diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,25977 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.999089345340484, + "eval_steps": 500, + "global_step": 3705, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00080947080845897, + "grad_norm": 25.965896606445312, + "learning_rate": 1.7857142857142858e-07, + "loss": 0.524, + "step": 1 + }, + { + "epoch": 0.00161894161691794, + "grad_norm": 27.55011749267578, + "learning_rate": 3.5714285714285716e-07, + "loss": 0.612, + "step": 2 + }, + { + "epoch": 0.00242841242537691, + "grad_norm": 14.853368759155273, + "learning_rate": 5.357142857142857e-07, + "loss": 0.5822, + "step": 3 + }, + { + "epoch": 0.00323788323383588, + "grad_norm": 7.189858436584473, + "learning_rate": 7.142857142857143e-07, + "loss": 0.493, + "step": 4 + }, + { + "epoch": 0.0040473540422948494, + "grad_norm": 6.2043681144714355, + "learning_rate": 8.928571428571429e-07, + "loss": 0.4953, + "step": 5 + }, + { + "epoch": 0.00485682485075382, + "grad_norm": 5.497591495513916, + "learning_rate": 1.0714285714285714e-06, + "loss": 0.4924, + "step": 6 + }, + { + "epoch": 0.005666295659212789, + "grad_norm": 4.930486679077148, + "learning_rate": 1.25e-06, + "loss": 0.4446, + "step": 7 + }, + { + "epoch": 0.00647576646767176, + "grad_norm": 6.2868523597717285, + "learning_rate": 1.4285714285714286e-06, + "loss": 0.4466, + "step": 8 + }, + { + "epoch": 0.007285237276130729, + "grad_norm": 5.715777397155762, + "learning_rate": 1.6071428571428574e-06, + "loss": 0.4365, + "step": 9 + }, + { + "epoch": 0.008094708084589699, + "grad_norm": 6.952973365783691, + "learning_rate": 1.7857142857142859e-06, + "loss": 0.4767, + "step": 10 + }, + { + "epoch": 0.00890417889304867, + "grad_norm": 6.250865936279297, + "learning_rate": 1.9642857142857144e-06, + "loss": 0.4569, + "step": 11 + }, + { + "epoch": 0.00971364970150764, + "grad_norm": 5.12192440032959, + "learning_rate": 2.1428571428571427e-06, + "loss": 0.4595, + "step": 12 + }, + { + "epoch": 0.01052312050996661, + "grad_norm": 5.387632846832275, + "learning_rate": 2.321428571428572e-06, + "loss": 0.446, + "step": 13 + }, + { + "epoch": 0.011332591318425579, + "grad_norm": 4.477903842926025, + "learning_rate": 2.5e-06, + "loss": 0.4035, + "step": 14 + }, + { + "epoch": 0.01214206212688455, + "grad_norm": 4.688737392425537, + "learning_rate": 2.6785714285714285e-06, + "loss": 0.4171, + "step": 15 + }, + { + "epoch": 0.01295153293534352, + "grad_norm": 4.3428778648376465, + "learning_rate": 2.8571428571428573e-06, + "loss": 0.3885, + "step": 16 + }, + { + "epoch": 0.01376100374380249, + "grad_norm": 4.60419225692749, + "learning_rate": 3.0357142857142856e-06, + "loss": 0.4251, + "step": 17 + }, + { + "epoch": 0.014570474552261459, + "grad_norm": 4.534536361694336, + "learning_rate": 3.2142857142857147e-06, + "loss": 0.4386, + "step": 18 + }, + { + "epoch": 0.015379945360720428, + "grad_norm": 4.956528663635254, + "learning_rate": 3.3928571428571435e-06, + "loss": 0.4685, + "step": 19 + }, + { + "epoch": 0.016189416169179398, + "grad_norm": 4.456263542175293, + "learning_rate": 3.5714285714285718e-06, + "loss": 0.4176, + "step": 20 + }, + { + "epoch": 0.016998886977638367, + "grad_norm": 4.5580549240112305, + "learning_rate": 3.7500000000000005e-06, + "loss": 0.4391, + "step": 21 + }, + { + "epoch": 0.01780835778609734, + "grad_norm": 4.370490074157715, + "learning_rate": 3.928571428571429e-06, + "loss": 0.4047, + "step": 22 + }, + { + "epoch": 0.01861782859455631, + "grad_norm": 4.739868640899658, + "learning_rate": 4.107142857142857e-06, + "loss": 0.4661, + "step": 23 + }, + { + "epoch": 0.01942729940301528, + "grad_norm": 4.903687477111816, + "learning_rate": 4.2857142857142855e-06, + "loss": 0.4459, + "step": 24 + }, + { + "epoch": 0.02023677021147425, + "grad_norm": 4.594234943389893, + "learning_rate": 4.464285714285715e-06, + "loss": 0.4023, + "step": 25 + }, + { + "epoch": 0.02104624101993322, + "grad_norm": 4.621077537536621, + "learning_rate": 4.642857142857144e-06, + "loss": 0.4305, + "step": 26 + }, + { + "epoch": 0.021855711828392188, + "grad_norm": 4.734155178070068, + "learning_rate": 4.821428571428572e-06, + "loss": 0.4111, + "step": 27 + }, + { + "epoch": 0.022665182636851158, + "grad_norm": 4.65170955657959, + "learning_rate": 5e-06, + "loss": 0.3966, + "step": 28 + }, + { + "epoch": 0.023474653445310127, + "grad_norm": 4.4155073165893555, + "learning_rate": 5.1785714285714296e-06, + "loss": 0.4325, + "step": 29 + }, + { + "epoch": 0.0242841242537691, + "grad_norm": 4.626680850982666, + "learning_rate": 5.357142857142857e-06, + "loss": 0.4281, + "step": 30 + }, + { + "epoch": 0.02509359506222807, + "grad_norm": 4.295886993408203, + "learning_rate": 5.535714285714286e-06, + "loss": 0.4232, + "step": 31 + }, + { + "epoch": 0.02590306587068704, + "grad_norm": 4.283386707305908, + "learning_rate": 5.7142857142857145e-06, + "loss": 0.3882, + "step": 32 + }, + { + "epoch": 0.02671253667914601, + "grad_norm": 4.097749710083008, + "learning_rate": 5.892857142857144e-06, + "loss": 0.4045, + "step": 33 + }, + { + "epoch": 0.02752200748760498, + "grad_norm": 4.104689121246338, + "learning_rate": 6.071428571428571e-06, + "loss": 0.4327, + "step": 34 + }, + { + "epoch": 0.028331478296063948, + "grad_norm": 4.376183032989502, + "learning_rate": 6.25e-06, + "loss": 0.4241, + "step": 35 + }, + { + "epoch": 0.029140949104522917, + "grad_norm": 4.425595283508301, + "learning_rate": 6.4285714285714295e-06, + "loss": 0.4346, + "step": 36 + }, + { + "epoch": 0.029950419912981887, + "grad_norm": 4.04736328125, + "learning_rate": 6.607142857142858e-06, + "loss": 0.4141, + "step": 37 + }, + { + "epoch": 0.030759890721440857, + "grad_norm": 4.193057060241699, + "learning_rate": 6.785714285714287e-06, + "loss": 0.4637, + "step": 38 + }, + { + "epoch": 0.03156936152989983, + "grad_norm": 3.9479832649230957, + "learning_rate": 6.964285714285714e-06, + "loss": 0.4218, + "step": 39 + }, + { + "epoch": 0.032378832338358796, + "grad_norm": 4.074609279632568, + "learning_rate": 7.1428571428571436e-06, + "loss": 0.4168, + "step": 40 + }, + { + "epoch": 0.03318830314681777, + "grad_norm": 4.31352424621582, + "learning_rate": 7.321428571428572e-06, + "loss": 0.4075, + "step": 41 + }, + { + "epoch": 0.033997773955276735, + "grad_norm": 4.356870174407959, + "learning_rate": 7.500000000000001e-06, + "loss": 0.4262, + "step": 42 + }, + { + "epoch": 0.03480724476373571, + "grad_norm": 4.042210102081299, + "learning_rate": 7.67857142857143e-06, + "loss": 0.4331, + "step": 43 + }, + { + "epoch": 0.03561671557219468, + "grad_norm": 3.9707818031311035, + "learning_rate": 7.857142857142858e-06, + "loss": 0.3851, + "step": 44 + }, + { + "epoch": 0.03642618638065365, + "grad_norm": 3.9531960487365723, + "learning_rate": 8.035714285714286e-06, + "loss": 0.4244, + "step": 45 + }, + { + "epoch": 0.03723565718911262, + "grad_norm": 4.326679229736328, + "learning_rate": 8.214285714285714e-06, + "loss": 0.4448, + "step": 46 + }, + { + "epoch": 0.038045127997571586, + "grad_norm": 3.8601601123809814, + "learning_rate": 8.392857142857144e-06, + "loss": 0.4132, + "step": 47 + }, + { + "epoch": 0.03885459880603056, + "grad_norm": 4.144084930419922, + "learning_rate": 8.571428571428571e-06, + "loss": 0.4711, + "step": 48 + }, + { + "epoch": 0.039664069614489525, + "grad_norm": 3.7318038940429688, + "learning_rate": 8.750000000000001e-06, + "loss": 0.4289, + "step": 49 + }, + { + "epoch": 0.0404735404229485, + "grad_norm": 3.4740703105926514, + "learning_rate": 8.92857142857143e-06, + "loss": 0.4173, + "step": 50 + }, + { + "epoch": 0.041283011231407464, + "grad_norm": 3.6853647232055664, + "learning_rate": 9.107142857142858e-06, + "loss": 0.4411, + "step": 51 + }, + { + "epoch": 0.04209248203986644, + "grad_norm": 3.7039976119995117, + "learning_rate": 9.285714285714288e-06, + "loss": 0.4536, + "step": 52 + }, + { + "epoch": 0.04290195284832541, + "grad_norm": 3.8168394565582275, + "learning_rate": 9.464285714285714e-06, + "loss": 0.4528, + "step": 53 + }, + { + "epoch": 0.043711423656784376, + "grad_norm": 4.134790897369385, + "learning_rate": 9.642857142857144e-06, + "loss": 0.4361, + "step": 54 + }, + { + "epoch": 0.04452089446524335, + "grad_norm": 4.15737247467041, + "learning_rate": 9.821428571428573e-06, + "loss": 0.4517, + "step": 55 + }, + { + "epoch": 0.045330365273702315, + "grad_norm": 3.777024507522583, + "learning_rate": 1e-05, + "loss": 0.4004, + "step": 56 + }, + { + "epoch": 0.04613983608216129, + "grad_norm": 3.8548386096954346, + "learning_rate": 1.0178571428571429e-05, + "loss": 0.399, + "step": 57 + }, + { + "epoch": 0.046949306890620254, + "grad_norm": 3.527252674102783, + "learning_rate": 1.0357142857142859e-05, + "loss": 0.3869, + "step": 58 + }, + { + "epoch": 0.04775877769907923, + "grad_norm": 3.802211284637451, + "learning_rate": 1.0535714285714287e-05, + "loss": 0.4365, + "step": 59 + }, + { + "epoch": 0.0485682485075382, + "grad_norm": 3.7647664546966553, + "learning_rate": 1.0714285714285714e-05, + "loss": 0.3968, + "step": 60 + }, + { + "epoch": 0.049377719315997166, + "grad_norm": 3.7366995811462402, + "learning_rate": 1.0892857142857142e-05, + "loss": 0.3991, + "step": 61 + }, + { + "epoch": 0.05018719012445614, + "grad_norm": 3.863279104232788, + "learning_rate": 1.1071428571428572e-05, + "loss": 0.4507, + "step": 62 + }, + { + "epoch": 0.050996660932915105, + "grad_norm": 4.250333786010742, + "learning_rate": 1.125e-05, + "loss": 0.4596, + "step": 63 + }, + { + "epoch": 0.05180613174137408, + "grad_norm": 3.708585739135742, + "learning_rate": 1.1428571428571429e-05, + "loss": 0.4501, + "step": 64 + }, + { + "epoch": 0.052615602549833045, + "grad_norm": 4.241486072540283, + "learning_rate": 1.1607142857142859e-05, + "loss": 0.4539, + "step": 65 + }, + { + "epoch": 0.05342507335829202, + "grad_norm": 3.551532030105591, + "learning_rate": 1.1785714285714287e-05, + "loss": 0.4293, + "step": 66 + }, + { + "epoch": 0.054234544166750984, + "grad_norm": 3.902594804763794, + "learning_rate": 1.1964285714285716e-05, + "loss": 0.4608, + "step": 67 + }, + { + "epoch": 0.05504401497520996, + "grad_norm": 3.5106375217437744, + "learning_rate": 1.2142857142857142e-05, + "loss": 0.4293, + "step": 68 + }, + { + "epoch": 0.05585348578366893, + "grad_norm": 3.576026201248169, + "learning_rate": 1.2321428571428572e-05, + "loss": 0.427, + "step": 69 + }, + { + "epoch": 0.056662956592127896, + "grad_norm": 3.9693775177001953, + "learning_rate": 1.25e-05, + "loss": 0.4372, + "step": 70 + }, + { + "epoch": 0.05747242740058687, + "grad_norm": 3.8853824138641357, + "learning_rate": 1.2678571428571429e-05, + "loss": 0.4493, + "step": 71 + }, + { + "epoch": 0.058281898209045835, + "grad_norm": 4.4115118980407715, + "learning_rate": 1.2857142857142859e-05, + "loss": 0.4416, + "step": 72 + }, + { + "epoch": 0.05909136901750481, + "grad_norm": 3.5301685333251953, + "learning_rate": 1.3035714285714287e-05, + "loss": 0.4526, + "step": 73 + }, + { + "epoch": 0.059900839825963774, + "grad_norm": 3.967383623123169, + "learning_rate": 1.3214285714285716e-05, + "loss": 0.4412, + "step": 74 + }, + { + "epoch": 0.06071031063442275, + "grad_norm": 3.5556814670562744, + "learning_rate": 1.3392857142857142e-05, + "loss": 0.4354, + "step": 75 + }, + { + "epoch": 0.06151978144288171, + "grad_norm": 3.7715063095092773, + "learning_rate": 1.3571428571428574e-05, + "loss": 0.4968, + "step": 76 + }, + { + "epoch": 0.062329252251340686, + "grad_norm": 3.7490086555480957, + "learning_rate": 1.375e-05, + "loss": 0.4542, + "step": 77 + }, + { + "epoch": 0.06313872305979966, + "grad_norm": 3.463604211807251, + "learning_rate": 1.3928571428571429e-05, + "loss": 0.4435, + "step": 78 + }, + { + "epoch": 0.06394819386825863, + "grad_norm": 3.7461092472076416, + "learning_rate": 1.4107142857142859e-05, + "loss": 0.4902, + "step": 79 + }, + { + "epoch": 0.06475766467671759, + "grad_norm": 3.443920612335205, + "learning_rate": 1.4285714285714287e-05, + "loss": 0.4771, + "step": 80 + }, + { + "epoch": 0.06556713548517656, + "grad_norm": 3.6136393547058105, + "learning_rate": 1.4464285714285715e-05, + "loss": 0.4517, + "step": 81 + }, + { + "epoch": 0.06637660629363554, + "grad_norm": 3.3156392574310303, + "learning_rate": 1.4642857142857144e-05, + "loss": 0.4545, + "step": 82 + }, + { + "epoch": 0.06718607710209451, + "grad_norm": 3.871445894241333, + "learning_rate": 1.4821428571428574e-05, + "loss": 0.4997, + "step": 83 + }, + { + "epoch": 0.06799554791055347, + "grad_norm": 3.905748128890991, + "learning_rate": 1.5000000000000002e-05, + "loss": 0.5078, + "step": 84 + }, + { + "epoch": 0.06880501871901244, + "grad_norm": 3.601673126220703, + "learning_rate": 1.5178571428571429e-05, + "loss": 0.433, + "step": 85 + }, + { + "epoch": 0.06961448952747142, + "grad_norm": 4.054693222045898, + "learning_rate": 1.535714285714286e-05, + "loss": 0.4851, + "step": 86 + }, + { + "epoch": 0.07042396033593039, + "grad_norm": 3.696444034576416, + "learning_rate": 1.553571428571429e-05, + "loss": 0.4707, + "step": 87 + }, + { + "epoch": 0.07123343114438936, + "grad_norm": 3.4664371013641357, + "learning_rate": 1.5714285714285715e-05, + "loss": 0.4802, + "step": 88 + }, + { + "epoch": 0.07204290195284832, + "grad_norm": 4.0117034912109375, + "learning_rate": 1.5892857142857142e-05, + "loss": 0.4658, + "step": 89 + }, + { + "epoch": 0.0728523727613073, + "grad_norm": 4.0240397453308105, + "learning_rate": 1.6071428571428572e-05, + "loss": 0.4704, + "step": 90 + }, + { + "epoch": 0.07366184356976627, + "grad_norm": 3.936882257461548, + "learning_rate": 1.6250000000000002e-05, + "loss": 0.48, + "step": 91 + }, + { + "epoch": 0.07447131437822524, + "grad_norm": 3.8120343685150146, + "learning_rate": 1.642857142857143e-05, + "loss": 0.4838, + "step": 92 + }, + { + "epoch": 0.0752807851866842, + "grad_norm": 4.005762100219727, + "learning_rate": 1.660714285714286e-05, + "loss": 0.492, + "step": 93 + }, + { + "epoch": 0.07609025599514317, + "grad_norm": 4.179708480834961, + "learning_rate": 1.678571428571429e-05, + "loss": 0.4649, + "step": 94 + }, + { + "epoch": 0.07689972680360214, + "grad_norm": 3.949793815612793, + "learning_rate": 1.6964285714285715e-05, + "loss": 0.4604, + "step": 95 + }, + { + "epoch": 0.07770919761206112, + "grad_norm": 4.026256084442139, + "learning_rate": 1.7142857142857142e-05, + "loss": 0.4841, + "step": 96 + }, + { + "epoch": 0.07851866842052009, + "grad_norm": 3.620859146118164, + "learning_rate": 1.7321428571428572e-05, + "loss": 0.4786, + "step": 97 + }, + { + "epoch": 0.07932813922897905, + "grad_norm": 3.746005058288574, + "learning_rate": 1.7500000000000002e-05, + "loss": 0.4388, + "step": 98 + }, + { + "epoch": 0.08013761003743802, + "grad_norm": 3.7392969131469727, + "learning_rate": 1.767857142857143e-05, + "loss": 0.4886, + "step": 99 + }, + { + "epoch": 0.080947080845897, + "grad_norm": 4.000370025634766, + "learning_rate": 1.785714285714286e-05, + "loss": 0.4976, + "step": 100 + }, + { + "epoch": 0.08175655165435597, + "grad_norm": 3.526796579360962, + "learning_rate": 1.803571428571429e-05, + "loss": 0.4895, + "step": 101 + }, + { + "epoch": 0.08256602246281493, + "grad_norm": 3.9062228202819824, + "learning_rate": 1.8214285714285715e-05, + "loss": 0.5131, + "step": 102 + }, + { + "epoch": 0.0833754932712739, + "grad_norm": 3.9170544147491455, + "learning_rate": 1.8392857142857142e-05, + "loss": 0.5187, + "step": 103 + }, + { + "epoch": 0.08418496407973287, + "grad_norm": 3.716252326965332, + "learning_rate": 1.8571428571428575e-05, + "loss": 0.4592, + "step": 104 + }, + { + "epoch": 0.08499443488819185, + "grad_norm": 3.9220058917999268, + "learning_rate": 1.8750000000000002e-05, + "loss": 0.5446, + "step": 105 + }, + { + "epoch": 0.08580390569665082, + "grad_norm": 3.5998971462249756, + "learning_rate": 1.892857142857143e-05, + "loss": 0.4562, + "step": 106 + }, + { + "epoch": 0.08661337650510978, + "grad_norm": 3.8255178928375244, + "learning_rate": 1.910714285714286e-05, + "loss": 0.5199, + "step": 107 + }, + { + "epoch": 0.08742284731356875, + "grad_norm": 3.45347261428833, + "learning_rate": 1.928571428571429e-05, + "loss": 0.4568, + "step": 108 + }, + { + "epoch": 0.08823231812202773, + "grad_norm": 3.709826946258545, + "learning_rate": 1.9464285714285715e-05, + "loss": 0.484, + "step": 109 + }, + { + "epoch": 0.0890417889304867, + "grad_norm": 3.884552240371704, + "learning_rate": 1.9642857142857145e-05, + "loss": 0.4862, + "step": 110 + }, + { + "epoch": 0.08985125973894566, + "grad_norm": 3.6887741088867188, + "learning_rate": 1.9821428571428575e-05, + "loss": 0.5265, + "step": 111 + }, + { + "epoch": 0.09066073054740463, + "grad_norm": 3.7118799686431885, + "learning_rate": 2e-05, + "loss": 0.4912, + "step": 112 + }, + { + "epoch": 0.0914702013558636, + "grad_norm": 3.58782958984375, + "learning_rate": 1.9999996177431404e-05, + "loss": 0.4847, + "step": 113 + }, + { + "epoch": 0.09227967216432258, + "grad_norm": 3.5130598545074463, + "learning_rate": 1.999998470972854e-05, + "loss": 0.4179, + "step": 114 + }, + { + "epoch": 0.09308914297278155, + "grad_norm": 3.696159839630127, + "learning_rate": 1.9999965596900167e-05, + "loss": 0.505, + "step": 115 + }, + { + "epoch": 0.09389861378124051, + "grad_norm": 3.353013038635254, + "learning_rate": 1.99999388389609e-05, + "loss": 0.4514, + "step": 116 + }, + { + "epoch": 0.09470808458969948, + "grad_norm": 4.047017574310303, + "learning_rate": 1.9999904435931204e-05, + "loss": 0.5279, + "step": 117 + }, + { + "epoch": 0.09551755539815845, + "grad_norm": 3.649754047393799, + "learning_rate": 1.9999862387837373e-05, + "loss": 0.5121, + "step": 118 + }, + { + "epoch": 0.09632702620661743, + "grad_norm": 3.6381752490997314, + "learning_rate": 1.9999812694711555e-05, + "loss": 0.5115, + "step": 119 + }, + { + "epoch": 0.0971364970150764, + "grad_norm": 3.7405974864959717, + "learning_rate": 1.999975535659174e-05, + "loss": 0.492, + "step": 120 + }, + { + "epoch": 0.09794596782353536, + "grad_norm": 3.6721608638763428, + "learning_rate": 1.9999690373521767e-05, + "loss": 0.4924, + "step": 121 + }, + { + "epoch": 0.09875543863199433, + "grad_norm": 3.562528610229492, + "learning_rate": 1.9999617745551313e-05, + "loss": 0.5015, + "step": 122 + }, + { + "epoch": 0.0995649094404533, + "grad_norm": 3.351188898086548, + "learning_rate": 1.9999537472735903e-05, + "loss": 0.5057, + "step": 123 + }, + { + "epoch": 0.10037438024891228, + "grad_norm": 4.203858852386475, + "learning_rate": 1.999944955513691e-05, + "loss": 0.4771, + "step": 124 + }, + { + "epoch": 0.10118385105737124, + "grad_norm": 3.87392520904541, + "learning_rate": 1.999935399282155e-05, + "loss": 0.4987, + "step": 125 + }, + { + "epoch": 0.10199332186583021, + "grad_norm": 4.470154285430908, + "learning_rate": 1.999925078586287e-05, + "loss": 0.5757, + "step": 126 + }, + { + "epoch": 0.10280279267428918, + "grad_norm": 3.961130142211914, + "learning_rate": 1.9999139934339787e-05, + "loss": 0.4806, + "step": 127 + }, + { + "epoch": 0.10361226348274816, + "grad_norm": 3.5203630924224854, + "learning_rate": 1.999902143833704e-05, + "loss": 0.4652, + "step": 128 + }, + { + "epoch": 0.10442173429120713, + "grad_norm": 3.7868943214416504, + "learning_rate": 1.9998895297945224e-05, + "loss": 0.4979, + "step": 129 + }, + { + "epoch": 0.10523120509966609, + "grad_norm": 4.392692565917969, + "learning_rate": 1.9998761513260776e-05, + "loss": 0.5299, + "step": 130 + }, + { + "epoch": 0.10604067590812506, + "grad_norm": 4.192699432373047, + "learning_rate": 1.9998620084385972e-05, + "loss": 0.5128, + "step": 131 + }, + { + "epoch": 0.10685014671658404, + "grad_norm": 3.793673038482666, + "learning_rate": 1.999847101142894e-05, + "loss": 0.4832, + "step": 132 + }, + { + "epoch": 0.10765961752504301, + "grad_norm": 3.9756815433502197, + "learning_rate": 1.999831429450365e-05, + "loss": 0.5046, + "step": 133 + }, + { + "epoch": 0.10846908833350197, + "grad_norm": 3.8226804733276367, + "learning_rate": 1.999814993372991e-05, + "loss": 0.5195, + "step": 134 + }, + { + "epoch": 0.10927855914196094, + "grad_norm": 3.4495139122009277, + "learning_rate": 1.9997977929233376e-05, + "loss": 0.4899, + "step": 135 + }, + { + "epoch": 0.11008802995041991, + "grad_norm": 3.65653133392334, + "learning_rate": 1.999779828114555e-05, + "loss": 0.5232, + "step": 136 + }, + { + "epoch": 0.11089750075887889, + "grad_norm": 3.755401611328125, + "learning_rate": 1.9997610989603775e-05, + "loss": 0.5108, + "step": 137 + }, + { + "epoch": 0.11170697156733786, + "grad_norm": 3.914477825164795, + "learning_rate": 1.9997416054751237e-05, + "loss": 0.5288, + "step": 138 + }, + { + "epoch": 0.11251644237579682, + "grad_norm": 3.6683566570281982, + "learning_rate": 1.999721347673697e-05, + "loss": 0.4824, + "step": 139 + }, + { + "epoch": 0.11332591318425579, + "grad_norm": 4.117203712463379, + "learning_rate": 1.9997003255715842e-05, + "loss": 0.5156, + "step": 140 + }, + { + "epoch": 0.11413538399271476, + "grad_norm": 3.967971086502075, + "learning_rate": 1.9996785391848575e-05, + "loss": 0.5231, + "step": 141 + }, + { + "epoch": 0.11494485480117374, + "grad_norm": 3.574307441711426, + "learning_rate": 1.999655988530173e-05, + "loss": 0.5272, + "step": 142 + }, + { + "epoch": 0.1157543256096327, + "grad_norm": 3.4860217571258545, + "learning_rate": 1.99963267362477e-05, + "loss": 0.5427, + "step": 143 + }, + { + "epoch": 0.11656379641809167, + "grad_norm": 3.520836353302002, + "learning_rate": 1.999608594486474e-05, + "loss": 0.516, + "step": 144 + }, + { + "epoch": 0.11737326722655064, + "grad_norm": 3.4383227825164795, + "learning_rate": 1.9995837511336935e-05, + "loss": 0.5188, + "step": 145 + }, + { + "epoch": 0.11818273803500962, + "grad_norm": 3.535409450531006, + "learning_rate": 1.9995581435854217e-05, + "loss": 0.5333, + "step": 146 + }, + { + "epoch": 0.11899220884346859, + "grad_norm": 3.61165189743042, + "learning_rate": 1.9995317718612356e-05, + "loss": 0.5247, + "step": 147 + }, + { + "epoch": 0.11980167965192755, + "grad_norm": 3.825261354446411, + "learning_rate": 1.9995046359812976e-05, + "loss": 0.5334, + "step": 148 + }, + { + "epoch": 0.12061115046038652, + "grad_norm": 3.8002312183380127, + "learning_rate": 1.9994767359663522e-05, + "loss": 0.4839, + "step": 149 + }, + { + "epoch": 0.1214206212688455, + "grad_norm": 4.288856506347656, + "learning_rate": 1.99944807183773e-05, + "loss": 0.533, + "step": 150 + }, + { + "epoch": 0.12223009207730447, + "grad_norm": 4.199477195739746, + "learning_rate": 1.9994186436173457e-05, + "loss": 0.5327, + "step": 151 + }, + { + "epoch": 0.12303956288576343, + "grad_norm": 3.9389615058898926, + "learning_rate": 1.999388451327697e-05, + "loss": 0.5323, + "step": 152 + }, + { + "epoch": 0.1238490336942224, + "grad_norm": 3.945695161819458, + "learning_rate": 1.9993574949918658e-05, + "loss": 0.4382, + "step": 153 + }, + { + "epoch": 0.12465850450268137, + "grad_norm": 3.522667646408081, + "learning_rate": 1.9993257746335193e-05, + "loss": 0.5191, + "step": 154 + }, + { + "epoch": 0.12546797531114035, + "grad_norm": 3.9494996070861816, + "learning_rate": 1.999293290276908e-05, + "loss": 0.4875, + "step": 155 + }, + { + "epoch": 0.12627744611959932, + "grad_norm": 3.666208267211914, + "learning_rate": 1.9992600419468665e-05, + "loss": 0.5026, + "step": 156 + }, + { + "epoch": 0.1270869169280583, + "grad_norm": 3.6311309337615967, + "learning_rate": 1.9992260296688142e-05, + "loss": 0.5302, + "step": 157 + }, + { + "epoch": 0.12789638773651726, + "grad_norm": 3.7941880226135254, + "learning_rate": 1.9991912534687536e-05, + "loss": 0.4942, + "step": 158 + }, + { + "epoch": 0.1287058585449762, + "grad_norm": 3.8308632373809814, + "learning_rate": 1.999155713373271e-05, + "loss": 0.5187, + "step": 159 + }, + { + "epoch": 0.12951532935343518, + "grad_norm": 3.8208775520324707, + "learning_rate": 1.999119409409538e-05, + "loss": 0.5362, + "step": 160 + }, + { + "epoch": 0.13032480016189416, + "grad_norm": 4.133471488952637, + "learning_rate": 1.9990823416053088e-05, + "loss": 0.5825, + "step": 161 + }, + { + "epoch": 0.13113427097035313, + "grad_norm": 3.283757209777832, + "learning_rate": 1.9990445099889234e-05, + "loss": 0.4774, + "step": 162 + }, + { + "epoch": 0.1319437417788121, + "grad_norm": 4.221770763397217, + "learning_rate": 1.9990059145893034e-05, + "loss": 0.4999, + "step": 163 + }, + { + "epoch": 0.13275321258727107, + "grad_norm": 3.3088607788085938, + "learning_rate": 1.998966555435956e-05, + "loss": 0.52, + "step": 164 + }, + { + "epoch": 0.13356268339573005, + "grad_norm": 3.409208297729492, + "learning_rate": 1.9989264325589717e-05, + "loss": 0.4913, + "step": 165 + }, + { + "epoch": 0.13437215420418902, + "grad_norm": 4.081577301025391, + "learning_rate": 1.998885545989026e-05, + "loss": 0.4974, + "step": 166 + }, + { + "epoch": 0.135181625012648, + "grad_norm": 3.5190975666046143, + "learning_rate": 1.9988438957573756e-05, + "loss": 0.4779, + "step": 167 + }, + { + "epoch": 0.13599109582110694, + "grad_norm": 5.7688374519348145, + "learning_rate": 1.998801481895863e-05, + "loss": 0.5469, + "step": 168 + }, + { + "epoch": 0.1368005666295659, + "grad_norm": 3.4816677570343018, + "learning_rate": 1.9987583044369153e-05, + "loss": 0.5261, + "step": 169 + }, + { + "epoch": 0.13761003743802488, + "grad_norm": 11.632960319519043, + "learning_rate": 1.9987143634135415e-05, + "loss": 0.5171, + "step": 170 + }, + { + "epoch": 0.13841950824648386, + "grad_norm": 3.4345312118530273, + "learning_rate": 1.998669658859335e-05, + "loss": 0.4862, + "step": 171 + }, + { + "epoch": 0.13922897905494283, + "grad_norm": 3.7332069873809814, + "learning_rate": 1.9986241908084735e-05, + "loss": 0.518, + "step": 172 + }, + { + "epoch": 0.1400384498634018, + "grad_norm": 3.6792213916778564, + "learning_rate": 1.998577959295717e-05, + "loss": 0.4631, + "step": 173 + }, + { + "epoch": 0.14084792067186078, + "grad_norm": 4.0718183517456055, + "learning_rate": 1.9985309643564115e-05, + "loss": 0.4894, + "step": 174 + }, + { + "epoch": 0.14165739148031975, + "grad_norm": 3.7480387687683105, + "learning_rate": 1.9984832060264842e-05, + "loss": 0.5423, + "step": 175 + }, + { + "epoch": 0.14246686228877872, + "grad_norm": 8.112882614135742, + "learning_rate": 1.9984346843424475e-05, + "loss": 0.5176, + "step": 176 + }, + { + "epoch": 0.14327633309723767, + "grad_norm": 4.150725364685059, + "learning_rate": 1.9983853993413968e-05, + "loss": 0.5485, + "step": 177 + }, + { + "epoch": 0.14408580390569664, + "grad_norm": 3.3512656688690186, + "learning_rate": 1.998335351061011e-05, + "loss": 0.5185, + "step": 178 + }, + { + "epoch": 0.14489527471415561, + "grad_norm": 3.825383424758911, + "learning_rate": 1.998284539539553e-05, + "loss": 0.5001, + "step": 179 + }, + { + "epoch": 0.1457047455226146, + "grad_norm": 3.7189595699310303, + "learning_rate": 1.9982329648158684e-05, + "loss": 0.5313, + "step": 180 + }, + { + "epoch": 0.14651421633107356, + "grad_norm": 3.7283647060394287, + "learning_rate": 1.9981806269293873e-05, + "loss": 0.5261, + "step": 181 + }, + { + "epoch": 0.14732368713953253, + "grad_norm": 3.611877918243408, + "learning_rate": 1.9981275259201226e-05, + "loss": 0.5252, + "step": 182 + }, + { + "epoch": 0.1481331579479915, + "grad_norm": 4.817241191864014, + "learning_rate": 1.9980736618286708e-05, + "loss": 0.4978, + "step": 183 + }, + { + "epoch": 0.14894262875645048, + "grad_norm": 3.6421382427215576, + "learning_rate": 1.9980190346962115e-05, + "loss": 0.4785, + "step": 184 + }, + { + "epoch": 0.14975209956490945, + "grad_norm": 7.843369483947754, + "learning_rate": 1.9979636445645077e-05, + "loss": 0.5072, + "step": 185 + }, + { + "epoch": 0.1505615703733684, + "grad_norm": 3.699031114578247, + "learning_rate": 1.9979074914759064e-05, + "loss": 0.4839, + "step": 186 + }, + { + "epoch": 0.15137104118182737, + "grad_norm": 14.572242736816406, + "learning_rate": 1.9978505754733375e-05, + "loss": 0.5585, + "step": 187 + }, + { + "epoch": 0.15218051199028634, + "grad_norm": 4.346083641052246, + "learning_rate": 1.9977928966003137e-05, + "loss": 0.4839, + "step": 188 + }, + { + "epoch": 0.15298998279874532, + "grad_norm": 4.056962966918945, + "learning_rate": 1.997734454900931e-05, + "loss": 0.5076, + "step": 189 + }, + { + "epoch": 0.1537994536072043, + "grad_norm": 3.6381850242614746, + "learning_rate": 1.99767525041987e-05, + "loss": 0.5489, + "step": 190 + }, + { + "epoch": 0.15460892441566326, + "grad_norm": 4.114363670349121, + "learning_rate": 1.9976152832023922e-05, + "loss": 0.4941, + "step": 191 + }, + { + "epoch": 0.15541839522412224, + "grad_norm": 4.045076370239258, + "learning_rate": 1.9975545532943435e-05, + "loss": 0.5062, + "step": 192 + }, + { + "epoch": 0.1562278660325812, + "grad_norm": 3.6190226078033447, + "learning_rate": 1.9974930607421536e-05, + "loss": 0.4949, + "step": 193 + }, + { + "epoch": 0.15703733684104018, + "grad_norm": 3.66159725189209, + "learning_rate": 1.9974308055928337e-05, + "loss": 0.5126, + "step": 194 + }, + { + "epoch": 0.15784680764949913, + "grad_norm": 4.371744155883789, + "learning_rate": 1.9973677878939783e-05, + "loss": 0.4722, + "step": 195 + }, + { + "epoch": 0.1586562784579581, + "grad_norm": 3.7384731769561768, + "learning_rate": 1.9973040076937666e-05, + "loss": 0.4876, + "step": 196 + }, + { + "epoch": 0.15946574926641707, + "grad_norm": 106.97744750976562, + "learning_rate": 1.997239465040958e-05, + "loss": 0.5055, + "step": 197 + }, + { + "epoch": 0.16027522007487605, + "grad_norm": 3.7254958152770996, + "learning_rate": 1.9971741599848973e-05, + "loss": 0.4879, + "step": 198 + }, + { + "epoch": 0.16108469088333502, + "grad_norm": 3.733030080795288, + "learning_rate": 1.9971080925755103e-05, + "loss": 0.5021, + "step": 199 + }, + { + "epoch": 0.161894161691794, + "grad_norm": 3.445021629333496, + "learning_rate": 1.9970412628633075e-05, + "loss": 0.5327, + "step": 200 + }, + { + "epoch": 0.16270363250025296, + "grad_norm": 3.9219253063201904, + "learning_rate": 1.9969736708993802e-05, + "loss": 0.4788, + "step": 201 + }, + { + "epoch": 0.16351310330871194, + "grad_norm": 3.873002767562866, + "learning_rate": 1.9969053167354035e-05, + "loss": 0.4909, + "step": 202 + }, + { + "epoch": 0.1643225741171709, + "grad_norm": 3.2730400562286377, + "learning_rate": 1.9968362004236353e-05, + "loss": 0.4759, + "step": 203 + }, + { + "epoch": 0.16513204492562986, + "grad_norm": 5.282098770141602, + "learning_rate": 1.9967663220169163e-05, + "loss": 0.5199, + "step": 204 + }, + { + "epoch": 0.16594151573408883, + "grad_norm": 3.5372684001922607, + "learning_rate": 1.9966956815686687e-05, + "loss": 0.4888, + "step": 205 + }, + { + "epoch": 0.1667509865425478, + "grad_norm": 3.7745816707611084, + "learning_rate": 1.9966242791328988e-05, + "loss": 0.5149, + "step": 206 + }, + { + "epoch": 0.16756045735100678, + "grad_norm": 3.853789806365967, + "learning_rate": 1.9965521147641947e-05, + "loss": 0.5304, + "step": 207 + }, + { + "epoch": 0.16836992815946575, + "grad_norm": 3.7542545795440674, + "learning_rate": 1.9964791885177264e-05, + "loss": 0.5422, + "step": 208 + }, + { + "epoch": 0.16917939896792472, + "grad_norm": 3.68860125541687, + "learning_rate": 1.9964055004492477e-05, + "loss": 0.5446, + "step": 209 + }, + { + "epoch": 0.1699888697763837, + "grad_norm": 3.206403970718384, + "learning_rate": 1.9963310506150938e-05, + "loss": 0.4676, + "step": 210 + }, + { + "epoch": 0.17079834058484267, + "grad_norm": 3.46980881690979, + "learning_rate": 1.9962558390721827e-05, + "loss": 0.4868, + "step": 211 + }, + { + "epoch": 0.17160781139330164, + "grad_norm": 3.532799243927002, + "learning_rate": 1.9961798658780146e-05, + "loss": 0.5164, + "step": 212 + }, + { + "epoch": 0.17241728220176059, + "grad_norm": 3.123750925064087, + "learning_rate": 1.9961031310906722e-05, + "loss": 0.4598, + "step": 213 + }, + { + "epoch": 0.17322675301021956, + "grad_norm": 3.6903374195098877, + "learning_rate": 1.9960256347688203e-05, + "loss": 0.4687, + "step": 214 + }, + { + "epoch": 0.17403622381867853, + "grad_norm": 4.252986907958984, + "learning_rate": 1.9959473769717058e-05, + "loss": 0.4981, + "step": 215 + }, + { + "epoch": 0.1748456946271375, + "grad_norm": 3.49993634223938, + "learning_rate": 1.9958683577591576e-05, + "loss": 0.5365, + "step": 216 + }, + { + "epoch": 0.17565516543559648, + "grad_norm": 3.6228742599487305, + "learning_rate": 1.9957885771915873e-05, + "loss": 0.5312, + "step": 217 + }, + { + "epoch": 0.17646463624405545, + "grad_norm": 3.20147442817688, + "learning_rate": 1.9957080353299886e-05, + "loss": 0.5004, + "step": 218 + }, + { + "epoch": 0.17727410705251442, + "grad_norm": 3.1047797203063965, + "learning_rate": 1.995626732235936e-05, + "loss": 0.4848, + "step": 219 + }, + { + "epoch": 0.1780835778609734, + "grad_norm": 3.4907708168029785, + "learning_rate": 1.9955446679715874e-05, + "loss": 0.4997, + "step": 220 + }, + { + "epoch": 0.17889304866943237, + "grad_norm": 3.4959888458251953, + "learning_rate": 1.995461842599682e-05, + "loss": 0.4981, + "step": 221 + }, + { + "epoch": 0.17970251947789131, + "grad_norm": 3.5294129848480225, + "learning_rate": 1.9953782561835405e-05, + "loss": 0.5253, + "step": 222 + }, + { + "epoch": 0.1805119902863503, + "grad_norm": 3.1695666313171387, + "learning_rate": 1.995293908787067e-05, + "loss": 0.5008, + "step": 223 + }, + { + "epoch": 0.18132146109480926, + "grad_norm": 3.4390206336975098, + "learning_rate": 1.995208800474745e-05, + "loss": 0.476, + "step": 224 + }, + { + "epoch": 0.18213093190326823, + "grad_norm": 3.670191526412964, + "learning_rate": 1.995122931311641e-05, + "loss": 0.5473, + "step": 225 + }, + { + "epoch": 0.1829404027117272, + "grad_norm": 3.6432971954345703, + "learning_rate": 1.995036301363404e-05, + "loss": 0.5181, + "step": 226 + }, + { + "epoch": 0.18374987352018618, + "grad_norm": 3.7165093421936035, + "learning_rate": 1.9949489106962635e-05, + "loss": 0.4969, + "step": 227 + }, + { + "epoch": 0.18455934432864515, + "grad_norm": 3.372722864151001, + "learning_rate": 1.9948607593770303e-05, + "loss": 0.508, + "step": 228 + }, + { + "epoch": 0.18536881513710413, + "grad_norm": 3.27115797996521, + "learning_rate": 1.994771847473098e-05, + "loss": 0.5113, + "step": 229 + }, + { + "epoch": 0.1861782859455631, + "grad_norm": 3.4903652667999268, + "learning_rate": 1.9946821750524406e-05, + "loss": 0.4931, + "step": 230 + }, + { + "epoch": 0.18698775675402204, + "grad_norm": 3.400618076324463, + "learning_rate": 1.9945917421836145e-05, + "loss": 0.4641, + "step": 231 + }, + { + "epoch": 0.18779722756248102, + "grad_norm": 3.658874988555908, + "learning_rate": 1.994500548935756e-05, + "loss": 0.4994, + "step": 232 + }, + { + "epoch": 0.18860669837094, + "grad_norm": 3.7507288455963135, + "learning_rate": 1.9944085953785838e-05, + "loss": 0.5032, + "step": 233 + }, + { + "epoch": 0.18941616917939896, + "grad_norm": 3.266350507736206, + "learning_rate": 1.994315881582398e-05, + "loss": 0.5108, + "step": 234 + }, + { + "epoch": 0.19022563998785794, + "grad_norm": 3.011080741882324, + "learning_rate": 1.994222407618079e-05, + "loss": 0.4405, + "step": 235 + }, + { + "epoch": 0.1910351107963169, + "grad_norm": 3.1277341842651367, + "learning_rate": 1.99412817355709e-05, + "loss": 0.4943, + "step": 236 + }, + { + "epoch": 0.19184458160477588, + "grad_norm": 3.8438057899475098, + "learning_rate": 1.9940331794714732e-05, + "loss": 0.4984, + "step": 237 + }, + { + "epoch": 0.19265405241323486, + "grad_norm": 3.7747750282287598, + "learning_rate": 1.9939374254338532e-05, + "loss": 0.5203, + "step": 238 + }, + { + "epoch": 0.19346352322169383, + "grad_norm": 3.4485630989074707, + "learning_rate": 1.993840911517435e-05, + "loss": 0.5217, + "step": 239 + }, + { + "epoch": 0.1942729940301528, + "grad_norm": 3.4376626014709473, + "learning_rate": 1.9937436377960055e-05, + "loss": 0.4869, + "step": 240 + }, + { + "epoch": 0.19508246483861175, + "grad_norm": 3.3969924449920654, + "learning_rate": 1.9936456043439313e-05, + "loss": 0.4952, + "step": 241 + }, + { + "epoch": 0.19589193564707072, + "grad_norm": 3.342012643814087, + "learning_rate": 1.9935468112361605e-05, + "loss": 0.4656, + "step": 242 + }, + { + "epoch": 0.1967014064555297, + "grad_norm": 3.5664021968841553, + "learning_rate": 1.9934472585482214e-05, + "loss": 0.5012, + "step": 243 + }, + { + "epoch": 0.19751087726398867, + "grad_norm": 3.5945332050323486, + "learning_rate": 1.993346946356224e-05, + "loss": 0.479, + "step": 244 + }, + { + "epoch": 0.19832034807244764, + "grad_norm": 3.466883897781372, + "learning_rate": 1.9932458747368582e-05, + "loss": 0.5228, + "step": 245 + }, + { + "epoch": 0.1991298188809066, + "grad_norm": 3.456101417541504, + "learning_rate": 1.9931440437673938e-05, + "loss": 0.4992, + "step": 246 + }, + { + "epoch": 0.19993928968936558, + "grad_norm": 3.761019706726074, + "learning_rate": 1.9930414535256833e-05, + "loss": 0.4963, + "step": 247 + }, + { + "epoch": 0.20074876049782456, + "grad_norm": 3.4094269275665283, + "learning_rate": 1.9929381040901575e-05, + "loss": 0.4884, + "step": 248 + }, + { + "epoch": 0.20155823130628353, + "grad_norm": 3.3968751430511475, + "learning_rate": 1.9928339955398287e-05, + "loss": 0.4751, + "step": 249 + }, + { + "epoch": 0.20236770211474248, + "grad_norm": 3.2722604274749756, + "learning_rate": 1.9927291279542892e-05, + "loss": 0.4788, + "step": 250 + }, + { + "epoch": 0.20317717292320145, + "grad_norm": 8.106240272521973, + "learning_rate": 1.992623501413712e-05, + "loss": 0.4863, + "step": 251 + }, + { + "epoch": 0.20398664373166042, + "grad_norm": 3.8677289485931396, + "learning_rate": 1.9925171159988495e-05, + "loss": 0.5466, + "step": 252 + }, + { + "epoch": 0.2047961145401194, + "grad_norm": 3.7176785469055176, + "learning_rate": 1.9924099717910356e-05, + "loss": 0.5757, + "step": 253 + }, + { + "epoch": 0.20560558534857837, + "grad_norm": 3.5725886821746826, + "learning_rate": 1.992302068872183e-05, + "loss": 0.519, + "step": 254 + }, + { + "epoch": 0.20641505615703734, + "grad_norm": 3.014592170715332, + "learning_rate": 1.9921934073247844e-05, + "loss": 0.491, + "step": 255 + }, + { + "epoch": 0.20722452696549631, + "grad_norm": 2.9971296787261963, + "learning_rate": 1.9920839872319145e-05, + "loss": 0.4915, + "step": 256 + }, + { + "epoch": 0.2080339977739553, + "grad_norm": 3.3932244777679443, + "learning_rate": 1.991973808677225e-05, + "loss": 0.5618, + "step": 257 + }, + { + "epoch": 0.20884346858241426, + "grad_norm": 3.043693780899048, + "learning_rate": 1.99186287174495e-05, + "loss": 0.4617, + "step": 258 + }, + { + "epoch": 0.2096529393908732, + "grad_norm": 3.396212100982666, + "learning_rate": 1.9917511765199018e-05, + "loss": 0.4953, + "step": 259 + }, + { + "epoch": 0.21046241019933218, + "grad_norm": 3.554068088531494, + "learning_rate": 1.9916387230874726e-05, + "loss": 0.527, + "step": 260 + }, + { + "epoch": 0.21127188100779115, + "grad_norm": 3.4166386127471924, + "learning_rate": 1.9915255115336352e-05, + "loss": 0.5326, + "step": 261 + }, + { + "epoch": 0.21208135181625012, + "grad_norm": 3.383317470550537, + "learning_rate": 1.9914115419449415e-05, + "loss": 0.5004, + "step": 262 + }, + { + "epoch": 0.2128908226247091, + "grad_norm": 3.6410481929779053, + "learning_rate": 1.9912968144085217e-05, + "loss": 0.4893, + "step": 263 + }, + { + "epoch": 0.21370029343316807, + "grad_norm": 3.121694326400757, + "learning_rate": 1.991181329012088e-05, + "loss": 0.4756, + "step": 264 + }, + { + "epoch": 0.21450976424162704, + "grad_norm": 3.3198859691619873, + "learning_rate": 1.99106508584393e-05, + "loss": 0.5105, + "step": 265 + }, + { + "epoch": 0.21531923505008602, + "grad_norm": 3.53363037109375, + "learning_rate": 1.9909480849929165e-05, + "loss": 0.5349, + "step": 266 + }, + { + "epoch": 0.216128705858545, + "grad_norm": 3.67458438873291, + "learning_rate": 1.9908303265484973e-05, + "loss": 0.5121, + "step": 267 + }, + { + "epoch": 0.21693817666700393, + "grad_norm": 3.628939390182495, + "learning_rate": 1.9907118106006997e-05, + "loss": 0.4937, + "step": 268 + }, + { + "epoch": 0.2177476474754629, + "grad_norm": 3.7784292697906494, + "learning_rate": 1.9905925372401312e-05, + "loss": 0.5474, + "step": 269 + }, + { + "epoch": 0.21855711828392188, + "grad_norm": 3.2850699424743652, + "learning_rate": 1.9904725065579775e-05, + "loss": 0.5266, + "step": 270 + }, + { + "epoch": 0.21936658909238085, + "grad_norm": 3.4028384685516357, + "learning_rate": 1.9903517186460043e-05, + "loss": 0.5533, + "step": 271 + }, + { + "epoch": 0.22017605990083983, + "grad_norm": 3.1334877014160156, + "learning_rate": 1.990230173596555e-05, + "loss": 0.4899, + "step": 272 + }, + { + "epoch": 0.2209855307092988, + "grad_norm": 3.079172134399414, + "learning_rate": 1.9901078715025527e-05, + "loss": 0.4671, + "step": 273 + }, + { + "epoch": 0.22179500151775777, + "grad_norm": 3.255831480026245, + "learning_rate": 1.9899848124574986e-05, + "loss": 0.5069, + "step": 274 + }, + { + "epoch": 0.22260447232621675, + "grad_norm": 3.5909717082977295, + "learning_rate": 1.989860996555474e-05, + "loss": 0.5416, + "step": 275 + }, + { + "epoch": 0.22341394313467572, + "grad_norm": 3.1524808406829834, + "learning_rate": 1.989736423891137e-05, + "loss": 0.4517, + "step": 276 + }, + { + "epoch": 0.22422341394313466, + "grad_norm": 3.2088425159454346, + "learning_rate": 1.9896110945597257e-05, + "loss": 0.502, + "step": 277 + }, + { + "epoch": 0.22503288475159364, + "grad_norm": 3.2425472736358643, + "learning_rate": 1.9894850086570553e-05, + "loss": 0.5086, + "step": 278 + }, + { + "epoch": 0.2258423555600526, + "grad_norm": 3.2855324745178223, + "learning_rate": 1.989358166279521e-05, + "loss": 0.4882, + "step": 279 + }, + { + "epoch": 0.22665182636851158, + "grad_norm": 3.2881808280944824, + "learning_rate": 1.9892305675240953e-05, + "loss": 0.4993, + "step": 280 + }, + { + "epoch": 0.22746129717697056, + "grad_norm": 3.413475513458252, + "learning_rate": 1.9891022124883292e-05, + "loss": 0.5077, + "step": 281 + }, + { + "epoch": 0.22827076798542953, + "grad_norm": 3.602837085723877, + "learning_rate": 1.9889731012703517e-05, + "loss": 0.5884, + "step": 282 + }, + { + "epoch": 0.2290802387938885, + "grad_norm": 3.406947374343872, + "learning_rate": 1.9888432339688704e-05, + "loss": 0.5162, + "step": 283 + }, + { + "epoch": 0.22988970960234748, + "grad_norm": 3.5940887928009033, + "learning_rate": 1.98871261068317e-05, + "loss": 0.5132, + "step": 284 + }, + { + "epoch": 0.23069918041080645, + "grad_norm": 3.387770414352417, + "learning_rate": 1.988581231513115e-05, + "loss": 0.5241, + "step": 285 + }, + { + "epoch": 0.2315086512192654, + "grad_norm": 3.3971402645111084, + "learning_rate": 1.9884490965591453e-05, + "loss": 0.4977, + "step": 286 + }, + { + "epoch": 0.23231812202772437, + "grad_norm": 3.2195136547088623, + "learning_rate": 1.9883162059222805e-05, + "loss": 0.5215, + "step": 287 + }, + { + "epoch": 0.23312759283618334, + "grad_norm": 3.3422725200653076, + "learning_rate": 1.988182559704117e-05, + "loss": 0.5337, + "step": 288 + }, + { + "epoch": 0.2339370636446423, + "grad_norm": 3.0852036476135254, + "learning_rate": 1.98804815800683e-05, + "loss": 0.474, + "step": 289 + }, + { + "epoch": 0.23474653445310129, + "grad_norm": 3.190284013748169, + "learning_rate": 1.9879130009331704e-05, + "loss": 0.4646, + "step": 290 + }, + { + "epoch": 0.23555600526156026, + "grad_norm": 3.214292049407959, + "learning_rate": 1.987777088586468e-05, + "loss": 0.5096, + "step": 291 + }, + { + "epoch": 0.23636547607001923, + "grad_norm": 3.4834420680999756, + "learning_rate": 1.98764042107063e-05, + "loss": 0.5006, + "step": 292 + }, + { + "epoch": 0.2371749468784782, + "grad_norm": 3.480073928833008, + "learning_rate": 1.98750299849014e-05, + "loss": 0.5287, + "step": 293 + }, + { + "epoch": 0.23798441768693718, + "grad_norm": 3.238118886947632, + "learning_rate": 1.98736482095006e-05, + "loss": 0.4952, + "step": 294 + }, + { + "epoch": 0.23879388849539612, + "grad_norm": 3.2850289344787598, + "learning_rate": 1.9872258885560286e-05, + "loss": 0.5195, + "step": 295 + }, + { + "epoch": 0.2396033593038551, + "grad_norm": 3.53340482711792, + "learning_rate": 1.987086201414261e-05, + "loss": 0.5529, + "step": 296 + }, + { + "epoch": 0.24041283011231407, + "grad_norm": 3.3625895977020264, + "learning_rate": 1.9869457596315504e-05, + "loss": 0.5064, + "step": 297 + }, + { + "epoch": 0.24122230092077304, + "grad_norm": 3.1558659076690674, + "learning_rate": 1.9868045633152662e-05, + "loss": 0.5313, + "step": 298 + }, + { + "epoch": 0.24203177172923201, + "grad_norm": 3.220930576324463, + "learning_rate": 1.9866626125733552e-05, + "loss": 0.5096, + "step": 299 + }, + { + "epoch": 0.242841242537691, + "grad_norm": 3.03218150138855, + "learning_rate": 1.9865199075143404e-05, + "loss": 0.4781, + "step": 300 + }, + { + "epoch": 0.24365071334614996, + "grad_norm": 3.110546112060547, + "learning_rate": 1.986376448247322e-05, + "loss": 0.4955, + "step": 301 + }, + { + "epoch": 0.24446018415460893, + "grad_norm": 3.1861066818237305, + "learning_rate": 1.9862322348819763e-05, + "loss": 0.4887, + "step": 302 + }, + { + "epoch": 0.2452696549630679, + "grad_norm": 3.418475866317749, + "learning_rate": 1.9860872675285567e-05, + "loss": 0.5036, + "step": 303 + }, + { + "epoch": 0.24607912577152685, + "grad_norm": 3.186245918273926, + "learning_rate": 1.9859415462978928e-05, + "loss": 0.4997, + "step": 304 + }, + { + "epoch": 0.24688859657998583, + "grad_norm": 3.483713150024414, + "learning_rate": 1.9857950713013902e-05, + "loss": 0.5084, + "step": 305 + }, + { + "epoch": 0.2476980673884448, + "grad_norm": 4.463476181030273, + "learning_rate": 1.985647842651031e-05, + "loss": 0.5166, + "step": 306 + }, + { + "epoch": 0.24850753819690377, + "grad_norm": 3.4903650283813477, + "learning_rate": 1.9854998604593737e-05, + "loss": 0.511, + "step": 307 + }, + { + "epoch": 0.24931700900536274, + "grad_norm": 3.4662067890167236, + "learning_rate": 1.9853511248395528e-05, + "loss": 0.5094, + "step": 308 + }, + { + "epoch": 0.2501264798138217, + "grad_norm": 3.54398775100708, + "learning_rate": 1.9852016359052783e-05, + "loss": 0.5826, + "step": 309 + }, + { + "epoch": 0.2509359506222807, + "grad_norm": 3.1070716381073, + "learning_rate": 1.985051393770837e-05, + "loss": 0.4635, + "step": 310 + }, + { + "epoch": 0.25174542143073964, + "grad_norm": 4.327908992767334, + "learning_rate": 1.9849003985510912e-05, + "loss": 0.5423, + "step": 311 + }, + { + "epoch": 0.25255489223919864, + "grad_norm": 3.8026461601257324, + "learning_rate": 1.984748650361478e-05, + "loss": 0.5134, + "step": 312 + }, + { + "epoch": 0.2533643630476576, + "grad_norm": 3.129201650619507, + "learning_rate": 1.984596149318012e-05, + "loss": 0.5381, + "step": 313 + }, + { + "epoch": 0.2541738338561166, + "grad_norm": 2.771766424179077, + "learning_rate": 1.9844428955372814e-05, + "loss": 0.4692, + "step": 314 + }, + { + "epoch": 0.2549833046645755, + "grad_norm": 2.888716459274292, + "learning_rate": 1.9842888891364517e-05, + "loss": 0.5005, + "step": 315 + }, + { + "epoch": 0.25579277547303453, + "grad_norm": 3.0885682106018066, + "learning_rate": 1.9841341302332623e-05, + "loss": 0.4827, + "step": 316 + }, + { + "epoch": 0.2566022462814935, + "grad_norm": 3.048676013946533, + "learning_rate": 1.9839786189460285e-05, + "loss": 0.4832, + "step": 317 + }, + { + "epoch": 0.2574117170899524, + "grad_norm": 3.4262969493865967, + "learning_rate": 1.9838223553936412e-05, + "loss": 0.5234, + "step": 318 + }, + { + "epoch": 0.2582211878984114, + "grad_norm": 3.3831844329833984, + "learning_rate": 1.9836653396955656e-05, + "loss": 0.4945, + "step": 319 + }, + { + "epoch": 0.25903065870687036, + "grad_norm": 3.482727527618408, + "learning_rate": 1.9835075719718428e-05, + "loss": 0.4988, + "step": 320 + }, + { + "epoch": 0.25984012951532937, + "grad_norm": 3.104116916656494, + "learning_rate": 1.983349052343088e-05, + "loss": 0.5078, + "step": 321 + }, + { + "epoch": 0.2606496003237883, + "grad_norm": 3.024611473083496, + "learning_rate": 1.9831897809304918e-05, + "loss": 0.4845, + "step": 322 + }, + { + "epoch": 0.2614590711322473, + "grad_norm": 3.584728717803955, + "learning_rate": 1.9830297578558195e-05, + "loss": 0.5175, + "step": 323 + }, + { + "epoch": 0.26226854194070626, + "grad_norm": 3.009638786315918, + "learning_rate": 1.9828689832414106e-05, + "loss": 0.4624, + "step": 324 + }, + { + "epoch": 0.26307801274916526, + "grad_norm": 3.415754795074463, + "learning_rate": 1.9827074572101798e-05, + "loss": 0.55, + "step": 325 + }, + { + "epoch": 0.2638874835576242, + "grad_norm": 3.273866653442383, + "learning_rate": 1.982545179885616e-05, + "loss": 0.5329, + "step": 326 + }, + { + "epoch": 0.26469695436608315, + "grad_norm": 3.075985908508301, + "learning_rate": 1.982382151391782e-05, + "loss": 0.5156, + "step": 327 + }, + { + "epoch": 0.26550642517454215, + "grad_norm": 3.4343605041503906, + "learning_rate": 1.9822183718533158e-05, + "loss": 0.4774, + "step": 328 + }, + { + "epoch": 0.2663158959830011, + "grad_norm": 3.4272046089172363, + "learning_rate": 1.982053841395429e-05, + "loss": 0.5361, + "step": 329 + }, + { + "epoch": 0.2671253667914601, + "grad_norm": 3.149395704269409, + "learning_rate": 1.981888560143907e-05, + "loss": 0.5562, + "step": 330 + }, + { + "epoch": 0.26793483759991904, + "grad_norm": 3.3542280197143555, + "learning_rate": 1.9817225282251106e-05, + "loss": 0.5092, + "step": 331 + }, + { + "epoch": 0.26874430840837804, + "grad_norm": 3.2784645557403564, + "learning_rate": 1.981555745765972e-05, + "loss": 0.5116, + "step": 332 + }, + { + "epoch": 0.269553779216837, + "grad_norm": 3.37007999420166, + "learning_rate": 1.981388212894e-05, + "loss": 0.553, + "step": 333 + }, + { + "epoch": 0.270363250025296, + "grad_norm": 3.2556002140045166, + "learning_rate": 1.9812199297372753e-05, + "loss": 0.5065, + "step": 334 + }, + { + "epoch": 0.27117272083375493, + "grad_norm": 3.0719475746154785, + "learning_rate": 1.9810508964244524e-05, + "loss": 0.5072, + "step": 335 + }, + { + "epoch": 0.2719821916422139, + "grad_norm": 3.549400568008423, + "learning_rate": 1.9808811130847596e-05, + "loss": 0.5374, + "step": 336 + }, + { + "epoch": 0.2727916624506729, + "grad_norm": 4.0547380447387695, + "learning_rate": 1.9807105798479995e-05, + "loss": 0.5426, + "step": 337 + }, + { + "epoch": 0.2736011332591318, + "grad_norm": 3.1792750358581543, + "learning_rate": 1.9805392968445457e-05, + "loss": 0.4901, + "step": 338 + }, + { + "epoch": 0.2744106040675908, + "grad_norm": 3.119729518890381, + "learning_rate": 1.9803672642053473e-05, + "loss": 0.5382, + "step": 339 + }, + { + "epoch": 0.27522007487604977, + "grad_norm": 2.8987486362457275, + "learning_rate": 1.9801944820619256e-05, + "loss": 0.4542, + "step": 340 + }, + { + "epoch": 0.27602954568450877, + "grad_norm": 3.3594210147857666, + "learning_rate": 1.980020950546375e-05, + "loss": 0.5702, + "step": 341 + }, + { + "epoch": 0.2768390164929677, + "grad_norm": 3.0946030616760254, + "learning_rate": 1.9798466697913617e-05, + "loss": 0.4651, + "step": 342 + }, + { + "epoch": 0.2776484873014267, + "grad_norm": 2.976400136947632, + "learning_rate": 1.979671639930127e-05, + "loss": 0.4613, + "step": 343 + }, + { + "epoch": 0.27845795810988566, + "grad_norm": 4.004617691040039, + "learning_rate": 1.9794958610964832e-05, + "loss": 0.5407, + "step": 344 + }, + { + "epoch": 0.2792674289183446, + "grad_norm": 3.1723101139068604, + "learning_rate": 1.9793193334248152e-05, + "loss": 0.5101, + "step": 345 + }, + { + "epoch": 0.2800768997268036, + "grad_norm": 3.180631399154663, + "learning_rate": 1.9791420570500813e-05, + "loss": 0.4922, + "step": 346 + }, + { + "epoch": 0.28088637053526255, + "grad_norm": 3.011983871459961, + "learning_rate": 1.978964032107812e-05, + "loss": 0.4795, + "step": 347 + }, + { + "epoch": 0.28169584134372155, + "grad_norm": 3.2634246349334717, + "learning_rate": 1.9787852587341088e-05, + "loss": 0.5428, + "step": 348 + }, + { + "epoch": 0.2825053121521805, + "grad_norm": 3.423222064971924, + "learning_rate": 1.9786057370656474e-05, + "loss": 0.5505, + "step": 349 + }, + { + "epoch": 0.2833147829606395, + "grad_norm": 3.534621238708496, + "learning_rate": 1.9784254672396743e-05, + "loss": 0.5215, + "step": 350 + }, + { + "epoch": 0.28412425376909844, + "grad_norm": 3.1066412925720215, + "learning_rate": 1.978244449394008e-05, + "loss": 0.4772, + "step": 351 + }, + { + "epoch": 0.28493372457755745, + "grad_norm": 3.148733615875244, + "learning_rate": 1.978062683667039e-05, + "loss": 0.5146, + "step": 352 + }, + { + "epoch": 0.2857431953860164, + "grad_norm": 3.1937990188598633, + "learning_rate": 1.9778801701977305e-05, + "loss": 0.5179, + "step": 353 + }, + { + "epoch": 0.28655266619447534, + "grad_norm": 3.123727798461914, + "learning_rate": 1.9776969091256158e-05, + "loss": 0.5512, + "step": 354 + }, + { + "epoch": 0.28736213700293434, + "grad_norm": 2.855808973312378, + "learning_rate": 1.9775129005908006e-05, + "loss": 0.4772, + "step": 355 + }, + { + "epoch": 0.2881716078113933, + "grad_norm": 3.017282009124756, + "learning_rate": 1.977328144733962e-05, + "loss": 0.52, + "step": 356 + }, + { + "epoch": 0.2889810786198523, + "grad_norm": 3.1748759746551514, + "learning_rate": 1.9771426416963488e-05, + "loss": 0.4828, + "step": 357 + }, + { + "epoch": 0.28979054942831123, + "grad_norm": 2.846700668334961, + "learning_rate": 1.97695639161978e-05, + "loss": 0.4951, + "step": 358 + }, + { + "epoch": 0.29060002023677023, + "grad_norm": 2.7805535793304443, + "learning_rate": 1.9767693946466465e-05, + "loss": 0.5116, + "step": 359 + }, + { + "epoch": 0.2914094910452292, + "grad_norm": 5.455881118774414, + "learning_rate": 1.9765816509199105e-05, + "loss": 0.4676, + "step": 360 + }, + { + "epoch": 0.2922189618536882, + "grad_norm": 3.217540740966797, + "learning_rate": 1.9763931605831043e-05, + "loss": 0.4893, + "step": 361 + }, + { + "epoch": 0.2930284326621471, + "grad_norm": 3.29526948928833, + "learning_rate": 1.976203923780331e-05, + "loss": 0.4945, + "step": 362 + }, + { + "epoch": 0.29383790347060607, + "grad_norm": 3.7660229206085205, + "learning_rate": 1.9760139406562656e-05, + "loss": 0.541, + "step": 363 + }, + { + "epoch": 0.29464737427906507, + "grad_norm": 3.15836763381958, + "learning_rate": 1.975823211356152e-05, + "loss": 0.5015, + "step": 364 + }, + { + "epoch": 0.295456845087524, + "grad_norm": 3.3420376777648926, + "learning_rate": 1.9756317360258057e-05, + "loss": 0.5416, + "step": 365 + }, + { + "epoch": 0.296266315895983, + "grad_norm": 3.321540594100952, + "learning_rate": 1.975439514811612e-05, + "loss": 0.5618, + "step": 366 + }, + { + "epoch": 0.29707578670444196, + "grad_norm": 3.7524049282073975, + "learning_rate": 1.9752465478605272e-05, + "loss": 0.5337, + "step": 367 + }, + { + "epoch": 0.29788525751290096, + "grad_norm": 3.4210922718048096, + "learning_rate": 1.9750528353200765e-05, + "loss": 0.5519, + "step": 368 + }, + { + "epoch": 0.2986947283213599, + "grad_norm": 3.680934429168701, + "learning_rate": 1.9748583773383564e-05, + "loss": 0.5261, + "step": 369 + }, + { + "epoch": 0.2995041991298189, + "grad_norm": 2.890639066696167, + "learning_rate": 1.9746631740640323e-05, + "loss": 0.5085, + "step": 370 + }, + { + "epoch": 0.30031366993827785, + "grad_norm": 2.9700279235839844, + "learning_rate": 1.97446722564634e-05, + "loss": 0.5058, + "step": 371 + }, + { + "epoch": 0.3011231407467368, + "grad_norm": 3.1853365898132324, + "learning_rate": 1.9742705322350846e-05, + "loss": 0.5572, + "step": 372 + }, + { + "epoch": 0.3019326115551958, + "grad_norm": 2.932831287384033, + "learning_rate": 1.974073093980641e-05, + "loss": 0.5119, + "step": 373 + }, + { + "epoch": 0.30274208236365474, + "grad_norm": 3.095426559448242, + "learning_rate": 1.973874911033953e-05, + "loss": 0.5136, + "step": 374 + }, + { + "epoch": 0.30355155317211374, + "grad_norm": 3.01311993598938, + "learning_rate": 1.9736759835465352e-05, + "loss": 0.4674, + "step": 375 + }, + { + "epoch": 0.3043610239805727, + "grad_norm": 3.1134865283966064, + "learning_rate": 1.9734763116704692e-05, + "loss": 0.5151, + "step": 376 + }, + { + "epoch": 0.3051704947890317, + "grad_norm": 3.2785098552703857, + "learning_rate": 1.9732758955584082e-05, + "loss": 0.5099, + "step": 377 + }, + { + "epoch": 0.30597996559749063, + "grad_norm": 3.00992488861084, + "learning_rate": 1.973074735363572e-05, + "loss": 0.5003, + "step": 378 + }, + { + "epoch": 0.30678943640594963, + "grad_norm": 3.189427614212036, + "learning_rate": 1.9728728312397507e-05, + "loss": 0.506, + "step": 379 + }, + { + "epoch": 0.3075989072144086, + "grad_norm": 3.337754249572754, + "learning_rate": 1.9726701833413028e-05, + "loss": 0.5356, + "step": 380 + }, + { + "epoch": 0.3084083780228675, + "grad_norm": 3.1951828002929688, + "learning_rate": 1.972466791823155e-05, + "loss": 0.5199, + "step": 381 + }, + { + "epoch": 0.3092178488313265, + "grad_norm": 3.415829658508301, + "learning_rate": 1.9722626568408033e-05, + "loss": 0.5384, + "step": 382 + }, + { + "epoch": 0.31002731963978547, + "grad_norm": 2.920175313949585, + "learning_rate": 1.9720577785503118e-05, + "loss": 0.5237, + "step": 383 + }, + { + "epoch": 0.31083679044824447, + "grad_norm": 3.0148444175720215, + "learning_rate": 1.971852157108313e-05, + "loss": 0.4914, + "step": 384 + }, + { + "epoch": 0.3116462612567034, + "grad_norm": 3.236902952194214, + "learning_rate": 1.9716457926720064e-05, + "loss": 0.5581, + "step": 385 + }, + { + "epoch": 0.3124557320651624, + "grad_norm": 2.9506475925445557, + "learning_rate": 1.971438685399161e-05, + "loss": 0.4892, + "step": 386 + }, + { + "epoch": 0.31326520287362136, + "grad_norm": 2.9730887413024902, + "learning_rate": 1.9712308354481133e-05, + "loss": 0.4843, + "step": 387 + }, + { + "epoch": 0.31407467368208036, + "grad_norm": 3.362948417663574, + "learning_rate": 1.9710222429777673e-05, + "loss": 0.5036, + "step": 388 + }, + { + "epoch": 0.3148841444905393, + "grad_norm": 2.965210437774658, + "learning_rate": 1.9708129081475945e-05, + "loss": 0.4788, + "step": 389 + }, + { + "epoch": 0.31569361529899825, + "grad_norm": 2.8389463424682617, + "learning_rate": 1.9706028311176348e-05, + "loss": 0.4969, + "step": 390 + }, + { + "epoch": 0.31650308610745725, + "grad_norm": 3.218151569366455, + "learning_rate": 1.970392012048495e-05, + "loss": 0.5227, + "step": 391 + }, + { + "epoch": 0.3173125569159162, + "grad_norm": 4.279620170593262, + "learning_rate": 1.970180451101348e-05, + "loss": 0.4862, + "step": 392 + }, + { + "epoch": 0.3181220277243752, + "grad_norm": 2.993753671646118, + "learning_rate": 1.9699681484379364e-05, + "loss": 0.5183, + "step": 393 + }, + { + "epoch": 0.31893149853283415, + "grad_norm": 3.1236953735351562, + "learning_rate": 1.9697551042205682e-05, + "loss": 0.4869, + "step": 394 + }, + { + "epoch": 0.31974096934129315, + "grad_norm": 2.9094808101654053, + "learning_rate": 1.969541318612118e-05, + "loss": 0.5308, + "step": 395 + }, + { + "epoch": 0.3205504401497521, + "grad_norm": 3.1053497791290283, + "learning_rate": 1.9693267917760283e-05, + "loss": 0.4756, + "step": 396 + }, + { + "epoch": 0.3213599109582111, + "grad_norm": 3.341804265975952, + "learning_rate": 1.969111523876308e-05, + "loss": 0.5092, + "step": 397 + }, + { + "epoch": 0.32216938176667004, + "grad_norm": 3.0341951847076416, + "learning_rate": 1.968895515077532e-05, + "loss": 0.5122, + "step": 398 + }, + { + "epoch": 0.322978852575129, + "grad_norm": 2.8221347332000732, + "learning_rate": 1.968678765544842e-05, + "loss": 0.4686, + "step": 399 + }, + { + "epoch": 0.323788323383588, + "grad_norm": 3.785186290740967, + "learning_rate": 1.9684612754439463e-05, + "loss": 0.5152, + "step": 400 + }, + { + "epoch": 0.32459779419204693, + "grad_norm": 2.838338851928711, + "learning_rate": 1.968243044941119e-05, + "loss": 0.4669, + "step": 401 + }, + { + "epoch": 0.32540726500050593, + "grad_norm": 3.022641897201538, + "learning_rate": 1.9680240742032e-05, + "loss": 0.5382, + "step": 402 + }, + { + "epoch": 0.3262167358089649, + "grad_norm": 2.7124810218811035, + "learning_rate": 1.9678043633975956e-05, + "loss": 0.4907, + "step": 403 + }, + { + "epoch": 0.3270262066174239, + "grad_norm": 3.1172637939453125, + "learning_rate": 1.9675839126922777e-05, + "loss": 0.5176, + "step": 404 + }, + { + "epoch": 0.3278356774258828, + "grad_norm": 3.1733920574188232, + "learning_rate": 1.9673627222557844e-05, + "loss": 0.5234, + "step": 405 + }, + { + "epoch": 0.3286451482343418, + "grad_norm": 3.2248072624206543, + "learning_rate": 1.9671407922572178e-05, + "loss": 0.4625, + "step": 406 + }, + { + "epoch": 0.32945461904280077, + "grad_norm": 2.9852468967437744, + "learning_rate": 1.9669181228662473e-05, + "loss": 0.5082, + "step": 407 + }, + { + "epoch": 0.3302640898512597, + "grad_norm": 3.1525166034698486, + "learning_rate": 1.9666947142531062e-05, + "loss": 0.5159, + "step": 408 + }, + { + "epoch": 0.3310735606597187, + "grad_norm": 3.0281875133514404, + "learning_rate": 1.9664705665885938e-05, + "loss": 0.5464, + "step": 409 + }, + { + "epoch": 0.33188303146817766, + "grad_norm": 3.3612561225891113, + "learning_rate": 1.9662456800440738e-05, + "loss": 0.5266, + "step": 410 + }, + { + "epoch": 0.33269250227663666, + "grad_norm": 3.0698964595794678, + "learning_rate": 1.9660200547914753e-05, + "loss": 0.5257, + "step": 411 + }, + { + "epoch": 0.3335019730850956, + "grad_norm": 3.357966661453247, + "learning_rate": 1.9657936910032915e-05, + "loss": 0.5629, + "step": 412 + }, + { + "epoch": 0.3343114438935546, + "grad_norm": 3.102520227432251, + "learning_rate": 1.965566588852581e-05, + "loss": 0.5188, + "step": 413 + }, + { + "epoch": 0.33512091470201355, + "grad_norm": 2.9197115898132324, + "learning_rate": 1.9653387485129665e-05, + "loss": 0.5107, + "step": 414 + }, + { + "epoch": 0.33593038551047255, + "grad_norm": 3.3753085136413574, + "learning_rate": 1.965110170158635e-05, + "loss": 0.536, + "step": 415 + }, + { + "epoch": 0.3367398563189315, + "grad_norm": 3.6513419151306152, + "learning_rate": 1.9648808539643375e-05, + "loss": 0.5169, + "step": 416 + }, + { + "epoch": 0.33754932712739044, + "grad_norm": 3.224358558654785, + "learning_rate": 1.96465080010539e-05, + "loss": 0.5196, + "step": 417 + }, + { + "epoch": 0.33835879793584944, + "grad_norm": 3.370121955871582, + "learning_rate": 1.964420008757671e-05, + "loss": 0.5493, + "step": 418 + }, + { + "epoch": 0.3391682687443084, + "grad_norm": 3.018153429031372, + "learning_rate": 1.9641884800976244e-05, + "loss": 0.5026, + "step": 419 + }, + { + "epoch": 0.3399777395527674, + "grad_norm": 3.3562841415405273, + "learning_rate": 1.9639562143022567e-05, + "loss": 0.5259, + "step": 420 + }, + { + "epoch": 0.34078721036122633, + "grad_norm": 2.9442625045776367, + "learning_rate": 1.9637232115491384e-05, + "loss": 0.5063, + "step": 421 + }, + { + "epoch": 0.34159668116968533, + "grad_norm": 3.495152711868286, + "learning_rate": 1.9634894720164028e-05, + "loss": 0.5322, + "step": 422 + }, + { + "epoch": 0.3424061519781443, + "grad_norm": 2.857074737548828, + "learning_rate": 1.9632549958827476e-05, + "loss": 0.5078, + "step": 423 + }, + { + "epoch": 0.3432156227866033, + "grad_norm": 3.1433255672454834, + "learning_rate": 1.963019783327433e-05, + "loss": 0.5815, + "step": 424 + }, + { + "epoch": 0.3440250935950622, + "grad_norm": 3.2642955780029297, + "learning_rate": 1.962783834530282e-05, + "loss": 0.5111, + "step": 425 + }, + { + "epoch": 0.34483456440352117, + "grad_norm": 3.3745622634887695, + "learning_rate": 1.9625471496716808e-05, + "loss": 0.5249, + "step": 426 + }, + { + "epoch": 0.34564403521198017, + "grad_norm": 2.9430830478668213, + "learning_rate": 1.962309728932578e-05, + "loss": 0.4799, + "step": 427 + }, + { + "epoch": 0.3464535060204391, + "grad_norm": 3.7242469787597656, + "learning_rate": 1.9620715724944855e-05, + "loss": 0.5164, + "step": 428 + }, + { + "epoch": 0.3472629768288981, + "grad_norm": 3.4125308990478516, + "learning_rate": 1.9618326805394765e-05, + "loss": 0.53, + "step": 429 + }, + { + "epoch": 0.34807244763735706, + "grad_norm": 3.112462043762207, + "learning_rate": 1.9615930532501877e-05, + "loss": 0.5046, + "step": 430 + }, + { + "epoch": 0.34888191844581606, + "grad_norm": 3.205561876296997, + "learning_rate": 1.9613526908098174e-05, + "loss": 0.5012, + "step": 431 + }, + { + "epoch": 0.349691389254275, + "grad_norm": 3.172207832336426, + "learning_rate": 1.9611115934021256e-05, + "loss": 0.5393, + "step": 432 + }, + { + "epoch": 0.350500860062734, + "grad_norm": 2.681241512298584, + "learning_rate": 1.960869761211435e-05, + "loss": 0.5041, + "step": 433 + }, + { + "epoch": 0.35131033087119296, + "grad_norm": 3.0296666622161865, + "learning_rate": 1.9606271944226295e-05, + "loss": 0.5209, + "step": 434 + }, + { + "epoch": 0.3521198016796519, + "grad_norm": 3.640713691711426, + "learning_rate": 1.9603838932211544e-05, + "loss": 0.5262, + "step": 435 + }, + { + "epoch": 0.3529292724881109, + "grad_norm": 2.7593679428100586, + "learning_rate": 1.960139857793017e-05, + "loss": 0.4599, + "step": 436 + }, + { + "epoch": 0.35373874329656985, + "grad_norm": 3.012739896774292, + "learning_rate": 1.959895088324786e-05, + "loss": 0.5153, + "step": 437 + }, + { + "epoch": 0.35454821410502885, + "grad_norm": 3.0251686573028564, + "learning_rate": 1.9596495850035914e-05, + "loss": 0.5146, + "step": 438 + }, + { + "epoch": 0.3553576849134878, + "grad_norm": 3.0073513984680176, + "learning_rate": 1.9594033480171227e-05, + "loss": 0.5281, + "step": 439 + }, + { + "epoch": 0.3561671557219468, + "grad_norm": 2.956883668899536, + "learning_rate": 1.9591563775536322e-05, + "loss": 0.5289, + "step": 440 + }, + { + "epoch": 0.35697662653040574, + "grad_norm": 3.270981788635254, + "learning_rate": 1.9589086738019323e-05, + "loss": 0.4838, + "step": 441 + }, + { + "epoch": 0.35778609733886474, + "grad_norm": 2.974882125854492, + "learning_rate": 1.9586602369513958e-05, + "loss": 0.4912, + "step": 442 + }, + { + "epoch": 0.3585955681473237, + "grad_norm": 3.2653326988220215, + "learning_rate": 1.9584110671919553e-05, + "loss": 0.5446, + "step": 443 + }, + { + "epoch": 0.35940503895578263, + "grad_norm": 2.765235185623169, + "learning_rate": 1.9581611647141055e-05, + "loss": 0.5029, + "step": 444 + }, + { + "epoch": 0.36021450976424163, + "grad_norm": 4.9633989334106445, + "learning_rate": 1.9579105297089e-05, + "loss": 0.5058, + "step": 445 + }, + { + "epoch": 0.3610239805727006, + "grad_norm": 3.063628911972046, + "learning_rate": 1.9576591623679528e-05, + "loss": 0.4862, + "step": 446 + }, + { + "epoch": 0.3618334513811596, + "grad_norm": 3.501770496368408, + "learning_rate": 1.957407062883437e-05, + "loss": 0.4882, + "step": 447 + }, + { + "epoch": 0.3626429221896185, + "grad_norm": 3.1768949031829834, + "learning_rate": 1.9571542314480872e-05, + "loss": 0.4876, + "step": 448 + }, + { + "epoch": 0.3634523929980775, + "grad_norm": 3.163228750228882, + "learning_rate": 1.9569006682551958e-05, + "loss": 0.5101, + "step": 449 + }, + { + "epoch": 0.36426186380653647, + "grad_norm": 2.935380458831787, + "learning_rate": 1.9566463734986154e-05, + "loss": 0.5316, + "step": 450 + }, + { + "epoch": 0.36507133461499547, + "grad_norm": 3.0391714572906494, + "learning_rate": 1.9563913473727577e-05, + "loss": 0.5334, + "step": 451 + }, + { + "epoch": 0.3658808054234544, + "grad_norm": 3.0552303791046143, + "learning_rate": 1.956135590072594e-05, + "loss": 0.4982, + "step": 452 + }, + { + "epoch": 0.36669027623191336, + "grad_norm": 5.1559014320373535, + "learning_rate": 1.9558791017936545e-05, + "loss": 0.5239, + "step": 453 + }, + { + "epoch": 0.36749974704037236, + "grad_norm": 3.650991678237915, + "learning_rate": 1.9556218827320275e-05, + "loss": 0.542, + "step": 454 + }, + { + "epoch": 0.3683092178488313, + "grad_norm": 3.2566704750061035, + "learning_rate": 1.9553639330843604e-05, + "loss": 0.506, + "step": 455 + }, + { + "epoch": 0.3691186886572903, + "grad_norm": 3.3874852657318115, + "learning_rate": 1.9551052530478597e-05, + "loss": 0.5498, + "step": 456 + }, + { + "epoch": 0.36992815946574925, + "grad_norm": 3.275639057159424, + "learning_rate": 1.9548458428202896e-05, + "loss": 0.5203, + "step": 457 + }, + { + "epoch": 0.37073763027420825, + "grad_norm": 3.1248674392700195, + "learning_rate": 1.9545857025999728e-05, + "loss": 0.5046, + "step": 458 + }, + { + "epoch": 0.3715471010826672, + "grad_norm": 3.0372114181518555, + "learning_rate": 1.95432483258579e-05, + "loss": 0.4733, + "step": 459 + }, + { + "epoch": 0.3723565718911262, + "grad_norm": 3.0337607860565186, + "learning_rate": 1.95406323297718e-05, + "loss": 0.5251, + "step": 460 + }, + { + "epoch": 0.37316604269958514, + "grad_norm": 3.0167458057403564, + "learning_rate": 1.9538009039741393e-05, + "loss": 0.5169, + "step": 461 + }, + { + "epoch": 0.3739755135080441, + "grad_norm": 2.819174289703369, + "learning_rate": 1.9535378457772223e-05, + "loss": 0.5136, + "step": 462 + }, + { + "epoch": 0.3747849843165031, + "grad_norm": 2.858234167098999, + "learning_rate": 1.95327405858754e-05, + "loss": 0.4704, + "step": 463 + }, + { + "epoch": 0.37559445512496203, + "grad_norm": 3.044003486633301, + "learning_rate": 1.9530095426067616e-05, + "loss": 0.4971, + "step": 464 + }, + { + "epoch": 0.37640392593342104, + "grad_norm": 2.906550407409668, + "learning_rate": 1.9527442980371133e-05, + "loss": 0.4773, + "step": 465 + }, + { + "epoch": 0.37721339674188, + "grad_norm": 3.275589942932129, + "learning_rate": 1.952478325081378e-05, + "loss": 0.5058, + "step": 466 + }, + { + "epoch": 0.378022867550339, + "grad_norm": 2.7474701404571533, + "learning_rate": 1.952211623942896e-05, + "loss": 0.4554, + "step": 467 + }, + { + "epoch": 0.3788323383587979, + "grad_norm": 3.026256561279297, + "learning_rate": 1.9519441948255638e-05, + "loss": 0.4991, + "step": 468 + }, + { + "epoch": 0.3796418091672569, + "grad_norm": 2.939652681350708, + "learning_rate": 1.9516760379338345e-05, + "loss": 0.4845, + "step": 469 + }, + { + "epoch": 0.3804512799757159, + "grad_norm": 3.081845998764038, + "learning_rate": 1.951407153472718e-05, + "loss": 0.5195, + "step": 470 + }, + { + "epoch": 0.3812607507841749, + "grad_norm": 3.2728562355041504, + "learning_rate": 1.95113754164778e-05, + "loss": 0.5479, + "step": 471 + }, + { + "epoch": 0.3820702215926338, + "grad_norm": 3.010256767272949, + "learning_rate": 1.950867202665142e-05, + "loss": 0.5413, + "step": 472 + }, + { + "epoch": 0.38287969240109276, + "grad_norm": 2.7573351860046387, + "learning_rate": 1.9505961367314824e-05, + "loss": 0.506, + "step": 473 + }, + { + "epoch": 0.38368916320955176, + "grad_norm": 2.7718589305877686, + "learning_rate": 1.950324344054035e-05, + "loss": 0.4481, + "step": 474 + }, + { + "epoch": 0.3844986340180107, + "grad_norm": 3.138713836669922, + "learning_rate": 1.9500518248405886e-05, + "loss": 0.5114, + "step": 475 + }, + { + "epoch": 0.3853081048264697, + "grad_norm": 2.6779539585113525, + "learning_rate": 1.9497785792994882e-05, + "loss": 0.4406, + "step": 476 + }, + { + "epoch": 0.38611757563492866, + "grad_norm": 3.110743761062622, + "learning_rate": 1.9495046076396336e-05, + "loss": 0.5096, + "step": 477 + }, + { + "epoch": 0.38692704644338766, + "grad_norm": 3.0556135177612305, + "learning_rate": 1.9492299100704798e-05, + "loss": 0.5002, + "step": 478 + }, + { + "epoch": 0.3877365172518466, + "grad_norm": 3.0118000507354736, + "learning_rate": 1.9489544868020368e-05, + "loss": 0.5024, + "step": 479 + }, + { + "epoch": 0.3885459880603056, + "grad_norm": 3.0608954429626465, + "learning_rate": 1.94867833804487e-05, + "loss": 0.4923, + "step": 480 + }, + { + "epoch": 0.38935545886876455, + "grad_norm": 2.895240306854248, + "learning_rate": 1.9484014640100982e-05, + "loss": 0.4833, + "step": 481 + }, + { + "epoch": 0.3901649296772235, + "grad_norm": 2.92136812210083, + "learning_rate": 1.9481238649093958e-05, + "loss": 0.5245, + "step": 482 + }, + { + "epoch": 0.3909744004856825, + "grad_norm": 2.94343900680542, + "learning_rate": 1.9478455409549907e-05, + "loss": 0.4965, + "step": 483 + }, + { + "epoch": 0.39178387129414144, + "grad_norm": 3.0315639972686768, + "learning_rate": 1.947566492359666e-05, + "loss": 0.5107, + "step": 484 + }, + { + "epoch": 0.39259334210260044, + "grad_norm": 3.2842588424682617, + "learning_rate": 1.947286719336758e-05, + "loss": 0.5205, + "step": 485 + }, + { + "epoch": 0.3934028129110594, + "grad_norm": 2.779125928878784, + "learning_rate": 1.947006222100157e-05, + "loss": 0.4923, + "step": 486 + }, + { + "epoch": 0.3942122837195184, + "grad_norm": 3.4183132648468018, + "learning_rate": 1.946725000864307e-05, + "loss": 0.4873, + "step": 487 + }, + { + "epoch": 0.39502175452797733, + "grad_norm": 3.0404224395751953, + "learning_rate": 1.9464430558442047e-05, + "loss": 0.5223, + "step": 488 + }, + { + "epoch": 0.39583122533643633, + "grad_norm": 2.976628303527832, + "learning_rate": 1.9461603872554025e-05, + "loss": 0.5082, + "step": 489 + }, + { + "epoch": 0.3966406961448953, + "grad_norm": 2.8417575359344482, + "learning_rate": 1.9458769953140032e-05, + "loss": 0.4823, + "step": 490 + }, + { + "epoch": 0.3974501669533542, + "grad_norm": 2.8879899978637695, + "learning_rate": 1.945592880236664e-05, + "loss": 0.4556, + "step": 491 + }, + { + "epoch": 0.3982596377618132, + "grad_norm": 3.1300411224365234, + "learning_rate": 1.9453080422405953e-05, + "loss": 0.4615, + "step": 492 + }, + { + "epoch": 0.39906910857027217, + "grad_norm": 3.2415857315063477, + "learning_rate": 1.945022481543559e-05, + "loss": 0.4941, + "step": 493 + }, + { + "epoch": 0.39987857937873117, + "grad_norm": 2.7292397022247314, + "learning_rate": 1.9447361983638707e-05, + "loss": 0.4919, + "step": 494 + }, + { + "epoch": 0.4006880501871901, + "grad_norm": 2.9145712852478027, + "learning_rate": 1.9444491929203978e-05, + "loss": 0.4943, + "step": 495 + }, + { + "epoch": 0.4014975209956491, + "grad_norm": 2.9586963653564453, + "learning_rate": 1.944161465432559e-05, + "loss": 0.5353, + "step": 496 + }, + { + "epoch": 0.40230699180410806, + "grad_norm": 3.7078516483306885, + "learning_rate": 1.943873016120327e-05, + "loss": 0.5084, + "step": 497 + }, + { + "epoch": 0.40311646261256706, + "grad_norm": 3.036256790161133, + "learning_rate": 1.943583845204225e-05, + "loss": 0.5078, + "step": 498 + }, + { + "epoch": 0.403925933421026, + "grad_norm": 2.8531179428100586, + "learning_rate": 1.9432939529053277e-05, + "loss": 0.5252, + "step": 499 + }, + { + "epoch": 0.40473540422948495, + "grad_norm": 3.0719563961029053, + "learning_rate": 1.9430033394452624e-05, + "loss": 0.4974, + "step": 500 + }, + { + "epoch": 0.40554487503794395, + "grad_norm": 2.836137533187866, + "learning_rate": 1.9427120050462066e-05, + "loss": 0.4992, + "step": 501 + }, + { + "epoch": 0.4063543458464029, + "grad_norm": 2.8626761436462402, + "learning_rate": 1.9424199499308894e-05, + "loss": 0.4877, + "step": 502 + }, + { + "epoch": 0.4071638166548619, + "grad_norm": 2.907041072845459, + "learning_rate": 1.9421271743225913e-05, + "loss": 0.4834, + "step": 503 + }, + { + "epoch": 0.40797328746332084, + "grad_norm": 3.1465632915496826, + "learning_rate": 1.9418336784451428e-05, + "loss": 0.4872, + "step": 504 + }, + { + "epoch": 0.40878275827177984, + "grad_norm": 3.217764377593994, + "learning_rate": 1.941539462522926e-05, + "loss": 0.5484, + "step": 505 + }, + { + "epoch": 0.4095922290802388, + "grad_norm": 3.4946095943450928, + "learning_rate": 1.941244526780873e-05, + "loss": 0.4872, + "step": 506 + }, + { + "epoch": 0.4104016998886978, + "grad_norm": 5.309779644012451, + "learning_rate": 1.940948871444466e-05, + "loss": 0.4743, + "step": 507 + }, + { + "epoch": 0.41121117069715674, + "grad_norm": 4.0193610191345215, + "learning_rate": 1.940652496739737e-05, + "loss": 0.5463, + "step": 508 + }, + { + "epoch": 0.4120206415056157, + "grad_norm": 3.301046133041382, + "learning_rate": 1.9403554028932696e-05, + "loss": 0.5206, + "step": 509 + }, + { + "epoch": 0.4128301123140747, + "grad_norm": 3.6537790298461914, + "learning_rate": 1.9400575901321955e-05, + "loss": 0.5186, + "step": 510 + }, + { + "epoch": 0.4136395831225336, + "grad_norm": 2.8580868244171143, + "learning_rate": 1.9397590586841968e-05, + "loss": 0.5319, + "step": 511 + }, + { + "epoch": 0.41444905393099263, + "grad_norm": 2.752870798110962, + "learning_rate": 1.939459808777505e-05, + "loss": 0.5192, + "step": 512 + }, + { + "epoch": 0.4152585247394516, + "grad_norm": 2.8357656002044678, + "learning_rate": 1.9391598406409004e-05, + "loss": 0.4879, + "step": 513 + }, + { + "epoch": 0.4160679955479106, + "grad_norm": 2.779935836791992, + "learning_rate": 1.938859154503713e-05, + "loss": 0.5049, + "step": 514 + }, + { + "epoch": 0.4168774663563695, + "grad_norm": 2.5529801845550537, + "learning_rate": 1.9385577505958217e-05, + "loss": 0.4735, + "step": 515 + }, + { + "epoch": 0.4176869371648285, + "grad_norm": 2.94710111618042, + "learning_rate": 1.9382556291476535e-05, + "loss": 0.4751, + "step": 516 + }, + { + "epoch": 0.41849640797328747, + "grad_norm": 2.853198289871216, + "learning_rate": 1.937952790390185e-05, + "loss": 0.5177, + "step": 517 + }, + { + "epoch": 0.4193058787817464, + "grad_norm": 2.757351875305176, + "learning_rate": 1.9376492345549396e-05, + "loss": 0.5125, + "step": 518 + }, + { + "epoch": 0.4201153495902054, + "grad_norm": 2.7711410522460938, + "learning_rate": 1.9373449618739903e-05, + "loss": 0.5284, + "step": 519 + }, + { + "epoch": 0.42092482039866436, + "grad_norm": 3.054570198059082, + "learning_rate": 1.9370399725799585e-05, + "loss": 0.5206, + "step": 520 + }, + { + "epoch": 0.42173429120712336, + "grad_norm": 2.8329365253448486, + "learning_rate": 1.9367342669060115e-05, + "loss": 0.4664, + "step": 521 + }, + { + "epoch": 0.4225437620155823, + "grad_norm": 5.834301948547363, + "learning_rate": 1.9364278450858663e-05, + "loss": 0.5361, + "step": 522 + }, + { + "epoch": 0.4233532328240413, + "grad_norm": 2.825716257095337, + "learning_rate": 1.9361207073537865e-05, + "loss": 0.4779, + "step": 523 + }, + { + "epoch": 0.42416270363250025, + "grad_norm": 3.0324203968048096, + "learning_rate": 1.935812853944583e-05, + "loss": 0.4645, + "step": 524 + }, + { + "epoch": 0.42497217444095925, + "grad_norm": 2.834280252456665, + "learning_rate": 1.935504285093614e-05, + "loss": 0.4818, + "step": 525 + }, + { + "epoch": 0.4257816452494182, + "grad_norm": 3.0398454666137695, + "learning_rate": 1.9351950010367843e-05, + "loss": 0.4853, + "step": 526 + }, + { + "epoch": 0.42659111605787714, + "grad_norm": 2.780987024307251, + "learning_rate": 1.9348850020105462e-05, + "loss": 0.5207, + "step": 527 + }, + { + "epoch": 0.42740058686633614, + "grad_norm": 8.881518363952637, + "learning_rate": 1.934574288251898e-05, + "loss": 0.4991, + "step": 528 + }, + { + "epoch": 0.4282100576747951, + "grad_norm": 3.454258918762207, + "learning_rate": 1.9342628599983852e-05, + "loss": 0.5074, + "step": 529 + }, + { + "epoch": 0.4290195284832541, + "grad_norm": 2.9717166423797607, + "learning_rate": 1.933950717488098e-05, + "loss": 0.5176, + "step": 530 + }, + { + "epoch": 0.42982899929171303, + "grad_norm": 3.314215660095215, + "learning_rate": 1.9336378609596737e-05, + "loss": 0.4858, + "step": 531 + }, + { + "epoch": 0.43063847010017203, + "grad_norm": 3.178234100341797, + "learning_rate": 1.9333242906522963e-05, + "loss": 0.5122, + "step": 532 + }, + { + "epoch": 0.431447940908631, + "grad_norm": 2.8029940128326416, + "learning_rate": 1.933010006805694e-05, + "loss": 0.5188, + "step": 533 + }, + { + "epoch": 0.43225741171709, + "grad_norm": 3.007542610168457, + "learning_rate": 1.932695009660141e-05, + "loss": 0.4833, + "step": 534 + }, + { + "epoch": 0.4330668825255489, + "grad_norm": 2.933992385864258, + "learning_rate": 1.9323792994564576e-05, + "loss": 0.5049, + "step": 535 + }, + { + "epoch": 0.43387635333400787, + "grad_norm": 3.024954080581665, + "learning_rate": 1.9320628764360078e-05, + "loss": 0.524, + "step": 536 + }, + { + "epoch": 0.43468582414246687, + "grad_norm": 2.8064894676208496, + "learning_rate": 1.931745740840702e-05, + "loss": 0.5235, + "step": 537 + }, + { + "epoch": 0.4354952949509258, + "grad_norm": 3.199556827545166, + "learning_rate": 1.931427892912994e-05, + "loss": 0.5329, + "step": 538 + }, + { + "epoch": 0.4363047657593848, + "grad_norm": 2.762144088745117, + "learning_rate": 1.9311093328958838e-05, + "loss": 0.5007, + "step": 539 + }, + { + "epoch": 0.43711423656784376, + "grad_norm": 2.891429901123047, + "learning_rate": 1.930790061032914e-05, + "loss": 0.4524, + "step": 540 + }, + { + "epoch": 0.43792370737630276, + "grad_norm": 3.191375732421875, + "learning_rate": 1.9304700775681726e-05, + "loss": 0.5532, + "step": 541 + }, + { + "epoch": 0.4387331781847617, + "grad_norm": 3.210228681564331, + "learning_rate": 1.9301493827462917e-05, + "loss": 0.5476, + "step": 542 + }, + { + "epoch": 0.4395426489932207, + "grad_norm": 3.1300623416900635, + "learning_rate": 1.9298279768124466e-05, + "loss": 0.5197, + "step": 543 + }, + { + "epoch": 0.44035211980167965, + "grad_norm": 2.911524534225464, + "learning_rate": 1.9295058600123568e-05, + "loss": 0.5207, + "step": 544 + }, + { + "epoch": 0.4411615906101386, + "grad_norm": 2.7098217010498047, + "learning_rate": 1.9291830325922848e-05, + "loss": 0.5079, + "step": 545 + }, + { + "epoch": 0.4419710614185976, + "grad_norm": 4.508588790893555, + "learning_rate": 1.928859494799036e-05, + "loss": 0.5167, + "step": 546 + }, + { + "epoch": 0.44278053222705654, + "grad_norm": 2.7942216396331787, + "learning_rate": 1.9285352468799606e-05, + "loss": 0.493, + "step": 547 + }, + { + "epoch": 0.44359000303551555, + "grad_norm": 2.807497024536133, + "learning_rate": 1.9282102890829502e-05, + "loss": 0.4973, + "step": 548 + }, + { + "epoch": 0.4443994738439745, + "grad_norm": 3.100604772567749, + "learning_rate": 1.927884621656439e-05, + "loss": 0.5127, + "step": 549 + }, + { + "epoch": 0.4452089446524335, + "grad_norm": 2.9301061630249023, + "learning_rate": 1.9275582448494052e-05, + "loss": 0.4857, + "step": 550 + }, + { + "epoch": 0.44601841546089244, + "grad_norm": 3.113266706466675, + "learning_rate": 1.9272311589113673e-05, + "loss": 0.51, + "step": 551 + }, + { + "epoch": 0.44682788626935144, + "grad_norm": 2.4865670204162598, + "learning_rate": 1.9269033640923873e-05, + "loss": 0.4372, + "step": 552 + }, + { + "epoch": 0.4476373570778104, + "grad_norm": 2.669595241546631, + "learning_rate": 1.926574860643069e-05, + "loss": 0.5086, + "step": 553 + }, + { + "epoch": 0.44844682788626933, + "grad_norm": 2.7644054889678955, + "learning_rate": 1.926245648814558e-05, + "loss": 0.5028, + "step": 554 + }, + { + "epoch": 0.44925629869472833, + "grad_norm": 2.8249318599700928, + "learning_rate": 1.925915728858541e-05, + "loss": 0.5064, + "step": 555 + }, + { + "epoch": 0.4500657695031873, + "grad_norm": 2.716909170150757, + "learning_rate": 1.925585101027246e-05, + "loss": 0.5024, + "step": 556 + }, + { + "epoch": 0.4508752403116463, + "grad_norm": 2.973722219467163, + "learning_rate": 1.9252537655734432e-05, + "loss": 0.5309, + "step": 557 + }, + { + "epoch": 0.4516847111201052, + "grad_norm": 2.7612786293029785, + "learning_rate": 1.9249217227504425e-05, + "loss": 0.4871, + "step": 558 + }, + { + "epoch": 0.4524941819285642, + "grad_norm": 2.964996576309204, + "learning_rate": 1.9245889728120954e-05, + "loss": 0.5003, + "step": 559 + }, + { + "epoch": 0.45330365273702317, + "grad_norm": 2.9560186862945557, + "learning_rate": 1.924255516012794e-05, + "loss": 0.495, + "step": 560 + }, + { + "epoch": 0.45411312354548217, + "grad_norm": 3.0016977787017822, + "learning_rate": 1.9239213526074704e-05, + "loss": 0.489, + "step": 561 + }, + { + "epoch": 0.4549225943539411, + "grad_norm": 2.7454395294189453, + "learning_rate": 1.923586482851597e-05, + "loss": 0.4873, + "step": 562 + }, + { + "epoch": 0.45573206516240006, + "grad_norm": 2.6640262603759766, + "learning_rate": 1.9232509070011868e-05, + "loss": 0.4642, + "step": 563 + }, + { + "epoch": 0.45654153597085906, + "grad_norm": 2.82383394241333, + "learning_rate": 1.9229146253127917e-05, + "loss": 0.4651, + "step": 564 + }, + { + "epoch": 0.457351006779318, + "grad_norm": 2.7471587657928467, + "learning_rate": 1.922577638043504e-05, + "loss": 0.524, + "step": 565 + }, + { + "epoch": 0.458160477587777, + "grad_norm": 2.9681692123413086, + "learning_rate": 1.9222399454509543e-05, + "loss": 0.539, + "step": 566 + }, + { + "epoch": 0.45896994839623595, + "grad_norm": 2.827547550201416, + "learning_rate": 1.921901547793314e-05, + "loss": 0.506, + "step": 567 + }, + { + "epoch": 0.45977941920469495, + "grad_norm": 2.6503727436065674, + "learning_rate": 1.9215624453292926e-05, + "loss": 0.4917, + "step": 568 + }, + { + "epoch": 0.4605888900131539, + "grad_norm": 2.928602457046509, + "learning_rate": 1.9212226383181385e-05, + "loss": 0.4899, + "step": 569 + }, + { + "epoch": 0.4613983608216129, + "grad_norm": 2.739837169647217, + "learning_rate": 1.9208821270196387e-05, + "loss": 0.4698, + "step": 570 + }, + { + "epoch": 0.46220783163007184, + "grad_norm": 2.9058072566986084, + "learning_rate": 1.9205409116941195e-05, + "loss": 0.5102, + "step": 571 + }, + { + "epoch": 0.4630173024385308, + "grad_norm": 2.9524176120758057, + "learning_rate": 1.9201989926024435e-05, + "loss": 0.5032, + "step": 572 + }, + { + "epoch": 0.4638267732469898, + "grad_norm": 4.0150628089904785, + "learning_rate": 1.9198563700060132e-05, + "loss": 0.5632, + "step": 573 + }, + { + "epoch": 0.46463624405544873, + "grad_norm": 2.689819812774658, + "learning_rate": 1.9195130441667686e-05, + "loss": 0.4647, + "step": 574 + }, + { + "epoch": 0.46544571486390773, + "grad_norm": 3.0407373905181885, + "learning_rate": 1.919169015347186e-05, + "loss": 0.5093, + "step": 575 + }, + { + "epoch": 0.4662551856723667, + "grad_norm": 2.759039878845215, + "learning_rate": 1.918824283810281e-05, + "loss": 0.5141, + "step": 576 + }, + { + "epoch": 0.4670646564808257, + "grad_norm": 45.922149658203125, + "learning_rate": 1.9184788498196055e-05, + "loss": 0.5274, + "step": 577 + }, + { + "epoch": 0.4678741272892846, + "grad_norm": 8.53474235534668, + "learning_rate": 1.9181327136392484e-05, + "loss": 0.492, + "step": 578 + }, + { + "epoch": 0.4686835980977436, + "grad_norm": 3.0021677017211914, + "learning_rate": 1.9177858755338358e-05, + "loss": 0.4857, + "step": 579 + }, + { + "epoch": 0.46949306890620257, + "grad_norm": 3.056568145751953, + "learning_rate": 1.9174383357685294e-05, + "loss": 0.5002, + "step": 580 + }, + { + "epoch": 0.4703025397146615, + "grad_norm": 2.980512857437134, + "learning_rate": 1.9170900946090292e-05, + "loss": 0.511, + "step": 581 + }, + { + "epoch": 0.4711120105231205, + "grad_norm": 2.797494411468506, + "learning_rate": 1.9167411523215693e-05, + "loss": 0.4852, + "step": 582 + }, + { + "epoch": 0.47192148133157946, + "grad_norm": 2.9639394283294678, + "learning_rate": 1.916391509172922e-05, + "loss": 0.5023, + "step": 583 + }, + { + "epoch": 0.47273095214003846, + "grad_norm": 3.005307912826538, + "learning_rate": 1.916041165430393e-05, + "loss": 0.5458, + "step": 584 + }, + { + "epoch": 0.4735404229484974, + "grad_norm": 2.7839038372039795, + "learning_rate": 1.915690121361826e-05, + "loss": 0.4714, + "step": 585 + }, + { + "epoch": 0.4743498937569564, + "grad_norm": 2.8450124263763428, + "learning_rate": 1.9153383772355984e-05, + "loss": 0.4941, + "step": 586 + }, + { + "epoch": 0.47515936456541535, + "grad_norm": 2.703827142715454, + "learning_rate": 1.9149859333206233e-05, + "loss": 0.5009, + "step": 587 + }, + { + "epoch": 0.47596883537387435, + "grad_norm": 2.8442089557647705, + "learning_rate": 1.9146327898863493e-05, + "loss": 0.5086, + "step": 588 + }, + { + "epoch": 0.4767783061823333, + "grad_norm": 2.8923816680908203, + "learning_rate": 1.9142789472027594e-05, + "loss": 0.5058, + "step": 589 + }, + { + "epoch": 0.47758777699079225, + "grad_norm": 2.6244001388549805, + "learning_rate": 1.9139244055403712e-05, + "loss": 0.4739, + "step": 590 + }, + { + "epoch": 0.47839724779925125, + "grad_norm": 2.707812786102295, + "learning_rate": 1.9135691651702365e-05, + "loss": 0.4933, + "step": 591 + }, + { + "epoch": 0.4792067186077102, + "grad_norm": 2.5218496322631836, + "learning_rate": 1.9132132263639415e-05, + "loss": 0.4981, + "step": 592 + }, + { + "epoch": 0.4800161894161692, + "grad_norm": 2.877865791320801, + "learning_rate": 1.9128565893936058e-05, + "loss": 0.5033, + "step": 593 + }, + { + "epoch": 0.48082566022462814, + "grad_norm": 2.891901731491089, + "learning_rate": 1.912499254531884e-05, + "loss": 0.5328, + "step": 594 + }, + { + "epoch": 0.48163513103308714, + "grad_norm": 2.7808306217193604, + "learning_rate": 1.9121412220519633e-05, + "loss": 0.4864, + "step": 595 + }, + { + "epoch": 0.4824446018415461, + "grad_norm": 2.690511465072632, + "learning_rate": 1.9117824922275644e-05, + "loss": 0.4738, + "step": 596 + }, + { + "epoch": 0.4832540726500051, + "grad_norm": 2.738534927368164, + "learning_rate": 1.911423065332941e-05, + "loss": 0.479, + "step": 597 + }, + { + "epoch": 0.48406354345846403, + "grad_norm": 2.920228958129883, + "learning_rate": 1.9110629416428802e-05, + "loss": 0.5306, + "step": 598 + }, + { + "epoch": 0.484873014266923, + "grad_norm": 2.7999422550201416, + "learning_rate": 1.9107021214327013e-05, + "loss": 0.4935, + "step": 599 + }, + { + "epoch": 0.485682485075382, + "grad_norm": 2.555917263031006, + "learning_rate": 1.9103406049782563e-05, + "loss": 0.4588, + "step": 600 + }, + { + "epoch": 0.4864919558838409, + "grad_norm": 2.672825336456299, + "learning_rate": 1.9099783925559295e-05, + "loss": 0.4854, + "step": 601 + }, + { + "epoch": 0.4873014266922999, + "grad_norm": 2.677549123764038, + "learning_rate": 1.9096154844426372e-05, + "loss": 0.4966, + "step": 602 + }, + { + "epoch": 0.48811089750075887, + "grad_norm": 2.9087677001953125, + "learning_rate": 1.909251880915828e-05, + "loss": 0.4727, + "step": 603 + }, + { + "epoch": 0.48892036830921787, + "grad_norm": 2.7944176197052, + "learning_rate": 1.908887582253481e-05, + "loss": 0.4534, + "step": 604 + }, + { + "epoch": 0.4897298391176768, + "grad_norm": 2.746797800064087, + "learning_rate": 1.9085225887341087e-05, + "loss": 0.4949, + "step": 605 + }, + { + "epoch": 0.4905393099261358, + "grad_norm": 2.894016742706299, + "learning_rate": 1.9081569006367525e-05, + "loss": 0.4769, + "step": 606 + }, + { + "epoch": 0.49134878073459476, + "grad_norm": 3.3397300243377686, + "learning_rate": 1.9077905182409866e-05, + "loss": 0.5352, + "step": 607 + }, + { + "epoch": 0.4921582515430537, + "grad_norm": 2.9913175106048584, + "learning_rate": 1.9074234418269153e-05, + "loss": 0.5174, + "step": 608 + }, + { + "epoch": 0.4929677223515127, + "grad_norm": 2.635820150375366, + "learning_rate": 1.9070556716751734e-05, + "loss": 0.4656, + "step": 609 + }, + { + "epoch": 0.49377719315997165, + "grad_norm": 2.9530270099639893, + "learning_rate": 1.9066872080669262e-05, + "loss": 0.5093, + "step": 610 + }, + { + "epoch": 0.49458666396843065, + "grad_norm": 2.6430680751800537, + "learning_rate": 1.9063180512838696e-05, + "loss": 0.501, + "step": 611 + }, + { + "epoch": 0.4953961347768896, + "grad_norm": 2.7677369117736816, + "learning_rate": 1.9059482016082286e-05, + "loss": 0.4967, + "step": 612 + }, + { + "epoch": 0.4962056055853486, + "grad_norm": 2.5820152759552, + "learning_rate": 1.9055776593227584e-05, + "loss": 0.4754, + "step": 613 + }, + { + "epoch": 0.49701507639380754, + "grad_norm": 2.848311185836792, + "learning_rate": 1.9052064247107437e-05, + "loss": 0.5295, + "step": 614 + }, + { + "epoch": 0.49782454720226654, + "grad_norm": 2.770975351333618, + "learning_rate": 1.9048344980559986e-05, + "loss": 0.4617, + "step": 615 + }, + { + "epoch": 0.4986340180107255, + "grad_norm": 2.7413315773010254, + "learning_rate": 1.9044618796428658e-05, + "loss": 0.5009, + "step": 616 + }, + { + "epoch": 0.49944348881918443, + "grad_norm": 2.8134844303131104, + "learning_rate": 1.9040885697562174e-05, + "loss": 0.4967, + "step": 617 + }, + { + "epoch": 0.5002529596276434, + "grad_norm": 2.6783790588378906, + "learning_rate": 1.903714568681454e-05, + "loss": 0.4959, + "step": 618 + }, + { + "epoch": 0.5010624304361024, + "grad_norm": 2.8186450004577637, + "learning_rate": 1.9033398767045045e-05, + "loss": 0.4918, + "step": 619 + }, + { + "epoch": 0.5018719012445614, + "grad_norm": 3.23675537109375, + "learning_rate": 1.9029644941118262e-05, + "loss": 0.5145, + "step": 620 + }, + { + "epoch": 0.5026813720530203, + "grad_norm": 2.542092800140381, + "learning_rate": 1.9025884211904038e-05, + "loss": 0.478, + "step": 621 + }, + { + "epoch": 0.5034908428614793, + "grad_norm": 3.0922930240631104, + "learning_rate": 1.9022116582277505e-05, + "loss": 0.4965, + "step": 622 + }, + { + "epoch": 0.5043003136699383, + "grad_norm": 3.3059985637664795, + "learning_rate": 1.901834205511907e-05, + "loss": 0.5125, + "step": 623 + }, + { + "epoch": 0.5051097844783973, + "grad_norm": 2.7035140991210938, + "learning_rate": 1.901456063331441e-05, + "loss": 0.4992, + "step": 624 + }, + { + "epoch": 0.5059192552868562, + "grad_norm": 2.9304909706115723, + "learning_rate": 1.9010772319754464e-05, + "loss": 0.5222, + "step": 625 + }, + { + "epoch": 0.5067287260953152, + "grad_norm": 2.8924059867858887, + "learning_rate": 1.9006977117335466e-05, + "loss": 0.5258, + "step": 626 + }, + { + "epoch": 0.5075381969037741, + "grad_norm": 9.207565307617188, + "learning_rate": 1.9003175028958887e-05, + "loss": 0.4793, + "step": 627 + }, + { + "epoch": 0.5083476677122332, + "grad_norm": 2.809640407562256, + "learning_rate": 1.899936605753148e-05, + "loss": 0.4903, + "step": 628 + }, + { + "epoch": 0.5091571385206921, + "grad_norm": 2.9886538982391357, + "learning_rate": 1.899555020596526e-05, + "loss": 0.5298, + "step": 629 + }, + { + "epoch": 0.509966609329151, + "grad_norm": 2.721402168273926, + "learning_rate": 1.8991727477177492e-05, + "loss": 0.4858, + "step": 630 + }, + { + "epoch": 0.51077608013761, + "grad_norm": 2.742741107940674, + "learning_rate": 1.8987897874090713e-05, + "loss": 0.4552, + "step": 631 + }, + { + "epoch": 0.5115855509460691, + "grad_norm": 2.8818297386169434, + "learning_rate": 1.8984061399632696e-05, + "loss": 0.5268, + "step": 632 + }, + { + "epoch": 0.512395021754528, + "grad_norm": 2.74894642829895, + "learning_rate": 1.8980218056736484e-05, + "loss": 0.5065, + "step": 633 + }, + { + "epoch": 0.513204492562987, + "grad_norm": 2.8183043003082275, + "learning_rate": 1.8976367848340366e-05, + "loss": 0.4744, + "step": 634 + }, + { + "epoch": 0.5140139633714459, + "grad_norm": 2.7935359477996826, + "learning_rate": 1.8972510777387878e-05, + "loss": 0.4812, + "step": 635 + }, + { + "epoch": 0.5148234341799048, + "grad_norm": 2.913808822631836, + "learning_rate": 1.8968646846827802e-05, + "loss": 0.5206, + "step": 636 + }, + { + "epoch": 0.5156329049883639, + "grad_norm": 2.6514201164245605, + "learning_rate": 1.896477605961417e-05, + "loss": 0.4793, + "step": 637 + }, + { + "epoch": 0.5164423757968228, + "grad_norm": 2.93558931350708, + "learning_rate": 1.896089841870625e-05, + "loss": 0.5523, + "step": 638 + }, + { + "epoch": 0.5172518466052818, + "grad_norm": 2.6415984630584717, + "learning_rate": 1.895701392706855e-05, + "loss": 0.4748, + "step": 639 + }, + { + "epoch": 0.5180613174137407, + "grad_norm": 2.7673261165618896, + "learning_rate": 1.895312258767082e-05, + "loss": 0.4991, + "step": 640 + }, + { + "epoch": 0.5188707882221998, + "grad_norm": 2.6211867332458496, + "learning_rate": 1.8949224403488038e-05, + "loss": 0.5184, + "step": 641 + }, + { + "epoch": 0.5196802590306587, + "grad_norm": 2.7133524417877197, + "learning_rate": 1.8945319377500425e-05, + "loss": 0.4689, + "step": 642 + }, + { + "epoch": 0.5204897298391177, + "grad_norm": 2.8613154888153076, + "learning_rate": 1.894140751269342e-05, + "loss": 0.4784, + "step": 643 + }, + { + "epoch": 0.5212992006475766, + "grad_norm": 2.9250259399414062, + "learning_rate": 1.8937488812057703e-05, + "loss": 0.5221, + "step": 644 + }, + { + "epoch": 0.5221086714560356, + "grad_norm": 2.7674083709716797, + "learning_rate": 1.8933563278589177e-05, + "loss": 0.5039, + "step": 645 + }, + { + "epoch": 0.5229181422644946, + "grad_norm": 6.899701118469238, + "learning_rate": 1.8929630915288957e-05, + "loss": 0.5005, + "step": 646 + }, + { + "epoch": 0.5237276130729536, + "grad_norm": 2.670905590057373, + "learning_rate": 1.8925691725163395e-05, + "loss": 0.4793, + "step": 647 + }, + { + "epoch": 0.5245370838814125, + "grad_norm": 2.829854965209961, + "learning_rate": 1.8921745711224057e-05, + "loss": 0.4884, + "step": 648 + }, + { + "epoch": 0.5253465546898715, + "grad_norm": 2.9477953910827637, + "learning_rate": 1.891779287648772e-05, + "loss": 0.5018, + "step": 649 + }, + { + "epoch": 0.5261560254983305, + "grad_norm": 3.05583119392395, + "learning_rate": 1.8913833223976384e-05, + "loss": 0.5078, + "step": 650 + }, + { + "epoch": 0.5269654963067895, + "grad_norm": 2.7035939693450928, + "learning_rate": 1.8909866756717257e-05, + "loss": 0.4697, + "step": 651 + }, + { + "epoch": 0.5277749671152484, + "grad_norm": 2.854893922805786, + "learning_rate": 1.8905893477742753e-05, + "loss": 0.4878, + "step": 652 + }, + { + "epoch": 0.5285844379237074, + "grad_norm": 2.7845396995544434, + "learning_rate": 1.8901913390090505e-05, + "loss": 0.5011, + "step": 653 + }, + { + "epoch": 0.5293939087321663, + "grad_norm": 2.9878756999969482, + "learning_rate": 1.8897926496803345e-05, + "loss": 0.5217, + "step": 654 + }, + { + "epoch": 0.5302033795406254, + "grad_norm": 2.82242488861084, + "learning_rate": 1.88939328009293e-05, + "loss": 0.5067, + "step": 655 + }, + { + "epoch": 0.5310128503490843, + "grad_norm": 2.8448617458343506, + "learning_rate": 1.8889932305521614e-05, + "loss": 0.4846, + "step": 656 + }, + { + "epoch": 0.5318223211575432, + "grad_norm": 2.6906208992004395, + "learning_rate": 1.888592501363871e-05, + "loss": 0.5037, + "step": 657 + }, + { + "epoch": 0.5326317919660022, + "grad_norm": 2.570463180541992, + "learning_rate": 1.8881910928344228e-05, + "loss": 0.4717, + "step": 658 + }, + { + "epoch": 0.5334412627744612, + "grad_norm": 2.739077091217041, + "learning_rate": 1.8877890052706988e-05, + "loss": 0.5174, + "step": 659 + }, + { + "epoch": 0.5342507335829202, + "grad_norm": 2.607880115509033, + "learning_rate": 1.8873862389801003e-05, + "loss": 0.4935, + "step": 660 + }, + { + "epoch": 0.5350602043913791, + "grad_norm": 2.5568504333496094, + "learning_rate": 1.886982794270548e-05, + "loss": 0.4746, + "step": 661 + }, + { + "epoch": 0.5358696751998381, + "grad_norm": 2.7721047401428223, + "learning_rate": 1.8865786714504803e-05, + "loss": 0.531, + "step": 662 + }, + { + "epoch": 0.536679146008297, + "grad_norm": 2.7260801792144775, + "learning_rate": 1.886173870828855e-05, + "loss": 0.4823, + "step": 663 + }, + { + "epoch": 0.5374886168167561, + "grad_norm": 3.3157103061676025, + "learning_rate": 1.885768392715148e-05, + "loss": 0.4653, + "step": 664 + }, + { + "epoch": 0.538298087625215, + "grad_norm": 2.7138195037841797, + "learning_rate": 1.8853622374193527e-05, + "loss": 0.4559, + "step": 665 + }, + { + "epoch": 0.539107558433674, + "grad_norm": 2.6963934898376465, + "learning_rate": 1.8849554052519798e-05, + "loss": 0.522, + "step": 666 + }, + { + "epoch": 0.5399170292421329, + "grad_norm": 2.7139594554901123, + "learning_rate": 1.884547896524059e-05, + "loss": 0.4557, + "step": 667 + }, + { + "epoch": 0.540726500050592, + "grad_norm": 2.9217121601104736, + "learning_rate": 1.8841397115471357e-05, + "loss": 0.5075, + "step": 668 + }, + { + "epoch": 0.5415359708590509, + "grad_norm": 2.823925495147705, + "learning_rate": 1.8837308506332733e-05, + "loss": 0.4711, + "step": 669 + }, + { + "epoch": 0.5423454416675099, + "grad_norm": 2.835688591003418, + "learning_rate": 1.883321314095051e-05, + "loss": 0.5001, + "step": 670 + }, + { + "epoch": 0.5431549124759688, + "grad_norm": 2.974424362182617, + "learning_rate": 1.882911102245566e-05, + "loss": 0.4951, + "step": 671 + }, + { + "epoch": 0.5439643832844278, + "grad_norm": 2.5430188179016113, + "learning_rate": 1.88250021539843e-05, + "loss": 0.4956, + "step": 672 + }, + { + "epoch": 0.5447738540928868, + "grad_norm": 2.5322115421295166, + "learning_rate": 1.8820886538677724e-05, + "loss": 0.5018, + "step": 673 + }, + { + "epoch": 0.5455833249013458, + "grad_norm": 2.732851028442383, + "learning_rate": 1.881676417968237e-05, + "loss": 0.4851, + "step": 674 + }, + { + "epoch": 0.5463927957098047, + "grad_norm": 3.0305449962615967, + "learning_rate": 1.881263508014984e-05, + "loss": 0.5282, + "step": 675 + }, + { + "epoch": 0.5472022665182636, + "grad_norm": 3.4129183292388916, + "learning_rate": 1.880849924323689e-05, + "loss": 0.4799, + "step": 676 + }, + { + "epoch": 0.5480117373267227, + "grad_norm": 2.719212293624878, + "learning_rate": 1.880435667210542e-05, + "loss": 0.4583, + "step": 677 + }, + { + "epoch": 0.5488212081351816, + "grad_norm": 2.7941267490386963, + "learning_rate": 1.8800207369922486e-05, + "loss": 0.4941, + "step": 678 + }, + { + "epoch": 0.5496306789436406, + "grad_norm": 2.6392111778259277, + "learning_rate": 1.879605133986028e-05, + "loss": 0.4829, + "step": 679 + }, + { + "epoch": 0.5504401497520995, + "grad_norm": 2.5776116847991943, + "learning_rate": 1.8791888585096153e-05, + "loss": 0.5162, + "step": 680 + }, + { + "epoch": 0.5512496205605585, + "grad_norm": 2.3474395275115967, + "learning_rate": 1.8787719108812585e-05, + "loss": 0.4747, + "step": 681 + }, + { + "epoch": 0.5520590913690175, + "grad_norm": 2.6783103942871094, + "learning_rate": 1.8783542914197194e-05, + "loss": 0.4709, + "step": 682 + }, + { + "epoch": 0.5528685621774765, + "grad_norm": 2.5588040351867676, + "learning_rate": 1.877936000444274e-05, + "loss": 0.4669, + "step": 683 + }, + { + "epoch": 0.5536780329859354, + "grad_norm": 2.728844404220581, + "learning_rate": 1.8775170382747115e-05, + "loss": 0.4645, + "step": 684 + }, + { + "epoch": 0.5544875037943944, + "grad_norm": 2.3917770385742188, + "learning_rate": 1.8770974052313347e-05, + "loss": 0.4721, + "step": 685 + }, + { + "epoch": 0.5552969746028534, + "grad_norm": 2.472500801086426, + "learning_rate": 1.8766771016349584e-05, + "loss": 0.4793, + "step": 686 + }, + { + "epoch": 0.5561064454113124, + "grad_norm": 2.517792224884033, + "learning_rate": 1.87625612780691e-05, + "loss": 0.4431, + "step": 687 + }, + { + "epoch": 0.5569159162197713, + "grad_norm": 2.8730733394622803, + "learning_rate": 1.8758344840690307e-05, + "loss": 0.4861, + "step": 688 + }, + { + "epoch": 0.5577253870282303, + "grad_norm": 2.7957956790924072, + "learning_rate": 1.875412170743672e-05, + "loss": 0.5073, + "step": 689 + }, + { + "epoch": 0.5585348578366892, + "grad_norm": 2.6837213039398193, + "learning_rate": 1.874989188153699e-05, + "loss": 0.4901, + "step": 690 + }, + { + "epoch": 0.5593443286451483, + "grad_norm": 2.6683971881866455, + "learning_rate": 1.8745655366224872e-05, + "loss": 0.5141, + "step": 691 + }, + { + "epoch": 0.5601537994536072, + "grad_norm": 2.5476932525634766, + "learning_rate": 1.8741412164739244e-05, + "loss": 0.5026, + "step": 692 + }, + { + "epoch": 0.5609632702620662, + "grad_norm": 2.5970840454101562, + "learning_rate": 1.873716228032409e-05, + "loss": 0.5117, + "step": 693 + }, + { + "epoch": 0.5617727410705251, + "grad_norm": 2.6278839111328125, + "learning_rate": 1.87329057162285e-05, + "loss": 0.551, + "step": 694 + }, + { + "epoch": 0.5625822118789842, + "grad_norm": 2.5984420776367188, + "learning_rate": 1.8728642475706687e-05, + "loss": 0.4716, + "step": 695 + }, + { + "epoch": 0.5633916826874431, + "grad_norm": 2.5453920364379883, + "learning_rate": 1.8724372562017948e-05, + "loss": 0.4751, + "step": 696 + }, + { + "epoch": 0.564201153495902, + "grad_norm": 2.753445863723755, + "learning_rate": 1.872009597842669e-05, + "loss": 0.4741, + "step": 697 + }, + { + "epoch": 0.565010624304361, + "grad_norm": 2.630821466445923, + "learning_rate": 1.8715812728202425e-05, + "loss": 0.5068, + "step": 698 + }, + { + "epoch": 0.5658200951128199, + "grad_norm": 2.7918753623962402, + "learning_rate": 1.8711522814619753e-05, + "loss": 0.5018, + "step": 699 + }, + { + "epoch": 0.566629565921279, + "grad_norm": 2.615885019302368, + "learning_rate": 1.8707226240958373e-05, + "loss": 0.492, + "step": 700 + }, + { + "epoch": 0.567439036729738, + "grad_norm": 3.6492109298706055, + "learning_rate": 1.8702923010503076e-05, + "loss": 0.4989, + "step": 701 + }, + { + "epoch": 0.5682485075381969, + "grad_norm": 2.8151535987854004, + "learning_rate": 1.8698613126543737e-05, + "loss": 0.5389, + "step": 702 + }, + { + "epoch": 0.5690579783466558, + "grad_norm": 2.373912811279297, + "learning_rate": 1.8694296592375325e-05, + "loss": 0.4661, + "step": 703 + }, + { + "epoch": 0.5698674491551149, + "grad_norm": 2.8038759231567383, + "learning_rate": 1.868997341129789e-05, + "loss": 0.4762, + "step": 704 + }, + { + "epoch": 0.5706769199635738, + "grad_norm": 2.7435057163238525, + "learning_rate": 1.8685643586616558e-05, + "loss": 0.5156, + "step": 705 + }, + { + "epoch": 0.5714863907720328, + "grad_norm": 2.583699941635132, + "learning_rate": 1.8681307121641544e-05, + "loss": 0.4916, + "step": 706 + }, + { + "epoch": 0.5722958615804917, + "grad_norm": 2.7841830253601074, + "learning_rate": 1.867696401968813e-05, + "loss": 0.5306, + "step": 707 + }, + { + "epoch": 0.5731053323889507, + "grad_norm": 2.663886308670044, + "learning_rate": 1.8672614284076686e-05, + "loss": 0.4911, + "step": 708 + }, + { + "epoch": 0.5739148031974097, + "grad_norm": 3.390315055847168, + "learning_rate": 1.866825791813264e-05, + "loss": 0.4708, + "step": 709 + }, + { + "epoch": 0.5747242740058687, + "grad_norm": 3.3997390270233154, + "learning_rate": 1.866389492518649e-05, + "loss": 0.5394, + "step": 710 + }, + { + "epoch": 0.5755337448143276, + "grad_norm": 2.57348370552063, + "learning_rate": 1.8659525308573806e-05, + "loss": 0.4677, + "step": 711 + }, + { + "epoch": 0.5763432156227866, + "grad_norm": 3.006875991821289, + "learning_rate": 1.8655149071635224e-05, + "loss": 0.5197, + "step": 712 + }, + { + "epoch": 0.5771526864312456, + "grad_norm": 2.504939079284668, + "learning_rate": 1.865076621771643e-05, + "loss": 0.464, + "step": 713 + }, + { + "epoch": 0.5779621572397046, + "grad_norm": 2.6375057697296143, + "learning_rate": 1.8646376750168183e-05, + "loss": 0.4962, + "step": 714 + }, + { + "epoch": 0.5787716280481635, + "grad_norm": 2.6042494773864746, + "learning_rate": 1.8641980672346284e-05, + "loss": 0.4479, + "step": 715 + }, + { + "epoch": 0.5795810988566225, + "grad_norm": 8.531290054321289, + "learning_rate": 1.8637577987611603e-05, + "loss": 0.4859, + "step": 716 + }, + { + "epoch": 0.5803905696650814, + "grad_norm": 2.884265899658203, + "learning_rate": 1.8633168699330048e-05, + "loss": 0.4896, + "step": 717 + }, + { + "epoch": 0.5812000404735405, + "grad_norm": 2.772880792617798, + "learning_rate": 1.8628752810872578e-05, + "loss": 0.485, + "step": 718 + }, + { + "epoch": 0.5820095112819994, + "grad_norm": 2.873414993286133, + "learning_rate": 1.8624330325615205e-05, + "loss": 0.4901, + "step": 719 + }, + { + "epoch": 0.5828189820904583, + "grad_norm": 2.93843412399292, + "learning_rate": 1.861990124693898e-05, + "loss": 0.5405, + "step": 720 + }, + { + "epoch": 0.5836284528989173, + "grad_norm": 2.535895824432373, + "learning_rate": 1.861546557822999e-05, + "loss": 0.4497, + "step": 721 + }, + { + "epoch": 0.5844379237073764, + "grad_norm": 2.721081495285034, + "learning_rate": 1.8611023322879366e-05, + "loss": 0.4844, + "step": 722 + }, + { + "epoch": 0.5852473945158353, + "grad_norm": 2.6494431495666504, + "learning_rate": 1.8606574484283276e-05, + "loss": 0.5349, + "step": 723 + }, + { + "epoch": 0.5860568653242942, + "grad_norm": 2.9973833560943604, + "learning_rate": 1.8602119065842918e-05, + "loss": 0.5382, + "step": 724 + }, + { + "epoch": 0.5868663361327532, + "grad_norm": 2.715162754058838, + "learning_rate": 1.8597657070964517e-05, + "loss": 0.4714, + "step": 725 + }, + { + "epoch": 0.5876758069412121, + "grad_norm": 2.5887551307678223, + "learning_rate": 1.8593188503059333e-05, + "loss": 0.4764, + "step": 726 + }, + { + "epoch": 0.5884852777496712, + "grad_norm": 2.535773515701294, + "learning_rate": 1.8588713365543644e-05, + "loss": 0.4683, + "step": 727 + }, + { + "epoch": 0.5892947485581301, + "grad_norm": 2.7653520107269287, + "learning_rate": 1.8584231661838753e-05, + "loss": 0.5267, + "step": 728 + }, + { + "epoch": 0.5901042193665891, + "grad_norm": 2.5949947834014893, + "learning_rate": 1.857974339537099e-05, + "loss": 0.4949, + "step": 729 + }, + { + "epoch": 0.590913690175048, + "grad_norm": 2.7669570446014404, + "learning_rate": 1.8575248569571692e-05, + "loss": 0.4741, + "step": 730 + }, + { + "epoch": 0.5917231609835071, + "grad_norm": 2.921645164489746, + "learning_rate": 1.8570747187877214e-05, + "loss": 0.4934, + "step": 731 + }, + { + "epoch": 0.592532631791966, + "grad_norm": 2.5315074920654297, + "learning_rate": 1.8566239253728928e-05, + "loss": 0.4997, + "step": 732 + }, + { + "epoch": 0.593342102600425, + "grad_norm": 2.5222811698913574, + "learning_rate": 1.856172477057321e-05, + "loss": 0.4872, + "step": 733 + }, + { + "epoch": 0.5941515734088839, + "grad_norm": 2.5237374305725098, + "learning_rate": 1.8557203741861444e-05, + "loss": 0.4602, + "step": 734 + }, + { + "epoch": 0.5949610442173429, + "grad_norm": 2.860875368118286, + "learning_rate": 1.8552676171050013e-05, + "loss": 0.506, + "step": 735 + }, + { + "epoch": 0.5957705150258019, + "grad_norm": 3.0834386348724365, + "learning_rate": 1.854814206160032e-05, + "loss": 0.5122, + "step": 736 + }, + { + "epoch": 0.5965799858342609, + "grad_norm": 2.918351650238037, + "learning_rate": 1.854360141697874e-05, + "loss": 0.5262, + "step": 737 + }, + { + "epoch": 0.5973894566427198, + "grad_norm": 2.5834996700286865, + "learning_rate": 1.8539054240656667e-05, + "loss": 0.4759, + "step": 738 + }, + { + "epoch": 0.5981989274511788, + "grad_norm": 2.790189027786255, + "learning_rate": 1.8534500536110476e-05, + "loss": 0.4824, + "step": 739 + }, + { + "epoch": 0.5990083982596378, + "grad_norm": 2.823833703994751, + "learning_rate": 1.8529940306821537e-05, + "loss": 0.4754, + "step": 740 + }, + { + "epoch": 0.5998178690680968, + "grad_norm": 2.6931285858154297, + "learning_rate": 1.852537355627621e-05, + "loss": 0.4937, + "step": 741 + }, + { + "epoch": 0.6006273398765557, + "grad_norm": 2.6135287284851074, + "learning_rate": 1.8520800287965836e-05, + "loss": 0.4614, + "step": 742 + }, + { + "epoch": 0.6014368106850146, + "grad_norm": 2.5539612770080566, + "learning_rate": 1.851622050538674e-05, + "loss": 0.4758, + "step": 743 + }, + { + "epoch": 0.6022462814934736, + "grad_norm": 2.52500581741333, + "learning_rate": 1.8511634212040235e-05, + "loss": 0.4899, + "step": 744 + }, + { + "epoch": 0.6030557523019326, + "grad_norm": 2.547438144683838, + "learning_rate": 1.85070414114326e-05, + "loss": 0.4458, + "step": 745 + }, + { + "epoch": 0.6038652231103916, + "grad_norm": 2.816446542739868, + "learning_rate": 1.850244210707509e-05, + "loss": 0.4986, + "step": 746 + }, + { + "epoch": 0.6046746939188505, + "grad_norm": 2.7901217937469482, + "learning_rate": 1.8497836302483945e-05, + "loss": 0.5049, + "step": 747 + }, + { + "epoch": 0.6054841647273095, + "grad_norm": 2.8099441528320312, + "learning_rate": 1.849322400118036e-05, + "loss": 0.5015, + "step": 748 + }, + { + "epoch": 0.6062936355357685, + "grad_norm": 2.7697811126708984, + "learning_rate": 1.8488605206690508e-05, + "loss": 0.502, + "step": 749 + }, + { + "epoch": 0.6071031063442275, + "grad_norm": 2.7742323875427246, + "learning_rate": 1.8483979922545515e-05, + "loss": 0.5132, + "step": 750 + }, + { + "epoch": 0.6079125771526864, + "grad_norm": 2.5849907398223877, + "learning_rate": 1.847934815228148e-05, + "loss": 0.4317, + "step": 751 + }, + { + "epoch": 0.6087220479611454, + "grad_norm": 2.8223841190338135, + "learning_rate": 1.847470989943945e-05, + "loss": 0.4805, + "step": 752 + }, + { + "epoch": 0.6095315187696043, + "grad_norm": 2.6553561687469482, + "learning_rate": 1.8470065167565433e-05, + "loss": 0.4865, + "step": 753 + }, + { + "epoch": 0.6103409895780634, + "grad_norm": 2.582911252975464, + "learning_rate": 1.8465413960210393e-05, + "loss": 0.457, + "step": 754 + }, + { + "epoch": 0.6111504603865223, + "grad_norm": 2.826479196548462, + "learning_rate": 1.8460756280930238e-05, + "loss": 0.4931, + "step": 755 + }, + { + "epoch": 0.6119599311949813, + "grad_norm": 3.2288568019866943, + "learning_rate": 1.8456092133285834e-05, + "loss": 0.4582, + "step": 756 + }, + { + "epoch": 0.6127694020034402, + "grad_norm": 2.6174867153167725, + "learning_rate": 1.845142152084298e-05, + "loss": 0.4757, + "step": 757 + }, + { + "epoch": 0.6135788728118993, + "grad_norm": 2.913802146911621, + "learning_rate": 1.8446744447172428e-05, + "loss": 0.4967, + "step": 758 + }, + { + "epoch": 0.6143883436203582, + "grad_norm": 2.6669833660125732, + "learning_rate": 1.8442060915849856e-05, + "loss": 0.5117, + "step": 759 + }, + { + "epoch": 0.6151978144288172, + "grad_norm": 2.7747111320495605, + "learning_rate": 1.84373709304559e-05, + "loss": 0.5008, + "step": 760 + }, + { + "epoch": 0.6160072852372761, + "grad_norm": 3.265331506729126, + "learning_rate": 1.8432674494576113e-05, + "loss": 0.52, + "step": 761 + }, + { + "epoch": 0.616816756045735, + "grad_norm": 2.597304344177246, + "learning_rate": 1.8427971611800982e-05, + "loss": 0.4798, + "step": 762 + }, + { + "epoch": 0.6176262268541941, + "grad_norm": 2.5773816108703613, + "learning_rate": 1.842326228572593e-05, + "loss": 0.4235, + "step": 763 + }, + { + "epoch": 0.618435697662653, + "grad_norm": 2.735212802886963, + "learning_rate": 1.8418546519951296e-05, + "loss": 0.5354, + "step": 764 + }, + { + "epoch": 0.619245168471112, + "grad_norm": 2.6665799617767334, + "learning_rate": 1.841382431808235e-05, + "loss": 0.5326, + "step": 765 + }, + { + "epoch": 0.6200546392795709, + "grad_norm": 2.664706230163574, + "learning_rate": 1.8409095683729284e-05, + "loss": 0.4867, + "step": 766 + }, + { + "epoch": 0.62086411008803, + "grad_norm": 2.5538930892944336, + "learning_rate": 1.84043606205072e-05, + "loss": 0.4972, + "step": 767 + }, + { + "epoch": 0.6216735808964889, + "grad_norm": 2.680497884750366, + "learning_rate": 1.8399619132036117e-05, + "loss": 0.4513, + "step": 768 + }, + { + "epoch": 0.6224830517049479, + "grad_norm": 2.504300355911255, + "learning_rate": 1.8394871221940972e-05, + "loss": 0.4804, + "step": 769 + }, + { + "epoch": 0.6232925225134068, + "grad_norm": 2.7403011322021484, + "learning_rate": 1.8390116893851605e-05, + "loss": 0.5391, + "step": 770 + }, + { + "epoch": 0.6241019933218658, + "grad_norm": 2.578341484069824, + "learning_rate": 1.838535615140277e-05, + "loss": 0.4899, + "step": 771 + }, + { + "epoch": 0.6249114641303248, + "grad_norm": 2.7620601654052734, + "learning_rate": 1.8380588998234107e-05, + "loss": 0.4923, + "step": 772 + }, + { + "epoch": 0.6257209349387838, + "grad_norm": 2.7333803176879883, + "learning_rate": 1.8375815437990185e-05, + "loss": 0.5142, + "step": 773 + }, + { + "epoch": 0.6265304057472427, + "grad_norm": 3.3537027835845947, + "learning_rate": 1.8371035474320453e-05, + "loss": 0.5088, + "step": 774 + }, + { + "epoch": 0.6273398765557017, + "grad_norm": 2.6102631092071533, + "learning_rate": 1.836624911087925e-05, + "loss": 0.4532, + "step": 775 + }, + { + "epoch": 0.6281493473641607, + "grad_norm": 2.8307254314422607, + "learning_rate": 1.8361456351325824e-05, + "loss": 0.5363, + "step": 776 + }, + { + "epoch": 0.6289588181726197, + "grad_norm": 2.806307554244995, + "learning_rate": 1.8356657199324304e-05, + "loss": 0.5039, + "step": 777 + }, + { + "epoch": 0.6297682889810786, + "grad_norm": 2.688058376312256, + "learning_rate": 1.8351851658543707e-05, + "loss": 0.5194, + "step": 778 + }, + { + "epoch": 0.6305777597895376, + "grad_norm": 2.8022544384002686, + "learning_rate": 1.8347039732657935e-05, + "loss": 0.5137, + "step": 779 + }, + { + "epoch": 0.6313872305979965, + "grad_norm": 2.3672170639038086, + "learning_rate": 1.8342221425345773e-05, + "loss": 0.4656, + "step": 780 + }, + { + "epoch": 0.6321967014064556, + "grad_norm": 2.6000335216522217, + "learning_rate": 1.833739674029088e-05, + "loss": 0.4964, + "step": 781 + }, + { + "epoch": 0.6330061722149145, + "grad_norm": 2.580603837966919, + "learning_rate": 1.8332565681181794e-05, + "loss": 0.5081, + "step": 782 + }, + { + "epoch": 0.6338156430233735, + "grad_norm": 2.692257881164551, + "learning_rate": 1.832772825171193e-05, + "loss": 0.5167, + "step": 783 + }, + { + "epoch": 0.6346251138318324, + "grad_norm": 2.742710590362549, + "learning_rate": 1.8322884455579567e-05, + "loss": 0.5062, + "step": 784 + }, + { + "epoch": 0.6354345846402915, + "grad_norm": 2.385173797607422, + "learning_rate": 1.8318034296487853e-05, + "loss": 0.4359, + "step": 785 + }, + { + "epoch": 0.6362440554487504, + "grad_norm": 2.5691487789154053, + "learning_rate": 1.83131777781448e-05, + "loss": 0.488, + "step": 786 + }, + { + "epoch": 0.6370535262572093, + "grad_norm": 2.5901083946228027, + "learning_rate": 1.8308314904263288e-05, + "loss": 0.4943, + "step": 787 + }, + { + "epoch": 0.6378629970656683, + "grad_norm": 2.5402183532714844, + "learning_rate": 1.8303445678561043e-05, + "loss": 0.4949, + "step": 788 + }, + { + "epoch": 0.6386724678741272, + "grad_norm": 2.5075602531433105, + "learning_rate": 1.829857010476066e-05, + "loss": 0.4468, + "step": 789 + }, + { + "epoch": 0.6394819386825863, + "grad_norm": 3.0251071453094482, + "learning_rate": 1.8293688186589582e-05, + "loss": 0.4695, + "step": 790 + }, + { + "epoch": 0.6402914094910452, + "grad_norm": 2.7452828884124756, + "learning_rate": 1.82887999277801e-05, + "loss": 0.4441, + "step": 791 + }, + { + "epoch": 0.6411008802995042, + "grad_norm": 2.7380993366241455, + "learning_rate": 1.8283905332069357e-05, + "loss": 0.4881, + "step": 792 + }, + { + "epoch": 0.6419103511079631, + "grad_norm": 2.5989420413970947, + "learning_rate": 1.827900440319934e-05, + "loss": 0.4616, + "step": 793 + }, + { + "epoch": 0.6427198219164222, + "grad_norm": 2.7274820804595947, + "learning_rate": 1.827409714491687e-05, + "loss": 0.4571, + "step": 794 + }, + { + "epoch": 0.6435292927248811, + "grad_norm": 3.005904197692871, + "learning_rate": 1.8269183560973622e-05, + "loss": 0.4825, + "step": 795 + }, + { + "epoch": 0.6443387635333401, + "grad_norm": 2.7077443599700928, + "learning_rate": 1.8264263655126092e-05, + "loss": 0.4999, + "step": 796 + }, + { + "epoch": 0.645148234341799, + "grad_norm": 2.608409881591797, + "learning_rate": 1.8259337431135617e-05, + "loss": 0.4212, + "step": 797 + }, + { + "epoch": 0.645957705150258, + "grad_norm": 2.7499301433563232, + "learning_rate": 1.8254404892768366e-05, + "loss": 0.4807, + "step": 798 + }, + { + "epoch": 0.646767175958717, + "grad_norm": 2.953740358352661, + "learning_rate": 1.8249466043795326e-05, + "loss": 0.4935, + "step": 799 + }, + { + "epoch": 0.647576646767176, + "grad_norm": 3.0464391708374023, + "learning_rate": 1.8244520887992322e-05, + "loss": 0.5478, + "step": 800 + }, + { + "epoch": 0.6483861175756349, + "grad_norm": 2.947700262069702, + "learning_rate": 1.8239569429139986e-05, + "loss": 0.5135, + "step": 801 + }, + { + "epoch": 0.6491955883840939, + "grad_norm": 2.7347915172576904, + "learning_rate": 1.823461167102378e-05, + "loss": 0.5181, + "step": 802 + }, + { + "epoch": 0.6500050591925529, + "grad_norm": 2.6112380027770996, + "learning_rate": 1.8229647617433982e-05, + "loss": 0.4729, + "step": 803 + }, + { + "epoch": 0.6508145300010119, + "grad_norm": 2.7688145637512207, + "learning_rate": 1.822467727216567e-05, + "loss": 0.5211, + "step": 804 + }, + { + "epoch": 0.6516240008094708, + "grad_norm": 2.541776418685913, + "learning_rate": 1.821970063901875e-05, + "loss": 0.4955, + "step": 805 + }, + { + "epoch": 0.6524334716179297, + "grad_norm": 2.6302928924560547, + "learning_rate": 1.821471772179792e-05, + "loss": 0.4573, + "step": 806 + }, + { + "epoch": 0.6532429424263887, + "grad_norm": 2.774416208267212, + "learning_rate": 1.8209728524312696e-05, + "loss": 0.4881, + "step": 807 + }, + { + "epoch": 0.6540524132348478, + "grad_norm": 2.7171905040740967, + "learning_rate": 1.820473305037738e-05, + "loss": 0.5226, + "step": 808 + }, + { + "epoch": 0.6548618840433067, + "grad_norm": 2.8347058296203613, + "learning_rate": 1.8199731303811083e-05, + "loss": 0.5266, + "step": 809 + }, + { + "epoch": 0.6556713548517656, + "grad_norm": 2.756420612335205, + "learning_rate": 1.819472328843771e-05, + "loss": 0.5367, + "step": 810 + }, + { + "epoch": 0.6564808256602246, + "grad_norm": 2.55179500579834, + "learning_rate": 1.8189709008085957e-05, + "loss": 0.447, + "step": 811 + }, + { + "epoch": 0.6572902964686836, + "grad_norm": 2.640458583831787, + "learning_rate": 1.8184688466589308e-05, + "loss": 0.4778, + "step": 812 + }, + { + "epoch": 0.6580997672771426, + "grad_norm": 2.487119197845459, + "learning_rate": 1.8179661667786042e-05, + "loss": 0.4927, + "step": 813 + }, + { + "epoch": 0.6589092380856015, + "grad_norm": 2.7604458332061768, + "learning_rate": 1.817462861551921e-05, + "loss": 0.4943, + "step": 814 + }, + { + "epoch": 0.6597187088940605, + "grad_norm": 2.514354705810547, + "learning_rate": 1.816958931363665e-05, + "loss": 0.4623, + "step": 815 + }, + { + "epoch": 0.6605281797025194, + "grad_norm": 2.4265823364257812, + "learning_rate": 1.8164543765990976e-05, + "loss": 0.4539, + "step": 816 + }, + { + "epoch": 0.6613376505109785, + "grad_norm": 2.4754762649536133, + "learning_rate": 1.8159491976439585e-05, + "loss": 0.4733, + "step": 817 + }, + { + "epoch": 0.6621471213194374, + "grad_norm": 2.669853687286377, + "learning_rate": 1.815443394884463e-05, + "loss": 0.4952, + "step": 818 + }, + { + "epoch": 0.6629565921278964, + "grad_norm": 2.4806864261627197, + "learning_rate": 1.8149369687073053e-05, + "loss": 0.4677, + "step": 819 + }, + { + "epoch": 0.6637660629363553, + "grad_norm": 2.684678316116333, + "learning_rate": 1.8144299194996545e-05, + "loss": 0.4979, + "step": 820 + }, + { + "epoch": 0.6645755337448144, + "grad_norm": 2.601315975189209, + "learning_rate": 1.813922247649157e-05, + "loss": 0.4856, + "step": 821 + }, + { + "epoch": 0.6653850045532733, + "grad_norm": 2.712395429611206, + "learning_rate": 1.813413953543934e-05, + "loss": 0.5221, + "step": 822 + }, + { + "epoch": 0.6661944753617323, + "grad_norm": 2.569307565689087, + "learning_rate": 1.8129050375725844e-05, + "loss": 0.4694, + "step": 823 + }, + { + "epoch": 0.6670039461701912, + "grad_norm": 2.686992645263672, + "learning_rate": 1.8123955001241813e-05, + "loss": 0.5117, + "step": 824 + }, + { + "epoch": 0.6678134169786502, + "grad_norm": 2.740213632583618, + "learning_rate": 1.8118853415882726e-05, + "loss": 0.4945, + "step": 825 + }, + { + "epoch": 0.6686228877871092, + "grad_norm": 2.9983811378479004, + "learning_rate": 1.811374562354882e-05, + "loss": 0.4789, + "step": 826 + }, + { + "epoch": 0.6694323585955682, + "grad_norm": 2.332531452178955, + "learning_rate": 1.8108631628145068e-05, + "loss": 0.4514, + "step": 827 + }, + { + "epoch": 0.6702418294040271, + "grad_norm": 2.678171396255493, + "learning_rate": 1.810351143358119e-05, + "loss": 0.5176, + "step": 828 + }, + { + "epoch": 0.671051300212486, + "grad_norm": 2.6020078659057617, + "learning_rate": 1.8098385043771646e-05, + "loss": 0.493, + "step": 829 + }, + { + "epoch": 0.6718607710209451, + "grad_norm": 2.7989866733551025, + "learning_rate": 1.8093252462635634e-05, + "loss": 0.4615, + "step": 830 + }, + { + "epoch": 0.672670241829404, + "grad_norm": 2.5469205379486084, + "learning_rate": 1.808811369409708e-05, + "loss": 0.4727, + "step": 831 + }, + { + "epoch": 0.673479712637863, + "grad_norm": 2.71799635887146, + "learning_rate": 1.808296874208464e-05, + "loss": 0.496, + "step": 832 + }, + { + "epoch": 0.6742891834463219, + "grad_norm": 2.6747519969940186, + "learning_rate": 1.8077817610531704e-05, + "loss": 0.5412, + "step": 833 + }, + { + "epoch": 0.6750986542547809, + "grad_norm": 2.481365442276001, + "learning_rate": 1.8072660303376384e-05, + "loss": 0.4972, + "step": 834 + }, + { + "epoch": 0.6759081250632399, + "grad_norm": 2.4019603729248047, + "learning_rate": 1.8067496824561514e-05, + "loss": 0.5012, + "step": 835 + }, + { + "epoch": 0.6767175958716989, + "grad_norm": 2.5155980587005615, + "learning_rate": 1.8062327178034635e-05, + "loss": 0.4581, + "step": 836 + }, + { + "epoch": 0.6775270666801578, + "grad_norm": 2.473384141921997, + "learning_rate": 1.8057151367748024e-05, + "loss": 0.4618, + "step": 837 + }, + { + "epoch": 0.6783365374886168, + "grad_norm": 2.502716302871704, + "learning_rate": 1.805196939765865e-05, + "loss": 0.4715, + "step": 838 + }, + { + "epoch": 0.6791460082970758, + "grad_norm": 2.631765365600586, + "learning_rate": 1.8046781271728203e-05, + "loss": 0.4887, + "step": 839 + }, + { + "epoch": 0.6799554791055348, + "grad_norm": 2.832674264907837, + "learning_rate": 1.804158699392308e-05, + "loss": 0.4991, + "step": 840 + }, + { + "epoch": 0.6807649499139937, + "grad_norm": 2.4482016563415527, + "learning_rate": 1.8036386568214374e-05, + "loss": 0.4644, + "step": 841 + }, + { + "epoch": 0.6815744207224527, + "grad_norm": 2.5509285926818848, + "learning_rate": 1.8031179998577882e-05, + "loss": 0.475, + "step": 842 + }, + { + "epoch": 0.6823838915309116, + "grad_norm": 2.447124481201172, + "learning_rate": 1.80259672889941e-05, + "loss": 0.4497, + "step": 843 + }, + { + "epoch": 0.6831933623393707, + "grad_norm": 2.6824610233306885, + "learning_rate": 1.8020748443448216e-05, + "loss": 0.537, + "step": 844 + }, + { + "epoch": 0.6840028331478296, + "grad_norm": 2.573214530944824, + "learning_rate": 1.8015523465930104e-05, + "loss": 0.5196, + "step": 845 + }, + { + "epoch": 0.6848123039562886, + "grad_norm": 2.6556625366210938, + "learning_rate": 1.8010292360434336e-05, + "loss": 0.5168, + "step": 846 + }, + { + "epoch": 0.6856217747647475, + "grad_norm": 2.4084174633026123, + "learning_rate": 1.8005055130960162e-05, + "loss": 0.4507, + "step": 847 + }, + { + "epoch": 0.6864312455732066, + "grad_norm": 2.484358310699463, + "learning_rate": 1.7999811781511517e-05, + "loss": 0.4701, + "step": 848 + }, + { + "epoch": 0.6872407163816655, + "grad_norm": 2.684661865234375, + "learning_rate": 1.7994562316097014e-05, + "loss": 0.5471, + "step": 849 + }, + { + "epoch": 0.6880501871901245, + "grad_norm": 4.29740571975708, + "learning_rate": 1.798930673872994e-05, + "loss": 0.5454, + "step": 850 + }, + { + "epoch": 0.6888596579985834, + "grad_norm": 2.6054797172546387, + "learning_rate": 1.7984045053428256e-05, + "loss": 0.4793, + "step": 851 + }, + { + "epoch": 0.6896691288070423, + "grad_norm": 2.6472408771514893, + "learning_rate": 1.797877726421459e-05, + "loss": 0.4905, + "step": 852 + }, + { + "epoch": 0.6904785996155014, + "grad_norm": 2.427823066711426, + "learning_rate": 1.7973503375116243e-05, + "loss": 0.4545, + "step": 853 + }, + { + "epoch": 0.6912880704239603, + "grad_norm": 2.723539352416992, + "learning_rate": 1.796822339016518e-05, + "loss": 0.4999, + "step": 854 + }, + { + "epoch": 0.6920975412324193, + "grad_norm": 3.703444242477417, + "learning_rate": 1.796293731339801e-05, + "loss": 0.5097, + "step": 855 + }, + { + "epoch": 0.6929070120408782, + "grad_norm": 2.4358720779418945, + "learning_rate": 1.795764514885602e-05, + "loss": 0.4586, + "step": 856 + }, + { + "epoch": 0.6937164828493373, + "grad_norm": 2.4158782958984375, + "learning_rate": 1.7952346900585142e-05, + "loss": 0.4662, + "step": 857 + }, + { + "epoch": 0.6945259536577962, + "grad_norm": 2.424586772918701, + "learning_rate": 1.7947042572635954e-05, + "loss": 0.4795, + "step": 858 + }, + { + "epoch": 0.6953354244662552, + "grad_norm": 2.5223560333251953, + "learning_rate": 1.7941732169063692e-05, + "loss": 0.5042, + "step": 859 + }, + { + "epoch": 0.6961448952747141, + "grad_norm": 2.359363555908203, + "learning_rate": 1.7936415693928233e-05, + "loss": 0.444, + "step": 860 + }, + { + "epoch": 0.6969543660831731, + "grad_norm": 2.4504969120025635, + "learning_rate": 1.7931093151294094e-05, + "loss": 0.4677, + "step": 861 + }, + { + "epoch": 0.6977638368916321, + "grad_norm": 2.4936771392822266, + "learning_rate": 1.7925764545230428e-05, + "loss": 0.4363, + "step": 862 + }, + { + "epoch": 0.6985733077000911, + "grad_norm": 2.515625476837158, + "learning_rate": 1.7920429879811037e-05, + "loss": 0.469, + "step": 863 + }, + { + "epoch": 0.69938277850855, + "grad_norm": 2.6376399993896484, + "learning_rate": 1.7915089159114333e-05, + "loss": 0.5012, + "step": 864 + }, + { + "epoch": 0.700192249317009, + "grad_norm": 2.5924456119537354, + "learning_rate": 1.7909742387223382e-05, + "loss": 0.4664, + "step": 865 + }, + { + "epoch": 0.701001720125468, + "grad_norm": 2.6718382835388184, + "learning_rate": 1.7904389568225857e-05, + "loss": 0.5227, + "step": 866 + }, + { + "epoch": 0.701811190933927, + "grad_norm": 2.608126401901245, + "learning_rate": 1.7899030706214067e-05, + "loss": 0.5101, + "step": 867 + }, + { + "epoch": 0.7026206617423859, + "grad_norm": 2.4242279529571533, + "learning_rate": 1.789366580528493e-05, + "loss": 0.4521, + "step": 868 + }, + { + "epoch": 0.7034301325508449, + "grad_norm": 2.6490864753723145, + "learning_rate": 1.788829486953999e-05, + "loss": 0.5026, + "step": 869 + }, + { + "epoch": 0.7042396033593038, + "grad_norm": 2.5904784202575684, + "learning_rate": 1.78829179030854e-05, + "loss": 0.5028, + "step": 870 + }, + { + "epoch": 0.7050490741677629, + "grad_norm": 2.573307752609253, + "learning_rate": 1.7877534910031923e-05, + "loss": 0.4891, + "step": 871 + }, + { + "epoch": 0.7058585449762218, + "grad_norm": 2.40058970451355, + "learning_rate": 1.7872145894494935e-05, + "loss": 0.4577, + "step": 872 + }, + { + "epoch": 0.7066680157846807, + "grad_norm": 2.6595945358276367, + "learning_rate": 1.7866750860594412e-05, + "loss": 0.4598, + "step": 873 + }, + { + "epoch": 0.7074774865931397, + "grad_norm": 2.4909627437591553, + "learning_rate": 1.7861349812454926e-05, + "loss": 0.4715, + "step": 874 + }, + { + "epoch": 0.7082869574015987, + "grad_norm": 2.9337961673736572, + "learning_rate": 1.785594275420566e-05, + "loss": 0.4227, + "step": 875 + }, + { + "epoch": 0.7090964282100577, + "grad_norm": 2.5457868576049805, + "learning_rate": 1.7850529689980375e-05, + "loss": 0.5166, + "step": 876 + }, + { + "epoch": 0.7099058990185166, + "grad_norm": 2.3930580615997314, + "learning_rate": 1.7845110623917436e-05, + "loss": 0.4652, + "step": 877 + }, + { + "epoch": 0.7107153698269756, + "grad_norm": 2.743706226348877, + "learning_rate": 1.7839685560159802e-05, + "loss": 0.4509, + "step": 878 + }, + { + "epoch": 0.7115248406354345, + "grad_norm": 2.5328824520111084, + "learning_rate": 1.7834254502855e-05, + "loss": 0.5212, + "step": 879 + }, + { + "epoch": 0.7123343114438936, + "grad_norm": 2.8872995376586914, + "learning_rate": 1.7828817456155144e-05, + "loss": 0.4783, + "step": 880 + }, + { + "epoch": 0.7131437822523525, + "grad_norm": 2.614563465118408, + "learning_rate": 1.7823374424216944e-05, + "loss": 0.4818, + "step": 881 + }, + { + "epoch": 0.7139532530608115, + "grad_norm": 2.5072665214538574, + "learning_rate": 1.7817925411201664e-05, + "loss": 0.4967, + "step": 882 + }, + { + "epoch": 0.7147627238692704, + "grad_norm": 2.5186476707458496, + "learning_rate": 1.7812470421275147e-05, + "loss": 0.4589, + "step": 883 + }, + { + "epoch": 0.7155721946777295, + "grad_norm": 2.500885486602783, + "learning_rate": 1.7807009458607814e-05, + "loss": 0.483, + "step": 884 + }, + { + "epoch": 0.7163816654861884, + "grad_norm": 2.3186795711517334, + "learning_rate": 1.7801542527374643e-05, + "loss": 0.4403, + "step": 885 + }, + { + "epoch": 0.7171911362946474, + "grad_norm": 2.7155909538269043, + "learning_rate": 1.7796069631755176e-05, + "loss": 0.4948, + "step": 886 + }, + { + "epoch": 0.7180006071031063, + "grad_norm": 2.5053658485412598, + "learning_rate": 1.779059077593352e-05, + "loss": 0.4722, + "step": 887 + }, + { + "epoch": 0.7188100779115653, + "grad_norm": 2.584824800491333, + "learning_rate": 1.7785105964098336e-05, + "loss": 0.4866, + "step": 888 + }, + { + "epoch": 0.7196195487200243, + "grad_norm": 2.6166129112243652, + "learning_rate": 1.777961520044283e-05, + "loss": 0.5367, + "step": 889 + }, + { + "epoch": 0.7204290195284833, + "grad_norm": 2.4438273906707764, + "learning_rate": 1.7774118489164775e-05, + "loss": 0.4484, + "step": 890 + }, + { + "epoch": 0.7212384903369422, + "grad_norm": 2.605755567550659, + "learning_rate": 1.7768615834466484e-05, + "loss": 0.5272, + "step": 891 + }, + { + "epoch": 0.7220479611454012, + "grad_norm": 2.622849941253662, + "learning_rate": 1.77631072405548e-05, + "loss": 0.4795, + "step": 892 + }, + { + "epoch": 0.7228574319538602, + "grad_norm": 2.590240955352783, + "learning_rate": 1.775759271164113e-05, + "loss": 0.5322, + "step": 893 + }, + { + "epoch": 0.7236669027623192, + "grad_norm": 2.5132153034210205, + "learning_rate": 1.7752072251941404e-05, + "loss": 0.4679, + "step": 894 + }, + { + "epoch": 0.7244763735707781, + "grad_norm": 2.7585039138793945, + "learning_rate": 1.7746545865676085e-05, + "loss": 0.5454, + "step": 895 + }, + { + "epoch": 0.725285844379237, + "grad_norm": 2.55354905128479, + "learning_rate": 1.7741013557070175e-05, + "loss": 0.5317, + "step": 896 + }, + { + "epoch": 0.726095315187696, + "grad_norm": 3.4242117404937744, + "learning_rate": 1.7735475330353198e-05, + "loss": 0.436, + "step": 897 + }, + { + "epoch": 0.726904785996155, + "grad_norm": 2.5605053901672363, + "learning_rate": 1.7729931189759208e-05, + "loss": 0.4944, + "step": 898 + }, + { + "epoch": 0.727714256804614, + "grad_norm": 2.9254157543182373, + "learning_rate": 1.7724381139526774e-05, + "loss": 0.5017, + "step": 899 + }, + { + "epoch": 0.7285237276130729, + "grad_norm": 2.9272122383117676, + "learning_rate": 1.7718825183898983e-05, + "loss": 0.5155, + "step": 900 + }, + { + "epoch": 0.7293331984215319, + "grad_norm": 2.607419967651367, + "learning_rate": 1.771326332712344e-05, + "loss": 0.5232, + "step": 901 + }, + { + "epoch": 0.7301426692299909, + "grad_norm": 2.6773085594177246, + "learning_rate": 1.770769557345227e-05, + "loss": 0.5113, + "step": 902 + }, + { + "epoch": 0.7309521400384499, + "grad_norm": 2.4693233966827393, + "learning_rate": 1.7702121927142083e-05, + "loss": 0.4833, + "step": 903 + }, + { + "epoch": 0.7317616108469088, + "grad_norm": 2.5243523120880127, + "learning_rate": 1.7696542392454018e-05, + "loss": 0.489, + "step": 904 + }, + { + "epoch": 0.7325710816553678, + "grad_norm": 2.6101861000061035, + "learning_rate": 1.76909569736537e-05, + "loss": 0.514, + "step": 905 + }, + { + "epoch": 0.7333805524638267, + "grad_norm": 2.507066011428833, + "learning_rate": 1.7685365675011265e-05, + "loss": 0.479, + "step": 906 + }, + { + "epoch": 0.7341900232722858, + "grad_norm": 2.5869712829589844, + "learning_rate": 1.767976850080133e-05, + "loss": 0.472, + "step": 907 + }, + { + "epoch": 0.7349994940807447, + "grad_norm": 2.6980485916137695, + "learning_rate": 1.7674165455303013e-05, + "loss": 0.4685, + "step": 908 + }, + { + "epoch": 0.7358089648892037, + "grad_norm": 2.562955141067505, + "learning_rate": 1.7668556542799928e-05, + "loss": 0.5035, + "step": 909 + }, + { + "epoch": 0.7366184356976626, + "grad_norm": 2.504657030105591, + "learning_rate": 1.766294176758015e-05, + "loss": 0.4925, + "step": 910 + }, + { + "epoch": 0.7374279065061217, + "grad_norm": 2.5493977069854736, + "learning_rate": 1.7657321133936263e-05, + "loss": 0.4523, + "step": 911 + }, + { + "epoch": 0.7382373773145806, + "grad_norm": 2.6682395935058594, + "learning_rate": 1.7651694646165317e-05, + "loss": 0.4983, + "step": 912 + }, + { + "epoch": 0.7390468481230396, + "grad_norm": 2.5802791118621826, + "learning_rate": 1.764606230856884e-05, + "loss": 0.4919, + "step": 913 + }, + { + "epoch": 0.7398563189314985, + "grad_norm": 2.644697427749634, + "learning_rate": 1.7640424125452827e-05, + "loss": 0.4833, + "step": 914 + }, + { + "epoch": 0.7406657897399574, + "grad_norm": 2.717024803161621, + "learning_rate": 1.763478010112775e-05, + "loss": 0.5056, + "step": 915 + }, + { + "epoch": 0.7414752605484165, + "grad_norm": 2.619049072265625, + "learning_rate": 1.762913023990854e-05, + "loss": 0.4956, + "step": 916 + }, + { + "epoch": 0.7422847313568754, + "grad_norm": 2.4587295055389404, + "learning_rate": 1.7623474546114593e-05, + "loss": 0.4599, + "step": 917 + }, + { + "epoch": 0.7430942021653344, + "grad_norm": 4.636078834533691, + "learning_rate": 1.761781302406977e-05, + "loss": 0.5046, + "step": 918 + }, + { + "epoch": 0.7439036729737933, + "grad_norm": 2.77103853225708, + "learning_rate": 1.7612145678102377e-05, + "loss": 0.5006, + "step": 919 + }, + { + "epoch": 0.7447131437822524, + "grad_norm": 2.5714871883392334, + "learning_rate": 1.7606472512545178e-05, + "loss": 0.4839, + "step": 920 + }, + { + "epoch": 0.7455226145907113, + "grad_norm": 2.419672966003418, + "learning_rate": 1.7600793531735392e-05, + "loss": 0.436, + "step": 921 + }, + { + "epoch": 0.7463320853991703, + "grad_norm": 2.5261123180389404, + "learning_rate": 1.759510874001467e-05, + "loss": 0.4959, + "step": 922 + }, + { + "epoch": 0.7471415562076292, + "grad_norm": 2.628504753112793, + "learning_rate": 1.7589418141729114e-05, + "loss": 0.4819, + "step": 923 + }, + { + "epoch": 0.7479510270160882, + "grad_norm": 2.4579715728759766, + "learning_rate": 1.7583721741229272e-05, + "loss": 0.4762, + "step": 924 + }, + { + "epoch": 0.7487604978245472, + "grad_norm": 2.4755959510803223, + "learning_rate": 1.7578019542870113e-05, + "loss": 0.4937, + "step": 925 + }, + { + "epoch": 0.7495699686330062, + "grad_norm": 2.5187032222747803, + "learning_rate": 1.7572311551011047e-05, + "loss": 0.4784, + "step": 926 + }, + { + "epoch": 0.7503794394414651, + "grad_norm": 2.599585771560669, + "learning_rate": 1.7566597770015915e-05, + "loss": 0.5189, + "step": 927 + }, + { + "epoch": 0.7511889102499241, + "grad_norm": 2.531540632247925, + "learning_rate": 1.756087820425298e-05, + "loss": 0.4886, + "step": 928 + }, + { + "epoch": 0.7519983810583831, + "grad_norm": 2.785987377166748, + "learning_rate": 1.7555152858094923e-05, + "loss": 0.4986, + "step": 929 + }, + { + "epoch": 0.7528078518668421, + "grad_norm": 2.6085269451141357, + "learning_rate": 1.7549421735918856e-05, + "loss": 0.4592, + "step": 930 + }, + { + "epoch": 0.753617322675301, + "grad_norm": 3.138561964035034, + "learning_rate": 1.7543684842106304e-05, + "loss": 0.4996, + "step": 931 + }, + { + "epoch": 0.75442679348376, + "grad_norm": 2.53794002532959, + "learning_rate": 1.753794218104319e-05, + "loss": 0.4711, + "step": 932 + }, + { + "epoch": 0.755236264292219, + "grad_norm": 2.3540737628936768, + "learning_rate": 1.7532193757119867e-05, + "loss": 0.4415, + "step": 933 + }, + { + "epoch": 0.756045735100678, + "grad_norm": 2.5483851432800293, + "learning_rate": 1.7526439574731074e-05, + "loss": 0.4974, + "step": 934 + }, + { + "epoch": 0.7568552059091369, + "grad_norm": 2.629504919052124, + "learning_rate": 1.7520679638275974e-05, + "loss": 0.4852, + "step": 935 + }, + { + "epoch": 0.7576646767175959, + "grad_norm": 2.547236919403076, + "learning_rate": 1.751491395215811e-05, + "loss": 0.4743, + "step": 936 + }, + { + "epoch": 0.7584741475260548, + "grad_norm": 2.4229674339294434, + "learning_rate": 1.750914252078543e-05, + "loss": 0.4377, + "step": 937 + }, + { + "epoch": 0.7592836183345139, + "grad_norm": 2.737112045288086, + "learning_rate": 1.7503365348570273e-05, + "loss": 0.4662, + "step": 938 + }, + { + "epoch": 0.7600930891429728, + "grad_norm": 2.973440170288086, + "learning_rate": 1.7497582439929367e-05, + "loss": 0.4756, + "step": 939 + }, + { + "epoch": 0.7609025599514317, + "grad_norm": 2.6873090267181396, + "learning_rate": 1.749179379928382e-05, + "loss": 0.4712, + "step": 940 + }, + { + "epoch": 0.7617120307598907, + "grad_norm": 2.757565498352051, + "learning_rate": 1.7485999431059135e-05, + "loss": 0.5018, + "step": 941 + }, + { + "epoch": 0.7625215015683497, + "grad_norm": 2.621570587158203, + "learning_rate": 1.748019933968518e-05, + "loss": 0.4482, + "step": 942 + }, + { + "epoch": 0.7633309723768087, + "grad_norm": 2.6197359561920166, + "learning_rate": 1.7474393529596208e-05, + "loss": 0.4398, + "step": 943 + }, + { + "epoch": 0.7641404431852676, + "grad_norm": 2.456888437271118, + "learning_rate": 1.7468582005230837e-05, + "loss": 0.4616, + "step": 944 + }, + { + "epoch": 0.7649499139937266, + "grad_norm": 2.628403663635254, + "learning_rate": 1.746276477103206e-05, + "loss": 0.4894, + "step": 945 + }, + { + "epoch": 0.7657593848021855, + "grad_norm": 2.7005257606506348, + "learning_rate": 1.745694183144723e-05, + "loss": 0.5153, + "step": 946 + }, + { + "epoch": 0.7665688556106446, + "grad_norm": 2.5277342796325684, + "learning_rate": 1.7451113190928065e-05, + "loss": 0.4987, + "step": 947 + }, + { + "epoch": 0.7673783264191035, + "grad_norm": 2.272611141204834, + "learning_rate": 1.7445278853930645e-05, + "loss": 0.4604, + "step": 948 + }, + { + "epoch": 0.7681877972275625, + "grad_norm": 2.444906234741211, + "learning_rate": 1.7439438824915394e-05, + "loss": 0.509, + "step": 949 + }, + { + "epoch": 0.7689972680360214, + "grad_norm": 2.554558515548706, + "learning_rate": 1.7433593108347097e-05, + "loss": 0.4965, + "step": 950 + }, + { + "epoch": 0.7698067388444805, + "grad_norm": 2.3335328102111816, + "learning_rate": 1.7427741708694887e-05, + "loss": 0.4489, + "step": 951 + }, + { + "epoch": 0.7706162096529394, + "grad_norm": 2.3146164417266846, + "learning_rate": 1.7421884630432236e-05, + "loss": 0.4795, + "step": 952 + }, + { + "epoch": 0.7714256804613984, + "grad_norm": 2.6994271278381348, + "learning_rate": 1.741602187803696e-05, + "loss": 0.508, + "step": 953 + }, + { + "epoch": 0.7722351512698573, + "grad_norm": 3.058566093444824, + "learning_rate": 1.7410153455991218e-05, + "loss": 0.4571, + "step": 954 + }, + { + "epoch": 0.7730446220783163, + "grad_norm": 2.7912354469299316, + "learning_rate": 1.7404279368781494e-05, + "loss": 0.496, + "step": 955 + }, + { + "epoch": 0.7738540928867753, + "grad_norm": 2.5112459659576416, + "learning_rate": 1.7398399620898613e-05, + "loss": 0.5066, + "step": 956 + }, + { + "epoch": 0.7746635636952343, + "grad_norm": 2.389423370361328, + "learning_rate": 1.739251421683772e-05, + "loss": 0.4508, + "step": 957 + }, + { + "epoch": 0.7754730345036932, + "grad_norm": 2.4844539165496826, + "learning_rate": 1.7386623161098287e-05, + "loss": 0.4707, + "step": 958 + }, + { + "epoch": 0.7762825053121521, + "grad_norm": 2.694202423095703, + "learning_rate": 1.7380726458184108e-05, + "loss": 0.5195, + "step": 959 + }, + { + "epoch": 0.7770919761206112, + "grad_norm": 2.5364737510681152, + "learning_rate": 1.7374824112603294e-05, + "loss": 0.4432, + "step": 960 + }, + { + "epoch": 0.7779014469290702, + "grad_norm": 2.779428720474243, + "learning_rate": 1.7368916128868266e-05, + "loss": 0.4772, + "step": 961 + }, + { + "epoch": 0.7787109177375291, + "grad_norm": 2.7698585987091064, + "learning_rate": 1.7363002511495765e-05, + "loss": 0.4752, + "step": 962 + }, + { + "epoch": 0.779520388545988, + "grad_norm": 2.7890233993530273, + "learning_rate": 1.7357083265006825e-05, + "loss": 0.5134, + "step": 963 + }, + { + "epoch": 0.780329859354447, + "grad_norm": 2.647568702697754, + "learning_rate": 1.7351158393926793e-05, + "loss": 0.4708, + "step": 964 + }, + { + "epoch": 0.781139330162906, + "grad_norm": 3.2525548934936523, + "learning_rate": 1.7345227902785318e-05, + "loss": 0.4524, + "step": 965 + }, + { + "epoch": 0.781948800971365, + "grad_norm": 2.3511910438537598, + "learning_rate": 1.733929179611634e-05, + "loss": 0.4481, + "step": 966 + }, + { + "epoch": 0.7827582717798239, + "grad_norm": 2.4978339672088623, + "learning_rate": 1.7333350078458096e-05, + "loss": 0.4805, + "step": 967 + }, + { + "epoch": 0.7835677425882829, + "grad_norm": 2.3907647132873535, + "learning_rate": 1.7327402754353108e-05, + "loss": 0.4689, + "step": 968 + }, + { + "epoch": 0.7843772133967419, + "grad_norm": 2.478454113006592, + "learning_rate": 1.7321449828348184e-05, + "loss": 0.4486, + "step": 969 + }, + { + "epoch": 0.7851866842052009, + "grad_norm": 7.012077331542969, + "learning_rate": 1.7315491304994422e-05, + "loss": 0.4613, + "step": 970 + }, + { + "epoch": 0.7859961550136598, + "grad_norm": 2.615460157394409, + "learning_rate": 1.7309527188847193e-05, + "loss": 0.5036, + "step": 971 + }, + { + "epoch": 0.7868056258221188, + "grad_norm": 2.6299996376037598, + "learning_rate": 1.7303557484466146e-05, + "loss": 0.4927, + "step": 972 + }, + { + "epoch": 0.7876150966305777, + "grad_norm": 2.7502124309539795, + "learning_rate": 1.7297582196415203e-05, + "loss": 0.451, + "step": 973 + }, + { + "epoch": 0.7884245674390368, + "grad_norm": 2.6889915466308594, + "learning_rate": 1.7291601329262552e-05, + "loss": 0.5058, + "step": 974 + }, + { + "epoch": 0.7892340382474957, + "grad_norm": 2.5078210830688477, + "learning_rate": 1.7285614887580653e-05, + "loss": 0.5082, + "step": 975 + }, + { + "epoch": 0.7900435090559547, + "grad_norm": 2.359553337097168, + "learning_rate": 1.7279622875946214e-05, + "loss": 0.4622, + "step": 976 + }, + { + "epoch": 0.7908529798644136, + "grad_norm": 2.610729217529297, + "learning_rate": 1.7273625298940216e-05, + "loss": 0.4975, + "step": 977 + }, + { + "epoch": 0.7916624506728727, + "grad_norm": 3.2304341793060303, + "learning_rate": 1.7267622161147887e-05, + "loss": 0.5193, + "step": 978 + }, + { + "epoch": 0.7924719214813316, + "grad_norm": 2.3112738132476807, + "learning_rate": 1.726161346715871e-05, + "loss": 0.469, + "step": 979 + }, + { + "epoch": 0.7932813922897906, + "grad_norm": 2.304487943649292, + "learning_rate": 1.7255599221566416e-05, + "loss": 0.4735, + "step": 980 + }, + { + "epoch": 0.7940908630982495, + "grad_norm": 2.5640969276428223, + "learning_rate": 1.724957942896897e-05, + "loss": 0.5289, + "step": 981 + }, + { + "epoch": 0.7949003339067084, + "grad_norm": 2.6010193824768066, + "learning_rate": 1.7243554093968595e-05, + "loss": 0.5051, + "step": 982 + }, + { + "epoch": 0.7957098047151675, + "grad_norm": 2.974827527999878, + "learning_rate": 1.7237523221171733e-05, + "loss": 0.5286, + "step": 983 + }, + { + "epoch": 0.7965192755236264, + "grad_norm": 2.4988255500793457, + "learning_rate": 1.7231486815189077e-05, + "loss": 0.4741, + "step": 984 + }, + { + "epoch": 0.7973287463320854, + "grad_norm": 2.688586711883545, + "learning_rate": 1.7225444880635544e-05, + "loss": 0.4942, + "step": 985 + }, + { + "epoch": 0.7981382171405443, + "grad_norm": 3.000936269760132, + "learning_rate": 1.7219397422130265e-05, + "loss": 0.4825, + "step": 986 + }, + { + "epoch": 0.7989476879490034, + "grad_norm": 2.332054376602173, + "learning_rate": 1.7213344444296615e-05, + "loss": 0.4423, + "step": 987 + }, + { + "epoch": 0.7997571587574623, + "grad_norm": 2.4883697032928467, + "learning_rate": 1.7207285951762177e-05, + "loss": 0.502, + "step": 988 + }, + { + "epoch": 0.8005666295659213, + "grad_norm": 2.464693546295166, + "learning_rate": 1.7201221949158745e-05, + "loss": 0.4939, + "step": 989 + }, + { + "epoch": 0.8013761003743802, + "grad_norm": 2.4625542163848877, + "learning_rate": 1.7195152441122338e-05, + "loss": 0.4608, + "step": 990 + }, + { + "epoch": 0.8021855711828392, + "grad_norm": 2.4094271659851074, + "learning_rate": 1.7189077432293176e-05, + "loss": 0.4505, + "step": 991 + }, + { + "epoch": 0.8029950419912982, + "grad_norm": 2.582942008972168, + "learning_rate": 1.7182996927315693e-05, + "loss": 0.4645, + "step": 992 + }, + { + "epoch": 0.8038045127997572, + "grad_norm": 2.5438919067382812, + "learning_rate": 1.7176910930838508e-05, + "loss": 0.5093, + "step": 993 + }, + { + "epoch": 0.8046139836082161, + "grad_norm": 2.642756700515747, + "learning_rate": 1.7170819447514456e-05, + "loss": 0.4825, + "step": 994 + }, + { + "epoch": 0.8054234544166751, + "grad_norm": 2.453993320465088, + "learning_rate": 1.7164722482000557e-05, + "loss": 0.4634, + "step": 995 + }, + { + "epoch": 0.8062329252251341, + "grad_norm": 2.4292731285095215, + "learning_rate": 1.7158620038958023e-05, + "loss": 0.455, + "step": 996 + }, + { + "epoch": 0.8070423960335931, + "grad_norm": 2.55206298828125, + "learning_rate": 1.7152512123052262e-05, + "loss": 0.4526, + "step": 997 + }, + { + "epoch": 0.807851866842052, + "grad_norm": 2.441127061843872, + "learning_rate": 1.714639873895285e-05, + "loss": 0.4366, + "step": 998 + }, + { + "epoch": 0.808661337650511, + "grad_norm": 2.432224988937378, + "learning_rate": 1.7140279891333565e-05, + "loss": 0.492, + "step": 999 + }, + { + "epoch": 0.8094708084589699, + "grad_norm": 2.471465826034546, + "learning_rate": 1.713415558487234e-05, + "loss": 0.4936, + "step": 1000 + }, + { + "epoch": 0.810280279267429, + "grad_norm": 2.581582546234131, + "learning_rate": 1.7128025824251296e-05, + "loss": 0.493, + "step": 1001 + }, + { + "epoch": 0.8110897500758879, + "grad_norm": 2.3795430660247803, + "learning_rate": 1.7121890614156714e-05, + "loss": 0.4818, + "step": 1002 + }, + { + "epoch": 0.8118992208843469, + "grad_norm": 2.6020090579986572, + "learning_rate": 1.711574995927905e-05, + "loss": 0.4675, + "step": 1003 + }, + { + "epoch": 0.8127086916928058, + "grad_norm": 2.58920955657959, + "learning_rate": 1.710960386431292e-05, + "loss": 0.4684, + "step": 1004 + }, + { + "epoch": 0.8135181625012649, + "grad_norm": 2.45979905128479, + "learning_rate": 1.7103452333957097e-05, + "loss": 0.4487, + "step": 1005 + }, + { + "epoch": 0.8143276333097238, + "grad_norm": 2.623964786529541, + "learning_rate": 1.7097295372914515e-05, + "loss": 0.5056, + "step": 1006 + }, + { + "epoch": 0.8151371041181827, + "grad_norm": 2.4810256958007812, + "learning_rate": 1.709113298589224e-05, + "loss": 0.4697, + "step": 1007 + }, + { + "epoch": 0.8159465749266417, + "grad_norm": 2.358898162841797, + "learning_rate": 1.708496517760152e-05, + "loss": 0.4573, + "step": 1008 + }, + { + "epoch": 0.8167560457351006, + "grad_norm": 2.430300235748291, + "learning_rate": 1.7078791952757718e-05, + "loss": 0.4569, + "step": 1009 + }, + { + "epoch": 0.8175655165435597, + "grad_norm": 3.092907428741455, + "learning_rate": 1.707261331608035e-05, + "loss": 0.4655, + "step": 1010 + }, + { + "epoch": 0.8183749873520186, + "grad_norm": 2.5205323696136475, + "learning_rate": 1.7066429272293072e-05, + "loss": 0.4518, + "step": 1011 + }, + { + "epoch": 0.8191844581604776, + "grad_norm": 2.5230295658111572, + "learning_rate": 1.7060239826123668e-05, + "loss": 0.4657, + "step": 1012 + }, + { + "epoch": 0.8199939289689365, + "grad_norm": 2.4902048110961914, + "learning_rate": 1.7054044982304058e-05, + "loss": 0.49, + "step": 1013 + }, + { + "epoch": 0.8208033997773956, + "grad_norm": 2.6711537837982178, + "learning_rate": 1.7047844745570276e-05, + "loss": 0.4954, + "step": 1014 + }, + { + "epoch": 0.8216128705858545, + "grad_norm": 2.483548879623413, + "learning_rate": 1.70416391206625e-05, + "loss": 0.4741, + "step": 1015 + }, + { + "epoch": 0.8224223413943135, + "grad_norm": 2.5759265422821045, + "learning_rate": 1.7035428112325008e-05, + "loss": 0.4907, + "step": 1016 + }, + { + "epoch": 0.8232318122027724, + "grad_norm": 2.545593738555908, + "learning_rate": 1.7029211725306205e-05, + "loss": 0.4781, + "step": 1017 + }, + { + "epoch": 0.8240412830112314, + "grad_norm": 2.641831874847412, + "learning_rate": 1.7022989964358597e-05, + "loss": 0.4614, + "step": 1018 + }, + { + "epoch": 0.8248507538196904, + "grad_norm": 2.385359525680542, + "learning_rate": 1.7016762834238815e-05, + "loss": 0.5, + "step": 1019 + }, + { + "epoch": 0.8256602246281494, + "grad_norm": 2.481903314590454, + "learning_rate": 1.701053033970758e-05, + "loss": 0.4246, + "step": 1020 + }, + { + "epoch": 0.8264696954366083, + "grad_norm": 2.4436490535736084, + "learning_rate": 1.7004292485529718e-05, + "loss": 0.5029, + "step": 1021 + }, + { + "epoch": 0.8272791662450673, + "grad_norm": 2.5008347034454346, + "learning_rate": 1.699804927647416e-05, + "loss": 0.4617, + "step": 1022 + }, + { + "epoch": 0.8280886370535263, + "grad_norm": 2.608060836791992, + "learning_rate": 1.6991800717313913e-05, + "loss": 0.4851, + "step": 1023 + }, + { + "epoch": 0.8288981078619853, + "grad_norm": 2.632477045059204, + "learning_rate": 1.6985546812826102e-05, + "loss": 0.4793, + "step": 1024 + }, + { + "epoch": 0.8297075786704442, + "grad_norm": 2.7196977138519287, + "learning_rate": 1.6979287567791917e-05, + "loss": 0.453, + "step": 1025 + }, + { + "epoch": 0.8305170494789031, + "grad_norm": 2.5669376850128174, + "learning_rate": 1.6973022986996633e-05, + "loss": 0.5181, + "step": 1026 + }, + { + "epoch": 0.8313265202873621, + "grad_norm": 2.5459508895874023, + "learning_rate": 1.696675307522961e-05, + "loss": 0.4961, + "step": 1027 + }, + { + "epoch": 0.8321359910958211, + "grad_norm": 2.3130640983581543, + "learning_rate": 1.696047783728428e-05, + "loss": 0.4659, + "step": 1028 + }, + { + "epoch": 0.8329454619042801, + "grad_norm": 3.1561124324798584, + "learning_rate": 1.6954197277958154e-05, + "loss": 0.4801, + "step": 1029 + }, + { + "epoch": 0.833754932712739, + "grad_norm": 2.231194496154785, + "learning_rate": 1.69479114020528e-05, + "loss": 0.4373, + "step": 1030 + }, + { + "epoch": 0.834564403521198, + "grad_norm": 2.497173309326172, + "learning_rate": 1.6941620214373856e-05, + "loss": 0.533, + "step": 1031 + }, + { + "epoch": 0.835373874329657, + "grad_norm": 2.559159994125366, + "learning_rate": 1.693532371973103e-05, + "loss": 0.4315, + "step": 1032 + }, + { + "epoch": 0.836183345138116, + "grad_norm": 2.5158908367156982, + "learning_rate": 1.6929021922938064e-05, + "loss": 0.4702, + "step": 1033 + }, + { + "epoch": 0.8369928159465749, + "grad_norm": 2.2467987537384033, + "learning_rate": 1.6922714828812785e-05, + "loss": 0.4365, + "step": 1034 + }, + { + "epoch": 0.8378022867550339, + "grad_norm": 4.310861587524414, + "learning_rate": 1.691640244217704e-05, + "loss": 0.4772, + "step": 1035 + }, + { + "epoch": 0.8386117575634928, + "grad_norm": 2.10097336769104, + "learning_rate": 1.6910084767856738e-05, + "loss": 0.4389, + "step": 1036 + }, + { + "epoch": 0.8394212283719519, + "grad_norm": 2.3843936920166016, + "learning_rate": 1.6903761810681833e-05, + "loss": 0.4984, + "step": 1037 + }, + { + "epoch": 0.8402306991804108, + "grad_norm": 2.374861240386963, + "learning_rate": 1.6897433575486302e-05, + "loss": 0.4344, + "step": 1038 + }, + { + "epoch": 0.8410401699888698, + "grad_norm": 2.501530885696411, + "learning_rate": 1.6891100067108182e-05, + "loss": 0.4297, + "step": 1039 + }, + { + "epoch": 0.8418496407973287, + "grad_norm": 2.6958343982696533, + "learning_rate": 1.6884761290389516e-05, + "loss": 0.473, + "step": 1040 + }, + { + "epoch": 0.8426591116057878, + "grad_norm": 2.320976495742798, + "learning_rate": 1.6878417250176393e-05, + "loss": 0.4744, + "step": 1041 + }, + { + "epoch": 0.8434685824142467, + "grad_norm": 2.6825976371765137, + "learning_rate": 1.687206795131891e-05, + "loss": 0.4792, + "step": 1042 + }, + { + "epoch": 0.8442780532227057, + "grad_norm": 2.869171619415283, + "learning_rate": 1.6865713398671197e-05, + "loss": 0.5007, + "step": 1043 + }, + { + "epoch": 0.8450875240311646, + "grad_norm": 2.424920082092285, + "learning_rate": 1.68593535970914e-05, + "loss": 0.4503, + "step": 1044 + }, + { + "epoch": 0.8458969948396236, + "grad_norm": 2.540224552154541, + "learning_rate": 1.6852988551441674e-05, + "loss": 0.4592, + "step": 1045 + }, + { + "epoch": 0.8467064656480826, + "grad_norm": 2.5557565689086914, + "learning_rate": 1.6846618266588182e-05, + "loss": 0.4335, + "step": 1046 + }, + { + "epoch": 0.8475159364565416, + "grad_norm": 2.5456111431121826, + "learning_rate": 1.684024274740109e-05, + "loss": 0.4466, + "step": 1047 + }, + { + "epoch": 0.8483254072650005, + "grad_norm": 2.557997941970825, + "learning_rate": 1.6833861998754578e-05, + "loss": 0.4874, + "step": 1048 + }, + { + "epoch": 0.8491348780734594, + "grad_norm": 2.3289544582366943, + "learning_rate": 1.6827476025526807e-05, + "loss": 0.4428, + "step": 1049 + }, + { + "epoch": 0.8499443488819185, + "grad_norm": 2.4546475410461426, + "learning_rate": 1.6821084832599953e-05, + "loss": 0.4917, + "step": 1050 + }, + { + "epoch": 0.8507538196903774, + "grad_norm": 2.339095115661621, + "learning_rate": 1.681468842486016e-05, + "loss": 0.4663, + "step": 1051 + }, + { + "epoch": 0.8515632904988364, + "grad_norm": 2.4625139236450195, + "learning_rate": 1.6808286807197572e-05, + "loss": 0.4854, + "step": 1052 + }, + { + "epoch": 0.8523727613072953, + "grad_norm": 2.41728138923645, + "learning_rate": 1.680187998450632e-05, + "loss": 0.4507, + "step": 1053 + }, + { + "epoch": 0.8531822321157543, + "grad_norm": 2.5671286582946777, + "learning_rate": 1.6795467961684496e-05, + "loss": 0.519, + "step": 1054 + }, + { + "epoch": 0.8539917029242133, + "grad_norm": 2.491149663925171, + "learning_rate": 1.6789050743634187e-05, + "loss": 0.4777, + "step": 1055 + }, + { + "epoch": 0.8548011737326723, + "grad_norm": 2.3900294303894043, + "learning_rate": 1.6782628335261445e-05, + "loss": 0.4882, + "step": 1056 + }, + { + "epoch": 0.8556106445411312, + "grad_norm": 2.4009006023406982, + "learning_rate": 1.6776200741476288e-05, + "loss": 0.4478, + "step": 1057 + }, + { + "epoch": 0.8564201153495902, + "grad_norm": 2.542780876159668, + "learning_rate": 1.6769767967192702e-05, + "loss": 0.4741, + "step": 1058 + }, + { + "epoch": 0.8572295861580492, + "grad_norm": 2.5278592109680176, + "learning_rate": 1.6763330017328622e-05, + "loss": 0.4974, + "step": 1059 + }, + { + "epoch": 0.8580390569665082, + "grad_norm": 2.5964314937591553, + "learning_rate": 1.675688689680596e-05, + "loss": 0.4809, + "step": 1060 + }, + { + "epoch": 0.8588485277749671, + "grad_norm": 2.541757583618164, + "learning_rate": 1.6750438610550563e-05, + "loss": 0.4655, + "step": 1061 + }, + { + "epoch": 0.8596579985834261, + "grad_norm": 2.3481216430664062, + "learning_rate": 1.6743985163492236e-05, + "loss": 0.496, + "step": 1062 + }, + { + "epoch": 0.860467469391885, + "grad_norm": 2.606701135635376, + "learning_rate": 1.6737526560564732e-05, + "loss": 0.4736, + "step": 1063 + }, + { + "epoch": 0.8612769402003441, + "grad_norm": 2.2494258880615234, + "learning_rate": 1.6731062806705735e-05, + "loss": 0.4503, + "step": 1064 + }, + { + "epoch": 0.862086411008803, + "grad_norm": 2.7492141723632812, + "learning_rate": 1.672459390685688e-05, + "loss": 0.4898, + "step": 1065 + }, + { + "epoch": 0.862895881817262, + "grad_norm": 2.2967073917388916, + "learning_rate": 1.671811986596372e-05, + "loss": 0.4687, + "step": 1066 + }, + { + "epoch": 0.8637053526257209, + "grad_norm": 2.5024197101593018, + "learning_rate": 1.671164068897576e-05, + "loss": 0.4907, + "step": 1067 + }, + { + "epoch": 0.86451482343418, + "grad_norm": 2.4369518756866455, + "learning_rate": 1.6705156380846413e-05, + "loss": 0.4207, + "step": 1068 + }, + { + "epoch": 0.8653242942426389, + "grad_norm": 2.598294258117676, + "learning_rate": 1.669866694653302e-05, + "loss": 0.487, + "step": 1069 + }, + { + "epoch": 0.8661337650510978, + "grad_norm": 2.452103853225708, + "learning_rate": 1.669217239099685e-05, + "loss": 0.4891, + "step": 1070 + }, + { + "epoch": 0.8669432358595568, + "grad_norm": 2.51023268699646, + "learning_rate": 1.6685672719203067e-05, + "loss": 0.4748, + "step": 1071 + }, + { + "epoch": 0.8677527066680157, + "grad_norm": 2.5311238765716553, + "learning_rate": 1.6679167936120773e-05, + "loss": 0.4653, + "step": 1072 + }, + { + "epoch": 0.8685621774764748, + "grad_norm": 2.2431399822235107, + "learning_rate": 1.667265804672296e-05, + "loss": 0.4307, + "step": 1073 + }, + { + "epoch": 0.8693716482849337, + "grad_norm": 2.590728282928467, + "learning_rate": 1.666614305598652e-05, + "loss": 0.4516, + "step": 1074 + }, + { + "epoch": 0.8701811190933927, + "grad_norm": 2.811022996902466, + "learning_rate": 1.6659622968892263e-05, + "loss": 0.4619, + "step": 1075 + }, + { + "epoch": 0.8709905899018516, + "grad_norm": 2.646606683731079, + "learning_rate": 1.6653097790424878e-05, + "loss": 0.4876, + "step": 1076 + }, + { + "epoch": 0.8718000607103107, + "grad_norm": 2.3428947925567627, + "learning_rate": 1.664656752557296e-05, + "loss": 0.4817, + "step": 1077 + }, + { + "epoch": 0.8726095315187696, + "grad_norm": 2.5434818267822266, + "learning_rate": 1.6640032179328978e-05, + "loss": 0.4976, + "step": 1078 + }, + { + "epoch": 0.8734190023272286, + "grad_norm": 2.3956143856048584, + "learning_rate": 1.66334917566893e-05, + "loss": 0.4803, + "step": 1079 + }, + { + "epoch": 0.8742284731356875, + "grad_norm": 2.469965696334839, + "learning_rate": 1.6626946262654167e-05, + "loss": 0.4645, + "step": 1080 + }, + { + "epoch": 0.8750379439441465, + "grad_norm": 2.619140863418579, + "learning_rate": 1.6620395702227695e-05, + "loss": 0.4539, + "step": 1081 + }, + { + "epoch": 0.8758474147526055, + "grad_norm": 2.6556339263916016, + "learning_rate": 1.6613840080417887e-05, + "loss": 0.495, + "step": 1082 + }, + { + "epoch": 0.8766568855610645, + "grad_norm": 2.348163366317749, + "learning_rate": 1.66072794022366e-05, + "loss": 0.4529, + "step": 1083 + }, + { + "epoch": 0.8774663563695234, + "grad_norm": 2.182230234146118, + "learning_rate": 1.6600713672699558e-05, + "loss": 0.4708, + "step": 1084 + }, + { + "epoch": 0.8782758271779824, + "grad_norm": 2.59633731842041, + "learning_rate": 1.6594142896826358e-05, + "loss": 0.502, + "step": 1085 + }, + { + "epoch": 0.8790852979864414, + "grad_norm": 2.687954902648926, + "learning_rate": 1.6587567079640447e-05, + "loss": 0.5406, + "step": 1086 + }, + { + "epoch": 0.8798947687949004, + "grad_norm": 2.4950804710388184, + "learning_rate": 1.6580986226169128e-05, + "loss": 0.4978, + "step": 1087 + }, + { + "epoch": 0.8807042396033593, + "grad_norm": 2.349104404449463, + "learning_rate": 1.657440034144355e-05, + "loss": 0.457, + "step": 1088 + }, + { + "epoch": 0.8815137104118183, + "grad_norm": 2.41629695892334, + "learning_rate": 1.6567809430498714e-05, + "loss": 0.4726, + "step": 1089 + }, + { + "epoch": 0.8823231812202772, + "grad_norm": 2.557931423187256, + "learning_rate": 1.656121349837346e-05, + "loss": 0.4523, + "step": 1090 + }, + { + "epoch": 0.8831326520287363, + "grad_norm": 2.5173518657684326, + "learning_rate": 1.6554612550110476e-05, + "loss": 0.518, + "step": 1091 + }, + { + "epoch": 0.8839421228371952, + "grad_norm": 2.375762462615967, + "learning_rate": 1.654800659075627e-05, + "loss": 0.4637, + "step": 1092 + }, + { + "epoch": 0.8847515936456541, + "grad_norm": 2.414553642272949, + "learning_rate": 1.6541395625361193e-05, + "loss": 0.4394, + "step": 1093 + }, + { + "epoch": 0.8855610644541131, + "grad_norm": 2.312469720840454, + "learning_rate": 1.6534779658979414e-05, + "loss": 0.4511, + "step": 1094 + }, + { + "epoch": 0.8863705352625721, + "grad_norm": 2.4642269611358643, + "learning_rate": 1.6528158696668936e-05, + "loss": 0.4912, + "step": 1095 + }, + { + "epoch": 0.8871800060710311, + "grad_norm": 2.3398821353912354, + "learning_rate": 1.652153274349157e-05, + "loss": 0.4615, + "step": 1096 + }, + { + "epoch": 0.88798947687949, + "grad_norm": 2.5376386642456055, + "learning_rate": 1.6514901804512953e-05, + "loss": 0.5186, + "step": 1097 + }, + { + "epoch": 0.888798947687949, + "grad_norm": 2.2921273708343506, + "learning_rate": 1.6508265884802523e-05, + "loss": 0.4536, + "step": 1098 + }, + { + "epoch": 0.8896084184964079, + "grad_norm": 2.3653886318206787, + "learning_rate": 1.6501624989433538e-05, + "loss": 0.4807, + "step": 1099 + }, + { + "epoch": 0.890417889304867, + "grad_norm": 2.475956678390503, + "learning_rate": 1.6494979123483048e-05, + "loss": 0.4438, + "step": 1100 + }, + { + "epoch": 0.8912273601133259, + "grad_norm": 2.3942854404449463, + "learning_rate": 1.648832829203191e-05, + "loss": 0.4505, + "step": 1101 + }, + { + "epoch": 0.8920368309217849, + "grad_norm": 2.332777261734009, + "learning_rate": 1.648167250016478e-05, + "loss": 0.4709, + "step": 1102 + }, + { + "epoch": 0.8928463017302438, + "grad_norm": 2.3613274097442627, + "learning_rate": 1.6475011752970102e-05, + "loss": 0.4272, + "step": 1103 + }, + { + "epoch": 0.8936557725387029, + "grad_norm": 2.752133846282959, + "learning_rate": 1.64683460555401e-05, + "loss": 0.4666, + "step": 1104 + }, + { + "epoch": 0.8944652433471618, + "grad_norm": 2.540388345718384, + "learning_rate": 1.6461675412970802e-05, + "loss": 0.4939, + "step": 1105 + }, + { + "epoch": 0.8952747141556208, + "grad_norm": 2.3702893257141113, + "learning_rate": 1.6454999830362e-05, + "loss": 0.4749, + "step": 1106 + }, + { + "epoch": 0.8960841849640797, + "grad_norm": 2.4737656116485596, + "learning_rate": 1.644831931281727e-05, + "loss": 0.4677, + "step": 1107 + }, + { + "epoch": 0.8968936557725387, + "grad_norm": 2.712149143218994, + "learning_rate": 1.6441633865443956e-05, + "loss": 0.4886, + "step": 1108 + }, + { + "epoch": 0.8977031265809977, + "grad_norm": 2.509884834289551, + "learning_rate": 1.6434943493353176e-05, + "loss": 0.4608, + "step": 1109 + }, + { + "epoch": 0.8985125973894567, + "grad_norm": 2.475060224533081, + "learning_rate": 1.6428248201659816e-05, + "loss": 0.4508, + "step": 1110 + }, + { + "epoch": 0.8993220681979156, + "grad_norm": 2.3813722133636475, + "learning_rate": 1.6421547995482516e-05, + "loss": 0.4503, + "step": 1111 + }, + { + "epoch": 0.9001315390063745, + "grad_norm": 2.1945178508758545, + "learning_rate": 1.641484287994367e-05, + "loss": 0.4522, + "step": 1112 + }, + { + "epoch": 0.9009410098148336, + "grad_norm": 2.4047982692718506, + "learning_rate": 1.6408132860169433e-05, + "loss": 0.4772, + "step": 1113 + }, + { + "epoch": 0.9017504806232925, + "grad_norm": 2.2190158367156982, + "learning_rate": 1.640141794128971e-05, + "loss": 0.4841, + "step": 1114 + }, + { + "epoch": 0.9025599514317515, + "grad_norm": 2.2752647399902344, + "learning_rate": 1.6394698128438152e-05, + "loss": 0.4265, + "step": 1115 + }, + { + "epoch": 0.9033694222402104, + "grad_norm": 2.3116421699523926, + "learning_rate": 1.638797342675214e-05, + "loss": 0.43, + "step": 1116 + }, + { + "epoch": 0.9041788930486694, + "grad_norm": 2.4428725242614746, + "learning_rate": 1.6381243841372808e-05, + "loss": 0.4601, + "step": 1117 + }, + { + "epoch": 0.9049883638571284, + "grad_norm": 2.420809030532837, + "learning_rate": 1.637450937744501e-05, + "loss": 0.4168, + "step": 1118 + }, + { + "epoch": 0.9057978346655874, + "grad_norm": 2.3211169242858887, + "learning_rate": 1.636777004011734e-05, + "loss": 0.4502, + "step": 1119 + }, + { + "epoch": 0.9066073054740463, + "grad_norm": 2.323817014694214, + "learning_rate": 1.636102583454211e-05, + "loss": 0.5085, + "step": 1120 + }, + { + "epoch": 0.9074167762825053, + "grad_norm": 2.560737371444702, + "learning_rate": 1.6354276765875364e-05, + "loss": 0.4692, + "step": 1121 + }, + { + "epoch": 0.9082262470909643, + "grad_norm": 2.696718454360962, + "learning_rate": 1.6347522839276852e-05, + "loss": 0.5133, + "step": 1122 + }, + { + "epoch": 0.9090357178994233, + "grad_norm": 2.2378759384155273, + "learning_rate": 1.6340764059910047e-05, + "loss": 0.4387, + "step": 1123 + }, + { + "epoch": 0.9098451887078822, + "grad_norm": 2.225743532180786, + "learning_rate": 1.6334000432942127e-05, + "loss": 0.4537, + "step": 1124 + }, + { + "epoch": 0.9106546595163412, + "grad_norm": 2.4611008167266846, + "learning_rate": 1.6327231963543976e-05, + "loss": 0.4865, + "step": 1125 + }, + { + "epoch": 0.9114641303248001, + "grad_norm": 2.4171793460845947, + "learning_rate": 1.6320458656890187e-05, + "loss": 0.4639, + "step": 1126 + }, + { + "epoch": 0.9122736011332592, + "grad_norm": 2.221613883972168, + "learning_rate": 1.6313680518159037e-05, + "loss": 0.488, + "step": 1127 + }, + { + "epoch": 0.9130830719417181, + "grad_norm": 2.2258172035217285, + "learning_rate": 1.6306897552532517e-05, + "loss": 0.4611, + "step": 1128 + }, + { + "epoch": 0.9138925427501771, + "grad_norm": 2.4187533855438232, + "learning_rate": 1.6300109765196288e-05, + "loss": 0.4908, + "step": 1129 + }, + { + "epoch": 0.914702013558636, + "grad_norm": 2.3296492099761963, + "learning_rate": 1.6293317161339714e-05, + "loss": 0.459, + "step": 1130 + }, + { + "epoch": 0.9155114843670951, + "grad_norm": 2.547128200531006, + "learning_rate": 1.6286519746155825e-05, + "loss": 0.4904, + "step": 1131 + }, + { + "epoch": 0.916320955175554, + "grad_norm": 2.3638195991516113, + "learning_rate": 1.627971752484135e-05, + "loss": 0.5254, + "step": 1132 + }, + { + "epoch": 0.917130425984013, + "grad_norm": 2.4238646030426025, + "learning_rate": 1.6272910502596664e-05, + "loss": 0.4286, + "step": 1133 + }, + { + "epoch": 0.9179398967924719, + "grad_norm": 2.325035333633423, + "learning_rate": 1.6266098684625846e-05, + "loss": 0.4682, + "step": 1134 + }, + { + "epoch": 0.9187493676009308, + "grad_norm": 2.2326245307922363, + "learning_rate": 1.6259282076136616e-05, + "loss": 0.4377, + "step": 1135 + }, + { + "epoch": 0.9195588384093899, + "grad_norm": 2.3872227668762207, + "learning_rate": 1.625246068234036e-05, + "loss": 0.4475, + "step": 1136 + }, + { + "epoch": 0.9203683092178488, + "grad_norm": 2.277620792388916, + "learning_rate": 1.624563450845214e-05, + "loss": 0.4559, + "step": 1137 + }, + { + "epoch": 0.9211777800263078, + "grad_norm": 2.3244476318359375, + "learning_rate": 1.6238803559690646e-05, + "loss": 0.4714, + "step": 1138 + }, + { + "epoch": 0.9219872508347667, + "grad_norm": 2.2856132984161377, + "learning_rate": 1.6231967841278246e-05, + "loss": 0.4698, + "step": 1139 + }, + { + "epoch": 0.9227967216432258, + "grad_norm": 2.3583779335021973, + "learning_rate": 1.622512735844093e-05, + "loss": 0.4548, + "step": 1140 + }, + { + "epoch": 0.9236061924516847, + "grad_norm": 2.537989616394043, + "learning_rate": 1.6218282116408342e-05, + "loss": 0.4868, + "step": 1141 + }, + { + "epoch": 0.9244156632601437, + "grad_norm": 2.4814610481262207, + "learning_rate": 1.6211432120413766e-05, + "loss": 0.4804, + "step": 1142 + }, + { + "epoch": 0.9252251340686026, + "grad_norm": 2.9801652431488037, + "learning_rate": 1.6204577375694116e-05, + "loss": 0.5004, + "step": 1143 + }, + { + "epoch": 0.9260346048770616, + "grad_norm": 2.8574349880218506, + "learning_rate": 1.619771788748994e-05, + "loss": 0.5139, + "step": 1144 + }, + { + "epoch": 0.9268440756855206, + "grad_norm": 2.4583914279937744, + "learning_rate": 1.6190853661045413e-05, + "loss": 0.4842, + "step": 1145 + }, + { + "epoch": 0.9276535464939796, + "grad_norm": 2.6337690353393555, + "learning_rate": 1.6183984701608326e-05, + "loss": 0.5294, + "step": 1146 + }, + { + "epoch": 0.9284630173024385, + "grad_norm": 2.6366004943847656, + "learning_rate": 1.6177111014430092e-05, + "loss": 0.4983, + "step": 1147 + }, + { + "epoch": 0.9292724881108975, + "grad_norm": 2.3664913177490234, + "learning_rate": 1.6170232604765746e-05, + "loss": 0.4838, + "step": 1148 + }, + { + "epoch": 0.9300819589193565, + "grad_norm": 2.2880051136016846, + "learning_rate": 1.616334947787392e-05, + "loss": 0.4242, + "step": 1149 + }, + { + "epoch": 0.9308914297278155, + "grad_norm": 2.65151047706604, + "learning_rate": 1.615646163901686e-05, + "loss": 0.5074, + "step": 1150 + }, + { + "epoch": 0.9317009005362744, + "grad_norm": 2.4227969646453857, + "learning_rate": 1.6149569093460417e-05, + "loss": 0.4777, + "step": 1151 + }, + { + "epoch": 0.9325103713447334, + "grad_norm": 2.3962512016296387, + "learning_rate": 1.6142671846474032e-05, + "loss": 0.4689, + "step": 1152 + }, + { + "epoch": 0.9333198421531923, + "grad_norm": 2.4187710285186768, + "learning_rate": 1.6135769903330746e-05, + "loss": 0.4643, + "step": 1153 + }, + { + "epoch": 0.9341293129616514, + "grad_norm": 2.514302968978882, + "learning_rate": 1.6128863269307193e-05, + "loss": 0.4727, + "step": 1154 + }, + { + "epoch": 0.9349387837701103, + "grad_norm": 2.114122152328491, + "learning_rate": 1.6121951949683584e-05, + "loss": 0.4467, + "step": 1155 + }, + { + "epoch": 0.9357482545785692, + "grad_norm": 2.534583568572998, + "learning_rate": 1.6115035949743725e-05, + "loss": 0.4871, + "step": 1156 + }, + { + "epoch": 0.9365577253870282, + "grad_norm": 2.5609772205352783, + "learning_rate": 1.610811527477498e-05, + "loss": 0.4685, + "step": 1157 + }, + { + "epoch": 0.9373671961954873, + "grad_norm": 2.541923999786377, + "learning_rate": 1.6101189930068316e-05, + "loss": 0.4762, + "step": 1158 + }, + { + "epoch": 0.9381766670039462, + "grad_norm": 2.4497127532958984, + "learning_rate": 1.6094259920918246e-05, + "loss": 0.451, + "step": 1159 + }, + { + "epoch": 0.9389861378124051, + "grad_norm": 2.1860172748565674, + "learning_rate": 1.6087325252622853e-05, + "loss": 0.4377, + "step": 1160 + }, + { + "epoch": 0.9397956086208641, + "grad_norm": 2.2513699531555176, + "learning_rate": 1.608038593048379e-05, + "loss": 0.4601, + "step": 1161 + }, + { + "epoch": 0.940605079429323, + "grad_norm": 2.569270133972168, + "learning_rate": 1.607344195980627e-05, + "loss": 0.5112, + "step": 1162 + }, + { + "epoch": 0.9414145502377821, + "grad_norm": 2.416106700897217, + "learning_rate": 1.6066493345899043e-05, + "loss": 0.4672, + "step": 1163 + }, + { + "epoch": 0.942224021046241, + "grad_norm": 2.44903302192688, + "learning_rate": 1.6059540094074427e-05, + "loss": 0.507, + "step": 1164 + }, + { + "epoch": 0.9430334918547, + "grad_norm": 3.203815460205078, + "learning_rate": 1.6052582209648275e-05, + "loss": 0.4661, + "step": 1165 + }, + { + "epoch": 0.9438429626631589, + "grad_norm": 2.4812278747558594, + "learning_rate": 1.6045619697939986e-05, + "loss": 0.4907, + "step": 1166 + }, + { + "epoch": 0.944652433471618, + "grad_norm": 2.50327730178833, + "learning_rate": 1.6038652564272497e-05, + "loss": 0.4684, + "step": 1167 + }, + { + "epoch": 0.9454619042800769, + "grad_norm": 2.2611701488494873, + "learning_rate": 1.603168081397228e-05, + "loss": 0.4347, + "step": 1168 + }, + { + "epoch": 0.9462713750885359, + "grad_norm": 2.4070281982421875, + "learning_rate": 1.6024704452369323e-05, + "loss": 0.4884, + "step": 1169 + }, + { + "epoch": 0.9470808458969948, + "grad_norm": 2.5659537315368652, + "learning_rate": 1.6017723484797163e-05, + "loss": 0.4917, + "step": 1170 + }, + { + "epoch": 0.9478903167054538, + "grad_norm": 2.4707388877868652, + "learning_rate": 1.6010737916592835e-05, + "loss": 0.4827, + "step": 1171 + }, + { + "epoch": 0.9486997875139128, + "grad_norm": 2.3992793560028076, + "learning_rate": 1.6003747753096908e-05, + "loss": 0.4813, + "step": 1172 + }, + { + "epoch": 0.9495092583223718, + "grad_norm": 2.468963861465454, + "learning_rate": 1.599675299965346e-05, + "loss": 0.4879, + "step": 1173 + }, + { + "epoch": 0.9503187291308307, + "grad_norm": 2.2783689498901367, + "learning_rate": 1.5989753661610068e-05, + "loss": 0.3978, + "step": 1174 + }, + { + "epoch": 0.9511281999392897, + "grad_norm": 2.2595999240875244, + "learning_rate": 1.5982749744317825e-05, + "loss": 0.476, + "step": 1175 + }, + { + "epoch": 0.9519376707477487, + "grad_norm": 2.5539584159851074, + "learning_rate": 1.5975741253131326e-05, + "loss": 0.4815, + "step": 1176 + }, + { + "epoch": 0.9527471415562077, + "grad_norm": 2.4220964908599854, + "learning_rate": 1.596872819340866e-05, + "loss": 0.4677, + "step": 1177 + }, + { + "epoch": 0.9535566123646666, + "grad_norm": 2.3030552864074707, + "learning_rate": 1.5961710570511395e-05, + "loss": 0.4826, + "step": 1178 + }, + { + "epoch": 0.9543660831731255, + "grad_norm": 2.375501871109009, + "learning_rate": 1.5954688389804612e-05, + "loss": 0.4411, + "step": 1179 + }, + { + "epoch": 0.9551755539815845, + "grad_norm": 2.3813629150390625, + "learning_rate": 1.594766165665686e-05, + "loss": 0.4767, + "step": 1180 + }, + { + "epoch": 0.9559850247900435, + "grad_norm": 2.1959753036499023, + "learning_rate": 1.5940630376440175e-05, + "loss": 0.4383, + "step": 1181 + }, + { + "epoch": 0.9567944955985025, + "grad_norm": 2.0717129707336426, + "learning_rate": 1.5933594554530064e-05, + "loss": 0.4529, + "step": 1182 + }, + { + "epoch": 0.9576039664069614, + "grad_norm": 2.255031108856201, + "learning_rate": 1.592655419630551e-05, + "loss": 0.4388, + "step": 1183 + }, + { + "epoch": 0.9584134372154204, + "grad_norm": 2.390521764755249, + "learning_rate": 1.5919509307148967e-05, + "loss": 0.464, + "step": 1184 + }, + { + "epoch": 0.9592229080238794, + "grad_norm": 2.5235776901245117, + "learning_rate": 1.5912459892446346e-05, + "loss": 0.4539, + "step": 1185 + }, + { + "epoch": 0.9600323788323384, + "grad_norm": 2.33115553855896, + "learning_rate": 1.5905405957587022e-05, + "loss": 0.4637, + "step": 1186 + }, + { + "epoch": 0.9608418496407973, + "grad_norm": 2.5281829833984375, + "learning_rate": 1.5898347507963827e-05, + "loss": 0.4952, + "step": 1187 + }, + { + "epoch": 0.9616513204492563, + "grad_norm": 2.3663227558135986, + "learning_rate": 1.589128454897304e-05, + "loss": 0.48, + "step": 1188 + }, + { + "epoch": 0.9624607912577152, + "grad_norm": 2.538421154022217, + "learning_rate": 1.588421708601439e-05, + "loss": 0.5201, + "step": 1189 + }, + { + "epoch": 0.9632702620661743, + "grad_norm": 2.4426865577697754, + "learning_rate": 1.587714512449105e-05, + "loss": 0.4655, + "step": 1190 + }, + { + "epoch": 0.9640797328746332, + "grad_norm": 2.3822338581085205, + "learning_rate": 1.5870068669809634e-05, + "loss": 0.5155, + "step": 1191 + }, + { + "epoch": 0.9648892036830922, + "grad_norm": 2.4676313400268555, + "learning_rate": 1.586298772738018e-05, + "loss": 0.5058, + "step": 1192 + }, + { + "epoch": 0.9656986744915511, + "grad_norm": 2.137993574142456, + "learning_rate": 1.585590230261618e-05, + "loss": 0.4393, + "step": 1193 + }, + { + "epoch": 0.9665081453000102, + "grad_norm": 2.5882534980773926, + "learning_rate": 1.5848812400934527e-05, + "loss": 0.4799, + "step": 1194 + }, + { + "epoch": 0.9673176161084691, + "grad_norm": 2.3306941986083984, + "learning_rate": 1.584171802775555e-05, + "loss": 0.4357, + "step": 1195 + }, + { + "epoch": 0.9681270869169281, + "grad_norm": 2.4501194953918457, + "learning_rate": 1.5834619188503e-05, + "loss": 0.4508, + "step": 1196 + }, + { + "epoch": 0.968936557725387, + "grad_norm": 2.4384043216705322, + "learning_rate": 1.582751588860403e-05, + "loss": 0.472, + "step": 1197 + }, + { + "epoch": 0.969746028533846, + "grad_norm": 2.3286690711975098, + "learning_rate": 1.5820408133489215e-05, + "loss": 0.468, + "step": 1198 + }, + { + "epoch": 0.970555499342305, + "grad_norm": 2.6863343715667725, + "learning_rate": 1.5813295928592533e-05, + "loss": 0.5073, + "step": 1199 + }, + { + "epoch": 0.971364970150764, + "grad_norm": 2.47938871383667, + "learning_rate": 1.580617927935135e-05, + "loss": 0.4848, + "step": 1200 + }, + { + "epoch": 0.9721744409592229, + "grad_norm": 2.2257871627807617, + "learning_rate": 1.579905819120646e-05, + "loss": 0.4346, + "step": 1201 + }, + { + "epoch": 0.9729839117676818, + "grad_norm": 2.295557737350464, + "learning_rate": 1.5791932669602023e-05, + "loss": 0.4454, + "step": 1202 + }, + { + "epoch": 0.9737933825761409, + "grad_norm": 2.436694860458374, + "learning_rate": 1.5784802719985595e-05, + "loss": 0.4712, + "step": 1203 + }, + { + "epoch": 0.9746028533845998, + "grad_norm": 2.480098009109497, + "learning_rate": 1.5777668347808122e-05, + "loss": 0.4671, + "step": 1204 + }, + { + "epoch": 0.9754123241930588, + "grad_norm": 2.4333441257476807, + "learning_rate": 1.5770529558523937e-05, + "loss": 0.4544, + "step": 1205 + }, + { + "epoch": 0.9762217950015177, + "grad_norm": 2.287688732147217, + "learning_rate": 1.5763386357590734e-05, + "loss": 0.4161, + "step": 1206 + }, + { + "epoch": 0.9770312658099767, + "grad_norm": 2.352911949157715, + "learning_rate": 1.5756238750469593e-05, + "loss": 0.4585, + "step": 1207 + }, + { + "epoch": 0.9778407366184357, + "grad_norm": 2.4072136878967285, + "learning_rate": 1.5749086742624952e-05, + "loss": 0.4617, + "step": 1208 + }, + { + "epoch": 0.9786502074268947, + "grad_norm": 2.5311672687530518, + "learning_rate": 1.5741930339524623e-05, + "loss": 0.5463, + "step": 1209 + }, + { + "epoch": 0.9794596782353536, + "grad_norm": 2.396151542663574, + "learning_rate": 1.5734769546639776e-05, + "loss": 0.4456, + "step": 1210 + }, + { + "epoch": 0.9802691490438126, + "grad_norm": 2.2795047760009766, + "learning_rate": 1.5727604369444932e-05, + "loss": 0.4495, + "step": 1211 + }, + { + "epoch": 0.9810786198522716, + "grad_norm": 2.6787383556365967, + "learning_rate": 1.572043481341797e-05, + "loss": 0.4939, + "step": 1212 + }, + { + "epoch": 0.9818880906607306, + "grad_norm": 2.1660823822021484, + "learning_rate": 1.571326088404011e-05, + "loss": 0.4637, + "step": 1213 + }, + { + "epoch": 0.9826975614691895, + "grad_norm": 2.0883231163024902, + "learning_rate": 1.5706082586795924e-05, + "loss": 0.4038, + "step": 1214 + }, + { + "epoch": 0.9835070322776485, + "grad_norm": 2.360609531402588, + "learning_rate": 1.5698899927173316e-05, + "loss": 0.4916, + "step": 1215 + }, + { + "epoch": 0.9843165030861074, + "grad_norm": 2.3815298080444336, + "learning_rate": 1.5691712910663526e-05, + "loss": 0.529, + "step": 1216 + }, + { + "epoch": 0.9851259738945665, + "grad_norm": 2.1789352893829346, + "learning_rate": 1.5684521542761134e-05, + "loss": 0.4669, + "step": 1217 + }, + { + "epoch": 0.9859354447030254, + "grad_norm": 2.321683406829834, + "learning_rate": 1.5677325828964035e-05, + "loss": 0.4711, + "step": 1218 + }, + { + "epoch": 0.9867449155114844, + "grad_norm": 2.2128498554229736, + "learning_rate": 1.5670125774773447e-05, + "loss": 0.4311, + "step": 1219 + }, + { + "epoch": 0.9875543863199433, + "grad_norm": 2.2772912979125977, + "learning_rate": 1.5662921385693914e-05, + "loss": 0.4529, + "step": 1220 + }, + { + "epoch": 0.9883638571284024, + "grad_norm": 2.255398750305176, + "learning_rate": 1.5655712667233293e-05, + "loss": 0.4756, + "step": 1221 + }, + { + "epoch": 0.9891733279368613, + "grad_norm": 2.372117519378662, + "learning_rate": 1.5648499624902743e-05, + "loss": 0.4725, + "step": 1222 + }, + { + "epoch": 0.9899827987453202, + "grad_norm": 2.4989511966705322, + "learning_rate": 1.564128226421674e-05, + "loss": 0.4495, + "step": 1223 + }, + { + "epoch": 0.9907922695537792, + "grad_norm": 2.5313796997070312, + "learning_rate": 1.5634060590693046e-05, + "loss": 0.4603, + "step": 1224 + }, + { + "epoch": 0.9916017403622381, + "grad_norm": 2.423219680786133, + "learning_rate": 1.5626834609852738e-05, + "loss": 0.4536, + "step": 1225 + }, + { + "epoch": 0.9924112111706972, + "grad_norm": 2.4433045387268066, + "learning_rate": 1.561960432722017e-05, + "loss": 0.5129, + "step": 1226 + }, + { + "epoch": 0.9932206819791561, + "grad_norm": 2.3895390033721924, + "learning_rate": 1.5612369748323e-05, + "loss": 0.4512, + "step": 1227 + }, + { + "epoch": 0.9940301527876151, + "grad_norm": 2.4500772953033447, + "learning_rate": 1.5605130878692162e-05, + "loss": 0.4596, + "step": 1228 + }, + { + "epoch": 0.994839623596074, + "grad_norm": 2.2707269191741943, + "learning_rate": 1.5597887723861864e-05, + "loss": 0.4324, + "step": 1229 + }, + { + "epoch": 0.9956490944045331, + "grad_norm": 2.3958611488342285, + "learning_rate": 1.5590640289369605e-05, + "loss": 0.4972, + "step": 1230 + }, + { + "epoch": 0.996458565212992, + "grad_norm": 2.3361308574676514, + "learning_rate": 1.5583388580756144e-05, + "loss": 0.4181, + "step": 1231 + }, + { + "epoch": 0.997268036021451, + "grad_norm": 2.4059741497039795, + "learning_rate": 1.557613260356551e-05, + "loss": 0.4924, + "step": 1232 + }, + { + "epoch": 0.9980775068299099, + "grad_norm": 2.4069783687591553, + "learning_rate": 1.5568872363345e-05, + "loss": 0.484, + "step": 1233 + }, + { + "epoch": 0.9988869776383689, + "grad_norm": 2.1920418739318848, + "learning_rate": 1.5561607865645174e-05, + "loss": 0.4198, + "step": 1234 + }, + { + "epoch": 0.9996964484468279, + "grad_norm": 2.3612093925476074, + "learning_rate": 1.555433911601983e-05, + "loss": 0.5021, + "step": 1235 + }, + { + "epoch": 1.0005059192552868, + "grad_norm": 2.0400185585021973, + "learning_rate": 1.5547066120026025e-05, + "loss": 0.4195, + "step": 1236 + }, + { + "epoch": 1.001315390063746, + "grad_norm": 1.9714311361312866, + "learning_rate": 1.553978888322407e-05, + "loss": 0.3427, + "step": 1237 + }, + { + "epoch": 1.0021248608722049, + "grad_norm": 1.9711785316467285, + "learning_rate": 1.5532507411177514e-05, + "loss": 0.336, + "step": 1238 + }, + { + "epoch": 1.0029343316806638, + "grad_norm": 2.0740904808044434, + "learning_rate": 1.5525221709453142e-05, + "loss": 0.3054, + "step": 1239 + }, + { + "epoch": 1.0037438024891228, + "grad_norm": 2.2183778285980225, + "learning_rate": 1.5517931783620968e-05, + "loss": 0.3019, + "step": 1240 + }, + { + "epoch": 1.0045532732975817, + "grad_norm": 2.0694656372070312, + "learning_rate": 1.5510637639254245e-05, + "loss": 0.301, + "step": 1241 + }, + { + "epoch": 1.0053627441060407, + "grad_norm": 2.2894346714019775, + "learning_rate": 1.5503339281929442e-05, + "loss": 0.3423, + "step": 1242 + }, + { + "epoch": 1.0061722149144996, + "grad_norm": 2.0487759113311768, + "learning_rate": 1.5496036717226257e-05, + "loss": 0.2873, + "step": 1243 + }, + { + "epoch": 1.0069816857229585, + "grad_norm": 2.5916972160339355, + "learning_rate": 1.54887299507276e-05, + "loss": 0.3122, + "step": 1244 + }, + { + "epoch": 1.0077911565314175, + "grad_norm": 2.801248073577881, + "learning_rate": 1.5481418988019592e-05, + "loss": 0.3099, + "step": 1245 + }, + { + "epoch": 1.0086006273398767, + "grad_norm": 2.2621991634368896, + "learning_rate": 1.547410383469157e-05, + "loss": 0.3179, + "step": 1246 + }, + { + "epoch": 1.0094100981483356, + "grad_norm": 2.4925899505615234, + "learning_rate": 1.546678449633606e-05, + "loss": 0.3266, + "step": 1247 + }, + { + "epoch": 1.0102195689567945, + "grad_norm": 2.4119088649749756, + "learning_rate": 1.5459460978548806e-05, + "loss": 0.3149, + "step": 1248 + }, + { + "epoch": 1.0110290397652535, + "grad_norm": 2.3748562335968018, + "learning_rate": 1.5452133286928732e-05, + "loss": 0.3175, + "step": 1249 + }, + { + "epoch": 1.0118385105737124, + "grad_norm": 2.689814329147339, + "learning_rate": 1.5444801427077964e-05, + "loss": 0.2732, + "step": 1250 + }, + { + "epoch": 1.0126479813821714, + "grad_norm": 2.3829123973846436, + "learning_rate": 1.54374654046018e-05, + "loss": 0.3215, + "step": 1251 + }, + { + "epoch": 1.0134574521906303, + "grad_norm": 2.3560941219329834, + "learning_rate": 1.543012522510874e-05, + "loss": 0.3133, + "step": 1252 + }, + { + "epoch": 1.0142669229990893, + "grad_norm": 2.357658624649048, + "learning_rate": 1.5422780894210446e-05, + "loss": 0.3099, + "step": 1253 + }, + { + "epoch": 1.0150763938075482, + "grad_norm": 2.2116565704345703, + "learning_rate": 1.5415432417521766e-05, + "loss": 0.314, + "step": 1254 + }, + { + "epoch": 1.0158858646160074, + "grad_norm": 2.0837275981903076, + "learning_rate": 1.5408079800660697e-05, + "loss": 0.3058, + "step": 1255 + }, + { + "epoch": 1.0166953354244663, + "grad_norm": 2.235020160675049, + "learning_rate": 1.5400723049248433e-05, + "loss": 0.3103, + "step": 1256 + }, + { + "epoch": 1.0175048062329253, + "grad_norm": 2.2880759239196777, + "learning_rate": 1.5393362168909303e-05, + "loss": 0.3402, + "step": 1257 + }, + { + "epoch": 1.0183142770413842, + "grad_norm": 2.062124013900757, + "learning_rate": 1.53859971652708e-05, + "loss": 0.3154, + "step": 1258 + }, + { + "epoch": 1.0191237478498432, + "grad_norm": 2.3065574169158936, + "learning_rate": 1.5378628043963567e-05, + "loss": 0.3243, + "step": 1259 + }, + { + "epoch": 1.019933218658302, + "grad_norm": 2.1475465297698975, + "learning_rate": 1.5371254810621414e-05, + "loss": 0.2854, + "step": 1260 + }, + { + "epoch": 1.020742689466761, + "grad_norm": 2.161513328552246, + "learning_rate": 1.536387747088126e-05, + "loss": 0.3113, + "step": 1261 + }, + { + "epoch": 1.02155216027522, + "grad_norm": 2.1765151023864746, + "learning_rate": 1.535649603038319e-05, + "loss": 0.2902, + "step": 1262 + }, + { + "epoch": 1.022361631083679, + "grad_norm": 2.2902815341949463, + "learning_rate": 1.5349110494770423e-05, + "loss": 0.2929, + "step": 1263 + }, + { + "epoch": 1.0231711018921381, + "grad_norm": 2.3274900913238525, + "learning_rate": 1.5341720869689294e-05, + "loss": 0.3043, + "step": 1264 + }, + { + "epoch": 1.023980572700597, + "grad_norm": 2.7106592655181885, + "learning_rate": 1.5334327160789273e-05, + "loss": 0.2865, + "step": 1265 + }, + { + "epoch": 1.024790043509056, + "grad_norm": 2.4650166034698486, + "learning_rate": 1.5326929373722957e-05, + "loss": 0.302, + "step": 1266 + }, + { + "epoch": 1.025599514317515, + "grad_norm": 2.196438789367676, + "learning_rate": 1.531952751414605e-05, + "loss": 0.3086, + "step": 1267 + }, + { + "epoch": 1.026408985125974, + "grad_norm": 2.3652966022491455, + "learning_rate": 1.531212158771738e-05, + "loss": 0.3066, + "step": 1268 + }, + { + "epoch": 1.0272184559344328, + "grad_norm": 2.3252246379852295, + "learning_rate": 1.5304711600098882e-05, + "loss": 0.3044, + "step": 1269 + }, + { + "epoch": 1.0280279267428918, + "grad_norm": 2.2796199321746826, + "learning_rate": 1.529729755695558e-05, + "loss": 0.3044, + "step": 1270 + }, + { + "epoch": 1.0288373975513507, + "grad_norm": 2.0896389484405518, + "learning_rate": 1.5289879463955622e-05, + "loss": 0.2799, + "step": 1271 + }, + { + "epoch": 1.0296468683598097, + "grad_norm": 2.3632619380950928, + "learning_rate": 1.5282457326770245e-05, + "loss": 0.3206, + "step": 1272 + }, + { + "epoch": 1.0304563391682688, + "grad_norm": 2.067262649536133, + "learning_rate": 1.5275031151073768e-05, + "loss": 0.2842, + "step": 1273 + }, + { + "epoch": 1.0312658099767278, + "grad_norm": 2.3579492568969727, + "learning_rate": 1.5267600942543606e-05, + "loss": 0.3168, + "step": 1274 + }, + { + "epoch": 1.0320752807851867, + "grad_norm": 2.3882062435150146, + "learning_rate": 1.5260166706860254e-05, + "loss": 0.3131, + "step": 1275 + }, + { + "epoch": 1.0328847515936457, + "grad_norm": 2.5632646083831787, + "learning_rate": 1.5252728449707288e-05, + "loss": 0.3437, + "step": 1276 + }, + { + "epoch": 1.0336942224021046, + "grad_norm": 2.306560516357422, + "learning_rate": 1.5245286176771361e-05, + "loss": 0.3244, + "step": 1277 + }, + { + "epoch": 1.0345036932105636, + "grad_norm": 2.1407761573791504, + "learning_rate": 1.5237839893742188e-05, + "loss": 0.2903, + "step": 1278 + }, + { + "epoch": 1.0353131640190225, + "grad_norm": 2.2571778297424316, + "learning_rate": 1.5230389606312557e-05, + "loss": 0.281, + "step": 1279 + }, + { + "epoch": 1.0361226348274815, + "grad_norm": 2.3582472801208496, + "learning_rate": 1.5222935320178315e-05, + "loss": 0.311, + "step": 1280 + }, + { + "epoch": 1.0369321056359404, + "grad_norm": 2.218230724334717, + "learning_rate": 1.5215477041038364e-05, + "loss": 0.2911, + "step": 1281 + }, + { + "epoch": 1.0377415764443996, + "grad_norm": 2.718247890472412, + "learning_rate": 1.5208014774594663e-05, + "loss": 0.3275, + "step": 1282 + }, + { + "epoch": 1.0385510472528585, + "grad_norm": 2.1194896697998047, + "learning_rate": 1.5200548526552219e-05, + "loss": 0.2827, + "step": 1283 + }, + { + "epoch": 1.0393605180613175, + "grad_norm": 2.2449536323547363, + "learning_rate": 1.5193078302619073e-05, + "loss": 0.3023, + "step": 1284 + }, + { + "epoch": 1.0401699888697764, + "grad_norm": 2.4261317253112793, + "learning_rate": 1.5185604108506327e-05, + "loss": 0.3195, + "step": 1285 + }, + { + "epoch": 1.0409794596782354, + "grad_norm": 2.375431776046753, + "learning_rate": 1.517812594992809e-05, + "loss": 0.3439, + "step": 1286 + }, + { + "epoch": 1.0417889304866943, + "grad_norm": 2.256706953048706, + "learning_rate": 1.5170643832601526e-05, + "loss": 0.2827, + "step": 1287 + }, + { + "epoch": 1.0425984012951532, + "grad_norm": 2.3537724018096924, + "learning_rate": 1.5163157762246814e-05, + "loss": 0.3166, + "step": 1288 + }, + { + "epoch": 1.0434078721036122, + "grad_norm": 2.138995885848999, + "learning_rate": 1.515566774458716e-05, + "loss": 0.2894, + "step": 1289 + }, + { + "epoch": 1.0442173429120711, + "grad_norm": 2.3536109924316406, + "learning_rate": 1.5148173785348777e-05, + "loss": 0.3407, + "step": 1290 + }, + { + "epoch": 1.0450268137205303, + "grad_norm": 2.1912403106689453, + "learning_rate": 1.5140675890260912e-05, + "loss": 0.2936, + "step": 1291 + }, + { + "epoch": 1.0458362845289892, + "grad_norm": 2.0768282413482666, + "learning_rate": 1.51331740650558e-05, + "loss": 0.2816, + "step": 1292 + }, + { + "epoch": 1.0466457553374482, + "grad_norm": 2.2444100379943848, + "learning_rate": 1.5125668315468689e-05, + "loss": 0.3012, + "step": 1293 + }, + { + "epoch": 1.0474552261459071, + "grad_norm": 2.0570876598358154, + "learning_rate": 1.5118158647237833e-05, + "loss": 0.2819, + "step": 1294 + }, + { + "epoch": 1.048264696954366, + "grad_norm": 2.2215828895568848, + "learning_rate": 1.5110645066104473e-05, + "loss": 0.2993, + "step": 1295 + }, + { + "epoch": 1.049074167762825, + "grad_norm": 2.2804620265960693, + "learning_rate": 1.5103127577812844e-05, + "loss": 0.2971, + "step": 1296 + }, + { + "epoch": 1.049883638571284, + "grad_norm": 2.385284662246704, + "learning_rate": 1.5095606188110173e-05, + "loss": 0.2919, + "step": 1297 + }, + { + "epoch": 1.050693109379743, + "grad_norm": 2.4513211250305176, + "learning_rate": 1.508808090274666e-05, + "loss": 0.3237, + "step": 1298 + }, + { + "epoch": 1.0515025801882019, + "grad_norm": 2.194413185119629, + "learning_rate": 1.5080551727475494e-05, + "loss": 0.2989, + "step": 1299 + }, + { + "epoch": 1.052312050996661, + "grad_norm": 2.278304100036621, + "learning_rate": 1.507301866805283e-05, + "loss": 0.3189, + "step": 1300 + }, + { + "epoch": 1.05312152180512, + "grad_norm": 2.0453970432281494, + "learning_rate": 1.5065481730237795e-05, + "loss": 0.2937, + "step": 1301 + }, + { + "epoch": 1.053930992613579, + "grad_norm": 2.4068217277526855, + "learning_rate": 1.5057940919792482e-05, + "loss": 0.3092, + "step": 1302 + }, + { + "epoch": 1.0547404634220379, + "grad_norm": 2.3076963424682617, + "learning_rate": 1.5050396242481946e-05, + "loss": 0.3099, + "step": 1303 + }, + { + "epoch": 1.0555499342304968, + "grad_norm": 2.2794859409332275, + "learning_rate": 1.504284770407419e-05, + "loss": 0.3152, + "step": 1304 + }, + { + "epoch": 1.0563594050389558, + "grad_norm": 2.1736249923706055, + "learning_rate": 1.5035295310340186e-05, + "loss": 0.3285, + "step": 1305 + }, + { + "epoch": 1.0571688758474147, + "grad_norm": 2.281214714050293, + "learning_rate": 1.5027739067053832e-05, + "loss": 0.3035, + "step": 1306 + }, + { + "epoch": 1.0579783466558736, + "grad_norm": 2.2812411785125732, + "learning_rate": 1.5020178979991986e-05, + "loss": 0.3035, + "step": 1307 + }, + { + "epoch": 1.0587878174643326, + "grad_norm": 2.0556745529174805, + "learning_rate": 1.5012615054934436e-05, + "loss": 0.3138, + "step": 1308 + }, + { + "epoch": 1.0595972882727918, + "grad_norm": 2.2270169258117676, + "learning_rate": 1.5005047297663905e-05, + "loss": 0.3029, + "step": 1309 + }, + { + "epoch": 1.0604067590812507, + "grad_norm": 2.1479833126068115, + "learning_rate": 1.499747571396605e-05, + "loss": 0.2976, + "step": 1310 + }, + { + "epoch": 1.0612162298897097, + "grad_norm": 2.242250919342041, + "learning_rate": 1.4989900309629453e-05, + "loss": 0.3283, + "step": 1311 + }, + { + "epoch": 1.0620257006981686, + "grad_norm": 2.033079147338867, + "learning_rate": 1.4982321090445609e-05, + "loss": 0.2747, + "step": 1312 + }, + { + "epoch": 1.0628351715066275, + "grad_norm": 2.1253302097320557, + "learning_rate": 1.4974738062208935e-05, + "loss": 0.291, + "step": 1313 + }, + { + "epoch": 1.0636446423150865, + "grad_norm": 2.1776559352874756, + "learning_rate": 1.4967151230716763e-05, + "loss": 0.2872, + "step": 1314 + }, + { + "epoch": 1.0644541131235454, + "grad_norm": 2.304260730743408, + "learning_rate": 1.495956060176933e-05, + "loss": 0.3009, + "step": 1315 + }, + { + "epoch": 1.0652635839320044, + "grad_norm": 2.357372522354126, + "learning_rate": 1.4951966181169775e-05, + "loss": 0.2932, + "step": 1316 + }, + { + "epoch": 1.0660730547404633, + "grad_norm": 2.164155960083008, + "learning_rate": 1.4944367974724137e-05, + "loss": 0.2908, + "step": 1317 + }, + { + "epoch": 1.0668825255489225, + "grad_norm": 2.2770445346832275, + "learning_rate": 1.4936765988241346e-05, + "loss": 0.2999, + "step": 1318 + }, + { + "epoch": 1.0676919963573814, + "grad_norm": 2.161355495452881, + "learning_rate": 1.4929160227533233e-05, + "loss": 0.288, + "step": 1319 + }, + { + "epoch": 1.0685014671658404, + "grad_norm": 2.2771058082580566, + "learning_rate": 1.4921550698414497e-05, + "loss": 0.2926, + "step": 1320 + }, + { + "epoch": 1.0693109379742993, + "grad_norm": 2.2670955657958984, + "learning_rate": 1.4913937406702735e-05, + "loss": 0.3033, + "step": 1321 + }, + { + "epoch": 1.0701204087827583, + "grad_norm": 2.316364288330078, + "learning_rate": 1.4906320358218407e-05, + "loss": 0.3049, + "step": 1322 + }, + { + "epoch": 1.0709298795912172, + "grad_norm": 2.241234064102173, + "learning_rate": 1.4898699558784858e-05, + "loss": 0.3227, + "step": 1323 + }, + { + "epoch": 1.0717393503996762, + "grad_norm": 2.226961374282837, + "learning_rate": 1.4891075014228282e-05, + "loss": 0.2734, + "step": 1324 + }, + { + "epoch": 1.072548821208135, + "grad_norm": 2.371670722961426, + "learning_rate": 1.4883446730377762e-05, + "loss": 0.2996, + "step": 1325 + }, + { + "epoch": 1.073358292016594, + "grad_norm": 2.1056933403015137, + "learning_rate": 1.4875814713065216e-05, + "loss": 0.2985, + "step": 1326 + }, + { + "epoch": 1.0741677628250532, + "grad_norm": 2.095703601837158, + "learning_rate": 1.4868178968125427e-05, + "loss": 0.3131, + "step": 1327 + }, + { + "epoch": 1.0749772336335122, + "grad_norm": 2.4207797050476074, + "learning_rate": 1.4860539501396031e-05, + "loss": 0.2995, + "step": 1328 + }, + { + "epoch": 1.075786704441971, + "grad_norm": 2.3861448764801025, + "learning_rate": 1.4852896318717504e-05, + "loss": 0.3097, + "step": 1329 + }, + { + "epoch": 1.07659617525043, + "grad_norm": 2.264679431915283, + "learning_rate": 1.484524942593316e-05, + "loss": 0.2854, + "step": 1330 + }, + { + "epoch": 1.077405646058889, + "grad_norm": 2.2002522945404053, + "learning_rate": 1.4837598828889159e-05, + "loss": 0.2666, + "step": 1331 + }, + { + "epoch": 1.078215116867348, + "grad_norm": 2.415808916091919, + "learning_rate": 1.4829944533434484e-05, + "loss": 0.3274, + "step": 1332 + }, + { + "epoch": 1.079024587675807, + "grad_norm": 2.41100811958313, + "learning_rate": 1.4822286545420948e-05, + "loss": 0.2808, + "step": 1333 + }, + { + "epoch": 1.0798340584842658, + "grad_norm": 2.6017239093780518, + "learning_rate": 1.4814624870703194e-05, + "loss": 0.3315, + "step": 1334 + }, + { + "epoch": 1.0806435292927248, + "grad_norm": 2.2788467407226562, + "learning_rate": 1.4806959515138667e-05, + "loss": 0.3098, + "step": 1335 + }, + { + "epoch": 1.081453000101184, + "grad_norm": 2.352787971496582, + "learning_rate": 1.4799290484587646e-05, + "loss": 0.3114, + "step": 1336 + }, + { + "epoch": 1.082262470909643, + "grad_norm": 2.292006015777588, + "learning_rate": 1.4791617784913203e-05, + "loss": 0.3075, + "step": 1337 + }, + { + "epoch": 1.0830719417181018, + "grad_norm": 2.357194662094116, + "learning_rate": 1.478394142198123e-05, + "loss": 0.3118, + "step": 1338 + }, + { + "epoch": 1.0838814125265608, + "grad_norm": 2.0139942169189453, + "learning_rate": 1.4776261401660404e-05, + "loss": 0.2666, + "step": 1339 + }, + { + "epoch": 1.0846908833350197, + "grad_norm": 2.136094331741333, + "learning_rate": 1.4768577729822214e-05, + "loss": 0.2937, + "step": 1340 + }, + { + "epoch": 1.0855003541434787, + "grad_norm": 2.2587780952453613, + "learning_rate": 1.476089041234092e-05, + "loss": 0.2759, + "step": 1341 + }, + { + "epoch": 1.0863098249519376, + "grad_norm": 2.2803754806518555, + "learning_rate": 1.4753199455093596e-05, + "loss": 0.3158, + "step": 1342 + }, + { + "epoch": 1.0871192957603966, + "grad_norm": 2.311800241470337, + "learning_rate": 1.4745504863960076e-05, + "loss": 0.3084, + "step": 1343 + }, + { + "epoch": 1.0879287665688555, + "grad_norm": 2.2190463542938232, + "learning_rate": 1.4737806644822984e-05, + "loss": 0.2915, + "step": 1344 + }, + { + "epoch": 1.0887382373773147, + "grad_norm": 2.268876791000366, + "learning_rate": 1.4730104803567708e-05, + "loss": 0.313, + "step": 1345 + }, + { + "epoch": 1.0895477081857736, + "grad_norm": 2.2732291221618652, + "learning_rate": 1.4722399346082421e-05, + "loss": 0.3179, + "step": 1346 + }, + { + "epoch": 1.0903571789942326, + "grad_norm": 2.3804373741149902, + "learning_rate": 1.4714690278258039e-05, + "loss": 0.3068, + "step": 1347 + }, + { + "epoch": 1.0911666498026915, + "grad_norm": 2.2066843509674072, + "learning_rate": 1.4706977605988264e-05, + "loss": 0.2937, + "step": 1348 + }, + { + "epoch": 1.0919761206111505, + "grad_norm": 2.285022497177124, + "learning_rate": 1.469926133516953e-05, + "loss": 0.3279, + "step": 1349 + }, + { + "epoch": 1.0927855914196094, + "grad_norm": 2.2723982334136963, + "learning_rate": 1.4691541471701034e-05, + "loss": 0.2992, + "step": 1350 + }, + { + "epoch": 1.0935950622280683, + "grad_norm": 2.227325677871704, + "learning_rate": 1.4683818021484717e-05, + "loss": 0.3022, + "step": 1351 + }, + { + "epoch": 1.0944045330365273, + "grad_norm": 2.074115753173828, + "learning_rate": 1.4676090990425265e-05, + "loss": 0.2912, + "step": 1352 + }, + { + "epoch": 1.0952140038449862, + "grad_norm": 2.051661729812622, + "learning_rate": 1.4668360384430098e-05, + "loss": 0.2947, + "step": 1353 + }, + { + "epoch": 1.0960234746534454, + "grad_norm": 2.31601881980896, + "learning_rate": 1.4660626209409371e-05, + "loss": 0.3006, + "step": 1354 + }, + { + "epoch": 1.0968329454619044, + "grad_norm": 2.3512141704559326, + "learning_rate": 1.4652888471275963e-05, + "loss": 0.3282, + "step": 1355 + }, + { + "epoch": 1.0976424162703633, + "grad_norm": 2.2796595096588135, + "learning_rate": 1.4645147175945489e-05, + "loss": 0.3088, + "step": 1356 + }, + { + "epoch": 1.0984518870788222, + "grad_norm": 2.228437900543213, + "learning_rate": 1.4637402329336268e-05, + "loss": 0.3135, + "step": 1357 + }, + { + "epoch": 1.0992613578872812, + "grad_norm": 2.1651415824890137, + "learning_rate": 1.4629653937369343e-05, + "loss": 0.2812, + "step": 1358 + }, + { + "epoch": 1.1000708286957401, + "grad_norm": 2.4091193675994873, + "learning_rate": 1.4621902005968469e-05, + "loss": 0.3401, + "step": 1359 + }, + { + "epoch": 1.100880299504199, + "grad_norm": 2.331089496612549, + "learning_rate": 1.4614146541060103e-05, + "loss": 0.2996, + "step": 1360 + }, + { + "epoch": 1.101689770312658, + "grad_norm": 2.0837929248809814, + "learning_rate": 1.46063875485734e-05, + "loss": 0.2729, + "step": 1361 + }, + { + "epoch": 1.102499241121117, + "grad_norm": 2.224663496017456, + "learning_rate": 1.4598625034440224e-05, + "loss": 0.2927, + "step": 1362 + }, + { + "epoch": 1.1033087119295761, + "grad_norm": 2.193199396133423, + "learning_rate": 1.4590859004595115e-05, + "loss": 0.3093, + "step": 1363 + }, + { + "epoch": 1.104118182738035, + "grad_norm": 2.2144150733947754, + "learning_rate": 1.4583089464975315e-05, + "loss": 0.3123, + "step": 1364 + }, + { + "epoch": 1.104927653546494, + "grad_norm": 2.0453903675079346, + "learning_rate": 1.4575316421520742e-05, + "loss": 0.3106, + "step": 1365 + }, + { + "epoch": 1.105737124354953, + "grad_norm": 2.1915595531463623, + "learning_rate": 1.4567539880173995e-05, + "loss": 0.3163, + "step": 1366 + }, + { + "epoch": 1.106546595163412, + "grad_norm": 2.006150722503662, + "learning_rate": 1.4559759846880344e-05, + "loss": 0.3107, + "step": 1367 + }, + { + "epoch": 1.1073560659718709, + "grad_norm": 2.2080488204956055, + "learning_rate": 1.4551976327587736e-05, + "loss": 0.3227, + "step": 1368 + }, + { + "epoch": 1.1081655367803298, + "grad_norm": 2.284022331237793, + "learning_rate": 1.4544189328246774e-05, + "loss": 0.3354, + "step": 1369 + }, + { + "epoch": 1.1089750075887888, + "grad_norm": 2.430936813354492, + "learning_rate": 1.4536398854810727e-05, + "loss": 0.3384, + "step": 1370 + }, + { + "epoch": 1.1097844783972477, + "grad_norm": 2.188613176345825, + "learning_rate": 1.452860491323552e-05, + "loss": 0.2893, + "step": 1371 + }, + { + "epoch": 1.1105939492057069, + "grad_norm": 2.134183883666992, + "learning_rate": 1.4520807509479726e-05, + "loss": 0.2934, + "step": 1372 + }, + { + "epoch": 1.1114034200141658, + "grad_norm": 2.30029034614563, + "learning_rate": 1.4513006649504569e-05, + "loss": 0.305, + "step": 1373 + }, + { + "epoch": 1.1122128908226248, + "grad_norm": 2.146151065826416, + "learning_rate": 1.4505202339273916e-05, + "loss": 0.2994, + "step": 1374 + }, + { + "epoch": 1.1130223616310837, + "grad_norm": 2.190295457839966, + "learning_rate": 1.449739458475426e-05, + "loss": 0.2877, + "step": 1375 + }, + { + "epoch": 1.1138318324395426, + "grad_norm": 2.2233071327209473, + "learning_rate": 1.448958339191475e-05, + "loss": 0.2934, + "step": 1376 + }, + { + "epoch": 1.1146413032480016, + "grad_norm": 2.4796159267425537, + "learning_rate": 1.4481768766727138e-05, + "loss": 0.3117, + "step": 1377 + }, + { + "epoch": 1.1154507740564605, + "grad_norm": 2.350609064102173, + "learning_rate": 1.4473950715165817e-05, + "loss": 0.3062, + "step": 1378 + }, + { + "epoch": 1.1162602448649195, + "grad_norm": 2.1366775035858154, + "learning_rate": 1.4466129243207796e-05, + "loss": 0.2754, + "step": 1379 + }, + { + "epoch": 1.1170697156733784, + "grad_norm": 2.223400115966797, + "learning_rate": 1.4458304356832695e-05, + "loss": 0.2979, + "step": 1380 + }, + { + "epoch": 1.1178791864818376, + "grad_norm": 2.087167263031006, + "learning_rate": 1.4450476062022749e-05, + "loss": 0.2833, + "step": 1381 + }, + { + "epoch": 1.1186886572902965, + "grad_norm": 2.344595432281494, + "learning_rate": 1.4442644364762797e-05, + "loss": 0.3106, + "step": 1382 + }, + { + "epoch": 1.1194981280987555, + "grad_norm": 2.394104480743408, + "learning_rate": 1.443480927104028e-05, + "loss": 0.3127, + "step": 1383 + }, + { + "epoch": 1.1203075989072144, + "grad_norm": 2.1381983757019043, + "learning_rate": 1.4426970786845228e-05, + "loss": 0.2786, + "step": 1384 + }, + { + "epoch": 1.1211170697156734, + "grad_norm": 2.227415084838867, + "learning_rate": 1.4419128918170278e-05, + "loss": 0.3155, + "step": 1385 + }, + { + "epoch": 1.1219265405241323, + "grad_norm": 2.1897544860839844, + "learning_rate": 1.4411283671010638e-05, + "loss": 0.2873, + "step": 1386 + }, + { + "epoch": 1.1227360113325913, + "grad_norm": 2.0919973850250244, + "learning_rate": 1.4403435051364116e-05, + "loss": 0.2721, + "step": 1387 + }, + { + "epoch": 1.1235454821410502, + "grad_norm": 2.2888474464416504, + "learning_rate": 1.4395583065231083e-05, + "loss": 0.3347, + "step": 1388 + }, + { + "epoch": 1.1243549529495092, + "grad_norm": 2.3464243412017822, + "learning_rate": 1.4387727718614492e-05, + "loss": 0.3225, + "step": 1389 + }, + { + "epoch": 1.1251644237579683, + "grad_norm": 2.1729278564453125, + "learning_rate": 1.4379869017519865e-05, + "loss": 0.2973, + "step": 1390 + }, + { + "epoch": 1.1259738945664273, + "grad_norm": 2.3403799533843994, + "learning_rate": 1.4372006967955285e-05, + "loss": 0.2971, + "step": 1391 + }, + { + "epoch": 1.1267833653748862, + "grad_norm": 2.136782169342041, + "learning_rate": 1.4364141575931395e-05, + "loss": 0.2776, + "step": 1392 + }, + { + "epoch": 1.1275928361833452, + "grad_norm": 2.3993210792541504, + "learning_rate": 1.4356272847461399e-05, + "loss": 0.3179, + "step": 1393 + }, + { + "epoch": 1.128402306991804, + "grad_norm": 2.1702611446380615, + "learning_rate": 1.4348400788561046e-05, + "loss": 0.2873, + "step": 1394 + }, + { + "epoch": 1.129211777800263, + "grad_norm": 2.3093981742858887, + "learning_rate": 1.434052540524863e-05, + "loss": 0.3015, + "step": 1395 + }, + { + "epoch": 1.130021248608722, + "grad_norm": 2.387298107147217, + "learning_rate": 1.4332646703544995e-05, + "loss": 0.3165, + "step": 1396 + }, + { + "epoch": 1.130830719417181, + "grad_norm": 2.1455044746398926, + "learning_rate": 1.4324764689473517e-05, + "loss": 0.2771, + "step": 1397 + }, + { + "epoch": 1.13164019022564, + "grad_norm": 2.241785764694214, + "learning_rate": 1.4316879369060095e-05, + "loss": 0.3099, + "step": 1398 + }, + { + "epoch": 1.132449661034099, + "grad_norm": 2.3026483058929443, + "learning_rate": 1.4308990748333176e-05, + "loss": 0.2811, + "step": 1399 + }, + { + "epoch": 1.133259131842558, + "grad_norm": 2.3456287384033203, + "learning_rate": 1.430109883332371e-05, + "loss": 0.3219, + "step": 1400 + }, + { + "epoch": 1.134068602651017, + "grad_norm": 2.0744516849517822, + "learning_rate": 1.4293203630065177e-05, + "loss": 0.2972, + "step": 1401 + }, + { + "epoch": 1.134878073459476, + "grad_norm": 2.067948818206787, + "learning_rate": 1.4285305144593571e-05, + "loss": 0.2989, + "step": 1402 + }, + { + "epoch": 1.1356875442679348, + "grad_norm": 2.173191785812378, + "learning_rate": 1.427740338294739e-05, + "loss": 0.3306, + "step": 1403 + }, + { + "epoch": 1.1364970150763938, + "grad_norm": 2.4132962226867676, + "learning_rate": 1.426949835116764e-05, + "loss": 0.3115, + "step": 1404 + }, + { + "epoch": 1.1373064858848527, + "grad_norm": 2.097028970718384, + "learning_rate": 1.4261590055297828e-05, + "loss": 0.308, + "step": 1405 + }, + { + "epoch": 1.1381159566933117, + "grad_norm": 2.133607864379883, + "learning_rate": 1.425367850138395e-05, + "loss": 0.2983, + "step": 1406 + }, + { + "epoch": 1.1389254275017708, + "grad_norm": 2.3114657402038574, + "learning_rate": 1.4245763695474498e-05, + "loss": 0.3178, + "step": 1407 + }, + { + "epoch": 1.1397348983102298, + "grad_norm": 2.173034429550171, + "learning_rate": 1.4237845643620454e-05, + "loss": 0.2863, + "step": 1408 + }, + { + "epoch": 1.1405443691186887, + "grad_norm": 2.159102439880371, + "learning_rate": 1.4229924351875272e-05, + "loss": 0.2856, + "step": 1409 + }, + { + "epoch": 1.1413538399271477, + "grad_norm": 2.308511972427368, + "learning_rate": 1.4221999826294897e-05, + "loss": 0.3235, + "step": 1410 + }, + { + "epoch": 1.1421633107356066, + "grad_norm": 2.4909510612487793, + "learning_rate": 1.421407207293773e-05, + "loss": 0.3343, + "step": 1411 + }, + { + "epoch": 1.1429727815440656, + "grad_norm": 2.1198372840881348, + "learning_rate": 1.4206141097864645e-05, + "loss": 0.285, + "step": 1412 + }, + { + "epoch": 1.1437822523525245, + "grad_norm": 2.314197540283203, + "learning_rate": 1.419820690713899e-05, + "loss": 0.2908, + "step": 1413 + }, + { + "epoch": 1.1445917231609835, + "grad_norm": 2.216062545776367, + "learning_rate": 1.4190269506826556e-05, + "loss": 0.2984, + "step": 1414 + }, + { + "epoch": 1.1454011939694424, + "grad_norm": 2.324465036392212, + "learning_rate": 1.4182328902995594e-05, + "loss": 0.3445, + "step": 1415 + }, + { + "epoch": 1.1462106647779016, + "grad_norm": 2.288498640060425, + "learning_rate": 1.417438510171681e-05, + "loss": 0.3012, + "step": 1416 + }, + { + "epoch": 1.1470201355863605, + "grad_norm": 2.0858850479125977, + "learning_rate": 1.4166438109063347e-05, + "loss": 0.2941, + "step": 1417 + }, + { + "epoch": 1.1478296063948195, + "grad_norm": 2.2059552669525146, + "learning_rate": 1.4158487931110786e-05, + "loss": 0.3108, + "step": 1418 + }, + { + "epoch": 1.1486390772032784, + "grad_norm": 2.285161018371582, + "learning_rate": 1.415053457393715e-05, + "loss": 0.3271, + "step": 1419 + }, + { + "epoch": 1.1494485480117373, + "grad_norm": 2.187830686569214, + "learning_rate": 1.4142578043622892e-05, + "loss": 0.2849, + "step": 1420 + }, + { + "epoch": 1.1502580188201963, + "grad_norm": 2.0700981616973877, + "learning_rate": 1.4134618346250883e-05, + "loss": 0.2874, + "step": 1421 + }, + { + "epoch": 1.1510674896286552, + "grad_norm": 2.320561408996582, + "learning_rate": 1.4126655487906426e-05, + "loss": 0.3223, + "step": 1422 + }, + { + "epoch": 1.1518769604371142, + "grad_norm": 2.3682754039764404, + "learning_rate": 1.4118689474677234e-05, + "loss": 0.3396, + "step": 1423 + }, + { + "epoch": 1.1526864312455731, + "grad_norm": 2.261134624481201, + "learning_rate": 1.4110720312653432e-05, + "loss": 0.3237, + "step": 1424 + }, + { + "epoch": 1.1534959020540323, + "grad_norm": 2.3450088500976562, + "learning_rate": 1.4102748007927555e-05, + "loss": 0.3073, + "step": 1425 + }, + { + "epoch": 1.1543053728624912, + "grad_norm": 2.2834434509277344, + "learning_rate": 1.4094772566594537e-05, + "loss": 0.3299, + "step": 1426 + }, + { + "epoch": 1.1551148436709502, + "grad_norm": 2.2099032402038574, + "learning_rate": 1.408679399475172e-05, + "loss": 0.2912, + "step": 1427 + }, + { + "epoch": 1.1559243144794091, + "grad_norm": 2.0691452026367188, + "learning_rate": 1.4078812298498821e-05, + "loss": 0.3003, + "step": 1428 + }, + { + "epoch": 1.156733785287868, + "grad_norm": 2.2858176231384277, + "learning_rate": 1.4070827483937963e-05, + "loss": 0.3462, + "step": 1429 + }, + { + "epoch": 1.157543256096327, + "grad_norm": 2.1472790241241455, + "learning_rate": 1.4062839557173641e-05, + "loss": 0.2964, + "step": 1430 + }, + { + "epoch": 1.158352726904786, + "grad_norm": 2.134711742401123, + "learning_rate": 1.4054848524312741e-05, + "loss": 0.3061, + "step": 1431 + }, + { + "epoch": 1.159162197713245, + "grad_norm": 2.0681235790252686, + "learning_rate": 1.4046854391464515e-05, + "loss": 0.2895, + "step": 1432 + }, + { + "epoch": 1.1599716685217039, + "grad_norm": 2.3649094104766846, + "learning_rate": 1.4038857164740584e-05, + "loss": 0.3381, + "step": 1433 + }, + { + "epoch": 1.160781139330163, + "grad_norm": 2.3517186641693115, + "learning_rate": 1.4030856850254946e-05, + "loss": 0.3318, + "step": 1434 + }, + { + "epoch": 1.161590610138622, + "grad_norm": 2.3080835342407227, + "learning_rate": 1.4022853454123939e-05, + "loss": 0.3155, + "step": 1435 + }, + { + "epoch": 1.162400080947081, + "grad_norm": 2.4445419311523438, + "learning_rate": 1.401484698246628e-05, + "loss": 0.3233, + "step": 1436 + }, + { + "epoch": 1.1632095517555399, + "grad_norm": 2.2327065467834473, + "learning_rate": 1.4006837441403018e-05, + "loss": 0.3087, + "step": 1437 + }, + { + "epoch": 1.1640190225639988, + "grad_norm": 2.2465951442718506, + "learning_rate": 1.3998824837057563e-05, + "loss": 0.3295, + "step": 1438 + }, + { + "epoch": 1.1648284933724578, + "grad_norm": 2.124689817428589, + "learning_rate": 1.3990809175555657e-05, + "loss": 0.3051, + "step": 1439 + }, + { + "epoch": 1.1656379641809167, + "grad_norm": 2.194650173187256, + "learning_rate": 1.3982790463025389e-05, + "loss": 0.3219, + "step": 1440 + }, + { + "epoch": 1.1664474349893756, + "grad_norm": 2.1732029914855957, + "learning_rate": 1.3974768705597166e-05, + "loss": 0.2997, + "step": 1441 + }, + { + "epoch": 1.1672569057978346, + "grad_norm": 2.1485719680786133, + "learning_rate": 1.396674390940374e-05, + "loss": 0.3066, + "step": 1442 + }, + { + "epoch": 1.1680663766062938, + "grad_norm": 2.3911197185516357, + "learning_rate": 1.395871608058017e-05, + "loss": 0.3086, + "step": 1443 + }, + { + "epoch": 1.1688758474147527, + "grad_norm": 2.2635796070098877, + "learning_rate": 1.395068522526385e-05, + "loss": 0.3047, + "step": 1444 + }, + { + "epoch": 1.1696853182232116, + "grad_norm": 2.1306238174438477, + "learning_rate": 1.3942651349594472e-05, + "loss": 0.2958, + "step": 1445 + }, + { + "epoch": 1.1704947890316706, + "grad_norm": 2.1929373741149902, + "learning_rate": 1.3934614459714046e-05, + "loss": 0.3113, + "step": 1446 + }, + { + "epoch": 1.1713042598401295, + "grad_norm": 2.180983066558838, + "learning_rate": 1.3926574561766888e-05, + "loss": 0.3149, + "step": 1447 + }, + { + "epoch": 1.1721137306485885, + "grad_norm": 2.282118082046509, + "learning_rate": 1.3918531661899606e-05, + "loss": 0.3147, + "step": 1448 + }, + { + "epoch": 1.1729232014570474, + "grad_norm": 2.8747406005859375, + "learning_rate": 1.3910485766261106e-05, + "loss": 0.3278, + "step": 1449 + }, + { + "epoch": 1.1737326722655064, + "grad_norm": 2.0672929286956787, + "learning_rate": 1.3902436881002595e-05, + "loss": 0.2704, + "step": 1450 + }, + { + "epoch": 1.1745421430739653, + "grad_norm": 2.150176763534546, + "learning_rate": 1.3894385012277545e-05, + "loss": 0.3025, + "step": 1451 + }, + { + "epoch": 1.1753516138824245, + "grad_norm": 2.521544933319092, + "learning_rate": 1.3886330166241725e-05, + "loss": 0.3399, + "step": 1452 + }, + { + "epoch": 1.1761610846908834, + "grad_norm": 2.364291191101074, + "learning_rate": 1.3878272349053176e-05, + "loss": 0.321, + "step": 1453 + }, + { + "epoch": 1.1769705554993424, + "grad_norm": 2.176023244857788, + "learning_rate": 1.3870211566872209e-05, + "loss": 0.3016, + "step": 1454 + }, + { + "epoch": 1.1777800263078013, + "grad_norm": 2.2523159980773926, + "learning_rate": 1.38621478258614e-05, + "loss": 0.2997, + "step": 1455 + }, + { + "epoch": 1.1785894971162603, + "grad_norm": 2.2909271717071533, + "learning_rate": 1.3854081132185597e-05, + "loss": 0.2923, + "step": 1456 + }, + { + "epoch": 1.1793989679247192, + "grad_norm": 2.3282344341278076, + "learning_rate": 1.3846011492011889e-05, + "loss": 0.2926, + "step": 1457 + }, + { + "epoch": 1.1802084387331782, + "grad_norm": 2.1609296798706055, + "learning_rate": 1.3837938911509631e-05, + "loss": 0.2926, + "step": 1458 + }, + { + "epoch": 1.181017909541637, + "grad_norm": 2.25118088722229, + "learning_rate": 1.3829863396850423e-05, + "loss": 0.2985, + "step": 1459 + }, + { + "epoch": 1.181827380350096, + "grad_norm": 2.1945571899414062, + "learning_rate": 1.3821784954208104e-05, + "loss": 0.3023, + "step": 1460 + }, + { + "epoch": 1.1826368511585552, + "grad_norm": 2.3955390453338623, + "learning_rate": 1.3813703589758754e-05, + "loss": 0.3132, + "step": 1461 + }, + { + "epoch": 1.1834463219670142, + "grad_norm": 2.1458489894866943, + "learning_rate": 1.3805619309680691e-05, + "loss": 0.302, + "step": 1462 + }, + { + "epoch": 1.184255792775473, + "grad_norm": 2.332533359527588, + "learning_rate": 1.3797532120154453e-05, + "loss": 0.2971, + "step": 1463 + }, + { + "epoch": 1.185065263583932, + "grad_norm": 2.301321029663086, + "learning_rate": 1.3789442027362809e-05, + "loss": 0.3288, + "step": 1464 + }, + { + "epoch": 1.185874734392391, + "grad_norm": 2.259166717529297, + "learning_rate": 1.3781349037490746e-05, + "loss": 0.2816, + "step": 1465 + }, + { + "epoch": 1.18668420520085, + "grad_norm": 2.2314181327819824, + "learning_rate": 1.3773253156725468e-05, + "loss": 0.3041, + "step": 1466 + }, + { + "epoch": 1.1874936760093089, + "grad_norm": 2.0948009490966797, + "learning_rate": 1.3765154391256386e-05, + "loss": 0.2953, + "step": 1467 + }, + { + "epoch": 1.1883031468177678, + "grad_norm": 2.354736328125, + "learning_rate": 1.3757052747275115e-05, + "loss": 0.3087, + "step": 1468 + }, + { + "epoch": 1.1891126176262268, + "grad_norm": 2.2797865867614746, + "learning_rate": 1.3748948230975472e-05, + "loss": 0.3073, + "step": 1469 + }, + { + "epoch": 1.189922088434686, + "grad_norm": 2.284743547439575, + "learning_rate": 1.3740840848553478e-05, + "loss": 0.2935, + "step": 1470 + }, + { + "epoch": 1.1907315592431449, + "grad_norm": 2.2780027389526367, + "learning_rate": 1.3732730606207332e-05, + "loss": 0.3149, + "step": 1471 + }, + { + "epoch": 1.1915410300516038, + "grad_norm": 2.222621202468872, + "learning_rate": 1.3724617510137426e-05, + "loss": 0.3101, + "step": 1472 + }, + { + "epoch": 1.1923505008600628, + "grad_norm": 2.1285455226898193, + "learning_rate": 1.3716501566546334e-05, + "loss": 0.3091, + "step": 1473 + }, + { + "epoch": 1.1931599716685217, + "grad_norm": 2.248628854751587, + "learning_rate": 1.3708382781638809e-05, + "loss": 0.3023, + "step": 1474 + }, + { + "epoch": 1.1939694424769807, + "grad_norm": 2.242053985595703, + "learning_rate": 1.3700261161621767e-05, + "loss": 0.3111, + "step": 1475 + }, + { + "epoch": 1.1947789132854396, + "grad_norm": 2.172844171524048, + "learning_rate": 1.3692136712704305e-05, + "loss": 0.2924, + "step": 1476 + }, + { + "epoch": 1.1955883840938986, + "grad_norm": 2.3949623107910156, + "learning_rate": 1.368400944109767e-05, + "loss": 0.3209, + "step": 1477 + }, + { + "epoch": 1.1963978549023575, + "grad_norm": 2.318692207336426, + "learning_rate": 1.3675879353015275e-05, + "loss": 0.3151, + "step": 1478 + }, + { + "epoch": 1.1972073257108167, + "grad_norm": 2.34869384765625, + "learning_rate": 1.3667746454672685e-05, + "loss": 0.3136, + "step": 1479 + }, + { + "epoch": 1.1980167965192756, + "grad_norm": 2.337177276611328, + "learning_rate": 1.3659610752287614e-05, + "loss": 0.3089, + "step": 1480 + }, + { + "epoch": 1.1988262673277346, + "grad_norm": 2.272796154022217, + "learning_rate": 1.3651472252079912e-05, + "loss": 0.3108, + "step": 1481 + }, + { + "epoch": 1.1996357381361935, + "grad_norm": 1.9725879430770874, + "learning_rate": 1.3643330960271577e-05, + "loss": 0.2708, + "step": 1482 + }, + { + "epoch": 1.2004452089446525, + "grad_norm": 2.1585986614227295, + "learning_rate": 1.3635186883086742e-05, + "loss": 0.3179, + "step": 1483 + }, + { + "epoch": 1.2012546797531114, + "grad_norm": 2.196004867553711, + "learning_rate": 1.3627040026751658e-05, + "loss": 0.3065, + "step": 1484 + }, + { + "epoch": 1.2020641505615703, + "grad_norm": 2.287330389022827, + "learning_rate": 1.3618890397494719e-05, + "loss": 0.3229, + "step": 1485 + }, + { + "epoch": 1.2028736213700293, + "grad_norm": 2.1885428428649902, + "learning_rate": 1.3610738001546417e-05, + "loss": 0.3176, + "step": 1486 + }, + { + "epoch": 1.2036830921784882, + "grad_norm": 2.068270683288574, + "learning_rate": 1.360258284513938e-05, + "loss": 0.3092, + "step": 1487 + }, + { + "epoch": 1.2044925629869474, + "grad_norm": 2.039956569671631, + "learning_rate": 1.359442493450833e-05, + "loss": 0.2914, + "step": 1488 + }, + { + "epoch": 1.2053020337954063, + "grad_norm": 2.233844518661499, + "learning_rate": 1.3586264275890107e-05, + "loss": 0.31, + "step": 1489 + }, + { + "epoch": 1.2061115046038653, + "grad_norm": 2.0161311626434326, + "learning_rate": 1.357810087552364e-05, + "loss": 0.2842, + "step": 1490 + }, + { + "epoch": 1.2069209754123242, + "grad_norm": 2.2422313690185547, + "learning_rate": 1.3569934739649968e-05, + "loss": 0.2986, + "step": 1491 + }, + { + "epoch": 1.2077304462207832, + "grad_norm": 2.248048782348633, + "learning_rate": 1.3561765874512208e-05, + "loss": 0.3094, + "step": 1492 + }, + { + "epoch": 1.2085399170292421, + "grad_norm": 2.3673503398895264, + "learning_rate": 1.3553594286355575e-05, + "loss": 0.3251, + "step": 1493 + }, + { + "epoch": 1.209349387837701, + "grad_norm": 2.0512101650238037, + "learning_rate": 1.3545419981427355e-05, + "loss": 0.278, + "step": 1494 + }, + { + "epoch": 1.21015885864616, + "grad_norm": 2.2214488983154297, + "learning_rate": 1.3537242965976918e-05, + "loss": 0.303, + "step": 1495 + }, + { + "epoch": 1.210968329454619, + "grad_norm": 2.1482245922088623, + "learning_rate": 1.3529063246255704e-05, + "loss": 0.2859, + "step": 1496 + }, + { + "epoch": 1.2117778002630781, + "grad_norm": 2.1298515796661377, + "learning_rate": 1.3520880828517224e-05, + "loss": 0.2835, + "step": 1497 + }, + { + "epoch": 1.212587271071537, + "grad_norm": 2.34360671043396, + "learning_rate": 1.3512695719017046e-05, + "loss": 0.321, + "step": 1498 + }, + { + "epoch": 1.213396741879996, + "grad_norm": 2.2806777954101562, + "learning_rate": 1.3504507924012802e-05, + "loss": 0.3068, + "step": 1499 + }, + { + "epoch": 1.214206212688455, + "grad_norm": 2.1825296878814697, + "learning_rate": 1.3496317449764168e-05, + "loss": 0.3139, + "step": 1500 + }, + { + "epoch": 1.215015683496914, + "grad_norm": 2.34944486618042, + "learning_rate": 1.3488124302532875e-05, + "loss": 0.3281, + "step": 1501 + }, + { + "epoch": 1.2158251543053729, + "grad_norm": 2.402270793914795, + "learning_rate": 1.3479928488582696e-05, + "loss": 0.3358, + "step": 1502 + }, + { + "epoch": 1.2166346251138318, + "grad_norm": 2.0941572189331055, + "learning_rate": 1.3471730014179449e-05, + "loss": 0.275, + "step": 1503 + }, + { + "epoch": 1.2174440959222907, + "grad_norm": 2.1930835247039795, + "learning_rate": 1.3463528885590975e-05, + "loss": 0.3093, + "step": 1504 + }, + { + "epoch": 1.2182535667307497, + "grad_norm": 2.341721534729004, + "learning_rate": 1.3455325109087154e-05, + "loss": 0.3277, + "step": 1505 + }, + { + "epoch": 1.2190630375392089, + "grad_norm": 2.2872767448425293, + "learning_rate": 1.3447118690939876e-05, + "loss": 0.3032, + "step": 1506 + }, + { + "epoch": 1.2198725083476678, + "grad_norm": 2.2081079483032227, + "learning_rate": 1.3438909637423071e-05, + "loss": 0.312, + "step": 1507 + }, + { + "epoch": 1.2206819791561268, + "grad_norm": 2.5270915031433105, + "learning_rate": 1.3430697954812669e-05, + "loss": 0.3105, + "step": 1508 + }, + { + "epoch": 1.2214914499645857, + "grad_norm": 2.332853078842163, + "learning_rate": 1.342248364938661e-05, + "loss": 0.2964, + "step": 1509 + }, + { + "epoch": 1.2223009207730446, + "grad_norm": 2.366656541824341, + "learning_rate": 1.341426672742485e-05, + "loss": 0.3103, + "step": 1510 + }, + { + "epoch": 1.2231103915815036, + "grad_norm": 2.202455997467041, + "learning_rate": 1.3406047195209336e-05, + "loss": 0.3312, + "step": 1511 + }, + { + "epoch": 1.2239198623899625, + "grad_norm": 2.1658222675323486, + "learning_rate": 1.3397825059024008e-05, + "loss": 0.3101, + "step": 1512 + }, + { + "epoch": 1.2247293331984215, + "grad_norm": 2.303131341934204, + "learning_rate": 1.3389600325154813e-05, + "loss": 0.299, + "step": 1513 + }, + { + "epoch": 1.2255388040068804, + "grad_norm": 2.3384716510772705, + "learning_rate": 1.3381372999889662e-05, + "loss": 0.3122, + "step": 1514 + }, + { + "epoch": 1.2263482748153396, + "grad_norm": 2.1086459159851074, + "learning_rate": 1.337314308951846e-05, + "loss": 0.2933, + "step": 1515 + }, + { + "epoch": 1.2271577456237985, + "grad_norm": 2.246309280395508, + "learning_rate": 1.3364910600333094e-05, + "loss": 0.2964, + "step": 1516 + }, + { + "epoch": 1.2279672164322575, + "grad_norm": 2.198324680328369, + "learning_rate": 1.3356675538627406e-05, + "loss": 0.2781, + "step": 1517 + }, + { + "epoch": 1.2287766872407164, + "grad_norm": 2.3406851291656494, + "learning_rate": 1.3348437910697221e-05, + "loss": 0.321, + "step": 1518 + }, + { + "epoch": 1.2295861580491754, + "grad_norm": 2.1824891567230225, + "learning_rate": 1.3340197722840314e-05, + "loss": 0.3038, + "step": 1519 + }, + { + "epoch": 1.2303956288576343, + "grad_norm": 2.2134945392608643, + "learning_rate": 1.333195498135642e-05, + "loss": 0.2819, + "step": 1520 + }, + { + "epoch": 1.2312050996660933, + "grad_norm": 2.2317569255828857, + "learning_rate": 1.3323709692547233e-05, + "loss": 0.3345, + "step": 1521 + }, + { + "epoch": 1.2320145704745522, + "grad_norm": 2.301758050918579, + "learning_rate": 1.3315461862716385e-05, + "loss": 0.3239, + "step": 1522 + }, + { + "epoch": 1.2328240412830112, + "grad_norm": 2.121061086654663, + "learning_rate": 1.3307211498169459e-05, + "loss": 0.286, + "step": 1523 + }, + { + "epoch": 1.2336335120914703, + "grad_norm": 2.272188425064087, + "learning_rate": 1.329895860521397e-05, + "loss": 0.3337, + "step": 1524 + }, + { + "epoch": 1.2344429828999293, + "grad_norm": 1.9752271175384521, + "learning_rate": 1.3290703190159367e-05, + "loss": 0.3002, + "step": 1525 + }, + { + "epoch": 1.2352524537083882, + "grad_norm": 2.0006113052368164, + "learning_rate": 1.3282445259317028e-05, + "loss": 0.2832, + "step": 1526 + }, + { + "epoch": 1.2360619245168472, + "grad_norm": 2.16455340385437, + "learning_rate": 1.3274184819000258e-05, + "loss": 0.2915, + "step": 1527 + }, + { + "epoch": 1.236871395325306, + "grad_norm": 2.183262586593628, + "learning_rate": 1.326592187552427e-05, + "loss": 0.3049, + "step": 1528 + }, + { + "epoch": 1.237680866133765, + "grad_norm": 2.236722707748413, + "learning_rate": 1.3257656435206204e-05, + "loss": 0.3296, + "step": 1529 + }, + { + "epoch": 1.238490336942224, + "grad_norm": 2.229301929473877, + "learning_rate": 1.3249388504365102e-05, + "loss": 0.3233, + "step": 1530 + }, + { + "epoch": 1.239299807750683, + "grad_norm": 2.184511423110962, + "learning_rate": 1.3241118089321905e-05, + "loss": 0.3121, + "step": 1531 + }, + { + "epoch": 1.2401092785591419, + "grad_norm": 2.14225172996521, + "learning_rate": 1.3232845196399466e-05, + "loss": 0.2943, + "step": 1532 + }, + { + "epoch": 1.240918749367601, + "grad_norm": 2.009190320968628, + "learning_rate": 1.322456983192252e-05, + "loss": 0.2729, + "step": 1533 + }, + { + "epoch": 1.24172822017606, + "grad_norm": 2.071439504623413, + "learning_rate": 1.32162920022177e-05, + "loss": 0.291, + "step": 1534 + }, + { + "epoch": 1.242537690984519, + "grad_norm": 2.108459949493408, + "learning_rate": 1.3208011713613518e-05, + "loss": 0.314, + "step": 1535 + }, + { + "epoch": 1.2433471617929779, + "grad_norm": 2.1672096252441406, + "learning_rate": 1.3199728972440373e-05, + "loss": 0.3323, + "step": 1536 + }, + { + "epoch": 1.2441566326014368, + "grad_norm": 2.3528082370758057, + "learning_rate": 1.3191443785030527e-05, + "loss": 0.3143, + "step": 1537 + }, + { + "epoch": 1.2449661034098958, + "grad_norm": 2.0628294944763184, + "learning_rate": 1.3183156157718122e-05, + "loss": 0.2967, + "step": 1538 + }, + { + "epoch": 1.2457755742183547, + "grad_norm": 2.3161022663116455, + "learning_rate": 1.3174866096839165e-05, + "loss": 0.3026, + "step": 1539 + }, + { + "epoch": 1.2465850450268137, + "grad_norm": 2.1901862621307373, + "learning_rate": 1.3166573608731518e-05, + "loss": 0.3276, + "step": 1540 + }, + { + "epoch": 1.2473945158352726, + "grad_norm": 2.137298583984375, + "learning_rate": 1.3158278699734906e-05, + "loss": 0.3042, + "step": 1541 + }, + { + "epoch": 1.2482039866437318, + "grad_norm": 2.2842936515808105, + "learning_rate": 1.3149981376190899e-05, + "loss": 0.3298, + "step": 1542 + }, + { + "epoch": 1.2490134574521907, + "grad_norm": 2.3114664554595947, + "learning_rate": 1.3141681644442912e-05, + "loss": 0.3216, + "step": 1543 + }, + { + "epoch": 1.2498229282606497, + "grad_norm": 2.236826181411743, + "learning_rate": 1.3133379510836209e-05, + "loss": 0.3154, + "step": 1544 + }, + { + "epoch": 1.2506323990691086, + "grad_norm": 2.236846685409546, + "learning_rate": 1.3125074981717879e-05, + "loss": 0.322, + "step": 1545 + }, + { + "epoch": 1.2514418698775676, + "grad_norm": 2.1719350814819336, + "learning_rate": 1.311676806343685e-05, + "loss": 0.2892, + "step": 1546 + }, + { + "epoch": 1.2522513406860265, + "grad_norm": 2.153785467147827, + "learning_rate": 1.3108458762343881e-05, + "loss": 0.3119, + "step": 1547 + }, + { + "epoch": 1.2530608114944854, + "grad_norm": 2.3871212005615234, + "learning_rate": 1.3100147084791544e-05, + "loss": 0.3244, + "step": 1548 + }, + { + "epoch": 1.2538702823029444, + "grad_norm": 2.234424114227295, + "learning_rate": 1.309183303713422e-05, + "loss": 0.2985, + "step": 1549 + }, + { + "epoch": 1.2546797531114033, + "grad_norm": 2.3523244857788086, + "learning_rate": 1.308351662572813e-05, + "loss": 0.3013, + "step": 1550 + }, + { + "epoch": 1.2554892239198625, + "grad_norm": 2.339329242706299, + "learning_rate": 1.3075197856931273e-05, + "loss": 0.2997, + "step": 1551 + }, + { + "epoch": 1.2562986947283212, + "grad_norm": 2.266200542449951, + "learning_rate": 1.3066876737103463e-05, + "loss": 0.3093, + "step": 1552 + }, + { + "epoch": 1.2571081655367804, + "grad_norm": 2.2158608436584473, + "learning_rate": 1.3058553272606314e-05, + "loss": 0.3145, + "step": 1553 + }, + { + "epoch": 1.2579176363452393, + "grad_norm": 2.1328935623168945, + "learning_rate": 1.3050227469803222e-05, + "loss": 0.286, + "step": 1554 + }, + { + "epoch": 1.2587271071536983, + "grad_norm": 1.9953495264053345, + "learning_rate": 1.3041899335059387e-05, + "loss": 0.2713, + "step": 1555 + }, + { + "epoch": 1.2595365779621572, + "grad_norm": 2.253742218017578, + "learning_rate": 1.3033568874741778e-05, + "loss": 0.2986, + "step": 1556 + }, + { + "epoch": 1.2603460487706162, + "grad_norm": 2.4024033546447754, + "learning_rate": 1.3025236095219145e-05, + "loss": 0.3019, + "step": 1557 + }, + { + "epoch": 1.2611555195790751, + "grad_norm": 2.34830641746521, + "learning_rate": 1.3016901002862013e-05, + "loss": 0.2958, + "step": 1558 + }, + { + "epoch": 1.261964990387534, + "grad_norm": 2.2518084049224854, + "learning_rate": 1.3008563604042673e-05, + "loss": 0.3108, + "step": 1559 + }, + { + "epoch": 1.2627744611959932, + "grad_norm": 2.271491765975952, + "learning_rate": 1.3000223905135183e-05, + "loss": 0.2972, + "step": 1560 + }, + { + "epoch": 1.263583932004452, + "grad_norm": 3.403878927230835, + "learning_rate": 1.2991881912515357e-05, + "loss": 0.3045, + "step": 1561 + }, + { + "epoch": 1.2643934028129111, + "grad_norm": 2.2436161041259766, + "learning_rate": 1.298353763256076e-05, + "loss": 0.3197, + "step": 1562 + }, + { + "epoch": 1.26520287362137, + "grad_norm": 2.204932928085327, + "learning_rate": 1.297519107165071e-05, + "loss": 0.3034, + "step": 1563 + }, + { + "epoch": 1.266012344429829, + "grad_norm": 2.2177276611328125, + "learning_rate": 1.2966842236166273e-05, + "loss": 0.304, + "step": 1564 + }, + { + "epoch": 1.266821815238288, + "grad_norm": 2.1493470668792725, + "learning_rate": 1.295849113249024e-05, + "loss": 0.323, + "step": 1565 + }, + { + "epoch": 1.267631286046747, + "grad_norm": 2.4118430614471436, + "learning_rate": 1.295013776700715e-05, + "loss": 0.3383, + "step": 1566 + }, + { + "epoch": 1.2684407568552059, + "grad_norm": 2.071516990661621, + "learning_rate": 1.294178214610326e-05, + "loss": 0.2798, + "step": 1567 + }, + { + "epoch": 1.2692502276636648, + "grad_norm": 2.3298585414886475, + "learning_rate": 1.2933424276166564e-05, + "loss": 0.317, + "step": 1568 + }, + { + "epoch": 1.270059698472124, + "grad_norm": 2.2027149200439453, + "learning_rate": 1.292506416358676e-05, + "loss": 0.3108, + "step": 1569 + }, + { + "epoch": 1.2708691692805827, + "grad_norm": 2.2976880073547363, + "learning_rate": 1.2916701814755275e-05, + "loss": 0.303, + "step": 1570 + }, + { + "epoch": 1.2716786400890419, + "grad_norm": 2.127467632293701, + "learning_rate": 1.2908337236065237e-05, + "loss": 0.282, + "step": 1571 + }, + { + "epoch": 1.2724881108975008, + "grad_norm": 2.3236083984375, + "learning_rate": 1.2899970433911482e-05, + "loss": 0.2887, + "step": 1572 + }, + { + "epoch": 1.2732975817059597, + "grad_norm": 2.2045531272888184, + "learning_rate": 1.2891601414690544e-05, + "loss": 0.2897, + "step": 1573 + }, + { + "epoch": 1.2741070525144187, + "grad_norm": 2.3968234062194824, + "learning_rate": 1.2883230184800655e-05, + "loss": 0.3431, + "step": 1574 + }, + { + "epoch": 1.2749165233228776, + "grad_norm": 2.116654872894287, + "learning_rate": 1.2874856750641734e-05, + "loss": 0.3005, + "step": 1575 + }, + { + "epoch": 1.2757259941313366, + "grad_norm": 2.3160862922668457, + "learning_rate": 1.2866481118615385e-05, + "loss": 0.3074, + "step": 1576 + }, + { + "epoch": 1.2765354649397955, + "grad_norm": 2.2567412853240967, + "learning_rate": 1.2858103295124891e-05, + "loss": 0.3064, + "step": 1577 + }, + { + "epoch": 1.2773449357482547, + "grad_norm": 2.272589921951294, + "learning_rate": 1.2849723286575222e-05, + "loss": 0.2988, + "step": 1578 + }, + { + "epoch": 1.2781544065567134, + "grad_norm": 2.467829465866089, + "learning_rate": 1.2841341099373001e-05, + "loss": 0.3146, + "step": 1579 + }, + { + "epoch": 1.2789638773651726, + "grad_norm": 2.168470621109009, + "learning_rate": 1.2832956739926527e-05, + "loss": 0.3017, + "step": 1580 + }, + { + "epoch": 1.2797733481736315, + "grad_norm": 2.2351019382476807, + "learning_rate": 1.2824570214645764e-05, + "loss": 0.3061, + "step": 1581 + }, + { + "epoch": 1.2805828189820905, + "grad_norm": 2.2978219985961914, + "learning_rate": 1.281618152994232e-05, + "loss": 0.2994, + "step": 1582 + }, + { + "epoch": 1.2813922897905494, + "grad_norm": 2.300835132598877, + "learning_rate": 1.2807790692229455e-05, + "loss": 0.325, + "step": 1583 + }, + { + "epoch": 1.2822017605990084, + "grad_norm": 2.1345391273498535, + "learning_rate": 1.279939770792209e-05, + "loss": 0.3139, + "step": 1584 + }, + { + "epoch": 1.2830112314074673, + "grad_norm": 2.270106315612793, + "learning_rate": 1.279100258343677e-05, + "loss": 0.3033, + "step": 1585 + }, + { + "epoch": 1.2838207022159263, + "grad_norm": 2.246824026107788, + "learning_rate": 1.2782605325191682e-05, + "loss": 0.3049, + "step": 1586 + }, + { + "epoch": 1.2846301730243854, + "grad_norm": 2.3306939601898193, + "learning_rate": 1.277420593960665e-05, + "loss": 0.3237, + "step": 1587 + }, + { + "epoch": 1.2854396438328441, + "grad_norm": 2.2579445838928223, + "learning_rate": 1.2765804433103118e-05, + "loss": 0.2927, + "step": 1588 + }, + { + "epoch": 1.2862491146413033, + "grad_norm": 2.16991925239563, + "learning_rate": 1.2757400812104149e-05, + "loss": 0.2897, + "step": 1589 + }, + { + "epoch": 1.2870585854497623, + "grad_norm": 2.267364501953125, + "learning_rate": 1.274899508303443e-05, + "loss": 0.2996, + "step": 1590 + }, + { + "epoch": 1.2878680562582212, + "grad_norm": 2.1422572135925293, + "learning_rate": 1.2740587252320255e-05, + "loss": 0.306, + "step": 1591 + }, + { + "epoch": 1.2886775270666802, + "grad_norm": 2.191147804260254, + "learning_rate": 1.2732177326389528e-05, + "loss": 0.3112, + "step": 1592 + }, + { + "epoch": 1.289486997875139, + "grad_norm": 2.1869122982025146, + "learning_rate": 1.2723765311671752e-05, + "loss": 0.2856, + "step": 1593 + }, + { + "epoch": 1.290296468683598, + "grad_norm": 2.155820369720459, + "learning_rate": 1.2715351214598026e-05, + "loss": 0.3058, + "step": 1594 + }, + { + "epoch": 1.291105939492057, + "grad_norm": 2.146437883377075, + "learning_rate": 1.2706935041601043e-05, + "loss": 0.3226, + "step": 1595 + }, + { + "epoch": 1.2919154103005162, + "grad_norm": 2.282118797302246, + "learning_rate": 1.2698516799115083e-05, + "loss": 0.307, + "step": 1596 + }, + { + "epoch": 1.2927248811089749, + "grad_norm": 2.1759696006774902, + "learning_rate": 1.269009649357601e-05, + "loss": 0.3004, + "step": 1597 + }, + { + "epoch": 1.293534351917434, + "grad_norm": 2.3346357345581055, + "learning_rate": 1.268167413142126e-05, + "loss": 0.3171, + "step": 1598 + }, + { + "epoch": 1.294343822725893, + "grad_norm": 2.0977821350097656, + "learning_rate": 1.2673249719089849e-05, + "loss": 0.2809, + "step": 1599 + }, + { + "epoch": 1.295153293534352, + "grad_norm": 2.2012879848480225, + "learning_rate": 1.2664823263022345e-05, + "loss": 0.3187, + "step": 1600 + }, + { + "epoch": 1.2959627643428109, + "grad_norm": 1.8981571197509766, + "learning_rate": 1.2656394769660904e-05, + "loss": 0.2664, + "step": 1601 + }, + { + "epoch": 1.2967722351512698, + "grad_norm": 1.8491209745407104, + "learning_rate": 1.2647964245449216e-05, + "loss": 0.2565, + "step": 1602 + }, + { + "epoch": 1.2975817059597288, + "grad_norm": 2.1662180423736572, + "learning_rate": 1.2639531696832531e-05, + "loss": 0.3027, + "step": 1603 + }, + { + "epoch": 1.2983911767681877, + "grad_norm": 2.143266201019287, + "learning_rate": 1.2631097130257652e-05, + "loss": 0.2781, + "step": 1604 + }, + { + "epoch": 1.2992006475766469, + "grad_norm": 2.1209585666656494, + "learning_rate": 1.2622660552172926e-05, + "loss": 0.3024, + "step": 1605 + }, + { + "epoch": 1.3000101183851056, + "grad_norm": 2.2236521244049072, + "learning_rate": 1.2614221969028217e-05, + "loss": 0.3148, + "step": 1606 + }, + { + "epoch": 1.3008195891935648, + "grad_norm": 2.219003677368164, + "learning_rate": 1.2605781387274956e-05, + "loss": 0.3116, + "step": 1607 + }, + { + "epoch": 1.3016290600020237, + "grad_norm": 2.09478497505188, + "learning_rate": 1.259733881336607e-05, + "loss": 0.2867, + "step": 1608 + }, + { + "epoch": 1.3024385308104827, + "grad_norm": 2.263580322265625, + "learning_rate": 1.2588894253756027e-05, + "loss": 0.3131, + "step": 1609 + }, + { + "epoch": 1.3032480016189416, + "grad_norm": 2.1255271434783936, + "learning_rate": 1.2580447714900813e-05, + "loss": 0.291, + "step": 1610 + }, + { + "epoch": 1.3040574724274006, + "grad_norm": 2.08858060836792, + "learning_rate": 1.2571999203257917e-05, + "loss": 0.2982, + "step": 1611 + }, + { + "epoch": 1.3048669432358595, + "grad_norm": 2.134596109390259, + "learning_rate": 1.2563548725286343e-05, + "loss": 0.3096, + "step": 1612 + }, + { + "epoch": 1.3056764140443184, + "grad_norm": 2.358646869659424, + "learning_rate": 1.2555096287446599e-05, + "loss": 0.3295, + "step": 1613 + }, + { + "epoch": 1.3064858848527776, + "grad_norm": 2.1440682411193848, + "learning_rate": 1.2546641896200686e-05, + "loss": 0.3072, + "step": 1614 + }, + { + "epoch": 1.3072953556612363, + "grad_norm": 2.1152946949005127, + "learning_rate": 1.2538185558012111e-05, + "loss": 0.2931, + "step": 1615 + }, + { + "epoch": 1.3081048264696955, + "grad_norm": 2.097304344177246, + "learning_rate": 1.2529727279345852e-05, + "loss": 0.296, + "step": 1616 + }, + { + "epoch": 1.3089142972781544, + "grad_norm": 2.2314112186431885, + "learning_rate": 1.252126706666838e-05, + "loss": 0.32, + "step": 1617 + }, + { + "epoch": 1.3097237680866134, + "grad_norm": 2.193524122238159, + "learning_rate": 1.2512804926447649e-05, + "loss": 0.2659, + "step": 1618 + }, + { + "epoch": 1.3105332388950723, + "grad_norm": 2.2027130126953125, + "learning_rate": 1.2504340865153077e-05, + "loss": 0.3022, + "step": 1619 + }, + { + "epoch": 1.3113427097035313, + "grad_norm": 2.2154715061187744, + "learning_rate": 1.249587488925555e-05, + "loss": 0.298, + "step": 1620 + }, + { + "epoch": 1.3121521805119902, + "grad_norm": 2.2317724227905273, + "learning_rate": 1.2487407005227436e-05, + "loss": 0.2794, + "step": 1621 + }, + { + "epoch": 1.3129616513204492, + "grad_norm": 2.5904760360717773, + "learning_rate": 1.2478937219542537e-05, + "loss": 0.3246, + "step": 1622 + }, + { + "epoch": 1.3137711221289083, + "grad_norm": 2.063204050064087, + "learning_rate": 1.2470465538676125e-05, + "loss": 0.3075, + "step": 1623 + }, + { + "epoch": 1.314580592937367, + "grad_norm": 2.3084206581115723, + "learning_rate": 1.2461991969104914e-05, + "loss": 0.3211, + "step": 1624 + }, + { + "epoch": 1.3153900637458262, + "grad_norm": 2.2174882888793945, + "learning_rate": 1.2453516517307067e-05, + "loss": 0.3306, + "step": 1625 + }, + { + "epoch": 1.3161995345542852, + "grad_norm": 2.337237596511841, + "learning_rate": 1.2445039189762182e-05, + "loss": 0.3178, + "step": 1626 + }, + { + "epoch": 1.3170090053627441, + "grad_norm": 2.3498501777648926, + "learning_rate": 1.2436559992951296e-05, + "loss": 0.3102, + "step": 1627 + }, + { + "epoch": 1.317818476171203, + "grad_norm": 2.1572329998016357, + "learning_rate": 1.2428078933356862e-05, + "loss": 0.3006, + "step": 1628 + }, + { + "epoch": 1.318627946979662, + "grad_norm": 2.289576292037964, + "learning_rate": 1.2419596017462775e-05, + "loss": 0.3157, + "step": 1629 + }, + { + "epoch": 1.319437417788121, + "grad_norm": 2.268357753753662, + "learning_rate": 1.2411111251754334e-05, + "loss": 0.3093, + "step": 1630 + }, + { + "epoch": 1.32024688859658, + "grad_norm": 2.3139634132385254, + "learning_rate": 1.2402624642718264e-05, + "loss": 0.3216, + "step": 1631 + }, + { + "epoch": 1.321056359405039, + "grad_norm": 2.063363790512085, + "learning_rate": 1.2394136196842692e-05, + "loss": 0.2964, + "step": 1632 + }, + { + "epoch": 1.3218658302134978, + "grad_norm": 2.4171969890594482, + "learning_rate": 1.2385645920617151e-05, + "loss": 0.2941, + "step": 1633 + }, + { + "epoch": 1.322675301021957, + "grad_norm": 2.2151312828063965, + "learning_rate": 1.2377153820532575e-05, + "loss": 0.3008, + "step": 1634 + }, + { + "epoch": 1.323484771830416, + "grad_norm": 2.167680501937866, + "learning_rate": 1.236865990308129e-05, + "loss": 0.2847, + "step": 1635 + }, + { + "epoch": 1.3242942426388749, + "grad_norm": 2.328291893005371, + "learning_rate": 1.2360164174757014e-05, + "loss": 0.2837, + "step": 1636 + }, + { + "epoch": 1.3251037134473338, + "grad_norm": 2.407090187072754, + "learning_rate": 1.2351666642054844e-05, + "loss": 0.3253, + "step": 1637 + }, + { + "epoch": 1.3259131842557927, + "grad_norm": 2.277446985244751, + "learning_rate": 1.2343167311471267e-05, + "loss": 0.29, + "step": 1638 + }, + { + "epoch": 1.3267226550642517, + "grad_norm": 2.1066746711730957, + "learning_rate": 1.233466618950413e-05, + "loss": 0.2688, + "step": 1639 + }, + { + "epoch": 1.3275321258727106, + "grad_norm": 2.341109275817871, + "learning_rate": 1.2326163282652663e-05, + "loss": 0.3232, + "step": 1640 + }, + { + "epoch": 1.3283415966811698, + "grad_norm": 2.185593366622925, + "learning_rate": 1.231765859741745e-05, + "loss": 0.2793, + "step": 1641 + }, + { + "epoch": 1.3291510674896285, + "grad_norm": 2.2156968116760254, + "learning_rate": 1.2309152140300446e-05, + "loss": 0.2971, + "step": 1642 + }, + { + "epoch": 1.3299605382980877, + "grad_norm": 2.1514577865600586, + "learning_rate": 1.2300643917804949e-05, + "loss": 0.3001, + "step": 1643 + }, + { + "epoch": 1.3307700091065466, + "grad_norm": 2.076521635055542, + "learning_rate": 1.2292133936435615e-05, + "loss": 0.2761, + "step": 1644 + }, + { + "epoch": 1.3315794799150056, + "grad_norm": 2.1462433338165283, + "learning_rate": 1.2283622202698439e-05, + "loss": 0.2954, + "step": 1645 + }, + { + "epoch": 1.3323889507234645, + "grad_norm": 2.2681198120117188, + "learning_rate": 1.2275108723100759e-05, + "loss": 0.3257, + "step": 1646 + }, + { + "epoch": 1.3331984215319235, + "grad_norm": 2.34364652633667, + "learning_rate": 1.2266593504151247e-05, + "loss": 0.3161, + "step": 1647 + }, + { + "epoch": 1.3340078923403824, + "grad_norm": 2.057530403137207, + "learning_rate": 1.2258076552359905e-05, + "loss": 0.2812, + "step": 1648 + }, + { + "epoch": 1.3348173631488414, + "grad_norm": 2.3150227069854736, + "learning_rate": 1.2249557874238058e-05, + "loss": 0.3184, + "step": 1649 + }, + { + "epoch": 1.3356268339573005, + "grad_norm": 2.185727119445801, + "learning_rate": 1.2241037476298359e-05, + "loss": 0.2904, + "step": 1650 + }, + { + "epoch": 1.3364363047657593, + "grad_norm": 2.3936753273010254, + "learning_rate": 1.223251536505476e-05, + "loss": 0.3284, + "step": 1651 + }, + { + "epoch": 1.3372457755742184, + "grad_norm": 2.1365787982940674, + "learning_rate": 1.2223991547022536e-05, + "loss": 0.2962, + "step": 1652 + }, + { + "epoch": 1.3380552463826774, + "grad_norm": 2.1316652297973633, + "learning_rate": 1.2215466028718261e-05, + "loss": 0.2929, + "step": 1653 + }, + { + "epoch": 1.3388647171911363, + "grad_norm": 2.322967052459717, + "learning_rate": 1.2206938816659815e-05, + "loss": 0.3055, + "step": 1654 + }, + { + "epoch": 1.3396741879995953, + "grad_norm": 2.30020809173584, + "learning_rate": 1.2198409917366365e-05, + "loss": 0.2992, + "step": 1655 + }, + { + "epoch": 1.3404836588080542, + "grad_norm": 2.028245449066162, + "learning_rate": 1.2189879337358372e-05, + "loss": 0.2827, + "step": 1656 + }, + { + "epoch": 1.3412931296165131, + "grad_norm": 2.314232349395752, + "learning_rate": 1.2181347083157582e-05, + "loss": 0.3149, + "step": 1657 + }, + { + "epoch": 1.342102600424972, + "grad_norm": 2.1831841468811035, + "learning_rate": 1.217281316128702e-05, + "loss": 0.2883, + "step": 1658 + }, + { + "epoch": 1.3429120712334313, + "grad_norm": 2.0856051445007324, + "learning_rate": 1.2164277578270989e-05, + "loss": 0.2806, + "step": 1659 + }, + { + "epoch": 1.34372154204189, + "grad_norm": 2.2805399894714355, + "learning_rate": 1.2155740340635054e-05, + "loss": 0.3183, + "step": 1660 + }, + { + "epoch": 1.3445310128503491, + "grad_norm": 2.0656015872955322, + "learning_rate": 1.2147201454906055e-05, + "loss": 0.2734, + "step": 1661 + }, + { + "epoch": 1.345340483658808, + "grad_norm": 2.2178189754486084, + "learning_rate": 1.2138660927612084e-05, + "loss": 0.312, + "step": 1662 + }, + { + "epoch": 1.346149954467267, + "grad_norm": 2.2001097202301025, + "learning_rate": 1.2130118765282493e-05, + "loss": 0.2845, + "step": 1663 + }, + { + "epoch": 1.346959425275726, + "grad_norm": 2.2469375133514404, + "learning_rate": 1.2121574974447885e-05, + "loss": 0.3068, + "step": 1664 + }, + { + "epoch": 1.347768896084185, + "grad_norm": 2.287724018096924, + "learning_rate": 1.21130295616401e-05, + "loss": 0.3019, + "step": 1665 + }, + { + "epoch": 1.3485783668926439, + "grad_norm": 2.3241655826568604, + "learning_rate": 1.2104482533392225e-05, + "loss": 0.3361, + "step": 1666 + }, + { + "epoch": 1.3493878377011028, + "grad_norm": 2.2313406467437744, + "learning_rate": 1.2095933896238583e-05, + "loss": 0.3113, + "step": 1667 + }, + { + "epoch": 1.350197308509562, + "grad_norm": 2.2057199478149414, + "learning_rate": 1.2087383656714721e-05, + "loss": 0.3057, + "step": 1668 + }, + { + "epoch": 1.3510067793180207, + "grad_norm": 2.080547571182251, + "learning_rate": 1.2078831821357417e-05, + "loss": 0.2876, + "step": 1669 + }, + { + "epoch": 1.3518162501264799, + "grad_norm": 2.1973211765289307, + "learning_rate": 1.2070278396704667e-05, + "loss": 0.296, + "step": 1670 + }, + { + "epoch": 1.3526257209349388, + "grad_norm": 2.1871321201324463, + "learning_rate": 1.2061723389295676e-05, + "loss": 0.2798, + "step": 1671 + }, + { + "epoch": 1.3534351917433978, + "grad_norm": 2.0501973628997803, + "learning_rate": 1.205316680567087e-05, + "loss": 0.271, + "step": 1672 + }, + { + "epoch": 1.3542446625518567, + "grad_norm": 2.0726170539855957, + "learning_rate": 1.2044608652371874e-05, + "loss": 0.2738, + "step": 1673 + }, + { + "epoch": 1.3550541333603157, + "grad_norm": 2.020134687423706, + "learning_rate": 1.203604893594151e-05, + "loss": 0.2701, + "step": 1674 + }, + { + "epoch": 1.3558636041687746, + "grad_norm": 2.4557061195373535, + "learning_rate": 1.2027487662923805e-05, + "loss": 0.2918, + "step": 1675 + }, + { + "epoch": 1.3566730749772336, + "grad_norm": 2.2276992797851562, + "learning_rate": 1.2018924839863965e-05, + "loss": 0.3135, + "step": 1676 + }, + { + "epoch": 1.3574825457856927, + "grad_norm": 2.2597742080688477, + "learning_rate": 1.2010360473308384e-05, + "loss": 0.3078, + "step": 1677 + }, + { + "epoch": 1.3582920165941514, + "grad_norm": 2.243450880050659, + "learning_rate": 1.2001794569804643e-05, + "loss": 0.3209, + "step": 1678 + }, + { + "epoch": 1.3591014874026106, + "grad_norm": 2.073199510574341, + "learning_rate": 1.1993227135901491e-05, + "loss": 0.2869, + "step": 1679 + }, + { + "epoch": 1.3599109582110696, + "grad_norm": 2.3520519733428955, + "learning_rate": 1.1984658178148844e-05, + "loss": 0.3093, + "step": 1680 + }, + { + "epoch": 1.3607204290195285, + "grad_norm": 2.25093150138855, + "learning_rate": 1.1976087703097794e-05, + "loss": 0.2851, + "step": 1681 + }, + { + "epoch": 1.3615298998279874, + "grad_norm": 2.218745231628418, + "learning_rate": 1.1967515717300584e-05, + "loss": 0.3071, + "step": 1682 + }, + { + "epoch": 1.3623393706364464, + "grad_norm": 2.115436553955078, + "learning_rate": 1.1958942227310615e-05, + "loss": 0.2899, + "step": 1683 + }, + { + "epoch": 1.3631488414449053, + "grad_norm": 2.219733715057373, + "learning_rate": 1.1950367239682436e-05, + "loss": 0.3005, + "step": 1684 + }, + { + "epoch": 1.3639583122533643, + "grad_norm": 2.022203207015991, + "learning_rate": 1.1941790760971748e-05, + "loss": 0.2959, + "step": 1685 + }, + { + "epoch": 1.3647677830618234, + "grad_norm": 2.676274538040161, + "learning_rate": 1.1933212797735381e-05, + "loss": 0.3081, + "step": 1686 + }, + { + "epoch": 1.3655772538702822, + "grad_norm": 2.1343793869018555, + "learning_rate": 1.192463335653131e-05, + "loss": 0.3, + "step": 1687 + }, + { + "epoch": 1.3663867246787413, + "grad_norm": 2.1723380088806152, + "learning_rate": 1.191605244391863e-05, + "loss": 0.3026, + "step": 1688 + }, + { + "epoch": 1.3671961954872003, + "grad_norm": 2.2606654167175293, + "learning_rate": 1.1907470066457572e-05, + "loss": 0.3254, + "step": 1689 + }, + { + "epoch": 1.3680056662956592, + "grad_norm": 2.3029162883758545, + "learning_rate": 1.1898886230709476e-05, + "loss": 0.2903, + "step": 1690 + }, + { + "epoch": 1.3688151371041182, + "grad_norm": 2.168123960494995, + "learning_rate": 1.1890300943236808e-05, + "loss": 0.2921, + "step": 1691 + }, + { + "epoch": 1.3696246079125771, + "grad_norm": 2.078742027282715, + "learning_rate": 1.1881714210603135e-05, + "loss": 0.2725, + "step": 1692 + }, + { + "epoch": 1.370434078721036, + "grad_norm": 2.2479448318481445, + "learning_rate": 1.1873126039373133e-05, + "loss": 0.2941, + "step": 1693 + }, + { + "epoch": 1.371243549529495, + "grad_norm": 2.3248889446258545, + "learning_rate": 1.1864536436112573e-05, + "loss": 0.2983, + "step": 1694 + }, + { + "epoch": 1.3720530203379542, + "grad_norm": 2.1690187454223633, + "learning_rate": 1.185594540738833e-05, + "loss": 0.2887, + "step": 1695 + }, + { + "epoch": 1.372862491146413, + "grad_norm": 2.080679178237915, + "learning_rate": 1.1847352959768358e-05, + "loss": 0.299, + "step": 1696 + }, + { + "epoch": 1.373671961954872, + "grad_norm": 2.338538646697998, + "learning_rate": 1.1838759099821707e-05, + "loss": 0.3243, + "step": 1697 + }, + { + "epoch": 1.374481432763331, + "grad_norm": 2.081068277359009, + "learning_rate": 1.1830163834118496e-05, + "loss": 0.2991, + "step": 1698 + }, + { + "epoch": 1.37529090357179, + "grad_norm": 2.257132053375244, + "learning_rate": 1.1821567169229926e-05, + "loss": 0.3166, + "step": 1699 + }, + { + "epoch": 1.376100374380249, + "grad_norm": 2.110198497772217, + "learning_rate": 1.1812969111728263e-05, + "loss": 0.2826, + "step": 1700 + }, + { + "epoch": 1.3769098451887078, + "grad_norm": 2.2251241207122803, + "learning_rate": 1.1804369668186843e-05, + "loss": 0.322, + "step": 1701 + }, + { + "epoch": 1.3777193159971668, + "grad_norm": 2.1608479022979736, + "learning_rate": 1.1795768845180055e-05, + "loss": 0.3292, + "step": 1702 + }, + { + "epoch": 1.3785287868056257, + "grad_norm": 2.0784552097320557, + "learning_rate": 1.178716664928335e-05, + "loss": 0.2726, + "step": 1703 + }, + { + "epoch": 1.379338257614085, + "grad_norm": 2.141090154647827, + "learning_rate": 1.1778563087073223e-05, + "loss": 0.3096, + "step": 1704 + }, + { + "epoch": 1.3801477284225436, + "grad_norm": 2.235459804534912, + "learning_rate": 1.1769958165127214e-05, + "loss": 0.3225, + "step": 1705 + }, + { + "epoch": 1.3809571992310028, + "grad_norm": 2.237542152404785, + "learning_rate": 1.1761351890023904e-05, + "loss": 0.2788, + "step": 1706 + }, + { + "epoch": 1.3817666700394617, + "grad_norm": 2.131882429122925, + "learning_rate": 1.1752744268342913e-05, + "loss": 0.285, + "step": 1707 + }, + { + "epoch": 1.3825761408479207, + "grad_norm": 2.0875606536865234, + "learning_rate": 1.1744135306664876e-05, + "loss": 0.289, + "step": 1708 + }, + { + "epoch": 1.3833856116563796, + "grad_norm": 2.2048161029815674, + "learning_rate": 1.1735525011571476e-05, + "loss": 0.3103, + "step": 1709 + }, + { + "epoch": 1.3841950824648386, + "grad_norm": 2.0756278038024902, + "learning_rate": 1.1726913389645391e-05, + "loss": 0.2678, + "step": 1710 + }, + { + "epoch": 1.3850045532732975, + "grad_norm": 2.362520456314087, + "learning_rate": 1.1718300447470327e-05, + "loss": 0.3534, + "step": 1711 + }, + { + "epoch": 1.3858140240817565, + "grad_norm": 2.278099298477173, + "learning_rate": 1.1709686191630998e-05, + "loss": 0.3192, + "step": 1712 + }, + { + "epoch": 1.3866234948902156, + "grad_norm": 2.1943137645721436, + "learning_rate": 1.170107062871312e-05, + "loss": 0.3002, + "step": 1713 + }, + { + "epoch": 1.3874329656986744, + "grad_norm": 2.082613468170166, + "learning_rate": 1.1692453765303406e-05, + "loss": 0.2731, + "step": 1714 + }, + { + "epoch": 1.3882424365071335, + "grad_norm": 2.0744917392730713, + "learning_rate": 1.1683835607989576e-05, + "loss": 0.2838, + "step": 1715 + }, + { + "epoch": 1.3890519073155925, + "grad_norm": 2.302658796310425, + "learning_rate": 1.1675216163360316e-05, + "loss": 0.312, + "step": 1716 + }, + { + "epoch": 1.3898613781240514, + "grad_norm": 2.1336326599121094, + "learning_rate": 1.1666595438005317e-05, + "loss": 0.2919, + "step": 1717 + }, + { + "epoch": 1.3906708489325104, + "grad_norm": 2.1932566165924072, + "learning_rate": 1.1657973438515242e-05, + "loss": 0.303, + "step": 1718 + }, + { + "epoch": 1.3914803197409693, + "grad_norm": 2.0392885208129883, + "learning_rate": 1.1649350171481728e-05, + "loss": 0.2946, + "step": 1719 + }, + { + "epoch": 1.3922897905494283, + "grad_norm": 2.2038958072662354, + "learning_rate": 1.164072564349738e-05, + "loss": 0.3204, + "step": 1720 + }, + { + "epoch": 1.3930992613578872, + "grad_norm": 2.0998027324676514, + "learning_rate": 1.1632099861155767e-05, + "loss": 0.2854, + "step": 1721 + }, + { + "epoch": 1.3939087321663464, + "grad_norm": 2.212007761001587, + "learning_rate": 1.1623472831051419e-05, + "loss": 0.2851, + "step": 1722 + }, + { + "epoch": 1.394718202974805, + "grad_norm": 2.4322383403778076, + "learning_rate": 1.1614844559779819e-05, + "loss": 0.3071, + "step": 1723 + }, + { + "epoch": 1.3955276737832643, + "grad_norm": 2.466329336166382, + "learning_rate": 1.1606215053937398e-05, + "loss": 0.3051, + "step": 1724 + }, + { + "epoch": 1.3963371445917232, + "grad_norm": 2.128221273422241, + "learning_rate": 1.159758432012153e-05, + "loss": 0.2851, + "step": 1725 + }, + { + "epoch": 1.3971466154001821, + "grad_norm": 2.315929889678955, + "learning_rate": 1.1588952364930534e-05, + "loss": 0.2875, + "step": 1726 + }, + { + "epoch": 1.397956086208641, + "grad_norm": 2.180180788040161, + "learning_rate": 1.1580319194963656e-05, + "loss": 0.2868, + "step": 1727 + }, + { + "epoch": 1.3987655570171, + "grad_norm": 2.1455373764038086, + "learning_rate": 1.1571684816821072e-05, + "loss": 0.2969, + "step": 1728 + }, + { + "epoch": 1.399575027825559, + "grad_norm": 2.152639150619507, + "learning_rate": 1.1563049237103883e-05, + "loss": 0.2609, + "step": 1729 + }, + { + "epoch": 1.400384498634018, + "grad_norm": 2.167268991470337, + "learning_rate": 1.1554412462414109e-05, + "loss": 0.317, + "step": 1730 + }, + { + "epoch": 1.401193969442477, + "grad_norm": 2.261333465576172, + "learning_rate": 1.154577449935468e-05, + "loss": 0.284, + "step": 1731 + }, + { + "epoch": 1.4020034402509358, + "grad_norm": 2.2488715648651123, + "learning_rate": 1.153713535452944e-05, + "loss": 0.3023, + "step": 1732 + }, + { + "epoch": 1.402812911059395, + "grad_norm": 2.1157219409942627, + "learning_rate": 1.1528495034543133e-05, + "loss": 0.2943, + "step": 1733 + }, + { + "epoch": 1.403622381867854, + "grad_norm": 2.1364150047302246, + "learning_rate": 1.1519853546001403e-05, + "loss": 0.286, + "step": 1734 + }, + { + "epoch": 1.4044318526763129, + "grad_norm": 2.303741693496704, + "learning_rate": 1.1511210895510782e-05, + "loss": 0.2981, + "step": 1735 + }, + { + "epoch": 1.4052413234847718, + "grad_norm": 2.1988391876220703, + "learning_rate": 1.1502567089678704e-05, + "loss": 0.2927, + "step": 1736 + }, + { + "epoch": 1.4060507942932308, + "grad_norm": 2.3248050212860107, + "learning_rate": 1.1493922135113467e-05, + "loss": 0.3171, + "step": 1737 + }, + { + "epoch": 1.4068602651016897, + "grad_norm": 2.2236204147338867, + "learning_rate": 1.1485276038424267e-05, + "loss": 0.3132, + "step": 1738 + }, + { + "epoch": 1.4076697359101487, + "grad_norm": 2.310819149017334, + "learning_rate": 1.1476628806221154e-05, + "loss": 0.3004, + "step": 1739 + }, + { + "epoch": 1.4084792067186078, + "grad_norm": 2.24983549118042, + "learning_rate": 1.1467980445115063e-05, + "loss": 0.2949, + "step": 1740 + }, + { + "epoch": 1.4092886775270665, + "grad_norm": 2.4373857975006104, + "learning_rate": 1.1459330961717782e-05, + "loss": 0.2678, + "step": 1741 + }, + { + "epoch": 1.4100981483355257, + "grad_norm": 2.1245174407958984, + "learning_rate": 1.145068036264196e-05, + "loss": 0.2946, + "step": 1742 + }, + { + "epoch": 1.4109076191439847, + "grad_norm": 2.1035351753234863, + "learning_rate": 1.1442028654501096e-05, + "loss": 0.3127, + "step": 1743 + }, + { + "epoch": 1.4117170899524436, + "grad_norm": 2.4441516399383545, + "learning_rate": 1.1433375843909547e-05, + "loss": 0.3044, + "step": 1744 + }, + { + "epoch": 1.4125265607609025, + "grad_norm": 2.163444757461548, + "learning_rate": 1.1424721937482495e-05, + "loss": 0.2944, + "step": 1745 + }, + { + "epoch": 1.4133360315693615, + "grad_norm": 2.190870523452759, + "learning_rate": 1.1416066941835978e-05, + "loss": 0.271, + "step": 1746 + }, + { + "epoch": 1.4141455023778204, + "grad_norm": 2.209620475769043, + "learning_rate": 1.1407410863586854e-05, + "loss": 0.2945, + "step": 1747 + }, + { + "epoch": 1.4149549731862794, + "grad_norm": 2.1045875549316406, + "learning_rate": 1.1398753709352819e-05, + "loss": 0.2909, + "step": 1748 + }, + { + "epoch": 1.4157644439947386, + "grad_norm": 2.3982651233673096, + "learning_rate": 1.1390095485752381e-05, + "loss": 0.3124, + "step": 1749 + }, + { + "epoch": 1.4165739148031973, + "grad_norm": 2.187883138656616, + "learning_rate": 1.1381436199404876e-05, + "loss": 0.3085, + "step": 1750 + }, + { + "epoch": 1.4173833856116564, + "grad_norm": 2.139913320541382, + "learning_rate": 1.137277585693044e-05, + "loss": 0.3174, + "step": 1751 + }, + { + "epoch": 1.4181928564201154, + "grad_norm": 2.1135921478271484, + "learning_rate": 1.136411446495003e-05, + "loss": 0.2699, + "step": 1752 + }, + { + "epoch": 1.4190023272285743, + "grad_norm": 2.1186842918395996, + "learning_rate": 1.13554520300854e-05, + "loss": 0.3033, + "step": 1753 + }, + { + "epoch": 1.4198117980370333, + "grad_norm": 2.2323269844055176, + "learning_rate": 1.1346788558959092e-05, + "loss": 0.3167, + "step": 1754 + }, + { + "epoch": 1.4206212688454922, + "grad_norm": 2.2396891117095947, + "learning_rate": 1.1338124058194455e-05, + "loss": 0.3104, + "step": 1755 + }, + { + "epoch": 1.4214307396539512, + "grad_norm": 2.084141254425049, + "learning_rate": 1.132945853441562e-05, + "loss": 0.2825, + "step": 1756 + }, + { + "epoch": 1.4222402104624101, + "grad_norm": 2.1281750202178955, + "learning_rate": 1.132079199424749e-05, + "loss": 0.3071, + "step": 1757 + }, + { + "epoch": 1.4230496812708693, + "grad_norm": 2.293630599975586, + "learning_rate": 1.1312124444315766e-05, + "loss": 0.3053, + "step": 1758 + }, + { + "epoch": 1.423859152079328, + "grad_norm": 2.0334064960479736, + "learning_rate": 1.1303455891246901e-05, + "loss": 0.2833, + "step": 1759 + }, + { + "epoch": 1.4246686228877872, + "grad_norm": 2.219479560852051, + "learning_rate": 1.1294786341668125e-05, + "loss": 0.3008, + "step": 1760 + }, + { + "epoch": 1.4254780936962461, + "grad_norm": 2.274202346801758, + "learning_rate": 1.1286115802207426e-05, + "loss": 0.3076, + "step": 1761 + }, + { + "epoch": 1.426287564504705, + "grad_norm": 2.0195059776306152, + "learning_rate": 1.1277444279493553e-05, + "loss": 0.2668, + "step": 1762 + }, + { + "epoch": 1.427097035313164, + "grad_norm": 2.017672300338745, + "learning_rate": 1.1268771780156e-05, + "loss": 0.2835, + "step": 1763 + }, + { + "epoch": 1.427906506121623, + "grad_norm": 2.2773489952087402, + "learning_rate": 1.1260098310825019e-05, + "loss": 0.3238, + "step": 1764 + }, + { + "epoch": 1.428715976930082, + "grad_norm": 2.107102632522583, + "learning_rate": 1.1251423878131586e-05, + "loss": 0.2836, + "step": 1765 + }, + { + "epoch": 1.4295254477385408, + "grad_norm": 2.2105538845062256, + "learning_rate": 1.1242748488707436e-05, + "loss": 0.2772, + "step": 1766 + }, + { + "epoch": 1.430334918547, + "grad_norm": 2.02182674407959, + "learning_rate": 1.1234072149185011e-05, + "loss": 0.3041, + "step": 1767 + }, + { + "epoch": 1.4311443893554587, + "grad_norm": 2.2123377323150635, + "learning_rate": 1.12253948661975e-05, + "loss": 0.2856, + "step": 1768 + }, + { + "epoch": 1.431953860163918, + "grad_norm": 2.056273937225342, + "learning_rate": 1.12167166463788e-05, + "loss": 0.2966, + "step": 1769 + }, + { + "epoch": 1.4327633309723768, + "grad_norm": 2.193716049194336, + "learning_rate": 1.1208037496363534e-05, + "loss": 0.2919, + "step": 1770 + }, + { + "epoch": 1.4335728017808358, + "grad_norm": 2.1852705478668213, + "learning_rate": 1.1199357422787026e-05, + "loss": 0.3094, + "step": 1771 + }, + { + "epoch": 1.4343822725892947, + "grad_norm": 2.195061445236206, + "learning_rate": 1.1190676432285317e-05, + "loss": 0.3093, + "step": 1772 + }, + { + "epoch": 1.4351917433977537, + "grad_norm": 2.2795543670654297, + "learning_rate": 1.118199453149514e-05, + "loss": 0.3121, + "step": 1773 + }, + { + "epoch": 1.4360012142062126, + "grad_norm": 2.1138217449188232, + "learning_rate": 1.1173311727053924e-05, + "loss": 0.2857, + "step": 1774 + }, + { + "epoch": 1.4368106850146716, + "grad_norm": 2.1993322372436523, + "learning_rate": 1.1164628025599797e-05, + "loss": 0.34, + "step": 1775 + }, + { + "epoch": 1.4376201558231307, + "grad_norm": 2.159982442855835, + "learning_rate": 1.1155943433771568e-05, + "loss": 0.3055, + "step": 1776 + }, + { + "epoch": 1.4384296266315895, + "grad_norm": 2.107720136642456, + "learning_rate": 1.1147257958208723e-05, + "loss": 0.2813, + "step": 1777 + }, + { + "epoch": 1.4392390974400486, + "grad_norm": 2.0468804836273193, + "learning_rate": 1.1138571605551434e-05, + "loss": 0.2729, + "step": 1778 + }, + { + "epoch": 1.4400485682485076, + "grad_norm": 2.232010841369629, + "learning_rate": 1.1129884382440525e-05, + "loss": 0.3248, + "step": 1779 + }, + { + "epoch": 1.4408580390569665, + "grad_norm": 2.281235456466675, + "learning_rate": 1.1121196295517512e-05, + "loss": 0.3185, + "step": 1780 + }, + { + "epoch": 1.4416675098654255, + "grad_norm": 2.2508654594421387, + "learning_rate": 1.1112507351424544e-05, + "loss": 0.3003, + "step": 1781 + }, + { + "epoch": 1.4424769806738844, + "grad_norm": 2.397904396057129, + "learning_rate": 1.1103817556804443e-05, + "loss": 0.2967, + "step": 1782 + }, + { + "epoch": 1.4432864514823434, + "grad_norm": 2.0830023288726807, + "learning_rate": 1.1095126918300676e-05, + "loss": 0.2879, + "step": 1783 + }, + { + "epoch": 1.4440959222908023, + "grad_norm": 2.2062270641326904, + "learning_rate": 1.1086435442557358e-05, + "loss": 0.3076, + "step": 1784 + }, + { + "epoch": 1.4449053930992615, + "grad_norm": 2.251624345779419, + "learning_rate": 1.1077743136219237e-05, + "loss": 0.2951, + "step": 1785 + }, + { + "epoch": 1.4457148639077202, + "grad_norm": 2.0944013595581055, + "learning_rate": 1.1069050005931703e-05, + "loss": 0.3114, + "step": 1786 + }, + { + "epoch": 1.4465243347161794, + "grad_norm": 2.218242645263672, + "learning_rate": 1.1060356058340776e-05, + "loss": 0.2979, + "step": 1787 + }, + { + "epoch": 1.4473338055246383, + "grad_norm": 2.151261806488037, + "learning_rate": 1.1051661300093088e-05, + "loss": 0.3074, + "step": 1788 + }, + { + "epoch": 1.4481432763330973, + "grad_norm": 2.1252424716949463, + "learning_rate": 1.1042965737835913e-05, + "loss": 0.2825, + "step": 1789 + }, + { + "epoch": 1.4489527471415562, + "grad_norm": 2.140530824661255, + "learning_rate": 1.1034269378217122e-05, + "loss": 0.2801, + "step": 1790 + }, + { + "epoch": 1.4497622179500151, + "grad_norm": 2.2501673698425293, + "learning_rate": 1.1025572227885202e-05, + "loss": 0.3058, + "step": 1791 + }, + { + "epoch": 1.450571688758474, + "grad_norm": 2.143719434738159, + "learning_rate": 1.1016874293489244e-05, + "loss": 0.2859, + "step": 1792 + }, + { + "epoch": 1.451381159566933, + "grad_norm": 2.2959024906158447, + "learning_rate": 1.1008175581678937e-05, + "loss": 0.3142, + "step": 1793 + }, + { + "epoch": 1.4521906303753922, + "grad_norm": 2.384709119796753, + "learning_rate": 1.0999476099104564e-05, + "loss": 0.3057, + "step": 1794 + }, + { + "epoch": 1.453000101183851, + "grad_norm": 2.104250192642212, + "learning_rate": 1.0990775852417005e-05, + "loss": 0.285, + "step": 1795 + }, + { + "epoch": 1.45380957199231, + "grad_norm": 1.9732557535171509, + "learning_rate": 1.0982074848267711e-05, + "loss": 0.2743, + "step": 1796 + }, + { + "epoch": 1.454619042800769, + "grad_norm": 2.208747148513794, + "learning_rate": 1.0973373093308726e-05, + "loss": 0.2899, + "step": 1797 + }, + { + "epoch": 1.455428513609228, + "grad_norm": 2.2221765518188477, + "learning_rate": 1.0964670594192652e-05, + "loss": 0.2889, + "step": 1798 + }, + { + "epoch": 1.456237984417687, + "grad_norm": 2.269122362136841, + "learning_rate": 1.0955967357572676e-05, + "loss": 0.2971, + "step": 1799 + }, + { + "epoch": 1.4570474552261459, + "grad_norm": 2.221320390701294, + "learning_rate": 1.0947263390102543e-05, + "loss": 0.3156, + "step": 1800 + }, + { + "epoch": 1.4578569260346048, + "grad_norm": 2.182011127471924, + "learning_rate": 1.093855869843655e-05, + "loss": 0.3011, + "step": 1801 + }, + { + "epoch": 1.4586663968430638, + "grad_norm": 2.1817612648010254, + "learning_rate": 1.0929853289229552e-05, + "loss": 0.2909, + "step": 1802 + }, + { + "epoch": 1.459475867651523, + "grad_norm": 2.209073543548584, + "learning_rate": 1.0921147169136965e-05, + "loss": 0.3035, + "step": 1803 + }, + { + "epoch": 1.4602853384599817, + "grad_norm": 2.0652170181274414, + "learning_rate": 1.0912440344814726e-05, + "loss": 0.3018, + "step": 1804 + }, + { + "epoch": 1.4610948092684408, + "grad_norm": 2.121267557144165, + "learning_rate": 1.0903732822919325e-05, + "loss": 0.2867, + "step": 1805 + }, + { + "epoch": 1.4619042800768998, + "grad_norm": 2.0957894325256348, + "learning_rate": 1.089502461010778e-05, + "loss": 0.2688, + "step": 1806 + }, + { + "epoch": 1.4627137508853587, + "grad_norm": 4.109768390655518, + "learning_rate": 1.0886315713037648e-05, + "loss": 0.28, + "step": 1807 + }, + { + "epoch": 1.4635232216938177, + "grad_norm": 2.210744619369507, + "learning_rate": 1.0877606138366985e-05, + "loss": 0.2916, + "step": 1808 + }, + { + "epoch": 1.4643326925022766, + "grad_norm": 2.2269599437713623, + "learning_rate": 1.0868895892754397e-05, + "loss": 0.2937, + "step": 1809 + }, + { + "epoch": 1.4651421633107355, + "grad_norm": 2.1294398307800293, + "learning_rate": 1.0860184982858974e-05, + "loss": 0.2909, + "step": 1810 + }, + { + "epoch": 1.4659516341191945, + "grad_norm": 2.13653826713562, + "learning_rate": 1.085147341534033e-05, + "loss": 0.3183, + "step": 1811 + }, + { + "epoch": 1.4667611049276537, + "grad_norm": 2.1977784633636475, + "learning_rate": 1.084276119685858e-05, + "loss": 0.2967, + "step": 1812 + }, + { + "epoch": 1.4675705757361124, + "grad_norm": 2.1586081981658936, + "learning_rate": 1.0834048334074332e-05, + "loss": 0.2913, + "step": 1813 + }, + { + "epoch": 1.4683800465445715, + "grad_norm": 2.086836576461792, + "learning_rate": 1.082533483364869e-05, + "loss": 0.3045, + "step": 1814 + }, + { + "epoch": 1.4691895173530305, + "grad_norm": 2.1630196571350098, + "learning_rate": 1.0816620702243248e-05, + "loss": 0.2871, + "step": 1815 + }, + { + "epoch": 1.4699989881614894, + "grad_norm": 2.152423620223999, + "learning_rate": 1.0807905946520073e-05, + "loss": 0.2996, + "step": 1816 + }, + { + "epoch": 1.4708084589699484, + "grad_norm": 2.1143691539764404, + "learning_rate": 1.0799190573141716e-05, + "loss": 0.3178, + "step": 1817 + }, + { + "epoch": 1.4716179297784073, + "grad_norm": 2.195939064025879, + "learning_rate": 1.0790474588771202e-05, + "loss": 0.3366, + "step": 1818 + }, + { + "epoch": 1.4724274005868663, + "grad_norm": 2.190502166748047, + "learning_rate": 1.0781758000072022e-05, + "loss": 0.2934, + "step": 1819 + }, + { + "epoch": 1.4732368713953252, + "grad_norm": 2.3172359466552734, + "learning_rate": 1.0773040813708124e-05, + "loss": 0.3022, + "step": 1820 + }, + { + "epoch": 1.4740463422037844, + "grad_norm": 2.0385868549346924, + "learning_rate": 1.0764323036343921e-05, + "loss": 0.2826, + "step": 1821 + }, + { + "epoch": 1.474855813012243, + "grad_norm": 2.2652952671051025, + "learning_rate": 1.0755604674644267e-05, + "loss": 0.2811, + "step": 1822 + }, + { + "epoch": 1.4756652838207023, + "grad_norm": 2.086310386657715, + "learning_rate": 1.0746885735274478e-05, + "loss": 0.2851, + "step": 1823 + }, + { + "epoch": 1.4764747546291612, + "grad_norm": 2.147878646850586, + "learning_rate": 1.0738166224900292e-05, + "loss": 0.2858, + "step": 1824 + }, + { + "epoch": 1.4772842254376202, + "grad_norm": 2.063559055328369, + "learning_rate": 1.0729446150187903e-05, + "loss": 0.2734, + "step": 1825 + }, + { + "epoch": 1.4780936962460791, + "grad_norm": 2.3385508060455322, + "learning_rate": 1.0720725517803925e-05, + "loss": 0.3058, + "step": 1826 + }, + { + "epoch": 1.478903167054538, + "grad_norm": 2.3047540187835693, + "learning_rate": 1.07120043344154e-05, + "loss": 0.3158, + "step": 1827 + }, + { + "epoch": 1.479712637862997, + "grad_norm": 2.295437812805176, + "learning_rate": 1.0703282606689792e-05, + "loss": 0.3106, + "step": 1828 + }, + { + "epoch": 1.480522108671456, + "grad_norm": 2.3844125270843506, + "learning_rate": 1.0694560341294987e-05, + "loss": 0.2878, + "step": 1829 + }, + { + "epoch": 1.4813315794799151, + "grad_norm": 2.107119083404541, + "learning_rate": 1.0685837544899266e-05, + "loss": 0.3052, + "step": 1830 + }, + { + "epoch": 1.4821410502883738, + "grad_norm": 2.124030828475952, + "learning_rate": 1.0677114224171335e-05, + "loss": 0.2841, + "step": 1831 + }, + { + "epoch": 1.482950521096833, + "grad_norm": 2.2345383167266846, + "learning_rate": 1.0668390385780292e-05, + "loss": 0.3049, + "step": 1832 + }, + { + "epoch": 1.483759991905292, + "grad_norm": 1.948248267173767, + "learning_rate": 1.0659666036395626e-05, + "loss": 0.2546, + "step": 1833 + }, + { + "epoch": 1.484569462713751, + "grad_norm": 2.2135074138641357, + "learning_rate": 1.0650941182687226e-05, + "loss": 0.2877, + "step": 1834 + }, + { + "epoch": 1.4853789335222098, + "grad_norm": 2.0912673473358154, + "learning_rate": 1.0642215831325362e-05, + "loss": 0.2822, + "step": 1835 + }, + { + "epoch": 1.4861884043306688, + "grad_norm": 2.053443670272827, + "learning_rate": 1.0633489988980682e-05, + "loss": 0.2787, + "step": 1836 + }, + { + "epoch": 1.4869978751391277, + "grad_norm": 1.9678865671157837, + "learning_rate": 1.0624763662324217e-05, + "loss": 0.2988, + "step": 1837 + }, + { + "epoch": 1.4878073459475867, + "grad_norm": 2.245508909225464, + "learning_rate": 1.0616036858027364e-05, + "loss": 0.3064, + "step": 1838 + }, + { + "epoch": 1.4886168167560458, + "grad_norm": 2.3571505546569824, + "learning_rate": 1.0607309582761878e-05, + "loss": 0.3106, + "step": 1839 + }, + { + "epoch": 1.4894262875645046, + "grad_norm": 2.102570056915283, + "learning_rate": 1.0598581843199883e-05, + "loss": 0.291, + "step": 1840 + }, + { + "epoch": 1.4902357583729637, + "grad_norm": 2.200453281402588, + "learning_rate": 1.0589853646013856e-05, + "loss": 0.3086, + "step": 1841 + }, + { + "epoch": 1.4910452291814227, + "grad_norm": 2.0059173107147217, + "learning_rate": 1.0581124997876627e-05, + "loss": 0.2808, + "step": 1842 + }, + { + "epoch": 1.4918546999898816, + "grad_norm": 2.217362880706787, + "learning_rate": 1.0572395905461363e-05, + "loss": 0.2954, + "step": 1843 + }, + { + "epoch": 1.4926641707983406, + "grad_norm": 2.279816150665283, + "learning_rate": 1.056366637544158e-05, + "loss": 0.2912, + "step": 1844 + }, + { + "epoch": 1.4934736416067995, + "grad_norm": 1.9861000776290894, + "learning_rate": 1.0554936414491112e-05, + "loss": 0.2742, + "step": 1845 + }, + { + "epoch": 1.4942831124152585, + "grad_norm": 2.2216663360595703, + "learning_rate": 1.0546206029284149e-05, + "loss": 0.3073, + "step": 1846 + }, + { + "epoch": 1.4950925832237174, + "grad_norm": 2.282381772994995, + "learning_rate": 1.0537475226495179e-05, + "loss": 0.3126, + "step": 1847 + }, + { + "epoch": 1.4959020540321766, + "grad_norm": 2.226245880126953, + "learning_rate": 1.0528744012799025e-05, + "loss": 0.2826, + "step": 1848 + }, + { + "epoch": 1.4967115248406353, + "grad_norm": 2.143547773361206, + "learning_rate": 1.0520012394870823e-05, + "loss": 0.2921, + "step": 1849 + }, + { + "epoch": 1.4975209956490945, + "grad_norm": 2.0648608207702637, + "learning_rate": 1.0511280379386009e-05, + "loss": 0.2715, + "step": 1850 + }, + { + "epoch": 1.4983304664575534, + "grad_norm": 2.07137393951416, + "learning_rate": 1.0502547973020328e-05, + "loss": 0.2937, + "step": 1851 + }, + { + "epoch": 1.4991399372660124, + "grad_norm": 2.2993004322052, + "learning_rate": 1.0493815182449828e-05, + "loss": 0.2763, + "step": 1852 + }, + { + "epoch": 1.4999494080744713, + "grad_norm": 2.130993366241455, + "learning_rate": 1.0485082014350847e-05, + "loss": 0.2715, + "step": 1853 + }, + { + "epoch": 1.5007588788829302, + "grad_norm": 2.243314027786255, + "learning_rate": 1.0476348475400011e-05, + "loss": 0.2958, + "step": 1854 + }, + { + "epoch": 1.5015683496913894, + "grad_norm": 2.102174758911133, + "learning_rate": 1.0467614572274229e-05, + "loss": 0.2886, + "step": 1855 + }, + { + "epoch": 1.5023778204998481, + "grad_norm": 2.0660698413848877, + "learning_rate": 1.0458880311650692e-05, + "loss": 0.2905, + "step": 1856 + }, + { + "epoch": 1.5031872913083073, + "grad_norm": 2.0928986072540283, + "learning_rate": 1.0450145700206862e-05, + "loss": 0.2946, + "step": 1857 + }, + { + "epoch": 1.503996762116766, + "grad_norm": 2.1515510082244873, + "learning_rate": 1.044141074462047e-05, + "loss": 0.2994, + "step": 1858 + }, + { + "epoch": 1.5048062329252252, + "grad_norm": 2.163512706756592, + "learning_rate": 1.0432675451569503e-05, + "loss": 0.2945, + "step": 1859 + }, + { + "epoch": 1.5056157037336841, + "grad_norm": 2.175233840942383, + "learning_rate": 1.0423939827732221e-05, + "loss": 0.3073, + "step": 1860 + }, + { + "epoch": 1.506425174542143, + "grad_norm": 2.29620361328125, + "learning_rate": 1.0415203879787123e-05, + "loss": 0.2759, + "step": 1861 + }, + { + "epoch": 1.507234645350602, + "grad_norm": 2.409715175628662, + "learning_rate": 1.0406467614412966e-05, + "loss": 0.3159, + "step": 1862 + }, + { + "epoch": 1.508044116159061, + "grad_norm": 2.144482374191284, + "learning_rate": 1.0397731038288737e-05, + "loss": 0.3005, + "step": 1863 + }, + { + "epoch": 1.5088535869675201, + "grad_norm": 2.0775158405303955, + "learning_rate": 1.0388994158093675e-05, + "loss": 0.2765, + "step": 1864 + }, + { + "epoch": 1.5096630577759789, + "grad_norm": 2.1981968879699707, + "learning_rate": 1.038025698050724e-05, + "loss": 0.3085, + "step": 1865 + }, + { + "epoch": 1.510472528584438, + "grad_norm": 2.2332935333251953, + "learning_rate": 1.037151951220913e-05, + "loss": 0.3102, + "step": 1866 + }, + { + "epoch": 1.5112819993928968, + "grad_norm": 2.318138599395752, + "learning_rate": 1.0362781759879252e-05, + "loss": 0.2938, + "step": 1867 + }, + { + "epoch": 1.512091470201356, + "grad_norm": 2.382509708404541, + "learning_rate": 1.0354043730197744e-05, + "loss": 0.3118, + "step": 1868 + }, + { + "epoch": 1.5129009410098149, + "grad_norm": 2.1262760162353516, + "learning_rate": 1.0345305429844946e-05, + "loss": 0.3264, + "step": 1869 + }, + { + "epoch": 1.5137104118182738, + "grad_norm": 2.127385139465332, + "learning_rate": 1.033656686550141e-05, + "loss": 0.313, + "step": 1870 + }, + { + "epoch": 1.5145198826267328, + "grad_norm": 2.1820154190063477, + "learning_rate": 1.0327828043847887e-05, + "loss": 0.3195, + "step": 1871 + }, + { + "epoch": 1.5153293534351917, + "grad_norm": 2.2033090591430664, + "learning_rate": 1.031908897156533e-05, + "loss": 0.3196, + "step": 1872 + }, + { + "epoch": 1.5161388242436509, + "grad_norm": 2.1344101428985596, + "learning_rate": 1.0310349655334872e-05, + "loss": 0.2962, + "step": 1873 + }, + { + "epoch": 1.5169482950521096, + "grad_norm": 2.005427360534668, + "learning_rate": 1.0301610101837848e-05, + "loss": 0.3004, + "step": 1874 + }, + { + "epoch": 1.5177577658605688, + "grad_norm": 2.0217959880828857, + "learning_rate": 1.0292870317755763e-05, + "loss": 0.3147, + "step": 1875 + }, + { + "epoch": 1.5185672366690275, + "grad_norm": 2.0925583839416504, + "learning_rate": 1.0284130309770302e-05, + "loss": 0.2975, + "step": 1876 + }, + { + "epoch": 1.5193767074774867, + "grad_norm": 2.1170215606689453, + "learning_rate": 1.027539008456332e-05, + "loss": 0.2877, + "step": 1877 + }, + { + "epoch": 1.5201861782859456, + "grad_norm": 2.099351644515991, + "learning_rate": 1.0266649648816843e-05, + "loss": 0.2905, + "step": 1878 + }, + { + "epoch": 1.5209956490944045, + "grad_norm": 2.2729759216308594, + "learning_rate": 1.0257909009213045e-05, + "loss": 0.3011, + "step": 1879 + }, + { + "epoch": 1.5218051199028635, + "grad_norm": 2.283637046813965, + "learning_rate": 1.0249168172434276e-05, + "loss": 0.3208, + "step": 1880 + }, + { + "epoch": 1.5226145907113224, + "grad_norm": 2.2793595790863037, + "learning_rate": 1.0240427145163019e-05, + "loss": 0.3047, + "step": 1881 + }, + { + "epoch": 1.5234240615197816, + "grad_norm": 1.9943989515304565, + "learning_rate": 1.0231685934081911e-05, + "loss": 0.243, + "step": 1882 + }, + { + "epoch": 1.5242335323282403, + "grad_norm": 2.278045654296875, + "learning_rate": 1.0222944545873729e-05, + "loss": 0.3163, + "step": 1883 + }, + { + "epoch": 1.5250430031366995, + "grad_norm": 2.195065975189209, + "learning_rate": 1.0214202987221383e-05, + "loss": 0.2928, + "step": 1884 + }, + { + "epoch": 1.5258524739451582, + "grad_norm": 2.366610527038574, + "learning_rate": 1.0205461264807914e-05, + "loss": 0.3047, + "step": 1885 + }, + { + "epoch": 1.5266619447536174, + "grad_norm": 2.2794370651245117, + "learning_rate": 1.019671938531649e-05, + "loss": 0.3284, + "step": 1886 + }, + { + "epoch": 1.5274714155620763, + "grad_norm": 2.148815155029297, + "learning_rate": 1.0187977355430397e-05, + "loss": 0.2844, + "step": 1887 + }, + { + "epoch": 1.5282808863705353, + "grad_norm": 2.2019779682159424, + "learning_rate": 1.0179235181833034e-05, + "loss": 0.3156, + "step": 1888 + }, + { + "epoch": 1.5290903571789942, + "grad_norm": 2.0003409385681152, + "learning_rate": 1.017049287120792e-05, + "loss": 0.2994, + "step": 1889 + }, + { + "epoch": 1.5298998279874532, + "grad_norm": 2.0247139930725098, + "learning_rate": 1.0161750430238664e-05, + "loss": 0.2736, + "step": 1890 + }, + { + "epoch": 1.5307092987959123, + "grad_norm": 2.008197784423828, + "learning_rate": 1.0153007865608986e-05, + "loss": 0.2852, + "step": 1891 + }, + { + "epoch": 1.531518769604371, + "grad_norm": 2.12471604347229, + "learning_rate": 1.0144265184002695e-05, + "loss": 0.2886, + "step": 1892 + }, + { + "epoch": 1.5323282404128302, + "grad_norm": 2.3199222087860107, + "learning_rate": 1.0135522392103691e-05, + "loss": 0.3149, + "step": 1893 + }, + { + "epoch": 1.533137711221289, + "grad_norm": 2.277864694595337, + "learning_rate": 1.012677949659596e-05, + "loss": 0.2927, + "step": 1894 + }, + { + "epoch": 1.5339471820297481, + "grad_norm": 2.2358646392822266, + "learning_rate": 1.0118036504163567e-05, + "loss": 0.3056, + "step": 1895 + }, + { + "epoch": 1.534756652838207, + "grad_norm": 2.3235061168670654, + "learning_rate": 1.0109293421490641e-05, + "loss": 0.3071, + "step": 1896 + }, + { + "epoch": 1.535566123646666, + "grad_norm": 2.2979137897491455, + "learning_rate": 1.0100550255261401e-05, + "loss": 0.3039, + "step": 1897 + }, + { + "epoch": 1.536375594455125, + "grad_norm": 2.2862579822540283, + "learning_rate": 1.0091807012160106e-05, + "loss": 0.3022, + "step": 1898 + }, + { + "epoch": 1.537185065263584, + "grad_norm": 2.172452926635742, + "learning_rate": 1.0083063698871093e-05, + "loss": 0.2685, + "step": 1899 + }, + { + "epoch": 1.537994536072043, + "grad_norm": 2.165916919708252, + "learning_rate": 1.0074320322078743e-05, + "loss": 0.3003, + "step": 1900 + }, + { + "epoch": 1.5388040068805018, + "grad_norm": 2.189218759536743, + "learning_rate": 1.006557688846749e-05, + "loss": 0.3026, + "step": 1901 + }, + { + "epoch": 1.539613477688961, + "grad_norm": 2.504699468612671, + "learning_rate": 1.0056833404721802e-05, + "loss": 0.3364, + "step": 1902 + }, + { + "epoch": 1.5404229484974197, + "grad_norm": 2.1300907135009766, + "learning_rate": 1.0048089877526199e-05, + "loss": 0.3045, + "step": 1903 + }, + { + "epoch": 1.5412324193058788, + "grad_norm": 2.2250430583953857, + "learning_rate": 1.0039346313565224e-05, + "loss": 0.3064, + "step": 1904 + }, + { + "epoch": 1.5420418901143378, + "grad_norm": 2.3059027194976807, + "learning_rate": 1.003060271952345e-05, + "loss": 0.3004, + "step": 1905 + }, + { + "epoch": 1.5428513609227967, + "grad_norm": 2.601947069168091, + "learning_rate": 1.0021859102085479e-05, + "loss": 0.296, + "step": 1906 + }, + { + "epoch": 1.5436608317312557, + "grad_norm": 2.3519208431243896, + "learning_rate": 1.0013115467935925e-05, + "loss": 0.3037, + "step": 1907 + }, + { + "epoch": 1.5444703025397146, + "grad_norm": 2.0808191299438477, + "learning_rate": 1.0004371823759415e-05, + "loss": 0.2985, + "step": 1908 + }, + { + "epoch": 1.5452797733481738, + "grad_norm": 2.11346435546875, + "learning_rate": 9.995628176240588e-06, + "loss": 0.3085, + "step": 1909 + }, + { + "epoch": 1.5460892441566325, + "grad_norm": 2.0242879390716553, + "learning_rate": 9.986884532064078e-06, + "loss": 0.2851, + "step": 1910 + }, + { + "epoch": 1.5468987149650917, + "grad_norm": 1.9867854118347168, + "learning_rate": 9.978140897914521e-06, + "loss": 0.2798, + "step": 1911 + }, + { + "epoch": 1.5477081857735504, + "grad_norm": 2.1201162338256836, + "learning_rate": 9.969397280476553e-06, + "loss": 0.3061, + "step": 1912 + }, + { + "epoch": 1.5485176565820096, + "grad_norm": 2.1784965991973877, + "learning_rate": 9.96065368643478e-06, + "loss": 0.3099, + "step": 1913 + }, + { + "epoch": 1.5493271273904685, + "grad_norm": 2.1150128841400146, + "learning_rate": 9.951910122473806e-06, + "loss": 0.3165, + "step": 1914 + }, + { + "epoch": 1.5501365981989275, + "grad_norm": 2.1257009506225586, + "learning_rate": 9.9431665952782e-06, + "loss": 0.2932, + "step": 1915 + }, + { + "epoch": 1.5509460690073864, + "grad_norm": 2.0725865364074707, + "learning_rate": 9.934423111532513e-06, + "loss": 0.2762, + "step": 1916 + }, + { + "epoch": 1.5517555398158454, + "grad_norm": 2.0905420780181885, + "learning_rate": 9.925679677921259e-06, + "loss": 0.2654, + "step": 1917 + }, + { + "epoch": 1.5525650106243045, + "grad_norm": 2.157318592071533, + "learning_rate": 9.916936301128908e-06, + "loss": 0.2695, + "step": 1918 + }, + { + "epoch": 1.5533744814327632, + "grad_norm": 2.2466237545013428, + "learning_rate": 9.908192987839894e-06, + "loss": 0.2911, + "step": 1919 + }, + { + "epoch": 1.5541839522412224, + "grad_norm": 2.202291488647461, + "learning_rate": 9.899449744738602e-06, + "loss": 0.2815, + "step": 1920 + }, + { + "epoch": 1.5549934230496811, + "grad_norm": 2.1678457260131836, + "learning_rate": 9.890706578509362e-06, + "loss": 0.3081, + "step": 1921 + }, + { + "epoch": 1.5558028938581403, + "grad_norm": 2.1658737659454346, + "learning_rate": 9.88196349583644e-06, + "loss": 0.2933, + "step": 1922 + }, + { + "epoch": 1.5566123646665992, + "grad_norm": 2.210695266723633, + "learning_rate": 9.873220503404042e-06, + "loss": 0.3136, + "step": 1923 + }, + { + "epoch": 1.5574218354750582, + "grad_norm": 2.1016845703125, + "learning_rate": 9.864477607896312e-06, + "loss": 0.3075, + "step": 1924 + }, + { + "epoch": 1.5582313062835171, + "grad_norm": 2.295900344848633, + "learning_rate": 9.855734815997309e-06, + "loss": 0.3187, + "step": 1925 + }, + { + "epoch": 1.559040777091976, + "grad_norm": 2.131289482116699, + "learning_rate": 9.84699213439102e-06, + "loss": 0.3025, + "step": 1926 + }, + { + "epoch": 1.5598502479004352, + "grad_norm": 2.7916295528411865, + "learning_rate": 9.83824956976134e-06, + "loss": 0.3065, + "step": 1927 + }, + { + "epoch": 1.560659718708894, + "grad_norm": 2.168250799179077, + "learning_rate": 9.829507128792083e-06, + "loss": 0.305, + "step": 1928 + }, + { + "epoch": 1.5614691895173531, + "grad_norm": 2.0440051555633545, + "learning_rate": 9.82076481816697e-06, + "loss": 0.2685, + "step": 1929 + }, + { + "epoch": 1.5622786603258119, + "grad_norm": 2.0716490745544434, + "learning_rate": 9.812022644569608e-06, + "loss": 0.2937, + "step": 1930 + }, + { + "epoch": 1.563088131134271, + "grad_norm": 2.112592935562134, + "learning_rate": 9.803280614683512e-06, + "loss": 0.2936, + "step": 1931 + }, + { + "epoch": 1.56389760194273, + "grad_norm": 2.2660083770751953, + "learning_rate": 9.794538735192089e-06, + "loss": 0.286, + "step": 1932 + }, + { + "epoch": 1.564707072751189, + "grad_norm": 2.3266208171844482, + "learning_rate": 9.785797012778619e-06, + "loss": 0.31, + "step": 1933 + }, + { + "epoch": 1.5655165435596479, + "grad_norm": 2.0559678077697754, + "learning_rate": 9.777055454126271e-06, + "loss": 0.2753, + "step": 1934 + }, + { + "epoch": 1.5663260143681068, + "grad_norm": 2.131547451019287, + "learning_rate": 9.76831406591809e-06, + "loss": 0.2775, + "step": 1935 + }, + { + "epoch": 1.567135485176566, + "grad_norm": 2.0968339443206787, + "learning_rate": 9.759572854836983e-06, + "loss": 0.3124, + "step": 1936 + }, + { + "epoch": 1.5679449559850247, + "grad_norm": 2.2517435550689697, + "learning_rate": 9.75083182756573e-06, + "loss": 0.3058, + "step": 1937 + }, + { + "epoch": 1.5687544267934839, + "grad_norm": 2.204279899597168, + "learning_rate": 9.742090990786956e-06, + "loss": 0.2844, + "step": 1938 + }, + { + "epoch": 1.5695638976019426, + "grad_norm": 2.1707873344421387, + "learning_rate": 9.733350351183162e-06, + "loss": 0.2862, + "step": 1939 + }, + { + "epoch": 1.5703733684104018, + "grad_norm": 2.2440407276153564, + "learning_rate": 9.724609915436685e-06, + "loss": 0.3005, + "step": 1940 + }, + { + "epoch": 1.5711828392188607, + "grad_norm": 2.1505918502807617, + "learning_rate": 9.715869690229703e-06, + "loss": 0.2898, + "step": 1941 + }, + { + "epoch": 1.5719923100273197, + "grad_norm": 2.2827060222625732, + "learning_rate": 9.707129682244239e-06, + "loss": 0.2994, + "step": 1942 + }, + { + "epoch": 1.5728017808357786, + "grad_norm": 2.1930325031280518, + "learning_rate": 9.698389898162153e-06, + "loss": 0.3145, + "step": 1943 + }, + { + "epoch": 1.5736112516442375, + "grad_norm": 1.938471794128418, + "learning_rate": 9.689650344665131e-06, + "loss": 0.2731, + "step": 1944 + }, + { + "epoch": 1.5744207224526967, + "grad_norm": 2.230184555053711, + "learning_rate": 9.680911028434675e-06, + "loss": 0.285, + "step": 1945 + }, + { + "epoch": 1.5752301932611554, + "grad_norm": 2.1433019638061523, + "learning_rate": 9.672171956152113e-06, + "loss": 0.2766, + "step": 1946 + }, + { + "epoch": 1.5760396640696146, + "grad_norm": 2.269101619720459, + "learning_rate": 9.663433134498593e-06, + "loss": 0.3021, + "step": 1947 + }, + { + "epoch": 1.5768491348780733, + "grad_norm": 2.3493664264678955, + "learning_rate": 9.654694570155056e-06, + "loss": 0.3187, + "step": 1948 + }, + { + "epoch": 1.5776586056865325, + "grad_norm": 2.1690940856933594, + "learning_rate": 9.64595626980226e-06, + "loss": 0.2744, + "step": 1949 + }, + { + "epoch": 1.5784680764949914, + "grad_norm": 2.113142251968384, + "learning_rate": 9.63721824012075e-06, + "loss": 0.2952, + "step": 1950 + }, + { + "epoch": 1.5792775473034504, + "grad_norm": 2.1931698322296143, + "learning_rate": 9.628480487790873e-06, + "loss": 0.283, + "step": 1951 + }, + { + "epoch": 1.5800870181119093, + "grad_norm": 2.4252266883850098, + "learning_rate": 9.619743019492764e-06, + "loss": 0.2872, + "step": 1952 + }, + { + "epoch": 1.5808964889203683, + "grad_norm": 2.1185452938079834, + "learning_rate": 9.61100584190633e-06, + "loss": 0.3027, + "step": 1953 + }, + { + "epoch": 1.5817059597288274, + "grad_norm": 2.0954809188842773, + "learning_rate": 9.602268961711264e-06, + "loss": 0.264, + "step": 1954 + }, + { + "epoch": 1.5825154305372862, + "grad_norm": 2.2048146724700928, + "learning_rate": 9.59353238558704e-06, + "loss": 0.2956, + "step": 1955 + }, + { + "epoch": 1.5833249013457453, + "grad_norm": 2.184283494949341, + "learning_rate": 9.584796120212878e-06, + "loss": 0.2835, + "step": 1956 + }, + { + "epoch": 1.584134372154204, + "grad_norm": 2.1046695709228516, + "learning_rate": 9.576060172267784e-06, + "loss": 0.2728, + "step": 1957 + }, + { + "epoch": 1.5849438429626632, + "grad_norm": 2.682138204574585, + "learning_rate": 9.567324548430497e-06, + "loss": 0.3143, + "step": 1958 + }, + { + "epoch": 1.5857533137711222, + "grad_norm": 2.1448261737823486, + "learning_rate": 9.558589255379534e-06, + "loss": 0.293, + "step": 1959 + }, + { + "epoch": 1.586562784579581, + "grad_norm": 2.149736166000366, + "learning_rate": 9.549854299793141e-06, + "loss": 0.2831, + "step": 1960 + }, + { + "epoch": 1.58737225538804, + "grad_norm": 2.1954097747802734, + "learning_rate": 9.54111968834931e-06, + "loss": 0.2896, + "step": 1961 + }, + { + "epoch": 1.588181726196499, + "grad_norm": 2.183539867401123, + "learning_rate": 9.532385427725771e-06, + "loss": 0.2957, + "step": 1962 + }, + { + "epoch": 1.5889911970049582, + "grad_norm": 1.9144871234893799, + "learning_rate": 9.523651524599992e-06, + "loss": 0.2629, + "step": 1963 + }, + { + "epoch": 1.589800667813417, + "grad_norm": 2.232121229171753, + "learning_rate": 9.514917985649156e-06, + "loss": 0.2892, + "step": 1964 + }, + { + "epoch": 1.590610138621876, + "grad_norm": 2.2588913440704346, + "learning_rate": 9.506184817550177e-06, + "loss": 0.3033, + "step": 1965 + }, + { + "epoch": 1.5914196094303348, + "grad_norm": 1.9229258298873901, + "learning_rate": 9.497452026979674e-06, + "loss": 0.2636, + "step": 1966 + }, + { + "epoch": 1.592229080238794, + "grad_norm": 2.0721275806427, + "learning_rate": 9.488719620613995e-06, + "loss": 0.296, + "step": 1967 + }, + { + "epoch": 1.593038551047253, + "grad_norm": 2.061227321624756, + "learning_rate": 9.479987605129182e-06, + "loss": 0.3057, + "step": 1968 + }, + { + "epoch": 1.5938480218557118, + "grad_norm": 2.0762016773223877, + "learning_rate": 9.47125598720098e-06, + "loss": 0.2825, + "step": 1969 + }, + { + "epoch": 1.5946574926641708, + "grad_norm": 2.127113103866577, + "learning_rate": 9.462524773504823e-06, + "loss": 0.2623, + "step": 1970 + }, + { + "epoch": 1.5954669634726297, + "grad_norm": 2.192781686782837, + "learning_rate": 9.453793970715855e-06, + "loss": 0.2606, + "step": 1971 + }, + { + "epoch": 1.596276434281089, + "grad_norm": 2.257842779159546, + "learning_rate": 9.44506358550889e-06, + "loss": 0.3064, + "step": 1972 + }, + { + "epoch": 1.5970859050895476, + "grad_norm": 2.15505051612854, + "learning_rate": 9.436333624558427e-06, + "loss": 0.2887, + "step": 1973 + }, + { + "epoch": 1.5978953758980068, + "grad_norm": 2.0413577556610107, + "learning_rate": 9.427604094538638e-06, + "loss": 0.2665, + "step": 1974 + }, + { + "epoch": 1.5987048467064655, + "grad_norm": 2.1276040077209473, + "learning_rate": 9.418875002123375e-06, + "loss": 0.2662, + "step": 1975 + }, + { + "epoch": 1.5995143175149247, + "grad_norm": 2.224900007247925, + "learning_rate": 9.410146353986145e-06, + "loss": 0.2873, + "step": 1976 + }, + { + "epoch": 1.6003237883233836, + "grad_norm": 2.212540626525879, + "learning_rate": 9.401418156800122e-06, + "loss": 0.2961, + "step": 1977 + }, + { + "epoch": 1.6011332591318426, + "grad_norm": 2.115434169769287, + "learning_rate": 9.392690417238127e-06, + "loss": 0.2896, + "step": 1978 + }, + { + "epoch": 1.6019427299403015, + "grad_norm": 2.0728821754455566, + "learning_rate": 9.38396314197264e-06, + "loss": 0.2808, + "step": 1979 + }, + { + "epoch": 1.6027522007487605, + "grad_norm": 2.1660044193267822, + "learning_rate": 9.375236337675785e-06, + "loss": 0.3078, + "step": 1980 + }, + { + "epoch": 1.6035616715572196, + "grad_norm": 2.0360488891601562, + "learning_rate": 9.366510011019319e-06, + "loss": 0.2626, + "step": 1981 + }, + { + "epoch": 1.6043711423656783, + "grad_norm": 2.0983877182006836, + "learning_rate": 9.357784168674638e-06, + "loss": 0.2968, + "step": 1982 + }, + { + "epoch": 1.6051806131741375, + "grad_norm": 2.2431774139404297, + "learning_rate": 9.349058817312775e-06, + "loss": 0.2885, + "step": 1983 + }, + { + "epoch": 1.6059900839825962, + "grad_norm": 2.1032657623291016, + "learning_rate": 9.340333963604377e-06, + "loss": 0.2703, + "step": 1984 + }, + { + "epoch": 1.6067995547910554, + "grad_norm": 2.2832143306732178, + "learning_rate": 9.33160961421971e-06, + "loss": 0.3072, + "step": 1985 + }, + { + "epoch": 1.6076090255995144, + "grad_norm": 2.078213691711426, + "learning_rate": 9.322885775828667e-06, + "loss": 0.2788, + "step": 1986 + }, + { + "epoch": 1.6084184964079733, + "grad_norm": 1.9921523332595825, + "learning_rate": 9.314162455100737e-06, + "loss": 0.2697, + "step": 1987 + }, + { + "epoch": 1.6092279672164322, + "grad_norm": 2.0540366172790527, + "learning_rate": 9.30543965870502e-06, + "loss": 0.2818, + "step": 1988 + }, + { + "epoch": 1.6100374380248912, + "grad_norm": 2.223742723464966, + "learning_rate": 9.29671739331021e-06, + "loss": 0.297, + "step": 1989 + }, + { + "epoch": 1.6108469088333504, + "grad_norm": 2.2907145023345947, + "learning_rate": 9.287995665584601e-06, + "loss": 0.3419, + "step": 1990 + }, + { + "epoch": 1.611656379641809, + "grad_norm": 2.0558223724365234, + "learning_rate": 9.279274482196079e-06, + "loss": 0.2858, + "step": 1991 + }, + { + "epoch": 1.6124658504502682, + "grad_norm": 1.884829044342041, + "learning_rate": 9.270553849812102e-06, + "loss": 0.2745, + "step": 1992 + }, + { + "epoch": 1.613275321258727, + "grad_norm": 2.097365140914917, + "learning_rate": 9.26183377509971e-06, + "loss": 0.2821, + "step": 1993 + }, + { + "epoch": 1.6140847920671861, + "grad_norm": 2.048264741897583, + "learning_rate": 9.253114264725526e-06, + "loss": 0.2715, + "step": 1994 + }, + { + "epoch": 1.614894262875645, + "grad_norm": 2.0882718563079834, + "learning_rate": 9.244395325355735e-06, + "loss": 0.2862, + "step": 1995 + }, + { + "epoch": 1.615703733684104, + "grad_norm": 2.049567699432373, + "learning_rate": 9.235676963656082e-06, + "loss": 0.2924, + "step": 1996 + }, + { + "epoch": 1.616513204492563, + "grad_norm": 2.0108001232147217, + "learning_rate": 9.226959186291876e-06, + "loss": 0.2932, + "step": 1997 + }, + { + "epoch": 1.617322675301022, + "grad_norm": 1.9663859605789185, + "learning_rate": 9.218241999927981e-06, + "loss": 0.2661, + "step": 1998 + }, + { + "epoch": 1.618132146109481, + "grad_norm": 2.1495392322540283, + "learning_rate": 9.209525411228801e-06, + "loss": 0.2764, + "step": 1999 + }, + { + "epoch": 1.6189416169179398, + "grad_norm": 2.2811386585235596, + "learning_rate": 9.20080942685829e-06, + "loss": 0.295, + "step": 2000 + }, + { + "epoch": 1.619751087726399, + "grad_norm": 2.0374226570129395, + "learning_rate": 9.19209405347993e-06, + "loss": 0.2829, + "step": 2001 + }, + { + "epoch": 1.6205605585348577, + "grad_norm": 2.251338243484497, + "learning_rate": 9.183379297756755e-06, + "loss": 0.3203, + "step": 2002 + }, + { + "epoch": 1.6213700293433169, + "grad_norm": 2.0725882053375244, + "learning_rate": 9.174665166351312e-06, + "loss": 0.2814, + "step": 2003 + }, + { + "epoch": 1.6221795001517758, + "grad_norm": 2.1812164783477783, + "learning_rate": 9.165951665925671e-06, + "loss": 0.3085, + "step": 2004 + }, + { + "epoch": 1.6229889709602348, + "grad_norm": 2.21926212310791, + "learning_rate": 9.157238803141421e-06, + "loss": 0.3168, + "step": 2005 + }, + { + "epoch": 1.6237984417686937, + "grad_norm": 2.2158713340759277, + "learning_rate": 9.148526584659672e-06, + "loss": 0.2839, + "step": 2006 + }, + { + "epoch": 1.6246079125771526, + "grad_norm": 2.0124945640563965, + "learning_rate": 9.13981501714103e-06, + "loss": 0.2844, + "step": 2007 + }, + { + "epoch": 1.6254173833856118, + "grad_norm": 2.2634456157684326, + "learning_rate": 9.131104107245609e-06, + "loss": 0.321, + "step": 2008 + }, + { + "epoch": 1.6262268541940705, + "grad_norm": 2.0122389793395996, + "learning_rate": 9.122393861633015e-06, + "loss": 0.2666, + "step": 2009 + }, + { + "epoch": 1.6270363250025297, + "grad_norm": 2.143702507019043, + "learning_rate": 9.113684286962357e-06, + "loss": 0.3026, + "step": 2010 + }, + { + "epoch": 1.6278457958109884, + "grad_norm": 2.0468873977661133, + "learning_rate": 9.104975389892222e-06, + "loss": 0.2663, + "step": 2011 + }, + { + "epoch": 1.6286552666194476, + "grad_norm": 2.1933066844940186, + "learning_rate": 9.096267177080679e-06, + "loss": 0.3185, + "step": 2012 + }, + { + "epoch": 1.6294647374279065, + "grad_norm": 2.2349910736083984, + "learning_rate": 9.087559655185276e-06, + "loss": 0.2883, + "step": 2013 + }, + { + "epoch": 1.6302742082363655, + "grad_norm": 2.338697671890259, + "learning_rate": 9.078852830863038e-06, + "loss": 0.2927, + "step": 2014 + }, + { + "epoch": 1.6310836790448244, + "grad_norm": 2.029578685760498, + "learning_rate": 9.07014671077045e-06, + "loss": 0.2762, + "step": 2015 + }, + { + "epoch": 1.6318931498532834, + "grad_norm": 2.0559613704681396, + "learning_rate": 9.061441301563455e-06, + "loss": 0.2556, + "step": 2016 + }, + { + "epoch": 1.6327026206617425, + "grad_norm": 2.262845993041992, + "learning_rate": 9.052736609897458e-06, + "loss": 0.3123, + "step": 2017 + }, + { + "epoch": 1.6335120914702013, + "grad_norm": 2.2485406398773193, + "learning_rate": 9.044032642427325e-06, + "loss": 0.2833, + "step": 2018 + }, + { + "epoch": 1.6343215622786604, + "grad_norm": 2.089289665222168, + "learning_rate": 9.03532940580735e-06, + "loss": 0.2711, + "step": 2019 + }, + { + "epoch": 1.6351310330871192, + "grad_norm": 2.2130727767944336, + "learning_rate": 9.02662690669128e-06, + "loss": 0.2999, + "step": 2020 + }, + { + "epoch": 1.6359405038955783, + "grad_norm": 2.1156375408172607, + "learning_rate": 9.01792515173229e-06, + "loss": 0.2851, + "step": 2021 + }, + { + "epoch": 1.6367499747040373, + "grad_norm": 2.12872052192688, + "learning_rate": 9.009224147582998e-06, + "loss": 0.2866, + "step": 2022 + }, + { + "epoch": 1.6375594455124962, + "grad_norm": 2.1995866298675537, + "learning_rate": 9.000523900895438e-06, + "loss": 0.2906, + "step": 2023 + }, + { + "epoch": 1.6383689163209552, + "grad_norm": 2.144336223602295, + "learning_rate": 8.991824418321068e-06, + "loss": 0.277, + "step": 2024 + }, + { + "epoch": 1.639178387129414, + "grad_norm": 2.0567173957824707, + "learning_rate": 8.983125706510758e-06, + "loss": 0.2743, + "step": 2025 + }, + { + "epoch": 1.6399878579378733, + "grad_norm": 1.9711045026779175, + "learning_rate": 8.974427772114801e-06, + "loss": 0.2652, + "step": 2026 + }, + { + "epoch": 1.640797328746332, + "grad_norm": 2.1071393489837646, + "learning_rate": 8.96573062178288e-06, + "loss": 0.2537, + "step": 2027 + }, + { + "epoch": 1.6416067995547912, + "grad_norm": 2.1644861698150635, + "learning_rate": 8.95703426216409e-06, + "loss": 0.2956, + "step": 2028 + }, + { + "epoch": 1.6424162703632499, + "grad_norm": 2.203782081604004, + "learning_rate": 8.948338699906912e-06, + "loss": 0.2965, + "step": 2029 + }, + { + "epoch": 1.643225741171709, + "grad_norm": 2.2450039386749268, + "learning_rate": 8.939643941659229e-06, + "loss": 0.293, + "step": 2030 + }, + { + "epoch": 1.644035211980168, + "grad_norm": 2.1587746143341064, + "learning_rate": 8.930949994068299e-06, + "loss": 0.2791, + "step": 2031 + }, + { + "epoch": 1.644844682788627, + "grad_norm": 2.2109522819519043, + "learning_rate": 8.922256863780766e-06, + "loss": 0.2985, + "step": 2032 + }, + { + "epoch": 1.645654153597086, + "grad_norm": 2.1115529537200928, + "learning_rate": 8.913564557442644e-06, + "loss": 0.2836, + "step": 2033 + }, + { + "epoch": 1.6464636244055448, + "grad_norm": 2.062091588973999, + "learning_rate": 8.904873081699326e-06, + "loss": 0.2728, + "step": 2034 + }, + { + "epoch": 1.647273095214004, + "grad_norm": 1.9952820539474487, + "learning_rate": 8.89618244319556e-06, + "loss": 0.2717, + "step": 2035 + }, + { + "epoch": 1.6480825660224627, + "grad_norm": 2.1595141887664795, + "learning_rate": 8.887492648575458e-06, + "loss": 0.2739, + "step": 2036 + }, + { + "epoch": 1.648892036830922, + "grad_norm": 2.12192964553833, + "learning_rate": 8.878803704482493e-06, + "loss": 0.2592, + "step": 2037 + }, + { + "epoch": 1.6497015076393806, + "grad_norm": 1.9523380994796753, + "learning_rate": 8.870115617559476e-06, + "loss": 0.2585, + "step": 2038 + }, + { + "epoch": 1.6505109784478398, + "grad_norm": 2.3498666286468506, + "learning_rate": 8.86142839444857e-06, + "loss": 0.301, + "step": 2039 + }, + { + "epoch": 1.6513204492562987, + "grad_norm": 2.2162935733795166, + "learning_rate": 8.852742041791276e-06, + "loss": 0.3032, + "step": 2040 + }, + { + "epoch": 1.6521299200647577, + "grad_norm": 2.1916608810424805, + "learning_rate": 8.844056566228435e-06, + "loss": 0.2989, + "step": 2041 + }, + { + "epoch": 1.6529393908732166, + "grad_norm": 2.171529531478882, + "learning_rate": 8.835371974400205e-06, + "loss": 0.3078, + "step": 2042 + }, + { + "epoch": 1.6537488616816756, + "grad_norm": 2.0770514011383057, + "learning_rate": 8.82668827294608e-06, + "loss": 0.2786, + "step": 2043 + }, + { + "epoch": 1.6545583324901347, + "grad_norm": 2.2998664379119873, + "learning_rate": 8.818005468504863e-06, + "loss": 0.2928, + "step": 2044 + }, + { + "epoch": 1.6553678032985935, + "grad_norm": 2.1669228076934814, + "learning_rate": 8.809323567714685e-06, + "loss": 0.3041, + "step": 2045 + }, + { + "epoch": 1.6561772741070526, + "grad_norm": 2.1752068996429443, + "learning_rate": 8.800642577212977e-06, + "loss": 0.3041, + "step": 2046 + }, + { + "epoch": 1.6569867449155113, + "grad_norm": 2.225858449935913, + "learning_rate": 8.791962503636469e-06, + "loss": 0.3035, + "step": 2047 + }, + { + "epoch": 1.6577962157239705, + "grad_norm": 2.145186424255371, + "learning_rate": 8.7832833536212e-06, + "loss": 0.2937, + "step": 2048 + }, + { + "epoch": 1.6586056865324295, + "grad_norm": 2.3225882053375244, + "learning_rate": 8.774605133802502e-06, + "loss": 0.2876, + "step": 2049 + }, + { + "epoch": 1.6594151573408884, + "grad_norm": 2.3251938819885254, + "learning_rate": 8.76592785081499e-06, + "loss": 0.2894, + "step": 2050 + }, + { + "epoch": 1.6602246281493473, + "grad_norm": 2.097527503967285, + "learning_rate": 8.757251511292571e-06, + "loss": 0.3001, + "step": 2051 + }, + { + "epoch": 1.6610340989578063, + "grad_norm": 2.205291509628296, + "learning_rate": 8.748576121868414e-06, + "loss": 0.2863, + "step": 2052 + }, + { + "epoch": 1.6618435697662655, + "grad_norm": 2.1367275714874268, + "learning_rate": 8.739901689174983e-06, + "loss": 0.2934, + "step": 2053 + }, + { + "epoch": 1.6626530405747242, + "grad_norm": 2.0559794902801514, + "learning_rate": 8.731228219844002e-06, + "loss": 0.274, + "step": 2054 + }, + { + "epoch": 1.6634625113831834, + "grad_norm": 2.1395490169525146, + "learning_rate": 8.72255572050645e-06, + "loss": 0.2852, + "step": 2055 + }, + { + "epoch": 1.664271982191642, + "grad_norm": 2.1958835124969482, + "learning_rate": 8.713884197792575e-06, + "loss": 0.3066, + "step": 2056 + }, + { + "epoch": 1.6650814530001012, + "grad_norm": 2.0747861862182617, + "learning_rate": 8.705213658331879e-06, + "loss": 0.2877, + "step": 2057 + }, + { + "epoch": 1.6658909238085602, + "grad_norm": 2.2235426902770996, + "learning_rate": 8.6965441087531e-06, + "loss": 0.2767, + "step": 2058 + }, + { + "epoch": 1.6667003946170191, + "grad_norm": 2.1586201190948486, + "learning_rate": 8.687875555684237e-06, + "loss": 0.2899, + "step": 2059 + }, + { + "epoch": 1.667509865425478, + "grad_norm": 2.1613950729370117, + "learning_rate": 8.679208005752509e-06, + "loss": 0.2712, + "step": 2060 + }, + { + "epoch": 1.668319336233937, + "grad_norm": 2.3145039081573486, + "learning_rate": 8.670541465584383e-06, + "loss": 0.3085, + "step": 2061 + }, + { + "epoch": 1.6691288070423962, + "grad_norm": 2.2922956943511963, + "learning_rate": 8.661875941805547e-06, + "loss": 0.3034, + "step": 2062 + }, + { + "epoch": 1.669938277850855, + "grad_norm": 2.0975279808044434, + "learning_rate": 8.653211441040913e-06, + "loss": 0.2904, + "step": 2063 + }, + { + "epoch": 1.670747748659314, + "grad_norm": 2.1269915103912354, + "learning_rate": 8.644547969914606e-06, + "loss": 0.2755, + "step": 2064 + }, + { + "epoch": 1.6715572194677728, + "grad_norm": 2.178309679031372, + "learning_rate": 8.635885535049971e-06, + "loss": 0.2879, + "step": 2065 + }, + { + "epoch": 1.672366690276232, + "grad_norm": 2.1274890899658203, + "learning_rate": 8.627224143069564e-06, + "loss": 0.2989, + "step": 2066 + }, + { + "epoch": 1.673176161084691, + "grad_norm": 2.1382195949554443, + "learning_rate": 8.61856380059513e-06, + "loss": 0.2844, + "step": 2067 + }, + { + "epoch": 1.6739856318931499, + "grad_norm": 2.1696126461029053, + "learning_rate": 8.60990451424762e-06, + "loss": 0.2776, + "step": 2068 + }, + { + "epoch": 1.6747951027016088, + "grad_norm": 2.0055642127990723, + "learning_rate": 8.601246290647185e-06, + "loss": 0.2704, + "step": 2069 + }, + { + "epoch": 1.6756045735100678, + "grad_norm": 2.179130792617798, + "learning_rate": 8.592589136413148e-06, + "loss": 0.2762, + "step": 2070 + }, + { + "epoch": 1.676414044318527, + "grad_norm": 2.244837999343872, + "learning_rate": 8.583933058164027e-06, + "loss": 0.3055, + "step": 2071 + }, + { + "epoch": 1.6772235151269856, + "grad_norm": 1.9937951564788818, + "learning_rate": 8.575278062517508e-06, + "loss": 0.2818, + "step": 2072 + }, + { + "epoch": 1.6780329859354448, + "grad_norm": 2.030060291290283, + "learning_rate": 8.566624156090457e-06, + "loss": 0.272, + "step": 2073 + }, + { + "epoch": 1.6788424567439035, + "grad_norm": 2.142340660095215, + "learning_rate": 8.557971345498907e-06, + "loss": 0.2758, + "step": 2074 + }, + { + "epoch": 1.6796519275523627, + "grad_norm": 2.15877628326416, + "learning_rate": 8.549319637358044e-06, + "loss": 0.2777, + "step": 2075 + }, + { + "epoch": 1.6804613983608216, + "grad_norm": 2.2616870403289795, + "learning_rate": 8.54066903828222e-06, + "loss": 0.2935, + "step": 2076 + }, + { + "epoch": 1.6812708691692806, + "grad_norm": 2.2128894329071045, + "learning_rate": 8.532019554884939e-06, + "loss": 0.3075, + "step": 2077 + }, + { + "epoch": 1.6820803399777395, + "grad_norm": 1.9969114065170288, + "learning_rate": 8.523371193778847e-06, + "loss": 0.2762, + "step": 2078 + }, + { + "epoch": 1.6828898107861985, + "grad_norm": 2.085120439529419, + "learning_rate": 8.514723961575738e-06, + "loss": 0.284, + "step": 2079 + }, + { + "epoch": 1.6836992815946576, + "grad_norm": 2.191542625427246, + "learning_rate": 8.506077864886533e-06, + "loss": 0.2868, + "step": 2080 + }, + { + "epoch": 1.6845087524031164, + "grad_norm": 2.0763320922851562, + "learning_rate": 8.497432910321298e-06, + "loss": 0.2818, + "step": 2081 + }, + { + "epoch": 1.6853182232115755, + "grad_norm": 2.056068181991577, + "learning_rate": 8.48878910448922e-06, + "loss": 0.2869, + "step": 2082 + }, + { + "epoch": 1.6861276940200343, + "grad_norm": 2.2703475952148438, + "learning_rate": 8.4801464539986e-06, + "loss": 0.2973, + "step": 2083 + }, + { + "epoch": 1.6869371648284934, + "grad_norm": 2.1496872901916504, + "learning_rate": 8.471504965456869e-06, + "loss": 0.2743, + "step": 2084 + }, + { + "epoch": 1.6877466356369524, + "grad_norm": 2.355111598968506, + "learning_rate": 8.462864645470563e-06, + "loss": 0.2977, + "step": 2085 + }, + { + "epoch": 1.6885561064454113, + "grad_norm": 2.3398826122283936, + "learning_rate": 8.454225500645325e-06, + "loss": 0.3026, + "step": 2086 + }, + { + "epoch": 1.6893655772538703, + "grad_norm": 2.1006901264190674, + "learning_rate": 8.445587537585896e-06, + "loss": 0.2934, + "step": 2087 + }, + { + "epoch": 1.6901750480623292, + "grad_norm": 2.2302968502044678, + "learning_rate": 8.43695076289612e-06, + "loss": 0.2956, + "step": 2088 + }, + { + "epoch": 1.6909845188707884, + "grad_norm": 2.086404800415039, + "learning_rate": 8.428315183178931e-06, + "loss": 0.2947, + "step": 2089 + }, + { + "epoch": 1.691793989679247, + "grad_norm": 2.0981264114379883, + "learning_rate": 8.419680805036348e-06, + "loss": 0.2753, + "step": 2090 + }, + { + "epoch": 1.6926034604877063, + "grad_norm": 2.196150064468384, + "learning_rate": 8.411047635069467e-06, + "loss": 0.2821, + "step": 2091 + }, + { + "epoch": 1.693412931296165, + "grad_norm": 2.163135051727295, + "learning_rate": 8.402415679878471e-06, + "loss": 0.2715, + "step": 2092 + }, + { + "epoch": 1.6942224021046242, + "grad_norm": 2.159428834915161, + "learning_rate": 8.393784946062605e-06, + "loss": 0.2943, + "step": 2093 + }, + { + "epoch": 1.695031872913083, + "grad_norm": 2.2756001949310303, + "learning_rate": 8.385155440220186e-06, + "loss": 0.2782, + "step": 2094 + }, + { + "epoch": 1.695841343721542, + "grad_norm": 2.0851054191589355, + "learning_rate": 8.376527168948584e-06, + "loss": 0.2842, + "step": 2095 + }, + { + "epoch": 1.696650814530001, + "grad_norm": 2.0070548057556152, + "learning_rate": 8.367900138844235e-06, + "loss": 0.2448, + "step": 2096 + }, + { + "epoch": 1.69746028533846, + "grad_norm": 2.2275431156158447, + "learning_rate": 8.359274356502625e-06, + "loss": 0.2702, + "step": 2097 + }, + { + "epoch": 1.698269756146919, + "grad_norm": 2.19362211227417, + "learning_rate": 8.350649828518274e-06, + "loss": 0.2957, + "step": 2098 + }, + { + "epoch": 1.6990792269553778, + "grad_norm": 2.18839168548584, + "learning_rate": 8.342026561484758e-06, + "loss": 0.284, + "step": 2099 + }, + { + "epoch": 1.699888697763837, + "grad_norm": 2.1769869327545166, + "learning_rate": 8.333404561994685e-06, + "loss": 0.2802, + "step": 2100 + }, + { + "epoch": 1.7006981685722957, + "grad_norm": 2.0374484062194824, + "learning_rate": 8.324783836639687e-06, + "loss": 0.2536, + "step": 2101 + }, + { + "epoch": 1.7015076393807549, + "grad_norm": 2.06109356880188, + "learning_rate": 8.316164392010431e-06, + "loss": 0.2872, + "step": 2102 + }, + { + "epoch": 1.7023171101892138, + "grad_norm": 2.063825845718384, + "learning_rate": 8.307546234696594e-06, + "loss": 0.2842, + "step": 2103 + }, + { + "epoch": 1.7031265809976728, + "grad_norm": 2.1202480792999268, + "learning_rate": 8.298929371286881e-06, + "loss": 0.2978, + "step": 2104 + }, + { + "epoch": 1.7039360518061317, + "grad_norm": 2.1758601665496826, + "learning_rate": 8.290313808369005e-06, + "loss": 0.2964, + "step": 2105 + }, + { + "epoch": 1.7047455226145907, + "grad_norm": 2.0782368183135986, + "learning_rate": 8.281699552529676e-06, + "loss": 0.2727, + "step": 2106 + }, + { + "epoch": 1.7055549934230498, + "grad_norm": 2.228055477142334, + "learning_rate": 8.27308661035461e-06, + "loss": 0.3084, + "step": 2107 + }, + { + "epoch": 1.7063644642315086, + "grad_norm": 2.1552958488464355, + "learning_rate": 8.264474988428527e-06, + "loss": 0.3074, + "step": 2108 + }, + { + "epoch": 1.7071739350399677, + "grad_norm": 2.1679954528808594, + "learning_rate": 8.255864693335126e-06, + "loss": 0.3033, + "step": 2109 + }, + { + "epoch": 1.7079834058484265, + "grad_norm": 2.297316789627075, + "learning_rate": 8.247255731657092e-06, + "loss": 0.2875, + "step": 2110 + }, + { + "epoch": 1.7087928766568856, + "grad_norm": 2.1566898822784424, + "learning_rate": 8.238648109976097e-06, + "loss": 0.2796, + "step": 2111 + }, + { + "epoch": 1.7096023474653446, + "grad_norm": 2.117708444595337, + "learning_rate": 8.230041834872791e-06, + "loss": 0.2929, + "step": 2112 + }, + { + "epoch": 1.7104118182738035, + "grad_norm": 2.141310453414917, + "learning_rate": 8.221436912926782e-06, + "loss": 0.2709, + "step": 2113 + }, + { + "epoch": 1.7112212890822625, + "grad_norm": 2.1076929569244385, + "learning_rate": 8.212833350716654e-06, + "loss": 0.3126, + "step": 2114 + }, + { + "epoch": 1.7120307598907214, + "grad_norm": 1.9901472330093384, + "learning_rate": 8.204231154819947e-06, + "loss": 0.2885, + "step": 2115 + }, + { + "epoch": 1.7128402306991806, + "grad_norm": 2.0827877521514893, + "learning_rate": 8.195630331813158e-06, + "loss": 0.3005, + "step": 2116 + }, + { + "epoch": 1.7136497015076393, + "grad_norm": 2.149812936782837, + "learning_rate": 8.18703088827174e-06, + "loss": 0.2818, + "step": 2117 + }, + { + "epoch": 1.7144591723160985, + "grad_norm": 2.162121295928955, + "learning_rate": 8.178432830770078e-06, + "loss": 0.2899, + "step": 2118 + }, + { + "epoch": 1.7152686431245572, + "grad_norm": 2.009563446044922, + "learning_rate": 8.169836165881504e-06, + "loss": 0.2603, + "step": 2119 + }, + { + "epoch": 1.7160781139330163, + "grad_norm": 2.0104637145996094, + "learning_rate": 8.161240900178295e-06, + "loss": 0.2843, + "step": 2120 + }, + { + "epoch": 1.7168875847414753, + "grad_norm": 2.064567804336548, + "learning_rate": 8.152647040231643e-06, + "loss": 0.291, + "step": 2121 + }, + { + "epoch": 1.7176970555499342, + "grad_norm": 2.096221923828125, + "learning_rate": 8.144054592611675e-06, + "loss": 0.2881, + "step": 2122 + }, + { + "epoch": 1.7185065263583932, + "grad_norm": 2.1507444381713867, + "learning_rate": 8.135463563887428e-06, + "loss": 0.2897, + "step": 2123 + }, + { + "epoch": 1.7193159971668521, + "grad_norm": 1.9931999444961548, + "learning_rate": 8.126873960626868e-06, + "loss": 0.2457, + "step": 2124 + }, + { + "epoch": 1.7201254679753113, + "grad_norm": 2.1899101734161377, + "learning_rate": 8.118285789396869e-06, + "loss": 0.2907, + "step": 2125 + }, + { + "epoch": 1.72093493878377, + "grad_norm": 2.270683526992798, + "learning_rate": 8.109699056763193e-06, + "loss": 0.3001, + "step": 2126 + }, + { + "epoch": 1.7217444095922292, + "grad_norm": 2.07289719581604, + "learning_rate": 8.101113769290524e-06, + "loss": 0.2845, + "step": 2127 + }, + { + "epoch": 1.722553880400688, + "grad_norm": 2.1417009830474854, + "learning_rate": 8.092529933542432e-06, + "loss": 0.2674, + "step": 2128 + }, + { + "epoch": 1.723363351209147, + "grad_norm": 2.2653861045837402, + "learning_rate": 8.083947556081371e-06, + "loss": 0.2893, + "step": 2129 + }, + { + "epoch": 1.724172822017606, + "grad_norm": 2.0617613792419434, + "learning_rate": 8.075366643468695e-06, + "loss": 0.259, + "step": 2130 + }, + { + "epoch": 1.724982292826065, + "grad_norm": 2.2043392658233643, + "learning_rate": 8.06678720226462e-06, + "loss": 0.3047, + "step": 2131 + }, + { + "epoch": 1.725791763634524, + "grad_norm": 2.0861616134643555, + "learning_rate": 8.058209239028256e-06, + "loss": 0.2733, + "step": 2132 + }, + { + "epoch": 1.7266012344429829, + "grad_norm": 2.0533688068389893, + "learning_rate": 8.049632760317566e-06, + "loss": 0.2755, + "step": 2133 + }, + { + "epoch": 1.727410705251442, + "grad_norm": 2.142582416534424, + "learning_rate": 8.041057772689391e-06, + "loss": 0.2957, + "step": 2134 + }, + { + "epoch": 1.7282201760599007, + "grad_norm": 2.1128785610198975, + "learning_rate": 8.03248428269942e-06, + "loss": 0.2856, + "step": 2135 + }, + { + "epoch": 1.72902964686836, + "grad_norm": 2.189969301223755, + "learning_rate": 8.023912296902208e-06, + "loss": 0.2929, + "step": 2136 + }, + { + "epoch": 1.7298391176768186, + "grad_norm": 2.170779228210449, + "learning_rate": 8.01534182185116e-06, + "loss": 0.2809, + "step": 2137 + }, + { + "epoch": 1.7306485884852778, + "grad_norm": 2.2644848823547363, + "learning_rate": 8.006772864098512e-06, + "loss": 0.2787, + "step": 2138 + }, + { + "epoch": 1.7314580592937368, + "grad_norm": 2.0082314014434814, + "learning_rate": 7.998205430195358e-06, + "loss": 0.2742, + "step": 2139 + }, + { + "epoch": 1.7322675301021957, + "grad_norm": 2.032003879547119, + "learning_rate": 7.98963952669162e-06, + "loss": 0.2906, + "step": 2140 + }, + { + "epoch": 1.7330770009106546, + "grad_norm": 2.2177202701568604, + "learning_rate": 7.98107516013604e-06, + "loss": 0.2753, + "step": 2141 + }, + { + "epoch": 1.7338864717191136, + "grad_norm": 2.204486608505249, + "learning_rate": 7.972512337076195e-06, + "loss": 0.27, + "step": 2142 + }, + { + "epoch": 1.7346959425275728, + "grad_norm": 2.10298490524292, + "learning_rate": 7.963951064058491e-06, + "loss": 0.2964, + "step": 2143 + }, + { + "epoch": 1.7355054133360315, + "grad_norm": 2.2007288932800293, + "learning_rate": 7.95539134762813e-06, + "loss": 0.2946, + "step": 2144 + }, + { + "epoch": 1.7363148841444906, + "grad_norm": 2.1236329078674316, + "learning_rate": 7.946833194329133e-06, + "loss": 0.2835, + "step": 2145 + }, + { + "epoch": 1.7371243549529494, + "grad_norm": 2.1248602867126465, + "learning_rate": 7.938276610704325e-06, + "loss": 0.2785, + "step": 2146 + }, + { + "epoch": 1.7379338257614085, + "grad_norm": 2.187263250350952, + "learning_rate": 7.929721603295336e-06, + "loss": 0.2958, + "step": 2147 + }, + { + "epoch": 1.7387432965698675, + "grad_norm": 2.2049314975738525, + "learning_rate": 7.921168178642586e-06, + "loss": 0.3051, + "step": 2148 + }, + { + "epoch": 1.7395527673783264, + "grad_norm": 2.019287586212158, + "learning_rate": 7.912616343285282e-06, + "loss": 0.2833, + "step": 2149 + }, + { + "epoch": 1.7403622381867854, + "grad_norm": 2.0730462074279785, + "learning_rate": 7.904066103761417e-06, + "loss": 0.2827, + "step": 2150 + }, + { + "epoch": 1.7411717089952443, + "grad_norm": 2.1061978340148926, + "learning_rate": 7.895517466607778e-06, + "loss": 0.2726, + "step": 2151 + }, + { + "epoch": 1.7419811798037035, + "grad_norm": 2.3088438510894775, + "learning_rate": 7.886970438359902e-06, + "loss": 0.3048, + "step": 2152 + }, + { + "epoch": 1.7427906506121622, + "grad_norm": 2.285433053970337, + "learning_rate": 7.87842502555212e-06, + "loss": 0.2963, + "step": 2153 + }, + { + "epoch": 1.7436001214206214, + "grad_norm": 2.124560594558716, + "learning_rate": 7.869881234717507e-06, + "loss": 0.2939, + "step": 2154 + }, + { + "epoch": 1.74440959222908, + "grad_norm": 2.3237366676330566, + "learning_rate": 7.861339072387917e-06, + "loss": 0.3038, + "step": 2155 + }, + { + "epoch": 1.7452190630375393, + "grad_norm": 1.9966607093811035, + "learning_rate": 7.85279854509395e-06, + "loss": 0.272, + "step": 2156 + }, + { + "epoch": 1.7460285338459982, + "grad_norm": 2.170196056365967, + "learning_rate": 7.84425965936495e-06, + "loss": 0.2809, + "step": 2157 + }, + { + "epoch": 1.7468380046544572, + "grad_norm": 2.10599946975708, + "learning_rate": 7.835722421729014e-06, + "loss": 0.2602, + "step": 2158 + }, + { + "epoch": 1.747647475462916, + "grad_norm": 2.084071636199951, + "learning_rate": 7.827186838712982e-06, + "loss": 0.2775, + "step": 2159 + }, + { + "epoch": 1.748456946271375, + "grad_norm": 2.119438409805298, + "learning_rate": 7.818652916842422e-06, + "loss": 0.2801, + "step": 2160 + }, + { + "epoch": 1.7492664170798342, + "grad_norm": 2.0939483642578125, + "learning_rate": 7.810120662641631e-06, + "loss": 0.284, + "step": 2161 + }, + { + "epoch": 1.750075887888293, + "grad_norm": 2.2205440998077393, + "learning_rate": 7.801590082633637e-06, + "loss": 0.2752, + "step": 2162 + }, + { + "epoch": 1.750885358696752, + "grad_norm": 2.1999289989471436, + "learning_rate": 7.79306118334019e-06, + "loss": 0.2741, + "step": 2163 + }, + { + "epoch": 1.7516948295052108, + "grad_norm": 2.106261730194092, + "learning_rate": 7.784533971281742e-06, + "loss": 0.2759, + "step": 2164 + }, + { + "epoch": 1.75250430031367, + "grad_norm": 2.6372907161712646, + "learning_rate": 7.77600845297747e-06, + "loss": 0.2848, + "step": 2165 + }, + { + "epoch": 1.753313771122129, + "grad_norm": 2.1821625232696533, + "learning_rate": 7.767484634945244e-06, + "loss": 0.2627, + "step": 2166 + }, + { + "epoch": 1.7541232419305879, + "grad_norm": 2.2120347023010254, + "learning_rate": 7.758962523701644e-06, + "loss": 0.2799, + "step": 2167 + }, + { + "epoch": 1.7549327127390468, + "grad_norm": 2.1152756214141846, + "learning_rate": 7.750442125761944e-06, + "loss": 0.2721, + "step": 2168 + }, + { + "epoch": 1.7557421835475058, + "grad_norm": 1.9770961999893188, + "learning_rate": 7.741923447640098e-06, + "loss": 0.2686, + "step": 2169 + }, + { + "epoch": 1.756551654355965, + "grad_norm": 2.2488245964050293, + "learning_rate": 7.733406495848755e-06, + "loss": 0.2903, + "step": 2170 + }, + { + "epoch": 1.7573611251644237, + "grad_norm": 2.133975028991699, + "learning_rate": 7.724891276899245e-06, + "loss": 0.2772, + "step": 2171 + }, + { + "epoch": 1.7581705959728828, + "grad_norm": 2.169708490371704, + "learning_rate": 7.716377797301564e-06, + "loss": 0.2932, + "step": 2172 + }, + { + "epoch": 1.7589800667813416, + "grad_norm": 2.2380127906799316, + "learning_rate": 7.70786606356439e-06, + "loss": 0.2924, + "step": 2173 + }, + { + "epoch": 1.7597895375898007, + "grad_norm": 2.0813217163085938, + "learning_rate": 7.699356082195051e-06, + "loss": 0.2822, + "step": 2174 + }, + { + "epoch": 1.7605990083982597, + "grad_norm": 2.2274789810180664, + "learning_rate": 7.690847859699555e-06, + "loss": 0.2867, + "step": 2175 + }, + { + "epoch": 1.7614084792067186, + "grad_norm": 1.9812572002410889, + "learning_rate": 7.682341402582552e-06, + "loss": 0.2742, + "step": 2176 + }, + { + "epoch": 1.7622179500151776, + "grad_norm": 2.1978535652160645, + "learning_rate": 7.67383671734734e-06, + "loss": 0.3048, + "step": 2177 + }, + { + "epoch": 1.7630274208236365, + "grad_norm": 2.1486942768096924, + "learning_rate": 7.665333810495869e-06, + "loss": 0.2851, + "step": 2178 + }, + { + "epoch": 1.7638368916320957, + "grad_norm": 2.6770498752593994, + "learning_rate": 7.656832688528736e-06, + "loss": 0.2938, + "step": 2179 + }, + { + "epoch": 1.7646463624405544, + "grad_norm": 2.17732572555542, + "learning_rate": 7.648333357945157e-06, + "loss": 0.2677, + "step": 2180 + }, + { + "epoch": 1.7654558332490136, + "grad_norm": 2.1852622032165527, + "learning_rate": 7.639835825242989e-06, + "loss": 0.2861, + "step": 2181 + }, + { + "epoch": 1.7662653040574723, + "grad_norm": 2.23689341545105, + "learning_rate": 7.63134009691871e-06, + "loss": 0.2769, + "step": 2182 + }, + { + "epoch": 1.7670747748659315, + "grad_norm": 2.146369218826294, + "learning_rate": 7.622846179467427e-06, + "loss": 0.2646, + "step": 2183 + }, + { + "epoch": 1.7678842456743904, + "grad_norm": 2.1605353355407715, + "learning_rate": 7.614354079382851e-06, + "loss": 0.285, + "step": 2184 + }, + { + "epoch": 1.7686937164828493, + "grad_norm": 1.9894274473190308, + "learning_rate": 7.605863803157313e-06, + "loss": 0.2753, + "step": 2185 + }, + { + "epoch": 1.7695031872913083, + "grad_norm": 2.2730958461761475, + "learning_rate": 7.597375357281739e-06, + "loss": 0.3033, + "step": 2186 + }, + { + "epoch": 1.7703126580997672, + "grad_norm": 2.233332395553589, + "learning_rate": 7.588888748245669e-06, + "loss": 0.2777, + "step": 2187 + }, + { + "epoch": 1.7711221289082264, + "grad_norm": 2.1601099967956543, + "learning_rate": 7.580403982537231e-06, + "loss": 0.2644, + "step": 2188 + }, + { + "epoch": 1.7719315997166851, + "grad_norm": 2.1985902786254883, + "learning_rate": 7.5719210666431405e-06, + "loss": 0.2885, + "step": 2189 + }, + { + "epoch": 1.7727410705251443, + "grad_norm": 2.1182379722595215, + "learning_rate": 7.563440007048708e-06, + "loss": 0.2753, + "step": 2190 + }, + { + "epoch": 1.773550541333603, + "grad_norm": 2.2867298126220703, + "learning_rate": 7.55496081023782e-06, + "loss": 0.2714, + "step": 2191 + }, + { + "epoch": 1.7743600121420622, + "grad_norm": 2.183664321899414, + "learning_rate": 7.5464834826929345e-06, + "loss": 0.2915, + "step": 2192 + }, + { + "epoch": 1.7751694829505211, + "grad_norm": 2.2840230464935303, + "learning_rate": 7.538008030895085e-06, + "loss": 0.2795, + "step": 2193 + }, + { + "epoch": 1.77597895375898, + "grad_norm": 2.164673089981079, + "learning_rate": 7.5295344613238775e-06, + "loss": 0.2724, + "step": 2194 + }, + { + "epoch": 1.776788424567439, + "grad_norm": 2.131239175796509, + "learning_rate": 7.521062780457466e-06, + "loss": 0.3043, + "step": 2195 + }, + { + "epoch": 1.777597895375898, + "grad_norm": 2.131805181503296, + "learning_rate": 7.51259299477257e-06, + "loss": 0.2959, + "step": 2196 + }, + { + "epoch": 1.7784073661843571, + "grad_norm": 2.2453370094299316, + "learning_rate": 7.50412511074445e-06, + "loss": 0.2801, + "step": 2197 + }, + { + "epoch": 1.7792168369928159, + "grad_norm": 2.123540163040161, + "learning_rate": 7.4956591348469265e-06, + "loss": 0.2705, + "step": 2198 + }, + { + "epoch": 1.780026307801275, + "grad_norm": 2.247647523880005, + "learning_rate": 7.4871950735523556e-06, + "loss": 0.3097, + "step": 2199 + }, + { + "epoch": 1.7808357786097337, + "grad_norm": 2.1610631942749023, + "learning_rate": 7.478732933331623e-06, + "loss": 0.2773, + "step": 2200 + }, + { + "epoch": 1.781645249418193, + "grad_norm": 2.171480417251587, + "learning_rate": 7.47027272065415e-06, + "loss": 0.2827, + "step": 2201 + }, + { + "epoch": 1.7824547202266519, + "grad_norm": 2.3553545475006104, + "learning_rate": 7.461814441987892e-06, + "loss": 0.3081, + "step": 2202 + }, + { + "epoch": 1.7832641910351108, + "grad_norm": 2.0512616634368896, + "learning_rate": 7.453358103799315e-06, + "loss": 0.2731, + "step": 2203 + }, + { + "epoch": 1.7840736618435697, + "grad_norm": 2.202101230621338, + "learning_rate": 7.444903712553405e-06, + "loss": 0.2878, + "step": 2204 + }, + { + "epoch": 1.7848831326520287, + "grad_norm": 2.3612966537475586, + "learning_rate": 7.436451274713658e-06, + "loss": 0.2991, + "step": 2205 + }, + { + "epoch": 1.7856926034604879, + "grad_norm": 2.271810293197632, + "learning_rate": 7.428000796742088e-06, + "loss": 0.2818, + "step": 2206 + }, + { + "epoch": 1.7865020742689466, + "grad_norm": 2.3051598072052, + "learning_rate": 7.419552285099191e-06, + "loss": 0.3015, + "step": 2207 + }, + { + "epoch": 1.7873115450774057, + "grad_norm": 2.0346782207489014, + "learning_rate": 7.411105746243976e-06, + "loss": 0.2701, + "step": 2208 + }, + { + "epoch": 1.7881210158858645, + "grad_norm": 2.0258195400238037, + "learning_rate": 7.402661186633933e-06, + "loss": 0.292, + "step": 2209 + }, + { + "epoch": 1.7889304866943236, + "grad_norm": 2.1862940788269043, + "learning_rate": 7.394218612725047e-06, + "loss": 0.2769, + "step": 2210 + }, + { + "epoch": 1.7897399575027826, + "grad_norm": 2.061384439468384, + "learning_rate": 7.385778030971784e-06, + "loss": 0.2585, + "step": 2211 + }, + { + "epoch": 1.7905494283112415, + "grad_norm": 2.087458848953247, + "learning_rate": 7.3773394478270795e-06, + "loss": 0.2757, + "step": 2212 + }, + { + "epoch": 1.7913588991197005, + "grad_norm": 2.2887160778045654, + "learning_rate": 7.368902869742347e-06, + "loss": 0.2951, + "step": 2213 + }, + { + "epoch": 1.7921683699281594, + "grad_norm": 2.157862663269043, + "learning_rate": 7.360468303167472e-06, + "loss": 0.2765, + "step": 2214 + }, + { + "epoch": 1.7929778407366186, + "grad_norm": 2.0872809886932373, + "learning_rate": 7.352035754550788e-06, + "loss": 0.2708, + "step": 2215 + }, + { + "epoch": 1.7937873115450773, + "grad_norm": 1.9581794738769531, + "learning_rate": 7.343605230339101e-06, + "loss": 0.2574, + "step": 2216 + }, + { + "epoch": 1.7945967823535365, + "grad_norm": 2.2087111473083496, + "learning_rate": 7.335176736977655e-06, + "loss": 0.29, + "step": 2217 + }, + { + "epoch": 1.7954062531619952, + "grad_norm": 2.227135181427002, + "learning_rate": 7.3267502809101555e-06, + "loss": 0.2962, + "step": 2218 + }, + { + "epoch": 1.7962157239704544, + "grad_norm": 2.191321611404419, + "learning_rate": 7.318325868578744e-06, + "loss": 0.2897, + "step": 2219 + }, + { + "epoch": 1.7970251947789133, + "grad_norm": 2.0689857006073, + "learning_rate": 7.309903506423994e-06, + "loss": 0.2554, + "step": 2220 + }, + { + "epoch": 1.7978346655873723, + "grad_norm": 2.0863142013549805, + "learning_rate": 7.301483200884918e-06, + "loss": 0.2569, + "step": 2221 + }, + { + "epoch": 1.7986441363958312, + "grad_norm": 2.1628034114837646, + "learning_rate": 7.29306495839896e-06, + "loss": 0.2711, + "step": 2222 + }, + { + "epoch": 1.7994536072042902, + "grad_norm": 2.248117208480835, + "learning_rate": 7.284648785401977e-06, + "loss": 0.2826, + "step": 2223 + }, + { + "epoch": 1.8002630780127493, + "grad_norm": 2.037733793258667, + "learning_rate": 7.276234688328253e-06, + "loss": 0.2507, + "step": 2224 + }, + { + "epoch": 1.801072548821208, + "grad_norm": 2.583833932876587, + "learning_rate": 7.267822673610474e-06, + "loss": 0.2944, + "step": 2225 + }, + { + "epoch": 1.8018820196296672, + "grad_norm": 2.018542528152466, + "learning_rate": 7.259412747679748e-06, + "loss": 0.278, + "step": 2226 + }, + { + "epoch": 1.802691490438126, + "grad_norm": 4.123225688934326, + "learning_rate": 7.251004916965574e-06, + "loss": 0.2717, + "step": 2227 + }, + { + "epoch": 1.803500961246585, + "grad_norm": 2.097001314163208, + "learning_rate": 7.242599187895857e-06, + "loss": 0.2861, + "step": 2228 + }, + { + "epoch": 1.804310432055044, + "grad_norm": 2.3437464237213135, + "learning_rate": 7.234195566896886e-06, + "loss": 0.2697, + "step": 2229 + }, + { + "epoch": 1.805119902863503, + "grad_norm": 2.257843017578125, + "learning_rate": 7.2257940603933515e-06, + "loss": 0.2958, + "step": 2230 + }, + { + "epoch": 1.805929373671962, + "grad_norm": 2.1625003814697266, + "learning_rate": 7.21739467480832e-06, + "loss": 0.3029, + "step": 2231 + }, + { + "epoch": 1.8067388444804209, + "grad_norm": 2.1276981830596924, + "learning_rate": 7.2089974165632344e-06, + "loss": 0.285, + "step": 2232 + }, + { + "epoch": 1.80754831528888, + "grad_norm": 2.200935125350952, + "learning_rate": 7.200602292077911e-06, + "loss": 0.2907, + "step": 2233 + }, + { + "epoch": 1.8083577860973388, + "grad_norm": 2.055274248123169, + "learning_rate": 7.192209307770545e-06, + "loss": 0.2757, + "step": 2234 + }, + { + "epoch": 1.809167256905798, + "grad_norm": 2.122598648071289, + "learning_rate": 7.183818470057684e-06, + "loss": 0.2785, + "step": 2235 + }, + { + "epoch": 1.8099767277142567, + "grad_norm": 2.047024965286255, + "learning_rate": 7.17542978535424e-06, + "loss": 0.2829, + "step": 2236 + }, + { + "epoch": 1.8107861985227158, + "grad_norm": 2.160977602005005, + "learning_rate": 7.167043260073472e-06, + "loss": 0.2886, + "step": 2237 + }, + { + "epoch": 1.8115956693311748, + "grad_norm": 2.0835394859313965, + "learning_rate": 7.158658900627e-06, + "loss": 0.2889, + "step": 2238 + }, + { + "epoch": 1.8124051401396337, + "grad_norm": 2.2439563274383545, + "learning_rate": 7.150276713424782e-06, + "loss": 0.2856, + "step": 2239 + }, + { + "epoch": 1.8132146109480927, + "grad_norm": 2.077322483062744, + "learning_rate": 7.141896704875109e-06, + "loss": 0.271, + "step": 2240 + }, + { + "epoch": 1.8140240817565516, + "grad_norm": 2.2923576831817627, + "learning_rate": 7.133518881384618e-06, + "loss": 0.2935, + "step": 2241 + }, + { + "epoch": 1.8148335525650108, + "grad_norm": 2.1053781509399414, + "learning_rate": 7.125143249358271e-06, + "loss": 0.2812, + "step": 2242 + }, + { + "epoch": 1.8156430233734695, + "grad_norm": 2.169262647628784, + "learning_rate": 7.116769815199348e-06, + "loss": 0.2629, + "step": 2243 + }, + { + "epoch": 1.8164524941819287, + "grad_norm": 2.2919161319732666, + "learning_rate": 7.108398585309455e-06, + "loss": 0.2961, + "step": 2244 + }, + { + "epoch": 1.8172619649903874, + "grad_norm": 2.099179267883301, + "learning_rate": 7.10002956608852e-06, + "loss": 0.2731, + "step": 2245 + }, + { + "epoch": 1.8180714357988466, + "grad_norm": 2.174734354019165, + "learning_rate": 7.091662763934764e-06, + "loss": 0.2816, + "step": 2246 + }, + { + "epoch": 1.8188809066073055, + "grad_norm": 2.124054431915283, + "learning_rate": 7.083298185244729e-06, + "loss": 0.2672, + "step": 2247 + }, + { + "epoch": 1.8196903774157644, + "grad_norm": 2.045927047729492, + "learning_rate": 7.074935836413242e-06, + "loss": 0.2687, + "step": 2248 + }, + { + "epoch": 1.8204998482242234, + "grad_norm": 2.19193959236145, + "learning_rate": 7.0665757238334384e-06, + "loss": 0.2821, + "step": 2249 + }, + { + "epoch": 1.8213093190326823, + "grad_norm": 1.9448604583740234, + "learning_rate": 7.058217853896743e-06, + "loss": 0.247, + "step": 2250 + }, + { + "epoch": 1.8221187898411415, + "grad_norm": 2.2897539138793945, + "learning_rate": 7.049862232992856e-06, + "loss": 0.3054, + "step": 2251 + }, + { + "epoch": 1.8229282606496002, + "grad_norm": 2.1833279132843018, + "learning_rate": 7.041508867509762e-06, + "loss": 0.2807, + "step": 2252 + }, + { + "epoch": 1.8237377314580594, + "grad_norm": 2.0329525470733643, + "learning_rate": 7.033157763833729e-06, + "loss": 0.2552, + "step": 2253 + }, + { + "epoch": 1.8245472022665181, + "grad_norm": 2.0433309078216553, + "learning_rate": 7.024808928349291e-06, + "loss": 0.2898, + "step": 2254 + }, + { + "epoch": 1.8253566730749773, + "grad_norm": 2.3651742935180664, + "learning_rate": 7.0164623674392425e-06, + "loss": 0.2895, + "step": 2255 + }, + { + "epoch": 1.8261661438834362, + "grad_norm": 2.1589841842651367, + "learning_rate": 7.008118087484645e-06, + "loss": 0.2602, + "step": 2256 + }, + { + "epoch": 1.8269756146918952, + "grad_norm": 2.1226391792297363, + "learning_rate": 6.999776094864821e-06, + "loss": 0.2721, + "step": 2257 + }, + { + "epoch": 1.8277850855003541, + "grad_norm": 2.1111502647399902, + "learning_rate": 6.991436395957331e-06, + "loss": 0.2818, + "step": 2258 + }, + { + "epoch": 1.828594556308813, + "grad_norm": 2.0960891246795654, + "learning_rate": 6.983098997137993e-06, + "loss": 0.2689, + "step": 2259 + }, + { + "epoch": 1.8294040271172722, + "grad_norm": 2.186408519744873, + "learning_rate": 6.974763904780858e-06, + "loss": 0.3062, + "step": 2260 + }, + { + "epoch": 1.830213497925731, + "grad_norm": 2.2094905376434326, + "learning_rate": 6.966431125258224e-06, + "loss": 0.2922, + "step": 2261 + }, + { + "epoch": 1.8310229687341901, + "grad_norm": 2.23893666267395, + "learning_rate": 6.958100664940615e-06, + "loss": 0.2814, + "step": 2262 + }, + { + "epoch": 1.8318324395426488, + "grad_norm": 1.9696898460388184, + "learning_rate": 6.9497725301967785e-06, + "loss": 0.2594, + "step": 2263 + }, + { + "epoch": 1.832641910351108, + "grad_norm": 2.078864097595215, + "learning_rate": 6.941446727393688e-06, + "loss": 0.2884, + "step": 2264 + }, + { + "epoch": 1.833451381159567, + "grad_norm": 2.239295244216919, + "learning_rate": 6.933123262896539e-06, + "loss": 0.2724, + "step": 2265 + }, + { + "epoch": 1.834260851968026, + "grad_norm": 2.056583881378174, + "learning_rate": 6.924802143068731e-06, + "loss": 0.2716, + "step": 2266 + }, + { + "epoch": 1.8350703227764849, + "grad_norm": 2.123631477355957, + "learning_rate": 6.916483374271875e-06, + "loss": 0.2836, + "step": 2267 + }, + { + "epoch": 1.8358797935849438, + "grad_norm": 2.133690357208252, + "learning_rate": 6.908166962865779e-06, + "loss": 0.2872, + "step": 2268 + }, + { + "epoch": 1.836689264393403, + "grad_norm": 2.1753058433532715, + "learning_rate": 6.89985291520846e-06, + "loss": 0.2719, + "step": 2269 + }, + { + "epoch": 1.8374987352018617, + "grad_norm": 1.9706343412399292, + "learning_rate": 6.891541237656121e-06, + "loss": 0.2578, + "step": 2270 + }, + { + "epoch": 1.8383082060103209, + "grad_norm": 2.026224136352539, + "learning_rate": 6.883231936563151e-06, + "loss": 0.2858, + "step": 2271 + }, + { + "epoch": 1.8391176768187796, + "grad_norm": 1.9437229633331299, + "learning_rate": 6.874925018282121e-06, + "loss": 0.2462, + "step": 2272 + }, + { + "epoch": 1.8399271476272387, + "grad_norm": 2.222057580947876, + "learning_rate": 6.8666204891637956e-06, + "loss": 0.3162, + "step": 2273 + }, + { + "epoch": 1.8407366184356977, + "grad_norm": 2.0690293312072754, + "learning_rate": 6.858318355557092e-06, + "loss": 0.2946, + "step": 2274 + }, + { + "epoch": 1.8415460892441566, + "grad_norm": 2.2411043643951416, + "learning_rate": 6.850018623809106e-06, + "loss": 0.2924, + "step": 2275 + }, + { + "epoch": 1.8423555600526156, + "grad_norm": 2.062793016433716, + "learning_rate": 6.841721300265097e-06, + "loss": 0.2668, + "step": 2276 + }, + { + "epoch": 1.8431650308610745, + "grad_norm": 2.111402750015259, + "learning_rate": 6.833426391268485e-06, + "loss": 0.2619, + "step": 2277 + }, + { + "epoch": 1.8439745016695337, + "grad_norm": 2.0430872440338135, + "learning_rate": 6.825133903160839e-06, + "loss": 0.2756, + "step": 2278 + }, + { + "epoch": 1.8447839724779924, + "grad_norm": 2.198821544647217, + "learning_rate": 6.816843842281884e-06, + "loss": 0.2654, + "step": 2279 + }, + { + "epoch": 1.8455934432864516, + "grad_norm": 2.1086838245391846, + "learning_rate": 6.808556214969478e-06, + "loss": 0.279, + "step": 2280 + }, + { + "epoch": 1.8464029140949103, + "grad_norm": 2.27058482170105, + "learning_rate": 6.80027102755963e-06, + "loss": 0.308, + "step": 2281 + }, + { + "epoch": 1.8472123849033695, + "grad_norm": 2.1309657096862793, + "learning_rate": 6.791988286386484e-06, + "loss": 0.2722, + "step": 2282 + }, + { + "epoch": 1.8480218557118284, + "grad_norm": 2.0972304344177246, + "learning_rate": 6.783707997782303e-06, + "loss": 0.2748, + "step": 2283 + }, + { + "epoch": 1.8488313265202874, + "grad_norm": 2.1097512245178223, + "learning_rate": 6.7754301680774795e-06, + "loss": 0.2934, + "step": 2284 + }, + { + "epoch": 1.8496407973287463, + "grad_norm": 2.140956401824951, + "learning_rate": 6.767154803600538e-06, + "loss": 0.2868, + "step": 2285 + }, + { + "epoch": 1.8504502681372053, + "grad_norm": 2.2758779525756836, + "learning_rate": 6.758881910678097e-06, + "loss": 0.2805, + "step": 2286 + }, + { + "epoch": 1.8512597389456644, + "grad_norm": 2.031975746154785, + "learning_rate": 6.750611495634905e-06, + "loss": 0.2895, + "step": 2287 + }, + { + "epoch": 1.8520692097541231, + "grad_norm": 2.2596452236175537, + "learning_rate": 6.7423435647937985e-06, + "loss": 0.2759, + "step": 2288 + }, + { + "epoch": 1.8528786805625823, + "grad_norm": 2.0335280895233154, + "learning_rate": 6.734078124475731e-06, + "loss": 0.2708, + "step": 2289 + }, + { + "epoch": 1.853688151371041, + "grad_norm": 2.164616584777832, + "learning_rate": 6.725815180999748e-06, + "loss": 0.2942, + "step": 2290 + }, + { + "epoch": 1.8544976221795002, + "grad_norm": 2.1926019191741943, + "learning_rate": 6.717554740682974e-06, + "loss": 0.2628, + "step": 2291 + }, + { + "epoch": 1.8553070929879591, + "grad_norm": 2.0628998279571533, + "learning_rate": 6.709296809840635e-06, + "loss": 0.2834, + "step": 2292 + }, + { + "epoch": 1.856116563796418, + "grad_norm": 2.1739001274108887, + "learning_rate": 6.701041394786035e-06, + "loss": 0.2864, + "step": 2293 + }, + { + "epoch": 1.856926034604877, + "grad_norm": 2.1777443885803223, + "learning_rate": 6.692788501830544e-06, + "loss": 0.2866, + "step": 2294 + }, + { + "epoch": 1.857735505413336, + "grad_norm": 1.9187604188919067, + "learning_rate": 6.684538137283614e-06, + "loss": 0.2401, + "step": 2295 + }, + { + "epoch": 1.8585449762217952, + "grad_norm": 2.12962007522583, + "learning_rate": 6.676290307452769e-06, + "loss": 0.2755, + "step": 2296 + }, + { + "epoch": 1.8593544470302539, + "grad_norm": 2.2156991958618164, + "learning_rate": 6.668045018643583e-06, + "loss": 0.2814, + "step": 2297 + }, + { + "epoch": 1.860163917838713, + "grad_norm": 2.277411460876465, + "learning_rate": 6.65980227715969e-06, + "loss": 0.2985, + "step": 2298 + }, + { + "epoch": 1.8609733886471718, + "grad_norm": 2.1597812175750732, + "learning_rate": 6.651562089302781e-06, + "loss": 0.2858, + "step": 2299 + }, + { + "epoch": 1.861782859455631, + "grad_norm": 2.133389711380005, + "learning_rate": 6.6433244613725955e-06, + "loss": 0.2821, + "step": 2300 + }, + { + "epoch": 1.8625923302640899, + "grad_norm": 2.0447285175323486, + "learning_rate": 6.635089399666908e-06, + "loss": 0.2796, + "step": 2301 + }, + { + "epoch": 1.8634018010725488, + "grad_norm": 1.9670519828796387, + "learning_rate": 6.626856910481543e-06, + "loss": 0.277, + "step": 2302 + }, + { + "epoch": 1.8642112718810078, + "grad_norm": 2.185389995574951, + "learning_rate": 6.6186270001103414e-06, + "loss": 0.2857, + "step": 2303 + }, + { + "epoch": 1.8650207426894667, + "grad_norm": 2.224663734436035, + "learning_rate": 6.610399674845191e-06, + "loss": 0.2954, + "step": 2304 + }, + { + "epoch": 1.8658302134979259, + "grad_norm": 2.1781768798828125, + "learning_rate": 6.602174940975994e-06, + "loss": 0.2685, + "step": 2305 + }, + { + "epoch": 1.8666396843063846, + "grad_norm": 2.1550655364990234, + "learning_rate": 6.593952804790669e-06, + "loss": 0.2937, + "step": 2306 + }, + { + "epoch": 1.8674491551148438, + "grad_norm": 2.126499652862549, + "learning_rate": 6.585733272575151e-06, + "loss": 0.2715, + "step": 2307 + }, + { + "epoch": 1.8682586259233025, + "grad_norm": 2.099956750869751, + "learning_rate": 6.577516350613392e-06, + "loss": 0.278, + "step": 2308 + }, + { + "epoch": 1.8690680967317617, + "grad_norm": 2.0305402278900146, + "learning_rate": 6.569302045187335e-06, + "loss": 0.2759, + "step": 2309 + }, + { + "epoch": 1.8698775675402206, + "grad_norm": 2.08272123336792, + "learning_rate": 6.561090362576934e-06, + "loss": 0.2724, + "step": 2310 + }, + { + "epoch": 1.8706870383486796, + "grad_norm": 1.779908537864685, + "learning_rate": 6.552881309060125e-06, + "loss": 0.2405, + "step": 2311 + }, + { + "epoch": 1.8714965091571385, + "grad_norm": 2.1014063358306885, + "learning_rate": 6.5446748909128495e-06, + "loss": 0.2752, + "step": 2312 + }, + { + "epoch": 1.8723059799655974, + "grad_norm": 2.2386200428009033, + "learning_rate": 6.536471114409027e-06, + "loss": 0.2766, + "step": 2313 + }, + { + "epoch": 1.8731154507740566, + "grad_norm": 2.2669332027435303, + "learning_rate": 6.528269985820554e-06, + "loss": 0.2897, + "step": 2314 + }, + { + "epoch": 1.8739249215825153, + "grad_norm": 2.196129322052002, + "learning_rate": 6.520071511417303e-06, + "loss": 0.2893, + "step": 2315 + }, + { + "epoch": 1.8747343923909745, + "grad_norm": 2.236703634262085, + "learning_rate": 6.511875697467129e-06, + "loss": 0.2742, + "step": 2316 + }, + { + "epoch": 1.8755438631994332, + "grad_norm": 2.0641026496887207, + "learning_rate": 6.5036825502358365e-06, + "loss": 0.2605, + "step": 2317 + }, + { + "epoch": 1.8763533340078924, + "grad_norm": 2.093029260635376, + "learning_rate": 6.495492075987204e-06, + "loss": 0.2803, + "step": 2318 + }, + { + "epoch": 1.8771628048163513, + "grad_norm": 2.0012223720550537, + "learning_rate": 6.487304280982955e-06, + "loss": 0.2612, + "step": 2319 + }, + { + "epoch": 1.8779722756248103, + "grad_norm": 1.9819083213806152, + "learning_rate": 6.479119171482777e-06, + "loss": 0.243, + "step": 2320 + }, + { + "epoch": 1.8787817464332692, + "grad_norm": 2.304042339324951, + "learning_rate": 6.470936753744297e-06, + "loss": 0.286, + "step": 2321 + }, + { + "epoch": 1.8795912172417282, + "grad_norm": 2.026416540145874, + "learning_rate": 6.462757034023086e-06, + "loss": 0.2632, + "step": 2322 + }, + { + "epoch": 1.8804006880501873, + "grad_norm": 2.2740976810455322, + "learning_rate": 6.454580018572647e-06, + "loss": 0.2813, + "step": 2323 + }, + { + "epoch": 1.881210158858646, + "grad_norm": 2.198915958404541, + "learning_rate": 6.446405713644428e-06, + "loss": 0.2749, + "step": 2324 + }, + { + "epoch": 1.8820196296671052, + "grad_norm": 2.030503034591675, + "learning_rate": 6.438234125487795e-06, + "loss": 0.2595, + "step": 2325 + }, + { + "epoch": 1.882829100475564, + "grad_norm": 2.2215118408203125, + "learning_rate": 6.430065260350036e-06, + "loss": 0.2955, + "step": 2326 + }, + { + "epoch": 1.8836385712840231, + "grad_norm": 2.1656641960144043, + "learning_rate": 6.42189912447636e-06, + "loss": 0.2668, + "step": 2327 + }, + { + "epoch": 1.884448042092482, + "grad_norm": 2.0146028995513916, + "learning_rate": 6.413735724109897e-06, + "loss": 0.255, + "step": 2328 + }, + { + "epoch": 1.885257512900941, + "grad_norm": 2.2438910007476807, + "learning_rate": 6.405575065491672e-06, + "loss": 0.3052, + "step": 2329 + }, + { + "epoch": 1.8860669837094, + "grad_norm": 2.052661895751953, + "learning_rate": 6.397417154860625e-06, + "loss": 0.2882, + "step": 2330 + }, + { + "epoch": 1.886876454517859, + "grad_norm": 2.1695728302001953, + "learning_rate": 6.389261998453584e-06, + "loss": 0.2739, + "step": 2331 + }, + { + "epoch": 1.887685925326318, + "grad_norm": 2.082711935043335, + "learning_rate": 6.3811096025052845e-06, + "loss": 0.2812, + "step": 2332 + }, + { + "epoch": 1.8884953961347768, + "grad_norm": 2.133394479751587, + "learning_rate": 6.372959973248343e-06, + "loss": 0.2746, + "step": 2333 + }, + { + "epoch": 1.889304866943236, + "grad_norm": 2.239548444747925, + "learning_rate": 6.364813116913262e-06, + "loss": 0.2754, + "step": 2334 + }, + { + "epoch": 1.8901143377516947, + "grad_norm": 2.0659308433532715, + "learning_rate": 6.356669039728423e-06, + "loss": 0.2774, + "step": 2335 + }, + { + "epoch": 1.8909238085601539, + "grad_norm": 1.9359850883483887, + "learning_rate": 6.3485277479200915e-06, + "loss": 0.2552, + "step": 2336 + }, + { + "epoch": 1.8917332793686128, + "grad_norm": 2.092707395553589, + "learning_rate": 6.340389247712391e-06, + "loss": 0.2987, + "step": 2337 + }, + { + "epoch": 1.8925427501770717, + "grad_norm": 2.0866682529449463, + "learning_rate": 6.332253545327318e-06, + "loss": 0.2727, + "step": 2338 + }, + { + "epoch": 1.8933522209855307, + "grad_norm": 2.0720674991607666, + "learning_rate": 6.324120646984726e-06, + "loss": 0.2775, + "step": 2339 + }, + { + "epoch": 1.8941616917939896, + "grad_norm": 2.1538314819335938, + "learning_rate": 6.315990558902333e-06, + "loss": 0.2894, + "step": 2340 + }, + { + "epoch": 1.8949711626024488, + "grad_norm": 2.0269508361816406, + "learning_rate": 6.3078632872957014e-06, + "loss": 0.2608, + "step": 2341 + }, + { + "epoch": 1.8957806334109075, + "grad_norm": 2.1473517417907715, + "learning_rate": 6.2997388383782354e-06, + "loss": 0.2522, + "step": 2342 + }, + { + "epoch": 1.8965901042193667, + "grad_norm": 2.116306781768799, + "learning_rate": 6.291617218361195e-06, + "loss": 0.2634, + "step": 2343 + }, + { + "epoch": 1.8973995750278254, + "grad_norm": 2.093594551086426, + "learning_rate": 6.28349843345367e-06, + "loss": 0.2709, + "step": 2344 + }, + { + "epoch": 1.8982090458362846, + "grad_norm": 2.1637964248657227, + "learning_rate": 6.275382489862581e-06, + "loss": 0.2675, + "step": 2345 + }, + { + "epoch": 1.8990185166447435, + "grad_norm": 2.1597166061401367, + "learning_rate": 6.2672693937926724e-06, + "loss": 0.2596, + "step": 2346 + }, + { + "epoch": 1.8998279874532025, + "grad_norm": 2.135725736618042, + "learning_rate": 6.259159151446524e-06, + "loss": 0.2687, + "step": 2347 + }, + { + "epoch": 1.9006374582616614, + "grad_norm": 2.0694642066955566, + "learning_rate": 6.25105176902453e-06, + "loss": 0.2674, + "step": 2348 + }, + { + "epoch": 1.9014469290701204, + "grad_norm": 2.365304470062256, + "learning_rate": 6.242947252724889e-06, + "loss": 0.3091, + "step": 2349 + }, + { + "epoch": 1.9022563998785795, + "grad_norm": 2.1861650943756104, + "learning_rate": 6.234845608743616e-06, + "loss": 0.2644, + "step": 2350 + }, + { + "epoch": 1.9030658706870383, + "grad_norm": 2.165543794631958, + "learning_rate": 6.226746843274535e-06, + "loss": 0.2751, + "step": 2351 + }, + { + "epoch": 1.9038753414954974, + "grad_norm": 2.295109987258911, + "learning_rate": 6.218650962509256e-06, + "loss": 0.2717, + "step": 2352 + }, + { + "epoch": 1.9046848123039561, + "grad_norm": 2.083728075027466, + "learning_rate": 6.2105579726371965e-06, + "loss": 0.2856, + "step": 2353 + }, + { + "epoch": 1.9054942831124153, + "grad_norm": 2.1810948848724365, + "learning_rate": 6.202467879845551e-06, + "loss": 0.2616, + "step": 2354 + }, + { + "epoch": 1.9063037539208743, + "grad_norm": 2.398364782333374, + "learning_rate": 6.194380690319312e-06, + "loss": 0.2839, + "step": 2355 + }, + { + "epoch": 1.9071132247293332, + "grad_norm": 2.027559280395508, + "learning_rate": 6.186296410241249e-06, + "loss": 0.2686, + "step": 2356 + }, + { + "epoch": 1.9079226955377921, + "grad_norm": 2.2549338340759277, + "learning_rate": 6.1782150457919e-06, + "loss": 0.2861, + "step": 2357 + }, + { + "epoch": 1.908732166346251, + "grad_norm": 2.2459051609039307, + "learning_rate": 6.170136603149579e-06, + "loss": 0.3111, + "step": 2358 + }, + { + "epoch": 1.9095416371547103, + "grad_norm": 2.1524622440338135, + "learning_rate": 6.162061088490371e-06, + "loss": 0.2537, + "step": 2359 + }, + { + "epoch": 1.910351107963169, + "grad_norm": 2.3109076023101807, + "learning_rate": 6.153988507988114e-06, + "loss": 0.2879, + "step": 2360 + }, + { + "epoch": 1.9111605787716281, + "grad_norm": 2.2176480293273926, + "learning_rate": 6.1459188678144085e-06, + "loss": 0.2874, + "step": 2361 + }, + { + "epoch": 1.9119700495800869, + "grad_norm": 2.1447489261627197, + "learning_rate": 6.1378521741386e-06, + "loss": 0.2742, + "step": 2362 + }, + { + "epoch": 1.912779520388546, + "grad_norm": 2.228905200958252, + "learning_rate": 6.129788433127793e-06, + "loss": 0.2961, + "step": 2363 + }, + { + "epoch": 1.913588991197005, + "grad_norm": 2.1456429958343506, + "learning_rate": 6.121727650946827e-06, + "loss": 0.273, + "step": 2364 + }, + { + "epoch": 1.914398462005464, + "grad_norm": 2.01604962348938, + "learning_rate": 6.113669833758278e-06, + "loss": 0.2496, + "step": 2365 + }, + { + "epoch": 1.9152079328139229, + "grad_norm": 2.07493257522583, + "learning_rate": 6.1056149877224555e-06, + "loss": 0.2804, + "step": 2366 + }, + { + "epoch": 1.9160174036223818, + "grad_norm": 2.0364673137664795, + "learning_rate": 6.0975631189974085e-06, + "loss": 0.2619, + "step": 2367 + }, + { + "epoch": 1.916826874430841, + "grad_norm": 2.08185076713562, + "learning_rate": 6.089514233738896e-06, + "loss": 0.2638, + "step": 2368 + }, + { + "epoch": 1.9176363452392997, + "grad_norm": 2.1193225383758545, + "learning_rate": 6.081468338100398e-06, + "loss": 0.2852, + "step": 2369 + }, + { + "epoch": 1.9184458160477589, + "grad_norm": 2.077718496322632, + "learning_rate": 6.073425438233114e-06, + "loss": 0.2584, + "step": 2370 + }, + { + "epoch": 1.9192552868562176, + "grad_norm": 2.101452112197876, + "learning_rate": 6.065385540285957e-06, + "loss": 0.2581, + "step": 2371 + }, + { + "epoch": 1.9200647576646768, + "grad_norm": 2.1649112701416016, + "learning_rate": 6.057348650405531e-06, + "loss": 0.2878, + "step": 2372 + }, + { + "epoch": 1.9208742284731357, + "grad_norm": 2.039034128189087, + "learning_rate": 6.049314774736156e-06, + "loss": 0.2631, + "step": 2373 + }, + { + "epoch": 1.9216836992815947, + "grad_norm": 2.045480966567993, + "learning_rate": 6.041283919419832e-06, + "loss": 0.2693, + "step": 2374 + }, + { + "epoch": 1.9224931700900536, + "grad_norm": 2.3989763259887695, + "learning_rate": 6.033256090596265e-06, + "loss": 0.2864, + "step": 2375 + }, + { + "epoch": 1.9233026408985125, + "grad_norm": 2.18249249458313, + "learning_rate": 6.025231294402838e-06, + "loss": 0.2913, + "step": 2376 + }, + { + "epoch": 1.9241121117069717, + "grad_norm": 2.1020963191986084, + "learning_rate": 6.017209536974618e-06, + "loss": 0.2673, + "step": 2377 + }, + { + "epoch": 1.9249215825154304, + "grad_norm": 2.201119899749756, + "learning_rate": 6.009190824444344e-06, + "loss": 0.2866, + "step": 2378 + }, + { + "epoch": 1.9257310533238896, + "grad_norm": 2.0938305854797363, + "learning_rate": 6.001175162942441e-06, + "loss": 0.2845, + "step": 2379 + }, + { + "epoch": 1.9265405241323483, + "grad_norm": 2.175389289855957, + "learning_rate": 5.993162558596986e-06, + "loss": 0.3055, + "step": 2380 + }, + { + "epoch": 1.9273499949408075, + "grad_norm": 2.068923234939575, + "learning_rate": 5.9851530175337265e-06, + "loss": 0.2969, + "step": 2381 + }, + { + "epoch": 1.9281594657492664, + "grad_norm": 2.1029162406921387, + "learning_rate": 5.977146545876062e-06, + "loss": 0.2939, + "step": 2382 + }, + { + "epoch": 1.9289689365577254, + "grad_norm": 2.1807918548583984, + "learning_rate": 5.969143149745058e-06, + "loss": 0.2973, + "step": 2383 + }, + { + "epoch": 1.9297784073661843, + "grad_norm": 2.027874708175659, + "learning_rate": 5.961142835259417e-06, + "loss": 0.2655, + "step": 2384 + }, + { + "epoch": 1.9305878781746433, + "grad_norm": 1.946837067604065, + "learning_rate": 5.9531456085354875e-06, + "loss": 0.2577, + "step": 2385 + }, + { + "epoch": 1.9313973489831024, + "grad_norm": 2.074690580368042, + "learning_rate": 5.9451514756872585e-06, + "loss": 0.2624, + "step": 2386 + }, + { + "epoch": 1.9322068197915612, + "grad_norm": 1.9160209894180298, + "learning_rate": 5.937160442826359e-06, + "loss": 0.2813, + "step": 2387 + }, + { + "epoch": 1.9330162906000203, + "grad_norm": 2.1088290214538574, + "learning_rate": 5.929172516062041e-06, + "loss": 0.278, + "step": 2388 + }, + { + "epoch": 1.933825761408479, + "grad_norm": 2.13334321975708, + "learning_rate": 5.921187701501183e-06, + "loss": 0.2703, + "step": 2389 + }, + { + "epoch": 1.9346352322169382, + "grad_norm": 2.0986831188201904, + "learning_rate": 5.913206005248284e-06, + "loss": 0.3019, + "step": 2390 + }, + { + "epoch": 1.9354447030253972, + "grad_norm": 1.9678962230682373, + "learning_rate": 5.905227433405465e-06, + "loss": 0.2652, + "step": 2391 + }, + { + "epoch": 1.9362541738338561, + "grad_norm": 2.0996243953704834, + "learning_rate": 5.897251992072448e-06, + "loss": 0.2813, + "step": 2392 + }, + { + "epoch": 1.937063644642315, + "grad_norm": 2.1879453659057617, + "learning_rate": 5.88927968734657e-06, + "loss": 0.2753, + "step": 2393 + }, + { + "epoch": 1.937873115450774, + "grad_norm": 2.1438722610473633, + "learning_rate": 5.881310525322769e-06, + "loss": 0.2881, + "step": 2394 + }, + { + "epoch": 1.9386825862592332, + "grad_norm": 2.1382410526275635, + "learning_rate": 5.873344512093576e-06, + "loss": 0.2849, + "step": 2395 + }, + { + "epoch": 1.939492057067692, + "grad_norm": 2.1884121894836426, + "learning_rate": 5.86538165374912e-06, + "loss": 0.2885, + "step": 2396 + }, + { + "epoch": 1.940301527876151, + "grad_norm": 2.273175001144409, + "learning_rate": 5.857421956377111e-06, + "loss": 0.2509, + "step": 2397 + }, + { + "epoch": 1.9411109986846098, + "grad_norm": 2.197890520095825, + "learning_rate": 5.849465426062852e-06, + "loss": 0.2722, + "step": 2398 + }, + { + "epoch": 1.941920469493069, + "grad_norm": 2.2589025497436523, + "learning_rate": 5.841512068889218e-06, + "loss": 0.2887, + "step": 2399 + }, + { + "epoch": 1.942729940301528, + "grad_norm": 1.9885386228561401, + "learning_rate": 5.8335618909366575e-06, + "loss": 0.2462, + "step": 2400 + }, + { + "epoch": 1.9435394111099868, + "grad_norm": 1.9374454021453857, + "learning_rate": 5.8256148982831904e-06, + "loss": 0.2476, + "step": 2401 + }, + { + "epoch": 1.9443488819184458, + "grad_norm": 2.14209246635437, + "learning_rate": 5.817671097004407e-06, + "loss": 0.2889, + "step": 2402 + }, + { + "epoch": 1.9451583527269047, + "grad_norm": 2.2639925479888916, + "learning_rate": 5.809730493173449e-06, + "loss": 0.3052, + "step": 2403 + }, + { + "epoch": 1.945967823535364, + "grad_norm": 2.110755681991577, + "learning_rate": 5.801793092861012e-06, + "loss": 0.2697, + "step": 2404 + }, + { + "epoch": 1.9467772943438226, + "grad_norm": 2.0751750469207764, + "learning_rate": 5.793858902135354e-06, + "loss": 0.2926, + "step": 2405 + }, + { + "epoch": 1.9475867651522818, + "grad_norm": 1.9452046155929565, + "learning_rate": 5.785927927062274e-06, + "loss": 0.2638, + "step": 2406 + }, + { + "epoch": 1.9483962359607405, + "grad_norm": 2.147329568862915, + "learning_rate": 5.778000173705108e-06, + "loss": 0.2811, + "step": 2407 + }, + { + "epoch": 1.9492057067691997, + "grad_norm": 1.850246787071228, + "learning_rate": 5.770075648124731e-06, + "loss": 0.2379, + "step": 2408 + }, + { + "epoch": 1.9500151775776586, + "grad_norm": 2.022179365158081, + "learning_rate": 5.762154356379549e-06, + "loss": 0.2631, + "step": 2409 + }, + { + "epoch": 1.9508246483861176, + "grad_norm": 2.136902093887329, + "learning_rate": 5.754236304525502e-06, + "loss": 0.2835, + "step": 2410 + }, + { + "epoch": 1.9516341191945765, + "grad_norm": 2.018768072128296, + "learning_rate": 5.746321498616054e-06, + "loss": 0.2927, + "step": 2411 + }, + { + "epoch": 1.9524435900030355, + "grad_norm": 2.25108003616333, + "learning_rate": 5.738409944702178e-06, + "loss": 0.3014, + "step": 2412 + }, + { + "epoch": 1.9532530608114946, + "grad_norm": 2.078582763671875, + "learning_rate": 5.73050164883236e-06, + "loss": 0.268, + "step": 2413 + }, + { + "epoch": 1.9540625316199534, + "grad_norm": 2.2694931030273438, + "learning_rate": 5.7225966170526126e-06, + "loss": 0.2909, + "step": 2414 + }, + { + "epoch": 1.9548720024284125, + "grad_norm": 2.1132640838623047, + "learning_rate": 5.714694855406433e-06, + "loss": 0.2481, + "step": 2415 + }, + { + "epoch": 1.9556814732368712, + "grad_norm": 2.1646275520324707, + "learning_rate": 5.706796369934825e-06, + "loss": 0.2742, + "step": 2416 + }, + { + "epoch": 1.9564909440453304, + "grad_norm": 1.9764515161514282, + "learning_rate": 5.698901166676291e-06, + "loss": 0.2476, + "step": 2417 + }, + { + "epoch": 1.9573004148537894, + "grad_norm": 2.2011826038360596, + "learning_rate": 5.691009251666828e-06, + "loss": 0.2899, + "step": 2418 + }, + { + "epoch": 1.9581098856622483, + "grad_norm": 2.2578701972961426, + "learning_rate": 5.683120630939908e-06, + "loss": 0.284, + "step": 2419 + }, + { + "epoch": 1.9589193564707073, + "grad_norm": 2.2151238918304443, + "learning_rate": 5.67523531052649e-06, + "loss": 0.2808, + "step": 2420 + }, + { + "epoch": 1.9597288272791662, + "grad_norm": 2.1864371299743652, + "learning_rate": 5.667353296455007e-06, + "loss": 0.2774, + "step": 2421 + }, + { + "epoch": 1.9605382980876254, + "grad_norm": 2.0950417518615723, + "learning_rate": 5.65947459475137e-06, + "loss": 0.2563, + "step": 2422 + }, + { + "epoch": 1.961347768896084, + "grad_norm": 2.2098519802093506, + "learning_rate": 5.651599211438959e-06, + "loss": 0.2952, + "step": 2423 + }, + { + "epoch": 1.9621572397045433, + "grad_norm": 1.958470344543457, + "learning_rate": 5.643727152538605e-06, + "loss": 0.2507, + "step": 2424 + }, + { + "epoch": 1.962966710513002, + "grad_norm": 2.088920831680298, + "learning_rate": 5.635858424068607e-06, + "loss": 0.2746, + "step": 2425 + }, + { + "epoch": 1.9637761813214611, + "grad_norm": 2.187926769256592, + "learning_rate": 5.627993032044719e-06, + "loss": 0.29, + "step": 2426 + }, + { + "epoch": 1.96458565212992, + "grad_norm": 2.052572011947632, + "learning_rate": 5.620130982480136e-06, + "loss": 0.2676, + "step": 2427 + }, + { + "epoch": 1.965395122938379, + "grad_norm": 2.0612266063690186, + "learning_rate": 5.612272281385511e-06, + "loss": 0.2598, + "step": 2428 + }, + { + "epoch": 1.966204593746838, + "grad_norm": 2.157381534576416, + "learning_rate": 5.604416934768917e-06, + "loss": 0.2814, + "step": 2429 + }, + { + "epoch": 1.967014064555297, + "grad_norm": 2.054642915725708, + "learning_rate": 5.596564948635887e-06, + "loss": 0.2759, + "step": 2430 + }, + { + "epoch": 1.967823535363756, + "grad_norm": 2.118021011352539, + "learning_rate": 5.588716328989365e-06, + "loss": 0.2847, + "step": 2431 + }, + { + "epoch": 1.9686330061722148, + "grad_norm": 2.1194803714752197, + "learning_rate": 5.580871081829726e-06, + "loss": 0.258, + "step": 2432 + }, + { + "epoch": 1.969442476980674, + "grad_norm": 2.2362263202667236, + "learning_rate": 5.573029213154772e-06, + "loss": 0.2986, + "step": 2433 + }, + { + "epoch": 1.9702519477891327, + "grad_norm": 2.024343490600586, + "learning_rate": 5.565190728959724e-06, + "loss": 0.2674, + "step": 2434 + }, + { + "epoch": 1.9710614185975919, + "grad_norm": 1.9038434028625488, + "learning_rate": 5.557355635237205e-06, + "loss": 0.2533, + "step": 2435 + }, + { + "epoch": 1.9718708894060508, + "grad_norm": 2.1973581314086914, + "learning_rate": 5.549523937977255e-06, + "loss": 0.2879, + "step": 2436 + }, + { + "epoch": 1.9726803602145098, + "grad_norm": 2.0526511669158936, + "learning_rate": 5.541695643167306e-06, + "loss": 0.255, + "step": 2437 + }, + { + "epoch": 1.9734898310229687, + "grad_norm": 2.35564923286438, + "learning_rate": 5.533870756792208e-06, + "loss": 0.2944, + "step": 2438 + }, + { + "epoch": 1.9742993018314277, + "grad_norm": 2.040198564529419, + "learning_rate": 5.526049284834185e-06, + "loss": 0.2944, + "step": 2439 + }, + { + "epoch": 1.9751087726398868, + "grad_norm": 2.158759593963623, + "learning_rate": 5.5182312332728616e-06, + "loss": 0.2642, + "step": 2440 + }, + { + "epoch": 1.9759182434483455, + "grad_norm": 2.001843214035034, + "learning_rate": 5.510416608085254e-06, + "loss": 0.2575, + "step": 2441 + }, + { + "epoch": 1.9767277142568047, + "grad_norm": 1.9253160953521729, + "learning_rate": 5.502605415245741e-06, + "loss": 0.2606, + "step": 2442 + }, + { + "epoch": 1.9775371850652634, + "grad_norm": 1.9545522928237915, + "learning_rate": 5.49479766072609e-06, + "loss": 0.2533, + "step": 2443 + }, + { + "epoch": 1.9783466558737226, + "grad_norm": 1.9362475872039795, + "learning_rate": 5.486993350495433e-06, + "loss": 0.2492, + "step": 2444 + }, + { + "epoch": 1.9791561266821815, + "grad_norm": 2.027722120285034, + "learning_rate": 5.479192490520274e-06, + "loss": 0.2781, + "step": 2445 + }, + { + "epoch": 1.9799655974906405, + "grad_norm": 2.1071300506591797, + "learning_rate": 5.471395086764483e-06, + "loss": 0.2914, + "step": 2446 + }, + { + "epoch": 1.9807750682990994, + "grad_norm": 1.9793752431869507, + "learning_rate": 5.463601145189277e-06, + "loss": 0.2697, + "step": 2447 + }, + { + "epoch": 1.9815845391075584, + "grad_norm": 2.1461567878723145, + "learning_rate": 5.455810671753228e-06, + "loss": 0.2748, + "step": 2448 + }, + { + "epoch": 1.9823940099160176, + "grad_norm": 2.0006039142608643, + "learning_rate": 5.448023672412268e-06, + "loss": 0.254, + "step": 2449 + }, + { + "epoch": 1.9832034807244763, + "grad_norm": 2.163480520248413, + "learning_rate": 5.4402401531196556e-06, + "loss": 0.2661, + "step": 2450 + }, + { + "epoch": 1.9840129515329354, + "grad_norm": 2.245497703552246, + "learning_rate": 5.432460119826009e-06, + "loss": 0.2714, + "step": 2451 + }, + { + "epoch": 1.9848224223413942, + "grad_norm": 2.0115485191345215, + "learning_rate": 5.424683578479259e-06, + "loss": 0.2563, + "step": 2452 + }, + { + "epoch": 1.9856318931498533, + "grad_norm": 2.1395263671875, + "learning_rate": 5.416910535024687e-06, + "loss": 0.2692, + "step": 2453 + }, + { + "epoch": 1.9864413639583123, + "grad_norm": 2.091540813446045, + "learning_rate": 5.409140995404889e-06, + "loss": 0.2656, + "step": 2454 + }, + { + "epoch": 1.9872508347667712, + "grad_norm": 2.1761462688446045, + "learning_rate": 5.401374965559779e-06, + "loss": 0.2867, + "step": 2455 + }, + { + "epoch": 1.9880603055752302, + "grad_norm": 2.1796507835388184, + "learning_rate": 5.393612451426599e-06, + "loss": 0.2791, + "step": 2456 + }, + { + "epoch": 1.9888697763836891, + "grad_norm": 2.3402316570281982, + "learning_rate": 5.385853458939899e-06, + "loss": 0.2812, + "step": 2457 + }, + { + "epoch": 1.9896792471921483, + "grad_norm": 2.135530948638916, + "learning_rate": 5.378097994031533e-06, + "loss": 0.274, + "step": 2458 + }, + { + "epoch": 1.990488718000607, + "grad_norm": 2.142026901245117, + "learning_rate": 5.3703460626306606e-06, + "loss": 0.2664, + "step": 2459 + }, + { + "epoch": 1.9912981888090662, + "grad_norm": 2.122676134109497, + "learning_rate": 5.362597670663734e-06, + "loss": 0.2561, + "step": 2460 + }, + { + "epoch": 1.992107659617525, + "grad_norm": 2.1640491485595703, + "learning_rate": 5.354852824054515e-06, + "loss": 0.2571, + "step": 2461 + }, + { + "epoch": 1.992917130425984, + "grad_norm": 2.1654131412506104, + "learning_rate": 5.347111528724037e-06, + "loss": 0.2831, + "step": 2462 + }, + { + "epoch": 1.993726601234443, + "grad_norm": 1.8599685430526733, + "learning_rate": 5.339373790590633e-06, + "loss": 0.2555, + "step": 2463 + }, + { + "epoch": 1.994536072042902, + "grad_norm": 2.0832908153533936, + "learning_rate": 5.331639615569903e-06, + "loss": 0.2574, + "step": 2464 + }, + { + "epoch": 1.995345542851361, + "grad_norm": 2.2036991119384766, + "learning_rate": 5.323909009574737e-06, + "loss": 0.2759, + "step": 2465 + }, + { + "epoch": 1.9961550136598198, + "grad_norm": 2.3514668941497803, + "learning_rate": 5.316181978515287e-06, + "loss": 0.2638, + "step": 2466 + }, + { + "epoch": 1.996964484468279, + "grad_norm": 2.1264941692352295, + "learning_rate": 5.30845852829897e-06, + "loss": 0.2698, + "step": 2467 + }, + { + "epoch": 1.9977739552767377, + "grad_norm": 2.030647039413452, + "learning_rate": 5.300738664830471e-06, + "loss": 0.2674, + "step": 2468 + }, + { + "epoch": 1.998583426085197, + "grad_norm": 2.2002739906311035, + "learning_rate": 5.293022394011739e-06, + "loss": 0.2991, + "step": 2469 + }, + { + "epoch": 1.9993928968936556, + "grad_norm": 2.0529372692108154, + "learning_rate": 5.285309721741963e-06, + "loss": 0.2605, + "step": 2470 + }, + { + "epoch": 2.000202367702115, + "grad_norm": 1.8294931650161743, + "learning_rate": 5.277600653917586e-06, + "loss": 0.2176, + "step": 2471 + }, + { + "epoch": 2.0010118385105735, + "grad_norm": 1.5261549949645996, + "learning_rate": 5.269895196432294e-06, + "loss": 0.1334, + "step": 2472 + }, + { + "epoch": 2.0018213093190327, + "grad_norm": 1.483446717262268, + "learning_rate": 5.262193355177019e-06, + "loss": 0.1201, + "step": 2473 + }, + { + "epoch": 2.002630780127492, + "grad_norm": 1.5387529134750366, + "learning_rate": 5.254495136039926e-06, + "loss": 0.1263, + "step": 2474 + }, + { + "epoch": 2.0034402509359506, + "grad_norm": 1.5015264749526978, + "learning_rate": 5.246800544906409e-06, + "loss": 0.1193, + "step": 2475 + }, + { + "epoch": 2.0042497217444097, + "grad_norm": 1.4047256708145142, + "learning_rate": 5.23910958765908e-06, + "loss": 0.1183, + "step": 2476 + }, + { + "epoch": 2.0050591925528685, + "grad_norm": 1.323002815246582, + "learning_rate": 5.231422270177793e-06, + "loss": 0.1068, + "step": 2477 + }, + { + "epoch": 2.0058686633613276, + "grad_norm": 1.4716951847076416, + "learning_rate": 5.223738598339596e-06, + "loss": 0.1128, + "step": 2478 + }, + { + "epoch": 2.0066781341697864, + "grad_norm": 1.4531677961349487, + "learning_rate": 5.216058578018775e-06, + "loss": 0.1073, + "step": 2479 + }, + { + "epoch": 2.0074876049782455, + "grad_norm": 1.5842974185943604, + "learning_rate": 5.208382215086796e-06, + "loss": 0.1107, + "step": 2480 + }, + { + "epoch": 2.0082970757867042, + "grad_norm": 1.8806239366531372, + "learning_rate": 5.200709515412358e-06, + "loss": 0.113, + "step": 2481 + }, + { + "epoch": 2.0091065465951634, + "grad_norm": 2.0215392112731934, + "learning_rate": 5.193040484861337e-06, + "loss": 0.1133, + "step": 2482 + }, + { + "epoch": 2.0099160174036226, + "grad_norm": 2.0342442989349365, + "learning_rate": 5.185375129296815e-06, + "loss": 0.1049, + "step": 2483 + }, + { + "epoch": 2.0107254882120813, + "grad_norm": 2.265446662902832, + "learning_rate": 5.177713454579051e-06, + "loss": 0.114, + "step": 2484 + }, + { + "epoch": 2.0115349590205405, + "grad_norm": 1.941333293914795, + "learning_rate": 5.170055466565518e-06, + "loss": 0.093, + "step": 2485 + }, + { + "epoch": 2.012344429828999, + "grad_norm": 2.1051292419433594, + "learning_rate": 5.1624011711108444e-06, + "loss": 0.1096, + "step": 2486 + }, + { + "epoch": 2.0131539006374584, + "grad_norm": 2.0482287406921387, + "learning_rate": 5.1547505740668434e-06, + "loss": 0.0956, + "step": 2487 + }, + { + "epoch": 2.013963371445917, + "grad_norm": 1.9178968667984009, + "learning_rate": 5.147103681282499e-06, + "loss": 0.102, + "step": 2488 + }, + { + "epoch": 2.0147728422543763, + "grad_norm": 1.9433512687683105, + "learning_rate": 5.139460498603972e-06, + "loss": 0.1027, + "step": 2489 + }, + { + "epoch": 2.015582313062835, + "grad_norm": 2.0065925121307373, + "learning_rate": 5.1318210318745745e-06, + "loss": 0.1008, + "step": 2490 + }, + { + "epoch": 2.016391783871294, + "grad_norm": 1.907593011856079, + "learning_rate": 5.124185286934785e-06, + "loss": 0.1074, + "step": 2491 + }, + { + "epoch": 2.0172012546797533, + "grad_norm": 1.8393498659133911, + "learning_rate": 5.116553269622242e-06, + "loss": 0.1079, + "step": 2492 + }, + { + "epoch": 2.018010725488212, + "grad_norm": 2.1212046146392822, + "learning_rate": 5.1089249857717205e-06, + "loss": 0.0944, + "step": 2493 + }, + { + "epoch": 2.018820196296671, + "grad_norm": 1.7024526596069336, + "learning_rate": 5.10130044121515e-06, + "loss": 0.0934, + "step": 2494 + }, + { + "epoch": 2.01962966710513, + "grad_norm": 1.8046730756759644, + "learning_rate": 5.093679641781595e-06, + "loss": 0.0997, + "step": 2495 + }, + { + "epoch": 2.020439137913589, + "grad_norm": 1.939642310142517, + "learning_rate": 5.086062593297267e-06, + "loss": 0.1107, + "step": 2496 + }, + { + "epoch": 2.021248608722048, + "grad_norm": 2.1145126819610596, + "learning_rate": 5.078449301585504e-06, + "loss": 0.1138, + "step": 2497 + }, + { + "epoch": 2.022058079530507, + "grad_norm": 1.6778801679611206, + "learning_rate": 5.0708397724667716e-06, + "loss": 0.0998, + "step": 2498 + }, + { + "epoch": 2.0228675503389657, + "grad_norm": 1.857100248336792, + "learning_rate": 5.063234011758654e-06, + "loss": 0.1092, + "step": 2499 + }, + { + "epoch": 2.023677021147425, + "grad_norm": 1.785500168800354, + "learning_rate": 5.0556320252758676e-06, + "loss": 0.1031, + "step": 2500 + }, + { + "epoch": 2.024486491955884, + "grad_norm": 2.2368059158325195, + "learning_rate": 5.048033818830227e-06, + "loss": 0.0991, + "step": 2501 + }, + { + "epoch": 2.0252959627643428, + "grad_norm": 1.770681619644165, + "learning_rate": 5.040439398230673e-06, + "loss": 0.1004, + "step": 2502 + }, + { + "epoch": 2.026105433572802, + "grad_norm": 1.8625531196594238, + "learning_rate": 5.032848769283238e-06, + "loss": 0.1009, + "step": 2503 + }, + { + "epoch": 2.0269149043812607, + "grad_norm": 1.746651530265808, + "learning_rate": 5.025261937791068e-06, + "loss": 0.0952, + "step": 2504 + }, + { + "epoch": 2.02772437518972, + "grad_norm": 1.7430765628814697, + "learning_rate": 5.017678909554397e-06, + "loss": 0.108, + "step": 2505 + }, + { + "epoch": 2.0285338459981785, + "grad_norm": 1.6383172273635864, + "learning_rate": 5.0100996903705515e-06, + "loss": 0.0906, + "step": 2506 + }, + { + "epoch": 2.0293433168066377, + "grad_norm": 1.741434097290039, + "learning_rate": 5.002524286033948e-06, + "loss": 0.1007, + "step": 2507 + }, + { + "epoch": 2.0301527876150964, + "grad_norm": 1.7886006832122803, + "learning_rate": 4.994952702336095e-06, + "loss": 0.1076, + "step": 2508 + }, + { + "epoch": 2.0309622584235556, + "grad_norm": 1.7242480516433716, + "learning_rate": 4.987384945065568e-06, + "loss": 0.0923, + "step": 2509 + }, + { + "epoch": 2.0317717292320148, + "grad_norm": 1.694351077079773, + "learning_rate": 4.979821020008019e-06, + "loss": 0.1013, + "step": 2510 + }, + { + "epoch": 2.0325812000404735, + "grad_norm": 2.0034894943237305, + "learning_rate": 4.9722609329461705e-06, + "loss": 0.1054, + "step": 2511 + }, + { + "epoch": 2.0333906708489327, + "grad_norm": 1.8571476936340332, + "learning_rate": 4.964704689659818e-06, + "loss": 0.0958, + "step": 2512 + }, + { + "epoch": 2.0342001416573914, + "grad_norm": 1.9573204517364502, + "learning_rate": 4.95715229592581e-06, + "loss": 0.1008, + "step": 2513 + }, + { + "epoch": 2.0350096124658505, + "grad_norm": 2.020095109939575, + "learning_rate": 4.949603757518059e-06, + "loss": 0.0981, + "step": 2514 + }, + { + "epoch": 2.0358190832743093, + "grad_norm": 1.9765684604644775, + "learning_rate": 4.942059080207518e-06, + "loss": 0.1092, + "step": 2515 + }, + { + "epoch": 2.0366285540827684, + "grad_norm": 1.9841855764389038, + "learning_rate": 4.9345182697622086e-06, + "loss": 0.0883, + "step": 2516 + }, + { + "epoch": 2.037438024891227, + "grad_norm": 1.7726657390594482, + "learning_rate": 4.926981331947175e-06, + "loss": 0.1, + "step": 2517 + }, + { + "epoch": 2.0382474956996863, + "grad_norm": 1.8251235485076904, + "learning_rate": 4.919448272524509e-06, + "loss": 0.0885, + "step": 2518 + }, + { + "epoch": 2.0390569665081455, + "grad_norm": 1.641751766204834, + "learning_rate": 4.9119190972533396e-06, + "loss": 0.0923, + "step": 2519 + }, + { + "epoch": 2.039866437316604, + "grad_norm": 1.9251255989074707, + "learning_rate": 4.904393811889829e-06, + "loss": 0.1023, + "step": 2520 + }, + { + "epoch": 2.0406759081250634, + "grad_norm": 1.84998619556427, + "learning_rate": 4.896872422187158e-06, + "loss": 0.0976, + "step": 2521 + }, + { + "epoch": 2.041485378933522, + "grad_norm": 1.8820695877075195, + "learning_rate": 4.8893549338955316e-06, + "loss": 0.0984, + "step": 2522 + }, + { + "epoch": 2.0422948497419813, + "grad_norm": 1.7917249202728271, + "learning_rate": 4.881841352762169e-06, + "loss": 0.0927, + "step": 2523 + }, + { + "epoch": 2.04310432055044, + "grad_norm": 1.739382028579712, + "learning_rate": 4.87433168453131e-06, + "loss": 0.0923, + "step": 2524 + }, + { + "epoch": 2.043913791358899, + "grad_norm": 1.9256110191345215, + "learning_rate": 4.866825934944203e-06, + "loss": 0.0989, + "step": 2525 + }, + { + "epoch": 2.044723262167358, + "grad_norm": 1.9168742895126343, + "learning_rate": 4.859324109739092e-06, + "loss": 0.0943, + "step": 2526 + }, + { + "epoch": 2.045532732975817, + "grad_norm": 1.8972927331924438, + "learning_rate": 4.851826214651223e-06, + "loss": 0.0903, + "step": 2527 + }, + { + "epoch": 2.0463422037842762, + "grad_norm": 1.7373346090316772, + "learning_rate": 4.844332255412845e-06, + "loss": 0.0986, + "step": 2528 + }, + { + "epoch": 2.047151674592735, + "grad_norm": 1.8570533990859985, + "learning_rate": 4.836842237753187e-06, + "loss": 0.0907, + "step": 2529 + }, + { + "epoch": 2.047961145401194, + "grad_norm": 1.8979943990707397, + "learning_rate": 4.829356167398477e-06, + "loss": 0.0926, + "step": 2530 + }, + { + "epoch": 2.048770616209653, + "grad_norm": 1.8695825338363647, + "learning_rate": 4.821874050071911e-06, + "loss": 0.0977, + "step": 2531 + }, + { + "epoch": 2.049580087018112, + "grad_norm": 1.7912495136260986, + "learning_rate": 4.8143958914936775e-06, + "loss": 0.1035, + "step": 2532 + }, + { + "epoch": 2.0503895578265707, + "grad_norm": 1.9020262956619263, + "learning_rate": 4.806921697380928e-06, + "loss": 0.1026, + "step": 2533 + }, + { + "epoch": 2.05119902863503, + "grad_norm": 1.5442593097686768, + "learning_rate": 4.799451473447787e-06, + "loss": 0.08, + "step": 2534 + }, + { + "epoch": 2.0520084994434886, + "grad_norm": 1.8599036931991577, + "learning_rate": 4.791985225405339e-06, + "loss": 0.1019, + "step": 2535 + }, + { + "epoch": 2.052817970251948, + "grad_norm": 1.8300751447677612, + "learning_rate": 4.784522958961636e-06, + "loss": 0.1085, + "step": 2536 + }, + { + "epoch": 2.053627441060407, + "grad_norm": 1.6940358877182007, + "learning_rate": 4.777064679821688e-06, + "loss": 0.096, + "step": 2537 + }, + { + "epoch": 2.0544369118688657, + "grad_norm": 1.8170047998428345, + "learning_rate": 4.769610393687447e-06, + "loss": 0.0895, + "step": 2538 + }, + { + "epoch": 2.055246382677325, + "grad_norm": 2.0196962356567383, + "learning_rate": 4.762160106257815e-06, + "loss": 0.099, + "step": 2539 + }, + { + "epoch": 2.0560558534857836, + "grad_norm": 1.827735185623169, + "learning_rate": 4.754713823228644e-06, + "loss": 0.1025, + "step": 2540 + }, + { + "epoch": 2.0568653242942427, + "grad_norm": 1.7728147506713867, + "learning_rate": 4.747271550292713e-06, + "loss": 0.0973, + "step": 2541 + }, + { + "epoch": 2.0576747951027015, + "grad_norm": 1.8611178398132324, + "learning_rate": 4.739833293139746e-06, + "loss": 0.1074, + "step": 2542 + }, + { + "epoch": 2.0584842659111606, + "grad_norm": 2.250262498855591, + "learning_rate": 4.7323990574563975e-06, + "loss": 0.0953, + "step": 2543 + }, + { + "epoch": 2.0592937367196193, + "grad_norm": 1.8677440881729126, + "learning_rate": 4.724968848926236e-06, + "loss": 0.0965, + "step": 2544 + }, + { + "epoch": 2.0601032075280785, + "grad_norm": 1.7938926219940186, + "learning_rate": 4.7175426732297605e-06, + "loss": 0.0921, + "step": 2545 + }, + { + "epoch": 2.0609126783365377, + "grad_norm": 1.8761329650878906, + "learning_rate": 4.710120536044378e-06, + "loss": 0.1005, + "step": 2546 + }, + { + "epoch": 2.0617221491449964, + "grad_norm": 1.866632342338562, + "learning_rate": 4.7027024430444205e-06, + "loss": 0.0927, + "step": 2547 + }, + { + "epoch": 2.0625316199534556, + "grad_norm": 1.8003506660461426, + "learning_rate": 4.695288399901123e-06, + "loss": 0.0927, + "step": 2548 + }, + { + "epoch": 2.0633410907619143, + "grad_norm": 1.8278974294662476, + "learning_rate": 4.687878412282622e-06, + "loss": 0.0887, + "step": 2549 + }, + { + "epoch": 2.0641505615703735, + "grad_norm": 2.1157047748565674, + "learning_rate": 4.680472485853951e-06, + "loss": 0.1159, + "step": 2550 + }, + { + "epoch": 2.064960032378832, + "grad_norm": 2.076115608215332, + "learning_rate": 4.673070626277047e-06, + "loss": 0.0916, + "step": 2551 + }, + { + "epoch": 2.0657695031872914, + "grad_norm": 1.7969961166381836, + "learning_rate": 4.665672839210728e-06, + "loss": 0.093, + "step": 2552 + }, + { + "epoch": 2.06657897399575, + "grad_norm": 1.8319323062896729, + "learning_rate": 4.65827913031071e-06, + "loss": 0.0937, + "step": 2553 + }, + { + "epoch": 2.0673884448042092, + "grad_norm": 1.7355996370315552, + "learning_rate": 4.6508895052295774e-06, + "loss": 0.0958, + "step": 2554 + }, + { + "epoch": 2.0681979156126684, + "grad_norm": 1.951331377029419, + "learning_rate": 4.643503969616809e-06, + "loss": 0.0902, + "step": 2555 + }, + { + "epoch": 2.069007386421127, + "grad_norm": 1.9663786888122559, + "learning_rate": 4.636122529118744e-06, + "loss": 0.0843, + "step": 2556 + }, + { + "epoch": 2.0698168572295863, + "grad_norm": 1.7957744598388672, + "learning_rate": 4.628745189378594e-06, + "loss": 0.0975, + "step": 2557 + }, + { + "epoch": 2.070626328038045, + "grad_norm": 1.8923978805541992, + "learning_rate": 4.621371956036432e-06, + "loss": 0.0929, + "step": 2558 + }, + { + "epoch": 2.071435798846504, + "grad_norm": 2.0479109287261963, + "learning_rate": 4.614002834729202e-06, + "loss": 0.1024, + "step": 2559 + }, + { + "epoch": 2.072245269654963, + "grad_norm": 1.8317805528640747, + "learning_rate": 4.606637831090701e-06, + "loss": 0.0886, + "step": 2560 + }, + { + "epoch": 2.073054740463422, + "grad_norm": 2.0156784057617188, + "learning_rate": 4.599276950751571e-06, + "loss": 0.0993, + "step": 2561 + }, + { + "epoch": 2.073864211271881, + "grad_norm": 1.8082611560821533, + "learning_rate": 4.591920199339303e-06, + "loss": 0.0906, + "step": 2562 + }, + { + "epoch": 2.07467368208034, + "grad_norm": 2.030212163925171, + "learning_rate": 4.584567582478241e-06, + "loss": 0.1042, + "step": 2563 + }, + { + "epoch": 2.075483152888799, + "grad_norm": 1.9346489906311035, + "learning_rate": 4.577219105789554e-06, + "loss": 0.098, + "step": 2564 + }, + { + "epoch": 2.076292623697258, + "grad_norm": 1.8094979524612427, + "learning_rate": 4.569874774891263e-06, + "loss": 0.1021, + "step": 2565 + }, + { + "epoch": 2.077102094505717, + "grad_norm": 1.8523356914520264, + "learning_rate": 4.562534595398199e-06, + "loss": 0.0968, + "step": 2566 + }, + { + "epoch": 2.0779115653141758, + "grad_norm": 1.8476094007492065, + "learning_rate": 4.555198572922041e-06, + "loss": 0.0985, + "step": 2567 + }, + { + "epoch": 2.078721036122635, + "grad_norm": 1.9571882486343384, + "learning_rate": 4.547866713071271e-06, + "loss": 0.1142, + "step": 2568 + }, + { + "epoch": 2.0795305069310936, + "grad_norm": 1.8949851989746094, + "learning_rate": 4.540539021451196e-06, + "loss": 0.0952, + "step": 2569 + }, + { + "epoch": 2.080339977739553, + "grad_norm": 1.8484987020492554, + "learning_rate": 4.533215503663938e-06, + "loss": 0.0898, + "step": 2570 + }, + { + "epoch": 2.0811494485480115, + "grad_norm": 2.023568630218506, + "learning_rate": 4.525896165308433e-06, + "loss": 0.1044, + "step": 2571 + }, + { + "epoch": 2.0819589193564707, + "grad_norm": 1.8515580892562866, + "learning_rate": 4.518581011980411e-06, + "loss": 0.0857, + "step": 2572 + }, + { + "epoch": 2.08276839016493, + "grad_norm": 1.8924628496170044, + "learning_rate": 4.511270049272406e-06, + "loss": 0.0885, + "step": 2573 + }, + { + "epoch": 2.0835778609733886, + "grad_norm": 1.9573137760162354, + "learning_rate": 4.503963282773746e-06, + "loss": 0.0984, + "step": 2574 + }, + { + "epoch": 2.0843873317818478, + "grad_norm": 2.0727148056030273, + "learning_rate": 4.496660718070559e-06, + "loss": 0.0955, + "step": 2575 + }, + { + "epoch": 2.0851968025903065, + "grad_norm": 1.7888051271438599, + "learning_rate": 4.489362360745758e-06, + "loss": 0.093, + "step": 2576 + }, + { + "epoch": 2.0860062733987657, + "grad_norm": 1.6437965631484985, + "learning_rate": 4.482068216379035e-06, + "loss": 0.0884, + "step": 2577 + }, + { + "epoch": 2.0868157442072244, + "grad_norm": 1.8353931903839111, + "learning_rate": 4.4747782905468594e-06, + "loss": 0.0969, + "step": 2578 + }, + { + "epoch": 2.0876252150156835, + "grad_norm": 1.9645878076553345, + "learning_rate": 4.467492588822487e-06, + "loss": 0.1068, + "step": 2579 + }, + { + "epoch": 2.0884346858241423, + "grad_norm": 1.809080719947815, + "learning_rate": 4.460211116775932e-06, + "loss": 0.0839, + "step": 2580 + }, + { + "epoch": 2.0892441566326014, + "grad_norm": 1.945007085800171, + "learning_rate": 4.452933879973978e-06, + "loss": 0.095, + "step": 2581 + }, + { + "epoch": 2.0900536274410606, + "grad_norm": 1.7426354885101318, + "learning_rate": 4.445660883980174e-06, + "loss": 0.088, + "step": 2582 + }, + { + "epoch": 2.0908630982495193, + "grad_norm": 1.9065150022506714, + "learning_rate": 4.43839213435483e-06, + "loss": 0.0953, + "step": 2583 + }, + { + "epoch": 2.0916725690579785, + "grad_norm": 2.072256565093994, + "learning_rate": 4.431127636655e-06, + "loss": 0.1184, + "step": 2584 + }, + { + "epoch": 2.092482039866437, + "grad_norm": 1.8454228639602661, + "learning_rate": 4.423867396434493e-06, + "loss": 0.0957, + "step": 2585 + }, + { + "epoch": 2.0932915106748964, + "grad_norm": 1.9113198518753052, + "learning_rate": 4.41661141924386e-06, + "loss": 0.0974, + "step": 2586 + }, + { + "epoch": 2.094100981483355, + "grad_norm": 1.7629326581954956, + "learning_rate": 4.409359710630396e-06, + "loss": 0.0893, + "step": 2587 + }, + { + "epoch": 2.0949104522918143, + "grad_norm": 2.033082962036133, + "learning_rate": 4.402112276138137e-06, + "loss": 0.1009, + "step": 2588 + }, + { + "epoch": 2.095719923100273, + "grad_norm": 1.9365968704223633, + "learning_rate": 4.394869121307843e-06, + "loss": 0.0992, + "step": 2589 + }, + { + "epoch": 2.096529393908732, + "grad_norm": 1.9708460569381714, + "learning_rate": 4.387630251677001e-06, + "loss": 0.0958, + "step": 2590 + }, + { + "epoch": 2.0973388647171913, + "grad_norm": 2.170088052749634, + "learning_rate": 4.380395672779833e-06, + "loss": 0.0889, + "step": 2591 + }, + { + "epoch": 2.09814833552565, + "grad_norm": 1.7374820709228516, + "learning_rate": 4.373165390147266e-06, + "loss": 0.0902, + "step": 2592 + }, + { + "epoch": 2.098957806334109, + "grad_norm": 1.8062101602554321, + "learning_rate": 4.365939409306954e-06, + "loss": 0.0841, + "step": 2593 + }, + { + "epoch": 2.099767277142568, + "grad_norm": 1.7921723127365112, + "learning_rate": 4.358717735783264e-06, + "loss": 0.096, + "step": 2594 + }, + { + "epoch": 2.100576747951027, + "grad_norm": 1.8842366933822632, + "learning_rate": 4.35150037509726e-06, + "loss": 0.1012, + "step": 2595 + }, + { + "epoch": 2.101386218759486, + "grad_norm": 1.7958980798721313, + "learning_rate": 4.344287332766712e-06, + "loss": 0.0944, + "step": 2596 + }, + { + "epoch": 2.102195689567945, + "grad_norm": 1.9543617963790894, + "learning_rate": 4.337078614306087e-06, + "loss": 0.0921, + "step": 2597 + }, + { + "epoch": 2.1030051603764037, + "grad_norm": 1.9840435981750488, + "learning_rate": 4.329874225226553e-06, + "loss": 0.0967, + "step": 2598 + }, + { + "epoch": 2.103814631184863, + "grad_norm": 1.8136100769042969, + "learning_rate": 4.322674171035969e-06, + "loss": 0.0959, + "step": 2599 + }, + { + "epoch": 2.104624101993322, + "grad_norm": 1.9529733657836914, + "learning_rate": 4.315478457238869e-06, + "loss": 0.0963, + "step": 2600 + }, + { + "epoch": 2.105433572801781, + "grad_norm": 2.2378573417663574, + "learning_rate": 4.308287089336473e-06, + "loss": 0.1017, + "step": 2601 + }, + { + "epoch": 2.10624304361024, + "grad_norm": 1.9880894422531128, + "learning_rate": 4.301100072826687e-06, + "loss": 0.1083, + "step": 2602 + }, + { + "epoch": 2.1070525144186987, + "grad_norm": 1.8294721841812134, + "learning_rate": 4.293917413204081e-06, + "loss": 0.0899, + "step": 2603 + }, + { + "epoch": 2.107861985227158, + "grad_norm": 1.875686526298523, + "learning_rate": 4.286739115959893e-06, + "loss": 0.0989, + "step": 2604 + }, + { + "epoch": 2.1086714560356166, + "grad_norm": 1.8923535346984863, + "learning_rate": 4.279565186582032e-06, + "loss": 0.0945, + "step": 2605 + }, + { + "epoch": 2.1094809268440757, + "grad_norm": 1.8434314727783203, + "learning_rate": 4.27239563055507e-06, + "loss": 0.0918, + "step": 2606 + }, + { + "epoch": 2.1102903976525345, + "grad_norm": 1.6979683637619019, + "learning_rate": 4.265230453360228e-06, + "loss": 0.0869, + "step": 2607 + }, + { + "epoch": 2.1110998684609936, + "grad_norm": 1.8444938659667969, + "learning_rate": 4.258069660475382e-06, + "loss": 0.0895, + "step": 2608 + }, + { + "epoch": 2.111909339269453, + "grad_norm": 2.0379316806793213, + "learning_rate": 4.250913257375052e-06, + "loss": 0.1058, + "step": 2609 + }, + { + "epoch": 2.1127188100779115, + "grad_norm": 1.9056406021118164, + "learning_rate": 4.243761249530411e-06, + "loss": 0.0962, + "step": 2610 + }, + { + "epoch": 2.1135282808863707, + "grad_norm": 2.178107261657715, + "learning_rate": 4.236613642409269e-06, + "loss": 0.1065, + "step": 2611 + }, + { + "epoch": 2.1143377516948294, + "grad_norm": 1.8277605772018433, + "learning_rate": 4.229470441476067e-06, + "loss": 0.0913, + "step": 2612 + }, + { + "epoch": 2.1151472225032886, + "grad_norm": 1.7604845762252808, + "learning_rate": 4.222331652191878e-06, + "loss": 0.0877, + "step": 2613 + }, + { + "epoch": 2.1159566933117473, + "grad_norm": 1.9069820642471313, + "learning_rate": 4.215197280014409e-06, + "loss": 0.0906, + "step": 2614 + }, + { + "epoch": 2.1167661641202065, + "grad_norm": 1.8559761047363281, + "learning_rate": 4.20806733039798e-06, + "loss": 0.093, + "step": 2615 + }, + { + "epoch": 2.117575634928665, + "grad_norm": 2.0406460762023926, + "learning_rate": 4.200941808793543e-06, + "loss": 0.0913, + "step": 2616 + }, + { + "epoch": 2.1183851057371244, + "grad_norm": 1.6208300590515137, + "learning_rate": 4.193820720648647e-06, + "loss": 0.0832, + "step": 2617 + }, + { + "epoch": 2.1191945765455835, + "grad_norm": 1.9039509296417236, + "learning_rate": 4.186704071407471e-06, + "loss": 0.0922, + "step": 2618 + }, + { + "epoch": 2.1200040473540422, + "grad_norm": 1.997240424156189, + "learning_rate": 4.1795918665107885e-06, + "loss": 0.0894, + "step": 2619 + }, + { + "epoch": 2.1208135181625014, + "grad_norm": 2.054901123046875, + "learning_rate": 4.172484111395972e-06, + "loss": 0.1045, + "step": 2620 + }, + { + "epoch": 2.12162298897096, + "grad_norm": 1.9834610223770142, + "learning_rate": 4.165380811497001e-06, + "loss": 0.093, + "step": 2621 + }, + { + "epoch": 2.1224324597794193, + "grad_norm": 1.860414743423462, + "learning_rate": 4.15828197224445e-06, + "loss": 0.0849, + "step": 2622 + }, + { + "epoch": 2.123241930587878, + "grad_norm": 1.837340235710144, + "learning_rate": 4.151187599065477e-06, + "loss": 0.0965, + "step": 2623 + }, + { + "epoch": 2.124051401396337, + "grad_norm": 1.7730510234832764, + "learning_rate": 4.144097697383826e-06, + "loss": 0.0842, + "step": 2624 + }, + { + "epoch": 2.124860872204796, + "grad_norm": 1.947758674621582, + "learning_rate": 4.137012272619821e-06, + "loss": 0.0891, + "step": 2625 + }, + { + "epoch": 2.125670343013255, + "grad_norm": 1.907871127128601, + "learning_rate": 4.129931330190373e-06, + "loss": 0.0947, + "step": 2626 + }, + { + "epoch": 2.1264798138217142, + "grad_norm": 2.1840672492980957, + "learning_rate": 4.122854875508952e-06, + "loss": 0.0973, + "step": 2627 + }, + { + "epoch": 2.127289284630173, + "grad_norm": 1.7846566438674927, + "learning_rate": 4.115782913985614e-06, + "loss": 0.0924, + "step": 2628 + }, + { + "epoch": 2.128098755438632, + "grad_norm": 1.8572288751602173, + "learning_rate": 4.108715451026962e-06, + "loss": 0.0845, + "step": 2629 + }, + { + "epoch": 2.128908226247091, + "grad_norm": 1.9222453832626343, + "learning_rate": 4.101652492036177e-06, + "loss": 0.0904, + "step": 2630 + }, + { + "epoch": 2.12971769705555, + "grad_norm": 1.9972676038742065, + "learning_rate": 4.094594042412982e-06, + "loss": 0.0984, + "step": 2631 + }, + { + "epoch": 2.1305271678640088, + "grad_norm": 1.9524338245391846, + "learning_rate": 4.087540107553656e-06, + "loss": 0.0983, + "step": 2632 + }, + { + "epoch": 2.131336638672468, + "grad_norm": 1.90904700756073, + "learning_rate": 4.080490692851033e-06, + "loss": 0.0981, + "step": 2633 + }, + { + "epoch": 2.1321461094809266, + "grad_norm": 1.8760076761245728, + "learning_rate": 4.073445803694492e-06, + "loss": 0.0915, + "step": 2634 + }, + { + "epoch": 2.132955580289386, + "grad_norm": 1.7651512622833252, + "learning_rate": 4.06640544546994e-06, + "loss": 0.0809, + "step": 2635 + }, + { + "epoch": 2.133765051097845, + "grad_norm": 1.9838407039642334, + "learning_rate": 4.059369623559831e-06, + "loss": 0.1038, + "step": 2636 + }, + { + "epoch": 2.1345745219063037, + "grad_norm": 1.905137300491333, + "learning_rate": 4.052338343343143e-06, + "loss": 0.0988, + "step": 2637 + }, + { + "epoch": 2.135383992714763, + "grad_norm": 1.8026025295257568, + "learning_rate": 4.045311610195389e-06, + "loss": 0.0914, + "step": 2638 + }, + { + "epoch": 2.1361934635232216, + "grad_norm": 1.8487508296966553, + "learning_rate": 4.038289429488608e-06, + "loss": 0.0901, + "step": 2639 + }, + { + "epoch": 2.1370029343316808, + "grad_norm": 1.8457449674606323, + "learning_rate": 4.031271806591348e-06, + "loss": 0.0895, + "step": 2640 + }, + { + "epoch": 2.1378124051401395, + "grad_norm": 1.8789719343185425, + "learning_rate": 4.024258746868675e-06, + "loss": 0.0951, + "step": 2641 + }, + { + "epoch": 2.1386218759485986, + "grad_norm": 2.1246821880340576, + "learning_rate": 4.017250255682177e-06, + "loss": 0.0938, + "step": 2642 + }, + { + "epoch": 2.1394313467570574, + "grad_norm": 1.8667241334915161, + "learning_rate": 4.0102463383899345e-06, + "loss": 0.0942, + "step": 2643 + }, + { + "epoch": 2.1402408175655165, + "grad_norm": 2.0159764289855957, + "learning_rate": 4.003247000346542e-06, + "loss": 0.0899, + "step": 2644 + }, + { + "epoch": 2.1410502883739757, + "grad_norm": 1.7406388521194458, + "learning_rate": 3.996252246903093e-06, + "loss": 0.0773, + "step": 2645 + }, + { + "epoch": 2.1418597591824344, + "grad_norm": 2.2176947593688965, + "learning_rate": 3.989262083407169e-06, + "loss": 0.0937, + "step": 2646 + }, + { + "epoch": 2.1426692299908936, + "grad_norm": 1.9978742599487305, + "learning_rate": 3.982276515202843e-06, + "loss": 0.097, + "step": 2647 + }, + { + "epoch": 2.1434787007993523, + "grad_norm": 2.1383368968963623, + "learning_rate": 3.975295547630679e-06, + "loss": 0.1069, + "step": 2648 + }, + { + "epoch": 2.1442881716078115, + "grad_norm": 1.9714694023132324, + "learning_rate": 3.968319186027727e-06, + "loss": 0.0965, + "step": 2649 + }, + { + "epoch": 2.14509764241627, + "grad_norm": 1.910153865814209, + "learning_rate": 3.961347435727504e-06, + "loss": 0.091, + "step": 2650 + }, + { + "epoch": 2.1459071132247294, + "grad_norm": 1.9794328212738037, + "learning_rate": 3.954380302060017e-06, + "loss": 0.0957, + "step": 2651 + }, + { + "epoch": 2.146716584033188, + "grad_norm": 1.8451275825500488, + "learning_rate": 3.947417790351726e-06, + "loss": 0.0916, + "step": 2652 + }, + { + "epoch": 2.1475260548416473, + "grad_norm": 1.9838416576385498, + "learning_rate": 3.940459905925577e-06, + "loss": 0.098, + "step": 2653 + }, + { + "epoch": 2.1483355256501064, + "grad_norm": 1.8887686729431152, + "learning_rate": 3.9335066541009616e-06, + "loss": 0.0883, + "step": 2654 + }, + { + "epoch": 2.149144996458565, + "grad_norm": 1.9989392757415771, + "learning_rate": 3.9265580401937334e-06, + "loss": 0.1061, + "step": 2655 + }, + { + "epoch": 2.1499544672670243, + "grad_norm": 1.8118600845336914, + "learning_rate": 3.919614069516208e-06, + "loss": 0.0924, + "step": 2656 + }, + { + "epoch": 2.150763938075483, + "grad_norm": 1.7273796796798706, + "learning_rate": 3.912674747377149e-06, + "loss": 0.0803, + "step": 2657 + }, + { + "epoch": 2.151573408883942, + "grad_norm": 1.9110602140426636, + "learning_rate": 3.905740079081759e-06, + "loss": 0.0967, + "step": 2658 + }, + { + "epoch": 2.152382879692401, + "grad_norm": 1.8333524465560913, + "learning_rate": 3.898810069931688e-06, + "loss": 0.0932, + "step": 2659 + }, + { + "epoch": 2.15319235050086, + "grad_norm": 1.9060814380645752, + "learning_rate": 3.891884725225019e-06, + "loss": 0.0985, + "step": 2660 + }, + { + "epoch": 2.154001821309319, + "grad_norm": 1.8501023054122925, + "learning_rate": 3.8849640502562765e-06, + "loss": 0.0896, + "step": 2661 + }, + { + "epoch": 2.154811292117778, + "grad_norm": 1.8158103227615356, + "learning_rate": 3.878048050316416e-06, + "loss": 0.0918, + "step": 2662 + }, + { + "epoch": 2.155620762926237, + "grad_norm": 1.9641982316970825, + "learning_rate": 3.87113673069281e-06, + "loss": 0.1022, + "step": 2663 + }, + { + "epoch": 2.156430233734696, + "grad_norm": 2.0094428062438965, + "learning_rate": 3.864230096669254e-06, + "loss": 0.095, + "step": 2664 + }, + { + "epoch": 2.157239704543155, + "grad_norm": 1.788661241531372, + "learning_rate": 3.857328153525972e-06, + "loss": 0.086, + "step": 2665 + }, + { + "epoch": 2.158049175351614, + "grad_norm": 2.038266181945801, + "learning_rate": 3.850430906539586e-06, + "loss": 0.0957, + "step": 2666 + }, + { + "epoch": 2.158858646160073, + "grad_norm": 1.9250860214233398, + "learning_rate": 3.8435383609831435e-06, + "loss": 0.0904, + "step": 2667 + }, + { + "epoch": 2.1596681169685317, + "grad_norm": 1.8413965702056885, + "learning_rate": 3.836650522126082e-06, + "loss": 0.0902, + "step": 2668 + }, + { + "epoch": 2.160477587776991, + "grad_norm": 2.0250656604766846, + "learning_rate": 3.829767395234258e-06, + "loss": 0.0957, + "step": 2669 + }, + { + "epoch": 2.1612870585854496, + "grad_norm": 1.9982463121414185, + "learning_rate": 3.822888985569911e-06, + "loss": 0.1005, + "step": 2670 + }, + { + "epoch": 2.1620965293939087, + "grad_norm": 1.8483494520187378, + "learning_rate": 3.816015298391681e-06, + "loss": 0.0906, + "step": 2671 + }, + { + "epoch": 2.162906000202368, + "grad_norm": 1.9515268802642822, + "learning_rate": 3.8091463389545878e-06, + "loss": 0.0965, + "step": 2672 + }, + { + "epoch": 2.1637154710108266, + "grad_norm": 1.7911432981491089, + "learning_rate": 3.8022821125100607e-06, + "loss": 0.0893, + "step": 2673 + }, + { + "epoch": 2.164524941819286, + "grad_norm": 1.7568262815475464, + "learning_rate": 3.7954226243058878e-06, + "loss": 0.0798, + "step": 2674 + }, + { + "epoch": 2.1653344126277445, + "grad_norm": 1.8316234350204468, + "learning_rate": 3.78856787958624e-06, + "loss": 0.087, + "step": 2675 + }, + { + "epoch": 2.1661438834362037, + "grad_norm": 2.1236515045166016, + "learning_rate": 3.781717883591662e-06, + "loss": 0.0945, + "step": 2676 + }, + { + "epoch": 2.1669533542446624, + "grad_norm": 2.0251522064208984, + "learning_rate": 3.7748726415590763e-06, + "loss": 0.0917, + "step": 2677 + }, + { + "epoch": 2.1677628250531216, + "grad_norm": 1.9179561138153076, + "learning_rate": 3.768032158721756e-06, + "loss": 0.0937, + "step": 2678 + }, + { + "epoch": 2.1685722958615803, + "grad_norm": 1.911206603050232, + "learning_rate": 3.7611964403093537e-06, + "loss": 0.0895, + "step": 2679 + }, + { + "epoch": 2.1693817666700395, + "grad_norm": 1.876673698425293, + "learning_rate": 3.754365491547861e-06, + "loss": 0.0916, + "step": 2680 + }, + { + "epoch": 2.1701912374784986, + "grad_norm": 2.008901357650757, + "learning_rate": 3.74753931765964e-06, + "loss": 0.0982, + "step": 2681 + }, + { + "epoch": 2.1710007082869573, + "grad_norm": 1.8682981729507446, + "learning_rate": 3.7407179238633895e-06, + "loss": 0.0734, + "step": 2682 + }, + { + "epoch": 2.1718101790954165, + "grad_norm": 1.8277274370193481, + "learning_rate": 3.7339013153741566e-06, + "loss": 0.0912, + "step": 2683 + }, + { + "epoch": 2.1726196499038752, + "grad_norm": 1.795392394065857, + "learning_rate": 3.7270894974033344e-06, + "loss": 0.0788, + "step": 2684 + }, + { + "epoch": 2.1734291207123344, + "grad_norm": 1.7404035329818726, + "learning_rate": 3.7202824751586555e-06, + "loss": 0.0824, + "step": 2685 + }, + { + "epoch": 2.174238591520793, + "grad_norm": 1.889543890953064, + "learning_rate": 3.713480253844177e-06, + "loss": 0.0994, + "step": 2686 + }, + { + "epoch": 2.1750480623292523, + "grad_norm": 1.9379239082336426, + "learning_rate": 3.7066828386602925e-06, + "loss": 0.0901, + "step": 2687 + }, + { + "epoch": 2.175857533137711, + "grad_norm": 1.9342714548110962, + "learning_rate": 3.6998902348037134e-06, + "loss": 0.0932, + "step": 2688 + }, + { + "epoch": 2.17666700394617, + "grad_norm": 1.921160101890564, + "learning_rate": 3.6931024474674838e-06, + "loss": 0.0841, + "step": 2689 + }, + { + "epoch": 2.1774764747546294, + "grad_norm": 1.7355329990386963, + "learning_rate": 3.686319481840963e-06, + "loss": 0.0859, + "step": 2690 + }, + { + "epoch": 2.178285945563088, + "grad_norm": 1.9171886444091797, + "learning_rate": 3.6795413431098172e-06, + "loss": 0.0914, + "step": 2691 + }, + { + "epoch": 2.1790954163715472, + "grad_norm": 1.798313021659851, + "learning_rate": 3.672768036456024e-06, + "loss": 0.0871, + "step": 2692 + }, + { + "epoch": 2.179904887180006, + "grad_norm": 1.7953184843063354, + "learning_rate": 3.6659995670578764e-06, + "loss": 0.0956, + "step": 2693 + }, + { + "epoch": 2.180714357988465, + "grad_norm": 1.9229546785354614, + "learning_rate": 3.6592359400899536e-06, + "loss": 0.0914, + "step": 2694 + }, + { + "epoch": 2.181523828796924, + "grad_norm": 1.81778085231781, + "learning_rate": 3.652477160723146e-06, + "loss": 0.0924, + "step": 2695 + }, + { + "epoch": 2.182333299605383, + "grad_norm": 1.698646903038025, + "learning_rate": 3.645723234124636e-06, + "loss": 0.0801, + "step": 2696 + }, + { + "epoch": 2.1831427704138417, + "grad_norm": 1.9736353158950806, + "learning_rate": 3.638974165457891e-06, + "loss": 0.1028, + "step": 2697 + }, + { + "epoch": 2.183952241222301, + "grad_norm": 1.8051334619522095, + "learning_rate": 3.6322299598826647e-06, + "loss": 0.0854, + "step": 2698 + }, + { + "epoch": 2.18476171203076, + "grad_norm": 1.912760853767395, + "learning_rate": 3.6254906225549924e-06, + "loss": 0.0926, + "step": 2699 + }, + { + "epoch": 2.185571182839219, + "grad_norm": 1.96225106716156, + "learning_rate": 3.618756158627197e-06, + "loss": 0.0932, + "step": 2700 + }, + { + "epoch": 2.186380653647678, + "grad_norm": 1.8408656120300293, + "learning_rate": 3.612026573247861e-06, + "loss": 0.0944, + "step": 2701 + }, + { + "epoch": 2.1871901244561367, + "grad_norm": 1.884421944618225, + "learning_rate": 3.6053018715618515e-06, + "loss": 0.0819, + "step": 2702 + }, + { + "epoch": 2.187999595264596, + "grad_norm": 2.047415018081665, + "learning_rate": 3.5985820587102893e-06, + "loss": 0.091, + "step": 2703 + }, + { + "epoch": 2.1888090660730546, + "grad_norm": 1.9923880100250244, + "learning_rate": 3.5918671398305694e-06, + "loss": 0.0967, + "step": 2704 + }, + { + "epoch": 2.1896185368815138, + "grad_norm": 1.945486307144165, + "learning_rate": 3.5851571200563373e-06, + "loss": 0.0909, + "step": 2705 + }, + { + "epoch": 2.1904280076899725, + "grad_norm": 2.044583320617676, + "learning_rate": 3.5784520045174898e-06, + "loss": 0.0979, + "step": 2706 + }, + { + "epoch": 2.1912374784984316, + "grad_norm": 1.8966721296310425, + "learning_rate": 3.571751798340185e-06, + "loss": 0.0814, + "step": 2707 + }, + { + "epoch": 2.192046949306891, + "grad_norm": 2.010061264038086, + "learning_rate": 3.5650565066468244e-06, + "loss": 0.0928, + "step": 2708 + }, + { + "epoch": 2.1928564201153495, + "grad_norm": 1.891133189201355, + "learning_rate": 3.558366134556048e-06, + "loss": 0.0864, + "step": 2709 + }, + { + "epoch": 2.1936658909238087, + "grad_norm": 2.076646327972412, + "learning_rate": 3.551680687182737e-06, + "loss": 0.0988, + "step": 2710 + }, + { + "epoch": 2.1944753617322674, + "grad_norm": 2.0097086429595947, + "learning_rate": 3.545000169638003e-06, + "loss": 0.0971, + "step": 2711 + }, + { + "epoch": 2.1952848325407266, + "grad_norm": 2.124657392501831, + "learning_rate": 3.538324587029198e-06, + "loss": 0.0907, + "step": 2712 + }, + { + "epoch": 2.1960943033491853, + "grad_norm": 1.8047164678573608, + "learning_rate": 3.5316539444599006e-06, + "loss": 0.0827, + "step": 2713 + }, + { + "epoch": 2.1969037741576445, + "grad_norm": 2.031362533569336, + "learning_rate": 3.5249882470299023e-06, + "loss": 0.0866, + "step": 2714 + }, + { + "epoch": 2.197713244966103, + "grad_norm": 1.8846752643585205, + "learning_rate": 3.51832749983522e-06, + "loss": 0.0845, + "step": 2715 + }, + { + "epoch": 2.1985227157745624, + "grad_norm": 2.024038314819336, + "learning_rate": 3.5116717079680906e-06, + "loss": 0.0957, + "step": 2716 + }, + { + "epoch": 2.1993321865830215, + "grad_norm": 1.97605299949646, + "learning_rate": 3.5050208765169533e-06, + "loss": 0.0887, + "step": 2717 + }, + { + "epoch": 2.2001416573914803, + "grad_norm": 2.270236015319824, + "learning_rate": 3.498375010566466e-06, + "loss": 0.0925, + "step": 2718 + }, + { + "epoch": 2.2009511281999394, + "grad_norm": 1.888261079788208, + "learning_rate": 3.4917341151974783e-06, + "loss": 0.0827, + "step": 2719 + }, + { + "epoch": 2.201760599008398, + "grad_norm": 1.8780378103256226, + "learning_rate": 3.485098195487051e-06, + "loss": 0.0929, + "step": 2720 + }, + { + "epoch": 2.2025700698168573, + "grad_norm": 1.99210524559021, + "learning_rate": 3.478467256508434e-06, + "loss": 0.0989, + "step": 2721 + }, + { + "epoch": 2.203379540625316, + "grad_norm": 1.9381688833236694, + "learning_rate": 3.4718413033310704e-06, + "loss": 0.0962, + "step": 2722 + }, + { + "epoch": 2.204189011433775, + "grad_norm": 2.0248677730560303, + "learning_rate": 3.4652203410205887e-06, + "loss": 0.0937, + "step": 2723 + }, + { + "epoch": 2.204998482242234, + "grad_norm": 1.9394561052322388, + "learning_rate": 3.4586043746388087e-06, + "loss": 0.0919, + "step": 2724 + }, + { + "epoch": 2.205807953050693, + "grad_norm": 1.8317484855651855, + "learning_rate": 3.4519934092437323e-06, + "loss": 0.0925, + "step": 2725 + }, + { + "epoch": 2.2066174238591523, + "grad_norm": 2.0071353912353516, + "learning_rate": 3.445387449889528e-06, + "loss": 0.0885, + "step": 2726 + }, + { + "epoch": 2.207426894667611, + "grad_norm": 2.1794867515563965, + "learning_rate": 3.438786501626541e-06, + "loss": 0.0902, + "step": 2727 + }, + { + "epoch": 2.20823636547607, + "grad_norm": 1.8351812362670898, + "learning_rate": 3.4321905695012914e-06, + "loss": 0.082, + "step": 2728 + }, + { + "epoch": 2.209045836284529, + "grad_norm": 1.8412020206451416, + "learning_rate": 3.4255996585564546e-06, + "loss": 0.0911, + "step": 2729 + }, + { + "epoch": 2.209855307092988, + "grad_norm": 1.9055275917053223, + "learning_rate": 3.419013773830877e-06, + "loss": 0.0897, + "step": 2730 + }, + { + "epoch": 2.2106647779014468, + "grad_norm": 1.9117176532745361, + "learning_rate": 3.412432920359554e-06, + "loss": 0.0952, + "step": 2731 + }, + { + "epoch": 2.211474248709906, + "grad_norm": 2.463092088699341, + "learning_rate": 3.405857103173644e-06, + "loss": 0.0996, + "step": 2732 + }, + { + "epoch": 2.2122837195183647, + "grad_norm": 1.9702363014221191, + "learning_rate": 3.399286327300445e-06, + "loss": 0.0869, + "step": 2733 + }, + { + "epoch": 2.213093190326824, + "grad_norm": 1.7708128690719604, + "learning_rate": 3.3927205977634035e-06, + "loss": 0.0784, + "step": 2734 + }, + { + "epoch": 2.213902661135283, + "grad_norm": 1.78065025806427, + "learning_rate": 3.3861599195821127e-06, + "loss": 0.0916, + "step": 2735 + }, + { + "epoch": 2.2147121319437417, + "grad_norm": 1.9055955410003662, + "learning_rate": 3.3796042977723043e-06, + "loss": 0.0892, + "step": 2736 + }, + { + "epoch": 2.215521602752201, + "grad_norm": 2.074922561645508, + "learning_rate": 3.3730537373458374e-06, + "loss": 0.1014, + "step": 2737 + }, + { + "epoch": 2.2163310735606596, + "grad_norm": 1.8657325506210327, + "learning_rate": 3.3665082433107055e-06, + "loss": 0.0863, + "step": 2738 + }, + { + "epoch": 2.217140544369119, + "grad_norm": 1.854893684387207, + "learning_rate": 3.359967820671025e-06, + "loss": 0.0857, + "step": 2739 + }, + { + "epoch": 2.2179500151775775, + "grad_norm": 1.7861332893371582, + "learning_rate": 3.353432474427043e-06, + "loss": 0.0924, + "step": 2740 + }, + { + "epoch": 2.2187594859860367, + "grad_norm": 2.0864531993865967, + "learning_rate": 3.3469022095751235e-06, + "loss": 0.0842, + "step": 2741 + }, + { + "epoch": 2.2195689567944954, + "grad_norm": 1.8364391326904297, + "learning_rate": 3.340377031107741e-06, + "loss": 0.0853, + "step": 2742 + }, + { + "epoch": 2.2203784276029546, + "grad_norm": 1.9386601448059082, + "learning_rate": 3.3338569440134806e-06, + "loss": 0.0975, + "step": 2743 + }, + { + "epoch": 2.2211878984114137, + "grad_norm": 1.9513109922409058, + "learning_rate": 3.3273419532770458e-06, + "loss": 0.0938, + "step": 2744 + }, + { + "epoch": 2.2219973692198725, + "grad_norm": 1.9756001234054565, + "learning_rate": 3.320832063879231e-06, + "loss": 0.0965, + "step": 2745 + }, + { + "epoch": 2.2228068400283316, + "grad_norm": 2.0334572792053223, + "learning_rate": 3.314327280796934e-06, + "loss": 0.0967, + "step": 2746 + }, + { + "epoch": 2.2236163108367903, + "grad_norm": 1.825202226638794, + "learning_rate": 3.3078276090031535e-06, + "loss": 0.0904, + "step": 2747 + }, + { + "epoch": 2.2244257816452495, + "grad_norm": 1.8933473825454712, + "learning_rate": 3.301333053466982e-06, + "loss": 0.0941, + "step": 2748 + }, + { + "epoch": 2.2252352524537082, + "grad_norm": 1.894253134727478, + "learning_rate": 3.2948436191535926e-06, + "loss": 0.0963, + "step": 2749 + }, + { + "epoch": 2.2260447232621674, + "grad_norm": 1.7676235437393188, + "learning_rate": 3.288359311024243e-06, + "loss": 0.0878, + "step": 2750 + }, + { + "epoch": 2.226854194070626, + "grad_norm": 1.9981290102005005, + "learning_rate": 3.2818801340362827e-06, + "loss": 0.0884, + "step": 2751 + }, + { + "epoch": 2.2276636648790853, + "grad_norm": 1.9764878749847412, + "learning_rate": 3.275406093143124e-06, + "loss": 0.0872, + "step": 2752 + }, + { + "epoch": 2.2284731356875445, + "grad_norm": 1.8420839309692383, + "learning_rate": 3.268937193294268e-06, + "loss": 0.0882, + "step": 2753 + }, + { + "epoch": 2.229282606496003, + "grad_norm": 1.8907479047775269, + "learning_rate": 3.2624734394352707e-06, + "loss": 0.0886, + "step": 2754 + }, + { + "epoch": 2.2300920773044624, + "grad_norm": 1.9710021018981934, + "learning_rate": 3.256014836507766e-06, + "loss": 0.0913, + "step": 2755 + }, + { + "epoch": 2.230901548112921, + "grad_norm": 1.9730639457702637, + "learning_rate": 3.2495613894494414e-06, + "loss": 0.0881, + "step": 2756 + }, + { + "epoch": 2.2317110189213802, + "grad_norm": 1.94786536693573, + "learning_rate": 3.243113103194043e-06, + "loss": 0.0864, + "step": 2757 + }, + { + "epoch": 2.232520489729839, + "grad_norm": 1.9586559534072876, + "learning_rate": 3.2366699826713787e-06, + "loss": 0.0945, + "step": 2758 + }, + { + "epoch": 2.233329960538298, + "grad_norm": 1.9276880025863647, + "learning_rate": 3.230232032807302e-06, + "loss": 0.0878, + "step": 2759 + }, + { + "epoch": 2.234139431346757, + "grad_norm": 1.877465009689331, + "learning_rate": 3.223799258523714e-06, + "loss": 0.0968, + "step": 2760 + }, + { + "epoch": 2.234948902155216, + "grad_norm": 1.9110994338989258, + "learning_rate": 3.217371664738559e-06, + "loss": 0.0916, + "step": 2761 + }, + { + "epoch": 2.235758372963675, + "grad_norm": 1.844465970993042, + "learning_rate": 3.2109492563658153e-06, + "loss": 0.0951, + "step": 2762 + }, + { + "epoch": 2.236567843772134, + "grad_norm": 1.8695050477981567, + "learning_rate": 3.2045320383155053e-06, + "loss": 0.0833, + "step": 2763 + }, + { + "epoch": 2.237377314580593, + "grad_norm": 1.8690173625946045, + "learning_rate": 3.1981200154936853e-06, + "loss": 0.0864, + "step": 2764 + }, + { + "epoch": 2.238186785389052, + "grad_norm": 1.9372035264968872, + "learning_rate": 3.1917131928024314e-06, + "loss": 0.096, + "step": 2765 + }, + { + "epoch": 2.238996256197511, + "grad_norm": 2.2574219703674316, + "learning_rate": 3.185311575139842e-06, + "loss": 0.0975, + "step": 2766 + }, + { + "epoch": 2.2398057270059697, + "grad_norm": 1.9640743732452393, + "learning_rate": 3.1789151674000518e-06, + "loss": 0.0874, + "step": 2767 + }, + { + "epoch": 2.240615197814429, + "grad_norm": 1.8647022247314453, + "learning_rate": 3.172523974473195e-06, + "loss": 0.087, + "step": 2768 + }, + { + "epoch": 2.2414246686228876, + "grad_norm": 1.9332929849624634, + "learning_rate": 3.166138001245426e-06, + "loss": 0.0932, + "step": 2769 + }, + { + "epoch": 2.2422341394313468, + "grad_norm": 1.8353947401046753, + "learning_rate": 3.1597572525989105e-06, + "loss": 0.0798, + "step": 2770 + }, + { + "epoch": 2.243043610239806, + "grad_norm": 2.0981578826904297, + "learning_rate": 3.153381733411822e-06, + "loss": 0.1038, + "step": 2771 + }, + { + "epoch": 2.2438530810482646, + "grad_norm": 1.778098702430725, + "learning_rate": 3.1470114485583304e-06, + "loss": 0.0835, + "step": 2772 + }, + { + "epoch": 2.244662551856724, + "grad_norm": 1.9872926473617554, + "learning_rate": 3.140646402908604e-06, + "loss": 0.0976, + "step": 2773 + }, + { + "epoch": 2.2454720226651825, + "grad_norm": 1.8441587686538696, + "learning_rate": 3.1342866013288053e-06, + "loss": 0.0915, + "step": 2774 + }, + { + "epoch": 2.2462814934736417, + "grad_norm": 1.8429256677627563, + "learning_rate": 3.1279320486810928e-06, + "loss": 0.0878, + "step": 2775 + }, + { + "epoch": 2.2470909642821004, + "grad_norm": 1.8548458814620972, + "learning_rate": 3.121582749823613e-06, + "loss": 0.0943, + "step": 2776 + }, + { + "epoch": 2.2479004350905596, + "grad_norm": 1.8649523258209229, + "learning_rate": 3.1152387096104875e-06, + "loss": 0.0873, + "step": 2777 + }, + { + "epoch": 2.2487099058990183, + "grad_norm": 2.04670786857605, + "learning_rate": 3.10889993289182e-06, + "loss": 0.0957, + "step": 2778 + }, + { + "epoch": 2.2495193767074775, + "grad_norm": 1.9086638689041138, + "learning_rate": 3.102566424513698e-06, + "loss": 0.083, + "step": 2779 + }, + { + "epoch": 2.2503288475159366, + "grad_norm": 1.9705802202224731, + "learning_rate": 3.09623818931817e-06, + "loss": 0.0811, + "step": 2780 + }, + { + "epoch": 2.2511383183243954, + "grad_norm": 1.870055079460144, + "learning_rate": 3.0899152321432647e-06, + "loss": 0.0842, + "step": 2781 + }, + { + "epoch": 2.2519477891328545, + "grad_norm": 1.6333681344985962, + "learning_rate": 3.0835975578229627e-06, + "loss": 0.0778, + "step": 2782 + }, + { + "epoch": 2.2527572599413133, + "grad_norm": 1.8491376638412476, + "learning_rate": 3.0772851711872187e-06, + "loss": 0.0882, + "step": 2783 + }, + { + "epoch": 2.2535667307497724, + "grad_norm": 1.839331030845642, + "learning_rate": 3.0709780770619366e-06, + "loss": 0.0843, + "step": 2784 + }, + { + "epoch": 2.254376201558231, + "grad_norm": 1.8893144130706787, + "learning_rate": 3.064676280268973e-06, + "loss": 0.0845, + "step": 2785 + }, + { + "epoch": 2.2551856723666903, + "grad_norm": 2.014094114303589, + "learning_rate": 3.058379785626142e-06, + "loss": 0.0985, + "step": 2786 + }, + { + "epoch": 2.2559951431751495, + "grad_norm": 1.968076229095459, + "learning_rate": 3.0520885979472026e-06, + "loss": 0.0927, + "step": 2787 + }, + { + "epoch": 2.256804613983608, + "grad_norm": 1.8457478284835815, + "learning_rate": 3.0458027220418494e-06, + "loss": 0.088, + "step": 2788 + }, + { + "epoch": 2.2576140847920674, + "grad_norm": 1.9214904308319092, + "learning_rate": 3.0395221627157234e-06, + "loss": 0.0889, + "step": 2789 + }, + { + "epoch": 2.258423555600526, + "grad_norm": 1.8584851026535034, + "learning_rate": 3.0332469247703934e-06, + "loss": 0.0981, + "step": 2790 + }, + { + "epoch": 2.2592330264089853, + "grad_norm": 1.8711340427398682, + "learning_rate": 3.0269770130033715e-06, + "loss": 0.092, + "step": 2791 + }, + { + "epoch": 2.260042497217444, + "grad_norm": 1.946076512336731, + "learning_rate": 3.0207124322080863e-06, + "loss": 0.0845, + "step": 2792 + }, + { + "epoch": 2.260851968025903, + "grad_norm": 2.193852186203003, + "learning_rate": 3.0144531871739003e-06, + "loss": 0.1022, + "step": 2793 + }, + { + "epoch": 2.261661438834362, + "grad_norm": 1.8088206052780151, + "learning_rate": 3.008199282686086e-06, + "loss": 0.0841, + "step": 2794 + }, + { + "epoch": 2.262470909642821, + "grad_norm": 1.8207679986953735, + "learning_rate": 3.0019507235258473e-06, + "loss": 0.083, + "step": 2795 + }, + { + "epoch": 2.26328038045128, + "grad_norm": 2.1222000122070312, + "learning_rate": 2.9957075144702887e-06, + "loss": 0.088, + "step": 2796 + }, + { + "epoch": 2.264089851259739, + "grad_norm": 2.0200557708740234, + "learning_rate": 2.9894696602924254e-06, + "loss": 0.082, + "step": 2797 + }, + { + "epoch": 2.264899322068198, + "grad_norm": 1.8341455459594727, + "learning_rate": 2.983237165761188e-06, + "loss": 0.0825, + "step": 2798 + }, + { + "epoch": 2.265708792876657, + "grad_norm": 2.120244264602661, + "learning_rate": 2.9770100356414044e-06, + "loss": 0.1044, + "step": 2799 + }, + { + "epoch": 2.266518263685116, + "grad_norm": 1.7892802953720093, + "learning_rate": 2.970788274693801e-06, + "loss": 0.0871, + "step": 2800 + }, + { + "epoch": 2.2673277344935747, + "grad_norm": 2.0893406867980957, + "learning_rate": 2.964571887674993e-06, + "loss": 0.0978, + "step": 2801 + }, + { + "epoch": 2.268137205302034, + "grad_norm": 2.006868362426758, + "learning_rate": 2.958360879337503e-06, + "loss": 0.0939, + "step": 2802 + }, + { + "epoch": 2.2689466761104926, + "grad_norm": 1.8910102844238281, + "learning_rate": 2.9521552544297227e-06, + "loss": 0.0941, + "step": 2803 + }, + { + "epoch": 2.269756146918952, + "grad_norm": 1.837114930152893, + "learning_rate": 2.945955017695946e-06, + "loss": 0.0871, + "step": 2804 + }, + { + "epoch": 2.270565617727411, + "grad_norm": 1.9383083581924438, + "learning_rate": 2.9397601738763314e-06, + "loss": 0.0935, + "step": 2805 + }, + { + "epoch": 2.2713750885358697, + "grad_norm": 1.7246918678283691, + "learning_rate": 2.933570727706929e-06, + "loss": 0.0785, + "step": 2806 + }, + { + "epoch": 2.272184559344329, + "grad_norm": 1.9113712310791016, + "learning_rate": 2.927386683919653e-06, + "loss": 0.098, + "step": 2807 + }, + { + "epoch": 2.2729940301527876, + "grad_norm": 2.0279736518859863, + "learning_rate": 2.921208047242284e-06, + "loss": 0.095, + "step": 2808 + }, + { + "epoch": 2.2738035009612467, + "grad_norm": 1.7939518690109253, + "learning_rate": 2.9150348223984802e-06, + "loss": 0.0834, + "step": 2809 + }, + { + "epoch": 2.2746129717697054, + "grad_norm": 1.8208882808685303, + "learning_rate": 2.908867014107759e-06, + "loss": 0.0876, + "step": 2810 + }, + { + "epoch": 2.2754224425781646, + "grad_norm": 1.8134597539901733, + "learning_rate": 2.9027046270854907e-06, + "loss": 0.0797, + "step": 2811 + }, + { + "epoch": 2.2762319133866233, + "grad_norm": 1.9979907274246216, + "learning_rate": 2.896547666042906e-06, + "loss": 0.0795, + "step": 2812 + }, + { + "epoch": 2.2770413841950825, + "grad_norm": 1.8885722160339355, + "learning_rate": 2.89039613568708e-06, + "loss": 0.0834, + "step": 2813 + }, + { + "epoch": 2.2778508550035417, + "grad_norm": 1.8130698204040527, + "learning_rate": 2.8842500407209526e-06, + "loss": 0.0911, + "step": 2814 + }, + { + "epoch": 2.2786603258120004, + "grad_norm": 1.9274588823318481, + "learning_rate": 2.8781093858432883e-06, + "loss": 0.0845, + "step": 2815 + }, + { + "epoch": 2.2794697966204596, + "grad_norm": 1.9587591886520386, + "learning_rate": 2.8719741757487096e-06, + "loss": 0.0926, + "step": 2816 + }, + { + "epoch": 2.2802792674289183, + "grad_norm": 1.9633991718292236, + "learning_rate": 2.8658444151276623e-06, + "loss": 0.0877, + "step": 2817 + }, + { + "epoch": 2.2810887382373775, + "grad_norm": 1.7725844383239746, + "learning_rate": 2.8597201086664383e-06, + "loss": 0.0862, + "step": 2818 + }, + { + "epoch": 2.281898209045836, + "grad_norm": 1.9035371541976929, + "learning_rate": 2.853601261047151e-06, + "loss": 0.0814, + "step": 2819 + }, + { + "epoch": 2.2827076798542953, + "grad_norm": 1.9928545951843262, + "learning_rate": 2.84748787694774e-06, + "loss": 0.0859, + "step": 2820 + }, + { + "epoch": 2.283517150662754, + "grad_norm": 1.9655340909957886, + "learning_rate": 2.8413799610419756e-06, + "loss": 0.0979, + "step": 2821 + }, + { + "epoch": 2.2843266214712132, + "grad_norm": 1.78617525100708, + "learning_rate": 2.8352775179994454e-06, + "loss": 0.0871, + "step": 2822 + }, + { + "epoch": 2.2851360922796724, + "grad_norm": 2.0127460956573486, + "learning_rate": 2.8291805524855474e-06, + "loss": 0.0918, + "step": 2823 + }, + { + "epoch": 2.285945563088131, + "grad_norm": 1.8558895587921143, + "learning_rate": 2.8230890691614963e-06, + "loss": 0.0848, + "step": 2824 + }, + { + "epoch": 2.2867550338965903, + "grad_norm": 1.9862706661224365, + "learning_rate": 2.817003072684311e-06, + "loss": 0.0806, + "step": 2825 + }, + { + "epoch": 2.287564504705049, + "grad_norm": 1.9072022438049316, + "learning_rate": 2.810922567706823e-06, + "loss": 0.0839, + "step": 2826 + }, + { + "epoch": 2.288373975513508, + "grad_norm": 1.867297887802124, + "learning_rate": 2.804847558877665e-06, + "loss": 0.0898, + "step": 2827 + }, + { + "epoch": 2.289183446321967, + "grad_norm": 1.7279740571975708, + "learning_rate": 2.79877805084126e-06, + "loss": 0.0824, + "step": 2828 + }, + { + "epoch": 2.289992917130426, + "grad_norm": 1.9668251276016235, + "learning_rate": 2.792714048237828e-06, + "loss": 0.0899, + "step": 2829 + }, + { + "epoch": 2.290802387938885, + "grad_norm": 1.7540323734283447, + "learning_rate": 2.786655555703389e-06, + "loss": 0.0838, + "step": 2830 + }, + { + "epoch": 2.291611858747344, + "grad_norm": 1.777409553527832, + "learning_rate": 2.7806025778697364e-06, + "loss": 0.0771, + "step": 2831 + }, + { + "epoch": 2.292421329555803, + "grad_norm": 1.9239623546600342, + "learning_rate": 2.7745551193644613e-06, + "loss": 0.0877, + "step": 2832 + }, + { + "epoch": 2.293230800364262, + "grad_norm": 2.0637497901916504, + "learning_rate": 2.7685131848109225e-06, + "loss": 0.0866, + "step": 2833 + }, + { + "epoch": 2.294040271172721, + "grad_norm": 1.9828383922576904, + "learning_rate": 2.762476778828269e-06, + "loss": 0.0898, + "step": 2834 + }, + { + "epoch": 2.2948497419811797, + "grad_norm": 1.8598941564559937, + "learning_rate": 2.756445906031411e-06, + "loss": 0.0865, + "step": 2835 + }, + { + "epoch": 2.295659212789639, + "grad_norm": 1.6684592962265015, + "learning_rate": 2.7504205710310363e-06, + "loss": 0.0828, + "step": 2836 + }, + { + "epoch": 2.2964686835980976, + "grad_norm": 1.9017168283462524, + "learning_rate": 2.744400778433587e-06, + "loss": 0.0884, + "step": 2837 + }, + { + "epoch": 2.297278154406557, + "grad_norm": 1.973301649093628, + "learning_rate": 2.7383865328412907e-06, + "loss": 0.0964, + "step": 2838 + }, + { + "epoch": 2.2980876252150155, + "grad_norm": 2.151437520980835, + "learning_rate": 2.732377838852115e-06, + "loss": 0.0896, + "step": 2839 + }, + { + "epoch": 2.2988970960234747, + "grad_norm": 1.9263279438018799, + "learning_rate": 2.7263747010597873e-06, + "loss": 0.0892, + "step": 2840 + }, + { + "epoch": 2.299706566831934, + "grad_norm": 1.929074764251709, + "learning_rate": 2.720377124053788e-06, + "loss": 0.0947, + "step": 2841 + }, + { + "epoch": 2.3005160376403926, + "grad_norm": 1.9586818218231201, + "learning_rate": 2.714385112419352e-06, + "loss": 0.0854, + "step": 2842 + }, + { + "epoch": 2.3013255084488518, + "grad_norm": 1.764746069908142, + "learning_rate": 2.708398670737448e-06, + "loss": 0.0839, + "step": 2843 + }, + { + "epoch": 2.3021349792573105, + "grad_norm": 1.8097349405288696, + "learning_rate": 2.7024178035847992e-06, + "loss": 0.0819, + "step": 2844 + }, + { + "epoch": 2.3029444500657696, + "grad_norm": 2.0664327144622803, + "learning_rate": 2.6964425155338547e-06, + "loss": 0.0849, + "step": 2845 + }, + { + "epoch": 2.3037539208742284, + "grad_norm": 2.0515666007995605, + "learning_rate": 2.69047281115281e-06, + "loss": 0.0962, + "step": 2846 + }, + { + "epoch": 2.3045633916826875, + "grad_norm": 1.8852027654647827, + "learning_rate": 2.6845086950055834e-06, + "loss": 0.0854, + "step": 2847 + }, + { + "epoch": 2.3053728624911463, + "grad_norm": 1.9183576107025146, + "learning_rate": 2.6785501716518192e-06, + "loss": 0.0873, + "step": 2848 + }, + { + "epoch": 2.3061823332996054, + "grad_norm": 1.7793123722076416, + "learning_rate": 2.6725972456468953e-06, + "loss": 0.0845, + "step": 2849 + }, + { + "epoch": 2.3069918041080646, + "grad_norm": 1.8519957065582275, + "learning_rate": 2.666649921541906e-06, + "loss": 0.0843, + "step": 2850 + }, + { + "epoch": 2.3078012749165233, + "grad_norm": 2.0787291526794434, + "learning_rate": 2.6607082038836616e-06, + "loss": 0.0938, + "step": 2851 + }, + { + "epoch": 2.3086107457249825, + "grad_norm": 1.7447445392608643, + "learning_rate": 2.6547720972146816e-06, + "loss": 0.0847, + "step": 2852 + }, + { + "epoch": 2.309420216533441, + "grad_norm": 1.852427363395691, + "learning_rate": 2.648841606073208e-06, + "loss": 0.08, + "step": 2853 + }, + { + "epoch": 2.3102296873419004, + "grad_norm": 2.0734732151031494, + "learning_rate": 2.642916734993177e-06, + "loss": 0.0825, + "step": 2854 + }, + { + "epoch": 2.311039158150359, + "grad_norm": 1.776775598526001, + "learning_rate": 2.6369974885042393e-06, + "loss": 0.0812, + "step": 2855 + }, + { + "epoch": 2.3118486289588183, + "grad_norm": 1.9330451488494873, + "learning_rate": 2.6310838711317345e-06, + "loss": 0.0875, + "step": 2856 + }, + { + "epoch": 2.312658099767277, + "grad_norm": 1.780428409576416, + "learning_rate": 2.6251758873967093e-06, + "loss": 0.0791, + "step": 2857 + }, + { + "epoch": 2.313467570575736, + "grad_norm": 1.9802789688110352, + "learning_rate": 2.6192735418158953e-06, + "loss": 0.0796, + "step": 2858 + }, + { + "epoch": 2.3142770413841953, + "grad_norm": 2.0297985076904297, + "learning_rate": 2.6133768389017177e-06, + "loss": 0.0922, + "step": 2859 + }, + { + "epoch": 2.315086512192654, + "grad_norm": 1.8801472187042236, + "learning_rate": 2.607485783162281e-06, + "loss": 0.0804, + "step": 2860 + }, + { + "epoch": 2.315895983001113, + "grad_norm": 1.9306038618087769, + "learning_rate": 2.601600379101389e-06, + "loss": 0.0932, + "step": 2861 + }, + { + "epoch": 2.316705453809572, + "grad_norm": 1.9895156621932983, + "learning_rate": 2.5957206312185078e-06, + "loss": 0.0946, + "step": 2862 + }, + { + "epoch": 2.317514924618031, + "grad_norm": 2.031255006790161, + "learning_rate": 2.589846544008787e-06, + "loss": 0.0841, + "step": 2863 + }, + { + "epoch": 2.31832439542649, + "grad_norm": 1.9931083917617798, + "learning_rate": 2.583978121963042e-06, + "loss": 0.0876, + "step": 2864 + }, + { + "epoch": 2.319133866234949, + "grad_norm": 2.0801737308502197, + "learning_rate": 2.5781153695677684e-06, + "loss": 0.0932, + "step": 2865 + }, + { + "epoch": 2.3199433370434077, + "grad_norm": 2.26285982131958, + "learning_rate": 2.572258291305115e-06, + "loss": 0.089, + "step": 2866 + }, + { + "epoch": 2.320752807851867, + "grad_norm": 2.0050580501556396, + "learning_rate": 2.566406891652905e-06, + "loss": 0.0977, + "step": 2867 + }, + { + "epoch": 2.321562278660326, + "grad_norm": 1.955724835395813, + "learning_rate": 2.560561175084607e-06, + "loss": 0.0859, + "step": 2868 + }, + { + "epoch": 2.3223717494687848, + "grad_norm": 1.7076606750488281, + "learning_rate": 2.554721146069359e-06, + "loss": 0.0815, + "step": 2869 + }, + { + "epoch": 2.323181220277244, + "grad_norm": 1.980772852897644, + "learning_rate": 2.548886809071938e-06, + "loss": 0.0959, + "step": 2870 + }, + { + "epoch": 2.3239906910857027, + "grad_norm": 2.072810173034668, + "learning_rate": 2.543058168552772e-06, + "loss": 0.0836, + "step": 2871 + }, + { + "epoch": 2.324800161894162, + "grad_norm": 1.925207495689392, + "learning_rate": 2.5372352289679413e-06, + "loss": 0.0917, + "step": 2872 + }, + { + "epoch": 2.3256096327026206, + "grad_norm": 1.8515655994415283, + "learning_rate": 2.531417994769165e-06, + "loss": 0.0786, + "step": 2873 + }, + { + "epoch": 2.3264191035110797, + "grad_norm": 1.8863904476165771, + "learning_rate": 2.5256064704037963e-06, + "loss": 0.0813, + "step": 2874 + }, + { + "epoch": 2.3272285743195384, + "grad_norm": 1.9213892221450806, + "learning_rate": 2.5198006603148238e-06, + "loss": 0.0936, + "step": 2875 + }, + { + "epoch": 2.3280380451279976, + "grad_norm": 1.9148366451263428, + "learning_rate": 2.5140005689408676e-06, + "loss": 0.0932, + "step": 2876 + }, + { + "epoch": 2.328847515936457, + "grad_norm": 1.9782136678695679, + "learning_rate": 2.50820620071618e-06, + "loss": 0.0946, + "step": 2877 + }, + { + "epoch": 2.3296569867449155, + "grad_norm": 2.065927028656006, + "learning_rate": 2.502417560070636e-06, + "loss": 0.0821, + "step": 2878 + }, + { + "epoch": 2.3304664575533747, + "grad_norm": 2.0044052600860596, + "learning_rate": 2.4966346514297303e-06, + "loss": 0.0906, + "step": 2879 + }, + { + "epoch": 2.3312759283618334, + "grad_norm": 1.8533389568328857, + "learning_rate": 2.49085747921457e-06, + "loss": 0.089, + "step": 2880 + }, + { + "epoch": 2.3320853991702926, + "grad_norm": 1.9593684673309326, + "learning_rate": 2.4850860478418928e-06, + "loss": 0.0843, + "step": 2881 + }, + { + "epoch": 2.3328948699787513, + "grad_norm": 1.9414558410644531, + "learning_rate": 2.479320361724027e-06, + "loss": 0.0856, + "step": 2882 + }, + { + "epoch": 2.3337043407872105, + "grad_norm": 1.8111214637756348, + "learning_rate": 2.473560425268926e-06, + "loss": 0.0881, + "step": 2883 + }, + { + "epoch": 2.334513811595669, + "grad_norm": 1.793941617012024, + "learning_rate": 2.4678062428801353e-06, + "loss": 0.0923, + "step": 2884 + }, + { + "epoch": 2.3353232824041283, + "grad_norm": 1.8438383340835571, + "learning_rate": 2.4620578189568124e-06, + "loss": 0.0878, + "step": 2885 + }, + { + "epoch": 2.3361327532125875, + "grad_norm": 1.9753698110580444, + "learning_rate": 2.4563151578937006e-06, + "loss": 0.0896, + "step": 2886 + }, + { + "epoch": 2.3369422240210462, + "grad_norm": 1.9809436798095703, + "learning_rate": 2.450578264081146e-06, + "loss": 0.0924, + "step": 2887 + }, + { + "epoch": 2.3377516948295054, + "grad_norm": 1.908180594444275, + "learning_rate": 2.444847141905079e-06, + "loss": 0.0896, + "step": 2888 + }, + { + "epoch": 2.338561165637964, + "grad_norm": 1.9213956594467163, + "learning_rate": 2.439121795747024e-06, + "loss": 0.0899, + "step": 2889 + }, + { + "epoch": 2.3393706364464233, + "grad_norm": 1.9327483177185059, + "learning_rate": 2.4334022299840878e-06, + "loss": 0.089, + "step": 2890 + }, + { + "epoch": 2.340180107254882, + "grad_norm": 1.817255973815918, + "learning_rate": 2.4276884489889564e-06, + "loss": 0.0818, + "step": 2891 + }, + { + "epoch": 2.340989578063341, + "grad_norm": 1.9887306690216064, + "learning_rate": 2.42198045712989e-06, + "loss": 0.0935, + "step": 2892 + }, + { + "epoch": 2.3417990488718, + "grad_norm": 1.9271197319030762, + "learning_rate": 2.4162782587707324e-06, + "loss": 0.0883, + "step": 2893 + }, + { + "epoch": 2.342608519680259, + "grad_norm": 1.889544129371643, + "learning_rate": 2.410581858270886e-06, + "loss": 0.0844, + "step": 2894 + }, + { + "epoch": 2.3434179904887182, + "grad_norm": 1.9806177616119385, + "learning_rate": 2.4048912599853337e-06, + "loss": 0.0987, + "step": 2895 + }, + { + "epoch": 2.344227461297177, + "grad_norm": 1.924872636795044, + "learning_rate": 2.399206468264611e-06, + "loss": 0.0893, + "step": 2896 + }, + { + "epoch": 2.345036932105636, + "grad_norm": 1.9854763746261597, + "learning_rate": 2.393527487454823e-06, + "loss": 0.083, + "step": 2897 + }, + { + "epoch": 2.345846402914095, + "grad_norm": 1.8994393348693848, + "learning_rate": 2.387854321897627e-06, + "loss": 0.0844, + "step": 2898 + }, + { + "epoch": 2.346655873722554, + "grad_norm": 1.7490770816802979, + "learning_rate": 2.382186975930233e-06, + "loss": 0.0861, + "step": 2899 + }, + { + "epoch": 2.3474653445310127, + "grad_norm": 1.9198600053787231, + "learning_rate": 2.3765254538854064e-06, + "loss": 0.0881, + "step": 2900 + }, + { + "epoch": 2.348274815339472, + "grad_norm": 1.885817289352417, + "learning_rate": 2.370869760091463e-06, + "loss": 0.0887, + "step": 2901 + }, + { + "epoch": 2.3490842861479306, + "grad_norm": 1.9150387048721313, + "learning_rate": 2.3652198988722553e-06, + "loss": 0.083, + "step": 2902 + }, + { + "epoch": 2.34989375695639, + "grad_norm": 1.7473728656768799, + "learning_rate": 2.359575874547174e-06, + "loss": 0.0847, + "step": 2903 + }, + { + "epoch": 2.350703227764849, + "grad_norm": 1.7903951406478882, + "learning_rate": 2.3539376914311627e-06, + "loss": 0.0807, + "step": 2904 + }, + { + "epoch": 2.3515126985733077, + "grad_norm": 1.758960247039795, + "learning_rate": 2.3483053538346822e-06, + "loss": 0.0812, + "step": 2905 + }, + { + "epoch": 2.352322169381767, + "grad_norm": 1.784450888633728, + "learning_rate": 2.342678866063737e-06, + "loss": 0.0841, + "step": 2906 + }, + { + "epoch": 2.3531316401902256, + "grad_norm": 1.8086222410202026, + "learning_rate": 2.33705823241985e-06, + "loss": 0.0764, + "step": 2907 + }, + { + "epoch": 2.3539411109986847, + "grad_norm": 1.828744888305664, + "learning_rate": 2.3314434572000766e-06, + "loss": 0.0871, + "step": 2908 + }, + { + "epoch": 2.3547505818071435, + "grad_norm": 1.8538408279418945, + "learning_rate": 2.325834544696989e-06, + "loss": 0.0848, + "step": 2909 + }, + { + "epoch": 2.3555600526156026, + "grad_norm": 2.006896734237671, + "learning_rate": 2.3202314991986752e-06, + "loss": 0.089, + "step": 2910 + }, + { + "epoch": 2.3563695234240614, + "grad_norm": 1.8956083059310913, + "learning_rate": 2.3146343249887393e-06, + "loss": 0.076, + "step": 2911 + }, + { + "epoch": 2.3571789942325205, + "grad_norm": 1.8410786390304565, + "learning_rate": 2.3090430263463004e-06, + "loss": 0.0871, + "step": 2912 + }, + { + "epoch": 2.3579884650409797, + "grad_norm": 1.9447009563446045, + "learning_rate": 2.303457607545985e-06, + "loss": 0.0889, + "step": 2913 + }, + { + "epoch": 2.3587979358494384, + "grad_norm": 1.7803914546966553, + "learning_rate": 2.2978780728579196e-06, + "loss": 0.0837, + "step": 2914 + }, + { + "epoch": 2.3596074066578976, + "grad_norm": 1.8948922157287598, + "learning_rate": 2.2923044265477332e-06, + "loss": 0.0912, + "step": 2915 + }, + { + "epoch": 2.3604168774663563, + "grad_norm": 1.7932016849517822, + "learning_rate": 2.2867366728765606e-06, + "loss": 0.0894, + "step": 2916 + }, + { + "epoch": 2.3612263482748155, + "grad_norm": 2.1542701721191406, + "learning_rate": 2.2811748161010184e-06, + "loss": 0.0861, + "step": 2917 + }, + { + "epoch": 2.362035819083274, + "grad_norm": 1.846718430519104, + "learning_rate": 2.2756188604732298e-06, + "loss": 0.084, + "step": 2918 + }, + { + "epoch": 2.3628452898917334, + "grad_norm": 1.9592006206512451, + "learning_rate": 2.270068810240793e-06, + "loss": 0.0857, + "step": 2919 + }, + { + "epoch": 2.363654760700192, + "grad_norm": 1.858391284942627, + "learning_rate": 2.2645246696468038e-06, + "loss": 0.0965, + "step": 2920 + }, + { + "epoch": 2.3644642315086513, + "grad_norm": 1.938528299331665, + "learning_rate": 2.258986442929829e-06, + "loss": 0.0843, + "step": 2921 + }, + { + "epoch": 2.3652737023171104, + "grad_norm": 2.075718641281128, + "learning_rate": 2.253454134323918e-06, + "loss": 0.0961, + "step": 2922 + }, + { + "epoch": 2.366083173125569, + "grad_norm": 1.7885947227478027, + "learning_rate": 2.247927748058598e-06, + "loss": 0.0824, + "step": 2923 + }, + { + "epoch": 2.3668926439340283, + "grad_norm": 2.186699867248535, + "learning_rate": 2.2424072883588723e-06, + "loss": 0.0833, + "step": 2924 + }, + { + "epoch": 2.367702114742487, + "grad_norm": 1.912595272064209, + "learning_rate": 2.2368927594452016e-06, + "loss": 0.0891, + "step": 2925 + }, + { + "epoch": 2.368511585550946, + "grad_norm": 1.8580678701400757, + "learning_rate": 2.2313841655335223e-06, + "loss": 0.0774, + "step": 2926 + }, + { + "epoch": 2.369321056359405, + "grad_norm": 1.9898111820220947, + "learning_rate": 2.225881510835226e-06, + "loss": 0.0813, + "step": 2927 + }, + { + "epoch": 2.370130527167864, + "grad_norm": 1.7827166318893433, + "learning_rate": 2.22038479955717e-06, + "loss": 0.071, + "step": 2928 + }, + { + "epoch": 2.370939997976323, + "grad_norm": 1.9177371263504028, + "learning_rate": 2.214894035901669e-06, + "loss": 0.0877, + "step": 2929 + }, + { + "epoch": 2.371749468784782, + "grad_norm": 1.941414713859558, + "learning_rate": 2.209409224066483e-06, + "loss": 0.0917, + "step": 2930 + }, + { + "epoch": 2.372558939593241, + "grad_norm": 1.9011470079421997, + "learning_rate": 2.2039303682448254e-06, + "loss": 0.0933, + "step": 2931 + }, + { + "epoch": 2.3733684104017, + "grad_norm": 1.9127817153930664, + "learning_rate": 2.1984574726253603e-06, + "loss": 0.0798, + "step": 2932 + }, + { + "epoch": 2.374177881210159, + "grad_norm": 1.9834080934524536, + "learning_rate": 2.1929905413921906e-06, + "loss": 0.0826, + "step": 2933 + }, + { + "epoch": 2.3749873520186178, + "grad_norm": 1.9732588529586792, + "learning_rate": 2.187529578724854e-06, + "loss": 0.0886, + "step": 2934 + }, + { + "epoch": 2.375796822827077, + "grad_norm": 2.0487606525421143, + "learning_rate": 2.1820745887983374e-06, + "loss": 0.0902, + "step": 2935 + }, + { + "epoch": 2.3766062936355357, + "grad_norm": 1.82564377784729, + "learning_rate": 2.176625575783058e-06, + "loss": 0.08, + "step": 2936 + }, + { + "epoch": 2.377415764443995, + "grad_norm": 1.97940194606781, + "learning_rate": 2.1711825438448564e-06, + "loss": 0.0872, + "step": 2937 + }, + { + "epoch": 2.3782252352524536, + "grad_norm": 1.9743996858596802, + "learning_rate": 2.1657454971450066e-06, + "loss": 0.0891, + "step": 2938 + }, + { + "epoch": 2.3790347060609127, + "grad_norm": 2.125206232070923, + "learning_rate": 2.160314439840202e-06, + "loss": 0.0936, + "step": 2939 + }, + { + "epoch": 2.379844176869372, + "grad_norm": 1.7750777006149292, + "learning_rate": 2.154889376082564e-06, + "loss": 0.0768, + "step": 2940 + }, + { + "epoch": 2.3806536476778306, + "grad_norm": 1.8579097986221313, + "learning_rate": 2.149470310019629e-06, + "loss": 0.0816, + "step": 2941 + }, + { + "epoch": 2.3814631184862898, + "grad_norm": 1.9925457239151, + "learning_rate": 2.1440572457943477e-06, + "loss": 0.0877, + "step": 2942 + }, + { + "epoch": 2.3822725892947485, + "grad_norm": 1.9482556581497192, + "learning_rate": 2.1386501875450773e-06, + "loss": 0.0877, + "step": 2943 + }, + { + "epoch": 2.3830820601032077, + "grad_norm": 1.9376554489135742, + "learning_rate": 2.133249139405593e-06, + "loss": 0.0879, + "step": 2944 + }, + { + "epoch": 2.3838915309116664, + "grad_norm": 1.907719373703003, + "learning_rate": 2.127854105505066e-06, + "loss": 0.0733, + "step": 2945 + }, + { + "epoch": 2.3847010017201256, + "grad_norm": 1.8877384662628174, + "learning_rate": 2.1224650899680755e-06, + "loss": 0.0802, + "step": 2946 + }, + { + "epoch": 2.3855104725285843, + "grad_norm": 2.0126540660858154, + "learning_rate": 2.117082096914601e-06, + "loss": 0.0905, + "step": 2947 + }, + { + "epoch": 2.3863199433370434, + "grad_norm": 1.9676190614700317, + "learning_rate": 2.1117051304600124e-06, + "loss": 0.0859, + "step": 2948 + }, + { + "epoch": 2.3871294141455026, + "grad_norm": 1.7298475503921509, + "learning_rate": 2.1063341947150726e-06, + "loss": 0.0775, + "step": 2949 + }, + { + "epoch": 2.3879388849539613, + "grad_norm": 1.9757049083709717, + "learning_rate": 2.100969293785935e-06, + "loss": 0.0918, + "step": 2950 + }, + { + "epoch": 2.3887483557624205, + "grad_norm": 1.696722149848938, + "learning_rate": 2.095610431774142e-06, + "loss": 0.0752, + "step": 2951 + }, + { + "epoch": 2.3895578265708792, + "grad_norm": 2.0841410160064697, + "learning_rate": 2.0902576127766196e-06, + "loss": 0.0875, + "step": 2952 + }, + { + "epoch": 2.3903672973793384, + "grad_norm": 1.7933954000473022, + "learning_rate": 2.084910840885668e-06, + "loss": 0.0807, + "step": 2953 + }, + { + "epoch": 2.391176768187797, + "grad_norm": 2.20745849609375, + "learning_rate": 2.079570120188965e-06, + "loss": 0.096, + "step": 2954 + }, + { + "epoch": 2.3919862389962563, + "grad_norm": 1.799983263015747, + "learning_rate": 2.074235454769572e-06, + "loss": 0.0756, + "step": 2955 + }, + { + "epoch": 2.392795709804715, + "grad_norm": 1.8222624063491821, + "learning_rate": 2.0689068487059093e-06, + "loss": 0.0827, + "step": 2956 + }, + { + "epoch": 2.393605180613174, + "grad_norm": 1.7967501878738403, + "learning_rate": 2.0635843060717687e-06, + "loss": 0.0912, + "step": 2957 + }, + { + "epoch": 2.3944146514216333, + "grad_norm": 1.9511682987213135, + "learning_rate": 2.058267830936307e-06, + "loss": 0.0864, + "step": 2958 + }, + { + "epoch": 2.395224122230092, + "grad_norm": 1.7715281248092651, + "learning_rate": 2.0529574273640475e-06, + "loss": 0.0807, + "step": 2959 + }, + { + "epoch": 2.3960335930385512, + "grad_norm": 1.9428633451461792, + "learning_rate": 2.0476530994148625e-06, + "loss": 0.0813, + "step": 2960 + }, + { + "epoch": 2.39684306384701, + "grad_norm": 1.727120041847229, + "learning_rate": 2.042354851143984e-06, + "loss": 0.0803, + "step": 2961 + }, + { + "epoch": 2.397652534655469, + "grad_norm": 1.8833268880844116, + "learning_rate": 2.0370626866019926e-06, + "loss": 0.0885, + "step": 2962 + }, + { + "epoch": 2.398462005463928, + "grad_norm": 1.9469451904296875, + "learning_rate": 2.0317766098348233e-06, + "loss": 0.0855, + "step": 2963 + }, + { + "epoch": 2.399271476272387, + "grad_norm": 1.924999475479126, + "learning_rate": 2.026496624883758e-06, + "loss": 0.0839, + "step": 2964 + }, + { + "epoch": 2.4000809470808457, + "grad_norm": 1.8907008171081543, + "learning_rate": 2.0212227357854132e-06, + "loss": 0.0886, + "step": 2965 + }, + { + "epoch": 2.400890417889305, + "grad_norm": 1.9772402048110962, + "learning_rate": 2.0159549465717478e-06, + "loss": 0.0807, + "step": 2966 + }, + { + "epoch": 2.401699888697764, + "grad_norm": 1.9782763719558716, + "learning_rate": 2.0106932612700647e-06, + "loss": 0.0853, + "step": 2967 + }, + { + "epoch": 2.402509359506223, + "grad_norm": 1.8918628692626953, + "learning_rate": 2.0054376839029887e-06, + "loss": 0.0824, + "step": 2968 + }, + { + "epoch": 2.403318830314682, + "grad_norm": 1.8643217086791992, + "learning_rate": 2.0001882184884858e-06, + "loss": 0.0829, + "step": 2969 + }, + { + "epoch": 2.4041283011231407, + "grad_norm": 2.0114455223083496, + "learning_rate": 1.99494486903984e-06, + "loss": 0.0884, + "step": 2970 + }, + { + "epoch": 2.4049377719316, + "grad_norm": 1.9691333770751953, + "learning_rate": 1.9897076395656677e-06, + "loss": 0.0885, + "step": 2971 + }, + { + "epoch": 2.4057472427400586, + "grad_norm": 2.0602457523345947, + "learning_rate": 1.9844765340699e-06, + "loss": 0.0846, + "step": 2972 + }, + { + "epoch": 2.4065567135485177, + "grad_norm": 2.0082437992095947, + "learning_rate": 1.979251556551789e-06, + "loss": 0.0911, + "step": 2973 + }, + { + "epoch": 2.4073661843569765, + "grad_norm": 1.9338845014572144, + "learning_rate": 1.974032711005901e-06, + "loss": 0.0887, + "step": 2974 + }, + { + "epoch": 2.4081756551654356, + "grad_norm": 1.8631622791290283, + "learning_rate": 1.9688200014221194e-06, + "loss": 0.0799, + "step": 2975 + }, + { + "epoch": 2.408985125973895, + "grad_norm": 1.7735637426376343, + "learning_rate": 1.9636134317856292e-06, + "loss": 0.0873, + "step": 2976 + }, + { + "epoch": 2.4097945967823535, + "grad_norm": 2.0488553047180176, + "learning_rate": 1.9584130060769247e-06, + "loss": 0.0871, + "step": 2977 + }, + { + "epoch": 2.4106040675908127, + "grad_norm": 1.827134132385254, + "learning_rate": 1.9532187282717984e-06, + "loss": 0.0762, + "step": 2978 + }, + { + "epoch": 2.4114135383992714, + "grad_norm": 2.009617328643799, + "learning_rate": 1.9480306023413544e-06, + "loss": 0.0807, + "step": 2979 + }, + { + "epoch": 2.4122230092077306, + "grad_norm": 1.6750550270080566, + "learning_rate": 1.94284863225198e-06, + "loss": 0.0809, + "step": 2980 + }, + { + "epoch": 2.4130324800161893, + "grad_norm": 2.2406303882598877, + "learning_rate": 1.9376728219653673e-06, + "loss": 0.0987, + "step": 2981 + }, + { + "epoch": 2.4138419508246485, + "grad_norm": 1.8973578214645386, + "learning_rate": 1.9325031754384894e-06, + "loss": 0.0847, + "step": 2982 + }, + { + "epoch": 2.414651421633107, + "grad_norm": 2.0232653617858887, + "learning_rate": 1.927339696623618e-06, + "loss": 0.0831, + "step": 2983 + }, + { + "epoch": 2.4154608924415664, + "grad_norm": 1.872723937034607, + "learning_rate": 1.922182389468299e-06, + "loss": 0.0867, + "step": 2984 + }, + { + "epoch": 2.4162703632500255, + "grad_norm": 1.9064781665802002, + "learning_rate": 1.9170312579153637e-06, + "loss": 0.083, + "step": 2985 + }, + { + "epoch": 2.4170798340584843, + "grad_norm": 1.9112921953201294, + "learning_rate": 1.9118863059029236e-06, + "loss": 0.087, + "step": 2986 + }, + { + "epoch": 2.4178893048669434, + "grad_norm": 2.043367862701416, + "learning_rate": 1.9067475373643695e-06, + "loss": 0.0949, + "step": 2987 + }, + { + "epoch": 2.418698775675402, + "grad_norm": 1.878126621246338, + "learning_rate": 1.901614956228357e-06, + "loss": 0.0821, + "step": 2988 + }, + { + "epoch": 2.4195082464838613, + "grad_norm": 1.8718503713607788, + "learning_rate": 1.8964885664188148e-06, + "loss": 0.0859, + "step": 2989 + }, + { + "epoch": 2.42031771729232, + "grad_norm": 2.041044235229492, + "learning_rate": 1.891368371854937e-06, + "loss": 0.0961, + "step": 2990 + }, + { + "epoch": 2.421127188100779, + "grad_norm": 1.8582855463027954, + "learning_rate": 1.886254376451182e-06, + "loss": 0.0851, + "step": 2991 + }, + { + "epoch": 2.421936658909238, + "grad_norm": 1.9624714851379395, + "learning_rate": 1.881146584117275e-06, + "loss": 0.0855, + "step": 2992 + }, + { + "epoch": 2.422746129717697, + "grad_norm": 1.8715063333511353, + "learning_rate": 1.8760449987581897e-06, + "loss": 0.0844, + "step": 2993 + }, + { + "epoch": 2.4235556005261563, + "grad_norm": 1.8429315090179443, + "learning_rate": 1.8709496242741553e-06, + "loss": 0.0781, + "step": 2994 + }, + { + "epoch": 2.424365071334615, + "grad_norm": 1.999250054359436, + "learning_rate": 1.8658604645606614e-06, + "loss": 0.0911, + "step": 2995 + }, + { + "epoch": 2.425174542143074, + "grad_norm": 1.8311150074005127, + "learning_rate": 1.8607775235084346e-06, + "loss": 0.0795, + "step": 2996 + }, + { + "epoch": 2.425984012951533, + "grad_norm": 1.9518768787384033, + "learning_rate": 1.8557008050034552e-06, + "loss": 0.083, + "step": 2997 + }, + { + "epoch": 2.426793483759992, + "grad_norm": 1.904239296913147, + "learning_rate": 1.8506303129269477e-06, + "loss": 0.0807, + "step": 2998 + }, + { + "epoch": 2.4276029545684508, + "grad_norm": 1.9226311445236206, + "learning_rate": 1.8455660511553697e-06, + "loss": 0.0897, + "step": 2999 + }, + { + "epoch": 2.42841242537691, + "grad_norm": 1.9318524599075317, + "learning_rate": 1.84050802356042e-06, + "loss": 0.0915, + "step": 3000 + }, + { + "epoch": 2.4292218961853687, + "grad_norm": 1.9277716875076294, + "learning_rate": 1.8354562340090255e-06, + "loss": 0.0775, + "step": 3001 + }, + { + "epoch": 2.430031366993828, + "grad_norm": 2.0829834938049316, + "learning_rate": 1.8304106863633552e-06, + "loss": 0.081, + "step": 3002 + }, + { + "epoch": 2.430840837802287, + "grad_norm": 2.083512544631958, + "learning_rate": 1.825371384480793e-06, + "loss": 0.0829, + "step": 3003 + }, + { + "epoch": 2.4316503086107457, + "grad_norm": 2.0435636043548584, + "learning_rate": 1.8203383322139613e-06, + "loss": 0.0916, + "step": 3004 + }, + { + "epoch": 2.432459779419205, + "grad_norm": 1.8139296770095825, + "learning_rate": 1.8153115334106918e-06, + "loss": 0.0822, + "step": 3005 + }, + { + "epoch": 2.4332692502276636, + "grad_norm": 2.0256123542785645, + "learning_rate": 1.8102909919140454e-06, + "loss": 0.0822, + "step": 3006 + }, + { + "epoch": 2.4340787210361228, + "grad_norm": 1.7914050817489624, + "learning_rate": 1.8052767115622937e-06, + "loss": 0.0834, + "step": 3007 + }, + { + "epoch": 2.4348881918445815, + "grad_norm": 1.903643012046814, + "learning_rate": 1.8002686961889194e-06, + "loss": 0.0896, + "step": 3008 + }, + { + "epoch": 2.4356976626530407, + "grad_norm": 1.7872087955474854, + "learning_rate": 1.7952669496226216e-06, + "loss": 0.0844, + "step": 3009 + }, + { + "epoch": 2.4365071334614994, + "grad_norm": 1.8275396823883057, + "learning_rate": 1.7902714756873063e-06, + "loss": 0.0874, + "step": 3010 + }, + { + "epoch": 2.4373166042699586, + "grad_norm": 1.9392116069793701, + "learning_rate": 1.7852822782020806e-06, + "loss": 0.0882, + "step": 3011 + }, + { + "epoch": 2.4381260750784177, + "grad_norm": 1.8765087127685547, + "learning_rate": 1.7802993609812535e-06, + "loss": 0.08, + "step": 3012 + }, + { + "epoch": 2.4389355458868764, + "grad_norm": 1.800761342048645, + "learning_rate": 1.775322727834331e-06, + "loss": 0.0827, + "step": 3013 + }, + { + "epoch": 2.4397450166953356, + "grad_norm": 1.9943658113479614, + "learning_rate": 1.7703523825660207e-06, + "loss": 0.0828, + "step": 3014 + }, + { + "epoch": 2.4405544875037943, + "grad_norm": 1.9836757183074951, + "learning_rate": 1.7653883289762208e-06, + "loss": 0.0927, + "step": 3015 + }, + { + "epoch": 2.4413639583122535, + "grad_norm": 1.9626820087432861, + "learning_rate": 1.7604305708600178e-06, + "loss": 0.0841, + "step": 3016 + }, + { + "epoch": 2.4421734291207122, + "grad_norm": 2.0761396884918213, + "learning_rate": 1.755479112007682e-06, + "loss": 0.0872, + "step": 3017 + }, + { + "epoch": 2.4429828999291714, + "grad_norm": 2.147408962249756, + "learning_rate": 1.7505339562046763e-06, + "loss": 0.0839, + "step": 3018 + }, + { + "epoch": 2.44379237073763, + "grad_norm": 1.844359040260315, + "learning_rate": 1.7455951072316368e-06, + "loss": 0.0828, + "step": 3019 + }, + { + "epoch": 2.4446018415460893, + "grad_norm": 1.737823247909546, + "learning_rate": 1.740662568864384e-06, + "loss": 0.0766, + "step": 3020 + }, + { + "epoch": 2.4454113123545484, + "grad_norm": 1.868481159210205, + "learning_rate": 1.7357363448739083e-06, + "loss": 0.0822, + "step": 3021 + }, + { + "epoch": 2.446220783163007, + "grad_norm": 2.039017677307129, + "learning_rate": 1.7308164390263794e-06, + "loss": 0.0872, + "step": 3022 + }, + { + "epoch": 2.4470302539714663, + "grad_norm": 1.9112192392349243, + "learning_rate": 1.7259028550831314e-06, + "loss": 0.0887, + "step": 3023 + }, + { + "epoch": 2.447839724779925, + "grad_norm": 1.816733717918396, + "learning_rate": 1.7209955968006653e-06, + "loss": 0.0775, + "step": 3024 + }, + { + "epoch": 2.4486491955883842, + "grad_norm": 1.8072538375854492, + "learning_rate": 1.7160946679306423e-06, + "loss": 0.083, + "step": 3025 + }, + { + "epoch": 2.449458666396843, + "grad_norm": 1.6930925846099854, + "learning_rate": 1.7112000722199006e-06, + "loss": 0.0732, + "step": 3026 + }, + { + "epoch": 2.450268137205302, + "grad_norm": 1.8576524257659912, + "learning_rate": 1.7063118134104205e-06, + "loss": 0.077, + "step": 3027 + }, + { + "epoch": 2.451077608013761, + "grad_norm": 2.0286154747009277, + "learning_rate": 1.7014298952393426e-06, + "loss": 0.0824, + "step": 3028 + }, + { + "epoch": 2.45188707882222, + "grad_norm": 2.0010907649993896, + "learning_rate": 1.6965543214389591e-06, + "loss": 0.0871, + "step": 3029 + }, + { + "epoch": 2.452696549630679, + "grad_norm": 1.8191273212432861, + "learning_rate": 1.691685095736717e-06, + "loss": 0.0767, + "step": 3030 + }, + { + "epoch": 2.453506020439138, + "grad_norm": 2.1199326515197754, + "learning_rate": 1.6868222218552e-06, + "loss": 0.0993, + "step": 3031 + }, + { + "epoch": 2.454315491247597, + "grad_norm": 2.0045390129089355, + "learning_rate": 1.6819657035121495e-06, + "loss": 0.0869, + "step": 3032 + }, + { + "epoch": 2.455124962056056, + "grad_norm": 1.8878406286239624, + "learning_rate": 1.6771155444204335e-06, + "loss": 0.0851, + "step": 3033 + }, + { + "epoch": 2.455934432864515, + "grad_norm": 1.8070582151412964, + "learning_rate": 1.6722717482880713e-06, + "loss": 0.084, + "step": 3034 + }, + { + "epoch": 2.4567439036729737, + "grad_norm": 2.0142321586608887, + "learning_rate": 1.6674343188182075e-06, + "loss": 0.0907, + "step": 3035 + }, + { + "epoch": 2.457553374481433, + "grad_norm": 1.9672812223434448, + "learning_rate": 1.6626032597091225e-06, + "loss": 0.077, + "step": 3036 + }, + { + "epoch": 2.4583628452898916, + "grad_norm": 1.963647484779358, + "learning_rate": 1.6577785746542286e-06, + "loss": 0.0959, + "step": 3037 + }, + { + "epoch": 2.4591723160983507, + "grad_norm": 1.8736863136291504, + "learning_rate": 1.6529602673420663e-06, + "loss": 0.0786, + "step": 3038 + }, + { + "epoch": 2.45998178690681, + "grad_norm": 1.8918837308883667, + "learning_rate": 1.648148341456296e-06, + "loss": 0.0783, + "step": 3039 + }, + { + "epoch": 2.4607912577152686, + "grad_norm": 1.9471391439437866, + "learning_rate": 1.6433428006757002e-06, + "loss": 0.0882, + "step": 3040 + }, + { + "epoch": 2.461600728523728, + "grad_norm": 1.886621356010437, + "learning_rate": 1.638543648674179e-06, + "loss": 0.0869, + "step": 3041 + }, + { + "epoch": 2.4624101993321865, + "grad_norm": 1.7392033338546753, + "learning_rate": 1.6337508891207521e-06, + "loss": 0.0803, + "step": 3042 + }, + { + "epoch": 2.4632196701406457, + "grad_norm": 1.843409538269043, + "learning_rate": 1.6289645256795516e-06, + "loss": 0.0848, + "step": 3043 + }, + { + "epoch": 2.4640291409491044, + "grad_norm": 1.9883002042770386, + "learning_rate": 1.6241845620098162e-06, + "loss": 0.0836, + "step": 3044 + }, + { + "epoch": 2.4648386117575636, + "grad_norm": 1.9262112379074097, + "learning_rate": 1.6194110017658926e-06, + "loss": 0.0873, + "step": 3045 + }, + { + "epoch": 2.4656480825660223, + "grad_norm": 1.7228769063949585, + "learning_rate": 1.614643848597236e-06, + "loss": 0.0763, + "step": 3046 + }, + { + "epoch": 2.4664575533744815, + "grad_norm": 1.9842746257781982, + "learning_rate": 1.6098831061483978e-06, + "loss": 0.0993, + "step": 3047 + }, + { + "epoch": 2.4672670241829406, + "grad_norm": 1.7610316276550293, + "learning_rate": 1.6051287780590275e-06, + "loss": 0.0843, + "step": 3048 + }, + { + "epoch": 2.4680764949913994, + "grad_norm": 1.8451064825057983, + "learning_rate": 1.6003808679638832e-06, + "loss": 0.0869, + "step": 3049 + }, + { + "epoch": 2.4688859657998585, + "grad_norm": 1.894405722618103, + "learning_rate": 1.5956393794928027e-06, + "loss": 0.0821, + "step": 3050 + }, + { + "epoch": 2.4696954366083173, + "grad_norm": 1.9139518737792969, + "learning_rate": 1.5909043162707194e-06, + "loss": 0.0899, + "step": 3051 + }, + { + "epoch": 2.4705049074167764, + "grad_norm": 1.938070297241211, + "learning_rate": 1.58617568191765e-06, + "loss": 0.0808, + "step": 3052 + }, + { + "epoch": 2.471314378225235, + "grad_norm": 1.9388099908828735, + "learning_rate": 1.581453480048707e-06, + "loss": 0.0941, + "step": 3053 + }, + { + "epoch": 2.4721238490336943, + "grad_norm": 2.046314239501953, + "learning_rate": 1.5767377142740724e-06, + "loss": 0.0842, + "step": 3054 + }, + { + "epoch": 2.472933319842153, + "grad_norm": 1.9282928705215454, + "learning_rate": 1.5720283881990206e-06, + "loss": 0.077, + "step": 3055 + }, + { + "epoch": 2.473742790650612, + "grad_norm": 3.0226001739501953, + "learning_rate": 1.5673255054238878e-06, + "loss": 0.0856, + "step": 3056 + }, + { + "epoch": 2.4745522614590714, + "grad_norm": 2.044685125350952, + "learning_rate": 1.5626290695441005e-06, + "loss": 0.079, + "step": 3057 + }, + { + "epoch": 2.47536173226753, + "grad_norm": 1.8571784496307373, + "learning_rate": 1.557939084150144e-06, + "loss": 0.0809, + "step": 3058 + }, + { + "epoch": 2.4761712030759893, + "grad_norm": 1.8629544973373413, + "learning_rate": 1.553255552827575e-06, + "loss": 0.0825, + "step": 3059 + }, + { + "epoch": 2.476980673884448, + "grad_norm": 1.8847112655639648, + "learning_rate": 1.5485784791570203e-06, + "loss": 0.0774, + "step": 3060 + }, + { + "epoch": 2.477790144692907, + "grad_norm": 1.8560924530029297, + "learning_rate": 1.543907866714167e-06, + "loss": 0.0883, + "step": 3061 + }, + { + "epoch": 2.478599615501366, + "grad_norm": 1.8440399169921875, + "learning_rate": 1.5392437190697628e-06, + "loss": 0.0674, + "step": 3062 + }, + { + "epoch": 2.479409086309825, + "grad_norm": 2.0902416706085205, + "learning_rate": 1.5345860397896117e-06, + "loss": 0.0848, + "step": 3063 + }, + { + "epoch": 2.4802185571182838, + "grad_norm": 1.8429791927337646, + "learning_rate": 1.5299348324345697e-06, + "loss": 0.0837, + "step": 3064 + }, + { + "epoch": 2.481028027926743, + "grad_norm": 1.8363887071609497, + "learning_rate": 1.5252901005605513e-06, + "loss": 0.0862, + "step": 3065 + }, + { + "epoch": 2.481837498735202, + "grad_norm": 1.9301354885101318, + "learning_rate": 1.5206518477185218e-06, + "loss": 0.0885, + "step": 3066 + }, + { + "epoch": 2.482646969543661, + "grad_norm": 1.890079379081726, + "learning_rate": 1.5160200774544864e-06, + "loss": 0.0799, + "step": 3067 + }, + { + "epoch": 2.48345644035212, + "grad_norm": 1.9057129621505737, + "learning_rate": 1.511394793309493e-06, + "loss": 0.0919, + "step": 3068 + }, + { + "epoch": 2.4842659111605787, + "grad_norm": 1.9289841651916504, + "learning_rate": 1.5067759988196406e-06, + "loss": 0.0759, + "step": 3069 + }, + { + "epoch": 2.485075381969038, + "grad_norm": 1.9098232984542847, + "learning_rate": 1.5021636975160558e-06, + "loss": 0.0812, + "step": 3070 + }, + { + "epoch": 2.4858848527774966, + "grad_norm": 2.102118492126465, + "learning_rate": 1.4975578929249123e-06, + "loss": 0.0796, + "step": 3071 + }, + { + "epoch": 2.4866943235859558, + "grad_norm": 1.8880397081375122, + "learning_rate": 1.4929585885674035e-06, + "loss": 0.0757, + "step": 3072 + }, + { + "epoch": 2.4875037943944145, + "grad_norm": 1.8363213539123535, + "learning_rate": 1.4883657879597691e-06, + "loss": 0.0844, + "step": 3073 + }, + { + "epoch": 2.4883132652028737, + "grad_norm": 2.10137677192688, + "learning_rate": 1.4837794946132623e-06, + "loss": 0.091, + "step": 3074 + }, + { + "epoch": 2.489122736011333, + "grad_norm": 1.93231201171875, + "learning_rate": 1.4791997120341695e-06, + "loss": 0.0858, + "step": 3075 + }, + { + "epoch": 2.4899322068197915, + "grad_norm": 1.9800245761871338, + "learning_rate": 1.4746264437237934e-06, + "loss": 0.0865, + "step": 3076 + }, + { + "epoch": 2.4907416776282507, + "grad_norm": 2.013294219970703, + "learning_rate": 1.4700596931784638e-06, + "loss": 0.0806, + "step": 3077 + }, + { + "epoch": 2.4915511484367094, + "grad_norm": 1.919503927230835, + "learning_rate": 1.465499463889527e-06, + "loss": 0.0779, + "step": 3078 + }, + { + "epoch": 2.4923606192451686, + "grad_norm": 1.9437743425369263, + "learning_rate": 1.4609457593433373e-06, + "loss": 0.0843, + "step": 3079 + }, + { + "epoch": 2.4931700900536273, + "grad_norm": 1.8206787109375, + "learning_rate": 1.4563985830212624e-06, + "loss": 0.0717, + "step": 3080 + }, + { + "epoch": 2.4939795608620865, + "grad_norm": 1.8345191478729248, + "learning_rate": 1.4518579383996855e-06, + "loss": 0.0704, + "step": 3081 + }, + { + "epoch": 2.494789031670545, + "grad_norm": 1.8307349681854248, + "learning_rate": 1.447323828949987e-06, + "loss": 0.0778, + "step": 3082 + }, + { + "epoch": 2.4955985024790044, + "grad_norm": 1.764757752418518, + "learning_rate": 1.4427962581385612e-06, + "loss": 0.0736, + "step": 3083 + }, + { + "epoch": 2.4964079732874636, + "grad_norm": 1.757617473602295, + "learning_rate": 1.438275229426792e-06, + "loss": 0.0799, + "step": 3084 + }, + { + "epoch": 2.4972174440959223, + "grad_norm": 1.7763340473175049, + "learning_rate": 1.4337607462710745e-06, + "loss": 0.0776, + "step": 3085 + }, + { + "epoch": 2.4980269149043814, + "grad_norm": 1.9023385047912598, + "learning_rate": 1.4292528121227888e-06, + "loss": 0.0816, + "step": 3086 + }, + { + "epoch": 2.49883638571284, + "grad_norm": 1.8766916990280151, + "learning_rate": 1.4247514304283106e-06, + "loss": 0.0854, + "step": 3087 + }, + { + "epoch": 2.4996458565212993, + "grad_norm": 2.080613136291504, + "learning_rate": 1.4202566046290112e-06, + "loss": 0.0912, + "step": 3088 + }, + { + "epoch": 2.500455327329758, + "grad_norm": 1.807839274406433, + "learning_rate": 1.4157683381612486e-06, + "loss": 0.0801, + "step": 3089 + }, + { + "epoch": 2.5012647981382172, + "grad_norm": 2.019901752471924, + "learning_rate": 1.4112866344563603e-06, + "loss": 0.0898, + "step": 3090 + }, + { + "epoch": 2.502074268946676, + "grad_norm": 2.052527666091919, + "learning_rate": 1.406811496940672e-06, + "loss": 0.0844, + "step": 3091 + }, + { + "epoch": 2.502883739755135, + "grad_norm": 1.903828740119934, + "learning_rate": 1.402342929035485e-06, + "loss": 0.0739, + "step": 3092 + }, + { + "epoch": 2.5036932105635943, + "grad_norm": 1.943083643913269, + "learning_rate": 1.3978809341570821e-06, + "loss": 0.0762, + "step": 3093 + }, + { + "epoch": 2.504502681372053, + "grad_norm": 1.996870994567871, + "learning_rate": 1.3934255157167243e-06, + "loss": 0.0692, + "step": 3094 + }, + { + "epoch": 2.5053121521805117, + "grad_norm": 1.988072395324707, + "learning_rate": 1.3889766771206348e-06, + "loss": 0.0876, + "step": 3095 + }, + { + "epoch": 2.506121622988971, + "grad_norm": 1.9547436237335205, + "learning_rate": 1.3845344217700108e-06, + "loss": 0.0858, + "step": 3096 + }, + { + "epoch": 2.50693109379743, + "grad_norm": 1.8583790063858032, + "learning_rate": 1.3800987530610232e-06, + "loss": 0.0801, + "step": 3097 + }, + { + "epoch": 2.507740564605889, + "grad_norm": 1.836418867111206, + "learning_rate": 1.3756696743847975e-06, + "loss": 0.0717, + "step": 3098 + }, + { + "epoch": 2.508550035414348, + "grad_norm": 1.8016287088394165, + "learning_rate": 1.3712471891274237e-06, + "loss": 0.0755, + "step": 3099 + }, + { + "epoch": 2.5093595062228067, + "grad_norm": 2.0239200592041016, + "learning_rate": 1.3668313006699542e-06, + "loss": 0.0876, + "step": 3100 + }, + { + "epoch": 2.510168977031266, + "grad_norm": 1.8295882940292358, + "learning_rate": 1.362422012388399e-06, + "loss": 0.0762, + "step": 3101 + }, + { + "epoch": 2.510978447839725, + "grad_norm": 1.9768253564834595, + "learning_rate": 1.3580193276537179e-06, + "loss": 0.0899, + "step": 3102 + }, + { + "epoch": 2.5117879186481837, + "grad_norm": 1.8130985498428345, + "learning_rate": 1.3536232498318202e-06, + "loss": 0.0919, + "step": 3103 + }, + { + "epoch": 2.5125973894566425, + "grad_norm": 2.0467734336853027, + "learning_rate": 1.349233782283572e-06, + "loss": 0.0859, + "step": 3104 + }, + { + "epoch": 2.5134068602651016, + "grad_norm": 1.8564404249191284, + "learning_rate": 1.3448509283647793e-06, + "loss": 0.0806, + "step": 3105 + }, + { + "epoch": 2.514216331073561, + "grad_norm": 2.063575267791748, + "learning_rate": 1.3404746914261958e-06, + "loss": 0.0811, + "step": 3106 + }, + { + "epoch": 2.5150258018820195, + "grad_norm": 1.7549986839294434, + "learning_rate": 1.3361050748135118e-06, + "loss": 0.0777, + "step": 3107 + }, + { + "epoch": 2.5158352726904787, + "grad_norm": 1.8853144645690918, + "learning_rate": 1.3317420818673632e-06, + "loss": 0.0771, + "step": 3108 + }, + { + "epoch": 2.5166447434989374, + "grad_norm": 2.0058462619781494, + "learning_rate": 1.327385715923316e-06, + "loss": 0.0851, + "step": 3109 + }, + { + "epoch": 2.5174542143073966, + "grad_norm": 1.842678427696228, + "learning_rate": 1.3230359803118699e-06, + "loss": 0.0741, + "step": 3110 + }, + { + "epoch": 2.5182636851158557, + "grad_norm": 1.9132338762283325, + "learning_rate": 1.3186928783584573e-06, + "loss": 0.0823, + "step": 3111 + }, + { + "epoch": 2.5190731559243145, + "grad_norm": 1.753975749015808, + "learning_rate": 1.314356413383444e-06, + "loss": 0.0804, + "step": 3112 + }, + { + "epoch": 2.519882626732773, + "grad_norm": 1.8905909061431885, + "learning_rate": 1.3100265887021146e-06, + "loss": 0.0793, + "step": 3113 + }, + { + "epoch": 2.5206920975412324, + "grad_norm": 1.9037917852401733, + "learning_rate": 1.3057034076246788e-06, + "loss": 0.0826, + "step": 3114 + }, + { + "epoch": 2.5215015683496915, + "grad_norm": 1.8455674648284912, + "learning_rate": 1.3013868734562652e-06, + "loss": 0.0835, + "step": 3115 + }, + { + "epoch": 2.5223110391581502, + "grad_norm": 1.8813096284866333, + "learning_rate": 1.297076989496925e-06, + "loss": 0.0831, + "step": 3116 + }, + { + "epoch": 2.5231205099666094, + "grad_norm": 1.926782250404358, + "learning_rate": 1.2927737590416278e-06, + "loss": 0.0793, + "step": 3117 + }, + { + "epoch": 2.523929980775068, + "grad_norm": 1.9606469869613647, + "learning_rate": 1.2884771853802491e-06, + "loss": 0.0862, + "step": 3118 + }, + { + "epoch": 2.5247394515835273, + "grad_norm": 1.8966920375823975, + "learning_rate": 1.2841872717975768e-06, + "loss": 0.0728, + "step": 3119 + }, + { + "epoch": 2.5255489223919865, + "grad_norm": 1.9750540256500244, + "learning_rate": 1.2799040215733116e-06, + "loss": 0.0839, + "step": 3120 + }, + { + "epoch": 2.526358393200445, + "grad_norm": 1.8847723007202148, + "learning_rate": 1.2756274379820566e-06, + "loss": 0.0878, + "step": 3121 + }, + { + "epoch": 2.527167864008904, + "grad_norm": 2.1157729625701904, + "learning_rate": 1.2713575242933151e-06, + "loss": 0.0865, + "step": 3122 + }, + { + "epoch": 2.527977334817363, + "grad_norm": 1.9402300119400024, + "learning_rate": 1.2670942837714984e-06, + "loss": 0.0856, + "step": 3123 + }, + { + "epoch": 2.5287868056258223, + "grad_norm": 1.931015133857727, + "learning_rate": 1.2628377196759123e-06, + "loss": 0.0749, + "step": 3124 + }, + { + "epoch": 2.529596276434281, + "grad_norm": 1.8675508499145508, + "learning_rate": 1.2585878352607584e-06, + "loss": 0.0752, + "step": 3125 + }, + { + "epoch": 2.53040574724274, + "grad_norm": 2.0913360118865967, + "learning_rate": 1.2543446337751309e-06, + "loss": 0.0945, + "step": 3126 + }, + { + "epoch": 2.531215218051199, + "grad_norm": 2.019024133682251, + "learning_rate": 1.2501081184630127e-06, + "loss": 0.0868, + "step": 3127 + }, + { + "epoch": 2.532024688859658, + "grad_norm": 2.213289260864258, + "learning_rate": 1.2458782925632805e-06, + "loss": 0.0957, + "step": 3128 + }, + { + "epoch": 2.532834159668117, + "grad_norm": 2.1766276359558105, + "learning_rate": 1.241655159309697e-06, + "loss": 0.0858, + "step": 3129 + }, + { + "epoch": 2.533643630476576, + "grad_norm": 1.9654102325439453, + "learning_rate": 1.2374387219309025e-06, + "loss": 0.0851, + "step": 3130 + }, + { + "epoch": 2.5344531012850346, + "grad_norm": 2.047818899154663, + "learning_rate": 1.2332289836504208e-06, + "loss": 0.0851, + "step": 3131 + }, + { + "epoch": 2.535262572093494, + "grad_norm": 2.0104317665100098, + "learning_rate": 1.2290259476866562e-06, + "loss": 0.0915, + "step": 3132 + }, + { + "epoch": 2.536072042901953, + "grad_norm": 1.8230055570602417, + "learning_rate": 1.2248296172528839e-06, + "loss": 0.0778, + "step": 3133 + }, + { + "epoch": 2.5368815137104117, + "grad_norm": 1.9411780834197998, + "learning_rate": 1.2206399955572623e-06, + "loss": 0.0783, + "step": 3134 + }, + { + "epoch": 2.537690984518871, + "grad_norm": 1.9959756135940552, + "learning_rate": 1.2164570858028079e-06, + "loss": 0.0845, + "step": 3135 + }, + { + "epoch": 2.5385004553273296, + "grad_norm": 1.8684245347976685, + "learning_rate": 1.2122808911874184e-06, + "loss": 0.0749, + "step": 3136 + }, + { + "epoch": 2.5393099261357888, + "grad_norm": 1.9102678298950195, + "learning_rate": 1.2081114149038486e-06, + "loss": 0.0815, + "step": 3137 + }, + { + "epoch": 2.540119396944248, + "grad_norm": 1.9346346855163574, + "learning_rate": 1.20394866013972e-06, + "loss": 0.0829, + "step": 3138 + }, + { + "epoch": 2.5409288677527067, + "grad_norm": 1.8643752336502075, + "learning_rate": 1.1997926300775165e-06, + "loss": 0.0849, + "step": 3139 + }, + { + "epoch": 2.5417383385611654, + "grad_norm": 1.928281545639038, + "learning_rate": 1.1956433278945811e-06, + "loss": 0.0799, + "step": 3140 + }, + { + "epoch": 2.5425478093696245, + "grad_norm": 1.8981578350067139, + "learning_rate": 1.191500756763112e-06, + "loss": 0.0769, + "step": 3141 + }, + { + "epoch": 2.5433572801780837, + "grad_norm": 1.8364908695220947, + "learning_rate": 1.1873649198501625e-06, + "loss": 0.074, + "step": 3142 + }, + { + "epoch": 2.5441667509865424, + "grad_norm": 1.73519766330719, + "learning_rate": 1.183235820317632e-06, + "loss": 0.0745, + "step": 3143 + }, + { + "epoch": 2.5449762217950016, + "grad_norm": 1.9699985980987549, + "learning_rate": 1.1791134613222788e-06, + "loss": 0.0846, + "step": 3144 + }, + { + "epoch": 2.5457856926034603, + "grad_norm": 2.019347667694092, + "learning_rate": 1.1749978460156998e-06, + "loss": 0.0945, + "step": 3145 + }, + { + "epoch": 2.5465951634119195, + "grad_norm": 2.0117125511169434, + "learning_rate": 1.1708889775443421e-06, + "loss": 0.0848, + "step": 3146 + }, + { + "epoch": 2.5474046342203787, + "grad_norm": 2.151951789855957, + "learning_rate": 1.166786859049489e-06, + "loss": 0.0796, + "step": 3147 + }, + { + "epoch": 2.5482141050288374, + "grad_norm": 1.7833185195922852, + "learning_rate": 1.1626914936672695e-06, + "loss": 0.0794, + "step": 3148 + }, + { + "epoch": 2.549023575837296, + "grad_norm": 1.9599018096923828, + "learning_rate": 1.1586028845286456e-06, + "loss": 0.0849, + "step": 3149 + }, + { + "epoch": 2.5498330466457553, + "grad_norm": 1.8535600900650024, + "learning_rate": 1.1545210347594127e-06, + "loss": 0.0735, + "step": 3150 + }, + { + "epoch": 2.5506425174542144, + "grad_norm": 1.82524836063385, + "learning_rate": 1.1504459474802022e-06, + "loss": 0.074, + "step": 3151 + }, + { + "epoch": 2.551451988262673, + "grad_norm": 1.7788665294647217, + "learning_rate": 1.1463776258064774e-06, + "loss": 0.0848, + "step": 3152 + }, + { + "epoch": 2.5522614590711323, + "grad_norm": 1.9932734966278076, + "learning_rate": 1.1423160728485228e-06, + "loss": 0.0887, + "step": 3153 + }, + { + "epoch": 2.553070929879591, + "grad_norm": 1.973741888999939, + "learning_rate": 1.138261291711451e-06, + "loss": 0.0852, + "step": 3154 + }, + { + "epoch": 2.5538804006880502, + "grad_norm": 2.0645554065704346, + "learning_rate": 1.1342132854952003e-06, + "loss": 0.0839, + "step": 3155 + }, + { + "epoch": 2.5546898714965094, + "grad_norm": 1.8189691305160522, + "learning_rate": 1.130172057294523e-06, + "loss": 0.0803, + "step": 3156 + }, + { + "epoch": 2.555499342304968, + "grad_norm": 1.9976301193237305, + "learning_rate": 1.1261376101989985e-06, + "loss": 0.0849, + "step": 3157 + }, + { + "epoch": 2.556308813113427, + "grad_norm": 2.2173051834106445, + "learning_rate": 1.122109947293013e-06, + "loss": 0.0971, + "step": 3158 + }, + { + "epoch": 2.557118283921886, + "grad_norm": 1.898952603340149, + "learning_rate": 1.1180890716557735e-06, + "loss": 0.0847, + "step": 3159 + }, + { + "epoch": 2.557927754730345, + "grad_norm": 1.8862676620483398, + "learning_rate": 1.1140749863612921e-06, + "loss": 0.0827, + "step": 3160 + }, + { + "epoch": 2.558737225538804, + "grad_norm": 1.795983076095581, + "learning_rate": 1.1100676944783905e-06, + "loss": 0.0805, + "step": 3161 + }, + { + "epoch": 2.559546696347263, + "grad_norm": 1.8983920812606812, + "learning_rate": 1.1060671990707006e-06, + "loss": 0.0819, + "step": 3162 + }, + { + "epoch": 2.560356167155722, + "grad_norm": 1.789292573928833, + "learning_rate": 1.1020735031966579e-06, + "loss": 0.0766, + "step": 3163 + }, + { + "epoch": 2.561165637964181, + "grad_norm": 2.1328821182250977, + "learning_rate": 1.0980866099094956e-06, + "loss": 0.0917, + "step": 3164 + }, + { + "epoch": 2.56197510877264, + "grad_norm": 1.9130382537841797, + "learning_rate": 1.094106522257249e-06, + "loss": 0.0844, + "step": 3165 + }, + { + "epoch": 2.562784579581099, + "grad_norm": 1.744791030883789, + "learning_rate": 1.090133243282746e-06, + "loss": 0.0843, + "step": 3166 + }, + { + "epoch": 2.5635940503895576, + "grad_norm": 1.9690933227539062, + "learning_rate": 1.0861667760236194e-06, + "loss": 0.0817, + "step": 3167 + }, + { + "epoch": 2.5644035211980167, + "grad_norm": 1.956362009048462, + "learning_rate": 1.0822071235122812e-06, + "loss": 0.0738, + "step": 3168 + }, + { + "epoch": 2.565212992006476, + "grad_norm": 1.9702086448669434, + "learning_rate": 1.078254288775945e-06, + "loss": 0.0759, + "step": 3169 + }, + { + "epoch": 2.5660224628149346, + "grad_norm": 1.8697761297225952, + "learning_rate": 1.074308274836604e-06, + "loss": 0.082, + "step": 3170 + }, + { + "epoch": 2.566831933623394, + "grad_norm": 1.8417905569076538, + "learning_rate": 1.0703690847110437e-06, + "loss": 0.0771, + "step": 3171 + }, + { + "epoch": 2.5676414044318525, + "grad_norm": 1.750064492225647, + "learning_rate": 1.0664367214108262e-06, + "loss": 0.0852, + "step": 3172 + }, + { + "epoch": 2.5684508752403117, + "grad_norm": 1.8871173858642578, + "learning_rate": 1.0625111879422966e-06, + "loss": 0.0741, + "step": 3173 + }, + { + "epoch": 2.569260346048771, + "grad_norm": 1.938151478767395, + "learning_rate": 1.05859248730658e-06, + "loss": 0.082, + "step": 3174 + }, + { + "epoch": 2.5700698168572296, + "grad_norm": 1.8408821821212769, + "learning_rate": 1.0546806224995786e-06, + "loss": 0.074, + "step": 3175 + }, + { + "epoch": 2.5708792876656883, + "grad_norm": 2.0165529251098633, + "learning_rate": 1.0507755965119649e-06, + "loss": 0.076, + "step": 3176 + }, + { + "epoch": 2.5716887584741475, + "grad_norm": 1.8598451614379883, + "learning_rate": 1.046877412329186e-06, + "loss": 0.08, + "step": 3177 + }, + { + "epoch": 2.5724982292826066, + "grad_norm": 1.9244662523269653, + "learning_rate": 1.0429860729314534e-06, + "loss": 0.0804, + "step": 3178 + }, + { + "epoch": 2.5733077000910654, + "grad_norm": 1.890440583229065, + "learning_rate": 1.0391015812937521e-06, + "loss": 0.0762, + "step": 3179 + }, + { + "epoch": 2.5741171708995245, + "grad_norm": 2.0052366256713867, + "learning_rate": 1.0352239403858323e-06, + "loss": 0.0855, + "step": 3180 + }, + { + "epoch": 2.5749266417079832, + "grad_norm": 2.0203773975372314, + "learning_rate": 1.0313531531721999e-06, + "loss": 0.0892, + "step": 3181 + }, + { + "epoch": 2.5757361125164424, + "grad_norm": 2.015434741973877, + "learning_rate": 1.0274892226121246e-06, + "loss": 0.0838, + "step": 3182 + }, + { + "epoch": 2.5765455833249016, + "grad_norm": 1.831428050994873, + "learning_rate": 1.023632151659637e-06, + "loss": 0.0817, + "step": 3183 + }, + { + "epoch": 2.5773550541333603, + "grad_norm": 1.8833180665969849, + "learning_rate": 1.019781943263518e-06, + "loss": 0.0835, + "step": 3184 + }, + { + "epoch": 2.578164524941819, + "grad_norm": 1.9859877824783325, + "learning_rate": 1.0159386003673077e-06, + "loss": 0.0813, + "step": 3185 + }, + { + "epoch": 2.578973995750278, + "grad_norm": 1.973945140838623, + "learning_rate": 1.01210212590929e-06, + "loss": 0.0877, + "step": 3186 + }, + { + "epoch": 2.5797834665587374, + "grad_norm": 1.866123914718628, + "learning_rate": 1.0082725228225077e-06, + "loss": 0.0711, + "step": 3187 + }, + { + "epoch": 2.580592937367196, + "grad_norm": 1.8692243099212646, + "learning_rate": 1.004449794034742e-06, + "loss": 0.069, + "step": 3188 + }, + { + "epoch": 2.5814024081756552, + "grad_norm": 1.9229217767715454, + "learning_rate": 1.0006339424685218e-06, + "loss": 0.0844, + "step": 3189 + }, + { + "epoch": 2.582211878984114, + "grad_norm": 1.847390055656433, + "learning_rate": 9.96824971041116e-07, + "loss": 0.0841, + "step": 3190 + }, + { + "epoch": 2.583021349792573, + "grad_norm": 1.9405722618103027, + "learning_rate": 9.930228826645383e-07, + "loss": 0.0818, + "step": 3191 + }, + { + "epoch": 2.5838308206010323, + "grad_norm": 1.899722695350647, + "learning_rate": 9.892276802455369e-07, + "loss": 0.0756, + "step": 3192 + }, + { + "epoch": 2.584640291409491, + "grad_norm": 1.711672306060791, + "learning_rate": 9.854393666855965e-07, + "loss": 0.0723, + "step": 3193 + }, + { + "epoch": 2.5854497622179498, + "grad_norm": 1.9854412078857422, + "learning_rate": 9.816579448809327e-07, + "loss": 0.0903, + "step": 3194 + }, + { + "epoch": 2.586259233026409, + "grad_norm": 1.9510711431503296, + "learning_rate": 9.77883417722496e-07, + "loss": 0.0803, + "step": 3195 + }, + { + "epoch": 2.587068703834868, + "grad_norm": 1.949691653251648, + "learning_rate": 9.74115788095963e-07, + "loss": 0.0892, + "step": 3196 + }, + { + "epoch": 2.587878174643327, + "grad_norm": 1.981431007385254, + "learning_rate": 9.703550588817413e-07, + "loss": 0.0836, + "step": 3197 + }, + { + "epoch": 2.588687645451786, + "grad_norm": 2.0906083583831787, + "learning_rate": 9.666012329549558e-07, + "loss": 0.0827, + "step": 3198 + }, + { + "epoch": 2.5894971162602447, + "grad_norm": 2.1692652702331543, + "learning_rate": 9.628543131854618e-07, + "loss": 0.094, + "step": 3199 + }, + { + "epoch": 2.590306587068704, + "grad_norm": 1.9802311658859253, + "learning_rate": 9.59114302437829e-07, + "loss": 0.0862, + "step": 3200 + }, + { + "epoch": 2.591116057877163, + "grad_norm": 2.0424160957336426, + "learning_rate": 9.55381203571345e-07, + "loss": 0.0932, + "step": 3201 + }, + { + "epoch": 2.5919255286856218, + "grad_norm": 1.9724071025848389, + "learning_rate": 9.516550194400165e-07, + "loss": 0.0929, + "step": 3202 + }, + { + "epoch": 2.5927349994940805, + "grad_norm": 1.7491130828857422, + "learning_rate": 9.479357528925648e-07, + "loss": 0.0762, + "step": 3203 + }, + { + "epoch": 2.5935444703025397, + "grad_norm": 1.8790130615234375, + "learning_rate": 9.442234067724188e-07, + "loss": 0.0751, + "step": 3204 + }, + { + "epoch": 2.594353941110999, + "grad_norm": 1.9470752477645874, + "learning_rate": 9.405179839177159e-07, + "loss": 0.0882, + "step": 3205 + }, + { + "epoch": 2.5951634119194575, + "grad_norm": 2.017338752746582, + "learning_rate": 9.368194871613057e-07, + "loss": 0.0867, + "step": 3206 + }, + { + "epoch": 2.5959728827279167, + "grad_norm": 1.7570725679397583, + "learning_rate": 9.331279193307374e-07, + "loss": 0.0748, + "step": 3207 + }, + { + "epoch": 2.5967823535363754, + "grad_norm": 1.8240690231323242, + "learning_rate": 9.29443283248267e-07, + "loss": 0.0823, + "step": 3208 + }, + { + "epoch": 2.5975918243448346, + "grad_norm": 2.007695198059082, + "learning_rate": 9.257655817308475e-07, + "loss": 0.0911, + "step": 3209 + }, + { + "epoch": 2.5984012951532938, + "grad_norm": 1.833555817604065, + "learning_rate": 9.220948175901345e-07, + "loss": 0.0821, + "step": 3210 + }, + { + "epoch": 2.5992107659617525, + "grad_norm": 1.8857390880584717, + "learning_rate": 9.184309936324765e-07, + "loss": 0.0881, + "step": 3211 + }, + { + "epoch": 2.600020236770211, + "grad_norm": 1.9068756103515625, + "learning_rate": 9.147741126589171e-07, + "loss": 0.0812, + "step": 3212 + }, + { + "epoch": 2.6008297075786704, + "grad_norm": 1.940076470375061, + "learning_rate": 9.11124177465188e-07, + "loss": 0.0871, + "step": 3213 + }, + { + "epoch": 2.6016391783871295, + "grad_norm": 1.8180162906646729, + "learning_rate": 9.074811908417225e-07, + "loss": 0.069, + "step": 3214 + }, + { + "epoch": 2.6024486491955883, + "grad_norm": 1.98991060256958, + "learning_rate": 9.038451555736294e-07, + "loss": 0.0817, + "step": 3215 + }, + { + "epoch": 2.6032581200040474, + "grad_norm": 1.8278461694717407, + "learning_rate": 9.002160744407084e-07, + "loss": 0.0715, + "step": 3216 + }, + { + "epoch": 2.604067590812506, + "grad_norm": 1.900182843208313, + "learning_rate": 8.965939502174392e-07, + "loss": 0.0747, + "step": 3217 + }, + { + "epoch": 2.6048770616209653, + "grad_norm": 1.975672721862793, + "learning_rate": 8.929787856729899e-07, + "loss": 0.0871, + "step": 3218 + }, + { + "epoch": 2.6056865324294245, + "grad_norm": 2.059126138687134, + "learning_rate": 8.893705835711997e-07, + "loss": 0.0825, + "step": 3219 + }, + { + "epoch": 2.606496003237883, + "grad_norm": 1.8161860704421997, + "learning_rate": 8.857693466705919e-07, + "loss": 0.0772, + "step": 3220 + }, + { + "epoch": 2.607305474046342, + "grad_norm": 1.967315673828125, + "learning_rate": 8.821750777243576e-07, + "loss": 0.0879, + "step": 3221 + }, + { + "epoch": 2.608114944854801, + "grad_norm": 1.8702328205108643, + "learning_rate": 8.785877794803688e-07, + "loss": 0.075, + "step": 3222 + }, + { + "epoch": 2.6089244156632603, + "grad_norm": 1.7315990924835205, + "learning_rate": 8.750074546811626e-07, + "loss": 0.0725, + "step": 3223 + }, + { + "epoch": 2.609733886471719, + "grad_norm": 1.9375969171524048, + "learning_rate": 8.714341060639442e-07, + "loss": 0.0739, + "step": 3224 + }, + { + "epoch": 2.610543357280178, + "grad_norm": 1.9697284698486328, + "learning_rate": 8.678677363605892e-07, + "loss": 0.0819, + "step": 3225 + }, + { + "epoch": 2.611352828088637, + "grad_norm": 1.9243870973587036, + "learning_rate": 8.643083482976378e-07, + "loss": 0.0812, + "step": 3226 + }, + { + "epoch": 2.612162298897096, + "grad_norm": 1.9545117616653442, + "learning_rate": 8.60755944596291e-07, + "loss": 0.0951, + "step": 3227 + }, + { + "epoch": 2.6129717697055552, + "grad_norm": 1.9525104761123657, + "learning_rate": 8.572105279724085e-07, + "loss": 0.077, + "step": 3228 + }, + { + "epoch": 2.613781240514014, + "grad_norm": 1.8725157976150513, + "learning_rate": 8.536721011365079e-07, + "loss": 0.0847, + "step": 3229 + }, + { + "epoch": 2.6145907113224727, + "grad_norm": 1.8429111242294312, + "learning_rate": 8.501406667937684e-07, + "loss": 0.0829, + "step": 3230 + }, + { + "epoch": 2.615400182130932, + "grad_norm": 1.8951866626739502, + "learning_rate": 8.466162276440204e-07, + "loss": 0.0783, + "step": 3231 + }, + { + "epoch": 2.616209652939391, + "grad_norm": 1.9374504089355469, + "learning_rate": 8.430987863817452e-07, + "loss": 0.0768, + "step": 3232 + }, + { + "epoch": 2.6170191237478497, + "grad_norm": 2.0487239360809326, + "learning_rate": 8.39588345696073e-07, + "loss": 0.0827, + "step": 3233 + }, + { + "epoch": 2.617828594556309, + "grad_norm": 2.2200942039489746, + "learning_rate": 8.360849082707856e-07, + "loss": 0.0986, + "step": 3234 + }, + { + "epoch": 2.6186380653647676, + "grad_norm": 1.9394195079803467, + "learning_rate": 8.325884767843095e-07, + "loss": 0.0742, + "step": 3235 + }, + { + "epoch": 2.619447536173227, + "grad_norm": 1.9573806524276733, + "learning_rate": 8.290990539097121e-07, + "loss": 0.0879, + "step": 3236 + }, + { + "epoch": 2.620257006981686, + "grad_norm": 2.0881707668304443, + "learning_rate": 8.256166423147061e-07, + "loss": 0.0982, + "step": 3237 + }, + { + "epoch": 2.6210664777901447, + "grad_norm": 2.0049214363098145, + "learning_rate": 8.221412446616462e-07, + "loss": 0.0842, + "step": 3238 + }, + { + "epoch": 2.6218759485986034, + "grad_norm": 1.7649911642074585, + "learning_rate": 8.186728636075181e-07, + "loss": 0.0796, + "step": 3239 + }, + { + "epoch": 2.6226854194070626, + "grad_norm": 1.7730791568756104, + "learning_rate": 8.152115018039475e-07, + "loss": 0.0729, + "step": 3240 + }, + { + "epoch": 2.6234948902155217, + "grad_norm": 1.8667387962341309, + "learning_rate": 8.117571618971909e-07, + "loss": 0.0772, + "step": 3241 + }, + { + "epoch": 2.6243043610239805, + "grad_norm": 1.8956987857818604, + "learning_rate": 8.083098465281413e-07, + "loss": 0.083, + "step": 3242 + }, + { + "epoch": 2.6251138318324396, + "grad_norm": 2.0110175609588623, + "learning_rate": 8.048695583323196e-07, + "loss": 0.0917, + "step": 3243 + }, + { + "epoch": 2.6259233026408983, + "grad_norm": 2.0262248516082764, + "learning_rate": 8.014362999398717e-07, + "loss": 0.0871, + "step": 3244 + }, + { + "epoch": 2.6267327734493575, + "grad_norm": 1.8190375566482544, + "learning_rate": 7.980100739755681e-07, + "loss": 0.0754, + "step": 3245 + }, + { + "epoch": 2.6275422442578167, + "grad_norm": 1.8147135972976685, + "learning_rate": 7.945908830588101e-07, + "loss": 0.0769, + "step": 3246 + }, + { + "epoch": 2.6283517150662754, + "grad_norm": 1.7291878461837769, + "learning_rate": 7.911787298036134e-07, + "loss": 0.077, + "step": 3247 + }, + { + "epoch": 2.629161185874734, + "grad_norm": 1.9009652137756348, + "learning_rate": 7.877736168186179e-07, + "loss": 0.0815, + "step": 3248 + }, + { + "epoch": 2.6299706566831933, + "grad_norm": 2.0139904022216797, + "learning_rate": 7.843755467070758e-07, + "loss": 0.0868, + "step": 3249 + }, + { + "epoch": 2.6307801274916525, + "grad_norm": 2.1452484130859375, + "learning_rate": 7.809845220668621e-07, + "loss": 0.0919, + "step": 3250 + }, + { + "epoch": 2.631589598300111, + "grad_norm": 2.2936043739318848, + "learning_rate": 7.776005454904601e-07, + "loss": 0.0929, + "step": 3251 + }, + { + "epoch": 2.6323990691085704, + "grad_norm": 1.8168753385543823, + "learning_rate": 7.742236195649655e-07, + "loss": 0.0767, + "step": 3252 + }, + { + "epoch": 2.633208539917029, + "grad_norm": 1.9279241561889648, + "learning_rate": 7.708537468720845e-07, + "loss": 0.0736, + "step": 3253 + }, + { + "epoch": 2.6340180107254882, + "grad_norm": 1.9805740118026733, + "learning_rate": 7.674909299881339e-07, + "loss": 0.0762, + "step": 3254 + }, + { + "epoch": 2.6348274815339474, + "grad_norm": 1.7986804246902466, + "learning_rate": 7.641351714840306e-07, + "loss": 0.0734, + "step": 3255 + }, + { + "epoch": 2.635636952342406, + "grad_norm": 1.7399113178253174, + "learning_rate": 7.607864739252968e-07, + "loss": 0.0691, + "step": 3256 + }, + { + "epoch": 2.636446423150865, + "grad_norm": 1.9186763763427734, + "learning_rate": 7.574448398720613e-07, + "loss": 0.0729, + "step": 3257 + }, + { + "epoch": 2.637255893959324, + "grad_norm": 1.8595455884933472, + "learning_rate": 7.541102718790472e-07, + "loss": 0.0811, + "step": 3258 + }, + { + "epoch": 2.638065364767783, + "grad_norm": 1.9744853973388672, + "learning_rate": 7.507827724955774e-07, + "loss": 0.083, + "step": 3259 + }, + { + "epoch": 2.638874835576242, + "grad_norm": 1.9282712936401367, + "learning_rate": 7.474623442655704e-07, + "loss": 0.0808, + "step": 3260 + }, + { + "epoch": 2.639684306384701, + "grad_norm": 1.8902528285980225, + "learning_rate": 7.441489897275423e-07, + "loss": 0.0757, + "step": 3261 + }, + { + "epoch": 2.64049377719316, + "grad_norm": 1.9379042387008667, + "learning_rate": 7.408427114145944e-07, + "loss": 0.0731, + "step": 3262 + }, + { + "epoch": 2.641303248001619, + "grad_norm": 1.8919955492019653, + "learning_rate": 7.375435118544238e-07, + "loss": 0.0805, + "step": 3263 + }, + { + "epoch": 2.642112718810078, + "grad_norm": 1.8018670082092285, + "learning_rate": 7.342513935693107e-07, + "loss": 0.0651, + "step": 3264 + }, + { + "epoch": 2.642922189618537, + "grad_norm": 1.745866298675537, + "learning_rate": 7.309663590761285e-07, + "loss": 0.0752, + "step": 3265 + }, + { + "epoch": 2.6437316604269956, + "grad_norm": 1.6793015003204346, + "learning_rate": 7.276884108863302e-07, + "loss": 0.0742, + "step": 3266 + }, + { + "epoch": 2.6445411312354548, + "grad_norm": 1.8033363819122314, + "learning_rate": 7.244175515059514e-07, + "loss": 0.0675, + "step": 3267 + }, + { + "epoch": 2.645350602043914, + "grad_norm": 1.9812811613082886, + "learning_rate": 7.211537834356086e-07, + "loss": 0.0754, + "step": 3268 + }, + { + "epoch": 2.6461600728523726, + "grad_norm": 1.8746434450149536, + "learning_rate": 7.178971091705e-07, + "loss": 0.0739, + "step": 3269 + }, + { + "epoch": 2.646969543660832, + "grad_norm": 1.9943737983703613, + "learning_rate": 7.146475312003931e-07, + "loss": 0.0847, + "step": 3270 + }, + { + "epoch": 2.6477790144692905, + "grad_norm": 1.9746605157852173, + "learning_rate": 7.114050520096394e-07, + "loss": 0.0791, + "step": 3271 + }, + { + "epoch": 2.6485884852777497, + "grad_norm": 1.8641917705535889, + "learning_rate": 7.081696740771549e-07, + "loss": 0.0825, + "step": 3272 + }, + { + "epoch": 2.649397956086209, + "grad_norm": 1.9768635034561157, + "learning_rate": 7.049413998764343e-07, + "loss": 0.0776, + "step": 3273 + }, + { + "epoch": 2.6502074268946676, + "grad_norm": 1.8881038427352905, + "learning_rate": 7.017202318755356e-07, + "loss": 0.0779, + "step": 3274 + }, + { + "epoch": 2.6510168977031263, + "grad_norm": 1.8678479194641113, + "learning_rate": 6.985061725370845e-07, + "loss": 0.0854, + "step": 3275 + }, + { + "epoch": 2.6518263685115855, + "grad_norm": 1.941008448600769, + "learning_rate": 6.952992243182743e-07, + "loss": 0.0782, + "step": 3276 + }, + { + "epoch": 2.6526358393200447, + "grad_norm": 1.8818292617797852, + "learning_rate": 6.920993896708628e-07, + "loss": 0.0765, + "step": 3277 + }, + { + "epoch": 2.6534453101285034, + "grad_norm": 1.9703333377838135, + "learning_rate": 6.889066710411663e-07, + "loss": 0.0747, + "step": 3278 + }, + { + "epoch": 2.6542547809369625, + "grad_norm": 1.9786368608474731, + "learning_rate": 6.857210708700623e-07, + "loss": 0.0817, + "step": 3279 + }, + { + "epoch": 2.6550642517454213, + "grad_norm": 1.8928227424621582, + "learning_rate": 6.825425915929828e-07, + "loss": 0.0761, + "step": 3280 + }, + { + "epoch": 2.6558737225538804, + "grad_norm": 1.8331146240234375, + "learning_rate": 6.793712356399229e-07, + "loss": 0.0719, + "step": 3281 + }, + { + "epoch": 2.6566831933623396, + "grad_norm": 2.1029624938964844, + "learning_rate": 6.762070054354264e-07, + "loss": 0.0808, + "step": 3282 + }, + { + "epoch": 2.6574926641707983, + "grad_norm": 1.8249082565307617, + "learning_rate": 6.730499033985904e-07, + "loss": 0.0697, + "step": 3283 + }, + { + "epoch": 2.658302134979257, + "grad_norm": 2.008965492248535, + "learning_rate": 6.698999319430621e-07, + "loss": 0.0897, + "step": 3284 + }, + { + "epoch": 2.659111605787716, + "grad_norm": 2.04976487159729, + "learning_rate": 6.667570934770395e-07, + "loss": 0.0921, + "step": 3285 + }, + { + "epoch": 2.6599210765961754, + "grad_norm": 2.0258471965789795, + "learning_rate": 6.636213904032652e-07, + "loss": 0.081, + "step": 3286 + }, + { + "epoch": 2.660730547404634, + "grad_norm": 1.9521265029907227, + "learning_rate": 6.604928251190257e-07, + "loss": 0.0792, + "step": 3287 + }, + { + "epoch": 2.6615400182130933, + "grad_norm": 1.8802385330200195, + "learning_rate": 6.573714000161525e-07, + "loss": 0.0717, + "step": 3288 + }, + { + "epoch": 2.662349489021552, + "grad_norm": 2.0145461559295654, + "learning_rate": 6.542571174810208e-07, + "loss": 0.078, + "step": 3289 + }, + { + "epoch": 2.663158959830011, + "grad_norm": 1.9155247211456299, + "learning_rate": 6.511499798945398e-07, + "loss": 0.081, + "step": 3290 + }, + { + "epoch": 2.6639684306384703, + "grad_norm": 2.075070381164551, + "learning_rate": 6.480499896321591e-07, + "loss": 0.0991, + "step": 3291 + }, + { + "epoch": 2.664777901446929, + "grad_norm": 1.9385182857513428, + "learning_rate": 6.449571490638618e-07, + "loss": 0.0777, + "step": 3292 + }, + { + "epoch": 2.6655873722553878, + "grad_norm": 1.8968099355697632, + "learning_rate": 6.418714605541698e-07, + "loss": 0.0737, + "step": 3293 + }, + { + "epoch": 2.666396843063847, + "grad_norm": 1.7786742448806763, + "learning_rate": 6.387929264621351e-07, + "loss": 0.0723, + "step": 3294 + }, + { + "epoch": 2.667206313872306, + "grad_norm": 1.7761585712432861, + "learning_rate": 6.357215491413371e-07, + "loss": 0.0757, + "step": 3295 + }, + { + "epoch": 2.668015784680765, + "grad_norm": 1.8842600584030151, + "learning_rate": 6.326573309398853e-07, + "loss": 0.0767, + "step": 3296 + }, + { + "epoch": 2.668825255489224, + "grad_norm": 1.9926681518554688, + "learning_rate": 6.296002742004192e-07, + "loss": 0.0883, + "step": 3297 + }, + { + "epoch": 2.6696347262976827, + "grad_norm": 1.8119494915008545, + "learning_rate": 6.26550381260097e-07, + "loss": 0.0813, + "step": 3298 + }, + { + "epoch": 2.670444197106142, + "grad_norm": 2.0495212078094482, + "learning_rate": 6.23507654450608e-07, + "loss": 0.0818, + "step": 3299 + }, + { + "epoch": 2.671253667914601, + "grad_norm": 1.934613585472107, + "learning_rate": 6.204720960981547e-07, + "loss": 0.0917, + "step": 3300 + }, + { + "epoch": 2.67206313872306, + "grad_norm": 2.0053763389587402, + "learning_rate": 6.174437085234663e-07, + "loss": 0.0818, + "step": 3301 + }, + { + "epoch": 2.6728726095315185, + "grad_norm": 2.0365090370178223, + "learning_rate": 6.144224940417854e-07, + "loss": 0.0772, + "step": 3302 + }, + { + "epoch": 2.6736820803399777, + "grad_norm": 1.9033055305480957, + "learning_rate": 6.114084549628707e-07, + "loss": 0.0778, + "step": 3303 + }, + { + "epoch": 2.674491551148437, + "grad_norm": 1.9633246660232544, + "learning_rate": 6.084015935909959e-07, + "loss": 0.0754, + "step": 3304 + }, + { + "epoch": 2.6753010219568956, + "grad_norm": 2.025421142578125, + "learning_rate": 6.054019122249522e-07, + "loss": 0.0759, + "step": 3305 + }, + { + "epoch": 2.6761104927653547, + "grad_norm": 2.0823163986206055, + "learning_rate": 6.024094131580338e-07, + "loss": 0.0788, + "step": 3306 + }, + { + "epoch": 2.6769199635738135, + "grad_norm": 1.8584526777267456, + "learning_rate": 5.994240986780464e-07, + "loss": 0.0727, + "step": 3307 + }, + { + "epoch": 2.6777294343822726, + "grad_norm": 1.8788436651229858, + "learning_rate": 5.964459710673065e-07, + "loss": 0.0782, + "step": 3308 + }, + { + "epoch": 2.678538905190732, + "grad_norm": 1.82304048538208, + "learning_rate": 5.934750326026317e-07, + "loss": 0.0701, + "step": 3309 + }, + { + "epoch": 2.6793483759991905, + "grad_norm": 1.9539531469345093, + "learning_rate": 5.905112855553453e-07, + "loss": 0.0799, + "step": 3310 + }, + { + "epoch": 2.6801578468076492, + "grad_norm": 1.9266644716262817, + "learning_rate": 5.875547321912722e-07, + "loss": 0.0863, + "step": 3311 + }, + { + "epoch": 2.6809673176161084, + "grad_norm": 1.8945074081420898, + "learning_rate": 5.846053747707403e-07, + "loss": 0.0786, + "step": 3312 + }, + { + "epoch": 2.6817767884245676, + "grad_norm": 2.0739586353302, + "learning_rate": 5.816632155485724e-07, + "loss": 0.0871, + "step": 3313 + }, + { + "epoch": 2.6825862592330263, + "grad_norm": 1.9057435989379883, + "learning_rate": 5.787282567740904e-07, + "loss": 0.0812, + "step": 3314 + }, + { + "epoch": 2.6833957300414855, + "grad_norm": 1.9711346626281738, + "learning_rate": 5.758005006911072e-07, + "loss": 0.0868, + "step": 3315 + }, + { + "epoch": 2.684205200849944, + "grad_norm": 1.9181876182556152, + "learning_rate": 5.728799495379355e-07, + "loss": 0.0692, + "step": 3316 + }, + { + "epoch": 2.6850146716584034, + "grad_norm": 1.93299400806427, + "learning_rate": 5.699666055473774e-07, + "loss": 0.0742, + "step": 3317 + }, + { + "epoch": 2.6858241424668625, + "grad_norm": 1.8523931503295898, + "learning_rate": 5.670604709467231e-07, + "loss": 0.0765, + "step": 3318 + }, + { + "epoch": 2.6866336132753212, + "grad_norm": 1.9909532070159912, + "learning_rate": 5.641615479577512e-07, + "loss": 0.0771, + "step": 3319 + }, + { + "epoch": 2.68744308408378, + "grad_norm": 2.0557963848114014, + "learning_rate": 5.612698387967308e-07, + "loss": 0.0859, + "step": 3320 + }, + { + "epoch": 2.688252554892239, + "grad_norm": 1.9933315515518188, + "learning_rate": 5.583853456744104e-07, + "loss": 0.072, + "step": 3321 + }, + { + "epoch": 2.6890620257006983, + "grad_norm": 2.235741138458252, + "learning_rate": 5.55508070796027e-07, + "loss": 0.0741, + "step": 3322 + }, + { + "epoch": 2.689871496509157, + "grad_norm": 2.075761079788208, + "learning_rate": 5.526380163612943e-07, + "loss": 0.095, + "step": 3323 + }, + { + "epoch": 2.690680967317616, + "grad_norm": 1.7902977466583252, + "learning_rate": 5.497751845644105e-07, + "loss": 0.0777, + "step": 3324 + }, + { + "epoch": 2.691490438126075, + "grad_norm": 1.7989822626113892, + "learning_rate": 5.469195775940495e-07, + "loss": 0.0721, + "step": 3325 + }, + { + "epoch": 2.692299908934534, + "grad_norm": 1.9305813312530518, + "learning_rate": 5.440711976333601e-07, + "loss": 0.0817, + "step": 3326 + }, + { + "epoch": 2.6931093797429932, + "grad_norm": 1.8135226964950562, + "learning_rate": 5.412300468599696e-07, + "loss": 0.0818, + "step": 3327 + }, + { + "epoch": 2.693918850551452, + "grad_norm": 1.9794269800186157, + "learning_rate": 5.383961274459771e-07, + "loss": 0.076, + "step": 3328 + }, + { + "epoch": 2.6947283213599107, + "grad_norm": 1.916674017906189, + "learning_rate": 5.355694415579515e-07, + "loss": 0.076, + "step": 3329 + }, + { + "epoch": 2.69553779216837, + "grad_norm": 2.14229416847229, + "learning_rate": 5.327499913569345e-07, + "loss": 0.0803, + "step": 3330 + }, + { + "epoch": 2.696347262976829, + "grad_norm": 1.794236421585083, + "learning_rate": 5.299377789984316e-07, + "loss": 0.0672, + "step": 3331 + }, + { + "epoch": 2.6971567337852878, + "grad_norm": 2.038097381591797, + "learning_rate": 5.271328066324211e-07, + "loss": 0.0788, + "step": 3332 + }, + { + "epoch": 2.697966204593747, + "grad_norm": 1.7536190748214722, + "learning_rate": 5.243350764033405e-07, + "loss": 0.0671, + "step": 3333 + }, + { + "epoch": 2.6987756754022056, + "grad_norm": 1.9153467416763306, + "learning_rate": 5.215445904500949e-07, + "loss": 0.083, + "step": 3334 + }, + { + "epoch": 2.699585146210665, + "grad_norm": 1.8422192335128784, + "learning_rate": 5.187613509060463e-07, + "loss": 0.0745, + "step": 3335 + }, + { + "epoch": 2.700394617019124, + "grad_norm": 2.006178855895996, + "learning_rate": 5.159853598990227e-07, + "loss": 0.0786, + "step": 3336 + }, + { + "epoch": 2.7012040878275827, + "grad_norm": 2.0464086532592773, + "learning_rate": 5.132166195513055e-07, + "loss": 0.0806, + "step": 3337 + }, + { + "epoch": 2.7020135586360414, + "grad_norm": 2.1011557579040527, + "learning_rate": 5.104551319796347e-07, + "loss": 0.0813, + "step": 3338 + }, + { + "epoch": 2.7028230294445006, + "grad_norm": 2.0979578495025635, + "learning_rate": 5.077008992952049e-07, + "loss": 0.0794, + "step": 3339 + }, + { + "epoch": 2.7036325002529598, + "grad_norm": 1.865708351135254, + "learning_rate": 5.049539236036671e-07, + "loss": 0.082, + "step": 3340 + }, + { + "epoch": 2.7044419710614185, + "grad_norm": 1.8453295230865479, + "learning_rate": 5.022142070051205e-07, + "loss": 0.0685, + "step": 3341 + }, + { + "epoch": 2.7052514418698776, + "grad_norm": 1.8258355855941772, + "learning_rate": 4.994817515941153e-07, + "loss": 0.0743, + "step": 3342 + }, + { + "epoch": 2.7060609126783364, + "grad_norm": 2.0706541538238525, + "learning_rate": 4.967565594596512e-07, + "loss": 0.0894, + "step": 3343 + }, + { + "epoch": 2.7068703834867955, + "grad_norm": 1.868679404258728, + "learning_rate": 4.940386326851753e-07, + "loss": 0.0763, + "step": 3344 + }, + { + "epoch": 2.7076798542952547, + "grad_norm": 1.8831230401992798, + "learning_rate": 4.913279733485821e-07, + "loss": 0.0836, + "step": 3345 + }, + { + "epoch": 2.7084893251037134, + "grad_norm": 1.973752737045288, + "learning_rate": 4.886245835222059e-07, + "loss": 0.0824, + "step": 3346 + }, + { + "epoch": 2.709298795912172, + "grad_norm": 1.8348760604858398, + "learning_rate": 4.859284652728236e-07, + "loss": 0.0782, + "step": 3347 + }, + { + "epoch": 2.7101082667206313, + "grad_norm": 1.7267906665802002, + "learning_rate": 4.832396206616574e-07, + "loss": 0.0664, + "step": 3348 + }, + { + "epoch": 2.7109177375290905, + "grad_norm": 1.987746238708496, + "learning_rate": 4.805580517443642e-07, + "loss": 0.0801, + "step": 3349 + }, + { + "epoch": 2.711727208337549, + "grad_norm": 1.8700426816940308, + "learning_rate": 4.778837605710418e-07, + "loss": 0.0765, + "step": 3350 + }, + { + "epoch": 2.7125366791460084, + "grad_norm": 1.9483146667480469, + "learning_rate": 4.752167491862203e-07, + "loss": 0.0808, + "step": 3351 + }, + { + "epoch": 2.713346149954467, + "grad_norm": 1.9466547966003418, + "learning_rate": 4.725570196288687e-07, + "loss": 0.0768, + "step": 3352 + }, + { + "epoch": 2.7141556207629263, + "grad_norm": 2.081078052520752, + "learning_rate": 4.69904573932386e-07, + "loss": 0.0864, + "step": 3353 + }, + { + "epoch": 2.7149650915713854, + "grad_norm": 1.869705080986023, + "learning_rate": 4.672594141246023e-07, + "loss": 0.0778, + "step": 3354 + }, + { + "epoch": 2.715774562379844, + "grad_norm": 1.9531344175338745, + "learning_rate": 4.6462154222778e-07, + "loss": 0.0748, + "step": 3355 + }, + { + "epoch": 2.716584033188303, + "grad_norm": 2.084768295288086, + "learning_rate": 4.619909602586059e-07, + "loss": 0.0865, + "step": 3356 + }, + { + "epoch": 2.717393503996762, + "grad_norm": 1.8469882011413574, + "learning_rate": 4.593676702282002e-07, + "loss": 0.0768, + "step": 3357 + }, + { + "epoch": 2.718202974805221, + "grad_norm": 1.962125539779663, + "learning_rate": 4.5675167414210073e-07, + "loss": 0.0774, + "step": 3358 + }, + { + "epoch": 2.71901244561368, + "grad_norm": 1.7526605129241943, + "learning_rate": 4.541429740002734e-07, + "loss": 0.0788, + "step": 3359 + }, + { + "epoch": 2.719821916422139, + "grad_norm": 1.9244959354400635, + "learning_rate": 4.5154157179710614e-07, + "loss": 0.0832, + "step": 3360 + }, + { + "epoch": 2.720631387230598, + "grad_norm": 1.8838040828704834, + "learning_rate": 4.4894746952140467e-07, + "loss": 0.0814, + "step": 3361 + }, + { + "epoch": 2.721440858039057, + "grad_norm": 1.8750816583633423, + "learning_rate": 4.46360669156396e-07, + "loss": 0.0776, + "step": 3362 + }, + { + "epoch": 2.722250328847516, + "grad_norm": 1.9816977977752686, + "learning_rate": 4.43781172679727e-07, + "loss": 0.0831, + "step": 3363 + }, + { + "epoch": 2.723059799655975, + "grad_norm": 1.8821675777435303, + "learning_rate": 4.412089820634569e-07, + "loss": 0.0827, + "step": 3364 + }, + { + "epoch": 2.7238692704644336, + "grad_norm": 1.8935624361038208, + "learning_rate": 4.3864409927405927e-07, + "loss": 0.0842, + "step": 3365 + }, + { + "epoch": 2.724678741272893, + "grad_norm": 1.7043288946151733, + "learning_rate": 4.360865262724223e-07, + "loss": 0.0713, + "step": 3366 + }, + { + "epoch": 2.725488212081352, + "grad_norm": 2.0001611709594727, + "learning_rate": 4.335362650138475e-07, + "loss": 0.0823, + "step": 3367 + }, + { + "epoch": 2.7262976828898107, + "grad_norm": 1.9384355545043945, + "learning_rate": 4.3099331744804404e-07, + "loss": 0.0818, + "step": 3368 + }, + { + "epoch": 2.72710715369827, + "grad_norm": 1.9076591730117798, + "learning_rate": 4.2845768551912916e-07, + "loss": 0.0773, + "step": 3369 + }, + { + "epoch": 2.7279166245067286, + "grad_norm": 1.8981099128723145, + "learning_rate": 4.259293711656287e-07, + "loss": 0.0794, + "step": 3370 + }, + { + "epoch": 2.7287260953151877, + "grad_norm": 1.9865577220916748, + "learning_rate": 4.234083763204755e-07, + "loss": 0.0804, + "step": 3371 + }, + { + "epoch": 2.729535566123647, + "grad_norm": 1.8678960800170898, + "learning_rate": 4.208947029110011e-07, + "loss": 0.0717, + "step": 3372 + }, + { + "epoch": 2.7303450369321056, + "grad_norm": 1.913780689239502, + "learning_rate": 4.183883528589461e-07, + "loss": 0.0778, + "step": 3373 + }, + { + "epoch": 2.7311545077405643, + "grad_norm": 2.038790464401245, + "learning_rate": 4.158893280804477e-07, + "loss": 0.0865, + "step": 3374 + }, + { + "epoch": 2.7319639785490235, + "grad_norm": 1.9504194259643555, + "learning_rate": 4.1339763048604764e-07, + "loss": 0.0757, + "step": 3375 + }, + { + "epoch": 2.7327734493574827, + "grad_norm": 1.988416075706482, + "learning_rate": 4.109132619806799e-07, + "loss": 0.0829, + "step": 3376 + }, + { + "epoch": 2.7335829201659414, + "grad_norm": 1.805504322052002, + "learning_rate": 4.084362244636797e-07, + "loss": 0.0745, + "step": 3377 + }, + { + "epoch": 2.7343923909744006, + "grad_norm": 1.905112862586975, + "learning_rate": 4.059665198287743e-07, + "loss": 0.0775, + "step": 3378 + }, + { + "epoch": 2.7352018617828593, + "grad_norm": 2.0920276641845703, + "learning_rate": 4.035041499640879e-07, + "loss": 0.0814, + "step": 3379 + }, + { + "epoch": 2.7360113325913185, + "grad_norm": 1.9528201818466187, + "learning_rate": 4.010491167521391e-07, + "loss": 0.0714, + "step": 3380 + }, + { + "epoch": 2.7368208033997776, + "grad_norm": 1.7633929252624512, + "learning_rate": 3.986014220698309e-07, + "loss": 0.0745, + "step": 3381 + }, + { + "epoch": 2.7376302742082363, + "grad_norm": 1.9347256422042847, + "learning_rate": 3.9616106778845863e-07, + "loss": 0.0794, + "step": 3382 + }, + { + "epoch": 2.738439745016695, + "grad_norm": 1.9771931171417236, + "learning_rate": 3.9372805577370997e-07, + "loss": 0.0864, + "step": 3383 + }, + { + "epoch": 2.7392492158251542, + "grad_norm": 2.0302586555480957, + "learning_rate": 3.9130238788565256e-07, + "loss": 0.08, + "step": 3384 + }, + { + "epoch": 2.7400586866336134, + "grad_norm": 2.00186824798584, + "learning_rate": 3.888840659787463e-07, + "loss": 0.0837, + "step": 3385 + }, + { + "epoch": 2.740868157442072, + "grad_norm": 1.7474604845046997, + "learning_rate": 3.864730919018289e-07, + "loss": 0.0752, + "step": 3386 + }, + { + "epoch": 2.7416776282505313, + "grad_norm": 1.8628708124160767, + "learning_rate": 3.840694674981249e-07, + "loss": 0.077, + "step": 3387 + }, + { + "epoch": 2.74248709905899, + "grad_norm": 1.965781331062317, + "learning_rate": 3.8167319460523653e-07, + "loss": 0.0782, + "step": 3388 + }, + { + "epoch": 2.743296569867449, + "grad_norm": 2.092222213745117, + "learning_rate": 3.792842750551473e-07, + "loss": 0.0839, + "step": 3389 + }, + { + "epoch": 2.7441060406759084, + "grad_norm": 1.8017605543136597, + "learning_rate": 3.7690271067421957e-07, + "loss": 0.0701, + "step": 3390 + }, + { + "epoch": 2.744915511484367, + "grad_norm": 1.861026644706726, + "learning_rate": 3.745285032831936e-07, + "loss": 0.0768, + "step": 3391 + }, + { + "epoch": 2.745724982292826, + "grad_norm": 2.1091067790985107, + "learning_rate": 3.721616546971818e-07, + "loss": 0.0777, + "step": 3392 + }, + { + "epoch": 2.746534453101285, + "grad_norm": 2.2249951362609863, + "learning_rate": 3.6980216672567235e-07, + "loss": 0.0815, + "step": 3393 + }, + { + "epoch": 2.747343923909744, + "grad_norm": 1.8852616548538208, + "learning_rate": 3.674500411725246e-07, + "loss": 0.0801, + "step": 3394 + }, + { + "epoch": 2.748153394718203, + "grad_norm": 1.9622523784637451, + "learning_rate": 3.6510527983597334e-07, + "loss": 0.0779, + "step": 3395 + }, + { + "epoch": 2.748962865526662, + "grad_norm": 1.808554768562317, + "learning_rate": 3.6276788450862133e-07, + "loss": 0.0721, + "step": 3396 + }, + { + "epoch": 2.7497723363351207, + "grad_norm": 1.9335891008377075, + "learning_rate": 3.604378569774369e-07, + "loss": 0.0802, + "step": 3397 + }, + { + "epoch": 2.75058180714358, + "grad_norm": 1.9388731718063354, + "learning_rate": 3.5811519902375857e-07, + "loss": 0.0817, + "step": 3398 + }, + { + "epoch": 2.751391277952039, + "grad_norm": 2.1183152198791504, + "learning_rate": 3.557999124232925e-07, + "loss": 0.0768, + "step": 3399 + }, + { + "epoch": 2.752200748760498, + "grad_norm": 1.8575383424758911, + "learning_rate": 3.5349199894610496e-07, + "loss": 0.0855, + "step": 3400 + }, + { + "epoch": 2.7530102195689565, + "grad_norm": 1.8268486261367798, + "learning_rate": 3.511914603566269e-07, + "loss": 0.0849, + "step": 3401 + }, + { + "epoch": 2.7538196903774157, + "grad_norm": 1.8394050598144531, + "learning_rate": 3.4889829841365243e-07, + "loss": 0.0819, + "step": 3402 + }, + { + "epoch": 2.754629161185875, + "grad_norm": 2.0226922035217285, + "learning_rate": 3.4661251487033696e-07, + "loss": 0.0731, + "step": 3403 + }, + { + "epoch": 2.7554386319943336, + "grad_norm": 1.8868749141693115, + "learning_rate": 3.443341114741916e-07, + "loss": 0.0837, + "step": 3404 + }, + { + "epoch": 2.7562481028027928, + "grad_norm": 1.8690674304962158, + "learning_rate": 3.420630899670863e-07, + "loss": 0.0764, + "step": 3405 + }, + { + "epoch": 2.7570575736112515, + "grad_norm": 2.186863899230957, + "learning_rate": 3.39799452085251e-07, + "loss": 0.093, + "step": 3406 + }, + { + "epoch": 2.7578670444197106, + "grad_norm": 1.7671945095062256, + "learning_rate": 3.375431995592637e-07, + "loss": 0.0731, + "step": 3407 + }, + { + "epoch": 2.75867651522817, + "grad_norm": 1.9746793508529663, + "learning_rate": 3.3529433411406466e-07, + "loss": 0.0836, + "step": 3408 + }, + { + "epoch": 2.7594859860366285, + "grad_norm": 1.912804126739502, + "learning_rate": 3.3305285746893954e-07, + "loss": 0.076, + "step": 3409 + }, + { + "epoch": 2.7602954568450873, + "grad_norm": 1.9972152709960938, + "learning_rate": 3.308187713375299e-07, + "loss": 0.0924, + "step": 3410 + }, + { + "epoch": 2.7611049276535464, + "grad_norm": 2.0249578952789307, + "learning_rate": 3.28592077427824e-07, + "loss": 0.0792, + "step": 3411 + }, + { + "epoch": 2.7619143984620056, + "grad_norm": 2.1599841117858887, + "learning_rate": 3.2637277744215987e-07, + "loss": 0.0997, + "step": 3412 + }, + { + "epoch": 2.7627238692704643, + "grad_norm": 1.9847511053085327, + "learning_rate": 3.241608730772228e-07, + "loss": 0.0825, + "step": 3413 + }, + { + "epoch": 2.7635333400789235, + "grad_norm": 1.7137846946716309, + "learning_rate": 3.2195636602404547e-07, + "loss": 0.078, + "step": 3414 + }, + { + "epoch": 2.764342810887382, + "grad_norm": 2.0172767639160156, + "learning_rate": 3.19759257968002e-07, + "loss": 0.0791, + "step": 3415 + }, + { + "epoch": 2.7651522816958414, + "grad_norm": 2.090937852859497, + "learning_rate": 3.175695505888132e-07, + "loss": 0.0747, + "step": 3416 + }, + { + "epoch": 2.7659617525043005, + "grad_norm": 1.8762423992156982, + "learning_rate": 3.153872455605378e-07, + "loss": 0.0827, + "step": 3417 + }, + { + "epoch": 2.7667712233127593, + "grad_norm": 1.8672865629196167, + "learning_rate": 3.1321234455158e-07, + "loss": 0.0672, + "step": 3418 + }, + { + "epoch": 2.767580694121218, + "grad_norm": 1.9550220966339111, + "learning_rate": 3.11044849224682e-07, + "loss": 0.0847, + "step": 3419 + }, + { + "epoch": 2.768390164929677, + "grad_norm": 2.1616694927215576, + "learning_rate": 3.088847612369217e-07, + "loss": 0.0845, + "step": 3420 + }, + { + "epoch": 2.7691996357381363, + "grad_norm": 1.9068938493728638, + "learning_rate": 3.06732082239718e-07, + "loss": 0.0715, + "step": 3421 + }, + { + "epoch": 2.770009106546595, + "grad_norm": 1.729872703552246, + "learning_rate": 3.045868138788222e-07, + "loss": 0.0702, + "step": 3422 + }, + { + "epoch": 2.770818577355054, + "grad_norm": 2.0457587242126465, + "learning_rate": 3.024489577943224e-07, + "loss": 0.0801, + "step": 3423 + }, + { + "epoch": 2.771628048163513, + "grad_norm": 2.0106201171875, + "learning_rate": 3.0031851562063676e-07, + "loss": 0.0768, + "step": 3424 + }, + { + "epoch": 2.772437518971972, + "grad_norm": 2.1421802043914795, + "learning_rate": 2.9819548898652016e-07, + "loss": 0.0824, + "step": 3425 + }, + { + "epoch": 2.7732469897804313, + "grad_norm": 1.605584740638733, + "learning_rate": 2.9607987951505434e-07, + "loss": 0.0678, + "step": 3426 + }, + { + "epoch": 2.77405646058889, + "grad_norm": 1.998680591583252, + "learning_rate": 2.939716888236532e-07, + "loss": 0.0879, + "step": 3427 + }, + { + "epoch": 2.7748659313973487, + "grad_norm": 1.9061836004257202, + "learning_rate": 2.9187091852405534e-07, + "loss": 0.0766, + "step": 3428 + }, + { + "epoch": 2.775675402205808, + "grad_norm": 2.594526767730713, + "learning_rate": 2.897775702223282e-07, + "loss": 0.074, + "step": 3429 + }, + { + "epoch": 2.776484873014267, + "grad_norm": 1.9006608724594116, + "learning_rate": 2.8769164551886717e-07, + "loss": 0.0732, + "step": 3430 + }, + { + "epoch": 2.7772943438227258, + "grad_norm": 1.6783177852630615, + "learning_rate": 2.8561314600838997e-07, + "loss": 0.0685, + "step": 3431 + }, + { + "epoch": 2.778103814631185, + "grad_norm": 1.7513998746871948, + "learning_rate": 2.835420732799377e-07, + "loss": 0.0667, + "step": 3432 + }, + { + "epoch": 2.7789132854396437, + "grad_norm": 1.8435324430465698, + "learning_rate": 2.814784289168737e-07, + "loss": 0.0726, + "step": 3433 + }, + { + "epoch": 2.779722756248103, + "grad_norm": 1.7977373600006104, + "learning_rate": 2.794222144968828e-07, + "loss": 0.0733, + "step": 3434 + }, + { + "epoch": 2.780532227056562, + "grad_norm": 1.8310524225234985, + "learning_rate": 2.773734315919674e-07, + "loss": 0.0768, + "step": 3435 + }, + { + "epoch": 2.7813416978650207, + "grad_norm": 1.878934621810913, + "learning_rate": 2.7533208176845237e-07, + "loss": 0.0848, + "step": 3436 + }, + { + "epoch": 2.7821511686734794, + "grad_norm": 1.8448275327682495, + "learning_rate": 2.7329816658697605e-07, + "loss": 0.0754, + "step": 3437 + }, + { + "epoch": 2.7829606394819386, + "grad_norm": 1.87752366065979, + "learning_rate": 2.7127168760249676e-07, + "loss": 0.0756, + "step": 3438 + }, + { + "epoch": 2.783770110290398, + "grad_norm": 1.999161720275879, + "learning_rate": 2.6925264636428306e-07, + "loss": 0.0816, + "step": 3439 + }, + { + "epoch": 2.7845795810988565, + "grad_norm": 2.004373788833618, + "learning_rate": 2.6724104441592015e-07, + "loss": 0.0767, + "step": 3440 + }, + { + "epoch": 2.7853890519073157, + "grad_norm": 2.060027599334717, + "learning_rate": 2.6523688329530563e-07, + "loss": 0.0912, + "step": 3441 + }, + { + "epoch": 2.7861985227157744, + "grad_norm": 1.8395438194274902, + "learning_rate": 2.632401645346494e-07, + "loss": 0.0709, + "step": 3442 + }, + { + "epoch": 2.7870079935242336, + "grad_norm": 1.8548438549041748, + "learning_rate": 2.612508896604693e-07, + "loss": 0.0795, + "step": 3443 + }, + { + "epoch": 2.7878174643326927, + "grad_norm": 1.9192556142807007, + "learning_rate": 2.592690601935932e-07, + "loss": 0.0772, + "step": 3444 + }, + { + "epoch": 2.7886269351411515, + "grad_norm": 1.9428927898406982, + "learning_rate": 2.5729467764915585e-07, + "loss": 0.0757, + "step": 3445 + }, + { + "epoch": 2.78943640594961, + "grad_norm": 1.8935883045196533, + "learning_rate": 2.5532774353660084e-07, + "loss": 0.079, + "step": 3446 + }, + { + "epoch": 2.7902458767580693, + "grad_norm": 1.9204962253570557, + "learning_rate": 2.533682593596776e-07, + "loss": 0.0732, + "step": 3447 + }, + { + "epoch": 2.7910553475665285, + "grad_norm": 2.037198066711426, + "learning_rate": 2.5141622661643774e-07, + "loss": 0.0835, + "step": 3448 + }, + { + "epoch": 2.7918648183749872, + "grad_norm": 1.9466896057128906, + "learning_rate": 2.494716467992353e-07, + "loss": 0.0828, + "step": 3449 + }, + { + "epoch": 2.7926742891834464, + "grad_norm": 1.881928563117981, + "learning_rate": 2.4753452139473e-07, + "loss": 0.0834, + "step": 3450 + }, + { + "epoch": 2.793483759991905, + "grad_norm": 1.972734808921814, + "learning_rate": 2.4560485188388163e-07, + "loss": 0.074, + "step": 3451 + }, + { + "epoch": 2.7942932308003643, + "grad_norm": 1.9788212776184082, + "learning_rate": 2.436826397419456e-07, + "loss": 0.0864, + "step": 3452 + }, + { + "epoch": 2.7951027016088235, + "grad_norm": 1.8061633110046387, + "learning_rate": 2.4176788643848205e-07, + "loss": 0.0727, + "step": 3453 + }, + { + "epoch": 2.795912172417282, + "grad_norm": 1.9645919799804688, + "learning_rate": 2.398605934373477e-07, + "loss": 0.0833, + "step": 3454 + }, + { + "epoch": 2.796721643225741, + "grad_norm": 1.8695896863937378, + "learning_rate": 2.379607621966906e-07, + "loss": 0.0739, + "step": 3455 + }, + { + "epoch": 2.7975311140342, + "grad_norm": 1.9630117416381836, + "learning_rate": 2.360683941689601e-07, + "loss": 0.0684, + "step": 3456 + }, + { + "epoch": 2.7983405848426592, + "grad_norm": 1.9637010097503662, + "learning_rate": 2.3418349080089663e-07, + "loss": 0.0724, + "step": 3457 + }, + { + "epoch": 2.799150055651118, + "grad_norm": 1.7701715230941772, + "learning_rate": 2.323060535335353e-07, + "loss": 0.0719, + "step": 3458 + }, + { + "epoch": 2.799959526459577, + "grad_norm": 1.8289519548416138, + "learning_rate": 2.304360838022024e-07, + "loss": 0.076, + "step": 3459 + }, + { + "epoch": 2.800768997268036, + "grad_norm": 1.9949322938919067, + "learning_rate": 2.2857358303651433e-07, + "loss": 0.082, + "step": 3460 + }, + { + "epoch": 2.801578468076495, + "grad_norm": 1.9749974012374878, + "learning_rate": 2.26718552660381e-07, + "loss": 0.0745, + "step": 3461 + }, + { + "epoch": 2.802387938884954, + "grad_norm": 2.0284533500671387, + "learning_rate": 2.2487099409199686e-07, + "loss": 0.0912, + "step": 3462 + }, + { + "epoch": 2.803197409693413, + "grad_norm": 1.9058609008789062, + "learning_rate": 2.2303090874384426e-07, + "loss": 0.0803, + "step": 3463 + }, + { + "epoch": 2.8040068805018716, + "grad_norm": 1.9215221405029297, + "learning_rate": 2.2119829802269677e-07, + "loss": 0.0799, + "step": 3464 + }, + { + "epoch": 2.804816351310331, + "grad_norm": 1.9013926982879639, + "learning_rate": 2.1937316332960924e-07, + "loss": 0.0785, + "step": 3465 + }, + { + "epoch": 2.80562582211879, + "grad_norm": 1.7817100286483765, + "learning_rate": 2.1755550605992216e-07, + "loss": 0.0778, + "step": 3466 + }, + { + "epoch": 2.8064352929272487, + "grad_norm": 1.9394621849060059, + "learning_rate": 2.157453276032595e-07, + "loss": 0.081, + "step": 3467 + }, + { + "epoch": 2.807244763735708, + "grad_norm": 1.9406628608703613, + "learning_rate": 2.139426293435265e-07, + "loss": 0.0756, + "step": 3468 + }, + { + "epoch": 2.8080542345441666, + "grad_norm": 2.021958351135254, + "learning_rate": 2.121474126589118e-07, + "loss": 0.0833, + "step": 3469 + }, + { + "epoch": 2.8088637053526258, + "grad_norm": 1.9084668159484863, + "learning_rate": 2.1035967892188314e-07, + "loss": 0.0806, + "step": 3470 + }, + { + "epoch": 2.809673176161085, + "grad_norm": 1.8481040000915527, + "learning_rate": 2.085794294991872e-07, + "loss": 0.0735, + "step": 3471 + }, + { + "epoch": 2.8104826469695436, + "grad_norm": 1.9322901964187622, + "learning_rate": 2.068066657518486e-07, + "loss": 0.0738, + "step": 3472 + }, + { + "epoch": 2.8112921177780024, + "grad_norm": 1.9235726594924927, + "learning_rate": 2.05041389035171e-07, + "loss": 0.0836, + "step": 3473 + }, + { + "epoch": 2.8121015885864615, + "grad_norm": 2.0797343254089355, + "learning_rate": 2.0328360069873266e-07, + "loss": 0.0872, + "step": 3474 + }, + { + "epoch": 2.8129110593949207, + "grad_norm": 1.8538960218429565, + "learning_rate": 2.0153330208638412e-07, + "loss": 0.0734, + "step": 3475 + }, + { + "epoch": 2.8137205302033794, + "grad_norm": 1.6988213062286377, + "learning_rate": 1.9979049453625498e-07, + "loss": 0.0654, + "step": 3476 + }, + { + "epoch": 2.8145300010118386, + "grad_norm": 1.7093678712844849, + "learning_rate": 1.9805517938074615e-07, + "loss": 0.0715, + "step": 3477 + }, + { + "epoch": 2.8153394718202973, + "grad_norm": 1.7742198705673218, + "learning_rate": 1.9632735794652857e-07, + "loss": 0.0685, + "step": 3478 + }, + { + "epoch": 2.8161489426287565, + "grad_norm": 2.0381946563720703, + "learning_rate": 1.9460703155454564e-07, + "loss": 0.0844, + "step": 3479 + }, + { + "epoch": 2.8169584134372156, + "grad_norm": 1.8216747045516968, + "learning_rate": 1.928942015200097e-07, + "loss": 0.0826, + "step": 3480 + }, + { + "epoch": 2.8177678842456744, + "grad_norm": 1.9380648136138916, + "learning_rate": 1.9118886915240442e-07, + "loss": 0.0714, + "step": 3481 + }, + { + "epoch": 2.818577355054133, + "grad_norm": 1.848543643951416, + "learning_rate": 1.8949103575547912e-07, + "loss": 0.0766, + "step": 3482 + }, + { + "epoch": 2.8193868258625923, + "grad_norm": 2.4003846645355225, + "learning_rate": 1.8780070262724993e-07, + "loss": 0.0938, + "step": 3483 + }, + { + "epoch": 2.8201962966710514, + "grad_norm": 1.9345378875732422, + "learning_rate": 1.861178710600009e-07, + "loss": 0.0916, + "step": 3484 + }, + { + "epoch": 2.82100576747951, + "grad_norm": 1.9023882150650024, + "learning_rate": 1.844425423402796e-07, + "loss": 0.0783, + "step": 3485 + }, + { + "epoch": 2.8218152382879693, + "grad_norm": 1.8425052165985107, + "learning_rate": 1.8277471774889698e-07, + "loss": 0.0773, + "step": 3486 + }, + { + "epoch": 2.822624709096428, + "grad_norm": 1.9461959600448608, + "learning_rate": 1.8111439856092984e-07, + "loss": 0.0812, + "step": 3487 + }, + { + "epoch": 2.823434179904887, + "grad_norm": 2.0066752433776855, + "learning_rate": 1.7946158604571162e-07, + "loss": 0.0851, + "step": 3488 + }, + { + "epoch": 2.8242436507133464, + "grad_norm": 1.8375251293182373, + "learning_rate": 1.7781628146684382e-07, + "loss": 0.0698, + "step": 3489 + }, + { + "epoch": 2.825053121521805, + "grad_norm": 1.9881941080093384, + "learning_rate": 1.7617848608218247e-07, + "loss": 0.0751, + "step": 3490 + }, + { + "epoch": 2.825862592330264, + "grad_norm": 1.8756792545318604, + "learning_rate": 1.7454820114384374e-07, + "loss": 0.0761, + "step": 3491 + }, + { + "epoch": 2.826672063138723, + "grad_norm": 1.9936344623565674, + "learning_rate": 1.72925427898204e-07, + "loss": 0.0812, + "step": 3492 + }, + { + "epoch": 2.827481533947182, + "grad_norm": 1.8127198219299316, + "learning_rate": 1.7131016758589636e-07, + "loss": 0.0775, + "step": 3493 + }, + { + "epoch": 2.828291004755641, + "grad_norm": 2.0459959506988525, + "learning_rate": 1.6970242144180859e-07, + "loss": 0.0859, + "step": 3494 + }, + { + "epoch": 2.8291004755641, + "grad_norm": 1.916442632675171, + "learning_rate": 1.6810219069508527e-07, + "loss": 0.075, + "step": 3495 + }, + { + "epoch": 2.8299099463725588, + "grad_norm": 1.7589393854141235, + "learning_rate": 1.665094765691222e-07, + "loss": 0.0811, + "step": 3496 + }, + { + "epoch": 2.830719417181018, + "grad_norm": 1.7405444383621216, + "learning_rate": 1.6492428028157536e-07, + "loss": 0.0774, + "step": 3497 + }, + { + "epoch": 2.831528887989477, + "grad_norm": 2.001612424850464, + "learning_rate": 1.6334660304434535e-07, + "loss": 0.0848, + "step": 3498 + }, + { + "epoch": 2.832338358797936, + "grad_norm": 2.0691614151000977, + "learning_rate": 1.6177644606359066e-07, + "loss": 0.0855, + "step": 3499 + }, + { + "epoch": 2.8331478296063946, + "grad_norm": 1.8450762033462524, + "learning_rate": 1.6021381053971664e-07, + "loss": 0.0793, + "step": 3500 + }, + { + "epoch": 2.8339573004148537, + "grad_norm": 1.9585386514663696, + "learning_rate": 1.5865869766737986e-07, + "loss": 0.0897, + "step": 3501 + }, + { + "epoch": 2.834766771223313, + "grad_norm": 1.797786831855774, + "learning_rate": 1.5711110863548594e-07, + "loss": 0.0765, + "step": 3502 + }, + { + "epoch": 2.8355762420317716, + "grad_norm": 1.8074800968170166, + "learning_rate": 1.555710446271863e-07, + "loss": 0.0786, + "step": 3503 + }, + { + "epoch": 2.8363857128402308, + "grad_norm": 1.839112639427185, + "learning_rate": 1.5403850681988242e-07, + "loss": 0.0785, + "step": 3504 + }, + { + "epoch": 2.8371951836486895, + "grad_norm": 1.8675745725631714, + "learning_rate": 1.525134963852204e-07, + "loss": 0.0835, + "step": 3505 + }, + { + "epoch": 2.8380046544571487, + "grad_norm": 1.9390770196914673, + "learning_rate": 1.5099601448909097e-07, + "loss": 0.079, + "step": 3506 + }, + { + "epoch": 2.838814125265608, + "grad_norm": 1.9703547954559326, + "learning_rate": 1.4948606229163054e-07, + "loss": 0.0904, + "step": 3507 + }, + { + "epoch": 2.8396235960740666, + "grad_norm": 1.7114065885543823, + "learning_rate": 1.4798364094721797e-07, + "loss": 0.0769, + "step": 3508 + }, + { + "epoch": 2.8404330668825253, + "grad_norm": 1.9542559385299683, + "learning_rate": 1.464887516044744e-07, + "loss": 0.0842, + "step": 3509 + }, + { + "epoch": 2.8412425376909844, + "grad_norm": 1.7902026176452637, + "learning_rate": 1.4500139540626458e-07, + "loss": 0.0733, + "step": 3510 + }, + { + "epoch": 2.8420520084994436, + "grad_norm": 1.7940454483032227, + "learning_rate": 1.435215734896922e-07, + "loss": 0.074, + "step": 3511 + }, + { + "epoch": 2.8428614793079023, + "grad_norm": 1.8185328245162964, + "learning_rate": 1.420492869861012e-07, + "loss": 0.0734, + "step": 3512 + }, + { + "epoch": 2.8436709501163615, + "grad_norm": 1.7391959428787231, + "learning_rate": 1.4058453702107455e-07, + "loss": 0.0666, + "step": 3513 + }, + { + "epoch": 2.8444804209248202, + "grad_norm": 1.8359476327896118, + "learning_rate": 1.3912732471443313e-07, + "loss": 0.0746, + "step": 3514 + }, + { + "epoch": 2.8452898917332794, + "grad_norm": 1.8736023902893066, + "learning_rate": 1.376776511802369e-07, + "loss": 0.0736, + "step": 3515 + }, + { + "epoch": 2.8460993625417386, + "grad_norm": 2.085463523864746, + "learning_rate": 1.3623551752678154e-07, + "loss": 0.0838, + "step": 3516 + }, + { + "epoch": 2.8469088333501973, + "grad_norm": 1.9128777980804443, + "learning_rate": 1.3480092485659736e-07, + "loss": 0.0847, + "step": 3517 + }, + { + "epoch": 2.847718304158656, + "grad_norm": 1.9302642345428467, + "learning_rate": 1.3337387426645031e-07, + "loss": 0.0765, + "step": 3518 + }, + { + "epoch": 2.848527774967115, + "grad_norm": 1.9896522760391235, + "learning_rate": 1.3195436684733885e-07, + "loss": 0.0885, + "step": 3519 + }, + { + "epoch": 2.8493372457755743, + "grad_norm": 1.9617047309875488, + "learning_rate": 1.3054240368449821e-07, + "loss": 0.0808, + "step": 3520 + }, + { + "epoch": 2.850146716584033, + "grad_norm": 1.8002290725708008, + "learning_rate": 1.2913798585739158e-07, + "loss": 0.0758, + "step": 3521 + }, + { + "epoch": 2.8509561873924922, + "grad_norm": 1.7590972185134888, + "learning_rate": 1.2774111443971672e-07, + "loss": 0.0714, + "step": 3522 + }, + { + "epoch": 2.851765658200951, + "grad_norm": 1.7985774278640747, + "learning_rate": 1.2635179049940049e-07, + "loss": 0.0672, + "step": 3523 + }, + { + "epoch": 2.85257512900941, + "grad_norm": 1.9801772832870483, + "learning_rate": 1.2497001509859997e-07, + "loss": 0.0922, + "step": 3524 + }, + { + "epoch": 2.8533845998178693, + "grad_norm": 1.9024227857589722, + "learning_rate": 1.2359578929370231e-07, + "loss": 0.0797, + "step": 3525 + }, + { + "epoch": 2.854194070626328, + "grad_norm": 1.9259132146835327, + "learning_rate": 1.2222911413532158e-07, + "loss": 0.0782, + "step": 3526 + }, + { + "epoch": 2.8550035414347867, + "grad_norm": 1.9188920259475708, + "learning_rate": 1.2086999066829863e-07, + "loss": 0.079, + "step": 3527 + }, + { + "epoch": 2.855813012243246, + "grad_norm": 1.9162348508834839, + "learning_rate": 1.1951841993170343e-07, + "loss": 0.0751, + "step": 3528 + }, + { + "epoch": 2.856622483051705, + "grad_norm": 1.8251134157180786, + "learning_rate": 1.1817440295883054e-07, + "loss": 0.0733, + "step": 3529 + }, + { + "epoch": 2.857431953860164, + "grad_norm": 1.9522722959518433, + "learning_rate": 1.1683794077719802e-07, + "loss": 0.077, + "step": 3530 + }, + { + "epoch": 2.858241424668623, + "grad_norm": 2.056112289428711, + "learning_rate": 1.1550903440854966e-07, + "loss": 0.0843, + "step": 3531 + }, + { + "epoch": 2.8590508954770817, + "grad_norm": 1.7863446474075317, + "learning_rate": 1.141876848688539e-07, + "loss": 0.0742, + "step": 3532 + }, + { + "epoch": 2.859860366285541, + "grad_norm": 2.0439293384552, + "learning_rate": 1.1287389316829934e-07, + "loss": 0.0854, + "step": 3533 + }, + { + "epoch": 2.860669837094, + "grad_norm": 1.8242279291152954, + "learning_rate": 1.1156766031129917e-07, + "loss": 0.0771, + "step": 3534 + }, + { + "epoch": 2.8614793079024587, + "grad_norm": 1.8393011093139648, + "learning_rate": 1.1026898729648461e-07, + "loss": 0.0703, + "step": 3535 + }, + { + "epoch": 2.8622887787109175, + "grad_norm": 1.8759177923202515, + "learning_rate": 1.0897787511671032e-07, + "loss": 0.0808, + "step": 3536 + }, + { + "epoch": 2.8630982495193766, + "grad_norm": 1.8068702220916748, + "learning_rate": 1.0769432475904785e-07, + "loss": 0.0735, + "step": 3537 + }, + { + "epoch": 2.863907720327836, + "grad_norm": 1.9034488201141357, + "learning_rate": 1.0641833720479e-07, + "loss": 0.0794, + "step": 3538 + }, + { + "epoch": 2.8647171911362945, + "grad_norm": 2.0366086959838867, + "learning_rate": 1.0514991342944758e-07, + "loss": 0.0886, + "step": 3539 + }, + { + "epoch": 2.8655266619447537, + "grad_norm": 1.841306209564209, + "learning_rate": 1.038890544027471e-07, + "loss": 0.0817, + "step": 3540 + }, + { + "epoch": 2.8663361327532124, + "grad_norm": 1.8680078983306885, + "learning_rate": 1.0263576108863194e-07, + "loss": 0.0779, + "step": 3541 + }, + { + "epoch": 2.8671456035616716, + "grad_norm": 1.9264686107635498, + "learning_rate": 1.0139003444526341e-07, + "loss": 0.0824, + "step": 3542 + }, + { + "epoch": 2.8679550743701308, + "grad_norm": 2.052473306655884, + "learning_rate": 1.0015187542501525e-07, + "loss": 0.0771, + "step": 3543 + }, + { + "epoch": 2.8687645451785895, + "grad_norm": 2.058781623840332, + "learning_rate": 9.892128497447696e-08, + "loss": 0.085, + "step": 3544 + }, + { + "epoch": 2.869574015987048, + "grad_norm": 1.9010498523712158, + "learning_rate": 9.769826403445371e-08, + "loss": 0.0816, + "step": 3545 + }, + { + "epoch": 2.8703834867955074, + "grad_norm": 1.9000492095947266, + "learning_rate": 9.648281353996091e-08, + "loss": 0.0813, + "step": 3546 + }, + { + "epoch": 2.8711929576039665, + "grad_norm": 1.8288404941558838, + "learning_rate": 9.52749344202264e-08, + "loss": 0.0724, + "step": 3547 + }, + { + "epoch": 2.8720024284124253, + "grad_norm": 1.9630149602890015, + "learning_rate": 9.407462759869035e-08, + "loss": 0.0816, + "step": 3548 + }, + { + "epoch": 2.8728118992208844, + "grad_norm": 1.939529538154602, + "learning_rate": 9.288189399300429e-08, + "loss": 0.0793, + "step": 3549 + }, + { + "epoch": 2.873621370029343, + "grad_norm": 1.9486548900604248, + "learning_rate": 9.169673451502992e-08, + "loss": 0.0757, + "step": 3550 + }, + { + "epoch": 2.8744308408378023, + "grad_norm": 2.0139758586883545, + "learning_rate": 9.05191500708369e-08, + "loss": 0.0786, + "step": 3551 + }, + { + "epoch": 2.8752403116462615, + "grad_norm": 2.0944249629974365, + "learning_rate": 8.934914156070506e-08, + "loss": 0.0831, + "step": 3552 + }, + { + "epoch": 2.87604978245472, + "grad_norm": 2.011934757232666, + "learning_rate": 8.818670987912226e-08, + "loss": 0.0846, + "step": 3553 + }, + { + "epoch": 2.876859253263179, + "grad_norm": 1.9506481885910034, + "learning_rate": 8.703185591478314e-08, + "loss": 0.0834, + "step": 3554 + }, + { + "epoch": 2.877668724071638, + "grad_norm": 2.0650415420532227, + "learning_rate": 8.588458055058923e-08, + "loss": 0.0846, + "step": 3555 + }, + { + "epoch": 2.8784781948800973, + "grad_norm": 1.8324848413467407, + "learning_rate": 8.474488466364894e-08, + "loss": 0.0728, + "step": 3556 + }, + { + "epoch": 2.879287665688556, + "grad_norm": 2.0325632095336914, + "learning_rate": 8.361276912527528e-08, + "loss": 0.077, + "step": 3557 + }, + { + "epoch": 2.880097136497015, + "grad_norm": 1.9899486303329468, + "learning_rate": 8.248823480098477e-08, + "loss": 0.0815, + "step": 3558 + }, + { + "epoch": 2.880906607305474, + "grad_norm": 1.7435157299041748, + "learning_rate": 8.137128255050197e-08, + "loss": 0.0776, + "step": 3559 + }, + { + "epoch": 2.881716078113933, + "grad_norm": 1.797224998474121, + "learning_rate": 8.026191322774934e-08, + "loss": 0.0737, + "step": 3560 + }, + { + "epoch": 2.882525548922392, + "grad_norm": 1.9771441221237183, + "learning_rate": 7.916012768085735e-08, + "loss": 0.0672, + "step": 3561 + }, + { + "epoch": 2.883335019730851, + "grad_norm": 1.9322302341461182, + "learning_rate": 7.80659267521533e-08, + "loss": 0.0844, + "step": 3562 + }, + { + "epoch": 2.8841444905393097, + "grad_norm": 1.9366923570632935, + "learning_rate": 7.697931127817248e-08, + "loss": 0.0798, + "step": 3563 + }, + { + "epoch": 2.884953961347769, + "grad_norm": 1.7694432735443115, + "learning_rate": 7.590028208964483e-08, + "loss": 0.0762, + "step": 3564 + }, + { + "epoch": 2.885763432156228, + "grad_norm": 2.0923423767089844, + "learning_rate": 7.482884001150381e-08, + "loss": 0.0855, + "step": 3565 + }, + { + "epoch": 2.8865729029646867, + "grad_norm": 1.953178882598877, + "learning_rate": 7.376498586288084e-08, + "loss": 0.0809, + "step": 3566 + }, + { + "epoch": 2.887382373773146, + "grad_norm": 1.7268513441085815, + "learning_rate": 7.270872045710753e-08, + "loss": 0.0716, + "step": 3567 + }, + { + "epoch": 2.8881918445816046, + "grad_norm": 1.8840540647506714, + "learning_rate": 7.16600446017135e-08, + "loss": 0.0738, + "step": 3568 + }, + { + "epoch": 2.8890013153900638, + "grad_norm": 1.8790191411972046, + "learning_rate": 7.061895909842632e-08, + "loss": 0.076, + "step": 3569 + }, + { + "epoch": 2.889810786198523, + "grad_norm": 1.9826503992080688, + "learning_rate": 6.958546474316818e-08, + "loss": 0.0859, + "step": 3570 + }, + { + "epoch": 2.8906202570069817, + "grad_norm": 1.9763182401657104, + "learning_rate": 6.855956232606153e-08, + "loss": 0.0775, + "step": 3571 + }, + { + "epoch": 2.8914297278154404, + "grad_norm": 1.7756454944610596, + "learning_rate": 6.75412526314212e-08, + "loss": 0.0725, + "step": 3572 + }, + { + "epoch": 2.8922391986238996, + "grad_norm": 2.3303139209747314, + "learning_rate": 6.653053643776108e-08, + "loss": 0.0842, + "step": 3573 + }, + { + "epoch": 2.8930486694323587, + "grad_norm": 1.8075859546661377, + "learning_rate": 6.552741451778643e-08, + "loss": 0.0762, + "step": 3574 + }, + { + "epoch": 2.8938581402408174, + "grad_norm": 1.985229253768921, + "learning_rate": 6.453188763839824e-08, + "loss": 0.0838, + "step": 3575 + }, + { + "epoch": 2.8946676110492766, + "grad_norm": 1.8757413625717163, + "learning_rate": 6.354395656068879e-08, + "loss": 0.0676, + "step": 3576 + }, + { + "epoch": 2.8954770818577353, + "grad_norm": 1.7401223182678223, + "learning_rate": 6.256362203994615e-08, + "loss": 0.0766, + "step": 3577 + }, + { + "epoch": 2.8962865526661945, + "grad_norm": 1.8051989078521729, + "learning_rate": 6.159088482564968e-08, + "loss": 0.0709, + "step": 3578 + }, + { + "epoch": 2.8970960234746537, + "grad_norm": 1.9177215099334717, + "learning_rate": 6.062574566147117e-08, + "loss": 0.0778, + "step": 3579 + }, + { + "epoch": 2.8979054942831124, + "grad_norm": 1.9627326726913452, + "learning_rate": 5.96682052852715e-08, + "loss": 0.077, + "step": 3580 + }, + { + "epoch": 2.898714965091571, + "grad_norm": 1.9226256608963013, + "learning_rate": 5.871826442910289e-08, + "loss": 0.0766, + "step": 3581 + }, + { + "epoch": 2.8995244359000303, + "grad_norm": 1.893538475036621, + "learning_rate": 5.777592381920882e-08, + "loss": 0.0747, + "step": 3582 + }, + { + "epoch": 2.9003339067084895, + "grad_norm": 2.0273947715759277, + "learning_rate": 5.6841184176021915e-08, + "loss": 0.0787, + "step": 3583 + }, + { + "epoch": 2.901143377516948, + "grad_norm": 1.9856191873550415, + "learning_rate": 5.591404621416274e-08, + "loss": 0.0734, + "step": 3584 + }, + { + "epoch": 2.9019528483254073, + "grad_norm": 1.8693398237228394, + "learning_rate": 5.499451064244321e-08, + "loss": 0.0769, + "step": 3585 + }, + { + "epoch": 2.902762319133866, + "grad_norm": 1.811806321144104, + "learning_rate": 5.408257816385764e-08, + "loss": 0.0778, + "step": 3586 + }, + { + "epoch": 2.9035717899423252, + "grad_norm": 1.960249423980713, + "learning_rate": 5.3178249475592805e-08, + "loss": 0.0722, + "step": 3587 + }, + { + "epoch": 2.9043812607507844, + "grad_norm": 1.863107442855835, + "learning_rate": 5.228152526902008e-08, + "loss": 0.0748, + "step": 3588 + }, + { + "epoch": 2.905190731559243, + "grad_norm": 1.816367745399475, + "learning_rate": 5.139240622969666e-08, + "loss": 0.0674, + "step": 3589 + }, + { + "epoch": 2.906000202367702, + "grad_norm": 1.9556515216827393, + "learning_rate": 5.051089303736767e-08, + "loss": 0.0788, + "step": 3590 + }, + { + "epoch": 2.906809673176161, + "grad_norm": 1.8568001985549927, + "learning_rate": 4.9636986365961814e-08, + "loss": 0.0815, + "step": 3591 + }, + { + "epoch": 2.90761914398462, + "grad_norm": 1.8013601303100586, + "learning_rate": 4.8770686883591326e-08, + "loss": 0.071, + "step": 3592 + }, + { + "epoch": 2.908428614793079, + "grad_norm": 1.8855005502700806, + "learning_rate": 4.79119952525553e-08, + "loss": 0.0734, + "step": 3593 + }, + { + "epoch": 2.909238085601538, + "grad_norm": 1.985625982284546, + "learning_rate": 4.706091212933528e-08, + "loss": 0.0804, + "step": 3594 + }, + { + "epoch": 2.910047556409997, + "grad_norm": 1.9673521518707275, + "learning_rate": 4.621743816459523e-08, + "loss": 0.0779, + "step": 3595 + }, + { + "epoch": 2.910857027218456, + "grad_norm": 1.8802547454833984, + "learning_rate": 4.538157400318266e-08, + "loss": 0.0862, + "step": 3596 + }, + { + "epoch": 2.911666498026915, + "grad_norm": 1.898134708404541, + "learning_rate": 4.455332028412862e-08, + "loss": 0.081, + "step": 3597 + }, + { + "epoch": 2.912475968835374, + "grad_norm": 2.203346014022827, + "learning_rate": 4.373267764064215e-08, + "loss": 0.0832, + "step": 3598 + }, + { + "epoch": 2.9132854396438326, + "grad_norm": 2.146695137023926, + "learning_rate": 4.2919646700116945e-08, + "loss": 0.0758, + "step": 3599 + }, + { + "epoch": 2.9140949104522917, + "grad_norm": 1.9870094060897827, + "learning_rate": 4.2114228084126904e-08, + "loss": 0.0778, + "step": 3600 + }, + { + "epoch": 2.914904381260751, + "grad_norm": 1.974664568901062, + "learning_rate": 4.131642240842504e-08, + "loss": 0.0763, + "step": 3601 + }, + { + "epoch": 2.9157138520692096, + "grad_norm": 1.922515630722046, + "learning_rate": 4.0526230282944556e-08, + "loss": 0.0744, + "step": 3602 + }, + { + "epoch": 2.916523322877669, + "grad_norm": 1.932558536529541, + "learning_rate": 3.9743652311798886e-08, + "loss": 0.0805, + "step": 3603 + }, + { + "epoch": 2.9173327936861275, + "grad_norm": 1.9623072147369385, + "learning_rate": 3.8968689093279444e-08, + "loss": 0.0859, + "step": 3604 + }, + { + "epoch": 2.9181422644945867, + "grad_norm": 1.9618598222732544, + "learning_rate": 3.820134121985453e-08, + "loss": 0.082, + "step": 3605 + }, + { + "epoch": 2.918951735303046, + "grad_norm": 1.9189358949661255, + "learning_rate": 3.744160927817486e-08, + "loss": 0.0775, + "step": 3606 + }, + { + "epoch": 2.9197612061115046, + "grad_norm": 1.8080579042434692, + "learning_rate": 3.6689493849063616e-08, + "loss": 0.0759, + "step": 3607 + }, + { + "epoch": 2.9205706769199633, + "grad_norm": 2.184316873550415, + "learning_rate": 3.5944995507525284e-08, + "loss": 0.0888, + "step": 3608 + }, + { + "epoch": 2.9213801477284225, + "grad_norm": 1.727544903755188, + "learning_rate": 3.520811482273678e-08, + "loss": 0.06, + "step": 3609 + }, + { + "epoch": 2.9221896185368816, + "grad_norm": 2.0043351650238037, + "learning_rate": 3.447885235805637e-08, + "loss": 0.0823, + "step": 3610 + }, + { + "epoch": 2.9229990893453404, + "grad_norm": 1.862843632698059, + "learning_rate": 3.375720867101251e-08, + "loss": 0.079, + "step": 3611 + }, + { + "epoch": 2.9238085601537995, + "grad_norm": 1.940791368484497, + "learning_rate": 3.304318431331277e-08, + "loss": 0.083, + "step": 3612 + }, + { + "epoch": 2.9246180309622583, + "grad_norm": 1.9732791185379028, + "learning_rate": 3.2336779830839384e-08, + "loss": 0.0759, + "step": 3613 + }, + { + "epoch": 2.9254275017707174, + "grad_norm": 1.8733102083206177, + "learning_rate": 3.163799576364701e-08, + "loss": 0.0733, + "step": 3614 + }, + { + "epoch": 2.9262369725791766, + "grad_norm": 1.8994005918502808, + "learning_rate": 3.09468326459661e-08, + "loss": 0.0784, + "step": 3615 + }, + { + "epoch": 2.9270464433876353, + "grad_norm": 2.116297483444214, + "learning_rate": 3.026329100620173e-08, + "loss": 0.0772, + "step": 3616 + }, + { + "epoch": 2.927855914196094, + "grad_norm": 1.9148296117782593, + "learning_rate": 2.958737136692813e-08, + "loss": 0.081, + "step": 3617 + }, + { + "epoch": 2.928665385004553, + "grad_norm": 1.7786005735397339, + "learning_rate": 2.8919074244896372e-08, + "loss": 0.0723, + "step": 3618 + }, + { + "epoch": 2.9294748558130124, + "grad_norm": 1.9845788478851318, + "learning_rate": 2.8258400151029987e-08, + "loss": 0.0851, + "step": 3619 + }, + { + "epoch": 2.930284326621471, + "grad_norm": 1.8352128267288208, + "learning_rate": 2.7605349590421603e-08, + "loss": 0.0733, + "step": 3620 + }, + { + "epoch": 2.9310937974299303, + "grad_norm": 1.9144186973571777, + "learning_rate": 2.695992306233852e-08, + "loss": 0.0804, + "step": 3621 + }, + { + "epoch": 2.931903268238389, + "grad_norm": 1.8999292850494385, + "learning_rate": 2.6322121060217142e-08, + "loss": 0.0747, + "step": 3622 + }, + { + "epoch": 2.932712739046848, + "grad_norm": 1.8880465030670166, + "learning_rate": 2.569194407166742e-08, + "loss": 0.0774, + "step": 3623 + }, + { + "epoch": 2.9335222098553073, + "grad_norm": 1.9338951110839844, + "learning_rate": 2.506939257846619e-08, + "loss": 0.087, + "step": 3624 + }, + { + "epoch": 2.934331680663766, + "grad_norm": 1.9416056871414185, + "learning_rate": 2.4454467056564958e-08, + "loss": 0.0779, + "step": 3625 + }, + { + "epoch": 2.9351411514722248, + "grad_norm": 1.8942091464996338, + "learning_rate": 2.3847167976081e-08, + "loss": 0.0715, + "step": 3626 + }, + { + "epoch": 2.935950622280684, + "grad_norm": 1.9759517908096313, + "learning_rate": 2.324749580130292e-08, + "loss": 0.0824, + "step": 3627 + }, + { + "epoch": 2.936760093089143, + "grad_norm": 1.859450340270996, + "learning_rate": 2.2655450990688443e-08, + "loss": 0.074, + "step": 3628 + }, + { + "epoch": 2.937569563897602, + "grad_norm": 1.9984499216079712, + "learning_rate": 2.2071033996864387e-08, + "loss": 0.0787, + "step": 3629 + }, + { + "epoch": 2.938379034706061, + "grad_norm": 1.791815996170044, + "learning_rate": 2.149424526662669e-08, + "loss": 0.0762, + "step": 3630 + }, + { + "epoch": 2.9391885055145197, + "grad_norm": 1.98390793800354, + "learning_rate": 2.092508524093595e-08, + "loss": 0.0826, + "step": 3631 + }, + { + "epoch": 2.939997976322979, + "grad_norm": 1.9028316736221313, + "learning_rate": 2.03635543549241e-08, + "loss": 0.0755, + "step": 3632 + }, + { + "epoch": 2.940807447131438, + "grad_norm": 2.127866744995117, + "learning_rate": 1.9809653037888844e-08, + "loss": 0.0842, + "step": 3633 + }, + { + "epoch": 2.9416169179398968, + "grad_norm": 1.783174991607666, + "learning_rate": 1.926338171329478e-08, + "loss": 0.0642, + "step": 3634 + }, + { + "epoch": 2.9424263887483555, + "grad_norm": 1.757351040840149, + "learning_rate": 1.872474079877451e-08, + "loss": 0.071, + "step": 3635 + }, + { + "epoch": 2.9432358595568147, + "grad_norm": 1.956337571144104, + "learning_rate": 1.819373070612751e-08, + "loss": 0.0842, + "step": 3636 + }, + { + "epoch": 2.944045330365274, + "grad_norm": 1.833010196685791, + "learning_rate": 1.7670351841315715e-08, + "loss": 0.074, + "step": 3637 + }, + { + "epoch": 2.9448548011737326, + "grad_norm": 1.9207520484924316, + "learning_rate": 1.715460460447238e-08, + "loss": 0.0725, + "step": 3638 + }, + { + "epoch": 2.9456642719821917, + "grad_norm": 1.7338045835494995, + "learning_rate": 1.6646489389890995e-08, + "loss": 0.0753, + "step": 3639 + }, + { + "epoch": 2.9464737427906504, + "grad_norm": 2.000628709793091, + "learning_rate": 1.614600658603305e-08, + "loss": 0.083, + "step": 3640 + }, + { + "epoch": 2.9472832135991096, + "grad_norm": 1.9100686311721802, + "learning_rate": 1.5653156575525795e-08, + "loss": 0.0832, + "step": 3641 + }, + { + "epoch": 2.9480926844075688, + "grad_norm": 1.9522784948349, + "learning_rate": 1.5167939735157842e-08, + "loss": 0.0707, + "step": 3642 + }, + { + "epoch": 2.9489021552160275, + "grad_norm": 1.9434032440185547, + "learning_rate": 1.4690356435886899e-08, + "loss": 0.0871, + "step": 3643 + }, + { + "epoch": 2.949711626024486, + "grad_norm": 1.8791226148605347, + "learning_rate": 1.4220407042828677e-08, + "loss": 0.0747, + "step": 3644 + }, + { + "epoch": 2.9505210968329454, + "grad_norm": 2.0520567893981934, + "learning_rate": 1.3758091915268002e-08, + "loss": 0.0802, + "step": 3645 + }, + { + "epoch": 2.9513305676414046, + "grad_norm": 2.235368013381958, + "learning_rate": 1.3303411406651034e-08, + "loss": 0.0894, + "step": 3646 + }, + { + "epoch": 2.9521400384498633, + "grad_norm": 1.9867382049560547, + "learning_rate": 1.2856365864586385e-08, + "loss": 0.08, + "step": 3647 + }, + { + "epoch": 2.9529495092583224, + "grad_norm": 1.9352195262908936, + "learning_rate": 1.2416955630847327e-08, + "loss": 0.0855, + "step": 3648 + }, + { + "epoch": 2.953758980066781, + "grad_norm": 1.9544976949691772, + "learning_rate": 1.1985181041368476e-08, + "loss": 0.0791, + "step": 3649 + }, + { + "epoch": 2.9545684508752403, + "grad_norm": 1.8957321643829346, + "learning_rate": 1.1561042426248003e-08, + "loss": 0.0753, + "step": 3650 + }, + { + "epoch": 2.9553779216836995, + "grad_norm": 1.922845482826233, + "learning_rate": 1.1144540109744306e-08, + "loss": 0.0867, + "step": 3651 + }, + { + "epoch": 2.9561873924921582, + "grad_norm": 1.9976263046264648, + "learning_rate": 1.073567441028156e-08, + "loss": 0.09, + "step": 3652 + }, + { + "epoch": 2.956996863300617, + "grad_norm": 1.8465843200683594, + "learning_rate": 1.0334445640440839e-08, + "loss": 0.079, + "step": 3653 + }, + { + "epoch": 2.957806334109076, + "grad_norm": 1.8183223009109497, + "learning_rate": 9.94085410696899e-09, + "loss": 0.0734, + "step": 3654 + }, + { + "epoch": 2.9586158049175353, + "grad_norm": 2.0470528602600098, + "learning_rate": 9.55490011076976e-09, + "loss": 0.0746, + "step": 3655 + }, + { + "epoch": 2.959425275725994, + "grad_norm": 1.8334193229675293, + "learning_rate": 9.176583946912676e-09, + "loss": 0.0756, + "step": 3656 + }, + { + "epoch": 2.960234746534453, + "grad_norm": 1.925060749053955, + "learning_rate": 8.805905904624157e-09, + "loss": 0.0728, + "step": 3657 + }, + { + "epoch": 2.961044217342912, + "grad_norm": 2.007845401763916, + "learning_rate": 8.44286626729307e-09, + "loss": 0.081, + "step": 3658 + }, + { + "epoch": 2.961853688151371, + "grad_norm": 1.9787217378616333, + "learning_rate": 8.087465312468513e-09, + "loss": 0.085, + "step": 3659 + }, + { + "epoch": 2.9626631589598302, + "grad_norm": 1.820393681526184, + "learning_rate": 7.739703311859803e-09, + "loss": 0.0752, + "step": 3660 + }, + { + "epoch": 2.963472629768289, + "grad_norm": 1.9719642400741577, + "learning_rate": 7.399580531334272e-09, + "loss": 0.0803, + "step": 3661 + }, + { + "epoch": 2.9642821005767477, + "grad_norm": 1.8526790142059326, + "learning_rate": 7.067097230920583e-09, + "loss": 0.0756, + "step": 3662 + }, + { + "epoch": 2.965091571385207, + "grad_norm": 1.8761849403381348, + "learning_rate": 6.74225366480874e-09, + "loss": 0.0832, + "step": 3663 + }, + { + "epoch": 2.965901042193666, + "grad_norm": 1.970446228981018, + "learning_rate": 6.42505008134453e-09, + "loss": 0.0791, + "step": 3664 + }, + { + "epoch": 2.9667105130021247, + "grad_norm": 1.8309025764465332, + "learning_rate": 6.11548672303397e-09, + "loss": 0.0769, + "step": 3665 + }, + { + "epoch": 2.967519983810584, + "grad_norm": 1.7394295930862427, + "learning_rate": 5.81356382654441e-09, + "loss": 0.0753, + "step": 3666 + }, + { + "epoch": 2.9683294546190426, + "grad_norm": 1.9777212142944336, + "learning_rate": 5.519281622697881e-09, + "loss": 0.0746, + "step": 3667 + }, + { + "epoch": 2.969138925427502, + "grad_norm": 2.1101503372192383, + "learning_rate": 5.232640336478856e-09, + "loss": 0.0797, + "step": 3668 + }, + { + "epoch": 2.969948396235961, + "grad_norm": 1.8973703384399414, + "learning_rate": 4.953640187027597e-09, + "loss": 0.0726, + "step": 3669 + }, + { + "epoch": 2.9707578670444197, + "grad_norm": 1.7964732646942139, + "learning_rate": 4.682281387643484e-09, + "loss": 0.0834, + "step": 3670 + }, + { + "epoch": 2.9715673378528784, + "grad_norm": 2.0136489868164062, + "learning_rate": 4.418564145785009e-09, + "loss": 0.0734, + "step": 3671 + }, + { + "epoch": 2.9723768086613376, + "grad_norm": 1.7337669134140015, + "learning_rate": 4.162488663066455e-09, + "loss": 0.0701, + "step": 3672 + }, + { + "epoch": 2.9731862794697967, + "grad_norm": 1.959366798400879, + "learning_rate": 3.914055135261219e-09, + "loss": 0.0799, + "step": 3673 + }, + { + "epoch": 2.9739957502782555, + "grad_norm": 1.7301225662231445, + "learning_rate": 3.6732637523018143e-09, + "loss": 0.0722, + "step": 3674 + }, + { + "epoch": 2.9748052210867146, + "grad_norm": 2.0277650356292725, + "learning_rate": 3.4401146982743216e-09, + "loss": 0.0794, + "step": 3675 + }, + { + "epoch": 2.9756146918951734, + "grad_norm": 1.8242539167404175, + "learning_rate": 3.214608151425047e-09, + "loss": 0.0698, + "step": 3676 + }, + { + "epoch": 2.9764241627036325, + "grad_norm": 1.970010757446289, + "learning_rate": 2.9967442841571936e-09, + "loss": 0.0825, + "step": 3677 + }, + { + "epoch": 2.9772336335120917, + "grad_norm": 1.7936547994613647, + "learning_rate": 2.78652326303086e-09, + "loss": 0.0732, + "step": 3678 + }, + { + "epoch": 2.9780431043205504, + "grad_norm": 1.9521154165267944, + "learning_rate": 2.5839452487630425e-09, + "loss": 0.0759, + "step": 3679 + }, + { + "epoch": 2.978852575129009, + "grad_norm": 2.0076730251312256, + "learning_rate": 2.389010396226521e-09, + "loss": 0.081, + "step": 3680 + }, + { + "epoch": 2.9796620459374683, + "grad_norm": 1.903939127922058, + "learning_rate": 2.201718854452084e-09, + "loss": 0.0742, + "step": 3681 + }, + { + "epoch": 2.9804715167459275, + "grad_norm": 1.9086863994598389, + "learning_rate": 2.022070766626305e-09, + "loss": 0.0717, + "step": 3682 + }, + { + "epoch": 2.981280987554386, + "grad_norm": 1.9153650999069214, + "learning_rate": 1.8500662700937644e-09, + "loss": 0.0793, + "step": 3683 + }, + { + "epoch": 2.9820904583628454, + "grad_norm": 1.9087774753570557, + "learning_rate": 1.6857054963537177e-09, + "loss": 0.0818, + "step": 3684 + }, + { + "epoch": 2.982899929171304, + "grad_norm": 1.988550066947937, + "learning_rate": 1.5289885710612073e-09, + "loss": 0.0838, + "step": 3685 + }, + { + "epoch": 2.9837093999797633, + "grad_norm": 1.9354054927825928, + "learning_rate": 1.3799156140292814e-09, + "loss": 0.0805, + "step": 3686 + }, + { + "epoch": 2.9845188707882224, + "grad_norm": 2.0531704425811768, + "learning_rate": 1.2384867392267742e-09, + "loss": 0.0915, + "step": 3687 + }, + { + "epoch": 2.985328341596681, + "grad_norm": 1.84806227684021, + "learning_rate": 1.1047020547771959e-09, + "loss": 0.0726, + "step": 3688 + }, + { + "epoch": 2.98613781240514, + "grad_norm": 1.9062418937683105, + "learning_rate": 9.785616629620632e-10, + "loss": 0.0774, + "step": 3689 + }, + { + "epoch": 2.986947283213599, + "grad_norm": 2.049729585647583, + "learning_rate": 8.600656602153479e-10, + "loss": 0.0829, + "step": 3690 + }, + { + "epoch": 2.987756754022058, + "grad_norm": 2.0688631534576416, + "learning_rate": 7.492141371301387e-10, + "loss": 0.086, + "step": 3691 + }, + { + "epoch": 2.988566224830517, + "grad_norm": 1.8065394163131714, + "learning_rate": 6.460071784541999e-10, + "loss": 0.0757, + "step": 3692 + }, + { + "epoch": 2.989375695638976, + "grad_norm": 1.9245977401733398, + "learning_rate": 5.504448630899717e-10, + "loss": 0.0794, + "step": 3693 + }, + { + "epoch": 2.990185166447435, + "grad_norm": 1.868876338005066, + "learning_rate": 4.625272640967904e-10, + "loss": 0.0779, + "step": 3694 + }, + { + "epoch": 2.990994637255894, + "grad_norm": 1.8760489225387573, + "learning_rate": 3.822544486886681e-10, + "loss": 0.0791, + "step": 3695 + }, + { + "epoch": 2.991804108064353, + "grad_norm": 2.0247747898101807, + "learning_rate": 3.096264782342928e-10, + "loss": 0.0834, + "step": 3696 + }, + { + "epoch": 2.992613578872812, + "grad_norm": 1.886752724647522, + "learning_rate": 2.446434082603588e-10, + "loss": 0.0834, + "step": 3697 + }, + { + "epoch": 2.9934230496812706, + "grad_norm": 1.887201189994812, + "learning_rate": 1.873052884460158e-10, + "loss": 0.074, + "step": 3698 + }, + { + "epoch": 2.9942325204897298, + "grad_norm": 1.9398181438446045, + "learning_rate": 1.3761216262841991e-10, + "loss": 0.0753, + "step": 3699 + }, + { + "epoch": 2.995041991298189, + "grad_norm": 1.7916117906570435, + "learning_rate": 9.556406879718261e-11, + "loss": 0.0831, + "step": 3700 + }, + { + "epoch": 2.9958514621066477, + "grad_norm": 1.9150363206863403, + "learning_rate": 6.116103909881154e-11, + "loss": 0.0768, + "step": 3701 + }, + { + "epoch": 2.996660932915107, + "grad_norm": 2.0668771266937256, + "learning_rate": 3.440309983560042e-11, + "loss": 0.0816, + "step": 3702 + }, + { + "epoch": 2.9974704037235655, + "grad_norm": 1.9825881719589233, + "learning_rate": 1.529027146451867e-11, + "loss": 0.0787, + "step": 3703 + }, + { + "epoch": 2.9982798745320247, + "grad_norm": 2.0460188388824463, + "learning_rate": 3.822568597211529e-12, + "loss": 0.0831, + "step": 3704 + }, + { + "epoch": 2.999089345340484, + "grad_norm": 1.965240240097046, + "learning_rate": 0.0, + "loss": 0.0806, + "step": 3705 + }, + { + "epoch": 2.999089345340484, + "step": 3705, + "total_flos": 4.6047708864682394e+18, + "train_loss": 0.2889727547463135, + "train_runtime": 58434.9858, + "train_samples_per_second": 8.117, + "train_steps_per_second": 0.063 + } + ], + "logging_steps": 1.0, + "max_steps": 3705, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 4.6047708864682394e+18, + "train_batch_size": 4, + "trial_name": null, + "trial_params": null +}