "step": 25382 + }, + { + "epoch": 2.6843274111675126, + "grad_norm": 8.75, + "learning_rate": 2.3231545127867134e-06, + "loss": 2.5855, + "step": 25383 + }, + { + "epoch": 2.6844331641285955, + "grad_norm": 8.375, + "learning_rate": 2.3229836859901705e-06, + "loss": 2.6822, + "step": 25384 + }, + { + "epoch": 2.6845389170896787, + "grad_norm": 8.375, + "learning_rate": 2.322812860024292e-06, + "loss": 2.7038, + "step": 25385 + }, + { + "epoch": 2.6846446700507616, + "grad_norm": 8.75, + "learning_rate": 2.3226420348898816e-06, + "loss": 2.4559, + "step": 25386 + }, + { + "epoch": 2.6847504230118444, + "grad_norm": 8.5625, + "learning_rate": 2.3224712105877377e-06, + "loss": 2.5793, + "step": 25387 + }, + { + "epoch": 2.6848561759729273, + "grad_norm": 8.125, + "learning_rate": 2.322300387118666e-06, + "loss": 2.4847, + "step": 25388 + }, + { + "epoch": 2.68496192893401, + "grad_norm": 8.4375, + "learning_rate": 2.3221295644834642e-06, + "loss": 2.4919, + "step": 25389 + }, + { + "epoch": 2.685067681895093, + "grad_norm": 7.875, + "learning_rate": 2.321958742682936e-06, + "loss": 2.4241, + "step": 25390 + }, + { + "epoch": 2.6851734348561758, + "grad_norm": 9.125, + "learning_rate": 2.3217879217178834e-06, + "loss": 2.685, + "step": 25391 + }, + { + "epoch": 2.685279187817259, + "grad_norm": 8.5, + "learning_rate": 2.321617101589106e-06, + "loss": 2.6884, + "step": 25392 + }, + { + "epoch": 2.685384940778342, + "grad_norm": 8.5, + "learning_rate": 2.3214462822974075e-06, + "loss": 2.5219, + "step": 25393 + }, + { + "epoch": 2.6854906937394247, + "grad_norm": 8.4375, + "learning_rate": 2.321275463843588e-06, + "loss": 2.6345, + "step": 25394 + }, + { + "epoch": 2.6855964467005076, + "grad_norm": 8.1875, + "learning_rate": 2.3211046462284505e-06, + "loss": 2.5226, + "step": 25395 + }, + { + "epoch": 2.6857021996615904, + "grad_norm": 8.6875, + "learning_rate": 2.320933829452795e-06, + "loss": 2.8233, + "step": 25396 + }, + { + "epoch": 2.6858079526226737, + "grad_norm": 8.625, + "learning_rate": 2.320763013517424e-06, + "loss": 2.415, + "step": 25397 + }, + { + "epoch": 2.685913705583756, + "grad_norm": 8.375, + "learning_rate": 2.3205921984231384e-06, + "loss": 2.6577, + "step": 25398 + }, + { + "epoch": 2.6860194585448394, + "grad_norm": 8.5625, + "learning_rate": 2.3204213841707406e-06, + "loss": 2.6361, + "step": 25399 + }, + { + "epoch": 2.686125211505922, + "grad_norm": 8.125, + "learning_rate": 2.3202505707610317e-06, + "loss": 2.4218, + "step": 25400 + }, + { + "epoch": 2.686230964467005, + "grad_norm": 8.3125, + "learning_rate": 2.320079758194813e-06, + "loss": 2.6555, + "step": 25401 + }, + { + "epoch": 2.686336717428088, + "grad_norm": 8.4375, + "learning_rate": 2.319908946472888e-06, + "loss": 2.4932, + "step": 25402 + }, + { + "epoch": 2.6864424703891707, + "grad_norm": 7.8125, + "learning_rate": 2.3197381355960545e-06, + "loss": 2.4403, + "step": 25403 + }, + { + "epoch": 2.686548223350254, + "grad_norm": 8.5625, + "learning_rate": 2.319567325565118e-06, + "loss": 2.5002, + "step": 25404 + }, + { + "epoch": 2.686653976311337, + "grad_norm": 8.6875, + "learning_rate": 2.3193965163808775e-06, + "loss": 2.5247, + "step": 25405 + }, + { + "epoch": 2.6867597292724197, + "grad_norm": 7.875, + "learning_rate": 2.3192257080441353e-06, + "loss": 2.5415, + "step": 25406 + }, + { + "epoch": 2.6868654822335025, + "grad_norm": 9.0625, + "learning_rate": 2.319054900555693e-06, + "loss": 2.6808, + "step": 25407 + }, + { + "epoch": 2.6869712351945854, + "grad_norm": 8.0625, + "learning_rate": 2.3188840939163513e-06, + "loss": 2.2566, + "step": 25408 + }, + { + "epoch": 2.687076988155668, + "grad_norm": 8.5, + "learning_rate": 2.3187132881269133e-06, + "loss": 2.5974, + "step": 25409 + }, + { + "epoch": 2.687182741116751, + "grad_norm": 8.3125, + "learning_rate": 2.3185424831881793e-06, + "loss": 2.6629, + "step": 25410 + }, + { + "epoch": 2.6872884940778343, + "grad_norm": 8.75, + "learning_rate": 2.3183716791009523e-06, + "loss": 2.6175, + "step": 25411 + }, + { + "epoch": 2.687394247038917, + "grad_norm": 8.1875, + "learning_rate": 2.3182008758660315e-06, + "loss": 2.6345, + "step": 25412 + }, + { + "epoch": 2.6875, + "grad_norm": 8.625, + "learning_rate": 2.3180300734842206e-06, + "loss": 2.6662, + "step": 25413 + }, + { + "epoch": 2.687605752961083, + "grad_norm": 8.375, + "learning_rate": 2.317859271956319e-06, + "loss": 2.5265, + "step": 25414 + }, + { + "epoch": 2.6877115059221657, + "grad_norm": 8.1875, + "learning_rate": 2.3176884712831297e-06, + "loss": 2.5324, + "step": 25415 + }, + { + "epoch": 2.687817258883249, + "grad_norm": 7.96875, + "learning_rate": 2.317517671465454e-06, + "loss": 2.524, + "step": 25416 + }, + { + "epoch": 2.687923011844332, + "grad_norm": 9.8125, + "learning_rate": 2.317346872504093e-06, + "loss": 2.5582, + "step": 25417 + }, + { + "epoch": 2.6880287648054146, + "grad_norm": 8.375, + "learning_rate": 2.3171760743998494e-06, + "loss": 2.543, + "step": 25418 + }, + { + "epoch": 2.6881345177664975, + "grad_norm": 8.3125, + "learning_rate": 2.317005277153522e-06, + "loss": 2.5978, + "step": 25419 + }, + { + "epoch": 2.6882402707275803, + "grad_norm": 8.5625, + "learning_rate": 2.316834480765916e-06, + "loss": 2.4759, + "step": 25420 + }, + { + "epoch": 2.688346023688663, + "grad_norm": 8.3125, + "learning_rate": 2.3166636852378296e-06, + "loss": 2.42, + "step": 25421 + }, + { + "epoch": 2.688451776649746, + "grad_norm": 8.4375, + "learning_rate": 2.316492890570066e-06, + "loss": 2.6866, + "step": 25422 + }, + { + "epoch": 2.6885575296108293, + "grad_norm": 8.5625, + "learning_rate": 2.3163220967634256e-06, + "loss": 2.5392, + "step": 25423 + }, + { + "epoch": 2.688663282571912, + "grad_norm": 8.1875, + "learning_rate": 2.3161513038187116e-06, + "loss": 2.3085, + "step": 25424 + }, + { + "epoch": 2.688769035532995, + "grad_norm": 8.375, + "learning_rate": 2.315980511736723e-06, + "loss": 2.6859, + "step": 25425 + }, + { + "epoch": 2.688874788494078, + "grad_norm": 8.25, + "learning_rate": 2.3158097205182626e-06, + "loss": 2.5056, + "step": 25426 + }, + { + "epoch": 2.6889805414551606, + "grad_norm": 8.5, + "learning_rate": 2.3156389301641337e-06, + "loss": 2.7343, + "step": 25427 + }, + { + "epoch": 2.689086294416244, + "grad_norm": 8.4375, + "learning_rate": 2.315468140675134e-06, + "loss": 2.6622, + "step": 25428 + }, + { + "epoch": 2.6891920473773263, + "grad_norm": 8.6875, + "learning_rate": 2.315297352052068e-06, + "loss": 2.2678, + "step": 25429 + }, + { + "epoch": 2.6892978003384096, + "grad_norm": 8.5625, + "learning_rate": 2.315126564295735e-06, + "loss": 2.6042, + "step": 25430 + }, + { + "epoch": 2.6894035532994924, + "grad_norm": 8.5, + "learning_rate": 2.314955777406938e-06, + "loss": 2.4311, + "step": 25431 + }, + { + "epoch": 2.6895093062605753, + "grad_norm": 8.5625, + "learning_rate": 2.314784991386478e-06, + "loss": 2.5524, + "step": 25432 + }, + { + "epoch": 2.689615059221658, + "grad_norm": 8.0625, + "learning_rate": 2.3146142062351566e-06, + "loss": 2.2685, + "step": 25433 + }, + { + "epoch": 2.689720812182741, + "grad_norm": 7.9375, + "learning_rate": 2.3144434219537744e-06, + "loss": 2.6097, + "step": 25434 + }, + { + "epoch": 2.6898265651438242, + "grad_norm": 8.875, + "learning_rate": 2.314272638543132e-06, + "loss": 2.6271, + "step": 25435 + }, + { + "epoch": 2.689932318104907, + "grad_norm": 8.5625, + "learning_rate": 2.3141018560040344e-06, + "loss": 2.3544, + "step": 25436 + }, + { + "epoch": 2.69003807106599, + "grad_norm": 8.25, + "learning_rate": 2.3139310743372795e-06, + "loss": 2.4772, + "step": 25437 + }, + { + "epoch": 2.6901438240270727, + "grad_norm": 8.375, + "learning_rate": 2.3137602935436707e-06, + "loss": 2.4649, + "step": 25438 + }, + { + "epoch": 2.6902495769881556, + "grad_norm": 8.25, + "learning_rate": 2.313589513624008e-06, + "loss": 2.5753, + "step": 25439 + }, + { + "epoch": 2.6903553299492384, + "grad_norm": 8.5, + "learning_rate": 2.3134187345790945e-06, + "loss": 2.7398, + "step": 25440 + }, + { + "epoch": 2.6904610829103213, + "grad_norm": 8.4375, + "learning_rate": 2.313247956409729e-06, + "loss": 2.5649, + "step": 25441 + }, + { + "epoch": 2.6905668358714045, + "grad_norm": 7.9375, + "learning_rate": 2.3130771791167146e-06, + "loss": 2.4834, + "step": 25442 + }, + { + "epoch": 2.6906725888324874, + "grad_norm": 8.5625, + "learning_rate": 2.312906402700854e-06, + "loss": 2.7156, + "step": 25443 + }, + { + "epoch": 2.6907783417935702, + "grad_norm": 8.3125, + "learning_rate": 2.3127356271629463e-06, + "loss": 2.6458, + "step": 25444 + }, + { + "epoch": 2.690884094754653, + "grad_norm": 9.0625, + "learning_rate": 2.312564852503794e-06, + "loss": 2.7517, + "step": 25445 + }, + { + "epoch": 2.690989847715736, + "grad_norm": 8.6875, + "learning_rate": 2.312394078724198e-06, + "loss": 2.7101, + "step": 25446 + }, + { + "epoch": 2.691095600676819, + "grad_norm": 8.625, + "learning_rate": 2.3122233058249595e-06, + "loss": 2.5055, + "step": 25447 + }, + { + "epoch": 2.691201353637902, + "grad_norm": 8.75, + "learning_rate": 2.3120525338068806e-06, + "loss": 2.808, + "step": 25448 + }, + { + "epoch": 2.691307106598985, + "grad_norm": 8.1875, + "learning_rate": 2.311881762670763e-06, + "loss": 2.3962, + "step": 25449 + }, + { + "epoch": 2.6914128595600677, + "grad_norm": 8.375, + "learning_rate": 2.3117109924174063e-06, + "loss": 2.6588, + "step": 25450 + }, + { + "epoch": 2.6915186125211505, + "grad_norm": 8.0625, + "learning_rate": 2.3115402230476123e-06, + "loss": 2.6002, + "step": 25451 + }, + { + "epoch": 2.6916243654822334, + "grad_norm": 8.625, + "learning_rate": 2.3113694545621846e-06, + "loss": 2.4435, + "step": 25452 + }, + { + "epoch": 2.691730118443316, + "grad_norm": 8.5, + "learning_rate": 2.311198686961922e-06, + "loss": 2.6114, + "step": 25453 + }, + { + "epoch": 2.6918358714043995, + "grad_norm": 7.9375, + "learning_rate": 2.3110279202476267e-06, + "loss": 2.4532, + "step": 25454 + }, + { + "epoch": 2.6919416243654823, + "grad_norm": 8.4375, + "learning_rate": 2.3108571544201004e-06, + "loss": 2.5624, + "step": 25455 + }, + { + "epoch": 2.692047377326565, + "grad_norm": 8.625, + "learning_rate": 2.3106863894801447e-06, + "loss": 2.586, + "step": 25456 + }, + { + "epoch": 2.692153130287648, + "grad_norm": 8.9375, + "learning_rate": 2.3105156254285584e-06, + "loss": 2.6795, + "step": 25457 + }, + { + "epoch": 2.692258883248731, + "grad_norm": 8.8125, + "learning_rate": 2.310344862266147e-06, + "loss": 2.6378, + "step": 25458 + }, + { + "epoch": 2.692364636209814, + "grad_norm": 8.5, + "learning_rate": 2.3101740999937082e-06, + "loss": 2.651, + "step": 25459 + }, + { + "epoch": 2.6924703891708965, + "grad_norm": 8.5, + "learning_rate": 2.3100033386120448e-06, + "loss": 2.5382, + "step": 25460 + }, + { + "epoch": 2.69257614213198, + "grad_norm": 8.0, + "learning_rate": 2.309832578121959e-06, + "loss": 2.567, + "step": 25461 + }, + { + "epoch": 2.6926818950930627, + "grad_norm": 8.0625, + "learning_rate": 2.3096618185242497e-06, + "loss": 2.4269, + "step": 25462 + }, + { + "epoch": 2.6927876480541455, + "grad_norm": 8.25, + "learning_rate": 2.3094910598197207e-06, + "loss": 2.7264, + "step": 25463 + }, + { + "epoch": 2.6928934010152283, + "grad_norm": 8.6875, + "learning_rate": 2.309320302009172e-06, + "loss": 2.6047, + "step": 25464 + }, + { + "epoch": 2.692999153976311, + "grad_norm": 8.625, + "learning_rate": 2.309149545093406e-06, + "loss": 2.729, + "step": 25465 + }, + { + "epoch": 2.6931049069373945, + "grad_norm": 8.375, + "learning_rate": 2.3089787890732216e-06, + "loss": 2.3825, + "step": 25466 + }, + { + "epoch": 2.6932106598984773, + "grad_norm": 8.9375, + "learning_rate": 2.3088080339494228e-06, + "loss": 2.5053, + "step": 25467 + }, + { + "epoch": 2.69331641285956, + "grad_norm": 8.3125, + "learning_rate": 2.3086372797228084e-06, + "loss": 2.2528, + "step": 25468 + }, + { + "epoch": 2.693422165820643, + "grad_norm": 8.9375, + "learning_rate": 2.3084665263941815e-06, + "loss": 2.46, + "step": 25469 + }, + { + "epoch": 2.693527918781726, + "grad_norm": 8.25, + "learning_rate": 2.3082957739643434e-06, + "loss": 2.5369, + "step": 25470 + }, + { + "epoch": 2.6936336717428087, + "grad_norm": 8.1875, + "learning_rate": 2.308125022434094e-06, + "loss": 2.5898, + "step": 25471 + }, + { + "epoch": 2.6937394247038915, + "grad_norm": 8.4375, + "learning_rate": 2.307954271804237e-06, + "loss": 2.5512, + "step": 25472 + }, + { + "epoch": 2.6938451776649748, + "grad_norm": 9.1875, + "learning_rate": 2.30778352207557e-06, + "loss": 2.5655, + "step": 25473 + }, + { + "epoch": 2.6939509306260576, + "grad_norm": 8.375, + "learning_rate": 2.3076127732488983e-06, + "loss": 2.5418, + "step": 25474 + }, + { + "epoch": 2.6940566835871405, + "grad_norm": 8.375, + "learning_rate": 2.3074420253250197e-06, + "loss": 2.497, + "step": 25475 + }, + { + "epoch": 2.6941624365482233, + "grad_norm": 8.5625, + "learning_rate": 2.3072712783047376e-06, + "loss": 2.569, + "step": 25476 + }, + { + "epoch": 2.694268189509306, + "grad_norm": 8.3125, + "learning_rate": 2.307100532188852e-06, + "loss": 2.5924, + "step": 25477 + }, + { + "epoch": 2.6943739424703894, + "grad_norm": 8.25, + "learning_rate": 2.306929786978165e-06, + "loss": 2.465, + "step": 25478 + }, + { + "epoch": 2.694479695431472, + "grad_norm": 8.75, + "learning_rate": 2.3067590426734775e-06, + "loss": 2.7099, + "step": 25479 + }, + { + "epoch": 2.694585448392555, + "grad_norm": 9.0, + "learning_rate": 2.3065882992755907e-06, + "loss": 2.5576, + "step": 25480 + }, + { + "epoch": 2.694691201353638, + "grad_norm": 8.6875, + "learning_rate": 2.306417556785307e-06, + "loss": 2.4858, + "step": 25481 + }, + { + "epoch": 2.6947969543147208, + "grad_norm": 8.25, + "learning_rate": 2.3062468152034253e-06, + "loss": 2.611, + "step": 25482 + }, + { + "epoch": 2.6949027072758036, + "grad_norm": 9.1875, + "learning_rate": 2.3060760745307484e-06, + "loss": 2.6023, + "step": 25483 + }, + { + "epoch": 2.6950084602368864, + "grad_norm": 8.8125, + "learning_rate": 2.3059053347680767e-06, + "loss": 2.5159, + "step": 25484 + }, + { + "epoch": 2.6951142131979697, + "grad_norm": 7.8125, + "learning_rate": 2.3057345959162127e-06, + "loss": 2.3483, + "step": 25485 + }, + { + "epoch": 2.6952199661590526, + "grad_norm": 8.625, + "learning_rate": 2.305563857975956e-06, + "loss": 2.6678, + "step": 25486 + }, + { + "epoch": 2.6953257191201354, + "grad_norm": 8.25, + "learning_rate": 2.305393120948109e-06, + "loss": 2.3724, + "step": 25487 + }, + { + "epoch": 2.6954314720812182, + "grad_norm": 8.4375, + "learning_rate": 2.305222384833473e-06, + "loss": 2.5178, + "step": 25488 + }, + { + "epoch": 2.695537225042301, + "grad_norm": 8.25, + "learning_rate": 2.3050516496328475e-06, + "loss": 2.5066, + "step": 25489 + }, + { + "epoch": 2.6956429780033844, + "grad_norm": 8.1875, + "learning_rate": 2.3048809153470362e-06, + "loss": 2.4352, + "step": 25490 + }, + { + "epoch": 2.6957487309644668, + "grad_norm": 8.375, + "learning_rate": 2.3047101819768383e-06, + "loss": 2.614, + "step": 25491 + }, + { + "epoch": 2.69585448392555, + "grad_norm": 8.5, + "learning_rate": 2.304539449523056e-06, + "loss": 2.4391, + "step": 25492 + }, + { + "epoch": 2.695960236886633, + "grad_norm": 8.0, + "learning_rate": 2.304368717986489e-06, + "loss": 2.5692, + "step": 25493 + }, + { + "epoch": 2.6960659898477157, + "grad_norm": 8.625, + "learning_rate": 2.3041979873679417e-06, + "loss": 2.5882, + "step": 25494 + }, + { + "epoch": 2.6961717428087986, + "grad_norm": 8.1875, + "learning_rate": 2.3040272576682107e-06, + "loss": 2.5517, + "step": 25495 + }, + { + "epoch": 2.6962774957698814, + "grad_norm": 8.0625, + "learning_rate": 2.303856528888101e-06, + "loss": 2.6886, + "step": 25496 + }, + { + "epoch": 2.6963832487309647, + "grad_norm": 8.9375, + "learning_rate": 2.3036858010284126e-06, + "loss": 2.2292, + "step": 25497 + }, + { + "epoch": 2.6964890016920475, + "grad_norm": 8.3125, + "learning_rate": 2.3035150740899455e-06, + "loss": 2.4715, + "step": 25498 + }, + { + "epoch": 2.6965947546531304, + "grad_norm": 8.8125, + "learning_rate": 2.303344348073503e-06, + "loss": 2.4178, + "step": 25499 + }, + { + "epoch": 2.696700507614213, + "grad_norm": 8.1875, + "learning_rate": 2.3031736229798838e-06, + "loss": 2.6771, + "step": 25500 + }, + { + "epoch": 2.696806260575296, + "grad_norm": 8.9375, + "learning_rate": 2.3030028988098913e-06, + "loss": 2.353, + "step": 25501 + }, + { + "epoch": 2.696912013536379, + "grad_norm": 8.1875, + "learning_rate": 2.3028321755643244e-06, + "loss": 2.4716, + "step": 25502 + }, + { + "epoch": 2.6970177664974617, + "grad_norm": 9.0, + "learning_rate": 2.3026614532439872e-06, + "loss": 2.6834, + "step": 25503 + }, + { + "epoch": 2.697123519458545, + "grad_norm": 8.6875, + "learning_rate": 2.302490731849678e-06, + "loss": 2.303, + "step": 25504 + }, + { + "epoch": 2.697229272419628, + "grad_norm": 8.6875, + "learning_rate": 2.3023200113821984e-06, + "loss": 2.5344, + "step": 25505 + }, + { + "epoch": 2.6973350253807107, + "grad_norm": 8.5625, + "learning_rate": 2.3021492918423515e-06, + "loss": 2.5991, + "step": 25506 + }, + { + "epoch": 2.6974407783417935, + "grad_norm": 8.5, + "learning_rate": 2.301978573230936e-06, + "loss": 2.5432, + "step": 25507 + }, + { + "epoch": 2.6975465313028764, + "grad_norm": 8.4375, + "learning_rate": 2.301807855548755e-06, + "loss": 2.6002, + "step": 25508 + }, + { + "epoch": 2.6976522842639596, + "grad_norm": 8.25, + "learning_rate": 2.301637138796608e-06, + "loss": 2.6089, + "step": 25509 + }, + { + "epoch": 2.697758037225042, + "grad_norm": 9.125, + "learning_rate": 2.3014664229752977e-06, + "loss": 2.439, + "step": 25510 + }, + { + "epoch": 2.6978637901861253, + "grad_norm": 7.65625, + "learning_rate": 2.3012957080856227e-06, + "loss": 2.4657, + "step": 25511 + }, + { + "epoch": 2.697969543147208, + "grad_norm": 8.6875, + "learning_rate": 2.3011249941283874e-06, + "loss": 2.3962, + "step": 25512 + }, + { + "epoch": 2.698075296108291, + "grad_norm": 8.8125, + "learning_rate": 2.3009542811043898e-06, + "loss": 2.8544, + "step": 25513 + }, + { + "epoch": 2.698181049069374, + "grad_norm": 8.1875, + "learning_rate": 2.3007835690144327e-06, + "loss": 2.4612, + "step": 25514 + }, + { + "epoch": 2.6982868020304567, + "grad_norm": 8.375, + "learning_rate": 2.300612857859317e-06, + "loss": 2.4967, + "step": 25515 + }, + { + "epoch": 2.69839255499154, + "grad_norm": 8.125, + "learning_rate": 2.3004421476398433e-06, + "loss": 2.3914, + "step": 25516 + }, + { + "epoch": 2.698498307952623, + "grad_norm": 8.9375, + "learning_rate": 2.3002714383568137e-06, + "loss": 2.6122, + "step": 25517 + }, + { + "epoch": 2.6986040609137056, + "grad_norm": 8.0625, + "learning_rate": 2.3001007300110276e-06, + "loss": 2.4115, + "step": 25518 + }, + { + "epoch": 2.6987098138747885, + "grad_norm": 8.8125, + "learning_rate": 2.2999300226032883e-06, + "loss": 2.3605, + "step": 25519 + }, + { + "epoch": 2.6988155668358713, + "grad_norm": 7.875, + "learning_rate": 2.2997593161343947e-06, + "loss": 2.4879, + "step": 25520 + }, + { + "epoch": 2.698921319796954, + "grad_norm": 8.1875, + "learning_rate": 2.299588610605149e-06, + "loss": 2.6276, + "step": 25521 + }, + { + "epoch": 2.699027072758037, + "grad_norm": 8.5625, + "learning_rate": 2.2994179060163514e-06, + "loss": 2.577, + "step": 25522 + }, + { + "epoch": 2.6991328257191203, + "grad_norm": 8.3125, + "learning_rate": 2.299247202368804e-06, + "loss": 2.4014, + "step": 25523 + }, + { + "epoch": 2.699238578680203, + "grad_norm": 8.3125, + "learning_rate": 2.2990764996633074e-06, + "loss": 2.4822, + "step": 25524 + }, + { + "epoch": 2.699344331641286, + "grad_norm": 8.25, + "learning_rate": 2.2989057979006623e-06, + "loss": 2.5675, + "step": 25525 + }, + { + "epoch": 2.699450084602369, + "grad_norm": 9.1875, + "learning_rate": 2.298735097081671e-06, + "loss": 2.5211, + "step": 25526 + }, + { + "epoch": 2.6995558375634516, + "grad_norm": 8.375, + "learning_rate": 2.298564397207132e-06, + "loss": 2.2746, + "step": 25527 + }, + { + "epoch": 2.699661590524535, + "grad_norm": 8.25, + "learning_rate": 2.2983936982778498e-06, + "loss": 2.6433, + "step": 25528 + }, + { + "epoch": 2.6997673434856178, + "grad_norm": 8.625, + "learning_rate": 2.298223000294622e-06, + "loss": 2.6895, + "step": 25529 + }, + { + "epoch": 2.6998730964467006, + "grad_norm": 8.5, + "learning_rate": 2.298052303258252e-06, + "loss": 2.6444, + "step": 25530 + }, + { + "epoch": 2.6999788494077834, + "grad_norm": 9.125, + "learning_rate": 2.297881607169539e-06, + "loss": 2.6227, + "step": 25531 + }, + { + "epoch": 2.7000846023688663, + "grad_norm": 8.125, + "learning_rate": 2.2977109120292856e-06, + "loss": 2.5504, + "step": 25532 + }, + { + "epoch": 2.700190355329949, + "grad_norm": 9.25, + "learning_rate": 2.297540217838292e-06, + "loss": 2.4631, + "step": 25533 + }, + { + "epoch": 2.700296108291032, + "grad_norm": 8.8125, + "learning_rate": 2.297369524597359e-06, + "loss": 2.5462, + "step": 25534 + }, + { + "epoch": 2.7004018612521152, + "grad_norm": 8.3125, + "learning_rate": 2.2971988323072894e-06, + "loss": 2.4706, + "step": 25535 + }, + { + "epoch": 2.700507614213198, + "grad_norm": 8.4375, + "learning_rate": 2.297028140968882e-06, + "loss": 2.5129, + "step": 25536 + }, + { + "epoch": 2.700613367174281, + "grad_norm": 8.1875, + "learning_rate": 2.296857450582938e-06, + "loss": 2.6249, + "step": 25537 + }, + { + "epoch": 2.7007191201353637, + "grad_norm": 8.6875, + "learning_rate": 2.2966867611502587e-06, + "loss": 2.46, + "step": 25538 + }, + { + "epoch": 2.7008248730964466, + "grad_norm": 8.5625, + "learning_rate": 2.296516072671646e-06, + "loss": 2.3747, + "step": 25539 + }, + { + "epoch": 2.70093062605753, + "grad_norm": 7.8125, + "learning_rate": 2.2963453851478997e-06, + "loss": 2.6144, + "step": 25540 + }, + { + "epoch": 2.7010363790186123, + "grad_norm": 8.9375, + "learning_rate": 2.296174698579821e-06, + "loss": 2.8863, + "step": 25541 + }, + { + "epoch": 2.7011421319796955, + "grad_norm": 9.0, + "learning_rate": 2.2960040129682123e-06, + "loss": 2.4365, + "step": 25542 + }, + { + "epoch": 2.7012478849407784, + "grad_norm": 8.875, + "learning_rate": 2.2958333283138718e-06, + "loss": 2.5433, + "step": 25543 + }, + { + "epoch": 2.7013536379018612, + "grad_norm": 8.6875, + "learning_rate": 2.2956626446176034e-06, + "loss": 2.529, + "step": 25544 + }, + { + "epoch": 2.701459390862944, + "grad_norm": 8.8125, + "learning_rate": 2.2954919618802056e-06, + "loss": 2.6193, + "step": 25545 + }, + { + "epoch": 2.701565143824027, + "grad_norm": 8.25, + "learning_rate": 2.2953212801024812e-06, + "loss": 2.5417, + "step": 25546 + }, + { + "epoch": 2.70167089678511, + "grad_norm": 8.3125, + "learning_rate": 2.2951505992852296e-06, + "loss": 2.4601, + "step": 25547 + }, + { + "epoch": 2.701776649746193, + "grad_norm": 9.9375, + "learning_rate": 2.2949799194292523e-06, + "loss": 2.5146, + "step": 25548 + }, + { + "epoch": 2.701882402707276, + "grad_norm": 8.375, + "learning_rate": 2.2948092405353508e-06, + "loss": 2.4126, + "step": 25549 + }, + { + "epoch": 2.7019881556683587, + "grad_norm": 8.0625, + "learning_rate": 2.2946385626043253e-06, + "loss": 2.4586, + "step": 25550 + }, + { + "epoch": 2.7020939086294415, + "grad_norm": 8.25, + "learning_rate": 2.294467885636978e-06, + "loss": 2.6472, + "step": 25551 + }, + { + "epoch": 2.7021996615905244, + "grad_norm": 9.0, + "learning_rate": 2.294297209634108e-06, + "loss": 2.2723, + "step": 25552 + }, + { + "epoch": 2.702305414551607, + "grad_norm": 8.1875, + "learning_rate": 2.2941265345965175e-06, + "loss": 2.4597, + "step": 25553 + }, + { + "epoch": 2.7024111675126905, + "grad_norm": 8.125, + "learning_rate": 2.2939558605250066e-06, + "loss": 2.4584, + "step": 25554 + }, + { + "epoch": 2.7025169204737733, + "grad_norm": 9.125, + "learning_rate": 2.2937851874203766e-06, + "loss": 2.4521, + "step": 25555 + }, + { + "epoch": 2.702622673434856, + "grad_norm": 8.4375, + "learning_rate": 2.2936145152834284e-06, + "loss": 2.7344, + "step": 25556 + }, + { + "epoch": 2.702728426395939, + "grad_norm": 8.3125, + "learning_rate": 2.2934438441149624e-06, + "loss": 2.5212, + "step": 25557 + }, + { + "epoch": 2.702834179357022, + "grad_norm": 8.5625, + "learning_rate": 2.2932731739157817e-06, + "loss": 2.5162, + "step": 25558 + }, + { + "epoch": 2.702939932318105, + "grad_norm": 8.25, + "learning_rate": 2.2931025046866836e-06, + "loss": 2.4226, + "step": 25559 + }, + { + "epoch": 2.703045685279188, + "grad_norm": 9.1875, + "learning_rate": 2.292931836428472e-06, + "loss": 2.5975, + "step": 25560 + }, + { + "epoch": 2.703151438240271, + "grad_norm": 8.0625, + "learning_rate": 2.2927611691419457e-06, + "loss": 2.4965, + "step": 25561 + }, + { + "epoch": 2.7032571912013537, + "grad_norm": 8.3125, + "learning_rate": 2.2925905028279073e-06, + "loss": 2.5894, + "step": 25562 + }, + { + "epoch": 2.7033629441624365, + "grad_norm": 8.3125, + "learning_rate": 2.292419837487156e-06, + "loss": 2.3408, + "step": 25563 + }, + { + "epoch": 2.7034686971235193, + "grad_norm": 8.0625, + "learning_rate": 2.2922491731204945e-06, + "loss": 2.3995, + "step": 25564 + }, + { + "epoch": 2.703574450084602, + "grad_norm": 8.6875, + "learning_rate": 2.2920785097287214e-06, + "loss": 2.6657, + "step": 25565 + }, + { + "epoch": 2.7036802030456855, + "grad_norm": 8.875, + "learning_rate": 2.2919078473126392e-06, + "loss": 2.7327, + "step": 25566 + }, + { + "epoch": 2.7037859560067683, + "grad_norm": 8.625, + "learning_rate": 2.2917371858730493e-06, + "loss": 2.6021, + "step": 25567 + }, + { + "epoch": 2.703891708967851, + "grad_norm": 8.625, + "learning_rate": 2.2915665254107506e-06, + "loss": 2.5792, + "step": 25568 + }, + { + "epoch": 2.703997461928934, + "grad_norm": 8.8125, + "learning_rate": 2.2913958659265455e-06, + "loss": 2.5223, + "step": 25569 + }, + { + "epoch": 2.704103214890017, + "grad_norm": 8.8125, + "learning_rate": 2.291225207421234e-06, + "loss": 2.5047, + "step": 25570 + }, + { + "epoch": 2.7042089678511, + "grad_norm": 8.25, + "learning_rate": 2.2910545498956176e-06, + "loss": 2.5459, + "step": 25571 + }, + { + "epoch": 2.7043147208121825, + "grad_norm": 8.25, + "learning_rate": 2.290883893350496e-06, + "loss": 2.4649, + "step": 25572 + }, + { + "epoch": 2.7044204737732658, + "grad_norm": 8.1875, + "learning_rate": 2.290713237786672e-06, + "loss": 2.5291, + "step": 25573 + }, + { + "epoch": 2.7045262267343486, + "grad_norm": 8.6875, + "learning_rate": 2.290542583204944e-06, + "loss": 2.6267, + "step": 25574 + }, + { + "epoch": 2.7046319796954315, + "grad_norm": 8.75, + "learning_rate": 2.2903719296061146e-06, + "loss": 2.5061, + "step": 25575 + }, + { + "epoch": 2.7047377326565143, + "grad_norm": 8.5625, + "learning_rate": 2.2902012769909837e-06, + "loss": 2.4099, + "step": 25576 + }, + { + "epoch": 2.704843485617597, + "grad_norm": 7.96875, + "learning_rate": 2.290030625360352e-06, + "loss": 2.4166, + "step": 25577 + }, + { + "epoch": 2.7049492385786804, + "grad_norm": 8.125, + "learning_rate": 2.2898599747150218e-06, + "loss": 2.5412, + "step": 25578 + }, + { + "epoch": 2.7050549915397633, + "grad_norm": 8.1875, + "learning_rate": 2.2896893250557915e-06, + "loss": 2.5494, + "step": 25579 + }, + { + "epoch": 2.705160744500846, + "grad_norm": 7.71875, + "learning_rate": 2.289518676383465e-06, + "loss": 2.4132, + "step": 25580 + }, + { + "epoch": 2.705266497461929, + "grad_norm": 9.1875, + "learning_rate": 2.2893480286988394e-06, + "loss": 2.5993, + "step": 25581 + }, + { + "epoch": 2.7053722504230118, + "grad_norm": 8.125, + "learning_rate": 2.289177382002719e-06, + "loss": 2.448, + "step": 25582 + }, + { + "epoch": 2.7054780033840946, + "grad_norm": 8.375, + "learning_rate": 2.2890067362959015e-06, + "loss": 2.5138, + "step": 25583 + }, + { + "epoch": 2.7055837563451774, + "grad_norm": 7.9375, + "learning_rate": 2.2888360915791894e-06, + "loss": 2.5496, + "step": 25584 + }, + { + "epoch": 2.7056895093062607, + "grad_norm": 8.75, + "learning_rate": 2.2886654478533835e-06, + "loss": 2.6466, + "step": 25585 + }, + { + "epoch": 2.7057952622673436, + "grad_norm": 8.25, + "learning_rate": 2.2884948051192837e-06, + "loss": 2.496, + "step": 25586 + }, + { + "epoch": 2.7059010152284264, + "grad_norm": 8.3125, + "learning_rate": 2.288324163377692e-06, + "loss": 2.5472, + "step": 25587 + }, + { + "epoch": 2.7060067681895092, + "grad_norm": 8.3125, + "learning_rate": 2.288153522629408e-06, + "loss": 2.4968, + "step": 25588 + }, + { + "epoch": 2.706112521150592, + "grad_norm": 8.125, + "learning_rate": 2.2879828828752335e-06, + "loss": 2.4432, + "step": 25589 + }, + { + "epoch": 2.7062182741116754, + "grad_norm": 8.6875, + "learning_rate": 2.287812244115968e-06, + "loss": 2.5781, + "step": 25590 + }, + { + "epoch": 2.7063240270727578, + "grad_norm": 8.25, + "learning_rate": 2.287641606352413e-06, + "loss": 2.29, + "step": 25591 + }, + { + "epoch": 2.706429780033841, + "grad_norm": 9.25, + "learning_rate": 2.2874709695853685e-06, + "loss": 2.5506, + "step": 25592 + }, + { + "epoch": 2.706535532994924, + "grad_norm": 7.875, + "learning_rate": 2.2873003338156363e-06, + "loss": 2.3919, + "step": 25593 + }, + { + "epoch": 2.7066412859560067, + "grad_norm": 8.875, + "learning_rate": 2.287129699044017e-06, + "loss": 2.9134, + "step": 25594 + }, + { + "epoch": 2.7067470389170896, + "grad_norm": 8.75, + "learning_rate": 2.28695906527131e-06, + "loss": 2.5777, + "step": 25595 + }, + { + "epoch": 2.7068527918781724, + "grad_norm": 8.4375, + "learning_rate": 2.2867884324983186e-06, + "loss": 2.7396, + "step": 25596 + }, + { + "epoch": 2.7069585448392557, + "grad_norm": 8.375, + "learning_rate": 2.2866178007258404e-06, + "loss": 2.0633, + "step": 25597 + }, + { + "epoch": 2.7070642978003385, + "grad_norm": 7.71875, + "learning_rate": 2.2864471699546787e-06, + "loss": 2.4232, + "step": 25598 + }, + { + "epoch": 2.7071700507614214, + "grad_norm": 8.3125, + "learning_rate": 2.286276540185632e-06, + "loss": 2.55, + "step": 25599 + }, + { + "epoch": 2.707275803722504, + "grad_norm": 8.25, + "learning_rate": 2.286105911419503e-06, + "loss": 2.5408, + "step": 25600 + }, + { + "epoch": 2.707381556683587, + "grad_norm": 8.4375, + "learning_rate": 2.2859352836570907e-06, + "loss": 2.4864, + "step": 25601 + }, + { + "epoch": 2.7074873096446703, + "grad_norm": 8.625, + "learning_rate": 2.285764656899197e-06, + "loss": 2.581, + "step": 25602 + }, + { + "epoch": 2.7075930626057527, + "grad_norm": 8.3125, + "learning_rate": 2.2855940311466216e-06, + "loss": 2.5189, + "step": 25603 + }, + { + "epoch": 2.707698815566836, + "grad_norm": 8.8125, + "learning_rate": 2.285423406400166e-06, + "loss": 2.6144, + "step": 25604 + }, + { + "epoch": 2.707804568527919, + "grad_norm": 7.96875, + "learning_rate": 2.285252782660632e-06, + "loss": 2.6308, + "step": 25605 + }, + { + "epoch": 2.7079103214890017, + "grad_norm": 8.375, + "learning_rate": 2.285082159928817e-06, + "loss": 2.6015, + "step": 25606 + }, + { + "epoch": 2.7080160744500845, + "grad_norm": 9.0, + "learning_rate": 2.2849115382055246e-06, + "loss": 2.6023, + "step": 25607 + }, + { + "epoch": 2.7081218274111674, + "grad_norm": 8.25, + "learning_rate": 2.2847409174915535e-06, + "loss": 2.5219, + "step": 25608 + }, + { + "epoch": 2.7082275803722506, + "grad_norm": 8.5625, + "learning_rate": 2.284570297787706e-06, + "loss": 2.4448, + "step": 25609 + }, + { + "epoch": 2.7083333333333335, + "grad_norm": 8.75, + "learning_rate": 2.2843996790947816e-06, + "loss": 2.4516, + "step": 25610 + }, + { + "epoch": 2.7084390862944163, + "grad_norm": 8.1875, + "learning_rate": 2.2842290614135807e-06, + "loss": 2.5582, + "step": 25611 + }, + { + "epoch": 2.708544839255499, + "grad_norm": 8.1875, + "learning_rate": 2.284058444744906e-06, + "loss": 2.653, + "step": 25612 + }, + { + "epoch": 2.708650592216582, + "grad_norm": 8.0, + "learning_rate": 2.2838878290895556e-06, + "loss": 2.4185, + "step": 25613 + }, + { + "epoch": 2.708756345177665, + "grad_norm": 8.0625, + "learning_rate": 2.2837172144483322e-06, + "loss": 2.4015, + "step": 25614 + }, + { + "epoch": 2.7088620981387477, + "grad_norm": 8.625, + "learning_rate": 2.2835466008220348e-06, + "loss": 2.517, + "step": 25615 + }, + { + "epoch": 2.708967851099831, + "grad_norm": 8.3125, + "learning_rate": 2.283375988211465e-06, + "loss": 2.5521, + "step": 25616 + }, + { + "epoch": 2.709073604060914, + "grad_norm": 8.125, + "learning_rate": 2.2832053766174223e-06, + "loss": 2.5436, + "step": 25617 + }, + { + "epoch": 2.7091793570219966, + "grad_norm": 8.375, + "learning_rate": 2.2830347660407094e-06, + "loss": 2.3454, + "step": 25618 + }, + { + "epoch": 2.7092851099830795, + "grad_norm": 8.1875, + "learning_rate": 2.282864156482124e-06, + "loss": 2.5009, + "step": 25619 + }, + { + "epoch": 2.7093908629441623, + "grad_norm": 9.0625, + "learning_rate": 2.2826935479424687e-06, + "loss": 2.8383, + "step": 25620 + }, + { + "epoch": 2.7094966159052456, + "grad_norm": 8.625, + "learning_rate": 2.2825229404225453e-06, + "loss": 2.6386, + "step": 25621 + }, + { + "epoch": 2.709602368866328, + "grad_norm": 8.25, + "learning_rate": 2.2823523339231513e-06, + "loss": 2.4687, + "step": 25622 + }, + { + "epoch": 2.7097081218274113, + "grad_norm": 8.5, + "learning_rate": 2.2821817284450893e-06, + "loss": 2.5087, + "step": 25623 + }, + { + "epoch": 2.709813874788494, + "grad_norm": 8.5, + "learning_rate": 2.282011123989159e-06, + "loss": 2.6008, + "step": 25624 + }, + { + "epoch": 2.709919627749577, + "grad_norm": 8.375, + "learning_rate": 2.2818405205561618e-06, + "loss": 2.6249, + "step": 25625 + }, + { + "epoch": 2.71002538071066, + "grad_norm": 8.5625, + "learning_rate": 2.281669918146897e-06, + "loss": 2.6577, + "step": 25626 + }, + { + "epoch": 2.7101311336717426, + "grad_norm": 8.25, + "learning_rate": 2.2814993167621676e-06, + "loss": 2.6486, + "step": 25627 + }, + { + "epoch": 2.710236886632826, + "grad_norm": 8.5, + "learning_rate": 2.281328716402771e-06, + "loss": 2.5119, + "step": 25628 + }, + { + "epoch": 2.7103426395939088, + "grad_norm": 8.125, + "learning_rate": 2.2811581170695095e-06, + "loss": 2.5147, + "step": 25629 + }, + { + "epoch": 2.7104483925549916, + "grad_norm": 8.75, + "learning_rate": 2.280987518763184e-06, + "loss": 2.6395, + "step": 25630 + }, + { + "epoch": 2.7105541455160744, + "grad_norm": 8.875, + "learning_rate": 2.2808169214845936e-06, + "loss": 2.5882, + "step": 25631 + }, + { + "epoch": 2.7106598984771573, + "grad_norm": 8.8125, + "learning_rate": 2.280646325234541e-06, + "loss": 2.5889, + "step": 25632 + }, + { + "epoch": 2.71076565143824, + "grad_norm": 8.0625, + "learning_rate": 2.2804757300138243e-06, + "loss": 2.4396, + "step": 25633 + }, + { + "epoch": 2.710871404399323, + "grad_norm": 8.6875, + "learning_rate": 2.2803051358232467e-06, + "loss": 2.6174, + "step": 25634 + }, + { + "epoch": 2.7109771573604062, + "grad_norm": 9.125, + "learning_rate": 2.2801345426636056e-06, + "loss": 2.4521, + "step": 25635 + }, + { + "epoch": 2.711082910321489, + "grad_norm": 8.5625, + "learning_rate": 2.279963950535705e-06, + "loss": 2.6016, + "step": 25636 + }, + { + "epoch": 2.711188663282572, + "grad_norm": 8.4375, + "learning_rate": 2.279793359440342e-06, + "loss": 2.5968, + "step": 25637 + }, + { + "epoch": 2.7112944162436547, + "grad_norm": 8.875, + "learning_rate": 2.2796227693783184e-06, + "loss": 2.5716, + "step": 25638 + }, + { + "epoch": 2.7114001692047376, + "grad_norm": 8.1875, + "learning_rate": 2.2794521803504363e-06, + "loss": 2.6525, + "step": 25639 + }, + { + "epoch": 2.711505922165821, + "grad_norm": 8.25, + "learning_rate": 2.279281592357494e-06, + "loss": 2.6921, + "step": 25640 + }, + { + "epoch": 2.7116116751269037, + "grad_norm": 8.25, + "learning_rate": 2.2791110054002937e-06, + "loss": 2.5136, + "step": 25641 + }, + { + "epoch": 2.7117174280879865, + "grad_norm": 8.375, + "learning_rate": 2.2789404194796346e-06, + "loss": 2.6139, + "step": 25642 + }, + { + "epoch": 2.7118231810490694, + "grad_norm": 9.4375, + "learning_rate": 2.278769834596318e-06, + "loss": 2.4353, + "step": 25643 + }, + { + "epoch": 2.7119289340101522, + "grad_norm": 9.3125, + "learning_rate": 2.278599250751144e-06, + "loss": 2.6882, + "step": 25644 + }, + { + "epoch": 2.712034686971235, + "grad_norm": 8.375, + "learning_rate": 2.278428667944913e-06, + "loss": 2.6702, + "step": 25645 + }, + { + "epoch": 2.712140439932318, + "grad_norm": 8.4375, + "learning_rate": 2.278258086178426e-06, + "loss": 2.132, + "step": 25646 + }, + { + "epoch": 2.712246192893401, + "grad_norm": 8.375, + "learning_rate": 2.278087505452482e-06, + "loss": 2.5548, + "step": 25647 + }, + { + "epoch": 2.712351945854484, + "grad_norm": 8.6875, + "learning_rate": 2.277916925767884e-06, + "loss": 2.6317, + "step": 25648 + }, + { + "epoch": 2.712457698815567, + "grad_norm": 8.875, + "learning_rate": 2.2777463471254302e-06, + "loss": 2.4775, + "step": 25649 + }, + { + "epoch": 2.7125634517766497, + "grad_norm": 8.6875, + "learning_rate": 2.277575769525923e-06, + "loss": 2.6476, + "step": 25650 + }, + { + "epoch": 2.7126692047377325, + "grad_norm": 8.5, + "learning_rate": 2.27740519297016e-06, + "loss": 2.3542, + "step": 25651 + }, + { + "epoch": 2.712774957698816, + "grad_norm": 8.0, + "learning_rate": 2.277234617458945e-06, + "loss": 2.4047, + "step": 25652 + }, + { + "epoch": 2.712880710659898, + "grad_norm": 9.5, + "learning_rate": 2.277064042993076e-06, + "loss": 2.6346, + "step": 25653 + }, + { + "epoch": 2.7129864636209815, + "grad_norm": 8.4375, + "learning_rate": 2.2768934695733546e-06, + "loss": 2.4604, + "step": 25654 + }, + { + "epoch": 2.7130922165820643, + "grad_norm": 8.75, + "learning_rate": 2.2767228972005805e-06, + "loss": 2.4889, + "step": 25655 + }, + { + "epoch": 2.713197969543147, + "grad_norm": 8.3125, + "learning_rate": 2.2765523258755544e-06, + "loss": 2.3107, + "step": 25656 + }, + { + "epoch": 2.71330372250423, + "grad_norm": 8.6875, + "learning_rate": 2.276381755599077e-06, + "loss": 2.593, + "step": 25657 + }, + { + "epoch": 2.713409475465313, + "grad_norm": 8.5625, + "learning_rate": 2.2762111863719487e-06, + "loss": 2.5492, + "step": 25658 + }, + { + "epoch": 2.713515228426396, + "grad_norm": 9.3125, + "learning_rate": 2.2760406181949705e-06, + "loss": 2.644, + "step": 25659 + }, + { + "epoch": 2.713620981387479, + "grad_norm": 8.5625, + "learning_rate": 2.275870051068941e-06, + "loss": 2.7297, + "step": 25660 + }, + { + "epoch": 2.713726734348562, + "grad_norm": 8.25, + "learning_rate": 2.2756994849946622e-06, + "loss": 2.6769, + "step": 25661 + }, + { + "epoch": 2.7138324873096447, + "grad_norm": 8.3125, + "learning_rate": 2.275528919972934e-06, + "loss": 2.5057, + "step": 25662 + }, + { + "epoch": 2.7139382402707275, + "grad_norm": 8.875, + "learning_rate": 2.275358356004556e-06, + "loss": 2.6476, + "step": 25663 + }, + { + "epoch": 2.7140439932318103, + "grad_norm": 8.3125, + "learning_rate": 2.2751877930903305e-06, + "loss": 2.6159, + "step": 25664 + }, + { + "epoch": 2.714149746192893, + "grad_norm": 8.0, + "learning_rate": 2.2750172312310555e-06, + "loss": 2.3475, + "step": 25665 + }, + { + "epoch": 2.7142554991539765, + "grad_norm": 8.5625, + "learning_rate": 2.2748466704275343e-06, + "loss": 2.5971, + "step": 25666 + }, + { + "epoch": 2.7143612521150593, + "grad_norm": 8.5, + "learning_rate": 2.2746761106805642e-06, + "loss": 2.7801, + "step": 25667 + }, + { + "epoch": 2.714467005076142, + "grad_norm": 7.84375, + "learning_rate": 2.274505551990948e-06, + "loss": 2.4839, + "step": 25668 + }, + { + "epoch": 2.714572758037225, + "grad_norm": 8.1875, + "learning_rate": 2.274334994359484e-06, + "loss": 2.4866, + "step": 25669 + }, + { + "epoch": 2.714678510998308, + "grad_norm": 8.75, + "learning_rate": 2.2741644377869743e-06, + "loss": 2.5476, + "step": 25670 + }, + { + "epoch": 2.714784263959391, + "grad_norm": 8.0625, + "learning_rate": 2.273993882274218e-06, + "loss": 2.548, + "step": 25671 + }, + { + "epoch": 2.714890016920474, + "grad_norm": 9.375, + "learning_rate": 2.2738233278220156e-06, + "loss": 2.3499, + "step": 25672 + }, + { + "epoch": 2.7149957698815568, + "grad_norm": 9.0, + "learning_rate": 2.273652774431169e-06, + "loss": 2.7196, + "step": 25673 + }, + { + "epoch": 2.7151015228426396, + "grad_norm": 8.75, + "learning_rate": 2.273482222102476e-06, + "loss": 2.5393, + "step": 25674 + }, + { + "epoch": 2.7152072758037225, + "grad_norm": 8.625, + "learning_rate": 2.27331167083674e-06, + "loss": 2.8062, + "step": 25675 + }, + { + "epoch": 2.7153130287648053, + "grad_norm": 8.4375, + "learning_rate": 2.2731411206347584e-06, + "loss": 2.4237, + "step": 25676 + }, + { + "epoch": 2.715418781725888, + "grad_norm": 8.375, + "learning_rate": 2.2729705714973334e-06, + "loss": 2.6574, + "step": 25677 + }, + { + "epoch": 2.7155245346869714, + "grad_norm": 8.375, + "learning_rate": 2.272800023425264e-06, + "loss": 2.5142, + "step": 25678 + }, + { + "epoch": 2.7156302876480543, + "grad_norm": 8.375, + "learning_rate": 2.2726294764193514e-06, + "loss": 2.161, + "step": 25679 + }, + { + "epoch": 2.715736040609137, + "grad_norm": 8.25, + "learning_rate": 2.2724589304803957e-06, + "loss": 2.2555, + "step": 25680 + }, + { + "epoch": 2.71584179357022, + "grad_norm": 8.4375, + "learning_rate": 2.2722883856091966e-06, + "loss": 2.5461, + "step": 25681 + }, + { + "epoch": 2.7159475465313028, + "grad_norm": 8.625, + "learning_rate": 2.2721178418065567e-06, + "loss": 2.717, + "step": 25682 + }, + { + "epoch": 2.716053299492386, + "grad_norm": 8.4375, + "learning_rate": 2.271947299073273e-06, + "loss": 2.534, + "step": 25683 + }, + { + "epoch": 2.7161590524534684, + "grad_norm": 8.4375, + "learning_rate": 2.271776757410148e-06, + "loss": 2.5269, + "step": 25684 + }, + { + "epoch": 2.7162648054145517, + "grad_norm": 8.875, + "learning_rate": 2.271606216817981e-06, + "loss": 2.4126, + "step": 25685 + }, + { + "epoch": 2.7163705583756346, + "grad_norm": 8.4375, + "learning_rate": 2.271435677297573e-06, + "loss": 2.5896, + "step": 25686 + }, + { + "epoch": 2.7164763113367174, + "grad_norm": 8.9375, + "learning_rate": 2.2712651388497233e-06, + "loss": 2.8161, + "step": 25687 + }, + { + "epoch": 2.7165820642978002, + "grad_norm": 8.1875, + "learning_rate": 2.2710946014752337e-06, + "loss": 2.6105, + "step": 25688 + }, + { + "epoch": 2.716687817258883, + "grad_norm": 7.875, + "learning_rate": 2.2709240651749022e-06, + "loss": 2.4618, + "step": 25689 + }, + { + "epoch": 2.7167935702199664, + "grad_norm": 7.78125, + "learning_rate": 2.2707535299495305e-06, + "loss": 2.3969, + "step": 25690 + }, + { + "epoch": 2.716899323181049, + "grad_norm": 8.25, + "learning_rate": 2.2705829957999203e-06, + "loss": 2.6464, + "step": 25691 + }, + { + "epoch": 2.717005076142132, + "grad_norm": 8.6875, + "learning_rate": 2.270412462726869e-06, + "loss": 2.5209, + "step": 25692 + }, + { + "epoch": 2.717110829103215, + "grad_norm": 8.25, + "learning_rate": 2.270241930731179e-06, + "loss": 2.454, + "step": 25693 + }, + { + "epoch": 2.7172165820642977, + "grad_norm": 8.4375, + "learning_rate": 2.270071399813648e-06, + "loss": 2.6748, + "step": 25694 + }, + { + "epoch": 2.7173223350253806, + "grad_norm": 8.0625, + "learning_rate": 2.2699008699750795e-06, + "loss": 2.5711, + "step": 25695 + }, + { + "epoch": 2.7174280879864634, + "grad_norm": 8.0625, + "learning_rate": 2.2697303412162715e-06, + "loss": 2.1652, + "step": 25696 + }, + { + "epoch": 2.7175338409475467, + "grad_norm": 8.5, + "learning_rate": 2.2695598135380252e-06, + "loss": 2.2838, + "step": 25697 + }, + { + "epoch": 2.7176395939086295, + "grad_norm": 8.0625, + "learning_rate": 2.26938928694114e-06, + "loss": 2.4, + "step": 25698 + }, + { + "epoch": 2.7177453468697124, + "grad_norm": 9.4375, + "learning_rate": 2.2692187614264163e-06, + "loss": 2.5675, + "step": 25699 + }, + { + "epoch": 2.717851099830795, + "grad_norm": 8.5, + "learning_rate": 2.2690482369946554e-06, + "loss": 2.7787, + "step": 25700 + }, + { + "epoch": 2.717956852791878, + "grad_norm": 8.6875, + "learning_rate": 2.268877713646656e-06, + "loss": 2.2601, + "step": 25701 + }, + { + "epoch": 2.7180626057529613, + "grad_norm": 8.25, + "learning_rate": 2.2687071913832192e-06, + "loss": 2.5333, + "step": 25702 + }, + { + "epoch": 2.7181683587140437, + "grad_norm": 8.375, + "learning_rate": 2.268536670205145e-06, + "loss": 2.4305, + "step": 25703 + }, + { + "epoch": 2.718274111675127, + "grad_norm": 8.5, + "learning_rate": 2.268366150113234e-06, + "loss": 2.5813, + "step": 25704 + }, + { + "epoch": 2.71837986463621, + "grad_norm": 8.75, + "learning_rate": 2.2681956311082843e-06, + "loss": 2.5136, + "step": 25705 + }, + { + "epoch": 2.7184856175972927, + "grad_norm": 8.5625, + "learning_rate": 2.2680251131910996e-06, + "loss": 2.6307, + "step": 25706 + }, + { + "epoch": 2.7185913705583755, + "grad_norm": 8.1875, + "learning_rate": 2.2678545963624776e-06, + "loss": 2.4531, + "step": 25707 + }, + { + "epoch": 2.7186971235194584, + "grad_norm": 8.5625, + "learning_rate": 2.2676840806232182e-06, + "loss": 2.6787, + "step": 25708 + }, + { + "epoch": 2.7188028764805416, + "grad_norm": 8.3125, + "learning_rate": 2.2675135659741233e-06, + "loss": 2.7592, + "step": 25709 + }, + { + "epoch": 2.7189086294416245, + "grad_norm": 9.8125, + "learning_rate": 2.2673430524159916e-06, + "loss": 2.5043, + "step": 25710 + }, + { + "epoch": 2.7190143824027073, + "grad_norm": 7.84375, + "learning_rate": 2.267172539949624e-06, + "loss": 2.3948, + "step": 25711 + }, + { + "epoch": 2.71912013536379, + "grad_norm": 8.1875, + "learning_rate": 2.2670020285758204e-06, + "loss": 2.4677, + "step": 25712 + }, + { + "epoch": 2.719225888324873, + "grad_norm": 8.5, + "learning_rate": 2.266831518295382e-06, + "loss": 2.5344, + "step": 25713 + }, + { + "epoch": 2.7193316412859563, + "grad_norm": 8.5, + "learning_rate": 2.2666610091091065e-06, + "loss": 2.3352, + "step": 25714 + }, + { + "epoch": 2.7194373942470387, + "grad_norm": 8.375, + "learning_rate": 2.266490501017796e-06, + "loss": 2.6652, + "step": 25715 + }, + { + "epoch": 2.719543147208122, + "grad_norm": 8.125, + "learning_rate": 2.26631999402225e-06, + "loss": 2.5276, + "step": 25716 + }, + { + "epoch": 2.719648900169205, + "grad_norm": 8.625, + "learning_rate": 2.266149488123268e-06, + "loss": 2.5694, + "step": 25717 + }, + { + "epoch": 2.7197546531302876, + "grad_norm": 8.0, + "learning_rate": 2.2659789833216524e-06, + "loss": 2.5435, + "step": 25718 + }, + { + "epoch": 2.7198604060913705, + "grad_norm": 8.6875, + "learning_rate": 2.2658084796182e-06, + "loss": 2.719, + "step": 25719 + }, + { + "epoch": 2.7199661590524533, + "grad_norm": 8.125, + "learning_rate": 2.2656379770137147e-06, + "loss": 2.5484, + "step": 25720 + }, + { + "epoch": 2.7200719120135366, + "grad_norm": 8.25, + "learning_rate": 2.2654674755089927e-06, + "loss": 2.5814, + "step": 25721 + }, + { + "epoch": 2.7201776649746194, + "grad_norm": 7.90625, + "learning_rate": 2.265296975104837e-06, + "loss": 2.5083, + "step": 25722 + }, + { + "epoch": 2.7202834179357023, + "grad_norm": 8.375, + "learning_rate": 2.265126475802046e-06, + "loss": 2.6511, + "step": 25723 + }, + { + "epoch": 2.720389170896785, + "grad_norm": 8.5, + "learning_rate": 2.2649559776014203e-06, + "loss": 2.5452, + "step": 25724 + }, + { + "epoch": 2.720494923857868, + "grad_norm": 8.6875, + "learning_rate": 2.26478548050376e-06, + "loss": 2.4337, + "step": 25725 + }, + { + "epoch": 2.720600676818951, + "grad_norm": 8.25, + "learning_rate": 2.264614984509865e-06, + "loss": 2.5202, + "step": 25726 + }, + { + "epoch": 2.7207064297800336, + "grad_norm": 8.6875, + "learning_rate": 2.2644444896205365e-06, + "loss": 2.8308, + "step": 25727 + }, + { + "epoch": 2.720812182741117, + "grad_norm": 8.0625, + "learning_rate": 2.2642739958365732e-06, + "loss": 2.3611, + "step": 25728 + }, + { + "epoch": 2.7209179357021998, + "grad_norm": 8.5625, + "learning_rate": 2.2641035031587765e-06, + "loss": 2.6772, + "step": 25729 + }, + { + "epoch": 2.7210236886632826, + "grad_norm": 8.125, + "learning_rate": 2.2639330115879447e-06, + "loss": 2.5411, + "step": 25730 + }, + { + "epoch": 2.7211294416243654, + "grad_norm": 8.1875, + "learning_rate": 2.263762521124879e-06, + "loss": 2.3345, + "step": 25731 + }, + { + "epoch": 2.7212351945854483, + "grad_norm": 8.4375, + "learning_rate": 2.2635920317703784e-06, + "loss": 2.6603, + "step": 25732 + }, + { + "epoch": 2.7213409475465316, + "grad_norm": 9.125, + "learning_rate": 2.263421543525245e-06, + "loss": 2.5681, + "step": 25733 + }, + { + "epoch": 2.721446700507614, + "grad_norm": 8.25, + "learning_rate": 2.2632510563902763e-06, + "loss": 2.5163, + "step": 25734 + }, + { + "epoch": 2.7215524534686972, + "grad_norm": 7.875, + "learning_rate": 2.263080570366274e-06, + "loss": 2.4539, + "step": 25735 + }, + { + "epoch": 2.72165820642978, + "grad_norm": 8.5625, + "learning_rate": 2.2629100854540385e-06, + "loss": 2.5078, + "step": 25736 + }, + { + "epoch": 2.721763959390863, + "grad_norm": 8.375, + "learning_rate": 2.262739601654368e-06, + "loss": 2.5736, + "step": 25737 + }, + { + "epoch": 2.7218697123519457, + "grad_norm": 10.125, + "learning_rate": 2.262569118968064e-06, + "loss": 2.5358, + "step": 25738 + }, + { + "epoch": 2.7219754653130286, + "grad_norm": 8.375, + "learning_rate": 2.2623986373959254e-06, + "loss": 2.3521, + "step": 25739 + }, + { + "epoch": 2.722081218274112, + "grad_norm": 8.5625, + "learning_rate": 2.2622281569387537e-06, + "loss": 2.4356, + "step": 25740 + }, + { + "epoch": 2.7221869712351947, + "grad_norm": 8.25, + "learning_rate": 2.2620576775973476e-06, + "loss": 2.4784, + "step": 25741 + }, + { + "epoch": 2.7222927241962775, + "grad_norm": 8.625, + "learning_rate": 2.261887199372508e-06, + "loss": 2.4582, + "step": 25742 + }, + { + "epoch": 2.7223984771573604, + "grad_norm": 8.125, + "learning_rate": 2.261716722265033e-06, + "loss": 2.3304, + "step": 25743 + }, + { + "epoch": 2.7225042301184432, + "grad_norm": 8.625, + "learning_rate": 2.2615462462757246e-06, + "loss": 2.589, + "step": 25744 + }, + { + "epoch": 2.722609983079526, + "grad_norm": 9.125, + "learning_rate": 2.261375771405383e-06, + "loss": 2.3035, + "step": 25745 + }, + { + "epoch": 2.722715736040609, + "grad_norm": 8.0625, + "learning_rate": 2.2612052976548065e-06, + "loss": 2.5919, + "step": 25746 + }, + { + "epoch": 2.722821489001692, + "grad_norm": 8.125, + "learning_rate": 2.261034825024796e-06, + "loss": 2.4274, + "step": 25747 + }, + { + "epoch": 2.722927241962775, + "grad_norm": 8.625, + "learning_rate": 2.2608643535161514e-06, + "loss": 2.4562, + "step": 25748 + }, + { + "epoch": 2.723032994923858, + "grad_norm": 8.4375, + "learning_rate": 2.260693883129673e-06, + "loss": 2.5478, + "step": 25749 + }, + { + "epoch": 2.7231387478849407, + "grad_norm": 8.6875, + "learning_rate": 2.2605234138661593e-06, + "loss": 2.4537, + "step": 25750 + }, + { + "epoch": 2.7232445008460235, + "grad_norm": 8.625, + "learning_rate": 2.2603529457264128e-06, + "loss": 2.4484, + "step": 25751 + }, + { + "epoch": 2.723350253807107, + "grad_norm": 8.1875, + "learning_rate": 2.260182478711231e-06, + "loss": 2.425, + "step": 25752 + }, + { + "epoch": 2.7234560067681897, + "grad_norm": 8.6875, + "learning_rate": 2.2600120128214143e-06, + "loss": 2.507, + "step": 25753 + }, + { + "epoch": 2.7235617597292725, + "grad_norm": 8.4375, + "learning_rate": 2.2598415480577637e-06, + "loss": 2.4809, + "step": 25754 + }, + { + "epoch": 2.7236675126903553, + "grad_norm": 8.4375, + "learning_rate": 2.2596710844210777e-06, + "loss": 2.5982, + "step": 25755 + }, + { + "epoch": 2.723773265651438, + "grad_norm": 8.375, + "learning_rate": 2.259500621912158e-06, + "loss": 2.6057, + "step": 25756 + }, + { + "epoch": 2.723879018612521, + "grad_norm": 9.1875, + "learning_rate": 2.259330160531803e-06, + "loss": 2.5385, + "step": 25757 + }, + { + "epoch": 2.723984771573604, + "grad_norm": 9.0625, + "learning_rate": 2.259159700280814e-06, + "loss": 2.5526, + "step": 25758 + }, + { + "epoch": 2.724090524534687, + "grad_norm": 7.6875, + "learning_rate": 2.2589892411599886e-06, + "loss": 2.5228, + "step": 25759 + }, + { + "epoch": 2.72419627749577, + "grad_norm": 8.5625, + "learning_rate": 2.2588187831701295e-06, + "loss": 2.6648, + "step": 25760 + }, + { + "epoch": 2.724302030456853, + "grad_norm": 9.625, + "learning_rate": 2.2586483263120343e-06, + "loss": 2.4894, + "step": 25761 + }, + { + "epoch": 2.7244077834179357, + "grad_norm": 8.4375, + "learning_rate": 2.258477870586504e-06, + "loss": 2.5852, + "step": 25762 + }, + { + "epoch": 2.7245135363790185, + "grad_norm": 8.6875, + "learning_rate": 2.2583074159943383e-06, + "loss": 2.444, + "step": 25763 + }, + { + "epoch": 2.724619289340102, + "grad_norm": 8.375, + "learning_rate": 2.2581369625363367e-06, + "loss": 2.6528, + "step": 25764 + }, + { + "epoch": 2.724725042301184, + "grad_norm": 8.875, + "learning_rate": 2.2579665102133e-06, + "loss": 2.3034, + "step": 25765 + }, + { + "epoch": 2.7248307952622675, + "grad_norm": 8.0625, + "learning_rate": 2.2577960590260268e-06, + "loss": 2.4281, + "step": 25766 + }, + { + "epoch": 2.7249365482233503, + "grad_norm": 9.25, + "learning_rate": 2.257625608975319e-06, + "loss": 2.7369, + "step": 25767 + }, + { + "epoch": 2.725042301184433, + "grad_norm": 8.3125, + "learning_rate": 2.2574551600619735e-06, + "loss": 2.4753, + "step": 25768 + }, + { + "epoch": 2.725148054145516, + "grad_norm": 8.6875, + "learning_rate": 2.257284712286793e-06, + "loss": 2.656, + "step": 25769 + }, + { + "epoch": 2.725253807106599, + "grad_norm": 8.1875, + "learning_rate": 2.257114265650575e-06, + "loss": 2.3543, + "step": 25770 + }, + { + "epoch": 2.725359560067682, + "grad_norm": 8.3125, + "learning_rate": 2.2569438201541207e-06, + "loss": 2.5264, + "step": 25771 + }, + { + "epoch": 2.725465313028765, + "grad_norm": 8.1875, + "learning_rate": 2.25677337579823e-06, + "loss": 2.4505, + "step": 25772 + }, + { + "epoch": 2.7255710659898478, + "grad_norm": 8.75, + "learning_rate": 2.256602932583702e-06, + "loss": 2.4487, + "step": 25773 + }, + { + "epoch": 2.7256768189509306, + "grad_norm": 9.0, + "learning_rate": 2.256432490511338e-06, + "loss": 2.9131, + "step": 25774 + }, + { + "epoch": 2.7257825719120135, + "grad_norm": 8.75, + "learning_rate": 2.2562620495819355e-06, + "loss": 2.6425, + "step": 25775 + }, + { + "epoch": 2.7258883248730963, + "grad_norm": 8.25, + "learning_rate": 2.256091609796296e-06, + "loss": 2.5755, + "step": 25776 + }, + { + "epoch": 2.725994077834179, + "grad_norm": 8.6875, + "learning_rate": 2.2559211711552185e-06, + "loss": 2.6558, + "step": 25777 + }, + { + "epoch": 2.7260998307952624, + "grad_norm": 8.8125, + "learning_rate": 2.2557507336595032e-06, + "loss": 2.4454, + "step": 25778 + }, + { + "epoch": 2.7262055837563453, + "grad_norm": 8.5625, + "learning_rate": 2.2555802973099497e-06, + "loss": 2.5959, + "step": 25779 + }, + { + "epoch": 2.726311336717428, + "grad_norm": 8.1875, + "learning_rate": 2.2554098621073576e-06, + "loss": 2.396, + "step": 25780 + }, + { + "epoch": 2.726417089678511, + "grad_norm": 8.6875, + "learning_rate": 2.2552394280525282e-06, + "loss": 2.4646, + "step": 25781 + }, + { + "epoch": 2.7265228426395938, + "grad_norm": 8.125, + "learning_rate": 2.2550689951462585e-06, + "loss": 2.4211, + "step": 25782 + }, + { + "epoch": 2.726628595600677, + "grad_norm": 8.3125, + "learning_rate": 2.2548985633893516e-06, + "loss": 2.7145, + "step": 25783 + }, + { + "epoch": 2.72673434856176, + "grad_norm": 8.5, + "learning_rate": 2.2547281327826044e-06, + "loss": 2.7376, + "step": 25784 + }, + { + "epoch": 2.7268401015228427, + "grad_norm": 9.1875, + "learning_rate": 2.2545577033268183e-06, + "loss": 2.8199, + "step": 25785 + }, + { + "epoch": 2.7269458544839256, + "grad_norm": 8.25, + "learning_rate": 2.254387275022792e-06, + "loss": 2.4351, + "step": 25786 + }, + { + "epoch": 2.7270516074450084, + "grad_norm": 8.4375, + "learning_rate": 2.254216847871325e-06, + "loss": 2.5049, + "step": 25787 + }, + { + "epoch": 2.7271573604060912, + "grad_norm": 8.1875, + "learning_rate": 2.2540464218732195e-06, + "loss": 2.481, + "step": 25788 + }, + { + "epoch": 2.727263113367174, + "grad_norm": 8.875, + "learning_rate": 2.2538759970292724e-06, + "loss": 2.2951, + "step": 25789 + }, + { + "epoch": 2.7273688663282574, + "grad_norm": 8.125, + "learning_rate": 2.2537055733402857e-06, + "loss": 2.5032, + "step": 25790 + }, + { + "epoch": 2.72747461928934, + "grad_norm": 9.0625, + "learning_rate": 2.2535351508070567e-06, + "loss": 2.3713, + "step": 25791 + }, + { + "epoch": 2.727580372250423, + "grad_norm": 9.1875, + "learning_rate": 2.2533647294303873e-06, + "loss": 2.5234, + "step": 25792 + }, + { + "epoch": 2.727686125211506, + "grad_norm": 8.375, + "learning_rate": 2.253194309211076e-06, + "loss": 2.4699, + "step": 25793 + }, + { + "epoch": 2.7277918781725887, + "grad_norm": 8.375, + "learning_rate": 2.253023890149923e-06, + "loss": 2.6166, + "step": 25794 + }, + { + "epoch": 2.727897631133672, + "grad_norm": 8.375, + "learning_rate": 2.2528534722477273e-06, + "loss": 2.5899, + "step": 25795 + }, + { + "epoch": 2.7280033840947544, + "grad_norm": 8.6875, + "learning_rate": 2.2526830555052896e-06, + "loss": 2.5871, + "step": 25796 + }, + { + "epoch": 2.7281091370558377, + "grad_norm": 8.3125, + "learning_rate": 2.2525126399234093e-06, + "loss": 2.3323, + "step": 25797 + }, + { + "epoch": 2.7282148900169205, + "grad_norm": 8.6875, + "learning_rate": 2.2523422255028858e-06, + "loss": 2.584, + "step": 25798 + }, + { + "epoch": 2.7283206429780034, + "grad_norm": 8.75, + "learning_rate": 2.25217181224452e-06, + "loss": 2.7283, + "step": 25799 + }, + { + "epoch": 2.728426395939086, + "grad_norm": 8.4375, + "learning_rate": 2.252001400149109e-06, + "loss": 2.6574, + "step": 25800 + }, + { + "epoch": 2.728532148900169, + "grad_norm": 8.375, + "learning_rate": 2.251830989217455e-06, + "loss": 2.2836, + "step": 25801 + }, + { + "epoch": 2.7286379018612523, + "grad_norm": 8.5625, + "learning_rate": 2.2516605794503558e-06, + "loss": 2.7032, + "step": 25802 + }, + { + "epoch": 2.728743654822335, + "grad_norm": 8.0625, + "learning_rate": 2.2514901708486128e-06, + "loss": 2.5157, + "step": 25803 + }, + { + "epoch": 2.728849407783418, + "grad_norm": 8.3125, + "learning_rate": 2.2513197634130236e-06, + "loss": 2.6132, + "step": 25804 + }, + { + "epoch": 2.728955160744501, + "grad_norm": 8.4375, + "learning_rate": 2.25114935714439e-06, + "loss": 2.4915, + "step": 25805 + }, + { + "epoch": 2.7290609137055837, + "grad_norm": 9.5, + "learning_rate": 2.250978952043511e-06, + "loss": 2.6221, + "step": 25806 + }, + { + "epoch": 2.7291666666666665, + "grad_norm": 9.4375, + "learning_rate": 2.250808548111185e-06, + "loss": 2.8176, + "step": 25807 + }, + { + "epoch": 2.7292724196277494, + "grad_norm": 9.0, + "learning_rate": 2.2506381453482134e-06, + "loss": 2.4075, + "step": 25808 + }, + { + "epoch": 2.7293781725888326, + "grad_norm": 8.4375, + "learning_rate": 2.2504677437553944e-06, + "loss": 2.5627, + "step": 25809 + }, + { + "epoch": 2.7294839255499155, + "grad_norm": 8.3125, + "learning_rate": 2.2502973433335286e-06, + "loss": 2.4731, + "step": 25810 + }, + { + "epoch": 2.7295896785109983, + "grad_norm": 8.25, + "learning_rate": 2.2501269440834147e-06, + "loss": 2.6103, + "step": 25811 + }, + { + "epoch": 2.729695431472081, + "grad_norm": 8.4375, + "learning_rate": 2.249956546005854e-06, + "loss": 2.5213, + "step": 25812 + }, + { + "epoch": 2.729801184433164, + "grad_norm": 8.3125, + "learning_rate": 2.249786149101644e-06, + "loss": 2.4099, + "step": 25813 + }, + { + "epoch": 2.7299069373942473, + "grad_norm": 8.125, + "learning_rate": 2.2496157533715844e-06, + "loss": 2.4584, + "step": 25814 + }, + { + "epoch": 2.7300126903553297, + "grad_norm": 8.25, + "learning_rate": 2.249445358816478e-06, + "loss": 2.58, + "step": 25815 + }, + { + "epoch": 2.730118443316413, + "grad_norm": 8.75, + "learning_rate": 2.24927496543712e-06, + "loss": 2.707, + "step": 25816 + }, + { + "epoch": 2.730224196277496, + "grad_norm": 9.9375, + "learning_rate": 2.249104573234313e-06, + "loss": 2.5031, + "step": 25817 + }, + { + "epoch": 2.7303299492385786, + "grad_norm": 8.1875, + "learning_rate": 2.2489341822088553e-06, + "loss": 2.6668, + "step": 25818 + }, + { + "epoch": 2.7304357021996615, + "grad_norm": 8.3125, + "learning_rate": 2.2487637923615475e-06, + "loss": 2.556, + "step": 25819 + }, + { + "epoch": 2.7305414551607443, + "grad_norm": 8.9375, + "learning_rate": 2.2485934036931875e-06, + "loss": 2.8316, + "step": 25820 + }, + { + "epoch": 2.7306472081218276, + "grad_norm": 8.0, + "learning_rate": 2.248423016204577e-06, + "loss": 2.6053, + "step": 25821 + }, + { + "epoch": 2.7307529610829104, + "grad_norm": 8.0625, + "learning_rate": 2.2482526298965134e-06, + "loss": 2.4169, + "step": 25822 + }, + { + "epoch": 2.7308587140439933, + "grad_norm": 8.9375, + "learning_rate": 2.2480822447697968e-06, + "loss": 2.6168, + "step": 25823 + }, + { + "epoch": 2.730964467005076, + "grad_norm": 8.75, + "learning_rate": 2.2479118608252286e-06, + "loss": 2.5657, + "step": 25824 + }, + { + "epoch": 2.731070219966159, + "grad_norm": 8.5, + "learning_rate": 2.247741478063606e-06, + "loss": 2.5055, + "step": 25825 + }, + { + "epoch": 2.7311759729272422, + "grad_norm": 8.1875, + "learning_rate": 2.24757109648573e-06, + "loss": 2.4287, + "step": 25826 + }, + { + "epoch": 2.7312817258883246, + "grad_norm": 8.75, + "learning_rate": 2.2474007160923985e-06, + "loss": 2.6716, + "step": 25827 + }, + { + "epoch": 2.731387478849408, + "grad_norm": 8.875, + "learning_rate": 2.247230336884414e-06, + "loss": 2.7063, + "step": 25828 + }, + { + "epoch": 2.7314932318104908, + "grad_norm": 8.4375, + "learning_rate": 2.247059958862573e-06, + "loss": 2.2429, + "step": 25829 + }, + { + "epoch": 2.7315989847715736, + "grad_norm": 8.125, + "learning_rate": 2.246889582027676e-06, + "loss": 2.4722, + "step": 25830 + }, + { + "epoch": 2.7317047377326564, + "grad_norm": 8.75, + "learning_rate": 2.246719206380523e-06, + "loss": 2.518, + "step": 25831 + }, + { + "epoch": 2.7318104906937393, + "grad_norm": 8.6875, + "learning_rate": 2.2465488319219126e-06, + "loss": 2.573, + "step": 25832 + }, + { + "epoch": 2.7319162436548226, + "grad_norm": 8.5, + "learning_rate": 2.2463784586526454e-06, + "loss": 2.4948, + "step": 25833 + }, + { + "epoch": 2.7320219966159054, + "grad_norm": 8.25, + "learning_rate": 2.24620808657352e-06, + "loss": 2.4425, + "step": 25834 + }, + { + "epoch": 2.7321277495769882, + "grad_norm": 8.5625, + "learning_rate": 2.246037715685337e-06, + "loss": 2.6779, + "step": 25835 + }, + { + "epoch": 2.732233502538071, + "grad_norm": 8.1875, + "learning_rate": 2.2458673459888944e-06, + "loss": 2.3351, + "step": 25836 + }, + { + "epoch": 2.732339255499154, + "grad_norm": 8.0, + "learning_rate": 2.2456969774849936e-06, + "loss": 2.4617, + "step": 25837 + }, + { + "epoch": 2.7324450084602367, + "grad_norm": 8.1875, + "learning_rate": 2.2455266101744317e-06, + "loss": 2.4994, + "step": 25838 + }, + { + "epoch": 2.7325507614213196, + "grad_norm": 8.8125, + "learning_rate": 2.2453562440580097e-06, + "loss": 2.5176, + "step": 25839 + }, + { + "epoch": 2.732656514382403, + "grad_norm": 8.125, + "learning_rate": 2.2451858791365267e-06, + "loss": 2.4477, + "step": 25840 + }, + { + "epoch": 2.7327622673434857, + "grad_norm": 8.25, + "learning_rate": 2.2450155154107815e-06, + "loss": 2.5447, + "step": 25841 + }, + { + "epoch": 2.7328680203045685, + "grad_norm": 8.3125, + "learning_rate": 2.244845152881575e-06, + "loss": 2.6458, + "step": 25842 + }, + { + "epoch": 2.7329737732656514, + "grad_norm": 8.8125, + "learning_rate": 2.2446747915497053e-06, + "loss": 2.5463, + "step": 25843 + }, + { + "epoch": 2.7330795262267342, + "grad_norm": 8.25, + "learning_rate": 2.244504431415974e-06, + "loss": 2.4534, + "step": 25844 + }, + { + "epoch": 2.7331852791878175, + "grad_norm": 8.5, + "learning_rate": 2.244334072481177e-06, + "loss": 2.589, + "step": 25845 + }, + { + "epoch": 2.7332910321489, + "grad_norm": 8.3125, + "learning_rate": 2.244163714746117e-06, + "loss": 2.4179, + "step": 25846 + }, + { + "epoch": 2.733396785109983, + "grad_norm": 8.375, + "learning_rate": 2.243993358211591e-06, + "loss": 2.5835, + "step": 25847 + }, + { + "epoch": 2.733502538071066, + "grad_norm": 9.0, + "learning_rate": 2.2438230028783997e-06, + "loss": 2.6136, + "step": 25848 + }, + { + "epoch": 2.733608291032149, + "grad_norm": 9.0, + "learning_rate": 2.243652648747342e-06, + "loss": 2.4667, + "step": 25849 + }, + { + "epoch": 2.7337140439932317, + "grad_norm": 8.4375, + "learning_rate": 2.2434822958192177e-06, + "loss": 2.6081, + "step": 25850 + }, + { + "epoch": 2.7338197969543145, + "grad_norm": 8.25, + "learning_rate": 2.2433119440948265e-06, + "loss": 2.4305, + "step": 25851 + }, + { + "epoch": 2.733925549915398, + "grad_norm": 8.0, + "learning_rate": 2.243141593574967e-06, + "loss": 2.4827, + "step": 25852 + }, + { + "epoch": 2.7340313028764807, + "grad_norm": 8.25, + "learning_rate": 2.24297124426044e-06, + "loss": 2.5204, + "step": 25853 + }, + { + "epoch": 2.7341370558375635, + "grad_norm": 8.125, + "learning_rate": 2.2428008961520426e-06, + "loss": 2.5269, + "step": 25854 + }, + { + "epoch": 2.7342428087986463, + "grad_norm": 8.1875, + "learning_rate": 2.2426305492505757e-06, + "loss": 2.4653, + "step": 25855 + }, + { + "epoch": 2.734348561759729, + "grad_norm": 8.25, + "learning_rate": 2.242460203556838e-06, + "loss": 2.6167, + "step": 25856 + }, + { + "epoch": 2.734454314720812, + "grad_norm": 8.6875, + "learning_rate": 2.24228985907163e-06, + "loss": 2.263, + "step": 25857 + }, + { + "epoch": 2.734560067681895, + "grad_norm": 8.9375, + "learning_rate": 2.24211951579575e-06, + "loss": 2.8006, + "step": 25858 + }, + { + "epoch": 2.734665820642978, + "grad_norm": 8.75, + "learning_rate": 2.241949173729997e-06, + "loss": 2.4519, + "step": 25859 + }, + { + "epoch": 2.734771573604061, + "grad_norm": 7.71875, + "learning_rate": 2.241778832875173e-06, + "loss": 2.5104, + "step": 25860 + }, + { + "epoch": 2.734877326565144, + "grad_norm": 8.5, + "learning_rate": 2.2416084932320735e-06, + "loss": 2.4294, + "step": 25861 + }, + { + "epoch": 2.7349830795262267, + "grad_norm": 8.8125, + "learning_rate": 2.2414381548015002e-06, + "loss": 2.5802, + "step": 25862 + }, + { + "epoch": 2.7350888324873095, + "grad_norm": 9.0, + "learning_rate": 2.2412678175842516e-06, + "loss": 2.5011, + "step": 25863 + }, + { + "epoch": 2.735194585448393, + "grad_norm": 8.0625, + "learning_rate": 2.241097481581128e-06, + "loss": 2.4802, + "step": 25864 + }, + { + "epoch": 2.7353003384094756, + "grad_norm": 8.1875, + "learning_rate": 2.2409271467929266e-06, + "loss": 2.5965, + "step": 25865 + }, + { + "epoch": 2.7354060913705585, + "grad_norm": 8.5, + "learning_rate": 2.24075681322045e-06, + "loss": 2.5731, + "step": 25866 + }, + { + "epoch": 2.7355118443316413, + "grad_norm": 8.6875, + "learning_rate": 2.2405864808644944e-06, + "loss": 2.5796, + "step": 25867 + }, + { + "epoch": 2.735617597292724, + "grad_norm": 8.25, + "learning_rate": 2.2404161497258596e-06, + "loss": 2.4367, + "step": 25868 + }, + { + "epoch": 2.735723350253807, + "grad_norm": 7.625, + "learning_rate": 2.240245819805348e-06, + "loss": 2.3201, + "step": 25869 + }, + { + "epoch": 2.73582910321489, + "grad_norm": 8.0625, + "learning_rate": 2.240075491103755e-06, + "loss": 2.3979, + "step": 25870 + }, + { + "epoch": 2.735934856175973, + "grad_norm": 8.6875, + "learning_rate": 2.2399051636218816e-06, + "loss": 2.6591, + "step": 25871 + }, + { + "epoch": 2.736040609137056, + "grad_norm": 8.25, + "learning_rate": 2.2397348373605264e-06, + "loss": 2.6682, + "step": 25872 + }, + { + "epoch": 2.7361463620981388, + "grad_norm": 8.3125, + "learning_rate": 2.2395645123204903e-06, + "loss": 2.6688, + "step": 25873 + }, + { + "epoch": 2.7362521150592216, + "grad_norm": 8.125, + "learning_rate": 2.2393941885025707e-06, + "loss": 2.6578, + "step": 25874 + }, + { + "epoch": 2.7363578680203045, + "grad_norm": 8.5, + "learning_rate": 2.2392238659075684e-06, + "loss": 2.6778, + "step": 25875 + }, + { + "epoch": 2.7364636209813877, + "grad_norm": 8.3125, + "learning_rate": 2.2390535445362806e-06, + "loss": 2.5262, + "step": 25876 + }, + { + "epoch": 2.73656937394247, + "grad_norm": 8.125, + "learning_rate": 2.238883224389508e-06, + "loss": 2.5974, + "step": 25877 + }, + { + "epoch": 2.7366751269035534, + "grad_norm": 8.375, + "learning_rate": 2.2387129054680505e-06, + "loss": 2.5465, + "step": 25878 + }, + { + "epoch": 2.7367808798646363, + "grad_norm": 8.5, + "learning_rate": 2.2385425877727056e-06, + "loss": 2.587, + "step": 25879 + }, + { + "epoch": 2.736886632825719, + "grad_norm": 8.25, + "learning_rate": 2.2383722713042736e-06, + "loss": 2.4599, + "step": 25880 + }, + { + "epoch": 2.736992385786802, + "grad_norm": 8.5, + "learning_rate": 2.2382019560635536e-06, + "loss": 2.4768, + "step": 25881 + }, + { + "epoch": 2.7370981387478848, + "grad_norm": 8.5625, + "learning_rate": 2.238031642051346e-06, + "loss": 2.6159, + "step": 25882 + }, + { + "epoch": 2.737203891708968, + "grad_norm": 7.75, + "learning_rate": 2.237861329268447e-06, + "loss": 2.4799, + "step": 25883 + }, + { + "epoch": 2.737309644670051, + "grad_norm": 8.875, + "learning_rate": 2.2376910177156587e-06, + "loss": 2.7089, + "step": 25884 + }, + { + "epoch": 2.7374153976311337, + "grad_norm": 9.75, + "learning_rate": 2.2375207073937784e-06, + "loss": 2.6125, + "step": 25885 + }, + { + "epoch": 2.7375211505922166, + "grad_norm": 8.6875, + "learning_rate": 2.2373503983036056e-06, + "loss": 2.514, + "step": 25886 + }, + { + "epoch": 2.7376269035532994, + "grad_norm": 8.3125, + "learning_rate": 2.237180090445941e-06, + "loss": 2.4698, + "step": 25887 + }, + { + "epoch": 2.7377326565143822, + "grad_norm": 8.3125, + "learning_rate": 2.2370097838215825e-06, + "loss": 2.0641, + "step": 25888 + }, + { + "epoch": 2.737838409475465, + "grad_norm": 8.6875, + "learning_rate": 2.2368394784313293e-06, + "loss": 2.8452, + "step": 25889 + }, + { + "epoch": 2.7379441624365484, + "grad_norm": 8.5625, + "learning_rate": 2.2366691742759806e-06, + "loss": 2.5541, + "step": 25890 + }, + { + "epoch": 2.738049915397631, + "grad_norm": 9.1875, + "learning_rate": 2.236498871356337e-06, + "loss": 2.7604, + "step": 25891 + }, + { + "epoch": 2.738155668358714, + "grad_norm": 8.625, + "learning_rate": 2.2363285696731954e-06, + "loss": 2.5436, + "step": 25892 + }, + { + "epoch": 2.738261421319797, + "grad_norm": 9.1875, + "learning_rate": 2.236158269227356e-06, + "loss": 2.8597, + "step": 25893 + }, + { + "epoch": 2.7383671742808797, + "grad_norm": 8.25, + "learning_rate": 2.235987970019618e-06, + "loss": 2.56, + "step": 25894 + }, + { + "epoch": 2.738472927241963, + "grad_norm": 8.5, + "learning_rate": 2.2358176720507805e-06, + "loss": 2.4154, + "step": 25895 + }, + { + "epoch": 2.738578680203046, + "grad_norm": 7.84375, + "learning_rate": 2.235647375321643e-06, + "loss": 2.477, + "step": 25896 + }, + { + "epoch": 2.7386844331641287, + "grad_norm": 8.4375, + "learning_rate": 2.2354770798330035e-06, + "loss": 2.4973, + "step": 25897 + }, + { + "epoch": 2.7387901861252115, + "grad_norm": 8.375, + "learning_rate": 2.235306785585663e-06, + "loss": 2.6567, + "step": 25898 + }, + { + "epoch": 2.7388959390862944, + "grad_norm": 8.4375, + "learning_rate": 2.2351364925804185e-06, + "loss": 2.4852, + "step": 25899 + }, + { + "epoch": 2.739001692047377, + "grad_norm": 8.3125, + "learning_rate": 2.2349662008180708e-06, + "loss": 2.3137, + "step": 25900 + }, + { + "epoch": 2.73910744500846, + "grad_norm": 8.3125, + "learning_rate": 2.2347959102994172e-06, + "loss": 2.6261, + "step": 25901 + }, + { + "epoch": 2.7392131979695433, + "grad_norm": 8.25, + "learning_rate": 2.2346256210252588e-06, + "loss": 2.5676, + "step": 25902 + }, + { + "epoch": 2.739318950930626, + "grad_norm": 8.25, + "learning_rate": 2.2344553329963938e-06, + "loss": 2.5403, + "step": 25903 + }, + { + "epoch": 2.739424703891709, + "grad_norm": 8.5, + "learning_rate": 2.234285046213621e-06, + "loss": 2.5941, + "step": 25904 + }, + { + "epoch": 2.739530456852792, + "grad_norm": 8.5, + "learning_rate": 2.2341147606777404e-06, + "loss": 2.4705, + "step": 25905 + }, + { + "epoch": 2.7396362098138747, + "grad_norm": 8.5, + "learning_rate": 2.2339444763895495e-06, + "loss": 2.5819, + "step": 25906 + }, + { + "epoch": 2.739741962774958, + "grad_norm": 8.4375, + "learning_rate": 2.2337741933498496e-06, + "loss": 2.4528, + "step": 25907 + }, + { + "epoch": 2.7398477157360404, + "grad_norm": 9.3125, + "learning_rate": 2.2336039115594377e-06, + "loss": 2.7083, + "step": 25908 + }, + { + "epoch": 2.7399534686971236, + "grad_norm": 8.375, + "learning_rate": 2.233433631019114e-06, + "loss": 2.5505, + "step": 25909 + }, + { + "epoch": 2.7400592216582065, + "grad_norm": 8.375, + "learning_rate": 2.233263351729677e-06, + "loss": 2.6119, + "step": 25910 + }, + { + "epoch": 2.7401649746192893, + "grad_norm": 8.0625, + "learning_rate": 2.2330930736919256e-06, + "loss": 2.5441, + "step": 25911 + }, + { + "epoch": 2.740270727580372, + "grad_norm": 8.6875, + "learning_rate": 2.23292279690666e-06, + "loss": 2.5354, + "step": 25912 + }, + { + "epoch": 2.740376480541455, + "grad_norm": 8.5, + "learning_rate": 2.2327525213746783e-06, + "loss": 2.5831, + "step": 25913 + }, + { + "epoch": 2.7404822335025383, + "grad_norm": 8.1875, + "learning_rate": 2.2325822470967804e-06, + "loss": 2.5773, + "step": 25914 + }, + { + "epoch": 2.740587986463621, + "grad_norm": 9.1875, + "learning_rate": 2.2324119740737635e-06, + "loss": 2.4809, + "step": 25915 + }, + { + "epoch": 2.740693739424704, + "grad_norm": 8.4375, + "learning_rate": 2.2322417023064284e-06, + "loss": 2.4481, + "step": 25916 + }, + { + "epoch": 2.740799492385787, + "grad_norm": 8.4375, + "learning_rate": 2.232071431795573e-06, + "loss": 2.5795, + "step": 25917 + }, + { + "epoch": 2.7409052453468696, + "grad_norm": 9.0625, + "learning_rate": 2.2319011625419973e-06, + "loss": 2.6649, + "step": 25918 + }, + { + "epoch": 2.7410109983079525, + "grad_norm": 8.3125, + "learning_rate": 2.2317308945464995e-06, + "loss": 2.422, + "step": 25919 + }, + { + "epoch": 2.7411167512690353, + "grad_norm": 8.5625, + "learning_rate": 2.2315606278098782e-06, + "loss": 2.5109, + "step": 25920 + }, + { + "epoch": 2.7412225042301186, + "grad_norm": 9.0, + "learning_rate": 2.2313903623329347e-06, + "loss": 2.8107, + "step": 25921 + }, + { + "epoch": 2.7413282571912014, + "grad_norm": 8.625, + "learning_rate": 2.2312200981164647e-06, + "loss": 2.6804, + "step": 25922 + }, + { + "epoch": 2.7414340101522843, + "grad_norm": 8.5, + "learning_rate": 2.2310498351612704e-06, + "loss": 2.6157, + "step": 25923 + }, + { + "epoch": 2.741539763113367, + "grad_norm": 8.125, + "learning_rate": 2.2308795734681483e-06, + "loss": 2.4609, + "step": 25924 + }, + { + "epoch": 2.74164551607445, + "grad_norm": 8.5, + "learning_rate": 2.2307093130378987e-06, + "loss": 2.353, + "step": 25925 + }, + { + "epoch": 2.7417512690355332, + "grad_norm": 8.375, + "learning_rate": 2.2305390538713197e-06, + "loss": 2.5338, + "step": 25926 + }, + { + "epoch": 2.7418570219966156, + "grad_norm": 9.0625, + "learning_rate": 2.230368795969211e-06, + "loss": 2.7262, + "step": 25927 + }, + { + "epoch": 2.741962774957699, + "grad_norm": 8.3125, + "learning_rate": 2.2301985393323712e-06, + "loss": 2.4291, + "step": 25928 + }, + { + "epoch": 2.7420685279187818, + "grad_norm": 8.375, + "learning_rate": 2.230028283961599e-06, + "loss": 2.6343, + "step": 25929 + }, + { + "epoch": 2.7421742808798646, + "grad_norm": 9.0625, + "learning_rate": 2.2298580298576947e-06, + "loss": 2.2281, + "step": 25930 + }, + { + "epoch": 2.7422800338409474, + "grad_norm": 8.5, + "learning_rate": 2.229687777021455e-06, + "loss": 2.4924, + "step": 25931 + }, + { + "epoch": 2.7423857868020303, + "grad_norm": 8.9375, + "learning_rate": 2.229517525453681e-06, + "loss": 2.4028, + "step": 25932 + }, + { + "epoch": 2.7424915397631136, + "grad_norm": 8.5625, + "learning_rate": 2.2293472751551697e-06, + "loss": 2.4475, + "step": 25933 + }, + { + "epoch": 2.7425972927241964, + "grad_norm": 9.0, + "learning_rate": 2.2291770261267213e-06, + "loss": 2.6576, + "step": 25934 + }, + { + "epoch": 2.7427030456852792, + "grad_norm": 8.1875, + "learning_rate": 2.229006778369134e-06, + "loss": 2.5648, + "step": 25935 + }, + { + "epoch": 2.742808798646362, + "grad_norm": 8.375, + "learning_rate": 2.2288365318832082e-06, + "loss": 2.6461, + "step": 25936 + }, + { + "epoch": 2.742914551607445, + "grad_norm": 8.375, + "learning_rate": 2.2286662866697407e-06, + "loss": 2.6264, + "step": 25937 + }, + { + "epoch": 2.743020304568528, + "grad_norm": 9.1875, + "learning_rate": 2.2284960427295306e-06, + "loss": 2.8679, + "step": 25938 + }, + { + "epoch": 2.7431260575296106, + "grad_norm": 8.3125, + "learning_rate": 2.2283258000633796e-06, + "loss": 2.3721, + "step": 25939 + }, + { + "epoch": 2.743231810490694, + "grad_norm": 8.4375, + "learning_rate": 2.228155558672083e-06, + "loss": 2.4528, + "step": 25940 + }, + { + "epoch": 2.7433375634517767, + "grad_norm": 8.5, + "learning_rate": 2.2279853185564418e-06, + "loss": 2.4314, + "step": 25941 + }, + { + "epoch": 2.7434433164128595, + "grad_norm": 8.75, + "learning_rate": 2.2278150797172538e-06, + "loss": 2.6685, + "step": 25942 + }, + { + "epoch": 2.7435490693739424, + "grad_norm": 8.25, + "learning_rate": 2.2276448421553186e-06, + "loss": 2.5524, + "step": 25943 + }, + { + "epoch": 2.7436548223350252, + "grad_norm": 8.1875, + "learning_rate": 2.2274746058714346e-06, + "loss": 2.5922, + "step": 25944 + }, + { + "epoch": 2.7437605752961085, + "grad_norm": 8.5625, + "learning_rate": 2.2273043708664014e-06, + "loss": 2.513, + "step": 25945 + }, + { + "epoch": 2.7438663282571913, + "grad_norm": 8.8125, + "learning_rate": 2.2271341371410165e-06, + "loss": 2.4754, + "step": 25946 + }, + { + "epoch": 2.743972081218274, + "grad_norm": 8.0625, + "learning_rate": 2.2269639046960793e-06, + "loss": 2.3607, + "step": 25947 + }, + { + "epoch": 2.744077834179357, + "grad_norm": 8.375, + "learning_rate": 2.2267936735323897e-06, + "loss": 2.4912, + "step": 25948 + }, + { + "epoch": 2.74418358714044, + "grad_norm": 8.875, + "learning_rate": 2.226623443650745e-06, + "loss": 2.5092, + "step": 25949 + }, + { + "epoch": 2.7442893401015227, + "grad_norm": 8.0, + "learning_rate": 2.226453215051945e-06, + "loss": 2.5966, + "step": 25950 + }, + { + "epoch": 2.7443950930626055, + "grad_norm": 7.96875, + "learning_rate": 2.2262829877367878e-06, + "loss": 2.3232, + "step": 25951 + }, + { + "epoch": 2.744500846023689, + "grad_norm": 8.4375, + "learning_rate": 2.2261127617060737e-06, + "loss": 2.3964, + "step": 25952 + }, + { + "epoch": 2.7446065989847717, + "grad_norm": 8.5, + "learning_rate": 2.2259425369605996e-06, + "loss": 2.6341, + "step": 25953 + }, + { + "epoch": 2.7447123519458545, + "grad_norm": 11.6875, + "learning_rate": 2.225772313501165e-06, + "loss": 2.7092, + "step": 25954 + }, + { + "epoch": 2.7448181049069373, + "grad_norm": 8.5625, + "learning_rate": 2.2256020913285687e-06, + "loss": 2.7862, + "step": 25955 + }, + { + "epoch": 2.74492385786802, + "grad_norm": 8.4375, + "learning_rate": 2.225431870443609e-06, + "loss": 2.4939, + "step": 25956 + }, + { + "epoch": 2.7450296108291035, + "grad_norm": 7.96875, + "learning_rate": 2.2252616508470864e-06, + "loss": 2.4812, + "step": 25957 + }, + { + "epoch": 2.745135363790186, + "grad_norm": 8.6875, + "learning_rate": 2.225091432539798e-06, + "loss": 2.5667, + "step": 25958 + }, + { + "epoch": 2.745241116751269, + "grad_norm": 8.1875, + "learning_rate": 2.224921215522544e-06, + "loss": 2.4362, + "step": 25959 + }, + { + "epoch": 2.745346869712352, + "grad_norm": 8.6875, + "learning_rate": 2.2247509997961204e-06, + "loss": 2.4976, + "step": 25960 + }, + { + "epoch": 2.745452622673435, + "grad_norm": 8.625, + "learning_rate": 2.2245807853613293e-06, + "loss": 2.1625, + "step": 25961 + }, + { + "epoch": 2.7455583756345177, + "grad_norm": 8.8125, + "learning_rate": 2.224410572218967e-06, + "loss": 2.5477, + "step": 25962 + }, + { + "epoch": 2.7456641285956005, + "grad_norm": 8.125, + "learning_rate": 2.224240360369834e-06, + "loss": 2.5452, + "step": 25963 + }, + { + "epoch": 2.745769881556684, + "grad_norm": 8.25, + "learning_rate": 2.2240701498147276e-06, + "loss": 2.6269, + "step": 25964 + }, + { + "epoch": 2.7458756345177666, + "grad_norm": 8.25, + "learning_rate": 2.223899940554447e-06, + "loss": 2.5395, + "step": 25965 + }, + { + "epoch": 2.7459813874788495, + "grad_norm": 8.375, + "learning_rate": 2.2237297325897914e-06, + "loss": 2.4147, + "step": 25966 + }, + { + "epoch": 2.7460871404399323, + "grad_norm": 7.96875, + "learning_rate": 2.223559525921559e-06, + "loss": 2.5133, + "step": 25967 + }, + { + "epoch": 2.746192893401015, + "grad_norm": 7.78125, + "learning_rate": 2.2233893205505494e-06, + "loss": 2.3319, + "step": 25968 + }, + { + "epoch": 2.746298646362098, + "grad_norm": 8.5625, + "learning_rate": 2.2232191164775598e-06, + "loss": 2.5525, + "step": 25969 + }, + { + "epoch": 2.746404399323181, + "grad_norm": 8.5625, + "learning_rate": 2.22304891370339e-06, + "loss": 2.6665, + "step": 25970 + }, + { + "epoch": 2.746510152284264, + "grad_norm": 8.625, + "learning_rate": 2.222878712228838e-06, + "loss": 2.3113, + "step": 25971 + }, + { + "epoch": 2.746615905245347, + "grad_norm": 8.375, + "learning_rate": 2.2227085120547034e-06, + "loss": 2.4814, + "step": 25972 + }, + { + "epoch": 2.7467216582064298, + "grad_norm": 9.25, + "learning_rate": 2.2225383131817837e-06, + "loss": 2.8464, + "step": 25973 + }, + { + "epoch": 2.7468274111675126, + "grad_norm": 8.4375, + "learning_rate": 2.222368115610878e-06, + "loss": 2.6887, + "step": 25974 + }, + { + "epoch": 2.7469331641285955, + "grad_norm": 8.5625, + "learning_rate": 2.222197919342787e-06, + "loss": 2.5612, + "step": 25975 + }, + { + "epoch": 2.7470389170896787, + "grad_norm": 8.375, + "learning_rate": 2.2220277243783055e-06, + "loss": 2.4915, + "step": 25976 + }, + { + "epoch": 2.7471446700507616, + "grad_norm": 8.75, + "learning_rate": 2.221857530718236e-06, + "loss": 2.5461, + "step": 25977 + }, + { + "epoch": 2.7472504230118444, + "grad_norm": 8.75, + "learning_rate": 2.221687338363374e-06, + "loss": 2.4902, + "step": 25978 + }, + { + "epoch": 2.7473561759729273, + "grad_norm": 8.3125, + "learning_rate": 2.2215171473145197e-06, + "loss": 2.5225, + "step": 25979 + }, + { + "epoch": 2.74746192893401, + "grad_norm": 8.625, + "learning_rate": 2.2213469575724717e-06, + "loss": 2.2862, + "step": 25980 + }, + { + "epoch": 2.747567681895093, + "grad_norm": 7.96875, + "learning_rate": 2.2211767691380286e-06, + "loss": 2.5272, + "step": 25981 + }, + { + "epoch": 2.7476734348561758, + "grad_norm": 8.25, + "learning_rate": 2.2210065820119884e-06, + "loss": 2.5101, + "step": 25982 + }, + { + "epoch": 2.747779187817259, + "grad_norm": 8.625, + "learning_rate": 2.2208363961951503e-06, + "loss": 2.704, + "step": 25983 + }, + { + "epoch": 2.747884940778342, + "grad_norm": 8.9375, + "learning_rate": 2.220666211688314e-06, + "loss": 2.7438, + "step": 25984 + }, + { + "epoch": 2.7479906937394247, + "grad_norm": 7.9375, + "learning_rate": 2.2204960284922754e-06, + "loss": 2.499, + "step": 25985 + }, + { + "epoch": 2.7480964467005076, + "grad_norm": 8.625, + "learning_rate": 2.2203258466078356e-06, + "loss": 2.6099, + "step": 25986 + }, + { + "epoch": 2.7482021996615904, + "grad_norm": 8.375, + "learning_rate": 2.2201556660357916e-06, + "loss": 2.4682, + "step": 25987 + }, + { + "epoch": 2.7483079526226737, + "grad_norm": 8.625, + "learning_rate": 2.219985486776943e-06, + "loss": 2.7191, + "step": 25988 + }, + { + "epoch": 2.748413705583756, + "grad_norm": 8.1875, + "learning_rate": 2.2198153088320874e-06, + "loss": 2.5209, + "step": 25989 + }, + { + "epoch": 2.7485194585448394, + "grad_norm": 9.0, + "learning_rate": 2.2196451322020255e-06, + "loss": 2.5075, + "step": 25990 + }, + { + "epoch": 2.748625211505922, + "grad_norm": 9.125, + "learning_rate": 2.2194749568875528e-06, + "loss": 2.5473, + "step": 25991 + }, + { + "epoch": 2.748730964467005, + "grad_norm": 8.8125, + "learning_rate": 2.219304782889469e-06, + "loss": 2.5678, + "step": 25992 + }, + { + "epoch": 2.748836717428088, + "grad_norm": 8.5625, + "learning_rate": 2.2191346102085747e-06, + "loss": 2.5023, + "step": 25993 + }, + { + "epoch": 2.7489424703891707, + "grad_norm": 8.625, + "learning_rate": 2.2189644388456656e-06, + "loss": 2.5992, + "step": 25994 + }, + { + "epoch": 2.749048223350254, + "grad_norm": 8.6875, + "learning_rate": 2.2187942688015424e-06, + "loss": 2.8093, + "step": 25995 + }, + { + "epoch": 2.749153976311337, + "grad_norm": 8.0, + "learning_rate": 2.2186241000770016e-06, + "loss": 2.6815, + "step": 25996 + }, + { + "epoch": 2.7492597292724197, + "grad_norm": 8.125, + "learning_rate": 2.2184539326728434e-06, + "loss": 2.7018, + "step": 25997 + }, + { + "epoch": 2.7493654822335025, + "grad_norm": 8.5, + "learning_rate": 2.218283766589866e-06, + "loss": 2.4768, + "step": 25998 + }, + { + "epoch": 2.7494712351945854, + "grad_norm": 8.3125, + "learning_rate": 2.218113601828868e-06, + "loss": 2.4659, + "step": 25999 + }, + { + "epoch": 2.749576988155668, + "grad_norm": 8.25, + "learning_rate": 2.2179434383906463e-06, + "loss": 2.4937, + "step": 26000 + }, + { + "epoch": 2.749682741116751, + "grad_norm": 8.6875, + "learning_rate": 2.2177732762760013e-06, + "loss": 2.616, + "step": 26001 + }, + { + "epoch": 2.7497884940778343, + "grad_norm": 8.625, + "learning_rate": 2.217603115485731e-06, + "loss": 2.6001, + "step": 26002 + }, + { + "epoch": 2.749894247038917, + "grad_norm": 8.0, + "learning_rate": 2.2174329560206336e-06, + "loss": 2.7149, + "step": 26003 + }, + { + "epoch": 2.75, + "grad_norm": 8.375, + "learning_rate": 2.217262797881508e-06, + "loss": 2.5654, + "step": 26004 + }, + { + "epoch": 2.750105752961083, + "grad_norm": 8.5625, + "learning_rate": 2.217092641069152e-06, + "loss": 2.718, + "step": 26005 + }, + { + "epoch": 2.7502115059221657, + "grad_norm": 10.4375, + "learning_rate": 2.2169224855843655e-06, + "loss": 2.4869, + "step": 26006 + }, + { + "epoch": 2.750317258883249, + "grad_norm": 8.5, + "learning_rate": 2.2167523314279453e-06, + "loss": 2.5598, + "step": 26007 + }, + { + "epoch": 2.750423011844332, + "grad_norm": 8.6875, + "learning_rate": 2.2165821786006895e-06, + "loss": 2.7997, + "step": 26008 + }, + { + "epoch": 2.7505287648054146, + "grad_norm": 8.375, + "learning_rate": 2.2164120271033996e-06, + "loss": 2.4117, + "step": 26009 + }, + { + "epoch": 2.7506345177664975, + "grad_norm": 8.4375, + "learning_rate": 2.216241876936871e-06, + "loss": 2.6093, + "step": 26010 + }, + { + "epoch": 2.7507402707275803, + "grad_norm": 8.1875, + "learning_rate": 2.2160717281019038e-06, + "loss": 2.5964, + "step": 26011 + }, + { + "epoch": 2.750846023688663, + "grad_norm": 9.0625, + "learning_rate": 2.215901580599295e-06, + "loss": 2.7327, + "step": 26012 + }, + { + "epoch": 2.750951776649746, + "grad_norm": 8.875, + "learning_rate": 2.215731434429845e-06, + "loss": 2.8063, + "step": 26013 + }, + { + "epoch": 2.7510575296108293, + "grad_norm": 8.125, + "learning_rate": 2.21556128959435e-06, + "loss": 2.4652, + "step": 26014 + }, + { + "epoch": 2.751163282571912, + "grad_norm": 8.375, + "learning_rate": 2.2153911460936105e-06, + "loss": 2.3914, + "step": 26015 + }, + { + "epoch": 2.751269035532995, + "grad_norm": 8.375, + "learning_rate": 2.215221003928423e-06, + "loss": 2.4704, + "step": 26016 + }, + { + "epoch": 2.751374788494078, + "grad_norm": 8.1875, + "learning_rate": 2.2150508630995873e-06, + "loss": 2.441, + "step": 26017 + }, + { + "epoch": 2.7514805414551606, + "grad_norm": 8.1875, + "learning_rate": 2.214880723607902e-06, + "loss": 2.4939, + "step": 26018 + }, + { + "epoch": 2.751586294416244, + "grad_norm": 10.3125, + "learning_rate": 2.2147105854541637e-06, + "loss": 2.7988, + "step": 26019 + }, + { + "epoch": 2.7516920473773263, + "grad_norm": 8.5, + "learning_rate": 2.2145404486391727e-06, + "loss": 2.5135, + "step": 26020 + }, + { + "epoch": 2.7517978003384096, + "grad_norm": 8.25, + "learning_rate": 2.214370313163726e-06, + "loss": 2.5648, + "step": 26021 + }, + { + "epoch": 2.7519035532994924, + "grad_norm": 8.625, + "learning_rate": 2.214200179028624e-06, + "loss": 2.6003, + "step": 26022 + }, + { + "epoch": 2.7520093062605753, + "grad_norm": 8.4375, + "learning_rate": 2.2140300462346623e-06, + "loss": 2.6578, + "step": 26023 + }, + { + "epoch": 2.752115059221658, + "grad_norm": 8.125, + "learning_rate": 2.2138599147826412e-06, + "loss": 2.525, + "step": 26024 + }, + { + "epoch": 2.752220812182741, + "grad_norm": 8.3125, + "learning_rate": 2.213689784673358e-06, + "loss": 2.6974, + "step": 26025 + }, + { + "epoch": 2.7523265651438242, + "grad_norm": 8.25, + "learning_rate": 2.213519655907612e-06, + "loss": 2.5259, + "step": 26026 + }, + { + "epoch": 2.752432318104907, + "grad_norm": 8.3125, + "learning_rate": 2.213349528486201e-06, + "loss": 2.5052, + "step": 26027 + }, + { + "epoch": 2.75253807106599, + "grad_norm": 8.4375, + "learning_rate": 2.2131794024099227e-06, + "loss": 2.3355, + "step": 26028 + }, + { + "epoch": 2.7526438240270727, + "grad_norm": 8.4375, + "learning_rate": 2.2130092776795783e-06, + "loss": 2.5935, + "step": 26029 + }, + { + "epoch": 2.7527495769881556, + "grad_norm": 8.4375, + "learning_rate": 2.212839154295961e-06, + "loss": 2.5111, + "step": 26030 + }, + { + "epoch": 2.7528553299492384, + "grad_norm": 8.875, + "learning_rate": 2.2126690322598747e-06, + "loss": 2.7789, + "step": 26031 + }, + { + "epoch": 2.7529610829103213, + "grad_norm": 8.5625, + "learning_rate": 2.212498911572114e-06, + "loss": 2.6188, + "step": 26032 + }, + { + "epoch": 2.7530668358714045, + "grad_norm": 8.75, + "learning_rate": 2.2123287922334785e-06, + "loss": 2.5121, + "step": 26033 + }, + { + "epoch": 2.7531725888324874, + "grad_norm": 8.6875, + "learning_rate": 2.212158674244766e-06, + "loss": 2.6748, + "step": 26034 + }, + { + "epoch": 2.7532783417935702, + "grad_norm": 8.625, + "learning_rate": 2.2119885576067746e-06, + "loss": 2.5245, + "step": 26035 + }, + { + "epoch": 2.753384094754653, + "grad_norm": 8.8125, + "learning_rate": 2.211818442320304e-06, + "loss": 2.6766, + "step": 26036 + }, + { + "epoch": 2.753489847715736, + "grad_norm": 8.375, + "learning_rate": 2.2116483283861514e-06, + "loss": 2.5253, + "step": 26037 + }, + { + "epoch": 2.753595600676819, + "grad_norm": 8.375, + "learning_rate": 2.211478215805116e-06, + "loss": 2.6162, + "step": 26038 + }, + { + "epoch": 2.753701353637902, + "grad_norm": 8.25, + "learning_rate": 2.211308104577994e-06, + "loss": 2.5168, + "step": 26039 + }, + { + "epoch": 2.753807106598985, + "grad_norm": 8.25, + "learning_rate": 2.211137994705586e-06, + "loss": 2.5987, + "step": 26040 + }, + { + "epoch": 2.7539128595600677, + "grad_norm": 8.1875, + "learning_rate": 2.2109678861886887e-06, + "loss": 2.4074, + "step": 26041 + }, + { + "epoch": 2.7540186125211505, + "grad_norm": 8.125, + "learning_rate": 2.2107977790281013e-06, + "loss": 2.527, + "step": 26042 + }, + { + "epoch": 2.7541243654822334, + "grad_norm": 8.125, + "learning_rate": 2.2106276732246212e-06, + "loss": 2.2344, + "step": 26043 + }, + { + "epoch": 2.754230118443316, + "grad_norm": 8.5625, + "learning_rate": 2.2104575687790468e-06, + "loss": 2.7307, + "step": 26044 + }, + { + "epoch": 2.7543358714043995, + "grad_norm": 8.25, + "learning_rate": 2.210287465692178e-06, + "loss": 2.5456, + "step": 26045 + }, + { + "epoch": 2.7544416243654823, + "grad_norm": 8.1875, + "learning_rate": 2.2101173639648097e-06, + "loss": 2.6455, + "step": 26046 + }, + { + "epoch": 2.754547377326565, + "grad_norm": 8.1875, + "learning_rate": 2.2099472635977438e-06, + "loss": 2.3851, + "step": 26047 + }, + { + "epoch": 2.754653130287648, + "grad_norm": 8.375, + "learning_rate": 2.209777164591776e-06, + "loss": 2.704, + "step": 26048 + }, + { + "epoch": 2.754758883248731, + "grad_norm": 8.0, + "learning_rate": 2.2096070669477053e-06, + "loss": 2.488, + "step": 26049 + }, + { + "epoch": 2.754864636209814, + "grad_norm": 8.25, + "learning_rate": 2.2094369706663295e-06, + "loss": 2.3347, + "step": 26050 + }, + { + "epoch": 2.7549703891708965, + "grad_norm": 8.625, + "learning_rate": 2.209266875748448e-06, + "loss": 2.5027, + "step": 26051 + }, + { + "epoch": 2.75507614213198, + "grad_norm": 8.125, + "learning_rate": 2.209096782194857e-06, + "loss": 2.4415, + "step": 26052 + }, + { + "epoch": 2.7551818950930627, + "grad_norm": 9.125, + "learning_rate": 2.2089266900063567e-06, + "loss": 2.7664, + "step": 26053 + }, + { + "epoch": 2.7552876480541455, + "grad_norm": 8.8125, + "learning_rate": 2.208756599183745e-06, + "loss": 2.5257, + "step": 26054 + }, + { + "epoch": 2.7553934010152283, + "grad_norm": 8.8125, + "learning_rate": 2.2085865097278186e-06, + "loss": 2.7204, + "step": 26055 + }, + { + "epoch": 2.755499153976311, + "grad_norm": 8.5625, + "learning_rate": 2.2084164216393765e-06, + "loss": 2.6136, + "step": 26056 + }, + { + "epoch": 2.7556049069373945, + "grad_norm": 8.6875, + "learning_rate": 2.2082463349192167e-06, + "loss": 2.5459, + "step": 26057 + }, + { + "epoch": 2.7557106598984773, + "grad_norm": 8.5625, + "learning_rate": 2.208076249568138e-06, + "loss": 2.5836, + "step": 26058 + }, + { + "epoch": 2.75581641285956, + "grad_norm": 8.9375, + "learning_rate": 2.207906165586938e-06, + "loss": 2.844, + "step": 26059 + }, + { + "epoch": 2.755922165820643, + "grad_norm": 9.0, + "learning_rate": 2.2077360829764154e-06, + "loss": 2.3637, + "step": 26060 + }, + { + "epoch": 2.756027918781726, + "grad_norm": 8.5625, + "learning_rate": 2.207566001737367e-06, + "loss": 2.5167, + "step": 26061 + }, + { + "epoch": 2.7561336717428087, + "grad_norm": 8.5625, + "learning_rate": 2.207395921870591e-06, + "loss": 2.6222, + "step": 26062 + }, + { + "epoch": 2.7562394247038915, + "grad_norm": 8.6875, + "learning_rate": 2.207225843376888e-06, + "loss": 2.5231, + "step": 26063 + }, + { + "epoch": 2.7563451776649748, + "grad_norm": 8.4375, + "learning_rate": 2.207055766257053e-06, + "loss": 2.5574, + "step": 26064 + }, + { + "epoch": 2.7564509306260576, + "grad_norm": 8.625, + "learning_rate": 2.206885690511886e-06, + "loss": 2.4468, + "step": 26065 + }, + { + "epoch": 2.7565566835871405, + "grad_norm": 8.6875, + "learning_rate": 2.206715616142185e-06, + "loss": 2.7205, + "step": 26066 + }, + { + "epoch": 2.7566624365482233, + "grad_norm": 8.5, + "learning_rate": 2.206545543148748e-06, + "loss": 2.7391, + "step": 26067 + }, + { + "epoch": 2.756768189509306, + "grad_norm": 8.1875, + "learning_rate": 2.206375471532371e-06, + "loss": 2.551, + "step": 26068 + }, + { + "epoch": 2.7568739424703894, + "grad_norm": 9.0, + "learning_rate": 2.206205401293856e-06, + "loss": 2.3041, + "step": 26069 + }, + { + "epoch": 2.756979695431472, + "grad_norm": 8.375, + "learning_rate": 2.2060353324339972e-06, + "loss": 2.5975, + "step": 26070 + }, + { + "epoch": 2.757085448392555, + "grad_norm": 8.875, + "learning_rate": 2.205865264953595e-06, + "loss": 2.8337, + "step": 26071 + }, + { + "epoch": 2.757191201353638, + "grad_norm": 8.5625, + "learning_rate": 2.2056951988534467e-06, + "loss": 2.5572, + "step": 26072 + }, + { + "epoch": 2.7572969543147208, + "grad_norm": 9.6875, + "learning_rate": 2.2055251341343502e-06, + "loss": 2.651, + "step": 26073 + }, + { + "epoch": 2.7574027072758036, + "grad_norm": 8.1875, + "learning_rate": 2.2053550707971045e-06, + "loss": 2.7172, + "step": 26074 + }, + { + "epoch": 2.7575084602368864, + "grad_norm": 8.625, + "learning_rate": 2.2051850088425063e-06, + "loss": 2.4091, + "step": 26075 + }, + { + "epoch": 2.7576142131979697, + "grad_norm": 8.375, + "learning_rate": 2.205014948271355e-06, + "loss": 2.6397, + "step": 26076 + }, + { + "epoch": 2.7577199661590526, + "grad_norm": 8.125, + "learning_rate": 2.2048448890844472e-06, + "loss": 2.1842, + "step": 26077 + }, + { + "epoch": 2.7578257191201354, + "grad_norm": 8.3125, + "learning_rate": 2.204674831282582e-06, + "loss": 2.5829, + "step": 26078 + }, + { + "epoch": 2.7579314720812182, + "grad_norm": 8.125, + "learning_rate": 2.204504774866557e-06, + "loss": 2.4676, + "step": 26079 + }, + { + "epoch": 2.758037225042301, + "grad_norm": 8.625, + "learning_rate": 2.2043347198371697e-06, + "loss": 2.4326, + "step": 26080 + }, + { + "epoch": 2.7581429780033844, + "grad_norm": 8.6875, + "learning_rate": 2.2041646661952194e-06, + "loss": 2.4983, + "step": 26081 + }, + { + "epoch": 2.7582487309644668, + "grad_norm": 8.625, + "learning_rate": 2.2039946139415028e-06, + "loss": 2.5756, + "step": 26082 + }, + { + "epoch": 2.75835448392555, + "grad_norm": 9.0625, + "learning_rate": 2.2038245630768194e-06, + "loss": 2.6641, + "step": 26083 + }, + { + "epoch": 2.758460236886633, + "grad_norm": 8.1875, + "learning_rate": 2.2036545136019645e-06, + "loss": 2.4967, + "step": 26084 + }, + { + "epoch": 2.7585659898477157, + "grad_norm": 8.4375, + "learning_rate": 2.2034844655177396e-06, + "loss": 2.7658, + "step": 26085 + }, + { + "epoch": 2.7586717428087986, + "grad_norm": 8.25, + "learning_rate": 2.2033144188249395e-06, + "loss": 2.3524, + "step": 26086 + }, + { + "epoch": 2.7587774957698814, + "grad_norm": 7.875, + "learning_rate": 2.203144373524364e-06, + "loss": 2.3819, + "step": 26087 + }, + { + "epoch": 2.7588832487309647, + "grad_norm": 8.625, + "learning_rate": 2.20297432961681e-06, + "loss": 2.6689, + "step": 26088 + }, + { + "epoch": 2.7589890016920475, + "grad_norm": 8.4375, + "learning_rate": 2.202804287103076e-06, + "loss": 2.624, + "step": 26089 + }, + { + "epoch": 2.7590947546531304, + "grad_norm": 8.5, + "learning_rate": 2.2026342459839612e-06, + "loss": 2.5681, + "step": 26090 + }, + { + "epoch": 2.759200507614213, + "grad_norm": 8.75, + "learning_rate": 2.202464206260261e-06, + "loss": 2.4685, + "step": 26091 + }, + { + "epoch": 2.759306260575296, + "grad_norm": 8.125, + "learning_rate": 2.2022941679327757e-06, + "loss": 2.404, + "step": 26092 + }, + { + "epoch": 2.759412013536379, + "grad_norm": 8.125, + "learning_rate": 2.2021241310023014e-06, + "loss": 2.6059, + "step": 26093 + }, + { + "epoch": 2.7595177664974617, + "grad_norm": 8.5, + "learning_rate": 2.201954095469637e-06, + "loss": 2.6437, + "step": 26094 + }, + { + "epoch": 2.759623519458545, + "grad_norm": 8.375, + "learning_rate": 2.2017840613355796e-06, + "loss": 2.6362, + "step": 26095 + }, + { + "epoch": 2.759729272419628, + "grad_norm": 10.1875, + "learning_rate": 2.2016140286009282e-06, + "loss": 2.534, + "step": 26096 + }, + { + "epoch": 2.7598350253807107, + "grad_norm": 8.6875, + "learning_rate": 2.2014439972664796e-06, + "loss": 2.4714, + "step": 26097 + }, + { + "epoch": 2.7599407783417935, + "grad_norm": 8.25, + "learning_rate": 2.2012739673330323e-06, + "loss": 2.6384, + "step": 26098 + }, + { + "epoch": 2.7600465313028764, + "grad_norm": 9.1875, + "learning_rate": 2.201103938801385e-06, + "loss": 2.5696, + "step": 26099 + }, + { + "epoch": 2.7601522842639596, + "grad_norm": 8.25, + "learning_rate": 2.200933911672333e-06, + "loss": 2.6319, + "step": 26100 + }, + { + "epoch": 2.760258037225042, + "grad_norm": 8.8125, + "learning_rate": 2.2007638859466774e-06, + "loss": 2.5481, + "step": 26101 + }, + { + "epoch": 2.7603637901861253, + "grad_norm": 8.75, + "learning_rate": 2.200593861625214e-06, + "loss": 2.656, + "step": 26102 + }, + { + "epoch": 2.760469543147208, + "grad_norm": 8.5, + "learning_rate": 2.2004238387087413e-06, + "loss": 2.4512, + "step": 26103 + }, + { + "epoch": 2.760575296108291, + "grad_norm": 8.625, + "learning_rate": 2.200253817198056e-06, + "loss": 2.6013, + "step": 26104 + }, + { + "epoch": 2.760681049069374, + "grad_norm": 8.375, + "learning_rate": 2.2000837970939587e-06, + "loss": 2.4703, + "step": 26105 + }, + { + "epoch": 2.7607868020304567, + "grad_norm": 8.1875, + "learning_rate": 2.1999137783972433e-06, + "loss": 2.5387, + "step": 26106 + }, + { + "epoch": 2.76089255499154, + "grad_norm": 8.1875, + "learning_rate": 2.199743761108711e-06, + "loss": 2.2235, + "step": 26107 + }, + { + "epoch": 2.760998307952623, + "grad_norm": 8.9375, + "learning_rate": 2.1995737452291593e-06, + "loss": 2.743, + "step": 26108 + }, + { + "epoch": 2.7611040609137056, + "grad_norm": 8.5625, + "learning_rate": 2.199403730759384e-06, + "loss": 2.7062, + "step": 26109 + }, + { + "epoch": 2.7612098138747885, + "grad_norm": 8.4375, + "learning_rate": 2.1992337177001844e-06, + "loss": 2.4538, + "step": 26110 + }, + { + "epoch": 2.7613155668358713, + "grad_norm": 8.25, + "learning_rate": 2.1990637060523577e-06, + "loss": 2.6125, + "step": 26111 + }, + { + "epoch": 2.761421319796954, + "grad_norm": 8.4375, + "learning_rate": 2.1988936958167023e-06, + "loss": 2.6931, + "step": 26112 + }, + { + "epoch": 2.761527072758037, + "grad_norm": 8.875, + "learning_rate": 2.198723686994015e-06, + "loss": 2.5125, + "step": 26113 + }, + { + "epoch": 2.7616328257191203, + "grad_norm": 8.625, + "learning_rate": 2.1985536795850954e-06, + "loss": 2.4521, + "step": 26114 + }, + { + "epoch": 2.761738578680203, + "grad_norm": 8.5, + "learning_rate": 2.1983836735907394e-06, + "loss": 2.4699, + "step": 26115 + }, + { + "epoch": 2.761844331641286, + "grad_norm": 8.25, + "learning_rate": 2.1982136690117444e-06, + "loss": 2.5979, + "step": 26116 + }, + { + "epoch": 2.761950084602369, + "grad_norm": 8.8125, + "learning_rate": 2.198043665848911e-06, + "loss": 2.5708, + "step": 26117 + }, + { + "epoch": 2.7620558375634516, + "grad_norm": 8.8125, + "learning_rate": 2.197873664103034e-06, + "loss": 2.6542, + "step": 26118 + }, + { + "epoch": 2.762161590524535, + "grad_norm": 8.4375, + "learning_rate": 2.197703663774913e-06, + "loss": 2.6022, + "step": 26119 + }, + { + "epoch": 2.7622673434856178, + "grad_norm": 8.3125, + "learning_rate": 2.197533664865345e-06, + "loss": 2.4826, + "step": 26120 + }, + { + "epoch": 2.7623730964467006, + "grad_norm": 8.5, + "learning_rate": 2.197363667375128e-06, + "loss": 2.5243, + "step": 26121 + }, + { + "epoch": 2.7624788494077834, + "grad_norm": 8.6875, + "learning_rate": 2.1971936713050584e-06, + "loss": 2.5677, + "step": 26122 + }, + { + "epoch": 2.7625846023688663, + "grad_norm": 8.75, + "learning_rate": 2.1970236766559353e-06, + "loss": 2.5159, + "step": 26123 + }, + { + "epoch": 2.762690355329949, + "grad_norm": 9.3125, + "learning_rate": 2.196853683428558e-06, + "loss": 2.5049, + "step": 26124 + }, + { + "epoch": 2.762796108291032, + "grad_norm": 8.25, + "learning_rate": 2.1966836916237206e-06, + "loss": 2.584, + "step": 26125 + }, + { + "epoch": 2.7629018612521152, + "grad_norm": 8.0625, + "learning_rate": 2.196513701242223e-06, + "loss": 2.5023, + "step": 26126 + }, + { + "epoch": 2.763007614213198, + "grad_norm": 8.25, + "learning_rate": 2.1963437122848625e-06, + "loss": 2.5932, + "step": 26127 + }, + { + "epoch": 2.763113367174281, + "grad_norm": 9.1875, + "learning_rate": 2.196173724752437e-06, + "loss": 2.739, + "step": 26128 + }, + { + "epoch": 2.7632191201353637, + "grad_norm": 8.4375, + "learning_rate": 2.1960037386457436e-06, + "loss": 2.6235, + "step": 26129 + }, + { + "epoch": 2.7633248730964466, + "grad_norm": 8.4375, + "learning_rate": 2.195833753965581e-06, + "loss": 2.5632, + "step": 26130 + }, + { + "epoch": 2.76343062605753, + "grad_norm": 8.3125, + "learning_rate": 2.1956637707127457e-06, + "loss": 2.48, + "step": 26131 + }, + { + "epoch": 2.7635363790186123, + "grad_norm": 9.1875, + "learning_rate": 2.1954937888880347e-06, + "loss": 2.6347, + "step": 26132 + }, + { + "epoch": 2.7636421319796955, + "grad_norm": 9.0625, + "learning_rate": 2.195323808492249e-06, + "loss": 2.6526, + "step": 26133 + }, + { + "epoch": 2.7637478849407784, + "grad_norm": 8.5, + "learning_rate": 2.1951538295261827e-06, + "loss": 2.5619, + "step": 26134 + }, + { + "epoch": 2.7638536379018612, + "grad_norm": 8.625, + "learning_rate": 2.1949838519906352e-06, + "loss": 2.4749, + "step": 26135 + }, + { + "epoch": 2.763959390862944, + "grad_norm": 8.625, + "learning_rate": 2.1948138758864036e-06, + "loss": 2.4684, + "step": 26136 + }, + { + "epoch": 2.764065143824027, + "grad_norm": 7.9375, + "learning_rate": 2.1946439012142862e-06, + "loss": 2.3945, + "step": 26137 + }, + { + "epoch": 2.76417089678511, + "grad_norm": 9.3125, + "learning_rate": 2.194473927975079e-06, + "loss": 2.7288, + "step": 26138 + }, + { + "epoch": 2.764276649746193, + "grad_norm": 8.6875, + "learning_rate": 2.1943039561695818e-06, + "loss": 2.7188, + "step": 26139 + }, + { + "epoch": 2.764382402707276, + "grad_norm": 8.5625, + "learning_rate": 2.19413398579859e-06, + "loss": 2.4421, + "step": 26140 + }, + { + "epoch": 2.7644881556683587, + "grad_norm": 7.90625, + "learning_rate": 2.193964016862903e-06, + "loss": 2.466, + "step": 26141 + }, + { + "epoch": 2.7645939086294415, + "grad_norm": 9.0625, + "learning_rate": 2.193794049363318e-06, + "loss": 2.3673, + "step": 26142 + }, + { + "epoch": 2.7646996615905244, + "grad_norm": 8.6875, + "learning_rate": 2.1936240833006313e-06, + "loss": 2.6896, + "step": 26143 + }, + { + "epoch": 2.764805414551607, + "grad_norm": 8.375, + "learning_rate": 2.1934541186756427e-06, + "loss": 2.5871, + "step": 26144 + }, + { + "epoch": 2.7649111675126905, + "grad_norm": 8.375, + "learning_rate": 2.193284155489147e-06, + "loss": 2.4978, + "step": 26145 + }, + { + "epoch": 2.7650169204737733, + "grad_norm": 8.75, + "learning_rate": 2.1931141937419453e-06, + "loss": 2.6491, + "step": 26146 + }, + { + "epoch": 2.765122673434856, + "grad_norm": 8.8125, + "learning_rate": 2.192944233434832e-06, + "loss": 2.5392, + "step": 26147 + }, + { + "epoch": 2.765228426395939, + "grad_norm": 9.125, + "learning_rate": 2.192774274568606e-06, + "loss": 2.4494, + "step": 26148 + }, + { + "epoch": 2.765334179357022, + "grad_norm": 8.6875, + "learning_rate": 2.1926043171440645e-06, + "loss": 2.5972, + "step": 26149 + }, + { + "epoch": 2.765439932318105, + "grad_norm": 8.6875, + "learning_rate": 2.1924343611620048e-06, + "loss": 2.7493, + "step": 26150 + }, + { + "epoch": 2.765545685279188, + "grad_norm": 8.5625, + "learning_rate": 2.1922644066232257e-06, + "loss": 2.4796, + "step": 26151 + }, + { + "epoch": 2.765651438240271, + "grad_norm": 8.5, + "learning_rate": 2.1920944535285234e-06, + "loss": 2.6105, + "step": 26152 + }, + { + "epoch": 2.7657571912013537, + "grad_norm": 8.4375, + "learning_rate": 2.1919245018786967e-06, + "loss": 2.6569, + "step": 26153 + }, + { + "epoch": 2.7658629441624365, + "grad_norm": 7.9375, + "learning_rate": 2.191754551674541e-06, + "loss": 2.5831, + "step": 26154 + }, + { + "epoch": 2.7659686971235193, + "grad_norm": 8.0625, + "learning_rate": 2.1915846029168566e-06, + "loss": 2.4035, + "step": 26155 + }, + { + "epoch": 2.766074450084602, + "grad_norm": 8.75, + "learning_rate": 2.1914146556064385e-06, + "loss": 2.5516, + "step": 26156 + }, + { + "epoch": 2.7661802030456855, + "grad_norm": 8.4375, + "learning_rate": 2.1912447097440856e-06, + "loss": 2.3891, + "step": 26157 + }, + { + "epoch": 2.7662859560067683, + "grad_norm": 7.9375, + "learning_rate": 2.1910747653305943e-06, + "loss": 2.4603, + "step": 26158 + }, + { + "epoch": 2.766391708967851, + "grad_norm": 8.75, + "learning_rate": 2.190904822366763e-06, + "loss": 2.4924, + "step": 26159 + }, + { + "epoch": 2.766497461928934, + "grad_norm": 8.8125, + "learning_rate": 2.1907348808533895e-06, + "loss": 2.6033, + "step": 26160 + }, + { + "epoch": 2.766603214890017, + "grad_norm": 8.25, + "learning_rate": 2.1905649407912704e-06, + "loss": 2.5362, + "step": 26161 + }, + { + "epoch": 2.7667089678511, + "grad_norm": 8.3125, + "learning_rate": 2.190395002181204e-06, + "loss": 2.6486, + "step": 26162 + }, + { + "epoch": 2.7668147208121825, + "grad_norm": 8.5, + "learning_rate": 2.190225065023987e-06, + "loss": 2.6252, + "step": 26163 + }, + { + "epoch": 2.7669204737732658, + "grad_norm": 8.1875, + "learning_rate": 2.190055129320417e-06, + "loss": 2.3765, + "step": 26164 + }, + { + "epoch": 2.7670262267343486, + "grad_norm": 8.3125, + "learning_rate": 2.189885195071291e-06, + "loss": 2.5855, + "step": 26165 + }, + { + "epoch": 2.7671319796954315, + "grad_norm": 8.25, + "learning_rate": 2.1897152622774076e-06, + "loss": 2.4133, + "step": 26166 + }, + { + "epoch": 2.7672377326565143, + "grad_norm": 8.125, + "learning_rate": 2.189545330939563e-06, + "loss": 2.538, + "step": 26167 + }, + { + "epoch": 2.767343485617597, + "grad_norm": 8.8125, + "learning_rate": 2.189375401058555e-06, + "loss": 2.5935, + "step": 26168 + }, + { + "epoch": 2.7674492385786804, + "grad_norm": 8.75, + "learning_rate": 2.189205472635183e-06, + "loss": 2.4608, + "step": 26169 + }, + { + "epoch": 2.7675549915397633, + "grad_norm": 8.5625, + "learning_rate": 2.1890355456702404e-06, + "loss": 2.5155, + "step": 26170 + }, + { + "epoch": 2.767660744500846, + "grad_norm": 8.5, + "learning_rate": 2.1888656201645282e-06, + "loss": 2.5109, + "step": 26171 + }, + { + "epoch": 2.767766497461929, + "grad_norm": 8.75, + "learning_rate": 2.188695696118842e-06, + "loss": 2.5884, + "step": 26172 + }, + { + "epoch": 2.7678722504230118, + "grad_norm": 8.5, + "learning_rate": 2.1885257735339794e-06, + "loss": 2.4092, + "step": 26173 + }, + { + "epoch": 2.7679780033840946, + "grad_norm": 8.0625, + "learning_rate": 2.1883558524107377e-06, + "loss": 2.6117, + "step": 26174 + }, + { + "epoch": 2.7680837563451774, + "grad_norm": 7.78125, + "learning_rate": 2.188185932749916e-06, + "loss": 2.3664, + "step": 26175 + }, + { + "epoch": 2.7681895093062607, + "grad_norm": 8.1875, + "learning_rate": 2.188016014552308e-06, + "loss": 2.4845, + "step": 26176 + }, + { + "epoch": 2.7682952622673436, + "grad_norm": 8.5, + "learning_rate": 2.187846097818714e-06, + "loss": 2.6235, + "step": 26177 + }, + { + "epoch": 2.7684010152284264, + "grad_norm": 8.5625, + "learning_rate": 2.187676182549932e-06, + "loss": 2.5571, + "step": 26178 + }, + { + "epoch": 2.7685067681895092, + "grad_norm": 8.875, + "learning_rate": 2.187506268746756e-06, + "loss": 2.6993, + "step": 26179 + }, + { + "epoch": 2.768612521150592, + "grad_norm": 9.4375, + "learning_rate": 2.1873363564099868e-06, + "loss": 2.2548, + "step": 26180 + }, + { + "epoch": 2.7687182741116754, + "grad_norm": 8.75, + "learning_rate": 2.187166445540419e-06, + "loss": 2.3138, + "step": 26181 + }, + { + "epoch": 2.7688240270727578, + "grad_norm": 8.4375, + "learning_rate": 2.186996536138852e-06, + "loss": 2.7384, + "step": 26182 + }, + { + "epoch": 2.768929780033841, + "grad_norm": 8.875, + "learning_rate": 2.1868266282060814e-06, + "loss": 2.7517, + "step": 26183 + }, + { + "epoch": 2.769035532994924, + "grad_norm": 8.4375, + "learning_rate": 2.186656721742907e-06, + "loss": 2.5854, + "step": 26184 + }, + { + "epoch": 2.7691412859560067, + "grad_norm": 8.375, + "learning_rate": 2.186486816750123e-06, + "loss": 2.42, + "step": 26185 + }, + { + "epoch": 2.7692470389170896, + "grad_norm": 8.5, + "learning_rate": 2.1863169132285273e-06, + "loss": 2.6553, + "step": 26186 + }, + { + "epoch": 2.7693527918781724, + "grad_norm": 8.0625, + "learning_rate": 2.18614701117892e-06, + "loss": 2.4307, + "step": 26187 + }, + { + "epoch": 2.7694585448392557, + "grad_norm": 8.0, + "learning_rate": 2.1859771106020957e-06, + "loss": 2.4485, + "step": 26188 + }, + { + "epoch": 2.7695642978003385, + "grad_norm": 9.0, + "learning_rate": 2.185807211498852e-06, + "loss": 2.5102, + "step": 26189 + }, + { + "epoch": 2.7696700507614214, + "grad_norm": 8.5, + "learning_rate": 2.1856373138699867e-06, + "loss": 2.794, + "step": 26190 + }, + { + "epoch": 2.769775803722504, + "grad_norm": 10.8125, + "learning_rate": 2.185467417716298e-06, + "loss": 2.4933, + "step": 26191 + }, + { + "epoch": 2.769881556683587, + "grad_norm": 8.375, + "learning_rate": 2.1852975230385803e-06, + "loss": 2.5375, + "step": 26192 + }, + { + "epoch": 2.7699873096446703, + "grad_norm": 8.5625, + "learning_rate": 2.185127629837634e-06, + "loss": 2.5854, + "step": 26193 + }, + { + "epoch": 2.7700930626057527, + "grad_norm": 8.9375, + "learning_rate": 2.184957738114254e-06, + "loss": 2.5826, + "step": 26194 + }, + { + "epoch": 2.770198815566836, + "grad_norm": 8.25, + "learning_rate": 2.184787847869239e-06, + "loss": 2.5223, + "step": 26195 + }, + { + "epoch": 2.770304568527919, + "grad_norm": 8.125, + "learning_rate": 2.1846179591033857e-06, + "loss": 2.5681, + "step": 26196 + }, + { + "epoch": 2.7704103214890017, + "grad_norm": 8.9375, + "learning_rate": 2.1844480718174908e-06, + "loss": 2.4539, + "step": 26197 + }, + { + "epoch": 2.7705160744500845, + "grad_norm": 8.8125, + "learning_rate": 2.1842781860123526e-06, + "loss": 2.7449, + "step": 26198 + }, + { + "epoch": 2.7706218274111674, + "grad_norm": 8.125, + "learning_rate": 2.1841083016887677e-06, + "loss": 2.5239, + "step": 26199 + }, + { + "epoch": 2.7707275803722506, + "grad_norm": 8.8125, + "learning_rate": 2.1839384188475337e-06, + "loss": 2.4988, + "step": 26200 + }, + { + "epoch": 2.7708333333333335, + "grad_norm": 8.25, + "learning_rate": 2.183768537489447e-06, + "loss": 2.2794, + "step": 26201 + }, + { + "epoch": 2.7709390862944163, + "grad_norm": 8.25, + "learning_rate": 2.1835986576153053e-06, + "loss": 2.6538, + "step": 26202 + }, + { + "epoch": 2.771044839255499, + "grad_norm": 8.1875, + "learning_rate": 2.1834287792259054e-06, + "loss": 2.4606, + "step": 26203 + }, + { + "epoch": 2.771150592216582, + "grad_norm": 8.0625, + "learning_rate": 2.1832589023220446e-06, + "loss": 2.5765, + "step": 26204 + }, + { + "epoch": 2.771256345177665, + "grad_norm": 8.1875, + "learning_rate": 2.1830890269045208e-06, + "loss": 2.6242, + "step": 26205 + }, + { + "epoch": 2.7713620981387477, + "grad_norm": 8.625, + "learning_rate": 2.18291915297413e-06, + "loss": 2.5942, + "step": 26206 + }, + { + "epoch": 2.771467851099831, + "grad_norm": 8.125, + "learning_rate": 2.1827492805316715e-06, + "loss": 2.4878, + "step": 26207 + }, + { + "epoch": 2.771573604060914, + "grad_norm": 8.25, + "learning_rate": 2.1825794095779384e-06, + "loss": 2.5211, + "step": 26208 + }, + { + "epoch": 2.7716793570219966, + "grad_norm": 8.25, + "learning_rate": 2.1824095401137327e-06, + "loss": 2.5965, + "step": 26209 + }, + { + "epoch": 2.7717851099830795, + "grad_norm": 7.75, + "learning_rate": 2.1822396721398476e-06, + "loss": 2.4882, + "step": 26210 + }, + { + "epoch": 2.7718908629441623, + "grad_norm": 8.5, + "learning_rate": 2.1820698056570823e-06, + "loss": 2.4116, + "step": 26211 + }, + { + "epoch": 2.7719966159052456, + "grad_norm": 8.5, + "learning_rate": 2.181899940666233e-06, + "loss": 2.4981, + "step": 26212 + }, + { + "epoch": 2.772102368866328, + "grad_norm": 8.1875, + "learning_rate": 2.181730077168097e-06, + "loss": 2.7171, + "step": 26213 + }, + { + "epoch": 2.7722081218274113, + "grad_norm": 8.875, + "learning_rate": 2.1815602151634725e-06, + "loss": 2.8105, + "step": 26214 + }, + { + "epoch": 2.772313874788494, + "grad_norm": 9.0625, + "learning_rate": 2.181390354653155e-06, + "loss": 2.3942, + "step": 26215 + }, + { + "epoch": 2.772419627749577, + "grad_norm": 8.5, + "learning_rate": 2.181220495637943e-06, + "loss": 2.9587, + "step": 26216 + }, + { + "epoch": 2.77252538071066, + "grad_norm": 8.375, + "learning_rate": 2.181050638118632e-06, + "loss": 2.6951, + "step": 26217 + }, + { + "epoch": 2.7726311336717426, + "grad_norm": 8.1875, + "learning_rate": 2.1808807820960206e-06, + "loss": 2.6451, + "step": 26218 + }, + { + "epoch": 2.772736886632826, + "grad_norm": 8.375, + "learning_rate": 2.1807109275709045e-06, + "loss": 2.49, + "step": 26219 + }, + { + "epoch": 2.7728426395939088, + "grad_norm": 8.5625, + "learning_rate": 2.180541074544082e-06, + "loss": 2.562, + "step": 26220 + }, + { + "epoch": 2.7729483925549916, + "grad_norm": 8.3125, + "learning_rate": 2.180371223016349e-06, + "loss": 2.2958, + "step": 26221 + }, + { + "epoch": 2.7730541455160744, + "grad_norm": 8.3125, + "learning_rate": 2.1802013729885033e-06, + "loss": 2.4862, + "step": 26222 + }, + { + "epoch": 2.7731598984771573, + "grad_norm": 8.25, + "learning_rate": 2.1800315244613427e-06, + "loss": 2.5288, + "step": 26223 + }, + { + "epoch": 2.77326565143824, + "grad_norm": 7.90625, + "learning_rate": 2.1798616774356615e-06, + "loss": 2.4526, + "step": 26224 + }, + { + "epoch": 2.773371404399323, + "grad_norm": 8.8125, + "learning_rate": 2.17969183191226e-06, + "loss": 2.5921, + "step": 26225 + }, + { + "epoch": 2.7734771573604062, + "grad_norm": 8.875, + "learning_rate": 2.179521987891933e-06, + "loss": 2.6292, + "step": 26226 + }, + { + "epoch": 2.773582910321489, + "grad_norm": 8.6875, + "learning_rate": 2.179352145375479e-06, + "loss": 2.6174, + "step": 26227 + }, + { + "epoch": 2.773688663282572, + "grad_norm": 8.1875, + "learning_rate": 2.1791823043636935e-06, + "loss": 2.7078, + "step": 26228 + }, + { + "epoch": 2.7737944162436547, + "grad_norm": 8.75, + "learning_rate": 2.1790124648573753e-06, + "loss": 2.676, + "step": 26229 + }, + { + "epoch": 2.7739001692047376, + "grad_norm": 8.3125, + "learning_rate": 2.1788426268573185e-06, + "loss": 2.4107, + "step": 26230 + }, + { + "epoch": 2.774005922165821, + "grad_norm": 8.3125, + "learning_rate": 2.1786727903643226e-06, + "loss": 2.4688, + "step": 26231 + }, + { + "epoch": 2.7741116751269037, + "grad_norm": 8.0625, + "learning_rate": 2.1785029553791847e-06, + "loss": 2.4541, + "step": 26232 + }, + { + "epoch": 2.7742174280879865, + "grad_norm": 7.84375, + "learning_rate": 2.1783331219027005e-06, + "loss": 2.4797, + "step": 26233 + }, + { + "epoch": 2.7743231810490694, + "grad_norm": 8.1875, + "learning_rate": 2.1781632899356676e-06, + "loss": 2.445, + "step": 26234 + }, + { + "epoch": 2.7744289340101522, + "grad_norm": 8.25, + "learning_rate": 2.177993459478882e-06, + "loss": 2.5487, + "step": 26235 + }, + { + "epoch": 2.774534686971235, + "grad_norm": 8.5, + "learning_rate": 2.177823630533142e-06, + "loss": 2.6824, + "step": 26236 + }, + { + "epoch": 2.774640439932318, + "grad_norm": 8.375, + "learning_rate": 2.177653803099243e-06, + "loss": 2.7885, + "step": 26237 + }, + { + "epoch": 2.774746192893401, + "grad_norm": 8.5, + "learning_rate": 2.1774839771779838e-06, + "loss": 2.5875, + "step": 26238 + }, + { + "epoch": 2.774851945854484, + "grad_norm": 8.0625, + "learning_rate": 2.177314152770161e-06, + "loss": 2.5209, + "step": 26239 + }, + { + "epoch": 2.774957698815567, + "grad_norm": 8.4375, + "learning_rate": 2.1771443298765695e-06, + "loss": 2.5893, + "step": 26240 + }, + { + "epoch": 2.7750634517766497, + "grad_norm": 7.84375, + "learning_rate": 2.176974508498009e-06, + "loss": 2.3762, + "step": 26241 + }, + { + "epoch": 2.7751692047377325, + "grad_norm": 8.625, + "learning_rate": 2.176804688635274e-06, + "loss": 2.4422, + "step": 26242 + }, + { + "epoch": 2.775274957698816, + "grad_norm": 8.625, + "learning_rate": 2.1766348702891628e-06, + "loss": 2.5516, + "step": 26243 + }, + { + "epoch": 2.775380710659898, + "grad_norm": 8.25, + "learning_rate": 2.1764650534604717e-06, + "loss": 2.4754, + "step": 26244 + }, + { + "epoch": 2.7754864636209815, + "grad_norm": 8.9375, + "learning_rate": 2.1762952381499986e-06, + "loss": 2.6976, + "step": 26245 + }, + { + "epoch": 2.7755922165820643, + "grad_norm": 8.25, + "learning_rate": 2.176125424358538e-06, + "loss": 2.4974, + "step": 26246 + }, + { + "epoch": 2.775697969543147, + "grad_norm": 8.75, + "learning_rate": 2.175955612086889e-06, + "loss": 2.5289, + "step": 26247 + }, + { + "epoch": 2.77580372250423, + "grad_norm": 8.0, + "learning_rate": 2.1757858013358486e-06, + "loss": 2.4766, + "step": 26248 + }, + { + "epoch": 2.775909475465313, + "grad_norm": 8.75, + "learning_rate": 2.175615992106212e-06, + "loss": 2.802, + "step": 26249 + }, + { + "epoch": 2.776015228426396, + "grad_norm": 8.25, + "learning_rate": 2.1754461843987774e-06, + "loss": 2.3457, + "step": 26250 + }, + { + "epoch": 2.776120981387479, + "grad_norm": 8.1875, + "learning_rate": 2.1752763782143402e-06, + "loss": 2.448, + "step": 26251 + }, + { + "epoch": 2.776226734348562, + "grad_norm": 9.1875, + "learning_rate": 2.1751065735536993e-06, + "loss": 2.5544, + "step": 26252 + }, + { + "epoch": 2.7763324873096447, + "grad_norm": 8.625, + "learning_rate": 2.1749367704176494e-06, + "loss": 2.5425, + "step": 26253 + }, + { + "epoch": 2.7764382402707275, + "grad_norm": 8.75, + "learning_rate": 2.1747669688069894e-06, + "loss": 2.5478, + "step": 26254 + }, + { + "epoch": 2.7765439932318103, + "grad_norm": 8.125, + "learning_rate": 2.174597168722514e-06, + "loss": 2.5036, + "step": 26255 + }, + { + "epoch": 2.776649746192893, + "grad_norm": 8.125, + "learning_rate": 2.1744273701650205e-06, + "loss": 2.5309, + "step": 26256 + }, + { + "epoch": 2.7767554991539765, + "grad_norm": 8.1875, + "learning_rate": 2.174257573135308e-06, + "loss": 2.5653, + "step": 26257 + }, + { + "epoch": 2.7768612521150593, + "grad_norm": 8.25, + "learning_rate": 2.17408777763417e-06, + "loss": 2.4714, + "step": 26258 + }, + { + "epoch": 2.776967005076142, + "grad_norm": 13.9375, + "learning_rate": 2.1739179836624054e-06, + "loss": 2.4684, + "step": 26259 + }, + { + "epoch": 2.777072758037225, + "grad_norm": 8.1875, + "learning_rate": 2.1737481912208097e-06, + "loss": 2.3551, + "step": 26260 + }, + { + "epoch": 2.777178510998308, + "grad_norm": 8.125, + "learning_rate": 2.1735784003101816e-06, + "loss": 2.4815, + "step": 26261 + }, + { + "epoch": 2.777284263959391, + "grad_norm": 8.9375, + "learning_rate": 2.173408610931315e-06, + "loss": 2.5555, + "step": 26262 + }, + { + "epoch": 2.777390016920474, + "grad_norm": 8.9375, + "learning_rate": 2.17323882308501e-06, + "loss": 2.6616, + "step": 26263 + }, + { + "epoch": 2.7774957698815568, + "grad_norm": 8.25, + "learning_rate": 2.17306903677206e-06, + "loss": 2.4407, + "step": 26264 + }, + { + "epoch": 2.7776015228426396, + "grad_norm": 8.1875, + "learning_rate": 2.1728992519932634e-06, + "loss": 2.3575, + "step": 26265 + }, + { + "epoch": 2.7777072758037225, + "grad_norm": 8.75, + "learning_rate": 2.1727294687494176e-06, + "loss": 2.4453, + "step": 26266 + }, + { + "epoch": 2.7778130287648053, + "grad_norm": 8.0, + "learning_rate": 2.172559687041318e-06, + "loss": 2.4879, + "step": 26267 + }, + { + "epoch": 2.777918781725888, + "grad_norm": 8.375, + "learning_rate": 2.1723899068697624e-06, + "loss": 2.4514, + "step": 26268 + }, + { + "epoch": 2.7780245346869714, + "grad_norm": 8.875, + "learning_rate": 2.172220128235546e-06, + "loss": 2.5963, + "step": 26269 + }, + { + "epoch": 2.7781302876480543, + "grad_norm": 7.90625, + "learning_rate": 2.172050351139468e-06, + "loss": 2.4383, + "step": 26270 + }, + { + "epoch": 2.778236040609137, + "grad_norm": 8.625, + "learning_rate": 2.1718805755823227e-06, + "loss": 2.4237, + "step": 26271 + }, + { + "epoch": 2.77834179357022, + "grad_norm": 8.4375, + "learning_rate": 2.171710801564908e-06, + "loss": 2.7402, + "step": 26272 + }, + { + "epoch": 2.7784475465313028, + "grad_norm": 8.4375, + "learning_rate": 2.1715410290880194e-06, + "loss": 2.5796, + "step": 26273 + }, + { + "epoch": 2.778553299492386, + "grad_norm": 8.3125, + "learning_rate": 2.1713712581524547e-06, + "loss": 2.3346, + "step": 26274 + }, + { + "epoch": 2.7786590524534684, + "grad_norm": 9.0, + "learning_rate": 2.171201488759011e-06, + "loss": 2.6224, + "step": 26275 + }, + { + "epoch": 2.7787648054145517, + "grad_norm": 8.5625, + "learning_rate": 2.1710317209084834e-06, + "loss": 2.6104, + "step": 26276 + }, + { + "epoch": 2.7788705583756346, + "grad_norm": 8.6875, + "learning_rate": 2.1708619546016705e-06, + "loss": 2.511, + "step": 26277 + }, + { + "epoch": 2.7789763113367174, + "grad_norm": 8.75, + "learning_rate": 2.170692189839366e-06, + "loss": 2.6857, + "step": 26278 + }, + { + "epoch": 2.7790820642978002, + "grad_norm": 8.8125, + "learning_rate": 2.1705224266223706e-06, + "loss": 2.5162, + "step": 26279 + }, + { + "epoch": 2.779187817258883, + "grad_norm": 8.625, + "learning_rate": 2.1703526649514774e-06, + "loss": 2.6864, + "step": 26280 + }, + { + "epoch": 2.7792935702199664, + "grad_norm": 8.125, + "learning_rate": 2.1701829048274845e-06, + "loss": 2.6887, + "step": 26281 + }, + { + "epoch": 2.779399323181049, + "grad_norm": 8.4375, + "learning_rate": 2.170013146251188e-06, + "loss": 2.5413, + "step": 26282 + }, + { + "epoch": 2.779505076142132, + "grad_norm": 8.5, + "learning_rate": 2.169843389223385e-06, + "loss": 2.5323, + "step": 26283 + }, + { + "epoch": 2.779610829103215, + "grad_norm": 8.6875, + "learning_rate": 2.1696736337448724e-06, + "loss": 2.5047, + "step": 26284 + }, + { + "epoch": 2.7797165820642977, + "grad_norm": 8.625, + "learning_rate": 2.1695038798164457e-06, + "loss": 2.4803, + "step": 26285 + }, + { + "epoch": 2.7798223350253806, + "grad_norm": 9.1875, + "learning_rate": 2.169334127438903e-06, + "loss": 2.6077, + "step": 26286 + }, + { + "epoch": 2.7799280879864634, + "grad_norm": 8.75, + "learning_rate": 2.1691643766130395e-06, + "loss": 2.6197, + "step": 26287 + }, + { + "epoch": 2.7800338409475467, + "grad_norm": 9.3125, + "learning_rate": 2.168994627339652e-06, + "loss": 2.5897, + "step": 26288 + }, + { + "epoch": 2.7801395939086295, + "grad_norm": 8.4375, + "learning_rate": 2.1688248796195377e-06, + "loss": 2.5163, + "step": 26289 + }, + { + "epoch": 2.7802453468697124, + "grad_norm": 8.5, + "learning_rate": 2.168655133453493e-06, + "loss": 2.4759, + "step": 26290 + }, + { + "epoch": 2.780351099830795, + "grad_norm": 8.5625, + "learning_rate": 2.1684853888423137e-06, + "loss": 2.373, + "step": 26291 + }, + { + "epoch": 2.780456852791878, + "grad_norm": 8.8125, + "learning_rate": 2.1683156457867966e-06, + "loss": 2.3823, + "step": 26292 + }, + { + "epoch": 2.7805626057529613, + "grad_norm": 8.125, + "learning_rate": 2.1681459042877398e-06, + "loss": 2.7055, + "step": 26293 + }, + { + "epoch": 2.7806683587140437, + "grad_norm": 8.0625, + "learning_rate": 2.167976164345937e-06, + "loss": 2.4718, + "step": 26294 + }, + { + "epoch": 2.780774111675127, + "grad_norm": 7.78125, + "learning_rate": 2.1678064259621876e-06, + "loss": 2.2859, + "step": 26295 + }, + { + "epoch": 2.78087986463621, + "grad_norm": 8.625, + "learning_rate": 2.167636689137286e-06, + "loss": 2.4591, + "step": 26296 + }, + { + "epoch": 2.7809856175972927, + "grad_norm": 8.3125, + "learning_rate": 2.16746695387203e-06, + "loss": 2.6019, + "step": 26297 + }, + { + "epoch": 2.7810913705583755, + "grad_norm": 8.625, + "learning_rate": 2.167297220167215e-06, + "loss": 2.6982, + "step": 26298 + }, + { + "epoch": 2.7811971235194584, + "grad_norm": 8.625, + "learning_rate": 2.167127488023639e-06, + "loss": 2.6106, + "step": 26299 + }, + { + "epoch": 2.7813028764805416, + "grad_norm": 8.125, + "learning_rate": 2.1669577574420963e-06, + "loss": 2.5407, + "step": 26300 + }, + { + "epoch": 2.7814086294416245, + "grad_norm": 9.5, + "learning_rate": 2.166788028423385e-06, + "loss": 2.5562, + "step": 26301 + }, + { + "epoch": 2.7815143824027073, + "grad_norm": 8.4375, + "learning_rate": 2.1666183009683023e-06, + "loss": 2.7244, + "step": 26302 + }, + { + "epoch": 2.78162013536379, + "grad_norm": 8.9375, + "learning_rate": 2.1664485750776423e-06, + "loss": 2.5561, + "step": 26303 + }, + { + "epoch": 2.781725888324873, + "grad_norm": 8.0625, + "learning_rate": 2.1662788507522034e-06, + "loss": 2.3943, + "step": 26304 + }, + { + "epoch": 2.7818316412859563, + "grad_norm": 8.125, + "learning_rate": 2.166109127992781e-06, + "loss": 2.4276, + "step": 26305 + }, + { + "epoch": 2.7819373942470387, + "grad_norm": 9.1875, + "learning_rate": 2.165939406800172e-06, + "loss": 2.5898, + "step": 26306 + }, + { + "epoch": 2.782043147208122, + "grad_norm": 8.625, + "learning_rate": 2.1657696871751727e-06, + "loss": 2.5471, + "step": 26307 + }, + { + "epoch": 2.782148900169205, + "grad_norm": 8.25, + "learning_rate": 2.1655999691185802e-06, + "loss": 2.3223, + "step": 26308 + }, + { + "epoch": 2.7822546531302876, + "grad_norm": 8.25, + "learning_rate": 2.1654302526311896e-06, + "loss": 2.5637, + "step": 26309 + }, + { + "epoch": 2.7823604060913705, + "grad_norm": 8.5, + "learning_rate": 2.1652605377137974e-06, + "loss": 2.5383, + "step": 26310 + }, + { + "epoch": 2.7824661590524533, + "grad_norm": 8.5, + "learning_rate": 2.165090824367202e-06, + "loss": 2.3566, + "step": 26311 + }, + { + "epoch": 2.7825719120135366, + "grad_norm": 8.6875, + "learning_rate": 2.1649211125921973e-06, + "loss": 2.7868, + "step": 26312 + }, + { + "epoch": 2.7826776649746194, + "grad_norm": 8.0625, + "learning_rate": 2.1647514023895815e-06, + "loss": 2.4133, + "step": 26313 + }, + { + "epoch": 2.7827834179357023, + "grad_norm": 8.3125, + "learning_rate": 2.1645816937601494e-06, + "loss": 2.5958, + "step": 26314 + }, + { + "epoch": 2.782889170896785, + "grad_norm": 9.375, + "learning_rate": 2.1644119867046997e-06, + "loss": 2.7142, + "step": 26315 + }, + { + "epoch": 2.782994923857868, + "grad_norm": 8.375, + "learning_rate": 2.1642422812240253e-06, + "loss": 2.493, + "step": 26316 + }, + { + "epoch": 2.783100676818951, + "grad_norm": 8.8125, + "learning_rate": 2.1640725773189262e-06, + "loss": 2.6217, + "step": 26317 + }, + { + "epoch": 2.7832064297800336, + "grad_norm": 8.5625, + "learning_rate": 2.1639028749901965e-06, + "loss": 2.4646, + "step": 26318 + }, + { + "epoch": 2.783312182741117, + "grad_norm": 8.4375, + "learning_rate": 2.163733174238633e-06, + "loss": 2.4823, + "step": 26319 + }, + { + "epoch": 2.7834179357021998, + "grad_norm": 8.5625, + "learning_rate": 2.1635634750650323e-06, + "loss": 2.5628, + "step": 26320 + }, + { + "epoch": 2.7835236886632826, + "grad_norm": 8.8125, + "learning_rate": 2.1633937774701906e-06, + "loss": 2.6237, + "step": 26321 + }, + { + "epoch": 2.7836294416243654, + "grad_norm": 8.375, + "learning_rate": 2.1632240814549044e-06, + "loss": 2.5123, + "step": 26322 + }, + { + "epoch": 2.7837351945854483, + "grad_norm": 8.1875, + "learning_rate": 2.1630543870199696e-06, + "loss": 2.5863, + "step": 26323 + }, + { + "epoch": 2.7838409475465316, + "grad_norm": 7.9375, + "learning_rate": 2.1628846941661835e-06, + "loss": 2.3636, + "step": 26324 + }, + { + "epoch": 2.783946700507614, + "grad_norm": 8.25, + "learning_rate": 2.1627150028943406e-06, + "loss": 2.5929, + "step": 26325 + }, + { + "epoch": 2.7840524534686972, + "grad_norm": 9.125, + "learning_rate": 2.162545313205239e-06, + "loss": 2.7818, + "step": 26326 + }, + { + "epoch": 2.78415820642978, + "grad_norm": 8.75, + "learning_rate": 2.1623756250996736e-06, + "loss": 2.5063, + "step": 26327 + }, + { + "epoch": 2.784263959390863, + "grad_norm": 8.375, + "learning_rate": 2.1622059385784413e-06, + "loss": 2.5591, + "step": 26328 + }, + { + "epoch": 2.7843697123519457, + "grad_norm": 8.3125, + "learning_rate": 2.162036253642339e-06, + "loss": 2.8193, + "step": 26329 + }, + { + "epoch": 2.7844754653130286, + "grad_norm": 8.6875, + "learning_rate": 2.1618665702921616e-06, + "loss": 2.4973, + "step": 26330 + }, + { + "epoch": 2.784581218274112, + "grad_norm": 8.25, + "learning_rate": 2.1616968885287075e-06, + "loss": 2.2475, + "step": 26331 + }, + { + "epoch": 2.7846869712351947, + "grad_norm": 8.625, + "learning_rate": 2.1615272083527697e-06, + "loss": 2.7271, + "step": 26332 + }, + { + "epoch": 2.7847927241962775, + "grad_norm": 8.4375, + "learning_rate": 2.161357529765148e-06, + "loss": 2.648, + "step": 26333 + }, + { + "epoch": 2.7848984771573604, + "grad_norm": 8.3125, + "learning_rate": 2.1611878527666357e-06, + "loss": 2.634, + "step": 26334 + }, + { + "epoch": 2.7850042301184432, + "grad_norm": 8.8125, + "learning_rate": 2.1610181773580306e-06, + "loss": 2.6153, + "step": 26335 + }, + { + "epoch": 2.785109983079526, + "grad_norm": 8.8125, + "learning_rate": 2.1608485035401283e-06, + "loss": 2.8413, + "step": 26336 + }, + { + "epoch": 2.785215736040609, + "grad_norm": 8.9375, + "learning_rate": 2.1606788313137252e-06, + "loss": 2.399, + "step": 26337 + }, + { + "epoch": 2.785321489001692, + "grad_norm": 8.9375, + "learning_rate": 2.1605091606796182e-06, + "loss": 2.588, + "step": 26338 + }, + { + "epoch": 2.785427241962775, + "grad_norm": 8.5, + "learning_rate": 2.1603394916386024e-06, + "loss": 2.5701, + "step": 26339 + }, + { + "epoch": 2.785532994923858, + "grad_norm": 8.75, + "learning_rate": 2.1601698241914755e-06, + "loss": 2.7004, + "step": 26340 + }, + { + "epoch": 2.7856387478849407, + "grad_norm": 8.125, + "learning_rate": 2.1600001583390313e-06, + "loss": 2.6092, + "step": 26341 + }, + { + "epoch": 2.7857445008460235, + "grad_norm": 8.625, + "learning_rate": 2.159830494082068e-06, + "loss": 2.7124, + "step": 26342 + }, + { + "epoch": 2.785850253807107, + "grad_norm": 8.4375, + "learning_rate": 2.1596608314213804e-06, + "loss": 2.5553, + "step": 26343 + }, + { + "epoch": 2.7859560067681897, + "grad_norm": 8.3125, + "learning_rate": 2.1594911703577657e-06, + "loss": 2.6692, + "step": 26344 + }, + { + "epoch": 2.7860617597292725, + "grad_norm": 9.0, + "learning_rate": 2.1593215108920194e-06, + "loss": 2.6797, + "step": 26345 + }, + { + "epoch": 2.7861675126903553, + "grad_norm": 8.1875, + "learning_rate": 2.1591518530249385e-06, + "loss": 2.5111, + "step": 26346 + }, + { + "epoch": 2.786273265651438, + "grad_norm": 8.0625, + "learning_rate": 2.158982196757319e-06, + "loss": 2.6186, + "step": 26347 + }, + { + "epoch": 2.786379018612521, + "grad_norm": 8.875, + "learning_rate": 2.158812542089955e-06, + "loss": 2.7465, + "step": 26348 + }, + { + "epoch": 2.786484771573604, + "grad_norm": 8.3125, + "learning_rate": 2.1586428890236454e-06, + "loss": 2.427, + "step": 26349 + }, + { + "epoch": 2.786590524534687, + "grad_norm": 8.5, + "learning_rate": 2.1584732375591844e-06, + "loss": 2.5801, + "step": 26350 + }, + { + "epoch": 2.78669627749577, + "grad_norm": 8.375, + "learning_rate": 2.1583035876973696e-06, + "loss": 2.4617, + "step": 26351 + }, + { + "epoch": 2.786802030456853, + "grad_norm": 8.125, + "learning_rate": 2.1581339394389957e-06, + "loss": 2.5168, + "step": 26352 + }, + { + "epoch": 2.7869077834179357, + "grad_norm": 8.3125, + "learning_rate": 2.1579642927848594e-06, + "loss": 2.5395, + "step": 26353 + }, + { + "epoch": 2.7870135363790185, + "grad_norm": 7.8125, + "learning_rate": 2.157794647735757e-06, + "loss": 2.4523, + "step": 26354 + }, + { + "epoch": 2.787119289340102, + "grad_norm": 8.625, + "learning_rate": 2.157625004292484e-06, + "loss": 2.618, + "step": 26355 + }, + { + "epoch": 2.787225042301184, + "grad_norm": 8.1875, + "learning_rate": 2.157455362455838e-06, + "loss": 2.507, + "step": 26356 + }, + { + "epoch": 2.7873307952622675, + "grad_norm": 8.875, + "learning_rate": 2.157285722226613e-06, + "loss": 2.7223, + "step": 26357 + }, + { + "epoch": 2.7874365482233503, + "grad_norm": 8.3125, + "learning_rate": 2.157116083605606e-06, + "loss": 2.5263, + "step": 26358 + }, + { + "epoch": 2.787542301184433, + "grad_norm": 8.6875, + "learning_rate": 2.1569464465936127e-06, + "loss": 2.3149, + "step": 26359 + }, + { + "epoch": 2.787648054145516, + "grad_norm": 8.75, + "learning_rate": 2.15677681119143e-06, + "loss": 2.6284, + "step": 26360 + }, + { + "epoch": 2.787753807106599, + "grad_norm": 8.625, + "learning_rate": 2.1566071773998524e-06, + "loss": 2.4746, + "step": 26361 + }, + { + "epoch": 2.787859560067682, + "grad_norm": 8.4375, + "learning_rate": 2.1564375452196775e-06, + "loss": 2.6735, + "step": 26362 + }, + { + "epoch": 2.787965313028765, + "grad_norm": 8.5, + "learning_rate": 2.1562679146517015e-06, + "loss": 2.702, + "step": 26363 + }, + { + "epoch": 2.7880710659898478, + "grad_norm": 8.375, + "learning_rate": 2.156098285696718e-06, + "loss": 2.6402, + "step": 26364 + }, + { + "epoch": 2.7881768189509306, + "grad_norm": 8.6875, + "learning_rate": 2.155928658355526e-06, + "loss": 2.5245, + "step": 26365 + }, + { + "epoch": 2.7882825719120135, + "grad_norm": 8.375, + "learning_rate": 2.1557590326289192e-06, + "loss": 2.478, + "step": 26366 + }, + { + "epoch": 2.7883883248730963, + "grad_norm": 7.9375, + "learning_rate": 2.155589408517695e-06, + "loss": 2.4729, + "step": 26367 + }, + { + "epoch": 2.788494077834179, + "grad_norm": 7.875, + "learning_rate": 2.1554197860226485e-06, + "loss": 2.4298, + "step": 26368 + }, + { + "epoch": 2.7885998307952624, + "grad_norm": 8.3125, + "learning_rate": 2.1552501651445766e-06, + "loss": 2.4119, + "step": 26369 + }, + { + "epoch": 2.7887055837563453, + "grad_norm": 8.4375, + "learning_rate": 2.1550805458842736e-06, + "loss": 2.4814, + "step": 26370 + }, + { + "epoch": 2.788811336717428, + "grad_norm": 8.6875, + "learning_rate": 2.1549109282425366e-06, + "loss": 2.4313, + "step": 26371 + }, + { + "epoch": 2.788917089678511, + "grad_norm": 8.875, + "learning_rate": 2.154741312220163e-06, + "loss": 2.4452, + "step": 26372 + }, + { + "epoch": 2.7890228426395938, + "grad_norm": 8.0, + "learning_rate": 2.154571697817946e-06, + "loss": 2.4982, + "step": 26373 + }, + { + "epoch": 2.789128595600677, + "grad_norm": 8.8125, + "learning_rate": 2.154402085036683e-06, + "loss": 2.5381, + "step": 26374 + }, + { + "epoch": 2.78923434856176, + "grad_norm": 8.125, + "learning_rate": 2.1542324738771695e-06, + "loss": 2.5408, + "step": 26375 + }, + { + "epoch": 2.7893401015228427, + "grad_norm": 8.9375, + "learning_rate": 2.154062864340202e-06, + "loss": 2.4157, + "step": 26376 + }, + { + "epoch": 2.7894458544839256, + "grad_norm": 8.125, + "learning_rate": 2.1538932564265755e-06, + "loss": 2.4262, + "step": 26377 + }, + { + "epoch": 2.7895516074450084, + "grad_norm": 8.1875, + "learning_rate": 2.1537236501370875e-06, + "loss": 2.4991, + "step": 26378 + }, + { + "epoch": 2.7896573604060912, + "grad_norm": 9.6875, + "learning_rate": 2.1535540454725314e-06, + "loss": 2.4431, + "step": 26379 + }, + { + "epoch": 2.789763113367174, + "grad_norm": 8.75, + "learning_rate": 2.1533844424337043e-06, + "loss": 2.585, + "step": 26380 + }, + { + "epoch": 2.7898688663282574, + "grad_norm": 8.0625, + "learning_rate": 2.1532148410214036e-06, + "loss": 2.5255, + "step": 26381 + }, + { + "epoch": 2.78997461928934, + "grad_norm": 9.25, + "learning_rate": 2.1530452412364226e-06, + "loss": 2.471, + "step": 26382 + }, + { + "epoch": 2.790080372250423, + "grad_norm": 8.5625, + "learning_rate": 2.1528756430795593e-06, + "loss": 2.5985, + "step": 26383 + }, + { + "epoch": 2.790186125211506, + "grad_norm": 8.25, + "learning_rate": 2.152706046551608e-06, + "loss": 2.564, + "step": 26384 + }, + { + "epoch": 2.7902918781725887, + "grad_norm": 8.5, + "learning_rate": 2.1525364516533664e-06, + "loss": 2.6982, + "step": 26385 + }, + { + "epoch": 2.790397631133672, + "grad_norm": 8.5625, + "learning_rate": 2.152366858385627e-06, + "loss": 2.5827, + "step": 26386 + }, + { + "epoch": 2.7905033840947544, + "grad_norm": 8.5625, + "learning_rate": 2.15219726674919e-06, + "loss": 2.5487, + "step": 26387 + }, + { + "epoch": 2.7906091370558377, + "grad_norm": 8.125, + "learning_rate": 2.1520276767448477e-06, + "loss": 2.4678, + "step": 26388 + }, + { + "epoch": 2.7907148900169205, + "grad_norm": 8.8125, + "learning_rate": 2.151858088373397e-06, + "loss": 2.4903, + "step": 26389 + }, + { + "epoch": 2.7908206429780034, + "grad_norm": 8.5, + "learning_rate": 2.1516885016356345e-06, + "loss": 2.6694, + "step": 26390 + }, + { + "epoch": 2.790926395939086, + "grad_norm": 8.875, + "learning_rate": 2.1515189165323548e-06, + "loss": 2.6727, + "step": 26391 + }, + { + "epoch": 2.791032148900169, + "grad_norm": 8.0, + "learning_rate": 2.151349333064355e-06, + "loss": 2.4317, + "step": 26392 + }, + { + "epoch": 2.7911379018612523, + "grad_norm": 8.875, + "learning_rate": 2.15117975123243e-06, + "loss": 2.6601, + "step": 26393 + }, + { + "epoch": 2.791243654822335, + "grad_norm": 8.25, + "learning_rate": 2.151010171037376e-06, + "loss": 2.3573, + "step": 26394 + }, + { + "epoch": 2.791349407783418, + "grad_norm": 8.0625, + "learning_rate": 2.150840592479988e-06, + "loss": 2.216, + "step": 26395 + }, + { + "epoch": 2.791455160744501, + "grad_norm": 8.75, + "learning_rate": 2.1506710155610628e-06, + "loss": 2.7283, + "step": 26396 + }, + { + "epoch": 2.7915609137055837, + "grad_norm": 9.4375, + "learning_rate": 2.150501440281395e-06, + "loss": 2.5713, + "step": 26397 + }, + { + "epoch": 2.7916666666666665, + "grad_norm": 9.0625, + "learning_rate": 2.150331866641781e-06, + "loss": 2.5995, + "step": 26398 + }, + { + "epoch": 2.7917724196277494, + "grad_norm": 8.6875, + "learning_rate": 2.1501622946430174e-06, + "loss": 2.4492, + "step": 26399 + }, + { + "epoch": 2.7918781725888326, + "grad_norm": 8.8125, + "learning_rate": 2.1499927242858982e-06, + "loss": 2.5031, + "step": 26400 + }, + { + "epoch": 2.7919839255499155, + "grad_norm": 8.875, + "learning_rate": 2.149823155571221e-06, + "loss": 2.6553, + "step": 26401 + }, + { + "epoch": 2.7920896785109983, + "grad_norm": 8.9375, + "learning_rate": 2.1496535884997792e-06, + "loss": 2.5103, + "step": 26402 + }, + { + "epoch": 2.792195431472081, + "grad_norm": 8.5625, + "learning_rate": 2.1494840230723714e-06, + "loss": 2.7074, + "step": 26403 + }, + { + "epoch": 2.792301184433164, + "grad_norm": 8.625, + "learning_rate": 2.1493144592897907e-06, + "loss": 2.5678, + "step": 26404 + }, + { + "epoch": 2.7924069373942473, + "grad_norm": 8.5625, + "learning_rate": 2.149144897152834e-06, + "loss": 2.391, + "step": 26405 + }, + { + "epoch": 2.7925126903553297, + "grad_norm": 8.125, + "learning_rate": 2.1489753366622963e-06, + "loss": 2.3704, + "step": 26406 + }, + { + "epoch": 2.792618443316413, + "grad_norm": 8.25, + "learning_rate": 2.1488057778189743e-06, + "loss": 2.5955, + "step": 26407 + }, + { + "epoch": 2.792724196277496, + "grad_norm": 8.3125, + "learning_rate": 2.148636220623663e-06, + "loss": 2.3888, + "step": 26408 + }, + { + "epoch": 2.7928299492385786, + "grad_norm": 8.375, + "learning_rate": 2.148466665077158e-06, + "loss": 2.3959, + "step": 26409 + }, + { + "epoch": 2.7929357021996615, + "grad_norm": 8.3125, + "learning_rate": 2.1482971111802563e-06, + "loss": 2.6217, + "step": 26410 + }, + { + "epoch": 2.7930414551607443, + "grad_norm": 8.4375, + "learning_rate": 2.1481275589337514e-06, + "loss": 2.7053, + "step": 26411 + }, + { + "epoch": 2.7931472081218276, + "grad_norm": 8.25, + "learning_rate": 2.1479580083384402e-06, + "loss": 2.4601, + "step": 26412 + }, + { + "epoch": 2.7932529610829104, + "grad_norm": 8.8125, + "learning_rate": 2.1477884593951177e-06, + "loss": 2.4844, + "step": 26413 + }, + { + "epoch": 2.7933587140439933, + "grad_norm": 8.6875, + "learning_rate": 2.1476189121045806e-06, + "loss": 2.7274, + "step": 26414 + }, + { + "epoch": 2.793464467005076, + "grad_norm": 8.625, + "learning_rate": 2.1474493664676233e-06, + "loss": 2.4894, + "step": 26415 + }, + { + "epoch": 2.793570219966159, + "grad_norm": 7.75, + "learning_rate": 2.1472798224850416e-06, + "loss": 2.4628, + "step": 26416 + }, + { + "epoch": 2.7936759729272422, + "grad_norm": 8.375, + "learning_rate": 2.1471102801576326e-06, + "loss": 2.5451, + "step": 26417 + }, + { + "epoch": 2.7937817258883246, + "grad_norm": 9.5625, + "learning_rate": 2.1469407394861895e-06, + "loss": 2.6011, + "step": 26418 + }, + { + "epoch": 2.793887478849408, + "grad_norm": 8.625, + "learning_rate": 2.1467712004715103e-06, + "loss": 2.8988, + "step": 26419 + }, + { + "epoch": 2.7939932318104908, + "grad_norm": 8.1875, + "learning_rate": 2.1466016631143885e-06, + "loss": 2.2142, + "step": 26420 + }, + { + "epoch": 2.7940989847715736, + "grad_norm": 8.3125, + "learning_rate": 2.146432127415621e-06, + "loss": 2.5285, + "step": 26421 + }, + { + "epoch": 2.7942047377326564, + "grad_norm": 8.4375, + "learning_rate": 2.1462625933760024e-06, + "loss": 2.4478, + "step": 26422 + }, + { + "epoch": 2.7943104906937393, + "grad_norm": 8.0, + "learning_rate": 2.14609306099633e-06, + "loss": 2.53, + "step": 26423 + }, + { + "epoch": 2.7944162436548226, + "grad_norm": 8.625, + "learning_rate": 2.1459235302773966e-06, + "loss": 2.5143, + "step": 26424 + }, + { + "epoch": 2.7945219966159054, + "grad_norm": 8.0625, + "learning_rate": 2.1457540012199996e-06, + "loss": 2.5084, + "step": 26425 + }, + { + "epoch": 2.7946277495769882, + "grad_norm": 7.8125, + "learning_rate": 2.1455844738249353e-06, + "loss": 2.4845, + "step": 26426 + }, + { + "epoch": 2.794733502538071, + "grad_norm": 8.3125, + "learning_rate": 2.145414948092997e-06, + "loss": 2.3734, + "step": 26427 + }, + { + "epoch": 2.794839255499154, + "grad_norm": 8.8125, + "learning_rate": 2.145245424024982e-06, + "loss": 2.7699, + "step": 26428 + }, + { + "epoch": 2.7949450084602367, + "grad_norm": 9.1875, + "learning_rate": 2.1450759016216842e-06, + "loss": 2.4644, + "step": 26429 + }, + { + "epoch": 2.7950507614213196, + "grad_norm": 8.0625, + "learning_rate": 2.1449063808839007e-06, + "loss": 2.5381, + "step": 26430 + }, + { + "epoch": 2.795156514382403, + "grad_norm": 7.90625, + "learning_rate": 2.144736861812426e-06, + "loss": 2.5475, + "step": 26431 + }, + { + "epoch": 2.7952622673434857, + "grad_norm": 8.875, + "learning_rate": 2.144567344408057e-06, + "loss": 2.3857, + "step": 26432 + }, + { + "epoch": 2.7953680203045685, + "grad_norm": 8.125, + "learning_rate": 2.144397828671587e-06, + "loss": 2.445, + "step": 26433 + }, + { + "epoch": 2.7954737732656514, + "grad_norm": 8.6875, + "learning_rate": 2.144228314603812e-06, + "loss": 2.3197, + "step": 26434 + }, + { + "epoch": 2.7955795262267342, + "grad_norm": 8.25, + "learning_rate": 2.1440588022055294e-06, + "loss": 2.3496, + "step": 26435 + }, + { + "epoch": 2.7956852791878175, + "grad_norm": 9.5, + "learning_rate": 2.1438892914775323e-06, + "loss": 2.9023, + "step": 26436 + }, + { + "epoch": 2.7957910321489, + "grad_norm": 8.5625, + "learning_rate": 2.143719782420618e-06, + "loss": 2.5083, + "step": 26437 + }, + { + "epoch": 2.795896785109983, + "grad_norm": 8.25, + "learning_rate": 2.14355027503558e-06, + "loss": 2.7006, + "step": 26438 + }, + { + "epoch": 2.796002538071066, + "grad_norm": 8.4375, + "learning_rate": 2.143380769323216e-06, + "loss": 2.548, + "step": 26439 + }, + { + "epoch": 2.796108291032149, + "grad_norm": 8.5625, + "learning_rate": 2.1432112652843186e-06, + "loss": 2.5489, + "step": 26440 + }, + { + "epoch": 2.7962140439932317, + "grad_norm": 8.3125, + "learning_rate": 2.1430417629196865e-06, + "loss": 2.3497, + "step": 26441 + }, + { + "epoch": 2.7963197969543145, + "grad_norm": 8.4375, + "learning_rate": 2.1428722622301124e-06, + "loss": 2.464, + "step": 26442 + }, + { + "epoch": 2.796425549915398, + "grad_norm": 8.25, + "learning_rate": 2.1427027632163926e-06, + "loss": 2.6332, + "step": 26443 + }, + { + "epoch": 2.7965313028764807, + "grad_norm": 8.25, + "learning_rate": 2.1425332658793232e-06, + "loss": 2.4757, + "step": 26444 + }, + { + "epoch": 2.7966370558375635, + "grad_norm": 9.0, + "learning_rate": 2.1423637702196984e-06, + "loss": 2.7142, + "step": 26445 + }, + { + "epoch": 2.7967428087986463, + "grad_norm": 8.375, + "learning_rate": 2.142194276238315e-06, + "loss": 2.4387, + "step": 26446 + }, + { + "epoch": 2.796848561759729, + "grad_norm": 8.0625, + "learning_rate": 2.1420247839359665e-06, + "loss": 2.2739, + "step": 26447 + }, + { + "epoch": 2.796954314720812, + "grad_norm": 8.875, + "learning_rate": 2.1418552933134506e-06, + "loss": 2.3645, + "step": 26448 + }, + { + "epoch": 2.797060067681895, + "grad_norm": 8.9375, + "learning_rate": 2.1416858043715602e-06, + "loss": 2.5333, + "step": 26449 + }, + { + "epoch": 2.797165820642978, + "grad_norm": 8.5, + "learning_rate": 2.141516317111092e-06, + "loss": 2.5173, + "step": 26450 + }, + { + "epoch": 2.797271573604061, + "grad_norm": 8.125, + "learning_rate": 2.141346831532841e-06, + "loss": 2.5033, + "step": 26451 + }, + { + "epoch": 2.797377326565144, + "grad_norm": 8.125, + "learning_rate": 2.141177347637603e-06, + "loss": 2.5258, + "step": 26452 + }, + { + "epoch": 2.7974830795262267, + "grad_norm": 8.1875, + "learning_rate": 2.1410078654261727e-06, + "loss": 2.5216, + "step": 26453 + }, + { + "epoch": 2.7975888324873095, + "grad_norm": 8.8125, + "learning_rate": 2.1408383848993455e-06, + "loss": 2.5142, + "step": 26454 + }, + { + "epoch": 2.797694585448393, + "grad_norm": 8.5625, + "learning_rate": 2.1406689060579177e-06, + "loss": 2.6037, + "step": 26455 + }, + { + "epoch": 2.7978003384094756, + "grad_norm": 8.1875, + "learning_rate": 2.1404994289026827e-06, + "loss": 2.2319, + "step": 26456 + }, + { + "epoch": 2.7979060913705585, + "grad_norm": 8.3125, + "learning_rate": 2.140329953434438e-06, + "loss": 2.4913, + "step": 26457 + }, + { + "epoch": 2.7980118443316413, + "grad_norm": 8.8125, + "learning_rate": 2.140160479653977e-06, + "loss": 2.5035, + "step": 26458 + }, + { + "epoch": 2.798117597292724, + "grad_norm": 8.8125, + "learning_rate": 2.1399910075620957e-06, + "loss": 2.6435, + "step": 26459 + }, + { + "epoch": 2.798223350253807, + "grad_norm": 8.125, + "learning_rate": 2.1398215371595894e-06, + "loss": 2.6217, + "step": 26460 + }, + { + "epoch": 2.79832910321489, + "grad_norm": 9.25, + "learning_rate": 2.1396520684472527e-06, + "loss": 2.697, + "step": 26461 + }, + { + "epoch": 2.798434856175973, + "grad_norm": 8.375, + "learning_rate": 2.1394826014258826e-06, + "loss": 2.3406, + "step": 26462 + }, + { + "epoch": 2.798540609137056, + "grad_norm": 8.4375, + "learning_rate": 2.1393131360962723e-06, + "loss": 2.6422, + "step": 26463 + }, + { + "epoch": 2.7986463620981388, + "grad_norm": 10.5625, + "learning_rate": 2.139143672459219e-06, + "loss": 2.5173, + "step": 26464 + }, + { + "epoch": 2.7987521150592216, + "grad_norm": 8.5, + "learning_rate": 2.1389742105155157e-06, + "loss": 2.4974, + "step": 26465 + }, + { + "epoch": 2.7988578680203045, + "grad_norm": 8.375, + "learning_rate": 2.1388047502659597e-06, + "loss": 2.5711, + "step": 26466 + }, + { + "epoch": 2.7989636209813877, + "grad_norm": 9.1875, + "learning_rate": 2.138635291711345e-06, + "loss": 2.7763, + "step": 26467 + }, + { + "epoch": 2.79906937394247, + "grad_norm": 8.375, + "learning_rate": 2.1384658348524665e-06, + "loss": 2.6345, + "step": 26468 + }, + { + "epoch": 2.7991751269035534, + "grad_norm": 8.1875, + "learning_rate": 2.1382963796901205e-06, + "loss": 2.4267, + "step": 26469 + }, + { + "epoch": 2.7992808798646363, + "grad_norm": 8.75, + "learning_rate": 2.1381269262251015e-06, + "loss": 2.5235, + "step": 26470 + }, + { + "epoch": 2.799386632825719, + "grad_norm": 8.0625, + "learning_rate": 2.1379574744582057e-06, + "loss": 2.5472, + "step": 26471 + }, + { + "epoch": 2.799492385786802, + "grad_norm": 8.8125, + "learning_rate": 2.1377880243902254e-06, + "loss": 2.7078, + "step": 26472 + }, + { + "epoch": 2.7995981387478848, + "grad_norm": 8.6875, + "learning_rate": 2.13761857602196e-06, + "loss": 2.6175, + "step": 26473 + }, + { + "epoch": 2.799703891708968, + "grad_norm": 8.375, + "learning_rate": 2.137449129354201e-06, + "loss": 2.5895, + "step": 26474 + }, + { + "epoch": 2.799809644670051, + "grad_norm": 8.1875, + "learning_rate": 2.1372796843877457e-06, + "loss": 2.5384, + "step": 26475 + }, + { + "epoch": 2.7999153976311337, + "grad_norm": 8.3125, + "learning_rate": 2.1371102411233877e-06, + "loss": 2.38, + "step": 26476 + }, + { + "epoch": 2.8000211505922166, + "grad_norm": 8.9375, + "learning_rate": 2.136940799561923e-06, + "loss": 2.6452, + "step": 26477 + }, + { + "epoch": 2.8001269035532994, + "grad_norm": 8.125, + "learning_rate": 2.1367713597041474e-06, + "loss": 2.5112, + "step": 26478 + }, + { + "epoch": 2.8002326565143822, + "grad_norm": 8.1875, + "learning_rate": 2.1366019215508544e-06, + "loss": 2.4522, + "step": 26479 + }, + { + "epoch": 2.800338409475465, + "grad_norm": 8.375, + "learning_rate": 2.1364324851028416e-06, + "loss": 2.6251, + "step": 26480 + }, + { + "epoch": 2.8004441624365484, + "grad_norm": 8.0625, + "learning_rate": 2.1362630503609006e-06, + "loss": 2.5002, + "step": 26481 + }, + { + "epoch": 2.800549915397631, + "grad_norm": 8.0625, + "learning_rate": 2.1360936173258296e-06, + "loss": 2.6899, + "step": 26482 + }, + { + "epoch": 2.800655668358714, + "grad_norm": 8.125, + "learning_rate": 2.1359241859984216e-06, + "loss": 2.6182, + "step": 26483 + }, + { + "epoch": 2.800761421319797, + "grad_norm": 8.8125, + "learning_rate": 2.135754756379473e-06, + "loss": 2.6608, + "step": 26484 + }, + { + "epoch": 2.8008671742808797, + "grad_norm": 8.375, + "learning_rate": 2.1355853284697777e-06, + "loss": 2.5776, + "step": 26485 + }, + { + "epoch": 2.800972927241963, + "grad_norm": 8.125, + "learning_rate": 2.1354159022701313e-06, + "loss": 2.5955, + "step": 26486 + }, + { + "epoch": 2.801078680203046, + "grad_norm": 9.0, + "learning_rate": 2.13524647778133e-06, + "loss": 2.4922, + "step": 26487 + }, + { + "epoch": 2.8011844331641287, + "grad_norm": 8.6875, + "learning_rate": 2.1350770550041665e-06, + "loss": 2.4305, + "step": 26488 + }, + { + "epoch": 2.8012901861252115, + "grad_norm": 8.5, + "learning_rate": 2.134907633939439e-06, + "loss": 2.7441, + "step": 26489 + }, + { + "epoch": 2.8013959390862944, + "grad_norm": 8.625, + "learning_rate": 2.134738214587939e-06, + "loss": 2.7198, + "step": 26490 + }, + { + "epoch": 2.801501692047377, + "grad_norm": 8.625, + "learning_rate": 2.1345687969504643e-06, + "loss": 2.6013, + "step": 26491 + }, + { + "epoch": 2.80160744500846, + "grad_norm": 8.375, + "learning_rate": 2.1343993810278077e-06, + "loss": 2.7417, + "step": 26492 + }, + { + "epoch": 2.8017131979695433, + "grad_norm": 7.96875, + "learning_rate": 2.1342299668207668e-06, + "loss": 2.4377, + "step": 26493 + }, + { + "epoch": 2.801818950930626, + "grad_norm": 11.0625, + "learning_rate": 2.134060554330133e-06, + "loss": 2.6719, + "step": 26494 + }, + { + "epoch": 2.801924703891709, + "grad_norm": 8.4375, + "learning_rate": 2.133891143556704e-06, + "loss": 2.42, + "step": 26495 + }, + { + "epoch": 2.802030456852792, + "grad_norm": 8.125, + "learning_rate": 2.1337217345012752e-06, + "loss": 2.531, + "step": 26496 + }, + { + "epoch": 2.8021362098138747, + "grad_norm": 8.875, + "learning_rate": 2.1335523271646397e-06, + "loss": 2.3791, + "step": 26497 + }, + { + "epoch": 2.802241962774958, + "grad_norm": 8.375, + "learning_rate": 2.1333829215475936e-06, + "loss": 2.4362, + "step": 26498 + }, + { + "epoch": 2.8023477157360404, + "grad_norm": 8.0625, + "learning_rate": 2.1332135176509312e-06, + "loss": 2.5743, + "step": 26499 + }, + { + "epoch": 2.8024534686971236, + "grad_norm": 8.375, + "learning_rate": 2.133044115475448e-06, + "loss": 2.4891, + "step": 26500 + }, + { + "epoch": 2.8025592216582065, + "grad_norm": 8.3125, + "learning_rate": 2.132874715021938e-06, + "loss": 2.6273, + "step": 26501 + }, + { + "epoch": 2.8026649746192893, + "grad_norm": 9.0, + "learning_rate": 2.1327053162911982e-06, + "loss": 2.3239, + "step": 26502 + }, + { + "epoch": 2.802770727580372, + "grad_norm": 8.4375, + "learning_rate": 2.132535919284021e-06, + "loss": 2.4507, + "step": 26503 + }, + { + "epoch": 2.802876480541455, + "grad_norm": 8.125, + "learning_rate": 2.132366524001202e-06, + "loss": 2.5766, + "step": 26504 + }, + { + "epoch": 2.8029822335025383, + "grad_norm": 8.375, + "learning_rate": 2.132197130443538e-06, + "loss": 2.4647, + "step": 26505 + }, + { + "epoch": 2.803087986463621, + "grad_norm": 8.625, + "learning_rate": 2.1320277386118218e-06, + "loss": 2.5271, + "step": 26506 + }, + { + "epoch": 2.803193739424704, + "grad_norm": 8.9375, + "learning_rate": 2.131858348506849e-06, + "loss": 2.7373, + "step": 26507 + }, + { + "epoch": 2.803299492385787, + "grad_norm": 8.0625, + "learning_rate": 2.131688960129414e-06, + "loss": 2.4865, + "step": 26508 + }, + { + "epoch": 2.8034052453468696, + "grad_norm": 8.25, + "learning_rate": 2.131519573480313e-06, + "loss": 2.4774, + "step": 26509 + }, + { + "epoch": 2.8035109983079525, + "grad_norm": 8.3125, + "learning_rate": 2.1313501885603383e-06, + "loss": 2.4895, + "step": 26510 + }, + { + "epoch": 2.8036167512690353, + "grad_norm": 8.3125, + "learning_rate": 2.1311808053702882e-06, + "loss": 2.5665, + "step": 26511 + }, + { + "epoch": 2.8037225042301186, + "grad_norm": 8.3125, + "learning_rate": 2.1310114239109543e-06, + "loss": 2.3339, + "step": 26512 + }, + { + "epoch": 2.8038282571912014, + "grad_norm": 8.5625, + "learning_rate": 2.1308420441831336e-06, + "loss": 2.5408, + "step": 26513 + }, + { + "epoch": 2.8039340101522843, + "grad_norm": 7.90625, + "learning_rate": 2.1306726661876203e-06, + "loss": 2.5388, + "step": 26514 + }, + { + "epoch": 2.804039763113367, + "grad_norm": 8.5625, + "learning_rate": 2.1305032899252083e-06, + "loss": 2.6961, + "step": 26515 + }, + { + "epoch": 2.80414551607445, + "grad_norm": 8.625, + "learning_rate": 2.130333915396694e-06, + "loss": 2.4063, + "step": 26516 + }, + { + "epoch": 2.8042512690355332, + "grad_norm": 8.3125, + "learning_rate": 2.130164542602871e-06, + "loss": 2.6347, + "step": 26517 + }, + { + "epoch": 2.8043570219966156, + "grad_norm": 7.96875, + "learning_rate": 2.1299951715445358e-06, + "loss": 2.4732, + "step": 26518 + }, + { + "epoch": 2.804462774957699, + "grad_norm": 8.5, + "learning_rate": 2.1298258022224805e-06, + "loss": 2.5074, + "step": 26519 + }, + { + "epoch": 2.8045685279187818, + "grad_norm": 8.625, + "learning_rate": 2.1296564346375023e-06, + "loss": 2.5558, + "step": 26520 + }, + { + "epoch": 2.8046742808798646, + "grad_norm": 9.1875, + "learning_rate": 2.129487068790394e-06, + "loss": 2.9053, + "step": 26521 + }, + { + "epoch": 2.8047800338409474, + "grad_norm": 8.125, + "learning_rate": 2.1293177046819517e-06, + "loss": 2.5139, + "step": 26522 + }, + { + "epoch": 2.8048857868020303, + "grad_norm": 8.0625, + "learning_rate": 2.12914834231297e-06, + "loss": 2.5291, + "step": 26523 + }, + { + "epoch": 2.8049915397631136, + "grad_norm": 9.0, + "learning_rate": 2.128978981684243e-06, + "loss": 2.1893, + "step": 26524 + }, + { + "epoch": 2.8050972927241964, + "grad_norm": 8.3125, + "learning_rate": 2.128809622796567e-06, + "loss": 2.545, + "step": 26525 + }, + { + "epoch": 2.8052030456852792, + "grad_norm": 8.5625, + "learning_rate": 2.1286402656507344e-06, + "loss": 2.3975, + "step": 26526 + }, + { + "epoch": 2.805308798646362, + "grad_norm": 9.1875, + "learning_rate": 2.128470910247542e-06, + "loss": 2.6112, + "step": 26527 + }, + { + "epoch": 2.805414551607445, + "grad_norm": 8.75, + "learning_rate": 2.128301556587783e-06, + "loss": 2.4343, + "step": 26528 + }, + { + "epoch": 2.805520304568528, + "grad_norm": 8.5625, + "learning_rate": 2.1281322046722536e-06, + "loss": 2.5076, + "step": 26529 + }, + { + "epoch": 2.8056260575296106, + "grad_norm": 8.6875, + "learning_rate": 2.127962854501747e-06, + "loss": 2.6646, + "step": 26530 + }, + { + "epoch": 2.805731810490694, + "grad_norm": 8.5625, + "learning_rate": 2.127793506077058e-06, + "loss": 2.3986, + "step": 26531 + }, + { + "epoch": 2.8058375634517767, + "grad_norm": 8.0625, + "learning_rate": 2.1276241593989833e-06, + "loss": 2.608, + "step": 26532 + }, + { + "epoch": 2.8059433164128595, + "grad_norm": 8.5625, + "learning_rate": 2.127454814468315e-06, + "loss": 2.4315, + "step": 26533 + }, + { + "epoch": 2.8060490693739424, + "grad_norm": 8.75, + "learning_rate": 2.12728547128585e-06, + "loss": 2.5483, + "step": 26534 + }, + { + "epoch": 2.8061548223350252, + "grad_norm": 17.5, + "learning_rate": 2.127116129852381e-06, + "loss": 2.6303, + "step": 26535 + }, + { + "epoch": 2.8062605752961085, + "grad_norm": 8.1875, + "learning_rate": 2.126946790168704e-06, + "loss": 2.5997, + "step": 26536 + }, + { + "epoch": 2.8063663282571913, + "grad_norm": 8.1875, + "learning_rate": 2.126777452235612e-06, + "loss": 2.5332, + "step": 26537 + }, + { + "epoch": 2.806472081218274, + "grad_norm": 8.4375, + "learning_rate": 2.126608116053902e-06, + "loss": 2.5167, + "step": 26538 + }, + { + "epoch": 2.806577834179357, + "grad_norm": 8.9375, + "learning_rate": 2.1264387816243667e-06, + "loss": 2.5692, + "step": 26539 + }, + { + "epoch": 2.80668358714044, + "grad_norm": 8.4375, + "learning_rate": 2.1262694489478015e-06, + "loss": 2.5592, + "step": 26540 + }, + { + "epoch": 2.8067893401015227, + "grad_norm": 9.0625, + "learning_rate": 2.126100118025002e-06, + "loss": 2.8546, + "step": 26541 + }, + { + "epoch": 2.8068950930626055, + "grad_norm": 8.0, + "learning_rate": 2.12593078885676e-06, + "loss": 2.6676, + "step": 26542 + }, + { + "epoch": 2.807000846023689, + "grad_norm": 8.3125, + "learning_rate": 2.125761461443873e-06, + "loss": 2.4128, + "step": 26543 + }, + { + "epoch": 2.8071065989847717, + "grad_norm": 8.1875, + "learning_rate": 2.125592135787134e-06, + "loss": 2.2992, + "step": 26544 + }, + { + "epoch": 2.8072123519458545, + "grad_norm": 8.6875, + "learning_rate": 2.1254228118873384e-06, + "loss": 2.7723, + "step": 26545 + }, + { + "epoch": 2.8073181049069373, + "grad_norm": 8.375, + "learning_rate": 2.1252534897452797e-06, + "loss": 2.4953, + "step": 26546 + }, + { + "epoch": 2.80742385786802, + "grad_norm": 8.5625, + "learning_rate": 2.125084169361754e-06, + "loss": 2.525, + "step": 26547 + }, + { + "epoch": 2.8075296108291035, + "grad_norm": 8.4375, + "learning_rate": 2.1249148507375536e-06, + "loss": 2.5846, + "step": 26548 + }, + { + "epoch": 2.807635363790186, + "grad_norm": 8.375, + "learning_rate": 2.124745533873475e-06, + "loss": 2.629, + "step": 26549 + }, + { + "epoch": 2.807741116751269, + "grad_norm": 8.4375, + "learning_rate": 2.124576218770313e-06, + "loss": 2.7022, + "step": 26550 + }, + { + "epoch": 2.807846869712352, + "grad_norm": 8.6875, + "learning_rate": 2.1244069054288603e-06, + "loss": 2.6454, + "step": 26551 + }, + { + "epoch": 2.807952622673435, + "grad_norm": 8.3125, + "learning_rate": 2.1242375938499127e-06, + "loss": 2.3, + "step": 26552 + }, + { + "epoch": 2.8080583756345177, + "grad_norm": 8.8125, + "learning_rate": 2.124068284034264e-06, + "loss": 2.8174, + "step": 26553 + }, + { + "epoch": 2.8081641285956005, + "grad_norm": 8.4375, + "learning_rate": 2.1238989759827097e-06, + "loss": 2.3084, + "step": 26554 + }, + { + "epoch": 2.808269881556684, + "grad_norm": 9.4375, + "learning_rate": 2.123729669696043e-06, + "loss": 2.8214, + "step": 26555 + }, + { + "epoch": 2.8083756345177666, + "grad_norm": 8.5625, + "learning_rate": 2.12356036517506e-06, + "loss": 2.4352, + "step": 26556 + }, + { + "epoch": 2.8084813874788495, + "grad_norm": 8.4375, + "learning_rate": 2.123391062420554e-06, + "loss": 2.7443, + "step": 26557 + }, + { + "epoch": 2.8085871404399323, + "grad_norm": 7.78125, + "learning_rate": 2.1232217614333182e-06, + "loss": 2.4721, + "step": 26558 + }, + { + "epoch": 2.808692893401015, + "grad_norm": 8.625, + "learning_rate": 2.123052462214151e-06, + "loss": 2.7328, + "step": 26559 + }, + { + "epoch": 2.808798646362098, + "grad_norm": 8.9375, + "learning_rate": 2.1228831647638424e-06, + "loss": 2.5158, + "step": 26560 + }, + { + "epoch": 2.808904399323181, + "grad_norm": 8.5, + "learning_rate": 2.1227138690831903e-06, + "loss": 2.5035, + "step": 26561 + }, + { + "epoch": 2.809010152284264, + "grad_norm": 8.25, + "learning_rate": 2.1225445751729866e-06, + "loss": 2.661, + "step": 26562 + }, + { + "epoch": 2.809115905245347, + "grad_norm": 8.4375, + "learning_rate": 2.1223752830340283e-06, + "loss": 2.5504, + "step": 26563 + }, + { + "epoch": 2.8092216582064298, + "grad_norm": 8.0625, + "learning_rate": 2.1222059926671065e-06, + "loss": 2.5001, + "step": 26564 + }, + { + "epoch": 2.8093274111675126, + "grad_norm": 8.375, + "learning_rate": 2.1220367040730192e-06, + "loss": 2.5618, + "step": 26565 + }, + { + "epoch": 2.8094331641285955, + "grad_norm": 8.8125, + "learning_rate": 2.121867417252558e-06, + "loss": 2.4006, + "step": 26566 + }, + { + "epoch": 2.8095389170896787, + "grad_norm": 8.25, + "learning_rate": 2.1216981322065184e-06, + "loss": 2.6149, + "step": 26567 + }, + { + "epoch": 2.8096446700507616, + "grad_norm": 8.6875, + "learning_rate": 2.121528848935695e-06, + "loss": 2.6366, + "step": 26568 + }, + { + "epoch": 2.8097504230118444, + "grad_norm": 8.875, + "learning_rate": 2.1213595674408816e-06, + "loss": 2.4439, + "step": 26569 + }, + { + "epoch": 2.8098561759729273, + "grad_norm": 8.375, + "learning_rate": 2.1211902877228733e-06, + "loss": 2.5713, + "step": 26570 + }, + { + "epoch": 2.80996192893401, + "grad_norm": 8.25, + "learning_rate": 2.1210210097824636e-06, + "loss": 2.5184, + "step": 26571 + }, + { + "epoch": 2.810067681895093, + "grad_norm": 8.375, + "learning_rate": 2.1208517336204483e-06, + "loss": 2.6123, + "step": 26572 + }, + { + "epoch": 2.8101734348561758, + "grad_norm": 8.5, + "learning_rate": 2.1206824592376197e-06, + "loss": 2.524, + "step": 26573 + }, + { + "epoch": 2.810279187817259, + "grad_norm": 8.75, + "learning_rate": 2.1205131866347737e-06, + "loss": 2.488, + "step": 26574 + }, + { + "epoch": 2.810384940778342, + "grad_norm": 8.625, + "learning_rate": 2.120343915812704e-06, + "loss": 2.8877, + "step": 26575 + }, + { + "epoch": 2.8104906937394247, + "grad_norm": 8.5, + "learning_rate": 2.1201746467722046e-06, + "loss": 2.6816, + "step": 26576 + }, + { + "epoch": 2.8105964467005076, + "grad_norm": 8.5, + "learning_rate": 2.120005379514071e-06, + "loss": 2.5587, + "step": 26577 + }, + { + "epoch": 2.8107021996615904, + "grad_norm": 8.375, + "learning_rate": 2.119836114039096e-06, + "loss": 2.4722, + "step": 26578 + }, + { + "epoch": 2.8108079526226737, + "grad_norm": 8.1875, + "learning_rate": 2.119666850348076e-06, + "loss": 2.4753, + "step": 26579 + }, + { + "epoch": 2.810913705583756, + "grad_norm": 8.5, + "learning_rate": 2.1194975884418017e-06, + "loss": 2.6412, + "step": 26580 + }, + { + "epoch": 2.8110194585448394, + "grad_norm": 8.6875, + "learning_rate": 2.1193283283210717e-06, + "loss": 2.645, + "step": 26581 + }, + { + "epoch": 2.811125211505922, + "grad_norm": 8.3125, + "learning_rate": 2.119159069986677e-06, + "loss": 2.372, + "step": 26582 + }, + { + "epoch": 2.811230964467005, + "grad_norm": 8.625, + "learning_rate": 2.118989813439413e-06, + "loss": 2.6015, + "step": 26583 + }, + { + "epoch": 2.811336717428088, + "grad_norm": 8.75, + "learning_rate": 2.1188205586800745e-06, + "loss": 2.4026, + "step": 26584 + }, + { + "epoch": 2.8114424703891707, + "grad_norm": 8.6875, + "learning_rate": 2.1186513057094545e-06, + "loss": 2.4046, + "step": 26585 + }, + { + "epoch": 2.811548223350254, + "grad_norm": 8.4375, + "learning_rate": 2.1184820545283487e-06, + "loss": 2.4479, + "step": 26586 + }, + { + "epoch": 2.811653976311337, + "grad_norm": 8.4375, + "learning_rate": 2.11831280513755e-06, + "loss": 2.4907, + "step": 26587 + }, + { + "epoch": 2.8117597292724197, + "grad_norm": 8.4375, + "learning_rate": 2.1181435575378545e-06, + "loss": 2.4567, + "step": 26588 + }, + { + "epoch": 2.8118654822335025, + "grad_norm": 8.25, + "learning_rate": 2.117974311730054e-06, + "loss": 2.4236, + "step": 26589 + }, + { + "epoch": 2.8119712351945854, + "grad_norm": 8.375, + "learning_rate": 2.1178050677149442e-06, + "loss": 2.7291, + "step": 26590 + }, + { + "epoch": 2.812076988155668, + "grad_norm": 8.4375, + "learning_rate": 2.1176358254933184e-06, + "loss": 2.6791, + "step": 26591 + }, + { + "epoch": 2.812182741116751, + "grad_norm": 8.3125, + "learning_rate": 2.117466585065971e-06, + "loss": 2.3236, + "step": 26592 + }, + { + "epoch": 2.8122884940778343, + "grad_norm": 8.3125, + "learning_rate": 2.1172973464336973e-06, + "loss": 2.5821, + "step": 26593 + }, + { + "epoch": 2.812394247038917, + "grad_norm": 8.6875, + "learning_rate": 2.11712810959729e-06, + "loss": 2.6223, + "step": 26594 + }, + { + "epoch": 2.8125, + "grad_norm": 8.8125, + "learning_rate": 2.116958874557545e-06, + "loss": 2.6424, + "step": 26595 + }, + { + "epoch": 2.812605752961083, + "grad_norm": 8.375, + "learning_rate": 2.116789641315254e-06, + "loss": 2.5451, + "step": 26596 + }, + { + "epoch": 2.8127115059221657, + "grad_norm": 8.5, + "learning_rate": 2.1166204098712136e-06, + "loss": 2.712, + "step": 26597 + }, + { + "epoch": 2.812817258883249, + "grad_norm": 8.375, + "learning_rate": 2.116451180226216e-06, + "loss": 2.5153, + "step": 26598 + }, + { + "epoch": 2.812923011844332, + "grad_norm": 9.0625, + "learning_rate": 2.116281952381057e-06, + "loss": 2.4495, + "step": 26599 + }, + { + "epoch": 2.8130287648054146, + "grad_norm": 8.75, + "learning_rate": 2.116112726336529e-06, + "loss": 2.7158, + "step": 26600 + }, + { + "epoch": 2.8131345177664975, + "grad_norm": 8.0625, + "learning_rate": 2.115943502093427e-06, + "loss": 2.382, + "step": 26601 + }, + { + "epoch": 2.8132402707275803, + "grad_norm": 9.1875, + "learning_rate": 2.1157742796525454e-06, + "loss": 2.5418, + "step": 26602 + }, + { + "epoch": 2.813346023688663, + "grad_norm": 8.5625, + "learning_rate": 2.1156050590146778e-06, + "loss": 2.5166, + "step": 26603 + }, + { + "epoch": 2.813451776649746, + "grad_norm": 8.9375, + "learning_rate": 2.1154358401806193e-06, + "loss": 2.4956, + "step": 26604 + }, + { + "epoch": 2.8135575296108293, + "grad_norm": 11.1875, + "learning_rate": 2.115266623151162e-06, + "loss": 2.611, + "step": 26605 + }, + { + "epoch": 2.813663282571912, + "grad_norm": 8.0625, + "learning_rate": 2.1150974079271016e-06, + "loss": 2.2702, + "step": 26606 + }, + { + "epoch": 2.813769035532995, + "grad_norm": 8.25, + "learning_rate": 2.1149281945092314e-06, + "loss": 2.4585, + "step": 26607 + }, + { + "epoch": 2.813874788494078, + "grad_norm": 9.1875, + "learning_rate": 2.114758982898346e-06, + "loss": 2.6124, + "step": 26608 + }, + { + "epoch": 2.8139805414551606, + "grad_norm": 8.6875, + "learning_rate": 2.114589773095239e-06, + "loss": 2.7294, + "step": 26609 + }, + { + "epoch": 2.814086294416244, + "grad_norm": 8.875, + "learning_rate": 2.1144205651007037e-06, + "loss": 2.5667, + "step": 26610 + }, + { + "epoch": 2.8141920473773263, + "grad_norm": 9.0, + "learning_rate": 2.1142513589155365e-06, + "loss": 2.5499, + "step": 26611 + }, + { + "epoch": 2.8142978003384096, + "grad_norm": 8.1875, + "learning_rate": 2.114082154540529e-06, + "loss": 2.5801, + "step": 26612 + }, + { + "epoch": 2.8144035532994924, + "grad_norm": 9.4375, + "learning_rate": 2.113912951976477e-06, + "loss": 2.5306, + "step": 26613 + }, + { + "epoch": 2.8145093062605753, + "grad_norm": 8.3125, + "learning_rate": 2.113743751224172e-06, + "loss": 2.4265, + "step": 26614 + }, + { + "epoch": 2.814615059221658, + "grad_norm": 8.4375, + "learning_rate": 2.1135745522844108e-06, + "loss": 2.7007, + "step": 26615 + }, + { + "epoch": 2.814720812182741, + "grad_norm": 8.625, + "learning_rate": 2.1134053551579856e-06, + "loss": 2.5903, + "step": 26616 + }, + { + "epoch": 2.8148265651438242, + "grad_norm": 8.6875, + "learning_rate": 2.1132361598456924e-06, + "loss": 2.6729, + "step": 26617 + }, + { + "epoch": 2.814932318104907, + "grad_norm": 8.875, + "learning_rate": 2.1130669663483215e-06, + "loss": 2.5506, + "step": 26618 + }, + { + "epoch": 2.81503807106599, + "grad_norm": 8.4375, + "learning_rate": 2.1128977746666703e-06, + "loss": 2.5716, + "step": 26619 + }, + { + "epoch": 2.8151438240270727, + "grad_norm": 8.5, + "learning_rate": 2.112728584801532e-06, + "loss": 2.53, + "step": 26620 + }, + { + "epoch": 2.8152495769881556, + "grad_norm": 8.3125, + "learning_rate": 2.112559396753699e-06, + "loss": 2.6601, + "step": 26621 + }, + { + "epoch": 2.8153553299492384, + "grad_norm": 8.1875, + "learning_rate": 2.1123902105239667e-06, + "loss": 2.6221, + "step": 26622 + }, + { + "epoch": 2.8154610829103213, + "grad_norm": 8.625, + "learning_rate": 2.1122210261131286e-06, + "loss": 2.5429, + "step": 26623 + }, + { + "epoch": 2.8155668358714045, + "grad_norm": 8.875, + "learning_rate": 2.112051843521979e-06, + "loss": 2.5965, + "step": 26624 + }, + { + "epoch": 2.8156725888324874, + "grad_norm": 8.9375, + "learning_rate": 2.111882662751311e-06, + "loss": 2.5627, + "step": 26625 + }, + { + "epoch": 2.8157783417935702, + "grad_norm": 9.3125, + "learning_rate": 2.1117134838019195e-06, + "loss": 2.5157, + "step": 26626 + }, + { + "epoch": 2.815884094754653, + "grad_norm": 8.5625, + "learning_rate": 2.1115443066745974e-06, + "loss": 2.5006, + "step": 26627 + }, + { + "epoch": 2.815989847715736, + "grad_norm": 8.8125, + "learning_rate": 2.1113751313701385e-06, + "loss": 2.5462, + "step": 26628 + }, + { + "epoch": 2.816095600676819, + "grad_norm": 8.3125, + "learning_rate": 2.1112059578893377e-06, + "loss": 2.7143, + "step": 26629 + }, + { + "epoch": 2.816201353637902, + "grad_norm": 8.875, + "learning_rate": 2.111036786232988e-06, + "loss": 2.6922, + "step": 26630 + }, + { + "epoch": 2.816307106598985, + "grad_norm": 8.6875, + "learning_rate": 2.1108676164018843e-06, + "loss": 2.5497, + "step": 26631 + }, + { + "epoch": 2.8164128595600677, + "grad_norm": 8.4375, + "learning_rate": 2.110698448396819e-06, + "loss": 2.5595, + "step": 26632 + }, + { + "epoch": 2.8165186125211505, + "grad_norm": 8.8125, + "learning_rate": 2.1105292822185875e-06, + "loss": 2.5777, + "step": 26633 + }, + { + "epoch": 2.8166243654822334, + "grad_norm": 8.6875, + "learning_rate": 2.1103601178679816e-06, + "loss": 2.5701, + "step": 26634 + }, + { + "epoch": 2.816730118443316, + "grad_norm": 8.25, + "learning_rate": 2.110190955345798e-06, + "loss": 2.4842, + "step": 26635 + }, + { + "epoch": 2.8168358714043995, + "grad_norm": 8.6875, + "learning_rate": 2.110021794652827e-06, + "loss": 2.6994, + "step": 26636 + }, + { + "epoch": 2.8169416243654823, + "grad_norm": 8.75, + "learning_rate": 2.1098526357898647e-06, + "loss": 2.7636, + "step": 26637 + }, + { + "epoch": 2.817047377326565, + "grad_norm": 8.6875, + "learning_rate": 2.109683478757705e-06, + "loss": 2.5819, + "step": 26638 + }, + { + "epoch": 2.817153130287648, + "grad_norm": 8.4375, + "learning_rate": 2.1095143235571404e-06, + "loss": 2.4843, + "step": 26639 + }, + { + "epoch": 2.817258883248731, + "grad_norm": 8.5, + "learning_rate": 2.1093451701889663e-06, + "loss": 2.5885, + "step": 26640 + }, + { + "epoch": 2.817364636209814, + "grad_norm": 8.6875, + "learning_rate": 2.1091760186539743e-06, + "loss": 2.5453, + "step": 26641 + }, + { + "epoch": 2.8174703891708965, + "grad_norm": 8.8125, + "learning_rate": 2.109006868952961e-06, + "loss": 2.2979, + "step": 26642 + }, + { + "epoch": 2.81757614213198, + "grad_norm": 8.0625, + "learning_rate": 2.108837721086717e-06, + "loss": 2.2906, + "step": 26643 + }, + { + "epoch": 2.8176818950930627, + "grad_norm": 8.4375, + "learning_rate": 2.108668575056039e-06, + "loss": 2.4176, + "step": 26644 + }, + { + "epoch": 2.8177876480541455, + "grad_norm": 8.375, + "learning_rate": 2.108499430861718e-06, + "loss": 2.59, + "step": 26645 + }, + { + "epoch": 2.8178934010152283, + "grad_norm": 8.5, + "learning_rate": 2.1083302885045495e-06, + "loss": 2.5677, + "step": 26646 + }, + { + "epoch": 2.817999153976311, + "grad_norm": 8.3125, + "learning_rate": 2.1081611479853274e-06, + "loss": 2.6942, + "step": 26647 + }, + { + "epoch": 2.8181049069373945, + "grad_norm": 8.5, + "learning_rate": 2.107992009304844e-06, + "loss": 2.4563, + "step": 26648 + }, + { + "epoch": 2.8182106598984773, + "grad_norm": 8.875, + "learning_rate": 2.1078228724638945e-06, + "loss": 2.6805, + "step": 26649 + }, + { + "epoch": 2.81831641285956, + "grad_norm": 8.5, + "learning_rate": 2.1076537374632706e-06, + "loss": 2.754, + "step": 26650 + }, + { + "epoch": 2.818422165820643, + "grad_norm": 8.625, + "learning_rate": 2.107484604303769e-06, + "loss": 2.7787, + "step": 26651 + }, + { + "epoch": 2.818527918781726, + "grad_norm": 8.1875, + "learning_rate": 2.1073154729861805e-06, + "loss": 2.622, + "step": 26652 + }, + { + "epoch": 2.8186336717428087, + "grad_norm": 8.125, + "learning_rate": 2.1071463435113e-06, + "loss": 2.4389, + "step": 26653 + }, + { + "epoch": 2.8187394247038915, + "grad_norm": 8.3125, + "learning_rate": 2.106977215879921e-06, + "loss": 2.5079, + "step": 26654 + }, + { + "epoch": 2.8188451776649748, + "grad_norm": 8.5, + "learning_rate": 2.106808090092837e-06, + "loss": 2.5425, + "step": 26655 + }, + { + "epoch": 2.8189509306260576, + "grad_norm": 8.5, + "learning_rate": 2.1066389661508426e-06, + "loss": 1.9665, + "step": 26656 + }, + { + "epoch": 2.8190566835871405, + "grad_norm": 8.5, + "learning_rate": 2.1064698440547297e-06, + "loss": 2.7022, + "step": 26657 + }, + { + "epoch": 2.8191624365482233, + "grad_norm": 8.5, + "learning_rate": 2.106300723805294e-06, + "loss": 2.5093, + "step": 26658 + }, + { + "epoch": 2.819268189509306, + "grad_norm": 8.1875, + "learning_rate": 2.1061316054033273e-06, + "loss": 2.3314, + "step": 26659 + }, + { + "epoch": 2.8193739424703894, + "grad_norm": 8.375, + "learning_rate": 2.1059624888496237e-06, + "loss": 2.447, + "step": 26660 + }, + { + "epoch": 2.819479695431472, + "grad_norm": 8.25, + "learning_rate": 2.105793374144977e-06, + "loss": 2.5323, + "step": 26661 + }, + { + "epoch": 2.819585448392555, + "grad_norm": 8.4375, + "learning_rate": 2.1056242612901812e-06, + "loss": 2.5082, + "step": 26662 + }, + { + "epoch": 2.819691201353638, + "grad_norm": 8.25, + "learning_rate": 2.105455150286029e-06, + "loss": 2.4678, + "step": 26663 + }, + { + "epoch": 2.8197969543147208, + "grad_norm": 8.0625, + "learning_rate": 2.105286041133314e-06, + "loss": 2.5042, + "step": 26664 + }, + { + "epoch": 2.8199027072758036, + "grad_norm": 8.5, + "learning_rate": 2.105116933832832e-06, + "loss": 2.666, + "step": 26665 + }, + { + "epoch": 2.8200084602368864, + "grad_norm": 8.625, + "learning_rate": 2.104947828385373e-06, + "loss": 2.6155, + "step": 26666 + }, + { + "epoch": 2.8201142131979697, + "grad_norm": 8.5625, + "learning_rate": 2.104778724791733e-06, + "loss": 2.584, + "step": 26667 + }, + { + "epoch": 2.8202199661590526, + "grad_norm": 8.8125, + "learning_rate": 2.1046096230527045e-06, + "loss": 2.6933, + "step": 26668 + }, + { + "epoch": 2.8203257191201354, + "grad_norm": 10.1875, + "learning_rate": 2.1044405231690816e-06, + "loss": 2.4805, + "step": 26669 + }, + { + "epoch": 2.8204314720812182, + "grad_norm": 8.5, + "learning_rate": 2.1042714251416573e-06, + "loss": 2.7354, + "step": 26670 + }, + { + "epoch": 2.820537225042301, + "grad_norm": 8.9375, + "learning_rate": 2.104102328971226e-06, + "loss": 2.7555, + "step": 26671 + }, + { + "epoch": 2.8206429780033844, + "grad_norm": 8.375, + "learning_rate": 2.1039332346585795e-06, + "loss": 2.4976, + "step": 26672 + }, + { + "epoch": 2.8207487309644668, + "grad_norm": 8.5625, + "learning_rate": 2.1037641422045125e-06, + "loss": 2.526, + "step": 26673 + }, + { + "epoch": 2.82085448392555, + "grad_norm": 8.5625, + "learning_rate": 2.1035950516098193e-06, + "loss": 2.3341, + "step": 26674 + }, + { + "epoch": 2.820960236886633, + "grad_norm": 8.0625, + "learning_rate": 2.1034259628752917e-06, + "loss": 2.6125, + "step": 26675 + }, + { + "epoch": 2.8210659898477157, + "grad_norm": 8.6875, + "learning_rate": 2.1032568760017242e-06, + "loss": 2.7364, + "step": 26676 + }, + { + "epoch": 2.8211717428087986, + "grad_norm": 8.375, + "learning_rate": 2.103087790989909e-06, + "loss": 2.4711, + "step": 26677 + }, + { + "epoch": 2.8212774957698814, + "grad_norm": 8.75, + "learning_rate": 2.1029187078406417e-06, + "loss": 2.4704, + "step": 26678 + }, + { + "epoch": 2.8213832487309647, + "grad_norm": 8.375, + "learning_rate": 2.1027496265547134e-06, + "loss": 2.5914, + "step": 26679 + }, + { + "epoch": 2.8214890016920475, + "grad_norm": 8.25, + "learning_rate": 2.10258054713292e-06, + "loss": 2.1453, + "step": 26680 + }, + { + "epoch": 2.8215947546531304, + "grad_norm": 8.375, + "learning_rate": 2.1024114695760527e-06, + "loss": 2.1467, + "step": 26681 + }, + { + "epoch": 2.821700507614213, + "grad_norm": 8.0625, + "learning_rate": 2.102242393884905e-06, + "loss": 2.5795, + "step": 26682 + }, + { + "epoch": 2.821806260575296, + "grad_norm": 8.8125, + "learning_rate": 2.1020733200602723e-06, + "loss": 2.6021, + "step": 26683 + }, + { + "epoch": 2.821912013536379, + "grad_norm": 8.875, + "learning_rate": 2.101904248102946e-06, + "loss": 2.6669, + "step": 26684 + }, + { + "epoch": 2.8220177664974617, + "grad_norm": 8.625, + "learning_rate": 2.1017351780137207e-06, + "loss": 2.4485, + "step": 26685 + }, + { + "epoch": 2.822123519458545, + "grad_norm": 8.8125, + "learning_rate": 2.101566109793389e-06, + "loss": 2.7332, + "step": 26686 + }, + { + "epoch": 2.822229272419628, + "grad_norm": 8.25, + "learning_rate": 2.1013970434427457e-06, + "loss": 2.5007, + "step": 26687 + }, + { + "epoch": 2.8223350253807107, + "grad_norm": 9.3125, + "learning_rate": 2.1012279789625813e-06, + "loss": 2.6609, + "step": 26688 + }, + { + "epoch": 2.8224407783417935, + "grad_norm": 8.875, + "learning_rate": 2.1010589163536915e-06, + "loss": 2.2671, + "step": 26689 + }, + { + "epoch": 2.8225465313028764, + "grad_norm": 8.375, + "learning_rate": 2.1008898556168698e-06, + "loss": 2.6099, + "step": 26690 + }, + { + "epoch": 2.8226522842639596, + "grad_norm": 8.0, + "learning_rate": 2.1007207967529082e-06, + "loss": 2.5753, + "step": 26691 + }, + { + "epoch": 2.822758037225042, + "grad_norm": 8.4375, + "learning_rate": 2.100551739762601e-06, + "loss": 2.5107, + "step": 26692 + }, + { + "epoch": 2.8228637901861253, + "grad_norm": 8.0625, + "learning_rate": 2.1003826846467403e-06, + "loss": 2.6592, + "step": 26693 + }, + { + "epoch": 2.822969543147208, + "grad_norm": 9.5625, + "learning_rate": 2.1002136314061203e-06, + "loss": 2.4973, + "step": 26694 + }, + { + "epoch": 2.823075296108291, + "grad_norm": 8.5, + "learning_rate": 2.100044580041535e-06, + "loss": 2.4774, + "step": 26695 + }, + { + "epoch": 2.823181049069374, + "grad_norm": 8.375, + "learning_rate": 2.099875530553777e-06, + "loss": 2.4673, + "step": 26696 + }, + { + "epoch": 2.8232868020304567, + "grad_norm": 9.1875, + "learning_rate": 2.0997064829436387e-06, + "loss": 2.883, + "step": 26697 + }, + { + "epoch": 2.82339255499154, + "grad_norm": 9.0625, + "learning_rate": 2.0995374372119142e-06, + "loss": 2.7081, + "step": 26698 + }, + { + "epoch": 2.823498307952623, + "grad_norm": 8.25, + "learning_rate": 2.099368393359397e-06, + "loss": 2.43, + "step": 26699 + }, + { + "epoch": 2.8236040609137056, + "grad_norm": 9.5, + "learning_rate": 2.09919935138688e-06, + "loss": 2.606, + "step": 26700 + }, + { + "epoch": 2.8237098138747885, + "grad_norm": 8.0, + "learning_rate": 2.0990303112951568e-06, + "loss": 2.5631, + "step": 26701 + }, + { + "epoch": 2.8238155668358713, + "grad_norm": 9.625, + "learning_rate": 2.09886127308502e-06, + "loss": 2.4535, + "step": 26702 + }, + { + "epoch": 2.823921319796954, + "grad_norm": 9.3125, + "learning_rate": 2.098692236757264e-06, + "loss": 2.8574, + "step": 26703 + }, + { + "epoch": 2.824027072758037, + "grad_norm": 9.0, + "learning_rate": 2.0985232023126797e-06, + "loss": 2.5028, + "step": 26704 + }, + { + "epoch": 2.8241328257191203, + "grad_norm": 8.6875, + "learning_rate": 2.0983541697520634e-06, + "loss": 2.5883, + "step": 26705 + }, + { + "epoch": 2.824238578680203, + "grad_norm": 8.0625, + "learning_rate": 2.0981851390762055e-06, + "loss": 2.5546, + "step": 26706 + }, + { + "epoch": 2.824344331641286, + "grad_norm": 8.25, + "learning_rate": 2.098016110285901e-06, + "loss": 2.5501, + "step": 26707 + }, + { + "epoch": 2.824450084602369, + "grad_norm": 8.375, + "learning_rate": 2.0978470833819423e-06, + "loss": 2.519, + "step": 26708 + }, + { + "epoch": 2.8245558375634516, + "grad_norm": 9.3125, + "learning_rate": 2.0976780583651225e-06, + "loss": 2.6428, + "step": 26709 + }, + { + "epoch": 2.824661590524535, + "grad_norm": 8.4375, + "learning_rate": 2.0975090352362355e-06, + "loss": 2.5391, + "step": 26710 + }, + { + "epoch": 2.8247673434856178, + "grad_norm": 8.8125, + "learning_rate": 2.0973400139960735e-06, + "loss": 2.4994, + "step": 26711 + }, + { + "epoch": 2.8248730964467006, + "grad_norm": 8.9375, + "learning_rate": 2.0971709946454315e-06, + "loss": 2.6347, + "step": 26712 + }, + { + "epoch": 2.8249788494077834, + "grad_norm": 8.25, + "learning_rate": 2.0970019771851e-06, + "loss": 2.5012, + "step": 26713 + }, + { + "epoch": 2.8250846023688663, + "grad_norm": 8.25, + "learning_rate": 2.096832961615874e-06, + "loss": 2.6467, + "step": 26714 + }, + { + "epoch": 2.825190355329949, + "grad_norm": 8.6875, + "learning_rate": 2.096663947938545e-06, + "loss": 2.5265, + "step": 26715 + }, + { + "epoch": 2.825296108291032, + "grad_norm": 8.3125, + "learning_rate": 2.0964949361539076e-06, + "loss": 2.584, + "step": 26716 + }, + { + "epoch": 2.8254018612521152, + "grad_norm": 8.75, + "learning_rate": 2.096325926262755e-06, + "loss": 2.5335, + "step": 26717 + }, + { + "epoch": 2.825507614213198, + "grad_norm": 8.5625, + "learning_rate": 2.0961569182658787e-06, + "loss": 2.3494, + "step": 26718 + }, + { + "epoch": 2.825613367174281, + "grad_norm": 8.625, + "learning_rate": 2.0959879121640746e-06, + "loss": 2.6983, + "step": 26719 + }, + { + "epoch": 2.8257191201353637, + "grad_norm": 8.75, + "learning_rate": 2.0958189079581327e-06, + "loss": 2.7848, + "step": 26720 + }, + { + "epoch": 2.8258248730964466, + "grad_norm": 8.4375, + "learning_rate": 2.0956499056488475e-06, + "loss": 2.5934, + "step": 26721 + }, + { + "epoch": 2.82593062605753, + "grad_norm": 8.8125, + "learning_rate": 2.095480905237012e-06, + "loss": 2.5987, + "step": 26722 + }, + { + "epoch": 2.8260363790186123, + "grad_norm": 8.5, + "learning_rate": 2.095311906723419e-06, + "loss": 2.7402, + "step": 26723 + }, + { + "epoch": 2.8261421319796955, + "grad_norm": 8.5625, + "learning_rate": 2.0951429101088616e-06, + "loss": 2.3693, + "step": 26724 + }, + { + "epoch": 2.8262478849407784, + "grad_norm": 9.3125, + "learning_rate": 2.0949739153941327e-06, + "loss": 2.1284, + "step": 26725 + }, + { + "epoch": 2.8263536379018612, + "grad_norm": 8.125, + "learning_rate": 2.0948049225800267e-06, + "loss": 2.4974, + "step": 26726 + }, + { + "epoch": 2.826459390862944, + "grad_norm": 8.125, + "learning_rate": 2.0946359316673343e-06, + "loss": 2.4344, + "step": 26727 + }, + { + "epoch": 2.826565143824027, + "grad_norm": 8.5625, + "learning_rate": 2.094466942656851e-06, + "loss": 2.35, + "step": 26728 + }, + { + "epoch": 2.82667089678511, + "grad_norm": 8.5, + "learning_rate": 2.0942979555493673e-06, + "loss": 2.6166, + "step": 26729 + }, + { + "epoch": 2.826776649746193, + "grad_norm": 8.3125, + "learning_rate": 2.094128970345678e-06, + "loss": 2.6367, + "step": 26730 + }, + { + "epoch": 2.826882402707276, + "grad_norm": 8.625, + "learning_rate": 2.093959987046575e-06, + "loss": 2.5086, + "step": 26731 + }, + { + "epoch": 2.8269881556683587, + "grad_norm": 8.625, + "learning_rate": 2.093791005652852e-06, + "loss": 2.7225, + "step": 26732 + }, + { + "epoch": 2.8270939086294415, + "grad_norm": 8.375, + "learning_rate": 2.0936220261653013e-06, + "loss": 2.7077, + "step": 26733 + }, + { + "epoch": 2.8271996615905244, + "grad_norm": 8.6875, + "learning_rate": 2.0934530485847165e-06, + "loss": 2.6648, + "step": 26734 + }, + { + "epoch": 2.827305414551607, + "grad_norm": 8.3125, + "learning_rate": 2.093284072911891e-06, + "loss": 2.6249, + "step": 26735 + }, + { + "epoch": 2.8274111675126905, + "grad_norm": 8.375, + "learning_rate": 2.0931150991476156e-06, + "loss": 2.3204, + "step": 26736 + }, + { + "epoch": 2.8275169204737733, + "grad_norm": 8.0625, + "learning_rate": 2.0929461272926855e-06, + "loss": 2.3372, + "step": 26737 + }, + { + "epoch": 2.827622673434856, + "grad_norm": 9.5625, + "learning_rate": 2.0927771573478927e-06, + "loss": 2.6523, + "step": 26738 + }, + { + "epoch": 2.827728426395939, + "grad_norm": 8.625, + "learning_rate": 2.0926081893140303e-06, + "loss": 2.5204, + "step": 26739 + }, + { + "epoch": 2.827834179357022, + "grad_norm": 8.5625, + "learning_rate": 2.0924392231918904e-06, + "loss": 2.3543, + "step": 26740 + }, + { + "epoch": 2.827939932318105, + "grad_norm": 8.25, + "learning_rate": 2.0922702589822674e-06, + "loss": 2.616, + "step": 26741 + }, + { + "epoch": 2.828045685279188, + "grad_norm": 8.25, + "learning_rate": 2.092101296685952e-06, + "loss": 2.3544, + "step": 26742 + }, + { + "epoch": 2.828151438240271, + "grad_norm": 9.625, + "learning_rate": 2.091932336303739e-06, + "loss": 2.7913, + "step": 26743 + }, + { + "epoch": 2.8282571912013537, + "grad_norm": 8.4375, + "learning_rate": 2.0917633778364214e-06, + "loss": 2.658, + "step": 26744 + }, + { + "epoch": 2.8283629441624365, + "grad_norm": 8.125, + "learning_rate": 2.0915944212847907e-06, + "loss": 2.5275, + "step": 26745 + }, + { + "epoch": 2.8284686971235193, + "grad_norm": 9.5, + "learning_rate": 2.0914254666496407e-06, + "loss": 2.8112, + "step": 26746 + }, + { + "epoch": 2.828574450084602, + "grad_norm": 8.0625, + "learning_rate": 2.0912565139317632e-06, + "loss": 2.6451, + "step": 26747 + }, + { + "epoch": 2.8286802030456855, + "grad_norm": 8.5, + "learning_rate": 2.0910875631319526e-06, + "loss": 2.4687, + "step": 26748 + }, + { + "epoch": 2.8287859560067683, + "grad_norm": 8.375, + "learning_rate": 2.0909186142509997e-06, + "loss": 2.6204, + "step": 26749 + }, + { + "epoch": 2.828891708967851, + "grad_norm": 8.3125, + "learning_rate": 2.0907496672897e-06, + "loss": 2.5838, + "step": 26750 + }, + { + "epoch": 2.828997461928934, + "grad_norm": 8.375, + "learning_rate": 2.0905807222488436e-06, + "loss": 2.5926, + "step": 26751 + }, + { + "epoch": 2.829103214890017, + "grad_norm": 8.4375, + "learning_rate": 2.0904117791292246e-06, + "loss": 2.2796, + "step": 26752 + }, + { + "epoch": 2.8292089678511, + "grad_norm": 8.1875, + "learning_rate": 2.0902428379316356e-06, + "loss": 2.3812, + "step": 26753 + }, + { + "epoch": 2.8293147208121825, + "grad_norm": 8.25, + "learning_rate": 2.090073898656869e-06, + "loss": 2.4976, + "step": 26754 + }, + { + "epoch": 2.8294204737732658, + "grad_norm": 9.125, + "learning_rate": 2.0899049613057185e-06, + "loss": 2.8756, + "step": 26755 + }, + { + "epoch": 2.8295262267343486, + "grad_norm": 8.1875, + "learning_rate": 2.089736025878976e-06, + "loss": 2.5696, + "step": 26756 + }, + { + "epoch": 2.8296319796954315, + "grad_norm": 8.5625, + "learning_rate": 2.0895670923774357e-06, + "loss": 2.4421, + "step": 26757 + }, + { + "epoch": 2.8297377326565143, + "grad_norm": 8.25, + "learning_rate": 2.0893981608018877e-06, + "loss": 2.583, + "step": 26758 + }, + { + "epoch": 2.829843485617597, + "grad_norm": 8.4375, + "learning_rate": 2.0892292311531267e-06, + "loss": 2.5675, + "step": 26759 + }, + { + "epoch": 2.8299492385786804, + "grad_norm": 8.1875, + "learning_rate": 2.089060303431945e-06, + "loss": 2.5163, + "step": 26760 + }, + { + "epoch": 2.8300549915397633, + "grad_norm": 8.8125, + "learning_rate": 2.0888913776391346e-06, + "loss": 2.5246, + "step": 26761 + }, + { + "epoch": 2.830160744500846, + "grad_norm": 8.3125, + "learning_rate": 2.0887224537754896e-06, + "loss": 2.6598, + "step": 26762 + }, + { + "epoch": 2.830266497461929, + "grad_norm": 8.3125, + "learning_rate": 2.088553531841801e-06, + "loss": 2.2302, + "step": 26763 + }, + { + "epoch": 2.8303722504230118, + "grad_norm": 8.25, + "learning_rate": 2.0883846118388633e-06, + "loss": 2.5378, + "step": 26764 + }, + { + "epoch": 2.8304780033840946, + "grad_norm": 8.4375, + "learning_rate": 2.088215693767468e-06, + "loss": 2.5824, + "step": 26765 + }, + { + "epoch": 2.8305837563451774, + "grad_norm": 9.0625, + "learning_rate": 2.088046777628409e-06, + "loss": 2.813, + "step": 26766 + }, + { + "epoch": 2.8306895093062607, + "grad_norm": 8.9375, + "learning_rate": 2.0878778634224765e-06, + "loss": 2.6544, + "step": 26767 + }, + { + "epoch": 2.8307952622673436, + "grad_norm": 8.375, + "learning_rate": 2.0877089511504656e-06, + "loss": 2.7243, + "step": 26768 + }, + { + "epoch": 2.8309010152284264, + "grad_norm": 8.0625, + "learning_rate": 2.087540040813167e-06, + "loss": 2.6207, + "step": 26769 + }, + { + "epoch": 2.8310067681895092, + "grad_norm": 8.5, + "learning_rate": 2.0873711324113747e-06, + "loss": 2.4288, + "step": 26770 + }, + { + "epoch": 2.831112521150592, + "grad_norm": 8.625, + "learning_rate": 2.0872022259458813e-06, + "loss": 2.6143, + "step": 26771 + }, + { + "epoch": 2.8312182741116754, + "grad_norm": 8.125, + "learning_rate": 2.0870333214174786e-06, + "loss": 2.6481, + "step": 26772 + }, + { + "epoch": 2.8313240270727578, + "grad_norm": 8.3125, + "learning_rate": 2.0868644188269606e-06, + "loss": 2.6111, + "step": 26773 + }, + { + "epoch": 2.831429780033841, + "grad_norm": 8.9375, + "learning_rate": 2.0866955181751176e-06, + "loss": 2.5102, + "step": 26774 + }, + { + "epoch": 2.831535532994924, + "grad_norm": 8.875, + "learning_rate": 2.086526619462744e-06, + "loss": 2.4262, + "step": 26775 + }, + { + "epoch": 2.8316412859560067, + "grad_norm": 8.25, + "learning_rate": 2.086357722690632e-06, + "loss": 2.3484, + "step": 26776 + }, + { + "epoch": 2.8317470389170896, + "grad_norm": 7.875, + "learning_rate": 2.0861888278595744e-06, + "loss": 2.473, + "step": 26777 + }, + { + "epoch": 2.8318527918781724, + "grad_norm": 8.375, + "learning_rate": 2.0860199349703623e-06, + "loss": 2.4978, + "step": 26778 + }, + { + "epoch": 2.8319585448392557, + "grad_norm": 8.625, + "learning_rate": 2.08585104402379e-06, + "loss": 2.5195, + "step": 26779 + }, + { + "epoch": 2.8320642978003385, + "grad_norm": 8.9375, + "learning_rate": 2.0856821550206493e-06, + "loss": 2.6411, + "step": 26780 + }, + { + "epoch": 2.8321700507614214, + "grad_norm": 8.25, + "learning_rate": 2.0855132679617324e-06, + "loss": 2.6378, + "step": 26781 + }, + { + "epoch": 2.832275803722504, + "grad_norm": 8.875, + "learning_rate": 2.085344382847834e-06, + "loss": 2.6617, + "step": 26782 + }, + { + "epoch": 2.832381556683587, + "grad_norm": 9.0625, + "learning_rate": 2.085175499679743e-06, + "loss": 2.7783, + "step": 26783 + }, + { + "epoch": 2.8324873096446703, + "grad_norm": 8.5, + "learning_rate": 2.085006618458254e-06, + "loss": 2.4421, + "step": 26784 + }, + { + "epoch": 2.8325930626057527, + "grad_norm": 8.6875, + "learning_rate": 2.084837739184159e-06, + "loss": 2.6078, + "step": 26785 + }, + { + "epoch": 2.832698815566836, + "grad_norm": 8.625, + "learning_rate": 2.0846688618582516e-06, + "loss": 2.5424, + "step": 26786 + }, + { + "epoch": 2.832804568527919, + "grad_norm": 8.125, + "learning_rate": 2.0844999864813225e-06, + "loss": 2.5224, + "step": 26787 + }, + { + "epoch": 2.8329103214890017, + "grad_norm": 8.1875, + "learning_rate": 2.084331113054165e-06, + "loss": 2.4817, + "step": 26788 + }, + { + "epoch": 2.8330160744500845, + "grad_norm": 8.0625, + "learning_rate": 2.0841622415775727e-06, + "loss": 2.5201, + "step": 26789 + }, + { + "epoch": 2.8331218274111674, + "grad_norm": 8.8125, + "learning_rate": 2.083993372052336e-06, + "loss": 2.2205, + "step": 26790 + }, + { + "epoch": 2.8332275803722506, + "grad_norm": 8.8125, + "learning_rate": 2.083824504479249e-06, + "loss": 2.5662, + "step": 26791 + }, + { + "epoch": 2.8333333333333335, + "grad_norm": 8.125, + "learning_rate": 2.083655638859102e-06, + "loss": 2.4429, + "step": 26792 + }, + { + "epoch": 2.8334390862944163, + "grad_norm": 8.0, + "learning_rate": 2.0834867751926903e-06, + "loss": 2.3375, + "step": 26793 + }, + { + "epoch": 2.833544839255499, + "grad_norm": 8.375, + "learning_rate": 2.083317913480804e-06, + "loss": 2.4667, + "step": 26794 + }, + { + "epoch": 2.833650592216582, + "grad_norm": 8.875, + "learning_rate": 2.0831490537242373e-06, + "loss": 2.6631, + "step": 26795 + }, + { + "epoch": 2.833756345177665, + "grad_norm": 8.3125, + "learning_rate": 2.0829801959237797e-06, + "loss": 2.4409, + "step": 26796 + }, + { + "epoch": 2.8338620981387477, + "grad_norm": 8.3125, + "learning_rate": 2.082811340080226e-06, + "loss": 2.608, + "step": 26797 + }, + { + "epoch": 2.833967851099831, + "grad_norm": 8.1875, + "learning_rate": 2.08264248619437e-06, + "loss": 2.571, + "step": 26798 + }, + { + "epoch": 2.834073604060914, + "grad_norm": 7.9375, + "learning_rate": 2.0824736342670003e-06, + "loss": 2.5023, + "step": 26799 + }, + { + "epoch": 2.8341793570219966, + "grad_norm": 8.5625, + "learning_rate": 2.0823047842989115e-06, + "loss": 2.4966, + "step": 26800 + }, + { + "epoch": 2.8342851099830795, + "grad_norm": 8.25, + "learning_rate": 2.0821359362908955e-06, + "loss": 2.3953, + "step": 26801 + }, + { + "epoch": 2.8343908629441623, + "grad_norm": 8.625, + "learning_rate": 2.0819670902437447e-06, + "loss": 2.5151, + "step": 26802 + }, + { + "epoch": 2.8344966159052456, + "grad_norm": 8.9375, + "learning_rate": 2.081798246158251e-06, + "loss": 2.0061, + "step": 26803 + }, + { + "epoch": 2.834602368866328, + "grad_norm": 8.75, + "learning_rate": 2.0816294040352074e-06, + "loss": 2.6353, + "step": 26804 + }, + { + "epoch": 2.8347081218274113, + "grad_norm": 8.625, + "learning_rate": 2.0814605638754065e-06, + "loss": 2.4965, + "step": 26805 + }, + { + "epoch": 2.834813874788494, + "grad_norm": 8.375, + "learning_rate": 2.0812917256796396e-06, + "loss": 2.439, + "step": 26806 + }, + { + "epoch": 2.834919627749577, + "grad_norm": 8.25, + "learning_rate": 2.081122889448699e-06, + "loss": 2.5138, + "step": 26807 + }, + { + "epoch": 2.83502538071066, + "grad_norm": 8.0625, + "learning_rate": 2.0809540551833775e-06, + "loss": 2.5494, + "step": 26808 + }, + { + "epoch": 2.8351311336717426, + "grad_norm": 8.4375, + "learning_rate": 2.0807852228844676e-06, + "loss": 2.6383, + "step": 26809 + }, + { + "epoch": 2.835236886632826, + "grad_norm": 8.375, + "learning_rate": 2.0806163925527608e-06, + "loss": 2.6251, + "step": 26810 + }, + { + "epoch": 2.8353426395939088, + "grad_norm": 8.5, + "learning_rate": 2.0804475641890503e-06, + "loss": 2.6285, + "step": 26811 + }, + { + "epoch": 2.8354483925549916, + "grad_norm": 8.125, + "learning_rate": 2.0802787377941273e-06, + "loss": 2.239, + "step": 26812 + }, + { + "epoch": 2.8355541455160744, + "grad_norm": 8.375, + "learning_rate": 2.0801099133687834e-06, + "loss": 2.4908, + "step": 26813 + }, + { + "epoch": 2.8356598984771573, + "grad_norm": 8.6875, + "learning_rate": 2.0799410909138143e-06, + "loss": 2.8652, + "step": 26814 + }, + { + "epoch": 2.83576565143824, + "grad_norm": 8.75, + "learning_rate": 2.079772270430008e-06, + "loss": 2.7795, + "step": 26815 + }, + { + "epoch": 2.835871404399323, + "grad_norm": 8.3125, + "learning_rate": 2.07960345191816e-06, + "loss": 2.4648, + "step": 26816 + }, + { + "epoch": 2.8359771573604062, + "grad_norm": 8.6875, + "learning_rate": 2.07943463537906e-06, + "loss": 2.4443, + "step": 26817 + }, + { + "epoch": 2.836082910321489, + "grad_norm": 8.9375, + "learning_rate": 2.0792658208135017e-06, + "loss": 2.6409, + "step": 26818 + }, + { + "epoch": 2.836188663282572, + "grad_norm": 8.0625, + "learning_rate": 2.0790970082222766e-06, + "loss": 2.3461, + "step": 26819 + }, + { + "epoch": 2.8362944162436547, + "grad_norm": 8.6875, + "learning_rate": 2.0789281976061777e-06, + "loss": 2.6702, + "step": 26820 + }, + { + "epoch": 2.8364001692047376, + "grad_norm": 8.625, + "learning_rate": 2.078759388965996e-06, + "loss": 2.6533, + "step": 26821 + }, + { + "epoch": 2.836505922165821, + "grad_norm": 8.0625, + "learning_rate": 2.0785905823025236e-06, + "loss": 2.4931, + "step": 26822 + }, + { + "epoch": 2.8366116751269037, + "grad_norm": 8.4375, + "learning_rate": 2.0784217776165544e-06, + "loss": 2.4962, + "step": 26823 + }, + { + "epoch": 2.8367174280879865, + "grad_norm": 8.625, + "learning_rate": 2.0782529749088786e-06, + "loss": 2.6856, + "step": 26824 + }, + { + "epoch": 2.8368231810490694, + "grad_norm": 8.875, + "learning_rate": 2.0780841741802897e-06, + "loss": 2.674, + "step": 26825 + }, + { + "epoch": 2.8369289340101522, + "grad_norm": 9.5625, + "learning_rate": 2.0779153754315786e-06, + "loss": 2.6558, + "step": 26826 + }, + { + "epoch": 2.837034686971235, + "grad_norm": 8.1875, + "learning_rate": 2.077746578663539e-06, + "loss": 2.4499, + "step": 26827 + }, + { + "epoch": 2.837140439932318, + "grad_norm": 8.8125, + "learning_rate": 2.0775777838769614e-06, + "loss": 2.8825, + "step": 26828 + }, + { + "epoch": 2.837246192893401, + "grad_norm": 8.4375, + "learning_rate": 2.0774089910726385e-06, + "loss": 2.672, + "step": 26829 + }, + { + "epoch": 2.837351945854484, + "grad_norm": 9.0, + "learning_rate": 2.077240200251362e-06, + "loss": 2.7507, + "step": 26830 + }, + { + "epoch": 2.837457698815567, + "grad_norm": 9.0625, + "learning_rate": 2.0770714114139244e-06, + "loss": 2.7722, + "step": 26831 + }, + { + "epoch": 2.8375634517766497, + "grad_norm": 8.3125, + "learning_rate": 2.076902624561118e-06, + "loss": 2.4782, + "step": 26832 + }, + { + "epoch": 2.8376692047377325, + "grad_norm": 8.4375, + "learning_rate": 2.0767338396937344e-06, + "loss": 2.6676, + "step": 26833 + }, + { + "epoch": 2.837774957698816, + "grad_norm": 9.1875, + "learning_rate": 2.0765650568125664e-06, + "loss": 2.9506, + "step": 26834 + }, + { + "epoch": 2.837880710659898, + "grad_norm": 8.0625, + "learning_rate": 2.0763962759184042e-06, + "loss": 2.4485, + "step": 26835 + }, + { + "epoch": 2.8379864636209815, + "grad_norm": 8.75, + "learning_rate": 2.0762274970120425e-06, + "loss": 2.5434, + "step": 26836 + }, + { + "epoch": 2.8380922165820643, + "grad_norm": 8.0, + "learning_rate": 2.0760587200942706e-06, + "loss": 2.4568, + "step": 26837 + }, + { + "epoch": 2.838197969543147, + "grad_norm": 8.75, + "learning_rate": 2.0758899451658827e-06, + "loss": 2.6541, + "step": 26838 + }, + { + "epoch": 2.83830372250423, + "grad_norm": 8.875, + "learning_rate": 2.075721172227669e-06, + "loss": 2.3953, + "step": 26839 + }, + { + "epoch": 2.838409475465313, + "grad_norm": 8.3125, + "learning_rate": 2.0755524012804222e-06, + "loss": 2.5641, + "step": 26840 + }, + { + "epoch": 2.838515228426396, + "grad_norm": 9.0625, + "learning_rate": 2.075383632324935e-06, + "loss": 2.5645, + "step": 26841 + }, + { + "epoch": 2.838620981387479, + "grad_norm": 8.625, + "learning_rate": 2.0752148653619982e-06, + "loss": 2.3751, + "step": 26842 + }, + { + "epoch": 2.838726734348562, + "grad_norm": 8.3125, + "learning_rate": 2.0750461003924054e-06, + "loss": 2.6423, + "step": 26843 + }, + { + "epoch": 2.8388324873096447, + "grad_norm": 8.1875, + "learning_rate": 2.0748773374169466e-06, + "loss": 2.5032, + "step": 26844 + }, + { + "epoch": 2.8389382402707275, + "grad_norm": 9.375, + "learning_rate": 2.074708576436415e-06, + "loss": 2.5198, + "step": 26845 + }, + { + "epoch": 2.8390439932318103, + "grad_norm": 7.875, + "learning_rate": 2.0745398174516017e-06, + "loss": 2.666, + "step": 26846 + }, + { + "epoch": 2.839149746192893, + "grad_norm": 8.25, + "learning_rate": 2.074371060463299e-06, + "loss": 2.5264, + "step": 26847 + }, + { + "epoch": 2.8392554991539765, + "grad_norm": 8.5, + "learning_rate": 2.0742023054722986e-06, + "loss": 2.6004, + "step": 26848 + }, + { + "epoch": 2.8393612521150593, + "grad_norm": 8.1875, + "learning_rate": 2.0740335524793932e-06, + "loss": 2.4863, + "step": 26849 + }, + { + "epoch": 2.839467005076142, + "grad_norm": 8.4375, + "learning_rate": 2.0738648014853747e-06, + "loss": 2.4811, + "step": 26850 + }, + { + "epoch": 2.839572758037225, + "grad_norm": 8.25, + "learning_rate": 2.073696052491033e-06, + "loss": 2.5489, + "step": 26851 + }, + { + "epoch": 2.839678510998308, + "grad_norm": 8.5, + "learning_rate": 2.0735273054971626e-06, + "loss": 2.6266, + "step": 26852 + }, + { + "epoch": 2.839784263959391, + "grad_norm": 8.5, + "learning_rate": 2.073358560504553e-06, + "loss": 2.5651, + "step": 26853 + }, + { + "epoch": 2.839890016920474, + "grad_norm": 8.125, + "learning_rate": 2.073189817513998e-06, + "loss": 2.4179, + "step": 26854 + }, + { + "epoch": 2.8399957698815568, + "grad_norm": 9.625, + "learning_rate": 2.073021076526288e-06, + "loss": 2.5224, + "step": 26855 + }, + { + "epoch": 2.8401015228426396, + "grad_norm": 9.1875, + "learning_rate": 2.072852337542216e-06, + "loss": 2.5654, + "step": 26856 + }, + { + "epoch": 2.8402072758037225, + "grad_norm": 8.75, + "learning_rate": 2.072683600562573e-06, + "loss": 2.6412, + "step": 26857 + }, + { + "epoch": 2.8403130287648053, + "grad_norm": 8.4375, + "learning_rate": 2.072514865588151e-06, + "loss": 2.4934, + "step": 26858 + }, + { + "epoch": 2.840418781725888, + "grad_norm": 12.5625, + "learning_rate": 2.0723461326197428e-06, + "loss": 2.3412, + "step": 26859 + }, + { + "epoch": 2.8405245346869714, + "grad_norm": 8.75, + "learning_rate": 2.072177401658138e-06, + "loss": 2.5263, + "step": 26860 + }, + { + "epoch": 2.8406302876480543, + "grad_norm": 8.25, + "learning_rate": 2.072008672704131e-06, + "loss": 2.5612, + "step": 26861 + }, + { + "epoch": 2.840736040609137, + "grad_norm": 8.1875, + "learning_rate": 2.071839945758511e-06, + "loss": 2.5323, + "step": 26862 + }, + { + "epoch": 2.84084179357022, + "grad_norm": 8.8125, + "learning_rate": 2.071671220822072e-06, + "loss": 2.656, + "step": 26863 + }, + { + "epoch": 2.8409475465313028, + "grad_norm": 8.6875, + "learning_rate": 2.071502497895604e-06, + "loss": 2.5442, + "step": 26864 + }, + { + "epoch": 2.841053299492386, + "grad_norm": 9.0625, + "learning_rate": 2.0713337769799003e-06, + "loss": 2.3103, + "step": 26865 + }, + { + "epoch": 2.8411590524534684, + "grad_norm": 9.1875, + "learning_rate": 2.0711650580757513e-06, + "loss": 2.4628, + "step": 26866 + }, + { + "epoch": 2.8412648054145517, + "grad_norm": 8.3125, + "learning_rate": 2.0709963411839488e-06, + "loss": 2.5559, + "step": 26867 + }, + { + "epoch": 2.8413705583756346, + "grad_norm": 8.5, + "learning_rate": 2.070827626305286e-06, + "loss": 2.6391, + "step": 26868 + }, + { + "epoch": 2.8414763113367174, + "grad_norm": 8.4375, + "learning_rate": 2.070658913440553e-06, + "loss": 2.6009, + "step": 26869 + }, + { + "epoch": 2.8415820642978002, + "grad_norm": 9.1875, + "learning_rate": 2.070490202590543e-06, + "loss": 2.6084, + "step": 26870 + }, + { + "epoch": 2.841687817258883, + "grad_norm": 8.5625, + "learning_rate": 2.0703214937560458e-06, + "loss": 2.6707, + "step": 26871 + }, + { + "epoch": 2.8417935702199664, + "grad_norm": 8.75, + "learning_rate": 2.070152786937855e-06, + "loss": 2.8034, + "step": 26872 + }, + { + "epoch": 2.841899323181049, + "grad_norm": 8.6875, + "learning_rate": 2.0699840821367604e-06, + "loss": 2.4099, + "step": 26873 + }, + { + "epoch": 2.842005076142132, + "grad_norm": 8.3125, + "learning_rate": 2.069815379353556e-06, + "loss": 2.5887, + "step": 26874 + }, + { + "epoch": 2.842110829103215, + "grad_norm": 8.25, + "learning_rate": 2.0696466785890307e-06, + "loss": 2.55, + "step": 26875 + }, + { + "epoch": 2.8422165820642977, + "grad_norm": 10.3125, + "learning_rate": 2.069477979843978e-06, + "loss": 2.5011, + "step": 26876 + }, + { + "epoch": 2.8423223350253806, + "grad_norm": 8.25, + "learning_rate": 2.069309283119189e-06, + "loss": 2.581, + "step": 26877 + }, + { + "epoch": 2.8424280879864634, + "grad_norm": 8.4375, + "learning_rate": 2.0691405884154553e-06, + "loss": 2.5214, + "step": 26878 + }, + { + "epoch": 2.8425338409475467, + "grad_norm": 8.5625, + "learning_rate": 2.0689718957335693e-06, + "loss": 2.263, + "step": 26879 + }, + { + "epoch": 2.8426395939086295, + "grad_norm": 8.125, + "learning_rate": 2.068803205074321e-06, + "loss": 2.5111, + "step": 26880 + }, + { + "epoch": 2.8427453468697124, + "grad_norm": 8.4375, + "learning_rate": 2.068634516438504e-06, + "loss": 2.4264, + "step": 26881 + }, + { + "epoch": 2.842851099830795, + "grad_norm": 8.25, + "learning_rate": 2.0684658298269077e-06, + "loss": 2.6123, + "step": 26882 + }, + { + "epoch": 2.842956852791878, + "grad_norm": 8.1875, + "learning_rate": 2.0682971452403256e-06, + "loss": 2.4191, + "step": 26883 + }, + { + "epoch": 2.8430626057529613, + "grad_norm": 8.875, + "learning_rate": 2.0681284626795477e-06, + "loss": 2.6093, + "step": 26884 + }, + { + "epoch": 2.8431683587140437, + "grad_norm": 8.625, + "learning_rate": 2.067959782145366e-06, + "loss": 2.625, + "step": 26885 + }, + { + "epoch": 2.843274111675127, + "grad_norm": 8.3125, + "learning_rate": 2.0677911036385734e-06, + "loss": 2.3132, + "step": 26886 + }, + { + "epoch": 2.84337986463621, + "grad_norm": 8.625, + "learning_rate": 2.0676224271599597e-06, + "loss": 2.5103, + "step": 26887 + }, + { + "epoch": 2.8434856175972927, + "grad_norm": 8.1875, + "learning_rate": 2.0674537527103172e-06, + "loss": 2.6399, + "step": 26888 + }, + { + "epoch": 2.8435913705583755, + "grad_norm": 8.125, + "learning_rate": 2.067285080290437e-06, + "loss": 2.5092, + "step": 26889 + }, + { + "epoch": 2.8436971235194584, + "grad_norm": 8.625, + "learning_rate": 2.067116409901113e-06, + "loss": 2.7649, + "step": 26890 + }, + { + "epoch": 2.8438028764805416, + "grad_norm": 8.5, + "learning_rate": 2.0669477415431323e-06, + "loss": 2.7163, + "step": 26891 + }, + { + "epoch": 2.8439086294416245, + "grad_norm": 8.0625, + "learning_rate": 2.0667790752172896e-06, + "loss": 2.5693, + "step": 26892 + }, + { + "epoch": 2.8440143824027073, + "grad_norm": 8.0625, + "learning_rate": 2.0666104109243752e-06, + "loss": 2.523, + "step": 26893 + }, + { + "epoch": 2.84412013536379, + "grad_norm": 8.75, + "learning_rate": 2.066441748665181e-06, + "loss": 2.8085, + "step": 26894 + }, + { + "epoch": 2.844225888324873, + "grad_norm": 8.25, + "learning_rate": 2.0662730884404984e-06, + "loss": 2.515, + "step": 26895 + }, + { + "epoch": 2.8443316412859563, + "grad_norm": 8.375, + "learning_rate": 2.0661044302511184e-06, + "loss": 2.5486, + "step": 26896 + }, + { + "epoch": 2.8444373942470387, + "grad_norm": 8.5625, + "learning_rate": 2.0659357740978344e-06, + "loss": 2.4045, + "step": 26897 + }, + { + "epoch": 2.844543147208122, + "grad_norm": 8.9375, + "learning_rate": 2.065767119981435e-06, + "loss": 2.4769, + "step": 26898 + }, + { + "epoch": 2.844648900169205, + "grad_norm": 8.6875, + "learning_rate": 2.065598467902713e-06, + "loss": 2.379, + "step": 26899 + }, + { + "epoch": 2.8447546531302876, + "grad_norm": 8.75, + "learning_rate": 2.0654298178624598e-06, + "loss": 2.6492, + "step": 26900 + }, + { + "epoch": 2.8448604060913705, + "grad_norm": 8.4375, + "learning_rate": 2.065261169861467e-06, + "loss": 2.5711, + "step": 26901 + }, + { + "epoch": 2.8449661590524533, + "grad_norm": 8.25, + "learning_rate": 2.0650925239005252e-06, + "loss": 2.4886, + "step": 26902 + }, + { + "epoch": 2.8450719120135366, + "grad_norm": 8.0625, + "learning_rate": 2.064923879980426e-06, + "loss": 2.4176, + "step": 26903 + }, + { + "epoch": 2.8451776649746194, + "grad_norm": 9.3125, + "learning_rate": 2.0647552381019626e-06, + "loss": 2.7076, + "step": 26904 + }, + { + "epoch": 2.8452834179357023, + "grad_norm": 7.96875, + "learning_rate": 2.0645865982659234e-06, + "loss": 2.3058, + "step": 26905 + }, + { + "epoch": 2.845389170896785, + "grad_norm": 7.96875, + "learning_rate": 2.0644179604731022e-06, + "loss": 2.4471, + "step": 26906 + }, + { + "epoch": 2.845494923857868, + "grad_norm": 8.3125, + "learning_rate": 2.064249324724289e-06, + "loss": 2.5646, + "step": 26907 + }, + { + "epoch": 2.845600676818951, + "grad_norm": 9.0, + "learning_rate": 2.0640806910202752e-06, + "loss": 2.5023, + "step": 26908 + }, + { + "epoch": 2.8457064297800336, + "grad_norm": 8.25, + "learning_rate": 2.063912059361853e-06, + "loss": 2.5213, + "step": 26909 + }, + { + "epoch": 2.845812182741117, + "grad_norm": 8.1875, + "learning_rate": 2.0637434297498128e-06, + "loss": 2.6561, + "step": 26910 + }, + { + "epoch": 2.8459179357021998, + "grad_norm": 8.5, + "learning_rate": 2.0635748021849457e-06, + "loss": 2.6672, + "step": 26911 + }, + { + "epoch": 2.8460236886632826, + "grad_norm": 7.90625, + "learning_rate": 2.0634061766680443e-06, + "loss": 2.3698, + "step": 26912 + }, + { + "epoch": 2.8461294416243654, + "grad_norm": 8.625, + "learning_rate": 2.0632375531998995e-06, + "loss": 2.7322, + "step": 26913 + }, + { + "epoch": 2.8462351945854483, + "grad_norm": 8.5625, + "learning_rate": 2.0630689317813013e-06, + "loss": 2.4705, + "step": 26914 + }, + { + "epoch": 2.8463409475465316, + "grad_norm": 9.0, + "learning_rate": 2.062900312413043e-06, + "loss": 2.7105, + "step": 26915 + }, + { + "epoch": 2.846446700507614, + "grad_norm": 8.8125, + "learning_rate": 2.062731695095914e-06, + "loss": 2.5413, + "step": 26916 + }, + { + "epoch": 2.8465524534686972, + "grad_norm": 8.5, + "learning_rate": 2.0625630798307063e-06, + "loss": 2.5699, + "step": 26917 + }, + { + "epoch": 2.84665820642978, + "grad_norm": 8.4375, + "learning_rate": 2.062394466618211e-06, + "loss": 2.6877, + "step": 26918 + }, + { + "epoch": 2.846763959390863, + "grad_norm": 8.5625, + "learning_rate": 2.0622258554592196e-06, + "loss": 2.621, + "step": 26919 + }, + { + "epoch": 2.8468697123519457, + "grad_norm": 8.6875, + "learning_rate": 2.062057246354524e-06, + "loss": 2.7597, + "step": 26920 + }, + { + "epoch": 2.8469754653130286, + "grad_norm": 8.5625, + "learning_rate": 2.0618886393049133e-06, + "loss": 2.5678, + "step": 26921 + }, + { + "epoch": 2.847081218274112, + "grad_norm": 8.5625, + "learning_rate": 2.0617200343111814e-06, + "loss": 2.5512, + "step": 26922 + }, + { + "epoch": 2.8471869712351947, + "grad_norm": 8.4375, + "learning_rate": 2.061551431374117e-06, + "loss": 2.6846, + "step": 26923 + }, + { + "epoch": 2.8472927241962775, + "grad_norm": 8.625, + "learning_rate": 2.061382830494513e-06, + "loss": 2.6419, + "step": 26924 + }, + { + "epoch": 2.8473984771573604, + "grad_norm": 8.75, + "learning_rate": 2.06121423167316e-06, + "loss": 2.4497, + "step": 26925 + }, + { + "epoch": 2.8475042301184432, + "grad_norm": 8.25, + "learning_rate": 2.0610456349108492e-06, + "loss": 2.5238, + "step": 26926 + }, + { + "epoch": 2.847609983079526, + "grad_norm": 7.96875, + "learning_rate": 2.060877040208371e-06, + "loss": 2.5735, + "step": 26927 + }, + { + "epoch": 2.847715736040609, + "grad_norm": 8.25, + "learning_rate": 2.0607084475665175e-06, + "loss": 2.2835, + "step": 26928 + }, + { + "epoch": 2.847821489001692, + "grad_norm": 8.0625, + "learning_rate": 2.0605398569860808e-06, + "loss": 2.454, + "step": 26929 + }, + { + "epoch": 2.847927241962775, + "grad_norm": 8.25, + "learning_rate": 2.06037126846785e-06, + "loss": 2.45, + "step": 26930 + }, + { + "epoch": 2.848032994923858, + "grad_norm": 8.75, + "learning_rate": 2.060202682012617e-06, + "loss": 2.5547, + "step": 26931 + }, + { + "epoch": 2.8481387478849407, + "grad_norm": 8.125, + "learning_rate": 2.0600340976211726e-06, + "loss": 2.5394, + "step": 26932 + }, + { + "epoch": 2.8482445008460235, + "grad_norm": 9.1875, + "learning_rate": 2.0598655152943087e-06, + "loss": 2.7216, + "step": 26933 + }, + { + "epoch": 2.848350253807107, + "grad_norm": 8.5625, + "learning_rate": 2.0596969350328157e-06, + "loss": 2.5854, + "step": 26934 + }, + { + "epoch": 2.8484560067681897, + "grad_norm": 8.4375, + "learning_rate": 2.0595283568374855e-06, + "loss": 2.6441, + "step": 26935 + }, + { + "epoch": 2.8485617597292725, + "grad_norm": 8.75, + "learning_rate": 2.0593597807091075e-06, + "loss": 2.6023, + "step": 26936 + }, + { + "epoch": 2.8486675126903553, + "grad_norm": 8.125, + "learning_rate": 2.059191206648474e-06, + "loss": 2.483, + "step": 26937 + }, + { + "epoch": 2.848773265651438, + "grad_norm": 8.375, + "learning_rate": 2.0590226346563767e-06, + "loss": 2.4409, + "step": 26938 + }, + { + "epoch": 2.848879018612521, + "grad_norm": 7.9375, + "learning_rate": 2.058854064733605e-06, + "loss": 2.5054, + "step": 26939 + }, + { + "epoch": 2.848984771573604, + "grad_norm": 8.375, + "learning_rate": 2.0586854968809517e-06, + "loss": 2.6828, + "step": 26940 + }, + { + "epoch": 2.849090524534687, + "grad_norm": 8.875, + "learning_rate": 2.0585169310992057e-06, + "loss": 2.4375, + "step": 26941 + }, + { + "epoch": 2.84919627749577, + "grad_norm": 8.4375, + "learning_rate": 2.05834836738916e-06, + "loss": 2.5883, + "step": 26942 + }, + { + "epoch": 2.849302030456853, + "grad_norm": 8.4375, + "learning_rate": 2.0581798057516045e-06, + "loss": 2.6149, + "step": 26943 + }, + { + "epoch": 2.8494077834179357, + "grad_norm": 8.125, + "learning_rate": 2.0580112461873315e-06, + "loss": 2.6161, + "step": 26944 + }, + { + "epoch": 2.8495135363790185, + "grad_norm": 8.5625, + "learning_rate": 2.0578426886971296e-06, + "loss": 2.6409, + "step": 26945 + }, + { + "epoch": 2.849619289340102, + "grad_norm": 8.875, + "learning_rate": 2.0576741332817912e-06, + "loss": 2.606, + "step": 26946 + }, + { + "epoch": 2.849725042301184, + "grad_norm": 8.5625, + "learning_rate": 2.0575055799421075e-06, + "loss": 2.6257, + "step": 26947 + }, + { + "epoch": 2.8498307952622675, + "grad_norm": 8.6875, + "learning_rate": 2.057337028678869e-06, + "loss": 2.628, + "step": 26948 + }, + { + "epoch": 2.8499365482233503, + "grad_norm": 8.375, + "learning_rate": 2.057168479492867e-06, + "loss": 2.5365, + "step": 26949 + }, + { + "epoch": 2.850042301184433, + "grad_norm": 8.6875, + "learning_rate": 2.056999932384892e-06, + "loss": 2.4302, + "step": 26950 + }, + { + "epoch": 2.850148054145516, + "grad_norm": 8.1875, + "learning_rate": 2.056831387355736e-06, + "loss": 2.5573, + "step": 26951 + }, + { + "epoch": 2.850253807106599, + "grad_norm": 8.375, + "learning_rate": 2.0566628444061883e-06, + "loss": 2.5889, + "step": 26952 + }, + { + "epoch": 2.850359560067682, + "grad_norm": 8.625, + "learning_rate": 2.0564943035370404e-06, + "loss": 2.6352, + "step": 26953 + }, + { + "epoch": 2.850465313028765, + "grad_norm": 8.25, + "learning_rate": 2.0563257647490833e-06, + "loss": 2.5695, + "step": 26954 + }, + { + "epoch": 2.8505710659898478, + "grad_norm": 8.375, + "learning_rate": 2.0561572280431082e-06, + "loss": 2.4605, + "step": 26955 + }, + { + "epoch": 2.8506768189509306, + "grad_norm": 8.375, + "learning_rate": 2.055988693419906e-06, + "loss": 2.3958, + "step": 26956 + }, + { + "epoch": 2.8507825719120135, + "grad_norm": 8.375, + "learning_rate": 2.0558201608802664e-06, + "loss": 2.2272, + "step": 26957 + }, + { + "epoch": 2.8508883248730963, + "grad_norm": 8.625, + "learning_rate": 2.0556516304249824e-06, + "loss": 2.4482, + "step": 26958 + }, + { + "epoch": 2.850994077834179, + "grad_norm": 8.75, + "learning_rate": 2.055483102054842e-06, + "loss": 2.3353, + "step": 26959 + }, + { + "epoch": 2.8510998307952624, + "grad_norm": 8.6875, + "learning_rate": 2.055314575770639e-06, + "loss": 2.4776, + "step": 26960 + }, + { + "epoch": 2.8512055837563453, + "grad_norm": 8.9375, + "learning_rate": 2.055146051573162e-06, + "loss": 2.4836, + "step": 26961 + }, + { + "epoch": 2.851311336717428, + "grad_norm": 8.6875, + "learning_rate": 2.0549775294632033e-06, + "loss": 2.4693, + "step": 26962 + }, + { + "epoch": 2.851417089678511, + "grad_norm": 8.25, + "learning_rate": 2.054809009441552e-06, + "loss": 2.4752, + "step": 26963 + }, + { + "epoch": 2.8515228426395938, + "grad_norm": 8.6875, + "learning_rate": 2.0546404915090005e-06, + "loss": 2.6207, + "step": 26964 + }, + { + "epoch": 2.851628595600677, + "grad_norm": 8.4375, + "learning_rate": 2.0544719756663393e-06, + "loss": 2.514, + "step": 26965 + }, + { + "epoch": 2.85173434856176, + "grad_norm": 9.0, + "learning_rate": 2.0543034619143582e-06, + "loss": 2.9011, + "step": 26966 + }, + { + "epoch": 2.8518401015228427, + "grad_norm": 9.125, + "learning_rate": 2.05413495025385e-06, + "loss": 2.4471, + "step": 26967 + }, + { + "epoch": 2.8519458544839256, + "grad_norm": 8.5, + "learning_rate": 2.053966440685603e-06, + "loss": 2.349, + "step": 26968 + }, + { + "epoch": 2.8520516074450084, + "grad_norm": 8.1875, + "learning_rate": 2.0537979332104095e-06, + "loss": 2.4796, + "step": 26969 + }, + { + "epoch": 2.8521573604060912, + "grad_norm": 8.1875, + "learning_rate": 2.0536294278290596e-06, + "loss": 2.4945, + "step": 26970 + }, + { + "epoch": 2.852263113367174, + "grad_norm": 8.75, + "learning_rate": 2.0534609245423446e-06, + "loss": 2.5851, + "step": 26971 + }, + { + "epoch": 2.8523688663282574, + "grad_norm": 8.875, + "learning_rate": 2.053292423351054e-06, + "loss": 2.6372, + "step": 26972 + }, + { + "epoch": 2.85247461928934, + "grad_norm": 8.625, + "learning_rate": 2.0531239242559795e-06, + "loss": 2.4548, + "step": 26973 + }, + { + "epoch": 2.852580372250423, + "grad_norm": 8.375, + "learning_rate": 2.0529554272579133e-06, + "loss": 2.5533, + "step": 26974 + }, + { + "epoch": 2.852686125211506, + "grad_norm": 8.8125, + "learning_rate": 2.052786932357642e-06, + "loss": 2.5966, + "step": 26975 + }, + { + "epoch": 2.8527918781725887, + "grad_norm": 8.4375, + "learning_rate": 2.0526184395559612e-06, + "loss": 2.6771, + "step": 26976 + }, + { + "epoch": 2.852897631133672, + "grad_norm": 8.125, + "learning_rate": 2.052449948853657e-06, + "loss": 2.5315, + "step": 26977 + }, + { + "epoch": 2.8530033840947544, + "grad_norm": 8.1875, + "learning_rate": 2.052281460251524e-06, + "loss": 2.5723, + "step": 26978 + }, + { + "epoch": 2.8531091370558377, + "grad_norm": 9.0625, + "learning_rate": 2.0521129737503494e-06, + "loss": 2.5755, + "step": 26979 + }, + { + "epoch": 2.8532148900169205, + "grad_norm": 8.875, + "learning_rate": 2.051944489350926e-06, + "loss": 2.5582, + "step": 26980 + }, + { + "epoch": 2.8533206429780034, + "grad_norm": 8.8125, + "learning_rate": 2.051776007054044e-06, + "loss": 2.4332, + "step": 26981 + }, + { + "epoch": 2.853426395939086, + "grad_norm": 8.75, + "learning_rate": 2.051607526860493e-06, + "loss": 2.5832, + "step": 26982 + }, + { + "epoch": 2.853532148900169, + "grad_norm": 8.5625, + "learning_rate": 2.051439048771066e-06, + "loss": 2.5309, + "step": 26983 + }, + { + "epoch": 2.8536379018612523, + "grad_norm": 8.4375, + "learning_rate": 2.051270572786551e-06, + "loss": 2.5903, + "step": 26984 + }, + { + "epoch": 2.853743654822335, + "grad_norm": 8.3125, + "learning_rate": 2.0511020989077406e-06, + "loss": 2.6327, + "step": 26985 + }, + { + "epoch": 2.853849407783418, + "grad_norm": 8.125, + "learning_rate": 2.0509336271354235e-06, + "loss": 2.5594, + "step": 26986 + }, + { + "epoch": 2.853955160744501, + "grad_norm": 8.1875, + "learning_rate": 2.0507651574703917e-06, + "loss": 2.6124, + "step": 26987 + }, + { + "epoch": 2.8540609137055837, + "grad_norm": 8.1875, + "learning_rate": 2.050596689913435e-06, + "loss": 2.6764, + "step": 26988 + }, + { + "epoch": 2.8541666666666665, + "grad_norm": 8.375, + "learning_rate": 2.0504282244653453e-06, + "loss": 2.4194, + "step": 26989 + }, + { + "epoch": 2.8542724196277494, + "grad_norm": 8.4375, + "learning_rate": 2.050259761126911e-06, + "loss": 2.4459, + "step": 26990 + }, + { + "epoch": 2.8543781725888326, + "grad_norm": 10.3125, + "learning_rate": 2.050091299898923e-06, + "loss": 2.6708, + "step": 26991 + }, + { + "epoch": 2.8544839255499155, + "grad_norm": 8.0625, + "learning_rate": 2.0499228407821743e-06, + "loss": 2.5467, + "step": 26992 + }, + { + "epoch": 2.8545896785109983, + "grad_norm": 8.4375, + "learning_rate": 2.0497543837774522e-06, + "loss": 2.6406, + "step": 26993 + }, + { + "epoch": 2.854695431472081, + "grad_norm": 8.6875, + "learning_rate": 2.0495859288855497e-06, + "loss": 2.7163, + "step": 26994 + }, + { + "epoch": 2.854801184433164, + "grad_norm": 8.0625, + "learning_rate": 2.0494174761072552e-06, + "loss": 2.1657, + "step": 26995 + }, + { + "epoch": 2.8549069373942473, + "grad_norm": 8.3125, + "learning_rate": 2.0492490254433617e-06, + "loss": 2.6388, + "step": 26996 + }, + { + "epoch": 2.8550126903553297, + "grad_norm": 8.25, + "learning_rate": 2.0490805768946564e-06, + "loss": 2.6531, + "step": 26997 + }, + { + "epoch": 2.855118443316413, + "grad_norm": 8.0625, + "learning_rate": 2.048912130461933e-06, + "loss": 2.357, + "step": 26998 + }, + { + "epoch": 2.855224196277496, + "grad_norm": 8.3125, + "learning_rate": 2.0487436861459796e-06, + "loss": 2.3785, + "step": 26999 + }, + { + "epoch": 2.8553299492385786, + "grad_norm": 8.5, + "learning_rate": 2.0485752439475875e-06, + "loss": 2.4928, + "step": 27000 + }, + { + "epoch": 2.8554357021996615, + "grad_norm": 8.9375, + "learning_rate": 2.048406803867547e-06, + "loss": 2.4197, + "step": 27001 + }, + { + "epoch": 2.8555414551607443, + "grad_norm": 8.25, + "learning_rate": 2.048238365906649e-06, + "loss": 2.4724, + "step": 27002 + }, + { + "epoch": 2.8556472081218276, + "grad_norm": 8.125, + "learning_rate": 2.048069930065684e-06, + "loss": 2.4594, + "step": 27003 + }, + { + "epoch": 2.8557529610829104, + "grad_norm": 8.1875, + "learning_rate": 2.0479014963454415e-06, + "loss": 2.3081, + "step": 27004 + }, + { + "epoch": 2.8558587140439933, + "grad_norm": 9.0625, + "learning_rate": 2.047733064746713e-06, + "loss": 2.6104, + "step": 27005 + }, + { + "epoch": 2.855964467005076, + "grad_norm": 8.25, + "learning_rate": 2.0475646352702876e-06, + "loss": 2.6463, + "step": 27006 + }, + { + "epoch": 2.856070219966159, + "grad_norm": 8.5, + "learning_rate": 2.0473962079169568e-06, + "loss": 2.4909, + "step": 27007 + }, + { + "epoch": 2.8561759729272422, + "grad_norm": 8.25, + "learning_rate": 2.04722778268751e-06, + "loss": 2.5365, + "step": 27008 + }, + { + "epoch": 2.8562817258883246, + "grad_norm": 8.25, + "learning_rate": 2.0470593595827374e-06, + "loss": 2.4267, + "step": 27009 + }, + { + "epoch": 2.856387478849408, + "grad_norm": 8.875, + "learning_rate": 2.0468909386034316e-06, + "loss": 2.7005, + "step": 27010 + }, + { + "epoch": 2.8564932318104908, + "grad_norm": 8.625, + "learning_rate": 2.04672251975038e-06, + "loss": 2.648, + "step": 27011 + }, + { + "epoch": 2.8565989847715736, + "grad_norm": 8.75, + "learning_rate": 2.0465541030243756e-06, + "loss": 2.4295, + "step": 27012 + }, + { + "epoch": 2.8567047377326564, + "grad_norm": 8.6875, + "learning_rate": 2.0463856884262056e-06, + "loss": 2.5548, + "step": 27013 + }, + { + "epoch": 2.8568104906937393, + "grad_norm": 8.5625, + "learning_rate": 2.0462172759566632e-06, + "loss": 2.619, + "step": 27014 + }, + { + "epoch": 2.8569162436548226, + "grad_norm": 8.0625, + "learning_rate": 2.046048865616537e-06, + "loss": 2.4672, + "step": 27015 + }, + { + "epoch": 2.8570219966159054, + "grad_norm": 8.1875, + "learning_rate": 2.0458804574066185e-06, + "loss": 2.4513, + "step": 27016 + }, + { + "epoch": 2.8571277495769882, + "grad_norm": 8.25, + "learning_rate": 2.0457120513276967e-06, + "loss": 2.5816, + "step": 27017 + }, + { + "epoch": 2.857233502538071, + "grad_norm": 8.3125, + "learning_rate": 2.0455436473805624e-06, + "loss": 2.5074, + "step": 27018 + }, + { + "epoch": 2.857339255499154, + "grad_norm": 8.3125, + "learning_rate": 2.0453752455660063e-06, + "loss": 2.4827, + "step": 27019 + }, + { + "epoch": 2.8574450084602367, + "grad_norm": 8.75, + "learning_rate": 2.0452068458848175e-06, + "loss": 2.5759, + "step": 27020 + }, + { + "epoch": 2.8575507614213196, + "grad_norm": 8.25, + "learning_rate": 2.0450384483377883e-06, + "loss": 2.5497, + "step": 27021 + }, + { + "epoch": 2.857656514382403, + "grad_norm": 8.5, + "learning_rate": 2.0448700529257065e-06, + "loss": 2.617, + "step": 27022 + }, + { + "epoch": 2.8577622673434857, + "grad_norm": 8.625, + "learning_rate": 2.044701659649364e-06, + "loss": 2.3677, + "step": 27023 + }, + { + "epoch": 2.8578680203045685, + "grad_norm": 9.125, + "learning_rate": 2.0445332685095497e-06, + "loss": 2.7137, + "step": 27024 + }, + { + "epoch": 2.8579737732656514, + "grad_norm": 7.875, + "learning_rate": 2.044364879507055e-06, + "loss": 2.368, + "step": 27025 + }, + { + "epoch": 2.8580795262267342, + "grad_norm": 9.125, + "learning_rate": 2.0441964926426696e-06, + "loss": 2.5618, + "step": 27026 + }, + { + "epoch": 2.8581852791878175, + "grad_norm": 8.375, + "learning_rate": 2.0440281079171828e-06, + "loss": 2.566, + "step": 27027 + }, + { + "epoch": 2.8582910321489, + "grad_norm": 8.3125, + "learning_rate": 2.043859725331387e-06, + "loss": 2.5026, + "step": 27028 + }, + { + "epoch": 2.858396785109983, + "grad_norm": 8.125, + "learning_rate": 2.0436913448860697e-06, + "loss": 2.6189, + "step": 27029 + }, + { + "epoch": 2.858502538071066, + "grad_norm": 8.5, + "learning_rate": 2.0435229665820235e-06, + "loss": 2.4018, + "step": 27030 + }, + { + "epoch": 2.858608291032149, + "grad_norm": 8.5, + "learning_rate": 2.0433545904200365e-06, + "loss": 2.4078, + "step": 27031 + }, + { + "epoch": 2.8587140439932317, + "grad_norm": 9.0625, + "learning_rate": 2.0431862164008996e-06, + "loss": 2.5674, + "step": 27032 + }, + { + "epoch": 2.8588197969543145, + "grad_norm": 8.4375, + "learning_rate": 2.043017844525403e-06, + "loss": 2.681, + "step": 27033 + }, + { + "epoch": 2.858925549915398, + "grad_norm": 8.5625, + "learning_rate": 2.0428494747943368e-06, + "loss": 2.4337, + "step": 27034 + }, + { + "epoch": 2.8590313028764807, + "grad_norm": 8.3125, + "learning_rate": 2.0426811072084914e-06, + "loss": 2.7316, + "step": 27035 + }, + { + "epoch": 2.8591370558375635, + "grad_norm": 8.5, + "learning_rate": 2.042512741768656e-06, + "loss": 2.5711, + "step": 27036 + }, + { + "epoch": 2.8592428087986463, + "grad_norm": 8.375, + "learning_rate": 2.0423443784756223e-06, + "loss": 2.6583, + "step": 27037 + }, + { + "epoch": 2.859348561759729, + "grad_norm": 8.25, + "learning_rate": 2.042176017330178e-06, + "loss": 2.2718, + "step": 27038 + }, + { + "epoch": 2.859454314720812, + "grad_norm": 8.5, + "learning_rate": 2.0420076583331153e-06, + "loss": 2.5012, + "step": 27039 + }, + { + "epoch": 2.859560067681895, + "grad_norm": 8.6875, + "learning_rate": 2.0418393014852226e-06, + "loss": 2.3745, + "step": 27040 + }, + { + "epoch": 2.859665820642978, + "grad_norm": 8.25, + "learning_rate": 2.041670946787291e-06, + "loss": 2.3928, + "step": 27041 + }, + { + "epoch": 2.859771573604061, + "grad_norm": 7.84375, + "learning_rate": 2.0415025942401103e-06, + "loss": 2.3569, + "step": 27042 + }, + { + "epoch": 2.859877326565144, + "grad_norm": 9.0625, + "learning_rate": 2.04133424384447e-06, + "loss": 2.6543, + "step": 27043 + }, + { + "epoch": 2.8599830795262267, + "grad_norm": 8.9375, + "learning_rate": 2.0411658956011614e-06, + "loss": 2.5697, + "step": 27044 + }, + { + "epoch": 2.8600888324873095, + "grad_norm": 8.375, + "learning_rate": 2.040997549510972e-06, + "loss": 2.4615, + "step": 27045 + }, + { + "epoch": 2.860194585448393, + "grad_norm": 8.0, + "learning_rate": 2.0408292055746953e-06, + "loss": 2.596, + "step": 27046 + }, + { + "epoch": 2.8603003384094756, + "grad_norm": 8.875, + "learning_rate": 2.040660863793118e-06, + "loss": 2.6233, + "step": 27047 + }, + { + "epoch": 2.8604060913705585, + "grad_norm": 7.96875, + "learning_rate": 2.040492524167032e-06, + "loss": 2.5496, + "step": 27048 + }, + { + "epoch": 2.8605118443316413, + "grad_norm": 8.5, + "learning_rate": 2.0403241866972263e-06, + "loss": 2.5162, + "step": 27049 + }, + { + "epoch": 2.860617597292724, + "grad_norm": 8.9375, + "learning_rate": 2.040155851384492e-06, + "loss": 2.5521, + "step": 27050 + }, + { + "epoch": 2.860723350253807, + "grad_norm": 8.4375, + "learning_rate": 2.039987518229617e-06, + "loss": 2.643, + "step": 27051 + }, + { + "epoch": 2.86082910321489, + "grad_norm": 8.125, + "learning_rate": 2.0398191872333925e-06, + "loss": 2.5125, + "step": 27052 + }, + { + "epoch": 2.860934856175973, + "grad_norm": 8.625, + "learning_rate": 2.0396508583966096e-06, + "loss": 2.6208, + "step": 27053 + }, + { + "epoch": 2.861040609137056, + "grad_norm": 9.375, + "learning_rate": 2.039482531720056e-06, + "loss": 2.8016, + "step": 27054 + }, + { + "epoch": 2.8611463620981388, + "grad_norm": 8.625, + "learning_rate": 2.039314207204523e-06, + "loss": 2.5242, + "step": 27055 + }, + { + "epoch": 2.8612521150592216, + "grad_norm": 11.3125, + "learning_rate": 2.0391458848507994e-06, + "loss": 2.4529, + "step": 27056 + }, + { + "epoch": 2.8613578680203045, + "grad_norm": 8.25, + "learning_rate": 2.0389775646596763e-06, + "loss": 2.5782, + "step": 27057 + }, + { + "epoch": 2.8614636209813877, + "grad_norm": 8.5, + "learning_rate": 2.0388092466319422e-06, + "loss": 2.5701, + "step": 27058 + }, + { + "epoch": 2.86156937394247, + "grad_norm": 8.125, + "learning_rate": 2.038640930768389e-06, + "loss": 2.3763, + "step": 27059 + }, + { + "epoch": 2.8616751269035534, + "grad_norm": 8.125, + "learning_rate": 2.038472617069804e-06, + "loss": 2.3172, + "step": 27060 + }, + { + "epoch": 2.8617808798646363, + "grad_norm": 8.25, + "learning_rate": 2.0383043055369777e-06, + "loss": 2.3906, + "step": 27061 + }, + { + "epoch": 2.861886632825719, + "grad_norm": 8.625, + "learning_rate": 2.0381359961707023e-06, + "loss": 2.5862, + "step": 27062 + }, + { + "epoch": 2.861992385786802, + "grad_norm": 8.5, + "learning_rate": 2.0379676889717644e-06, + "loss": 2.571, + "step": 27063 + }, + { + "epoch": 2.8620981387478848, + "grad_norm": 8.5, + "learning_rate": 2.037799383940956e-06, + "loss": 2.2781, + "step": 27064 + }, + { + "epoch": 2.862203891708968, + "grad_norm": 8.125, + "learning_rate": 2.0376310810790656e-06, + "loss": 2.4761, + "step": 27065 + }, + { + "epoch": 2.862309644670051, + "grad_norm": 8.25, + "learning_rate": 2.037462780386884e-06, + "loss": 2.6131, + "step": 27066 + }, + { + "epoch": 2.8624153976311337, + "grad_norm": 9.3125, + "learning_rate": 2.037294481865199e-06, + "loss": 2.8746, + "step": 27067 + }, + { + "epoch": 2.8625211505922166, + "grad_norm": 8.6875, + "learning_rate": 2.0371261855148037e-06, + "loss": 2.5285, + "step": 27068 + }, + { + "epoch": 2.8626269035532994, + "grad_norm": 8.375, + "learning_rate": 2.0369578913364847e-06, + "loss": 2.4284, + "step": 27069 + }, + { + "epoch": 2.8627326565143822, + "grad_norm": 9.1875, + "learning_rate": 2.0367895993310325e-06, + "loss": 2.562, + "step": 27070 + }, + { + "epoch": 2.862838409475465, + "grad_norm": 8.5, + "learning_rate": 2.036621309499238e-06, + "loss": 2.6884, + "step": 27071 + }, + { + "epoch": 2.8629441624365484, + "grad_norm": 8.75, + "learning_rate": 2.0364530218418902e-06, + "loss": 2.615, + "step": 27072 + }, + { + "epoch": 2.863049915397631, + "grad_norm": 8.4375, + "learning_rate": 2.036284736359779e-06, + "loss": 2.511, + "step": 27073 + }, + { + "epoch": 2.863155668358714, + "grad_norm": 8.9375, + "learning_rate": 2.036116453053693e-06, + "loss": 2.7239, + "step": 27074 + }, + { + "epoch": 2.863261421319797, + "grad_norm": 8.5625, + "learning_rate": 2.035948171924424e-06, + "loss": 2.2985, + "step": 27075 + }, + { + "epoch": 2.8633671742808797, + "grad_norm": 8.25, + "learning_rate": 2.0357798929727597e-06, + "loss": 2.6186, + "step": 27076 + }, + { + "epoch": 2.863472927241963, + "grad_norm": 8.6875, + "learning_rate": 2.0356116161994905e-06, + "loss": 2.6123, + "step": 27077 + }, + { + "epoch": 2.863578680203046, + "grad_norm": 8.5625, + "learning_rate": 2.035443341605406e-06, + "loss": 2.5722, + "step": 27078 + }, + { + "epoch": 2.8636844331641287, + "grad_norm": 8.5625, + "learning_rate": 2.0352750691912955e-06, + "loss": 2.5388, + "step": 27079 + }, + { + "epoch": 2.8637901861252115, + "grad_norm": 8.0625, + "learning_rate": 2.03510679895795e-06, + "loss": 2.2997, + "step": 27080 + }, + { + "epoch": 2.8638959390862944, + "grad_norm": 8.375, + "learning_rate": 2.034938530906157e-06, + "loss": 2.4913, + "step": 27081 + }, + { + "epoch": 2.864001692047377, + "grad_norm": 9.0, + "learning_rate": 2.034770265036709e-06, + "loss": 2.7146, + "step": 27082 + }, + { + "epoch": 2.86410744500846, + "grad_norm": 8.25, + "learning_rate": 2.034602001350392e-06, + "loss": 2.456, + "step": 27083 + }, + { + "epoch": 2.8642131979695433, + "grad_norm": 8.6875, + "learning_rate": 2.034433739847999e-06, + "loss": 2.6402, + "step": 27084 + }, + { + "epoch": 2.864318950930626, + "grad_norm": 8.3125, + "learning_rate": 2.034265480530317e-06, + "loss": 2.512, + "step": 27085 + }, + { + "epoch": 2.864424703891709, + "grad_norm": 8.5, + "learning_rate": 2.034097223398137e-06, + "loss": 2.4699, + "step": 27086 + }, + { + "epoch": 2.864530456852792, + "grad_norm": 7.90625, + "learning_rate": 2.033928968452248e-06, + "loss": 2.5219, + "step": 27087 + }, + { + "epoch": 2.8646362098138747, + "grad_norm": 8.25, + "learning_rate": 2.03376071569344e-06, + "loss": 2.4522, + "step": 27088 + }, + { + "epoch": 2.864741962774958, + "grad_norm": 8.4375, + "learning_rate": 2.033592465122502e-06, + "loss": 2.6977, + "step": 27089 + }, + { + "epoch": 2.8648477157360404, + "grad_norm": 8.6875, + "learning_rate": 2.033424216740224e-06, + "loss": 2.5251, + "step": 27090 + }, + { + "epoch": 2.8649534686971236, + "grad_norm": 8.1875, + "learning_rate": 2.0332559705473957e-06, + "loss": 2.5854, + "step": 27091 + }, + { + "epoch": 2.8650592216582065, + "grad_norm": 7.75, + "learning_rate": 2.033087726544806e-06, + "loss": 2.3693, + "step": 27092 + }, + { + "epoch": 2.8651649746192893, + "grad_norm": 8.4375, + "learning_rate": 2.0329194847332447e-06, + "loss": 2.4838, + "step": 27093 + }, + { + "epoch": 2.865270727580372, + "grad_norm": 8.625, + "learning_rate": 2.0327512451135e-06, + "loss": 2.5404, + "step": 27094 + }, + { + "epoch": 2.865376480541455, + "grad_norm": 8.25, + "learning_rate": 2.032583007686364e-06, + "loss": 2.2642, + "step": 27095 + }, + { + "epoch": 2.8654822335025383, + "grad_norm": 8.3125, + "learning_rate": 2.0324147724526244e-06, + "loss": 1.9858, + "step": 27096 + }, + { + "epoch": 2.865587986463621, + "grad_norm": 8.625, + "learning_rate": 2.0322465394130706e-06, + "loss": 2.6256, + "step": 27097 + }, + { + "epoch": 2.865693739424704, + "grad_norm": 8.25, + "learning_rate": 2.0320783085684934e-06, + "loss": 2.4614, + "step": 27098 + }, + { + "epoch": 2.865799492385787, + "grad_norm": 8.25, + "learning_rate": 2.0319100799196802e-06, + "loss": 2.5541, + "step": 27099 + }, + { + "epoch": 2.8659052453468696, + "grad_norm": 8.75, + "learning_rate": 2.031741853467423e-06, + "loss": 2.4415, + "step": 27100 + }, + { + "epoch": 2.8660109983079525, + "grad_norm": 8.8125, + "learning_rate": 2.031573629212509e-06, + "loss": 2.5641, + "step": 27101 + }, + { + "epoch": 2.8661167512690353, + "grad_norm": 9.1875, + "learning_rate": 2.0314054071557282e-06, + "loss": 2.7242, + "step": 27102 + }, + { + "epoch": 2.8662225042301186, + "grad_norm": 8.9375, + "learning_rate": 2.03123718729787e-06, + "loss": 2.6025, + "step": 27103 + }, + { + "epoch": 2.8663282571912014, + "grad_norm": 8.0, + "learning_rate": 2.031068969639725e-06, + "loss": 2.3296, + "step": 27104 + }, + { + "epoch": 2.8664340101522843, + "grad_norm": 8.6875, + "learning_rate": 2.03090075418208e-06, + "loss": 2.4776, + "step": 27105 + }, + { + "epoch": 2.866539763113367, + "grad_norm": 8.5625, + "learning_rate": 2.0307325409257263e-06, + "loss": 2.5961, + "step": 27106 + }, + { + "epoch": 2.86664551607445, + "grad_norm": 8.0625, + "learning_rate": 2.0305643298714544e-06, + "loss": 2.6498, + "step": 27107 + }, + { + "epoch": 2.8667512690355332, + "grad_norm": 8.8125, + "learning_rate": 2.030396121020051e-06, + "loss": 2.467, + "step": 27108 + }, + { + "epoch": 2.8668570219966156, + "grad_norm": 8.5, + "learning_rate": 2.0302279143723064e-06, + "loss": 2.5492, + "step": 27109 + }, + { + "epoch": 2.866962774957699, + "grad_norm": 8.3125, + "learning_rate": 2.0300597099290098e-06, + "loss": 2.4999, + "step": 27110 + }, + { + "epoch": 2.8670685279187818, + "grad_norm": 8.4375, + "learning_rate": 2.0298915076909515e-06, + "loss": 2.4195, + "step": 27111 + }, + { + "epoch": 2.8671742808798646, + "grad_norm": 8.625, + "learning_rate": 2.0297233076589194e-06, + "loss": 2.4777, + "step": 27112 + }, + { + "epoch": 2.8672800338409474, + "grad_norm": 8.5625, + "learning_rate": 2.0295551098337046e-06, + "loss": 2.482, + "step": 27113 + }, + { + "epoch": 2.8673857868020303, + "grad_norm": 8.75, + "learning_rate": 2.0293869142160944e-06, + "loss": 2.6573, + "step": 27114 + }, + { + "epoch": 2.8674915397631136, + "grad_norm": 8.75, + "learning_rate": 2.029218720806878e-06, + "loss": 2.7398, + "step": 27115 + }, + { + "epoch": 2.8675972927241964, + "grad_norm": 8.5, + "learning_rate": 2.0290505296068476e-06, + "loss": 2.5334, + "step": 27116 + }, + { + "epoch": 2.8677030456852792, + "grad_norm": 8.1875, + "learning_rate": 2.028882340616789e-06, + "loss": 2.5408, + "step": 27117 + }, + { + "epoch": 2.867808798646362, + "grad_norm": 8.125, + "learning_rate": 2.028714153837494e-06, + "loss": 2.5657, + "step": 27118 + }, + { + "epoch": 2.867914551607445, + "grad_norm": 9.0625, + "learning_rate": 2.0285459692697494e-06, + "loss": 2.6518, + "step": 27119 + }, + { + "epoch": 2.868020304568528, + "grad_norm": 8.0, + "learning_rate": 2.028377786914347e-06, + "loss": 2.3473, + "step": 27120 + }, + { + "epoch": 2.8681260575296106, + "grad_norm": 8.25, + "learning_rate": 2.0282096067720737e-06, + "loss": 2.6304, + "step": 27121 + }, + { + "epoch": 2.868231810490694, + "grad_norm": 8.4375, + "learning_rate": 2.0280414288437213e-06, + "loss": 2.5406, + "step": 27122 + }, + { + "epoch": 2.8683375634517767, + "grad_norm": 8.4375, + "learning_rate": 2.027873253130076e-06, + "loss": 2.4768, + "step": 27123 + }, + { + "epoch": 2.8684433164128595, + "grad_norm": 8.25, + "learning_rate": 2.0277050796319285e-06, + "loss": 2.4996, + "step": 27124 + }, + { + "epoch": 2.8685490693739424, + "grad_norm": 8.0, + "learning_rate": 2.0275369083500683e-06, + "loss": 2.4945, + "step": 27125 + }, + { + "epoch": 2.8686548223350252, + "grad_norm": 8.4375, + "learning_rate": 2.027368739285284e-06, + "loss": 2.6735, + "step": 27126 + }, + { + "epoch": 2.8687605752961085, + "grad_norm": 8.5625, + "learning_rate": 2.027200572438365e-06, + "loss": 2.5645, + "step": 27127 + }, + { + "epoch": 2.8688663282571913, + "grad_norm": 8.625, + "learning_rate": 2.0270324078101008e-06, + "loss": 2.73, + "step": 27128 + }, + { + "epoch": 2.868972081218274, + "grad_norm": 8.4375, + "learning_rate": 2.02686424540128e-06, + "loss": 2.517, + "step": 27129 + }, + { + "epoch": 2.869077834179357, + "grad_norm": 8.5625, + "learning_rate": 2.026696085212691e-06, + "loss": 2.6177, + "step": 27130 + }, + { + "epoch": 2.86918358714044, + "grad_norm": 8.375, + "learning_rate": 2.0265279272451247e-06, + "loss": 2.5581, + "step": 27131 + }, + { + "epoch": 2.8692893401015227, + "grad_norm": 7.90625, + "learning_rate": 2.0263597714993683e-06, + "loss": 2.4498, + "step": 27132 + }, + { + "epoch": 2.8693950930626055, + "grad_norm": 8.4375, + "learning_rate": 2.0261916179762116e-06, + "loss": 2.5624, + "step": 27133 + }, + { + "epoch": 2.869500846023689, + "grad_norm": 8.375, + "learning_rate": 2.0260234666764447e-06, + "loss": 2.7647, + "step": 27134 + }, + { + "epoch": 2.8696065989847717, + "grad_norm": 8.25, + "learning_rate": 2.025855317600855e-06, + "loss": 2.5706, + "step": 27135 + }, + { + "epoch": 2.8697123519458545, + "grad_norm": 8.625, + "learning_rate": 2.0256871707502342e-06, + "loss": 2.6529, + "step": 27136 + }, + { + "epoch": 2.8698181049069373, + "grad_norm": 8.5, + "learning_rate": 2.025519026125367e-06, + "loss": 2.554, + "step": 27137 + }, + { + "epoch": 2.86992385786802, + "grad_norm": 7.875, + "learning_rate": 2.025350883727047e-06, + "loss": 2.5245, + "step": 27138 + }, + { + "epoch": 2.8700296108291035, + "grad_norm": 8.375, + "learning_rate": 2.02518274355606e-06, + "loss": 2.4758, + "step": 27139 + }, + { + "epoch": 2.870135363790186, + "grad_norm": 8.75, + "learning_rate": 2.025014605613197e-06, + "loss": 2.6376, + "step": 27140 + }, + { + "epoch": 2.870241116751269, + "grad_norm": 7.84375, + "learning_rate": 2.0248464698992453e-06, + "loss": 2.4847, + "step": 27141 + }, + { + "epoch": 2.870346869712352, + "grad_norm": 8.8125, + "learning_rate": 2.0246783364149945e-06, + "loss": 2.2811, + "step": 27142 + }, + { + "epoch": 2.870452622673435, + "grad_norm": 8.1875, + "learning_rate": 2.024510205161235e-06, + "loss": 2.4617, + "step": 27143 + }, + { + "epoch": 2.8705583756345177, + "grad_norm": 9.5, + "learning_rate": 2.0243420761387537e-06, + "loss": 2.6937, + "step": 27144 + }, + { + "epoch": 2.8706641285956005, + "grad_norm": 8.6875, + "learning_rate": 2.024173949348342e-06, + "loss": 2.5672, + "step": 27145 + }, + { + "epoch": 2.870769881556684, + "grad_norm": 8.375, + "learning_rate": 2.024005824790786e-06, + "loss": 2.5112, + "step": 27146 + }, + { + "epoch": 2.8708756345177666, + "grad_norm": 9.4375, + "learning_rate": 2.0238377024668767e-06, + "loss": 2.6448, + "step": 27147 + }, + { + "epoch": 2.8709813874788495, + "grad_norm": 8.375, + "learning_rate": 2.0236695823774017e-06, + "loss": 2.5561, + "step": 27148 + }, + { + "epoch": 2.8710871404399323, + "grad_norm": 8.875, + "learning_rate": 2.0235014645231504e-06, + "loss": 2.6858, + "step": 27149 + }, + { + "epoch": 2.871192893401015, + "grad_norm": 8.3125, + "learning_rate": 2.0233333489049124e-06, + "loss": 2.5542, + "step": 27150 + }, + { + "epoch": 2.871298646362098, + "grad_norm": 9.0625, + "learning_rate": 2.0231652355234754e-06, + "loss": 2.2555, + "step": 27151 + }, + { + "epoch": 2.871404399323181, + "grad_norm": 7.84375, + "learning_rate": 2.0229971243796305e-06, + "loss": 2.3605, + "step": 27152 + }, + { + "epoch": 2.871510152284264, + "grad_norm": 8.75, + "learning_rate": 2.022829015474163e-06, + "loss": 2.5753, + "step": 27153 + }, + { + "epoch": 2.871615905245347, + "grad_norm": 8.125, + "learning_rate": 2.022660908807865e-06, + "loss": 2.4653, + "step": 27154 + }, + { + "epoch": 2.8717216582064298, + "grad_norm": 8.1875, + "learning_rate": 2.022492804381524e-06, + "loss": 2.4203, + "step": 27155 + }, + { + "epoch": 2.8718274111675126, + "grad_norm": 8.4375, + "learning_rate": 2.022324702195929e-06, + "loss": 2.5, + "step": 27156 + }, + { + "epoch": 2.8719331641285955, + "grad_norm": 8.8125, + "learning_rate": 2.0221566022518684e-06, + "loss": 2.5406, + "step": 27157 + }, + { + "epoch": 2.8720389170896787, + "grad_norm": 8.8125, + "learning_rate": 2.021988504550131e-06, + "loss": 2.3728, + "step": 27158 + }, + { + "epoch": 2.8721446700507616, + "grad_norm": 8.5625, + "learning_rate": 2.021820409091507e-06, + "loss": 2.2589, + "step": 27159 + }, + { + "epoch": 2.8722504230118444, + "grad_norm": 7.96875, + "learning_rate": 2.0216523158767837e-06, + "loss": 2.4106, + "step": 27160 + }, + { + "epoch": 2.8723561759729273, + "grad_norm": 8.875, + "learning_rate": 2.0214842249067513e-06, + "loss": 2.5825, + "step": 27161 + }, + { + "epoch": 2.87246192893401, + "grad_norm": 8.3125, + "learning_rate": 2.021316136182197e-06, + "loss": 2.3837, + "step": 27162 + }, + { + "epoch": 2.872567681895093, + "grad_norm": 8.625, + "learning_rate": 2.0211480497039108e-06, + "loss": 2.7532, + "step": 27163 + }, + { + "epoch": 2.8726734348561758, + "grad_norm": 8.0625, + "learning_rate": 2.0209799654726802e-06, + "loss": 2.5449, + "step": 27164 + }, + { + "epoch": 2.872779187817259, + "grad_norm": 8.5625, + "learning_rate": 2.020811883489295e-06, + "loss": 2.6011, + "step": 27165 + }, + { + "epoch": 2.872884940778342, + "grad_norm": 8.25, + "learning_rate": 2.0206438037545437e-06, + "loss": 2.325, + "step": 27166 + }, + { + "epoch": 2.8729906937394247, + "grad_norm": 8.1875, + "learning_rate": 2.0204757262692144e-06, + "loss": 2.3554, + "step": 27167 + }, + { + "epoch": 2.8730964467005076, + "grad_norm": 8.0625, + "learning_rate": 2.020307651034098e-06, + "loss": 2.4208, + "step": 27168 + }, + { + "epoch": 2.8732021996615904, + "grad_norm": 8.5, + "learning_rate": 2.0201395780499797e-06, + "loss": 2.4136, + "step": 27169 + }, + { + "epoch": 2.8733079526226737, + "grad_norm": 8.375, + "learning_rate": 2.019971507317652e-06, + "loss": 2.4661, + "step": 27170 + }, + { + "epoch": 2.873413705583756, + "grad_norm": 8.5, + "learning_rate": 2.0198034388379003e-06, + "loss": 2.1122, + "step": 27171 + }, + { + "epoch": 2.8735194585448394, + "grad_norm": 8.4375, + "learning_rate": 2.019635372611515e-06, + "loss": 2.6279, + "step": 27172 + }, + { + "epoch": 2.873625211505922, + "grad_norm": 8.75, + "learning_rate": 2.0194673086392845e-06, + "loss": 2.734, + "step": 27173 + }, + { + "epoch": 2.873730964467005, + "grad_norm": 8.0, + "learning_rate": 2.0192992469219982e-06, + "loss": 2.5884, + "step": 27174 + }, + { + "epoch": 2.873836717428088, + "grad_norm": 8.6875, + "learning_rate": 2.019131187460442e-06, + "loss": 2.5628, + "step": 27175 + }, + { + "epoch": 2.8739424703891707, + "grad_norm": 8.625, + "learning_rate": 2.0189631302554075e-06, + "loss": 2.6562, + "step": 27176 + }, + { + "epoch": 2.874048223350254, + "grad_norm": 8.3125, + "learning_rate": 2.018795075307683e-06, + "loss": 2.6676, + "step": 27177 + }, + { + "epoch": 2.874153976311337, + "grad_norm": 8.5625, + "learning_rate": 2.018627022618056e-06, + "loss": 2.6416, + "step": 27178 + }, + { + "epoch": 2.8742597292724197, + "grad_norm": 9.0625, + "learning_rate": 2.018458972187315e-06, + "loss": 2.6672, + "step": 27179 + }, + { + "epoch": 2.8743654822335025, + "grad_norm": 8.25, + "learning_rate": 2.0182909240162494e-06, + "loss": 2.714, + "step": 27180 + }, + { + "epoch": 2.8744712351945854, + "grad_norm": 8.625, + "learning_rate": 2.018122878105648e-06, + "loss": 2.4983, + "step": 27181 + }, + { + "epoch": 2.874576988155668, + "grad_norm": 8.875, + "learning_rate": 2.0179548344562978e-06, + "loss": 2.5018, + "step": 27182 + }, + { + "epoch": 2.874682741116751, + "grad_norm": 8.3125, + "learning_rate": 2.01778679306899e-06, + "loss": 2.6729, + "step": 27183 + }, + { + "epoch": 2.8747884940778343, + "grad_norm": 8.125, + "learning_rate": 2.01761875394451e-06, + "loss": 2.4855, + "step": 27184 + }, + { + "epoch": 2.874894247038917, + "grad_norm": 8.1875, + "learning_rate": 2.0174507170836475e-06, + "loss": 2.424, + "step": 27185 + }, + { + "epoch": 2.875, + "grad_norm": 8.125, + "learning_rate": 2.017282682487193e-06, + "loss": 2.4568, + "step": 27186 + }, + { + "epoch": 2.875105752961083, + "grad_norm": 8.0, + "learning_rate": 2.0171146501559325e-06, + "loss": 2.4785, + "step": 27187 + }, + { + "epoch": 2.8752115059221657, + "grad_norm": 8.375, + "learning_rate": 2.016946620090656e-06, + "loss": 2.3829, + "step": 27188 + }, + { + "epoch": 2.875317258883249, + "grad_norm": 8.875, + "learning_rate": 2.016778592292151e-06, + "loss": 2.6812, + "step": 27189 + }, + { + "epoch": 2.875423011844332, + "grad_norm": 8.125, + "learning_rate": 2.0166105667612072e-06, + "loss": 2.4548, + "step": 27190 + }, + { + "epoch": 2.8755287648054146, + "grad_norm": 7.96875, + "learning_rate": 2.016442543498611e-06, + "loss": 2.3731, + "step": 27191 + }, + { + "epoch": 2.8756345177664975, + "grad_norm": 8.4375, + "learning_rate": 2.0162745225051537e-06, + "loss": 2.4349, + "step": 27192 + }, + { + "epoch": 2.8757402707275803, + "grad_norm": 8.4375, + "learning_rate": 2.0161065037816213e-06, + "loss": 2.478, + "step": 27193 + }, + { + "epoch": 2.875846023688663, + "grad_norm": 8.5, + "learning_rate": 2.015938487328803e-06, + "loss": 2.7007, + "step": 27194 + }, + { + "epoch": 2.875951776649746, + "grad_norm": 8.375, + "learning_rate": 2.0157704731474877e-06, + "loss": 2.6404, + "step": 27195 + }, + { + "epoch": 2.8760575296108293, + "grad_norm": 8.1875, + "learning_rate": 2.0156024612384632e-06, + "loss": 2.5217, + "step": 27196 + }, + { + "epoch": 2.876163282571912, + "grad_norm": 9.6875, + "learning_rate": 2.015434451602519e-06, + "loss": 2.7239, + "step": 27197 + }, + { + "epoch": 2.876269035532995, + "grad_norm": 8.125, + "learning_rate": 2.015266444240442e-06, + "loss": 2.5294, + "step": 27198 + }, + { + "epoch": 2.876374788494078, + "grad_norm": 8.125, + "learning_rate": 2.0150984391530224e-06, + "loss": 2.6151, + "step": 27199 + }, + { + "epoch": 2.8764805414551606, + "grad_norm": 8.6875, + "learning_rate": 2.0149304363410466e-06, + "loss": 2.2666, + "step": 27200 + }, + { + "epoch": 2.876586294416244, + "grad_norm": 8.125, + "learning_rate": 2.0147624358053043e-06, + "loss": 2.5382, + "step": 27201 + }, + { + "epoch": 2.8766920473773263, + "grad_norm": 8.0625, + "learning_rate": 2.0145944375465825e-06, + "loss": 2.4419, + "step": 27202 + }, + { + "epoch": 2.8767978003384096, + "grad_norm": 7.78125, + "learning_rate": 2.0144264415656713e-06, + "loss": 2.4006, + "step": 27203 + }, + { + "epoch": 2.8769035532994924, + "grad_norm": 8.3125, + "learning_rate": 2.0142584478633583e-06, + "loss": 2.5909, + "step": 27204 + }, + { + "epoch": 2.8770093062605753, + "grad_norm": 9.0, + "learning_rate": 2.014090456440431e-06, + "loss": 2.5152, + "step": 27205 + }, + { + "epoch": 2.877115059221658, + "grad_norm": 9.1875, + "learning_rate": 2.0139224672976797e-06, + "loss": 2.722, + "step": 27206 + }, + { + "epoch": 2.877220812182741, + "grad_norm": 8.25, + "learning_rate": 2.01375448043589e-06, + "loss": 2.612, + "step": 27207 + }, + { + "epoch": 2.8773265651438242, + "grad_norm": 8.75, + "learning_rate": 2.013586495855853e-06, + "loss": 2.331, + "step": 27208 + }, + { + "epoch": 2.877432318104907, + "grad_norm": 7.9375, + "learning_rate": 2.0134185135583546e-06, + "loss": 2.4995, + "step": 27209 + }, + { + "epoch": 2.87753807106599, + "grad_norm": 8.4375, + "learning_rate": 2.0132505335441847e-06, + "loss": 2.3667, + "step": 27210 + }, + { + "epoch": 2.8776438240270727, + "grad_norm": 8.0, + "learning_rate": 2.0130825558141306e-06, + "loss": 2.5443, + "step": 27211 + }, + { + "epoch": 2.8777495769881556, + "grad_norm": 9.1875, + "learning_rate": 2.012914580368981e-06, + "loss": 2.1393, + "step": 27212 + }, + { + "epoch": 2.8778553299492384, + "grad_norm": 8.625, + "learning_rate": 2.012746607209524e-06, + "loss": 2.5858, + "step": 27213 + }, + { + "epoch": 2.8779610829103213, + "grad_norm": 8.125, + "learning_rate": 2.0125786363365482e-06, + "loss": 2.3769, + "step": 27214 + }, + { + "epoch": 2.8780668358714045, + "grad_norm": 8.25, + "learning_rate": 2.012410667750842e-06, + "loss": 2.6471, + "step": 27215 + }, + { + "epoch": 2.8781725888324874, + "grad_norm": 8.9375, + "learning_rate": 2.0122427014531925e-06, + "loss": 2.3614, + "step": 27216 + }, + { + "epoch": 2.8782783417935702, + "grad_norm": 8.375, + "learning_rate": 2.0120747374443886e-06, + "loss": 2.5664, + "step": 27217 + }, + { + "epoch": 2.878384094754653, + "grad_norm": 8.3125, + "learning_rate": 2.011906775725218e-06, + "loss": 2.3864, + "step": 27218 + }, + { + "epoch": 2.878489847715736, + "grad_norm": 9.125, + "learning_rate": 2.01173881629647e-06, + "loss": 2.654, + "step": 27219 + }, + { + "epoch": 2.878595600676819, + "grad_norm": 8.3125, + "learning_rate": 2.011570859158931e-06, + "loss": 2.5842, + "step": 27220 + }, + { + "epoch": 2.878701353637902, + "grad_norm": 8.5, + "learning_rate": 2.011402904313391e-06, + "loss": 2.419, + "step": 27221 + }, + { + "epoch": 2.878807106598985, + "grad_norm": 8.8125, + "learning_rate": 2.011234951760638e-06, + "loss": 2.6154, + "step": 27222 + }, + { + "epoch": 2.8789128595600677, + "grad_norm": 8.375, + "learning_rate": 2.0110670015014576e-06, + "loss": 2.4615, + "step": 27223 + }, + { + "epoch": 2.8790186125211505, + "grad_norm": 8.375, + "learning_rate": 2.0108990535366415e-06, + "loss": 2.3091, + "step": 27224 + }, + { + "epoch": 2.8791243654822334, + "grad_norm": 8.6875, + "learning_rate": 2.0107311078669754e-06, + "loss": 2.6619, + "step": 27225 + }, + { + "epoch": 2.879230118443316, + "grad_norm": 8.625, + "learning_rate": 2.010563164493248e-06, + "loss": 2.5104, + "step": 27226 + }, + { + "epoch": 2.8793358714043995, + "grad_norm": 8.5625, + "learning_rate": 2.0103952234162476e-06, + "loss": 2.3717, + "step": 27227 + }, + { + "epoch": 2.8794416243654823, + "grad_norm": 8.5, + "learning_rate": 2.010227284636763e-06, + "loss": 2.5792, + "step": 27228 + }, + { + "epoch": 2.879547377326565, + "grad_norm": 8.3125, + "learning_rate": 2.01005934815558e-06, + "loss": 2.627, + "step": 27229 + }, + { + "epoch": 2.879653130287648, + "grad_norm": 9.0, + "learning_rate": 2.0098914139734883e-06, + "loss": 2.7144, + "step": 27230 + }, + { + "epoch": 2.879758883248731, + "grad_norm": 8.5, + "learning_rate": 2.009723482091277e-06, + "loss": 2.3782, + "step": 27231 + }, + { + "epoch": 2.879864636209814, + "grad_norm": 8.4375, + "learning_rate": 2.0095555525097316e-06, + "loss": 2.5351, + "step": 27232 + }, + { + "epoch": 2.8799703891708965, + "grad_norm": 7.9375, + "learning_rate": 2.0093876252296426e-06, + "loss": 2.4882, + "step": 27233 + }, + { + "epoch": 2.88007614213198, + "grad_norm": 8.875, + "learning_rate": 2.009219700251796e-06, + "loss": 2.5028, + "step": 27234 + }, + { + "epoch": 2.8801818950930627, + "grad_norm": 8.0625, + "learning_rate": 2.009051777576981e-06, + "loss": 2.5591, + "step": 27235 + }, + { + "epoch": 2.8802876480541455, + "grad_norm": 8.25, + "learning_rate": 2.008883857205984e-06, + "loss": 2.4067, + "step": 27236 + }, + { + "epoch": 2.8803934010152283, + "grad_norm": 8.875, + "learning_rate": 2.0087159391395964e-06, + "loss": 2.692, + "step": 27237 + }, + { + "epoch": 2.880499153976311, + "grad_norm": 7.90625, + "learning_rate": 2.0085480233786023e-06, + "loss": 2.5782, + "step": 27238 + }, + { + "epoch": 2.8806049069373945, + "grad_norm": 8.5, + "learning_rate": 2.008380109923791e-06, + "loss": 2.5232, + "step": 27239 + }, + { + "epoch": 2.8807106598984773, + "grad_norm": 8.4375, + "learning_rate": 2.008212198775952e-06, + "loss": 2.576, + "step": 27240 + }, + { + "epoch": 2.88081641285956, + "grad_norm": 7.96875, + "learning_rate": 2.0080442899358716e-06, + "loss": 2.2612, + "step": 27241 + }, + { + "epoch": 2.880922165820643, + "grad_norm": 8.1875, + "learning_rate": 2.0078763834043382e-06, + "loss": 2.5395, + "step": 27242 + }, + { + "epoch": 2.881027918781726, + "grad_norm": 8.875, + "learning_rate": 2.0077084791821394e-06, + "loss": 2.5462, + "step": 27243 + }, + { + "epoch": 2.8811336717428087, + "grad_norm": 7.75, + "learning_rate": 2.007540577270064e-06, + "loss": 2.5064, + "step": 27244 + }, + { + "epoch": 2.8812394247038915, + "grad_norm": 8.5, + "learning_rate": 2.0073726776688985e-06, + "loss": 2.6499, + "step": 27245 + }, + { + "epoch": 2.8813451776649748, + "grad_norm": 8.6875, + "learning_rate": 2.0072047803794326e-06, + "loss": 2.6045, + "step": 27246 + }, + { + "epoch": 2.8814509306260576, + "grad_norm": 8.5625, + "learning_rate": 2.007036885402452e-06, + "loss": 2.4511, + "step": 27247 + }, + { + "epoch": 2.8815566835871405, + "grad_norm": 8.8125, + "learning_rate": 2.0068689927387457e-06, + "loss": 2.8339, + "step": 27248 + }, + { + "epoch": 2.8816624365482233, + "grad_norm": 8.5625, + "learning_rate": 2.006701102389102e-06, + "loss": 2.6154, + "step": 27249 + }, + { + "epoch": 2.881768189509306, + "grad_norm": 8.625, + "learning_rate": 2.006533214354308e-06, + "loss": 2.4346, + "step": 27250 + }, + { + "epoch": 2.8818739424703894, + "grad_norm": 8.6875, + "learning_rate": 2.0063653286351523e-06, + "loss": 2.506, + "step": 27251 + }, + { + "epoch": 2.881979695431472, + "grad_norm": 8.0, + "learning_rate": 2.0061974452324215e-06, + "loss": 2.3521, + "step": 27252 + }, + { + "epoch": 2.882085448392555, + "grad_norm": 8.0625, + "learning_rate": 2.006029564146905e-06, + "loss": 2.4882, + "step": 27253 + }, + { + "epoch": 2.882191201353638, + "grad_norm": 8.5, + "learning_rate": 2.005861685379389e-06, + "loss": 2.5958, + "step": 27254 + }, + { + "epoch": 2.8822969543147208, + "grad_norm": 8.3125, + "learning_rate": 2.0056938089306627e-06, + "loss": 2.289, + "step": 27255 + }, + { + "epoch": 2.8824027072758036, + "grad_norm": 9.0625, + "learning_rate": 2.0055259348015125e-06, + "loss": 2.6811, + "step": 27256 + }, + { + "epoch": 2.8825084602368864, + "grad_norm": 8.5625, + "learning_rate": 2.005358062992727e-06, + "loss": 2.639, + "step": 27257 + }, + { + "epoch": 2.8826142131979697, + "grad_norm": 8.75, + "learning_rate": 2.005190193505094e-06, + "loss": 2.7785, + "step": 27258 + }, + { + "epoch": 2.8827199661590526, + "grad_norm": 8.8125, + "learning_rate": 2.0050223263394008e-06, + "loss": 2.7136, + "step": 27259 + }, + { + "epoch": 2.8828257191201354, + "grad_norm": 9.375, + "learning_rate": 2.004854461496436e-06, + "loss": 2.268, + "step": 27260 + }, + { + "epoch": 2.8829314720812182, + "grad_norm": 8.125, + "learning_rate": 2.0046865989769855e-06, + "loss": 2.5775, + "step": 27261 + }, + { + "epoch": 2.883037225042301, + "grad_norm": 8.1875, + "learning_rate": 2.00451873878184e-06, + "loss": 2.4656, + "step": 27262 + }, + { + "epoch": 2.8831429780033844, + "grad_norm": 8.25, + "learning_rate": 2.004350880911784e-06, + "loss": 2.3131, + "step": 27263 + }, + { + "epoch": 2.8832487309644668, + "grad_norm": 8.625, + "learning_rate": 2.0041830253676063e-06, + "loss": 2.5683, + "step": 27264 + }, + { + "epoch": 2.88335448392555, + "grad_norm": 8.125, + "learning_rate": 2.004015172150096e-06, + "loss": 2.6516, + "step": 27265 + }, + { + "epoch": 2.883460236886633, + "grad_norm": 8.25, + "learning_rate": 2.003847321260039e-06, + "loss": 2.4377, + "step": 27266 + }, + { + "epoch": 2.8835659898477157, + "grad_norm": 8.25, + "learning_rate": 2.003679472698224e-06, + "loss": 2.4286, + "step": 27267 + }, + { + "epoch": 2.8836717428087986, + "grad_norm": 8.625, + "learning_rate": 2.003511626465438e-06, + "loss": 2.6436, + "step": 27268 + }, + { + "epoch": 2.8837774957698814, + "grad_norm": 8.125, + "learning_rate": 2.00334378256247e-06, + "loss": 2.4935, + "step": 27269 + }, + { + "epoch": 2.8838832487309647, + "grad_norm": 8.3125, + "learning_rate": 2.0031759409901046e-06, + "loss": 2.5203, + "step": 27270 + }, + { + "epoch": 2.8839890016920475, + "grad_norm": 8.375, + "learning_rate": 2.0030081017491326e-06, + "loss": 2.6475, + "step": 27271 + }, + { + "epoch": 2.8840947546531304, + "grad_norm": 9.0, + "learning_rate": 2.0028402648403393e-06, + "loss": 2.6763, + "step": 27272 + }, + { + "epoch": 2.884200507614213, + "grad_norm": 8.1875, + "learning_rate": 2.0026724302645136e-06, + "loss": 2.4447, + "step": 27273 + }, + { + "epoch": 2.884306260575296, + "grad_norm": 8.125, + "learning_rate": 2.0025045980224432e-06, + "loss": 2.731, + "step": 27274 + }, + { + "epoch": 2.884412013536379, + "grad_norm": 8.5625, + "learning_rate": 2.002336768114915e-06, + "loss": 2.7023, + "step": 27275 + }, + { + "epoch": 2.8845177664974617, + "grad_norm": 8.375, + "learning_rate": 2.0021689405427178e-06, + "loss": 2.5811, + "step": 27276 + }, + { + "epoch": 2.884623519458545, + "grad_norm": 8.3125, + "learning_rate": 2.0020011153066365e-06, + "loss": 2.3077, + "step": 27277 + }, + { + "epoch": 2.884729272419628, + "grad_norm": 8.4375, + "learning_rate": 2.001833292407462e-06, + "loss": 2.4473, + "step": 27278 + }, + { + "epoch": 2.8848350253807107, + "grad_norm": 8.4375, + "learning_rate": 2.001665471845979e-06, + "loss": 2.4643, + "step": 27279 + }, + { + "epoch": 2.8849407783417935, + "grad_norm": 9.25, + "learning_rate": 2.0014976536229763e-06, + "loss": 2.8447, + "step": 27280 + }, + { + "epoch": 2.8850465313028764, + "grad_norm": 8.625, + "learning_rate": 2.001329837739241e-06, + "loss": 2.4391, + "step": 27281 + }, + { + "epoch": 2.8851522842639596, + "grad_norm": 8.25, + "learning_rate": 2.0011620241955607e-06, + "loss": 2.3651, + "step": 27282 + }, + { + "epoch": 2.885258037225042, + "grad_norm": 8.5, + "learning_rate": 2.0009942129927236e-06, + "loss": 2.6261, + "step": 27283 + }, + { + "epoch": 2.8853637901861253, + "grad_norm": 8.5, + "learning_rate": 2.0008264041315163e-06, + "loss": 2.6073, + "step": 27284 + }, + { + "epoch": 2.885469543147208, + "grad_norm": 8.6875, + "learning_rate": 2.000658597612727e-06, + "loss": 2.613, + "step": 27285 + }, + { + "epoch": 2.885575296108291, + "grad_norm": 8.8125, + "learning_rate": 2.000490793437142e-06, + "loss": 2.9239, + "step": 27286 + }, + { + "epoch": 2.885681049069374, + "grad_norm": 8.5, + "learning_rate": 2.00032299160555e-06, + "loss": 2.6307, + "step": 27287 + }, + { + "epoch": 2.8857868020304567, + "grad_norm": 8.1875, + "learning_rate": 2.0001551921187375e-06, + "loss": 2.4898, + "step": 27288 + }, + { + "epoch": 2.88589255499154, + "grad_norm": 8.5, + "learning_rate": 1.9999873949774922e-06, + "loss": 2.4006, + "step": 27289 + }, + { + "epoch": 2.885998307952623, + "grad_norm": 8.375, + "learning_rate": 1.9998196001826015e-06, + "loss": 2.5717, + "step": 27290 + }, + { + "epoch": 2.8861040609137056, + "grad_norm": 8.375, + "learning_rate": 1.9996518077348526e-06, + "loss": 2.7486, + "step": 27291 + }, + { + "epoch": 2.8862098138747885, + "grad_norm": 8.5, + "learning_rate": 1.999484017635034e-06, + "loss": 2.3785, + "step": 27292 + }, + { + "epoch": 2.8863155668358713, + "grad_norm": 8.3125, + "learning_rate": 1.999316229883931e-06, + "loss": 2.513, + "step": 27293 + }, + { + "epoch": 2.886421319796954, + "grad_norm": 8.4375, + "learning_rate": 1.9991484444823336e-06, + "loss": 2.4918, + "step": 27294 + }, + { + "epoch": 2.886527072758037, + "grad_norm": 8.875, + "learning_rate": 1.9989806614310266e-06, + "loss": 2.5714, + "step": 27295 + }, + { + "epoch": 2.8866328257191203, + "grad_norm": 8.625, + "learning_rate": 1.9988128807307992e-06, + "loss": 2.5668, + "step": 27296 + }, + { + "epoch": 2.886738578680203, + "grad_norm": 8.5625, + "learning_rate": 1.9986451023824372e-06, + "loss": 2.6218, + "step": 27297 + }, + { + "epoch": 2.886844331641286, + "grad_norm": 8.0625, + "learning_rate": 1.99847732638673e-06, + "loss": 2.3894, + "step": 27298 + }, + { + "epoch": 2.886950084602369, + "grad_norm": 8.0, + "learning_rate": 1.998309552744462e-06, + "loss": 2.4826, + "step": 27299 + }, + { + "epoch": 2.8870558375634516, + "grad_norm": 8.625, + "learning_rate": 1.9981417814564222e-06, + "loss": 2.5518, + "step": 27300 + }, + { + "epoch": 2.887161590524535, + "grad_norm": 8.875, + "learning_rate": 1.9979740125233994e-06, + "loss": 2.6088, + "step": 27301 + }, + { + "epoch": 2.8872673434856178, + "grad_norm": 9.125, + "learning_rate": 1.9978062459461776e-06, + "loss": 2.2604, + "step": 27302 + }, + { + "epoch": 2.8873730964467006, + "grad_norm": 8.25, + "learning_rate": 1.9976384817255463e-06, + "loss": 2.5375, + "step": 27303 + }, + { + "epoch": 2.8874788494077834, + "grad_norm": 8.3125, + "learning_rate": 1.997470719862292e-06, + "loss": 2.6517, + "step": 27304 + }, + { + "epoch": 2.8875846023688663, + "grad_norm": 8.5625, + "learning_rate": 1.9973029603572023e-06, + "loss": 2.6809, + "step": 27305 + }, + { + "epoch": 2.887690355329949, + "grad_norm": 8.75, + "learning_rate": 1.997135203211064e-06, + "loss": 2.6583, + "step": 27306 + }, + { + "epoch": 2.887796108291032, + "grad_norm": 8.875, + "learning_rate": 1.9969674484246652e-06, + "loss": 2.5161, + "step": 27307 + }, + { + "epoch": 2.8879018612521152, + "grad_norm": 8.1875, + "learning_rate": 1.9967996959987915e-06, + "loss": 2.5315, + "step": 27308 + }, + { + "epoch": 2.888007614213198, + "grad_norm": 8.375, + "learning_rate": 1.9966319459342306e-06, + "loss": 2.47, + "step": 27309 + }, + { + "epoch": 2.888113367174281, + "grad_norm": 8.125, + "learning_rate": 1.9964641982317717e-06, + "loss": 2.5272, + "step": 27310 + }, + { + "epoch": 2.8882191201353637, + "grad_norm": 8.25, + "learning_rate": 1.9962964528921994e-06, + "loss": 2.542, + "step": 27311 + }, + { + "epoch": 2.8883248730964466, + "grad_norm": 8.75, + "learning_rate": 1.9961287099163026e-06, + "loss": 2.6586, + "step": 27312 + }, + { + "epoch": 2.88843062605753, + "grad_norm": 8.3125, + "learning_rate": 1.9959609693048664e-06, + "loss": 2.6398, + "step": 27313 + }, + { + "epoch": 2.8885363790186123, + "grad_norm": 8.6875, + "learning_rate": 1.995793231058681e-06, + "loss": 2.7958, + "step": 27314 + }, + { + "epoch": 2.8886421319796955, + "grad_norm": 8.1875, + "learning_rate": 1.9956254951785294e-06, + "loss": 2.4506, + "step": 27315 + }, + { + "epoch": 2.8887478849407784, + "grad_norm": 8.0625, + "learning_rate": 1.995457761665203e-06, + "loss": 2.5736, + "step": 27316 + }, + { + "epoch": 2.8888536379018612, + "grad_norm": 8.875, + "learning_rate": 1.9952900305194864e-06, + "loss": 2.2846, + "step": 27317 + }, + { + "epoch": 2.888959390862944, + "grad_norm": 8.5, + "learning_rate": 1.9951223017421665e-06, + "loss": 2.2904, + "step": 27318 + }, + { + "epoch": 2.889065143824027, + "grad_norm": 8.25, + "learning_rate": 1.9949545753340326e-06, + "loss": 2.4053, + "step": 27319 + }, + { + "epoch": 2.88917089678511, + "grad_norm": 8.25, + "learning_rate": 1.994786851295869e-06, + "loss": 2.5716, + "step": 27320 + }, + { + "epoch": 2.889276649746193, + "grad_norm": 8.625, + "learning_rate": 1.9946191296284647e-06, + "loss": 2.5631, + "step": 27321 + }, + { + "epoch": 2.889382402707276, + "grad_norm": 8.4375, + "learning_rate": 1.994451410332606e-06, + "loss": 2.5757, + "step": 27322 + }, + { + "epoch": 2.8894881556683587, + "grad_norm": 8.5625, + "learning_rate": 1.994283693409081e-06, + "loss": 2.4551, + "step": 27323 + }, + { + "epoch": 2.8895939086294415, + "grad_norm": 8.375, + "learning_rate": 1.9941159788586744e-06, + "loss": 2.7076, + "step": 27324 + }, + { + "epoch": 2.8896996615905244, + "grad_norm": 8.3125, + "learning_rate": 1.9939482666821754e-06, + "loss": 2.5898, + "step": 27325 + }, + { + "epoch": 2.889805414551607, + "grad_norm": 8.1875, + "learning_rate": 1.9937805568803697e-06, + "loss": 2.589, + "step": 27326 + }, + { + "epoch": 2.8899111675126905, + "grad_norm": 8.8125, + "learning_rate": 1.9936128494540446e-06, + "loss": 2.7132, + "step": 27327 + }, + { + "epoch": 2.8900169204737733, + "grad_norm": 8.0625, + "learning_rate": 1.9934451444039882e-06, + "loss": 2.5335, + "step": 27328 + }, + { + "epoch": 2.890122673434856, + "grad_norm": 8.1875, + "learning_rate": 1.993277441730986e-06, + "loss": 2.4915, + "step": 27329 + }, + { + "epoch": 2.890228426395939, + "grad_norm": 8.5625, + "learning_rate": 1.9931097414358265e-06, + "loss": 2.6575, + "step": 27330 + }, + { + "epoch": 2.890334179357022, + "grad_norm": 8.3125, + "learning_rate": 1.9929420435192936e-06, + "loss": 2.6792, + "step": 27331 + }, + { + "epoch": 2.890439932318105, + "grad_norm": 8.5, + "learning_rate": 1.9927743479821783e-06, + "loss": 2.4516, + "step": 27332 + }, + { + "epoch": 2.890545685279188, + "grad_norm": 8.625, + "learning_rate": 1.9926066548252646e-06, + "loss": 2.6061, + "step": 27333 + }, + { + "epoch": 2.890651438240271, + "grad_norm": 8.375, + "learning_rate": 1.9924389640493406e-06, + "loss": 2.5431, + "step": 27334 + }, + { + "epoch": 2.8907571912013537, + "grad_norm": 8.75, + "learning_rate": 1.9922712756551927e-06, + "loss": 2.5006, + "step": 27335 + }, + { + "epoch": 2.8908629441624365, + "grad_norm": 8.0625, + "learning_rate": 1.992103589643608e-06, + "loss": 2.5337, + "step": 27336 + }, + { + "epoch": 2.8909686971235193, + "grad_norm": 8.6875, + "learning_rate": 1.991935906015374e-06, + "loss": 2.5853, + "step": 27337 + }, + { + "epoch": 2.891074450084602, + "grad_norm": 8.5, + "learning_rate": 1.991768224771276e-06, + "loss": 2.3953, + "step": 27338 + }, + { + "epoch": 2.8911802030456855, + "grad_norm": 8.125, + "learning_rate": 1.9916005459121036e-06, + "loss": 2.4572, + "step": 27339 + }, + { + "epoch": 2.8912859560067683, + "grad_norm": 8.6875, + "learning_rate": 1.9914328694386405e-06, + "loss": 2.3735, + "step": 27340 + }, + { + "epoch": 2.891391708967851, + "grad_norm": 8.0, + "learning_rate": 1.9912651953516753e-06, + "loss": 2.4159, + "step": 27341 + }, + { + "epoch": 2.891497461928934, + "grad_norm": 8.5, + "learning_rate": 1.9910975236519946e-06, + "loss": 2.4591, + "step": 27342 + }, + { + "epoch": 2.891603214890017, + "grad_norm": 8.625, + "learning_rate": 1.990929854340385e-06, + "loss": 2.3675, + "step": 27343 + }, + { + "epoch": 2.8917089678511, + "grad_norm": 8.3125, + "learning_rate": 1.990762187417633e-06, + "loss": 2.6524, + "step": 27344 + }, + { + "epoch": 2.8918147208121825, + "grad_norm": 8.25, + "learning_rate": 1.990594522884526e-06, + "loss": 2.653, + "step": 27345 + }, + { + "epoch": 2.8919204737732658, + "grad_norm": 8.1875, + "learning_rate": 1.990426860741851e-06, + "loss": 2.361, + "step": 27346 + }, + { + "epoch": 2.8920262267343486, + "grad_norm": 8.5, + "learning_rate": 1.9902592009903933e-06, + "loss": 2.7085, + "step": 27347 + }, + { + "epoch": 2.8921319796954315, + "grad_norm": 8.25, + "learning_rate": 1.9900915436309423e-06, + "loss": 2.4326, + "step": 27348 + }, + { + "epoch": 2.8922377326565143, + "grad_norm": 8.875, + "learning_rate": 1.989923888664282e-06, + "loss": 2.5929, + "step": 27349 + }, + { + "epoch": 2.892343485617597, + "grad_norm": 8.625, + "learning_rate": 1.9897562360912005e-06, + "loss": 2.7414, + "step": 27350 + }, + { + "epoch": 2.8924492385786804, + "grad_norm": 8.5, + "learning_rate": 1.989588585912484e-06, + "loss": 2.4787, + "step": 27351 + }, + { + "epoch": 2.8925549915397633, + "grad_norm": 8.4375, + "learning_rate": 1.989420938128921e-06, + "loss": 2.495, + "step": 27352 + }, + { + "epoch": 2.892660744500846, + "grad_norm": 8.25, + "learning_rate": 1.9892532927412946e-06, + "loss": 2.3685, + "step": 27353 + }, + { + "epoch": 2.892766497461929, + "grad_norm": 9.3125, + "learning_rate": 1.989085649750394e-06, + "loss": 2.4157, + "step": 27354 + }, + { + "epoch": 2.8928722504230118, + "grad_norm": 8.4375, + "learning_rate": 1.988918009157007e-06, + "loss": 2.4852, + "step": 27355 + }, + { + "epoch": 2.8929780033840946, + "grad_norm": 8.4375, + "learning_rate": 1.988750370961918e-06, + "loss": 2.7608, + "step": 27356 + }, + { + "epoch": 2.8930837563451774, + "grad_norm": 8.625, + "learning_rate": 1.988582735165914e-06, + "loss": 2.521, + "step": 27357 + }, + { + "epoch": 2.8931895093062607, + "grad_norm": 8.5, + "learning_rate": 1.9884151017697818e-06, + "loss": 2.3476, + "step": 27358 + }, + { + "epoch": 2.8932952622673436, + "grad_norm": 8.6875, + "learning_rate": 1.988247470774309e-06, + "loss": 2.514, + "step": 27359 + }, + { + "epoch": 2.8934010152284264, + "grad_norm": 8.875, + "learning_rate": 1.9880798421802812e-06, + "loss": 2.5539, + "step": 27360 + }, + { + "epoch": 2.8935067681895092, + "grad_norm": 8.6875, + "learning_rate": 1.987912215988486e-06, + "loss": 2.5894, + "step": 27361 + }, + { + "epoch": 2.893612521150592, + "grad_norm": 8.1875, + "learning_rate": 1.9877445921997088e-06, + "loss": 2.327, + "step": 27362 + }, + { + "epoch": 2.8937182741116754, + "grad_norm": 8.4375, + "learning_rate": 1.987576970814736e-06, + "loss": 2.5678, + "step": 27363 + }, + { + "epoch": 2.8938240270727578, + "grad_norm": 9.125, + "learning_rate": 1.987409351834356e-06, + "loss": 2.3365, + "step": 27364 + }, + { + "epoch": 2.893929780033841, + "grad_norm": 8.3125, + "learning_rate": 1.9872417352593535e-06, + "loss": 2.2158, + "step": 27365 + }, + { + "epoch": 2.894035532994924, + "grad_norm": 8.8125, + "learning_rate": 1.9870741210905165e-06, + "loss": 2.4496, + "step": 27366 + }, + { + "epoch": 2.8941412859560067, + "grad_norm": 8.3125, + "learning_rate": 1.9869065093286303e-06, + "loss": 2.6164, + "step": 27367 + }, + { + "epoch": 2.8942470389170896, + "grad_norm": 8.25, + "learning_rate": 1.986738899974483e-06, + "loss": 2.1242, + "step": 27368 + }, + { + "epoch": 2.8943527918781724, + "grad_norm": 8.5, + "learning_rate": 1.9865712930288585e-06, + "loss": 2.7575, + "step": 27369 + }, + { + "epoch": 2.8944585448392557, + "grad_norm": 8.3125, + "learning_rate": 1.986403688492547e-06, + "loss": 2.5012, + "step": 27370 + }, + { + "epoch": 2.8945642978003385, + "grad_norm": 8.5625, + "learning_rate": 1.986236086366331e-06, + "loss": 2.6517, + "step": 27371 + }, + { + "epoch": 2.8946700507614214, + "grad_norm": 8.5625, + "learning_rate": 1.9860684866509996e-06, + "loss": 2.6882, + "step": 27372 + }, + { + "epoch": 2.894775803722504, + "grad_norm": 7.84375, + "learning_rate": 1.985900889347339e-06, + "loss": 2.4813, + "step": 27373 + }, + { + "epoch": 2.894881556683587, + "grad_norm": 9.0, + "learning_rate": 1.9857332944561346e-06, + "loss": 2.5178, + "step": 27374 + }, + { + "epoch": 2.8949873096446703, + "grad_norm": 8.6875, + "learning_rate": 1.985565701978174e-06, + "loss": 2.4055, + "step": 27375 + }, + { + "epoch": 2.8950930626057527, + "grad_norm": 8.6875, + "learning_rate": 1.985398111914243e-06, + "loss": 2.353, + "step": 27376 + }, + { + "epoch": 2.895198815566836, + "grad_norm": 8.375, + "learning_rate": 1.9852305242651288e-06, + "loss": 2.5519, + "step": 27377 + }, + { + "epoch": 2.895304568527919, + "grad_norm": 8.1875, + "learning_rate": 1.9850629390316165e-06, + "loss": 2.5641, + "step": 27378 + }, + { + "epoch": 2.8954103214890017, + "grad_norm": 8.9375, + "learning_rate": 1.9848953562144924e-06, + "loss": 2.3434, + "step": 27379 + }, + { + "epoch": 2.8955160744500845, + "grad_norm": 7.96875, + "learning_rate": 1.9847277758145457e-06, + "loss": 2.176, + "step": 27380 + }, + { + "epoch": 2.8956218274111674, + "grad_norm": 9.0, + "learning_rate": 1.9845601978325592e-06, + "loss": 2.6326, + "step": 27381 + }, + { + "epoch": 2.8957275803722506, + "grad_norm": 12.375, + "learning_rate": 1.984392622269322e-06, + "loss": 2.7003, + "step": 27382 + }, + { + "epoch": 2.8958333333333335, + "grad_norm": 8.5, + "learning_rate": 1.9842250491256185e-06, + "loss": 2.6257, + "step": 27383 + }, + { + "epoch": 2.8959390862944163, + "grad_norm": 7.8125, + "learning_rate": 1.984057478402237e-06, + "loss": 2.3218, + "step": 27384 + }, + { + "epoch": 2.896044839255499, + "grad_norm": 8.5, + "learning_rate": 1.983889910099961e-06, + "loss": 2.3732, + "step": 27385 + }, + { + "epoch": 2.896150592216582, + "grad_norm": 8.1875, + "learning_rate": 1.9837223442195805e-06, + "loss": 2.6143, + "step": 27386 + }, + { + "epoch": 2.896256345177665, + "grad_norm": 8.1875, + "learning_rate": 1.9835547807618784e-06, + "loss": 2.4691, + "step": 27387 + }, + { + "epoch": 2.8963620981387477, + "grad_norm": 8.8125, + "learning_rate": 1.983387219727643e-06, + "loss": 2.5341, + "step": 27388 + }, + { + "epoch": 2.896467851099831, + "grad_norm": 8.125, + "learning_rate": 1.98321966111766e-06, + "loss": 2.4526, + "step": 27389 + }, + { + "epoch": 2.896573604060914, + "grad_norm": 8.8125, + "learning_rate": 1.983052104932716e-06, + "loss": 2.6628, + "step": 27390 + }, + { + "epoch": 2.8966793570219966, + "grad_norm": 8.5625, + "learning_rate": 1.982884551173597e-06, + "loss": 2.4729, + "step": 27391 + }, + { + "epoch": 2.8967851099830795, + "grad_norm": 8.625, + "learning_rate": 1.982716999841089e-06, + "loss": 2.4987, + "step": 27392 + }, + { + "epoch": 2.8968908629441623, + "grad_norm": 8.75, + "learning_rate": 1.9825494509359796e-06, + "loss": 2.591, + "step": 27393 + }, + { + "epoch": 2.8969966159052456, + "grad_norm": 8.625, + "learning_rate": 1.9823819044590528e-06, + "loss": 2.5417, + "step": 27394 + }, + { + "epoch": 2.897102368866328, + "grad_norm": 8.4375, + "learning_rate": 1.9822143604110967e-06, + "loss": 2.506, + "step": 27395 + }, + { + "epoch": 2.8972081218274113, + "grad_norm": 8.375, + "learning_rate": 1.9820468187928966e-06, + "loss": 2.5461, + "step": 27396 + }, + { + "epoch": 2.897313874788494, + "grad_norm": 8.8125, + "learning_rate": 1.9818792796052384e-06, + "loss": 2.5498, + "step": 27397 + }, + { + "epoch": 2.897419627749577, + "grad_norm": 8.5625, + "learning_rate": 1.9817117428489096e-06, + "loss": 2.4893, + "step": 27398 + }, + { + "epoch": 2.89752538071066, + "grad_norm": 9.125, + "learning_rate": 1.981544208524695e-06, + "loss": 2.6406, + "step": 27399 + }, + { + "epoch": 2.8976311336717426, + "grad_norm": 8.5625, + "learning_rate": 1.9813766766333823e-06, + "loss": 2.7811, + "step": 27400 + }, + { + "epoch": 2.897736886632826, + "grad_norm": 8.625, + "learning_rate": 1.9812091471757555e-06, + "loss": 2.499, + "step": 27401 + }, + { + "epoch": 2.8978426395939088, + "grad_norm": 7.6875, + "learning_rate": 1.9810416201526035e-06, + "loss": 2.2275, + "step": 27402 + }, + { + "epoch": 2.8979483925549916, + "grad_norm": 8.125, + "learning_rate": 1.9808740955647098e-06, + "loss": 2.5908, + "step": 27403 + }, + { + "epoch": 2.8980541455160744, + "grad_norm": 8.4375, + "learning_rate": 1.980706573412862e-06, + "loss": 2.4679, + "step": 27404 + }, + { + "epoch": 2.8981598984771573, + "grad_norm": 8.9375, + "learning_rate": 1.980539053697846e-06, + "loss": 2.5734, + "step": 27405 + }, + { + "epoch": 2.89826565143824, + "grad_norm": 8.5625, + "learning_rate": 1.9803715364204475e-06, + "loss": 2.4364, + "step": 27406 + }, + { + "epoch": 2.898371404399323, + "grad_norm": 8.625, + "learning_rate": 1.980204021581453e-06, + "loss": 2.576, + "step": 27407 + }, + { + "epoch": 2.8984771573604062, + "grad_norm": 8.125, + "learning_rate": 1.980036509181648e-06, + "loss": 2.6638, + "step": 27408 + }, + { + "epoch": 2.898582910321489, + "grad_norm": 8.3125, + "learning_rate": 1.9798689992218203e-06, + "loss": 2.638, + "step": 27409 + }, + { + "epoch": 2.898688663282572, + "grad_norm": 8.4375, + "learning_rate": 1.9797014917027535e-06, + "loss": 2.5909, + "step": 27410 + }, + { + "epoch": 2.8987944162436547, + "grad_norm": 8.625, + "learning_rate": 1.9795339866252354e-06, + "loss": 2.5077, + "step": 27411 + }, + { + "epoch": 2.8989001692047376, + "grad_norm": 8.6875, + "learning_rate": 1.9793664839900508e-06, + "loss": 2.3355, + "step": 27412 + }, + { + "epoch": 2.899005922165821, + "grad_norm": 8.0, + "learning_rate": 1.979198983797987e-06, + "loss": 2.382, + "step": 27413 + }, + { + "epoch": 2.8991116751269037, + "grad_norm": 8.75, + "learning_rate": 1.9790314860498285e-06, + "loss": 2.4428, + "step": 27414 + }, + { + "epoch": 2.8992174280879865, + "grad_norm": 8.3125, + "learning_rate": 1.9788639907463625e-06, + "loss": 2.6169, + "step": 27415 + }, + { + "epoch": 2.8993231810490694, + "grad_norm": 8.5, + "learning_rate": 1.9786964978883755e-06, + "loss": 2.416, + "step": 27416 + }, + { + "epoch": 2.8994289340101522, + "grad_norm": 8.5625, + "learning_rate": 1.9785290074766513e-06, + "loss": 2.5818, + "step": 27417 + }, + { + "epoch": 2.899534686971235, + "grad_norm": 8.375, + "learning_rate": 1.9783615195119787e-06, + "loss": 2.5352, + "step": 27418 + }, + { + "epoch": 2.899640439932318, + "grad_norm": 8.0625, + "learning_rate": 1.978194033995141e-06, + "loss": 2.36, + "step": 27419 + }, + { + "epoch": 2.899746192893401, + "grad_norm": 8.6875, + "learning_rate": 1.978026550926926e-06, + "loss": 2.5222, + "step": 27420 + }, + { + "epoch": 2.899851945854484, + "grad_norm": 8.125, + "learning_rate": 1.977859070308118e-06, + "loss": 2.4689, + "step": 27421 + }, + { + "epoch": 2.899957698815567, + "grad_norm": 8.875, + "learning_rate": 1.977691592139505e-06, + "loss": 2.5105, + "step": 27422 + }, + { + "epoch": 2.9000634517766497, + "grad_norm": 8.1875, + "learning_rate": 1.9775241164218706e-06, + "loss": 2.4364, + "step": 27423 + }, + { + "epoch": 2.9001692047377325, + "grad_norm": 8.6875, + "learning_rate": 1.977356643156002e-06, + "loss": 2.4957, + "step": 27424 + }, + { + "epoch": 2.900274957698816, + "grad_norm": 8.5625, + "learning_rate": 1.9771891723426862e-06, + "loss": 2.525, + "step": 27425 + }, + { + "epoch": 2.900380710659898, + "grad_norm": 8.6875, + "learning_rate": 1.9770217039827064e-06, + "loss": 2.6055, + "step": 27426 + }, + { + "epoch": 2.9004864636209815, + "grad_norm": 8.5, + "learning_rate": 1.9768542380768508e-06, + "loss": 2.7018, + "step": 27427 + }, + { + "epoch": 2.9005922165820643, + "grad_norm": 8.625, + "learning_rate": 1.9766867746259033e-06, + "loss": 2.5996, + "step": 27428 + }, + { + "epoch": 2.900697969543147, + "grad_norm": 8.6875, + "learning_rate": 1.9765193136306517e-06, + "loss": 2.5902, + "step": 27429 + }, + { + "epoch": 2.90080372250423, + "grad_norm": 8.0, + "learning_rate": 1.97635185509188e-06, + "loss": 2.5079, + "step": 27430 + }, + { + "epoch": 2.900909475465313, + "grad_norm": 8.75, + "learning_rate": 1.976184399010376e-06, + "loss": 2.4924, + "step": 27431 + }, + { + "epoch": 2.901015228426396, + "grad_norm": 8.125, + "learning_rate": 1.9760169453869236e-06, + "loss": 2.3152, + "step": 27432 + }, + { + "epoch": 2.901120981387479, + "grad_norm": 8.875, + "learning_rate": 1.9758494942223084e-06, + "loss": 2.5618, + "step": 27433 + }, + { + "epoch": 2.901226734348562, + "grad_norm": 8.5625, + "learning_rate": 1.9756820455173194e-06, + "loss": 2.5394, + "step": 27434 + }, + { + "epoch": 2.9013324873096447, + "grad_norm": 8.9375, + "learning_rate": 1.975514599272738e-06, + "loss": 2.62, + "step": 27435 + }, + { + "epoch": 2.9014382402707275, + "grad_norm": 8.625, + "learning_rate": 1.9753471554893534e-06, + "loss": 2.6361, + "step": 27436 + }, + { + "epoch": 2.9015439932318103, + "grad_norm": 8.0, + "learning_rate": 1.9751797141679497e-06, + "loss": 2.6157, + "step": 27437 + }, + { + "epoch": 2.901649746192893, + "grad_norm": 8.5, + "learning_rate": 1.9750122753093135e-06, + "loss": 2.3921, + "step": 27438 + }, + { + "epoch": 2.9017554991539765, + "grad_norm": 8.1875, + "learning_rate": 1.9748448389142284e-06, + "loss": 2.3402, + "step": 27439 + }, + { + "epoch": 2.9018612521150593, + "grad_norm": 8.625, + "learning_rate": 1.9746774049834837e-06, + "loss": 2.6547, + "step": 27440 + }, + { + "epoch": 2.901967005076142, + "grad_norm": 8.375, + "learning_rate": 1.9745099735178615e-06, + "loss": 2.5048, + "step": 27441 + }, + { + "epoch": 2.902072758037225, + "grad_norm": 8.0, + "learning_rate": 1.9743425445181496e-06, + "loss": 2.3399, + "step": 27442 + }, + { + "epoch": 2.902178510998308, + "grad_norm": 8.75, + "learning_rate": 1.974175117985133e-06, + "loss": 2.3531, + "step": 27443 + }, + { + "epoch": 2.902284263959391, + "grad_norm": 8.25, + "learning_rate": 1.9740076939195978e-06, + "loss": 2.5423, + "step": 27444 + }, + { + "epoch": 2.902390016920474, + "grad_norm": 8.75, + "learning_rate": 1.9738402723223295e-06, + "loss": 2.5023, + "step": 27445 + }, + { + "epoch": 2.9024957698815568, + "grad_norm": 8.25, + "learning_rate": 1.9736728531941128e-06, + "loss": 2.1614, + "step": 27446 + }, + { + "epoch": 2.9026015228426396, + "grad_norm": 7.84375, + "learning_rate": 1.9735054365357357e-06, + "loss": 2.3837, + "step": 27447 + }, + { + "epoch": 2.9027072758037225, + "grad_norm": 8.3125, + "learning_rate": 1.973338022347981e-06, + "loss": 2.2257, + "step": 27448 + }, + { + "epoch": 2.9028130287648053, + "grad_norm": 8.3125, + "learning_rate": 1.9731706106316358e-06, + "loss": 2.65, + "step": 27449 + }, + { + "epoch": 2.902918781725888, + "grad_norm": 8.3125, + "learning_rate": 1.9730032013874855e-06, + "loss": 2.4599, + "step": 27450 + }, + { + "epoch": 2.9030245346869714, + "grad_norm": 8.25, + "learning_rate": 1.972835794616315e-06, + "loss": 2.4515, + "step": 27451 + }, + { + "epoch": 2.9031302876480543, + "grad_norm": 8.6875, + "learning_rate": 1.972668390318912e-06, + "loss": 2.1779, + "step": 27452 + }, + { + "epoch": 2.903236040609137, + "grad_norm": 8.5, + "learning_rate": 1.9725009884960593e-06, + "loss": 2.3003, + "step": 27453 + }, + { + "epoch": 2.90334179357022, + "grad_norm": 8.875, + "learning_rate": 1.9723335891485445e-06, + "loss": 2.7243, + "step": 27454 + }, + { + "epoch": 2.9034475465313028, + "grad_norm": 8.0, + "learning_rate": 1.9721661922771514e-06, + "loss": 2.545, + "step": 27455 + }, + { + "epoch": 2.903553299492386, + "grad_norm": 8.3125, + "learning_rate": 1.9719987978826677e-06, + "loss": 2.5998, + "step": 27456 + }, + { + "epoch": 2.9036590524534684, + "grad_norm": 8.75, + "learning_rate": 1.971831405965877e-06, + "loss": 2.5732, + "step": 27457 + }, + { + "epoch": 2.9037648054145517, + "grad_norm": 8.0, + "learning_rate": 1.9716640165275656e-06, + "loss": 2.5669, + "step": 27458 + }, + { + "epoch": 2.9038705583756346, + "grad_norm": 8.625, + "learning_rate": 1.971496629568519e-06, + "loss": 2.355, + "step": 27459 + }, + { + "epoch": 2.9039763113367174, + "grad_norm": 8.125, + "learning_rate": 1.971329245089522e-06, + "loss": 2.5184, + "step": 27460 + }, + { + "epoch": 2.9040820642978002, + "grad_norm": 8.1875, + "learning_rate": 1.971161863091361e-06, + "loss": 2.4026, + "step": 27461 + }, + { + "epoch": 2.904187817258883, + "grad_norm": 8.625, + "learning_rate": 1.970994483574821e-06, + "loss": 2.5309, + "step": 27462 + }, + { + "epoch": 2.9042935702199664, + "grad_norm": 8.625, + "learning_rate": 1.9708271065406884e-06, + "loss": 2.5589, + "step": 27463 + }, + { + "epoch": 2.904399323181049, + "grad_norm": 8.5, + "learning_rate": 1.9706597319897467e-06, + "loss": 2.3805, + "step": 27464 + }, + { + "epoch": 2.904505076142132, + "grad_norm": 8.3125, + "learning_rate": 1.970492359922783e-06, + "loss": 2.1825, + "step": 27465 + }, + { + "epoch": 2.904610829103215, + "grad_norm": 8.8125, + "learning_rate": 1.9703249903405816e-06, + "loss": 2.5852, + "step": 27466 + }, + { + "epoch": 2.9047165820642977, + "grad_norm": 8.25, + "learning_rate": 1.970157623243929e-06, + "loss": 2.4113, + "step": 27467 + }, + { + "epoch": 2.9048223350253806, + "grad_norm": 8.625, + "learning_rate": 1.9699902586336092e-06, + "loss": 2.5476, + "step": 27468 + }, + { + "epoch": 2.9049280879864634, + "grad_norm": 8.0, + "learning_rate": 1.9698228965104087e-06, + "loss": 2.486, + "step": 27469 + }, + { + "epoch": 2.9050338409475467, + "grad_norm": 8.1875, + "learning_rate": 1.969655536875113e-06, + "loss": 2.4819, + "step": 27470 + }, + { + "epoch": 2.9051395939086295, + "grad_norm": 8.5625, + "learning_rate": 1.969488179728506e-06, + "loss": 2.4525, + "step": 27471 + }, + { + "epoch": 2.9052453468697124, + "grad_norm": 8.1875, + "learning_rate": 1.9693208250713752e-06, + "loss": 2.5754, + "step": 27472 + }, + { + "epoch": 2.905351099830795, + "grad_norm": 8.625, + "learning_rate": 1.9691534729045036e-06, + "loss": 2.6037, + "step": 27473 + }, + { + "epoch": 2.905456852791878, + "grad_norm": 8.625, + "learning_rate": 1.9689861232286784e-06, + "loss": 2.6544, + "step": 27474 + }, + { + "epoch": 2.9055626057529613, + "grad_norm": 8.4375, + "learning_rate": 1.9688187760446836e-06, + "loss": 2.5162, + "step": 27475 + }, + { + "epoch": 2.9056683587140437, + "grad_norm": 8.6875, + "learning_rate": 1.968651431353306e-06, + "loss": 2.3119, + "step": 27476 + }, + { + "epoch": 2.905774111675127, + "grad_norm": 8.6875, + "learning_rate": 1.9684840891553285e-06, + "loss": 2.5639, + "step": 27477 + }, + { + "epoch": 2.90587986463621, + "grad_norm": 8.25, + "learning_rate": 1.968316749451538e-06, + "loss": 2.5375, + "step": 27478 + }, + { + "epoch": 2.9059856175972927, + "grad_norm": 8.5625, + "learning_rate": 1.9681494122427206e-06, + "loss": 2.5322, + "step": 27479 + }, + { + "epoch": 2.9060913705583755, + "grad_norm": 8.3125, + "learning_rate": 1.96798207752966e-06, + "loss": 2.612, + "step": 27480 + }, + { + "epoch": 2.9061971235194584, + "grad_norm": 8.9375, + "learning_rate": 1.9678147453131418e-06, + "loss": 2.6197, + "step": 27481 + }, + { + "epoch": 2.9063028764805416, + "grad_norm": 8.375, + "learning_rate": 1.967647415593951e-06, + "loss": 2.504, + "step": 27482 + }, + { + "epoch": 2.9064086294416245, + "grad_norm": 8.9375, + "learning_rate": 1.967480088372874e-06, + "loss": 2.6609, + "step": 27483 + }, + { + "epoch": 2.9065143824027073, + "grad_norm": 8.875, + "learning_rate": 1.9673127636506945e-06, + "loss": 2.5384, + "step": 27484 + }, + { + "epoch": 2.90662013536379, + "grad_norm": 8.5, + "learning_rate": 1.9671454414281995e-06, + "loss": 2.5286, + "step": 27485 + }, + { + "epoch": 2.906725888324873, + "grad_norm": 8.5625, + "learning_rate": 1.9669781217061715e-06, + "loss": 2.616, + "step": 27486 + }, + { + "epoch": 2.9068316412859563, + "grad_norm": 8.0625, + "learning_rate": 1.9668108044853966e-06, + "loss": 2.3166, + "step": 27487 + }, + { + "epoch": 2.9069373942470387, + "grad_norm": 8.5, + "learning_rate": 1.9666434897666627e-06, + "loss": 2.2065, + "step": 27488 + }, + { + "epoch": 2.907043147208122, + "grad_norm": 8.5, + "learning_rate": 1.9664761775507514e-06, + "loss": 2.6026, + "step": 27489 + }, + { + "epoch": 2.907148900169205, + "grad_norm": 8.375, + "learning_rate": 1.9663088678384498e-06, + "loss": 2.5341, + "step": 27490 + }, + { + "epoch": 2.9072546531302876, + "grad_norm": 8.75, + "learning_rate": 1.9661415606305416e-06, + "loss": 2.5401, + "step": 27491 + }, + { + "epoch": 2.9073604060913705, + "grad_norm": 8.3125, + "learning_rate": 1.9659742559278143e-06, + "loss": 2.5221, + "step": 27492 + }, + { + "epoch": 2.9074661590524533, + "grad_norm": 8.25, + "learning_rate": 1.965806953731049e-06, + "loss": 2.4086, + "step": 27493 + }, + { + "epoch": 2.9075719120135366, + "grad_norm": 8.4375, + "learning_rate": 1.9656396540410344e-06, + "loss": 2.3918, + "step": 27494 + }, + { + "epoch": 2.9076776649746194, + "grad_norm": 8.4375, + "learning_rate": 1.9654723568585553e-06, + "loss": 2.4534, + "step": 27495 + }, + { + "epoch": 2.9077834179357023, + "grad_norm": 8.9375, + "learning_rate": 1.965305062184395e-06, + "loss": 2.7312, + "step": 27496 + }, + { + "epoch": 2.907889170896785, + "grad_norm": 8.25, + "learning_rate": 1.9651377700193395e-06, + "loss": 2.427, + "step": 27497 + }, + { + "epoch": 2.907994923857868, + "grad_norm": 8.6875, + "learning_rate": 1.9649704803641732e-06, + "loss": 2.7653, + "step": 27498 + }, + { + "epoch": 2.908100676818951, + "grad_norm": 8.5625, + "learning_rate": 1.9648031932196825e-06, + "loss": 2.5413, + "step": 27499 + }, + { + "epoch": 2.9082064297800336, + "grad_norm": 7.84375, + "learning_rate": 1.9646359085866504e-06, + "loss": 2.4583, + "step": 27500 + }, + { + "epoch": 2.908312182741117, + "grad_norm": 8.6875, + "learning_rate": 1.964468626465864e-06, + "loss": 2.6387, + "step": 27501 + }, + { + "epoch": 2.9084179357021998, + "grad_norm": 8.5625, + "learning_rate": 1.964301346858107e-06, + "loss": 2.6693, + "step": 27502 + }, + { + "epoch": 2.9085236886632826, + "grad_norm": 8.6875, + "learning_rate": 1.9641340697641645e-06, + "loss": 2.5474, + "step": 27503 + }, + { + "epoch": 2.9086294416243654, + "grad_norm": 8.25, + "learning_rate": 1.963966795184822e-06, + "loss": 2.4944, + "step": 27504 + }, + { + "epoch": 2.9087351945854483, + "grad_norm": 8.8125, + "learning_rate": 1.9637995231208634e-06, + "loss": 2.5112, + "step": 27505 + }, + { + "epoch": 2.9088409475465316, + "grad_norm": 9.25, + "learning_rate": 1.963632253573075e-06, + "loss": 2.7753, + "step": 27506 + }, + { + "epoch": 2.908946700507614, + "grad_norm": 8.375, + "learning_rate": 1.9634649865422406e-06, + "loss": 2.2024, + "step": 27507 + }, + { + "epoch": 2.9090524534686972, + "grad_norm": 8.25, + "learning_rate": 1.9632977220291467e-06, + "loss": 2.6106, + "step": 27508 + }, + { + "epoch": 2.90915820642978, + "grad_norm": 8.1875, + "learning_rate": 1.9631304600345753e-06, + "loss": 2.5268, + "step": 27509 + }, + { + "epoch": 2.909263959390863, + "grad_norm": 8.625, + "learning_rate": 1.962963200559315e-06, + "loss": 2.5379, + "step": 27510 + }, + { + "epoch": 2.9093697123519457, + "grad_norm": 8.1875, + "learning_rate": 1.9627959436041474e-06, + "loss": 2.4282, + "step": 27511 + }, + { + "epoch": 2.9094754653130286, + "grad_norm": 8.125, + "learning_rate": 1.9626286891698587e-06, + "loss": 2.1791, + "step": 27512 + }, + { + "epoch": 2.909581218274112, + "grad_norm": 9.0625, + "learning_rate": 1.9624614372572346e-06, + "loss": 2.7963, + "step": 27513 + }, + { + "epoch": 2.9096869712351947, + "grad_norm": 8.4375, + "learning_rate": 1.9622941878670585e-06, + "loss": 2.7237, + "step": 27514 + }, + { + "epoch": 2.9097927241962775, + "grad_norm": 8.5, + "learning_rate": 1.962126941000116e-06, + "loss": 2.1157, + "step": 27515 + }, + { + "epoch": 2.9098984771573604, + "grad_norm": 7.8125, + "learning_rate": 1.961959696657192e-06, + "loss": 2.4118, + "step": 27516 + }, + { + "epoch": 2.9100042301184432, + "grad_norm": 8.0625, + "learning_rate": 1.9617924548390716e-06, + "loss": 2.4996, + "step": 27517 + }, + { + "epoch": 2.910109983079526, + "grad_norm": 8.25, + "learning_rate": 1.9616252155465384e-06, + "loss": 2.5503, + "step": 27518 + }, + { + "epoch": 2.910215736040609, + "grad_norm": 8.375, + "learning_rate": 1.961457978780378e-06, + "loss": 2.5867, + "step": 27519 + }, + { + "epoch": 2.910321489001692, + "grad_norm": 8.3125, + "learning_rate": 1.9612907445413744e-06, + "loss": 2.3994, + "step": 27520 + }, + { + "epoch": 2.910427241962775, + "grad_norm": 8.4375, + "learning_rate": 1.9611235128303137e-06, + "loss": 2.5906, + "step": 27521 + }, + { + "epoch": 2.910532994923858, + "grad_norm": 8.9375, + "learning_rate": 1.9609562836479796e-06, + "loss": 2.5872, + "step": 27522 + }, + { + "epoch": 2.9106387478849407, + "grad_norm": 8.5625, + "learning_rate": 1.9607890569951575e-06, + "loss": 2.5275, + "step": 27523 + }, + { + "epoch": 2.9107445008460235, + "grad_norm": 8.5625, + "learning_rate": 1.9606218328726327e-06, + "loss": 2.775, + "step": 27524 + }, + { + "epoch": 2.910850253807107, + "grad_norm": 8.75, + "learning_rate": 1.960454611281187e-06, + "loss": 2.3276, + "step": 27525 + }, + { + "epoch": 2.9109560067681897, + "grad_norm": 8.5, + "learning_rate": 1.9602873922216093e-06, + "loss": 2.5697, + "step": 27526 + }, + { + "epoch": 2.9110617597292725, + "grad_norm": 8.3125, + "learning_rate": 1.9601201756946806e-06, + "loss": 2.4951, + "step": 27527 + }, + { + "epoch": 2.9111675126903553, + "grad_norm": 8.0625, + "learning_rate": 1.9599529617011877e-06, + "loss": 2.4789, + "step": 27528 + }, + { + "epoch": 2.911273265651438, + "grad_norm": 8.6875, + "learning_rate": 1.9597857502419143e-06, + "loss": 2.1926, + "step": 27529 + }, + { + "epoch": 2.911379018612521, + "grad_norm": 8.3125, + "learning_rate": 1.9596185413176455e-06, + "loss": 2.4726, + "step": 27530 + }, + { + "epoch": 2.911484771573604, + "grad_norm": 8.1875, + "learning_rate": 1.9594513349291662e-06, + "loss": 2.5528, + "step": 27531 + }, + { + "epoch": 2.911590524534687, + "grad_norm": 7.875, + "learning_rate": 1.95928413107726e-06, + "loss": 2.5433, + "step": 27532 + }, + { + "epoch": 2.91169627749577, + "grad_norm": 8.6875, + "learning_rate": 1.959116929762714e-06, + "loss": 2.6689, + "step": 27533 + }, + { + "epoch": 2.911802030456853, + "grad_norm": 8.0, + "learning_rate": 1.95894973098631e-06, + "loss": 2.4742, + "step": 27534 + }, + { + "epoch": 2.9119077834179357, + "grad_norm": 7.90625, + "learning_rate": 1.9587825347488336e-06, + "loss": 2.4466, + "step": 27535 + }, + { + "epoch": 2.9120135363790185, + "grad_norm": 8.5, + "learning_rate": 1.9586153410510687e-06, + "loss": 2.6314, + "step": 27536 + }, + { + "epoch": 2.912119289340102, + "grad_norm": 8.4375, + "learning_rate": 1.9584481498938017e-06, + "loss": 2.4066, + "step": 27537 + }, + { + "epoch": 2.912225042301184, + "grad_norm": 8.75, + "learning_rate": 1.958280961277815e-06, + "loss": 2.3459, + "step": 27538 + }, + { + "epoch": 2.9123307952622675, + "grad_norm": 9.0, + "learning_rate": 1.9581137752038944e-06, + "loss": 2.6964, + "step": 27539 + }, + { + "epoch": 2.9124365482233503, + "grad_norm": 8.25, + "learning_rate": 1.9579465916728254e-06, + "loss": 2.4043, + "step": 27540 + }, + { + "epoch": 2.912542301184433, + "grad_norm": 8.1875, + "learning_rate": 1.9577794106853896e-06, + "loss": 2.3205, + "step": 27541 + }, + { + "epoch": 2.912648054145516, + "grad_norm": 8.0625, + "learning_rate": 1.957612232242375e-06, + "loss": 2.5098, + "step": 27542 + }, + { + "epoch": 2.912753807106599, + "grad_norm": 8.5, + "learning_rate": 1.9574450563445633e-06, + "loss": 2.521, + "step": 27543 + }, + { + "epoch": 2.912859560067682, + "grad_norm": 8.1875, + "learning_rate": 1.9572778829927407e-06, + "loss": 2.5146, + "step": 27544 + }, + { + "epoch": 2.912965313028765, + "grad_norm": 7.75, + "learning_rate": 1.9571107121876904e-06, + "loss": 2.4657, + "step": 27545 + }, + { + "epoch": 2.9130710659898478, + "grad_norm": 8.5, + "learning_rate": 1.9569435439301984e-06, + "loss": 2.571, + "step": 27546 + }, + { + "epoch": 2.9131768189509306, + "grad_norm": 7.8125, + "learning_rate": 1.9567763782210465e-06, + "loss": 2.3949, + "step": 27547 + }, + { + "epoch": 2.9132825719120135, + "grad_norm": 8.1875, + "learning_rate": 1.956609215061022e-06, + "loss": 2.5018, + "step": 27548 + }, + { + "epoch": 2.9133883248730963, + "grad_norm": 8.625, + "learning_rate": 1.956442054450909e-06, + "loss": 2.4038, + "step": 27549 + }, + { + "epoch": 2.913494077834179, + "grad_norm": 8.25, + "learning_rate": 1.95627489639149e-06, + "loss": 2.6727, + "step": 27550 + }, + { + "epoch": 2.9135998307952624, + "grad_norm": 8.25, + "learning_rate": 1.956107740883551e-06, + "loss": 2.4855, + "step": 27551 + }, + { + "epoch": 2.9137055837563453, + "grad_norm": 11.875, + "learning_rate": 1.955940587927875e-06, + "loss": 2.8217, + "step": 27552 + }, + { + "epoch": 2.913811336717428, + "grad_norm": 9.0625, + "learning_rate": 1.9557734375252486e-06, + "loss": 2.6507, + "step": 27553 + }, + { + "epoch": 2.913917089678511, + "grad_norm": 8.5, + "learning_rate": 1.9556062896764543e-06, + "loss": 2.5244, + "step": 27554 + }, + { + "epoch": 2.9140228426395938, + "grad_norm": 8.25, + "learning_rate": 1.9554391443822775e-06, + "loss": 2.4376, + "step": 27555 + }, + { + "epoch": 2.914128595600677, + "grad_norm": 8.6875, + "learning_rate": 1.9552720016435016e-06, + "loss": 2.6232, + "step": 27556 + }, + { + "epoch": 2.91423434856176, + "grad_norm": 8.4375, + "learning_rate": 1.955104861460911e-06, + "loss": 2.5221, + "step": 27557 + }, + { + "epoch": 2.9143401015228427, + "grad_norm": 8.75, + "learning_rate": 1.9549377238352912e-06, + "loss": 2.3641, + "step": 27558 + }, + { + "epoch": 2.9144458544839256, + "grad_norm": 9.0, + "learning_rate": 1.954770588767425e-06, + "loss": 2.7618, + "step": 27559 + }, + { + "epoch": 2.9145516074450084, + "grad_norm": 8.625, + "learning_rate": 1.9546034562580983e-06, + "loss": 2.6856, + "step": 27560 + }, + { + "epoch": 2.9146573604060912, + "grad_norm": 8.5625, + "learning_rate": 1.954436326308094e-06, + "loss": 2.5012, + "step": 27561 + }, + { + "epoch": 2.914763113367174, + "grad_norm": 8.5625, + "learning_rate": 1.9542691989181976e-06, + "loss": 2.6047, + "step": 27562 + }, + { + "epoch": 2.9148688663282574, + "grad_norm": 8.3125, + "learning_rate": 1.954102074089191e-06, + "loss": 2.4109, + "step": 27563 + }, + { + "epoch": 2.91497461928934, + "grad_norm": 8.3125, + "learning_rate": 1.953934951821862e-06, + "loss": 2.4906, + "step": 27564 + }, + { + "epoch": 2.915080372250423, + "grad_norm": 8.3125, + "learning_rate": 1.953767832116992e-06, + "loss": 2.316, + "step": 27565 + }, + { + "epoch": 2.915186125211506, + "grad_norm": 8.625, + "learning_rate": 1.9536007149753657e-06, + "loss": 2.7114, + "step": 27566 + }, + { + "epoch": 2.9152918781725887, + "grad_norm": 9.0, + "learning_rate": 1.953433600397769e-06, + "loss": 2.6839, + "step": 27567 + }, + { + "epoch": 2.915397631133672, + "grad_norm": 8.0625, + "learning_rate": 1.9532664883849837e-06, + "loss": 2.5656, + "step": 27568 + }, + { + "epoch": 2.9155033840947544, + "grad_norm": 8.625, + "learning_rate": 1.953099378937796e-06, + "loss": 2.5767, + "step": 27569 + }, + { + "epoch": 2.9156091370558377, + "grad_norm": 7.96875, + "learning_rate": 1.9529322720569885e-06, + "loss": 2.4787, + "step": 27570 + }, + { + "epoch": 2.9157148900169205, + "grad_norm": 8.6875, + "learning_rate": 1.9527651677433474e-06, + "loss": 2.6981, + "step": 27571 + }, + { + "epoch": 2.9158206429780034, + "grad_norm": 8.875, + "learning_rate": 1.952598065997654e-06, + "loss": 2.8064, + "step": 27572 + }, + { + "epoch": 2.915926395939086, + "grad_norm": 8.375, + "learning_rate": 1.952430966820695e-06, + "loss": 2.5777, + "step": 27573 + }, + { + "epoch": 2.916032148900169, + "grad_norm": 7.9375, + "learning_rate": 1.9522638702132528e-06, + "loss": 2.4807, + "step": 27574 + }, + { + "epoch": 2.9161379018612523, + "grad_norm": 9.0625, + "learning_rate": 1.9520967761761126e-06, + "loss": 2.7495, + "step": 27575 + }, + { + "epoch": 2.916243654822335, + "grad_norm": 8.5625, + "learning_rate": 1.9519296847100584e-06, + "loss": 2.3934, + "step": 27576 + }, + { + "epoch": 2.916349407783418, + "grad_norm": 9.1875, + "learning_rate": 1.9517625958158736e-06, + "loss": 2.825, + "step": 27577 + }, + { + "epoch": 2.916455160744501, + "grad_norm": 8.5, + "learning_rate": 1.9515955094943435e-06, + "loss": 2.4545, + "step": 27578 + }, + { + "epoch": 2.9165609137055837, + "grad_norm": 8.375, + "learning_rate": 1.9514284257462497e-06, + "loss": 2.6198, + "step": 27579 + }, + { + "epoch": 2.9166666666666665, + "grad_norm": 8.375, + "learning_rate": 1.9512613445723798e-06, + "loss": 2.6299, + "step": 27580 + }, + { + "epoch": 2.9167724196277494, + "grad_norm": 8.8125, + "learning_rate": 1.9510942659735154e-06, + "loss": 2.5924, + "step": 27581 + }, + { + "epoch": 2.9168781725888326, + "grad_norm": 8.625, + "learning_rate": 1.950927189950441e-06, + "loss": 2.4648, + "step": 27582 + }, + { + "epoch": 2.9169839255499155, + "grad_norm": 8.125, + "learning_rate": 1.9507601165039403e-06, + "loss": 2.3959, + "step": 27583 + }, + { + "epoch": 2.9170896785109983, + "grad_norm": 8.5, + "learning_rate": 1.950593045634798e-06, + "loss": 2.5702, + "step": 27584 + }, + { + "epoch": 2.917195431472081, + "grad_norm": 8.1875, + "learning_rate": 1.950425977343798e-06, + "loss": 2.6166, + "step": 27585 + }, + { + "epoch": 2.917301184433164, + "grad_norm": 8.5, + "learning_rate": 1.950258911631724e-06, + "loss": 2.4559, + "step": 27586 + }, + { + "epoch": 2.9174069373942473, + "grad_norm": 8.3125, + "learning_rate": 1.950091848499361e-06, + "loss": 2.37, + "step": 27587 + }, + { + "epoch": 2.9175126903553297, + "grad_norm": 8.625, + "learning_rate": 1.949924787947491e-06, + "loss": 2.5594, + "step": 27588 + }, + { + "epoch": 2.917618443316413, + "grad_norm": 8.625, + "learning_rate": 1.949757729976899e-06, + "loss": 2.641, + "step": 27589 + }, + { + "epoch": 2.917724196277496, + "grad_norm": 8.9375, + "learning_rate": 1.949590674588369e-06, + "loss": 2.5606, + "step": 27590 + }, + { + "epoch": 2.9178299492385786, + "grad_norm": 8.0, + "learning_rate": 1.949423621782685e-06, + "loss": 2.3722, + "step": 27591 + }, + { + "epoch": 2.9179357021996615, + "grad_norm": 7.8125, + "learning_rate": 1.9492565715606306e-06, + "loss": 2.281, + "step": 27592 + }, + { + "epoch": 2.9180414551607443, + "grad_norm": 8.25, + "learning_rate": 1.9490895239229897e-06, + "loss": 2.7516, + "step": 27593 + }, + { + "epoch": 2.9181472081218276, + "grad_norm": 8.4375, + "learning_rate": 1.9489224788705475e-06, + "loss": 2.5547, + "step": 27594 + }, + { + "epoch": 2.9182529610829104, + "grad_norm": 8.6875, + "learning_rate": 1.948755436404085e-06, + "loss": 2.4206, + "step": 27595 + }, + { + "epoch": 2.9183587140439933, + "grad_norm": 8.3125, + "learning_rate": 1.9485883965243897e-06, + "loss": 2.4891, + "step": 27596 + }, + { + "epoch": 2.918464467005076, + "grad_norm": 8.5, + "learning_rate": 1.948421359232242e-06, + "loss": 2.6473, + "step": 27597 + }, + { + "epoch": 2.918570219966159, + "grad_norm": 8.1875, + "learning_rate": 1.9482543245284282e-06, + "loss": 2.3784, + "step": 27598 + }, + { + "epoch": 2.9186759729272422, + "grad_norm": 8.3125, + "learning_rate": 1.948087292413731e-06, + "loss": 2.5209, + "step": 27599 + }, + { + "epoch": 2.9187817258883246, + "grad_norm": 8.8125, + "learning_rate": 1.9479202628889344e-06, + "loss": 2.6977, + "step": 27600 + }, + { + "epoch": 2.918887478849408, + "grad_norm": 8.25, + "learning_rate": 1.9477532359548214e-06, + "loss": 2.4772, + "step": 27601 + }, + { + "epoch": 2.9189932318104908, + "grad_norm": 9.0, + "learning_rate": 1.947586211612177e-06, + "loss": 2.5393, + "step": 27602 + }, + { + "epoch": 2.9190989847715736, + "grad_norm": 8.125, + "learning_rate": 1.947419189861785e-06, + "loss": 2.3882, + "step": 27603 + }, + { + "epoch": 2.9192047377326564, + "grad_norm": 8.25, + "learning_rate": 1.9472521707044285e-06, + "loss": 2.2688, + "step": 27604 + }, + { + "epoch": 2.9193104906937393, + "grad_norm": 8.4375, + "learning_rate": 1.9470851541408918e-06, + "loss": 2.7425, + "step": 27605 + }, + { + "epoch": 2.9194162436548226, + "grad_norm": 8.375, + "learning_rate": 1.946918140171958e-06, + "loss": 2.5223, + "step": 27606 + }, + { + "epoch": 2.9195219966159054, + "grad_norm": 9.0625, + "learning_rate": 1.9467511287984115e-06, + "loss": 2.4349, + "step": 27607 + }, + { + "epoch": 2.9196277495769882, + "grad_norm": 8.1875, + "learning_rate": 1.946584120021035e-06, + "loss": 2.409, + "step": 27608 + }, + { + "epoch": 2.919733502538071, + "grad_norm": 8.6875, + "learning_rate": 1.946417113840613e-06, + "loss": 2.4498, + "step": 27609 + }, + { + "epoch": 2.919839255499154, + "grad_norm": 8.4375, + "learning_rate": 1.94625011025793e-06, + "loss": 2.585, + "step": 27610 + }, + { + "epoch": 2.9199450084602367, + "grad_norm": 8.125, + "learning_rate": 1.946083109273768e-06, + "loss": 2.4896, + "step": 27611 + }, + { + "epoch": 2.9200507614213196, + "grad_norm": 8.25, + "learning_rate": 1.945916110888912e-06, + "loss": 2.757, + "step": 27612 + }, + { + "epoch": 2.920156514382403, + "grad_norm": 8.125, + "learning_rate": 1.9457491151041442e-06, + "loss": 2.493, + "step": 27613 + }, + { + "epoch": 2.9202622673434857, + "grad_norm": 8.5625, + "learning_rate": 1.94558212192025e-06, + "loss": 2.4074, + "step": 27614 + }, + { + "epoch": 2.9203680203045685, + "grad_norm": 8.25, + "learning_rate": 1.9454151313380114e-06, + "loss": 2.5055, + "step": 27615 + }, + { + "epoch": 2.9204737732656514, + "grad_norm": 8.25, + "learning_rate": 1.945248143358214e-06, + "loss": 2.5507, + "step": 27616 + }, + { + "epoch": 2.9205795262267342, + "grad_norm": 8.375, + "learning_rate": 1.9450811579816383e-06, + "loss": 2.5822, + "step": 27617 + }, + { + "epoch": 2.9206852791878175, + "grad_norm": 9.0, + "learning_rate": 1.9449141752090703e-06, + "loss": 2.5749, + "step": 27618 + }, + { + "epoch": 2.9207910321489, + "grad_norm": 9.1875, + "learning_rate": 1.9447471950412945e-06, + "loss": 2.816, + "step": 27619 + }, + { + "epoch": 2.920896785109983, + "grad_norm": 8.375, + "learning_rate": 1.9445802174790917e-06, + "loss": 2.5717, + "step": 27620 + }, + { + "epoch": 2.921002538071066, + "grad_norm": 8.5625, + "learning_rate": 1.944413242523247e-06, + "loss": 2.6465, + "step": 27621 + }, + { + "epoch": 2.921108291032149, + "grad_norm": 7.84375, + "learning_rate": 1.9442462701745433e-06, + "loss": 2.4652, + "step": 27622 + }, + { + "epoch": 2.9212140439932317, + "grad_norm": 8.25, + "learning_rate": 1.9440793004337656e-06, + "loss": 2.4773, + "step": 27623 + }, + { + "epoch": 2.9213197969543145, + "grad_norm": 8.0625, + "learning_rate": 1.9439123333016957e-06, + "loss": 2.3836, + "step": 27624 + }, + { + "epoch": 2.921425549915398, + "grad_norm": 8.0, + "learning_rate": 1.9437453687791186e-06, + "loss": 2.632, + "step": 27625 + }, + { + "epoch": 2.9215313028764807, + "grad_norm": 8.1875, + "learning_rate": 1.9435784068668164e-06, + "loss": 2.653, + "step": 27626 + }, + { + "epoch": 2.9216370558375635, + "grad_norm": 8.9375, + "learning_rate": 1.9434114475655724e-06, + "loss": 2.6798, + "step": 27627 + }, + { + "epoch": 2.9217428087986463, + "grad_norm": 8.4375, + "learning_rate": 1.943244490876172e-06, + "loss": 2.3007, + "step": 27628 + }, + { + "epoch": 2.921848561759729, + "grad_norm": 8.4375, + "learning_rate": 1.9430775367993966e-06, + "loss": 2.3051, + "step": 27629 + }, + { + "epoch": 2.921954314720812, + "grad_norm": 9.375, + "learning_rate": 1.9429105853360313e-06, + "loss": 2.5928, + "step": 27630 + }, + { + "epoch": 2.922060067681895, + "grad_norm": 8.3125, + "learning_rate": 1.942743636486858e-06, + "loss": 2.3217, + "step": 27631 + }, + { + "epoch": 2.922165820642978, + "grad_norm": 8.6875, + "learning_rate": 1.942576690252662e-06, + "loss": 2.6887, + "step": 27632 + }, + { + "epoch": 2.922271573604061, + "grad_norm": 9.1875, + "learning_rate": 1.9424097466342246e-06, + "loss": 2.3705, + "step": 27633 + }, + { + "epoch": 2.922377326565144, + "grad_norm": 8.375, + "learning_rate": 1.9422428056323312e-06, + "loss": 2.6574, + "step": 27634 + }, + { + "epoch": 2.9224830795262267, + "grad_norm": 8.625, + "learning_rate": 1.9420758672477633e-06, + "loss": 2.5527, + "step": 27635 + }, + { + "epoch": 2.9225888324873095, + "grad_norm": 9.25, + "learning_rate": 1.941908931481305e-06, + "loss": 2.4827, + "step": 27636 + }, + { + "epoch": 2.922694585448393, + "grad_norm": 8.3125, + "learning_rate": 1.941741998333741e-06, + "loss": 2.5272, + "step": 27637 + }, + { + "epoch": 2.9228003384094756, + "grad_norm": 8.8125, + "learning_rate": 1.941575067805852e-06, + "loss": 2.5835, + "step": 27638 + }, + { + "epoch": 2.9229060913705585, + "grad_norm": 7.9375, + "learning_rate": 1.9414081398984243e-06, + "loss": 2.4495, + "step": 27639 + }, + { + "epoch": 2.9230118443316413, + "grad_norm": 8.1875, + "learning_rate": 1.9412412146122383e-06, + "loss": 2.2611, + "step": 27640 + }, + { + "epoch": 2.923117597292724, + "grad_norm": 8.1875, + "learning_rate": 1.9410742919480805e-06, + "loss": 2.5079, + "step": 27641 + }, + { + "epoch": 2.923223350253807, + "grad_norm": 8.0625, + "learning_rate": 1.940907371906731e-06, + "loss": 2.4174, + "step": 27642 + }, + { + "epoch": 2.92332910321489, + "grad_norm": 9.0, + "learning_rate": 1.940740454488975e-06, + "loss": 2.25, + "step": 27643 + }, + { + "epoch": 2.923434856175973, + "grad_norm": 8.375, + "learning_rate": 1.9405735396955957e-06, + "loss": 2.4964, + "step": 27644 + }, + { + "epoch": 2.923540609137056, + "grad_norm": 8.875, + "learning_rate": 1.940406627527375e-06, + "loss": 2.6661, + "step": 27645 + }, + { + "epoch": 2.9236463620981388, + "grad_norm": 9.0, + "learning_rate": 1.940239717985098e-06, + "loss": 2.5373, + "step": 27646 + }, + { + "epoch": 2.9237521150592216, + "grad_norm": 9.375, + "learning_rate": 1.9400728110695465e-06, + "loss": 2.5879, + "step": 27647 + }, + { + "epoch": 2.9238578680203045, + "grad_norm": 8.0625, + "learning_rate": 1.939905906781505e-06, + "loss": 2.4064, + "step": 27648 + }, + { + "epoch": 2.9239636209813877, + "grad_norm": 8.4375, + "learning_rate": 1.9397390051217557e-06, + "loss": 2.5245, + "step": 27649 + }, + { + "epoch": 2.92406937394247, + "grad_norm": 8.5625, + "learning_rate": 1.939572106091082e-06, + "loss": 2.5646, + "step": 27650 + }, + { + "epoch": 2.9241751269035534, + "grad_norm": 9.0, + "learning_rate": 1.9394052096902672e-06, + "loss": 2.7972, + "step": 27651 + }, + { + "epoch": 2.9242808798646363, + "grad_norm": 8.5625, + "learning_rate": 1.9392383159200944e-06, + "loss": 2.3968, + "step": 27652 + }, + { + "epoch": 2.924386632825719, + "grad_norm": 8.5625, + "learning_rate": 1.939071424781347e-06, + "loss": 2.6469, + "step": 27653 + }, + { + "epoch": 2.924492385786802, + "grad_norm": 8.125, + "learning_rate": 1.9389045362748073e-06, + "loss": 2.5194, + "step": 27654 + }, + { + "epoch": 2.9245981387478848, + "grad_norm": 8.375, + "learning_rate": 1.93873765040126e-06, + "loss": 2.5883, + "step": 27655 + }, + { + "epoch": 2.924703891708968, + "grad_norm": 8.5625, + "learning_rate": 1.938570767161487e-06, + "loss": 2.5241, + "step": 27656 + }, + { + "epoch": 2.924809644670051, + "grad_norm": 8.1875, + "learning_rate": 1.9384038865562726e-06, + "loss": 2.5748, + "step": 27657 + }, + { + "epoch": 2.9249153976311337, + "grad_norm": 8.3125, + "learning_rate": 1.938237008586398e-06, + "loss": 2.3579, + "step": 27658 + }, + { + "epoch": 2.9250211505922166, + "grad_norm": 8.0625, + "learning_rate": 1.938070133252648e-06, + "loss": 2.4222, + "step": 27659 + }, + { + "epoch": 2.9251269035532994, + "grad_norm": 8.5625, + "learning_rate": 1.9379032605558045e-06, + "loss": 2.8041, + "step": 27660 + }, + { + "epoch": 2.9252326565143822, + "grad_norm": 8.9375, + "learning_rate": 1.9377363904966516e-06, + "loss": 2.8237, + "step": 27661 + }, + { + "epoch": 2.925338409475465, + "grad_norm": 8.3125, + "learning_rate": 1.937569523075971e-06, + "loss": 2.4026, + "step": 27662 + }, + { + "epoch": 2.9254441624365484, + "grad_norm": 9.0, + "learning_rate": 1.9374026582945472e-06, + "loss": 2.5813, + "step": 27663 + }, + { + "epoch": 2.925549915397631, + "grad_norm": 8.125, + "learning_rate": 1.9372357961531636e-06, + "loss": 2.4875, + "step": 27664 + }, + { + "epoch": 2.925655668358714, + "grad_norm": 8.8125, + "learning_rate": 1.9370689366526012e-06, + "loss": 2.5873, + "step": 27665 + }, + { + "epoch": 2.925761421319797, + "grad_norm": 8.5625, + "learning_rate": 1.936902079793644e-06, + "loss": 2.247, + "step": 27666 + }, + { + "epoch": 2.9258671742808797, + "grad_norm": 8.4375, + "learning_rate": 1.936735225577075e-06, + "loss": 2.6962, + "step": 27667 + }, + { + "epoch": 2.925972927241963, + "grad_norm": 8.5, + "learning_rate": 1.936568374003678e-06, + "loss": 2.449, + "step": 27668 + }, + { + "epoch": 2.926078680203046, + "grad_norm": 9.0, + "learning_rate": 1.936401525074234e-06, + "loss": 2.2782, + "step": 27669 + }, + { + "epoch": 2.9261844331641287, + "grad_norm": 8.5625, + "learning_rate": 1.9362346787895287e-06, + "loss": 2.5523, + "step": 27670 + }, + { + "epoch": 2.9262901861252115, + "grad_norm": 8.5, + "learning_rate": 1.9360678351503414e-06, + "loss": 2.7089, + "step": 27671 + }, + { + "epoch": 2.9263959390862944, + "grad_norm": 8.3125, + "learning_rate": 1.935900994157458e-06, + "loss": 2.5214, + "step": 27672 + }, + { + "epoch": 2.926501692047377, + "grad_norm": 8.25, + "learning_rate": 1.9357341558116616e-06, + "loss": 2.694, + "step": 27673 + }, + { + "epoch": 2.92660744500846, + "grad_norm": 8.375, + "learning_rate": 1.935567320113733e-06, + "loss": 2.3886, + "step": 27674 + }, + { + "epoch": 2.9267131979695433, + "grad_norm": 8.625, + "learning_rate": 1.9354004870644566e-06, + "loss": 2.5218, + "step": 27675 + }, + { + "epoch": 2.926818950930626, + "grad_norm": 8.0625, + "learning_rate": 1.935233656664614e-06, + "loss": 2.6573, + "step": 27676 + }, + { + "epoch": 2.926924703891709, + "grad_norm": 9.0, + "learning_rate": 1.93506682891499e-06, + "loss": 2.7611, + "step": 27677 + }, + { + "epoch": 2.927030456852792, + "grad_norm": 8.0, + "learning_rate": 1.9349000038163653e-06, + "loss": 2.5029, + "step": 27678 + }, + { + "epoch": 2.9271362098138747, + "grad_norm": 8.5, + "learning_rate": 1.9347331813695248e-06, + "loss": 2.7607, + "step": 27679 + }, + { + "epoch": 2.927241962774958, + "grad_norm": 8.9375, + "learning_rate": 1.9345663615752492e-06, + "loss": 1.8662, + "step": 27680 + }, + { + "epoch": 2.9273477157360404, + "grad_norm": 8.1875, + "learning_rate": 1.9343995444343226e-06, + "loss": 2.5511, + "step": 27681 + }, + { + "epoch": 2.9274534686971236, + "grad_norm": 8.125, + "learning_rate": 1.9342327299475284e-06, + "loss": 2.3982, + "step": 27682 + }, + { + "epoch": 2.9275592216582065, + "grad_norm": 8.3125, + "learning_rate": 1.9340659181156476e-06, + "loss": 2.3093, + "step": 27683 + }, + { + "epoch": 2.9276649746192893, + "grad_norm": 8.1875, + "learning_rate": 1.933899108939465e-06, + "loss": 2.3509, + "step": 27684 + }, + { + "epoch": 2.927770727580372, + "grad_norm": 8.8125, + "learning_rate": 1.9337323024197614e-06, + "loss": 2.5531, + "step": 27685 + }, + { + "epoch": 2.927876480541455, + "grad_norm": 8.75, + "learning_rate": 1.9335654985573214e-06, + "loss": 2.5997, + "step": 27686 + }, + { + "epoch": 2.9279822335025383, + "grad_norm": 9.0625, + "learning_rate": 1.933398697352926e-06, + "loss": 2.4369, + "step": 27687 + }, + { + "epoch": 2.928087986463621, + "grad_norm": 8.4375, + "learning_rate": 1.93323189880736e-06, + "loss": 2.5789, + "step": 27688 + }, + { + "epoch": 2.928193739424704, + "grad_norm": 8.0625, + "learning_rate": 1.9330651029214033e-06, + "loss": 2.3912, + "step": 27689 + }, + { + "epoch": 2.928299492385787, + "grad_norm": 8.625, + "learning_rate": 1.932898309695841e-06, + "loss": 2.4016, + "step": 27690 + }, + { + "epoch": 2.9284052453468696, + "grad_norm": 7.96875, + "learning_rate": 1.932731519131455e-06, + "loss": 2.5174, + "step": 27691 + }, + { + "epoch": 2.9285109983079525, + "grad_norm": 8.3125, + "learning_rate": 1.9325647312290275e-06, + "loss": 2.2565, + "step": 27692 + }, + { + "epoch": 2.9286167512690353, + "grad_norm": 7.90625, + "learning_rate": 1.9323979459893425e-06, + "loss": 2.2663, + "step": 27693 + }, + { + "epoch": 2.9287225042301186, + "grad_norm": 7.9375, + "learning_rate": 1.9322311634131814e-06, + "loss": 2.464, + "step": 27694 + }, + { + "epoch": 2.9288282571912014, + "grad_norm": 7.8125, + "learning_rate": 1.9320643835013276e-06, + "loss": 2.4464, + "step": 27695 + }, + { + "epoch": 2.9289340101522843, + "grad_norm": 7.875, + "learning_rate": 1.931897606254563e-06, + "loss": 2.4049, + "step": 27696 + }, + { + "epoch": 2.929039763113367, + "grad_norm": 8.4375, + "learning_rate": 1.931730831673671e-06, + "loss": 2.3914, + "step": 27697 + }, + { + "epoch": 2.92914551607445, + "grad_norm": 8.3125, + "learning_rate": 1.931564059759433e-06, + "loss": 2.4968, + "step": 27698 + }, + { + "epoch": 2.9292512690355332, + "grad_norm": 8.3125, + "learning_rate": 1.9313972905126323e-06, + "loss": 2.3694, + "step": 27699 + }, + { + "epoch": 2.9293570219966156, + "grad_norm": 7.96875, + "learning_rate": 1.931230523934052e-06, + "loss": 2.2867, + "step": 27700 + }, + { + "epoch": 2.929462774957699, + "grad_norm": 8.5625, + "learning_rate": 1.931063760024474e-06, + "loss": 2.5888, + "step": 27701 + }, + { + "epoch": 2.9295685279187818, + "grad_norm": 8.5625, + "learning_rate": 1.9308969987846824e-06, + "loss": 2.8055, + "step": 27702 + }, + { + "epoch": 2.9296742808798646, + "grad_norm": 8.9375, + "learning_rate": 1.9307302402154572e-06, + "loss": 2.5179, + "step": 27703 + }, + { + "epoch": 2.9297800338409474, + "grad_norm": 8.1875, + "learning_rate": 1.9305634843175826e-06, + "loss": 2.577, + "step": 27704 + }, + { + "epoch": 2.9298857868020303, + "grad_norm": 8.4375, + "learning_rate": 1.9303967310918403e-06, + "loss": 2.6401, + "step": 27705 + }, + { + "epoch": 2.9299915397631136, + "grad_norm": 8.875, + "learning_rate": 1.9302299805390133e-06, + "loss": 2.7293, + "step": 27706 + }, + { + "epoch": 2.9300972927241964, + "grad_norm": 7.90625, + "learning_rate": 1.9300632326598835e-06, + "loss": 2.3909, + "step": 27707 + }, + { + "epoch": 2.9302030456852792, + "grad_norm": 8.875, + "learning_rate": 1.929896487455234e-06, + "loss": 2.5315, + "step": 27708 + }, + { + "epoch": 2.930308798646362, + "grad_norm": 8.3125, + "learning_rate": 1.929729744925848e-06, + "loss": 2.4143, + "step": 27709 + }, + { + "epoch": 2.930414551607445, + "grad_norm": 7.90625, + "learning_rate": 1.929563005072506e-06, + "loss": 2.3724, + "step": 27710 + }, + { + "epoch": 2.930520304568528, + "grad_norm": 8.625, + "learning_rate": 1.929396267895993e-06, + "loss": 2.6943, + "step": 27711 + }, + { + "epoch": 2.9306260575296106, + "grad_norm": 8.3125, + "learning_rate": 1.9292295333970886e-06, + "loss": 2.635, + "step": 27712 + }, + { + "epoch": 2.930731810490694, + "grad_norm": 8.5, + "learning_rate": 1.9290628015765774e-06, + "loss": 2.5063, + "step": 27713 + }, + { + "epoch": 2.9308375634517767, + "grad_norm": 8.25, + "learning_rate": 1.92889607243524e-06, + "loss": 2.4681, + "step": 27714 + }, + { + "epoch": 2.9309433164128595, + "grad_norm": 8.3125, + "learning_rate": 1.92872934597386e-06, + "loss": 2.3961, + "step": 27715 + }, + { + "epoch": 2.9310490693739424, + "grad_norm": 8.375, + "learning_rate": 1.92856262219322e-06, + "loss": 2.466, + "step": 27716 + }, + { + "epoch": 2.9311548223350252, + "grad_norm": 8.25, + "learning_rate": 1.9283959010941014e-06, + "loss": 2.6832, + "step": 27717 + }, + { + "epoch": 2.9312605752961085, + "grad_norm": 8.25, + "learning_rate": 1.928229182677288e-06, + "loss": 2.518, + "step": 27718 + }, + { + "epoch": 2.9313663282571913, + "grad_norm": 8.375, + "learning_rate": 1.9280624669435606e-06, + "loss": 2.0659, + "step": 27719 + }, + { + "epoch": 2.931472081218274, + "grad_norm": 8.5625, + "learning_rate": 1.9278957538937023e-06, + "loss": 2.5678, + "step": 27720 + }, + { + "epoch": 2.931577834179357, + "grad_norm": 8.625, + "learning_rate": 1.9277290435284945e-06, + "loss": 2.5766, + "step": 27721 + }, + { + "epoch": 2.93168358714044, + "grad_norm": 7.9375, + "learning_rate": 1.927562335848721e-06, + "loss": 2.3437, + "step": 27722 + }, + { + "epoch": 2.9317893401015227, + "grad_norm": 8.0625, + "learning_rate": 1.9273956308551628e-06, + "loss": 2.4545, + "step": 27723 + }, + { + "epoch": 2.9318950930626055, + "grad_norm": 7.875, + "learning_rate": 1.9272289285486028e-06, + "loss": 2.4533, + "step": 27724 + }, + { + "epoch": 2.932000846023689, + "grad_norm": 8.5, + "learning_rate": 1.927062228929824e-06, + "loss": 2.3031, + "step": 27725 + }, + { + "epoch": 2.9321065989847717, + "grad_norm": 9.625, + "learning_rate": 1.926895531999607e-06, + "loss": 2.5809, + "step": 27726 + }, + { + "epoch": 2.9322123519458545, + "grad_norm": 7.875, + "learning_rate": 1.9267288377587357e-06, + "loss": 2.5317, + "step": 27727 + }, + { + "epoch": 2.9323181049069373, + "grad_norm": 8.625, + "learning_rate": 1.9265621462079907e-06, + "loss": 2.7035, + "step": 27728 + }, + { + "epoch": 2.93242385786802, + "grad_norm": 8.5, + "learning_rate": 1.9263954573481553e-06, + "loss": 2.6477, + "step": 27729 + }, + { + "epoch": 2.9325296108291035, + "grad_norm": 8.6875, + "learning_rate": 1.926228771180011e-06, + "loss": 2.629, + "step": 27730 + }, + { + "epoch": 2.932635363790186, + "grad_norm": 8.0, + "learning_rate": 1.9260620877043412e-06, + "loss": 2.4696, + "step": 27731 + }, + { + "epoch": 2.932741116751269, + "grad_norm": 8.6875, + "learning_rate": 1.9258954069219265e-06, + "loss": 2.5479, + "step": 27732 + }, + { + "epoch": 2.932846869712352, + "grad_norm": 8.8125, + "learning_rate": 1.9257287288335504e-06, + "loss": 2.7165, + "step": 27733 + }, + { + "epoch": 2.932952622673435, + "grad_norm": 8.75, + "learning_rate": 1.9255620534399948e-06, + "loss": 2.4193, + "step": 27734 + }, + { + "epoch": 2.9330583756345177, + "grad_norm": 9.0625, + "learning_rate": 1.9253953807420407e-06, + "loss": 2.5315, + "step": 27735 + }, + { + "epoch": 2.9331641285956005, + "grad_norm": 8.375, + "learning_rate": 1.925228710740472e-06, + "loss": 2.3458, + "step": 27736 + }, + { + "epoch": 2.933269881556684, + "grad_norm": 8.5, + "learning_rate": 1.925062043436069e-06, + "loss": 2.5156, + "step": 27737 + }, + { + "epoch": 2.9333756345177666, + "grad_norm": 8.5, + "learning_rate": 1.924895378829615e-06, + "loss": 2.4868, + "step": 27738 + }, + { + "epoch": 2.9334813874788495, + "grad_norm": 7.84375, + "learning_rate": 1.9247287169218916e-06, + "loss": 2.3548, + "step": 27739 + }, + { + "epoch": 2.9335871404399323, + "grad_norm": 8.375, + "learning_rate": 1.924562057713682e-06, + "loss": 2.5824, + "step": 27740 + }, + { + "epoch": 2.933692893401015, + "grad_norm": 8.1875, + "learning_rate": 1.924395401205766e-06, + "loss": 2.4234, + "step": 27741 + }, + { + "epoch": 2.933798646362098, + "grad_norm": 8.0625, + "learning_rate": 1.924228747398927e-06, + "loss": 2.4191, + "step": 27742 + }, + { + "epoch": 2.933904399323181, + "grad_norm": 8.5625, + "learning_rate": 1.9240620962939485e-06, + "loss": 2.4995, + "step": 27743 + }, + { + "epoch": 2.934010152284264, + "grad_norm": 8.375, + "learning_rate": 1.92389544789161e-06, + "loss": 2.506, + "step": 27744 + }, + { + "epoch": 2.934115905245347, + "grad_norm": 9.3125, + "learning_rate": 1.923728802192695e-06, + "loss": 2.6197, + "step": 27745 + }, + { + "epoch": 2.9342216582064298, + "grad_norm": 8.625, + "learning_rate": 1.9235621591979844e-06, + "loss": 2.5207, + "step": 27746 + }, + { + "epoch": 2.9343274111675126, + "grad_norm": 8.6875, + "learning_rate": 1.9233955189082613e-06, + "loss": 2.6142, + "step": 27747 + }, + { + "epoch": 2.9344331641285955, + "grad_norm": 9.6875, + "learning_rate": 1.9232288813243066e-06, + "loss": 2.7912, + "step": 27748 + }, + { + "epoch": 2.9345389170896787, + "grad_norm": 8.25, + "learning_rate": 1.923062246446904e-06, + "loss": 2.3982, + "step": 27749 + }, + { + "epoch": 2.9346446700507616, + "grad_norm": 8.25, + "learning_rate": 1.9228956142768335e-06, + "loss": 2.2892, + "step": 27750 + }, + { + "epoch": 2.9347504230118444, + "grad_norm": 8.4375, + "learning_rate": 1.9227289848148774e-06, + "loss": 2.7227, + "step": 27751 + }, + { + "epoch": 2.9348561759729273, + "grad_norm": 8.375, + "learning_rate": 1.922562358061819e-06, + "loss": 2.536, + "step": 27752 + }, + { + "epoch": 2.93496192893401, + "grad_norm": 8.6875, + "learning_rate": 1.9223957340184386e-06, + "loss": 2.5083, + "step": 27753 + }, + { + "epoch": 2.935067681895093, + "grad_norm": 8.125, + "learning_rate": 1.9222291126855196e-06, + "loss": 2.5487, + "step": 27754 + }, + { + "epoch": 2.9351734348561758, + "grad_norm": 8.0625, + "learning_rate": 1.922062494063842e-06, + "loss": 2.3745, + "step": 27755 + }, + { + "epoch": 2.935279187817259, + "grad_norm": 8.125, + "learning_rate": 1.9218958781541903e-06, + "loss": 2.4617, + "step": 27756 + }, + { + "epoch": 2.935384940778342, + "grad_norm": 8.375, + "learning_rate": 1.9217292649573433e-06, + "loss": 2.3179, + "step": 27757 + }, + { + "epoch": 2.9354906937394247, + "grad_norm": 8.3125, + "learning_rate": 1.9215626544740854e-06, + "loss": 2.4989, + "step": 27758 + }, + { + "epoch": 2.9355964467005076, + "grad_norm": 8.875, + "learning_rate": 1.9213960467051966e-06, + "loss": 2.3901, + "step": 27759 + }, + { + "epoch": 2.9357021996615904, + "grad_norm": 8.9375, + "learning_rate": 1.9212294416514593e-06, + "loss": 2.842, + "step": 27760 + }, + { + "epoch": 2.9358079526226737, + "grad_norm": 8.25, + "learning_rate": 1.9210628393136564e-06, + "loss": 2.5323, + "step": 27761 + }, + { + "epoch": 2.935913705583756, + "grad_norm": 8.1875, + "learning_rate": 1.920896239692568e-06, + "loss": 2.2527, + "step": 27762 + }, + { + "epoch": 2.9360194585448394, + "grad_norm": 8.3125, + "learning_rate": 1.9207296427889776e-06, + "loss": 2.2072, + "step": 27763 + }, + { + "epoch": 2.936125211505922, + "grad_norm": 8.3125, + "learning_rate": 1.9205630486036652e-06, + "loss": 2.5046, + "step": 27764 + }, + { + "epoch": 2.936230964467005, + "grad_norm": 8.0625, + "learning_rate": 1.9203964571374146e-06, + "loss": 2.5803, + "step": 27765 + }, + { + "epoch": 2.936336717428088, + "grad_norm": 8.8125, + "learning_rate": 1.9202298683910054e-06, + "loss": 2.3253, + "step": 27766 + }, + { + "epoch": 2.9364424703891707, + "grad_norm": 8.6875, + "learning_rate": 1.920063282365221e-06, + "loss": 2.523, + "step": 27767 + }, + { + "epoch": 2.936548223350254, + "grad_norm": 8.4375, + "learning_rate": 1.9198966990608416e-06, + "loss": 2.4358, + "step": 27768 + }, + { + "epoch": 2.936653976311337, + "grad_norm": 8.0625, + "learning_rate": 1.91973011847865e-06, + "loss": 2.2392, + "step": 27769 + }, + { + "epoch": 2.9367597292724197, + "grad_norm": 8.5, + "learning_rate": 1.9195635406194277e-06, + "loss": 2.8696, + "step": 27770 + }, + { + "epoch": 2.9368654822335025, + "grad_norm": 8.1875, + "learning_rate": 1.919396965483956e-06, + "loss": 2.5942, + "step": 27771 + }, + { + "epoch": 2.9369712351945854, + "grad_norm": 8.125, + "learning_rate": 1.9192303930730183e-06, + "loss": 2.6686, + "step": 27772 + }, + { + "epoch": 2.937076988155668, + "grad_norm": 8.875, + "learning_rate": 1.919063823387393e-06, + "loss": 2.4528, + "step": 27773 + }, + { + "epoch": 2.937182741116751, + "grad_norm": 9.375, + "learning_rate": 1.918897256427865e-06, + "loss": 2.455, + "step": 27774 + }, + { + "epoch": 2.9372884940778343, + "grad_norm": 8.0, + "learning_rate": 1.9187306921952135e-06, + "loss": 2.3461, + "step": 27775 + }, + { + "epoch": 2.937394247038917, + "grad_norm": 9.125, + "learning_rate": 1.9185641306902216e-06, + "loss": 2.6328, + "step": 27776 + }, + { + "epoch": 2.9375, + "grad_norm": 8.875, + "learning_rate": 1.91839757191367e-06, + "loss": 2.2007, + "step": 27777 + }, + { + "epoch": 2.937605752961083, + "grad_norm": 8.4375, + "learning_rate": 1.9182310158663407e-06, + "loss": 2.5192, + "step": 27778 + }, + { + "epoch": 2.9377115059221657, + "grad_norm": 8.125, + "learning_rate": 1.918064462549016e-06, + "loss": 2.6748, + "step": 27779 + }, + { + "epoch": 2.937817258883249, + "grad_norm": 8.8125, + "learning_rate": 1.9178979119624764e-06, + "loss": 2.606, + "step": 27780 + }, + { + "epoch": 2.937923011844332, + "grad_norm": 8.875, + "learning_rate": 1.9177313641075047e-06, + "loss": 2.3984, + "step": 27781 + }, + { + "epoch": 2.9380287648054146, + "grad_norm": 8.5625, + "learning_rate": 1.91756481898488e-06, + "loss": 2.3068, + "step": 27782 + }, + { + "epoch": 2.9381345177664975, + "grad_norm": 8.5, + "learning_rate": 1.917398276595387e-06, + "loss": 2.5123, + "step": 27783 + }, + { + "epoch": 2.9382402707275803, + "grad_norm": 8.375, + "learning_rate": 1.917231736939804e-06, + "loss": 2.453, + "step": 27784 + }, + { + "epoch": 2.938346023688663, + "grad_norm": 7.9375, + "learning_rate": 1.9170652000189153e-06, + "loss": 2.3978, + "step": 27785 + }, + { + "epoch": 2.938451776649746, + "grad_norm": 8.125, + "learning_rate": 1.9168986658335003e-06, + "loss": 2.563, + "step": 27786 + }, + { + "epoch": 2.9385575296108293, + "grad_norm": 8.875, + "learning_rate": 1.9167321343843413e-06, + "loss": 2.41, + "step": 27787 + }, + { + "epoch": 2.938663282571912, + "grad_norm": 8.625, + "learning_rate": 1.9165656056722215e-06, + "loss": 2.5232, + "step": 27788 + }, + { + "epoch": 2.938769035532995, + "grad_norm": 8.0625, + "learning_rate": 1.9163990796979193e-06, + "loss": 2.3565, + "step": 27789 + }, + { + "epoch": 2.938874788494078, + "grad_norm": 8.125, + "learning_rate": 1.916232556462218e-06, + "loss": 2.4798, + "step": 27790 + }, + { + "epoch": 2.9389805414551606, + "grad_norm": 8.4375, + "learning_rate": 1.9160660359658982e-06, + "loss": 2.573, + "step": 27791 + }, + { + "epoch": 2.939086294416244, + "grad_norm": 8.8125, + "learning_rate": 1.915899518209742e-06, + "loss": 2.5718, + "step": 27792 + }, + { + "epoch": 2.9391920473773263, + "grad_norm": 8.0, + "learning_rate": 1.91573300319453e-06, + "loss": 2.3785, + "step": 27793 + }, + { + "epoch": 2.9392978003384096, + "grad_norm": 8.375, + "learning_rate": 1.915566490921045e-06, + "loss": 2.4774, + "step": 27794 + }, + { + "epoch": 2.9394035532994924, + "grad_norm": 8.375, + "learning_rate": 1.915399981390067e-06, + "loss": 2.5781, + "step": 27795 + }, + { + "epoch": 2.9395093062605753, + "grad_norm": 8.5625, + "learning_rate": 1.9152334746023764e-06, + "loss": 2.5691, + "step": 27796 + }, + { + "epoch": 2.939615059221658, + "grad_norm": 8.4375, + "learning_rate": 1.9150669705587583e-06, + "loss": 2.6363, + "step": 27797 + }, + { + "epoch": 2.939720812182741, + "grad_norm": 8.125, + "learning_rate": 1.91490046925999e-06, + "loss": 2.6455, + "step": 27798 + }, + { + "epoch": 2.9398265651438242, + "grad_norm": 8.625, + "learning_rate": 1.914733970706855e-06, + "loss": 2.3995, + "step": 27799 + }, + { + "epoch": 2.939932318104907, + "grad_norm": 8.375, + "learning_rate": 1.914567474900134e-06, + "loss": 2.5061, + "step": 27800 + }, + { + "epoch": 2.94003807106599, + "grad_norm": 8.125, + "learning_rate": 1.914400981840609e-06, + "loss": 2.5634, + "step": 27801 + }, + { + "epoch": 2.9401438240270727, + "grad_norm": 8.375, + "learning_rate": 1.91423449152906e-06, + "loss": 2.4575, + "step": 27802 + }, + { + "epoch": 2.9402495769881556, + "grad_norm": 8.25, + "learning_rate": 1.91406800396627e-06, + "loss": 2.3671, + "step": 27803 + }, + { + "epoch": 2.9403553299492384, + "grad_norm": 8.5625, + "learning_rate": 1.913901519153018e-06, + "loss": 2.5571, + "step": 27804 + }, + { + "epoch": 2.9404610829103213, + "grad_norm": 8.125, + "learning_rate": 1.9137350370900865e-06, + "loss": 2.6226, + "step": 27805 + }, + { + "epoch": 2.9405668358714045, + "grad_norm": 9.1875, + "learning_rate": 1.9135685577782576e-06, + "loss": 2.776, + "step": 27806 + }, + { + "epoch": 2.9406725888324874, + "grad_norm": 8.625, + "learning_rate": 1.913402081218311e-06, + "loss": 2.6463, + "step": 27807 + }, + { + "epoch": 2.9407783417935702, + "grad_norm": 8.6875, + "learning_rate": 1.9132356074110285e-06, + "loss": 2.7371, + "step": 27808 + }, + { + "epoch": 2.940884094754653, + "grad_norm": 8.375, + "learning_rate": 1.9130691363571915e-06, + "loss": 2.4608, + "step": 27809 + }, + { + "epoch": 2.940989847715736, + "grad_norm": 8.75, + "learning_rate": 1.912902668057581e-06, + "loss": 2.3292, + "step": 27810 + }, + { + "epoch": 2.941095600676819, + "grad_norm": 8.25, + "learning_rate": 1.912736202512978e-06, + "loss": 2.0715, + "step": 27811 + }, + { + "epoch": 2.941201353637902, + "grad_norm": 8.3125, + "learning_rate": 1.912569739724164e-06, + "loss": 2.5881, + "step": 27812 + }, + { + "epoch": 2.941307106598985, + "grad_norm": 8.5, + "learning_rate": 1.91240327969192e-06, + "loss": 2.7335, + "step": 27813 + }, + { + "epoch": 2.9414128595600677, + "grad_norm": 8.4375, + "learning_rate": 1.9122368224170265e-06, + "loss": 2.688, + "step": 27814 + }, + { + "epoch": 2.9415186125211505, + "grad_norm": 8.3125, + "learning_rate": 1.912070367900266e-06, + "loss": 2.3282, + "step": 27815 + }, + { + "epoch": 2.9416243654822334, + "grad_norm": 8.5625, + "learning_rate": 1.911903916142418e-06, + "loss": 2.5538, + "step": 27816 + }, + { + "epoch": 2.941730118443316, + "grad_norm": 8.0, + "learning_rate": 1.9117374671442653e-06, + "loss": 2.4159, + "step": 27817 + }, + { + "epoch": 2.9418358714043995, + "grad_norm": 8.375, + "learning_rate": 1.9115710209065874e-06, + "loss": 2.4537, + "step": 27818 + }, + { + "epoch": 2.9419416243654823, + "grad_norm": 8.625, + "learning_rate": 1.911404577430167e-06, + "loss": 2.7136, + "step": 27819 + }, + { + "epoch": 2.942047377326565, + "grad_norm": 8.3125, + "learning_rate": 1.911238136715783e-06, + "loss": 2.6244, + "step": 27820 + }, + { + "epoch": 2.942153130287648, + "grad_norm": 8.25, + "learning_rate": 1.911071698764219e-06, + "loss": 2.5283, + "step": 27821 + }, + { + "epoch": 2.942258883248731, + "grad_norm": 8.8125, + "learning_rate": 1.9109052635762534e-06, + "loss": 2.6302, + "step": 27822 + }, + { + "epoch": 2.942364636209814, + "grad_norm": 7.96875, + "learning_rate": 1.910738831152668e-06, + "loss": 2.6795, + "step": 27823 + }, + { + "epoch": 2.9424703891708965, + "grad_norm": 8.375, + "learning_rate": 1.9105724014942455e-06, + "loss": 2.7002, + "step": 27824 + }, + { + "epoch": 2.94257614213198, + "grad_norm": 8.3125, + "learning_rate": 1.910405974601765e-06, + "loss": 2.587, + "step": 27825 + }, + { + "epoch": 2.9426818950930627, + "grad_norm": 8.5625, + "learning_rate": 1.9102395504760093e-06, + "loss": 2.522, + "step": 27826 + }, + { + "epoch": 2.9427876480541455, + "grad_norm": 8.8125, + "learning_rate": 1.910073129117757e-06, + "loss": 2.6819, + "step": 27827 + }, + { + "epoch": 2.9428934010152283, + "grad_norm": 8.375, + "learning_rate": 1.9099067105277906e-06, + "loss": 2.5337, + "step": 27828 + }, + { + "epoch": 2.942999153976311, + "grad_norm": 8.4375, + "learning_rate": 1.90974029470689e-06, + "loss": 2.6878, + "step": 27829 + }, + { + "epoch": 2.9431049069373945, + "grad_norm": 8.75, + "learning_rate": 1.909573881655837e-06, + "loss": 2.3779, + "step": 27830 + }, + { + "epoch": 2.9432106598984773, + "grad_norm": 8.0, + "learning_rate": 1.909407471375413e-06, + "loss": 2.5611, + "step": 27831 + }, + { + "epoch": 2.94331641285956, + "grad_norm": 8.375, + "learning_rate": 1.9092410638663977e-06, + "loss": 2.5057, + "step": 27832 + }, + { + "epoch": 2.943422165820643, + "grad_norm": 8.1875, + "learning_rate": 1.909074659129573e-06, + "loss": 2.6494, + "step": 27833 + }, + { + "epoch": 2.943527918781726, + "grad_norm": 8.0625, + "learning_rate": 1.9089082571657184e-06, + "loss": 2.4374, + "step": 27834 + }, + { + "epoch": 2.9436336717428087, + "grad_norm": 8.25, + "learning_rate": 1.908741857975617e-06, + "loss": 2.4968, + "step": 27835 + }, + { + "epoch": 2.9437394247038915, + "grad_norm": 8.3125, + "learning_rate": 1.908575461560047e-06, + "loss": 2.329, + "step": 27836 + }, + { + "epoch": 2.9438451776649748, + "grad_norm": 8.5, + "learning_rate": 1.9084090679197907e-06, + "loss": 2.7724, + "step": 27837 + }, + { + "epoch": 2.9439509306260576, + "grad_norm": 8.4375, + "learning_rate": 1.908242677055629e-06, + "loss": 2.5041, + "step": 27838 + }, + { + "epoch": 2.9440566835871405, + "grad_norm": 8.875, + "learning_rate": 1.9080762889683416e-06, + "loss": 2.5241, + "step": 27839 + }, + { + "epoch": 2.9441624365482233, + "grad_norm": 8.75, + "learning_rate": 1.907909903658711e-06, + "loss": 2.684, + "step": 27840 + }, + { + "epoch": 2.944268189509306, + "grad_norm": 8.375, + "learning_rate": 1.9077435211275163e-06, + "loss": 2.6553, + "step": 27841 + }, + { + "epoch": 2.9443739424703894, + "grad_norm": 9.125, + "learning_rate": 1.9075771413755403e-06, + "loss": 2.6048, + "step": 27842 + }, + { + "epoch": 2.944479695431472, + "grad_norm": 8.25, + "learning_rate": 1.9074107644035615e-06, + "loss": 2.4265, + "step": 27843 + }, + { + "epoch": 2.944585448392555, + "grad_norm": 8.375, + "learning_rate": 1.907244390212362e-06, + "loss": 2.3439, + "step": 27844 + }, + { + "epoch": 2.944691201353638, + "grad_norm": 8.1875, + "learning_rate": 1.9070780188027215e-06, + "loss": 2.4227, + "step": 27845 + }, + { + "epoch": 2.9447969543147208, + "grad_norm": 8.8125, + "learning_rate": 1.906911650175422e-06, + "loss": 2.6084, + "step": 27846 + }, + { + "epoch": 2.9449027072758036, + "grad_norm": 9.125, + "learning_rate": 1.9067452843312429e-06, + "loss": 2.4246, + "step": 27847 + }, + { + "epoch": 2.9450084602368864, + "grad_norm": 8.5, + "learning_rate": 1.9065789212709656e-06, + "loss": 2.7513, + "step": 27848 + }, + { + "epoch": 2.9451142131979697, + "grad_norm": 7.96875, + "learning_rate": 1.906412560995372e-06, + "loss": 2.6335, + "step": 27849 + }, + { + "epoch": 2.9452199661590526, + "grad_norm": 8.875, + "learning_rate": 1.9062462035052398e-06, + "loss": 2.6109, + "step": 27850 + }, + { + "epoch": 2.9453257191201354, + "grad_norm": 8.5625, + "learning_rate": 1.9060798488013529e-06, + "loss": 2.6392, + "step": 27851 + }, + { + "epoch": 2.9454314720812182, + "grad_norm": 8.875, + "learning_rate": 1.9059134968844892e-06, + "loss": 2.719, + "step": 27852 + }, + { + "epoch": 2.945537225042301, + "grad_norm": 8.5625, + "learning_rate": 1.9057471477554313e-06, + "loss": 2.5432, + "step": 27853 + }, + { + "epoch": 2.9456429780033844, + "grad_norm": 9.125, + "learning_rate": 1.9055808014149582e-06, + "loss": 2.5243, + "step": 27854 + }, + { + "epoch": 2.9457487309644668, + "grad_norm": 8.5, + "learning_rate": 1.9054144578638517e-06, + "loss": 2.3982, + "step": 27855 + }, + { + "epoch": 2.94585448392555, + "grad_norm": 8.375, + "learning_rate": 1.9052481171028916e-06, + "loss": 2.3625, + "step": 27856 + }, + { + "epoch": 2.945960236886633, + "grad_norm": 8.3125, + "learning_rate": 1.905081779132859e-06, + "loss": 2.4911, + "step": 27857 + }, + { + "epoch": 2.9460659898477157, + "grad_norm": 8.75, + "learning_rate": 1.9049154439545354e-06, + "loss": 2.7723, + "step": 27858 + }, + { + "epoch": 2.9461717428087986, + "grad_norm": 8.8125, + "learning_rate": 1.9047491115686989e-06, + "loss": 2.5749, + "step": 27859 + }, + { + "epoch": 2.9462774957698814, + "grad_norm": 8.75, + "learning_rate": 1.904582781976132e-06, + "loss": 2.6476, + "step": 27860 + }, + { + "epoch": 2.9463832487309647, + "grad_norm": 8.125, + "learning_rate": 1.904416455177614e-06, + "loss": 2.6216, + "step": 27861 + }, + { + "epoch": 2.9464890016920475, + "grad_norm": 8.625, + "learning_rate": 1.9042501311739267e-06, + "loss": 2.2923, + "step": 27862 + }, + { + "epoch": 2.9465947546531304, + "grad_norm": 9.0, + "learning_rate": 1.9040838099658494e-06, + "loss": 2.7202, + "step": 27863 + }, + { + "epoch": 2.946700507614213, + "grad_norm": 8.625, + "learning_rate": 1.9039174915541636e-06, + "loss": 2.4825, + "step": 27864 + }, + { + "epoch": 2.946806260575296, + "grad_norm": 9.25, + "learning_rate": 1.9037511759396487e-06, + "loss": 2.5575, + "step": 27865 + }, + { + "epoch": 2.946912013536379, + "grad_norm": 8.6875, + "learning_rate": 1.903584863123085e-06, + "loss": 2.6508, + "step": 27866 + }, + { + "epoch": 2.9470177664974617, + "grad_norm": 8.0625, + "learning_rate": 1.9034185531052552e-06, + "loss": 2.4774, + "step": 27867 + }, + { + "epoch": 2.947123519458545, + "grad_norm": 8.375, + "learning_rate": 1.9032522458869368e-06, + "loss": 2.545, + "step": 27868 + }, + { + "epoch": 2.947229272419628, + "grad_norm": 8.125, + "learning_rate": 1.9030859414689123e-06, + "loss": 2.2753, + "step": 27869 + }, + { + "epoch": 2.9473350253807107, + "grad_norm": 8.75, + "learning_rate": 1.9029196398519608e-06, + "loss": 2.6275, + "step": 27870 + }, + { + "epoch": 2.9474407783417935, + "grad_norm": 7.96875, + "learning_rate": 1.902753341036864e-06, + "loss": 2.286, + "step": 27871 + }, + { + "epoch": 2.9475465313028764, + "grad_norm": 9.5625, + "learning_rate": 1.9025870450244008e-06, + "loss": 2.7236, + "step": 27872 + }, + { + "epoch": 2.9476522842639596, + "grad_norm": 8.625, + "learning_rate": 1.902420751815353e-06, + "loss": 2.6339, + "step": 27873 + }, + { + "epoch": 2.947758037225042, + "grad_norm": 7.875, + "learning_rate": 1.9022544614104995e-06, + "loss": 2.5376, + "step": 27874 + }, + { + "epoch": 2.9478637901861253, + "grad_norm": 8.1875, + "learning_rate": 1.9020881738106215e-06, + "loss": 2.377, + "step": 27875 + }, + { + "epoch": 2.947969543147208, + "grad_norm": 8.25, + "learning_rate": 1.9019218890164994e-06, + "loss": 2.444, + "step": 27876 + }, + { + "epoch": 2.948075296108291, + "grad_norm": 8.125, + "learning_rate": 1.9017556070289128e-06, + "loss": 2.4827, + "step": 27877 + }, + { + "epoch": 2.948181049069374, + "grad_norm": 8.5625, + "learning_rate": 1.901589327848643e-06, + "loss": 2.6916, + "step": 27878 + }, + { + "epoch": 2.9482868020304567, + "grad_norm": 8.4375, + "learning_rate": 1.9014230514764692e-06, + "loss": 2.5557, + "step": 27879 + }, + { + "epoch": 2.94839255499154, + "grad_norm": 8.625, + "learning_rate": 1.9012567779131734e-06, + "loss": 2.5254, + "step": 27880 + }, + { + "epoch": 2.948498307952623, + "grad_norm": 8.3125, + "learning_rate": 1.9010905071595334e-06, + "loss": 2.4292, + "step": 27881 + }, + { + "epoch": 2.9486040609137056, + "grad_norm": 8.5625, + "learning_rate": 1.9009242392163314e-06, + "loss": 2.8669, + "step": 27882 + }, + { + "epoch": 2.9487098138747885, + "grad_norm": 7.96875, + "learning_rate": 1.9007579740843462e-06, + "loss": 2.4924, + "step": 27883 + }, + { + "epoch": 2.9488155668358713, + "grad_norm": 7.6875, + "learning_rate": 1.900591711764359e-06, + "loss": 2.3557, + "step": 27884 + }, + { + "epoch": 2.948921319796954, + "grad_norm": 8.1875, + "learning_rate": 1.9004254522571501e-06, + "loss": 2.4821, + "step": 27885 + }, + { + "epoch": 2.949027072758037, + "grad_norm": 8.0625, + "learning_rate": 1.9002591955634988e-06, + "loss": 2.5594, + "step": 27886 + }, + { + "epoch": 2.9491328257191203, + "grad_norm": 8.25, + "learning_rate": 1.9000929416841866e-06, + "loss": 2.4805, + "step": 27887 + }, + { + "epoch": 2.949238578680203, + "grad_norm": 8.5, + "learning_rate": 1.899926690619992e-06, + "loss": 2.5725, + "step": 27888 + }, + { + "epoch": 2.949344331641286, + "grad_norm": 8.625, + "learning_rate": 1.8997604423716972e-06, + "loss": 2.6782, + "step": 27889 + }, + { + "epoch": 2.949450084602369, + "grad_norm": 8.625, + "learning_rate": 1.8995941969400803e-06, + "loss": 2.2579, + "step": 27890 + }, + { + "epoch": 2.9495558375634516, + "grad_norm": 9.75, + "learning_rate": 1.8994279543259226e-06, + "loss": 2.7229, + "step": 27891 + }, + { + "epoch": 2.949661590524535, + "grad_norm": 9.125, + "learning_rate": 1.8992617145300033e-06, + "loss": 2.7009, + "step": 27892 + }, + { + "epoch": 2.9497673434856178, + "grad_norm": 7.90625, + "learning_rate": 1.8990954775531032e-06, + "loss": 2.5453, + "step": 27893 + }, + { + "epoch": 2.9498730964467006, + "grad_norm": 8.625, + "learning_rate": 1.8989292433960028e-06, + "loss": 2.499, + "step": 27894 + }, + { + "epoch": 2.9499788494077834, + "grad_norm": 8.875, + "learning_rate": 1.898763012059481e-06, + "loss": 2.8221, + "step": 27895 + }, + { + "epoch": 2.9500846023688663, + "grad_norm": 8.5, + "learning_rate": 1.8985967835443197e-06, + "loss": 2.5687, + "step": 27896 + }, + { + "epoch": 2.950190355329949, + "grad_norm": 8.1875, + "learning_rate": 1.8984305578512966e-06, + "loss": 2.5153, + "step": 27897 + }, + { + "epoch": 2.950296108291032, + "grad_norm": 8.8125, + "learning_rate": 1.8982643349811933e-06, + "loss": 2.6325, + "step": 27898 + }, + { + "epoch": 2.9504018612521152, + "grad_norm": 9.4375, + "learning_rate": 1.8980981149347888e-06, + "loss": 2.7814, + "step": 27899 + }, + { + "epoch": 2.950507614213198, + "grad_norm": 8.4375, + "learning_rate": 1.8979318977128645e-06, + "loss": 2.5455, + "step": 27900 + }, + { + "epoch": 2.950613367174281, + "grad_norm": 8.375, + "learning_rate": 1.8977656833161987e-06, + "loss": 2.5649, + "step": 27901 + }, + { + "epoch": 2.9507191201353637, + "grad_norm": 8.4375, + "learning_rate": 1.8975994717455723e-06, + "loss": 2.6002, + "step": 27902 + }, + { + "epoch": 2.9508248730964466, + "grad_norm": 8.5, + "learning_rate": 1.8974332630017664e-06, + "loss": 2.4802, + "step": 27903 + }, + { + "epoch": 2.95093062605753, + "grad_norm": 8.0, + "learning_rate": 1.897267057085558e-06, + "loss": 2.4134, + "step": 27904 + }, + { + "epoch": 2.9510363790186123, + "grad_norm": 8.25, + "learning_rate": 1.8971008539977307e-06, + "loss": 2.6709, + "step": 27905 + }, + { + "epoch": 2.9511421319796955, + "grad_norm": 8.3125, + "learning_rate": 1.8969346537390613e-06, + "loss": 2.7235, + "step": 27906 + }, + { + "epoch": 2.9512478849407784, + "grad_norm": 8.0, + "learning_rate": 1.8967684563103314e-06, + "loss": 2.5069, + "step": 27907 + }, + { + "epoch": 2.9513536379018612, + "grad_norm": 8.625, + "learning_rate": 1.8966022617123198e-06, + "loss": 2.6878, + "step": 27908 + }, + { + "epoch": 2.951459390862944, + "grad_norm": 8.5625, + "learning_rate": 1.8964360699458076e-06, + "loss": 2.5385, + "step": 27909 + }, + { + "epoch": 2.951565143824027, + "grad_norm": 8.25, + "learning_rate": 1.8962698810115737e-06, + "loss": 2.4234, + "step": 27910 + }, + { + "epoch": 2.95167089678511, + "grad_norm": 8.5, + "learning_rate": 1.896103694910398e-06, + "loss": 2.3667, + "step": 27911 + }, + { + "epoch": 2.951776649746193, + "grad_norm": 8.625, + "learning_rate": 1.8959375116430623e-06, + "loss": 2.3067, + "step": 27912 + }, + { + "epoch": 2.951882402707276, + "grad_norm": 8.9375, + "learning_rate": 1.8957713312103433e-06, + "loss": 2.3847, + "step": 27913 + }, + { + "epoch": 2.9519881556683587, + "grad_norm": 9.3125, + "learning_rate": 1.895605153613023e-06, + "loss": 2.5692, + "step": 27914 + }, + { + "epoch": 2.9520939086294415, + "grad_norm": 8.6875, + "learning_rate": 1.89543897885188e-06, + "loss": 2.6641, + "step": 27915 + }, + { + "epoch": 2.9521996615905244, + "grad_norm": 8.8125, + "learning_rate": 1.895272806927695e-06, + "loss": 2.528, + "step": 27916 + }, + { + "epoch": 2.952305414551607, + "grad_norm": 8.3125, + "learning_rate": 1.895106637841247e-06, + "loss": 2.4816, + "step": 27917 + }, + { + "epoch": 2.9524111675126905, + "grad_norm": 7.9375, + "learning_rate": 1.894940471593317e-06, + "loss": 2.6376, + "step": 27918 + }, + { + "epoch": 2.9525169204737733, + "grad_norm": 8.625, + "learning_rate": 1.894774308184683e-06, + "loss": 2.5771, + "step": 27919 + }, + { + "epoch": 2.952622673434856, + "grad_norm": 8.5, + "learning_rate": 1.8946081476161253e-06, + "loss": 2.5168, + "step": 27920 + }, + { + "epoch": 2.952728426395939, + "grad_norm": 8.5625, + "learning_rate": 1.8944419898884253e-06, + "loss": 2.3942, + "step": 27921 + }, + { + "epoch": 2.952834179357022, + "grad_norm": 8.5625, + "learning_rate": 1.8942758350023602e-06, + "loss": 2.5861, + "step": 27922 + }, + { + "epoch": 2.952939932318105, + "grad_norm": 8.8125, + "learning_rate": 1.8941096829587117e-06, + "loss": 2.7154, + "step": 27923 + }, + { + "epoch": 2.953045685279188, + "grad_norm": 8.8125, + "learning_rate": 1.893943533758258e-06, + "loss": 2.4105, + "step": 27924 + }, + { + "epoch": 2.953151438240271, + "grad_norm": 8.5, + "learning_rate": 1.8937773874017801e-06, + "loss": 2.3215, + "step": 27925 + }, + { + "epoch": 2.9532571912013537, + "grad_norm": 8.875, + "learning_rate": 1.893611243890056e-06, + "loss": 2.5559, + "step": 27926 + }, + { + "epoch": 2.9533629441624365, + "grad_norm": 8.4375, + "learning_rate": 1.893445103223868e-06, + "loss": 2.5615, + "step": 27927 + }, + { + "epoch": 2.9534686971235193, + "grad_norm": 8.375, + "learning_rate": 1.8932789654039926e-06, + "loss": 2.5551, + "step": 27928 + }, + { + "epoch": 2.953574450084602, + "grad_norm": 7.96875, + "learning_rate": 1.893112830431211e-06, + "loss": 2.5079, + "step": 27929 + }, + { + "epoch": 2.9536802030456855, + "grad_norm": 8.1875, + "learning_rate": 1.8929466983063033e-06, + "loss": 2.5073, + "step": 27930 + }, + { + "epoch": 2.9537859560067683, + "grad_norm": 8.125, + "learning_rate": 1.8927805690300477e-06, + "loss": 2.4296, + "step": 27931 + }, + { + "epoch": 2.953891708967851, + "grad_norm": 8.25, + "learning_rate": 1.8926144426032253e-06, + "loss": 2.6174, + "step": 27932 + }, + { + "epoch": 2.953997461928934, + "grad_norm": 7.875, + "learning_rate": 1.8924483190266143e-06, + "loss": 2.52, + "step": 27933 + }, + { + "epoch": 2.954103214890017, + "grad_norm": 8.6875, + "learning_rate": 1.892282198300996e-06, + "loss": 2.6949, + "step": 27934 + }, + { + "epoch": 2.9542089678511, + "grad_norm": 8.625, + "learning_rate": 1.8921160804271477e-06, + "loss": 2.701, + "step": 27935 + }, + { + "epoch": 2.9543147208121825, + "grad_norm": 8.8125, + "learning_rate": 1.8919499654058505e-06, + "loss": 2.497, + "step": 27936 + }, + { + "epoch": 2.9544204737732658, + "grad_norm": 8.5625, + "learning_rate": 1.8917838532378829e-06, + "loss": 2.6194, + "step": 27937 + }, + { + "epoch": 2.9545262267343486, + "grad_norm": 8.6875, + "learning_rate": 1.8916177439240245e-06, + "loss": 2.6134, + "step": 27938 + }, + { + "epoch": 2.9546319796954315, + "grad_norm": 8.0625, + "learning_rate": 1.8914516374650565e-06, + "loss": 2.4528, + "step": 27939 + }, + { + "epoch": 2.9547377326565143, + "grad_norm": 8.3125, + "learning_rate": 1.8912855338617562e-06, + "loss": 2.5793, + "step": 27940 + }, + { + "epoch": 2.954843485617597, + "grad_norm": 8.6875, + "learning_rate": 1.891119433114905e-06, + "loss": 2.4856, + "step": 27941 + }, + { + "epoch": 2.9549492385786804, + "grad_norm": 8.1875, + "learning_rate": 1.8909533352252796e-06, + "loss": 2.5786, + "step": 27942 + }, + { + "epoch": 2.9550549915397633, + "grad_norm": 8.3125, + "learning_rate": 1.8907872401936627e-06, + "loss": 2.5547, + "step": 27943 + }, + { + "epoch": 2.955160744500846, + "grad_norm": 8.1875, + "learning_rate": 1.8906211480208315e-06, + "loss": 2.5064, + "step": 27944 + }, + { + "epoch": 2.955266497461929, + "grad_norm": 7.6875, + "learning_rate": 1.8904550587075653e-06, + "loss": 2.0037, + "step": 27945 + }, + { + "epoch": 2.9553722504230118, + "grad_norm": 8.375, + "learning_rate": 1.8902889722546452e-06, + "loss": 2.455, + "step": 27946 + }, + { + "epoch": 2.9554780033840946, + "grad_norm": 8.375, + "learning_rate": 1.8901228886628492e-06, + "loss": 2.4425, + "step": 27947 + }, + { + "epoch": 2.9555837563451774, + "grad_norm": 8.0, + "learning_rate": 1.8899568079329573e-06, + "loss": 2.5269, + "step": 27948 + }, + { + "epoch": 2.9556895093062607, + "grad_norm": 8.6875, + "learning_rate": 1.8897907300657481e-06, + "loss": 2.4444, + "step": 27949 + }, + { + "epoch": 2.9557952622673436, + "grad_norm": 8.125, + "learning_rate": 1.8896246550620027e-06, + "loss": 2.3686, + "step": 27950 + }, + { + "epoch": 2.9559010152284264, + "grad_norm": 8.9375, + "learning_rate": 1.8894585829224976e-06, + "loss": 2.5381, + "step": 27951 + }, + { + "epoch": 2.9560067681895092, + "grad_norm": 8.0625, + "learning_rate": 1.8892925136480148e-06, + "loss": 2.539, + "step": 27952 + }, + { + "epoch": 2.956112521150592, + "grad_norm": 7.8125, + "learning_rate": 1.8891264472393314e-06, + "loss": 2.5577, + "step": 27953 + }, + { + "epoch": 2.9562182741116754, + "grad_norm": 9.0625, + "learning_rate": 1.8889603836972282e-06, + "loss": 2.4852, + "step": 27954 + }, + { + "epoch": 2.9563240270727578, + "grad_norm": 8.125, + "learning_rate": 1.8887943230224841e-06, + "loss": 2.3079, + "step": 27955 + }, + { + "epoch": 2.956429780033841, + "grad_norm": 8.625, + "learning_rate": 1.8886282652158784e-06, + "loss": 2.6462, + "step": 27956 + }, + { + "epoch": 2.956535532994924, + "grad_norm": 8.375, + "learning_rate": 1.8884622102781905e-06, + "loss": 2.5008, + "step": 27957 + }, + { + "epoch": 2.9566412859560067, + "grad_norm": 8.375, + "learning_rate": 1.8882961582101982e-06, + "loss": 2.5934, + "step": 27958 + }, + { + "epoch": 2.9567470389170896, + "grad_norm": 8.375, + "learning_rate": 1.8881301090126833e-06, + "loss": 2.4528, + "step": 27959 + }, + { + "epoch": 2.9568527918781724, + "grad_norm": 8.625, + "learning_rate": 1.8879640626864226e-06, + "loss": 2.5633, + "step": 27960 + }, + { + "epoch": 2.9569585448392557, + "grad_norm": 9.0, + "learning_rate": 1.8877980192321966e-06, + "loss": 2.5335, + "step": 27961 + }, + { + "epoch": 2.9570642978003385, + "grad_norm": 9.0, + "learning_rate": 1.8876319786507838e-06, + "loss": 2.6052, + "step": 27962 + }, + { + "epoch": 2.9571700507614214, + "grad_norm": 7.96875, + "learning_rate": 1.8874659409429635e-06, + "loss": 2.4051, + "step": 27963 + }, + { + "epoch": 2.957275803722504, + "grad_norm": 8.9375, + "learning_rate": 1.8872999061095156e-06, + "loss": 2.6566, + "step": 27964 + }, + { + "epoch": 2.957381556683587, + "grad_norm": 9.0, + "learning_rate": 1.887133874151218e-06, + "loss": 2.8163, + "step": 27965 + }, + { + "epoch": 2.9574873096446703, + "grad_norm": 9.25, + "learning_rate": 1.8869678450688517e-06, + "loss": 2.4736, + "step": 27966 + }, + { + "epoch": 2.9575930626057527, + "grad_norm": 8.5625, + "learning_rate": 1.8868018188631936e-06, + "loss": 2.4934, + "step": 27967 + }, + { + "epoch": 2.957698815566836, + "grad_norm": 8.8125, + "learning_rate": 1.8866357955350243e-06, + "loss": 2.5571, + "step": 27968 + }, + { + "epoch": 2.957804568527919, + "grad_norm": 8.375, + "learning_rate": 1.8864697750851215e-06, + "loss": 2.5187, + "step": 27969 + }, + { + "epoch": 2.9579103214890017, + "grad_norm": 8.625, + "learning_rate": 1.886303757514266e-06, + "loss": 2.6077, + "step": 27970 + }, + { + "epoch": 2.9580160744500845, + "grad_norm": 8.625, + "learning_rate": 1.8861377428232353e-06, + "loss": 2.6293, + "step": 27971 + }, + { + "epoch": 2.9581218274111674, + "grad_norm": 8.375, + "learning_rate": 1.885971731012809e-06, + "loss": 2.5511, + "step": 27972 + }, + { + "epoch": 2.9582275803722506, + "grad_norm": 8.6875, + "learning_rate": 1.8858057220837676e-06, + "loss": 2.5649, + "step": 27973 + }, + { + "epoch": 2.9583333333333335, + "grad_norm": 8.4375, + "learning_rate": 1.8856397160368869e-06, + "loss": 2.5607, + "step": 27974 + }, + { + "epoch": 2.9584390862944163, + "grad_norm": 8.375, + "learning_rate": 1.8854737128729495e-06, + "loss": 2.597, + "step": 27975 + }, + { + "epoch": 2.958544839255499, + "grad_norm": 8.375, + "learning_rate": 1.8853077125927314e-06, + "loss": 2.3541, + "step": 27976 + }, + { + "epoch": 2.958650592216582, + "grad_norm": 8.375, + "learning_rate": 1.8851417151970137e-06, + "loss": 2.5272, + "step": 27977 + }, + { + "epoch": 2.958756345177665, + "grad_norm": 9.4375, + "learning_rate": 1.8849757206865735e-06, + "loss": 2.5714, + "step": 27978 + }, + { + "epoch": 2.9588620981387477, + "grad_norm": 8.4375, + "learning_rate": 1.8848097290621915e-06, + "loss": 2.5267, + "step": 27979 + }, + { + "epoch": 2.958967851099831, + "grad_norm": 8.625, + "learning_rate": 1.884643740324645e-06, + "loss": 2.4121, + "step": 27980 + }, + { + "epoch": 2.959073604060914, + "grad_norm": 8.625, + "learning_rate": 1.884477754474714e-06, + "loss": 2.6413, + "step": 27981 + }, + { + "epoch": 2.9591793570219966, + "grad_norm": 8.3125, + "learning_rate": 1.8843117715131782e-06, + "loss": 2.5514, + "step": 27982 + }, + { + "epoch": 2.9592851099830795, + "grad_norm": 8.75, + "learning_rate": 1.8841457914408142e-06, + "loss": 2.7005, + "step": 27983 + }, + { + "epoch": 2.9593908629441623, + "grad_norm": 7.9375, + "learning_rate": 1.8839798142584028e-06, + "loss": 2.5285, + "step": 27984 + }, + { + "epoch": 2.9594966159052456, + "grad_norm": 8.0, + "learning_rate": 1.8838138399667216e-06, + "loss": 2.418, + "step": 27985 + }, + { + "epoch": 2.959602368866328, + "grad_norm": 8.3125, + "learning_rate": 1.8836478685665507e-06, + "loss": 2.2566, + "step": 27986 + }, + { + "epoch": 2.9597081218274113, + "grad_norm": 8.375, + "learning_rate": 1.8834819000586679e-06, + "loss": 2.4976, + "step": 27987 + }, + { + "epoch": 2.959813874788494, + "grad_norm": 8.625, + "learning_rate": 1.8833159344438531e-06, + "loss": 2.6082, + "step": 27988 + }, + { + "epoch": 2.959919627749577, + "grad_norm": 8.625, + "learning_rate": 1.8831499717228835e-06, + "loss": 2.4391, + "step": 27989 + }, + { + "epoch": 2.96002538071066, + "grad_norm": 8.625, + "learning_rate": 1.8829840118965381e-06, + "loss": 2.2916, + "step": 27990 + }, + { + "epoch": 2.9601311336717426, + "grad_norm": 8.4375, + "learning_rate": 1.8828180549655983e-06, + "loss": 2.5519, + "step": 27991 + }, + { + "epoch": 2.960236886632826, + "grad_norm": 8.1875, + "learning_rate": 1.8826521009308395e-06, + "loss": 2.6524, + "step": 27992 + }, + { + "epoch": 2.9603426395939088, + "grad_norm": 8.3125, + "learning_rate": 1.882486149793043e-06, + "loss": 2.3691, + "step": 27993 + }, + { + "epoch": 2.9604483925549916, + "grad_norm": 7.8125, + "learning_rate": 1.8823202015529853e-06, + "loss": 2.4485, + "step": 27994 + }, + { + "epoch": 2.9605541455160744, + "grad_norm": 9.1875, + "learning_rate": 1.8821542562114478e-06, + "loss": 2.4232, + "step": 27995 + }, + { + "epoch": 2.9606598984771573, + "grad_norm": 8.5, + "learning_rate": 1.8819883137692059e-06, + "loss": 2.5053, + "step": 27996 + }, + { + "epoch": 2.96076565143824, + "grad_norm": 8.0, + "learning_rate": 1.8818223742270417e-06, + "loss": 2.5133, + "step": 27997 + }, + { + "epoch": 2.960871404399323, + "grad_norm": 8.125, + "learning_rate": 1.881656437585731e-06, + "loss": 2.5143, + "step": 27998 + }, + { + "epoch": 2.9609771573604062, + "grad_norm": 8.5625, + "learning_rate": 1.881490503846054e-06, + "loss": 2.7021, + "step": 27999 + }, + { + "epoch": 2.961082910321489, + "grad_norm": 8.3125, + "learning_rate": 1.8813245730087894e-06, + "loss": 2.4724, + "step": 28000 + }, + { + "epoch": 2.961188663282572, + "grad_norm": 8.8125, + "learning_rate": 1.8811586450747154e-06, + "loss": 2.242, + "step": 28001 + }, + { + "epoch": 2.9612944162436547, + "grad_norm": 8.6875, + "learning_rate": 1.880992720044611e-06, + "loss": 2.5645, + "step": 28002 + }, + { + "epoch": 2.9614001692047376, + "grad_norm": 8.4375, + "learning_rate": 1.8808267979192541e-06, + "loss": 2.6362, + "step": 28003 + }, + { + "epoch": 2.961505922165821, + "grad_norm": 8.5625, + "learning_rate": 1.8806608786994252e-06, + "loss": 2.6381, + "step": 28004 + }, + { + "epoch": 2.9616116751269037, + "grad_norm": 8.0, + "learning_rate": 1.8804949623859001e-06, + "loss": 2.5073, + "step": 28005 + }, + { + "epoch": 2.9617174280879865, + "grad_norm": 8.875, + "learning_rate": 1.8803290489794593e-06, + "loss": 2.4546, + "step": 28006 + }, + { + "epoch": 2.9618231810490694, + "grad_norm": 8.4375, + "learning_rate": 1.8801631384808802e-06, + "loss": 2.6443, + "step": 28007 + }, + { + "epoch": 2.9619289340101522, + "grad_norm": 8.0, + "learning_rate": 1.8799972308909422e-06, + "loss": 2.0936, + "step": 28008 + }, + { + "epoch": 2.962034686971235, + "grad_norm": 7.75, + "learning_rate": 1.8798313262104243e-06, + "loss": 2.5745, + "step": 28009 + }, + { + "epoch": 2.962140439932318, + "grad_norm": 8.125, + "learning_rate": 1.8796654244401036e-06, + "loss": 2.4686, + "step": 28010 + }, + { + "epoch": 2.962246192893401, + "grad_norm": 8.5625, + "learning_rate": 1.8794995255807601e-06, + "loss": 2.5903, + "step": 28011 + }, + { + "epoch": 2.962351945854484, + "grad_norm": 8.75, + "learning_rate": 1.8793336296331705e-06, + "loss": 2.8035, + "step": 28012 + }, + { + "epoch": 2.962457698815567, + "grad_norm": 8.5625, + "learning_rate": 1.8791677365981156e-06, + "loss": 2.6141, + "step": 28013 + }, + { + "epoch": 2.9625634517766497, + "grad_norm": 8.75, + "learning_rate": 1.8790018464763718e-06, + "loss": 2.7695, + "step": 28014 + }, + { + "epoch": 2.9626692047377325, + "grad_norm": 9.0, + "learning_rate": 1.8788359592687189e-06, + "loss": 2.6801, + "step": 28015 + }, + { + "epoch": 2.962774957698816, + "grad_norm": 7.84375, + "learning_rate": 1.878670074975934e-06, + "loss": 2.5347, + "step": 28016 + }, + { + "epoch": 2.962880710659898, + "grad_norm": 8.125, + "learning_rate": 1.8785041935987963e-06, + "loss": 2.2771, + "step": 28017 + }, + { + "epoch": 2.9629864636209815, + "grad_norm": 8.8125, + "learning_rate": 1.8783383151380848e-06, + "loss": 2.5139, + "step": 28018 + }, + { + "epoch": 2.9630922165820643, + "grad_norm": 8.5625, + "learning_rate": 1.8781724395945773e-06, + "loss": 2.3111, + "step": 28019 + }, + { + "epoch": 2.963197969543147, + "grad_norm": 8.8125, + "learning_rate": 1.8780065669690529e-06, + "loss": 2.7504, + "step": 28020 + }, + { + "epoch": 2.96330372250423, + "grad_norm": 8.5625, + "learning_rate": 1.8778406972622882e-06, + "loss": 2.5327, + "step": 28021 + }, + { + "epoch": 2.963409475465313, + "grad_norm": 8.5, + "learning_rate": 1.877674830475063e-06, + "loss": 2.766, + "step": 28022 + }, + { + "epoch": 2.963515228426396, + "grad_norm": 8.6875, + "learning_rate": 1.877508966608155e-06, + "loss": 2.4466, + "step": 28023 + }, + { + "epoch": 2.963620981387479, + "grad_norm": 9.25, + "learning_rate": 1.8773431056623431e-06, + "loss": 2.409, + "step": 28024 + }, + { + "epoch": 2.963726734348562, + "grad_norm": 8.0, + "learning_rate": 1.8771772476384053e-06, + "loss": 2.5179, + "step": 28025 + }, + { + "epoch": 2.9638324873096447, + "grad_norm": 8.125, + "learning_rate": 1.8770113925371197e-06, + "loss": 2.4085, + "step": 28026 + }, + { + "epoch": 2.9639382402707275, + "grad_norm": 8.3125, + "learning_rate": 1.8768455403592658e-06, + "loss": 2.5969, + "step": 28027 + }, + { + "epoch": 2.9640439932318103, + "grad_norm": 7.75, + "learning_rate": 1.8766796911056196e-06, + "loss": 2.3611, + "step": 28028 + }, + { + "epoch": 2.964149746192893, + "grad_norm": 8.4375, + "learning_rate": 1.8765138447769621e-06, + "loss": 2.5437, + "step": 28029 + }, + { + "epoch": 2.9642554991539765, + "grad_norm": 8.6875, + "learning_rate": 1.876348001374069e-06, + "loss": 2.6615, + "step": 28030 + }, + { + "epoch": 2.9643612521150593, + "grad_norm": 8.125, + "learning_rate": 1.8761821608977203e-06, + "loss": 2.4676, + "step": 28031 + }, + { + "epoch": 2.964467005076142, + "grad_norm": 8.875, + "learning_rate": 1.8760163233486927e-06, + "loss": 2.5394, + "step": 28032 + }, + { + "epoch": 2.964572758037225, + "grad_norm": 8.6875, + "learning_rate": 1.8758504887277668e-06, + "loss": 2.6616, + "step": 28033 + }, + { + "epoch": 2.964678510998308, + "grad_norm": 8.375, + "learning_rate": 1.8756846570357177e-06, + "loss": 2.464, + "step": 28034 + }, + { + "epoch": 2.964784263959391, + "grad_norm": 8.5625, + "learning_rate": 1.8755188282733254e-06, + "loss": 2.767, + "step": 28035 + }, + { + "epoch": 2.964890016920474, + "grad_norm": 8.0625, + "learning_rate": 1.875353002441369e-06, + "loss": 2.3951, + "step": 28036 + }, + { + "epoch": 2.9649957698815568, + "grad_norm": 9.0, + "learning_rate": 1.8751871795406245e-06, + "loss": 2.6086, + "step": 28037 + }, + { + "epoch": 2.9651015228426396, + "grad_norm": 9.0625, + "learning_rate": 1.8750213595718719e-06, + "loss": 2.6575, + "step": 28038 + }, + { + "epoch": 2.9652072758037225, + "grad_norm": 8.5, + "learning_rate": 1.8748555425358874e-06, + "loss": 2.5967, + "step": 28039 + }, + { + "epoch": 2.9653130287648053, + "grad_norm": 8.1875, + "learning_rate": 1.874689728433451e-06, + "loss": 2.3893, + "step": 28040 + }, + { + "epoch": 2.965418781725888, + "grad_norm": 8.25, + "learning_rate": 1.874523917265339e-06, + "loss": 2.4361, + "step": 28041 + }, + { + "epoch": 2.9655245346869714, + "grad_norm": 8.0625, + "learning_rate": 1.874358109032332e-06, + "loss": 2.482, + "step": 28042 + }, + { + "epoch": 2.9656302876480543, + "grad_norm": 8.375, + "learning_rate": 1.8741923037352052e-06, + "loss": 2.4812, + "step": 28043 + }, + { + "epoch": 2.965736040609137, + "grad_norm": 8.0625, + "learning_rate": 1.8740265013747378e-06, + "loss": 2.4814, + "step": 28044 + }, + { + "epoch": 2.96584179357022, + "grad_norm": 8.4375, + "learning_rate": 1.873860701951709e-06, + "loss": 2.6012, + "step": 28045 + }, + { + "epoch": 2.9659475465313028, + "grad_norm": 8.3125, + "learning_rate": 1.8736949054668951e-06, + "loss": 2.4805, + "step": 28046 + }, + { + "epoch": 2.966053299492386, + "grad_norm": 8.5, + "learning_rate": 1.8735291119210752e-06, + "loss": 2.5741, + "step": 28047 + }, + { + "epoch": 2.9661590524534684, + "grad_norm": 8.125, + "learning_rate": 1.8733633213150268e-06, + "loss": 2.64, + "step": 28048 + }, + { + "epoch": 2.9662648054145517, + "grad_norm": 8.25, + "learning_rate": 1.873197533649529e-06, + "loss": 2.4022, + "step": 28049 + }, + { + "epoch": 2.9663705583756346, + "grad_norm": 8.1875, + "learning_rate": 1.8730317489253569e-06, + "loss": 2.4415, + "step": 28050 + }, + { + "epoch": 2.9664763113367174, + "grad_norm": 8.3125, + "learning_rate": 1.8728659671432926e-06, + "loss": 2.3902, + "step": 28051 + }, + { + "epoch": 2.9665820642978002, + "grad_norm": 8.0625, + "learning_rate": 1.8727001883041102e-06, + "loss": 2.3553, + "step": 28052 + }, + { + "epoch": 2.966687817258883, + "grad_norm": 8.4375, + "learning_rate": 1.8725344124085892e-06, + "loss": 2.639, + "step": 28053 + }, + { + "epoch": 2.9667935702199664, + "grad_norm": 8.0625, + "learning_rate": 1.8723686394575087e-06, + "loss": 2.238, + "step": 28054 + }, + { + "epoch": 2.966899323181049, + "grad_norm": 7.6875, + "learning_rate": 1.8722028694516445e-06, + "loss": 2.4397, + "step": 28055 + }, + { + "epoch": 2.967005076142132, + "grad_norm": 8.25, + "learning_rate": 1.872037102391776e-06, + "loss": 2.4262, + "step": 28056 + }, + { + "epoch": 2.967110829103215, + "grad_norm": 8.25, + "learning_rate": 1.8718713382786802e-06, + "loss": 2.4786, + "step": 28057 + }, + { + "epoch": 2.9672165820642977, + "grad_norm": 8.25, + "learning_rate": 1.8717055771131365e-06, + "loss": 2.6546, + "step": 28058 + }, + { + "epoch": 2.9673223350253806, + "grad_norm": 8.6875, + "learning_rate": 1.8715398188959199e-06, + "loss": 2.4324, + "step": 28059 + }, + { + "epoch": 2.9674280879864634, + "grad_norm": 9.75, + "learning_rate": 1.8713740636278098e-06, + "loss": 2.8897, + "step": 28060 + }, + { + "epoch": 2.9675338409475467, + "grad_norm": 8.125, + "learning_rate": 1.8712083113095858e-06, + "loss": 2.4862, + "step": 28061 + }, + { + "epoch": 2.9676395939086295, + "grad_norm": 8.4375, + "learning_rate": 1.8710425619420231e-06, + "loss": 2.5362, + "step": 28062 + }, + { + "epoch": 2.9677453468697124, + "grad_norm": 8.1875, + "learning_rate": 1.8708768155259005e-06, + "loss": 2.4247, + "step": 28063 + }, + { + "epoch": 2.967851099830795, + "grad_norm": 8.25, + "learning_rate": 1.8707110720619953e-06, + "loss": 2.5048, + "step": 28064 + }, + { + "epoch": 2.967956852791878, + "grad_norm": 9.0625, + "learning_rate": 1.8705453315510868e-06, + "loss": 2.5194, + "step": 28065 + }, + { + "epoch": 2.9680626057529613, + "grad_norm": 8.25, + "learning_rate": 1.87037959399395e-06, + "loss": 2.6052, + "step": 28066 + }, + { + "epoch": 2.9681683587140437, + "grad_norm": 8.1875, + "learning_rate": 1.870213859391366e-06, + "loss": 2.4907, + "step": 28067 + }, + { + "epoch": 2.968274111675127, + "grad_norm": 8.9375, + "learning_rate": 1.8700481277441096e-06, + "loss": 2.4702, + "step": 28068 + }, + { + "epoch": 2.96837986463621, + "grad_norm": 8.9375, + "learning_rate": 1.8698823990529595e-06, + "loss": 2.3956, + "step": 28069 + }, + { + "epoch": 2.9684856175972927, + "grad_norm": 8.75, + "learning_rate": 1.8697166733186943e-06, + "loss": 2.3746, + "step": 28070 + }, + { + "epoch": 2.9685913705583755, + "grad_norm": 7.78125, + "learning_rate": 1.8695509505420905e-06, + "loss": 2.3897, + "step": 28071 + }, + { + "epoch": 2.9686971235194584, + "grad_norm": 8.75, + "learning_rate": 1.869385230723927e-06, + "loss": 2.6446, + "step": 28072 + }, + { + "epoch": 2.9688028764805416, + "grad_norm": 8.625, + "learning_rate": 1.8692195138649799e-06, + "loss": 2.6044, + "step": 28073 + }, + { + "epoch": 2.9689086294416245, + "grad_norm": 8.5625, + "learning_rate": 1.8690537999660286e-06, + "loss": 2.7119, + "step": 28074 + }, + { + "epoch": 2.9690143824027073, + "grad_norm": 9.375, + "learning_rate": 1.8688880890278489e-06, + "loss": 2.6443, + "step": 28075 + }, + { + "epoch": 2.96912013536379, + "grad_norm": 8.6875, + "learning_rate": 1.8687223810512198e-06, + "loss": 2.5742, + "step": 28076 + }, + { + "epoch": 2.969225888324873, + "grad_norm": 8.3125, + "learning_rate": 1.8685566760369179e-06, + "loss": 2.5715, + "step": 28077 + }, + { + "epoch": 2.9693316412859563, + "grad_norm": 8.375, + "learning_rate": 1.868390973985721e-06, + "loss": 2.8251, + "step": 28078 + }, + { + "epoch": 2.9694373942470387, + "grad_norm": 8.0625, + "learning_rate": 1.8682252748984076e-06, + "loss": 2.4843, + "step": 28079 + }, + { + "epoch": 2.969543147208122, + "grad_norm": 8.75, + "learning_rate": 1.8680595787757541e-06, + "loss": 2.5196, + "step": 28080 + }, + { + "epoch": 2.969648900169205, + "grad_norm": 9.5, + "learning_rate": 1.8678938856185398e-06, + "loss": 2.831, + "step": 28081 + }, + { + "epoch": 2.9697546531302876, + "grad_norm": 8.625, + "learning_rate": 1.8677281954275388e-06, + "loss": 2.5421, + "step": 28082 + }, + { + "epoch": 2.9698604060913705, + "grad_norm": 8.3125, + "learning_rate": 1.8675625082035328e-06, + "loss": 2.3352, + "step": 28083 + }, + { + "epoch": 2.9699661590524533, + "grad_norm": 8.25, + "learning_rate": 1.8673968239472962e-06, + "loss": 2.5188, + "step": 28084 + }, + { + "epoch": 2.9700719120135366, + "grad_norm": 8.5, + "learning_rate": 1.8672311426596083e-06, + "loss": 2.6015, + "step": 28085 + }, + { + "epoch": 2.9701776649746194, + "grad_norm": 8.4375, + "learning_rate": 1.867065464341245e-06, + "loss": 2.49, + "step": 28086 + }, + { + "epoch": 2.9702834179357023, + "grad_norm": 8.25, + "learning_rate": 1.8668997889929847e-06, + "loss": 2.5922, + "step": 28087 + }, + { + "epoch": 2.970389170896785, + "grad_norm": 8.25, + "learning_rate": 1.8667341166156055e-06, + "loss": 2.6703, + "step": 28088 + }, + { + "epoch": 2.970494923857868, + "grad_norm": 8.625, + "learning_rate": 1.8665684472098834e-06, + "loss": 2.5889, + "step": 28089 + }, + { + "epoch": 2.970600676818951, + "grad_norm": 8.1875, + "learning_rate": 1.8664027807765975e-06, + "loss": 2.282, + "step": 28090 + }, + { + "epoch": 2.9707064297800336, + "grad_norm": 8.3125, + "learning_rate": 1.8662371173165232e-06, + "loss": 2.5144, + "step": 28091 + }, + { + "epoch": 2.970812182741117, + "grad_norm": 8.1875, + "learning_rate": 1.8660714568304394e-06, + "loss": 2.4583, + "step": 28092 + }, + { + "epoch": 2.9709179357021998, + "grad_norm": 8.375, + "learning_rate": 1.8659057993191226e-06, + "loss": 2.5457, + "step": 28093 + }, + { + "epoch": 2.9710236886632826, + "grad_norm": 8.625, + "learning_rate": 1.8657401447833508e-06, + "loss": 2.5901, + "step": 28094 + }, + { + "epoch": 2.9711294416243654, + "grad_norm": 8.3125, + "learning_rate": 1.865574493223901e-06, + "loss": 2.5328, + "step": 28095 + }, + { + "epoch": 2.9712351945854483, + "grad_norm": 8.5, + "learning_rate": 1.8654088446415502e-06, + "loss": 2.3203, + "step": 28096 + }, + { + "epoch": 2.9713409475465316, + "grad_norm": 8.3125, + "learning_rate": 1.8652431990370772e-06, + "loss": 2.4801, + "step": 28097 + }, + { + "epoch": 2.971446700507614, + "grad_norm": 8.25, + "learning_rate": 1.8650775564112572e-06, + "loss": 2.5203, + "step": 28098 + }, + { + "epoch": 2.9715524534686972, + "grad_norm": 8.25, + "learning_rate": 1.8649119167648699e-06, + "loss": 2.5915, + "step": 28099 + }, + { + "epoch": 2.97165820642978, + "grad_norm": 8.875, + "learning_rate": 1.8647462800986898e-06, + "loss": 2.5144, + "step": 28100 + }, + { + "epoch": 2.971763959390863, + "grad_norm": 8.375, + "learning_rate": 1.8645806464134963e-06, + "loss": 2.6506, + "step": 28101 + }, + { + "epoch": 2.9718697123519457, + "grad_norm": 8.5625, + "learning_rate": 1.8644150157100656e-06, + "loss": 2.531, + "step": 28102 + }, + { + "epoch": 2.9719754653130286, + "grad_norm": 9.3125, + "learning_rate": 1.8642493879891765e-06, + "loss": 2.5379, + "step": 28103 + }, + { + "epoch": 2.972081218274112, + "grad_norm": 9.625, + "learning_rate": 1.8640837632516032e-06, + "loss": 2.4612, + "step": 28104 + }, + { + "epoch": 2.9721869712351947, + "grad_norm": 8.5625, + "learning_rate": 1.863918141498125e-06, + "loss": 2.6516, + "step": 28105 + }, + { + "epoch": 2.9722927241962775, + "grad_norm": 8.4375, + "learning_rate": 1.8637525227295202e-06, + "loss": 2.6139, + "step": 28106 + }, + { + "epoch": 2.9723984771573604, + "grad_norm": 8.6875, + "learning_rate": 1.8635869069465633e-06, + "loss": 2.4707, + "step": 28107 + }, + { + "epoch": 2.9725042301184432, + "grad_norm": 8.5, + "learning_rate": 1.8634212941500337e-06, + "loss": 2.6927, + "step": 28108 + }, + { + "epoch": 2.972609983079526, + "grad_norm": 8.4375, + "learning_rate": 1.8632556843407067e-06, + "loss": 2.556, + "step": 28109 + }, + { + "epoch": 2.972715736040609, + "grad_norm": 8.5625, + "learning_rate": 1.8630900775193611e-06, + "loss": 2.544, + "step": 28110 + }, + { + "epoch": 2.972821489001692, + "grad_norm": 8.25, + "learning_rate": 1.8629244736867728e-06, + "loss": 2.5258, + "step": 28111 + }, + { + "epoch": 2.972927241962775, + "grad_norm": 8.5, + "learning_rate": 1.86275887284372e-06, + "loss": 2.6901, + "step": 28112 + }, + { + "epoch": 2.973032994923858, + "grad_norm": 8.5, + "learning_rate": 1.8625932749909787e-06, + "loss": 2.5204, + "step": 28113 + }, + { + "epoch": 2.9731387478849407, + "grad_norm": 8.5, + "learning_rate": 1.8624276801293256e-06, + "loss": 2.5195, + "step": 28114 + }, + { + "epoch": 2.9732445008460235, + "grad_norm": 8.9375, + "learning_rate": 1.8622620882595405e-06, + "loss": 2.792, + "step": 28115 + }, + { + "epoch": 2.973350253807107, + "grad_norm": 8.9375, + "learning_rate": 1.8620964993823973e-06, + "loss": 2.6616, + "step": 28116 + }, + { + "epoch": 2.9734560067681897, + "grad_norm": 8.125, + "learning_rate": 1.861930913498675e-06, + "loss": 2.5516, + "step": 28117 + }, + { + "epoch": 2.9735617597292725, + "grad_norm": 8.625, + "learning_rate": 1.8617653306091494e-06, + "loss": 2.613, + "step": 28118 + }, + { + "epoch": 2.9736675126903553, + "grad_norm": 8.1875, + "learning_rate": 1.861599750714599e-06, + "loss": 2.5949, + "step": 28119 + }, + { + "epoch": 2.973773265651438, + "grad_norm": 8.5625, + "learning_rate": 1.8614341738157982e-06, + "loss": 2.4257, + "step": 28120 + }, + { + "epoch": 2.973879018612521, + "grad_norm": 8.25, + "learning_rate": 1.8612685999135274e-06, + "loss": 2.3876, + "step": 28121 + }, + { + "epoch": 2.973984771573604, + "grad_norm": 8.5, + "learning_rate": 1.8611030290085607e-06, + "loss": 2.5227, + "step": 28122 + }, + { + "epoch": 2.974090524534687, + "grad_norm": 8.0625, + "learning_rate": 1.8609374611016762e-06, + "loss": 2.4065, + "step": 28123 + }, + { + "epoch": 2.97419627749577, + "grad_norm": 8.8125, + "learning_rate": 1.8607718961936516e-06, + "loss": 2.4931, + "step": 28124 + }, + { + "epoch": 2.974302030456853, + "grad_norm": 9.25, + "learning_rate": 1.8606063342852624e-06, + "loss": 2.7613, + "step": 28125 + }, + { + "epoch": 2.9744077834179357, + "grad_norm": 8.125, + "learning_rate": 1.8604407753772865e-06, + "loss": 2.3194, + "step": 28126 + }, + { + "epoch": 2.9745135363790185, + "grad_norm": 8.5625, + "learning_rate": 1.8602752194705001e-06, + "loss": 2.4353, + "step": 28127 + }, + { + "epoch": 2.974619289340102, + "grad_norm": 8.125, + "learning_rate": 1.8601096665656815e-06, + "loss": 2.48, + "step": 28128 + }, + { + "epoch": 2.974725042301184, + "grad_norm": 9.5, + "learning_rate": 1.8599441166636055e-06, + "loss": 2.7645, + "step": 28129 + }, + { + "epoch": 2.9748307952622675, + "grad_norm": 8.3125, + "learning_rate": 1.8597785697650506e-06, + "loss": 2.3539, + "step": 28130 + }, + { + "epoch": 2.9749365482233503, + "grad_norm": 8.4375, + "learning_rate": 1.8596130258707923e-06, + "loss": 2.4217, + "step": 28131 + }, + { + "epoch": 2.975042301184433, + "grad_norm": 8.5625, + "learning_rate": 1.8594474849816083e-06, + "loss": 2.605, + "step": 28132 + }, + { + "epoch": 2.975148054145516, + "grad_norm": 8.5, + "learning_rate": 1.8592819470982757e-06, + "loss": 2.5689, + "step": 28133 + }, + { + "epoch": 2.975253807106599, + "grad_norm": 8.875, + "learning_rate": 1.8591164122215704e-06, + "loss": 2.3258, + "step": 28134 + }, + { + "epoch": 2.975359560067682, + "grad_norm": 8.3125, + "learning_rate": 1.8589508803522706e-06, + "loss": 2.8515, + "step": 28135 + }, + { + "epoch": 2.975465313028765, + "grad_norm": 8.375, + "learning_rate": 1.8587853514911508e-06, + "loss": 2.437, + "step": 28136 + }, + { + "epoch": 2.9755710659898478, + "grad_norm": 8.8125, + "learning_rate": 1.8586198256389903e-06, + "loss": 2.323, + "step": 28137 + }, + { + "epoch": 2.9756768189509306, + "grad_norm": 9.1875, + "learning_rate": 1.858454302796564e-06, + "loss": 2.9039, + "step": 28138 + }, + { + "epoch": 2.9757825719120135, + "grad_norm": 9.125, + "learning_rate": 1.8582887829646496e-06, + "loss": 2.4839, + "step": 28139 + }, + { + "epoch": 2.9758883248730963, + "grad_norm": 8.6875, + "learning_rate": 1.8581232661440229e-06, + "loss": 2.5952, + "step": 28140 + }, + { + "epoch": 2.975994077834179, + "grad_norm": 8.625, + "learning_rate": 1.8579577523354614e-06, + "loss": 2.6076, + "step": 28141 + }, + { + "epoch": 2.9760998307952624, + "grad_norm": 8.375, + "learning_rate": 1.857792241539742e-06, + "loss": 2.4361, + "step": 28142 + }, + { + "epoch": 2.9762055837563453, + "grad_norm": 8.1875, + "learning_rate": 1.8576267337576406e-06, + "loss": 2.5309, + "step": 28143 + }, + { + "epoch": 2.976311336717428, + "grad_norm": 8.3125, + "learning_rate": 1.8574612289899352e-06, + "loss": 2.2375, + "step": 28144 + }, + { + "epoch": 2.976417089678511, + "grad_norm": 8.1875, + "learning_rate": 1.8572957272374003e-06, + "loss": 2.7108, + "step": 28145 + }, + { + "epoch": 2.9765228426395938, + "grad_norm": 9.25, + "learning_rate": 1.857130228500814e-06, + "loss": 2.7264, + "step": 28146 + }, + { + "epoch": 2.976628595600677, + "grad_norm": 8.1875, + "learning_rate": 1.8569647327809527e-06, + "loss": 2.396, + "step": 28147 + }, + { + "epoch": 2.97673434856176, + "grad_norm": 8.75, + "learning_rate": 1.8567992400785927e-06, + "loss": 2.7105, + "step": 28148 + }, + { + "epoch": 2.9768401015228427, + "grad_norm": 8.25, + "learning_rate": 1.8566337503945109e-06, + "loss": 2.3489, + "step": 28149 + }, + { + "epoch": 2.9769458544839256, + "grad_norm": 9.25, + "learning_rate": 1.8564682637294835e-06, + "loss": 2.6522, + "step": 28150 + }, + { + "epoch": 2.9770516074450084, + "grad_norm": 8.625, + "learning_rate": 1.8563027800842883e-06, + "loss": 2.5534, + "step": 28151 + }, + { + "epoch": 2.9771573604060912, + "grad_norm": 8.0625, + "learning_rate": 1.8561372994596994e-06, + "loss": 2.5459, + "step": 28152 + }, + { + "epoch": 2.977263113367174, + "grad_norm": 8.0625, + "learning_rate": 1.8559718218564965e-06, + "loss": 2.304, + "step": 28153 + }, + { + "epoch": 2.9773688663282574, + "grad_norm": 8.3125, + "learning_rate": 1.8558063472754533e-06, + "loss": 2.4818, + "step": 28154 + }, + { + "epoch": 2.97747461928934, + "grad_norm": 8.375, + "learning_rate": 1.8556408757173484e-06, + "loss": 2.5226, + "step": 28155 + }, + { + "epoch": 2.977580372250423, + "grad_norm": 8.25, + "learning_rate": 1.8554754071829563e-06, + "loss": 2.5876, + "step": 28156 + }, + { + "epoch": 2.977686125211506, + "grad_norm": 8.625, + "learning_rate": 1.8553099416730556e-06, + "loss": 2.4556, + "step": 28157 + }, + { + "epoch": 2.9777918781725887, + "grad_norm": 8.25, + "learning_rate": 1.8551444791884199e-06, + "loss": 2.388, + "step": 28158 + }, + { + "epoch": 2.977897631133672, + "grad_norm": 8.625, + "learning_rate": 1.8549790197298284e-06, + "loss": 2.3957, + "step": 28159 + }, + { + "epoch": 2.9780033840947544, + "grad_norm": 8.375, + "learning_rate": 1.8548135632980573e-06, + "loss": 2.5489, + "step": 28160 + }, + { + "epoch": 2.9781091370558377, + "grad_norm": 7.9375, + "learning_rate": 1.8546481098938814e-06, + "loss": 2.1753, + "step": 28161 + }, + { + "epoch": 2.9782148900169205, + "grad_norm": 8.3125, + "learning_rate": 1.8544826595180787e-06, + "loss": 2.3787, + "step": 28162 + }, + { + "epoch": 2.9783206429780034, + "grad_norm": 8.3125, + "learning_rate": 1.854317212171424e-06, + "loss": 2.2366, + "step": 28163 + }, + { + "epoch": 2.978426395939086, + "grad_norm": 8.5, + "learning_rate": 1.854151767854695e-06, + "loss": 2.4566, + "step": 28164 + }, + { + "epoch": 2.978532148900169, + "grad_norm": 8.25, + "learning_rate": 1.8539863265686675e-06, + "loss": 2.5596, + "step": 28165 + }, + { + "epoch": 2.9786379018612523, + "grad_norm": 8.8125, + "learning_rate": 1.8538208883141187e-06, + "loss": 2.6632, + "step": 28166 + }, + { + "epoch": 2.978743654822335, + "grad_norm": 8.25, + "learning_rate": 1.8536554530918233e-06, + "loss": 2.447, + "step": 28167 + }, + { + "epoch": 2.978849407783418, + "grad_norm": 9.375, + "learning_rate": 1.8534900209025578e-06, + "loss": 2.506, + "step": 28168 + }, + { + "epoch": 2.978955160744501, + "grad_norm": 8.25, + "learning_rate": 1.853324591747101e-06, + "loss": 2.373, + "step": 28169 + }, + { + "epoch": 2.9790609137055837, + "grad_norm": 8.6875, + "learning_rate": 1.853159165626226e-06, + "loss": 2.4523, + "step": 28170 + }, + { + "epoch": 2.9791666666666665, + "grad_norm": 8.75, + "learning_rate": 1.8529937425407114e-06, + "loss": 2.4947, + "step": 28171 + }, + { + "epoch": 2.9792724196277494, + "grad_norm": 8.0625, + "learning_rate": 1.8528283224913317e-06, + "loss": 2.3115, + "step": 28172 + }, + { + "epoch": 2.9793781725888326, + "grad_norm": 8.375, + "learning_rate": 1.8526629054788653e-06, + "loss": 2.511, + "step": 28173 + }, + { + "epoch": 2.9794839255499155, + "grad_norm": 8.5625, + "learning_rate": 1.8524974915040856e-06, + "loss": 2.534, + "step": 28174 + }, + { + "epoch": 2.9795896785109983, + "grad_norm": 8.5, + "learning_rate": 1.8523320805677708e-06, + "loss": 2.67, + "step": 28175 + }, + { + "epoch": 2.979695431472081, + "grad_norm": 12.8125, + "learning_rate": 1.8521666726706977e-06, + "loss": 2.5936, + "step": 28176 + }, + { + "epoch": 2.979801184433164, + "grad_norm": 7.65625, + "learning_rate": 1.8520012678136404e-06, + "loss": 2.5739, + "step": 28177 + }, + { + "epoch": 2.9799069373942473, + "grad_norm": 8.1875, + "learning_rate": 1.8518358659973768e-06, + "loss": 2.4423, + "step": 28178 + }, + { + "epoch": 2.9800126903553297, + "grad_norm": 8.25, + "learning_rate": 1.8516704672226815e-06, + "loss": 2.4705, + "step": 28179 + }, + { + "epoch": 2.980118443316413, + "grad_norm": 8.25, + "learning_rate": 1.8515050714903322e-06, + "loss": 2.5008, + "step": 28180 + }, + { + "epoch": 2.980224196277496, + "grad_norm": 8.4375, + "learning_rate": 1.851339678801104e-06, + "loss": 2.681, + "step": 28181 + }, + { + "epoch": 2.9803299492385786, + "grad_norm": 8.0625, + "learning_rate": 1.8511742891557744e-06, + "loss": 2.5843, + "step": 28182 + }, + { + "epoch": 2.9804357021996615, + "grad_norm": 9.1875, + "learning_rate": 1.8510089025551175e-06, + "loss": 2.6196, + "step": 28183 + }, + { + "epoch": 2.9805414551607443, + "grad_norm": 8.25, + "learning_rate": 1.8508435189999097e-06, + "loss": 2.5071, + "step": 28184 + }, + { + "epoch": 2.9806472081218276, + "grad_norm": 8.5625, + "learning_rate": 1.8506781384909295e-06, + "loss": 2.6413, + "step": 28185 + }, + { + "epoch": 2.9807529610829104, + "grad_norm": 8.4375, + "learning_rate": 1.85051276102895e-06, + "loss": 2.5533, + "step": 28186 + }, + { + "epoch": 2.9808587140439933, + "grad_norm": 8.5, + "learning_rate": 1.8503473866147492e-06, + "loss": 2.4883, + "step": 28187 + }, + { + "epoch": 2.980964467005076, + "grad_norm": 8.5625, + "learning_rate": 1.8501820152491018e-06, + "loss": 2.7601, + "step": 28188 + }, + { + "epoch": 2.981070219966159, + "grad_norm": 8.4375, + "learning_rate": 1.8500166469327854e-06, + "loss": 2.5141, + "step": 28189 + }, + { + "epoch": 2.9811759729272422, + "grad_norm": 9.3125, + "learning_rate": 1.8498512816665737e-06, + "loss": 2.5384, + "step": 28190 + }, + { + "epoch": 2.9812817258883246, + "grad_norm": 8.3125, + "learning_rate": 1.8496859194512453e-06, + "loss": 2.6361, + "step": 28191 + }, + { + "epoch": 2.981387478849408, + "grad_norm": 8.375, + "learning_rate": 1.8495205602875741e-06, + "loss": 2.4945, + "step": 28192 + }, + { + "epoch": 2.9814932318104908, + "grad_norm": 7.9375, + "learning_rate": 1.8493552041763368e-06, + "loss": 2.4207, + "step": 28193 + }, + { + "epoch": 2.9815989847715736, + "grad_norm": 8.6875, + "learning_rate": 1.84918985111831e-06, + "loss": 2.5442, + "step": 28194 + }, + { + "epoch": 2.9817047377326564, + "grad_norm": 8.625, + "learning_rate": 1.8490245011142682e-06, + "loss": 2.6281, + "step": 28195 + }, + { + "epoch": 2.9818104906937393, + "grad_norm": 9.375, + "learning_rate": 1.8488591541649892e-06, + "loss": 2.5819, + "step": 28196 + }, + { + "epoch": 2.9819162436548226, + "grad_norm": 9.75, + "learning_rate": 1.848693810271247e-06, + "loss": 2.6652, + "step": 28197 + }, + { + "epoch": 2.9820219966159054, + "grad_norm": 8.5625, + "learning_rate": 1.8485284694338196e-06, + "loss": 2.6352, + "step": 28198 + }, + { + "epoch": 2.9821277495769882, + "grad_norm": 8.5625, + "learning_rate": 1.8483631316534806e-06, + "loss": 2.398, + "step": 28199 + }, + { + "epoch": 2.982233502538071, + "grad_norm": 8.25, + "learning_rate": 1.8481977969310076e-06, + "loss": 2.4463, + "step": 28200 + }, + { + "epoch": 2.982339255499154, + "grad_norm": 8.5625, + "learning_rate": 1.848032465267175e-06, + "loss": 2.5389, + "step": 28201 + }, + { + "epoch": 2.9824450084602367, + "grad_norm": 8.8125, + "learning_rate": 1.8478671366627594e-06, + "loss": 2.4875, + "step": 28202 + }, + { + "epoch": 2.9825507614213196, + "grad_norm": 8.625, + "learning_rate": 1.8477018111185368e-06, + "loss": 2.6719, + "step": 28203 + }, + { + "epoch": 2.982656514382403, + "grad_norm": 9.0625, + "learning_rate": 1.8475364886352829e-06, + "loss": 2.466, + "step": 28204 + }, + { + "epoch": 2.9827622673434857, + "grad_norm": 8.4375, + "learning_rate": 1.8473711692137742e-06, + "loss": 2.7536, + "step": 28205 + }, + { + "epoch": 2.9828680203045685, + "grad_norm": 8.0625, + "learning_rate": 1.8472058528547843e-06, + "loss": 2.5261, + "step": 28206 + }, + { + "epoch": 2.9829737732656514, + "grad_norm": 8.3125, + "learning_rate": 1.8470405395590918e-06, + "loss": 2.4689, + "step": 28207 + }, + { + "epoch": 2.9830795262267342, + "grad_norm": 8.3125, + "learning_rate": 1.8468752293274698e-06, + "loss": 2.6747, + "step": 28208 + }, + { + "epoch": 2.9831852791878175, + "grad_norm": 8.0, + "learning_rate": 1.846709922160696e-06, + "loss": 2.4493, + "step": 28209 + }, + { + "epoch": 2.9832910321489, + "grad_norm": 8.75, + "learning_rate": 1.8465446180595444e-06, + "loss": 2.4299, + "step": 28210 + }, + { + "epoch": 2.983396785109983, + "grad_norm": 8.1875, + "learning_rate": 1.846379317024792e-06, + "loss": 2.5436, + "step": 28211 + }, + { + "epoch": 2.983502538071066, + "grad_norm": 8.375, + "learning_rate": 1.8462140190572147e-06, + "loss": 2.6574, + "step": 28212 + }, + { + "epoch": 2.983608291032149, + "grad_norm": 8.3125, + "learning_rate": 1.846048724157587e-06, + "loss": 2.6667, + "step": 28213 + }, + { + "epoch": 2.9837140439932317, + "grad_norm": 8.25, + "learning_rate": 1.845883432326686e-06, + "loss": 2.7012, + "step": 28214 + }, + { + "epoch": 2.9838197969543145, + "grad_norm": 8.4375, + "learning_rate": 1.845718143565286e-06, + "loss": 2.5989, + "step": 28215 + }, + { + "epoch": 2.983925549915398, + "grad_norm": 8.4375, + "learning_rate": 1.8455528578741634e-06, + "loss": 2.6278, + "step": 28216 + }, + { + "epoch": 2.9840313028764807, + "grad_norm": 8.0625, + "learning_rate": 1.8453875752540928e-06, + "loss": 2.4602, + "step": 28217 + }, + { + "epoch": 2.9841370558375635, + "grad_norm": 8.5, + "learning_rate": 1.8452222957058514e-06, + "loss": 2.4286, + "step": 28218 + }, + { + "epoch": 2.9842428087986463, + "grad_norm": 8.25, + "learning_rate": 1.8450570192302136e-06, + "loss": 2.3214, + "step": 28219 + }, + { + "epoch": 2.984348561759729, + "grad_norm": 8.25, + "learning_rate": 1.844891745827955e-06, + "loss": 2.4678, + "step": 28220 + }, + { + "epoch": 2.984454314720812, + "grad_norm": 8.375, + "learning_rate": 1.8447264754998529e-06, + "loss": 2.5837, + "step": 28221 + }, + { + "epoch": 2.984560067681895, + "grad_norm": 8.375, + "learning_rate": 1.8445612082466797e-06, + "loss": 2.2521, + "step": 28222 + }, + { + "epoch": 2.984665820642978, + "grad_norm": 8.1875, + "learning_rate": 1.8443959440692138e-06, + "loss": 2.4684, + "step": 28223 + }, + { + "epoch": 2.984771573604061, + "grad_norm": 8.0, + "learning_rate": 1.8442306829682294e-06, + "loss": 2.5434, + "step": 28224 + }, + { + "epoch": 2.984877326565144, + "grad_norm": 9.75, + "learning_rate": 1.8440654249445023e-06, + "loss": 2.4759, + "step": 28225 + }, + { + "epoch": 2.9849830795262267, + "grad_norm": 9.75, + "learning_rate": 1.8439001699988074e-06, + "loss": 2.565, + "step": 28226 + }, + { + "epoch": 2.9850888324873095, + "grad_norm": 8.0625, + "learning_rate": 1.8437349181319214e-06, + "loss": 2.3227, + "step": 28227 + }, + { + "epoch": 2.985194585448393, + "grad_norm": 8.4375, + "learning_rate": 1.843569669344618e-06, + "loss": 2.5393, + "step": 28228 + }, + { + "epoch": 2.9853003384094756, + "grad_norm": 8.0625, + "learning_rate": 1.8434044236376738e-06, + "loss": 2.5479, + "step": 28229 + }, + { + "epoch": 2.9854060913705585, + "grad_norm": 8.5, + "learning_rate": 1.8432391810118656e-06, + "loss": 2.6149, + "step": 28230 + }, + { + "epoch": 2.9855118443316413, + "grad_norm": 8.0625, + "learning_rate": 1.8430739414679658e-06, + "loss": 2.2254, + "step": 28231 + }, + { + "epoch": 2.985617597292724, + "grad_norm": 8.0625, + "learning_rate": 1.8429087050067518e-06, + "loss": 2.4319, + "step": 28232 + }, + { + "epoch": 2.985723350253807, + "grad_norm": 8.0625, + "learning_rate": 1.8427434716289982e-06, + "loss": 2.3875, + "step": 28233 + }, + { + "epoch": 2.98582910321489, + "grad_norm": 8.0625, + "learning_rate": 1.8425782413354815e-06, + "loss": 2.4558, + "step": 28234 + }, + { + "epoch": 2.985934856175973, + "grad_norm": 8.1875, + "learning_rate": 1.8424130141269752e-06, + "loss": 2.4078, + "step": 28235 + }, + { + "epoch": 2.986040609137056, + "grad_norm": 8.1875, + "learning_rate": 1.8422477900042568e-06, + "loss": 2.3214, + "step": 28236 + }, + { + "epoch": 2.9861463620981388, + "grad_norm": 8.375, + "learning_rate": 1.8420825689681e-06, + "loss": 2.2958, + "step": 28237 + }, + { + "epoch": 2.9862521150592216, + "grad_norm": 8.25, + "learning_rate": 1.8419173510192796e-06, + "loss": 2.5379, + "step": 28238 + }, + { + "epoch": 2.9863578680203045, + "grad_norm": 8.25, + "learning_rate": 1.8417521361585736e-06, + "loss": 2.5164, + "step": 28239 + }, + { + "epoch": 2.9864636209813877, + "grad_norm": 7.75, + "learning_rate": 1.8415869243867545e-06, + "loss": 2.4852, + "step": 28240 + }, + { + "epoch": 2.98656937394247, + "grad_norm": 8.375, + "learning_rate": 1.8414217157045994e-06, + "loss": 2.5176, + "step": 28241 + }, + { + "epoch": 2.9866751269035534, + "grad_norm": 8.625, + "learning_rate": 1.841256510112882e-06, + "loss": 2.3435, + "step": 28242 + }, + { + "epoch": 2.9867808798646363, + "grad_norm": 8.375, + "learning_rate": 1.8410913076123796e-06, + "loss": 2.4693, + "step": 28243 + }, + { + "epoch": 2.986886632825719, + "grad_norm": 8.375, + "learning_rate": 1.8409261082038648e-06, + "loss": 2.5075, + "step": 28244 + }, + { + "epoch": 2.986992385786802, + "grad_norm": 9.6875, + "learning_rate": 1.8407609118881153e-06, + "loss": 2.6828, + "step": 28245 + }, + { + "epoch": 2.9870981387478848, + "grad_norm": 8.625, + "learning_rate": 1.8405957186659046e-06, + "loss": 2.7264, + "step": 28246 + }, + { + "epoch": 2.987203891708968, + "grad_norm": 8.1875, + "learning_rate": 1.8404305285380086e-06, + "loss": 2.5857, + "step": 28247 + }, + { + "epoch": 2.987309644670051, + "grad_norm": 8.875, + "learning_rate": 1.8402653415052028e-06, + "loss": 2.6452, + "step": 28248 + }, + { + "epoch": 2.9874153976311337, + "grad_norm": 8.75, + "learning_rate": 1.840100157568261e-06, + "loss": 2.5128, + "step": 28249 + }, + { + "epoch": 2.9875211505922166, + "grad_norm": 8.0, + "learning_rate": 1.8399349767279604e-06, + "loss": 2.1546, + "step": 28250 + }, + { + "epoch": 2.9876269035532994, + "grad_norm": 8.1875, + "learning_rate": 1.839769798985074e-06, + "loss": 2.5471, + "step": 28251 + }, + { + "epoch": 2.9877326565143822, + "grad_norm": 8.1875, + "learning_rate": 1.8396046243403793e-06, + "loss": 2.6703, + "step": 28252 + }, + { + "epoch": 2.987838409475465, + "grad_norm": 8.5625, + "learning_rate": 1.8394394527946487e-06, + "loss": 2.5977, + "step": 28253 + }, + { + "epoch": 2.9879441624365484, + "grad_norm": 8.3125, + "learning_rate": 1.8392742843486593e-06, + "loss": 2.5851, + "step": 28254 + }, + { + "epoch": 2.988049915397631, + "grad_norm": 8.25, + "learning_rate": 1.8391091190031849e-06, + "loss": 2.6109, + "step": 28255 + }, + { + "epoch": 2.988155668358714, + "grad_norm": 8.4375, + "learning_rate": 1.8389439567590011e-06, + "loss": 2.4122, + "step": 28256 + }, + { + "epoch": 2.988261421319797, + "grad_norm": 8.4375, + "learning_rate": 1.8387787976168832e-06, + "loss": 2.6449, + "step": 28257 + }, + { + "epoch": 2.9883671742808797, + "grad_norm": 8.5625, + "learning_rate": 1.8386136415776056e-06, + "loss": 2.6899, + "step": 28258 + }, + { + "epoch": 2.988472927241963, + "grad_norm": 9.125, + "learning_rate": 1.838448488641945e-06, + "loss": 2.3214, + "step": 28259 + }, + { + "epoch": 2.988578680203046, + "grad_norm": 8.625, + "learning_rate": 1.8382833388106736e-06, + "loss": 2.5102, + "step": 28260 + }, + { + "epoch": 2.9886844331641287, + "grad_norm": 9.0, + "learning_rate": 1.838118192084569e-06, + "loss": 2.5464, + "step": 28261 + }, + { + "epoch": 2.9887901861252115, + "grad_norm": 8.25, + "learning_rate": 1.8379530484644042e-06, + "loss": 2.6129, + "step": 28262 + }, + { + "epoch": 2.9888959390862944, + "grad_norm": 8.3125, + "learning_rate": 1.8377879079509553e-06, + "loss": 2.5583, + "step": 28263 + }, + { + "epoch": 2.989001692047377, + "grad_norm": 8.375, + "learning_rate": 1.8376227705449967e-06, + "loss": 2.6052, + "step": 28264 + }, + { + "epoch": 2.98910744500846, + "grad_norm": 8.5, + "learning_rate": 1.8374576362473032e-06, + "loss": 2.6095, + "step": 28265 + }, + { + "epoch": 2.9892131979695433, + "grad_norm": 7.9375, + "learning_rate": 1.8372925050586507e-06, + "loss": 2.5318, + "step": 28266 + }, + { + "epoch": 2.989318950930626, + "grad_norm": 8.875, + "learning_rate": 1.837127376979813e-06, + "loss": 2.5466, + "step": 28267 + }, + { + "epoch": 2.989424703891709, + "grad_norm": 8.1875, + "learning_rate": 1.8369622520115665e-06, + "loss": 2.364, + "step": 28268 + }, + { + "epoch": 2.989530456852792, + "grad_norm": 9.0, + "learning_rate": 1.8367971301546836e-06, + "loss": 2.8391, + "step": 28269 + }, + { + "epoch": 2.9896362098138747, + "grad_norm": 8.75, + "learning_rate": 1.8366320114099412e-06, + "loss": 2.7384, + "step": 28270 + }, + { + "epoch": 2.989741962774958, + "grad_norm": 8.4375, + "learning_rate": 1.8364668957781133e-06, + "loss": 2.4804, + "step": 28271 + }, + { + "epoch": 2.9898477157360404, + "grad_norm": 8.875, + "learning_rate": 1.836301783259975e-06, + "loss": 2.6186, + "step": 28272 + }, + { + "epoch": 2.9899534686971236, + "grad_norm": 8.6875, + "learning_rate": 1.8361366738563006e-06, + "loss": 2.7461, + "step": 28273 + }, + { + "epoch": 2.9900592216582065, + "grad_norm": 8.6875, + "learning_rate": 1.8359715675678652e-06, + "loss": 2.636, + "step": 28274 + }, + { + "epoch": 2.9901649746192893, + "grad_norm": 8.0, + "learning_rate": 1.8358064643954448e-06, + "loss": 2.2354, + "step": 28275 + }, + { + "epoch": 2.990270727580372, + "grad_norm": 8.5625, + "learning_rate": 1.8356413643398114e-06, + "loss": 2.7566, + "step": 28276 + }, + { + "epoch": 2.990376480541455, + "grad_norm": 8.125, + "learning_rate": 1.8354762674017429e-06, + "loss": 2.3799, + "step": 28277 + }, + { + "epoch": 2.9904822335025383, + "grad_norm": 8.75, + "learning_rate": 1.8353111735820114e-06, + "loss": 2.456, + "step": 28278 + }, + { + "epoch": 2.990587986463621, + "grad_norm": 8.125, + "learning_rate": 1.8351460828813934e-06, + "loss": 2.4409, + "step": 28279 + }, + { + "epoch": 2.990693739424704, + "grad_norm": 8.8125, + "learning_rate": 1.8349809953006622e-06, + "loss": 2.6071, + "step": 28280 + }, + { + "epoch": 2.990799492385787, + "grad_norm": 8.25, + "learning_rate": 1.8348159108405943e-06, + "loss": 2.6211, + "step": 28281 + }, + { + "epoch": 2.9909052453468696, + "grad_norm": 8.5625, + "learning_rate": 1.8346508295019621e-06, + "loss": 2.4621, + "step": 28282 + }, + { + "epoch": 2.9910109983079525, + "grad_norm": 9.3125, + "learning_rate": 1.8344857512855416e-06, + "loss": 2.4794, + "step": 28283 + }, + { + "epoch": 2.9911167512690353, + "grad_norm": 8.25, + "learning_rate": 1.8343206761921083e-06, + "loss": 2.3292, + "step": 28284 + }, + { + "epoch": 2.9912225042301186, + "grad_norm": 7.875, + "learning_rate": 1.834155604222435e-06, + "loss": 2.4095, + "step": 28285 + }, + { + "epoch": 2.9913282571912014, + "grad_norm": 8.3125, + "learning_rate": 1.8339905353772975e-06, + "loss": 2.4604, + "step": 28286 + }, + { + "epoch": 2.9914340101522843, + "grad_norm": 8.375, + "learning_rate": 1.8338254696574692e-06, + "loss": 2.3964, + "step": 28287 + }, + { + "epoch": 2.991539763113367, + "grad_norm": 8.4375, + "learning_rate": 1.8336604070637265e-06, + "loss": 2.5469, + "step": 28288 + }, + { + "epoch": 2.99164551607445, + "grad_norm": 8.5625, + "learning_rate": 1.8334953475968423e-06, + "loss": 2.7323, + "step": 28289 + }, + { + "epoch": 2.9917512690355332, + "grad_norm": 8.1875, + "learning_rate": 1.8333302912575918e-06, + "loss": 2.4687, + "step": 28290 + }, + { + "epoch": 2.9918570219966156, + "grad_norm": 8.3125, + "learning_rate": 1.8331652380467507e-06, + "loss": 2.5704, + "step": 28291 + }, + { + "epoch": 2.991962774957699, + "grad_norm": 8.4375, + "learning_rate": 1.8330001879650907e-06, + "loss": 2.3335, + "step": 28292 + }, + { + "epoch": 2.9920685279187818, + "grad_norm": 8.75, + "learning_rate": 1.8328351410133898e-06, + "loss": 2.2539, + "step": 28293 + }, + { + "epoch": 2.9921742808798646, + "grad_norm": 8.0625, + "learning_rate": 1.8326700971924193e-06, + "loss": 2.4676, + "step": 28294 + }, + { + "epoch": 2.9922800338409474, + "grad_norm": 8.3125, + "learning_rate": 1.8325050565029563e-06, + "loss": 2.5006, + "step": 28295 + }, + { + "epoch": 2.9923857868020303, + "grad_norm": 8.25, + "learning_rate": 1.832340018945773e-06, + "loss": 2.3408, + "step": 28296 + }, + { + "epoch": 2.9924915397631136, + "grad_norm": 9.0, + "learning_rate": 1.8321749845216464e-06, + "loss": 2.5371, + "step": 28297 + }, + { + "epoch": 2.9925972927241964, + "grad_norm": 8.5625, + "learning_rate": 1.8320099532313473e-06, + "loss": 2.5343, + "step": 28298 + }, + { + "epoch": 2.9927030456852792, + "grad_norm": 9.0, + "learning_rate": 1.8318449250756532e-06, + "loss": 2.5598, + "step": 28299 + }, + { + "epoch": 2.992808798646362, + "grad_norm": 8.5, + "learning_rate": 1.8316799000553387e-06, + "loss": 2.5437, + "step": 28300 + }, + { + "epoch": 2.992914551607445, + "grad_norm": 9.625, + "learning_rate": 1.8315148781711762e-06, + "loss": 2.7613, + "step": 28301 + }, + { + "epoch": 2.993020304568528, + "grad_norm": 8.75, + "learning_rate": 1.831349859423941e-06, + "loss": 2.6061, + "step": 28302 + }, + { + "epoch": 2.9931260575296106, + "grad_norm": 8.4375, + "learning_rate": 1.831184843814407e-06, + "loss": 2.6935, + "step": 28303 + }, + { + "epoch": 2.993231810490694, + "grad_norm": 9.75, + "learning_rate": 1.8310198313433496e-06, + "loss": 2.9245, + "step": 28304 + }, + { + "epoch": 2.9933375634517767, + "grad_norm": 8.5, + "learning_rate": 1.8308548220115424e-06, + "loss": 2.5445, + "step": 28305 + }, + { + "epoch": 2.9934433164128595, + "grad_norm": 8.875, + "learning_rate": 1.8306898158197605e-06, + "loss": 2.5215, + "step": 28306 + }, + { + "epoch": 2.9935490693739424, + "grad_norm": 8.5625, + "learning_rate": 1.8305248127687767e-06, + "loss": 2.5459, + "step": 28307 + }, + { + "epoch": 2.9936548223350252, + "grad_norm": 8.9375, + "learning_rate": 1.8303598128593653e-06, + "loss": 2.4933, + "step": 28308 + }, + { + "epoch": 2.9937605752961085, + "grad_norm": 9.3125, + "learning_rate": 1.8301948160923034e-06, + "loss": 2.5378, + "step": 28309 + }, + { + "epoch": 2.9938663282571913, + "grad_norm": 8.0625, + "learning_rate": 1.8300298224683618e-06, + "loss": 2.6694, + "step": 28310 + }, + { + "epoch": 2.993972081218274, + "grad_norm": 8.5, + "learning_rate": 1.8298648319883173e-06, + "loss": 2.4046, + "step": 28311 + }, + { + "epoch": 2.994077834179357, + "grad_norm": 8.25, + "learning_rate": 1.829699844652942e-06, + "loss": 2.721, + "step": 28312 + }, + { + "epoch": 2.99418358714044, + "grad_norm": 8.125, + "learning_rate": 1.8295348604630126e-06, + "loss": 2.5352, + "step": 28313 + }, + { + "epoch": 2.9942893401015227, + "grad_norm": 8.375, + "learning_rate": 1.8293698794193e-06, + "loss": 2.5548, + "step": 28314 + }, + { + "epoch": 2.9943950930626055, + "grad_norm": 8.6875, + "learning_rate": 1.8292049015225822e-06, + "loss": 2.5637, + "step": 28315 + }, + { + "epoch": 2.994500846023689, + "grad_norm": 8.375, + "learning_rate": 1.8290399267736303e-06, + "loss": 2.5002, + "step": 28316 + }, + { + "epoch": 2.9946065989847717, + "grad_norm": 8.6875, + "learning_rate": 1.8288749551732194e-06, + "loss": 2.659, + "step": 28317 + }, + { + "epoch": 2.9947123519458545, + "grad_norm": 8.8125, + "learning_rate": 1.8287099867221248e-06, + "loss": 2.621, + "step": 28318 + }, + { + "epoch": 2.9948181049069373, + "grad_norm": 8.375, + "learning_rate": 1.8285450214211186e-06, + "loss": 2.5364, + "step": 28319 + }, + { + "epoch": 2.99492385786802, + "grad_norm": 8.0625, + "learning_rate": 1.8283800592709772e-06, + "loss": 2.618, + "step": 28320 + }, + { + "epoch": 2.9950296108291035, + "grad_norm": 8.9375, + "learning_rate": 1.8282151002724723e-06, + "loss": 2.5652, + "step": 28321 + }, + { + "epoch": 2.995135363790186, + "grad_norm": 8.8125, + "learning_rate": 1.8280501444263806e-06, + "loss": 2.7168, + "step": 28322 + }, + { + "epoch": 2.995241116751269, + "grad_norm": 8.5, + "learning_rate": 1.8278851917334736e-06, + "loss": 2.5139, + "step": 28323 + }, + { + "epoch": 2.995346869712352, + "grad_norm": 8.0625, + "learning_rate": 1.8277202421945273e-06, + "loss": 2.471, + "step": 28324 + }, + { + "epoch": 2.995452622673435, + "grad_norm": 8.1875, + "learning_rate": 1.827555295810314e-06, + "loss": 2.4829, + "step": 28325 + }, + { + "epoch": 2.9955583756345177, + "grad_norm": 8.4375, + "learning_rate": 1.8273903525816088e-06, + "loss": 2.6246, + "step": 28326 + }, + { + "epoch": 2.9956641285956005, + "grad_norm": 8.6875, + "learning_rate": 1.8272254125091861e-06, + "loss": 2.4755, + "step": 28327 + }, + { + "epoch": 2.995769881556684, + "grad_norm": 8.625, + "learning_rate": 1.8270604755938194e-06, + "loss": 2.6905, + "step": 28328 + }, + { + "epoch": 2.9958756345177666, + "grad_norm": 8.375, + "learning_rate": 1.8268955418362829e-06, + "loss": 2.172, + "step": 28329 + }, + { + "epoch": 2.9959813874788495, + "grad_norm": 9.0, + "learning_rate": 1.8267306112373492e-06, + "loss": 2.7317, + "step": 28330 + }, + { + "epoch": 2.9960871404399323, + "grad_norm": 8.5625, + "learning_rate": 1.8265656837977949e-06, + "loss": 2.5336, + "step": 28331 + }, + { + "epoch": 2.996192893401015, + "grad_norm": 8.4375, + "learning_rate": 1.8264007595183913e-06, + "loss": 2.5424, + "step": 28332 + }, + { + "epoch": 2.996298646362098, + "grad_norm": 8.3125, + "learning_rate": 1.826235838399914e-06, + "loss": 2.5257, + "step": 28333 + }, + { + "epoch": 2.996404399323181, + "grad_norm": 8.6875, + "learning_rate": 1.826070920443136e-06, + "loss": 2.5538, + "step": 28334 + }, + { + "epoch": 2.996510152284264, + "grad_norm": 8.25, + "learning_rate": 1.8259060056488313e-06, + "loss": 2.3594, + "step": 28335 + }, + { + "epoch": 2.996615905245347, + "grad_norm": 8.1875, + "learning_rate": 1.8257410940177745e-06, + "loss": 2.348, + "step": 28336 + }, + { + "epoch": 2.9967216582064298, + "grad_norm": 8.4375, + "learning_rate": 1.8255761855507388e-06, + "loss": 2.5994, + "step": 28337 + }, + { + "epoch": 2.9968274111675126, + "grad_norm": 8.8125, + "learning_rate": 1.8254112802484991e-06, + "loss": 2.6018, + "step": 28338 + }, + { + "epoch": 2.9969331641285955, + "grad_norm": 8.625, + "learning_rate": 1.825246378111827e-06, + "loss": 2.3813, + "step": 28339 + }, + { + "epoch": 2.9970389170896787, + "grad_norm": 8.0625, + "learning_rate": 1.8250814791414987e-06, + "loss": 2.3668, + "step": 28340 + }, + { + "epoch": 2.9971446700507616, + "grad_norm": 8.625, + "learning_rate": 1.8249165833382861e-06, + "loss": 2.3241, + "step": 28341 + }, + { + "epoch": 2.9972504230118444, + "grad_norm": 8.625, + "learning_rate": 1.8247516907029645e-06, + "loss": 2.5294, + "step": 28342 + }, + { + "epoch": 2.9973561759729273, + "grad_norm": 8.375, + "learning_rate": 1.8245868012363067e-06, + "loss": 2.418, + "step": 28343 + }, + { + "epoch": 2.99746192893401, + "grad_norm": 8.5, + "learning_rate": 1.8244219149390867e-06, + "loss": 2.4668, + "step": 28344 + }, + { + "epoch": 2.997567681895093, + "grad_norm": 8.25, + "learning_rate": 1.8242570318120794e-06, + "loss": 2.6124, + "step": 28345 + }, + { + "epoch": 2.9976734348561758, + "grad_norm": 8.625, + "learning_rate": 1.8240921518560556e-06, + "loss": 2.7166, + "step": 28346 + }, + { + "epoch": 2.997779187817259, + "grad_norm": 8.0, + "learning_rate": 1.823927275071793e-06, + "loss": 2.5986, + "step": 28347 + }, + { + "epoch": 2.997884940778342, + "grad_norm": 8.875, + "learning_rate": 1.8237624014600618e-06, + "loss": 2.6173, + "step": 28348 + }, + { + "epoch": 2.9979906937394247, + "grad_norm": 8.9375, + "learning_rate": 1.8235975310216377e-06, + "loss": 2.488, + "step": 28349 + }, + { + "epoch": 2.9980964467005076, + "grad_norm": 8.4375, + "learning_rate": 1.8234326637572932e-06, + "loss": 2.6844, + "step": 28350 + }, + { + "epoch": 2.9982021996615904, + "grad_norm": 8.5625, + "learning_rate": 1.823267799667804e-06, + "loss": 2.5165, + "step": 28351 + }, + { + "epoch": 2.9983079526226737, + "grad_norm": 9.0625, + "learning_rate": 1.82310293875394e-06, + "loss": 2.5781, + "step": 28352 + }, + { + "epoch": 2.998413705583756, + "grad_norm": 8.5625, + "learning_rate": 1.8229380810164777e-06, + "loss": 2.4321, + "step": 28353 + }, + { + "epoch": 2.9985194585448394, + "grad_norm": 8.25, + "learning_rate": 1.8227732264561914e-06, + "loss": 2.5572, + "step": 28354 + }, + { + "epoch": 2.998625211505922, + "grad_norm": 8.4375, + "learning_rate": 1.822608375073852e-06, + "loss": 2.4408, + "step": 28355 + }, + { + "epoch": 2.998730964467005, + "grad_norm": 8.375, + "learning_rate": 1.8224435268702351e-06, + "loss": 2.7178, + "step": 28356 + }, + { + "epoch": 2.998836717428088, + "grad_norm": 8.25, + "learning_rate": 1.8222786818461132e-06, + "loss": 2.4953, + "step": 28357 + }, + { + "epoch": 2.9989424703891707, + "grad_norm": 8.375, + "learning_rate": 1.8221138400022608e-06, + "loss": 2.7222, + "step": 28358 + }, + { + "epoch": 2.999048223350254, + "grad_norm": 8.0625, + "learning_rate": 1.8219490013394502e-06, + "loss": 2.5533, + "step": 28359 + }, + { + "epoch": 2.999153976311337, + "grad_norm": 8.25, + "learning_rate": 1.821784165858457e-06, + "loss": 2.6481, + "step": 28360 + }, + { + "epoch": 2.9992597292724197, + "grad_norm": 8.625, + "learning_rate": 1.821619333560052e-06, + "loss": 2.6135, + "step": 28361 + }, + { + "epoch": 2.9993654822335025, + "grad_norm": 9.125, + "learning_rate": 1.8214545044450097e-06, + "loss": 2.6359, + "step": 28362 + }, + { + "epoch": 2.9994712351945854, + "grad_norm": 7.96875, + "learning_rate": 1.8212896785141055e-06, + "loss": 2.481, + "step": 28363 + }, + { + "epoch": 2.999576988155668, + "grad_norm": 8.3125, + "learning_rate": 1.8211248557681099e-06, + "loss": 2.4408, + "step": 28364 + }, + { + "epoch": 2.999682741116751, + "grad_norm": 8.375, + "learning_rate": 1.8209600362077985e-06, + "loss": 2.6428, + "step": 28365 + }, + { + "epoch": 2.9997884940778343, + "grad_norm": 8.25, + "learning_rate": 1.8207952198339428e-06, + "loss": 2.4904, + "step": 28366 + }, + { + "epoch": 2.999894247038917, + "grad_norm": 8.5625, + "learning_rate": 1.820630406647319e-06, + "loss": 2.3952, + "step": 28367 + }, + { + "epoch": 3.0, + "grad_norm": 8.3125, + "learning_rate": 1.8204655966486973e-06, + "loss": 2.4077, + "step": 28368 + } + ], + "logging_steps": 1, + "max_steps": 47280, + "num_input_tokens_seen": 0, + "num_train_epochs": 5, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 1.3817142586512507e+18, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}