diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,78330 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 500, + "global_step": 11184, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 8.94134477825465e-05, + "grad_norm": 0.6782906980599177, + "learning_rate": 5.952380952380952e-07, + "loss": 0.9843, + "step": 1 + }, + { + "epoch": 0.000178826895565093, + "grad_norm": 0.6463239114665619, + "learning_rate": 1.1904761904761904e-06, + "loss": 1.0091, + "step": 2 + }, + { + "epoch": 0.0002682403433476395, + "grad_norm": 0.7196377761245901, + "learning_rate": 1.7857142857142857e-06, + "loss": 1.021, + "step": 3 + }, + { + "epoch": 0.000357653791130186, + "grad_norm": 0.7285855823472192, + "learning_rate": 2.3809523809523808e-06, + "loss": 1.0525, + "step": 4 + }, + { + "epoch": 0.0004470672389127325, + "grad_norm": 0.7033175660597506, + "learning_rate": 2.9761904761904763e-06, + "loss": 0.9965, + "step": 5 + }, + { + "epoch": 0.000536480686695279, + "grad_norm": 0.6935004789442767, + "learning_rate": 3.5714285714285714e-06, + "loss": 0.9855, + "step": 6 + }, + { + "epoch": 0.0006258941344778255, + "grad_norm": 0.6352226543116061, + "learning_rate": 4.166666666666667e-06, + "loss": 1.0116, + "step": 7 + }, + { + "epoch": 0.000715307582260372, + "grad_norm": 0.5550226629353024, + "learning_rate": 4.7619047619047615e-06, + "loss": 0.9649, + "step": 8 + }, + { + "epoch": 0.0008047210300429185, + "grad_norm": 0.5187529868838625, + "learning_rate": 5.357142857142857e-06, + "loss": 0.985, + "step": 9 + }, + { + "epoch": 0.000894134477825465, + "grad_norm": 0.42828966697266957, + "learning_rate": 5.9523809523809525e-06, + "loss": 0.952, + "step": 10 + }, + { + "epoch": 0.0009835479256080114, + "grad_norm": 0.36969982094372605, + "learning_rate": 6.547619047619048e-06, + "loss": 0.9836, + "step": 11 + }, + { + "epoch": 0.001072961373390558, + "grad_norm": 0.30758798166410417, + "learning_rate": 7.142857142857143e-06, + "loss": 1.0235, + "step": 12 + }, + { + "epoch": 0.0011623748211731044, + "grad_norm": 0.22980369019690852, + "learning_rate": 7.738095238095238e-06, + "loss": 0.9155, + "step": 13 + }, + { + "epoch": 0.001251788268955651, + "grad_norm": 0.23409774405142605, + "learning_rate": 8.333333333333334e-06, + "loss": 0.9218, + "step": 14 + }, + { + "epoch": 0.0013412017167381974, + "grad_norm": 0.22422941915520278, + "learning_rate": 8.92857142857143e-06, + "loss": 0.9542, + "step": 15 + }, + { + "epoch": 0.001430615164520744, + "grad_norm": 0.21850962147437303, + "learning_rate": 9.523809523809523e-06, + "loss": 0.9494, + "step": 16 + }, + { + "epoch": 0.0015200286123032904, + "grad_norm": 0.22092020274672947, + "learning_rate": 1.011904761904762e-05, + "loss": 0.9741, + "step": 17 + }, + { + "epoch": 0.001609442060085837, + "grad_norm": 0.19239835694212362, + "learning_rate": 1.0714285714285714e-05, + "loss": 0.9049, + "step": 18 + }, + { + "epoch": 0.0016988555078683834, + "grad_norm": 0.21221529095156458, + "learning_rate": 1.130952380952381e-05, + "loss": 0.9813, + "step": 19 + }, + { + "epoch": 0.00178826895565093, + "grad_norm": 0.17655083244828304, + "learning_rate": 1.1904761904761905e-05, + "loss": 0.8747, + "step": 20 + }, + { + "epoch": 0.0018776824034334764, + "grad_norm": 0.17765077552666675, + "learning_rate": 1.25e-05, + "loss": 0.8891, + "step": 21 + }, + { + "epoch": 0.001967095851216023, + "grad_norm": 0.1667861096735661, + "learning_rate": 1.3095238095238096e-05, + "loss": 0.9316, + "step": 22 + }, + { + "epoch": 0.0020565092989985696, + "grad_norm": 0.14314335840621184, + "learning_rate": 1.3690476190476192e-05, + "loss": 0.8547, + "step": 23 + }, + { + "epoch": 0.002145922746781116, + "grad_norm": 0.14334678190346298, + "learning_rate": 1.4285714285714285e-05, + "loss": 0.9076, + "step": 24 + }, + { + "epoch": 0.0022353361945636626, + "grad_norm": 0.14665277185687053, + "learning_rate": 1.4880952380952381e-05, + "loss": 0.8927, + "step": 25 + }, + { + "epoch": 0.002324749642346209, + "grad_norm": 0.14104016615662898, + "learning_rate": 1.5476190476190476e-05, + "loss": 0.9121, + "step": 26 + }, + { + "epoch": 0.0024141630901287556, + "grad_norm": 0.1531863411096855, + "learning_rate": 1.6071428571428572e-05, + "loss": 0.9127, + "step": 27 + }, + { + "epoch": 0.002503576537911302, + "grad_norm": 0.15940823251609024, + "learning_rate": 1.6666666666666667e-05, + "loss": 0.9008, + "step": 28 + }, + { + "epoch": 0.0025929899856938486, + "grad_norm": 0.16159843376010974, + "learning_rate": 1.7261904761904763e-05, + "loss": 0.8836, + "step": 29 + }, + { + "epoch": 0.002682403433476395, + "grad_norm": 0.144925919172616, + "learning_rate": 1.785714285714286e-05, + "loss": 0.8941, + "step": 30 + }, + { + "epoch": 0.0027718168812589416, + "grad_norm": 0.1408116185352553, + "learning_rate": 1.8452380952380954e-05, + "loss": 0.8638, + "step": 31 + }, + { + "epoch": 0.002861230329041488, + "grad_norm": 0.17810338323051883, + "learning_rate": 1.9047619047619046e-05, + "loss": 0.8597, + "step": 32 + }, + { + "epoch": 0.0029506437768240345, + "grad_norm": 0.15283232037337707, + "learning_rate": 1.9642857142857145e-05, + "loss": 0.9124, + "step": 33 + }, + { + "epoch": 0.003040057224606581, + "grad_norm": 0.1595865304312526, + "learning_rate": 2.023809523809524e-05, + "loss": 0.8898, + "step": 34 + }, + { + "epoch": 0.0031294706723891275, + "grad_norm": 0.16063659088090979, + "learning_rate": 2.0833333333333336e-05, + "loss": 0.8774, + "step": 35 + }, + { + "epoch": 0.003218884120171674, + "grad_norm": 0.15169138779608896, + "learning_rate": 2.1428571428571428e-05, + "loss": 0.8775, + "step": 36 + }, + { + "epoch": 0.0033082975679542205, + "grad_norm": 0.13832242547614126, + "learning_rate": 2.2023809523809524e-05, + "loss": 0.8542, + "step": 37 + }, + { + "epoch": 0.003397711015736767, + "grad_norm": 0.13608478752072947, + "learning_rate": 2.261904761904762e-05, + "loss": 0.8601, + "step": 38 + }, + { + "epoch": 0.0034871244635193135, + "grad_norm": 0.13433063832630773, + "learning_rate": 2.3214285714285715e-05, + "loss": 0.8507, + "step": 39 + }, + { + "epoch": 0.00357653791130186, + "grad_norm": 0.10158511231272437, + "learning_rate": 2.380952380952381e-05, + "loss": 0.8233, + "step": 40 + }, + { + "epoch": 0.0036659513590844065, + "grad_norm": 0.15469834072832142, + "learning_rate": 2.4404761904761906e-05, + "loss": 0.8835, + "step": 41 + }, + { + "epoch": 0.0037553648068669528, + "grad_norm": 0.14292818924224057, + "learning_rate": 2.5e-05, + "loss": 0.8605, + "step": 42 + }, + { + "epoch": 0.0038447782546494995, + "grad_norm": 0.12086094248058331, + "learning_rate": 2.5595238095238093e-05, + "loss": 0.8293, + "step": 43 + }, + { + "epoch": 0.003934191702432046, + "grad_norm": 0.12744027873956576, + "learning_rate": 2.6190476190476192e-05, + "loss": 0.8603, + "step": 44 + }, + { + "epoch": 0.004023605150214592, + "grad_norm": 0.15030470289361023, + "learning_rate": 2.6785714285714288e-05, + "loss": 0.9014, + "step": 45 + }, + { + "epoch": 0.004113018597997139, + "grad_norm": 0.11429133064037213, + "learning_rate": 2.7380952380952383e-05, + "loss": 0.8185, + "step": 46 + }, + { + "epoch": 0.0042024320457796855, + "grad_norm": 0.12122139920666253, + "learning_rate": 2.797619047619048e-05, + "loss": 0.8326, + "step": 47 + }, + { + "epoch": 0.004291845493562232, + "grad_norm": 0.12372423531608305, + "learning_rate": 2.857142857142857e-05, + "loss": 0.8393, + "step": 48 + }, + { + "epoch": 0.004381258941344778, + "grad_norm": 0.13372615949564698, + "learning_rate": 2.916666666666667e-05, + "loss": 0.832, + "step": 49 + }, + { + "epoch": 0.004470672389127325, + "grad_norm": 0.14898367278203478, + "learning_rate": 2.9761904761904762e-05, + "loss": 0.8873, + "step": 50 + }, + { + "epoch": 0.0045600858369098714, + "grad_norm": 0.1382355773160592, + "learning_rate": 3.0357142857142857e-05, + "loss": 0.8489, + "step": 51 + }, + { + "epoch": 0.004649499284692418, + "grad_norm": 0.11460457328207474, + "learning_rate": 3.095238095238095e-05, + "loss": 0.8117, + "step": 52 + }, + { + "epoch": 0.004738912732474964, + "grad_norm": 0.12793886342260358, + "learning_rate": 3.154761904761905e-05, + "loss": 0.8225, + "step": 53 + }, + { + "epoch": 0.004828326180257511, + "grad_norm": 0.1307915190716159, + "learning_rate": 3.2142857142857144e-05, + "loss": 0.8606, + "step": 54 + }, + { + "epoch": 0.004917739628040057, + "grad_norm": 0.15142165374702826, + "learning_rate": 3.273809523809524e-05, + "loss": 0.8764, + "step": 55 + }, + { + "epoch": 0.005007153075822604, + "grad_norm": 0.14766208631916114, + "learning_rate": 3.3333333333333335e-05, + "loss": 0.8229, + "step": 56 + }, + { + "epoch": 0.00509656652360515, + "grad_norm": 0.15811938250608404, + "learning_rate": 3.392857142857143e-05, + "loss": 0.8306, + "step": 57 + }, + { + "epoch": 0.005185979971387697, + "grad_norm": 0.1821578067412281, + "learning_rate": 3.4523809523809526e-05, + "loss": 0.8842, + "step": 58 + }, + { + "epoch": 0.005275393419170243, + "grad_norm": 0.17198549093074655, + "learning_rate": 3.511904761904762e-05, + "loss": 0.834, + "step": 59 + }, + { + "epoch": 0.00536480686695279, + "grad_norm": 0.1749604557804795, + "learning_rate": 3.571428571428572e-05, + "loss": 0.8446, + "step": 60 + }, + { + "epoch": 0.005454220314735336, + "grad_norm": 0.22096368609110362, + "learning_rate": 3.630952380952381e-05, + "loss": 0.8236, + "step": 61 + }, + { + "epoch": 0.005543633762517883, + "grad_norm": 0.20423265725100614, + "learning_rate": 3.690476190476191e-05, + "loss": 0.843, + "step": 62 + }, + { + "epoch": 0.005633047210300429, + "grad_norm": 0.18347091263976512, + "learning_rate": 3.7500000000000003e-05, + "loss": 0.8163, + "step": 63 + }, + { + "epoch": 0.005722460658082976, + "grad_norm": 0.1614731762998425, + "learning_rate": 3.809523809523809e-05, + "loss": 0.2105, + "step": 64 + }, + { + "epoch": 0.005811874105865522, + "grad_norm": 0.2239348250005407, + "learning_rate": 3.8690476190476195e-05, + "loss": 0.8598, + "step": 65 + }, + { + "epoch": 0.005901287553648069, + "grad_norm": 0.16660836200268747, + "learning_rate": 3.928571428571429e-05, + "loss": 0.8309, + "step": 66 + }, + { + "epoch": 0.005990701001430615, + "grad_norm": 0.2578497220857537, + "learning_rate": 3.9880952380952386e-05, + "loss": 0.784, + "step": 67 + }, + { + "epoch": 0.006080114449213162, + "grad_norm": 0.16760399075951618, + "learning_rate": 4.047619047619048e-05, + "loss": 0.8088, + "step": 68 + }, + { + "epoch": 0.006169527896995708, + "grad_norm": 0.19967440303350734, + "learning_rate": 4.107142857142857e-05, + "loss": 0.8286, + "step": 69 + }, + { + "epoch": 0.006258941344778255, + "grad_norm": 0.16853987793242056, + "learning_rate": 4.166666666666667e-05, + "loss": 0.8092, + "step": 70 + }, + { + "epoch": 0.006348354792560801, + "grad_norm": 0.14288940066527756, + "learning_rate": 4.226190476190476e-05, + "loss": 0.7707, + "step": 71 + }, + { + "epoch": 0.006437768240343348, + "grad_norm": 0.17710172608924607, + "learning_rate": 4.2857142857142856e-05, + "loss": 0.7997, + "step": 72 + }, + { + "epoch": 0.006527181688125894, + "grad_norm": 0.15701610609153246, + "learning_rate": 4.345238095238096e-05, + "loss": 0.8164, + "step": 73 + }, + { + "epoch": 0.006616595135908441, + "grad_norm": 0.22346692040588528, + "learning_rate": 4.404761904761905e-05, + "loss": 0.7992, + "step": 74 + }, + { + "epoch": 0.006706008583690987, + "grad_norm": 0.1408060414158345, + "learning_rate": 4.464285714285715e-05, + "loss": 0.7908, + "step": 75 + }, + { + "epoch": 0.006795422031473534, + "grad_norm": 0.18190965085417046, + "learning_rate": 4.523809523809524e-05, + "loss": 0.7756, + "step": 76 + }, + { + "epoch": 0.00688483547925608, + "grad_norm": 0.17060149952371872, + "learning_rate": 4.5833333333333334e-05, + "loss": 0.7698, + "step": 77 + }, + { + "epoch": 0.006974248927038627, + "grad_norm": 0.15285247211057645, + "learning_rate": 4.642857142857143e-05, + "loss": 0.7619, + "step": 78 + }, + { + "epoch": 0.007063662374821173, + "grad_norm": 0.2015819620775906, + "learning_rate": 4.7023809523809525e-05, + "loss": 0.8179, + "step": 79 + }, + { + "epoch": 0.00715307582260372, + "grad_norm": 0.1559125618082628, + "learning_rate": 4.761904761904762e-05, + "loss": 0.8443, + "step": 80 + }, + { + "epoch": 0.007242489270386266, + "grad_norm": 0.15780023067115476, + "learning_rate": 4.8214285714285716e-05, + "loss": 0.7832, + "step": 81 + }, + { + "epoch": 0.007331902718168813, + "grad_norm": 0.18344312638285085, + "learning_rate": 4.880952380952381e-05, + "loss": 0.7783, + "step": 82 + }, + { + "epoch": 0.007421316165951359, + "grad_norm": 0.18591170879016408, + "learning_rate": 4.940476190476191e-05, + "loss": 0.7981, + "step": 83 + }, + { + "epoch": 0.0075107296137339056, + "grad_norm": 0.20308250904644207, + "learning_rate": 5e-05, + "loss": 0.7804, + "step": 84 + }, + { + "epoch": 0.007600143061516452, + "grad_norm": 0.17643158455099006, + "learning_rate": 5.05952380952381e-05, + "loss": 0.7662, + "step": 85 + }, + { + "epoch": 0.007689556509298999, + "grad_norm": 0.2050763768096218, + "learning_rate": 5.119047619047619e-05, + "loss": 0.7834, + "step": 86 + }, + { + "epoch": 0.007778969957081545, + "grad_norm": 0.08612407074473787, + "learning_rate": 5.1785714285714296e-05, + "loss": 0.1996, + "step": 87 + }, + { + "epoch": 0.007868383404864092, + "grad_norm": 0.24271061422034018, + "learning_rate": 5.2380952380952384e-05, + "loss": 0.7915, + "step": 88 + }, + { + "epoch": 0.007957796852646639, + "grad_norm": 0.16660868606799756, + "learning_rate": 5.297619047619048e-05, + "loss": 0.7647, + "step": 89 + }, + { + "epoch": 0.008047210300429184, + "grad_norm": 0.5794311605852219, + "learning_rate": 5.3571428571428575e-05, + "loss": 0.802, + "step": 90 + }, + { + "epoch": 0.008136623748211731, + "grad_norm": 0.2174655718939883, + "learning_rate": 5.4166666666666664e-05, + "loss": 0.8257, + "step": 91 + }, + { + "epoch": 0.008226037195994278, + "grad_norm": 0.31981825444151923, + "learning_rate": 5.4761904761904766e-05, + "loss": 0.7877, + "step": 92 + }, + { + "epoch": 0.008315450643776824, + "grad_norm": 0.20599788932251273, + "learning_rate": 5.535714285714286e-05, + "loss": 0.7832, + "step": 93 + }, + { + "epoch": 0.008404864091559371, + "grad_norm": 0.18409811646814828, + "learning_rate": 5.595238095238096e-05, + "loss": 0.7525, + "step": 94 + }, + { + "epoch": 0.008494277539341916, + "grad_norm": 0.14478272074108095, + "learning_rate": 5.6547619047619046e-05, + "loss": 0.8117, + "step": 95 + }, + { + "epoch": 0.008583690987124463, + "grad_norm": 0.17591872500500405, + "learning_rate": 5.714285714285714e-05, + "loss": 0.7849, + "step": 96 + }, + { + "epoch": 0.00867310443490701, + "grad_norm": 0.15770457206422925, + "learning_rate": 5.773809523809524e-05, + "loss": 0.7968, + "step": 97 + }, + { + "epoch": 0.008762517882689556, + "grad_norm": 0.1500820651815235, + "learning_rate": 5.833333333333334e-05, + "loss": 0.7779, + "step": 98 + }, + { + "epoch": 0.008851931330472103, + "grad_norm": 0.16943209919149313, + "learning_rate": 5.8928571428571435e-05, + "loss": 0.7563, + "step": 99 + }, + { + "epoch": 0.00894134477825465, + "grad_norm": 0.16200930984716413, + "learning_rate": 5.9523809523809524e-05, + "loss": 0.8021, + "step": 100 + }, + { + "epoch": 0.009030758226037196, + "grad_norm": 0.14324220428941942, + "learning_rate": 6.011904761904762e-05, + "loss": 0.7547, + "step": 101 + }, + { + "epoch": 0.009120171673819743, + "grad_norm": 0.1329139886721918, + "learning_rate": 6.0714285714285715e-05, + "loss": 0.7447, + "step": 102 + }, + { + "epoch": 0.009209585121602288, + "grad_norm": 0.13207412002243896, + "learning_rate": 6.130952380952381e-05, + "loss": 0.7965, + "step": 103 + }, + { + "epoch": 0.009298998569384835, + "grad_norm": 0.1580074421475962, + "learning_rate": 6.19047619047619e-05, + "loss": 0.782, + "step": 104 + }, + { + "epoch": 0.009388412017167383, + "grad_norm": 0.17831866936678828, + "learning_rate": 6.25e-05, + "loss": 0.7785, + "step": 105 + }, + { + "epoch": 0.009477825464949928, + "grad_norm": 0.14870167761991032, + "learning_rate": 6.30952380952381e-05, + "loss": 0.81, + "step": 106 + }, + { + "epoch": 0.009567238912732475, + "grad_norm": 0.15772271508483385, + "learning_rate": 6.369047619047619e-05, + "loss": 0.7596, + "step": 107 + }, + { + "epoch": 0.009656652360515022, + "grad_norm": 0.15278304631670642, + "learning_rate": 6.428571428571429e-05, + "loss": 0.778, + "step": 108 + }, + { + "epoch": 0.009746065808297568, + "grad_norm": 0.1629976020775258, + "learning_rate": 6.488095238095238e-05, + "loss": 0.7533, + "step": 109 + }, + { + "epoch": 0.009835479256080115, + "grad_norm": 0.13138069494542626, + "learning_rate": 6.547619047619048e-05, + "loss": 0.7651, + "step": 110 + }, + { + "epoch": 0.00992489270386266, + "grad_norm": 0.17062723128400617, + "learning_rate": 6.607142857142857e-05, + "loss": 0.7618, + "step": 111 + }, + { + "epoch": 0.010014306151645207, + "grad_norm": 0.1392407425779791, + "learning_rate": 6.666666666666667e-05, + "loss": 0.755, + "step": 112 + }, + { + "epoch": 0.010103719599427755, + "grad_norm": 0.14799006848786708, + "learning_rate": 6.726190476190477e-05, + "loss": 0.7691, + "step": 113 + }, + { + "epoch": 0.0101931330472103, + "grad_norm": 0.1409804595259366, + "learning_rate": 6.785714285714286e-05, + "loss": 0.8222, + "step": 114 + }, + { + "epoch": 0.010282546494992847, + "grad_norm": 0.13239977797742028, + "learning_rate": 6.845238095238096e-05, + "loss": 0.7311, + "step": 115 + }, + { + "epoch": 0.010371959942775394, + "grad_norm": 0.13854671761747736, + "learning_rate": 6.904761904761905e-05, + "loss": 0.7462, + "step": 116 + }, + { + "epoch": 0.01046137339055794, + "grad_norm": 0.1595561596971385, + "learning_rate": 6.964285714285715e-05, + "loss": 0.7521, + "step": 117 + }, + { + "epoch": 0.010550786838340487, + "grad_norm": 0.16785752125914366, + "learning_rate": 7.023809523809524e-05, + "loss": 0.7842, + "step": 118 + }, + { + "epoch": 0.010640200286123032, + "grad_norm": 0.164095116827372, + "learning_rate": 7.083333333333334e-05, + "loss": 0.7717, + "step": 119 + }, + { + "epoch": 0.01072961373390558, + "grad_norm": 0.16173180491989872, + "learning_rate": 7.142857142857143e-05, + "loss": 0.7546, + "step": 120 + }, + { + "epoch": 0.010819027181688127, + "grad_norm": 0.13368700969815156, + "learning_rate": 7.202380952380953e-05, + "loss": 0.7622, + "step": 121 + }, + { + "epoch": 0.010908440629470672, + "grad_norm": 0.18923003992775494, + "learning_rate": 7.261904761904762e-05, + "loss": 0.776, + "step": 122 + }, + { + "epoch": 0.010997854077253219, + "grad_norm": 0.12598688807272285, + "learning_rate": 7.321428571428571e-05, + "loss": 0.7558, + "step": 123 + }, + { + "epoch": 0.011087267525035766, + "grad_norm": 0.18304512026984385, + "learning_rate": 7.380952380952382e-05, + "loss": 0.7792, + "step": 124 + }, + { + "epoch": 0.011176680972818312, + "grad_norm": 0.19894669923026373, + "learning_rate": 7.440476190476191e-05, + "loss": 0.7929, + "step": 125 + }, + { + "epoch": 0.011266094420600859, + "grad_norm": 0.15136469255793872, + "learning_rate": 7.500000000000001e-05, + "loss": 0.7891, + "step": 126 + }, + { + "epoch": 0.011355507868383404, + "grad_norm": 0.1358592251476268, + "learning_rate": 7.55952380952381e-05, + "loss": 0.7391, + "step": 127 + }, + { + "epoch": 0.011444921316165951, + "grad_norm": 0.13643981556516377, + "learning_rate": 7.619047619047618e-05, + "loss": 0.7649, + "step": 128 + }, + { + "epoch": 0.011534334763948498, + "grad_norm": 0.19244183396255207, + "learning_rate": 7.67857142857143e-05, + "loss": 0.7573, + "step": 129 + }, + { + "epoch": 0.011623748211731044, + "grad_norm": 0.12457261447357702, + "learning_rate": 7.738095238095239e-05, + "loss": 0.728, + "step": 130 + }, + { + "epoch": 0.011713161659513591, + "grad_norm": 0.16366337503962425, + "learning_rate": 7.797619047619048e-05, + "loss": 0.7492, + "step": 131 + }, + { + "epoch": 0.011802575107296138, + "grad_norm": 0.1317582435853772, + "learning_rate": 7.857142857142858e-05, + "loss": 0.7252, + "step": 132 + }, + { + "epoch": 0.011891988555078684, + "grad_norm": 0.1787639159098099, + "learning_rate": 7.916666666666666e-05, + "loss": 0.7648, + "step": 133 + }, + { + "epoch": 0.01198140200286123, + "grad_norm": 0.13235417193806417, + "learning_rate": 7.976190476190477e-05, + "loss": 0.7248, + "step": 134 + }, + { + "epoch": 0.012070815450643776, + "grad_norm": 0.17246303736696683, + "learning_rate": 8.035714285714287e-05, + "loss": 0.7856, + "step": 135 + }, + { + "epoch": 0.012160228898426323, + "grad_norm": 0.14420132054020124, + "learning_rate": 8.095238095238096e-05, + "loss": 0.792, + "step": 136 + }, + { + "epoch": 0.01224964234620887, + "grad_norm": 0.18821969459311855, + "learning_rate": 8.154761904761904e-05, + "loss": 0.7008, + "step": 137 + }, + { + "epoch": 0.012339055793991416, + "grad_norm": 0.09788754549911195, + "learning_rate": 8.214285714285714e-05, + "loss": 0.2167, + "step": 138 + }, + { + "epoch": 0.012428469241773963, + "grad_norm": 0.28290128616154764, + "learning_rate": 8.273809523809524e-05, + "loss": 0.7728, + "step": 139 + }, + { + "epoch": 0.01251788268955651, + "grad_norm": 0.1387164277418416, + "learning_rate": 8.333333333333334e-05, + "loss": 0.7258, + "step": 140 + }, + { + "epoch": 0.012607296137339056, + "grad_norm": 0.2793562703598964, + "learning_rate": 8.392857142857144e-05, + "loss": 0.7911, + "step": 141 + }, + { + "epoch": 0.012696709585121603, + "grad_norm": 0.12844937956275618, + "learning_rate": 8.452380952380952e-05, + "loss": 0.7493, + "step": 142 + }, + { + "epoch": 0.012786123032904148, + "grad_norm": 0.14199491540183443, + "learning_rate": 8.511904761904762e-05, + "loss": 0.7474, + "step": 143 + }, + { + "epoch": 0.012875536480686695, + "grad_norm": 0.16027944078672884, + "learning_rate": 8.571428571428571e-05, + "loss": 0.792, + "step": 144 + }, + { + "epoch": 0.012964949928469242, + "grad_norm": 0.14981451954669323, + "learning_rate": 8.630952380952382e-05, + "loss": 0.762, + "step": 145 + }, + { + "epoch": 0.013054363376251788, + "grad_norm": 0.13606136078392758, + "learning_rate": 8.690476190476192e-05, + "loss": 0.7963, + "step": 146 + }, + { + "epoch": 0.013143776824034335, + "grad_norm": 0.12975481658295215, + "learning_rate": 8.75e-05, + "loss": 0.7825, + "step": 147 + }, + { + "epoch": 0.013233190271816882, + "grad_norm": 0.12730994894347875, + "learning_rate": 8.80952380952381e-05, + "loss": 0.729, + "step": 148 + }, + { + "epoch": 0.013322603719599427, + "grad_norm": 0.13249914192679824, + "learning_rate": 8.869047619047619e-05, + "loss": 0.7711, + "step": 149 + }, + { + "epoch": 0.013412017167381975, + "grad_norm": 0.1196144528286821, + "learning_rate": 8.92857142857143e-05, + "loss": 0.7352, + "step": 150 + }, + { + "epoch": 0.01350143061516452, + "grad_norm": 0.12797748979357648, + "learning_rate": 8.988095238095238e-05, + "loss": 0.7377, + "step": 151 + }, + { + "epoch": 0.013590844062947067, + "grad_norm": 0.11559339175779536, + "learning_rate": 9.047619047619048e-05, + "loss": 0.779, + "step": 152 + }, + { + "epoch": 0.013680257510729614, + "grad_norm": 0.14171490095896627, + "learning_rate": 9.107142857142857e-05, + "loss": 0.79, + "step": 153 + }, + { + "epoch": 0.01376967095851216, + "grad_norm": 0.108228040905301, + "learning_rate": 9.166666666666667e-05, + "loss": 0.757, + "step": 154 + }, + { + "epoch": 0.013859084406294707, + "grad_norm": 0.10580359201171059, + "learning_rate": 9.226190476190478e-05, + "loss": 0.7387, + "step": 155 + }, + { + "epoch": 0.013948497854077254, + "grad_norm": 0.11703743948204479, + "learning_rate": 9.285714285714286e-05, + "loss": 0.7622, + "step": 156 + }, + { + "epoch": 0.0140379113018598, + "grad_norm": 0.12162379936228199, + "learning_rate": 9.345238095238095e-05, + "loss": 0.7816, + "step": 157 + }, + { + "epoch": 0.014127324749642347, + "grad_norm": 0.12106447652260913, + "learning_rate": 9.404761904761905e-05, + "loss": 0.7581, + "step": 158 + }, + { + "epoch": 0.014216738197424892, + "grad_norm": 0.13274638415659498, + "learning_rate": 9.464285714285715e-05, + "loss": 0.7219, + "step": 159 + }, + { + "epoch": 0.01430615164520744, + "grad_norm": 0.1265435417920056, + "learning_rate": 9.523809523809524e-05, + "loss": 0.7129, + "step": 160 + }, + { + "epoch": 0.014395565092989986, + "grad_norm": 0.14959194065347278, + "learning_rate": 9.583333333333334e-05, + "loss": 0.7352, + "step": 161 + }, + { + "epoch": 0.014484978540772532, + "grad_norm": 0.1257461860656756, + "learning_rate": 9.642857142857143e-05, + "loss": 0.7619, + "step": 162 + }, + { + "epoch": 0.014574391988555079, + "grad_norm": 0.12585955748885735, + "learning_rate": 9.702380952380953e-05, + "loss": 0.7108, + "step": 163 + }, + { + "epoch": 0.014663805436337626, + "grad_norm": 0.1528051391696236, + "learning_rate": 9.761904761904762e-05, + "loss": 0.7537, + "step": 164 + }, + { + "epoch": 0.014753218884120171, + "grad_norm": 0.07751482793898605, + "learning_rate": 9.821428571428572e-05, + "loss": 0.1929, + "step": 165 + }, + { + "epoch": 0.014842632331902719, + "grad_norm": 0.13660483235498885, + "learning_rate": 9.880952380952381e-05, + "loss": 0.7596, + "step": 166 + }, + { + "epoch": 0.014932045779685264, + "grad_norm": 0.13488327721581267, + "learning_rate": 9.940476190476191e-05, + "loss": 0.7042, + "step": 167 + }, + { + "epoch": 0.015021459227467811, + "grad_norm": 0.14364703325762035, + "learning_rate": 0.0001, + "loss": 0.7704, + "step": 168 + }, + { + "epoch": 0.015110872675250358, + "grad_norm": 0.13907157021497774, + "learning_rate": 0.0001005952380952381, + "loss": 0.753, + "step": 169 + }, + { + "epoch": 0.015200286123032904, + "grad_norm": 0.14096093367432982, + "learning_rate": 0.0001011904761904762, + "loss": 0.7724, + "step": 170 + }, + { + "epoch": 0.01528969957081545, + "grad_norm": 0.12413871836552795, + "learning_rate": 0.00010178571428571428, + "loss": 0.7849, + "step": 171 + }, + { + "epoch": 0.015379113018597998, + "grad_norm": 0.13542287002248152, + "learning_rate": 0.00010238095238095237, + "loss": 0.7391, + "step": 172 + }, + { + "epoch": 0.015468526466380543, + "grad_norm": 0.15226449290819236, + "learning_rate": 0.00010297619047619047, + "loss": 0.7481, + "step": 173 + }, + { + "epoch": 0.01555793991416309, + "grad_norm": 0.13607857651663038, + "learning_rate": 0.00010357142857142859, + "loss": 0.7739, + "step": 174 + }, + { + "epoch": 0.015647353361945636, + "grad_norm": 0.1656460834274211, + "learning_rate": 0.00010416666666666667, + "loss": 0.7419, + "step": 175 + }, + { + "epoch": 0.015736766809728183, + "grad_norm": 0.1282486251045182, + "learning_rate": 0.00010476190476190477, + "loss": 0.7793, + "step": 176 + }, + { + "epoch": 0.01582618025751073, + "grad_norm": 0.14054396183259948, + "learning_rate": 0.00010535714285714286, + "loss": 0.7238, + "step": 177 + }, + { + "epoch": 0.015915593705293277, + "grad_norm": 0.1237026281020843, + "learning_rate": 0.00010595238095238096, + "loss": 0.7556, + "step": 178 + }, + { + "epoch": 0.01600500715307582, + "grad_norm": 0.16853796514428807, + "learning_rate": 0.00010654761904761906, + "loss": 0.8139, + "step": 179 + }, + { + "epoch": 0.016094420600858368, + "grad_norm": 0.12129016908382627, + "learning_rate": 0.00010714285714285715, + "loss": 0.7606, + "step": 180 + }, + { + "epoch": 0.016183834048640915, + "grad_norm": 0.1356355663983982, + "learning_rate": 0.00010773809523809523, + "loss": 0.705, + "step": 181 + }, + { + "epoch": 0.016273247496423462, + "grad_norm": 0.13265410789104798, + "learning_rate": 0.00010833333333333333, + "loss": 0.7209, + "step": 182 + }, + { + "epoch": 0.01636266094420601, + "grad_norm": 0.14040646972299922, + "learning_rate": 0.00010892857142857142, + "loss": 0.7454, + "step": 183 + }, + { + "epoch": 0.016452074391988557, + "grad_norm": 0.13556674145834827, + "learning_rate": 0.00010952380952380953, + "loss": 0.7421, + "step": 184 + }, + { + "epoch": 0.0165414878397711, + "grad_norm": 0.15195953792652905, + "learning_rate": 0.00011011904761904763, + "loss": 0.7824, + "step": 185 + }, + { + "epoch": 0.016630901287553648, + "grad_norm": 0.12634775757748393, + "learning_rate": 0.00011071428571428572, + "loss": 0.7342, + "step": 186 + }, + { + "epoch": 0.016720314735336195, + "grad_norm": 0.13565572474071894, + "learning_rate": 0.00011130952380952382, + "loss": 0.7537, + "step": 187 + }, + { + "epoch": 0.016809728183118742, + "grad_norm": 0.13072214918295508, + "learning_rate": 0.00011190476190476191, + "loss": 0.7643, + "step": 188 + }, + { + "epoch": 0.01689914163090129, + "grad_norm": 0.15987404630231764, + "learning_rate": 0.00011250000000000001, + "loss": 0.7493, + "step": 189 + }, + { + "epoch": 0.016988555078683833, + "grad_norm": 0.12117672321451946, + "learning_rate": 0.00011309523809523809, + "loss": 0.7473, + "step": 190 + }, + { + "epoch": 0.01707796852646638, + "grad_norm": 0.154543421313165, + "learning_rate": 0.00011369047619047619, + "loss": 0.7291, + "step": 191 + }, + { + "epoch": 0.017167381974248927, + "grad_norm": 0.14781763492138902, + "learning_rate": 0.00011428571428571428, + "loss": 0.7385, + "step": 192 + }, + { + "epoch": 0.017256795422031474, + "grad_norm": 0.14589946965149436, + "learning_rate": 0.00011488095238095238, + "loss": 0.7489, + "step": 193 + }, + { + "epoch": 0.01734620886981402, + "grad_norm": 0.13302171769381832, + "learning_rate": 0.00011547619047619047, + "loss": 0.7613, + "step": 194 + }, + { + "epoch": 0.017435622317596565, + "grad_norm": 0.13740644646958464, + "learning_rate": 0.00011607142857142858, + "loss": 0.7787, + "step": 195 + }, + { + "epoch": 0.017525035765379112, + "grad_norm": 0.11878070680962817, + "learning_rate": 0.00011666666666666668, + "loss": 0.753, + "step": 196 + }, + { + "epoch": 0.01761444921316166, + "grad_norm": 0.12138599845759415, + "learning_rate": 0.00011726190476190477, + "loss": 0.7698, + "step": 197 + }, + { + "epoch": 0.017703862660944206, + "grad_norm": 0.1391398279969926, + "learning_rate": 0.00011785714285714287, + "loss": 0.7593, + "step": 198 + }, + { + "epoch": 0.017793276108726754, + "grad_norm": 0.12359411765095017, + "learning_rate": 0.00011845238095238097, + "loss": 0.7418, + "step": 199 + }, + { + "epoch": 0.0178826895565093, + "grad_norm": 0.14897714828021416, + "learning_rate": 0.00011904761904761905, + "loss": 0.7924, + "step": 200 + }, + { + "epoch": 0.017972103004291844, + "grad_norm": 0.1290404214796986, + "learning_rate": 0.00011964285714285714, + "loss": 0.7884, + "step": 201 + }, + { + "epoch": 0.01806151645207439, + "grad_norm": 0.1337610525558041, + "learning_rate": 0.00012023809523809524, + "loss": 0.7534, + "step": 202 + }, + { + "epoch": 0.01815092989985694, + "grad_norm": 0.1373615283367947, + "learning_rate": 0.00012083333333333333, + "loss": 0.7288, + "step": 203 + }, + { + "epoch": 0.018240343347639486, + "grad_norm": 0.1378490034842756, + "learning_rate": 0.00012142857142857143, + "loss": 0.7471, + "step": 204 + }, + { + "epoch": 0.018329756795422033, + "grad_norm": 0.1255569911087147, + "learning_rate": 0.00012202380952380954, + "loss": 0.7236, + "step": 205 + }, + { + "epoch": 0.018419170243204577, + "grad_norm": 0.11936964870607598, + "learning_rate": 0.00012261904761904762, + "loss": 0.7356, + "step": 206 + }, + { + "epoch": 0.018508583690987124, + "grad_norm": 0.13331046357570073, + "learning_rate": 0.00012321428571428572, + "loss": 0.7354, + "step": 207 + }, + { + "epoch": 0.01859799713876967, + "grad_norm": 0.15818879783505116, + "learning_rate": 0.0001238095238095238, + "loss": 0.7665, + "step": 208 + }, + { + "epoch": 0.018687410586552218, + "grad_norm": 0.22536233948124632, + "learning_rate": 0.0001244047619047619, + "loss": 0.7604, + "step": 209 + }, + { + "epoch": 0.018776824034334765, + "grad_norm": 0.17574879739689883, + "learning_rate": 0.000125, + "loss": 0.7225, + "step": 210 + }, + { + "epoch": 0.01886623748211731, + "grad_norm": 0.12347129312941954, + "learning_rate": 0.0001255952380952381, + "loss": 0.7729, + "step": 211 + }, + { + "epoch": 0.018955650929899856, + "grad_norm": 0.2149225144465704, + "learning_rate": 0.0001261904761904762, + "loss": 0.7413, + "step": 212 + }, + { + "epoch": 0.019045064377682403, + "grad_norm": 0.13412475148544334, + "learning_rate": 0.0001267857142857143, + "loss": 0.7451, + "step": 213 + }, + { + "epoch": 0.01913447782546495, + "grad_norm": 0.16197274033985393, + "learning_rate": 0.00012738095238095238, + "loss": 0.7883, + "step": 214 + }, + { + "epoch": 0.019223891273247497, + "grad_norm": 0.13237551022691998, + "learning_rate": 0.00012797619047619048, + "loss": 0.719, + "step": 215 + }, + { + "epoch": 0.019313304721030045, + "grad_norm": 0.131562377625104, + "learning_rate": 0.00012857142857142858, + "loss": 0.7315, + "step": 216 + }, + { + "epoch": 0.019402718168812588, + "grad_norm": 0.136728077306639, + "learning_rate": 0.00012916666666666667, + "loss": 0.774, + "step": 217 + }, + { + "epoch": 0.019492131616595135, + "grad_norm": 0.10560079034401827, + "learning_rate": 0.00012976190476190477, + "loss": 0.7455, + "step": 218 + }, + { + "epoch": 0.019581545064377683, + "grad_norm": 0.11249293421648487, + "learning_rate": 0.00013035714285714286, + "loss": 0.7314, + "step": 219 + }, + { + "epoch": 0.01967095851216023, + "grad_norm": 0.11903801252103464, + "learning_rate": 0.00013095238095238096, + "loss": 0.7429, + "step": 220 + }, + { + "epoch": 0.019760371959942777, + "grad_norm": 0.11350806195735766, + "learning_rate": 0.00013154761904761905, + "loss": 0.7581, + "step": 221 + }, + { + "epoch": 0.01984978540772532, + "grad_norm": 0.11903584928545503, + "learning_rate": 0.00013214285714285715, + "loss": 0.7567, + "step": 222 + }, + { + "epoch": 0.019939198855507868, + "grad_norm": 0.1107704086711381, + "learning_rate": 0.00013273809523809524, + "loss": 0.7346, + "step": 223 + }, + { + "epoch": 0.020028612303290415, + "grad_norm": 0.13073193157434274, + "learning_rate": 0.00013333333333333334, + "loss": 0.7077, + "step": 224 + }, + { + "epoch": 0.020118025751072962, + "grad_norm": 0.11448147627936948, + "learning_rate": 0.00013392857142857144, + "loss": 0.7183, + "step": 225 + }, + { + "epoch": 0.02020743919885551, + "grad_norm": 0.13320871541592277, + "learning_rate": 0.00013452380952380953, + "loss": 0.7164, + "step": 226 + }, + { + "epoch": 0.020296852646638053, + "grad_norm": 0.11656220916458761, + "learning_rate": 0.00013511904761904763, + "loss": 0.7452, + "step": 227 + }, + { + "epoch": 0.0203862660944206, + "grad_norm": 0.12542212044852505, + "learning_rate": 0.00013571428571428572, + "loss": 0.718, + "step": 228 + }, + { + "epoch": 0.020475679542203147, + "grad_norm": 0.11822120118256729, + "learning_rate": 0.00013630952380952382, + "loss": 0.7124, + "step": 229 + }, + { + "epoch": 0.020565092989985694, + "grad_norm": 0.1224093443331628, + "learning_rate": 0.0001369047619047619, + "loss": 0.7493, + "step": 230 + }, + { + "epoch": 0.02065450643776824, + "grad_norm": 0.11410510698216417, + "learning_rate": 0.0001375, + "loss": 0.672, + "step": 231 + }, + { + "epoch": 0.02074391988555079, + "grad_norm": 0.16449353187904192, + "learning_rate": 0.0001380952380952381, + "loss": 0.7696, + "step": 232 + }, + { + "epoch": 0.020833333333333332, + "grad_norm": 0.14209766566128276, + "learning_rate": 0.0001386904761904762, + "loss": 0.739, + "step": 233 + }, + { + "epoch": 0.02092274678111588, + "grad_norm": 0.11404079252462487, + "learning_rate": 0.0001392857142857143, + "loss": 0.7398, + "step": 234 + }, + { + "epoch": 0.021012160228898426, + "grad_norm": 0.15588365446933602, + "learning_rate": 0.0001398809523809524, + "loss": 0.7369, + "step": 235 + }, + { + "epoch": 0.021101573676680974, + "grad_norm": 0.1407608771285815, + "learning_rate": 0.00014047619047619049, + "loss": 0.7467, + "step": 236 + }, + { + "epoch": 0.02119098712446352, + "grad_norm": 0.13811367779063655, + "learning_rate": 0.00014107142857142858, + "loss": 0.7745, + "step": 237 + }, + { + "epoch": 0.021280400572246064, + "grad_norm": 0.143458800029559, + "learning_rate": 0.00014166666666666668, + "loss": 0.7587, + "step": 238 + }, + { + "epoch": 0.02136981402002861, + "grad_norm": 0.13285383201149709, + "learning_rate": 0.00014226190476190477, + "loss": 0.7362, + "step": 239 + }, + { + "epoch": 0.02145922746781116, + "grad_norm": 0.10777370792127913, + "learning_rate": 0.00014285714285714287, + "loss": 0.7592, + "step": 240 + }, + { + "epoch": 0.021548640915593706, + "grad_norm": 0.11730059403481756, + "learning_rate": 0.00014345238095238096, + "loss": 0.714, + "step": 241 + }, + { + "epoch": 0.021638054363376253, + "grad_norm": 0.11578609928386505, + "learning_rate": 0.00014404761904761906, + "loss": 0.7235, + "step": 242 + }, + { + "epoch": 0.021727467811158797, + "grad_norm": 0.10329880397742973, + "learning_rate": 0.00014464285714285715, + "loss": 0.7216, + "step": 243 + }, + { + "epoch": 0.021816881258941344, + "grad_norm": 0.1337235719118983, + "learning_rate": 0.00014523809523809525, + "loss": 0.7493, + "step": 244 + }, + { + "epoch": 0.02190629470672389, + "grad_norm": 0.11444811139249986, + "learning_rate": 0.00014583333333333335, + "loss": 0.7324, + "step": 245 + }, + { + "epoch": 0.021995708154506438, + "grad_norm": 0.13233936026647528, + "learning_rate": 0.00014642857142857141, + "loss": 0.7462, + "step": 246 + }, + { + "epoch": 0.022085121602288985, + "grad_norm": 0.11496225017144215, + "learning_rate": 0.00014702380952380954, + "loss": 0.7506, + "step": 247 + }, + { + "epoch": 0.022174535050071532, + "grad_norm": 0.16404919010490004, + "learning_rate": 0.00014761904761904763, + "loss": 0.7409, + "step": 248 + }, + { + "epoch": 0.022263948497854076, + "grad_norm": 0.12800262148718047, + "learning_rate": 0.00014821428571428573, + "loss": 0.7448, + "step": 249 + }, + { + "epoch": 0.022353361945636623, + "grad_norm": 0.13392774659542356, + "learning_rate": 0.00014880952380952382, + "loss": 0.7741, + "step": 250 + }, + { + "epoch": 0.02244277539341917, + "grad_norm": 0.13332599430639042, + "learning_rate": 0.00014940476190476192, + "loss": 0.7398, + "step": 251 + }, + { + "epoch": 0.022532188841201718, + "grad_norm": 0.15228604381856575, + "learning_rate": 0.00015000000000000001, + "loss": 0.7077, + "step": 252 + }, + { + "epoch": 0.022621602288984265, + "grad_norm": 0.1354341911630598, + "learning_rate": 0.0001505952380952381, + "loss": 0.7145, + "step": 253 + }, + { + "epoch": 0.02271101573676681, + "grad_norm": 0.13777127566778816, + "learning_rate": 0.0001511904761904762, + "loss": 0.7233, + "step": 254 + }, + { + "epoch": 0.022800429184549355, + "grad_norm": 0.1330470351291351, + "learning_rate": 0.00015178571428571427, + "loss": 0.7097, + "step": 255 + }, + { + "epoch": 0.022889842632331903, + "grad_norm": 0.12924276227988632, + "learning_rate": 0.00015238095238095237, + "loss": 0.7151, + "step": 256 + }, + { + "epoch": 0.02297925608011445, + "grad_norm": 0.13726313087941044, + "learning_rate": 0.00015297619047619046, + "loss": 0.7284, + "step": 257 + }, + { + "epoch": 0.023068669527896997, + "grad_norm": 0.1324665269539585, + "learning_rate": 0.0001535714285714286, + "loss": 0.782, + "step": 258 + }, + { + "epoch": 0.02315808297567954, + "grad_norm": 0.11384380776078969, + "learning_rate": 0.00015416666666666668, + "loss": 0.7581, + "step": 259 + }, + { + "epoch": 0.023247496423462088, + "grad_norm": 0.12353044699141566, + "learning_rate": 0.00015476190476190478, + "loss": 0.7729, + "step": 260 + }, + { + "epoch": 0.023336909871244635, + "grad_norm": 0.09859086236758895, + "learning_rate": 0.00015535714285714287, + "loss": 0.6687, + "step": 261 + }, + { + "epoch": 0.023426323319027182, + "grad_norm": 0.08452824218011731, + "learning_rate": 0.00015595238095238097, + "loss": 0.216, + "step": 262 + }, + { + "epoch": 0.02351573676680973, + "grad_norm": 0.11085941053005569, + "learning_rate": 0.00015654761904761906, + "loss": 0.7514, + "step": 263 + }, + { + "epoch": 0.023605150214592276, + "grad_norm": 0.11753304180755175, + "learning_rate": 0.00015714285714285716, + "loss": 0.7832, + "step": 264 + }, + { + "epoch": 0.02369456366237482, + "grad_norm": 0.12087226217745185, + "learning_rate": 0.00015773809523809523, + "loss": 0.7248, + "step": 265 + }, + { + "epoch": 0.023783977110157367, + "grad_norm": 0.12260089521427592, + "learning_rate": 0.00015833333333333332, + "loss": 0.7425, + "step": 266 + }, + { + "epoch": 0.023873390557939914, + "grad_norm": 0.11754670137318607, + "learning_rate": 0.00015892857142857142, + "loss": 0.7249, + "step": 267 + }, + { + "epoch": 0.02396280400572246, + "grad_norm": 0.12296957064500091, + "learning_rate": 0.00015952380952380954, + "loss": 0.7251, + "step": 268 + }, + { + "epoch": 0.02405221745350501, + "grad_norm": 0.1401975618816341, + "learning_rate": 0.00016011904761904764, + "loss": 0.7819, + "step": 269 + }, + { + "epoch": 0.024141630901287552, + "grad_norm": 0.12208986323984777, + "learning_rate": 0.00016071428571428573, + "loss": 0.7248, + "step": 270 + }, + { + "epoch": 0.0242310443490701, + "grad_norm": 0.12890812488971884, + "learning_rate": 0.00016130952380952383, + "loss": 0.7197, + "step": 271 + }, + { + "epoch": 0.024320457796852647, + "grad_norm": 0.11674462170879474, + "learning_rate": 0.00016190476190476192, + "loss": 0.734, + "step": 272 + }, + { + "epoch": 0.024409871244635194, + "grad_norm": 0.12443383178289107, + "learning_rate": 0.00016250000000000002, + "loss": 0.7234, + "step": 273 + }, + { + "epoch": 0.02449928469241774, + "grad_norm": 0.12879328867445525, + "learning_rate": 0.0001630952380952381, + "loss": 0.728, + "step": 274 + }, + { + "epoch": 0.024588698140200285, + "grad_norm": 0.1232326193528575, + "learning_rate": 0.00016369047619047618, + "loss": 0.765, + "step": 275 + }, + { + "epoch": 0.02467811158798283, + "grad_norm": 0.15732183989065884, + "learning_rate": 0.00016428571428571428, + "loss": 0.7418, + "step": 276 + }, + { + "epoch": 0.02476752503576538, + "grad_norm": 0.16785729964789686, + "learning_rate": 0.00016488095238095237, + "loss": 0.2421, + "step": 277 + }, + { + "epoch": 0.024856938483547926, + "grad_norm": 0.18737567434155666, + "learning_rate": 0.00016547619047619047, + "loss": 0.7258, + "step": 278 + }, + { + "epoch": 0.024946351931330473, + "grad_norm": 0.11891006709248339, + "learning_rate": 0.0001660714285714286, + "loss": 0.7155, + "step": 279 + }, + { + "epoch": 0.02503576537911302, + "grad_norm": 0.2680207997256139, + "learning_rate": 0.0001666666666666667, + "loss": 0.7345, + "step": 280 + }, + { + "epoch": 0.025125178826895564, + "grad_norm": 0.11289986680198684, + "learning_rate": 0.00016726190476190478, + "loss": 0.7183, + "step": 281 + }, + { + "epoch": 0.02521459227467811, + "grad_norm": 0.14117456461860917, + "learning_rate": 0.00016785714285714288, + "loss": 0.7815, + "step": 282 + }, + { + "epoch": 0.025304005722460658, + "grad_norm": 0.12018254667517389, + "learning_rate": 0.00016845238095238097, + "loss": 0.2344, + "step": 283 + }, + { + "epoch": 0.025393419170243205, + "grad_norm": 0.13161546200958293, + "learning_rate": 0.00016904761904761904, + "loss": 0.7452, + "step": 284 + }, + { + "epoch": 0.025482832618025753, + "grad_norm": 0.1202687825894608, + "learning_rate": 0.00016964285714285714, + "loss": 0.7277, + "step": 285 + }, + { + "epoch": 0.025572246065808296, + "grad_norm": 0.15180219422011978, + "learning_rate": 0.00017023809523809523, + "loss": 0.7353, + "step": 286 + }, + { + "epoch": 0.025661659513590843, + "grad_norm": 0.11042089918045021, + "learning_rate": 0.00017083333333333333, + "loss": 0.7125, + "step": 287 + }, + { + "epoch": 0.02575107296137339, + "grad_norm": 0.11020971655276005, + "learning_rate": 0.00017142857142857143, + "loss": 0.697, + "step": 288 + }, + { + "epoch": 0.025840486409155938, + "grad_norm": 0.1248739257296159, + "learning_rate": 0.00017202380952380955, + "loss": 0.7647, + "step": 289 + }, + { + "epoch": 0.025929899856938485, + "grad_norm": 0.1779147543825648, + "learning_rate": 0.00017261904761904764, + "loss": 0.7189, + "step": 290 + }, + { + "epoch": 0.02601931330472103, + "grad_norm": 0.1522869096854565, + "learning_rate": 0.00017321428571428574, + "loss": 0.7663, + "step": 291 + }, + { + "epoch": 0.026108726752503576, + "grad_norm": 0.12569538970618935, + "learning_rate": 0.00017380952380952383, + "loss": 0.7398, + "step": 292 + }, + { + "epoch": 0.026198140200286123, + "grad_norm": 0.11098513622862206, + "learning_rate": 0.0001744047619047619, + "loss": 0.751, + "step": 293 + }, + { + "epoch": 0.02628755364806867, + "grad_norm": 0.11551880361854323, + "learning_rate": 0.000175, + "loss": 0.7502, + "step": 294 + }, + { + "epoch": 0.026376967095851217, + "grad_norm": 0.11760880993008917, + "learning_rate": 0.0001755952380952381, + "loss": 0.7266, + "step": 295 + }, + { + "epoch": 0.026466380543633764, + "grad_norm": 0.11393912164246002, + "learning_rate": 0.0001761904761904762, + "loss": 0.7516, + "step": 296 + }, + { + "epoch": 0.026555793991416308, + "grad_norm": 0.13238148167959923, + "learning_rate": 0.00017678571428571428, + "loss": 0.7574, + "step": 297 + }, + { + "epoch": 0.026645207439198855, + "grad_norm": 0.15770165808551959, + "learning_rate": 0.00017738095238095238, + "loss": 0.743, + "step": 298 + }, + { + "epoch": 0.026734620886981402, + "grad_norm": 0.12114941984925703, + "learning_rate": 0.00017797619047619048, + "loss": 0.7311, + "step": 299 + }, + { + "epoch": 0.02682403433476395, + "grad_norm": 0.13526515102667552, + "learning_rate": 0.0001785714285714286, + "loss": 0.7295, + "step": 300 + }, + { + "epoch": 0.026913447782546496, + "grad_norm": 0.11934507664065921, + "learning_rate": 0.0001791666666666667, + "loss": 0.7501, + "step": 301 + }, + { + "epoch": 0.02700286123032904, + "grad_norm": 0.11330019797215576, + "learning_rate": 0.00017976190476190476, + "loss": 0.7168, + "step": 302 + }, + { + "epoch": 0.027092274678111587, + "grad_norm": 0.11453654436561021, + "learning_rate": 0.00018035714285714286, + "loss": 0.7304, + "step": 303 + }, + { + "epoch": 0.027181688125894134, + "grad_norm": 0.11777299926278684, + "learning_rate": 0.00018095238095238095, + "loss": 0.7092, + "step": 304 + }, + { + "epoch": 0.02727110157367668, + "grad_norm": 0.12330708103071383, + "learning_rate": 0.00018154761904761905, + "loss": 0.7872, + "step": 305 + }, + { + "epoch": 0.02736051502145923, + "grad_norm": 0.11558045901686517, + "learning_rate": 0.00018214285714285714, + "loss": 0.7257, + "step": 306 + }, + { + "epoch": 0.027449928469241772, + "grad_norm": 0.11875104442745467, + "learning_rate": 0.00018273809523809524, + "loss": 0.7378, + "step": 307 + }, + { + "epoch": 0.02753934191702432, + "grad_norm": 0.38971470947618647, + "learning_rate": 0.00018333333333333334, + "loss": 0.2917, + "step": 308 + }, + { + "epoch": 0.027628755364806867, + "grad_norm": 0.13377153983939627, + "learning_rate": 0.00018392857142857143, + "loss": 0.7062, + "step": 309 + }, + { + "epoch": 0.027718168812589414, + "grad_norm": 0.13657162627701286, + "learning_rate": 0.00018452380952380955, + "loss": 0.7219, + "step": 310 + }, + { + "epoch": 0.02780758226037196, + "grad_norm": 0.12728853833650272, + "learning_rate": 0.00018511904761904765, + "loss": 0.7461, + "step": 311 + }, + { + "epoch": 0.027896995708154508, + "grad_norm": 0.12244251774292765, + "learning_rate": 0.00018571428571428572, + "loss": 0.7035, + "step": 312 + }, + { + "epoch": 0.027986409155937052, + "grad_norm": 0.14128394908503994, + "learning_rate": 0.0001863095238095238, + "loss": 0.7411, + "step": 313 + }, + { + "epoch": 0.0280758226037196, + "grad_norm": 0.1254620267079459, + "learning_rate": 0.0001869047619047619, + "loss": 0.7183, + "step": 314 + }, + { + "epoch": 0.028165236051502146, + "grad_norm": 0.11696352046766664, + "learning_rate": 0.0001875, + "loss": 0.7416, + "step": 315 + }, + { + "epoch": 0.028254649499284693, + "grad_norm": 0.11614105217866469, + "learning_rate": 0.0001880952380952381, + "loss": 0.7123, + "step": 316 + }, + { + "epoch": 0.02834406294706724, + "grad_norm": 0.11603734943866223, + "learning_rate": 0.0001886904761904762, + "loss": 0.6997, + "step": 317 + }, + { + "epoch": 0.028433476394849784, + "grad_norm": 0.1223248474171665, + "learning_rate": 0.0001892857142857143, + "loss": 0.6911, + "step": 318 + }, + { + "epoch": 0.02852288984263233, + "grad_norm": 0.1120065698938123, + "learning_rate": 0.00018988095238095239, + "loss": 0.7008, + "step": 319 + }, + { + "epoch": 0.02861230329041488, + "grad_norm": 0.12896039844187662, + "learning_rate": 0.00019047619047619048, + "loss": 0.7256, + "step": 320 + }, + { + "epoch": 0.028701716738197425, + "grad_norm": 0.10824455093047315, + "learning_rate": 0.00019107142857142858, + "loss": 0.7293, + "step": 321 + }, + { + "epoch": 0.028791130185979973, + "grad_norm": 0.12488857908655322, + "learning_rate": 0.00019166666666666667, + "loss": 0.7496, + "step": 322 + }, + { + "epoch": 0.028880543633762516, + "grad_norm": 0.1321786162655319, + "learning_rate": 0.00019226190476190477, + "loss": 0.7172, + "step": 323 + }, + { + "epoch": 0.028969957081545063, + "grad_norm": 0.11881800071233804, + "learning_rate": 0.00019285714285714286, + "loss": 0.7134, + "step": 324 + }, + { + "epoch": 0.02905937052932761, + "grad_norm": 0.12579620551562218, + "learning_rate": 0.00019345238095238096, + "loss": 0.7136, + "step": 325 + }, + { + "epoch": 0.029148783977110158, + "grad_norm": 0.12936483076088823, + "learning_rate": 0.00019404761904761905, + "loss": 0.7301, + "step": 326 + }, + { + "epoch": 0.029238197424892705, + "grad_norm": 0.1250682091889051, + "learning_rate": 0.00019464285714285715, + "loss": 0.7372, + "step": 327 + }, + { + "epoch": 0.029327610872675252, + "grad_norm": 0.1256659700956961, + "learning_rate": 0.00019523809523809525, + "loss": 0.7344, + "step": 328 + }, + { + "epoch": 0.029417024320457796, + "grad_norm": 0.12018191749678514, + "learning_rate": 0.00019583333333333334, + "loss": 0.7231, + "step": 329 + }, + { + "epoch": 0.029506437768240343, + "grad_norm": 0.1650862342656057, + "learning_rate": 0.00019642857142857144, + "loss": 0.7456, + "step": 330 + }, + { + "epoch": 0.02959585121602289, + "grad_norm": 0.11325014379309774, + "learning_rate": 0.00019702380952380953, + "loss": 0.7323, + "step": 331 + }, + { + "epoch": 0.029685264663805437, + "grad_norm": 0.1278108803678704, + "learning_rate": 0.00019761904761904763, + "loss": 0.729, + "step": 332 + }, + { + "epoch": 0.029774678111587984, + "grad_norm": 0.10570726672893542, + "learning_rate": 0.00019821428571428572, + "loss": 0.7364, + "step": 333 + }, + { + "epoch": 0.029864091559370528, + "grad_norm": 0.15728997309851345, + "learning_rate": 0.00019880952380952382, + "loss": 0.7338, + "step": 334 + }, + { + "epoch": 0.029953505007153075, + "grad_norm": 0.12458965213169466, + "learning_rate": 0.00019940476190476191, + "loss": 0.743, + "step": 335 + }, + { + "epoch": 0.030042918454935622, + "grad_norm": 0.15485741311910156, + "learning_rate": 0.0002, + "loss": 0.7297, + "step": 336 + }, + { + "epoch": 0.03013233190271817, + "grad_norm": 0.11646273917315782, + "learning_rate": 0.0001999999958065604, + "loss": 0.7212, + "step": 337 + }, + { + "epoch": 0.030221745350500717, + "grad_norm": 0.19083471783807998, + "learning_rate": 0.0001999999832262419, + "loss": 0.7016, + "step": 338 + }, + { + "epoch": 0.03031115879828326, + "grad_norm": 0.11631955216447437, + "learning_rate": 0.00019999996225904558, + "loss": 0.7567, + "step": 339 + }, + { + "epoch": 0.030400572246065807, + "grad_norm": 0.1341032337889414, + "learning_rate": 0.00019999993290497318, + "loss": 0.7539, + "step": 340 + }, + { + "epoch": 0.030489985693848354, + "grad_norm": 0.1116019150572653, + "learning_rate": 0.0001999998951640272, + "loss": 0.7092, + "step": 341 + }, + { + "epoch": 0.0305793991416309, + "grad_norm": 0.11030882790636407, + "learning_rate": 0.0001999998490362108, + "loss": 0.7239, + "step": 342 + }, + { + "epoch": 0.03066881258941345, + "grad_norm": 0.13101281893425468, + "learning_rate": 0.0001999997945215278, + "loss": 0.742, + "step": 343 + }, + { + "epoch": 0.030758226037195996, + "grad_norm": 0.10805489497082699, + "learning_rate": 0.0001999997316199828, + "loss": 0.7415, + "step": 344 + }, + { + "epoch": 0.03084763948497854, + "grad_norm": 0.12390908099122365, + "learning_rate": 0.00019999966033158108, + "loss": 0.7191, + "step": 345 + }, + { + "epoch": 0.030937052932761087, + "grad_norm": 0.1282648391330519, + "learning_rate": 0.00019999958065632862, + "loss": 0.7551, + "step": 346 + }, + { + "epoch": 0.031026466380543634, + "grad_norm": 0.12931102972534342, + "learning_rate": 0.0001999994925942321, + "loss": 0.7549, + "step": 347 + }, + { + "epoch": 0.03111587982832618, + "grad_norm": 0.12016938603980129, + "learning_rate": 0.00019999939614529893, + "loss": 0.744, + "step": 348 + }, + { + "epoch": 0.031205293276108728, + "grad_norm": 0.13244770942656758, + "learning_rate": 0.00019999929130953714, + "loss": 0.736, + "step": 349 + }, + { + "epoch": 0.03129470672389127, + "grad_norm": 0.10874197779649829, + "learning_rate": 0.00019999917808695558, + "loss": 0.7118, + "step": 350 + }, + { + "epoch": 0.03138412017167382, + "grad_norm": 0.14014871421039424, + "learning_rate": 0.0001999990564775637, + "loss": 0.7435, + "step": 351 + }, + { + "epoch": 0.031473533619456366, + "grad_norm": 0.12307161625419644, + "learning_rate": 0.00019999892648137174, + "loss": 0.7491, + "step": 352 + }, + { + "epoch": 0.03156294706723891, + "grad_norm": 0.14285250877835, + "learning_rate": 0.00019999878809839056, + "loss": 0.7037, + "step": 353 + }, + { + "epoch": 0.03165236051502146, + "grad_norm": 0.11837578709858497, + "learning_rate": 0.0001999986413286318, + "loss": 0.7108, + "step": 354 + }, + { + "epoch": 0.031741773962804004, + "grad_norm": 0.12509519276377967, + "learning_rate": 0.00019999848617210776, + "loss": 0.7691, + "step": 355 + }, + { + "epoch": 0.031831187410586555, + "grad_norm": 0.11220302455326481, + "learning_rate": 0.00019999832262883148, + "loss": 0.7243, + "step": 356 + }, + { + "epoch": 0.0319206008583691, + "grad_norm": 0.11059088582520307, + "learning_rate": 0.00019999815069881663, + "loss": 0.7232, + "step": 357 + }, + { + "epoch": 0.03201001430615164, + "grad_norm": 0.10273273875402593, + "learning_rate": 0.00019999797038207763, + "loss": 0.7598, + "step": 358 + }, + { + "epoch": 0.03209942775393419, + "grad_norm": 0.11092203601215612, + "learning_rate": 0.00019999778167862964, + "loss": 0.7565, + "step": 359 + }, + { + "epoch": 0.032188841201716736, + "grad_norm": 0.11719737953435666, + "learning_rate": 0.00019999758458848847, + "loss": 0.7157, + "step": 360 + }, + { + "epoch": 0.03227825464949929, + "grad_norm": 0.12563781586629327, + "learning_rate": 0.00019999737911167065, + "loss": 0.7652, + "step": 361 + }, + { + "epoch": 0.03236766809728183, + "grad_norm": 0.10998641223968911, + "learning_rate": 0.00019999716524819337, + "loss": 0.7082, + "step": 362 + }, + { + "epoch": 0.032457081545064374, + "grad_norm": 0.12227825025650838, + "learning_rate": 0.00019999694299807465, + "loss": 0.7408, + "step": 363 + }, + { + "epoch": 0.032546494992846925, + "grad_norm": 0.1274160102061918, + "learning_rate": 0.0001999967123613331, + "loss": 0.7397, + "step": 364 + }, + { + "epoch": 0.03263590844062947, + "grad_norm": 0.1094028297895593, + "learning_rate": 0.000199996473337988, + "loss": 0.7246, + "step": 365 + }, + { + "epoch": 0.03272532188841202, + "grad_norm": 0.12379037814946615, + "learning_rate": 0.0001999962259280595, + "loss": 0.7347, + "step": 366 + }, + { + "epoch": 0.03281473533619456, + "grad_norm": 0.11272655024197642, + "learning_rate": 0.00019999597013156824, + "loss": 0.7066, + "step": 367 + }, + { + "epoch": 0.032904148783977114, + "grad_norm": 0.12707157868725777, + "learning_rate": 0.00019999570594853575, + "loss": 0.7008, + "step": 368 + }, + { + "epoch": 0.03299356223175966, + "grad_norm": 0.10875945095299727, + "learning_rate": 0.0001999954333789842, + "loss": 0.7067, + "step": 369 + }, + { + "epoch": 0.0330829756795422, + "grad_norm": 0.12000516472482455, + "learning_rate": 0.00019999515242293637, + "loss": 0.7392, + "step": 370 + }, + { + "epoch": 0.03317238912732475, + "grad_norm": 0.124077510697105, + "learning_rate": 0.0001999948630804159, + "loss": 0.7633, + "step": 371 + }, + { + "epoch": 0.033261802575107295, + "grad_norm": 0.12602408584860128, + "learning_rate": 0.000199994565351447, + "loss": 0.7408, + "step": 372 + }, + { + "epoch": 0.033351216022889846, + "grad_norm": 0.1273306108117371, + "learning_rate": 0.00019999425923605468, + "loss": 0.7156, + "step": 373 + }, + { + "epoch": 0.03344062947067239, + "grad_norm": 0.11601226700571922, + "learning_rate": 0.0001999939447342646, + "loss": 0.7166, + "step": 374 + }, + { + "epoch": 0.03353004291845493, + "grad_norm": 0.10230660617628208, + "learning_rate": 0.00019999362184610316, + "loss": 0.6953, + "step": 375 + }, + { + "epoch": 0.033619456366237484, + "grad_norm": 0.12038011838512679, + "learning_rate": 0.00019999329057159736, + "loss": 0.742, + "step": 376 + }, + { + "epoch": 0.03370886981402003, + "grad_norm": 0.12400196663268573, + "learning_rate": 0.0001999929509107751, + "loss": 0.7492, + "step": 377 + }, + { + "epoch": 0.03379828326180258, + "grad_norm": 0.11374582691855968, + "learning_rate": 0.00019999260286366477, + "loss": 0.7085, + "step": 378 + }, + { + "epoch": 0.03388769670958512, + "grad_norm": 0.10976125231263266, + "learning_rate": 0.00019999224643029565, + "loss": 0.726, + "step": 379 + }, + { + "epoch": 0.033977110157367665, + "grad_norm": 0.11778009384473788, + "learning_rate": 0.0001999918816106975, + "loss": 0.7077, + "step": 380 + }, + { + "epoch": 0.034066523605150216, + "grad_norm": 0.12112441058766611, + "learning_rate": 0.00019999150840490105, + "loss": 0.7108, + "step": 381 + }, + { + "epoch": 0.03415593705293276, + "grad_norm": 0.12041762266285676, + "learning_rate": 0.00019999112681293757, + "loss": 0.713, + "step": 382 + }, + { + "epoch": 0.03424535050071531, + "grad_norm": 0.12453836275818993, + "learning_rate": 0.000199990736834839, + "loss": 0.7417, + "step": 383 + }, + { + "epoch": 0.034334763948497854, + "grad_norm": 0.11303772735129876, + "learning_rate": 0.00019999033847063811, + "loss": 0.7403, + "step": 384 + }, + { + "epoch": 0.0344241773962804, + "grad_norm": 0.1189184835426681, + "learning_rate": 0.00019998993172036828, + "loss": 0.7002, + "step": 385 + }, + { + "epoch": 0.03451359084406295, + "grad_norm": 0.12887161606488848, + "learning_rate": 0.00019998951658406364, + "loss": 0.7415, + "step": 386 + }, + { + "epoch": 0.03460300429184549, + "grad_norm": 0.12839330699377174, + "learning_rate": 0.000199989093061759, + "loss": 0.7786, + "step": 387 + }, + { + "epoch": 0.03469241773962804, + "grad_norm": 0.13128577742807626, + "learning_rate": 0.00019998866115348988, + "loss": 0.7375, + "step": 388 + }, + { + "epoch": 0.034781831187410586, + "grad_norm": 0.28743912494856, + "learning_rate": 0.0001999882208592925, + "loss": 0.2725, + "step": 389 + }, + { + "epoch": 0.03487124463519313, + "grad_norm": 0.1584433144201048, + "learning_rate": 0.00019998777217920385, + "loss": 0.7606, + "step": 390 + }, + { + "epoch": 0.03496065808297568, + "grad_norm": 0.17203550203827567, + "learning_rate": 0.0001999873151132614, + "loss": 0.7131, + "step": 391 + }, + { + "epoch": 0.035050071530758224, + "grad_norm": 0.12228720065787724, + "learning_rate": 0.00019998684966150365, + "loss": 0.7165, + "step": 392 + }, + { + "epoch": 0.035139484978540775, + "grad_norm": 0.11045693140413496, + "learning_rate": 0.00019998637582396958, + "loss": 0.7241, + "step": 393 + }, + { + "epoch": 0.03522889842632332, + "grad_norm": 0.14639512099023816, + "learning_rate": 0.0001999858936006989, + "loss": 0.7193, + "step": 394 + }, + { + "epoch": 0.03531831187410586, + "grad_norm": 0.12200798411881697, + "learning_rate": 0.00019998540299173207, + "loss": 0.7397, + "step": 395 + }, + { + "epoch": 0.03540772532188841, + "grad_norm": 2.371486834433153, + "learning_rate": 0.00019998490399711024, + "loss": 0.426, + "step": 396 + }, + { + "epoch": 0.035497138769670956, + "grad_norm": 0.13061540321615442, + "learning_rate": 0.0001999843966168753, + "loss": 0.7388, + "step": 397 + }, + { + "epoch": 0.03558655221745351, + "grad_norm": 0.11958270756363219, + "learning_rate": 0.00019998388085106972, + "loss": 0.7439, + "step": 398 + }, + { + "epoch": 0.03567596566523605, + "grad_norm": 0.13428539470653567, + "learning_rate": 0.00019998335669973682, + "loss": 0.7185, + "step": 399 + }, + { + "epoch": 0.0357653791130186, + "grad_norm": 0.13688168292776953, + "learning_rate": 0.00019998282416292055, + "loss": 0.7251, + "step": 400 + }, + { + "epoch": 0.035854792560801145, + "grad_norm": 0.14668982953149365, + "learning_rate": 0.00019998228324066557, + "loss": 0.6993, + "step": 401 + }, + { + "epoch": 0.03594420600858369, + "grad_norm": 0.12453105303530129, + "learning_rate": 0.00019998173393301723, + "loss": 0.7287, + "step": 402 + }, + { + "epoch": 0.03603361945636624, + "grad_norm": 0.11695004292936023, + "learning_rate": 0.0001999811762400216, + "loss": 0.7325, + "step": 403 + }, + { + "epoch": 0.03612303290414878, + "grad_norm": 0.13583146441994237, + "learning_rate": 0.0001999806101617255, + "loss": 0.7612, + "step": 404 + }, + { + "epoch": 0.036212446351931334, + "grad_norm": 0.12240976378372644, + "learning_rate": 0.00019998003569817637, + "loss": 0.7286, + "step": 405 + }, + { + "epoch": 0.03630185979971388, + "grad_norm": 0.14466304288738005, + "learning_rate": 0.00019997945284942235, + "loss": 0.741, + "step": 406 + }, + { + "epoch": 0.03639127324749642, + "grad_norm": 0.19442476828606042, + "learning_rate": 0.0001999788616155124, + "loss": 0.7747, + "step": 407 + }, + { + "epoch": 0.03648068669527897, + "grad_norm": 0.16613266173699107, + "learning_rate": 0.00019997826199649605, + "loss": 0.7109, + "step": 408 + }, + { + "epoch": 0.036570100143061515, + "grad_norm": 0.13087456564240638, + "learning_rate": 0.00019997765399242364, + "loss": 0.6841, + "step": 409 + }, + { + "epoch": 0.036659513590844066, + "grad_norm": 0.20988646261670446, + "learning_rate": 0.0001999770376033461, + "loss": 0.7412, + "step": 410 + }, + { + "epoch": 0.03674892703862661, + "grad_norm": 0.2283098019741191, + "learning_rate": 0.00019997641282931515, + "loss": 0.7332, + "step": 411 + }, + { + "epoch": 0.03683834048640915, + "grad_norm": 0.1708864883599509, + "learning_rate": 0.00019997577967038324, + "loss": 0.7113, + "step": 412 + }, + { + "epoch": 0.036927753934191704, + "grad_norm": 0.19187612031815332, + "learning_rate": 0.0001999751381266034, + "loss": 0.7762, + "step": 413 + }, + { + "epoch": 0.03701716738197425, + "grad_norm": 0.2401839122147698, + "learning_rate": 0.00019997448819802948, + "loss": 0.7538, + "step": 414 + }, + { + "epoch": 0.0371065808297568, + "grad_norm": 0.14334890379453943, + "learning_rate": 0.00019997382988471595, + "loss": 0.7423, + "step": 415 + }, + { + "epoch": 0.03719599427753934, + "grad_norm": 0.18222047597973284, + "learning_rate": 0.00019997316318671806, + "loss": 0.7195, + "step": 416 + }, + { + "epoch": 0.037285407725321885, + "grad_norm": 0.15542233251699122, + "learning_rate": 0.00019997248810409173, + "loss": 0.713, + "step": 417 + }, + { + "epoch": 0.037374821173104436, + "grad_norm": 0.1523246800390254, + "learning_rate": 0.0001999718046368935, + "loss": 0.7924, + "step": 418 + }, + { + "epoch": 0.03746423462088698, + "grad_norm": 0.13541042509554974, + "learning_rate": 0.0001999711127851808, + "loss": 0.6995, + "step": 419 + }, + { + "epoch": 0.03755364806866953, + "grad_norm": 0.16668191859284895, + "learning_rate": 0.0001999704125490116, + "loss": 0.7478, + "step": 420 + }, + { + "epoch": 0.037643061516452074, + "grad_norm": 0.11359182378482696, + "learning_rate": 0.00019996970392844462, + "loss": 0.7157, + "step": 421 + }, + { + "epoch": 0.03773247496423462, + "grad_norm": 0.1375820016627743, + "learning_rate": 0.0001999689869235393, + "loss": 0.7401, + "step": 422 + }, + { + "epoch": 0.03782188841201717, + "grad_norm": 0.13913399037025748, + "learning_rate": 0.00019996826153435582, + "loss": 0.7077, + "step": 423 + }, + { + "epoch": 0.03791130185979971, + "grad_norm": 0.12883482893282758, + "learning_rate": 0.00019996752776095495, + "loss": 0.7076, + "step": 424 + }, + { + "epoch": 0.03800071530758226, + "grad_norm": 0.13774655488848023, + "learning_rate": 0.00019996678560339824, + "loss": 0.7501, + "step": 425 + }, + { + "epoch": 0.038090128755364806, + "grad_norm": 0.1348736166220882, + "learning_rate": 0.00019996603506174795, + "loss": 0.6877, + "step": 426 + }, + { + "epoch": 0.03817954220314735, + "grad_norm": 0.12263485232315835, + "learning_rate": 0.00019996527613606708, + "loss": 0.6977, + "step": 427 + }, + { + "epoch": 0.0382689556509299, + "grad_norm": 0.1220441099422687, + "learning_rate": 0.00019996450882641916, + "loss": 0.7213, + "step": 428 + }, + { + "epoch": 0.038358369098712444, + "grad_norm": 0.1325727434701897, + "learning_rate": 0.00019996373313286867, + "loss": 0.7393, + "step": 429 + }, + { + "epoch": 0.038447782546494995, + "grad_norm": 0.14632397905057162, + "learning_rate": 0.00019996294905548056, + "loss": 0.7481, + "step": 430 + }, + { + "epoch": 0.03853719599427754, + "grad_norm": 0.12792586044731089, + "learning_rate": 0.00019996215659432066, + "loss": 0.7458, + "step": 431 + }, + { + "epoch": 0.03862660944206009, + "grad_norm": 0.155887800937757, + "learning_rate": 0.00019996135574945544, + "loss": 0.7192, + "step": 432 + }, + { + "epoch": 0.03871602288984263, + "grad_norm": 0.11486977318558894, + "learning_rate": 0.00019996054652095198, + "loss": 0.7042, + "step": 433 + }, + { + "epoch": 0.038805436337625177, + "grad_norm": 0.14442196476371066, + "learning_rate": 0.00019995972890887823, + "loss": 0.719, + "step": 434 + }, + { + "epoch": 0.03889484978540773, + "grad_norm": 0.14699772440447398, + "learning_rate": 0.00019995890291330272, + "loss": 0.7381, + "step": 435 + }, + { + "epoch": 0.03898426323319027, + "grad_norm": 0.1321941486981512, + "learning_rate": 0.00019995806853429477, + "loss": 0.6969, + "step": 436 + }, + { + "epoch": 0.03907367668097282, + "grad_norm": 0.14916029981505494, + "learning_rate": 0.0001999572257719243, + "loss": 0.7287, + "step": 437 + }, + { + "epoch": 0.039163090128755365, + "grad_norm": 0.12316420051594341, + "learning_rate": 0.00019995637462626205, + "loss": 0.7216, + "step": 438 + }, + { + "epoch": 0.03925250357653791, + "grad_norm": 0.13301187462820527, + "learning_rate": 0.00019995551509737936, + "loss": 0.7272, + "step": 439 + }, + { + "epoch": 0.03934191702432046, + "grad_norm": 0.13308266029056567, + "learning_rate": 0.00019995464718534835, + "loss": 0.7158, + "step": 440 + }, + { + "epoch": 0.039431330472103, + "grad_norm": 0.12152477153460256, + "learning_rate": 0.00019995377089024178, + "loss": 0.744, + "step": 441 + }, + { + "epoch": 0.039520743919885554, + "grad_norm": 0.13968542244298335, + "learning_rate": 0.00019995288621213318, + "loss": 0.6954, + "step": 442 + }, + { + "epoch": 0.0396101573676681, + "grad_norm": 0.1407717515899488, + "learning_rate": 0.0001999519931510967, + "loss": 0.776, + "step": 443 + }, + { + "epoch": 0.03969957081545064, + "grad_norm": 0.11389336099159703, + "learning_rate": 0.00019995109170720728, + "loss": 0.7438, + "step": 444 + }, + { + "epoch": 0.03978898426323319, + "grad_norm": 0.14142677384661, + "learning_rate": 0.0001999501818805405, + "loss": 0.7699, + "step": 445 + }, + { + "epoch": 0.039878397711015735, + "grad_norm": 0.12926131475817237, + "learning_rate": 0.0001999492636711727, + "loss": 0.7581, + "step": 446 + }, + { + "epoch": 0.039967811158798286, + "grad_norm": 0.1224600873447628, + "learning_rate": 0.00019994833707918084, + "loss": 0.6797, + "step": 447 + }, + { + "epoch": 0.04005722460658083, + "grad_norm": 0.14014525523472846, + "learning_rate": 0.00019994740210464268, + "loss": 0.7434, + "step": 448 + }, + { + "epoch": 0.04014663805436337, + "grad_norm": 0.10956001672196543, + "learning_rate": 0.00019994645874763658, + "loss": 0.7145, + "step": 449 + }, + { + "epoch": 0.040236051502145924, + "grad_norm": 0.11586099405548478, + "learning_rate": 0.00019994550700824172, + "loss": 0.7274, + "step": 450 + }, + { + "epoch": 0.04032546494992847, + "grad_norm": 0.12552482519469693, + "learning_rate": 0.00019994454688653784, + "loss": 0.7489, + "step": 451 + }, + { + "epoch": 0.04041487839771102, + "grad_norm": 0.11869758659780758, + "learning_rate": 0.00019994357838260557, + "loss": 0.736, + "step": 452 + }, + { + "epoch": 0.04050429184549356, + "grad_norm": 0.13868907740751055, + "learning_rate": 0.00019994260149652603, + "loss": 0.7706, + "step": 453 + }, + { + "epoch": 0.040593705293276106, + "grad_norm": 0.11851009640690603, + "learning_rate": 0.00019994161622838126, + "loss": 0.7162, + "step": 454 + }, + { + "epoch": 0.040683118741058656, + "grad_norm": 0.1112811242646738, + "learning_rate": 0.0001999406225782538, + "loss": 0.7395, + "step": 455 + }, + { + "epoch": 0.0407725321888412, + "grad_norm": 0.12330557675505842, + "learning_rate": 0.00019993962054622703, + "loss": 0.7275, + "step": 456 + }, + { + "epoch": 0.04086194563662375, + "grad_norm": 0.1236925949509374, + "learning_rate": 0.00019993861013238497, + "loss": 0.7102, + "step": 457 + }, + { + "epoch": 0.040951359084406294, + "grad_norm": 0.11543910886576259, + "learning_rate": 0.0001999375913368124, + "loss": 0.7229, + "step": 458 + }, + { + "epoch": 0.04104077253218884, + "grad_norm": 0.1329058200643324, + "learning_rate": 0.00019993656415959472, + "loss": 0.745, + "step": 459 + }, + { + "epoch": 0.04113018597997139, + "grad_norm": 0.18431801228650269, + "learning_rate": 0.00019993552860081814, + "loss": 0.702, + "step": 460 + }, + { + "epoch": 0.04121959942775393, + "grad_norm": 0.15044448817043213, + "learning_rate": 0.00019993448466056938, + "loss": 0.7379, + "step": 461 + }, + { + "epoch": 0.04130901287553648, + "grad_norm": 0.1321243773237953, + "learning_rate": 0.00019993343233893615, + "loss": 0.7239, + "step": 462 + }, + { + "epoch": 0.041398426323319026, + "grad_norm": 0.13604694416986995, + "learning_rate": 0.00019993237163600663, + "loss": 0.7921, + "step": 463 + }, + { + "epoch": 0.04148783977110158, + "grad_norm": 0.1266737128009209, + "learning_rate": 0.00019993130255186977, + "loss": 0.7378, + "step": 464 + }, + { + "epoch": 0.04157725321888412, + "grad_norm": 0.11469304084554299, + "learning_rate": 0.00019993022508661525, + "loss": 0.7264, + "step": 465 + }, + { + "epoch": 0.041666666666666664, + "grad_norm": 0.12084023567268108, + "learning_rate": 0.00019992913924033349, + "loss": 0.7321, + "step": 466 + }, + { + "epoch": 0.041756080114449215, + "grad_norm": 0.12449659864232121, + "learning_rate": 0.00019992804501311543, + "loss": 0.6676, + "step": 467 + }, + { + "epoch": 0.04184549356223176, + "grad_norm": 0.11357433616397794, + "learning_rate": 0.00019992694240505293, + "loss": 0.7122, + "step": 468 + }, + { + "epoch": 0.04193490701001431, + "grad_norm": 0.12305125069463874, + "learning_rate": 0.00019992583141623848, + "loss": 0.719, + "step": 469 + }, + { + "epoch": 0.04202432045779685, + "grad_norm": 0.11010671400463687, + "learning_rate": 0.00019992471204676525, + "loss": 0.6846, + "step": 470 + }, + { + "epoch": 0.0421137339055794, + "grad_norm": 0.2310744660177931, + "learning_rate": 0.00019992358429672704, + "loss": 0.3611, + "step": 471 + }, + { + "epoch": 0.04220314735336195, + "grad_norm": 0.13163228241519886, + "learning_rate": 0.00019992244816621852, + "loss": 0.6978, + "step": 472 + }, + { + "epoch": 0.04229256080114449, + "grad_norm": 0.14820400128521902, + "learning_rate": 0.00019992130365533497, + "loss": 0.7275, + "step": 473 + }, + { + "epoch": 0.04238197424892704, + "grad_norm": 0.1812068513469353, + "learning_rate": 0.00019992015076417233, + "loss": 0.6901, + "step": 474 + }, + { + "epoch": 0.042471387696709585, + "grad_norm": 0.1273447061072419, + "learning_rate": 0.00019991898949282732, + "loss": 0.7016, + "step": 475 + }, + { + "epoch": 0.04256080114449213, + "grad_norm": 0.14010106465379876, + "learning_rate": 0.00019991781984139736, + "loss": 0.7933, + "step": 476 + }, + { + "epoch": 0.04265021459227468, + "grad_norm": 0.11008677835728194, + "learning_rate": 0.00019991664180998048, + "loss": 0.7237, + "step": 477 + }, + { + "epoch": 0.04273962804005722, + "grad_norm": 0.10431221945178994, + "learning_rate": 0.00019991545539867556, + "loss": 0.6871, + "step": 478 + }, + { + "epoch": 0.042829041487839774, + "grad_norm": 0.1247797210703578, + "learning_rate": 0.00019991426060758202, + "loss": 0.7062, + "step": 479 + }, + { + "epoch": 0.04291845493562232, + "grad_norm": 0.12283178791263442, + "learning_rate": 0.00019991305743680013, + "loss": 0.7316, + "step": 480 + }, + { + "epoch": 0.04300786838340486, + "grad_norm": 0.12833594176413815, + "learning_rate": 0.00019991184588643077, + "loss": 0.7573, + "step": 481 + }, + { + "epoch": 0.04309728183118741, + "grad_norm": 0.11836160033815801, + "learning_rate": 0.00019991062595657558, + "loss": 0.7144, + "step": 482 + }, + { + "epoch": 0.043186695278969955, + "grad_norm": 0.11920522504333539, + "learning_rate": 0.00019990939764733684, + "loss": 0.7204, + "step": 483 + }, + { + "epoch": 0.043276108726752506, + "grad_norm": 0.12712212112152219, + "learning_rate": 0.0001999081609588176, + "loss": 0.7121, + "step": 484 + }, + { + "epoch": 0.04336552217453505, + "grad_norm": 0.12610777187232877, + "learning_rate": 0.0001999069158911215, + "loss": 0.7, + "step": 485 + }, + { + "epoch": 0.04345493562231759, + "grad_norm": 0.12031850460736526, + "learning_rate": 0.00019990566244435307, + "loss": 0.7343, + "step": 486 + }, + { + "epoch": 0.043544349070100144, + "grad_norm": 0.1291839385653691, + "learning_rate": 0.0001999044006186174, + "loss": 0.7255, + "step": 487 + }, + { + "epoch": 0.04363376251788269, + "grad_norm": 0.12443306778324026, + "learning_rate": 0.00019990313041402024, + "loss": 0.7843, + "step": 488 + }, + { + "epoch": 0.04372317596566524, + "grad_norm": 0.11610422333968685, + "learning_rate": 0.00019990185183066825, + "loss": 0.7136, + "step": 489 + }, + { + "epoch": 0.04381258941344778, + "grad_norm": 0.1112609153106568, + "learning_rate": 0.00019990056486866858, + "loss": 0.7144, + "step": 490 + }, + { + "epoch": 0.043902002861230326, + "grad_norm": 0.10674488707936858, + "learning_rate": 0.00019989926952812916, + "loss": 0.7034, + "step": 491 + }, + { + "epoch": 0.043991416309012876, + "grad_norm": 0.5826648129831681, + "learning_rate": 0.00019989796580915866, + "loss": 0.3274, + "step": 492 + }, + { + "epoch": 0.04408082975679542, + "grad_norm": 0.10589073614370012, + "learning_rate": 0.0001998966537118664, + "loss": 0.6737, + "step": 493 + }, + { + "epoch": 0.04417024320457797, + "grad_norm": 0.10899781381059528, + "learning_rate": 0.0001998953332363625, + "loss": 0.7367, + "step": 494 + }, + { + "epoch": 0.044259656652360514, + "grad_norm": 0.12020789925109322, + "learning_rate": 0.00019989400438275758, + "loss": 0.6964, + "step": 495 + }, + { + "epoch": 0.044349070100143065, + "grad_norm": 0.11371493621001, + "learning_rate": 0.00019989266715116316, + "loss": 0.7217, + "step": 496 + }, + { + "epoch": 0.04443848354792561, + "grad_norm": 0.11308713400300244, + "learning_rate": 0.0001998913215416914, + "loss": 0.7204, + "step": 497 + }, + { + "epoch": 0.04452789699570815, + "grad_norm": 0.11869541268617906, + "learning_rate": 0.00019988996755445517, + "loss": 0.7138, + "step": 498 + }, + { + "epoch": 0.0446173104434907, + "grad_norm": 0.13057712635712268, + "learning_rate": 0.00019988860518956796, + "loss": 0.7243, + "step": 499 + }, + { + "epoch": 0.044706723891273246, + "grad_norm": 0.1243698650148934, + "learning_rate": 0.0001998872344471441, + "loss": 0.738, + "step": 500 + }, + { + "epoch": 0.0447961373390558, + "grad_norm": 0.11461058475387588, + "learning_rate": 0.00019988585532729848, + "loss": 0.7497, + "step": 501 + }, + { + "epoch": 0.04488555078683834, + "grad_norm": 0.11564624117472005, + "learning_rate": 0.00019988446783014683, + "loss": 0.6778, + "step": 502 + }, + { + "epoch": 0.044974964234620884, + "grad_norm": 0.10779902581960145, + "learning_rate": 0.0001998830719558055, + "loss": 0.7433, + "step": 503 + }, + { + "epoch": 0.045064377682403435, + "grad_norm": 0.12792061696079826, + "learning_rate": 0.00019988166770439154, + "loss": 0.7386, + "step": 504 + }, + { + "epoch": 0.04515379113018598, + "grad_norm": 0.12525614797932708, + "learning_rate": 0.00019988025507602274, + "loss": 0.7324, + "step": 505 + }, + { + "epoch": 0.04524320457796853, + "grad_norm": 0.11996501709047881, + "learning_rate": 0.00019987883407081753, + "loss": 0.744, + "step": 506 + }, + { + "epoch": 0.04533261802575107, + "grad_norm": 0.12429492266208864, + "learning_rate": 0.00019987740468889519, + "loss": 0.7081, + "step": 507 + }, + { + "epoch": 0.04542203147353362, + "grad_norm": 0.13800495679697136, + "learning_rate": 0.00019987596693037552, + "loss": 0.7739, + "step": 508 + }, + { + "epoch": 0.04551144492131617, + "grad_norm": 0.11118652055680096, + "learning_rate": 0.00019987452079537913, + "loss": 0.7455, + "step": 509 + }, + { + "epoch": 0.04560085836909871, + "grad_norm": 0.1365138179174107, + "learning_rate": 0.00019987306628402727, + "loss": 0.7249, + "step": 510 + }, + { + "epoch": 0.04569027181688126, + "grad_norm": 0.11100194018838709, + "learning_rate": 0.00019987160339644198, + "loss": 0.6872, + "step": 511 + }, + { + "epoch": 0.045779685264663805, + "grad_norm": 0.11559172536025529, + "learning_rate": 0.00019987013213274593, + "loss": 0.7403, + "step": 512 + }, + { + "epoch": 0.04586909871244635, + "grad_norm": 0.10857148983945396, + "learning_rate": 0.0001998686524930625, + "loss": 0.7133, + "step": 513 + }, + { + "epoch": 0.0459585121602289, + "grad_norm": 0.1183005303307398, + "learning_rate": 0.0001998671644775158, + "loss": 0.718, + "step": 514 + }, + { + "epoch": 0.04604792560801144, + "grad_norm": 0.1187229162372755, + "learning_rate": 0.00019986566808623062, + "loss": 0.7295, + "step": 515 + }, + { + "epoch": 0.046137339055793994, + "grad_norm": 0.12783336057460412, + "learning_rate": 0.00019986416331933246, + "loss": 0.7616, + "step": 516 + }, + { + "epoch": 0.04622675250357654, + "grad_norm": 0.11384143083830242, + "learning_rate": 0.00019986265017694755, + "loss": 0.7404, + "step": 517 + }, + { + "epoch": 0.04631616595135908, + "grad_norm": 0.11207926468424693, + "learning_rate": 0.00019986112865920277, + "loss": 0.7325, + "step": 518 + }, + { + "epoch": 0.04640557939914163, + "grad_norm": 0.12632272937615543, + "learning_rate": 0.00019985959876622574, + "loss": 0.7604, + "step": 519 + }, + { + "epoch": 0.046494992846924176, + "grad_norm": 0.11425083584034283, + "learning_rate": 0.00019985806049814474, + "loss": 0.7158, + "step": 520 + }, + { + "epoch": 0.046584406294706726, + "grad_norm": 0.13202266839008797, + "learning_rate": 0.0001998565138550888, + "loss": 0.7546, + "step": 521 + }, + { + "epoch": 0.04667381974248927, + "grad_norm": 0.12217957402411199, + "learning_rate": 0.00019985495883718764, + "loss": 0.741, + "step": 522 + }, + { + "epoch": 0.04676323319027181, + "grad_norm": 0.11753392880374167, + "learning_rate": 0.0001998533954445717, + "loss": 0.6976, + "step": 523 + }, + { + "epoch": 0.046852646638054364, + "grad_norm": 0.11917412938504263, + "learning_rate": 0.00019985182367737202, + "loss": 0.7367, + "step": 524 + }, + { + "epoch": 0.04694206008583691, + "grad_norm": 0.11300707008905075, + "learning_rate": 0.00019985024353572054, + "loss": 0.7196, + "step": 525 + }, + { + "epoch": 0.04703147353361946, + "grad_norm": 0.12670253342094262, + "learning_rate": 0.0001998486550197497, + "loss": 0.731, + "step": 526 + }, + { + "epoch": 0.047120886981402, + "grad_norm": 0.11913373175880343, + "learning_rate": 0.00019984705812959276, + "loss": 0.7402, + "step": 527 + }, + { + "epoch": 0.04721030042918455, + "grad_norm": 0.11861670788725329, + "learning_rate": 0.0001998454528653836, + "loss": 0.7474, + "step": 528 + }, + { + "epoch": 0.047299713876967096, + "grad_norm": 0.12133552878659408, + "learning_rate": 0.00019984383922725695, + "loss": 0.7229, + "step": 529 + }, + { + "epoch": 0.04738912732474964, + "grad_norm": 0.1357948610041324, + "learning_rate": 0.00019984221721534805, + "loss": 0.7425, + "step": 530 + }, + { + "epoch": 0.04747854077253219, + "grad_norm": 0.10760063019550692, + "learning_rate": 0.00019984058682979297, + "loss": 0.7006, + "step": 531 + }, + { + "epoch": 0.047567954220314734, + "grad_norm": 0.11133961017796704, + "learning_rate": 0.00019983894807072848, + "loss": 0.7552, + "step": 532 + }, + { + "epoch": 0.047657367668097285, + "grad_norm": 0.10958678723641495, + "learning_rate": 0.00019983730093829194, + "loss": 0.693, + "step": 533 + }, + { + "epoch": 0.04774678111587983, + "grad_norm": 0.10682738586593554, + "learning_rate": 0.00019983564543262156, + "loss": 0.7142, + "step": 534 + }, + { + "epoch": 0.04783619456366237, + "grad_norm": 0.10349764955321224, + "learning_rate": 0.0001998339815538562, + "loss": 0.7147, + "step": 535 + }, + { + "epoch": 0.04792560801144492, + "grad_norm": 0.12383127199719264, + "learning_rate": 0.00019983230930213536, + "loss": 0.7213, + "step": 536 + }, + { + "epoch": 0.04801502145922747, + "grad_norm": 0.12118548786710415, + "learning_rate": 0.00019983062867759928, + "loss": 0.7561, + "step": 537 + }, + { + "epoch": 0.04810443490701002, + "grad_norm": 0.11978565940007246, + "learning_rate": 0.00019982893968038896, + "loss": 0.7463, + "step": 538 + }, + { + "epoch": 0.04819384835479256, + "grad_norm": 0.11728295552349778, + "learning_rate": 0.00019982724231064602, + "loss": 0.716, + "step": 539 + }, + { + "epoch": 0.048283261802575105, + "grad_norm": 0.13036533559079455, + "learning_rate": 0.00019982553656851284, + "loss": 0.7691, + "step": 540 + }, + { + "epoch": 0.048372675250357655, + "grad_norm": 0.11417041111922423, + "learning_rate": 0.00019982382245413248, + "loss": 0.7306, + "step": 541 + }, + { + "epoch": 0.0484620886981402, + "grad_norm": 0.11428266902053073, + "learning_rate": 0.00019982209996764866, + "loss": 0.7183, + "step": 542 + }, + { + "epoch": 0.04855150214592275, + "grad_norm": 0.12253808241783207, + "learning_rate": 0.0001998203691092059, + "loss": 0.719, + "step": 543 + }, + { + "epoch": 0.04864091559370529, + "grad_norm": 0.11015936922333894, + "learning_rate": 0.00019981862987894934, + "loss": 0.7331, + "step": 544 + }, + { + "epoch": 0.04873032904148784, + "grad_norm": 0.1258709792617412, + "learning_rate": 0.0001998168822770248, + "loss": 0.724, + "step": 545 + }, + { + "epoch": 0.04881974248927039, + "grad_norm": 0.11343786439192628, + "learning_rate": 0.0001998151263035789, + "loss": 0.6962, + "step": 546 + }, + { + "epoch": 0.04890915593705293, + "grad_norm": 0.11741581768865289, + "learning_rate": 0.00019981336195875894, + "loss": 0.7053, + "step": 547 + }, + { + "epoch": 0.04899856938483548, + "grad_norm": 0.11102650136253375, + "learning_rate": 0.00019981158924271283, + "loss": 0.6922, + "step": 548 + }, + { + "epoch": 0.049087982832618025, + "grad_norm": 0.11439168856417327, + "learning_rate": 0.00019980980815558925, + "loss": 0.7232, + "step": 549 + }, + { + "epoch": 0.04917739628040057, + "grad_norm": 0.12059398771398898, + "learning_rate": 0.00019980801869753765, + "loss": 0.7378, + "step": 550 + }, + { + "epoch": 0.04926680972818312, + "grad_norm": 0.10355396539362899, + "learning_rate": 0.00019980622086870803, + "loss": 0.7134, + "step": 551 + }, + { + "epoch": 0.04935622317596566, + "grad_norm": 0.10233412266879746, + "learning_rate": 0.00019980441466925118, + "loss": 0.7145, + "step": 552 + }, + { + "epoch": 0.049445636623748214, + "grad_norm": 0.11267698580180091, + "learning_rate": 0.00019980260009931864, + "loss": 0.7318, + "step": 553 + }, + { + "epoch": 0.04953505007153076, + "grad_norm": 0.1281553172546472, + "learning_rate": 0.00019980077715906256, + "loss": 0.7576, + "step": 554 + }, + { + "epoch": 0.0496244635193133, + "grad_norm": 0.12522467148760894, + "learning_rate": 0.0001997989458486358, + "loss": 0.7134, + "step": 555 + }, + { + "epoch": 0.04971387696709585, + "grad_norm": 0.13902897634113479, + "learning_rate": 0.000199797106168192, + "loss": 0.7418, + "step": 556 + }, + { + "epoch": 0.049803290414878396, + "grad_norm": 0.12003172306207267, + "learning_rate": 0.00019979525811788542, + "loss": 0.6763, + "step": 557 + }, + { + "epoch": 0.049892703862660946, + "grad_norm": 0.13146480793854576, + "learning_rate": 0.0001997934016978711, + "loss": 0.7226, + "step": 558 + }, + { + "epoch": 0.04998211731044349, + "grad_norm": 0.13074134422973932, + "learning_rate": 0.00019979153690830463, + "loss": 0.7295, + "step": 559 + }, + { + "epoch": 0.05007153075822604, + "grad_norm": 0.13523130211686687, + "learning_rate": 0.00019978966374934254, + "loss": 0.7631, + "step": 560 + }, + { + "epoch": 0.050160944206008584, + "grad_norm": 0.10209870672600624, + "learning_rate": 0.00019978778222114185, + "loss": 0.6572, + "step": 561 + }, + { + "epoch": 0.05025035765379113, + "grad_norm": 0.11435562568414748, + "learning_rate": 0.00019978589232386035, + "loss": 0.7019, + "step": 562 + }, + { + "epoch": 0.05033977110157368, + "grad_norm": 0.10281265524598004, + "learning_rate": 0.0001997839940576566, + "loss": 0.7185, + "step": 563 + }, + { + "epoch": 0.05042918454935622, + "grad_norm": 0.10545065073251603, + "learning_rate": 0.00019978208742268977, + "loss": 0.6838, + "step": 564 + }, + { + "epoch": 0.05051859799713877, + "grad_norm": 0.11555825879486703, + "learning_rate": 0.00019978017241911977, + "loss": 0.7211, + "step": 565 + }, + { + "epoch": 0.050608011444921316, + "grad_norm": 0.10505517254622675, + "learning_rate": 0.00019977824904710722, + "loss": 0.7113, + "step": 566 + }, + { + "epoch": 0.05069742489270386, + "grad_norm": 0.106354990559188, + "learning_rate": 0.00019977631730681343, + "loss": 0.6908, + "step": 567 + }, + { + "epoch": 0.05078683834048641, + "grad_norm": 0.130476231862864, + "learning_rate": 0.0001997743771984004, + "loss": 0.7359, + "step": 568 + }, + { + "epoch": 0.050876251788268954, + "grad_norm": 0.11052143543783513, + "learning_rate": 0.00019977242872203083, + "loss": 0.6875, + "step": 569 + }, + { + "epoch": 0.050965665236051505, + "grad_norm": 0.11625535120302236, + "learning_rate": 0.00019977047187786818, + "loss": 0.7511, + "step": 570 + }, + { + "epoch": 0.05105507868383405, + "grad_norm": 0.12011189551806827, + "learning_rate": 0.00019976850666607657, + "loss": 0.7322, + "step": 571 + }, + { + "epoch": 0.05114449213161659, + "grad_norm": 0.1165684011642952, + "learning_rate": 0.00019976653308682076, + "loss": 0.6842, + "step": 572 + }, + { + "epoch": 0.05123390557939914, + "grad_norm": 0.0964227025478255, + "learning_rate": 0.0001997645511402663, + "loss": 0.7176, + "step": 573 + }, + { + "epoch": 0.05132331902718169, + "grad_norm": 0.1226034645213074, + "learning_rate": 0.00019976256082657946, + "loss": 0.7608, + "step": 574 + }, + { + "epoch": 0.05141273247496424, + "grad_norm": 0.10974764503149916, + "learning_rate": 0.00019976056214592708, + "loss": 0.7447, + "step": 575 + }, + { + "epoch": 0.05150214592274678, + "grad_norm": 0.1209028485287643, + "learning_rate": 0.00019975855509847686, + "loss": 0.7021, + "step": 576 + }, + { + "epoch": 0.051591559370529325, + "grad_norm": 0.11565052392113946, + "learning_rate": 0.00019975653968439712, + "loss": 0.7009, + "step": 577 + }, + { + "epoch": 0.051680972818311875, + "grad_norm": 0.10487263513992738, + "learning_rate": 0.00019975451590385684, + "loss": 0.7077, + "step": 578 + }, + { + "epoch": 0.05177038626609442, + "grad_norm": 0.11955302933027727, + "learning_rate": 0.0001997524837570258, + "loss": 0.6753, + "step": 579 + }, + { + "epoch": 0.05185979971387697, + "grad_norm": 0.12828800368106652, + "learning_rate": 0.0001997504432440744, + "loss": 0.7206, + "step": 580 + }, + { + "epoch": 0.05194921316165951, + "grad_norm": 0.11941603588555633, + "learning_rate": 0.00019974839436517382, + "loss": 0.7012, + "step": 581 + }, + { + "epoch": 0.05203862660944206, + "grad_norm": 0.12051815052091772, + "learning_rate": 0.00019974633712049587, + "loss": 0.7044, + "step": 582 + }, + { + "epoch": 0.05212804005722461, + "grad_norm": 0.10924931546274173, + "learning_rate": 0.00019974427151021304, + "loss": 0.7398, + "step": 583 + }, + { + "epoch": 0.05221745350500715, + "grad_norm": 0.11454244371799958, + "learning_rate": 0.00019974219753449867, + "loss": 0.7423, + "step": 584 + }, + { + "epoch": 0.0523068669527897, + "grad_norm": 0.1141714267650137, + "learning_rate": 0.00019974011519352663, + "loss": 0.7097, + "step": 585 + }, + { + "epoch": 0.052396280400572245, + "grad_norm": 0.11711025039169974, + "learning_rate": 0.0001997380244874716, + "loss": 0.7379, + "step": 586 + }, + { + "epoch": 0.05248569384835479, + "grad_norm": 0.10319982662873195, + "learning_rate": 0.0001997359254165089, + "loss": 0.6977, + "step": 587 + }, + { + "epoch": 0.05257510729613734, + "grad_norm": 0.10991593389739647, + "learning_rate": 0.00019973381798081457, + "loss": 0.7081, + "step": 588 + }, + { + "epoch": 0.05266452074391988, + "grad_norm": 0.10913399128496833, + "learning_rate": 0.0001997317021805654, + "loss": 0.7193, + "step": 589 + }, + { + "epoch": 0.052753934191702434, + "grad_norm": 0.10916718253127075, + "learning_rate": 0.0001997295780159388, + "loss": 0.7108, + "step": 590 + }, + { + "epoch": 0.05284334763948498, + "grad_norm": 0.11088954321651731, + "learning_rate": 0.00019972744548711293, + "loss": 0.7698, + "step": 591 + }, + { + "epoch": 0.05293276108726753, + "grad_norm": 0.11957772683963319, + "learning_rate": 0.00019972530459426663, + "loss": 0.7279, + "step": 592 + }, + { + "epoch": 0.05302217453505007, + "grad_norm": 0.11633020013610394, + "learning_rate": 0.00019972315533757954, + "loss": 0.6944, + "step": 593 + }, + { + "epoch": 0.053111587982832616, + "grad_norm": 0.11226982417918474, + "learning_rate": 0.00019972099771723177, + "loss": 0.6966, + "step": 594 + }, + { + "epoch": 0.053201001430615166, + "grad_norm": 0.12031847774414826, + "learning_rate": 0.00019971883173340439, + "loss": 0.7248, + "step": 595 + }, + { + "epoch": 0.05329041487839771, + "grad_norm": 0.12164452771978108, + "learning_rate": 0.00019971665738627902, + "loss": 0.7328, + "step": 596 + }, + { + "epoch": 0.05337982832618026, + "grad_norm": 0.11919785135087896, + "learning_rate": 0.00019971447467603804, + "loss": 0.7108, + "step": 597 + }, + { + "epoch": 0.053469241773962804, + "grad_norm": 0.13140966008085206, + "learning_rate": 0.00019971228360286445, + "loss": 0.764, + "step": 598 + }, + { + "epoch": 0.05355865522174535, + "grad_norm": 0.1059109675375995, + "learning_rate": 0.00019971008416694208, + "loss": 0.6949, + "step": 599 + }, + { + "epoch": 0.0536480686695279, + "grad_norm": 0.1030946989556957, + "learning_rate": 0.00019970787636845535, + "loss": 0.7118, + "step": 600 + }, + { + "epoch": 0.05373748211731044, + "grad_norm": 0.10940864209239974, + "learning_rate": 0.00019970566020758947, + "loss": 0.7065, + "step": 601 + }, + { + "epoch": 0.05382689556509299, + "grad_norm": 0.1356128351551537, + "learning_rate": 0.0001997034356845303, + "loss": 0.7307, + "step": 602 + }, + { + "epoch": 0.053916309012875537, + "grad_norm": 0.13177307185035828, + "learning_rate": 0.00019970120279946436, + "loss": 0.6792, + "step": 603 + }, + { + "epoch": 0.05400572246065808, + "grad_norm": 0.1288289203873258, + "learning_rate": 0.00019969896155257896, + "loss": 0.7063, + "step": 604 + }, + { + "epoch": 0.05409513590844063, + "grad_norm": 0.1315765187618584, + "learning_rate": 0.00019969671194406205, + "loss": 0.7272, + "step": 605 + }, + { + "epoch": 0.054184549356223174, + "grad_norm": 0.11517914037274417, + "learning_rate": 0.0001996944539741023, + "loss": 0.6803, + "step": 606 + }, + { + "epoch": 0.054273962804005725, + "grad_norm": 0.1300176709097199, + "learning_rate": 0.00019969218764288914, + "loss": 0.7503, + "step": 607 + }, + { + "epoch": 0.05436337625178827, + "grad_norm": 0.10552141860878163, + "learning_rate": 0.0001996899129506126, + "loss": 0.717, + "step": 608 + }, + { + "epoch": 0.05445278969957081, + "grad_norm": 0.10673343627065872, + "learning_rate": 0.0001996876298974634, + "loss": 0.7029, + "step": 609 + }, + { + "epoch": 0.05454220314735336, + "grad_norm": 0.11044369394752275, + "learning_rate": 0.00019968533848363311, + "loss": 0.7231, + "step": 610 + }, + { + "epoch": 0.05463161659513591, + "grad_norm": 0.12831422853354982, + "learning_rate": 0.00019968303870931386, + "loss": 0.6809, + "step": 611 + }, + { + "epoch": 0.05472103004291846, + "grad_norm": 0.10050744744993807, + "learning_rate": 0.00019968073057469857, + "loss": 0.6826, + "step": 612 + }, + { + "epoch": 0.054810443490701, + "grad_norm": 0.10574733243860869, + "learning_rate": 0.00019967841407998076, + "loss": 0.7045, + "step": 613 + }, + { + "epoch": 0.054899856938483545, + "grad_norm": 0.11233222488269366, + "learning_rate": 0.00019967608922535476, + "loss": 0.731, + "step": 614 + }, + { + "epoch": 0.054989270386266095, + "grad_norm": 0.12601111473782725, + "learning_rate": 0.00019967375601101552, + "loss": 0.7278, + "step": 615 + }, + { + "epoch": 0.05507868383404864, + "grad_norm": 0.1012784704081863, + "learning_rate": 0.00019967141443715872, + "loss": 0.6842, + "step": 616 + }, + { + "epoch": 0.05516809728183119, + "grad_norm": 0.1182968014607365, + "learning_rate": 0.0001996690645039808, + "loss": 0.6934, + "step": 617 + }, + { + "epoch": 0.05525751072961373, + "grad_norm": 0.12346346571902439, + "learning_rate": 0.00019966670621167877, + "loss": 0.7485, + "step": 618 + }, + { + "epoch": 0.05534692417739628, + "grad_norm": 0.14070047577593403, + "learning_rate": 0.0001996643395604505, + "loss": 0.7542, + "step": 619 + }, + { + "epoch": 0.05543633762517883, + "grad_norm": 0.11556929887218749, + "learning_rate": 0.00019966196455049442, + "loss": 0.7153, + "step": 620 + }, + { + "epoch": 0.05552575107296137, + "grad_norm": 0.1492113996570166, + "learning_rate": 0.00019965958118200972, + "loss": 0.7543, + "step": 621 + }, + { + "epoch": 0.05561516452074392, + "grad_norm": 0.12084854692371828, + "learning_rate": 0.00019965718945519633, + "loss": 0.6757, + "step": 622 + }, + { + "epoch": 0.055704577968526466, + "grad_norm": 0.1262596228995325, + "learning_rate": 0.00019965478937025483, + "loss": 0.6951, + "step": 623 + }, + { + "epoch": 0.055793991416309016, + "grad_norm": 0.10926726958368663, + "learning_rate": 0.00019965238092738643, + "loss": 0.71, + "step": 624 + }, + { + "epoch": 0.05588340486409156, + "grad_norm": 0.10149379264344058, + "learning_rate": 0.00019964996412679325, + "loss": 0.679, + "step": 625 + }, + { + "epoch": 0.055972818311874104, + "grad_norm": 0.10776595317600692, + "learning_rate": 0.00019964753896867788, + "loss": 0.709, + "step": 626 + }, + { + "epoch": 0.056062231759656654, + "grad_norm": 0.09651104722160596, + "learning_rate": 0.00019964510545324382, + "loss": 0.696, + "step": 627 + }, + { + "epoch": 0.0561516452074392, + "grad_norm": 0.09853416782115001, + "learning_rate": 0.00019964266358069504, + "loss": 0.6642, + "step": 628 + }, + { + "epoch": 0.05624105865522175, + "grad_norm": 0.12858402576217884, + "learning_rate": 0.00019964021335123645, + "loss": 0.735, + "step": 629 + }, + { + "epoch": 0.05633047210300429, + "grad_norm": 0.12181016472736804, + "learning_rate": 0.00019963775476507348, + "loss": 0.7204, + "step": 630 + }, + { + "epoch": 0.056419885550786836, + "grad_norm": 0.11071425008604056, + "learning_rate": 0.00019963528782241237, + "loss": 0.7065, + "step": 631 + }, + { + "epoch": 0.056509298998569386, + "grad_norm": 0.1142391674478439, + "learning_rate": 0.00019963281252346, + "loss": 0.6917, + "step": 632 + }, + { + "epoch": 0.05659871244635193, + "grad_norm": 0.12010266695166508, + "learning_rate": 0.00019963032886842393, + "loss": 0.7276, + "step": 633 + }, + { + "epoch": 0.05668812589413448, + "grad_norm": 0.1057420179212418, + "learning_rate": 0.00019962783685751253, + "loss": 0.7371, + "step": 634 + }, + { + "epoch": 0.056777539341917024, + "grad_norm": 0.10846267883183804, + "learning_rate": 0.0001996253364909348, + "loss": 0.6842, + "step": 635 + }, + { + "epoch": 0.05686695278969957, + "grad_norm": 0.10588123883334041, + "learning_rate": 0.00019962282776890037, + "loss": 0.685, + "step": 636 + }, + { + "epoch": 0.05695636623748212, + "grad_norm": 0.10317255546535478, + "learning_rate": 0.0001996203106916197, + "loss": 0.6886, + "step": 637 + }, + { + "epoch": 0.05704577968526466, + "grad_norm": 0.10896973395095513, + "learning_rate": 0.00019961778525930387, + "loss": 0.7457, + "step": 638 + }, + { + "epoch": 0.05713519313304721, + "grad_norm": 0.17731257099495026, + "learning_rate": 0.00019961525147216475, + "loss": 0.3698, + "step": 639 + }, + { + "epoch": 0.05722460658082976, + "grad_norm": 0.10282058119675466, + "learning_rate": 0.00019961270933041477, + "loss": 0.681, + "step": 640 + }, + { + "epoch": 0.0573140200286123, + "grad_norm": 0.10487202399886368, + "learning_rate": 0.00019961015883426716, + "loss": 0.6926, + "step": 641 + }, + { + "epoch": 0.05740343347639485, + "grad_norm": 0.1271276720396297, + "learning_rate": 0.0001996075999839358, + "loss": 0.7286, + "step": 642 + }, + { + "epoch": 0.057492846924177395, + "grad_norm": 0.12414325815717936, + "learning_rate": 0.0001996050327796353, + "loss": 0.6559, + "step": 643 + }, + { + "epoch": 0.057582260371959945, + "grad_norm": 0.1182285446541165, + "learning_rate": 0.00019960245722158108, + "loss": 0.7044, + "step": 644 + }, + { + "epoch": 0.05767167381974249, + "grad_norm": 0.11741177057210497, + "learning_rate": 0.000199599873309989, + "loss": 0.6894, + "step": 645 + }, + { + "epoch": 0.05776108726752503, + "grad_norm": 0.12436590960473552, + "learning_rate": 0.00019959728104507586, + "loss": 0.6753, + "step": 646 + }, + { + "epoch": 0.05785050071530758, + "grad_norm": 0.1187517510352482, + "learning_rate": 0.00019959468042705903, + "loss": 0.7446, + "step": 647 + }, + { + "epoch": 0.05793991416309013, + "grad_norm": 0.11384450779812655, + "learning_rate": 0.00019959207145615665, + "loss": 0.7366, + "step": 648 + }, + { + "epoch": 0.05802932761087268, + "grad_norm": 0.1123309160577, + "learning_rate": 0.00019958945413258748, + "loss": 0.6869, + "step": 649 + }, + { + "epoch": 0.05811874105865522, + "grad_norm": 0.13387187123185781, + "learning_rate": 0.00019958682845657108, + "loss": 0.7532, + "step": 650 + }, + { + "epoch": 0.058208154506437765, + "grad_norm": 0.12229697675174298, + "learning_rate": 0.00019958419442832765, + "loss": 0.7104, + "step": 651 + }, + { + "epoch": 0.058297567954220315, + "grad_norm": 0.12133418576028704, + "learning_rate": 0.00019958155204807812, + "loss": 0.7439, + "step": 652 + }, + { + "epoch": 0.05838698140200286, + "grad_norm": 0.10076753550207558, + "learning_rate": 0.00019957890131604405, + "loss": 0.7017, + "step": 653 + }, + { + "epoch": 0.05847639484978541, + "grad_norm": 0.12154270812914209, + "learning_rate": 0.0001995762422324478, + "loss": 0.737, + "step": 654 + }, + { + "epoch": 0.05856580829756795, + "grad_norm": 0.13092701785649796, + "learning_rate": 0.00019957357479751236, + "loss": 0.7185, + "step": 655 + }, + { + "epoch": 0.058655221745350504, + "grad_norm": 0.11305947570040513, + "learning_rate": 0.00019957089901146148, + "loss": 0.7152, + "step": 656 + }, + { + "epoch": 0.05874463519313305, + "grad_norm": 0.12698687558093555, + "learning_rate": 0.00019956821487451953, + "loss": 0.7085, + "step": 657 + }, + { + "epoch": 0.05883404864091559, + "grad_norm": 0.11794452344495139, + "learning_rate": 0.00019956552238691166, + "loss": 0.7394, + "step": 658 + }, + { + "epoch": 0.05892346208869814, + "grad_norm": 0.11004638267530217, + "learning_rate": 0.00019956282154886369, + "loss": 0.7286, + "step": 659 + }, + { + "epoch": 0.059012875536480686, + "grad_norm": 0.10863897013520481, + "learning_rate": 0.00019956011236060207, + "loss": 0.7083, + "step": 660 + }, + { + "epoch": 0.059102288984263236, + "grad_norm": 0.1052749637518954, + "learning_rate": 0.0001995573948223541, + "loss": 0.6861, + "step": 661 + }, + { + "epoch": 0.05919170243204578, + "grad_norm": 0.12668146327847898, + "learning_rate": 0.00019955466893434767, + "loss": 0.7502, + "step": 662 + }, + { + "epoch": 0.059281115879828324, + "grad_norm": 0.10400503532709576, + "learning_rate": 0.00019955193469681137, + "loss": 0.7034, + "step": 663 + }, + { + "epoch": 0.059370529327610874, + "grad_norm": 0.11570984072686724, + "learning_rate": 0.00019954919210997453, + "loss": 0.7181, + "step": 664 + }, + { + "epoch": 0.05945994277539342, + "grad_norm": 0.12177012925083482, + "learning_rate": 0.00019954644117406718, + "loss": 0.7229, + "step": 665 + }, + { + "epoch": 0.05954935622317597, + "grad_norm": 0.10979409212057367, + "learning_rate": 0.00019954368188932002, + "loss": 0.6798, + "step": 666 + }, + { + "epoch": 0.05963876967095851, + "grad_norm": 0.11048085381239368, + "learning_rate": 0.0001995409142559645, + "loss": 0.7193, + "step": 667 + }, + { + "epoch": 0.059728183118741056, + "grad_norm": 0.1155352133139896, + "learning_rate": 0.0001995381382742327, + "loss": 0.7037, + "step": 668 + }, + { + "epoch": 0.059817596566523606, + "grad_norm": 0.11846201224866842, + "learning_rate": 0.00019953535394435744, + "loss": 0.6971, + "step": 669 + }, + { + "epoch": 0.05990701001430615, + "grad_norm": 0.11104955474465528, + "learning_rate": 0.0001995325612665723, + "loss": 0.7081, + "step": 670 + }, + { + "epoch": 0.0599964234620887, + "grad_norm": 0.12679044490778874, + "learning_rate": 0.00019952976024111143, + "loss": 0.7114, + "step": 671 + }, + { + "epoch": 0.060085836909871244, + "grad_norm": 0.12574112289809566, + "learning_rate": 0.00019952695086820975, + "loss": 0.7508, + "step": 672 + }, + { + "epoch": 0.06017525035765379, + "grad_norm": 0.10556097580054653, + "learning_rate": 0.0001995241331481029, + "loss": 0.6938, + "step": 673 + }, + { + "epoch": 0.06026466380543634, + "grad_norm": 0.12314668617666374, + "learning_rate": 0.00019952130708102722, + "loss": 0.7319, + "step": 674 + }, + { + "epoch": 0.06035407725321888, + "grad_norm": 0.1185744965696475, + "learning_rate": 0.0001995184726672197, + "loss": 0.6941, + "step": 675 + }, + { + "epoch": 0.06044349070100143, + "grad_norm": 0.11206266250830489, + "learning_rate": 0.00019951562990691807, + "loss": 0.7354, + "step": 676 + }, + { + "epoch": 0.06053290414878398, + "grad_norm": 0.12698502840264356, + "learning_rate": 0.00019951277880036073, + "loss": 0.7257, + "step": 677 + }, + { + "epoch": 0.06062231759656652, + "grad_norm": 0.12255110694984395, + "learning_rate": 0.0001995099193477868, + "loss": 0.7457, + "step": 678 + }, + { + "epoch": 0.06071173104434907, + "grad_norm": 0.12921242282796142, + "learning_rate": 0.00019950705154943613, + "loss": 0.2791, + "step": 679 + }, + { + "epoch": 0.060801144492131615, + "grad_norm": 0.12751380640273455, + "learning_rate": 0.00019950417540554925, + "loss": 0.6952, + "step": 680 + }, + { + "epoch": 0.060890557939914165, + "grad_norm": 0.10274639412072654, + "learning_rate": 0.00019950129091636732, + "loss": 0.3082, + "step": 681 + }, + { + "epoch": 0.06097997138769671, + "grad_norm": 0.15082807125174544, + "learning_rate": 0.00019949839808213227, + "loss": 0.7688, + "step": 682 + }, + { + "epoch": 0.06106938483547925, + "grad_norm": 0.1258675782529766, + "learning_rate": 0.00019949549690308677, + "loss": 0.7572, + "step": 683 + }, + { + "epoch": 0.0611587982832618, + "grad_norm": 0.1263568500009586, + "learning_rate": 0.0001994925873794741, + "loss": 0.7002, + "step": 684 + }, + { + "epoch": 0.06124821173104435, + "grad_norm": 0.13010152945817918, + "learning_rate": 0.00019948966951153824, + "loss": 0.7111, + "step": 685 + }, + { + "epoch": 0.0613376251788269, + "grad_norm": 0.10926266221059004, + "learning_rate": 0.000199486743299524, + "loss": 0.7405, + "step": 686 + }, + { + "epoch": 0.06142703862660944, + "grad_norm": 0.12643178082895828, + "learning_rate": 0.00019948380874367674, + "loss": 0.7787, + "step": 687 + }, + { + "epoch": 0.06151645207439199, + "grad_norm": 0.1086468243661553, + "learning_rate": 0.00019948086584424256, + "loss": 0.6941, + "step": 688 + }, + { + "epoch": 0.061605865522174535, + "grad_norm": 0.27076849029517364, + "learning_rate": 0.00019947791460146833, + "loss": 0.3067, + "step": 689 + }, + { + "epoch": 0.06169527896995708, + "grad_norm": 0.13956483887673393, + "learning_rate": 0.00019947495501560153, + "loss": 0.7227, + "step": 690 + }, + { + "epoch": 0.06178469241773963, + "grad_norm": 0.11670520415665588, + "learning_rate": 0.00019947198708689042, + "loss": 0.7429, + "step": 691 + }, + { + "epoch": 0.06187410586552217, + "grad_norm": 0.11892732218963382, + "learning_rate": 0.00019946901081558386, + "loss": 0.7303, + "step": 692 + }, + { + "epoch": 0.061963519313304724, + "grad_norm": 0.1292916735030595, + "learning_rate": 0.0001994660262019315, + "loss": 0.7763, + "step": 693 + }, + { + "epoch": 0.06205293276108727, + "grad_norm": 0.10909747916294513, + "learning_rate": 0.0001994630332461836, + "loss": 0.7168, + "step": 694 + }, + { + "epoch": 0.06214234620886981, + "grad_norm": 0.1245158041860307, + "learning_rate": 0.00019946003194859125, + "loss": 0.7173, + "step": 695 + }, + { + "epoch": 0.06223175965665236, + "grad_norm": 0.10571322318935426, + "learning_rate": 0.00019945702230940614, + "loss": 0.7158, + "step": 696 + }, + { + "epoch": 0.062321173104434906, + "grad_norm": 0.1353697527049124, + "learning_rate": 0.0001994540043288807, + "loss": 0.7321, + "step": 697 + }, + { + "epoch": 0.062410586552217456, + "grad_norm": 0.12204028192288369, + "learning_rate": 0.00019945097800726802, + "loss": 0.7256, + "step": 698 + }, + { + "epoch": 0.0625, + "grad_norm": 0.20323522085060386, + "learning_rate": 0.00019944794334482194, + "loss": 0.3161, + "step": 699 + }, + { + "epoch": 0.06258941344778254, + "grad_norm": 0.11454185889921673, + "learning_rate": 0.0001994449003417969, + "loss": 0.6943, + "step": 700 + }, + { + "epoch": 0.06267882689556509, + "grad_norm": 0.10596554400614061, + "learning_rate": 0.00019944184899844822, + "loss": 0.6784, + "step": 701 + }, + { + "epoch": 0.06276824034334764, + "grad_norm": 0.1386815537081674, + "learning_rate": 0.00019943878931503176, + "loss": 0.7783, + "step": 702 + }, + { + "epoch": 0.06285765379113019, + "grad_norm": 0.11593789077581639, + "learning_rate": 0.0001994357212918041, + "loss": 0.739, + "step": 703 + }, + { + "epoch": 0.06294706723891273, + "grad_norm": 0.11444407187962381, + "learning_rate": 0.00019943264492902258, + "loss": 0.7253, + "step": 704 + }, + { + "epoch": 0.06303648068669528, + "grad_norm": 0.11162510277411432, + "learning_rate": 0.00019942956022694523, + "loss": 0.6894, + "step": 705 + }, + { + "epoch": 0.06312589413447782, + "grad_norm": 0.1279649963957015, + "learning_rate": 0.00019942646718583076, + "loss": 0.701, + "step": 706 + }, + { + "epoch": 0.06321530758226038, + "grad_norm": 0.11934932094853835, + "learning_rate": 0.00019942336580593852, + "loss": 0.7047, + "step": 707 + }, + { + "epoch": 0.06330472103004292, + "grad_norm": 0.11352892351732356, + "learning_rate": 0.0001994202560875287, + "loss": 0.7416, + "step": 708 + }, + { + "epoch": 0.06339413447782546, + "grad_norm": 0.1202876008821823, + "learning_rate": 0.00019941713803086204, + "loss": 0.7023, + "step": 709 + }, + { + "epoch": 0.06348354792560801, + "grad_norm": 0.10879229271835981, + "learning_rate": 0.0001994140116362001, + "loss": 0.7076, + "step": 710 + }, + { + "epoch": 0.06357296137339055, + "grad_norm": 0.11879240801427894, + "learning_rate": 0.0001994108769038051, + "loss": 0.7301, + "step": 711 + }, + { + "epoch": 0.06366237482117311, + "grad_norm": 0.13575112531728897, + "learning_rate": 0.00019940773383393987, + "loss": 0.7558, + "step": 712 + }, + { + "epoch": 0.06375178826895565, + "grad_norm": 0.11419155953330484, + "learning_rate": 0.00019940458242686802, + "loss": 0.7436, + "step": 713 + }, + { + "epoch": 0.0638412017167382, + "grad_norm": 0.12580007538126606, + "learning_rate": 0.00019940142268285395, + "loss": 0.7263, + "step": 714 + }, + { + "epoch": 0.06393061516452074, + "grad_norm": 0.122550009893746, + "learning_rate": 0.0001993982546021626, + "loss": 0.7036, + "step": 715 + }, + { + "epoch": 0.06402002861230328, + "grad_norm": 0.1345760502688522, + "learning_rate": 0.00019939507818505966, + "loss": 0.7462, + "step": 716 + }, + { + "epoch": 0.06410944206008584, + "grad_norm": 0.11297934978227346, + "learning_rate": 0.00019939189343181157, + "loss": 0.694, + "step": 717 + }, + { + "epoch": 0.06419885550786839, + "grad_norm": 0.11302602809349127, + "learning_rate": 0.00019938870034268542, + "loss": 0.6836, + "step": 718 + }, + { + "epoch": 0.06428826895565093, + "grad_norm": 0.12123391776634203, + "learning_rate": 0.00019938549891794898, + "loss": 0.7382, + "step": 719 + }, + { + "epoch": 0.06437768240343347, + "grad_norm": 0.10871208025999943, + "learning_rate": 0.0001993822891578708, + "loss": 0.7292, + "step": 720 + }, + { + "epoch": 0.06446709585121602, + "grad_norm": 0.11748945420533358, + "learning_rate": 0.00019937907106272002, + "loss": 0.7106, + "step": 721 + }, + { + "epoch": 0.06455650929899857, + "grad_norm": 0.10246315742814112, + "learning_rate": 0.00019937584463276657, + "loss": 0.7086, + "step": 722 + }, + { + "epoch": 0.06464592274678112, + "grad_norm": 0.1322374728808258, + "learning_rate": 0.00019937260986828108, + "loss": 0.7324, + "step": 723 + }, + { + "epoch": 0.06473533619456366, + "grad_norm": 0.1071753209569447, + "learning_rate": 0.0001993693667695348, + "loss": 0.6884, + "step": 724 + }, + { + "epoch": 0.0648247496423462, + "grad_norm": 0.13239541676420957, + "learning_rate": 0.0001993661153367997, + "loss": 0.723, + "step": 725 + }, + { + "epoch": 0.06491416309012875, + "grad_norm": 0.12727923572636582, + "learning_rate": 0.00019936285557034858, + "loss": 0.7528, + "step": 726 + }, + { + "epoch": 0.0650035765379113, + "grad_norm": 0.11180535140396006, + "learning_rate": 0.00019935958747045472, + "loss": 0.6957, + "step": 727 + }, + { + "epoch": 0.06509298998569385, + "grad_norm": 0.13031931005076436, + "learning_rate": 0.00019935631103739225, + "loss": 0.701, + "step": 728 + }, + { + "epoch": 0.0651824034334764, + "grad_norm": 0.102436211501317, + "learning_rate": 0.00019935302627143594, + "loss": 0.6924, + "step": 729 + }, + { + "epoch": 0.06527181688125894, + "grad_norm": 0.1561590044725497, + "learning_rate": 0.00019934973317286138, + "loss": 0.2807, + "step": 730 + }, + { + "epoch": 0.0653612303290415, + "grad_norm": 0.13000728371763376, + "learning_rate": 0.00019934643174194462, + "loss": 0.7047, + "step": 731 + }, + { + "epoch": 0.06545064377682404, + "grad_norm": 0.13100754518560342, + "learning_rate": 0.00019934312197896262, + "loss": 0.7021, + "step": 732 + }, + { + "epoch": 0.06554005722460658, + "grad_norm": 0.12128402753555667, + "learning_rate": 0.00019933980388419297, + "loss": 0.7124, + "step": 733 + }, + { + "epoch": 0.06562947067238913, + "grad_norm": 0.13127660886264464, + "learning_rate": 0.00019933647745791393, + "loss": 0.7253, + "step": 734 + }, + { + "epoch": 0.06571888412017167, + "grad_norm": 0.13392719755133242, + "learning_rate": 0.0001993331427004045, + "loss": 0.7272, + "step": 735 + }, + { + "epoch": 0.06580829756795423, + "grad_norm": 0.11081614793303968, + "learning_rate": 0.00019932979961194435, + "loss": 0.6951, + "step": 736 + }, + { + "epoch": 0.06589771101573677, + "grad_norm": 0.11467504916675766, + "learning_rate": 0.00019932644819281389, + "loss": 0.711, + "step": 737 + }, + { + "epoch": 0.06598712446351931, + "grad_norm": 0.11584130537587219, + "learning_rate": 0.00019932308844329417, + "loss": 0.7069, + "step": 738 + }, + { + "epoch": 0.06607653791130186, + "grad_norm": 0.10802648487816047, + "learning_rate": 0.00019931972036366696, + "loss": 0.6657, + "step": 739 + }, + { + "epoch": 0.0661659513590844, + "grad_norm": 0.12104973100047736, + "learning_rate": 0.00019931634395421475, + "loss": 0.7105, + "step": 740 + }, + { + "epoch": 0.06625536480686696, + "grad_norm": 0.1291518627839904, + "learning_rate": 0.0001993129592152207, + "loss": 0.713, + "step": 741 + }, + { + "epoch": 0.0663447782546495, + "grad_norm": 0.12988633402209498, + "learning_rate": 0.00019930956614696874, + "loss": 0.7302, + "step": 742 + }, + { + "epoch": 0.06643419170243205, + "grad_norm": 0.12770303706949415, + "learning_rate": 0.0001993061647497434, + "loss": 0.746, + "step": 743 + }, + { + "epoch": 0.06652360515021459, + "grad_norm": 0.11141973116092994, + "learning_rate": 0.0001993027550238299, + "loss": 0.6985, + "step": 744 + }, + { + "epoch": 0.06661301859799713, + "grad_norm": 0.13568802772035435, + "learning_rate": 0.00019929933696951433, + "loss": 0.7459, + "step": 745 + }, + { + "epoch": 0.06670243204577969, + "grad_norm": 0.12512928891465577, + "learning_rate": 0.00019929591058708324, + "loss": 0.7411, + "step": 746 + }, + { + "epoch": 0.06679184549356224, + "grad_norm": 0.11702607042632575, + "learning_rate": 0.00019929247587682406, + "loss": 0.7145, + "step": 747 + }, + { + "epoch": 0.06688125894134478, + "grad_norm": 0.10694249420412473, + "learning_rate": 0.00019928903283902486, + "loss": 0.6483, + "step": 748 + }, + { + "epoch": 0.06697067238912732, + "grad_norm": 0.11976994355737157, + "learning_rate": 0.00019928558147397439, + "loss": 0.693, + "step": 749 + }, + { + "epoch": 0.06706008583690987, + "grad_norm": 0.12060326914946237, + "learning_rate": 0.0001992821217819621, + "loss": 0.711, + "step": 750 + }, + { + "epoch": 0.06714949928469242, + "grad_norm": 0.2171734317263738, + "learning_rate": 0.00019927865376327816, + "loss": 0.2884, + "step": 751 + }, + { + "epoch": 0.06723891273247497, + "grad_norm": 0.11361091197768486, + "learning_rate": 0.00019927517741821343, + "loss": 0.6895, + "step": 752 + }, + { + "epoch": 0.06732832618025751, + "grad_norm": 0.12060629961859654, + "learning_rate": 0.00019927169274705945, + "loss": 0.6962, + "step": 753 + }, + { + "epoch": 0.06741773962804005, + "grad_norm": 0.11467136124066674, + "learning_rate": 0.00019926819975010852, + "loss": 0.6763, + "step": 754 + }, + { + "epoch": 0.0675071530758226, + "grad_norm": 0.1216497828486784, + "learning_rate": 0.00019926469842765352, + "loss": 0.7249, + "step": 755 + }, + { + "epoch": 0.06759656652360516, + "grad_norm": 0.13626914920112956, + "learning_rate": 0.00019926118877998817, + "loss": 0.7461, + "step": 756 + }, + { + "epoch": 0.0676859799713877, + "grad_norm": 0.10654923540944328, + "learning_rate": 0.0001992576708074068, + "loss": 0.6965, + "step": 757 + }, + { + "epoch": 0.06777539341917024, + "grad_norm": 0.11342786693768103, + "learning_rate": 0.00019925414451020442, + "loss": 0.6745, + "step": 758 + }, + { + "epoch": 0.06786480686695279, + "grad_norm": 0.10888766048303178, + "learning_rate": 0.00019925060988867682, + "loss": 0.7109, + "step": 759 + }, + { + "epoch": 0.06795422031473533, + "grad_norm": 0.11099207067670487, + "learning_rate": 0.00019924706694312045, + "loss": 0.6741, + "step": 760 + }, + { + "epoch": 0.06804363376251789, + "grad_norm": 0.12685277979212797, + "learning_rate": 0.00019924351567383243, + "loss": 0.6913, + "step": 761 + }, + { + "epoch": 0.06813304721030043, + "grad_norm": 0.1412055321751857, + "learning_rate": 0.00019923995608111058, + "loss": 0.74, + "step": 762 + }, + { + "epoch": 0.06822246065808298, + "grad_norm": 0.1453848617714165, + "learning_rate": 0.0001992363881652535, + "loss": 0.7586, + "step": 763 + }, + { + "epoch": 0.06831187410586552, + "grad_norm": 0.13390199970618843, + "learning_rate": 0.0001992328119265604, + "loss": 0.7053, + "step": 764 + }, + { + "epoch": 0.06840128755364806, + "grad_norm": 0.10400630977195731, + "learning_rate": 0.0001992292273653312, + "loss": 0.6693, + "step": 765 + }, + { + "epoch": 0.06849070100143062, + "grad_norm": 0.11195713921351405, + "learning_rate": 0.00019922563448186652, + "loss": 0.7127, + "step": 766 + }, + { + "epoch": 0.06858011444921316, + "grad_norm": 0.11639679216866816, + "learning_rate": 0.00019922203327646772, + "loss": 0.7369, + "step": 767 + }, + { + "epoch": 0.06866952789699571, + "grad_norm": 0.118527951041702, + "learning_rate": 0.0001992184237494368, + "loss": 0.7151, + "step": 768 + }, + { + "epoch": 0.06875894134477825, + "grad_norm": 0.12556618919828047, + "learning_rate": 0.00019921480590107653, + "loss": 0.7284, + "step": 769 + }, + { + "epoch": 0.0688483547925608, + "grad_norm": 0.11276571892942718, + "learning_rate": 0.0001992111797316903, + "loss": 0.7203, + "step": 770 + }, + { + "epoch": 0.06893776824034335, + "grad_norm": 0.11411861201829823, + "learning_rate": 0.00019920754524158226, + "loss": 0.7052, + "step": 771 + }, + { + "epoch": 0.0690271816881259, + "grad_norm": 0.1225887940538595, + "learning_rate": 0.00019920390243105716, + "loss": 0.6818, + "step": 772 + }, + { + "epoch": 0.06911659513590844, + "grad_norm": 0.2096732764957758, + "learning_rate": 0.00019920025130042062, + "loss": 0.3123, + "step": 773 + }, + { + "epoch": 0.06920600858369098, + "grad_norm": 0.09752650929453743, + "learning_rate": 0.0001991965918499788, + "loss": 0.6728, + "step": 774 + }, + { + "epoch": 0.06929542203147353, + "grad_norm": 0.1088171454142821, + "learning_rate": 0.00019919292408003862, + "loss": 0.6929, + "step": 775 + }, + { + "epoch": 0.06938483547925609, + "grad_norm": 0.12027636115891209, + "learning_rate": 0.0001991892479909077, + "loss": 0.733, + "step": 776 + }, + { + "epoch": 0.06947424892703863, + "grad_norm": 0.12851046567189037, + "learning_rate": 0.0001991855635828943, + "loss": 0.7298, + "step": 777 + }, + { + "epoch": 0.06956366237482117, + "grad_norm": 0.10747451766433937, + "learning_rate": 0.00019918187085630752, + "loss": 0.7119, + "step": 778 + }, + { + "epoch": 0.06965307582260372, + "grad_norm": 0.12776264463757872, + "learning_rate": 0.000199178169811457, + "loss": 0.7513, + "step": 779 + }, + { + "epoch": 0.06974248927038626, + "grad_norm": 0.11500169727355654, + "learning_rate": 0.00019917446044865312, + "loss": 0.7104, + "step": 780 + }, + { + "epoch": 0.06983190271816882, + "grad_norm": 0.09680696110956487, + "learning_rate": 0.00019917074276820705, + "loss": 0.2554, + "step": 781 + }, + { + "epoch": 0.06992131616595136, + "grad_norm": 0.17058509599873686, + "learning_rate": 0.00019916701677043054, + "loss": 0.7076, + "step": 782 + }, + { + "epoch": 0.0700107296137339, + "grad_norm": 0.125522890287281, + "learning_rate": 0.00019916328245563611, + "loss": 0.7156, + "step": 783 + }, + { + "epoch": 0.07010014306151645, + "grad_norm": 0.14094426298141988, + "learning_rate": 0.0001991595398241369, + "loss": 0.7719, + "step": 784 + }, + { + "epoch": 0.07018955650929899, + "grad_norm": 0.09878823889080718, + "learning_rate": 0.0001991557888762469, + "loss": 0.705, + "step": 785 + }, + { + "epoch": 0.07027896995708155, + "grad_norm": 0.12437607790437163, + "learning_rate": 0.00019915202961228058, + "loss": 0.6895, + "step": 786 + }, + { + "epoch": 0.0703683834048641, + "grad_norm": 0.12512461417234466, + "learning_rate": 0.00019914826203255333, + "loss": 0.6912, + "step": 787 + }, + { + "epoch": 0.07045779685264664, + "grad_norm": 0.11272947279984495, + "learning_rate": 0.00019914448613738106, + "loss": 0.6959, + "step": 788 + }, + { + "epoch": 0.07054721030042918, + "grad_norm": 0.11272835418158479, + "learning_rate": 0.00019914070192708047, + "loss": 0.7182, + "step": 789 + }, + { + "epoch": 0.07063662374821172, + "grad_norm": 0.10862200609597249, + "learning_rate": 0.00019913690940196894, + "loss": 0.7198, + "step": 790 + }, + { + "epoch": 0.07072603719599428, + "grad_norm": 0.1270801815332772, + "learning_rate": 0.00019913310856236452, + "loss": 0.7359, + "step": 791 + }, + { + "epoch": 0.07081545064377683, + "grad_norm": 0.12285080032496072, + "learning_rate": 0.00019912929940858607, + "loss": 0.734, + "step": 792 + }, + { + "epoch": 0.07090486409155937, + "grad_norm": 0.11387003797166025, + "learning_rate": 0.00019912548194095297, + "loss": 0.6627, + "step": 793 + }, + { + "epoch": 0.07099427753934191, + "grad_norm": 0.09552715329689136, + "learning_rate": 0.0001991216561597854, + "loss": 0.7418, + "step": 794 + }, + { + "epoch": 0.07108369098712447, + "grad_norm": 0.11176613025477951, + "learning_rate": 0.00019911782206540423, + "loss": 0.7483, + "step": 795 + }, + { + "epoch": 0.07117310443490701, + "grad_norm": 0.11438308304319754, + "learning_rate": 0.00019911397965813107, + "loss": 0.7347, + "step": 796 + }, + { + "epoch": 0.07126251788268956, + "grad_norm": 0.10951880924938176, + "learning_rate": 0.0001991101289382881, + "loss": 0.721, + "step": 797 + }, + { + "epoch": 0.0713519313304721, + "grad_norm": 0.12401791985608758, + "learning_rate": 0.0001991062699061983, + "loss": 0.697, + "step": 798 + }, + { + "epoch": 0.07144134477825465, + "grad_norm": 0.12424305375479179, + "learning_rate": 0.00019910240256218535, + "loss": 0.6816, + "step": 799 + }, + { + "epoch": 0.0715307582260372, + "grad_norm": 0.12858124112694744, + "learning_rate": 0.00019909852690657359, + "loss": 0.7084, + "step": 800 + }, + { + "epoch": 0.07162017167381975, + "grad_norm": 0.1119533310905414, + "learning_rate": 0.00019909464293968804, + "loss": 0.7016, + "step": 801 + }, + { + "epoch": 0.07170958512160229, + "grad_norm": 0.11229402561328619, + "learning_rate": 0.0001990907506618545, + "loss": 0.7126, + "step": 802 + }, + { + "epoch": 0.07179899856938483, + "grad_norm": 0.11722474588031269, + "learning_rate": 0.00019908685007339932, + "loss": 0.7513, + "step": 803 + }, + { + "epoch": 0.07188841201716738, + "grad_norm": 0.10809150137137255, + "learning_rate": 0.00019908294117464975, + "loss": 0.6881, + "step": 804 + }, + { + "epoch": 0.07197782546494993, + "grad_norm": 0.10423523476682647, + "learning_rate": 0.00019907902396593352, + "loss": 0.6612, + "step": 805 + }, + { + "epoch": 0.07206723891273248, + "grad_norm": 0.11699964130731853, + "learning_rate": 0.00019907509844757925, + "loss": 0.7257, + "step": 806 + }, + { + "epoch": 0.07215665236051502, + "grad_norm": 0.1046725687517917, + "learning_rate": 0.00019907116461991605, + "loss": 0.6988, + "step": 807 + }, + { + "epoch": 0.07224606580829757, + "grad_norm": 0.11498412978464888, + "learning_rate": 0.00019906722248327397, + "loss": 0.7375, + "step": 808 + }, + { + "epoch": 0.07233547925608011, + "grad_norm": 0.11779836668244224, + "learning_rate": 0.0001990632720379836, + "loss": 0.7256, + "step": 809 + }, + { + "epoch": 0.07242489270386267, + "grad_norm": 0.11417659417137653, + "learning_rate": 0.00019905931328437624, + "loss": 0.7035, + "step": 810 + }, + { + "epoch": 0.07251430615164521, + "grad_norm": 0.11522720839868365, + "learning_rate": 0.00019905534622278388, + "loss": 0.7365, + "step": 811 + }, + { + "epoch": 0.07260371959942775, + "grad_norm": 0.12524950575858418, + "learning_rate": 0.00019905137085353926, + "loss": 0.707, + "step": 812 + }, + { + "epoch": 0.0726931330472103, + "grad_norm": 0.10433334639048773, + "learning_rate": 0.0001990473871769758, + "loss": 0.6707, + "step": 813 + }, + { + "epoch": 0.07278254649499284, + "grad_norm": 0.11065002645148805, + "learning_rate": 0.00019904339519342764, + "loss": 0.7047, + "step": 814 + }, + { + "epoch": 0.0728719599427754, + "grad_norm": 0.12135186790348351, + "learning_rate": 0.00019903939490322948, + "loss": 0.7197, + "step": 815 + }, + { + "epoch": 0.07296137339055794, + "grad_norm": 0.10588951743386193, + "learning_rate": 0.0001990353863067169, + "loss": 0.7175, + "step": 816 + }, + { + "epoch": 0.07305078683834049, + "grad_norm": 0.1043061905883348, + "learning_rate": 0.00019903136940422605, + "loss": 0.705, + "step": 817 + }, + { + "epoch": 0.07314020028612303, + "grad_norm": 0.10453329771639015, + "learning_rate": 0.00019902734419609389, + "loss": 0.7094, + "step": 818 + }, + { + "epoch": 0.07322961373390557, + "grad_norm": 0.11066267639129537, + "learning_rate": 0.00019902331068265793, + "loss": 0.7361, + "step": 819 + }, + { + "epoch": 0.07331902718168813, + "grad_norm": 0.11767755043355893, + "learning_rate": 0.00019901926886425653, + "loss": 0.721, + "step": 820 + }, + { + "epoch": 0.07340844062947068, + "grad_norm": 0.12616978255330857, + "learning_rate": 0.00019901521874122859, + "loss": 0.6925, + "step": 821 + }, + { + "epoch": 0.07349785407725322, + "grad_norm": 0.11283157467984306, + "learning_rate": 0.00019901116031391386, + "loss": 0.6961, + "step": 822 + }, + { + "epoch": 0.07358726752503576, + "grad_norm": 0.10761789787373853, + "learning_rate": 0.0001990070935826527, + "loss": 0.6872, + "step": 823 + }, + { + "epoch": 0.0736766809728183, + "grad_norm": 0.1184520345699307, + "learning_rate": 0.00019900301854778617, + "loss": 0.6726, + "step": 824 + }, + { + "epoch": 0.07376609442060086, + "grad_norm": 0.1147944668269074, + "learning_rate": 0.00019899893520965604, + "loss": 0.7313, + "step": 825 + }, + { + "epoch": 0.07385550786838341, + "grad_norm": 0.11345059805709111, + "learning_rate": 0.00019899484356860473, + "loss": 0.7356, + "step": 826 + }, + { + "epoch": 0.07394492131616595, + "grad_norm": 0.11825211017385515, + "learning_rate": 0.00019899074362497552, + "loss": 0.7432, + "step": 827 + }, + { + "epoch": 0.0740343347639485, + "grad_norm": 0.10979754083907678, + "learning_rate": 0.00019898663537911213, + "loss": 0.705, + "step": 828 + }, + { + "epoch": 0.07412374821173104, + "grad_norm": 0.10195802685257957, + "learning_rate": 0.00019898251883135922, + "loss": 0.6739, + "step": 829 + }, + { + "epoch": 0.0742131616595136, + "grad_norm": 0.1283557909424953, + "learning_rate": 0.00019897839398206197, + "loss": 0.7149, + "step": 830 + }, + { + "epoch": 0.07430257510729614, + "grad_norm": 0.11384825663644632, + "learning_rate": 0.00019897426083156634, + "loss": 0.7033, + "step": 831 + }, + { + "epoch": 0.07439198855507868, + "grad_norm": 0.11632496004265834, + "learning_rate": 0.000198970119380219, + "loss": 0.7243, + "step": 832 + }, + { + "epoch": 0.07448140200286123, + "grad_norm": 0.11290512291147677, + "learning_rate": 0.0001989659696283673, + "loss": 0.7446, + "step": 833 + }, + { + "epoch": 0.07457081545064377, + "grad_norm": 0.10981087527798433, + "learning_rate": 0.00019896181157635923, + "loss": 0.6946, + "step": 834 + }, + { + "epoch": 0.07466022889842633, + "grad_norm": 0.11164282623803001, + "learning_rate": 0.0001989576452245435, + "loss": 0.7013, + "step": 835 + }, + { + "epoch": 0.07474964234620887, + "grad_norm": 0.11855649381048719, + "learning_rate": 0.00019895347057326962, + "loss": 0.6768, + "step": 836 + }, + { + "epoch": 0.07483905579399142, + "grad_norm": 0.1265480259427198, + "learning_rate": 0.00019894928762288766, + "loss": 0.7163, + "step": 837 + }, + { + "epoch": 0.07492846924177396, + "grad_norm": 0.10758260255538375, + "learning_rate": 0.00019894509637374843, + "loss": 0.6832, + "step": 838 + }, + { + "epoch": 0.0750178826895565, + "grad_norm": 0.10629387388398602, + "learning_rate": 0.00019894089682620349, + "loss": 0.7151, + "step": 839 + }, + { + "epoch": 0.07510729613733906, + "grad_norm": 0.10702838166901853, + "learning_rate": 0.00019893668898060502, + "loss": 0.6846, + "step": 840 + }, + { + "epoch": 0.0751967095851216, + "grad_norm": 0.13937843338361441, + "learning_rate": 0.00019893247283730593, + "loss": 0.7303, + "step": 841 + }, + { + "epoch": 0.07528612303290415, + "grad_norm": 0.11267981924382499, + "learning_rate": 0.0001989282483966598, + "loss": 0.7072, + "step": 842 + }, + { + "epoch": 0.07537553648068669, + "grad_norm": 0.13032361684505453, + "learning_rate": 0.00019892401565902096, + "loss": 0.7217, + "step": 843 + }, + { + "epoch": 0.07546494992846924, + "grad_norm": 0.12105657683208099, + "learning_rate": 0.0001989197746247444, + "loss": 0.6951, + "step": 844 + }, + { + "epoch": 0.07555436337625179, + "grad_norm": 0.09349207450355397, + "learning_rate": 0.0001989155252941858, + "loss": 0.2662, + "step": 845 + }, + { + "epoch": 0.07564377682403434, + "grad_norm": 0.11685420867998601, + "learning_rate": 0.00019891126766770158, + "loss": 0.7087, + "step": 846 + }, + { + "epoch": 0.07573319027181688, + "grad_norm": 0.10308562177789556, + "learning_rate": 0.00019890700174564878, + "loss": 0.6709, + "step": 847 + }, + { + "epoch": 0.07582260371959942, + "grad_norm": 0.12074636839738866, + "learning_rate": 0.00019890272752838518, + "loss": 0.68, + "step": 848 + }, + { + "epoch": 0.07591201716738197, + "grad_norm": 0.11686524529634439, + "learning_rate": 0.00019889844501626928, + "loss": 0.6859, + "step": 849 + }, + { + "epoch": 0.07600143061516453, + "grad_norm": 0.11304390139038306, + "learning_rate": 0.00019889415420966026, + "loss": 0.7102, + "step": 850 + }, + { + "epoch": 0.07609084406294707, + "grad_norm": 0.11572231394735091, + "learning_rate": 0.00019888985510891792, + "loss": 0.7093, + "step": 851 + }, + { + "epoch": 0.07618025751072961, + "grad_norm": 0.10610832902761848, + "learning_rate": 0.00019888554771440288, + "loss": 0.7182, + "step": 852 + }, + { + "epoch": 0.07626967095851216, + "grad_norm": 0.10471481367869077, + "learning_rate": 0.00019888123202647636, + "loss": 0.6779, + "step": 853 + }, + { + "epoch": 0.0763590844062947, + "grad_norm": 0.12790113873420467, + "learning_rate": 0.00019887690804550035, + "loss": 0.7231, + "step": 854 + }, + { + "epoch": 0.07644849785407726, + "grad_norm": 0.1152680190878884, + "learning_rate": 0.00019887257577183744, + "loss": 0.6925, + "step": 855 + }, + { + "epoch": 0.0765379113018598, + "grad_norm": 0.11368643650389873, + "learning_rate": 0.00019886823520585105, + "loss": 0.6789, + "step": 856 + }, + { + "epoch": 0.07662732474964234, + "grad_norm": 0.10418181977323843, + "learning_rate": 0.00019886388634790517, + "loss": 0.7255, + "step": 857 + }, + { + "epoch": 0.07671673819742489, + "grad_norm": 0.10688967432826452, + "learning_rate": 0.0001988595291983645, + "loss": 0.6952, + "step": 858 + }, + { + "epoch": 0.07680615164520745, + "grad_norm": 0.1023693007920791, + "learning_rate": 0.00019885516375759457, + "loss": 0.7076, + "step": 859 + }, + { + "epoch": 0.07689556509298999, + "grad_norm": 0.11504038844981804, + "learning_rate": 0.00019885079002596138, + "loss": 0.7032, + "step": 860 + }, + { + "epoch": 0.07698497854077253, + "grad_norm": 0.11807855908531183, + "learning_rate": 0.00019884640800383186, + "loss": 0.7142, + "step": 861 + }, + { + "epoch": 0.07707439198855508, + "grad_norm": 0.12022282104609323, + "learning_rate": 0.00019884201769157346, + "loss": 0.7027, + "step": 862 + }, + { + "epoch": 0.07716380543633762, + "grad_norm": 0.10471588482916895, + "learning_rate": 0.0001988376190895544, + "loss": 0.7063, + "step": 863 + }, + { + "epoch": 0.07725321888412018, + "grad_norm": 0.09398079284492097, + "learning_rate": 0.0001988332121981436, + "loss": 0.6661, + "step": 864 + }, + { + "epoch": 0.07734263233190272, + "grad_norm": 0.09777100032453398, + "learning_rate": 0.00019882879701771063, + "loss": 0.7161, + "step": 865 + }, + { + "epoch": 0.07743204577968527, + "grad_norm": 0.11460453872192398, + "learning_rate": 0.00019882437354862585, + "loss": 0.7125, + "step": 866 + }, + { + "epoch": 0.07752145922746781, + "grad_norm": 0.09714925457774835, + "learning_rate": 0.00019881994179126017, + "loss": 0.675, + "step": 867 + }, + { + "epoch": 0.07761087267525035, + "grad_norm": 0.13729794031781195, + "learning_rate": 0.00019881550174598536, + "loss": 0.7359, + "step": 868 + }, + { + "epoch": 0.07770028612303291, + "grad_norm": 0.11861506433720831, + "learning_rate": 0.00019881105341317372, + "loss": 0.7173, + "step": 869 + }, + { + "epoch": 0.07778969957081545, + "grad_norm": 0.12967563939657376, + "learning_rate": 0.00019880659679319838, + "loss": 0.7476, + "step": 870 + }, + { + "epoch": 0.077879113018598, + "grad_norm": 0.13899966759703075, + "learning_rate": 0.00019880213188643307, + "loss": 0.7167, + "step": 871 + }, + { + "epoch": 0.07796852646638054, + "grad_norm": 0.13589215526518905, + "learning_rate": 0.00019879765869325233, + "loss": 0.6955, + "step": 872 + }, + { + "epoch": 0.07805793991416309, + "grad_norm": 0.11580825923509534, + "learning_rate": 0.00019879317721403124, + "loss": 0.6907, + "step": 873 + }, + { + "epoch": 0.07814735336194564, + "grad_norm": 0.10583394425165109, + "learning_rate": 0.00019878868744914569, + "loss": 0.702, + "step": 874 + }, + { + "epoch": 0.07823676680972819, + "grad_norm": 0.11288229098628381, + "learning_rate": 0.00019878418939897223, + "loss": 0.7205, + "step": 875 + }, + { + "epoch": 0.07832618025751073, + "grad_norm": 0.1401474603126658, + "learning_rate": 0.00019877968306388811, + "loss": 0.6826, + "step": 876 + }, + { + "epoch": 0.07841559370529327, + "grad_norm": 0.10165143840752397, + "learning_rate": 0.00019877516844427127, + "loss": 0.718, + "step": 877 + }, + { + "epoch": 0.07850500715307582, + "grad_norm": 0.1176308984123481, + "learning_rate": 0.00019877064554050036, + "loss": 0.6912, + "step": 878 + }, + { + "epoch": 0.07859442060085838, + "grad_norm": 0.11372402607571987, + "learning_rate": 0.00019876611435295466, + "loss": 0.7207, + "step": 879 + }, + { + "epoch": 0.07868383404864092, + "grad_norm": 0.12009234792274971, + "learning_rate": 0.00019876157488201424, + "loss": 0.7075, + "step": 880 + }, + { + "epoch": 0.07877324749642346, + "grad_norm": 0.10920569796892675, + "learning_rate": 0.00019875702712805984, + "loss": 0.7099, + "step": 881 + }, + { + "epoch": 0.078862660944206, + "grad_norm": 0.11127152536231488, + "learning_rate": 0.00019875247109147278, + "loss": 0.741, + "step": 882 + }, + { + "epoch": 0.07895207439198855, + "grad_norm": 0.12653438858504415, + "learning_rate": 0.0001987479067726353, + "loss": 0.7362, + "step": 883 + }, + { + "epoch": 0.07904148783977111, + "grad_norm": 0.12818638051667863, + "learning_rate": 0.00019874333417193007, + "loss": 0.708, + "step": 884 + }, + { + "epoch": 0.07913090128755365, + "grad_norm": 0.1156142311983435, + "learning_rate": 0.00019873875328974073, + "loss": 0.6852, + "step": 885 + }, + { + "epoch": 0.0792203147353362, + "grad_norm": 0.11417611984804105, + "learning_rate": 0.00019873416412645133, + "loss": 0.7047, + "step": 886 + }, + { + "epoch": 0.07930972818311874, + "grad_norm": 0.10895156416957248, + "learning_rate": 0.00019872956668244687, + "loss": 0.693, + "step": 887 + }, + { + "epoch": 0.07939914163090128, + "grad_norm": 0.11173278187992348, + "learning_rate": 0.00019872496095811286, + "loss": 0.7033, + "step": 888 + }, + { + "epoch": 0.07948855507868384, + "grad_norm": 0.12112519958412021, + "learning_rate": 0.00019872034695383558, + "loss": 0.7004, + "step": 889 + }, + { + "epoch": 0.07957796852646638, + "grad_norm": 0.1076057458360151, + "learning_rate": 0.0001987157246700021, + "loss": 0.6776, + "step": 890 + }, + { + "epoch": 0.07966738197424893, + "grad_norm": 0.12783784843583254, + "learning_rate": 0.00019871109410699996, + "loss": 0.6882, + "step": 891 + }, + { + "epoch": 0.07975679542203147, + "grad_norm": 0.11248699016661662, + "learning_rate": 0.00019870645526521758, + "loss": 0.6804, + "step": 892 + }, + { + "epoch": 0.07984620886981401, + "grad_norm": 0.11361613496901396, + "learning_rate": 0.000198701808145044, + "loss": 0.7092, + "step": 893 + }, + { + "epoch": 0.07993562231759657, + "grad_norm": 0.12614189189569888, + "learning_rate": 0.00019869715274686898, + "loss": 0.7192, + "step": 894 + }, + { + "epoch": 0.08002503576537912, + "grad_norm": 0.12757896551728806, + "learning_rate": 0.00019869248907108294, + "loss": 0.775, + "step": 895 + }, + { + "epoch": 0.08011444921316166, + "grad_norm": 0.10961591214010595, + "learning_rate": 0.00019868781711807705, + "loss": 0.7181, + "step": 896 + }, + { + "epoch": 0.0802038626609442, + "grad_norm": 0.1165032385875809, + "learning_rate": 0.0001986831368882431, + "loss": 0.7469, + "step": 897 + }, + { + "epoch": 0.08029327610872675, + "grad_norm": 0.11596310674590371, + "learning_rate": 0.00019867844838197365, + "loss": 0.732, + "step": 898 + }, + { + "epoch": 0.0803826895565093, + "grad_norm": 0.12505293604234227, + "learning_rate": 0.00019867375159966192, + "loss": 0.7059, + "step": 899 + }, + { + "epoch": 0.08047210300429185, + "grad_norm": 0.10564110899264803, + "learning_rate": 0.00019866904654170179, + "loss": 0.6741, + "step": 900 + }, + { + "epoch": 0.08056151645207439, + "grad_norm": 0.12216908409276865, + "learning_rate": 0.0001986643332084879, + "loss": 0.7228, + "step": 901 + }, + { + "epoch": 0.08065092989985694, + "grad_norm": 0.12630308119891576, + "learning_rate": 0.00019865961160041556, + "loss": 0.7154, + "step": 902 + }, + { + "epoch": 0.08074034334763948, + "grad_norm": 0.12839574137191942, + "learning_rate": 0.0001986548817178807, + "loss": 0.6797, + "step": 903 + }, + { + "epoch": 0.08082975679542204, + "grad_norm": 0.1348932276121236, + "learning_rate": 0.0001986501435612801, + "loss": 0.7446, + "step": 904 + }, + { + "epoch": 0.08091917024320458, + "grad_norm": 0.12568183162105914, + "learning_rate": 0.00019864539713101108, + "loss": 0.7333, + "step": 905 + }, + { + "epoch": 0.08100858369098712, + "grad_norm": 0.12889952286254808, + "learning_rate": 0.00019864064242747174, + "loss": 0.6939, + "step": 906 + }, + { + "epoch": 0.08109799713876967, + "grad_norm": 0.11760748884241455, + "learning_rate": 0.00019863587945106084, + "loss": 0.7203, + "step": 907 + }, + { + "epoch": 0.08118741058655221, + "grad_norm": 0.14054610211139637, + "learning_rate": 0.00019863110820217785, + "loss": 0.6939, + "step": 908 + }, + { + "epoch": 0.08127682403433477, + "grad_norm": 0.13778011718527491, + "learning_rate": 0.00019862632868122292, + "loss": 0.703, + "step": 909 + }, + { + "epoch": 0.08136623748211731, + "grad_norm": 0.099102231545081, + "learning_rate": 0.00019862154088859697, + "loss": 0.6954, + "step": 910 + }, + { + "epoch": 0.08145565092989986, + "grad_norm": 0.1195369226645288, + "learning_rate": 0.00019861674482470144, + "loss": 0.7081, + "step": 911 + }, + { + "epoch": 0.0815450643776824, + "grad_norm": 0.12410415780863396, + "learning_rate": 0.00019861194048993863, + "loss": 0.6923, + "step": 912 + }, + { + "epoch": 0.08163447782546494, + "grad_norm": 0.10949099699336698, + "learning_rate": 0.00019860712788471148, + "loss": 0.6835, + "step": 913 + }, + { + "epoch": 0.0817238912732475, + "grad_norm": 0.10743196664273957, + "learning_rate": 0.00019860230700942356, + "loss": 0.7576, + "step": 914 + }, + { + "epoch": 0.08181330472103004, + "grad_norm": 0.12271803656931238, + "learning_rate": 0.00019859747786447928, + "loss": 0.7316, + "step": 915 + }, + { + "epoch": 0.08190271816881259, + "grad_norm": 0.12520073468663856, + "learning_rate": 0.00019859264045028358, + "loss": 0.7544, + "step": 916 + }, + { + "epoch": 0.08199213161659513, + "grad_norm": 0.11598931392102207, + "learning_rate": 0.00019858779476724219, + "loss": 0.7039, + "step": 917 + }, + { + "epoch": 0.08208154506437768, + "grad_norm": 0.1011808577360062, + "learning_rate": 0.00019858294081576155, + "loss": 0.7199, + "step": 918 + }, + { + "epoch": 0.08217095851216023, + "grad_norm": 0.11220766762656365, + "learning_rate": 0.00019857807859624869, + "loss": 0.6988, + "step": 919 + }, + { + "epoch": 0.08226037195994278, + "grad_norm": 0.11744083202604934, + "learning_rate": 0.00019857320810911144, + "loss": 0.6828, + "step": 920 + }, + { + "epoch": 0.08234978540772532, + "grad_norm": 0.12601592493779687, + "learning_rate": 0.00019856832935475827, + "loss": 0.6894, + "step": 921 + }, + { + "epoch": 0.08243919885550786, + "grad_norm": 0.13542024303175354, + "learning_rate": 0.00019856344233359837, + "loss": 0.7248, + "step": 922 + }, + { + "epoch": 0.08252861230329042, + "grad_norm": 0.1145716709343715, + "learning_rate": 0.0001985585470460416, + "loss": 0.7032, + "step": 923 + }, + { + "epoch": 0.08261802575107297, + "grad_norm": 0.11469914743706378, + "learning_rate": 0.00019855364349249848, + "loss": 0.7445, + "step": 924 + }, + { + "epoch": 0.08270743919885551, + "grad_norm": 0.11134270041756354, + "learning_rate": 0.00019854873167338033, + "loss": 0.7403, + "step": 925 + }, + { + "epoch": 0.08279685264663805, + "grad_norm": 0.11546849978273574, + "learning_rate": 0.0001985438115890991, + "loss": 0.7303, + "step": 926 + }, + { + "epoch": 0.0828862660944206, + "grad_norm": 0.11226907885482024, + "learning_rate": 0.00019853888324006735, + "loss": 0.7366, + "step": 927 + }, + { + "epoch": 0.08297567954220315, + "grad_norm": 0.11687516383844204, + "learning_rate": 0.00019853394662669847, + "loss": 0.7149, + "step": 928 + }, + { + "epoch": 0.0830650929899857, + "grad_norm": 0.11832572205049056, + "learning_rate": 0.00019852900174940655, + "loss": 0.727, + "step": 929 + }, + { + "epoch": 0.08315450643776824, + "grad_norm": 0.1058176855750486, + "learning_rate": 0.00019852404860860618, + "loss": 0.7263, + "step": 930 + }, + { + "epoch": 0.08324391988555079, + "grad_norm": 0.10443619553572385, + "learning_rate": 0.00019851908720471285, + "loss": 0.6447, + "step": 931 + }, + { + "epoch": 0.08333333333333333, + "grad_norm": 0.12598394584895953, + "learning_rate": 0.0001985141175381427, + "loss": 0.6604, + "step": 932 + }, + { + "epoch": 0.08342274678111589, + "grad_norm": 0.12497731641085785, + "learning_rate": 0.00019850913960931243, + "loss": 0.7104, + "step": 933 + }, + { + "epoch": 0.08351216022889843, + "grad_norm": 0.11057595742019644, + "learning_rate": 0.00019850415341863962, + "loss": 0.6787, + "step": 934 + }, + { + "epoch": 0.08360157367668097, + "grad_norm": 0.10904385915765015, + "learning_rate": 0.00019849915896654242, + "loss": 0.6692, + "step": 935 + }, + { + "epoch": 0.08369098712446352, + "grad_norm": 0.11178677273381916, + "learning_rate": 0.0001984941562534397, + "loss": 0.6976, + "step": 936 + }, + { + "epoch": 0.08378040057224606, + "grad_norm": 0.13758027743953313, + "learning_rate": 0.00019848914527975108, + "loss": 0.6935, + "step": 937 + }, + { + "epoch": 0.08386981402002862, + "grad_norm": 0.12410328591325683, + "learning_rate": 0.00019848412604589678, + "loss": 0.7033, + "step": 938 + }, + { + "epoch": 0.08395922746781116, + "grad_norm": 0.12492653379531807, + "learning_rate": 0.00019847909855229775, + "loss": 0.7532, + "step": 939 + }, + { + "epoch": 0.0840486409155937, + "grad_norm": 0.10868361376771968, + "learning_rate": 0.00019847406279937567, + "loss": 0.2795, + "step": 940 + }, + { + "epoch": 0.08413805436337625, + "grad_norm": 0.1196142208766059, + "learning_rate": 0.00019846901878755287, + "loss": 0.6826, + "step": 941 + }, + { + "epoch": 0.0842274678111588, + "grad_norm": 0.12493014494537076, + "learning_rate": 0.00019846396651725237, + "loss": 0.7193, + "step": 942 + }, + { + "epoch": 0.08431688125894135, + "grad_norm": 0.12109241060970075, + "learning_rate": 0.0001984589059888979, + "loss": 0.6933, + "step": 943 + }, + { + "epoch": 0.0844062947067239, + "grad_norm": 0.11782646196217317, + "learning_rate": 0.00019845383720291392, + "loss": 0.7533, + "step": 944 + }, + { + "epoch": 0.08449570815450644, + "grad_norm": 0.10996788900415913, + "learning_rate": 0.00019844876015972552, + "loss": 0.7068, + "step": 945 + }, + { + "epoch": 0.08458512160228898, + "grad_norm": 0.10202778064623141, + "learning_rate": 0.0001984436748597585, + "loss": 0.7037, + "step": 946 + }, + { + "epoch": 0.08467453505007153, + "grad_norm": 0.1212986247528106, + "learning_rate": 0.00019843858130343933, + "loss": 0.6837, + "step": 947 + }, + { + "epoch": 0.08476394849785408, + "grad_norm": 0.11918174388801858, + "learning_rate": 0.00019843347949119526, + "loss": 0.7205, + "step": 948 + }, + { + "epoch": 0.08485336194563663, + "grad_norm": 0.12293098285424638, + "learning_rate": 0.00019842836942345415, + "loss": 0.6996, + "step": 949 + }, + { + "epoch": 0.08494277539341917, + "grad_norm": 0.1180667280231936, + "learning_rate": 0.00019842325110064454, + "loss": 0.7108, + "step": 950 + }, + { + "epoch": 0.08503218884120171, + "grad_norm": 0.11935421255033239, + "learning_rate": 0.00019841812452319575, + "loss": 0.6734, + "step": 951 + }, + { + "epoch": 0.08512160228898426, + "grad_norm": 0.12071026853783963, + "learning_rate": 0.0001984129896915377, + "loss": 0.7633, + "step": 952 + }, + { + "epoch": 0.08521101573676682, + "grad_norm": 0.12185738186215983, + "learning_rate": 0.00019840784660610106, + "loss": 0.676, + "step": 953 + }, + { + "epoch": 0.08530042918454936, + "grad_norm": 0.11141577989534279, + "learning_rate": 0.00019840269526731716, + "loss": 0.6815, + "step": 954 + }, + { + "epoch": 0.0853898426323319, + "grad_norm": 0.11767507323493905, + "learning_rate": 0.00019839753567561807, + "loss": 0.6943, + "step": 955 + }, + { + "epoch": 0.08547925608011445, + "grad_norm": 0.1225120649802503, + "learning_rate": 0.0001983923678314365, + "loss": 0.699, + "step": 956 + }, + { + "epoch": 0.08556866952789699, + "grad_norm": 0.11538037044594443, + "learning_rate": 0.00019838719173520585, + "loss": 0.7254, + "step": 957 + }, + { + "epoch": 0.08565808297567955, + "grad_norm": 0.13193851209146074, + "learning_rate": 0.00019838200738736027, + "loss": 0.7396, + "step": 958 + }, + { + "epoch": 0.08574749642346209, + "grad_norm": 0.11964853139558357, + "learning_rate": 0.0001983768147883345, + "loss": 0.7051, + "step": 959 + }, + { + "epoch": 0.08583690987124463, + "grad_norm": 0.1167936282143552, + "learning_rate": 0.0001983716139385641, + "loss": 0.7169, + "step": 960 + }, + { + "epoch": 0.08592632331902718, + "grad_norm": 0.11790374460168355, + "learning_rate": 0.00019836640483848528, + "loss": 0.7174, + "step": 961 + }, + { + "epoch": 0.08601573676680972, + "grad_norm": 0.123860977934636, + "learning_rate": 0.00019836118748853485, + "loss": 0.663, + "step": 962 + }, + { + "epoch": 0.08610515021459228, + "grad_norm": 0.12950181927249316, + "learning_rate": 0.00019835596188915044, + "loss": 0.7249, + "step": 963 + }, + { + "epoch": 0.08619456366237482, + "grad_norm": 0.11104500735058027, + "learning_rate": 0.00019835072804077027, + "loss": 0.6772, + "step": 964 + }, + { + "epoch": 0.08628397711015737, + "grad_norm": 0.1056195002360987, + "learning_rate": 0.00019834548594383332, + "loss": 0.6838, + "step": 965 + }, + { + "epoch": 0.08637339055793991, + "grad_norm": 0.1122358904451552, + "learning_rate": 0.00019834023559877923, + "loss": 0.6993, + "step": 966 + }, + { + "epoch": 0.08646280400572245, + "grad_norm": 0.10434652821505935, + "learning_rate": 0.00019833497700604835, + "loss": 0.677, + "step": 967 + }, + { + "epoch": 0.08655221745350501, + "grad_norm": 0.1016310818390034, + "learning_rate": 0.0001983297101660817, + "loss": 0.6826, + "step": 968 + }, + { + "epoch": 0.08664163090128756, + "grad_norm": 0.11132205528528374, + "learning_rate": 0.00019832443507932103, + "loss": 0.691, + "step": 969 + }, + { + "epoch": 0.0867310443490701, + "grad_norm": 0.1123740695348612, + "learning_rate": 0.00019831915174620872, + "loss": 0.665, + "step": 970 + }, + { + "epoch": 0.08682045779685264, + "grad_norm": 0.12034151666829344, + "learning_rate": 0.0001983138601671879, + "loss": 0.7379, + "step": 971 + }, + { + "epoch": 0.08690987124463519, + "grad_norm": 0.1024779740978403, + "learning_rate": 0.00019830856034270235, + "loss": 0.7215, + "step": 972 + }, + { + "epoch": 0.08699928469241774, + "grad_norm": 0.11182557441202441, + "learning_rate": 0.0001983032522731966, + "loss": 0.6579, + "step": 973 + }, + { + "epoch": 0.08708869814020029, + "grad_norm": 0.10613470619965619, + "learning_rate": 0.00019829793595911577, + "loss": 0.6747, + "step": 974 + }, + { + "epoch": 0.08717811158798283, + "grad_norm": 0.11120323148888717, + "learning_rate": 0.0001982926114009058, + "loss": 0.7091, + "step": 975 + }, + { + "epoch": 0.08726752503576538, + "grad_norm": 0.11214153446000909, + "learning_rate": 0.00019828727859901317, + "loss": 0.7029, + "step": 976 + }, + { + "epoch": 0.08735693848354792, + "grad_norm": 0.10440221113735543, + "learning_rate": 0.00019828193755388522, + "loss": 0.6623, + "step": 977 + }, + { + "epoch": 0.08744635193133048, + "grad_norm": 0.22372033806108144, + "learning_rate": 0.00019827658826596984, + "loss": 0.2882, + "step": 978 + }, + { + "epoch": 0.08753576537911302, + "grad_norm": 0.11978801684582921, + "learning_rate": 0.00019827123073571572, + "loss": 0.6921, + "step": 979 + }, + { + "epoch": 0.08762517882689556, + "grad_norm": 0.10892283103687446, + "learning_rate": 0.00019826586496357216, + "loss": 0.6848, + "step": 980 + }, + { + "epoch": 0.08771459227467811, + "grad_norm": 0.11173902430930845, + "learning_rate": 0.00019826049094998912, + "loss": 0.6695, + "step": 981 + }, + { + "epoch": 0.08780400572246065, + "grad_norm": 0.11044581735413149, + "learning_rate": 0.00019825510869541743, + "loss": 0.7308, + "step": 982 + }, + { + "epoch": 0.08789341917024321, + "grad_norm": 0.1000911502939639, + "learning_rate": 0.00019824971820030842, + "loss": 0.6901, + "step": 983 + }, + { + "epoch": 0.08798283261802575, + "grad_norm": 0.11469204676154737, + "learning_rate": 0.0001982443194651142, + "loss": 0.6957, + "step": 984 + }, + { + "epoch": 0.0880722460658083, + "grad_norm": 0.12203360364337429, + "learning_rate": 0.00019823891249028756, + "loss": 0.7198, + "step": 985 + }, + { + "epoch": 0.08816165951359084, + "grad_norm": 0.12489075214156875, + "learning_rate": 0.00019823349727628197, + "loss": 0.6716, + "step": 986 + }, + { + "epoch": 0.0882510729613734, + "grad_norm": 0.11607248298655107, + "learning_rate": 0.00019822807382355163, + "loss": 0.6787, + "step": 987 + }, + { + "epoch": 0.08834048640915594, + "grad_norm": 0.11063494069410205, + "learning_rate": 0.0001982226421325513, + "loss": 0.6868, + "step": 988 + }, + { + "epoch": 0.08842989985693848, + "grad_norm": 0.14552686310554752, + "learning_rate": 0.00019821720220373665, + "loss": 0.3075, + "step": 989 + }, + { + "epoch": 0.08851931330472103, + "grad_norm": 0.11084922329393045, + "learning_rate": 0.00019821175403756384, + "loss": 0.6825, + "step": 990 + }, + { + "epoch": 0.08860872675250357, + "grad_norm": 0.11215874043537734, + "learning_rate": 0.00019820629763448987, + "loss": 0.712, + "step": 991 + }, + { + "epoch": 0.08869814020028613, + "grad_norm": 0.10916027920724235, + "learning_rate": 0.00019820083299497228, + "loss": 0.7127, + "step": 992 + }, + { + "epoch": 0.08878755364806867, + "grad_norm": 0.12314819316131503, + "learning_rate": 0.00019819536011946945, + "loss": 0.7142, + "step": 993 + }, + { + "epoch": 0.08887696709585122, + "grad_norm": 0.1154109793406363, + "learning_rate": 0.00019818987900844032, + "loss": 0.6977, + "step": 994 + }, + { + "epoch": 0.08896638054363376, + "grad_norm": 0.11682975759788719, + "learning_rate": 0.00019818438966234464, + "loss": 0.7151, + "step": 995 + }, + { + "epoch": 0.0890557939914163, + "grad_norm": 0.11414265972410793, + "learning_rate": 0.00019817889208164277, + "loss": 0.6918, + "step": 996 + }, + { + "epoch": 0.08914520743919886, + "grad_norm": 0.12472410370756887, + "learning_rate": 0.0001981733862667958, + "loss": 0.7065, + "step": 997 + }, + { + "epoch": 0.0892346208869814, + "grad_norm": 0.10384304127818812, + "learning_rate": 0.00019816787221826548, + "loss": 0.6876, + "step": 998 + }, + { + "epoch": 0.08932403433476395, + "grad_norm": 0.11971484649010079, + "learning_rate": 0.0001981623499365143, + "loss": 0.7199, + "step": 999 + }, + { + "epoch": 0.08941344778254649, + "grad_norm": 0.1034786806777283, + "learning_rate": 0.00019815681942200535, + "loss": 0.6628, + "step": 1000 + }, + { + "epoch": 0.08950286123032904, + "grad_norm": 0.11862103970254362, + "learning_rate": 0.00019815128067520252, + "loss": 0.7034, + "step": 1001 + }, + { + "epoch": 0.0895922746781116, + "grad_norm": 0.13895487057828043, + "learning_rate": 0.0001981457336965703, + "loss": 0.7278, + "step": 1002 + }, + { + "epoch": 0.08968168812589414, + "grad_norm": 0.11563036673358958, + "learning_rate": 0.0001981401784865739, + "loss": 0.7083, + "step": 1003 + }, + { + "epoch": 0.08977110157367668, + "grad_norm": 0.12010633168431613, + "learning_rate": 0.00019813461504567933, + "loss": 0.6735, + "step": 1004 + }, + { + "epoch": 0.08986051502145923, + "grad_norm": 0.1269263266814388, + "learning_rate": 0.00019812904337435306, + "loss": 0.7396, + "step": 1005 + }, + { + "epoch": 0.08994992846924177, + "grad_norm": 0.11094564707078479, + "learning_rate": 0.00019812346347306242, + "loss": 0.6879, + "step": 1006 + }, + { + "epoch": 0.09003934191702433, + "grad_norm": 0.13090855137639734, + "learning_rate": 0.00019811787534227543, + "loss": 0.7089, + "step": 1007 + }, + { + "epoch": 0.09012875536480687, + "grad_norm": 0.1165300612279197, + "learning_rate": 0.0001981122789824607, + "loss": 0.724, + "step": 1008 + }, + { + "epoch": 0.09021816881258941, + "grad_norm": 0.1193162018970232, + "learning_rate": 0.00019810667439408767, + "loss": 0.6955, + "step": 1009 + }, + { + "epoch": 0.09030758226037196, + "grad_norm": 0.11967980974851998, + "learning_rate": 0.0001981010615776263, + "loss": 0.706, + "step": 1010 + }, + { + "epoch": 0.0903969957081545, + "grad_norm": 0.105886230158505, + "learning_rate": 0.00019809544053354738, + "loss": 0.6974, + "step": 1011 + }, + { + "epoch": 0.09048640915593706, + "grad_norm": 0.11079978522626091, + "learning_rate": 0.00019808981126232236, + "loss": 0.729, + "step": 1012 + }, + { + "epoch": 0.0905758226037196, + "grad_norm": 0.11672410331336246, + "learning_rate": 0.0001980841737644233, + "loss": 0.7011, + "step": 1013 + }, + { + "epoch": 0.09066523605150215, + "grad_norm": 0.12172898344395326, + "learning_rate": 0.00019807852804032305, + "loss": 0.7188, + "step": 1014 + }, + { + "epoch": 0.09075464949928469, + "grad_norm": 0.13375112068803205, + "learning_rate": 0.00019807287409049512, + "loss": 0.7043, + "step": 1015 + }, + { + "epoch": 0.09084406294706723, + "grad_norm": 0.09971918903566983, + "learning_rate": 0.00019806721191541367, + "loss": 0.6861, + "step": 1016 + }, + { + "epoch": 0.09093347639484979, + "grad_norm": 0.1321256814304062, + "learning_rate": 0.00019806154151555356, + "loss": 0.6951, + "step": 1017 + }, + { + "epoch": 0.09102288984263233, + "grad_norm": 0.11187151218596961, + "learning_rate": 0.00019805586289139043, + "loss": 0.7112, + "step": 1018 + }, + { + "epoch": 0.09111230329041488, + "grad_norm": 0.12376416714010513, + "learning_rate": 0.0001980501760434005, + "loss": 0.7201, + "step": 1019 + }, + { + "epoch": 0.09120171673819742, + "grad_norm": 0.12459377579265576, + "learning_rate": 0.00019804448097206068, + "loss": 0.7352, + "step": 1020 + }, + { + "epoch": 0.09129113018597997, + "grad_norm": 0.13560654667672972, + "learning_rate": 0.0001980387776778487, + "loss": 0.726, + "step": 1021 + }, + { + "epoch": 0.09138054363376252, + "grad_norm": 0.11252706797344579, + "learning_rate": 0.00019803306616124282, + "loss": 0.7207, + "step": 1022 + }, + { + "epoch": 0.09146995708154507, + "grad_norm": 0.1088288076481225, + "learning_rate": 0.00019802734642272206, + "loss": 0.7526, + "step": 1023 + }, + { + "epoch": 0.09155937052932761, + "grad_norm": 0.12025359022305346, + "learning_rate": 0.00019802161846276615, + "loss": 0.7261, + "step": 1024 + }, + { + "epoch": 0.09164878397711015, + "grad_norm": 0.12825899228388685, + "learning_rate": 0.0001980158822818555, + "loss": 0.6904, + "step": 1025 + }, + { + "epoch": 0.0917381974248927, + "grad_norm": 0.12370693416294953, + "learning_rate": 0.00019801013788047116, + "loss": 0.7449, + "step": 1026 + }, + { + "epoch": 0.09182761087267526, + "grad_norm": 0.13287727709678795, + "learning_rate": 0.00019800438525909491, + "loss": 0.7467, + "step": 1027 + }, + { + "epoch": 0.0919170243204578, + "grad_norm": 0.12204493380362601, + "learning_rate": 0.00019799862441820923, + "loss": 0.7222, + "step": 1028 + }, + { + "epoch": 0.09200643776824034, + "grad_norm": 0.1338376150033401, + "learning_rate": 0.0001979928553582973, + "loss": 0.7108, + "step": 1029 + }, + { + "epoch": 0.09209585121602289, + "grad_norm": 0.12251371555667952, + "learning_rate": 0.0001979870780798429, + "loss": 0.7101, + "step": 1030 + }, + { + "epoch": 0.09218526466380543, + "grad_norm": 0.12748922571882482, + "learning_rate": 0.00019798129258333065, + "loss": 0.7398, + "step": 1031 + }, + { + "epoch": 0.09227467811158799, + "grad_norm": 0.1597934685919047, + "learning_rate": 0.00019797549886924566, + "loss": 0.69, + "step": 1032 + }, + { + "epoch": 0.09236409155937053, + "grad_norm": 0.12372050468274952, + "learning_rate": 0.00019796969693807394, + "loss": 0.6863, + "step": 1033 + }, + { + "epoch": 0.09245350500715308, + "grad_norm": 0.10044384227813216, + "learning_rate": 0.00019796388679030205, + "loss": 0.6678, + "step": 1034 + }, + { + "epoch": 0.09254291845493562, + "grad_norm": 0.10711921767444298, + "learning_rate": 0.0001979580684264173, + "loss": 0.6786, + "step": 1035 + }, + { + "epoch": 0.09263233190271816, + "grad_norm": 0.11752145484756293, + "learning_rate": 0.00019795224184690764, + "loss": 0.6798, + "step": 1036 + }, + { + "epoch": 0.09272174535050072, + "grad_norm": 0.11354400661986086, + "learning_rate": 0.00019794640705226175, + "loss": 0.7004, + "step": 1037 + }, + { + "epoch": 0.09281115879828326, + "grad_norm": 0.1195576944568909, + "learning_rate": 0.00019794056404296898, + "loss": 0.715, + "step": 1038 + }, + { + "epoch": 0.09290057224606581, + "grad_norm": 0.10886777796514739, + "learning_rate": 0.0001979347128195194, + "loss": 0.7103, + "step": 1039 + }, + { + "epoch": 0.09298998569384835, + "grad_norm": 0.11402139111644959, + "learning_rate": 0.00019792885338240374, + "loss": 0.697, + "step": 1040 + }, + { + "epoch": 0.0930793991416309, + "grad_norm": 0.11516901161325745, + "learning_rate": 0.0001979229857321134, + "loss": 0.6804, + "step": 1041 + }, + { + "epoch": 0.09316881258941345, + "grad_norm": 0.10255598061917323, + "learning_rate": 0.00019791710986914051, + "loss": 0.6968, + "step": 1042 + }, + { + "epoch": 0.093258226037196, + "grad_norm": 0.11731096005281866, + "learning_rate": 0.00019791122579397789, + "loss": 0.7022, + "step": 1043 + }, + { + "epoch": 0.09334763948497854, + "grad_norm": 0.11856169954984853, + "learning_rate": 0.00019790533350711899, + "loss": 0.6891, + "step": 1044 + }, + { + "epoch": 0.09343705293276108, + "grad_norm": 0.1044248950219601, + "learning_rate": 0.000197899433009058, + "loss": 0.6813, + "step": 1045 + }, + { + "epoch": 0.09352646638054363, + "grad_norm": 0.12894205910684448, + "learning_rate": 0.0001978935243002898, + "loss": 0.7391, + "step": 1046 + }, + { + "epoch": 0.09361587982832618, + "grad_norm": 0.11660906975366912, + "learning_rate": 0.00019788760738130993, + "loss": 0.7236, + "step": 1047 + }, + { + "epoch": 0.09370529327610873, + "grad_norm": 0.10356774204970605, + "learning_rate": 0.00019788168225261469, + "loss": 0.69, + "step": 1048 + }, + { + "epoch": 0.09379470672389127, + "grad_norm": 0.11920040518227787, + "learning_rate": 0.00019787574891470095, + "loss": 0.7203, + "step": 1049 + }, + { + "epoch": 0.09388412017167382, + "grad_norm": 0.11624294260978775, + "learning_rate": 0.00019786980736806635, + "loss": 0.6835, + "step": 1050 + }, + { + "epoch": 0.09397353361945637, + "grad_norm": 0.1236838516308205, + "learning_rate": 0.0001978638576132092, + "loss": 0.7091, + "step": 1051 + }, + { + "epoch": 0.09406294706723892, + "grad_norm": 0.11613851860724816, + "learning_rate": 0.00019785789965062848, + "loss": 0.7457, + "step": 1052 + }, + { + "epoch": 0.09415236051502146, + "grad_norm": 0.11250189021644533, + "learning_rate": 0.00019785193348082394, + "loss": 0.7478, + "step": 1053 + }, + { + "epoch": 0.094241773962804, + "grad_norm": 0.11625505932806605, + "learning_rate": 0.0001978459591042959, + "loss": 0.7074, + "step": 1054 + }, + { + "epoch": 0.09433118741058655, + "grad_norm": 0.11133257536751832, + "learning_rate": 0.00019783997652154543, + "loss": 0.6894, + "step": 1055 + }, + { + "epoch": 0.0944206008583691, + "grad_norm": 0.1102519659309424, + "learning_rate": 0.00019783398573307428, + "loss": 0.6907, + "step": 1056 + }, + { + "epoch": 0.09451001430615165, + "grad_norm": 0.12594385927491294, + "learning_rate": 0.00019782798673938492, + "loss": 0.7082, + "step": 1057 + }, + { + "epoch": 0.09459942775393419, + "grad_norm": 0.11210927412438453, + "learning_rate": 0.00019782197954098046, + "loss": 0.7257, + "step": 1058 + }, + { + "epoch": 0.09468884120171674, + "grad_norm": 0.12283413963304286, + "learning_rate": 0.0001978159641383647, + "loss": 0.7397, + "step": 1059 + }, + { + "epoch": 0.09477825464949928, + "grad_norm": 0.11479707180953519, + "learning_rate": 0.00019780994053204216, + "loss": 0.7221, + "step": 1060 + }, + { + "epoch": 0.09486766809728184, + "grad_norm": 0.11764081265388891, + "learning_rate": 0.00019780390872251803, + "loss": 0.7044, + "step": 1061 + }, + { + "epoch": 0.09495708154506438, + "grad_norm": 0.11575045454125338, + "learning_rate": 0.00019779786871029819, + "loss": 0.2684, + "step": 1062 + }, + { + "epoch": 0.09504649499284692, + "grad_norm": 0.11557483172401833, + "learning_rate": 0.00019779182049588925, + "loss": 0.7086, + "step": 1063 + }, + { + "epoch": 0.09513590844062947, + "grad_norm": 0.12351287669011168, + "learning_rate": 0.0001977857640797984, + "loss": 0.6707, + "step": 1064 + }, + { + "epoch": 0.09522532188841201, + "grad_norm": 0.11022783910207709, + "learning_rate": 0.0001977796994625336, + "loss": 0.6989, + "step": 1065 + }, + { + "epoch": 0.09531473533619457, + "grad_norm": 0.11978517705266153, + "learning_rate": 0.0001977736266446035, + "loss": 0.724, + "step": 1066 + }, + { + "epoch": 0.09540414878397711, + "grad_norm": 0.109231354534863, + "learning_rate": 0.00019776754562651742, + "loss": 0.7054, + "step": 1067 + }, + { + "epoch": 0.09549356223175966, + "grad_norm": 0.11716888177982222, + "learning_rate": 0.00019776145640878538, + "loss": 0.6964, + "step": 1068 + }, + { + "epoch": 0.0955829756795422, + "grad_norm": 0.09569939757157649, + "learning_rate": 0.000197755358991918, + "loss": 0.6888, + "step": 1069 + }, + { + "epoch": 0.09567238912732474, + "grad_norm": 0.10811801258740004, + "learning_rate": 0.00019774925337642677, + "loss": 0.7193, + "step": 1070 + }, + { + "epoch": 0.0957618025751073, + "grad_norm": 0.1110942412343688, + "learning_rate": 0.0001977431395628237, + "loss": 0.7333, + "step": 1071 + }, + { + "epoch": 0.09585121602288985, + "grad_norm": 0.11772329957716476, + "learning_rate": 0.00019773701755162158, + "loss": 0.6846, + "step": 1072 + }, + { + "epoch": 0.09594062947067239, + "grad_norm": 0.11961180943316672, + "learning_rate": 0.0001977308873433338, + "loss": 0.6898, + "step": 1073 + }, + { + "epoch": 0.09603004291845493, + "grad_norm": 0.12058275177914522, + "learning_rate": 0.00019772474893847456, + "loss": 0.7317, + "step": 1074 + }, + { + "epoch": 0.09611945636623748, + "grad_norm": 0.12343941451877088, + "learning_rate": 0.00019771860233755862, + "loss": 0.7111, + "step": 1075 + }, + { + "epoch": 0.09620886981402003, + "grad_norm": 0.10604150378093655, + "learning_rate": 0.0001977124475411015, + "loss": 0.6759, + "step": 1076 + }, + { + "epoch": 0.09629828326180258, + "grad_norm": 0.1144018856871387, + "learning_rate": 0.00019770628454961946, + "loss": 0.6288, + "step": 1077 + }, + { + "epoch": 0.09638769670958512, + "grad_norm": 0.12042694293873107, + "learning_rate": 0.00019770011336362928, + "loss": 0.7006, + "step": 1078 + }, + { + "epoch": 0.09647711015736767, + "grad_norm": 0.11065520265159139, + "learning_rate": 0.00019769393398364865, + "loss": 0.7193, + "step": 1079 + }, + { + "epoch": 0.09656652360515021, + "grad_norm": 0.13093208803996592, + "learning_rate": 0.0001976877464101957, + "loss": 0.749, + "step": 1080 + }, + { + "epoch": 0.09665593705293277, + "grad_norm": 0.12140199291122508, + "learning_rate": 0.00019768155064378947, + "loss": 0.6835, + "step": 1081 + }, + { + "epoch": 0.09674535050071531, + "grad_norm": 0.12167187842362784, + "learning_rate": 0.00019767534668494954, + "loss": 0.7366, + "step": 1082 + }, + { + "epoch": 0.09683476394849785, + "grad_norm": 0.11266322560585844, + "learning_rate": 0.00019766913453419624, + "loss": 0.7056, + "step": 1083 + }, + { + "epoch": 0.0969241773962804, + "grad_norm": 0.11571736974204426, + "learning_rate": 0.0001976629141920506, + "loss": 0.6862, + "step": 1084 + }, + { + "epoch": 0.09701359084406294, + "grad_norm": 0.11573247504049684, + "learning_rate": 0.0001976566856590343, + "loss": 0.699, + "step": 1085 + }, + { + "epoch": 0.0971030042918455, + "grad_norm": 0.10727270125652823, + "learning_rate": 0.00019765044893566968, + "loss": 0.6688, + "step": 1086 + }, + { + "epoch": 0.09719241773962804, + "grad_norm": 0.09726466167232377, + "learning_rate": 0.00019764420402247987, + "loss": 0.6698, + "step": 1087 + }, + { + "epoch": 0.09728183118741059, + "grad_norm": 0.1323764026973944, + "learning_rate": 0.00019763795091998858, + "loss": 0.7179, + "step": 1088 + }, + { + "epoch": 0.09737124463519313, + "grad_norm": 0.11835360478985907, + "learning_rate": 0.00019763168962872026, + "loss": 0.2805, + "step": 1089 + }, + { + "epoch": 0.09746065808297567, + "grad_norm": 0.13103903085203983, + "learning_rate": 0.00019762542014920004, + "loss": 0.727, + "step": 1090 + }, + { + "epoch": 0.09755007153075823, + "grad_norm": 0.11335698526712235, + "learning_rate": 0.00019761914248195373, + "loss": 0.7147, + "step": 1091 + }, + { + "epoch": 0.09763948497854077, + "grad_norm": 0.10709138165510883, + "learning_rate": 0.00019761285662750787, + "loss": 0.6703, + "step": 1092 + }, + { + "epoch": 0.09772889842632332, + "grad_norm": 0.11609527689430603, + "learning_rate": 0.00019760656258638958, + "loss": 0.7137, + "step": 1093 + }, + { + "epoch": 0.09781831187410586, + "grad_norm": 0.1162747555240391, + "learning_rate": 0.0001976002603591268, + "loss": 0.7027, + "step": 1094 + }, + { + "epoch": 0.0979077253218884, + "grad_norm": 0.13096515554271987, + "learning_rate": 0.000197593949946248, + "loss": 0.7268, + "step": 1095 + }, + { + "epoch": 0.09799713876967096, + "grad_norm": 0.12022065386983642, + "learning_rate": 0.0001975876313482825, + "loss": 0.7042, + "step": 1096 + }, + { + "epoch": 0.09808655221745351, + "grad_norm": 0.12742226034920776, + "learning_rate": 0.00019758130456576023, + "loss": 0.7178, + "step": 1097 + }, + { + "epoch": 0.09817596566523605, + "grad_norm": 0.12753485571218628, + "learning_rate": 0.00019757496959921177, + "loss": 0.7153, + "step": 1098 + }, + { + "epoch": 0.0982653791130186, + "grad_norm": 0.12904216649912495, + "learning_rate": 0.00019756862644916846, + "loss": 0.7136, + "step": 1099 + }, + { + "epoch": 0.09835479256080114, + "grad_norm": 0.11343038649419486, + "learning_rate": 0.00019756227511616233, + "loss": 0.6715, + "step": 1100 + }, + { + "epoch": 0.0984442060085837, + "grad_norm": 0.11292078631601664, + "learning_rate": 0.00019755591560072596, + "loss": 0.7035, + "step": 1101 + }, + { + "epoch": 0.09853361945636624, + "grad_norm": 0.12601346490413912, + "learning_rate": 0.00019754954790339278, + "loss": 0.702, + "step": 1102 + }, + { + "epoch": 0.09862303290414878, + "grad_norm": 0.10615699445746007, + "learning_rate": 0.00019754317202469682, + "loss": 0.6624, + "step": 1103 + }, + { + "epoch": 0.09871244635193133, + "grad_norm": 0.11134662122971592, + "learning_rate": 0.00019753678796517282, + "loss": 0.7183, + "step": 1104 + }, + { + "epoch": 0.09880185979971387, + "grad_norm": 0.10685828667545864, + "learning_rate": 0.0001975303957253562, + "loss": 0.6815, + "step": 1105 + }, + { + "epoch": 0.09889127324749643, + "grad_norm": 0.12467020829718363, + "learning_rate": 0.00019752399530578312, + "loss": 0.7161, + "step": 1106 + }, + { + "epoch": 0.09898068669527897, + "grad_norm": 0.12837970109525793, + "learning_rate": 0.00019751758670699028, + "loss": 0.6993, + "step": 1107 + }, + { + "epoch": 0.09907010014306152, + "grad_norm": 0.11738476211363627, + "learning_rate": 0.00019751116992951527, + "loss": 0.6934, + "step": 1108 + }, + { + "epoch": 0.09915951359084406, + "grad_norm": 0.12016425013766559, + "learning_rate": 0.00019750474497389614, + "loss": 0.6975, + "step": 1109 + }, + { + "epoch": 0.0992489270386266, + "grad_norm": 0.1200521844013347, + "learning_rate": 0.00019749831184067185, + "loss": 0.6921, + "step": 1110 + }, + { + "epoch": 0.09933834048640916, + "grad_norm": 0.13934949293555657, + "learning_rate": 0.00019749187053038188, + "loss": 0.7451, + "step": 1111 + }, + { + "epoch": 0.0994277539341917, + "grad_norm": 0.12543073488199316, + "learning_rate": 0.00019748542104356648, + "loss": 0.7551, + "step": 1112 + }, + { + "epoch": 0.09951716738197425, + "grad_norm": 0.11368227183550267, + "learning_rate": 0.00019747896338076655, + "loss": 0.6862, + "step": 1113 + }, + { + "epoch": 0.09960658082975679, + "grad_norm": 0.10863217037112295, + "learning_rate": 0.00019747249754252367, + "loss": 0.707, + "step": 1114 + }, + { + "epoch": 0.09969599427753935, + "grad_norm": 0.11394912561079946, + "learning_rate": 0.00019746602352938014, + "loss": 0.7286, + "step": 1115 + }, + { + "epoch": 0.09978540772532189, + "grad_norm": 0.11972540682624132, + "learning_rate": 0.00019745954134187894, + "loss": 0.7073, + "step": 1116 + }, + { + "epoch": 0.09987482117310444, + "grad_norm": 0.12243023554104887, + "learning_rate": 0.00019745305098056368, + "loss": 0.7004, + "step": 1117 + }, + { + "epoch": 0.09996423462088698, + "grad_norm": 0.11921975125077842, + "learning_rate": 0.00019744655244597877, + "loss": 0.7267, + "step": 1118 + }, + { + "epoch": 0.10005364806866952, + "grad_norm": 0.12351083679287543, + "learning_rate": 0.00019744004573866915, + "loss": 0.7493, + "step": 1119 + }, + { + "epoch": 0.10014306151645208, + "grad_norm": 0.11199836364130246, + "learning_rate": 0.0001974335308591806, + "loss": 0.704, + "step": 1120 + }, + { + "epoch": 0.10023247496423462, + "grad_norm": 0.11436115667600712, + "learning_rate": 0.00019742700780805948, + "loss": 0.7254, + "step": 1121 + }, + { + "epoch": 0.10032188841201717, + "grad_norm": 0.10374792006283777, + "learning_rate": 0.00019742047658585286, + "loss": 0.6682, + "step": 1122 + }, + { + "epoch": 0.10041130185979971, + "grad_norm": 0.09511904019321307, + "learning_rate": 0.00019741393719310853, + "loss": 0.2522, + "step": 1123 + }, + { + "epoch": 0.10050071530758226, + "grad_norm": 0.12590492647683324, + "learning_rate": 0.00019740738963037495, + "loss": 0.6766, + "step": 1124 + }, + { + "epoch": 0.10059012875536481, + "grad_norm": 0.12984773615912676, + "learning_rate": 0.00019740083389820122, + "loss": 0.7205, + "step": 1125 + }, + { + "epoch": 0.10067954220314736, + "grad_norm": 0.13317182404081196, + "learning_rate": 0.0001973942699971372, + "loss": 0.6859, + "step": 1126 + }, + { + "epoch": 0.1007689556509299, + "grad_norm": 0.12736109936734338, + "learning_rate": 0.00019738769792773336, + "loss": 0.717, + "step": 1127 + }, + { + "epoch": 0.10085836909871244, + "grad_norm": 0.11033910140431222, + "learning_rate": 0.00019738111769054093, + "loss": 0.6958, + "step": 1128 + }, + { + "epoch": 0.10094778254649499, + "grad_norm": 0.11132471463984488, + "learning_rate": 0.00019737452928611176, + "loss": 0.6599, + "step": 1129 + }, + { + "epoch": 0.10103719599427755, + "grad_norm": 0.10008626123301406, + "learning_rate": 0.00019736793271499844, + "loss": 0.6843, + "step": 1130 + }, + { + "epoch": 0.10112660944206009, + "grad_norm": 0.12252856289604784, + "learning_rate": 0.00019736132797775415, + "loss": 0.6949, + "step": 1131 + }, + { + "epoch": 0.10121602288984263, + "grad_norm": 0.10786050099158463, + "learning_rate": 0.00019735471507493287, + "loss": 0.7221, + "step": 1132 + }, + { + "epoch": 0.10130543633762518, + "grad_norm": 0.11447056918207331, + "learning_rate": 0.00019734809400708922, + "loss": 0.6953, + "step": 1133 + }, + { + "epoch": 0.10139484978540772, + "grad_norm": 0.1085548431565893, + "learning_rate": 0.0001973414647747785, + "loss": 0.7442, + "step": 1134 + }, + { + "epoch": 0.10148426323319028, + "grad_norm": 0.11477328904168818, + "learning_rate": 0.0001973348273785567, + "loss": 0.7167, + "step": 1135 + }, + { + "epoch": 0.10157367668097282, + "grad_norm": 0.11235214254617756, + "learning_rate": 0.00019732818181898045, + "loss": 0.6988, + "step": 1136 + }, + { + "epoch": 0.10166309012875537, + "grad_norm": 0.11793533675897533, + "learning_rate": 0.00019732152809660716, + "loss": 0.7144, + "step": 1137 + }, + { + "epoch": 0.10175250357653791, + "grad_norm": 0.11866650533605111, + "learning_rate": 0.0001973148662119948, + "loss": 0.7166, + "step": 1138 + }, + { + "epoch": 0.10184191702432045, + "grad_norm": 0.10304239170450025, + "learning_rate": 0.0001973081961657022, + "loss": 0.7087, + "step": 1139 + }, + { + "epoch": 0.10193133047210301, + "grad_norm": 0.10050357182580368, + "learning_rate": 0.00019730151795828866, + "loss": 0.6634, + "step": 1140 + }, + { + "epoch": 0.10202074391988555, + "grad_norm": 0.11143012067557614, + "learning_rate": 0.0001972948315903143, + "loss": 0.6937, + "step": 1141 + }, + { + "epoch": 0.1021101573676681, + "grad_norm": 0.12571542799277655, + "learning_rate": 0.00019728813706233997, + "loss": 0.7132, + "step": 1142 + }, + { + "epoch": 0.10219957081545064, + "grad_norm": 0.12827107858640027, + "learning_rate": 0.00019728143437492706, + "loss": 0.7092, + "step": 1143 + }, + { + "epoch": 0.10228898426323318, + "grad_norm": 0.10543679812792303, + "learning_rate": 0.00019727472352863774, + "loss": 0.7095, + "step": 1144 + }, + { + "epoch": 0.10237839771101574, + "grad_norm": 0.10364743192910404, + "learning_rate": 0.00019726800452403483, + "loss": 0.6516, + "step": 1145 + }, + { + "epoch": 0.10246781115879829, + "grad_norm": 0.10942650078457908, + "learning_rate": 0.00019726127736168186, + "loss": 0.6709, + "step": 1146 + }, + { + "epoch": 0.10255722460658083, + "grad_norm": 0.1195135247662525, + "learning_rate": 0.000197254542042143, + "loss": 0.7077, + "step": 1147 + }, + { + "epoch": 0.10264663805436337, + "grad_norm": 0.12670708350201973, + "learning_rate": 0.00019724779856598317, + "loss": 0.7207, + "step": 1148 + }, + { + "epoch": 0.10273605150214592, + "grad_norm": 0.11283568625771598, + "learning_rate": 0.0001972410469337679, + "loss": 0.6928, + "step": 1149 + }, + { + "epoch": 0.10282546494992847, + "grad_norm": 0.09487975116404751, + "learning_rate": 0.00019723428714606348, + "loss": 0.7032, + "step": 1150 + }, + { + "epoch": 0.10291487839771102, + "grad_norm": 0.12400155419591148, + "learning_rate": 0.0001972275192034368, + "loss": 0.6848, + "step": 1151 + }, + { + "epoch": 0.10300429184549356, + "grad_norm": 0.12204505677011965, + "learning_rate": 0.00019722074310645553, + "loss": 0.7265, + "step": 1152 + }, + { + "epoch": 0.1030937052932761, + "grad_norm": 0.11447698210790597, + "learning_rate": 0.00019721395885568795, + "loss": 0.7086, + "step": 1153 + }, + { + "epoch": 0.10318311874105865, + "grad_norm": 0.1213724554653746, + "learning_rate": 0.00019720716645170303, + "loss": 0.7147, + "step": 1154 + }, + { + "epoch": 0.1032725321888412, + "grad_norm": 0.10448167183876085, + "learning_rate": 0.00019720036589507048, + "loss": 0.697, + "step": 1155 + }, + { + "epoch": 0.10336194563662375, + "grad_norm": 0.11214017495293666, + "learning_rate": 0.0001971935571863606, + "loss": 0.6529, + "step": 1156 + }, + { + "epoch": 0.1034513590844063, + "grad_norm": 0.10696516831902937, + "learning_rate": 0.00019718674032614448, + "loss": 0.733, + "step": 1157 + }, + { + "epoch": 0.10354077253218884, + "grad_norm": 0.10660809084215116, + "learning_rate": 0.0001971799153149938, + "loss": 0.6946, + "step": 1158 + }, + { + "epoch": 0.10363018597997138, + "grad_norm": 0.1232127448006562, + "learning_rate": 0.00019717308215348102, + "loss": 0.7213, + "step": 1159 + }, + { + "epoch": 0.10371959942775394, + "grad_norm": 0.11010129703550207, + "learning_rate": 0.00019716624084217918, + "loss": 0.6742, + "step": 1160 + }, + { + "epoch": 0.10380901287553648, + "grad_norm": 0.11331104759365129, + "learning_rate": 0.00019715939138166205, + "loss": 0.7049, + "step": 1161 + }, + { + "epoch": 0.10389842632331903, + "grad_norm": 0.11519045854586069, + "learning_rate": 0.00019715253377250411, + "loss": 0.6768, + "step": 1162 + }, + { + "epoch": 0.10398783977110157, + "grad_norm": 0.10491012175086208, + "learning_rate": 0.0001971456680152805, + "loss": 0.6867, + "step": 1163 + }, + { + "epoch": 0.10407725321888411, + "grad_norm": 0.10125078273369886, + "learning_rate": 0.00019713879411056704, + "loss": 0.6846, + "step": 1164 + }, + { + "epoch": 0.10416666666666667, + "grad_norm": 0.1181170804069778, + "learning_rate": 0.00019713191205894025, + "loss": 0.731, + "step": 1165 + }, + { + "epoch": 0.10425608011444921, + "grad_norm": 0.11001632795017643, + "learning_rate": 0.00019712502186097726, + "loss": 0.716, + "step": 1166 + }, + { + "epoch": 0.10434549356223176, + "grad_norm": 0.10886449620142223, + "learning_rate": 0.00019711812351725603, + "loss": 0.6758, + "step": 1167 + }, + { + "epoch": 0.1044349070100143, + "grad_norm": 0.10898188381206078, + "learning_rate": 0.00019711121702835504, + "loss": 0.6974, + "step": 1168 + }, + { + "epoch": 0.10452432045779685, + "grad_norm": 0.12715930626510313, + "learning_rate": 0.00019710430239485354, + "loss": 0.703, + "step": 1169 + }, + { + "epoch": 0.1046137339055794, + "grad_norm": 0.11676190279552061, + "learning_rate": 0.0001970973796173315, + "loss": 0.7163, + "step": 1170 + }, + { + "epoch": 0.10470314735336195, + "grad_norm": 0.10575561996244942, + "learning_rate": 0.00019709044869636947, + "loss": 0.6833, + "step": 1171 + }, + { + "epoch": 0.10479256080114449, + "grad_norm": 0.11995122450132073, + "learning_rate": 0.00019708350963254878, + "loss": 0.7261, + "step": 1172 + }, + { + "epoch": 0.10488197424892703, + "grad_norm": 0.12019285099434596, + "learning_rate": 0.0001970765624264514, + "loss": 0.7255, + "step": 1173 + }, + { + "epoch": 0.10497138769670958, + "grad_norm": 0.1253666000971538, + "learning_rate": 0.00019706960707865996, + "loss": 0.6752, + "step": 1174 + }, + { + "epoch": 0.10506080114449214, + "grad_norm": 0.12417329396288784, + "learning_rate": 0.00019706264358975779, + "loss": 0.7176, + "step": 1175 + }, + { + "epoch": 0.10515021459227468, + "grad_norm": 0.11054270472199185, + "learning_rate": 0.00019705567196032892, + "loss": 0.6664, + "step": 1176 + }, + { + "epoch": 0.10523962804005722, + "grad_norm": 0.10324696249694924, + "learning_rate": 0.0001970486921909581, + "loss": 0.6913, + "step": 1177 + }, + { + "epoch": 0.10532904148783977, + "grad_norm": 0.11259656358299654, + "learning_rate": 0.0001970417042822306, + "loss": 0.7438, + "step": 1178 + }, + { + "epoch": 0.10541845493562232, + "grad_norm": 0.11186860043609369, + "learning_rate": 0.00019703470823473262, + "loss": 0.6839, + "step": 1179 + }, + { + "epoch": 0.10550786838340487, + "grad_norm": 0.12711731912366964, + "learning_rate": 0.0001970277040490508, + "loss": 0.6962, + "step": 1180 + }, + { + "epoch": 0.10559728183118741, + "grad_norm": 0.11685555632229629, + "learning_rate": 0.0001970206917257727, + "loss": 0.7007, + "step": 1181 + }, + { + "epoch": 0.10568669527896996, + "grad_norm": 0.1113168755327013, + "learning_rate": 0.0001970136712654863, + "loss": 0.6927, + "step": 1182 + }, + { + "epoch": 0.1057761087267525, + "grad_norm": 0.12925066997874513, + "learning_rate": 0.00019700664266878045, + "loss": 0.7141, + "step": 1183 + }, + { + "epoch": 0.10586552217453506, + "grad_norm": 0.1116970623847749, + "learning_rate": 0.00019699960593624462, + "loss": 0.6946, + "step": 1184 + }, + { + "epoch": 0.1059549356223176, + "grad_norm": 0.11991861467026607, + "learning_rate": 0.00019699256106846903, + "loss": 0.6987, + "step": 1185 + }, + { + "epoch": 0.10604434907010014, + "grad_norm": 0.11181219698406358, + "learning_rate": 0.00019698550806604445, + "loss": 0.7132, + "step": 1186 + }, + { + "epoch": 0.10613376251788269, + "grad_norm": 0.10836094371495336, + "learning_rate": 0.00019697844692956245, + "loss": 0.6929, + "step": 1187 + }, + { + "epoch": 0.10622317596566523, + "grad_norm": 0.11994014083359124, + "learning_rate": 0.0001969713776596152, + "loss": 0.723, + "step": 1188 + }, + { + "epoch": 0.10631258941344779, + "grad_norm": 0.12520600534931187, + "learning_rate": 0.00019696430025679566, + "loss": 0.7368, + "step": 1189 + }, + { + "epoch": 0.10640200286123033, + "grad_norm": 0.10988892154608666, + "learning_rate": 0.00019695721472169734, + "loss": 0.6862, + "step": 1190 + }, + { + "epoch": 0.10649141630901288, + "grad_norm": 0.1075782189334719, + "learning_rate": 0.0001969501210549145, + "loss": 0.6781, + "step": 1191 + }, + { + "epoch": 0.10658082975679542, + "grad_norm": 0.10403788100719273, + "learning_rate": 0.0001969430192570421, + "loss": 0.7187, + "step": 1192 + }, + { + "epoch": 0.10667024320457796, + "grad_norm": 0.13090416541459943, + "learning_rate": 0.00019693590932867578, + "loss": 0.6895, + "step": 1193 + }, + { + "epoch": 0.10675965665236052, + "grad_norm": 0.10805190188273561, + "learning_rate": 0.0001969287912704118, + "loss": 0.659, + "step": 1194 + }, + { + "epoch": 0.10684907010014306, + "grad_norm": 0.11624515972969272, + "learning_rate": 0.00019692166508284716, + "loss": 0.7493, + "step": 1195 + }, + { + "epoch": 0.10693848354792561, + "grad_norm": 0.13105416528588545, + "learning_rate": 0.0001969145307665795, + "loss": 0.7565, + "step": 1196 + }, + { + "epoch": 0.10702789699570815, + "grad_norm": 0.12700184180579324, + "learning_rate": 0.0001969073883222072, + "loss": 0.7499, + "step": 1197 + }, + { + "epoch": 0.1071173104434907, + "grad_norm": 0.12294612729803947, + "learning_rate": 0.00019690023775032929, + "loss": 0.7007, + "step": 1198 + }, + { + "epoch": 0.10720672389127325, + "grad_norm": 0.1138218789466324, + "learning_rate": 0.00019689307905154543, + "loss": 0.7049, + "step": 1199 + }, + { + "epoch": 0.1072961373390558, + "grad_norm": 0.12018196901462116, + "learning_rate": 0.00019688591222645607, + "loss": 0.7149, + "step": 1200 + }, + { + "epoch": 0.10738555078683834, + "grad_norm": 0.11894539117808854, + "learning_rate": 0.00019687873727566226, + "loss": 0.7034, + "step": 1201 + }, + { + "epoch": 0.10747496423462088, + "grad_norm": 0.13866374699758754, + "learning_rate": 0.00019687155419976574, + "loss": 0.7272, + "step": 1202 + }, + { + "epoch": 0.10756437768240343, + "grad_norm": 0.1285999039351491, + "learning_rate": 0.000196864362999369, + "loss": 0.679, + "step": 1203 + }, + { + "epoch": 0.10765379113018599, + "grad_norm": 0.13887997482505882, + "learning_rate": 0.00019685716367507508, + "loss": 0.706, + "step": 1204 + }, + { + "epoch": 0.10774320457796853, + "grad_norm": 0.11296262922710712, + "learning_rate": 0.00019684995622748784, + "loss": 0.7038, + "step": 1205 + }, + { + "epoch": 0.10783261802575107, + "grad_norm": 0.14138232485281604, + "learning_rate": 0.00019684274065721172, + "loss": 0.7207, + "step": 1206 + }, + { + "epoch": 0.10792203147353362, + "grad_norm": 0.1142052614730536, + "learning_rate": 0.00019683551696485192, + "loss": 0.6589, + "step": 1207 + }, + { + "epoch": 0.10801144492131616, + "grad_norm": 0.12842300090840617, + "learning_rate": 0.00019682828515101423, + "loss": 0.7578, + "step": 1208 + }, + { + "epoch": 0.10810085836909872, + "grad_norm": 0.1084295369907099, + "learning_rate": 0.0001968210452163052, + "loss": 0.6724, + "step": 1209 + }, + { + "epoch": 0.10819027181688126, + "grad_norm": 0.12381589293416753, + "learning_rate": 0.00019681379716133206, + "loss": 0.6887, + "step": 1210 + }, + { + "epoch": 0.1082796852646638, + "grad_norm": 0.1101614627217912, + "learning_rate": 0.00019680654098670267, + "loss": 0.6692, + "step": 1211 + }, + { + "epoch": 0.10836909871244635, + "grad_norm": 0.1269803773354812, + "learning_rate": 0.0001967992766930256, + "loss": 0.7607, + "step": 1212 + }, + { + "epoch": 0.10845851216022889, + "grad_norm": 0.1253798908516139, + "learning_rate": 0.0001967920042809101, + "loss": 0.7293, + "step": 1213 + }, + { + "epoch": 0.10854792560801145, + "grad_norm": 0.10528718829484565, + "learning_rate": 0.0001967847237509661, + "loss": 0.7308, + "step": 1214 + }, + { + "epoch": 0.108637339055794, + "grad_norm": 0.11065271200237686, + "learning_rate": 0.00019677743510380417, + "loss": 0.7145, + "step": 1215 + }, + { + "epoch": 0.10872675250357654, + "grad_norm": 0.11973804082803842, + "learning_rate": 0.0001967701383400357, + "loss": 0.6826, + "step": 1216 + }, + { + "epoch": 0.10881616595135908, + "grad_norm": 0.10148395139584215, + "learning_rate": 0.00019676283346027254, + "loss": 0.6939, + "step": 1217 + }, + { + "epoch": 0.10890557939914162, + "grad_norm": 0.10938519426134258, + "learning_rate": 0.0001967555204651274, + "loss": 0.6717, + "step": 1218 + }, + { + "epoch": 0.10899499284692418, + "grad_norm": 0.1326979238330544, + "learning_rate": 0.00019674819935521366, + "loss": 0.724, + "step": 1219 + }, + { + "epoch": 0.10908440629470673, + "grad_norm": 0.12496801885972095, + "learning_rate": 0.0001967408701311452, + "loss": 0.7249, + "step": 1220 + }, + { + "epoch": 0.10917381974248927, + "grad_norm": 0.22664208521610468, + "learning_rate": 0.00019673353279353684, + "loss": 0.7208, + "step": 1221 + }, + { + "epoch": 0.10926323319027181, + "grad_norm": 0.12051229574000212, + "learning_rate": 0.00019672618734300392, + "loss": 0.6835, + "step": 1222 + }, + { + "epoch": 0.10935264663805436, + "grad_norm": 0.12714093673621846, + "learning_rate": 0.0001967188337801625, + "loss": 0.7032, + "step": 1223 + }, + { + "epoch": 0.10944206008583691, + "grad_norm": 0.11241804069544546, + "learning_rate": 0.00019671147210562927, + "loss": 0.6773, + "step": 1224 + }, + { + "epoch": 0.10953147353361946, + "grad_norm": 0.12465718086889087, + "learning_rate": 0.00019670410232002164, + "loss": 0.7346, + "step": 1225 + }, + { + "epoch": 0.109620886981402, + "grad_norm": 0.11852803912493205, + "learning_rate": 0.00019669672442395778, + "loss": 0.6831, + "step": 1226 + }, + { + "epoch": 0.10971030042918455, + "grad_norm": 0.09713923788391618, + "learning_rate": 0.00019668933841805644, + "loss": 0.6708, + "step": 1227 + }, + { + "epoch": 0.10979971387696709, + "grad_norm": 0.10443199040718543, + "learning_rate": 0.00019668194430293702, + "loss": 0.3034, + "step": 1228 + }, + { + "epoch": 0.10988912732474965, + "grad_norm": 0.12341031218645673, + "learning_rate": 0.0001966745420792197, + "loss": 0.7473, + "step": 1229 + }, + { + "epoch": 0.10997854077253219, + "grad_norm": 0.10774190447794786, + "learning_rate": 0.0001966671317475253, + "loss": 0.6603, + "step": 1230 + }, + { + "epoch": 0.11006795422031473, + "grad_norm": 0.11740955896867304, + "learning_rate": 0.00019665971330847532, + "loss": 0.7156, + "step": 1231 + }, + { + "epoch": 0.11015736766809728, + "grad_norm": 0.13328883937712693, + "learning_rate": 0.0001966522867626919, + "loss": 0.6977, + "step": 1232 + }, + { + "epoch": 0.11024678111587982, + "grad_norm": 0.13824270537705768, + "learning_rate": 0.00019664485211079793, + "loss": 0.7622, + "step": 1233 + }, + { + "epoch": 0.11033619456366238, + "grad_norm": 0.12024273125183096, + "learning_rate": 0.0001966374093534169, + "loss": 0.6952, + "step": 1234 + }, + { + "epoch": 0.11042560801144492, + "grad_norm": 0.11520224705199973, + "learning_rate": 0.00019662995849117307, + "loss": 0.694, + "step": 1235 + }, + { + "epoch": 0.11051502145922747, + "grad_norm": 0.10980867868489468, + "learning_rate": 0.00019662249952469133, + "loss": 0.6748, + "step": 1236 + }, + { + "epoch": 0.11060443490701001, + "grad_norm": 0.11710754989477411, + "learning_rate": 0.00019661503245459722, + "loss": 0.692, + "step": 1237 + }, + { + "epoch": 0.11069384835479255, + "grad_norm": 0.12927822880804807, + "learning_rate": 0.00019660755728151706, + "loss": 0.7556, + "step": 1238 + }, + { + "epoch": 0.11078326180257511, + "grad_norm": 0.11942537405761576, + "learning_rate": 0.00019660007400607772, + "loss": 0.7218, + "step": 1239 + }, + { + "epoch": 0.11087267525035766, + "grad_norm": 0.10432291039728915, + "learning_rate": 0.00019659258262890683, + "loss": 0.7122, + "step": 1240 + }, + { + "epoch": 0.1109620886981402, + "grad_norm": 0.09846088294109304, + "learning_rate": 0.00019658508315063272, + "loss": 0.6507, + "step": 1241 + }, + { + "epoch": 0.11105150214592274, + "grad_norm": 0.11680826551435233, + "learning_rate": 0.0001965775755718843, + "loss": 0.7308, + "step": 1242 + }, + { + "epoch": 0.1111409155937053, + "grad_norm": 0.1111590228010603, + "learning_rate": 0.00019657005989329128, + "loss": 0.7224, + "step": 1243 + }, + { + "epoch": 0.11123032904148784, + "grad_norm": 0.11563721747995066, + "learning_rate": 0.0001965625361154839, + "loss": 0.7283, + "step": 1244 + }, + { + "epoch": 0.11131974248927039, + "grad_norm": 0.1176016052693368, + "learning_rate": 0.0001965550042390933, + "loss": 0.6767, + "step": 1245 + }, + { + "epoch": 0.11140915593705293, + "grad_norm": 0.11423301244846722, + "learning_rate": 0.0001965474642647511, + "loss": 0.6938, + "step": 1246 + }, + { + "epoch": 0.11149856938483547, + "grad_norm": 0.09963848305362791, + "learning_rate": 0.00019653991619308965, + "loss": 0.6874, + "step": 1247 + }, + { + "epoch": 0.11158798283261803, + "grad_norm": 0.10939398646721915, + "learning_rate": 0.000196532360024742, + "loss": 0.6766, + "step": 1248 + }, + { + "epoch": 0.11167739628040058, + "grad_norm": 0.11513335203329117, + "learning_rate": 0.00019652479576034196, + "loss": 0.6999, + "step": 1249 + }, + { + "epoch": 0.11176680972818312, + "grad_norm": 0.11995005968457807, + "learning_rate": 0.0001965172234005238, + "loss": 0.6938, + "step": 1250 + }, + { + "epoch": 0.11185622317596566, + "grad_norm": 0.12408052414507244, + "learning_rate": 0.00019650964294592272, + "loss": 0.68, + "step": 1251 + }, + { + "epoch": 0.11194563662374821, + "grad_norm": 0.132084241808048, + "learning_rate": 0.0001965020543971744, + "loss": 0.7245, + "step": 1252 + }, + { + "epoch": 0.11203505007153076, + "grad_norm": 0.12035746485167009, + "learning_rate": 0.0001964944577549154, + "loss": 0.7092, + "step": 1253 + }, + { + "epoch": 0.11212446351931331, + "grad_norm": 0.13198127655275488, + "learning_rate": 0.00019648685301978271, + "loss": 0.745, + "step": 1254 + }, + { + "epoch": 0.11221387696709585, + "grad_norm": 0.13660856916149586, + "learning_rate": 0.00019647924019241416, + "loss": 0.7367, + "step": 1255 + }, + { + "epoch": 0.1123032904148784, + "grad_norm": 0.1292083931085304, + "learning_rate": 0.00019647161927344831, + "loss": 0.6797, + "step": 1256 + }, + { + "epoch": 0.11239270386266094, + "grad_norm": 0.10652094599727072, + "learning_rate": 0.00019646399026352422, + "loss": 0.6734, + "step": 1257 + }, + { + "epoch": 0.1124821173104435, + "grad_norm": 0.11540577857961776, + "learning_rate": 0.00019645635316328179, + "loss": 0.7151, + "step": 1258 + }, + { + "epoch": 0.11257153075822604, + "grad_norm": 0.11757484789595411, + "learning_rate": 0.0001964487079733615, + "loss": 0.704, + "step": 1259 + }, + { + "epoch": 0.11266094420600858, + "grad_norm": 0.12274031927742277, + "learning_rate": 0.00019644105469440455, + "loss": 0.7074, + "step": 1260 + }, + { + "epoch": 0.11275035765379113, + "grad_norm": 0.127218503509548, + "learning_rate": 0.00019643339332705282, + "loss": 0.7186, + "step": 1261 + }, + { + "epoch": 0.11283977110157367, + "grad_norm": 0.11514627131565887, + "learning_rate": 0.00019642572387194884, + "loss": 0.6802, + "step": 1262 + }, + { + "epoch": 0.11292918454935623, + "grad_norm": 0.1310357230409485, + "learning_rate": 0.00019641804632973585, + "loss": 0.7257, + "step": 1263 + }, + { + "epoch": 0.11301859799713877, + "grad_norm": 0.11291642185617894, + "learning_rate": 0.00019641036070105778, + "loss": 0.6836, + "step": 1264 + }, + { + "epoch": 0.11310801144492132, + "grad_norm": 0.12115336828825031, + "learning_rate": 0.00019640266698655917, + "loss": 0.6562, + "step": 1265 + }, + { + "epoch": 0.11319742489270386, + "grad_norm": 0.133690452839611, + "learning_rate": 0.00019639496518688532, + "loss": 0.7256, + "step": 1266 + }, + { + "epoch": 0.1132868383404864, + "grad_norm": 0.11035158256375152, + "learning_rate": 0.00019638725530268217, + "loss": 0.6738, + "step": 1267 + }, + { + "epoch": 0.11337625178826896, + "grad_norm": 0.11692949582942308, + "learning_rate": 0.00019637953733459628, + "loss": 0.7078, + "step": 1268 + }, + { + "epoch": 0.1134656652360515, + "grad_norm": 0.11157075855123512, + "learning_rate": 0.00019637181128327505, + "loss": 0.6272, + "step": 1269 + }, + { + "epoch": 0.11355507868383405, + "grad_norm": 0.10622698255544447, + "learning_rate": 0.00019636407714936636, + "loss": 0.702, + "step": 1270 + }, + { + "epoch": 0.11364449213161659, + "grad_norm": 0.11902178256631198, + "learning_rate": 0.00019635633493351887, + "loss": 0.7227, + "step": 1271 + }, + { + "epoch": 0.11373390557939914, + "grad_norm": 0.13245539462089703, + "learning_rate": 0.000196348584636382, + "loss": 0.2878, + "step": 1272 + }, + { + "epoch": 0.1138233190271817, + "grad_norm": 0.1445023859493238, + "learning_rate": 0.00019634082625860562, + "loss": 0.6932, + "step": 1273 + }, + { + "epoch": 0.11391273247496424, + "grad_norm": 0.12441878112149107, + "learning_rate": 0.00019633305980084055, + "loss": 0.6519, + "step": 1274 + }, + { + "epoch": 0.11400214592274678, + "grad_norm": 0.10350925324794416, + "learning_rate": 0.0001963252852637381, + "loss": 0.6872, + "step": 1275 + }, + { + "epoch": 0.11409155937052932, + "grad_norm": 0.12896116151013695, + "learning_rate": 0.00019631750264795028, + "loss": 0.7262, + "step": 1276 + }, + { + "epoch": 0.11418097281831187, + "grad_norm": 0.11877646472143952, + "learning_rate": 0.00019630971195412983, + "loss": 0.7307, + "step": 1277 + }, + { + "epoch": 0.11427038626609443, + "grad_norm": 0.11905939881346582, + "learning_rate": 0.00019630191318293017, + "loss": 0.7016, + "step": 1278 + }, + { + "epoch": 0.11435979971387697, + "grad_norm": 0.12712610553583445, + "learning_rate": 0.00019629410633500535, + "loss": 0.7154, + "step": 1279 + }, + { + "epoch": 0.11444921316165951, + "grad_norm": 0.10353728363168897, + "learning_rate": 0.00019628629141101012, + "loss": 0.6849, + "step": 1280 + }, + { + "epoch": 0.11453862660944206, + "grad_norm": 0.10535182977582803, + "learning_rate": 0.0001962784684115999, + "loss": 0.7016, + "step": 1281 + }, + { + "epoch": 0.1146280400572246, + "grad_norm": 0.10773673692888608, + "learning_rate": 0.00019627063733743084, + "loss": 0.6828, + "step": 1282 + }, + { + "epoch": 0.11471745350500716, + "grad_norm": 0.10560660822724055, + "learning_rate": 0.0001962627981891597, + "loss": 0.668, + "step": 1283 + }, + { + "epoch": 0.1148068669527897, + "grad_norm": 0.10987980951946305, + "learning_rate": 0.00019625495096744388, + "loss": 0.6655, + "step": 1284 + }, + { + "epoch": 0.11489628040057225, + "grad_norm": 0.11719802353903049, + "learning_rate": 0.00019624709567294158, + "loss": 0.7418, + "step": 1285 + }, + { + "epoch": 0.11498569384835479, + "grad_norm": 0.11622833504620805, + "learning_rate": 0.00019623923230631165, + "loss": 0.7196, + "step": 1286 + }, + { + "epoch": 0.11507510729613733, + "grad_norm": 0.1053228493464359, + "learning_rate": 0.0001962313608682135, + "loss": 0.7125, + "step": 1287 + }, + { + "epoch": 0.11516452074391989, + "grad_norm": 0.11728160214933293, + "learning_rate": 0.00019622348135930735, + "loss": 0.7282, + "step": 1288 + }, + { + "epoch": 0.11525393419170243, + "grad_norm": 0.10145056612037681, + "learning_rate": 0.00019621559378025401, + "loss": 0.2575, + "step": 1289 + }, + { + "epoch": 0.11534334763948498, + "grad_norm": 0.13641590306824836, + "learning_rate": 0.00019620769813171504, + "loss": 0.6889, + "step": 1290 + }, + { + "epoch": 0.11543276108726752, + "grad_norm": 0.12020107095226072, + "learning_rate": 0.0001961997944143526, + "loss": 0.6955, + "step": 1291 + }, + { + "epoch": 0.11552217453505007, + "grad_norm": 0.12558132291633162, + "learning_rate": 0.00019619188262882958, + "loss": 0.6956, + "step": 1292 + }, + { + "epoch": 0.11561158798283262, + "grad_norm": 0.11865060214169458, + "learning_rate": 0.00019618396277580952, + "loss": 0.7036, + "step": 1293 + }, + { + "epoch": 0.11570100143061517, + "grad_norm": 0.12402696943641586, + "learning_rate": 0.0001961760348559567, + "loss": 0.6795, + "step": 1294 + }, + { + "epoch": 0.11579041487839771, + "grad_norm": 0.14192681425203063, + "learning_rate": 0.00019616809886993596, + "loss": 0.7187, + "step": 1295 + }, + { + "epoch": 0.11587982832618025, + "grad_norm": 0.12601147412251476, + "learning_rate": 0.0001961601548184129, + "loss": 0.6845, + "step": 1296 + }, + { + "epoch": 0.1159692417739628, + "grad_norm": 0.1163661985375088, + "learning_rate": 0.00019615220270205382, + "loss": 0.675, + "step": 1297 + }, + { + "epoch": 0.11605865522174535, + "grad_norm": 0.12398661409618089, + "learning_rate": 0.00019614424252152558, + "loss": 0.7149, + "step": 1298 + }, + { + "epoch": 0.1161480686695279, + "grad_norm": 0.12079477636210345, + "learning_rate": 0.0001961362742774959, + "loss": 0.6981, + "step": 1299 + }, + { + "epoch": 0.11623748211731044, + "grad_norm": 0.1251898669530329, + "learning_rate": 0.00019612829797063295, + "loss": 0.7084, + "step": 1300 + }, + { + "epoch": 0.11632689556509299, + "grad_norm": 0.11661650693185259, + "learning_rate": 0.00019612031360160574, + "loss": 0.7096, + "step": 1301 + }, + { + "epoch": 0.11641630901287553, + "grad_norm": 0.11763289664576931, + "learning_rate": 0.00019611232117108395, + "loss": 0.7246, + "step": 1302 + }, + { + "epoch": 0.11650572246065809, + "grad_norm": 0.10961685597610349, + "learning_rate": 0.00019610432067973781, + "loss": 0.6698, + "step": 1303 + }, + { + "epoch": 0.11659513590844063, + "grad_norm": 0.11625723956709404, + "learning_rate": 0.0001960963121282384, + "loss": 0.6659, + "step": 1304 + }, + { + "epoch": 0.11668454935622317, + "grad_norm": 0.11576809868947356, + "learning_rate": 0.0001960882955172573, + "loss": 0.6733, + "step": 1305 + }, + { + "epoch": 0.11677396280400572, + "grad_norm": 0.11229284348494409, + "learning_rate": 0.00019608027084746694, + "loss": 0.6685, + "step": 1306 + }, + { + "epoch": 0.11686337625178828, + "grad_norm": 0.125684209576539, + "learning_rate": 0.0001960722381195403, + "loss": 0.7104, + "step": 1307 + }, + { + "epoch": 0.11695278969957082, + "grad_norm": 0.11262829074947935, + "learning_rate": 0.00019606419733415105, + "loss": 0.6734, + "step": 1308 + }, + { + "epoch": 0.11704220314735336, + "grad_norm": 0.10726991200223623, + "learning_rate": 0.00019605614849197358, + "loss": 0.7058, + "step": 1309 + }, + { + "epoch": 0.1171316165951359, + "grad_norm": 0.11128653926935062, + "learning_rate": 0.000196048091593683, + "loss": 0.6611, + "step": 1310 + }, + { + "epoch": 0.11722103004291845, + "grad_norm": 0.10135835762156749, + "learning_rate": 0.00019604002663995492, + "loss": 0.6897, + "step": 1311 + }, + { + "epoch": 0.11731044349070101, + "grad_norm": 0.11878123782881872, + "learning_rate": 0.0001960319536314658, + "loss": 0.6632, + "step": 1312 + }, + { + "epoch": 0.11739985693848355, + "grad_norm": 0.11098516914258622, + "learning_rate": 0.0001960238725688927, + "loss": 0.6863, + "step": 1313 + }, + { + "epoch": 0.1174892703862661, + "grad_norm": 0.1279052966100143, + "learning_rate": 0.0001960157834529134, + "loss": 0.7349, + "step": 1314 + }, + { + "epoch": 0.11757868383404864, + "grad_norm": 0.12058436864922713, + "learning_rate": 0.0001960076862842063, + "loss": 0.6652, + "step": 1315 + }, + { + "epoch": 0.11766809728183118, + "grad_norm": 0.09256535934406156, + "learning_rate": 0.00019599958106345045, + "loss": 0.6747, + "step": 1316 + }, + { + "epoch": 0.11775751072961374, + "grad_norm": 0.12826964398154214, + "learning_rate": 0.00019599146779132576, + "loss": 0.7011, + "step": 1317 + }, + { + "epoch": 0.11784692417739628, + "grad_norm": 0.14399884617895234, + "learning_rate": 0.00019598334646851254, + "loss": 0.2701, + "step": 1318 + }, + { + "epoch": 0.11793633762517883, + "grad_norm": 0.11256892574261569, + "learning_rate": 0.000195975217095692, + "loss": 0.6871, + "step": 1319 + }, + { + "epoch": 0.11802575107296137, + "grad_norm": 0.11987376201434492, + "learning_rate": 0.00019596707967354585, + "loss": 0.7056, + "step": 1320 + }, + { + "epoch": 0.11811516452074391, + "grad_norm": 0.1238942616508295, + "learning_rate": 0.00019595893420275667, + "loss": 0.7124, + "step": 1321 + }, + { + "epoch": 0.11820457796852647, + "grad_norm": 0.12839492060349186, + "learning_rate": 0.00019595078068400756, + "loss": 0.7116, + "step": 1322 + }, + { + "epoch": 0.11829399141630902, + "grad_norm": 0.12450196769437955, + "learning_rate": 0.00019594261911798236, + "loss": 0.6636, + "step": 1323 + }, + { + "epoch": 0.11838340486409156, + "grad_norm": 0.12552297948060112, + "learning_rate": 0.00019593444950536556, + "loss": 0.7027, + "step": 1324 + }, + { + "epoch": 0.1184728183118741, + "grad_norm": 0.13855282088247564, + "learning_rate": 0.00019592627184684234, + "loss": 0.749, + "step": 1325 + }, + { + "epoch": 0.11856223175965665, + "grad_norm": 0.11726929823823844, + "learning_rate": 0.00019591808614309854, + "loss": 0.6872, + "step": 1326 + }, + { + "epoch": 0.1186516452074392, + "grad_norm": 0.11655066553199171, + "learning_rate": 0.0001959098923948207, + "loss": 0.629, + "step": 1327 + }, + { + "epoch": 0.11874105865522175, + "grad_norm": 0.12154968533159438, + "learning_rate": 0.00019590169060269602, + "loss": 0.7105, + "step": 1328 + }, + { + "epoch": 0.11883047210300429, + "grad_norm": 0.1256428808173355, + "learning_rate": 0.0001958934807674124, + "loss": 0.6918, + "step": 1329 + }, + { + "epoch": 0.11891988555078684, + "grad_norm": 0.12581923880172083, + "learning_rate": 0.00019588526288965828, + "loss": 0.7236, + "step": 1330 + }, + { + "epoch": 0.11900929899856938, + "grad_norm": 0.10385785112174444, + "learning_rate": 0.00019587703697012302, + "loss": 0.6592, + "step": 1331 + }, + { + "epoch": 0.11909871244635194, + "grad_norm": 0.11615685730225582, + "learning_rate": 0.00019586880300949644, + "loss": 0.6891, + "step": 1332 + }, + { + "epoch": 0.11918812589413448, + "grad_norm": 0.10093117840700538, + "learning_rate": 0.00019586056100846916, + "loss": 0.7294, + "step": 1333 + }, + { + "epoch": 0.11927753934191702, + "grad_norm": 0.12948616224321244, + "learning_rate": 0.00019585231096773238, + "loss": 0.6999, + "step": 1334 + }, + { + "epoch": 0.11936695278969957, + "grad_norm": 0.11624198779675886, + "learning_rate": 0.00019584405288797802, + "loss": 0.6966, + "step": 1335 + }, + { + "epoch": 0.11945636623748211, + "grad_norm": 0.1254528547252989, + "learning_rate": 0.0001958357867698987, + "loss": 0.7492, + "step": 1336 + }, + { + "epoch": 0.11954577968526467, + "grad_norm": 0.1030797713987865, + "learning_rate": 0.0001958275126141877, + "loss": 0.6605, + "step": 1337 + }, + { + "epoch": 0.11963519313304721, + "grad_norm": 0.11234451753625661, + "learning_rate": 0.00019581923042153894, + "loss": 0.6971, + "step": 1338 + }, + { + "epoch": 0.11972460658082976, + "grad_norm": 0.11303629721556366, + "learning_rate": 0.00019581094019264705, + "loss": 0.7087, + "step": 1339 + }, + { + "epoch": 0.1198140200286123, + "grad_norm": 0.12473493868294776, + "learning_rate": 0.00019580264192820733, + "loss": 0.7224, + "step": 1340 + }, + { + "epoch": 0.11990343347639484, + "grad_norm": 0.124838521448301, + "learning_rate": 0.00019579433562891572, + "loss": 0.6768, + "step": 1341 + }, + { + "epoch": 0.1199928469241774, + "grad_norm": 0.11856320426605804, + "learning_rate": 0.00019578602129546885, + "loss": 0.6838, + "step": 1342 + }, + { + "epoch": 0.12008226037195995, + "grad_norm": 0.1185043899122802, + "learning_rate": 0.00019577769892856407, + "loss": 0.7258, + "step": 1343 + }, + { + "epoch": 0.12017167381974249, + "grad_norm": 0.1210448249831688, + "learning_rate": 0.00019576936852889936, + "loss": 0.6883, + "step": 1344 + }, + { + "epoch": 0.12026108726752503, + "grad_norm": 0.1357601086223499, + "learning_rate": 0.00019576103009717337, + "loss": 0.6754, + "step": 1345 + }, + { + "epoch": 0.12035050071530758, + "grad_norm": 0.13602534387706394, + "learning_rate": 0.00019575268363408542, + "loss": 0.7537, + "step": 1346 + }, + { + "epoch": 0.12043991416309013, + "grad_norm": 0.13374123150757194, + "learning_rate": 0.00019574432914033554, + "loss": 0.7196, + "step": 1347 + }, + { + "epoch": 0.12052932761087268, + "grad_norm": 0.12461187433244905, + "learning_rate": 0.00019573596661662438, + "loss": 0.7351, + "step": 1348 + }, + { + "epoch": 0.12061874105865522, + "grad_norm": 0.12143756561324717, + "learning_rate": 0.00019572759606365336, + "loss": 0.7291, + "step": 1349 + }, + { + "epoch": 0.12070815450643776, + "grad_norm": 0.09607438661544475, + "learning_rate": 0.00019571921748212447, + "loss": 0.6639, + "step": 1350 + }, + { + "epoch": 0.12079756795422031, + "grad_norm": 0.1345769131624991, + "learning_rate": 0.0001957108308727404, + "loss": 0.7364, + "step": 1351 + }, + { + "epoch": 0.12088698140200287, + "grad_norm": 0.11069661500472427, + "learning_rate": 0.0001957024362362045, + "loss": 0.708, + "step": 1352 + }, + { + "epoch": 0.12097639484978541, + "grad_norm": 0.12339661161744833, + "learning_rate": 0.0001956940335732209, + "loss": 0.7135, + "step": 1353 + }, + { + "epoch": 0.12106580829756795, + "grad_norm": 0.12122645158919086, + "learning_rate": 0.00019568562288449422, + "loss": 0.7119, + "step": 1354 + }, + { + "epoch": 0.1211552217453505, + "grad_norm": 0.11085274736212412, + "learning_rate": 0.00019567720417072997, + "loss": 0.6817, + "step": 1355 + }, + { + "epoch": 0.12124463519313304, + "grad_norm": 0.1426406135242438, + "learning_rate": 0.00019566877743263414, + "loss": 0.7054, + "step": 1356 + }, + { + "epoch": 0.1213340486409156, + "grad_norm": 0.11637108419033511, + "learning_rate": 0.00019566034267091346, + "loss": 0.682, + "step": 1357 + }, + { + "epoch": 0.12142346208869814, + "grad_norm": 0.13144963303115906, + "learning_rate": 0.0001956518998862754, + "loss": 0.7078, + "step": 1358 + }, + { + "epoch": 0.12151287553648069, + "grad_norm": 0.12032881079377375, + "learning_rate": 0.00019564344907942798, + "loss": 0.7099, + "step": 1359 + }, + { + "epoch": 0.12160228898426323, + "grad_norm": 0.11547296863764686, + "learning_rate": 0.00019563499025107998, + "loss": 0.7397, + "step": 1360 + }, + { + "epoch": 0.12169170243204577, + "grad_norm": 0.13552957385078165, + "learning_rate": 0.0001956265234019409, + "loss": 0.7641, + "step": 1361 + }, + { + "epoch": 0.12178111587982833, + "grad_norm": 0.12562473171906177, + "learning_rate": 0.00019561804853272075, + "loss": 0.714, + "step": 1362 + }, + { + "epoch": 0.12187052932761087, + "grad_norm": 0.12275691446774233, + "learning_rate": 0.00019560956564413035, + "loss": 0.6926, + "step": 1363 + }, + { + "epoch": 0.12195994277539342, + "grad_norm": 0.12278429143348328, + "learning_rate": 0.00019560107473688118, + "loss": 0.7028, + "step": 1364 + }, + { + "epoch": 0.12204935622317596, + "grad_norm": 0.12877898887338404, + "learning_rate": 0.0001955925758116853, + "loss": 0.7261, + "step": 1365 + }, + { + "epoch": 0.1221387696709585, + "grad_norm": 0.09625600379728264, + "learning_rate": 0.00019558406886925554, + "loss": 0.6701, + "step": 1366 + }, + { + "epoch": 0.12222818311874106, + "grad_norm": 0.1892060749409265, + "learning_rate": 0.00019557555391030537, + "loss": 0.2714, + "step": 1367 + }, + { + "epoch": 0.1223175965665236, + "grad_norm": 0.13834988092028228, + "learning_rate": 0.0001955670309355489, + "loss": 0.252, + "step": 1368 + }, + { + "epoch": 0.12240701001430615, + "grad_norm": 0.12892071241273875, + "learning_rate": 0.00019555849994570097, + "loss": 0.7099, + "step": 1369 + }, + { + "epoch": 0.1224964234620887, + "grad_norm": 0.12382927647028406, + "learning_rate": 0.00019554996094147707, + "loss": 0.6822, + "step": 1370 + }, + { + "epoch": 0.12258583690987125, + "grad_norm": 0.13570647063504523, + "learning_rate": 0.00019554141392359332, + "loss": 0.7315, + "step": 1371 + }, + { + "epoch": 0.1226752503576538, + "grad_norm": 0.12158639942596551, + "learning_rate": 0.00019553285889276656, + "loss": 0.6667, + "step": 1372 + }, + { + "epoch": 0.12276466380543634, + "grad_norm": 0.11590999351345593, + "learning_rate": 0.00019552429584971434, + "loss": 0.6916, + "step": 1373 + }, + { + "epoch": 0.12285407725321888, + "grad_norm": 0.14733274869355162, + "learning_rate": 0.00019551572479515478, + "loss": 0.2904, + "step": 1374 + }, + { + "epoch": 0.12294349070100143, + "grad_norm": 0.11695885545809437, + "learning_rate": 0.00019550714572980668, + "loss": 0.7, + "step": 1375 + }, + { + "epoch": 0.12303290414878398, + "grad_norm": 0.12117359231352268, + "learning_rate": 0.00019549855865438965, + "loss": 0.7205, + "step": 1376 + }, + { + "epoch": 0.12312231759656653, + "grad_norm": 0.12182697381675031, + "learning_rate": 0.00019548996356962386, + "loss": 0.6968, + "step": 1377 + }, + { + "epoch": 0.12321173104434907, + "grad_norm": 0.11517453597773013, + "learning_rate": 0.00019548136047623015, + "loss": 0.7169, + "step": 1378 + }, + { + "epoch": 0.12330114449213161, + "grad_norm": 0.11042115119897382, + "learning_rate": 0.00019547274937492998, + "loss": 0.7136, + "step": 1379 + }, + { + "epoch": 0.12339055793991416, + "grad_norm": 0.11970664039754336, + "learning_rate": 0.00019546413026644567, + "loss": 0.7182, + "step": 1380 + }, + { + "epoch": 0.12347997138769672, + "grad_norm": 0.11204362314966508, + "learning_rate": 0.00019545550315150004, + "loss": 0.6738, + "step": 1381 + }, + { + "epoch": 0.12356938483547926, + "grad_norm": 0.10729454597211419, + "learning_rate": 0.00019544686803081666, + "loss": 0.6685, + "step": 1382 + }, + { + "epoch": 0.1236587982832618, + "grad_norm": 0.10683668999883526, + "learning_rate": 0.00019543822490511974, + "loss": 0.66, + "step": 1383 + }, + { + "epoch": 0.12374821173104435, + "grad_norm": 0.13126201795680842, + "learning_rate": 0.00019542957377513412, + "loss": 0.7181, + "step": 1384 + }, + { + "epoch": 0.12383762517882689, + "grad_norm": 0.13485174052468551, + "learning_rate": 0.00019542091464158542, + "loss": 0.6563, + "step": 1385 + }, + { + "epoch": 0.12392703862660945, + "grad_norm": 0.12499281977272073, + "learning_rate": 0.00019541224750519983, + "loss": 0.7224, + "step": 1386 + }, + { + "epoch": 0.12401645207439199, + "grad_norm": 0.13186740874527494, + "learning_rate": 0.00019540357236670427, + "loss": 0.7054, + "step": 1387 + }, + { + "epoch": 0.12410586552217454, + "grad_norm": 0.1398935254378444, + "learning_rate": 0.00019539488922682633, + "loss": 0.7132, + "step": 1388 + }, + { + "epoch": 0.12419527896995708, + "grad_norm": 0.11466376319283143, + "learning_rate": 0.00019538619808629422, + "loss": 0.689, + "step": 1389 + }, + { + "epoch": 0.12428469241773962, + "grad_norm": 0.13662538815301117, + "learning_rate": 0.0001953774989458369, + "loss": 0.718, + "step": 1390 + }, + { + "epoch": 0.12437410586552218, + "grad_norm": 0.11429014891690144, + "learning_rate": 0.00019536879180618392, + "loss": 0.7096, + "step": 1391 + }, + { + "epoch": 0.12446351931330472, + "grad_norm": 0.10854969684349529, + "learning_rate": 0.00019536007666806556, + "loss": 0.654, + "step": 1392 + }, + { + "epoch": 0.12455293276108727, + "grad_norm": 0.11635123483465608, + "learning_rate": 0.00019535135353221272, + "loss": 0.6998, + "step": 1393 + }, + { + "epoch": 0.12464234620886981, + "grad_norm": 0.138056401852607, + "learning_rate": 0.000195342622399357, + "loss": 0.722, + "step": 1394 + }, + { + "epoch": 0.12473175965665236, + "grad_norm": 0.11122200365169384, + "learning_rate": 0.0001953338832702307, + "loss": 0.6917, + "step": 1395 + }, + { + "epoch": 0.12482117310443491, + "grad_norm": 0.10432983524118007, + "learning_rate": 0.00019532513614556673, + "loss": 0.6915, + "step": 1396 + }, + { + "epoch": 0.12491058655221746, + "grad_norm": 0.12750615220999406, + "learning_rate": 0.00019531638102609873, + "loss": 0.6832, + "step": 1397 + }, + { + "epoch": 0.125, + "grad_norm": 0.11001162417196433, + "learning_rate": 0.00019530761791256097, + "loss": 0.6843, + "step": 1398 + }, + { + "epoch": 0.12508941344778254, + "grad_norm": 0.1360017757059273, + "learning_rate": 0.0001952988468056884, + "loss": 0.7223, + "step": 1399 + }, + { + "epoch": 0.1251788268955651, + "grad_norm": 0.1232132089042788, + "learning_rate": 0.00019529006770621662, + "loss": 0.6909, + "step": 1400 + }, + { + "epoch": 0.12526824034334763, + "grad_norm": 0.11096665052395698, + "learning_rate": 0.00019528128061488195, + "loss": 0.6889, + "step": 1401 + }, + { + "epoch": 0.12535765379113017, + "grad_norm": 0.12335022666157545, + "learning_rate": 0.00019527248553242137, + "loss": 0.7137, + "step": 1402 + }, + { + "epoch": 0.12544706723891275, + "grad_norm": 0.13100991712245438, + "learning_rate": 0.00019526368245957246, + "loss": 0.7193, + "step": 1403 + }, + { + "epoch": 0.1255364806866953, + "grad_norm": 0.1115223478571511, + "learning_rate": 0.00019525487139707357, + "loss": 0.7016, + "step": 1404 + }, + { + "epoch": 0.12562589413447783, + "grad_norm": 0.15270860865426458, + "learning_rate": 0.00019524605234566363, + "loss": 0.7285, + "step": 1405 + }, + { + "epoch": 0.12571530758226038, + "grad_norm": 0.11674906371897689, + "learning_rate": 0.00019523722530608232, + "loss": 0.6788, + "step": 1406 + }, + { + "epoch": 0.12580472103004292, + "grad_norm": 0.14028711910229796, + "learning_rate": 0.00019522839027906995, + "loss": 0.6986, + "step": 1407 + }, + { + "epoch": 0.12589413447782546, + "grad_norm": 0.12775172131964993, + "learning_rate": 0.0001952195472653675, + "loss": 0.7256, + "step": 1408 + }, + { + "epoch": 0.125983547925608, + "grad_norm": 0.11706815288675582, + "learning_rate": 0.0001952106962657166, + "loss": 0.7137, + "step": 1409 + }, + { + "epoch": 0.12607296137339055, + "grad_norm": 0.1286457195699109, + "learning_rate": 0.0001952018372808596, + "loss": 0.6932, + "step": 1410 + }, + { + "epoch": 0.1261623748211731, + "grad_norm": 0.1310424943713035, + "learning_rate": 0.00019519297031153946, + "loss": 0.7094, + "step": 1411 + }, + { + "epoch": 0.12625178826895564, + "grad_norm": 0.11307633997326949, + "learning_rate": 0.0001951840953584999, + "loss": 0.6873, + "step": 1412 + }, + { + "epoch": 0.1263412017167382, + "grad_norm": 0.1313760852844871, + "learning_rate": 0.0001951752124224852, + "loss": 0.7069, + "step": 1413 + }, + { + "epoch": 0.12643061516452075, + "grad_norm": 0.13776091224696024, + "learning_rate": 0.00019516632150424034, + "loss": 0.7112, + "step": 1414 + }, + { + "epoch": 0.1265200286123033, + "grad_norm": 0.12364561373721411, + "learning_rate": 0.00019515742260451107, + "loss": 0.722, + "step": 1415 + }, + { + "epoch": 0.12660944206008584, + "grad_norm": 0.10988955911048962, + "learning_rate": 0.00019514851572404368, + "loss": 0.6634, + "step": 1416 + }, + { + "epoch": 0.12669885550786839, + "grad_norm": 0.1097000316784263, + "learning_rate": 0.0001951396008635852, + "loss": 0.2735, + "step": 1417 + }, + { + "epoch": 0.12678826895565093, + "grad_norm": 0.14394439685805202, + "learning_rate": 0.00019513067802388325, + "loss": 0.7004, + "step": 1418 + }, + { + "epoch": 0.12687768240343347, + "grad_norm": 0.12632133263107415, + "learning_rate": 0.00019512174720568627, + "loss": 0.6877, + "step": 1419 + }, + { + "epoch": 0.12696709585121602, + "grad_norm": 0.11432608588122568, + "learning_rate": 0.0001951128084097432, + "loss": 0.6847, + "step": 1420 + }, + { + "epoch": 0.12705650929899856, + "grad_norm": 0.12038360109048901, + "learning_rate": 0.00019510386163680375, + "loss": 0.6578, + "step": 1421 + }, + { + "epoch": 0.1271459227467811, + "grad_norm": 0.12201159289937066, + "learning_rate": 0.00019509490688761832, + "loss": 0.7006, + "step": 1422 + }, + { + "epoch": 0.12723533619456368, + "grad_norm": 0.12182174585237455, + "learning_rate": 0.0001950859441629379, + "loss": 0.6794, + "step": 1423 + }, + { + "epoch": 0.12732474964234622, + "grad_norm": 0.1047144269560641, + "learning_rate": 0.00019507697346351414, + "loss": 0.6475, + "step": 1424 + }, + { + "epoch": 0.12741416309012876, + "grad_norm": 0.11519728205724439, + "learning_rate": 0.00019506799479009944, + "loss": 0.6779, + "step": 1425 + }, + { + "epoch": 0.1275035765379113, + "grad_norm": 0.11319408443334907, + "learning_rate": 0.00019505900814344683, + "loss": 0.6758, + "step": 1426 + }, + { + "epoch": 0.12759298998569385, + "grad_norm": 0.1265673290642628, + "learning_rate": 0.00019505001352431003, + "loss": 0.7076, + "step": 1427 + }, + { + "epoch": 0.1276824034334764, + "grad_norm": 0.12822052537920103, + "learning_rate": 0.00019504101093344338, + "loss": 0.7195, + "step": 1428 + }, + { + "epoch": 0.12777181688125894, + "grad_norm": 0.1223222539234346, + "learning_rate": 0.00019503200037160193, + "loss": 0.7032, + "step": 1429 + }, + { + "epoch": 0.12786123032904148, + "grad_norm": 0.11899509879530583, + "learning_rate": 0.00019502298183954136, + "loss": 0.7298, + "step": 1430 + }, + { + "epoch": 0.12795064377682402, + "grad_norm": 0.12136463699438285, + "learning_rate": 0.00019501395533801807, + "loss": 0.6925, + "step": 1431 + }, + { + "epoch": 0.12804005722460657, + "grad_norm": 0.10814905401724123, + "learning_rate": 0.0001950049208677891, + "loss": 0.6598, + "step": 1432 + }, + { + "epoch": 0.12812947067238914, + "grad_norm": 0.10027123735751607, + "learning_rate": 0.00019499587842961214, + "loss": 0.6682, + "step": 1433 + }, + { + "epoch": 0.12821888412017168, + "grad_norm": 0.13373958561082666, + "learning_rate": 0.0001949868280242456, + "loss": 0.71, + "step": 1434 + }, + { + "epoch": 0.12830829756795423, + "grad_norm": 0.13076492698251707, + "learning_rate": 0.0001949777696524485, + "loss": 0.7573, + "step": 1435 + }, + { + "epoch": 0.12839771101573677, + "grad_norm": 0.11981281797382669, + "learning_rate": 0.00019496870331498056, + "loss": 0.6911, + "step": 1436 + }, + { + "epoch": 0.12848712446351931, + "grad_norm": 0.10571428197107414, + "learning_rate": 0.00019495962901260215, + "loss": 0.6972, + "step": 1437 + }, + { + "epoch": 0.12857653791130186, + "grad_norm": 0.1155165983249355, + "learning_rate": 0.00019495054674607438, + "loss": 0.6932, + "step": 1438 + }, + { + "epoch": 0.1286659513590844, + "grad_norm": 0.13087237647524505, + "learning_rate": 0.00019494145651615888, + "loss": 0.6957, + "step": 1439 + }, + { + "epoch": 0.12875536480686695, + "grad_norm": 0.11785929117958743, + "learning_rate": 0.0001949323583236181, + "loss": 0.7059, + "step": 1440 + }, + { + "epoch": 0.1288447782546495, + "grad_norm": 0.11955098785267515, + "learning_rate": 0.00019492325216921506, + "loss": 0.7289, + "step": 1441 + }, + { + "epoch": 0.12893419170243203, + "grad_norm": 0.11481873738437949, + "learning_rate": 0.00019491413805371356, + "loss": 0.6726, + "step": 1442 + }, + { + "epoch": 0.1290236051502146, + "grad_norm": 0.12135232137277892, + "learning_rate": 0.0001949050159778779, + "loss": 0.7346, + "step": 1443 + }, + { + "epoch": 0.12911301859799715, + "grad_norm": 0.1216708140328037, + "learning_rate": 0.00019489588594247313, + "loss": 0.7366, + "step": 1444 + }, + { + "epoch": 0.1292024320457797, + "grad_norm": 0.11495695016226866, + "learning_rate": 0.00019488674794826505, + "loss": 0.6822, + "step": 1445 + }, + { + "epoch": 0.12929184549356224, + "grad_norm": 0.1061052641855081, + "learning_rate": 0.00019487760199602, + "loss": 0.654, + "step": 1446 + }, + { + "epoch": 0.12938125894134478, + "grad_norm": 0.10618725893551444, + "learning_rate": 0.00019486844808650503, + "loss": 0.6787, + "step": 1447 + }, + { + "epoch": 0.12947067238912732, + "grad_norm": 0.11964577566153616, + "learning_rate": 0.00019485928622048793, + "loss": 0.7033, + "step": 1448 + }, + { + "epoch": 0.12956008583690987, + "grad_norm": 0.11682432571581401, + "learning_rate": 0.00019485011639873702, + "loss": 0.6849, + "step": 1449 + }, + { + "epoch": 0.1296494992846924, + "grad_norm": 0.11851915799863415, + "learning_rate": 0.0001948409386220214, + "loss": 0.6868, + "step": 1450 + }, + { + "epoch": 0.12973891273247495, + "grad_norm": 0.11002283928307778, + "learning_rate": 0.00019483175289111083, + "loss": 0.6632, + "step": 1451 + }, + { + "epoch": 0.1298283261802575, + "grad_norm": 0.12127835941094146, + "learning_rate": 0.00019482255920677565, + "loss": 0.6832, + "step": 1452 + }, + { + "epoch": 0.12991773962804007, + "grad_norm": 0.11947584819733502, + "learning_rate": 0.00019481335756978696, + "loss": 0.6877, + "step": 1453 + }, + { + "epoch": 0.1300071530758226, + "grad_norm": 0.13311908601330233, + "learning_rate": 0.00019480414798091647, + "loss": 0.6845, + "step": 1454 + }, + { + "epoch": 0.13009656652360516, + "grad_norm": 0.12061142312025516, + "learning_rate": 0.00019479493044093657, + "loss": 0.7176, + "step": 1455 + }, + { + "epoch": 0.1301859799713877, + "grad_norm": 0.11536564872272462, + "learning_rate": 0.00019478570495062037, + "loss": 0.7323, + "step": 1456 + }, + { + "epoch": 0.13027539341917024, + "grad_norm": 0.1132187839344571, + "learning_rate": 0.00019477647151074155, + "loss": 0.7097, + "step": 1457 + }, + { + "epoch": 0.1303648068669528, + "grad_norm": 0.11578546182092157, + "learning_rate": 0.0001947672301220745, + "loss": 0.7071, + "step": 1458 + }, + { + "epoch": 0.13045422031473533, + "grad_norm": 0.11818733110113316, + "learning_rate": 0.00019475798078539433, + "loss": 0.7219, + "step": 1459 + }, + { + "epoch": 0.13054363376251787, + "grad_norm": 0.1259640286291692, + "learning_rate": 0.00019474872350147676, + "loss": 0.6893, + "step": 1460 + }, + { + "epoch": 0.13063304721030042, + "grad_norm": 0.13638104452193192, + "learning_rate": 0.0001947394582710982, + "loss": 0.7414, + "step": 1461 + }, + { + "epoch": 0.130722460658083, + "grad_norm": 0.11275118328796549, + "learning_rate": 0.00019473018509503565, + "loss": 0.6332, + "step": 1462 + }, + { + "epoch": 0.13081187410586553, + "grad_norm": 0.11707156465739088, + "learning_rate": 0.00019472090397406686, + "loss": 0.6856, + "step": 1463 + }, + { + "epoch": 0.13090128755364808, + "grad_norm": 0.10550852541423597, + "learning_rate": 0.00019471161490897029, + "loss": 0.69, + "step": 1464 + }, + { + "epoch": 0.13099070100143062, + "grad_norm": 0.11812742312804057, + "learning_rate": 0.00019470231790052496, + "loss": 0.6946, + "step": 1465 + }, + { + "epoch": 0.13108011444921316, + "grad_norm": 0.11339175237404832, + "learning_rate": 0.0001946930129495106, + "loss": 0.6672, + "step": 1466 + }, + { + "epoch": 0.1311695278969957, + "grad_norm": 0.13177236791341465, + "learning_rate": 0.00019468370005670758, + "loss": 0.7231, + "step": 1467 + }, + { + "epoch": 0.13125894134477825, + "grad_norm": 0.13350912428821948, + "learning_rate": 0.00019467437922289697, + "loss": 0.7097, + "step": 1468 + }, + { + "epoch": 0.1313483547925608, + "grad_norm": 0.12071648835869163, + "learning_rate": 0.00019466505044886056, + "loss": 0.694, + "step": 1469 + }, + { + "epoch": 0.13143776824034334, + "grad_norm": 0.1230300298022115, + "learning_rate": 0.00019465571373538068, + "loss": 0.6647, + "step": 1470 + }, + { + "epoch": 0.13152718168812588, + "grad_norm": 0.11549434971877987, + "learning_rate": 0.00019464636908324038, + "loss": 0.7033, + "step": 1471 + }, + { + "epoch": 0.13161659513590845, + "grad_norm": 0.11953480934474589, + "learning_rate": 0.00019463701649322343, + "loss": 0.7133, + "step": 1472 + }, + { + "epoch": 0.131706008583691, + "grad_norm": 0.12726763903555938, + "learning_rate": 0.0001946276559661142, + "loss": 0.7197, + "step": 1473 + }, + { + "epoch": 0.13179542203147354, + "grad_norm": 0.10246269958737887, + "learning_rate": 0.00019461828750269775, + "loss": 0.6789, + "step": 1474 + }, + { + "epoch": 0.13188483547925609, + "grad_norm": 0.11075278331896468, + "learning_rate": 0.00019460891110375977, + "loss": 0.6931, + "step": 1475 + }, + { + "epoch": 0.13197424892703863, + "grad_norm": 0.09845353319170361, + "learning_rate": 0.00019459952677008672, + "loss": 0.6802, + "step": 1476 + }, + { + "epoch": 0.13206366237482117, + "grad_norm": 0.1283463524959499, + "learning_rate": 0.00019459013450246558, + "loss": 0.7085, + "step": 1477 + }, + { + "epoch": 0.13215307582260372, + "grad_norm": 0.1150957413269532, + "learning_rate": 0.0001945807343016841, + "loss": 0.6734, + "step": 1478 + }, + { + "epoch": 0.13224248927038626, + "grad_norm": 0.11206838674272698, + "learning_rate": 0.00019457132616853065, + "loss": 0.6674, + "step": 1479 + }, + { + "epoch": 0.1323319027181688, + "grad_norm": 0.11134923261888223, + "learning_rate": 0.00019456191010379427, + "loss": 0.2613, + "step": 1480 + }, + { + "epoch": 0.13242131616595135, + "grad_norm": 0.11757043138116899, + "learning_rate": 0.00019455248610826474, + "loss": 0.6858, + "step": 1481 + }, + { + "epoch": 0.13251072961373392, + "grad_norm": 0.11073004193918576, + "learning_rate": 0.00019454305418273234, + "loss": 0.7163, + "step": 1482 + }, + { + "epoch": 0.13260014306151646, + "grad_norm": 0.12746627763446436, + "learning_rate": 0.0001945336143279882, + "loss": 0.7015, + "step": 1483 + }, + { + "epoch": 0.132689556509299, + "grad_norm": 0.11575457785885657, + "learning_rate": 0.000194524166544824, + "loss": 0.6894, + "step": 1484 + }, + { + "epoch": 0.13277896995708155, + "grad_norm": 0.14373053164919591, + "learning_rate": 0.00019451471083403209, + "loss": 0.7409, + "step": 1485 + }, + { + "epoch": 0.1328683834048641, + "grad_norm": 0.1278277293237307, + "learning_rate": 0.0001945052471964055, + "loss": 0.6629, + "step": 1486 + }, + { + "epoch": 0.13295779685264664, + "grad_norm": 0.11282883366346426, + "learning_rate": 0.000194495775632738, + "loss": 0.6983, + "step": 1487 + }, + { + "epoch": 0.13304721030042918, + "grad_norm": 0.1270837764464861, + "learning_rate": 0.0001944862961438239, + "loss": 0.2824, + "step": 1488 + }, + { + "epoch": 0.13313662374821172, + "grad_norm": 0.12049933576600101, + "learning_rate": 0.0001944768087304583, + "loss": 0.7043, + "step": 1489 + }, + { + "epoch": 0.13322603719599427, + "grad_norm": 0.1256399869072594, + "learning_rate": 0.0001944673133934368, + "loss": 0.7396, + "step": 1490 + }, + { + "epoch": 0.1333154506437768, + "grad_norm": 0.10541399067960348, + "learning_rate": 0.00019445781013355582, + "loss": 0.7104, + "step": 1491 + }, + { + "epoch": 0.13340486409155938, + "grad_norm": 0.11811088284373562, + "learning_rate": 0.00019444829895161239, + "loss": 0.6763, + "step": 1492 + }, + { + "epoch": 0.13349427753934193, + "grad_norm": 0.1242755134622475, + "learning_rate": 0.0001944387798484042, + "loss": 0.7073, + "step": 1493 + }, + { + "epoch": 0.13358369098712447, + "grad_norm": 0.12105456764720958, + "learning_rate": 0.00019442925282472958, + "loss": 0.6819, + "step": 1494 + }, + { + "epoch": 0.13367310443490701, + "grad_norm": 0.12480183752532904, + "learning_rate": 0.00019441971788138756, + "loss": 0.7083, + "step": 1495 + }, + { + "epoch": 0.13376251788268956, + "grad_norm": 0.12389208478096753, + "learning_rate": 0.00019441017501917784, + "loss": 0.6763, + "step": 1496 + }, + { + "epoch": 0.1338519313304721, + "grad_norm": 0.1205631706793967, + "learning_rate": 0.0001944006242389008, + "loss": 0.7137, + "step": 1497 + }, + { + "epoch": 0.13394134477825465, + "grad_norm": 0.12554935209979773, + "learning_rate": 0.00019439106554135736, + "loss": 0.7245, + "step": 1498 + }, + { + "epoch": 0.1340307582260372, + "grad_norm": 0.12726145327076133, + "learning_rate": 0.00019438149892734926, + "loss": 0.7315, + "step": 1499 + }, + { + "epoch": 0.13412017167381973, + "grad_norm": 0.10874121189217771, + "learning_rate": 0.00019437192439767883, + "loss": 0.6702, + "step": 1500 + }, + { + "epoch": 0.13420958512160228, + "grad_norm": 0.12052542917044717, + "learning_rate": 0.00019436234195314907, + "loss": 0.7212, + "step": 1501 + }, + { + "epoch": 0.13429899856938485, + "grad_norm": 0.12573508403682696, + "learning_rate": 0.00019435275159456364, + "loss": 0.6772, + "step": 1502 + }, + { + "epoch": 0.1343884120171674, + "grad_norm": 0.11733600031444333, + "learning_rate": 0.00019434315332272692, + "loss": 0.701, + "step": 1503 + }, + { + "epoch": 0.13447782546494993, + "grad_norm": 0.12079646252894306, + "learning_rate": 0.00019433354713844386, + "loss": 0.6965, + "step": 1504 + }, + { + "epoch": 0.13456723891273248, + "grad_norm": 0.13611951981901607, + "learning_rate": 0.00019432393304252013, + "loss": 0.7451, + "step": 1505 + }, + { + "epoch": 0.13465665236051502, + "grad_norm": 0.1196660250000599, + "learning_rate": 0.00019431431103576202, + "loss": 0.708, + "step": 1506 + }, + { + "epoch": 0.13474606580829757, + "grad_norm": 0.12279575753068238, + "learning_rate": 0.00019430468111897656, + "loss": 0.7012, + "step": 1507 + }, + { + "epoch": 0.1348354792560801, + "grad_norm": 0.1249116600121028, + "learning_rate": 0.0001942950432929714, + "loss": 0.7255, + "step": 1508 + }, + { + "epoch": 0.13492489270386265, + "grad_norm": 0.09053597739573617, + "learning_rate": 0.00019428539755855483, + "loss": 0.6191, + "step": 1509 + }, + { + "epoch": 0.1350143061516452, + "grad_norm": 0.12581701393600966, + "learning_rate": 0.00019427574391653581, + "loss": 0.7283, + "step": 1510 + }, + { + "epoch": 0.13510371959942774, + "grad_norm": 0.10275362277736566, + "learning_rate": 0.00019426608236772404, + "loss": 0.6988, + "step": 1511 + }, + { + "epoch": 0.1351931330472103, + "grad_norm": 0.10747031266249872, + "learning_rate": 0.00019425641291292978, + "loss": 0.7124, + "step": 1512 + }, + { + "epoch": 0.13528254649499286, + "grad_norm": 0.11601553995635122, + "learning_rate": 0.000194246735552964, + "loss": 0.691, + "step": 1513 + }, + { + "epoch": 0.1353719599427754, + "grad_norm": 0.11627793899750222, + "learning_rate": 0.00019423705028863832, + "loss": 0.7002, + "step": 1514 + }, + { + "epoch": 0.13546137339055794, + "grad_norm": 0.12339512416391978, + "learning_rate": 0.00019422735712076506, + "loss": 0.699, + "step": 1515 + }, + { + "epoch": 0.1355507868383405, + "grad_norm": 0.13101157888793344, + "learning_rate": 0.00019421765605015713, + "loss": 0.6945, + "step": 1516 + }, + { + "epoch": 0.13564020028612303, + "grad_norm": 0.13859378824183455, + "learning_rate": 0.0001942079470776282, + "loss": 0.6882, + "step": 1517 + }, + { + "epoch": 0.13572961373390557, + "grad_norm": 0.1268948424727313, + "learning_rate": 0.0001941982302039925, + "loss": 0.6929, + "step": 1518 + }, + { + "epoch": 0.13581902718168812, + "grad_norm": 0.12307332899456473, + "learning_rate": 0.000194188505430065, + "loss": 0.6804, + "step": 1519 + }, + { + "epoch": 0.13590844062947066, + "grad_norm": 0.11653461124838034, + "learning_rate": 0.0001941787727566613, + "loss": 0.683, + "step": 1520 + }, + { + "epoch": 0.1359978540772532, + "grad_norm": 0.11400851330894964, + "learning_rate": 0.0001941690321845977, + "loss": 0.6629, + "step": 1521 + }, + { + "epoch": 0.13608726752503578, + "grad_norm": 0.12276327922800516, + "learning_rate": 0.00019415928371469105, + "loss": 0.7062, + "step": 1522 + }, + { + "epoch": 0.13617668097281832, + "grad_norm": 0.12162563180017896, + "learning_rate": 0.000194149527347759, + "loss": 0.6808, + "step": 1523 + }, + { + "epoch": 0.13626609442060086, + "grad_norm": 0.12517922487458535, + "learning_rate": 0.00019413976308461982, + "loss": 0.6668, + "step": 1524 + }, + { + "epoch": 0.1363555078683834, + "grad_norm": 0.1222171307674075, + "learning_rate": 0.0001941299909260924, + "loss": 0.6898, + "step": 1525 + }, + { + "epoch": 0.13644492131616595, + "grad_norm": 0.12042433652030107, + "learning_rate": 0.0001941202108729963, + "loss": 0.7263, + "step": 1526 + }, + { + "epoch": 0.1365343347639485, + "grad_norm": 0.13041694173759696, + "learning_rate": 0.0001941104229261518, + "loss": 0.7759, + "step": 1527 + }, + { + "epoch": 0.13662374821173104, + "grad_norm": 0.11589947284074426, + "learning_rate": 0.0001941006270863798, + "loss": 0.6582, + "step": 1528 + }, + { + "epoch": 0.13671316165951358, + "grad_norm": 0.12461084280541163, + "learning_rate": 0.0001940908233545018, + "loss": 0.6949, + "step": 1529 + }, + { + "epoch": 0.13680257510729613, + "grad_norm": 0.12550975227708494, + "learning_rate": 0.00019408101173134013, + "loss": 0.6744, + "step": 1530 + }, + { + "epoch": 0.1368919885550787, + "grad_norm": 0.12548525437091054, + "learning_rate": 0.00019407119221771758, + "loss": 0.7442, + "step": 1531 + }, + { + "epoch": 0.13698140200286124, + "grad_norm": 0.12094338532002966, + "learning_rate": 0.00019406136481445782, + "loss": 0.6602, + "step": 1532 + }, + { + "epoch": 0.13707081545064378, + "grad_norm": 0.11862299989986812, + "learning_rate": 0.0001940515295223849, + "loss": 0.6718, + "step": 1533 + }, + { + "epoch": 0.13716022889842633, + "grad_norm": 0.12680695663557542, + "learning_rate": 0.00019404168634232382, + "loss": 0.7255, + "step": 1534 + }, + { + "epoch": 0.13724964234620887, + "grad_norm": 0.10722808382475126, + "learning_rate": 0.0001940318352751001, + "loss": 0.6963, + "step": 1535 + }, + { + "epoch": 0.13733905579399142, + "grad_norm": 0.12654847488240686, + "learning_rate": 0.00019402197632153992, + "loss": 0.681, + "step": 1536 + }, + { + "epoch": 0.13742846924177396, + "grad_norm": 0.13029586598455895, + "learning_rate": 0.0001940121094824701, + "loss": 0.7093, + "step": 1537 + }, + { + "epoch": 0.1375178826895565, + "grad_norm": 0.1307313456849636, + "learning_rate": 0.00019400223475871825, + "loss": 0.7193, + "step": 1538 + }, + { + "epoch": 0.13760729613733905, + "grad_norm": 0.09989620672357703, + "learning_rate": 0.00019399235215111245, + "loss": 0.666, + "step": 1539 + }, + { + "epoch": 0.1376967095851216, + "grad_norm": 0.12843918241245192, + "learning_rate": 0.00019398246166048159, + "loss": 0.7041, + "step": 1540 + }, + { + "epoch": 0.13778612303290416, + "grad_norm": 0.10824084663381169, + "learning_rate": 0.00019397256328765517, + "loss": 0.6904, + "step": 1541 + }, + { + "epoch": 0.1378755364806867, + "grad_norm": 0.09992354195849573, + "learning_rate": 0.00019396265703346339, + "loss": 0.6568, + "step": 1542 + }, + { + "epoch": 0.13796494992846925, + "grad_norm": 0.11432648750964643, + "learning_rate": 0.00019395274289873705, + "loss": 0.6902, + "step": 1543 + }, + { + "epoch": 0.1380543633762518, + "grad_norm": 0.10448557474864632, + "learning_rate": 0.00019394282088430758, + "loss": 0.6588, + "step": 1544 + }, + { + "epoch": 0.13814377682403434, + "grad_norm": 0.11721762271273022, + "learning_rate": 0.0001939328909910072, + "loss": 0.6945, + "step": 1545 + }, + { + "epoch": 0.13823319027181688, + "grad_norm": 0.11258434560141939, + "learning_rate": 0.0001939229532196687, + "loss": 0.6719, + "step": 1546 + }, + { + "epoch": 0.13832260371959942, + "grad_norm": 0.11385578357803208, + "learning_rate": 0.00019391300757112557, + "loss": 0.7232, + "step": 1547 + }, + { + "epoch": 0.13841201716738197, + "grad_norm": 0.1276591440541223, + "learning_rate": 0.00019390305404621186, + "loss": 0.7185, + "step": 1548 + }, + { + "epoch": 0.1385014306151645, + "grad_norm": 0.13121729479634028, + "learning_rate": 0.00019389309264576242, + "loss": 0.7209, + "step": 1549 + }, + { + "epoch": 0.13859084406294706, + "grad_norm": 0.12536678397237028, + "learning_rate": 0.00019388312337061274, + "loss": 0.6711, + "step": 1550 + }, + { + "epoch": 0.13868025751072963, + "grad_norm": 0.10622010132973222, + "learning_rate": 0.00019387314622159885, + "loss": 0.6953, + "step": 1551 + }, + { + "epoch": 0.13876967095851217, + "grad_norm": 0.11585834964532364, + "learning_rate": 0.00019386316119955756, + "loss": 0.6355, + "step": 1552 + }, + { + "epoch": 0.1388590844062947, + "grad_norm": 0.10847660581354016, + "learning_rate": 0.0001938531683053263, + "loss": 0.6969, + "step": 1553 + }, + { + "epoch": 0.13894849785407726, + "grad_norm": 0.1187265652057057, + "learning_rate": 0.00019384316753974314, + "loss": 0.7116, + "step": 1554 + }, + { + "epoch": 0.1390379113018598, + "grad_norm": 0.13294935575835393, + "learning_rate": 0.00019383315890364689, + "loss": 0.7416, + "step": 1555 + }, + { + "epoch": 0.13912732474964234, + "grad_norm": 0.12716626955435048, + "learning_rate": 0.00019382314239787691, + "loss": 0.6599, + "step": 1556 + }, + { + "epoch": 0.1392167381974249, + "grad_norm": 0.12389740340706308, + "learning_rate": 0.00019381311802327327, + "loss": 0.6756, + "step": 1557 + }, + { + "epoch": 0.13930615164520743, + "grad_norm": 0.11830820378956655, + "learning_rate": 0.00019380308578067674, + "loss": 0.7126, + "step": 1558 + }, + { + "epoch": 0.13939556509298998, + "grad_norm": 0.13090599927506003, + "learning_rate": 0.00019379304567092867, + "loss": 0.7095, + "step": 1559 + }, + { + "epoch": 0.13948497854077252, + "grad_norm": 0.12988305236771172, + "learning_rate": 0.00019378299769487117, + "loss": 0.6841, + "step": 1560 + }, + { + "epoch": 0.1395743919885551, + "grad_norm": 0.12351121497197795, + "learning_rate": 0.0001937729418533469, + "loss": 0.6961, + "step": 1561 + }, + { + "epoch": 0.13966380543633763, + "grad_norm": 0.10521621790797256, + "learning_rate": 0.0001937628781471992, + "loss": 0.6748, + "step": 1562 + }, + { + "epoch": 0.13975321888412018, + "grad_norm": 0.11550501442683482, + "learning_rate": 0.0001937528065772722, + "loss": 0.7451, + "step": 1563 + }, + { + "epoch": 0.13984263233190272, + "grad_norm": 0.10706180095865842, + "learning_rate": 0.0001937427271444105, + "loss": 0.6531, + "step": 1564 + }, + { + "epoch": 0.13993204577968527, + "grad_norm": 0.12163227489377014, + "learning_rate": 0.00019373263984945953, + "loss": 0.7268, + "step": 1565 + }, + { + "epoch": 0.1400214592274678, + "grad_norm": 0.11079739006764149, + "learning_rate": 0.00019372254469326522, + "loss": 0.6736, + "step": 1566 + }, + { + "epoch": 0.14011087267525035, + "grad_norm": 0.126110339402405, + "learning_rate": 0.0001937124416766743, + "loss": 0.7104, + "step": 1567 + }, + { + "epoch": 0.1402002861230329, + "grad_norm": 0.13574278409029766, + "learning_rate": 0.00019370233080053407, + "loss": 0.7075, + "step": 1568 + }, + { + "epoch": 0.14028969957081544, + "grad_norm": 0.13478711528094103, + "learning_rate": 0.0001936922120656925, + "loss": 0.6897, + "step": 1569 + }, + { + "epoch": 0.14037911301859798, + "grad_norm": 0.1229593548734267, + "learning_rate": 0.00019368208547299826, + "loss": 0.6885, + "step": 1570 + }, + { + "epoch": 0.14046852646638056, + "grad_norm": 0.12141054183940989, + "learning_rate": 0.00019367195102330066, + "loss": 0.7095, + "step": 1571 + }, + { + "epoch": 0.1405579399141631, + "grad_norm": 0.12165256080894007, + "learning_rate": 0.00019366180871744964, + "loss": 0.6616, + "step": 1572 + }, + { + "epoch": 0.14064735336194564, + "grad_norm": 0.11782220861458287, + "learning_rate": 0.00019365165855629587, + "loss": 0.6789, + "step": 1573 + }, + { + "epoch": 0.1407367668097282, + "grad_norm": 0.12900811952932487, + "learning_rate": 0.00019364150054069059, + "loss": 0.6631, + "step": 1574 + }, + { + "epoch": 0.14082618025751073, + "grad_norm": 0.11388722711975893, + "learning_rate": 0.00019363133467148572, + "loss": 0.689, + "step": 1575 + }, + { + "epoch": 0.14091559370529327, + "grad_norm": 0.10764705091522628, + "learning_rate": 0.00019362116094953391, + "loss": 0.6746, + "step": 1576 + }, + { + "epoch": 0.14100500715307582, + "grad_norm": 0.10879213229596306, + "learning_rate": 0.0001936109793756884, + "loss": 0.6788, + "step": 1577 + }, + { + "epoch": 0.14109442060085836, + "grad_norm": 0.13174185837941063, + "learning_rate": 0.00019360078995080308, + "loss": 0.7154, + "step": 1578 + }, + { + "epoch": 0.1411838340486409, + "grad_norm": 0.11928993451613822, + "learning_rate": 0.0001935905926757326, + "loss": 0.7199, + "step": 1579 + }, + { + "epoch": 0.14127324749642345, + "grad_norm": 0.1129537905192582, + "learning_rate": 0.0001935803875513321, + "loss": 0.6624, + "step": 1580 + }, + { + "epoch": 0.14136266094420602, + "grad_norm": 0.13763555448490608, + "learning_rate": 0.0001935701745784575, + "loss": 0.7609, + "step": 1581 + }, + { + "epoch": 0.14145207439198856, + "grad_norm": 0.11849902042545403, + "learning_rate": 0.0001935599537579654, + "loss": 0.6849, + "step": 1582 + }, + { + "epoch": 0.1415414878397711, + "grad_norm": 0.12640660890147223, + "learning_rate": 0.00019354972509071295, + "loss": 0.7162, + "step": 1583 + }, + { + "epoch": 0.14163090128755365, + "grad_norm": 0.10585464781863359, + "learning_rate": 0.00019353948857755803, + "loss": 0.6886, + "step": 1584 + }, + { + "epoch": 0.1417203147353362, + "grad_norm": 0.11768013037371369, + "learning_rate": 0.00019352924421935916, + "loss": 0.6819, + "step": 1585 + }, + { + "epoch": 0.14180972818311874, + "grad_norm": 0.09851011921521935, + "learning_rate": 0.00019351899201697556, + "loss": 0.6783, + "step": 1586 + }, + { + "epoch": 0.14189914163090128, + "grad_norm": 0.11369548150133794, + "learning_rate": 0.00019350873197126705, + "loss": 0.6911, + "step": 1587 + }, + { + "epoch": 0.14198855507868383, + "grad_norm": 0.11141500842457122, + "learning_rate": 0.0001934984640830941, + "loss": 0.6943, + "step": 1588 + }, + { + "epoch": 0.14207796852646637, + "grad_norm": 0.10848267738726837, + "learning_rate": 0.00019348818835331788, + "loss": 0.6647, + "step": 1589 + }, + { + "epoch": 0.14216738197424894, + "grad_norm": 0.10720252451397826, + "learning_rate": 0.0001934779047828002, + "loss": 0.7106, + "step": 1590 + }, + { + "epoch": 0.14225679542203148, + "grad_norm": 0.11969500998994731, + "learning_rate": 0.00019346761337240355, + "loss": 0.7162, + "step": 1591 + }, + { + "epoch": 0.14234620886981403, + "grad_norm": 0.12412250770314144, + "learning_rate": 0.00019345731412299106, + "loss": 0.6848, + "step": 1592 + }, + { + "epoch": 0.14243562231759657, + "grad_norm": 0.09933117564510505, + "learning_rate": 0.0001934470070354265, + "loss": 0.6663, + "step": 1593 + }, + { + "epoch": 0.14252503576537912, + "grad_norm": 0.12673493645148226, + "learning_rate": 0.00019343669211057432, + "loss": 0.7577, + "step": 1594 + }, + { + "epoch": 0.14261444921316166, + "grad_norm": 0.10584181204926343, + "learning_rate": 0.00019342636934929959, + "loss": 0.6388, + "step": 1595 + }, + { + "epoch": 0.1427038626609442, + "grad_norm": 0.11950001946684302, + "learning_rate": 0.0001934160387524681, + "loss": 0.7182, + "step": 1596 + }, + { + "epoch": 0.14279327610872675, + "grad_norm": 0.11754517767435674, + "learning_rate": 0.00019340570032094626, + "loss": 0.6456, + "step": 1597 + }, + { + "epoch": 0.1428826895565093, + "grad_norm": 0.12052964028678, + "learning_rate": 0.00019339535405560115, + "loss": 0.706, + "step": 1598 + }, + { + "epoch": 0.14297210300429183, + "grad_norm": 0.13497113984233836, + "learning_rate": 0.00019338499995730048, + "loss": 0.7207, + "step": 1599 + }, + { + "epoch": 0.1430615164520744, + "grad_norm": 0.13086809808736038, + "learning_rate": 0.00019337463802691264, + "loss": 0.7234, + "step": 1600 + }, + { + "epoch": 0.14315092989985695, + "grad_norm": 0.13863554109404802, + "learning_rate": 0.00019336426826530668, + "loss": 0.7194, + "step": 1601 + }, + { + "epoch": 0.1432403433476395, + "grad_norm": 0.1413144739871027, + "learning_rate": 0.0001933538906733523, + "loss": 0.7037, + "step": 1602 + }, + { + "epoch": 0.14332975679542204, + "grad_norm": 0.11300331310944246, + "learning_rate": 0.00019334350525191987, + "loss": 0.6373, + "step": 1603 + }, + { + "epoch": 0.14341917024320458, + "grad_norm": 0.12803483048998365, + "learning_rate": 0.00019333311200188036, + "loss": 0.6885, + "step": 1604 + }, + { + "epoch": 0.14350858369098712, + "grad_norm": 0.12739354076448725, + "learning_rate": 0.00019332271092410545, + "loss": 0.7377, + "step": 1605 + }, + { + "epoch": 0.14359799713876967, + "grad_norm": 0.1202918669605036, + "learning_rate": 0.0001933123020194675, + "loss": 0.6936, + "step": 1606 + }, + { + "epoch": 0.1436874105865522, + "grad_norm": 0.11796961280767333, + "learning_rate": 0.00019330188528883947, + "loss": 0.714, + "step": 1607 + }, + { + "epoch": 0.14377682403433475, + "grad_norm": 0.1278214137139425, + "learning_rate": 0.00019329146073309504, + "loss": 0.708, + "step": 1608 + }, + { + "epoch": 0.1438662374821173, + "grad_norm": 0.0964961214591868, + "learning_rate": 0.00019328102835310842, + "loss": 0.6653, + "step": 1609 + }, + { + "epoch": 0.14395565092989987, + "grad_norm": 0.1261679898942874, + "learning_rate": 0.00019327058814975462, + "loss": 0.7015, + "step": 1610 + }, + { + "epoch": 0.1440450643776824, + "grad_norm": 0.1195820906895268, + "learning_rate": 0.00019326014012390922, + "loss": 0.7042, + "step": 1611 + }, + { + "epoch": 0.14413447782546496, + "grad_norm": 0.11555882571793086, + "learning_rate": 0.00019324968427644848, + "loss": 0.7018, + "step": 1612 + }, + { + "epoch": 0.1442238912732475, + "grad_norm": 0.12388427729173736, + "learning_rate": 0.00019323922060824939, + "loss": 0.6912, + "step": 1613 + }, + { + "epoch": 0.14431330472103004, + "grad_norm": 0.11546395514825804, + "learning_rate": 0.00019322874912018945, + "loss": 0.6895, + "step": 1614 + }, + { + "epoch": 0.1444027181688126, + "grad_norm": 0.11479497798688773, + "learning_rate": 0.00019321826981314691, + "loss": 0.6904, + "step": 1615 + }, + { + "epoch": 0.14449213161659513, + "grad_norm": 0.11638219018493944, + "learning_rate": 0.00019320778268800066, + "loss": 0.6821, + "step": 1616 + }, + { + "epoch": 0.14458154506437768, + "grad_norm": 0.1175648401924945, + "learning_rate": 0.00019319728774563023, + "loss": 0.6962, + "step": 1617 + }, + { + "epoch": 0.14467095851216022, + "grad_norm": 0.11219002544064562, + "learning_rate": 0.00019318678498691586, + "loss": 0.692, + "step": 1618 + }, + { + "epoch": 0.14476037195994276, + "grad_norm": 0.1308943991119446, + "learning_rate": 0.00019317627441273836, + "loss": 0.7345, + "step": 1619 + }, + { + "epoch": 0.14484978540772533, + "grad_norm": 0.14034170667559237, + "learning_rate": 0.00019316575602397923, + "loss": 0.6846, + "step": 1620 + }, + { + "epoch": 0.14493919885550788, + "grad_norm": 0.12571462822255072, + "learning_rate": 0.0001931552298215207, + "loss": 0.7159, + "step": 1621 + }, + { + "epoch": 0.14502861230329042, + "grad_norm": 0.10677807210153191, + "learning_rate": 0.0001931446958062455, + "loss": 0.6641, + "step": 1622 + }, + { + "epoch": 0.14511802575107297, + "grad_norm": 0.1356729542103235, + "learning_rate": 0.0001931341539790372, + "loss": 0.7339, + "step": 1623 + }, + { + "epoch": 0.1452074391988555, + "grad_norm": 0.12124273023700187, + "learning_rate": 0.00019312360434077985, + "loss": 0.7199, + "step": 1624 + }, + { + "epoch": 0.14529685264663805, + "grad_norm": 0.09739745997920558, + "learning_rate": 0.0001931130468923583, + "loss": 0.6418, + "step": 1625 + }, + { + "epoch": 0.1453862660944206, + "grad_norm": 0.13602249328858715, + "learning_rate": 0.00019310248163465795, + "loss": 0.7103, + "step": 1626 + }, + { + "epoch": 0.14547567954220314, + "grad_norm": 0.12279277862571172, + "learning_rate": 0.00019309190856856486, + "loss": 0.713, + "step": 1627 + }, + { + "epoch": 0.14556509298998568, + "grad_norm": 0.12196879766876602, + "learning_rate": 0.0001930813276949659, + "loss": 0.688, + "step": 1628 + }, + { + "epoch": 0.14565450643776823, + "grad_norm": 0.12679308207863135, + "learning_rate": 0.00019307073901474834, + "loss": 0.7247, + "step": 1629 + }, + { + "epoch": 0.1457439198855508, + "grad_norm": 0.13392121473732374, + "learning_rate": 0.00019306014252880034, + "loss": 0.7245, + "step": 1630 + }, + { + "epoch": 0.14583333333333334, + "grad_norm": 0.10350631795440726, + "learning_rate": 0.00019304953823801055, + "loss": 0.699, + "step": 1631 + }, + { + "epoch": 0.1459227467811159, + "grad_norm": 0.1096179659741716, + "learning_rate": 0.00019303892614326836, + "loss": 0.692, + "step": 1632 + }, + { + "epoch": 0.14601216022889843, + "grad_norm": 0.12183060423652922, + "learning_rate": 0.0001930283062454638, + "loss": 0.6769, + "step": 1633 + }, + { + "epoch": 0.14610157367668097, + "grad_norm": 0.1125051296293097, + "learning_rate": 0.00019301767854548756, + "loss": 0.6791, + "step": 1634 + }, + { + "epoch": 0.14619098712446352, + "grad_norm": 0.116328704128805, + "learning_rate": 0.00019300704304423094, + "loss": 0.6782, + "step": 1635 + }, + { + "epoch": 0.14628040057224606, + "grad_norm": 0.12035577690460862, + "learning_rate": 0.00019299639974258598, + "loss": 0.6711, + "step": 1636 + }, + { + "epoch": 0.1463698140200286, + "grad_norm": 0.12169500103775334, + "learning_rate": 0.00019298574864144523, + "loss": 0.7103, + "step": 1637 + }, + { + "epoch": 0.14645922746781115, + "grad_norm": 0.1145178883209945, + "learning_rate": 0.00019297508974170207, + "loss": 0.6608, + "step": 1638 + }, + { + "epoch": 0.1465486409155937, + "grad_norm": 0.10324382668756428, + "learning_rate": 0.0001929644230442504, + "loss": 0.6315, + "step": 1639 + }, + { + "epoch": 0.14663805436337626, + "grad_norm": 0.12172567703185704, + "learning_rate": 0.00019295374854998488, + "loss": 0.713, + "step": 1640 + }, + { + "epoch": 0.1467274678111588, + "grad_norm": 0.1264635558706153, + "learning_rate": 0.0001929430662598007, + "loss": 0.6838, + "step": 1641 + }, + { + "epoch": 0.14681688125894135, + "grad_norm": 0.10579027563923, + "learning_rate": 0.00019293237617459382, + "loss": 0.6584, + "step": 1642 + }, + { + "epoch": 0.1469062947067239, + "grad_norm": 0.1135044548400072, + "learning_rate": 0.00019292167829526076, + "loss": 0.7102, + "step": 1643 + }, + { + "epoch": 0.14699570815450644, + "grad_norm": 0.12494384482397522, + "learning_rate": 0.00019291097262269874, + "loss": 0.7105, + "step": 1644 + }, + { + "epoch": 0.14708512160228898, + "grad_norm": 0.10719700393074186, + "learning_rate": 0.0001929002591578057, + "loss": 0.6849, + "step": 1645 + }, + { + "epoch": 0.14717453505007153, + "grad_norm": 0.1232143742962914, + "learning_rate": 0.00019288953790148013, + "loss": 0.7371, + "step": 1646 + }, + { + "epoch": 0.14726394849785407, + "grad_norm": 0.11048201401068539, + "learning_rate": 0.00019287880885462115, + "loss": 0.674, + "step": 1647 + }, + { + "epoch": 0.1473533619456366, + "grad_norm": 0.11037686003662409, + "learning_rate": 0.00019286807201812867, + "loss": 0.6738, + "step": 1648 + }, + { + "epoch": 0.14744277539341916, + "grad_norm": 0.11767679445753639, + "learning_rate": 0.00019285732739290315, + "loss": 0.684, + "step": 1649 + }, + { + "epoch": 0.14753218884120173, + "grad_norm": 0.1076601413764181, + "learning_rate": 0.0001928465749798457, + "loss": 0.699, + "step": 1650 + }, + { + "epoch": 0.14762160228898427, + "grad_norm": 0.10951208629598043, + "learning_rate": 0.00019283581477985817, + "loss": 0.686, + "step": 1651 + }, + { + "epoch": 0.14771101573676682, + "grad_norm": 0.11688379627506593, + "learning_rate": 0.00019282504679384293, + "loss": 0.7068, + "step": 1652 + }, + { + "epoch": 0.14780042918454936, + "grad_norm": 0.12128497431333228, + "learning_rate": 0.00019281427102270314, + "loss": 0.6873, + "step": 1653 + }, + { + "epoch": 0.1478898426323319, + "grad_norm": 0.10883530997117362, + "learning_rate": 0.00019280348746734255, + "loss": 0.6804, + "step": 1654 + }, + { + "epoch": 0.14797925608011445, + "grad_norm": 0.12198670233204478, + "learning_rate": 0.00019279269612866554, + "loss": 0.7052, + "step": 1655 + }, + { + "epoch": 0.148068669527897, + "grad_norm": 0.12317062714991397, + "learning_rate": 0.00019278189700757715, + "loss": 0.6598, + "step": 1656 + }, + { + "epoch": 0.14815808297567953, + "grad_norm": 0.11974861352587242, + "learning_rate": 0.0001927710901049831, + "loss": 0.7154, + "step": 1657 + }, + { + "epoch": 0.14824749642346208, + "grad_norm": 0.12409049204631967, + "learning_rate": 0.00019276027542178978, + "loss": 0.7511, + "step": 1658 + }, + { + "epoch": 0.14833690987124465, + "grad_norm": 0.1356630219450362, + "learning_rate": 0.0001927494529589042, + "loss": 0.6865, + "step": 1659 + }, + { + "epoch": 0.1484263233190272, + "grad_norm": 0.10741228943421707, + "learning_rate": 0.000192738622717234, + "loss": 0.6612, + "step": 1660 + }, + { + "epoch": 0.14851573676680974, + "grad_norm": 0.1287440752767585, + "learning_rate": 0.0001927277846976875, + "loss": 0.705, + "step": 1661 + }, + { + "epoch": 0.14860515021459228, + "grad_norm": 0.12337365710458456, + "learning_rate": 0.00019271693890117372, + "loss": 0.6926, + "step": 1662 + }, + { + "epoch": 0.14869456366237482, + "grad_norm": 0.11095939355068538, + "learning_rate": 0.00019270608532860224, + "loss": 0.7143, + "step": 1663 + }, + { + "epoch": 0.14878397711015737, + "grad_norm": 0.11667482210972342, + "learning_rate": 0.00019269522398088332, + "loss": 0.6894, + "step": 1664 + }, + { + "epoch": 0.1488733905579399, + "grad_norm": 0.11454351845917907, + "learning_rate": 0.0001926843548589279, + "loss": 0.7027, + "step": 1665 + }, + { + "epoch": 0.14896280400572245, + "grad_norm": 0.12474200593992414, + "learning_rate": 0.0001926734779636476, + "loss": 0.6895, + "step": 1666 + }, + { + "epoch": 0.149052217453505, + "grad_norm": 0.10599476562341384, + "learning_rate": 0.00019266259329595462, + "loss": 0.6717, + "step": 1667 + }, + { + "epoch": 0.14914163090128754, + "grad_norm": 0.10713445979801794, + "learning_rate": 0.00019265170085676185, + "loss": 0.6781, + "step": 1668 + }, + { + "epoch": 0.1492310443490701, + "grad_norm": 0.1119414028955798, + "learning_rate": 0.00019264080064698282, + "loss": 0.691, + "step": 1669 + }, + { + "epoch": 0.14932045779685266, + "grad_norm": 0.12516766434361415, + "learning_rate": 0.00019262989266753173, + "loss": 0.7058, + "step": 1670 + }, + { + "epoch": 0.1494098712446352, + "grad_norm": 0.11410480201387796, + "learning_rate": 0.0001926189769193234, + "loss": 0.6823, + "step": 1671 + }, + { + "epoch": 0.14949928469241774, + "grad_norm": 0.1355059483576872, + "learning_rate": 0.00019260805340327335, + "loss": 0.3053, + "step": 1672 + }, + { + "epoch": 0.1495886981402003, + "grad_norm": 0.12610797034314136, + "learning_rate": 0.00019259712212029765, + "loss": 0.7165, + "step": 1673 + }, + { + "epoch": 0.14967811158798283, + "grad_norm": 0.11470101863357544, + "learning_rate": 0.0001925861830713132, + "loss": 0.6268, + "step": 1674 + }, + { + "epoch": 0.14976752503576538, + "grad_norm": 0.11949407015992508, + "learning_rate": 0.00019257523625723736, + "loss": 0.7339, + "step": 1675 + }, + { + "epoch": 0.14985693848354792, + "grad_norm": 0.12995640762587524, + "learning_rate": 0.0001925642816789883, + "loss": 0.7245, + "step": 1676 + }, + { + "epoch": 0.14994635193133046, + "grad_norm": 0.13090940791089656, + "learning_rate": 0.00019255331933748472, + "loss": 0.6882, + "step": 1677 + }, + { + "epoch": 0.150035765379113, + "grad_norm": 0.12470721380886783, + "learning_rate": 0.000192542349233646, + "loss": 0.7236, + "step": 1678 + }, + { + "epoch": 0.15012517882689558, + "grad_norm": 0.11505445400405362, + "learning_rate": 0.0001925313713683922, + "loss": 0.7066, + "step": 1679 + }, + { + "epoch": 0.15021459227467812, + "grad_norm": 0.11255768228760338, + "learning_rate": 0.00019252038574264405, + "loss": 0.6739, + "step": 1680 + }, + { + "epoch": 0.15030400572246067, + "grad_norm": 0.11996337337574427, + "learning_rate": 0.00019250939235732287, + "loss": 0.6607, + "step": 1681 + }, + { + "epoch": 0.1503934191702432, + "grad_norm": 0.12036149966562121, + "learning_rate": 0.00019249839121335068, + "loss": 0.6585, + "step": 1682 + }, + { + "epoch": 0.15048283261802575, + "grad_norm": 0.1117484713438509, + "learning_rate": 0.00019248738231165017, + "loss": 0.6549, + "step": 1683 + }, + { + "epoch": 0.1505722460658083, + "grad_norm": 0.1340257772559842, + "learning_rate": 0.00019247636565314453, + "loss": 0.7378, + "step": 1684 + }, + { + "epoch": 0.15066165951359084, + "grad_norm": 0.11936269312568466, + "learning_rate": 0.00019246534123875783, + "loss": 0.6738, + "step": 1685 + }, + { + "epoch": 0.15075107296137338, + "grad_norm": 0.12245812777229599, + "learning_rate": 0.00019245430906941464, + "loss": 0.6801, + "step": 1686 + }, + { + "epoch": 0.15084048640915593, + "grad_norm": 0.11171329109591985, + "learning_rate": 0.00019244326914604019, + "loss": 0.661, + "step": 1687 + }, + { + "epoch": 0.15092989985693847, + "grad_norm": 0.1096132486319703, + "learning_rate": 0.00019243222146956039, + "loss": 0.6594, + "step": 1688 + }, + { + "epoch": 0.15101931330472104, + "grad_norm": 0.11571405316446248, + "learning_rate": 0.0001924211660409018, + "loss": 0.6767, + "step": 1689 + }, + { + "epoch": 0.15110872675250359, + "grad_norm": 0.11546186677244163, + "learning_rate": 0.00019241010286099165, + "loss": 0.6797, + "step": 1690 + }, + { + "epoch": 0.15119814020028613, + "grad_norm": 0.11373129856907858, + "learning_rate": 0.00019239903193075776, + "loss": 0.6686, + "step": 1691 + }, + { + "epoch": 0.15128755364806867, + "grad_norm": 0.13793610298123565, + "learning_rate": 0.0001923879532511287, + "loss": 0.7416, + "step": 1692 + }, + { + "epoch": 0.15137696709585122, + "grad_norm": 0.17260341431214543, + "learning_rate": 0.0001923768668230335, + "loss": 0.2463, + "step": 1693 + }, + { + "epoch": 0.15146638054363376, + "grad_norm": 0.11562643993172193, + "learning_rate": 0.0001923657726474021, + "loss": 0.7055, + "step": 1694 + }, + { + "epoch": 0.1515557939914163, + "grad_norm": 0.1294526355705112, + "learning_rate": 0.00019235467072516488, + "loss": 0.7254, + "step": 1695 + }, + { + "epoch": 0.15164520743919885, + "grad_norm": 0.108914247388925, + "learning_rate": 0.00019234356105725297, + "loss": 0.6622, + "step": 1696 + }, + { + "epoch": 0.1517346208869814, + "grad_norm": 0.12091984918160036, + "learning_rate": 0.00019233244364459814, + "loss": 0.6894, + "step": 1697 + }, + { + "epoch": 0.15182403433476394, + "grad_norm": 0.12178549212756597, + "learning_rate": 0.00019232131848813272, + "loss": 0.6663, + "step": 1698 + }, + { + "epoch": 0.1519134477825465, + "grad_norm": 0.12296952670911442, + "learning_rate": 0.00019231018558878984, + "loss": 0.718, + "step": 1699 + }, + { + "epoch": 0.15200286123032905, + "grad_norm": 0.1213129385082206, + "learning_rate": 0.00019229904494750315, + "loss": 0.7076, + "step": 1700 + }, + { + "epoch": 0.1520922746781116, + "grad_norm": 0.11891326658841689, + "learning_rate": 0.00019228789656520708, + "loss": 0.7299, + "step": 1701 + }, + { + "epoch": 0.15218168812589414, + "grad_norm": 0.11622469991113143, + "learning_rate": 0.00019227674044283653, + "loss": 0.6628, + "step": 1702 + }, + { + "epoch": 0.15227110157367668, + "grad_norm": 0.11519318466862243, + "learning_rate": 0.00019226557658132723, + "loss": 0.6713, + "step": 1703 + }, + { + "epoch": 0.15236051502145923, + "grad_norm": 0.10641347078131391, + "learning_rate": 0.00019225440498161546, + "loss": 0.6711, + "step": 1704 + }, + { + "epoch": 0.15244992846924177, + "grad_norm": 0.13109106216027816, + "learning_rate": 0.00019224322564463813, + "loss": 0.7653, + "step": 1705 + }, + { + "epoch": 0.1525393419170243, + "grad_norm": 0.10974365993088024, + "learning_rate": 0.00019223203857133287, + "loss": 0.6765, + "step": 1706 + }, + { + "epoch": 0.15262875536480686, + "grad_norm": 0.10804965364602807, + "learning_rate": 0.00019222084376263794, + "loss": 0.6636, + "step": 1707 + }, + { + "epoch": 0.1527181688125894, + "grad_norm": 0.131597254870936, + "learning_rate": 0.0001922096412194922, + "loss": 0.7331, + "step": 1708 + }, + { + "epoch": 0.15280758226037197, + "grad_norm": 0.1112032121102995, + "learning_rate": 0.00019219843094283524, + "loss": 0.6463, + "step": 1709 + }, + { + "epoch": 0.15289699570815452, + "grad_norm": 0.13761098864977525, + "learning_rate": 0.00019218721293360718, + "loss": 0.724, + "step": 1710 + }, + { + "epoch": 0.15298640915593706, + "grad_norm": 0.12612121759108783, + "learning_rate": 0.00019217598719274896, + "loss": 0.679, + "step": 1711 + }, + { + "epoch": 0.1530758226037196, + "grad_norm": 0.10720358860665527, + "learning_rate": 0.00019216475372120197, + "loss": 0.6621, + "step": 1712 + }, + { + "epoch": 0.15316523605150215, + "grad_norm": 0.19205521322431504, + "learning_rate": 0.0001921535125199084, + "loss": 0.2835, + "step": 1713 + }, + { + "epoch": 0.1532546494992847, + "grad_norm": 0.13347814348229772, + "learning_rate": 0.00019214226358981105, + "loss": 0.7248, + "step": 1714 + }, + { + "epoch": 0.15334406294706723, + "grad_norm": 0.11617093861880455, + "learning_rate": 0.00019213100693185332, + "loss": 0.7028, + "step": 1715 + }, + { + "epoch": 0.15343347639484978, + "grad_norm": 0.11119316488428564, + "learning_rate": 0.00019211974254697932, + "loss": 0.6771, + "step": 1716 + }, + { + "epoch": 0.15352288984263232, + "grad_norm": 0.13155184566302336, + "learning_rate": 0.00019210847043613373, + "loss": 0.7101, + "step": 1717 + }, + { + "epoch": 0.1536123032904149, + "grad_norm": 0.12359289199904877, + "learning_rate": 0.000192097190600262, + "loss": 0.6912, + "step": 1718 + }, + { + "epoch": 0.15370171673819744, + "grad_norm": 0.11020017669869636, + "learning_rate": 0.0001920859030403101, + "loss": 0.6835, + "step": 1719 + }, + { + "epoch": 0.15379113018597998, + "grad_norm": 0.11856644592535787, + "learning_rate": 0.00019207460775722473, + "loss": 0.6836, + "step": 1720 + }, + { + "epoch": 0.15388054363376252, + "grad_norm": 0.12135155599414783, + "learning_rate": 0.00019206330475195319, + "loss": 0.6776, + "step": 1721 + }, + { + "epoch": 0.15396995708154507, + "grad_norm": 0.10846623642564485, + "learning_rate": 0.0001920519940254435, + "loss": 0.6967, + "step": 1722 + }, + { + "epoch": 0.1540593705293276, + "grad_norm": 0.13119697223504206, + "learning_rate": 0.0001920406755786442, + "loss": 0.2968, + "step": 1723 + }, + { + "epoch": 0.15414878397711015, + "grad_norm": 0.10989356193805411, + "learning_rate": 0.0001920293494125046, + "loss": 0.6614, + "step": 1724 + }, + { + "epoch": 0.1542381974248927, + "grad_norm": 0.12127329910966457, + "learning_rate": 0.00019201801552797462, + "loss": 0.7131, + "step": 1725 + }, + { + "epoch": 0.15432761087267524, + "grad_norm": 0.11952835322146453, + "learning_rate": 0.0001920066739260048, + "loss": 0.6517, + "step": 1726 + }, + { + "epoch": 0.15441702432045779, + "grad_norm": 0.12546668288121446, + "learning_rate": 0.0001919953246075464, + "loss": 0.7022, + "step": 1727 + }, + { + "epoch": 0.15450643776824036, + "grad_norm": 0.11438531807650926, + "learning_rate": 0.00019198396757355118, + "loss": 0.6966, + "step": 1728 + }, + { + "epoch": 0.1545958512160229, + "grad_norm": 0.11335841551788997, + "learning_rate": 0.00019197260282497171, + "loss": 0.6286, + "step": 1729 + }, + { + "epoch": 0.15468526466380544, + "grad_norm": 0.1163079674986109, + "learning_rate": 0.0001919612303627611, + "loss": 0.2612, + "step": 1730 + }, + { + "epoch": 0.154774678111588, + "grad_norm": 0.12614802378537723, + "learning_rate": 0.00019194985018787316, + "loss": 0.6877, + "step": 1731 + }, + { + "epoch": 0.15486409155937053, + "grad_norm": 0.12245004933054096, + "learning_rate": 0.00019193846230126233, + "loss": 0.7052, + "step": 1732 + }, + { + "epoch": 0.15495350500715308, + "grad_norm": 0.13469394925220396, + "learning_rate": 0.00019192706670388373, + "loss": 0.7322, + "step": 1733 + }, + { + "epoch": 0.15504291845493562, + "grad_norm": 0.10694649398785691, + "learning_rate": 0.00019191566339669302, + "loss": 0.6745, + "step": 1734 + }, + { + "epoch": 0.15513233190271816, + "grad_norm": 0.1096222816412736, + "learning_rate": 0.00019190425238064667, + "loss": 0.7179, + "step": 1735 + }, + { + "epoch": 0.1552217453505007, + "grad_norm": 0.1132287955859975, + "learning_rate": 0.00019189283365670163, + "loss": 0.6616, + "step": 1736 + }, + { + "epoch": 0.15531115879828325, + "grad_norm": 0.11575803309955662, + "learning_rate": 0.00019188140722581562, + "loss": 0.7049, + "step": 1737 + }, + { + "epoch": 0.15540057224606582, + "grad_norm": 0.14013902630397723, + "learning_rate": 0.00019186997308894696, + "loss": 0.7009, + "step": 1738 + }, + { + "epoch": 0.15548998569384836, + "grad_norm": 0.1275432578597959, + "learning_rate": 0.0001918585312470546, + "loss": 0.7309, + "step": 1739 + }, + { + "epoch": 0.1555793991416309, + "grad_norm": 0.11730883698776669, + "learning_rate": 0.00019184708170109818, + "loss": 0.7137, + "step": 1740 + }, + { + "epoch": 0.15566881258941345, + "grad_norm": 0.12815747519340265, + "learning_rate": 0.00019183562445203794, + "loss": 0.7064, + "step": 1741 + }, + { + "epoch": 0.155758226037196, + "grad_norm": 0.11050829762316085, + "learning_rate": 0.00019182415950083477, + "loss": 0.6794, + "step": 1742 + }, + { + "epoch": 0.15584763948497854, + "grad_norm": 0.12426286983007953, + "learning_rate": 0.0001918126868484502, + "loss": 0.6946, + "step": 1743 + }, + { + "epoch": 0.15593705293276108, + "grad_norm": 0.13224713923894502, + "learning_rate": 0.00019180120649584653, + "loss": 0.7383, + "step": 1744 + }, + { + "epoch": 0.15602646638054363, + "grad_norm": 0.12445792397764721, + "learning_rate": 0.00019178971844398653, + "loss": 0.6834, + "step": 1745 + }, + { + "epoch": 0.15611587982832617, + "grad_norm": 0.11309674442014214, + "learning_rate": 0.00019177822269383368, + "loss": 0.6847, + "step": 1746 + }, + { + "epoch": 0.15620529327610871, + "grad_norm": 0.11967831606765021, + "learning_rate": 0.00019176671924635215, + "loss": 0.7135, + "step": 1747 + }, + { + "epoch": 0.15629470672389129, + "grad_norm": 0.13061823604652045, + "learning_rate": 0.00019175520810250666, + "loss": 0.6876, + "step": 1748 + }, + { + "epoch": 0.15638412017167383, + "grad_norm": 0.13307552895512395, + "learning_rate": 0.00019174368926326273, + "loss": 0.6741, + "step": 1749 + }, + { + "epoch": 0.15647353361945637, + "grad_norm": 0.11069777233973276, + "learning_rate": 0.00019173216272958633, + "loss": 0.678, + "step": 1750 + }, + { + "epoch": 0.15656294706723892, + "grad_norm": 0.12502143589149975, + "learning_rate": 0.00019172062850244425, + "loss": 0.7213, + "step": 1751 + }, + { + "epoch": 0.15665236051502146, + "grad_norm": 0.1256493620844223, + "learning_rate": 0.00019170908658280386, + "loss": 0.6813, + "step": 1752 + }, + { + "epoch": 0.156741773962804, + "grad_norm": 0.13926079633256125, + "learning_rate": 0.0001916975369716331, + "loss": 0.7743, + "step": 1753 + }, + { + "epoch": 0.15683118741058655, + "grad_norm": 0.12035062960887621, + "learning_rate": 0.00019168597966990065, + "loss": 0.6925, + "step": 1754 + }, + { + "epoch": 0.1569206008583691, + "grad_norm": 0.11240717287829818, + "learning_rate": 0.00019167441467857584, + "loss": 0.6767, + "step": 1755 + }, + { + "epoch": 0.15701001430615164, + "grad_norm": 0.11377126879498531, + "learning_rate": 0.00019166284199862856, + "loss": 0.6711, + "step": 1756 + }, + { + "epoch": 0.15709942775393418, + "grad_norm": 0.10394273832784846, + "learning_rate": 0.00019165126163102943, + "loss": 0.6859, + "step": 1757 + }, + { + "epoch": 0.15718884120171675, + "grad_norm": 0.11250476939147157, + "learning_rate": 0.0001916396735767497, + "loss": 0.6905, + "step": 1758 + }, + { + "epoch": 0.1572782546494993, + "grad_norm": 0.12443571797739432, + "learning_rate": 0.00019162807783676118, + "loss": 0.6897, + "step": 1759 + }, + { + "epoch": 0.15736766809728184, + "grad_norm": 0.10605836607877296, + "learning_rate": 0.00019161647441203646, + "loss": 0.68, + "step": 1760 + }, + { + "epoch": 0.15745708154506438, + "grad_norm": 0.11003376332156369, + "learning_rate": 0.0001916048633035487, + "loss": 0.6957, + "step": 1761 + }, + { + "epoch": 0.15754649499284692, + "grad_norm": 0.10987884795880767, + "learning_rate": 0.00019159324451227164, + "loss": 0.6712, + "step": 1762 + }, + { + "epoch": 0.15763590844062947, + "grad_norm": 0.12792398856332476, + "learning_rate": 0.00019158161803917975, + "loss": 0.6911, + "step": 1763 + }, + { + "epoch": 0.157725321888412, + "grad_norm": 0.10655316987597205, + "learning_rate": 0.0001915699838852482, + "loss": 0.7234, + "step": 1764 + }, + { + "epoch": 0.15781473533619456, + "grad_norm": 0.11403793513236549, + "learning_rate": 0.0001915583420514527, + "loss": 0.6863, + "step": 1765 + }, + { + "epoch": 0.1579041487839771, + "grad_norm": 0.12407543191063304, + "learning_rate": 0.00019154669253876962, + "loss": 0.6831, + "step": 1766 + }, + { + "epoch": 0.15799356223175964, + "grad_norm": 0.12434335817823712, + "learning_rate": 0.000191535035348176, + "loss": 0.6957, + "step": 1767 + }, + { + "epoch": 0.15808297567954221, + "grad_norm": 0.1119578635445353, + "learning_rate": 0.00019152337048064947, + "loss": 0.6791, + "step": 1768 + }, + { + "epoch": 0.15817238912732476, + "grad_norm": 0.10813414378468275, + "learning_rate": 0.00019151169793716843, + "loss": 0.6604, + "step": 1769 + }, + { + "epoch": 0.1582618025751073, + "grad_norm": 0.13055837687068758, + "learning_rate": 0.0001915000177187118, + "loss": 0.7122, + "step": 1770 + }, + { + "epoch": 0.15835121602288985, + "grad_norm": 0.11656059515686934, + "learning_rate": 0.00019148832982625918, + "loss": 0.6426, + "step": 1771 + }, + { + "epoch": 0.1584406294706724, + "grad_norm": 0.11699584659654826, + "learning_rate": 0.00019147663426079083, + "loss": 0.7077, + "step": 1772 + }, + { + "epoch": 0.15853004291845493, + "grad_norm": 0.13864750029529074, + "learning_rate": 0.00019146493102328765, + "loss": 0.6467, + "step": 1773 + }, + { + "epoch": 0.15861945636623748, + "grad_norm": 0.11980729291349554, + "learning_rate": 0.00019145322011473117, + "loss": 0.6918, + "step": 1774 + }, + { + "epoch": 0.15870886981402002, + "grad_norm": 0.11180474978160337, + "learning_rate": 0.00019144150153610354, + "loss": 0.6915, + "step": 1775 + }, + { + "epoch": 0.15879828326180256, + "grad_norm": 0.1155201393186322, + "learning_rate": 0.00019142977528838762, + "loss": 0.6868, + "step": 1776 + }, + { + "epoch": 0.1588876967095851, + "grad_norm": 0.11778379319164653, + "learning_rate": 0.00019141804137256686, + "loss": 0.6738, + "step": 1777 + }, + { + "epoch": 0.15897711015736768, + "grad_norm": 0.11530758377796298, + "learning_rate": 0.0001914062997896254, + "loss": 0.6698, + "step": 1778 + }, + { + "epoch": 0.15906652360515022, + "grad_norm": 0.10704079640838476, + "learning_rate": 0.00019139455054054794, + "loss": 0.6488, + "step": 1779 + }, + { + "epoch": 0.15915593705293277, + "grad_norm": 0.11480013795221215, + "learning_rate": 0.0001913827936263199, + "loss": 0.7071, + "step": 1780 + }, + { + "epoch": 0.1592453505007153, + "grad_norm": 0.11711379866985537, + "learning_rate": 0.00019137102904792736, + "loss": 0.7018, + "step": 1781 + }, + { + "epoch": 0.15933476394849785, + "grad_norm": 0.1326575520003818, + "learning_rate": 0.00019135925680635694, + "loss": 0.689, + "step": 1782 + }, + { + "epoch": 0.1594241773962804, + "grad_norm": 0.13579643334375635, + "learning_rate": 0.00019134747690259597, + "loss": 0.2671, + "step": 1783 + }, + { + "epoch": 0.15951359084406294, + "grad_norm": 0.13853863608848796, + "learning_rate": 0.00019133568933763244, + "loss": 0.7346, + "step": 1784 + }, + { + "epoch": 0.15960300429184548, + "grad_norm": 0.132787149788848, + "learning_rate": 0.00019132389411245497, + "loss": 0.7586, + "step": 1785 + }, + { + "epoch": 0.15969241773962803, + "grad_norm": 0.10578155399029562, + "learning_rate": 0.00019131209122805277, + "loss": 0.6588, + "step": 1786 + }, + { + "epoch": 0.1597818311874106, + "grad_norm": 0.12080955803451345, + "learning_rate": 0.00019130028068541576, + "loss": 0.6685, + "step": 1787 + }, + { + "epoch": 0.15987124463519314, + "grad_norm": 0.11147845183429941, + "learning_rate": 0.0001912884624855345, + "loss": 0.68, + "step": 1788 + }, + { + "epoch": 0.1599606580829757, + "grad_norm": 0.11262300967072039, + "learning_rate": 0.0001912766366294001, + "loss": 0.7014, + "step": 1789 + }, + { + "epoch": 0.16005007153075823, + "grad_norm": 0.130038075899897, + "learning_rate": 0.00019126480311800444, + "loss": 0.6538, + "step": 1790 + }, + { + "epoch": 0.16013948497854077, + "grad_norm": 0.11908989686404553, + "learning_rate": 0.00019125296195233996, + "loss": 0.7034, + "step": 1791 + }, + { + "epoch": 0.16022889842632332, + "grad_norm": 0.11425627992864153, + "learning_rate": 0.00019124111313339976, + "loss": 0.6588, + "step": 1792 + }, + { + "epoch": 0.16031831187410586, + "grad_norm": 0.10669996673030831, + "learning_rate": 0.0001912292566621776, + "loss": 0.6929, + "step": 1793 + }, + { + "epoch": 0.1604077253218884, + "grad_norm": 0.11069339797999496, + "learning_rate": 0.00019121739253966785, + "loss": 0.6328, + "step": 1794 + }, + { + "epoch": 0.16049713876967095, + "grad_norm": 0.11381916617410061, + "learning_rate": 0.00019120552076686554, + "loss": 0.6591, + "step": 1795 + }, + { + "epoch": 0.1605865522174535, + "grad_norm": 0.12842137446863588, + "learning_rate": 0.0001911936413447664, + "loss": 0.7396, + "step": 1796 + }, + { + "epoch": 0.16067596566523606, + "grad_norm": 0.1281330320968276, + "learning_rate": 0.00019118175427436666, + "loss": 0.6857, + "step": 1797 + }, + { + "epoch": 0.1607653791130186, + "grad_norm": 0.10254543524126197, + "learning_rate": 0.0001911698595566633, + "loss": 0.6895, + "step": 1798 + }, + { + "epoch": 0.16085479256080115, + "grad_norm": 0.14527597291874023, + "learning_rate": 0.00019115795719265395, + "loss": 0.7322, + "step": 1799 + }, + { + "epoch": 0.1609442060085837, + "grad_norm": 0.11433302456330388, + "learning_rate": 0.0001911460471833368, + "loss": 0.6753, + "step": 1800 + }, + { + "epoch": 0.16103361945636624, + "grad_norm": 0.12472852960992482, + "learning_rate": 0.00019113412952971077, + "loss": 0.6969, + "step": 1801 + }, + { + "epoch": 0.16112303290414878, + "grad_norm": 0.11989553549307963, + "learning_rate": 0.00019112220423277534, + "loss": 0.6521, + "step": 1802 + }, + { + "epoch": 0.16121244635193133, + "grad_norm": 0.12734257446156286, + "learning_rate": 0.0001911102712935307, + "loss": 0.7088, + "step": 1803 + }, + { + "epoch": 0.16130185979971387, + "grad_norm": 0.1055221004007204, + "learning_rate": 0.00019109833071297763, + "loss": 0.6721, + "step": 1804 + }, + { + "epoch": 0.16139127324749641, + "grad_norm": 0.11000277996511225, + "learning_rate": 0.00019108638249211758, + "loss": 0.6998, + "step": 1805 + }, + { + "epoch": 0.16148068669527896, + "grad_norm": 0.12377999877903084, + "learning_rate": 0.00019107442663195265, + "loss": 0.6825, + "step": 1806 + }, + { + "epoch": 0.16157010014306153, + "grad_norm": 0.11550220276799744, + "learning_rate": 0.00019106246313348554, + "loss": 0.6893, + "step": 1807 + }, + { + "epoch": 0.16165951359084407, + "grad_norm": 0.11238592839677837, + "learning_rate": 0.00019105049199771962, + "loss": 0.708, + "step": 1808 + }, + { + "epoch": 0.16174892703862662, + "grad_norm": 0.14191199600178517, + "learning_rate": 0.00019103851322565892, + "loss": 0.2971, + "step": 1809 + }, + { + "epoch": 0.16183834048640916, + "grad_norm": 0.12274334264676583, + "learning_rate": 0.00019102652681830804, + "loss": 0.7078, + "step": 1810 + }, + { + "epoch": 0.1619277539341917, + "grad_norm": 0.13433496926235805, + "learning_rate": 0.00019101453277667226, + "loss": 0.7016, + "step": 1811 + }, + { + "epoch": 0.16201716738197425, + "grad_norm": 0.13602309060300172, + "learning_rate": 0.00019100253110175758, + "loss": 0.6766, + "step": 1812 + }, + { + "epoch": 0.1621065808297568, + "grad_norm": 0.12101749421086544, + "learning_rate": 0.00019099052179457054, + "loss": 0.7129, + "step": 1813 + }, + { + "epoch": 0.16219599427753933, + "grad_norm": 0.12688835452951058, + "learning_rate": 0.00019097850485611827, + "loss": 0.6832, + "step": 1814 + }, + { + "epoch": 0.16228540772532188, + "grad_norm": 0.12852663764881359, + "learning_rate": 0.00019096648028740868, + "loss": 0.6959, + "step": 1815 + }, + { + "epoch": 0.16237482117310442, + "grad_norm": 0.1218008701648276, + "learning_rate": 0.00019095444808945027, + "loss": 0.6802, + "step": 1816 + }, + { + "epoch": 0.162464234620887, + "grad_norm": 0.10556141241494164, + "learning_rate": 0.00019094240826325213, + "loss": 0.6429, + "step": 1817 + }, + { + "epoch": 0.16255364806866954, + "grad_norm": 0.12043395266246626, + "learning_rate": 0.00019093036080982404, + "loss": 0.7108, + "step": 1818 + }, + { + "epoch": 0.16264306151645208, + "grad_norm": 0.12746586616379435, + "learning_rate": 0.0001909183057301764, + "loss": 0.7045, + "step": 1819 + }, + { + "epoch": 0.16273247496423462, + "grad_norm": 0.12188653748090475, + "learning_rate": 0.0001909062430253203, + "loss": 0.699, + "step": 1820 + }, + { + "epoch": 0.16282188841201717, + "grad_norm": 0.11896710121120188, + "learning_rate": 0.00019089417269626733, + "loss": 0.7037, + "step": 1821 + }, + { + "epoch": 0.1629113018597997, + "grad_norm": 0.13061079057634808, + "learning_rate": 0.00019088209474402992, + "loss": 0.7086, + "step": 1822 + }, + { + "epoch": 0.16300071530758226, + "grad_norm": 0.12914965396101255, + "learning_rate": 0.00019087000916962095, + "loss": 0.6987, + "step": 1823 + }, + { + "epoch": 0.1630901287553648, + "grad_norm": 0.12174527371826448, + "learning_rate": 0.00019085791597405404, + "loss": 0.6899, + "step": 1824 + }, + { + "epoch": 0.16317954220314734, + "grad_norm": 0.10694182678764153, + "learning_rate": 0.00019084581515834347, + "loss": 0.6607, + "step": 1825 + }, + { + "epoch": 0.1632689556509299, + "grad_norm": 0.10807740441462976, + "learning_rate": 0.00019083370672350408, + "loss": 0.701, + "step": 1826 + }, + { + "epoch": 0.16335836909871246, + "grad_norm": 0.12601516200598928, + "learning_rate": 0.0001908215906705514, + "loss": 0.6772, + "step": 1827 + }, + { + "epoch": 0.163447782546495, + "grad_norm": 0.11013131813862798, + "learning_rate": 0.00019080946700050162, + "loss": 0.6613, + "step": 1828 + }, + { + "epoch": 0.16353719599427755, + "grad_norm": 0.12265842628829392, + "learning_rate": 0.00019079733571437154, + "loss": 0.6895, + "step": 1829 + }, + { + "epoch": 0.1636266094420601, + "grad_norm": 0.13555125201913948, + "learning_rate": 0.0001907851968131785, + "loss": 0.7409, + "step": 1830 + }, + { + "epoch": 0.16371602288984263, + "grad_norm": 0.12566543337306016, + "learning_rate": 0.00019077305029794068, + "loss": 0.7, + "step": 1831 + }, + { + "epoch": 0.16380543633762518, + "grad_norm": 0.13139956443218398, + "learning_rate": 0.00019076089616967677, + "loss": 0.6853, + "step": 1832 + }, + { + "epoch": 0.16389484978540772, + "grad_norm": 0.12234976858115502, + "learning_rate": 0.0001907487344294061, + "loss": 0.7021, + "step": 1833 + }, + { + "epoch": 0.16398426323319026, + "grad_norm": 0.12641946140786375, + "learning_rate": 0.00019073656507814866, + "loss": 0.7056, + "step": 1834 + }, + { + "epoch": 0.1640736766809728, + "grad_norm": 0.11622012176547453, + "learning_rate": 0.00019072438811692507, + "loss": 0.6963, + "step": 1835 + }, + { + "epoch": 0.16416309012875535, + "grad_norm": 0.11902338274120808, + "learning_rate": 0.00019071220354675665, + "loss": 0.7025, + "step": 1836 + }, + { + "epoch": 0.16425250357653792, + "grad_norm": 0.1334148259472465, + "learning_rate": 0.00019070001136866526, + "loss": 0.692, + "step": 1837 + }, + { + "epoch": 0.16434191702432047, + "grad_norm": 0.12738668648913765, + "learning_rate": 0.00019068781158367346, + "loss": 0.7218, + "step": 1838 + }, + { + "epoch": 0.164431330472103, + "grad_norm": 0.12553666691942714, + "learning_rate": 0.00019067560419280442, + "loss": 0.6854, + "step": 1839 + }, + { + "epoch": 0.16452074391988555, + "grad_norm": 0.14834700607556164, + "learning_rate": 0.00019066338919708197, + "loss": 0.696, + "step": 1840 + }, + { + "epoch": 0.1646101573676681, + "grad_norm": 0.11222328200791422, + "learning_rate": 0.00019065116659753054, + "loss": 0.6999, + "step": 1841 + }, + { + "epoch": 0.16469957081545064, + "grad_norm": 0.11884459943300574, + "learning_rate": 0.00019063893639517527, + "loss": 0.6772, + "step": 1842 + }, + { + "epoch": 0.16478898426323318, + "grad_norm": 0.10983061591334367, + "learning_rate": 0.00019062669859104187, + "loss": 0.7042, + "step": 1843 + }, + { + "epoch": 0.16487839771101573, + "grad_norm": 0.12745022856761226, + "learning_rate": 0.0001906144531861567, + "loss": 0.6681, + "step": 1844 + }, + { + "epoch": 0.16496781115879827, + "grad_norm": 0.1151117728634059, + "learning_rate": 0.00019060220018154677, + "loss": 0.6372, + "step": 1845 + }, + { + "epoch": 0.16505722460658084, + "grad_norm": 0.1082606998264293, + "learning_rate": 0.00019058993957823974, + "loss": 0.6706, + "step": 1846 + }, + { + "epoch": 0.1651466380543634, + "grad_norm": 0.11449777118420797, + "learning_rate": 0.00019057767137726388, + "loss": 0.6572, + "step": 1847 + }, + { + "epoch": 0.16523605150214593, + "grad_norm": 0.11925747277542763, + "learning_rate": 0.00019056539557964813, + "loss": 0.674, + "step": 1848 + }, + { + "epoch": 0.16532546494992847, + "grad_norm": 0.12322627672344537, + "learning_rate": 0.000190553112186422, + "loss": 0.7176, + "step": 1849 + }, + { + "epoch": 0.16541487839771102, + "grad_norm": 0.13614292727077237, + "learning_rate": 0.00019054082119861573, + "loss": 0.7186, + "step": 1850 + }, + { + "epoch": 0.16550429184549356, + "grad_norm": 0.13453519626011268, + "learning_rate": 0.0001905285226172601, + "loss": 0.6955, + "step": 1851 + }, + { + "epoch": 0.1655937052932761, + "grad_norm": 0.1211301065773757, + "learning_rate": 0.00019051621644338665, + "loss": 0.7175, + "step": 1852 + }, + { + "epoch": 0.16568311874105865, + "grad_norm": 0.11452385729010409, + "learning_rate": 0.0001905039026780274, + "loss": 0.7247, + "step": 1853 + }, + { + "epoch": 0.1657725321888412, + "grad_norm": 0.13667105337228255, + "learning_rate": 0.00019049158132221515, + "loss": 0.6894, + "step": 1854 + }, + { + "epoch": 0.16586194563662374, + "grad_norm": 0.11712185127505324, + "learning_rate": 0.0001904792523769833, + "loss": 0.7122, + "step": 1855 + }, + { + "epoch": 0.1659513590844063, + "grad_norm": 0.11043929147997786, + "learning_rate": 0.00019046691584336577, + "loss": 0.6759, + "step": 1856 + }, + { + "epoch": 0.16604077253218885, + "grad_norm": 0.11628023889301813, + "learning_rate": 0.0001904545717223973, + "loss": 0.7061, + "step": 1857 + }, + { + "epoch": 0.1661301859799714, + "grad_norm": 0.11826081016660321, + "learning_rate": 0.00019044222001511312, + "loss": 0.7122, + "step": 1858 + }, + { + "epoch": 0.16621959942775394, + "grad_norm": 0.11898242198212622, + "learning_rate": 0.00019042986072254919, + "loss": 0.6702, + "step": 1859 + }, + { + "epoch": 0.16630901287553648, + "grad_norm": 0.11669462226442971, + "learning_rate": 0.00019041749384574204, + "loss": 0.694, + "step": 1860 + }, + { + "epoch": 0.16639842632331903, + "grad_norm": 0.11600048113793035, + "learning_rate": 0.0001904051193857289, + "loss": 0.7181, + "step": 1861 + }, + { + "epoch": 0.16648783977110157, + "grad_norm": 0.10801742380921624, + "learning_rate": 0.00019039273734354755, + "loss": 0.644, + "step": 1862 + }, + { + "epoch": 0.1665772532188841, + "grad_norm": 0.1267140739177637, + "learning_rate": 0.00019038034772023648, + "loss": 0.7354, + "step": 1863 + }, + { + "epoch": 0.16666666666666666, + "grad_norm": 0.1164085053281927, + "learning_rate": 0.00019036795051683483, + "loss": 0.6578, + "step": 1864 + }, + { + "epoch": 0.1667560801144492, + "grad_norm": 0.10385484966855711, + "learning_rate": 0.0001903555457343823, + "loss": 0.6558, + "step": 1865 + }, + { + "epoch": 0.16684549356223177, + "grad_norm": 0.11683130181078293, + "learning_rate": 0.00019034313337391924, + "loss": 0.6723, + "step": 1866 + }, + { + "epoch": 0.16693490701001432, + "grad_norm": 0.1149737180253618, + "learning_rate": 0.00019033071343648673, + "loss": 0.6909, + "step": 1867 + }, + { + "epoch": 0.16702432045779686, + "grad_norm": 0.12425334291742444, + "learning_rate": 0.00019031828592312635, + "loss": 0.6886, + "step": 1868 + }, + { + "epoch": 0.1671137339055794, + "grad_norm": 0.11422953862286694, + "learning_rate": 0.00019030585083488043, + "loss": 0.6881, + "step": 1869 + }, + { + "epoch": 0.16720314735336195, + "grad_norm": 0.11625916269144815, + "learning_rate": 0.00019029340817279183, + "loss": 0.6673, + "step": 1870 + }, + { + "epoch": 0.1672925608011445, + "grad_norm": 0.11802066474955285, + "learning_rate": 0.00019028095793790416, + "loss": 0.6752, + "step": 1871 + }, + { + "epoch": 0.16738197424892703, + "grad_norm": 0.11695424062639746, + "learning_rate": 0.00019026850013126157, + "loss": 0.7056, + "step": 1872 + }, + { + "epoch": 0.16747138769670958, + "grad_norm": 0.11699052337154017, + "learning_rate": 0.0001902560347539089, + "loss": 0.6967, + "step": 1873 + }, + { + "epoch": 0.16756080114449212, + "grad_norm": 0.11825458974502198, + "learning_rate": 0.00019024356180689158, + "loss": 0.681, + "step": 1874 + }, + { + "epoch": 0.16765021459227467, + "grad_norm": 0.11612378776834918, + "learning_rate": 0.00019023108129125572, + "loss": 0.7234, + "step": 1875 + }, + { + "epoch": 0.16773962804005724, + "grad_norm": 0.14204192877506616, + "learning_rate": 0.00019021859320804806, + "loss": 0.7391, + "step": 1876 + }, + { + "epoch": 0.16782904148783978, + "grad_norm": 0.10313168010089004, + "learning_rate": 0.00019020609755831592, + "loss": 0.6859, + "step": 1877 + }, + { + "epoch": 0.16791845493562232, + "grad_norm": 0.12939499091302886, + "learning_rate": 0.00019019359434310738, + "loss": 0.6824, + "step": 1878 + }, + { + "epoch": 0.16800786838340487, + "grad_norm": 0.10751273319150763, + "learning_rate": 0.00019018108356347094, + "loss": 0.6508, + "step": 1879 + }, + { + "epoch": 0.1680972818311874, + "grad_norm": 0.11916830744589589, + "learning_rate": 0.00019016856522045597, + "loss": 0.6527, + "step": 1880 + }, + { + "epoch": 0.16818669527896996, + "grad_norm": 0.10594152147527344, + "learning_rate": 0.0001901560393151123, + "loss": 0.649, + "step": 1881 + }, + { + "epoch": 0.1682761087267525, + "grad_norm": 0.13589736353394655, + "learning_rate": 0.00019014350584849052, + "loss": 0.7052, + "step": 1882 + }, + { + "epoch": 0.16836552217453504, + "grad_norm": 0.13291718690589668, + "learning_rate": 0.00019013096482164177, + "loss": 0.7056, + "step": 1883 + }, + { + "epoch": 0.1684549356223176, + "grad_norm": 0.1301458178954311, + "learning_rate": 0.00019011841623561783, + "loss": 0.2603, + "step": 1884 + }, + { + "epoch": 0.16854434907010013, + "grad_norm": 0.13427371984319553, + "learning_rate": 0.00019010586009147117, + "loss": 0.2771, + "step": 1885 + }, + { + "epoch": 0.1686337625178827, + "grad_norm": 0.1168553338256289, + "learning_rate": 0.00019009329639025483, + "loss": 0.6874, + "step": 1886 + }, + { + "epoch": 0.16872317596566525, + "grad_norm": 0.13890114562958722, + "learning_rate": 0.00019008072513302255, + "loss": 0.7254, + "step": 1887 + }, + { + "epoch": 0.1688125894134478, + "grad_norm": 0.12072241221130657, + "learning_rate": 0.00019006814632082863, + "loss": 0.7203, + "step": 1888 + }, + { + "epoch": 0.16890200286123033, + "grad_norm": 0.14468452610811955, + "learning_rate": 0.00019005555995472805, + "loss": 0.7417, + "step": 1889 + }, + { + "epoch": 0.16899141630901288, + "grad_norm": 0.12106766600415382, + "learning_rate": 0.00019004296603577646, + "loss": 0.6631, + "step": 1890 + }, + { + "epoch": 0.16908082975679542, + "grad_norm": 0.14788132366537568, + "learning_rate": 0.00019003036456503, + "loss": 0.2522, + "step": 1891 + }, + { + "epoch": 0.16917024320457796, + "grad_norm": 0.12337494230205853, + "learning_rate": 0.0001900177555435456, + "loss": 0.6882, + "step": 1892 + }, + { + "epoch": 0.1692596566523605, + "grad_norm": 0.11795559221721491, + "learning_rate": 0.00019000513897238076, + "loss": 0.7301, + "step": 1893 + }, + { + "epoch": 0.16934907010014305, + "grad_norm": 0.10576836800829928, + "learning_rate": 0.00018999251485259363, + "loss": 0.6704, + "step": 1894 + }, + { + "epoch": 0.1694384835479256, + "grad_norm": 0.11510650370073108, + "learning_rate": 0.00018997988318524293, + "loss": 0.7127, + "step": 1895 + }, + { + "epoch": 0.16952789699570817, + "grad_norm": 0.11763815458665651, + "learning_rate": 0.00018996724397138813, + "loss": 0.7069, + "step": 1896 + }, + { + "epoch": 0.1696173104434907, + "grad_norm": 0.1214304401604972, + "learning_rate": 0.0001899545972120892, + "loss": 0.6594, + "step": 1897 + }, + { + "epoch": 0.16970672389127325, + "grad_norm": 0.12406224083724697, + "learning_rate": 0.00018994194290840687, + "loss": 0.7179, + "step": 1898 + }, + { + "epoch": 0.1697961373390558, + "grad_norm": 0.11528399537005808, + "learning_rate": 0.0001899292810614024, + "loss": 0.6788, + "step": 1899 + }, + { + "epoch": 0.16988555078683834, + "grad_norm": 0.12611319447964545, + "learning_rate": 0.00018991661167213773, + "loss": 0.7405, + "step": 1900 + }, + { + "epoch": 0.16997496423462088, + "grad_norm": 0.12396590941606385, + "learning_rate": 0.00018990393474167542, + "loss": 0.7119, + "step": 1901 + }, + { + "epoch": 0.17006437768240343, + "grad_norm": 0.11594510256259329, + "learning_rate": 0.0001898912502710787, + "loss": 0.7043, + "step": 1902 + }, + { + "epoch": 0.17015379113018597, + "grad_norm": 0.11428745257039895, + "learning_rate": 0.00018987855826141137, + "loss": 0.6924, + "step": 1903 + }, + { + "epoch": 0.17024320457796852, + "grad_norm": 0.1095633885992649, + "learning_rate": 0.0001898658587137379, + "loss": 0.6657, + "step": 1904 + }, + { + "epoch": 0.17033261802575106, + "grad_norm": 0.12434485817509691, + "learning_rate": 0.0001898531516291234, + "loss": 0.7053, + "step": 1905 + }, + { + "epoch": 0.17042203147353363, + "grad_norm": 0.10310173193584424, + "learning_rate": 0.00018984043700863356, + "loss": 0.6483, + "step": 1906 + }, + { + "epoch": 0.17051144492131617, + "grad_norm": 0.11795424839664828, + "learning_rate": 0.0001898277148533348, + "loss": 0.7162, + "step": 1907 + }, + { + "epoch": 0.17060085836909872, + "grad_norm": 0.12132720151286334, + "learning_rate": 0.0001898149851642941, + "loss": 0.6592, + "step": 1908 + }, + { + "epoch": 0.17069027181688126, + "grad_norm": 0.1125455022362896, + "learning_rate": 0.00018980224794257905, + "loss": 0.7238, + "step": 1909 + }, + { + "epoch": 0.1707796852646638, + "grad_norm": 0.18841000372231023, + "learning_rate": 0.0001897895031892579, + "loss": 0.3079, + "step": 1910 + }, + { + "epoch": 0.17086909871244635, + "grad_norm": 0.12300191196500047, + "learning_rate": 0.00018977675090539955, + "loss": 0.7182, + "step": 1911 + }, + { + "epoch": 0.1709585121602289, + "grad_norm": 0.12632167780791181, + "learning_rate": 0.00018976399109207353, + "loss": 0.6974, + "step": 1912 + }, + { + "epoch": 0.17104792560801144, + "grad_norm": 0.12825071822883163, + "learning_rate": 0.00018975122375035, + "loss": 0.7157, + "step": 1913 + }, + { + "epoch": 0.17113733905579398, + "grad_norm": 0.11263270068745176, + "learning_rate": 0.0001897384488812997, + "loss": 0.6989, + "step": 1914 + }, + { + "epoch": 0.17122675250357655, + "grad_norm": 0.11608559497333142, + "learning_rate": 0.00018972566648599407, + "loss": 0.6867, + "step": 1915 + }, + { + "epoch": 0.1713161659513591, + "grad_norm": 0.10266621466186321, + "learning_rate": 0.0001897128765655052, + "loss": 0.648, + "step": 1916 + }, + { + "epoch": 0.17140557939914164, + "grad_norm": 0.13791710589759676, + "learning_rate": 0.00018970007912090566, + "loss": 0.7091, + "step": 1917 + }, + { + "epoch": 0.17149499284692418, + "grad_norm": 0.11880545335646407, + "learning_rate": 0.00018968727415326884, + "loss": 0.7123, + "step": 1918 + }, + { + "epoch": 0.17158440629470673, + "grad_norm": 0.12380963851251645, + "learning_rate": 0.00018967446166366867, + "loss": 0.7005, + "step": 1919 + }, + { + "epoch": 0.17167381974248927, + "grad_norm": 0.12770174321500308, + "learning_rate": 0.00018966164165317966, + "loss": 0.7109, + "step": 1920 + }, + { + "epoch": 0.1717632331902718, + "grad_norm": 0.11513206277268913, + "learning_rate": 0.00018964881412287708, + "loss": 0.6778, + "step": 1921 + }, + { + "epoch": 0.17185264663805436, + "grad_norm": 0.12710163540454672, + "learning_rate": 0.00018963597907383672, + "loss": 0.6957, + "step": 1922 + }, + { + "epoch": 0.1719420600858369, + "grad_norm": 0.11718883769687244, + "learning_rate": 0.00018962313650713503, + "loss": 0.6687, + "step": 1923 + }, + { + "epoch": 0.17203147353361944, + "grad_norm": 0.13160546992569203, + "learning_rate": 0.00018961028642384915, + "loss": 0.6866, + "step": 1924 + }, + { + "epoch": 0.17212088698140202, + "grad_norm": 0.12364450999295856, + "learning_rate": 0.00018959742882505674, + "loss": 0.6971, + "step": 1925 + }, + { + "epoch": 0.17221030042918456, + "grad_norm": 0.11586714258437039, + "learning_rate": 0.00018958456371183618, + "loss": 0.6751, + "step": 1926 + }, + { + "epoch": 0.1722997138769671, + "grad_norm": 0.10842509813970815, + "learning_rate": 0.00018957169108526646, + "loss": 0.6643, + "step": 1927 + }, + { + "epoch": 0.17238912732474965, + "grad_norm": 0.12457449199120395, + "learning_rate": 0.00018955881094642721, + "loss": 0.692, + "step": 1928 + }, + { + "epoch": 0.1724785407725322, + "grad_norm": 0.12642201900361733, + "learning_rate": 0.00018954592329639862, + "loss": 0.6828, + "step": 1929 + }, + { + "epoch": 0.17256795422031473, + "grad_norm": 0.11981778864171488, + "learning_rate": 0.00018953302813626158, + "loss": 0.6796, + "step": 1930 + }, + { + "epoch": 0.17265736766809728, + "grad_norm": 0.12091924168259775, + "learning_rate": 0.00018952012546709764, + "loss": 0.6553, + "step": 1931 + }, + { + "epoch": 0.17274678111587982, + "grad_norm": 0.10453498183937204, + "learning_rate": 0.00018950721528998885, + "loss": 0.6384, + "step": 1932 + }, + { + "epoch": 0.17283619456366237, + "grad_norm": 0.1443120712212849, + "learning_rate": 0.00018949429760601802, + "loss": 0.7472, + "step": 1933 + }, + { + "epoch": 0.1729256080114449, + "grad_norm": 0.12191909962800651, + "learning_rate": 0.00018948137241626853, + "loss": 0.6836, + "step": 1934 + }, + { + "epoch": 0.17301502145922748, + "grad_norm": 0.11660090560595136, + "learning_rate": 0.0001894684397218244, + "loss": 0.6712, + "step": 1935 + }, + { + "epoch": 0.17310443490701002, + "grad_norm": 0.11438408515104383, + "learning_rate": 0.0001894554995237703, + "loss": 0.6686, + "step": 1936 + }, + { + "epoch": 0.17319384835479257, + "grad_norm": 0.13427170976727007, + "learning_rate": 0.00018944255182319148, + "loss": 0.7081, + "step": 1937 + }, + { + "epoch": 0.1732832618025751, + "grad_norm": 0.17223869682831958, + "learning_rate": 0.00018942959662117384, + "loss": 0.7481, + "step": 1938 + }, + { + "epoch": 0.17337267525035766, + "grad_norm": 0.12081030312534782, + "learning_rate": 0.00018941663391880396, + "loss": 0.6544, + "step": 1939 + }, + { + "epoch": 0.1734620886981402, + "grad_norm": 0.11833915661139312, + "learning_rate": 0.00018940366371716897, + "loss": 0.6868, + "step": 1940 + }, + { + "epoch": 0.17355150214592274, + "grad_norm": 0.12877785303904765, + "learning_rate": 0.00018939068601735666, + "loss": 0.7182, + "step": 1941 + }, + { + "epoch": 0.1736409155937053, + "grad_norm": 0.1226087073527401, + "learning_rate": 0.0001893777008204555, + "loss": 0.6937, + "step": 1942 + }, + { + "epoch": 0.17373032904148783, + "grad_norm": 0.1189044128025027, + "learning_rate": 0.0001893647081275545, + "loss": 0.6778, + "step": 1943 + }, + { + "epoch": 0.17381974248927037, + "grad_norm": 0.09638768459178762, + "learning_rate": 0.00018935170793974335, + "loss": 0.6707, + "step": 1944 + }, + { + "epoch": 0.17390915593705294, + "grad_norm": 0.11717108947612714, + "learning_rate": 0.00018933870025811237, + "loss": 0.6693, + "step": 1945 + }, + { + "epoch": 0.1739985693848355, + "grad_norm": 0.1230721919541804, + "learning_rate": 0.0001893256850837525, + "loss": 0.7017, + "step": 1946 + }, + { + "epoch": 0.17408798283261803, + "grad_norm": 0.11541704282432169, + "learning_rate": 0.0001893126624177553, + "loss": 0.6769, + "step": 1947 + }, + { + "epoch": 0.17417739628040058, + "grad_norm": 0.1150643509658155, + "learning_rate": 0.00018929963226121295, + "loss": 0.7077, + "step": 1948 + }, + { + "epoch": 0.17426680972818312, + "grad_norm": 0.12608954776748588, + "learning_rate": 0.0001892865946152183, + "loss": 0.7046, + "step": 1949 + }, + { + "epoch": 0.17435622317596566, + "grad_norm": 0.10561051760751211, + "learning_rate": 0.0001892735494808648, + "loss": 0.71, + "step": 1950 + }, + { + "epoch": 0.1744456366237482, + "grad_norm": 0.09643997367689452, + "learning_rate": 0.0001892604968592465, + "loss": 0.6354, + "step": 1951 + }, + { + "epoch": 0.17453505007153075, + "grad_norm": 0.10611842492382081, + "learning_rate": 0.00018924743675145813, + "loss": 0.6578, + "step": 1952 + }, + { + "epoch": 0.1746244635193133, + "grad_norm": 0.11108628164027949, + "learning_rate": 0.00018923436915859503, + "loss": 0.6381, + "step": 1953 + }, + { + "epoch": 0.17471387696709584, + "grad_norm": 0.17818046641664942, + "learning_rate": 0.00018922129408175314, + "loss": 0.32, + "step": 1954 + }, + { + "epoch": 0.1748032904148784, + "grad_norm": 0.12154001690320032, + "learning_rate": 0.0001892082115220291, + "loss": 0.7092, + "step": 1955 + }, + { + "epoch": 0.17489270386266095, + "grad_norm": 0.1310448549104356, + "learning_rate": 0.00018919512148052005, + "loss": 0.7262, + "step": 1956 + }, + { + "epoch": 0.1749821173104435, + "grad_norm": 0.11808890413879283, + "learning_rate": 0.0001891820239583239, + "loss": 0.6993, + "step": 1957 + }, + { + "epoch": 0.17507153075822604, + "grad_norm": 0.13261616149235378, + "learning_rate": 0.00018916891895653915, + "loss": 0.7123, + "step": 1958 + }, + { + "epoch": 0.17516094420600858, + "grad_norm": 0.11610953258948926, + "learning_rate": 0.0001891558064762648, + "loss": 0.6783, + "step": 1959 + }, + { + "epoch": 0.17525035765379113, + "grad_norm": 0.13097625321661244, + "learning_rate": 0.00018914268651860067, + "loss": 0.676, + "step": 1960 + }, + { + "epoch": 0.17533977110157367, + "grad_norm": 0.12985529353473307, + "learning_rate": 0.00018912955908464708, + "loss": 0.697, + "step": 1961 + }, + { + "epoch": 0.17542918454935622, + "grad_norm": 0.15575985309065107, + "learning_rate": 0.00018911642417550497, + "loss": 0.7266, + "step": 1962 + }, + { + "epoch": 0.17551859799713876, + "grad_norm": 0.17870649910599348, + "learning_rate": 0.00018910328179227605, + "loss": 0.2942, + "step": 1963 + }, + { + "epoch": 0.1756080114449213, + "grad_norm": 0.12886345489825782, + "learning_rate": 0.0001890901319360624, + "loss": 0.7166, + "step": 1964 + }, + { + "epoch": 0.17569742489270387, + "grad_norm": 0.1395921400826629, + "learning_rate": 0.00018907697460796707, + "loss": 0.6862, + "step": 1965 + }, + { + "epoch": 0.17578683834048642, + "grad_norm": 0.11853108765511218, + "learning_rate": 0.00018906380980909343, + "loss": 0.6911, + "step": 1966 + }, + { + "epoch": 0.17587625178826896, + "grad_norm": 0.12576695159625104, + "learning_rate": 0.00018905063754054563, + "loss": 0.6865, + "step": 1967 + }, + { + "epoch": 0.1759656652360515, + "grad_norm": 0.11970521518768175, + "learning_rate": 0.00018903745780342839, + "loss": 0.7064, + "step": 1968 + }, + { + "epoch": 0.17605507868383405, + "grad_norm": 0.11528435251078725, + "learning_rate": 0.00018902427059884708, + "loss": 0.6922, + "step": 1969 + }, + { + "epoch": 0.1761444921316166, + "grad_norm": 0.11604076044384082, + "learning_rate": 0.00018901107592790776, + "loss": 0.6778, + "step": 1970 + }, + { + "epoch": 0.17623390557939914, + "grad_norm": 0.11213686257665133, + "learning_rate": 0.00018899787379171693, + "loss": 0.6568, + "step": 1971 + }, + { + "epoch": 0.17632331902718168, + "grad_norm": 0.12334463803832439, + "learning_rate": 0.00018898466419138197, + "loss": 0.7051, + "step": 1972 + }, + { + "epoch": 0.17641273247496422, + "grad_norm": 0.11396286964113633, + "learning_rate": 0.00018897144712801066, + "loss": 0.6922, + "step": 1973 + }, + { + "epoch": 0.1765021459227468, + "grad_norm": 0.11844073864189718, + "learning_rate": 0.00018895822260271152, + "loss": 0.6719, + "step": 1974 + }, + { + "epoch": 0.17659155937052934, + "grad_norm": 0.12709094724108635, + "learning_rate": 0.0001889449906165937, + "loss": 0.7201, + "step": 1975 + }, + { + "epoch": 0.17668097281831188, + "grad_norm": 0.11570101480120625, + "learning_rate": 0.00018893175117076693, + "loss": 0.6738, + "step": 1976 + }, + { + "epoch": 0.17677038626609443, + "grad_norm": 0.13232494368624922, + "learning_rate": 0.00018891850426634162, + "loss": 0.7358, + "step": 1977 + }, + { + "epoch": 0.17685979971387697, + "grad_norm": 0.10950790352560692, + "learning_rate": 0.00018890524990442873, + "loss": 0.6586, + "step": 1978 + }, + { + "epoch": 0.1769492131616595, + "grad_norm": 0.1155862644460618, + "learning_rate": 0.00018889198808613985, + "loss": 0.6673, + "step": 1979 + }, + { + "epoch": 0.17703862660944206, + "grad_norm": 0.1286750405399078, + "learning_rate": 0.00018887871881258735, + "loss": 0.735, + "step": 1980 + }, + { + "epoch": 0.1771280400572246, + "grad_norm": 0.12227047905171491, + "learning_rate": 0.000188865442084884, + "loss": 0.6588, + "step": 1981 + }, + { + "epoch": 0.17721745350500714, + "grad_norm": 0.1211146677198949, + "learning_rate": 0.00018885215790414336, + "loss": 0.6881, + "step": 1982 + }, + { + "epoch": 0.1773068669527897, + "grad_norm": 0.12173209414922334, + "learning_rate": 0.00018883886627147955, + "loss": 0.7, + "step": 1983 + }, + { + "epoch": 0.17739628040057226, + "grad_norm": 0.1241121864439775, + "learning_rate": 0.0001888255671880073, + "loss": 0.6737, + "step": 1984 + }, + { + "epoch": 0.1774856938483548, + "grad_norm": 0.1197826497389157, + "learning_rate": 0.00018881226065484204, + "loss": 0.6684, + "step": 1985 + }, + { + "epoch": 0.17757510729613735, + "grad_norm": 0.12704227179026553, + "learning_rate": 0.0001887989466730997, + "loss": 0.718, + "step": 1986 + }, + { + "epoch": 0.1776645207439199, + "grad_norm": 0.13339241142852087, + "learning_rate": 0.00018878562524389696, + "loss": 0.7113, + "step": 1987 + }, + { + "epoch": 0.17775393419170243, + "grad_norm": 0.12544499468555292, + "learning_rate": 0.00018877229636835106, + "loss": 0.6812, + "step": 1988 + }, + { + "epoch": 0.17784334763948498, + "grad_norm": 0.13095333089008557, + "learning_rate": 0.00018875896004757984, + "loss": 0.7287, + "step": 1989 + }, + { + "epoch": 0.17793276108726752, + "grad_norm": 0.11701508232317133, + "learning_rate": 0.00018874561628270188, + "loss": 0.6854, + "step": 1990 + }, + { + "epoch": 0.17802217453505007, + "grad_norm": 0.14933076153959107, + "learning_rate": 0.00018873226507483623, + "loss": 0.7556, + "step": 1991 + }, + { + "epoch": 0.1781115879828326, + "grad_norm": 0.11929408961838235, + "learning_rate": 0.0001887189064251027, + "loss": 0.6431, + "step": 1992 + }, + { + "epoch": 0.17820100143061515, + "grad_norm": 0.11529227519359792, + "learning_rate": 0.00018870554033462159, + "loss": 0.7106, + "step": 1993 + }, + { + "epoch": 0.17829041487839772, + "grad_norm": 0.1384982345171805, + "learning_rate": 0.00018869216680451398, + "loss": 0.7142, + "step": 1994 + }, + { + "epoch": 0.17837982832618027, + "grad_norm": 0.10954139645978618, + "learning_rate": 0.0001886787858359014, + "loss": 0.6827, + "step": 1995 + }, + { + "epoch": 0.1784692417739628, + "grad_norm": 0.13250385396816466, + "learning_rate": 0.0001886653974299062, + "loss": 0.7134, + "step": 1996 + }, + { + "epoch": 0.17855865522174535, + "grad_norm": 0.11546744428457084, + "learning_rate": 0.0001886520015876512, + "loss": 0.651, + "step": 1997 + }, + { + "epoch": 0.1786480686695279, + "grad_norm": 0.10414755391372328, + "learning_rate": 0.00018863859831025988, + "loss": 0.6525, + "step": 1998 + }, + { + "epoch": 0.17873748211731044, + "grad_norm": 0.1194436559373267, + "learning_rate": 0.00018862518759885636, + "loss": 0.661, + "step": 1999 + }, + { + "epoch": 0.17882689556509299, + "grad_norm": 0.12192684087443896, + "learning_rate": 0.0001886117694545654, + "loss": 0.7056, + "step": 2000 + }, + { + "epoch": 0.17891630901287553, + "grad_norm": 0.11347178496607332, + "learning_rate": 0.00018859834387851233, + "loss": 0.6732, + "step": 2001 + }, + { + "epoch": 0.17900572246065807, + "grad_norm": 0.11153290491090181, + "learning_rate": 0.00018858491087182317, + "loss": 0.6868, + "step": 2002 + }, + { + "epoch": 0.17909513590844062, + "grad_norm": 0.12921461320560004, + "learning_rate": 0.00018857147043562452, + "loss": 0.694, + "step": 2003 + }, + { + "epoch": 0.1791845493562232, + "grad_norm": 0.11247204368165518, + "learning_rate": 0.00018855802257104363, + "loss": 0.6485, + "step": 2004 + }, + { + "epoch": 0.17927396280400573, + "grad_norm": 0.13142468215438394, + "learning_rate": 0.0001885445672792083, + "loss": 0.711, + "step": 2005 + }, + { + "epoch": 0.17936337625178828, + "grad_norm": 0.11043207853992618, + "learning_rate": 0.00018853110456124709, + "loss": 0.6937, + "step": 2006 + }, + { + "epoch": 0.17945278969957082, + "grad_norm": 0.1083358377682362, + "learning_rate": 0.00018851763441828903, + "loss": 0.6747, + "step": 2007 + }, + { + "epoch": 0.17954220314735336, + "grad_norm": 0.12018435582765194, + "learning_rate": 0.00018850415685146387, + "loss": 0.6953, + "step": 2008 + }, + { + "epoch": 0.1796316165951359, + "grad_norm": 0.13216283288419842, + "learning_rate": 0.00018849067186190198, + "loss": 0.6767, + "step": 2009 + }, + { + "epoch": 0.17972103004291845, + "grad_norm": 0.10984096337082117, + "learning_rate": 0.0001884771794507343, + "loss": 0.6977, + "step": 2010 + }, + { + "epoch": 0.179810443490701, + "grad_norm": 0.12506613008024034, + "learning_rate": 0.00018846367961909244, + "loss": 0.6848, + "step": 2011 + }, + { + "epoch": 0.17989985693848354, + "grad_norm": 0.13220872503593875, + "learning_rate": 0.0001884501723681086, + "loss": 0.7054, + "step": 2012 + }, + { + "epoch": 0.17998927038626608, + "grad_norm": 0.10817474959457696, + "learning_rate": 0.00018843665769891562, + "loss": 0.6541, + "step": 2013 + }, + { + "epoch": 0.18007868383404865, + "grad_norm": 0.12610832152515683, + "learning_rate": 0.00018842313561264696, + "loss": 0.6979, + "step": 2014 + }, + { + "epoch": 0.1801680972818312, + "grad_norm": 0.11398257200352761, + "learning_rate": 0.0001884096061104367, + "loss": 0.6836, + "step": 2015 + }, + { + "epoch": 0.18025751072961374, + "grad_norm": 0.12537066965418941, + "learning_rate": 0.0001883960691934196, + "loss": 0.747, + "step": 2016 + }, + { + "epoch": 0.18034692417739628, + "grad_norm": 0.1089965641416328, + "learning_rate": 0.00018838252486273087, + "loss": 0.6733, + "step": 2017 + }, + { + "epoch": 0.18043633762517883, + "grad_norm": 0.10885415707829742, + "learning_rate": 0.00018836897311950653, + "loss": 0.6702, + "step": 2018 + }, + { + "epoch": 0.18052575107296137, + "grad_norm": 0.11848487001794955, + "learning_rate": 0.00018835541396488315, + "loss": 0.6532, + "step": 2019 + }, + { + "epoch": 0.18061516452074391, + "grad_norm": 0.10737887204350077, + "learning_rate": 0.00018834184739999793, + "loss": 0.6427, + "step": 2020 + }, + { + "epoch": 0.18070457796852646, + "grad_norm": 0.11525622582198047, + "learning_rate": 0.00018832827342598861, + "loss": 0.6799, + "step": 2021 + }, + { + "epoch": 0.180793991416309, + "grad_norm": 0.12451404284086609, + "learning_rate": 0.0001883146920439937, + "loss": 0.6852, + "step": 2022 + }, + { + "epoch": 0.18088340486409155, + "grad_norm": 0.10661191830642246, + "learning_rate": 0.00018830110325515222, + "loss": 0.6671, + "step": 2023 + }, + { + "epoch": 0.18097281831187412, + "grad_norm": 0.12677890263803931, + "learning_rate": 0.00018828750706060385, + "loss": 0.6943, + "step": 2024 + }, + { + "epoch": 0.18106223175965666, + "grad_norm": 0.12070922174318834, + "learning_rate": 0.00018827390346148887, + "loss": 0.674, + "step": 2025 + }, + { + "epoch": 0.1811516452074392, + "grad_norm": 0.12519417920927614, + "learning_rate": 0.00018826029245894827, + "loss": 0.7224, + "step": 2026 + }, + { + "epoch": 0.18124105865522175, + "grad_norm": 0.14199705278742059, + "learning_rate": 0.00018824667405412348, + "loss": 0.7161, + "step": 2027 + }, + { + "epoch": 0.1813304721030043, + "grad_norm": 0.11645573952652243, + "learning_rate": 0.00018823304824815672, + "loss": 0.6976, + "step": 2028 + }, + { + "epoch": 0.18141988555078684, + "grad_norm": 0.111495005552564, + "learning_rate": 0.0001882194150421908, + "loss": 0.622, + "step": 2029 + }, + { + "epoch": 0.18150929899856938, + "grad_norm": 0.1203465237136251, + "learning_rate": 0.00018820577443736904, + "loss": 0.6961, + "step": 2030 + }, + { + "epoch": 0.18159871244635192, + "grad_norm": 0.13073063653322364, + "learning_rate": 0.0001881921264348355, + "loss": 0.6911, + "step": 2031 + }, + { + "epoch": 0.18168812589413447, + "grad_norm": 0.12257641324192861, + "learning_rate": 0.00018817847103573486, + "loss": 0.6721, + "step": 2032 + }, + { + "epoch": 0.181777539341917, + "grad_norm": 0.11395915974882388, + "learning_rate": 0.00018816480824121232, + "loss": 0.6693, + "step": 2033 + }, + { + "epoch": 0.18186695278969958, + "grad_norm": 0.1289812610524996, + "learning_rate": 0.0001881511380524138, + "loss": 0.6724, + "step": 2034 + }, + { + "epoch": 0.18195636623748213, + "grad_norm": 0.1236149936403225, + "learning_rate": 0.0001881374604704858, + "loss": 0.6965, + "step": 2035 + }, + { + "epoch": 0.18204577968526467, + "grad_norm": 0.1183720796270324, + "learning_rate": 0.0001881237754965754, + "loss": 0.6525, + "step": 2036 + }, + { + "epoch": 0.1821351931330472, + "grad_norm": 0.1276783530208454, + "learning_rate": 0.0001881100831318304, + "loss": 0.7187, + "step": 2037 + }, + { + "epoch": 0.18222460658082976, + "grad_norm": 0.12197370450019886, + "learning_rate": 0.00018809638337739915, + "loss": 0.7062, + "step": 2038 + }, + { + "epoch": 0.1823140200286123, + "grad_norm": 0.13451829344309701, + "learning_rate": 0.0001880826762344306, + "loss": 0.6945, + "step": 2039 + }, + { + "epoch": 0.18240343347639484, + "grad_norm": 0.11865292066330312, + "learning_rate": 0.00018806896170407437, + "loss": 0.6854, + "step": 2040 + }, + { + "epoch": 0.1824928469241774, + "grad_norm": 0.12037326348694048, + "learning_rate": 0.00018805523978748068, + "loss": 0.6947, + "step": 2041 + }, + { + "epoch": 0.18258226037195993, + "grad_norm": 0.1132324865810452, + "learning_rate": 0.0001880415104858004, + "loss": 0.6749, + "step": 2042 + }, + { + "epoch": 0.1826716738197425, + "grad_norm": 0.11393864869195477, + "learning_rate": 0.00018802777380018496, + "loss": 0.6984, + "step": 2043 + }, + { + "epoch": 0.18276108726752505, + "grad_norm": 0.11672269402128499, + "learning_rate": 0.00018801402973178642, + "loss": 0.6662, + "step": 2044 + }, + { + "epoch": 0.1828505007153076, + "grad_norm": 0.10932573617617145, + "learning_rate": 0.0001880002782817575, + "loss": 0.66, + "step": 2045 + }, + { + "epoch": 0.18293991416309013, + "grad_norm": 0.11240378400507145, + "learning_rate": 0.00018798651945125153, + "loss": 0.6425, + "step": 2046 + }, + { + "epoch": 0.18302932761087268, + "grad_norm": 0.12306376254004134, + "learning_rate": 0.00018797275324142242, + "loss": 0.6973, + "step": 2047 + }, + { + "epoch": 0.18311874105865522, + "grad_norm": 0.10667244097060088, + "learning_rate": 0.00018795897965342474, + "loss": 0.6863, + "step": 2048 + }, + { + "epoch": 0.18320815450643776, + "grad_norm": 0.12016164483952818, + "learning_rate": 0.00018794519868841367, + "loss": 0.6939, + "step": 2049 + }, + { + "epoch": 0.1832975679542203, + "grad_norm": 0.11739536757458227, + "learning_rate": 0.000187931410347545, + "loss": 0.6728, + "step": 2050 + }, + { + "epoch": 0.18338698140200285, + "grad_norm": 0.1126577804991417, + "learning_rate": 0.00018791761463197513, + "loss": 0.6797, + "step": 2051 + }, + { + "epoch": 0.1834763948497854, + "grad_norm": 0.11851918741996671, + "learning_rate": 0.00018790381154286113, + "loss": 0.6631, + "step": 2052 + }, + { + "epoch": 0.18356580829756797, + "grad_norm": 0.13147996981923793, + "learning_rate": 0.00018789000108136058, + "loss": 0.7032, + "step": 2053 + }, + { + "epoch": 0.1836552217453505, + "grad_norm": 0.11551888679723392, + "learning_rate": 0.0001878761832486318, + "loss": 0.6854, + "step": 2054 + }, + { + "epoch": 0.18374463519313305, + "grad_norm": 0.11797339066766298, + "learning_rate": 0.00018786235804583366, + "loss": 0.6895, + "step": 2055 + }, + { + "epoch": 0.1838340486409156, + "grad_norm": 0.11497814670776041, + "learning_rate": 0.00018784852547412565, + "loss": 0.7268, + "step": 2056 + }, + { + "epoch": 0.18392346208869814, + "grad_norm": 0.11421125306430319, + "learning_rate": 0.0001878346855346679, + "loss": 0.6992, + "step": 2057 + }, + { + "epoch": 0.18401287553648069, + "grad_norm": 0.10889300482609411, + "learning_rate": 0.00018782083822862114, + "loss": 0.6654, + "step": 2058 + }, + { + "epoch": 0.18410228898426323, + "grad_norm": 0.13222727854604616, + "learning_rate": 0.0001878069835571468, + "loss": 0.6783, + "step": 2059 + }, + { + "epoch": 0.18419170243204577, + "grad_norm": 0.10951904831211447, + "learning_rate": 0.00018779312152140674, + "loss": 0.6824, + "step": 2060 + }, + { + "epoch": 0.18428111587982832, + "grad_norm": 0.11129508368260747, + "learning_rate": 0.0001877792521225636, + "loss": 0.6726, + "step": 2061 + }, + { + "epoch": 0.18437052932761086, + "grad_norm": 0.14059375526462586, + "learning_rate": 0.00018776537536178064, + "loss": 0.668, + "step": 2062 + }, + { + "epoch": 0.18445994277539343, + "grad_norm": 0.10973653096802928, + "learning_rate": 0.00018775149124022162, + "loss": 0.6861, + "step": 2063 + }, + { + "epoch": 0.18454935622317598, + "grad_norm": 0.12398564787171538, + "learning_rate": 0.00018773759975905098, + "loss": 0.6722, + "step": 2064 + }, + { + "epoch": 0.18463876967095852, + "grad_norm": 0.11178171289081444, + "learning_rate": 0.00018772370091943384, + "loss": 0.6769, + "step": 2065 + }, + { + "epoch": 0.18472818311874106, + "grad_norm": 0.12397062170545527, + "learning_rate": 0.00018770979472253581, + "loss": 0.6984, + "step": 2066 + }, + { + "epoch": 0.1848175965665236, + "grad_norm": 0.12166022339719239, + "learning_rate": 0.0001876958811695233, + "loss": 0.6508, + "step": 2067 + }, + { + "epoch": 0.18490701001430615, + "grad_norm": 0.1256785645756714, + "learning_rate": 0.00018768196026156306, + "loss": 0.6688, + "step": 2068 + }, + { + "epoch": 0.1849964234620887, + "grad_norm": 0.1236121358742154, + "learning_rate": 0.00018766803199982273, + "loss": 0.6444, + "step": 2069 + }, + { + "epoch": 0.18508583690987124, + "grad_norm": 0.13077867413002525, + "learning_rate": 0.00018765409638547048, + "loss": 0.7031, + "step": 2070 + }, + { + "epoch": 0.18517525035765378, + "grad_norm": 0.11480493583768185, + "learning_rate": 0.00018764015341967498, + "loss": 0.65, + "step": 2071 + }, + { + "epoch": 0.18526466380543632, + "grad_norm": 0.1171198344519166, + "learning_rate": 0.00018762620310360567, + "loss": 0.6805, + "step": 2072 + }, + { + "epoch": 0.1853540772532189, + "grad_norm": 0.1264956171061027, + "learning_rate": 0.00018761224543843255, + "loss": 0.7385, + "step": 2073 + }, + { + "epoch": 0.18544349070100144, + "grad_norm": 0.12265237953563382, + "learning_rate": 0.00018759828042532616, + "loss": 0.6361, + "step": 2074 + }, + { + "epoch": 0.18553290414878398, + "grad_norm": 0.11277341330917284, + "learning_rate": 0.00018758430806545783, + "loss": 0.7109, + "step": 2075 + }, + { + "epoch": 0.18562231759656653, + "grad_norm": 0.12839452068675578, + "learning_rate": 0.00018757032835999931, + "loss": 0.7092, + "step": 2076 + }, + { + "epoch": 0.18571173104434907, + "grad_norm": 0.12367473633846386, + "learning_rate": 0.00018755634131012317, + "loss": 0.7034, + "step": 2077 + }, + { + "epoch": 0.18580114449213161, + "grad_norm": 0.13759261202918385, + "learning_rate": 0.00018754234691700238, + "loss": 0.7438, + "step": 2078 + }, + { + "epoch": 0.18589055793991416, + "grad_norm": 0.12301927405175342, + "learning_rate": 0.00018752834518181072, + "loss": 0.6832, + "step": 2079 + }, + { + "epoch": 0.1859799713876967, + "grad_norm": 0.1316013656171402, + "learning_rate": 0.00018751433610572242, + "loss": 0.7145, + "step": 2080 + }, + { + "epoch": 0.18606938483547925, + "grad_norm": 0.12842459433451206, + "learning_rate": 0.00018750031968991243, + "loss": 0.7016, + "step": 2081 + }, + { + "epoch": 0.1861587982832618, + "grad_norm": 0.14999854374768168, + "learning_rate": 0.00018748629593555633, + "loss": 0.6905, + "step": 2082 + }, + { + "epoch": 0.18624821173104436, + "grad_norm": 0.10968169908820198, + "learning_rate": 0.00018747226484383024, + "loss": 0.6543, + "step": 2083 + }, + { + "epoch": 0.1863376251788269, + "grad_norm": 0.11607882918011753, + "learning_rate": 0.00018745822641591094, + "loss": 0.6921, + "step": 2084 + }, + { + "epoch": 0.18642703862660945, + "grad_norm": 0.1215894735846465, + "learning_rate": 0.00018744418065297583, + "loss": 0.6878, + "step": 2085 + }, + { + "epoch": 0.186516452074392, + "grad_norm": 0.11857715210166284, + "learning_rate": 0.00018743012755620286, + "loss": 0.6741, + "step": 2086 + }, + { + "epoch": 0.18660586552217454, + "grad_norm": 0.11969943105736788, + "learning_rate": 0.0001874160671267707, + "loss": 0.7007, + "step": 2087 + }, + { + "epoch": 0.18669527896995708, + "grad_norm": 0.1071402786065064, + "learning_rate": 0.00018740199936585853, + "loss": 0.6789, + "step": 2088 + }, + { + "epoch": 0.18678469241773962, + "grad_norm": 0.11767771237971461, + "learning_rate": 0.00018738792427464625, + "loss": 0.672, + "step": 2089 + }, + { + "epoch": 0.18687410586552217, + "grad_norm": 0.11467685820851432, + "learning_rate": 0.00018737384185431432, + "loss": 0.6696, + "step": 2090 + }, + { + "epoch": 0.1869635193133047, + "grad_norm": 0.1239616070133864, + "learning_rate": 0.00018735975210604376, + "loss": 0.7138, + "step": 2091 + }, + { + "epoch": 0.18705293276108725, + "grad_norm": 0.12050085905029359, + "learning_rate": 0.00018734565503101636, + "loss": 0.7222, + "step": 2092 + }, + { + "epoch": 0.18714234620886983, + "grad_norm": 0.10731514628549013, + "learning_rate": 0.0001873315506304143, + "loss": 0.6373, + "step": 2093 + }, + { + "epoch": 0.18723175965665237, + "grad_norm": 0.10272501603254355, + "learning_rate": 0.00018731743890542058, + "loss": 0.6713, + "step": 2094 + }, + { + "epoch": 0.1873211731044349, + "grad_norm": 0.117248390696958, + "learning_rate": 0.0001873033198572187, + "loss": 0.6752, + "step": 2095 + }, + { + "epoch": 0.18741058655221746, + "grad_norm": 0.12299368792793841, + "learning_rate": 0.00018728919348699283, + "loss": 0.7051, + "step": 2096 + }, + { + "epoch": 0.1875, + "grad_norm": 0.11218848360087218, + "learning_rate": 0.0001872750597959277, + "loss": 0.7107, + "step": 2097 + }, + { + "epoch": 0.18758941344778254, + "grad_norm": 0.13094430064854448, + "learning_rate": 0.00018726091878520871, + "loss": 0.671, + "step": 2098 + }, + { + "epoch": 0.1876788268955651, + "grad_norm": 0.11334882037186338, + "learning_rate": 0.00018724677045602186, + "loss": 0.661, + "step": 2099 + }, + { + "epoch": 0.18776824034334763, + "grad_norm": 0.11584777189984079, + "learning_rate": 0.00018723261480955373, + "loss": 0.6848, + "step": 2100 + }, + { + "epoch": 0.18785765379113017, + "grad_norm": 0.11425084836907876, + "learning_rate": 0.00018721845184699158, + "loss": 0.6476, + "step": 2101 + }, + { + "epoch": 0.18794706723891275, + "grad_norm": 0.1109825870082431, + "learning_rate": 0.00018720428156952316, + "loss": 0.637, + "step": 2102 + }, + { + "epoch": 0.1880364806866953, + "grad_norm": 0.1285533357682346, + "learning_rate": 0.00018719010397833698, + "loss": 0.6775, + "step": 2103 + }, + { + "epoch": 0.18812589413447783, + "grad_norm": 0.11484318516693723, + "learning_rate": 0.00018717591907462208, + "loss": 0.6435, + "step": 2104 + }, + { + "epoch": 0.18821530758226038, + "grad_norm": 0.14069439099239242, + "learning_rate": 0.00018716172685956815, + "loss": 0.6729, + "step": 2105 + }, + { + "epoch": 0.18830472103004292, + "grad_norm": 0.10102537224107795, + "learning_rate": 0.0001871475273343654, + "loss": 0.6477, + "step": 2106 + }, + { + "epoch": 0.18839413447782546, + "grad_norm": 0.14267828034779848, + "learning_rate": 0.00018713332050020482, + "loss": 0.7052, + "step": 2107 + }, + { + "epoch": 0.188483547925608, + "grad_norm": 0.1134583809281424, + "learning_rate": 0.00018711910635827787, + "loss": 0.6547, + "step": 2108 + }, + { + "epoch": 0.18857296137339055, + "grad_norm": 0.12898823549513716, + "learning_rate": 0.0001871048849097767, + "loss": 0.7251, + "step": 2109 + }, + { + "epoch": 0.1886623748211731, + "grad_norm": 0.1194502216718756, + "learning_rate": 0.000187090656155894, + "loss": 0.6846, + "step": 2110 + }, + { + "epoch": 0.18875178826895564, + "grad_norm": 0.11962068013335615, + "learning_rate": 0.00018707642009782317, + "loss": 0.7165, + "step": 2111 + }, + { + "epoch": 0.1888412017167382, + "grad_norm": 0.13819554472912698, + "learning_rate": 0.00018706217673675811, + "loss": 0.7299, + "step": 2112 + }, + { + "epoch": 0.18893061516452075, + "grad_norm": 0.12482212741283001, + "learning_rate": 0.00018704792607389346, + "loss": 0.7203, + "step": 2113 + }, + { + "epoch": 0.1890200286123033, + "grad_norm": 0.11579494942285608, + "learning_rate": 0.00018703366811042438, + "loss": 0.6445, + "step": 2114 + }, + { + "epoch": 0.18910944206008584, + "grad_norm": 0.12570714358060783, + "learning_rate": 0.00018701940284754665, + "loss": 0.6903, + "step": 2115 + }, + { + "epoch": 0.18919885550786839, + "grad_norm": 0.1046931845695302, + "learning_rate": 0.00018700513028645672, + "loss": 0.6765, + "step": 2116 + }, + { + "epoch": 0.18928826895565093, + "grad_norm": 0.10556259454756706, + "learning_rate": 0.00018699085042835157, + "loss": 0.6603, + "step": 2117 + }, + { + "epoch": 0.18937768240343347, + "grad_norm": 0.13372043532165293, + "learning_rate": 0.00018697656327442888, + "loss": 0.2689, + "step": 2118 + }, + { + "epoch": 0.18946709585121602, + "grad_norm": 0.1160016619008874, + "learning_rate": 0.00018696226882588683, + "loss": 0.6787, + "step": 2119 + }, + { + "epoch": 0.18955650929899856, + "grad_norm": 0.11614496223521677, + "learning_rate": 0.00018694796708392436, + "loss": 0.6647, + "step": 2120 + }, + { + "epoch": 0.1896459227467811, + "grad_norm": 0.12203219327824058, + "learning_rate": 0.00018693365804974086, + "loss": 0.6968, + "step": 2121 + }, + { + "epoch": 0.18973533619456368, + "grad_norm": 0.11775590422353754, + "learning_rate": 0.00018691934172453646, + "loss": 0.6797, + "step": 2122 + }, + { + "epoch": 0.18982474964234622, + "grad_norm": 0.12994177611835503, + "learning_rate": 0.00018690501810951182, + "loss": 0.6739, + "step": 2123 + }, + { + "epoch": 0.18991416309012876, + "grad_norm": 0.12138538902857458, + "learning_rate": 0.0001868906872058683, + "loss": 0.6535, + "step": 2124 + }, + { + "epoch": 0.1900035765379113, + "grad_norm": 0.10784931503473241, + "learning_rate": 0.00018687634901480777, + "loss": 0.6608, + "step": 2125 + }, + { + "epoch": 0.19009298998569385, + "grad_norm": 0.1303832785695839, + "learning_rate": 0.00018686200353753275, + "loss": 0.6818, + "step": 2126 + }, + { + "epoch": 0.1901824034334764, + "grad_norm": 0.10920360963628667, + "learning_rate": 0.00018684765077524643, + "loss": 0.6821, + "step": 2127 + }, + { + "epoch": 0.19027181688125894, + "grad_norm": 0.1246493873340553, + "learning_rate": 0.00018683329072915252, + "loss": 0.6851, + "step": 2128 + }, + { + "epoch": 0.19036123032904148, + "grad_norm": 0.12399820731276776, + "learning_rate": 0.00018681892340045538, + "loss": 0.7251, + "step": 2129 + }, + { + "epoch": 0.19045064377682402, + "grad_norm": 0.11714517527411167, + "learning_rate": 0.00018680454879035997, + "loss": 0.6976, + "step": 2130 + }, + { + "epoch": 0.19054005722460657, + "grad_norm": 0.12312184555603943, + "learning_rate": 0.0001867901669000719, + "loss": 0.6515, + "step": 2131 + }, + { + "epoch": 0.19062947067238914, + "grad_norm": 0.12402214839944624, + "learning_rate": 0.00018677577773079733, + "loss": 0.6872, + "step": 2132 + }, + { + "epoch": 0.19071888412017168, + "grad_norm": 0.12361462850488579, + "learning_rate": 0.00018676138128374313, + "loss": 0.6951, + "step": 2133 + }, + { + "epoch": 0.19080829756795423, + "grad_norm": 0.13727775447959722, + "learning_rate": 0.0001867469775601166, + "loss": 0.7, + "step": 2134 + }, + { + "epoch": 0.19089771101573677, + "grad_norm": 0.11930825147864119, + "learning_rate": 0.00018673256656112584, + "loss": 0.6861, + "step": 2135 + }, + { + "epoch": 0.19098712446351931, + "grad_norm": 0.13359562204189754, + "learning_rate": 0.0001867181482879795, + "loss": 0.7034, + "step": 2136 + }, + { + "epoch": 0.19107653791130186, + "grad_norm": 0.15529860315585803, + "learning_rate": 0.00018670372274188677, + "loss": 0.3045, + "step": 2137 + }, + { + "epoch": 0.1911659513590844, + "grad_norm": 0.10608187779544179, + "learning_rate": 0.00018668928992405755, + "loss": 0.6458, + "step": 2138 + }, + { + "epoch": 0.19125536480686695, + "grad_norm": 0.13977992082458146, + "learning_rate": 0.00018667484983570223, + "loss": 0.6948, + "step": 2139 + }, + { + "epoch": 0.1913447782546495, + "grad_norm": 0.12355959298012625, + "learning_rate": 0.00018666040247803195, + "loss": 0.7104, + "step": 2140 + }, + { + "epoch": 0.19143419170243203, + "grad_norm": 0.11528597993527843, + "learning_rate": 0.0001866459478522584, + "loss": 0.623, + "step": 2141 + }, + { + "epoch": 0.1915236051502146, + "grad_norm": 0.10572106591221167, + "learning_rate": 0.0001866314859595938, + "loss": 0.6799, + "step": 2142 + }, + { + "epoch": 0.19161301859799715, + "grad_norm": 0.11417659717541001, + "learning_rate": 0.00018661701680125115, + "loss": 0.6622, + "step": 2143 + }, + { + "epoch": 0.1917024320457797, + "grad_norm": 0.13092311220190053, + "learning_rate": 0.00018660254037844388, + "loss": 0.7249, + "step": 2144 + }, + { + "epoch": 0.19179184549356224, + "grad_norm": 0.13532330784045207, + "learning_rate": 0.00018658805669238612, + "loss": 0.7092, + "step": 2145 + }, + { + "epoch": 0.19188125894134478, + "grad_norm": 0.11271119407448496, + "learning_rate": 0.00018657356574429266, + "loss": 0.6658, + "step": 2146 + }, + { + "epoch": 0.19197067238912732, + "grad_norm": 0.12970594022339205, + "learning_rate": 0.00018655906753537878, + "loss": 0.6729, + "step": 2147 + }, + { + "epoch": 0.19206008583690987, + "grad_norm": 0.11217702661359415, + "learning_rate": 0.00018654456206686042, + "loss": 0.6759, + "step": 2148 + }, + { + "epoch": 0.1921494992846924, + "grad_norm": 0.11667091560494959, + "learning_rate": 0.00018653004933995418, + "loss": 0.6988, + "step": 2149 + }, + { + "epoch": 0.19223891273247495, + "grad_norm": 0.12108995322662625, + "learning_rate": 0.00018651552935587717, + "loss": 0.6874, + "step": 2150 + }, + { + "epoch": 0.1923283261802575, + "grad_norm": 0.12329835130167624, + "learning_rate": 0.00018650100211584723, + "loss": 0.6559, + "step": 2151 + }, + { + "epoch": 0.19241773962804007, + "grad_norm": 0.12481788878627732, + "learning_rate": 0.00018648646762108273, + "loss": 0.6134, + "step": 2152 + }, + { + "epoch": 0.1925071530758226, + "grad_norm": 0.11475048248738222, + "learning_rate": 0.0001864719258728026, + "loss": 0.675, + "step": 2153 + }, + { + "epoch": 0.19259656652360516, + "grad_norm": 0.13084804171224373, + "learning_rate": 0.0001864573768722265, + "loss": 0.6818, + "step": 2154 + }, + { + "epoch": 0.1926859799713877, + "grad_norm": 0.12257867167528923, + "learning_rate": 0.0001864428206205746, + "loss": 0.6731, + "step": 2155 + }, + { + "epoch": 0.19277539341917024, + "grad_norm": 0.12003904886271301, + "learning_rate": 0.00018642825711906772, + "loss": 0.6825, + "step": 2156 + }, + { + "epoch": 0.1928648068669528, + "grad_norm": 0.11828267110177869, + "learning_rate": 0.00018641368636892734, + "loss": 0.6498, + "step": 2157 + }, + { + "epoch": 0.19295422031473533, + "grad_norm": 0.12827710648977397, + "learning_rate": 0.00018639910837137542, + "loss": 0.7202, + "step": 2158 + }, + { + "epoch": 0.19304363376251787, + "grad_norm": 0.11634581380319756, + "learning_rate": 0.0001863845231276346, + "loss": 0.6412, + "step": 2159 + }, + { + "epoch": 0.19313304721030042, + "grad_norm": 0.14164804662367353, + "learning_rate": 0.0001863699306389282, + "loss": 0.7169, + "step": 2160 + }, + { + "epoch": 0.193222460658083, + "grad_norm": 0.1259324388481718, + "learning_rate": 0.00018635533090647998, + "loss": 0.6656, + "step": 2161 + }, + { + "epoch": 0.19331187410586553, + "grad_norm": 0.11938033909086909, + "learning_rate": 0.00018634072393151446, + "loss": 0.6879, + "step": 2162 + }, + { + "epoch": 0.19340128755364808, + "grad_norm": 0.11674271002120613, + "learning_rate": 0.00018632610971525671, + "loss": 0.6686, + "step": 2163 + }, + { + "epoch": 0.19349070100143062, + "grad_norm": 0.13789278038589198, + "learning_rate": 0.00018631148825893238, + "loss": 0.694, + "step": 2164 + }, + { + "epoch": 0.19358011444921316, + "grad_norm": 0.11398190615583337, + "learning_rate": 0.00018629685956376779, + "loss": 0.6759, + "step": 2165 + }, + { + "epoch": 0.1936695278969957, + "grad_norm": 0.13271560191956944, + "learning_rate": 0.0001862822236309898, + "loss": 0.7165, + "step": 2166 + }, + { + "epoch": 0.19375894134477825, + "grad_norm": 0.1291486864014017, + "learning_rate": 0.0001862675804618259, + "loss": 0.6873, + "step": 2167 + }, + { + "epoch": 0.1938483547925608, + "grad_norm": 0.12595283569290597, + "learning_rate": 0.00018625293005750424, + "loss": 0.6799, + "step": 2168 + }, + { + "epoch": 0.19393776824034334, + "grad_norm": 0.13284961889422317, + "learning_rate": 0.00018623827241925347, + "loss": 0.6939, + "step": 2169 + }, + { + "epoch": 0.19402718168812588, + "grad_norm": 0.129088234481457, + "learning_rate": 0.000186223607548303, + "loss": 0.7075, + "step": 2170 + }, + { + "epoch": 0.19411659513590845, + "grad_norm": 0.13057246857187763, + "learning_rate": 0.00018620893544588264, + "loss": 0.6789, + "step": 2171 + }, + { + "epoch": 0.194206008583691, + "grad_norm": 0.1233999925699606, + "learning_rate": 0.00018619425611322298, + "loss": 0.6543, + "step": 2172 + }, + { + "epoch": 0.19429542203147354, + "grad_norm": 0.1271547678437462, + "learning_rate": 0.0001861795695515552, + "loss": 0.7084, + "step": 2173 + }, + { + "epoch": 0.19438483547925609, + "grad_norm": 0.12779912266122673, + "learning_rate": 0.00018616487576211092, + "loss": 0.7287, + "step": 2174 + }, + { + "epoch": 0.19447424892703863, + "grad_norm": 0.1211851755767301, + "learning_rate": 0.00018615017474612265, + "loss": 0.7045, + "step": 2175 + }, + { + "epoch": 0.19456366237482117, + "grad_norm": 0.12099319879123924, + "learning_rate": 0.00018613546650482322, + "loss": 0.7056, + "step": 2176 + }, + { + "epoch": 0.19465307582260372, + "grad_norm": 0.12953808520123425, + "learning_rate": 0.00018612075103944625, + "loss": 0.7287, + "step": 2177 + }, + { + "epoch": 0.19474248927038626, + "grad_norm": 0.11280208993065065, + "learning_rate": 0.00018610602835122592, + "loss": 0.6624, + "step": 2178 + }, + { + "epoch": 0.1948319027181688, + "grad_norm": 0.12082971799054473, + "learning_rate": 0.00018609129844139697, + "loss": 0.723, + "step": 2179 + }, + { + "epoch": 0.19492131616595135, + "grad_norm": 0.09954138207225698, + "learning_rate": 0.00018607656131119476, + "loss": 0.6681, + "step": 2180 + }, + { + "epoch": 0.19501072961373392, + "grad_norm": 0.11138645340841548, + "learning_rate": 0.00018606181696185535, + "loss": 0.6872, + "step": 2181 + }, + { + "epoch": 0.19510014306151646, + "grad_norm": 0.12178537862646713, + "learning_rate": 0.00018604706539461526, + "loss": 0.6766, + "step": 2182 + }, + { + "epoch": 0.195189556509299, + "grad_norm": 0.12194093586386003, + "learning_rate": 0.00018603230661071174, + "loss": 0.7022, + "step": 2183 + }, + { + "epoch": 0.19527896995708155, + "grad_norm": 0.12322195531068572, + "learning_rate": 0.00018601754061138256, + "loss": 0.6862, + "step": 2184 + }, + { + "epoch": 0.1953683834048641, + "grad_norm": 0.1334352551970223, + "learning_rate": 0.00018600276739786612, + "loss": 0.6765, + "step": 2185 + }, + { + "epoch": 0.19545779685264664, + "grad_norm": 0.13080040027604692, + "learning_rate": 0.00018598798697140145, + "loss": 0.6685, + "step": 2186 + }, + { + "epoch": 0.19554721030042918, + "grad_norm": 0.11828918198200694, + "learning_rate": 0.00018597319933322815, + "loss": 0.6604, + "step": 2187 + }, + { + "epoch": 0.19563662374821172, + "grad_norm": 0.11184401244199557, + "learning_rate": 0.0001859584044845865, + "loss": 0.639, + "step": 2188 + }, + { + "epoch": 0.19572603719599427, + "grad_norm": 0.12505309306715232, + "learning_rate": 0.0001859436024267172, + "loss": 0.7013, + "step": 2189 + }, + { + "epoch": 0.1958154506437768, + "grad_norm": 0.14632986345673263, + "learning_rate": 0.0001859287931608618, + "loss": 0.2875, + "step": 2190 + }, + { + "epoch": 0.19590486409155938, + "grad_norm": 0.13758154657980967, + "learning_rate": 0.00018591397668826228, + "loss": 0.7118, + "step": 2191 + }, + { + "epoch": 0.19599427753934193, + "grad_norm": 0.11041779179034607, + "learning_rate": 0.0001858991530101613, + "loss": 0.6673, + "step": 2192 + }, + { + "epoch": 0.19608369098712447, + "grad_norm": 0.12085657356086872, + "learning_rate": 0.00018588432212780212, + "loss": 0.6666, + "step": 2193 + }, + { + "epoch": 0.19617310443490701, + "grad_norm": 0.11568364942923723, + "learning_rate": 0.00018586948404242853, + "loss": 0.6723, + "step": 2194 + }, + { + "epoch": 0.19626251788268956, + "grad_norm": 0.1314494711055535, + "learning_rate": 0.00018585463875528505, + "loss": 0.6729, + "step": 2195 + }, + { + "epoch": 0.1963519313304721, + "grad_norm": 0.11498102740080547, + "learning_rate": 0.00018583978626761667, + "loss": 0.7008, + "step": 2196 + }, + { + "epoch": 0.19644134477825465, + "grad_norm": 0.12861330672230198, + "learning_rate": 0.00018582492658066909, + "loss": 0.7309, + "step": 2197 + }, + { + "epoch": 0.1965307582260372, + "grad_norm": 0.11580458148249599, + "learning_rate": 0.00018581005969568856, + "loss": 0.6445, + "step": 2198 + }, + { + "epoch": 0.19662017167381973, + "grad_norm": 0.13570499636621838, + "learning_rate": 0.00018579518561392198, + "loss": 0.7253, + "step": 2199 + }, + { + "epoch": 0.19670958512160228, + "grad_norm": 0.13060299098717798, + "learning_rate": 0.00018578030433661678, + "loss": 0.6955, + "step": 2200 + }, + { + "epoch": 0.19679899856938485, + "grad_norm": 0.133820259872612, + "learning_rate": 0.00018576541586502106, + "loss": 0.7363, + "step": 2201 + }, + { + "epoch": 0.1968884120171674, + "grad_norm": 0.11640220120568333, + "learning_rate": 0.00018575052020038352, + "loss": 0.6883, + "step": 2202 + }, + { + "epoch": 0.19697782546494993, + "grad_norm": 0.13608210680290378, + "learning_rate": 0.00018573561734395338, + "loss": 0.3007, + "step": 2203 + }, + { + "epoch": 0.19706723891273248, + "grad_norm": 0.11854434880731285, + "learning_rate": 0.0001857207072969805, + "loss": 0.6974, + "step": 2204 + }, + { + "epoch": 0.19715665236051502, + "grad_norm": 0.11883627993934075, + "learning_rate": 0.0001857057900607155, + "loss": 0.6909, + "step": 2205 + }, + { + "epoch": 0.19724606580829757, + "grad_norm": 0.11151737676132163, + "learning_rate": 0.0001856908656364094, + "loss": 0.6618, + "step": 2206 + }, + { + "epoch": 0.1973354792560801, + "grad_norm": 0.11408761665422348, + "learning_rate": 0.00018567593402531385, + "loss": 0.6664, + "step": 2207 + }, + { + "epoch": 0.19742489270386265, + "grad_norm": 0.12105422752751722, + "learning_rate": 0.00018566099522868119, + "loss": 0.6998, + "step": 2208 + }, + { + "epoch": 0.1975143061516452, + "grad_norm": 0.1347605989697998, + "learning_rate": 0.00018564604924776432, + "loss": 0.7438, + "step": 2209 + }, + { + "epoch": 0.19760371959942774, + "grad_norm": 0.12307361831992043, + "learning_rate": 0.00018563109608381675, + "loss": 0.6977, + "step": 2210 + }, + { + "epoch": 0.1976931330472103, + "grad_norm": 0.11505482162574462, + "learning_rate": 0.00018561613573809253, + "loss": 0.646, + "step": 2211 + }, + { + "epoch": 0.19778254649499286, + "grad_norm": 0.12593166277699677, + "learning_rate": 0.00018560116821184642, + "loss": 0.733, + "step": 2212 + }, + { + "epoch": 0.1978719599427754, + "grad_norm": 0.11330972258650607, + "learning_rate": 0.0001855861935063337, + "loss": 0.6354, + "step": 2213 + }, + { + "epoch": 0.19796137339055794, + "grad_norm": 0.11679181898161037, + "learning_rate": 0.00018557121162281033, + "loss": 0.6517, + "step": 2214 + }, + { + "epoch": 0.1980507868383405, + "grad_norm": 0.12674699779850374, + "learning_rate": 0.00018555622256253274, + "loss": 0.6964, + "step": 2215 + }, + { + "epoch": 0.19814020028612303, + "grad_norm": 0.10982670975659979, + "learning_rate": 0.00018554122632675815, + "loss": 0.6614, + "step": 2216 + }, + { + "epoch": 0.19822961373390557, + "grad_norm": 0.09908313603174916, + "learning_rate": 0.00018552622291674416, + "loss": 0.6717, + "step": 2217 + }, + { + "epoch": 0.19831902718168812, + "grad_norm": 0.12275108142452587, + "learning_rate": 0.00018551121233374915, + "loss": 0.6934, + "step": 2218 + }, + { + "epoch": 0.19840844062947066, + "grad_norm": 0.10540920496273533, + "learning_rate": 0.00018549619457903206, + "loss": 0.6543, + "step": 2219 + }, + { + "epoch": 0.1984978540772532, + "grad_norm": 0.10939580952493477, + "learning_rate": 0.00018548116965385236, + "loss": 0.6718, + "step": 2220 + }, + { + "epoch": 0.19858726752503578, + "grad_norm": 0.13238723869825184, + "learning_rate": 0.0001854661375594702, + "loss": 0.705, + "step": 2221 + }, + { + "epoch": 0.19867668097281832, + "grad_norm": 0.1485413218189658, + "learning_rate": 0.0001854510982971463, + "loss": 0.2714, + "step": 2222 + }, + { + "epoch": 0.19876609442060086, + "grad_norm": 0.12385308944599148, + "learning_rate": 0.000185436051868142, + "loss": 0.6668, + "step": 2223 + }, + { + "epoch": 0.1988555078683834, + "grad_norm": 0.10315119193721546, + "learning_rate": 0.0001854209982737192, + "loss": 0.6492, + "step": 2224 + }, + { + "epoch": 0.19894492131616595, + "grad_norm": 0.1270744819998033, + "learning_rate": 0.00018540593751514042, + "loss": 0.697, + "step": 2225 + }, + { + "epoch": 0.1990343347639485, + "grad_norm": 0.14042746442251028, + "learning_rate": 0.00018539086959366881, + "loss": 0.6624, + "step": 2226 + }, + { + "epoch": 0.19912374821173104, + "grad_norm": 0.13422158836044484, + "learning_rate": 0.00018537579451056811, + "loss": 0.667, + "step": 2227 + }, + { + "epoch": 0.19921316165951358, + "grad_norm": 0.11397344095354796, + "learning_rate": 0.00018536071226710267, + "loss": 0.2542, + "step": 2228 + }, + { + "epoch": 0.19930257510729613, + "grad_norm": 0.11535565476259065, + "learning_rate": 0.0001853456228645373, + "loss": 0.6784, + "step": 2229 + }, + { + "epoch": 0.1993919885550787, + "grad_norm": 0.13276242346757794, + "learning_rate": 0.00018533052630413766, + "loss": 0.6723, + "step": 2230 + }, + { + "epoch": 0.19948140200286124, + "grad_norm": 0.112305992908615, + "learning_rate": 0.00018531542258716982, + "loss": 0.6453, + "step": 2231 + }, + { + "epoch": 0.19957081545064378, + "grad_norm": 0.09829942668137528, + "learning_rate": 0.00018530031171490053, + "loss": 0.6903, + "step": 2232 + }, + { + "epoch": 0.19966022889842633, + "grad_norm": 0.11831212737098423, + "learning_rate": 0.0001852851936885971, + "loss": 0.7279, + "step": 2233 + }, + { + "epoch": 0.19974964234620887, + "grad_norm": 0.10751466533500735, + "learning_rate": 0.00018527006850952747, + "loss": 0.683, + "step": 2234 + }, + { + "epoch": 0.19983905579399142, + "grad_norm": 0.12394538223700734, + "learning_rate": 0.0001852549361789602, + "loss": 0.7174, + "step": 2235 + }, + { + "epoch": 0.19992846924177396, + "grad_norm": 0.13744828194299016, + "learning_rate": 0.00018523979669816438, + "loss": 0.6852, + "step": 2236 + }, + { + "epoch": 0.2000178826895565, + "grad_norm": 0.1280777995389496, + "learning_rate": 0.00018522465006840975, + "loss": 0.7186, + "step": 2237 + }, + { + "epoch": 0.20010729613733905, + "grad_norm": 0.1254112475522452, + "learning_rate": 0.00018520949629096664, + "loss": 0.7171, + "step": 2238 + }, + { + "epoch": 0.2001967095851216, + "grad_norm": 0.12135296611248832, + "learning_rate": 0.000185194335367106, + "loss": 0.6818, + "step": 2239 + }, + { + "epoch": 0.20028612303290416, + "grad_norm": 0.12355443105448846, + "learning_rate": 0.0001851791672980993, + "loss": 0.7073, + "step": 2240 + }, + { + "epoch": 0.2003755364806867, + "grad_norm": 0.11558514975642718, + "learning_rate": 0.0001851639920852188, + "loss": 0.6401, + "step": 2241 + }, + { + "epoch": 0.20046494992846925, + "grad_norm": 0.1302058783232784, + "learning_rate": 0.00018514880972973706, + "loss": 0.7, + "step": 2242 + }, + { + "epoch": 0.2005543633762518, + "grad_norm": 0.104016625716741, + "learning_rate": 0.0001851336202329275, + "loss": 0.6583, + "step": 2243 + }, + { + "epoch": 0.20064377682403434, + "grad_norm": 0.12424885808274497, + "learning_rate": 0.00018511842359606403, + "loss": 0.7323, + "step": 2244 + }, + { + "epoch": 0.20073319027181688, + "grad_norm": 0.13481514405353437, + "learning_rate": 0.00018510321982042116, + "loss": 0.7043, + "step": 2245 + }, + { + "epoch": 0.20082260371959942, + "grad_norm": 0.14918055438909028, + "learning_rate": 0.00018508800890727403, + "loss": 0.2887, + "step": 2246 + }, + { + "epoch": 0.20091201716738197, + "grad_norm": 0.1403337567668427, + "learning_rate": 0.00018507279085789834, + "loss": 0.7335, + "step": 2247 + }, + { + "epoch": 0.2010014306151645, + "grad_norm": 0.10841556368206934, + "learning_rate": 0.00018505756567357046, + "loss": 0.685, + "step": 2248 + }, + { + "epoch": 0.20109084406294706, + "grad_norm": 0.1266966230097476, + "learning_rate": 0.00018504233335556723, + "loss": 0.7007, + "step": 2249 + }, + { + "epoch": 0.20118025751072963, + "grad_norm": 0.12081819199309156, + "learning_rate": 0.00018502709390516624, + "loss": 0.6867, + "step": 2250 + }, + { + "epoch": 0.20126967095851217, + "grad_norm": 0.13465842274746267, + "learning_rate": 0.00018501184732364553, + "loss": 0.6748, + "step": 2251 + }, + { + "epoch": 0.2013590844062947, + "grad_norm": 0.12572161429233608, + "learning_rate": 0.0001849965936122839, + "loss": 0.6678, + "step": 2252 + }, + { + "epoch": 0.20144849785407726, + "grad_norm": 0.12624398312215274, + "learning_rate": 0.00018498133277236058, + "loss": 0.6724, + "step": 2253 + }, + { + "epoch": 0.2015379113018598, + "grad_norm": 0.1344216315127687, + "learning_rate": 0.00018496606480515552, + "loss": 0.6999, + "step": 2254 + }, + { + "epoch": 0.20162732474964234, + "grad_norm": 0.12112016656681843, + "learning_rate": 0.0001849507897119492, + "loss": 0.6771, + "step": 2255 + }, + { + "epoch": 0.2017167381974249, + "grad_norm": 0.10471241063542387, + "learning_rate": 0.00018493550749402278, + "loss": 0.6448, + "step": 2256 + }, + { + "epoch": 0.20180615164520743, + "grad_norm": 0.11736865052233778, + "learning_rate": 0.0001849202181526579, + "loss": 0.673, + "step": 2257 + }, + { + "epoch": 0.20189556509298998, + "grad_norm": 0.11796823936400788, + "learning_rate": 0.00018490492168913688, + "loss": 0.6804, + "step": 2258 + }, + { + "epoch": 0.20198497854077252, + "grad_norm": 0.1574185002663257, + "learning_rate": 0.00018488961810474264, + "loss": 0.2868, + "step": 2259 + }, + { + "epoch": 0.2020743919885551, + "grad_norm": 0.14058527645253693, + "learning_rate": 0.00018487430740075862, + "loss": 0.7175, + "step": 2260 + }, + { + "epoch": 0.20216380543633763, + "grad_norm": 0.12609736082472112, + "learning_rate": 0.00018485898957846896, + "loss": 0.6601, + "step": 2261 + }, + { + "epoch": 0.20225321888412018, + "grad_norm": 0.11327913430656038, + "learning_rate": 0.0001848436646391583, + "loss": 0.6724, + "step": 2262 + }, + { + "epoch": 0.20234263233190272, + "grad_norm": 0.12685856458885084, + "learning_rate": 0.000184828332584112, + "loss": 0.7148, + "step": 2263 + }, + { + "epoch": 0.20243204577968527, + "grad_norm": 0.13566381327709356, + "learning_rate": 0.00018481299341461583, + "loss": 0.6588, + "step": 2264 + }, + { + "epoch": 0.2025214592274678, + "grad_norm": 0.11413390844003321, + "learning_rate": 0.0001847976471319564, + "loss": 0.678, + "step": 2265 + }, + { + "epoch": 0.20261087267525035, + "grad_norm": 0.13647740020885768, + "learning_rate": 0.00018478229373742065, + "loss": 0.6847, + "step": 2266 + }, + { + "epoch": 0.2027002861230329, + "grad_norm": 0.1259604772612776, + "learning_rate": 0.00018476693323229637, + "loss": 0.6862, + "step": 2267 + }, + { + "epoch": 0.20278969957081544, + "grad_norm": 0.13769015458574793, + "learning_rate": 0.00018475156561787172, + "loss": 0.7117, + "step": 2268 + }, + { + "epoch": 0.20287911301859798, + "grad_norm": 0.11788344138287211, + "learning_rate": 0.00018473619089543565, + "loss": 0.6631, + "step": 2269 + }, + { + "epoch": 0.20296852646638056, + "grad_norm": 0.11811510942421688, + "learning_rate": 0.00018472080906627758, + "loss": 0.7095, + "step": 2270 + }, + { + "epoch": 0.2030579399141631, + "grad_norm": 0.12492510243117734, + "learning_rate": 0.00018470542013168757, + "loss": 0.6785, + "step": 2271 + }, + { + "epoch": 0.20314735336194564, + "grad_norm": 0.1256484582335202, + "learning_rate": 0.00018469002409295628, + "loss": 0.7367, + "step": 2272 + }, + { + "epoch": 0.2032367668097282, + "grad_norm": 0.10734464419904544, + "learning_rate": 0.00018467462095137494, + "loss": 0.6934, + "step": 2273 + }, + { + "epoch": 0.20332618025751073, + "grad_norm": 0.12360742109801291, + "learning_rate": 0.0001846592107082354, + "loss": 0.6926, + "step": 2274 + }, + { + "epoch": 0.20341559370529327, + "grad_norm": 0.12030225865436213, + "learning_rate": 0.0001846437933648301, + "loss": 0.7068, + "step": 2275 + }, + { + "epoch": 0.20350500715307582, + "grad_norm": 0.10462463182617215, + "learning_rate": 0.00018462836892245207, + "loss": 0.6996, + "step": 2276 + }, + { + "epoch": 0.20359442060085836, + "grad_norm": 0.10974190321793187, + "learning_rate": 0.00018461293738239495, + "loss": 0.7016, + "step": 2277 + }, + { + "epoch": 0.2036838340486409, + "grad_norm": 0.14183197332766756, + "learning_rate": 0.00018459749874595298, + "loss": 0.3237, + "step": 2278 + }, + { + "epoch": 0.20377324749642345, + "grad_norm": 0.12613572538421922, + "learning_rate": 0.00018458205301442093, + "loss": 0.692, + "step": 2279 + }, + { + "epoch": 0.20386266094420602, + "grad_norm": 0.11551004653670914, + "learning_rate": 0.00018456660018909425, + "loss": 0.665, + "step": 2280 + }, + { + "epoch": 0.20395207439198856, + "grad_norm": 0.1280867730569401, + "learning_rate": 0.0001845511402712689, + "loss": 0.7334, + "step": 2281 + }, + { + "epoch": 0.2040414878397711, + "grad_norm": 0.10898516631176446, + "learning_rate": 0.0001845356732622416, + "loss": 0.6819, + "step": 2282 + }, + { + "epoch": 0.20413090128755365, + "grad_norm": 0.10540748619629815, + "learning_rate": 0.00018452019916330944, + "loss": 0.6681, + "step": 2283 + }, + { + "epoch": 0.2042203147353362, + "grad_norm": 0.12255551831950233, + "learning_rate": 0.00018450471797577028, + "loss": 0.709, + "step": 2284 + }, + { + "epoch": 0.20430972818311874, + "grad_norm": 0.10979765980797744, + "learning_rate": 0.00018448922970092243, + "loss": 0.6612, + "step": 2285 + }, + { + "epoch": 0.20439914163090128, + "grad_norm": 0.10905622128405125, + "learning_rate": 0.00018447373434006496, + "loss": 0.7078, + "step": 2286 + }, + { + "epoch": 0.20448855507868383, + "grad_norm": 0.1109150280542635, + "learning_rate": 0.0001844582318944974, + "loss": 0.6581, + "step": 2287 + }, + { + "epoch": 0.20457796852646637, + "grad_norm": 0.12421270930966803, + "learning_rate": 0.0001844427223655199, + "loss": 0.6963, + "step": 2288 + }, + { + "epoch": 0.20466738197424894, + "grad_norm": 0.13180057479413254, + "learning_rate": 0.0001844272057544333, + "loss": 0.6865, + "step": 2289 + }, + { + "epoch": 0.20475679542203148, + "grad_norm": 0.10691679766860536, + "learning_rate": 0.00018441168206253893, + "loss": 0.619, + "step": 2290 + }, + { + "epoch": 0.20484620886981403, + "grad_norm": 0.14225992425930234, + "learning_rate": 0.00018439615129113866, + "loss": 0.734, + "step": 2291 + }, + { + "epoch": 0.20493562231759657, + "grad_norm": 0.14796422591638217, + "learning_rate": 0.00018438061344153517, + "loss": 0.2698, + "step": 2292 + }, + { + "epoch": 0.20502503576537912, + "grad_norm": 0.1321388027643194, + "learning_rate": 0.0001843650685150315, + "loss": 0.7082, + "step": 2293 + }, + { + "epoch": 0.20511444921316166, + "grad_norm": 0.1465122534637134, + "learning_rate": 0.00018434951651293143, + "loss": 0.2997, + "step": 2294 + }, + { + "epoch": 0.2052038626609442, + "grad_norm": 0.12918031808352776, + "learning_rate": 0.0001843339574365393, + "loss": 0.6593, + "step": 2295 + }, + { + "epoch": 0.20529327610872675, + "grad_norm": 0.09798067194328003, + "learning_rate": 0.00018431839128715997, + "loss": 0.666, + "step": 2296 + }, + { + "epoch": 0.2053826895565093, + "grad_norm": 0.11335240171558535, + "learning_rate": 0.000184302818066099, + "loss": 0.6521, + "step": 2297 + }, + { + "epoch": 0.20547210300429183, + "grad_norm": 0.13703315709169864, + "learning_rate": 0.00018428723777466253, + "loss": 0.7111, + "step": 2298 + }, + { + "epoch": 0.2055615164520744, + "grad_norm": 0.13664881823016894, + "learning_rate": 0.0001842716504141572, + "loss": 0.7103, + "step": 2299 + }, + { + "epoch": 0.20565092989985695, + "grad_norm": 0.12515255059120314, + "learning_rate": 0.00018425605598589031, + "loss": 0.6931, + "step": 2300 + }, + { + "epoch": 0.2057403433476395, + "grad_norm": 0.12464237441517004, + "learning_rate": 0.00018424045449116978, + "loss": 0.7389, + "step": 2301 + }, + { + "epoch": 0.20582975679542204, + "grad_norm": 0.1245096733096311, + "learning_rate": 0.000184224845931304, + "loss": 0.6654, + "step": 2302 + }, + { + "epoch": 0.20591917024320458, + "grad_norm": 0.1280030928990073, + "learning_rate": 0.0001842092303076022, + "loss": 0.6845, + "step": 2303 + }, + { + "epoch": 0.20600858369098712, + "grad_norm": 0.13122312315446763, + "learning_rate": 0.00018419360762137395, + "loss": 0.7136, + "step": 2304 + }, + { + "epoch": 0.20609799713876967, + "grad_norm": 0.12881602081218127, + "learning_rate": 0.00018417797787392948, + "loss": 0.3027, + "step": 2305 + }, + { + "epoch": 0.2061874105865522, + "grad_norm": 0.11551238980240634, + "learning_rate": 0.00018416234106657963, + "loss": 0.2574, + "step": 2306 + }, + { + "epoch": 0.20627682403433475, + "grad_norm": 0.14369090781209715, + "learning_rate": 0.00018414669720063592, + "loss": 0.6663, + "step": 2307 + }, + { + "epoch": 0.2063662374821173, + "grad_norm": 0.1273703860147821, + "learning_rate": 0.00018413104627741035, + "loss": 0.6426, + "step": 2308 + }, + { + "epoch": 0.20645565092989987, + "grad_norm": 0.13731580532811188, + "learning_rate": 0.00018411538829821552, + "loss": 0.6757, + "step": 2309 + }, + { + "epoch": 0.2065450643776824, + "grad_norm": 0.14438068899474854, + "learning_rate": 0.00018409972326436465, + "loss": 0.7373, + "step": 2310 + }, + { + "epoch": 0.20663447782546496, + "grad_norm": 0.1489694749518332, + "learning_rate": 0.00018408405117717154, + "loss": 0.6965, + "step": 2311 + }, + { + "epoch": 0.2067238912732475, + "grad_norm": 0.19361853370375967, + "learning_rate": 0.00018406837203795067, + "loss": 0.3166, + "step": 2312 + }, + { + "epoch": 0.20681330472103004, + "grad_norm": 0.1324152783672681, + "learning_rate": 0.0001840526858480169, + "loss": 0.6715, + "step": 2313 + }, + { + "epoch": 0.2069027181688126, + "grad_norm": 0.12659757547501205, + "learning_rate": 0.0001840369926086859, + "loss": 0.6703, + "step": 2314 + }, + { + "epoch": 0.20699213161659513, + "grad_norm": 0.12100453295222745, + "learning_rate": 0.00018402129232127383, + "loss": 0.6441, + "step": 2315 + }, + { + "epoch": 0.20708154506437768, + "grad_norm": 0.13163804154284012, + "learning_rate": 0.00018400558498709744, + "loss": 0.6866, + "step": 2316 + }, + { + "epoch": 0.20717095851216022, + "grad_norm": 0.11385217599394282, + "learning_rate": 0.00018398987060747407, + "loss": 0.6656, + "step": 2317 + }, + { + "epoch": 0.20726037195994276, + "grad_norm": 0.10272187151222267, + "learning_rate": 0.00018397414918372172, + "loss": 0.6792, + "step": 2318 + }, + { + "epoch": 0.20734978540772533, + "grad_norm": 0.12954470710594282, + "learning_rate": 0.00018395842071715888, + "loss": 0.7303, + "step": 2319 + }, + { + "epoch": 0.20743919885550788, + "grad_norm": 0.1161654922022769, + "learning_rate": 0.00018394268520910466, + "loss": 0.6233, + "step": 2320 + }, + { + "epoch": 0.20752861230329042, + "grad_norm": 0.11546364483976701, + "learning_rate": 0.00018392694266087885, + "loss": 0.6597, + "step": 2321 + }, + { + "epoch": 0.20761802575107297, + "grad_norm": 0.11914451325892346, + "learning_rate": 0.00018391119307380172, + "loss": 0.7033, + "step": 2322 + }, + { + "epoch": 0.2077074391988555, + "grad_norm": 0.13259492922480617, + "learning_rate": 0.00018389543644919414, + "loss": 0.7001, + "step": 2323 + }, + { + "epoch": 0.20779685264663805, + "grad_norm": 0.13745485888957434, + "learning_rate": 0.00018387967278837763, + "loss": 0.7182, + "step": 2324 + }, + { + "epoch": 0.2078862660944206, + "grad_norm": 0.12711972516989017, + "learning_rate": 0.00018386390209267428, + "loss": 0.7026, + "step": 2325 + }, + { + "epoch": 0.20797567954220314, + "grad_norm": 0.12409649829815621, + "learning_rate": 0.00018384812436340672, + "loss": 0.6838, + "step": 2326 + }, + { + "epoch": 0.20806509298998568, + "grad_norm": 0.11589406827435533, + "learning_rate": 0.00018383233960189826, + "loss": 0.6603, + "step": 2327 + }, + { + "epoch": 0.20815450643776823, + "grad_norm": 0.10782357118663273, + "learning_rate": 0.0001838165478094727, + "loss": 0.6158, + "step": 2328 + }, + { + "epoch": 0.2082439198855508, + "grad_norm": 0.10996942190200032, + "learning_rate": 0.0001838007489874545, + "loss": 0.6671, + "step": 2329 + }, + { + "epoch": 0.20833333333333334, + "grad_norm": 0.15200793155637787, + "learning_rate": 0.0001837849431371687, + "loss": 0.6783, + "step": 2330 + }, + { + "epoch": 0.2084227467811159, + "grad_norm": 0.12690122137035917, + "learning_rate": 0.0001837691302599409, + "loss": 0.6921, + "step": 2331 + }, + { + "epoch": 0.20851216022889843, + "grad_norm": 0.1416624731236192, + "learning_rate": 0.0001837533103570973, + "loss": 0.6944, + "step": 2332 + }, + { + "epoch": 0.20860157367668097, + "grad_norm": 0.1154553794067739, + "learning_rate": 0.00018373748342996474, + "loss": 0.6873, + "step": 2333 + }, + { + "epoch": 0.20869098712446352, + "grad_norm": 0.12966072640269916, + "learning_rate": 0.00018372164947987054, + "loss": 0.6613, + "step": 2334 + }, + { + "epoch": 0.20878040057224606, + "grad_norm": 0.12581144837711436, + "learning_rate": 0.00018370580850814272, + "loss": 0.6723, + "step": 2335 + }, + { + "epoch": 0.2088698140200286, + "grad_norm": 0.12018900304006519, + "learning_rate": 0.00018368996051610986, + "loss": 0.6516, + "step": 2336 + }, + { + "epoch": 0.20895922746781115, + "grad_norm": 0.12157197234670081, + "learning_rate": 0.00018367410550510104, + "loss": 0.6672, + "step": 2337 + }, + { + "epoch": 0.2090486409155937, + "grad_norm": 0.1254378528332755, + "learning_rate": 0.00018365824347644607, + "loss": 0.6991, + "step": 2338 + }, + { + "epoch": 0.20913805436337626, + "grad_norm": 0.11457818177838726, + "learning_rate": 0.00018364237443147525, + "loss": 0.6671, + "step": 2339 + }, + { + "epoch": 0.2092274678111588, + "grad_norm": 0.127649495427195, + "learning_rate": 0.00018362649837151947, + "loss": 0.6916, + "step": 2340 + }, + { + "epoch": 0.20931688125894135, + "grad_norm": 0.12048340723685758, + "learning_rate": 0.0001836106152979103, + "loss": 0.6717, + "step": 2341 + }, + { + "epoch": 0.2094062947067239, + "grad_norm": 0.12433566270020836, + "learning_rate": 0.0001835947252119798, + "loss": 0.6941, + "step": 2342 + }, + { + "epoch": 0.20949570815450644, + "grad_norm": 0.11690214967372745, + "learning_rate": 0.00018357882811506065, + "loss": 0.664, + "step": 2343 + }, + { + "epoch": 0.20958512160228898, + "grad_norm": 0.12438202137116093, + "learning_rate": 0.00018356292400848611, + "loss": 0.7209, + "step": 2344 + }, + { + "epoch": 0.20967453505007153, + "grad_norm": 0.12563110163856775, + "learning_rate": 0.00018354701289359005, + "loss": 0.6894, + "step": 2345 + }, + { + "epoch": 0.20976394849785407, + "grad_norm": 0.12091554082162337, + "learning_rate": 0.00018353109477170696, + "loss": 0.7, + "step": 2346 + }, + { + "epoch": 0.2098533619456366, + "grad_norm": 0.10156129198878062, + "learning_rate": 0.0001835151696441718, + "loss": 0.6821, + "step": 2347 + }, + { + "epoch": 0.20994277539341916, + "grad_norm": 0.11565282932294428, + "learning_rate": 0.00018349923751232022, + "loss": 0.6623, + "step": 2348 + }, + { + "epoch": 0.21003218884120173, + "grad_norm": 0.14358338036328977, + "learning_rate": 0.00018348329837748843, + "loss": 0.6956, + "step": 2349 + }, + { + "epoch": 0.21012160228898427, + "grad_norm": 0.12484876668218507, + "learning_rate": 0.00018346735224101325, + "loss": 0.6649, + "step": 2350 + }, + { + "epoch": 0.21021101573676682, + "grad_norm": 0.11409246699460678, + "learning_rate": 0.000183451399104232, + "loss": 0.6933, + "step": 2351 + }, + { + "epoch": 0.21030042918454936, + "grad_norm": 0.11816804932883553, + "learning_rate": 0.00018343543896848273, + "loss": 0.6724, + "step": 2352 + }, + { + "epoch": 0.2103898426323319, + "grad_norm": 0.11545512022870323, + "learning_rate": 0.00018341947183510393, + "loss": 0.6573, + "step": 2353 + }, + { + "epoch": 0.21047925608011445, + "grad_norm": 0.1332222890847795, + "learning_rate": 0.00018340349770543481, + "loss": 0.6982, + "step": 2354 + }, + { + "epoch": 0.210568669527897, + "grad_norm": 0.12216818237615279, + "learning_rate": 0.00018338751658081504, + "loss": 0.6712, + "step": 2355 + }, + { + "epoch": 0.21065808297567953, + "grad_norm": 0.14219689254750983, + "learning_rate": 0.00018337152846258493, + "loss": 0.7142, + "step": 2356 + }, + { + "epoch": 0.21074749642346208, + "grad_norm": 0.12430383098316003, + "learning_rate": 0.00018335553335208546, + "loss": 0.658, + "step": 2357 + }, + { + "epoch": 0.21083690987124465, + "grad_norm": 0.12668442785407083, + "learning_rate": 0.00018333953125065805, + "loss": 0.7023, + "step": 2358 + }, + { + "epoch": 0.2109263233190272, + "grad_norm": 0.1306496692111917, + "learning_rate": 0.0001833235221596448, + "loss": 0.6465, + "step": 2359 + }, + { + "epoch": 0.21101573676680974, + "grad_norm": 0.1238119473150666, + "learning_rate": 0.00018330750608038844, + "loss": 0.7063, + "step": 2360 + }, + { + "epoch": 0.21110515021459228, + "grad_norm": 0.11374259522461132, + "learning_rate": 0.0001832914830142321, + "loss": 0.6364, + "step": 2361 + }, + { + "epoch": 0.21119456366237482, + "grad_norm": 0.11192372225575102, + "learning_rate": 0.00018327545296251968, + "loss": 0.7091, + "step": 2362 + }, + { + "epoch": 0.21128397711015737, + "grad_norm": 0.11956607491492441, + "learning_rate": 0.00018325941592659553, + "loss": 0.6732, + "step": 2363 + }, + { + "epoch": 0.2113733905579399, + "grad_norm": 0.12118675926678218, + "learning_rate": 0.0001832433719078048, + "loss": 0.6964, + "step": 2364 + }, + { + "epoch": 0.21146280400572245, + "grad_norm": 0.1229728317186926, + "learning_rate": 0.00018322732090749296, + "loss": 0.6472, + "step": 2365 + }, + { + "epoch": 0.211552217453505, + "grad_norm": 0.15140862653197779, + "learning_rate": 0.00018321126292700628, + "loss": 0.6949, + "step": 2366 + }, + { + "epoch": 0.21164163090128754, + "grad_norm": 0.1445428716238848, + "learning_rate": 0.00018319519796769143, + "loss": 0.7238, + "step": 2367 + }, + { + "epoch": 0.2117310443490701, + "grad_norm": 0.1293484997573991, + "learning_rate": 0.0001831791260308958, + "loss": 0.7038, + "step": 2368 + }, + { + "epoch": 0.21182045779685266, + "grad_norm": 0.11175636855419786, + "learning_rate": 0.00018316304711796732, + "loss": 0.668, + "step": 2369 + }, + { + "epoch": 0.2119098712446352, + "grad_norm": 0.1092071683978567, + "learning_rate": 0.00018314696123025454, + "loss": 0.6945, + "step": 2370 + }, + { + "epoch": 0.21199928469241774, + "grad_norm": 0.12443132930104757, + "learning_rate": 0.0001831308683691065, + "loss": 0.6865, + "step": 2371 + }, + { + "epoch": 0.2120886981402003, + "grad_norm": 0.11772892009384432, + "learning_rate": 0.00018311476853587297, + "loss": 0.7048, + "step": 2372 + }, + { + "epoch": 0.21217811158798283, + "grad_norm": 0.10535756335742402, + "learning_rate": 0.00018309866173190416, + "loss": 0.6362, + "step": 2373 + }, + { + "epoch": 0.21226752503576538, + "grad_norm": 0.14044981242246238, + "learning_rate": 0.00018308254795855095, + "loss": 0.7242, + "step": 2374 + }, + { + "epoch": 0.21235693848354792, + "grad_norm": 0.10698319354951996, + "learning_rate": 0.00018306642721716476, + "loss": 0.6617, + "step": 2375 + }, + { + "epoch": 0.21244635193133046, + "grad_norm": 0.11625974142741634, + "learning_rate": 0.00018305029950909768, + "loss": 0.6475, + "step": 2376 + }, + { + "epoch": 0.212535765379113, + "grad_norm": 0.13114671844320583, + "learning_rate": 0.00018303416483570227, + "loss": 0.6808, + "step": 2377 + }, + { + "epoch": 0.21262517882689558, + "grad_norm": 0.1250438936238546, + "learning_rate": 0.0001830180231983317, + "loss": 0.6786, + "step": 2378 + }, + { + "epoch": 0.21271459227467812, + "grad_norm": 0.13467418029756628, + "learning_rate": 0.00018300187459833981, + "loss": 0.6758, + "step": 2379 + }, + { + "epoch": 0.21280400572246067, + "grad_norm": 0.12952204931783876, + "learning_rate": 0.00018298571903708092, + "loss": 0.6338, + "step": 2380 + }, + { + "epoch": 0.2128934191702432, + "grad_norm": 0.11953391541995198, + "learning_rate": 0.00018296955651591002, + "loss": 0.7012, + "step": 2381 + }, + { + "epoch": 0.21298283261802575, + "grad_norm": 0.1120255613296604, + "learning_rate": 0.00018295338703618258, + "loss": 0.6604, + "step": 2382 + }, + { + "epoch": 0.2130722460658083, + "grad_norm": 0.11846254709206644, + "learning_rate": 0.0001829372105992548, + "loss": 0.651, + "step": 2383 + }, + { + "epoch": 0.21316165951359084, + "grad_norm": 0.12144990669813821, + "learning_rate": 0.00018292102720648333, + "loss": 0.6798, + "step": 2384 + }, + { + "epoch": 0.21325107296137338, + "grad_norm": 0.13190762746665938, + "learning_rate": 0.0001829048368592254, + "loss": 0.6708, + "step": 2385 + }, + { + "epoch": 0.21334048640915593, + "grad_norm": 0.14314487259997874, + "learning_rate": 0.00018288863955883897, + "loss": 0.681, + "step": 2386 + }, + { + "epoch": 0.21342989985693847, + "grad_norm": 0.12425399563232989, + "learning_rate": 0.00018287243530668243, + "loss": 0.6803, + "step": 2387 + }, + { + "epoch": 0.21351931330472104, + "grad_norm": 0.12593436179054918, + "learning_rate": 0.00018285622410411484, + "loss": 0.6717, + "step": 2388 + }, + { + "epoch": 0.21360872675250359, + "grad_norm": 0.11921019429536658, + "learning_rate": 0.00018284000595249577, + "loss": 0.6659, + "step": 2389 + }, + { + "epoch": 0.21369814020028613, + "grad_norm": 0.12447291744584094, + "learning_rate": 0.00018282378085318545, + "loss": 0.6845, + "step": 2390 + }, + { + "epoch": 0.21378755364806867, + "grad_norm": 0.13209582724795677, + "learning_rate": 0.00018280754880754468, + "loss": 0.659, + "step": 2391 + }, + { + "epoch": 0.21387696709585122, + "grad_norm": 0.11210395883721672, + "learning_rate": 0.0001827913098169348, + "loss": 0.6384, + "step": 2392 + }, + { + "epoch": 0.21396638054363376, + "grad_norm": 0.11968394512562339, + "learning_rate": 0.00018277506388271773, + "loss": 0.6993, + "step": 2393 + }, + { + "epoch": 0.2140557939914163, + "grad_norm": 0.13621374973204212, + "learning_rate": 0.000182758811006256, + "loss": 0.7059, + "step": 2394 + }, + { + "epoch": 0.21414520743919885, + "grad_norm": 0.12566294311590323, + "learning_rate": 0.0001827425511889128, + "loss": 0.7125, + "step": 2395 + }, + { + "epoch": 0.2142346208869814, + "grad_norm": 0.11078255315982151, + "learning_rate": 0.00018272628443205172, + "loss": 0.668, + "step": 2396 + }, + { + "epoch": 0.21432403433476394, + "grad_norm": 0.10078674632597295, + "learning_rate": 0.00018271001073703706, + "loss": 0.6543, + "step": 2397 + }, + { + "epoch": 0.2144134477825465, + "grad_norm": 0.11589001836117074, + "learning_rate": 0.0001826937301052337, + "loss": 0.6763, + "step": 2398 + }, + { + "epoch": 0.21450286123032905, + "grad_norm": 0.13798529580510668, + "learning_rate": 0.00018267744253800707, + "loss": 0.6841, + "step": 2399 + }, + { + "epoch": 0.2145922746781116, + "grad_norm": 0.12590878985310044, + "learning_rate": 0.00018266114803672318, + "loss": 0.7059, + "step": 2400 + }, + { + "epoch": 0.21468168812589414, + "grad_norm": 0.11917463846993674, + "learning_rate": 0.00018264484660274866, + "loss": 0.7056, + "step": 2401 + }, + { + "epoch": 0.21477110157367668, + "grad_norm": 0.12542338219947927, + "learning_rate": 0.00018262853823745062, + "loss": 0.6793, + "step": 2402 + }, + { + "epoch": 0.21486051502145923, + "grad_norm": 0.10633838862708932, + "learning_rate": 0.0001826122229421969, + "loss": 0.6279, + "step": 2403 + }, + { + "epoch": 0.21494992846924177, + "grad_norm": 0.12667940030051575, + "learning_rate": 0.0001825959007183558, + "loss": 0.7429, + "step": 2404 + }, + { + "epoch": 0.2150393419170243, + "grad_norm": 0.1419807222891753, + "learning_rate": 0.00018257957156729625, + "loss": 0.7076, + "step": 2405 + }, + { + "epoch": 0.21512875536480686, + "grad_norm": 0.15006741058079198, + "learning_rate": 0.00018256323549038778, + "loss": 0.3024, + "step": 2406 + }, + { + "epoch": 0.2152181688125894, + "grad_norm": 0.12786165969581956, + "learning_rate": 0.00018254689248900047, + "loss": 0.6737, + "step": 2407 + }, + { + "epoch": 0.21530758226037197, + "grad_norm": 0.11650523225942325, + "learning_rate": 0.00018253054256450494, + "loss": 0.665, + "step": 2408 + }, + { + "epoch": 0.21539699570815452, + "grad_norm": 0.1113874110227416, + "learning_rate": 0.0001825141857182725, + "loss": 0.68, + "step": 2409 + }, + { + "epoch": 0.21548640915593706, + "grad_norm": 0.12289724775655068, + "learning_rate": 0.00018249782195167496, + "loss": 0.6693, + "step": 2410 + }, + { + "epoch": 0.2155758226037196, + "grad_norm": 0.1190500777580604, + "learning_rate": 0.0001824814512660847, + "loss": 0.6944, + "step": 2411 + }, + { + "epoch": 0.21566523605150215, + "grad_norm": 0.1398862792861051, + "learning_rate": 0.00018246507366287475, + "loss": 0.7747, + "step": 2412 + }, + { + "epoch": 0.2157546494992847, + "grad_norm": 0.12235600826674738, + "learning_rate": 0.0001824486891434187, + "loss": 0.6528, + "step": 2413 + }, + { + "epoch": 0.21584406294706723, + "grad_norm": 0.1259116063306729, + "learning_rate": 0.0001824322977090906, + "loss": 0.6948, + "step": 2414 + }, + { + "epoch": 0.21593347639484978, + "grad_norm": 0.12214749318888249, + "learning_rate": 0.0001824158993612653, + "loss": 0.6479, + "step": 2415 + }, + { + "epoch": 0.21602288984263232, + "grad_norm": 0.11542989082084314, + "learning_rate": 0.00018239949410131802, + "loss": 0.6639, + "step": 2416 + }, + { + "epoch": 0.2161123032904149, + "grad_norm": 0.13193363534780908, + "learning_rate": 0.0001823830819306247, + "loss": 0.676, + "step": 2417 + }, + { + "epoch": 0.21620171673819744, + "grad_norm": 0.145229059797414, + "learning_rate": 0.0001823666628505618, + "loss": 0.7097, + "step": 2418 + }, + { + "epoch": 0.21629113018597998, + "grad_norm": 0.13092407364116335, + "learning_rate": 0.00018235023686250635, + "loss": 0.6645, + "step": 2419 + }, + { + "epoch": 0.21638054363376252, + "grad_norm": 0.13444348862506897, + "learning_rate": 0.00018233380396783595, + "loss": 0.6598, + "step": 2420 + }, + { + "epoch": 0.21646995708154507, + "grad_norm": 0.15680389781686735, + "learning_rate": 0.0001823173641679289, + "loss": 0.2636, + "step": 2421 + }, + { + "epoch": 0.2165593705293276, + "grad_norm": 0.11815893884278171, + "learning_rate": 0.0001823009174641639, + "loss": 0.7157, + "step": 2422 + }, + { + "epoch": 0.21664878397711015, + "grad_norm": 0.11085175555848042, + "learning_rate": 0.00018228446385792037, + "loss": 0.6883, + "step": 2423 + }, + { + "epoch": 0.2167381974248927, + "grad_norm": 0.135738741061496, + "learning_rate": 0.00018226800335057822, + "loss": 0.7148, + "step": 2424 + }, + { + "epoch": 0.21682761087267524, + "grad_norm": 0.12353149173682226, + "learning_rate": 0.00018225153594351795, + "loss": 0.6756, + "step": 2425 + }, + { + "epoch": 0.21691702432045779, + "grad_norm": 0.12782294682483805, + "learning_rate": 0.00018223506163812076, + "loss": 0.6619, + "step": 2426 + }, + { + "epoch": 0.21700643776824036, + "grad_norm": 0.10979981971819529, + "learning_rate": 0.0001822185804357682, + "loss": 0.6651, + "step": 2427 + }, + { + "epoch": 0.2170958512160229, + "grad_norm": 0.14451145435311527, + "learning_rate": 0.00018220209233784266, + "loss": 0.7072, + "step": 2428 + }, + { + "epoch": 0.21718526466380544, + "grad_norm": 0.1293597641590559, + "learning_rate": 0.00018218559734572686, + "loss": 0.7016, + "step": 2429 + }, + { + "epoch": 0.217274678111588, + "grad_norm": 0.1059056989341081, + "learning_rate": 0.00018216909546080428, + "loss": 0.6878, + "step": 2430 + }, + { + "epoch": 0.21736409155937053, + "grad_norm": 0.12168894686415802, + "learning_rate": 0.00018215258668445892, + "loss": 0.6833, + "step": 2431 + }, + { + "epoch": 0.21745350500715308, + "grad_norm": 0.11564419736018566, + "learning_rate": 0.00018213607101807527, + "loss": 0.6671, + "step": 2432 + }, + { + "epoch": 0.21754291845493562, + "grad_norm": 0.14022212744699833, + "learning_rate": 0.0001821195484630386, + "loss": 0.7326, + "step": 2433 + }, + { + "epoch": 0.21763233190271816, + "grad_norm": 0.12405594713830158, + "learning_rate": 0.00018210301902073456, + "loss": 0.6959, + "step": 2434 + }, + { + "epoch": 0.2177217453505007, + "grad_norm": 0.1208220044068157, + "learning_rate": 0.00018208648269254946, + "loss": 0.67, + "step": 2435 + }, + { + "epoch": 0.21781115879828325, + "grad_norm": 0.12613433316289757, + "learning_rate": 0.0001820699394798702, + "loss": 0.6538, + "step": 2436 + }, + { + "epoch": 0.21790057224606582, + "grad_norm": 0.1205088537789745, + "learning_rate": 0.00018205338938408425, + "loss": 0.6709, + "step": 2437 + }, + { + "epoch": 0.21798998569384836, + "grad_norm": 0.13514546055922594, + "learning_rate": 0.0001820368324065796, + "loss": 0.7075, + "step": 2438 + }, + { + "epoch": 0.2180793991416309, + "grad_norm": 0.11810848373455145, + "learning_rate": 0.00018202026854874487, + "loss": 0.6778, + "step": 2439 + }, + { + "epoch": 0.21816881258941345, + "grad_norm": 0.12434791094839948, + "learning_rate": 0.00018200369781196934, + "loss": 0.6367, + "step": 2440 + }, + { + "epoch": 0.218258226037196, + "grad_norm": 0.13383956241088837, + "learning_rate": 0.00018198712019764266, + "loss": 0.6954, + "step": 2441 + }, + { + "epoch": 0.21834763948497854, + "grad_norm": 0.11493975731521122, + "learning_rate": 0.00018197053570715523, + "loss": 0.6732, + "step": 2442 + }, + { + "epoch": 0.21843705293276108, + "grad_norm": 0.12381456449560635, + "learning_rate": 0.00018195394434189797, + "loss": 0.6901, + "step": 2443 + }, + { + "epoch": 0.21852646638054363, + "grad_norm": 0.13140959906718874, + "learning_rate": 0.00018193734610326239, + "loss": 0.7135, + "step": 2444 + }, + { + "epoch": 0.21861587982832617, + "grad_norm": 0.11662911046562399, + "learning_rate": 0.0001819207409926405, + "loss": 0.6711, + "step": 2445 + }, + { + "epoch": 0.21870529327610871, + "grad_norm": 0.11673584042723037, + "learning_rate": 0.00018190412901142504, + "loss": 0.7091, + "step": 2446 + }, + { + "epoch": 0.21879470672389129, + "grad_norm": 0.126029898862576, + "learning_rate": 0.00018188751016100918, + "loss": 0.6809, + "step": 2447 + }, + { + "epoch": 0.21888412017167383, + "grad_norm": 0.11694560987447161, + "learning_rate": 0.00018187088444278674, + "loss": 0.7062, + "step": 2448 + }, + { + "epoch": 0.21897353361945637, + "grad_norm": 0.13016028393514054, + "learning_rate": 0.0001818542518581521, + "loss": 0.7108, + "step": 2449 + }, + { + "epoch": 0.21906294706723892, + "grad_norm": 0.12634224798057136, + "learning_rate": 0.0001818376124085002, + "loss": 0.6967, + "step": 2450 + }, + { + "epoch": 0.21915236051502146, + "grad_norm": 0.11398914227304038, + "learning_rate": 0.0001818209660952266, + "loss": 0.658, + "step": 2451 + }, + { + "epoch": 0.219241773962804, + "grad_norm": 0.11045888984213831, + "learning_rate": 0.00018180431291972738, + "loss": 0.675, + "step": 2452 + }, + { + "epoch": 0.21933118741058655, + "grad_norm": 0.09774506572762051, + "learning_rate": 0.00018178765288339924, + "loss": 0.6284, + "step": 2453 + }, + { + "epoch": 0.2194206008583691, + "grad_norm": 0.11657628725952347, + "learning_rate": 0.00018177098598763942, + "loss": 0.6795, + "step": 2454 + }, + { + "epoch": 0.21951001430615164, + "grad_norm": 0.1284224769542488, + "learning_rate": 0.00018175431223384575, + "loss": 0.6762, + "step": 2455 + }, + { + "epoch": 0.21959942775393418, + "grad_norm": 0.13339749449149702, + "learning_rate": 0.00018173763162341667, + "loss": 0.2803, + "step": 2456 + }, + { + "epoch": 0.21968884120171675, + "grad_norm": 0.13301202384339958, + "learning_rate": 0.00018172094415775113, + "loss": 0.7009, + "step": 2457 + }, + { + "epoch": 0.2197782546494993, + "grad_norm": 0.12880772281405295, + "learning_rate": 0.00018170424983824868, + "loss": 0.6975, + "step": 2458 + }, + { + "epoch": 0.21986766809728184, + "grad_norm": 0.13108487772624464, + "learning_rate": 0.00018168754866630947, + "loss": 0.6944, + "step": 2459 + }, + { + "epoch": 0.21995708154506438, + "grad_norm": 0.13106465579776339, + "learning_rate": 0.00018167084064333423, + "loss": 0.2813, + "step": 2460 + }, + { + "epoch": 0.22004649499284692, + "grad_norm": 0.13703895077201303, + "learning_rate": 0.0001816541257707242, + "loss": 0.685, + "step": 2461 + }, + { + "epoch": 0.22013590844062947, + "grad_norm": 0.12600834131212957, + "learning_rate": 0.00018163740404988126, + "loss": 0.7007, + "step": 2462 + }, + { + "epoch": 0.220225321888412, + "grad_norm": 0.1252367544609942, + "learning_rate": 0.00018162067548220786, + "loss": 0.6707, + "step": 2463 + }, + { + "epoch": 0.22031473533619456, + "grad_norm": 0.11567409565471043, + "learning_rate": 0.00018160394006910694, + "loss": 0.6597, + "step": 2464 + }, + { + "epoch": 0.2204041487839771, + "grad_norm": 0.139456865294758, + "learning_rate": 0.00018158719781198213, + "loss": 0.6574, + "step": 2465 + }, + { + "epoch": 0.22049356223175964, + "grad_norm": 0.13678566091251323, + "learning_rate": 0.00018157044871223757, + "loss": 0.7078, + "step": 2466 + }, + { + "epoch": 0.22058297567954221, + "grad_norm": 0.12590207781244786, + "learning_rate": 0.00018155369277127802, + "loss": 0.7033, + "step": 2467 + }, + { + "epoch": 0.22067238912732476, + "grad_norm": 0.1405460310617664, + "learning_rate": 0.00018153692999050872, + "loss": 0.6896, + "step": 2468 + }, + { + "epoch": 0.2207618025751073, + "grad_norm": 0.12007432669195475, + "learning_rate": 0.00018152016037133558, + "loss": 0.6933, + "step": 2469 + }, + { + "epoch": 0.22085121602288985, + "grad_norm": 0.11832205687771735, + "learning_rate": 0.00018150338391516505, + "loss": 0.6744, + "step": 2470 + }, + { + "epoch": 0.2209406294706724, + "grad_norm": 0.12256392058351946, + "learning_rate": 0.0001814866006234041, + "loss": 0.6701, + "step": 2471 + }, + { + "epoch": 0.22103004291845493, + "grad_norm": 0.11102332688468147, + "learning_rate": 0.00018146981049746043, + "loss": 0.6789, + "step": 2472 + }, + { + "epoch": 0.22111945636623748, + "grad_norm": 0.12950024619723208, + "learning_rate": 0.0001814530135387421, + "loss": 0.6794, + "step": 2473 + }, + { + "epoch": 0.22120886981402002, + "grad_norm": 0.19046199952479265, + "learning_rate": 0.0001814362097486579, + "loss": 0.3005, + "step": 2474 + }, + { + "epoch": 0.22129828326180256, + "grad_norm": 0.12302835552397919, + "learning_rate": 0.00018141939912861717, + "loss": 0.6653, + "step": 2475 + }, + { + "epoch": 0.2213876967095851, + "grad_norm": 0.1248181358046688, + "learning_rate": 0.00018140258168002971, + "loss": 0.6464, + "step": 2476 + }, + { + "epoch": 0.22147711015736768, + "grad_norm": 0.14346467039741326, + "learning_rate": 0.0001813857574043061, + "loss": 0.7068, + "step": 2477 + }, + { + "epoch": 0.22156652360515022, + "grad_norm": 0.12635058137305805, + "learning_rate": 0.00018136892630285726, + "loss": 0.6717, + "step": 2478 + }, + { + "epoch": 0.22165593705293277, + "grad_norm": 0.12629441172147537, + "learning_rate": 0.00018135208837709486, + "loss": 0.6791, + "step": 2479 + }, + { + "epoch": 0.2217453505007153, + "grad_norm": 0.11478838237745084, + "learning_rate": 0.00018133524362843104, + "loss": 0.6909, + "step": 2480 + }, + { + "epoch": 0.22183476394849785, + "grad_norm": 0.14781263134017952, + "learning_rate": 0.00018131839205827856, + "loss": 0.2766, + "step": 2481 + }, + { + "epoch": 0.2219241773962804, + "grad_norm": 0.12814950451716192, + "learning_rate": 0.00018130153366805075, + "loss": 0.6703, + "step": 2482 + }, + { + "epoch": 0.22201359084406294, + "grad_norm": 0.13579619039903948, + "learning_rate": 0.00018128466845916154, + "loss": 0.6777, + "step": 2483 + }, + { + "epoch": 0.22210300429184548, + "grad_norm": 0.1124197584774415, + "learning_rate": 0.00018126779643302528, + "loss": 0.6369, + "step": 2484 + }, + { + "epoch": 0.22219241773962803, + "grad_norm": 0.11098217393297823, + "learning_rate": 0.00018125091759105713, + "loss": 0.6872, + "step": 2485 + }, + { + "epoch": 0.2222818311874106, + "grad_norm": 0.1143559153717401, + "learning_rate": 0.00018123403193467266, + "loss": 0.6517, + "step": 2486 + }, + { + "epoch": 0.22237124463519314, + "grad_norm": 0.13651502461431145, + "learning_rate": 0.000181217139465288, + "loss": 0.7112, + "step": 2487 + }, + { + "epoch": 0.2224606580829757, + "grad_norm": 0.13081639883334029, + "learning_rate": 0.00018120024018431998, + "loss": 0.6797, + "step": 2488 + }, + { + "epoch": 0.22255007153075823, + "grad_norm": 0.10336958764692307, + "learning_rate": 0.00018118333409318583, + "loss": 0.628, + "step": 2489 + }, + { + "epoch": 0.22263948497854077, + "grad_norm": 0.13648288687739915, + "learning_rate": 0.00018116642119330354, + "loss": 0.6944, + "step": 2490 + }, + { + "epoch": 0.22272889842632332, + "grad_norm": 0.12529661079240942, + "learning_rate": 0.0001811495014860915, + "loss": 0.7037, + "step": 2491 + }, + { + "epoch": 0.22281831187410586, + "grad_norm": 0.12105661328011817, + "learning_rate": 0.00018113257497296879, + "loss": 0.7045, + "step": 2492 + }, + { + "epoch": 0.2229077253218884, + "grad_norm": 0.1215871511450143, + "learning_rate": 0.000181115641655355, + "loss": 0.6385, + "step": 2493 + }, + { + "epoch": 0.22299713876967095, + "grad_norm": 0.12128796829601947, + "learning_rate": 0.00018109870153467031, + "loss": 0.7129, + "step": 2494 + }, + { + "epoch": 0.2230865522174535, + "grad_norm": 0.13066685790246835, + "learning_rate": 0.00018108175461233544, + "loss": 0.7251, + "step": 2495 + }, + { + "epoch": 0.22317596566523606, + "grad_norm": 0.1277832426916807, + "learning_rate": 0.00018106480088977172, + "loss": 0.679, + "step": 2496 + }, + { + "epoch": 0.2232653791130186, + "grad_norm": 0.1315224549297112, + "learning_rate": 0.0001810478403684011, + "loss": 0.68, + "step": 2497 + }, + { + "epoch": 0.22335479256080115, + "grad_norm": 0.13912466385468164, + "learning_rate": 0.00018103087304964597, + "loss": 0.6929, + "step": 2498 + }, + { + "epoch": 0.2234442060085837, + "grad_norm": 0.13536972785876603, + "learning_rate": 0.00018101389893492937, + "loss": 0.711, + "step": 2499 + }, + { + "epoch": 0.22353361945636624, + "grad_norm": 0.1687167732642572, + "learning_rate": 0.0001809969180256749, + "loss": 0.2841, + "step": 2500 + }, + { + "epoch": 0.22362303290414878, + "grad_norm": 0.12044284599534377, + "learning_rate": 0.00018097993032330676, + "loss": 0.7089, + "step": 2501 + }, + { + "epoch": 0.22371244635193133, + "grad_norm": 0.1325641135899709, + "learning_rate": 0.00018096293582924963, + "loss": 0.6744, + "step": 2502 + }, + { + "epoch": 0.22380185979971387, + "grad_norm": 0.14534315315496682, + "learning_rate": 0.00018094593454492887, + "loss": 0.7141, + "step": 2503 + }, + { + "epoch": 0.22389127324749641, + "grad_norm": 0.1502929735536168, + "learning_rate": 0.00018092892647177035, + "loss": 0.7217, + "step": 2504 + }, + { + "epoch": 0.22398068669527896, + "grad_norm": 0.1326386057721552, + "learning_rate": 0.0001809119116112005, + "loss": 0.7033, + "step": 2505 + }, + { + "epoch": 0.22407010014306153, + "grad_norm": 0.1375901468800106, + "learning_rate": 0.00018089488996464632, + "loss": 0.6953, + "step": 2506 + }, + { + "epoch": 0.22415951359084407, + "grad_norm": 0.11075535818539727, + "learning_rate": 0.00018087786153353543, + "loss": 0.6562, + "step": 2507 + }, + { + "epoch": 0.22424892703862662, + "grad_norm": 0.13612696227719284, + "learning_rate": 0.00018086082631929595, + "loss": 0.6997, + "step": 2508 + }, + { + "epoch": 0.22433834048640916, + "grad_norm": 0.12113504382177367, + "learning_rate": 0.00018084378432335667, + "loss": 0.6618, + "step": 2509 + }, + { + "epoch": 0.2244277539341917, + "grad_norm": 0.13394341313568292, + "learning_rate": 0.00018082673554714677, + "loss": 0.6565, + "step": 2510 + }, + { + "epoch": 0.22451716738197425, + "grad_norm": 0.13440633270127153, + "learning_rate": 0.00018080967999209622, + "loss": 0.6768, + "step": 2511 + }, + { + "epoch": 0.2246065808297568, + "grad_norm": 0.12658733809063458, + "learning_rate": 0.00018079261765963537, + "loss": 0.7283, + "step": 2512 + }, + { + "epoch": 0.22469599427753933, + "grad_norm": 0.1254197101388256, + "learning_rate": 0.00018077554855119526, + "loss": 0.6903, + "step": 2513 + }, + { + "epoch": 0.22478540772532188, + "grad_norm": 0.1334516048859269, + "learning_rate": 0.00018075847266820746, + "loss": 0.6693, + "step": 2514 + }, + { + "epoch": 0.22487482117310442, + "grad_norm": 0.11644162206236518, + "learning_rate": 0.0001807413900121041, + "loss": 0.648, + "step": 2515 + }, + { + "epoch": 0.224964234620887, + "grad_norm": 0.12545760567634634, + "learning_rate": 0.00018072430058431783, + "loss": 0.6936, + "step": 2516 + }, + { + "epoch": 0.22505364806866954, + "grad_norm": 0.11099497631342035, + "learning_rate": 0.000180707204386282, + "loss": 0.6349, + "step": 2517 + }, + { + "epoch": 0.22514306151645208, + "grad_norm": 0.1307750632861993, + "learning_rate": 0.00018069010141943037, + "loss": 0.6939, + "step": 2518 + }, + { + "epoch": 0.22523247496423462, + "grad_norm": 0.11554943557094176, + "learning_rate": 0.00018067299168519741, + "loss": 0.669, + "step": 2519 + }, + { + "epoch": 0.22532188841201717, + "grad_norm": 0.13240720088549582, + "learning_rate": 0.00018065587518501804, + "loss": 0.7084, + "step": 2520 + }, + { + "epoch": 0.2254113018597997, + "grad_norm": 0.12409322230529503, + "learning_rate": 0.00018063875192032787, + "loss": 0.6579, + "step": 2521 + }, + { + "epoch": 0.22550071530758226, + "grad_norm": 0.11764055675417724, + "learning_rate": 0.00018062162189256292, + "loss": 0.6329, + "step": 2522 + }, + { + "epoch": 0.2255901287553648, + "grad_norm": 0.11691196898778333, + "learning_rate": 0.00018060448510315993, + "loss": 0.6412, + "step": 2523 + }, + { + "epoch": 0.22567954220314734, + "grad_norm": 0.12692481938784553, + "learning_rate": 0.00018058734155355612, + "loss": 0.6913, + "step": 2524 + }, + { + "epoch": 0.2257689556509299, + "grad_norm": 0.13018637736891994, + "learning_rate": 0.00018057019124518927, + "loss": 0.7023, + "step": 2525 + }, + { + "epoch": 0.22585836909871246, + "grad_norm": 0.11635724100896674, + "learning_rate": 0.00018055303417949782, + "loss": 0.6778, + "step": 2526 + }, + { + "epoch": 0.225947782546495, + "grad_norm": 0.12742877829960117, + "learning_rate": 0.00018053587035792067, + "loss": 0.661, + "step": 2527 + }, + { + "epoch": 0.22603719599427755, + "grad_norm": 0.12206043555770878, + "learning_rate": 0.00018051869978189731, + "loss": 0.6867, + "step": 2528 + }, + { + "epoch": 0.2261266094420601, + "grad_norm": 0.1288583099185624, + "learning_rate": 0.0001805015224528679, + "loss": 0.7325, + "step": 2529 + }, + { + "epoch": 0.22621602288984263, + "grad_norm": 0.11317749756139422, + "learning_rate": 0.00018048433837227295, + "loss": 0.619, + "step": 2530 + }, + { + "epoch": 0.22630543633762518, + "grad_norm": 0.12642524102931568, + "learning_rate": 0.0001804671475415538, + "loss": 0.6305, + "step": 2531 + }, + { + "epoch": 0.22639484978540772, + "grad_norm": 0.12337085311594521, + "learning_rate": 0.00018044994996215213, + "loss": 0.6879, + "step": 2532 + }, + { + "epoch": 0.22648426323319026, + "grad_norm": 0.121026978022724, + "learning_rate": 0.00018043274563551035, + "loss": 0.6778, + "step": 2533 + }, + { + "epoch": 0.2265736766809728, + "grad_norm": 0.15492944478042994, + "learning_rate": 0.00018041553456307128, + "loss": 0.3017, + "step": 2534 + }, + { + "epoch": 0.22666309012875535, + "grad_norm": 0.13566571043959139, + "learning_rate": 0.00018039831674627847, + "loss": 0.6832, + "step": 2535 + }, + { + "epoch": 0.22675250357653792, + "grad_norm": 0.13352235898237952, + "learning_rate": 0.00018038109218657594, + "loss": 0.6772, + "step": 2536 + }, + { + "epoch": 0.22684191702432047, + "grad_norm": 0.12994971432623303, + "learning_rate": 0.00018036386088540827, + "loss": 0.7174, + "step": 2537 + }, + { + "epoch": 0.226931330472103, + "grad_norm": 0.12874120590811713, + "learning_rate": 0.00018034662284422065, + "loss": 0.6589, + "step": 2538 + }, + { + "epoch": 0.22702074391988555, + "grad_norm": 0.12484692051048367, + "learning_rate": 0.00018032937806445882, + "loss": 0.6912, + "step": 2539 + }, + { + "epoch": 0.2271101573676681, + "grad_norm": 0.12849352076176435, + "learning_rate": 0.00018031212654756905, + "loss": 0.6886, + "step": 2540 + }, + { + "epoch": 0.22719957081545064, + "grad_norm": 0.13081382513864773, + "learning_rate": 0.00018029486829499822, + "loss": 0.7146, + "step": 2541 + }, + { + "epoch": 0.22728898426323318, + "grad_norm": 0.12888508582004918, + "learning_rate": 0.00018027760330819375, + "loss": 0.6954, + "step": 2542 + }, + { + "epoch": 0.22737839771101573, + "grad_norm": 0.12152698722223992, + "learning_rate": 0.00018026033158860365, + "loss": 0.7081, + "step": 2543 + }, + { + "epoch": 0.22746781115879827, + "grad_norm": 0.12584991015046865, + "learning_rate": 0.00018024305313767646, + "loss": 0.6995, + "step": 2544 + }, + { + "epoch": 0.22755722460658084, + "grad_norm": 0.12007106959076955, + "learning_rate": 0.00018022576795686133, + "loss": 0.6497, + "step": 2545 + }, + { + "epoch": 0.2276466380543634, + "grad_norm": 0.11859930361397909, + "learning_rate": 0.00018020847604760794, + "loss": 0.6962, + "step": 2546 + }, + { + "epoch": 0.22773605150214593, + "grad_norm": 0.1662946347917575, + "learning_rate": 0.00018019117741136648, + "loss": 0.2654, + "step": 2547 + }, + { + "epoch": 0.22782546494992847, + "grad_norm": 0.12608530369637178, + "learning_rate": 0.00018017387204958784, + "loss": 0.6707, + "step": 2548 + }, + { + "epoch": 0.22791487839771102, + "grad_norm": 0.1258359405438104, + "learning_rate": 0.0001801565599637234, + "loss": 0.6932, + "step": 2549 + }, + { + "epoch": 0.22800429184549356, + "grad_norm": 0.11771940493786653, + "learning_rate": 0.00018013924115522508, + "loss": 0.6944, + "step": 2550 + }, + { + "epoch": 0.2280937052932761, + "grad_norm": 0.11876005393847915, + "learning_rate": 0.00018012191562554537, + "loss": 0.6486, + "step": 2551 + }, + { + "epoch": 0.22818311874105865, + "grad_norm": 0.11972463799373789, + "learning_rate": 0.00018010458337613735, + "loss": 0.6845, + "step": 2552 + }, + { + "epoch": 0.2282725321888412, + "grad_norm": 0.14395991877316794, + "learning_rate": 0.00018008724440845468, + "loss": 0.7364, + "step": 2553 + }, + { + "epoch": 0.22836194563662374, + "grad_norm": 0.13331356290013538, + "learning_rate": 0.00018006989872395156, + "loss": 0.6957, + "step": 2554 + }, + { + "epoch": 0.2284513590844063, + "grad_norm": 0.11165241065929347, + "learning_rate": 0.0001800525463240827, + "loss": 0.6565, + "step": 2555 + }, + { + "epoch": 0.22854077253218885, + "grad_norm": 0.13286601681198507, + "learning_rate": 0.00018003518721030349, + "loss": 0.6295, + "step": 2556 + }, + { + "epoch": 0.2286301859799714, + "grad_norm": 0.1259205789492328, + "learning_rate": 0.00018001782138406976, + "loss": 0.6819, + "step": 2557 + }, + { + "epoch": 0.22871959942775394, + "grad_norm": 0.1164113005933035, + "learning_rate": 0.000180000448846838, + "loss": 0.6628, + "step": 2558 + }, + { + "epoch": 0.22880901287553648, + "grad_norm": 0.11317505571969094, + "learning_rate": 0.0001799830696000652, + "loss": 0.6819, + "step": 2559 + }, + { + "epoch": 0.22889842632331903, + "grad_norm": 0.12769578584139965, + "learning_rate": 0.00017996568364520897, + "loss": 0.6952, + "step": 2560 + }, + { + "epoch": 0.22898783977110157, + "grad_norm": 0.12588501634569285, + "learning_rate": 0.00017994829098372738, + "loss": 0.6848, + "step": 2561 + }, + { + "epoch": 0.2290772532188841, + "grad_norm": 0.12116428970227161, + "learning_rate": 0.0001799308916170792, + "loss": 0.6991, + "step": 2562 + }, + { + "epoch": 0.22916666666666666, + "grad_norm": 0.1101203984658687, + "learning_rate": 0.00017991348554672373, + "loss": 0.6797, + "step": 2563 + }, + { + "epoch": 0.2292560801144492, + "grad_norm": 0.13506911993149143, + "learning_rate": 0.00017989607277412066, + "loss": 0.7108, + "step": 2564 + }, + { + "epoch": 0.22934549356223177, + "grad_norm": 0.12972667506514068, + "learning_rate": 0.00017987865330073048, + "loss": 0.6801, + "step": 2565 + }, + { + "epoch": 0.22943490701001432, + "grad_norm": 0.13942375538156473, + "learning_rate": 0.00017986122712801414, + "loss": 0.2734, + "step": 2566 + }, + { + "epoch": 0.22952432045779686, + "grad_norm": 0.12563913599531454, + "learning_rate": 0.0001798437942574331, + "loss": 0.6892, + "step": 2567 + }, + { + "epoch": 0.2296137339055794, + "grad_norm": 0.1050676191025854, + "learning_rate": 0.0001798263546904495, + "loss": 0.6144, + "step": 2568 + }, + { + "epoch": 0.22970314735336195, + "grad_norm": 0.1269686776143616, + "learning_rate": 0.0001798089084285259, + "loss": 0.6589, + "step": 2569 + }, + { + "epoch": 0.2297925608011445, + "grad_norm": 0.13029913916428384, + "learning_rate": 0.00017979145547312555, + "loss": 0.704, + "step": 2570 + }, + { + "epoch": 0.22988197424892703, + "grad_norm": 0.13468180748804, + "learning_rate": 0.0001797739958257122, + "loss": 0.715, + "step": 2571 + }, + { + "epoch": 0.22997138769670958, + "grad_norm": 0.1487556970165325, + "learning_rate": 0.00017975652948775013, + "loss": 0.7254, + "step": 2572 + }, + { + "epoch": 0.23006080114449212, + "grad_norm": 0.13246101533456234, + "learning_rate": 0.0001797390564607043, + "loss": 0.7242, + "step": 2573 + }, + { + "epoch": 0.23015021459227467, + "grad_norm": 0.13432589604982542, + "learning_rate": 0.00017972157674604007, + "loss": 0.7024, + "step": 2574 + }, + { + "epoch": 0.23023962804005724, + "grad_norm": 0.11133704301475997, + "learning_rate": 0.00017970409034522348, + "loss": 0.6925, + "step": 2575 + }, + { + "epoch": 0.23032904148783978, + "grad_norm": 0.11941283122558924, + "learning_rate": 0.00017968659725972112, + "loss": 0.6798, + "step": 2576 + }, + { + "epoch": 0.23041845493562232, + "grad_norm": 0.11293494976876077, + "learning_rate": 0.00017966909749100006, + "loss": 0.6739, + "step": 2577 + }, + { + "epoch": 0.23050786838340487, + "grad_norm": 0.12329236760075847, + "learning_rate": 0.00017965159104052803, + "loss": 0.6598, + "step": 2578 + }, + { + "epoch": 0.2305972818311874, + "grad_norm": 0.13576512351137807, + "learning_rate": 0.00017963407790977322, + "loss": 0.7307, + "step": 2579 + }, + { + "epoch": 0.23068669527896996, + "grad_norm": 0.14652969418994874, + "learning_rate": 0.00017961655810020452, + "loss": 0.27, + "step": 2580 + }, + { + "epoch": 0.2307761087267525, + "grad_norm": 0.132927487935088, + "learning_rate": 0.00017959903161329118, + "loss": 0.7095, + "step": 2581 + }, + { + "epoch": 0.23086552217453504, + "grad_norm": 0.12180558816201914, + "learning_rate": 0.00017958149845050323, + "loss": 0.6895, + "step": 2582 + }, + { + "epoch": 0.2309549356223176, + "grad_norm": 0.11645737853118598, + "learning_rate": 0.0001795639586133111, + "loss": 0.6724, + "step": 2583 + }, + { + "epoch": 0.23104434907010013, + "grad_norm": 0.13315414123491112, + "learning_rate": 0.00017954641210318588, + "loss": 0.6922, + "step": 2584 + }, + { + "epoch": 0.2311337625178827, + "grad_norm": 0.11926226698729554, + "learning_rate": 0.0001795288589215991, + "loss": 0.6827, + "step": 2585 + }, + { + "epoch": 0.23122317596566525, + "grad_norm": 0.12766649967115104, + "learning_rate": 0.000179511299070023, + "loss": 0.68, + "step": 2586 + }, + { + "epoch": 0.2313125894134478, + "grad_norm": 0.11257136683516569, + "learning_rate": 0.00017949373254993027, + "loss": 0.6722, + "step": 2587 + }, + { + "epoch": 0.23140200286123033, + "grad_norm": 0.12131064653635677, + "learning_rate": 0.00017947615936279417, + "loss": 0.6849, + "step": 2588 + }, + { + "epoch": 0.23149141630901288, + "grad_norm": 0.11874897123045973, + "learning_rate": 0.00017945857951008859, + "loss": 0.6672, + "step": 2589 + }, + { + "epoch": 0.23158082975679542, + "grad_norm": 0.14126113732337803, + "learning_rate": 0.00017944099299328791, + "loss": 0.6852, + "step": 2590 + }, + { + "epoch": 0.23167024320457796, + "grad_norm": 0.1308507067917132, + "learning_rate": 0.00017942339981386708, + "loss": 0.6711, + "step": 2591 + }, + { + "epoch": 0.2317596566523605, + "grad_norm": 0.18491996597730598, + "learning_rate": 0.00017940579997330165, + "loss": 0.6913, + "step": 2592 + }, + { + "epoch": 0.23184907010014305, + "grad_norm": 0.1135098826429387, + "learning_rate": 0.00017938819347306764, + "loss": 0.6716, + "step": 2593 + }, + { + "epoch": 0.2319384835479256, + "grad_norm": 0.12430774743734001, + "learning_rate": 0.00017937058031464173, + "loss": 0.7137, + "step": 2594 + }, + { + "epoch": 0.23202789699570817, + "grad_norm": 0.12012359500596621, + "learning_rate": 0.0001793529604995011, + "loss": 0.6773, + "step": 2595 + }, + { + "epoch": 0.2321173104434907, + "grad_norm": 0.13446791227727353, + "learning_rate": 0.00017933533402912354, + "loss": 0.721, + "step": 2596 + }, + { + "epoch": 0.23220672389127325, + "grad_norm": 0.1233708902024786, + "learning_rate": 0.0001793177009049873, + "loss": 0.7138, + "step": 2597 + }, + { + "epoch": 0.2322961373390558, + "grad_norm": 0.12345438118852942, + "learning_rate": 0.00017930006112857127, + "loss": 0.6866, + "step": 2598 + }, + { + "epoch": 0.23238555078683834, + "grad_norm": 0.1333343386384798, + "learning_rate": 0.0001792824147013549, + "loss": 0.68, + "step": 2599 + }, + { + "epoch": 0.23247496423462088, + "grad_norm": 0.12099796742517148, + "learning_rate": 0.00017926476162481817, + "loss": 0.6606, + "step": 2600 + }, + { + "epoch": 0.23256437768240343, + "grad_norm": 0.1122402337502642, + "learning_rate": 0.00017924710190044156, + "loss": 0.6887, + "step": 2601 + }, + { + "epoch": 0.23265379113018597, + "grad_norm": 0.13877228040130016, + "learning_rate": 0.00017922943552970625, + "loss": 0.6869, + "step": 2602 + }, + { + "epoch": 0.23274320457796852, + "grad_norm": 0.15197300050299142, + "learning_rate": 0.0001792117625140939, + "loss": 0.6672, + "step": 2603 + }, + { + "epoch": 0.23283261802575106, + "grad_norm": 0.14028558681035233, + "learning_rate": 0.00017919408285508662, + "loss": 0.695, + "step": 2604 + }, + { + "epoch": 0.23292203147353363, + "grad_norm": 0.11735270530530814, + "learning_rate": 0.0001791763965541673, + "loss": 0.6383, + "step": 2605 + }, + { + "epoch": 0.23301144492131617, + "grad_norm": 0.11912524162635074, + "learning_rate": 0.00017915870361281922, + "loss": 0.6877, + "step": 2606 + }, + { + "epoch": 0.23310085836909872, + "grad_norm": 0.12971011139992483, + "learning_rate": 0.00017914100403252628, + "loss": 0.7228, + "step": 2607 + }, + { + "epoch": 0.23319027181688126, + "grad_norm": 0.11008661043559703, + "learning_rate": 0.00017912329781477287, + "loss": 0.6848, + "step": 2608 + }, + { + "epoch": 0.2332796852646638, + "grad_norm": 0.1097918963492545, + "learning_rate": 0.00017910558496104403, + "loss": 0.6871, + "step": 2609 + }, + { + "epoch": 0.23336909871244635, + "grad_norm": 0.1205565702662758, + "learning_rate": 0.00017908786547282538, + "loss": 0.6977, + "step": 2610 + }, + { + "epoch": 0.2334585121602289, + "grad_norm": 0.11477945559172562, + "learning_rate": 0.0001790701393516029, + "loss": 0.6776, + "step": 2611 + }, + { + "epoch": 0.23354792560801144, + "grad_norm": 0.1243152196256237, + "learning_rate": 0.00017905240659886335, + "loss": 0.6677, + "step": 2612 + }, + { + "epoch": 0.23363733905579398, + "grad_norm": 0.11976438871205462, + "learning_rate": 0.00017903466721609393, + "loss": 0.7157, + "step": 2613 + }, + { + "epoch": 0.23372675250357655, + "grad_norm": 0.13284435709329484, + "learning_rate": 0.0001790169212047824, + "loss": 0.6512, + "step": 2614 + }, + { + "epoch": 0.2338161659513591, + "grad_norm": 0.13495828791878153, + "learning_rate": 0.00017899916856641714, + "loss": 0.6964, + "step": 2615 + }, + { + "epoch": 0.23390557939914164, + "grad_norm": 0.14455098684518186, + "learning_rate": 0.00017898140930248704, + "loss": 0.7331, + "step": 2616 + }, + { + "epoch": 0.23399499284692418, + "grad_norm": 0.11808453779605804, + "learning_rate": 0.0001789636434144815, + "loss": 0.6398, + "step": 2617 + }, + { + "epoch": 0.23408440629470673, + "grad_norm": 0.1217192420990994, + "learning_rate": 0.00017894587090389052, + "loss": 0.6708, + "step": 2618 + }, + { + "epoch": 0.23417381974248927, + "grad_norm": 0.12520772345704298, + "learning_rate": 0.00017892809177220474, + "loss": 0.6969, + "step": 2619 + }, + { + "epoch": 0.2342632331902718, + "grad_norm": 0.14280193960859336, + "learning_rate": 0.00017891030602091519, + "loss": 0.7082, + "step": 2620 + }, + { + "epoch": 0.23435264663805436, + "grad_norm": 0.1375036096916246, + "learning_rate": 0.0001788925136515136, + "loss": 0.7256, + "step": 2621 + }, + { + "epoch": 0.2344420600858369, + "grad_norm": 0.14140926315883093, + "learning_rate": 0.00017887471466549216, + "loss": 0.7118, + "step": 2622 + }, + { + "epoch": 0.23453147353361944, + "grad_norm": 0.11273529347633497, + "learning_rate": 0.00017885690906434365, + "loss": 0.6685, + "step": 2623 + }, + { + "epoch": 0.23462088698140202, + "grad_norm": 0.12475851610360059, + "learning_rate": 0.0001788390968495614, + "loss": 0.6869, + "step": 2624 + }, + { + "epoch": 0.23471030042918456, + "grad_norm": 0.13014828880165782, + "learning_rate": 0.00017882127802263935, + "loss": 0.7083, + "step": 2625 + }, + { + "epoch": 0.2347997138769671, + "grad_norm": 0.1104295622463144, + "learning_rate": 0.00017880345258507188, + "loss": 0.6443, + "step": 2626 + }, + { + "epoch": 0.23488912732474965, + "grad_norm": 0.11748323853900279, + "learning_rate": 0.000178785620538354, + "loss": 0.6854, + "step": 2627 + }, + { + "epoch": 0.2349785407725322, + "grad_norm": 0.11165759437319153, + "learning_rate": 0.00017876778188398128, + "loss": 0.6661, + "step": 2628 + }, + { + "epoch": 0.23506795422031473, + "grad_norm": 0.1036313053292268, + "learning_rate": 0.00017874993662344983, + "loss": 0.6488, + "step": 2629 + }, + { + "epoch": 0.23515736766809728, + "grad_norm": 0.10745946340242589, + "learning_rate": 0.00017873208475825632, + "loss": 0.6588, + "step": 2630 + }, + { + "epoch": 0.23524678111587982, + "grad_norm": 0.11882115632695792, + "learning_rate": 0.0001787142262898979, + "loss": 0.6731, + "step": 2631 + }, + { + "epoch": 0.23533619456366237, + "grad_norm": 0.12574549561236606, + "learning_rate": 0.00017869636121987243, + "loss": 0.6747, + "step": 2632 + }, + { + "epoch": 0.2354256080114449, + "grad_norm": 0.1168694863890653, + "learning_rate": 0.00017867848954967815, + "loss": 0.6897, + "step": 2633 + }, + { + "epoch": 0.23551502145922748, + "grad_norm": 0.11567598295283615, + "learning_rate": 0.000178660611280814, + "loss": 0.6286, + "step": 2634 + }, + { + "epoch": 0.23560443490701002, + "grad_norm": 0.12665500109106767, + "learning_rate": 0.00017864272641477936, + "loss": 0.6669, + "step": 2635 + }, + { + "epoch": 0.23569384835479257, + "grad_norm": 0.1050295782884434, + "learning_rate": 0.00017862483495307424, + "loss": 0.6594, + "step": 2636 + }, + { + "epoch": 0.2357832618025751, + "grad_norm": 0.10483820610128552, + "learning_rate": 0.00017860693689719916, + "loss": 0.664, + "step": 2637 + }, + { + "epoch": 0.23587267525035766, + "grad_norm": 0.12994691161761787, + "learning_rate": 0.0001785890322486552, + "loss": 0.716, + "step": 2638 + }, + { + "epoch": 0.2359620886981402, + "grad_norm": 0.13121333151787018, + "learning_rate": 0.00017857112100894406, + "loss": 0.7161, + "step": 2639 + }, + { + "epoch": 0.23605150214592274, + "grad_norm": 0.1160792794807502, + "learning_rate": 0.00017855320317956784, + "loss": 0.6789, + "step": 2640 + }, + { + "epoch": 0.2361409155937053, + "grad_norm": 0.1284530473975064, + "learning_rate": 0.0001785352787620294, + "loss": 0.6728, + "step": 2641 + }, + { + "epoch": 0.23623032904148783, + "grad_norm": 0.10370444845398774, + "learning_rate": 0.00017851734775783194, + "loss": 0.6679, + "step": 2642 + }, + { + "epoch": 0.23631974248927037, + "grad_norm": 0.11456294749257585, + "learning_rate": 0.00017849941016847933, + "loss": 0.6637, + "step": 2643 + }, + { + "epoch": 0.23640915593705294, + "grad_norm": 0.12143004598960883, + "learning_rate": 0.000178481465995476, + "loss": 0.6672, + "step": 2644 + }, + { + "epoch": 0.2364985693848355, + "grad_norm": 0.12323334466160589, + "learning_rate": 0.00017846351524032693, + "loss": 0.6619, + "step": 2645 + }, + { + "epoch": 0.23658798283261803, + "grad_norm": 0.14338092773377617, + "learning_rate": 0.0001784455579045376, + "loss": 0.7027, + "step": 2646 + }, + { + "epoch": 0.23667739628040058, + "grad_norm": 0.13598651249249086, + "learning_rate": 0.00017842759398961405, + "loss": 0.7097, + "step": 2647 + }, + { + "epoch": 0.23676680972818312, + "grad_norm": 0.10817092903851565, + "learning_rate": 0.00017840962349706288, + "loss": 0.6641, + "step": 2648 + }, + { + "epoch": 0.23685622317596566, + "grad_norm": 0.1286756977781123, + "learning_rate": 0.00017839164642839133, + "loss": 0.6731, + "step": 2649 + }, + { + "epoch": 0.2369456366237482, + "grad_norm": 0.115374548459784, + "learning_rate": 0.000178373662785107, + "loss": 0.6828, + "step": 2650 + }, + { + "epoch": 0.23703505007153075, + "grad_norm": 0.11831205175119501, + "learning_rate": 0.00017835567256871827, + "loss": 0.6647, + "step": 2651 + }, + { + "epoch": 0.2371244635193133, + "grad_norm": 0.1066596052421221, + "learning_rate": 0.00017833767578073393, + "loss": 0.6556, + "step": 2652 + }, + { + "epoch": 0.23721387696709584, + "grad_norm": 0.12295950228405442, + "learning_rate": 0.0001783196724226633, + "loss": 0.6831, + "step": 2653 + }, + { + "epoch": 0.2373032904148784, + "grad_norm": 0.12700747297207962, + "learning_rate": 0.00017830166249601637, + "loss": 0.6972, + "step": 2654 + }, + { + "epoch": 0.23739270386266095, + "grad_norm": 0.13804142331181185, + "learning_rate": 0.00017828364600230352, + "loss": 0.6744, + "step": 2655 + }, + { + "epoch": 0.2374821173104435, + "grad_norm": 0.13288205637400882, + "learning_rate": 0.00017826562294303585, + "loss": 0.6944, + "step": 2656 + }, + { + "epoch": 0.23757153075822604, + "grad_norm": 0.1196689229240669, + "learning_rate": 0.0001782475933197249, + "loss": 0.7063, + "step": 2657 + }, + { + "epoch": 0.23766094420600858, + "grad_norm": 0.11790050645883018, + "learning_rate": 0.00017822955713388277, + "loss": 0.6561, + "step": 2658 + }, + { + "epoch": 0.23775035765379113, + "grad_norm": 0.1319133540415549, + "learning_rate": 0.0001782115143870222, + "loss": 0.6695, + "step": 2659 + }, + { + "epoch": 0.23783977110157367, + "grad_norm": 0.11752098209213437, + "learning_rate": 0.00017819346508065635, + "loss": 0.6872, + "step": 2660 + }, + { + "epoch": 0.23792918454935622, + "grad_norm": 0.13034633693398612, + "learning_rate": 0.00017817540921629904, + "loss": 0.7037, + "step": 2661 + }, + { + "epoch": 0.23801859799713876, + "grad_norm": 0.1121959436659483, + "learning_rate": 0.00017815734679546457, + "loss": 0.665, + "step": 2662 + }, + { + "epoch": 0.2381080114449213, + "grad_norm": 0.11339806338254078, + "learning_rate": 0.00017813927781966778, + "loss": 0.6333, + "step": 2663 + }, + { + "epoch": 0.23819742489270387, + "grad_norm": 0.12298746612010521, + "learning_rate": 0.00017812120229042416, + "loss": 0.6627, + "step": 2664 + }, + { + "epoch": 0.23828683834048642, + "grad_norm": 0.1262014728295087, + "learning_rate": 0.00017810312020924963, + "loss": 0.6939, + "step": 2665 + }, + { + "epoch": 0.23837625178826896, + "grad_norm": 0.12872214760518055, + "learning_rate": 0.00017808503157766073, + "loss": 0.6892, + "step": 2666 + }, + { + "epoch": 0.2384656652360515, + "grad_norm": 0.12968563056313095, + "learning_rate": 0.00017806693639717456, + "loss": 0.7088, + "step": 2667 + }, + { + "epoch": 0.23855507868383405, + "grad_norm": 0.11819199576864205, + "learning_rate": 0.0001780488346693087, + "loss": 0.68, + "step": 2668 + }, + { + "epoch": 0.2386444921316166, + "grad_norm": 0.12371369978539266, + "learning_rate": 0.00017803072639558133, + "loss": 0.6703, + "step": 2669 + }, + { + "epoch": 0.23873390557939914, + "grad_norm": 0.11581826073926511, + "learning_rate": 0.0001780126115775112, + "loss": 0.6631, + "step": 2670 + }, + { + "epoch": 0.23882331902718168, + "grad_norm": 0.10533070773821014, + "learning_rate": 0.00017799449021661752, + "loss": 0.6593, + "step": 2671 + }, + { + "epoch": 0.23891273247496422, + "grad_norm": 0.10192444795249936, + "learning_rate": 0.00017797636231442016, + "loss": 0.627, + "step": 2672 + }, + { + "epoch": 0.2390021459227468, + "grad_norm": 0.11976550987181209, + "learning_rate": 0.00017795822787243946, + "loss": 0.6808, + "step": 2673 + }, + { + "epoch": 0.23909155937052934, + "grad_norm": 0.12340712973998809, + "learning_rate": 0.0001779400868921963, + "loss": 0.7106, + "step": 2674 + }, + { + "epoch": 0.23918097281831188, + "grad_norm": 0.13673433086823408, + "learning_rate": 0.00017792193937521224, + "loss": 0.7095, + "step": 2675 + }, + { + "epoch": 0.23927038626609443, + "grad_norm": 0.13839356197640798, + "learning_rate": 0.0001779037853230092, + "loss": 0.7171, + "step": 2676 + }, + { + "epoch": 0.23935979971387697, + "grad_norm": 0.12199935847645997, + "learning_rate": 0.00017788562473710978, + "loss": 0.6897, + "step": 2677 + }, + { + "epoch": 0.2394492131616595, + "grad_norm": 0.11117737091726902, + "learning_rate": 0.00017786745761903708, + "loss": 0.6588, + "step": 2678 + }, + { + "epoch": 0.23953862660944206, + "grad_norm": 0.11638524269619345, + "learning_rate": 0.00017784928397031476, + "loss": 0.6746, + "step": 2679 + }, + { + "epoch": 0.2396280400572246, + "grad_norm": 0.11774815845820488, + "learning_rate": 0.00017783110379246696, + "loss": 0.6273, + "step": 2680 + }, + { + "epoch": 0.23971745350500714, + "grad_norm": 0.14886112878952676, + "learning_rate": 0.00017781291708701853, + "loss": 0.7021, + "step": 2681 + }, + { + "epoch": 0.2398068669527897, + "grad_norm": 0.14018043546449827, + "learning_rate": 0.0001777947238554947, + "loss": 0.7115, + "step": 2682 + }, + { + "epoch": 0.23989628040057226, + "grad_norm": 0.12678373852182215, + "learning_rate": 0.00017777652409942132, + "loss": 0.6709, + "step": 2683 + }, + { + "epoch": 0.2399856938483548, + "grad_norm": 0.12097441047890935, + "learning_rate": 0.00017775831782032483, + "loss": 0.7118, + "step": 2684 + }, + { + "epoch": 0.24007510729613735, + "grad_norm": 0.12078736035555868, + "learning_rate": 0.00017774010501973208, + "loss": 0.6833, + "step": 2685 + }, + { + "epoch": 0.2401645207439199, + "grad_norm": 0.1339853672451839, + "learning_rate": 0.00017772188569917065, + "loss": 0.2816, + "step": 2686 + }, + { + "epoch": 0.24025393419170243, + "grad_norm": 0.12441446456893156, + "learning_rate": 0.00017770365986016852, + "loss": 0.6651, + "step": 2687 + }, + { + "epoch": 0.24034334763948498, + "grad_norm": 0.12899512542308753, + "learning_rate": 0.00017768542750425426, + "loss": 0.6784, + "step": 2688 + }, + { + "epoch": 0.24043276108726752, + "grad_norm": 0.12805478972121828, + "learning_rate": 0.00017766718863295705, + "loss": 0.6789, + "step": 2689 + }, + { + "epoch": 0.24052217453505007, + "grad_norm": 0.12912073942539457, + "learning_rate": 0.00017764894324780653, + "loss": 0.7296, + "step": 2690 + }, + { + "epoch": 0.2406115879828326, + "grad_norm": 0.12015265505996814, + "learning_rate": 0.0001776306913503329, + "loss": 0.6729, + "step": 2691 + }, + { + "epoch": 0.24070100143061515, + "grad_norm": 0.12390733189060399, + "learning_rate": 0.00017761243294206694, + "loss": 0.6648, + "step": 2692 + }, + { + "epoch": 0.24079041487839772, + "grad_norm": 0.1213637898290781, + "learning_rate": 0.00017759416802453997, + "loss": 0.6678, + "step": 2693 + }, + { + "epoch": 0.24087982832618027, + "grad_norm": 0.13674554394151234, + "learning_rate": 0.0001775758965992838, + "loss": 0.6998, + "step": 2694 + }, + { + "epoch": 0.2409692417739628, + "grad_norm": 0.13559526357424304, + "learning_rate": 0.0001775576186678309, + "loss": 0.6889, + "step": 2695 + }, + { + "epoch": 0.24105865522174535, + "grad_norm": 0.13655042996004954, + "learning_rate": 0.00017753933423171421, + "loss": 0.7225, + "step": 2696 + }, + { + "epoch": 0.2411480686695279, + "grad_norm": 0.13746084929238422, + "learning_rate": 0.00017752104329246717, + "loss": 0.7006, + "step": 2697 + }, + { + "epoch": 0.24123748211731044, + "grad_norm": 0.10874861283490977, + "learning_rate": 0.00017750274585162385, + "loss": 0.6614, + "step": 2698 + }, + { + "epoch": 0.24132689556509299, + "grad_norm": 0.1273362992080128, + "learning_rate": 0.00017748444191071884, + "loss": 0.6825, + "step": 2699 + }, + { + "epoch": 0.24141630901287553, + "grad_norm": 0.11567562593067225, + "learning_rate": 0.00017746613147128726, + "loss": 0.6751, + "step": 2700 + }, + { + "epoch": 0.24150572246065807, + "grad_norm": 0.1552000212233341, + "learning_rate": 0.0001774478145348648, + "loss": 0.2969, + "step": 2701 + }, + { + "epoch": 0.24159513590844062, + "grad_norm": 0.12223502480998027, + "learning_rate": 0.00017742949110298767, + "loss": 0.6406, + "step": 2702 + }, + { + "epoch": 0.2416845493562232, + "grad_norm": 0.1220111771087652, + "learning_rate": 0.00017741116117719262, + "loss": 0.6605, + "step": 2703 + }, + { + "epoch": 0.24177396280400573, + "grad_norm": 0.13800203198301159, + "learning_rate": 0.000177392824759017, + "loss": 0.7348, + "step": 2704 + }, + { + "epoch": 0.24186337625178828, + "grad_norm": 0.1102879914890447, + "learning_rate": 0.0001773744818499986, + "loss": 0.6357, + "step": 2705 + }, + { + "epoch": 0.24195278969957082, + "grad_norm": 0.13230890321433045, + "learning_rate": 0.00017735613245167586, + "loss": 0.6857, + "step": 2706 + }, + { + "epoch": 0.24204220314735336, + "grad_norm": 0.13430193677013075, + "learning_rate": 0.00017733777656558773, + "loss": 0.6941, + "step": 2707 + }, + { + "epoch": 0.2421316165951359, + "grad_norm": 0.12117491971587853, + "learning_rate": 0.00017731941419327365, + "loss": 0.6722, + "step": 2708 + }, + { + "epoch": 0.24222103004291845, + "grad_norm": 0.1222730744007976, + "learning_rate": 0.0001773010453362737, + "loss": 0.6725, + "step": 2709 + }, + { + "epoch": 0.242310443490701, + "grad_norm": 0.11934927150341666, + "learning_rate": 0.00017728266999612844, + "loss": 0.6656, + "step": 2710 + }, + { + "epoch": 0.24239985693848354, + "grad_norm": 0.11446088182074318, + "learning_rate": 0.000177264288174379, + "loss": 0.6748, + "step": 2711 + }, + { + "epoch": 0.24248927038626608, + "grad_norm": 0.11935118667865277, + "learning_rate": 0.00017724589987256698, + "loss": 0.6726, + "step": 2712 + }, + { + "epoch": 0.24257868383404865, + "grad_norm": 0.1165511107904844, + "learning_rate": 0.00017722750509223465, + "loss": 0.6976, + "step": 2713 + }, + { + "epoch": 0.2426680972818312, + "grad_norm": 0.13919310252428013, + "learning_rate": 0.0001772091038349247, + "loss": 0.7043, + "step": 2714 + }, + { + "epoch": 0.24275751072961374, + "grad_norm": 0.12404582341948645, + "learning_rate": 0.00017719069610218048, + "loss": 0.6825, + "step": 2715 + }, + { + "epoch": 0.24284692417739628, + "grad_norm": 0.13064485681173552, + "learning_rate": 0.00017717228189554582, + "loss": 0.7061, + "step": 2716 + }, + { + "epoch": 0.24293633762517883, + "grad_norm": 0.10985284969940236, + "learning_rate": 0.00017715386121656507, + "loss": 0.6693, + "step": 2717 + }, + { + "epoch": 0.24302575107296137, + "grad_norm": 0.11908788881776304, + "learning_rate": 0.00017713543406678315, + "loss": 0.6602, + "step": 2718 + }, + { + "epoch": 0.24311516452074391, + "grad_norm": 0.135279735267145, + "learning_rate": 0.0001771170004477455, + "loss": 0.6726, + "step": 2719 + }, + { + "epoch": 0.24320457796852646, + "grad_norm": 0.11652672169828236, + "learning_rate": 0.0001770985603609982, + "loss": 0.6698, + "step": 2720 + }, + { + "epoch": 0.243293991416309, + "grad_norm": 0.11714426303518398, + "learning_rate": 0.00017708011380808774, + "loss": 0.67, + "step": 2721 + }, + { + "epoch": 0.24338340486409155, + "grad_norm": 0.13484778775885953, + "learning_rate": 0.00017706166079056124, + "loss": 0.6655, + "step": 2722 + }, + { + "epoch": 0.24347281831187412, + "grad_norm": 0.13775255362062197, + "learning_rate": 0.0001770432013099663, + "loss": 0.7462, + "step": 2723 + }, + { + "epoch": 0.24356223175965666, + "grad_norm": 0.12036489368436817, + "learning_rate": 0.0001770247353678511, + "loss": 0.6882, + "step": 2724 + }, + { + "epoch": 0.2436516452074392, + "grad_norm": 0.13223453389992273, + "learning_rate": 0.0001770062629657644, + "loss": 0.7027, + "step": 2725 + }, + { + "epoch": 0.24374105865522175, + "grad_norm": 0.12564105398605888, + "learning_rate": 0.0001769877841052554, + "loss": 0.6869, + "step": 2726 + }, + { + "epoch": 0.2438304721030043, + "grad_norm": 0.1361411291489478, + "learning_rate": 0.00017696929878787394, + "loss": 0.682, + "step": 2727 + }, + { + "epoch": 0.24391988555078684, + "grad_norm": 0.11954304837663669, + "learning_rate": 0.00017695080701517034, + "loss": 0.6759, + "step": 2728 + }, + { + "epoch": 0.24400929899856938, + "grad_norm": 0.21973185303858847, + "learning_rate": 0.00017693230878869547, + "loss": 0.3418, + "step": 2729 + }, + { + "epoch": 0.24409871244635192, + "grad_norm": 0.1282119043285775, + "learning_rate": 0.00017691380411000079, + "loss": 0.6798, + "step": 2730 + }, + { + "epoch": 0.24418812589413447, + "grad_norm": 0.13760746820437153, + "learning_rate": 0.00017689529298063822, + "loss": 0.6676, + "step": 2731 + }, + { + "epoch": 0.244277539341917, + "grad_norm": 0.11600086918686454, + "learning_rate": 0.00017687677540216033, + "loss": 0.6765, + "step": 2732 + }, + { + "epoch": 0.24436695278969958, + "grad_norm": 0.13036828979707282, + "learning_rate": 0.00017685825137612012, + "loss": 0.6816, + "step": 2733 + }, + { + "epoch": 0.24445636623748213, + "grad_norm": 0.1328888723937378, + "learning_rate": 0.00017683972090407123, + "loss": 0.6635, + "step": 2734 + }, + { + "epoch": 0.24454577968526467, + "grad_norm": 0.11891464105162525, + "learning_rate": 0.00017682118398756766, + "loss": 0.6365, + "step": 2735 + }, + { + "epoch": 0.2446351931330472, + "grad_norm": 0.13139530834490387, + "learning_rate": 0.0001768026406281642, + "loss": 0.7012, + "step": 2736 + }, + { + "epoch": 0.24472460658082976, + "grad_norm": 0.13708833058060174, + "learning_rate": 0.00017678409082741604, + "loss": 0.6951, + "step": 2737 + }, + { + "epoch": 0.2448140200286123, + "grad_norm": 0.1232800985948365, + "learning_rate": 0.00017676553458687892, + "loss": 0.6447, + "step": 2738 + }, + { + "epoch": 0.24490343347639484, + "grad_norm": 0.12335903695953301, + "learning_rate": 0.00017674697190810912, + "loss": 0.6901, + "step": 2739 + }, + { + "epoch": 0.2449928469241774, + "grad_norm": 0.17055903970453345, + "learning_rate": 0.00017672840279266345, + "loss": 0.3146, + "step": 2740 + }, + { + "epoch": 0.24508226037195993, + "grad_norm": 0.13366626752433594, + "learning_rate": 0.00017670982724209933, + "loss": 0.6949, + "step": 2741 + }, + { + "epoch": 0.2451716738197425, + "grad_norm": 0.153527765411321, + "learning_rate": 0.00017669124525797463, + "loss": 0.701, + "step": 2742 + }, + { + "epoch": 0.24526108726752505, + "grad_norm": 0.13990028624899503, + "learning_rate": 0.0001766726568418478, + "loss": 0.666, + "step": 2743 + }, + { + "epoch": 0.2453505007153076, + "grad_norm": 0.13449003708563184, + "learning_rate": 0.00017665406199527785, + "loss": 0.7022, + "step": 2744 + }, + { + "epoch": 0.24543991416309013, + "grad_norm": 0.13324379677984133, + "learning_rate": 0.00017663546071982432, + "loss": 0.666, + "step": 2745 + }, + { + "epoch": 0.24552932761087268, + "grad_norm": 0.10545548152404813, + "learning_rate": 0.0001766168530170472, + "loss": 0.642, + "step": 2746 + }, + { + "epoch": 0.24561874105865522, + "grad_norm": 0.11823742434865611, + "learning_rate": 0.00017659823888850715, + "loss": 0.6928, + "step": 2747 + }, + { + "epoch": 0.24570815450643776, + "grad_norm": 0.132790620650247, + "learning_rate": 0.00017657961833576535, + "loss": 0.6993, + "step": 2748 + }, + { + "epoch": 0.2457975679542203, + "grad_norm": 0.13761004369629604, + "learning_rate": 0.0001765609913603834, + "loss": 0.7199, + "step": 2749 + }, + { + "epoch": 0.24588698140200285, + "grad_norm": 0.12469087364218018, + "learning_rate": 0.00017654235796392363, + "loss": 0.6609, + "step": 2750 + }, + { + "epoch": 0.2459763948497854, + "grad_norm": 0.13698661517011101, + "learning_rate": 0.0001765237181479487, + "loss": 0.6649, + "step": 2751 + }, + { + "epoch": 0.24606580829756797, + "grad_norm": 0.13072394514756028, + "learning_rate": 0.00017650507191402194, + "loss": 0.684, + "step": 2752 + }, + { + "epoch": 0.2461552217453505, + "grad_norm": 0.12326498074372327, + "learning_rate": 0.0001764864192637072, + "loss": 0.6474, + "step": 2753 + }, + { + "epoch": 0.24624463519313305, + "grad_norm": 0.13496918787475093, + "learning_rate": 0.00017646776019856884, + "loss": 0.6751, + "step": 2754 + }, + { + "epoch": 0.2463340486409156, + "grad_norm": 0.14074096546050827, + "learning_rate": 0.0001764490947201718, + "loss": 0.7016, + "step": 2755 + }, + { + "epoch": 0.24642346208869814, + "grad_norm": 0.12919660106302142, + "learning_rate": 0.00017643042283008148, + "loss": 0.725, + "step": 2756 + }, + { + "epoch": 0.24651287553648069, + "grad_norm": 0.104230365414311, + "learning_rate": 0.00017641174452986396, + "loss": 0.6659, + "step": 2757 + }, + { + "epoch": 0.24660228898426323, + "grad_norm": 0.12013473013862959, + "learning_rate": 0.00017639305982108567, + "loss": 0.6728, + "step": 2758 + }, + { + "epoch": 0.24669170243204577, + "grad_norm": 0.14429921617891087, + "learning_rate": 0.0001763743687053137, + "loss": 0.7203, + "step": 2759 + }, + { + "epoch": 0.24678111587982832, + "grad_norm": 0.11474050065780422, + "learning_rate": 0.0001763556711841157, + "loss": 0.6816, + "step": 2760 + }, + { + "epoch": 0.24687052932761086, + "grad_norm": 0.11798113800940516, + "learning_rate": 0.00017633696725905974, + "loss": 0.6387, + "step": 2761 + }, + { + "epoch": 0.24695994277539343, + "grad_norm": 0.14154192846104668, + "learning_rate": 0.00017631825693171453, + "loss": 0.3172, + "step": 2762 + }, + { + "epoch": 0.24704935622317598, + "grad_norm": 0.13471257085161398, + "learning_rate": 0.0001762995402036493, + "loss": 0.6595, + "step": 2763 + }, + { + "epoch": 0.24713876967095852, + "grad_norm": 0.1272966751298662, + "learning_rate": 0.00017628081707643376, + "loss": 0.6633, + "step": 2764 + }, + { + "epoch": 0.24722818311874106, + "grad_norm": 0.1355399432772288, + "learning_rate": 0.00017626208755163822, + "loss": 0.692, + "step": 2765 + }, + { + "epoch": 0.2473175965665236, + "grad_norm": 0.12318375366231529, + "learning_rate": 0.0001762433516308335, + "loss": 0.6603, + "step": 2766 + }, + { + "epoch": 0.24740701001430615, + "grad_norm": 0.11066298937438657, + "learning_rate": 0.00017622460931559098, + "loss": 0.7001, + "step": 2767 + }, + { + "epoch": 0.2474964234620887, + "grad_norm": 0.1333339579414932, + "learning_rate": 0.00017620586060748252, + "loss": 0.7215, + "step": 2768 + }, + { + "epoch": 0.24758583690987124, + "grad_norm": 0.10700684022717591, + "learning_rate": 0.00017618710550808056, + "loss": 0.6479, + "step": 2769 + }, + { + "epoch": 0.24767525035765378, + "grad_norm": 0.12470249051023825, + "learning_rate": 0.00017616834401895805, + "loss": 0.6882, + "step": 2770 + }, + { + "epoch": 0.24776466380543632, + "grad_norm": 0.10906086267451783, + "learning_rate": 0.0001761495761416885, + "loss": 0.678, + "step": 2771 + }, + { + "epoch": 0.2478540772532189, + "grad_norm": 0.13320502670043763, + "learning_rate": 0.00017613080187784603, + "loss": 0.7136, + "step": 2772 + }, + { + "epoch": 0.24794349070100144, + "grad_norm": 0.11481579465977329, + "learning_rate": 0.00017611202122900512, + "loss": 0.6656, + "step": 2773 + }, + { + "epoch": 0.24803290414878398, + "grad_norm": 0.10765199887206192, + "learning_rate": 0.0001760932341967409, + "loss": 0.657, + "step": 2774 + }, + { + "epoch": 0.24812231759656653, + "grad_norm": 0.12940185786174943, + "learning_rate": 0.00017607444078262903, + "loss": 0.6584, + "step": 2775 + }, + { + "epoch": 0.24821173104434907, + "grad_norm": 0.1214821178600409, + "learning_rate": 0.00017605564098824568, + "loss": 0.7075, + "step": 2776 + }, + { + "epoch": 0.24830114449213161, + "grad_norm": 0.1271918407461164, + "learning_rate": 0.00017603683481516762, + "loss": 0.6615, + "step": 2777 + }, + { + "epoch": 0.24839055793991416, + "grad_norm": 0.22095038134957967, + "learning_rate": 0.000176018022264972, + "loss": 0.651, + "step": 2778 + }, + { + "epoch": 0.2484799713876967, + "grad_norm": 0.11966897820479791, + "learning_rate": 0.00017599920333923668, + "loss": 0.6398, + "step": 2779 + }, + { + "epoch": 0.24856938483547925, + "grad_norm": 0.1673916073678918, + "learning_rate": 0.00017598037803953994, + "loss": 0.6889, + "step": 2780 + }, + { + "epoch": 0.2486587982832618, + "grad_norm": 0.12870281650872267, + "learning_rate": 0.00017596154636746066, + "loss": 0.663, + "step": 2781 + }, + { + "epoch": 0.24874821173104436, + "grad_norm": 0.14645504020067907, + "learning_rate": 0.00017594270832457825, + "loss": 0.7073, + "step": 2782 + }, + { + "epoch": 0.2488376251788269, + "grad_norm": 0.1421598079977415, + "learning_rate": 0.0001759238639124726, + "loss": 0.7167, + "step": 2783 + }, + { + "epoch": 0.24892703862660945, + "grad_norm": 0.11539556366523919, + "learning_rate": 0.00017590501313272415, + "loss": 0.6914, + "step": 2784 + }, + { + "epoch": 0.249016452074392, + "grad_norm": 0.10795909290500529, + "learning_rate": 0.00017588615598691397, + "loss": 0.6752, + "step": 2785 + }, + { + "epoch": 0.24910586552217454, + "grad_norm": 0.11432943049883842, + "learning_rate": 0.00017586729247662345, + "loss": 0.6977, + "step": 2786 + }, + { + "epoch": 0.24919527896995708, + "grad_norm": 0.13495211580423297, + "learning_rate": 0.00017584842260343482, + "loss": 0.7183, + "step": 2787 + }, + { + "epoch": 0.24928469241773962, + "grad_norm": 0.11269680879058115, + "learning_rate": 0.00017582954636893055, + "loss": 0.6607, + "step": 2788 + }, + { + "epoch": 0.24937410586552217, + "grad_norm": 0.1158377861413361, + "learning_rate": 0.0001758106637746938, + "loss": 0.6332, + "step": 2789 + }, + { + "epoch": 0.2494635193133047, + "grad_norm": 0.1397596095499722, + "learning_rate": 0.00017579177482230824, + "loss": 0.7254, + "step": 2790 + }, + { + "epoch": 0.24955293276108725, + "grad_norm": 0.13252638369215755, + "learning_rate": 0.00017577287951335807, + "loss": 0.6605, + "step": 2791 + }, + { + "epoch": 0.24964234620886983, + "grad_norm": 0.1335935072546966, + "learning_rate": 0.00017575397784942799, + "loss": 0.7015, + "step": 2792 + }, + { + "epoch": 0.24973175965665237, + "grad_norm": 0.12914675393646782, + "learning_rate": 0.00017573506983210329, + "loss": 0.6428, + "step": 2793 + }, + { + "epoch": 0.2498211731044349, + "grad_norm": 0.1390122177713663, + "learning_rate": 0.00017571615546296972, + "loss": 0.6772, + "step": 2794 + }, + { + "epoch": 0.24991058655221746, + "grad_norm": 0.1387658654705644, + "learning_rate": 0.00017569723474361365, + "loss": 0.6973, + "step": 2795 + }, + { + "epoch": 0.25, + "grad_norm": 0.11034409389087017, + "learning_rate": 0.00017567830767562198, + "loss": 0.6823, + "step": 2796 + }, + { + "epoch": 0.25008941344778257, + "grad_norm": 0.1263471280218487, + "learning_rate": 0.00017565937426058196, + "loss": 0.7146, + "step": 2797 + }, + { + "epoch": 0.2501788268955651, + "grad_norm": 0.12399762107779158, + "learning_rate": 0.00017564043450008163, + "loss": 0.6801, + "step": 2798 + }, + { + "epoch": 0.25026824034334766, + "grad_norm": 0.12763142194788107, + "learning_rate": 0.0001756214883957094, + "loss": 0.692, + "step": 2799 + }, + { + "epoch": 0.2503576537911302, + "grad_norm": 0.18203467006594268, + "learning_rate": 0.00017560253594905425, + "loss": 0.3028, + "step": 2800 + }, + { + "epoch": 0.25044706723891275, + "grad_norm": 0.1282102396058687, + "learning_rate": 0.00017558357716170573, + "loss": 0.6393, + "step": 2801 + }, + { + "epoch": 0.25053648068669526, + "grad_norm": 0.13955769203043425, + "learning_rate": 0.00017556461203525387, + "loss": 0.6617, + "step": 2802 + }, + { + "epoch": 0.25062589413447783, + "grad_norm": 0.14201553707248216, + "learning_rate": 0.00017554564057128928, + "loss": 0.6549, + "step": 2803 + }, + { + "epoch": 0.25071530758226035, + "grad_norm": 0.12208662353524714, + "learning_rate": 0.00017552666277140304, + "loss": 0.6741, + "step": 2804 + }, + { + "epoch": 0.2508047210300429, + "grad_norm": 0.12457257218338204, + "learning_rate": 0.0001755076786371868, + "loss": 0.6551, + "step": 2805 + }, + { + "epoch": 0.2508941344778255, + "grad_norm": 0.12846283840765393, + "learning_rate": 0.00017548868817023275, + "loss": 0.6754, + "step": 2806 + }, + { + "epoch": 0.250983547925608, + "grad_norm": 0.1311752136825843, + "learning_rate": 0.00017546969137213357, + "loss": 0.6802, + "step": 2807 + }, + { + "epoch": 0.2510729613733906, + "grad_norm": 0.12068324226451042, + "learning_rate": 0.00017545068824448255, + "loss": 0.6758, + "step": 2808 + }, + { + "epoch": 0.2511623748211731, + "grad_norm": 0.12837276956689603, + "learning_rate": 0.0001754316787888734, + "loss": 0.648, + "step": 2809 + }, + { + "epoch": 0.25125178826895567, + "grad_norm": 0.11914241941743671, + "learning_rate": 0.00017541266300690047, + "loss": 0.6547, + "step": 2810 + }, + { + "epoch": 0.2513412017167382, + "grad_norm": 0.12053874664212741, + "learning_rate": 0.00017539364090015855, + "loss": 0.7048, + "step": 2811 + }, + { + "epoch": 0.25143061516452075, + "grad_norm": 0.11652349428520876, + "learning_rate": 0.00017537461247024304, + "loss": 0.6321, + "step": 2812 + }, + { + "epoch": 0.25152002861230327, + "grad_norm": 0.129247809986282, + "learning_rate": 0.0001753555777187498, + "loss": 0.7093, + "step": 2813 + }, + { + "epoch": 0.25160944206008584, + "grad_norm": 0.13890681497526786, + "learning_rate": 0.00017533653664727529, + "loss": 0.6728, + "step": 2814 + }, + { + "epoch": 0.25169885550786836, + "grad_norm": 0.12528666676626826, + "learning_rate": 0.0001753174892574164, + "loss": 0.7156, + "step": 2815 + }, + { + "epoch": 0.25178826895565093, + "grad_norm": 0.1442759705343363, + "learning_rate": 0.00017529843555077066, + "loss": 0.7206, + "step": 2816 + }, + { + "epoch": 0.2518776824034335, + "grad_norm": 0.1088708180727441, + "learning_rate": 0.00017527937552893605, + "loss": 0.6332, + "step": 2817 + }, + { + "epoch": 0.251967095851216, + "grad_norm": 0.12379641979762586, + "learning_rate": 0.00017526030919351113, + "loss": 0.6969, + "step": 2818 + }, + { + "epoch": 0.2520565092989986, + "grad_norm": 0.14490699319070482, + "learning_rate": 0.000175241236546095, + "loss": 0.7031, + "step": 2819 + }, + { + "epoch": 0.2521459227467811, + "grad_norm": 0.2781124047820011, + "learning_rate": 0.00017522215758828722, + "loss": 0.3236, + "step": 2820 + }, + { + "epoch": 0.2522353361945637, + "grad_norm": 0.13292196292130007, + "learning_rate": 0.0001752030723216879, + "loss": 0.7163, + "step": 2821 + }, + { + "epoch": 0.2523247496423462, + "grad_norm": 0.12396818804723529, + "learning_rate": 0.00017518398074789775, + "loss": 0.6754, + "step": 2822 + }, + { + "epoch": 0.25241416309012876, + "grad_norm": 0.12477270507473048, + "learning_rate": 0.00017516488286851794, + "loss": 0.6744, + "step": 2823 + }, + { + "epoch": 0.2525035765379113, + "grad_norm": 0.10916164513541529, + "learning_rate": 0.00017514577868515016, + "loss": 0.6798, + "step": 2824 + }, + { + "epoch": 0.25259298998569385, + "grad_norm": 0.1235698872065035, + "learning_rate": 0.0001751266681993967, + "loss": 0.6866, + "step": 2825 + }, + { + "epoch": 0.2526824034334764, + "grad_norm": 0.12425219304597582, + "learning_rate": 0.00017510755141286028, + "loss": 0.6563, + "step": 2826 + }, + { + "epoch": 0.25277181688125894, + "grad_norm": 0.11154661527338272, + "learning_rate": 0.00017508842832714426, + "loss": 0.6593, + "step": 2827 + }, + { + "epoch": 0.2528612303290415, + "grad_norm": 0.12953568967403573, + "learning_rate": 0.0001750692989438524, + "loss": 0.7006, + "step": 2828 + }, + { + "epoch": 0.252950643776824, + "grad_norm": 0.12400876290030917, + "learning_rate": 0.00017505016326458913, + "loss": 0.7036, + "step": 2829 + }, + { + "epoch": 0.2530400572246066, + "grad_norm": 0.1225217146084359, + "learning_rate": 0.00017503102129095928, + "loss": 0.7121, + "step": 2830 + }, + { + "epoch": 0.2531294706723891, + "grad_norm": 0.11336709506430946, + "learning_rate": 0.0001750118730245683, + "loss": 0.6724, + "step": 2831 + }, + { + "epoch": 0.2532188841201717, + "grad_norm": 0.1289305182981423, + "learning_rate": 0.00017499271846702213, + "loss": 0.6643, + "step": 2832 + }, + { + "epoch": 0.2533082975679542, + "grad_norm": 0.14142713051676442, + "learning_rate": 0.00017497355761992724, + "loss": 0.6653, + "step": 2833 + }, + { + "epoch": 0.25339771101573677, + "grad_norm": 0.13091519117999026, + "learning_rate": 0.00017495439048489063, + "loss": 0.7057, + "step": 2834 + }, + { + "epoch": 0.2534871244635193, + "grad_norm": 0.1555815775784396, + "learning_rate": 0.00017493521706351975, + "loss": 0.7134, + "step": 2835 + }, + { + "epoch": 0.25357653791130186, + "grad_norm": 0.1128471718871932, + "learning_rate": 0.00017491603735742277, + "loss": 0.6499, + "step": 2836 + }, + { + "epoch": 0.25366595135908443, + "grad_norm": 0.13170898420042923, + "learning_rate": 0.0001748968513682082, + "loss": 0.6705, + "step": 2837 + }, + { + "epoch": 0.25375536480686695, + "grad_norm": 0.11692978699890312, + "learning_rate": 0.00017487765909748513, + "loss": 0.6507, + "step": 2838 + }, + { + "epoch": 0.2538447782546495, + "grad_norm": 0.13295335446957668, + "learning_rate": 0.00017485846054686324, + "loss": 0.7065, + "step": 2839 + }, + { + "epoch": 0.25393419170243203, + "grad_norm": 0.12280098470151764, + "learning_rate": 0.00017483925571795268, + "loss": 0.6465, + "step": 2840 + }, + { + "epoch": 0.2540236051502146, + "grad_norm": 0.12356788619037636, + "learning_rate": 0.00017482004461236413, + "loss": 0.703, + "step": 2841 + }, + { + "epoch": 0.2541130185979971, + "grad_norm": 0.1184076892438362, + "learning_rate": 0.00017480082723170877, + "loss": 0.6716, + "step": 2842 + }, + { + "epoch": 0.2542024320457797, + "grad_norm": 0.11250241133646839, + "learning_rate": 0.00017478160357759838, + "loss": 0.6618, + "step": 2843 + }, + { + "epoch": 0.2542918454935622, + "grad_norm": 0.1430551226265974, + "learning_rate": 0.00017476237365164523, + "loss": 0.6861, + "step": 2844 + }, + { + "epoch": 0.2543812589413448, + "grad_norm": 0.13597901523829517, + "learning_rate": 0.00017474313745546204, + "loss": 0.6763, + "step": 2845 + }, + { + "epoch": 0.25447067238912735, + "grad_norm": 0.11206532639004985, + "learning_rate": 0.00017472389499066223, + "loss": 0.6767, + "step": 2846 + }, + { + "epoch": 0.25456008583690987, + "grad_norm": 0.14129068829100047, + "learning_rate": 0.00017470464625885958, + "loss": 0.6782, + "step": 2847 + }, + { + "epoch": 0.25464949928469244, + "grad_norm": 0.13548069125066714, + "learning_rate": 0.00017468539126166846, + "loss": 0.7028, + "step": 2848 + }, + { + "epoch": 0.25473891273247495, + "grad_norm": 0.14461307937542994, + "learning_rate": 0.0001746661300007038, + "loss": 0.6851, + "step": 2849 + }, + { + "epoch": 0.2548283261802575, + "grad_norm": 0.13316777161517848, + "learning_rate": 0.00017464686247758095, + "loss": 0.7005, + "step": 2850 + }, + { + "epoch": 0.25491773962804004, + "grad_norm": 0.1448690943697754, + "learning_rate": 0.00017462758869391591, + "loss": 0.6941, + "step": 2851 + }, + { + "epoch": 0.2550071530758226, + "grad_norm": 0.125290363871442, + "learning_rate": 0.00017460830865132513, + "loss": 0.706, + "step": 2852 + }, + { + "epoch": 0.25509656652360513, + "grad_norm": 0.13550904379354833, + "learning_rate": 0.00017458902235142562, + "loss": 0.6986, + "step": 2853 + }, + { + "epoch": 0.2551859799713877, + "grad_norm": 0.1446105668309678, + "learning_rate": 0.00017456972979583486, + "loss": 0.685, + "step": 2854 + }, + { + "epoch": 0.2552753934191702, + "grad_norm": 0.12943854663178767, + "learning_rate": 0.00017455043098617097, + "loss": 0.7117, + "step": 2855 + }, + { + "epoch": 0.2553648068669528, + "grad_norm": 0.11080096287318897, + "learning_rate": 0.00017453112592405242, + "loss": 0.6763, + "step": 2856 + }, + { + "epoch": 0.25545422031473536, + "grad_norm": 0.12823851281333026, + "learning_rate": 0.00017451181461109835, + "loss": 0.6933, + "step": 2857 + }, + { + "epoch": 0.2555436337625179, + "grad_norm": 0.10710029379486075, + "learning_rate": 0.0001744924970489284, + "loss": 0.6725, + "step": 2858 + }, + { + "epoch": 0.25563304721030045, + "grad_norm": 0.13449141254621538, + "learning_rate": 0.00017447317323916267, + "loss": 0.6865, + "step": 2859 + }, + { + "epoch": 0.25572246065808296, + "grad_norm": 0.13057280730913723, + "learning_rate": 0.00017445384318342185, + "loss": 0.6874, + "step": 2860 + }, + { + "epoch": 0.25581187410586553, + "grad_norm": 0.12860189126712376, + "learning_rate": 0.00017443450688332712, + "loss": 0.6495, + "step": 2861 + }, + { + "epoch": 0.25590128755364805, + "grad_norm": 0.12264056587445467, + "learning_rate": 0.00017441516434050017, + "loss": 0.6754, + "step": 2862 + }, + { + "epoch": 0.2559907010014306, + "grad_norm": 0.11094562239786357, + "learning_rate": 0.0001743958155565633, + "loss": 0.6631, + "step": 2863 + }, + { + "epoch": 0.25608011444921314, + "grad_norm": 0.12026189527698772, + "learning_rate": 0.0001743764605331392, + "loss": 0.6627, + "step": 2864 + }, + { + "epoch": 0.2561695278969957, + "grad_norm": 0.13668156581321408, + "learning_rate": 0.0001743570992718512, + "loss": 0.7487, + "step": 2865 + }, + { + "epoch": 0.2562589413447783, + "grad_norm": 0.13093960816098885, + "learning_rate": 0.00017433773177432307, + "loss": 0.682, + "step": 2866 + }, + { + "epoch": 0.2563483547925608, + "grad_norm": 0.11100956955999372, + "learning_rate": 0.00017431835804217912, + "loss": 0.6544, + "step": 2867 + }, + { + "epoch": 0.25643776824034337, + "grad_norm": 0.11007308563301843, + "learning_rate": 0.00017429897807704427, + "loss": 0.6249, + "step": 2868 + }, + { + "epoch": 0.2565271816881259, + "grad_norm": 0.12487874098541277, + "learning_rate": 0.00017427959188054385, + "loss": 0.7017, + "step": 2869 + }, + { + "epoch": 0.25661659513590845, + "grad_norm": 0.11719769042782661, + "learning_rate": 0.0001742601994543038, + "loss": 0.6714, + "step": 2870 + }, + { + "epoch": 0.25670600858369097, + "grad_norm": 0.11856021346643221, + "learning_rate": 0.00017424080079995045, + "loss": 0.6676, + "step": 2871 + }, + { + "epoch": 0.25679542203147354, + "grad_norm": 0.14115627260976718, + "learning_rate": 0.00017422139591911085, + "loss": 0.7191, + "step": 2872 + }, + { + "epoch": 0.25688483547925606, + "grad_norm": 0.12552724576219793, + "learning_rate": 0.00017420198481341237, + "loss": 0.6916, + "step": 2873 + }, + { + "epoch": 0.25697424892703863, + "grad_norm": 0.11176755614259479, + "learning_rate": 0.00017418256748448304, + "loss": 0.6409, + "step": 2874 + }, + { + "epoch": 0.2570636623748212, + "grad_norm": 0.13998073383173537, + "learning_rate": 0.0001741631439339514, + "loss": 0.6651, + "step": 2875 + }, + { + "epoch": 0.2571530758226037, + "grad_norm": 0.13575688147836484, + "learning_rate": 0.0001741437141634464, + "loss": 0.7054, + "step": 2876 + }, + { + "epoch": 0.2572424892703863, + "grad_norm": 0.1199712671041745, + "learning_rate": 0.00017412427817459767, + "loss": 0.6617, + "step": 2877 + }, + { + "epoch": 0.2573319027181688, + "grad_norm": 0.8724067120035623, + "learning_rate": 0.00017410483596903525, + "loss": 0.3649, + "step": 2878 + }, + { + "epoch": 0.2574213161659514, + "grad_norm": 0.13670482740780823, + "learning_rate": 0.0001740853875483897, + "loss": 0.6808, + "step": 2879 + }, + { + "epoch": 0.2575107296137339, + "grad_norm": 0.36766650230988457, + "learning_rate": 0.00017406593291429217, + "loss": 0.6921, + "step": 2880 + }, + { + "epoch": 0.25760014306151646, + "grad_norm": 0.11703377835384726, + "learning_rate": 0.00017404647206837432, + "loss": 0.6681, + "step": 2881 + }, + { + "epoch": 0.257689556509299, + "grad_norm": 0.7178272808657636, + "learning_rate": 0.00017402700501226826, + "loss": 0.7316, + "step": 2882 + }, + { + "epoch": 0.25777896995708155, + "grad_norm": 0.11879494789421126, + "learning_rate": 0.00017400753174760672, + "loss": 0.6483, + "step": 2883 + }, + { + "epoch": 0.25786838340486407, + "grad_norm": 0.19852847315888705, + "learning_rate": 0.0001739880522760229, + "loss": 0.6815, + "step": 2884 + }, + { + "epoch": 0.25795779685264664, + "grad_norm": 0.13272606633212675, + "learning_rate": 0.00017396856659915045, + "loss": 0.649, + "step": 2885 + }, + { + "epoch": 0.2580472103004292, + "grad_norm": 0.15631998003390127, + "learning_rate": 0.00017394907471862363, + "loss": 0.6645, + "step": 2886 + }, + { + "epoch": 0.2581366237482117, + "grad_norm": 0.1573392308984454, + "learning_rate": 0.00017392957663607723, + "loss": 0.7029, + "step": 2887 + }, + { + "epoch": 0.2582260371959943, + "grad_norm": 0.1333841489510521, + "learning_rate": 0.00017391007235314655, + "loss": 0.7051, + "step": 2888 + }, + { + "epoch": 0.2583154506437768, + "grad_norm": 0.13648135552919008, + "learning_rate": 0.00017389056187146733, + "loss": 0.6852, + "step": 2889 + }, + { + "epoch": 0.2584048640915594, + "grad_norm": 0.13353327232848686, + "learning_rate": 0.00017387104519267594, + "loss": 0.7113, + "step": 2890 + }, + { + "epoch": 0.2584942775393419, + "grad_norm": 0.1276286768009021, + "learning_rate": 0.0001738515223184092, + "loss": 0.6457, + "step": 2891 + }, + { + "epoch": 0.25858369098712447, + "grad_norm": 0.15795827908047094, + "learning_rate": 0.00017383199325030448, + "loss": 0.7011, + "step": 2892 + }, + { + "epoch": 0.258673104434907, + "grad_norm": 0.12504404612075246, + "learning_rate": 0.00017381245798999965, + "loss": 0.6656, + "step": 2893 + }, + { + "epoch": 0.25876251788268956, + "grad_norm": 0.1269218856790461, + "learning_rate": 0.00017379291653913311, + "loss": 0.6698, + "step": 2894 + }, + { + "epoch": 0.25885193133047213, + "grad_norm": 0.17420319259385733, + "learning_rate": 0.0001737733688993438, + "loss": 0.6613, + "step": 2895 + }, + { + "epoch": 0.25894134477825465, + "grad_norm": 0.13390389893301033, + "learning_rate": 0.00017375381507227108, + "loss": 0.6738, + "step": 2896 + }, + { + "epoch": 0.2590307582260372, + "grad_norm": 0.12156720318584263, + "learning_rate": 0.000173734255059555, + "loss": 0.6474, + "step": 2897 + }, + { + "epoch": 0.25912017167381973, + "grad_norm": 0.13848296162741267, + "learning_rate": 0.000173714688862836, + "loss": 0.7106, + "step": 2898 + }, + { + "epoch": 0.2592095851216023, + "grad_norm": 0.13035349156394402, + "learning_rate": 0.00017369511648375507, + "loss": 0.7017, + "step": 2899 + }, + { + "epoch": 0.2592989985693848, + "grad_norm": 0.1430832066517599, + "learning_rate": 0.00017367553792395373, + "loss": 0.6826, + "step": 2900 + }, + { + "epoch": 0.2593884120171674, + "grad_norm": 0.12924217675763078, + "learning_rate": 0.00017365595318507397, + "loss": 0.6774, + "step": 2901 + }, + { + "epoch": 0.2594778254649499, + "grad_norm": 0.1255750752423414, + "learning_rate": 0.00017363636226875836, + "loss": 0.6604, + "step": 2902 + }, + { + "epoch": 0.2595672389127325, + "grad_norm": 0.6395514922865448, + "learning_rate": 0.00017361676517665001, + "loss": 0.3568, + "step": 2903 + }, + { + "epoch": 0.259656652360515, + "grad_norm": 0.14815942682729552, + "learning_rate": 0.00017359716191039248, + "loss": 0.6823, + "step": 2904 + }, + { + "epoch": 0.25974606580829757, + "grad_norm": 0.14593391507033016, + "learning_rate": 0.00017357755247162984, + "loss": 0.7084, + "step": 2905 + }, + { + "epoch": 0.25983547925608014, + "grad_norm": 0.15622155461674236, + "learning_rate": 0.00017355793686200675, + "loss": 0.7338, + "step": 2906 + }, + { + "epoch": 0.25992489270386265, + "grad_norm": 0.1439562498678843, + "learning_rate": 0.00017353831508316834, + "loss": 0.6789, + "step": 2907 + }, + { + "epoch": 0.2600143061516452, + "grad_norm": 0.14097236156129772, + "learning_rate": 0.00017351868713676023, + "loss": 0.6747, + "step": 2908 + }, + { + "epoch": 0.26010371959942774, + "grad_norm": 0.14070758802140587, + "learning_rate": 0.00017349905302442863, + "loss": 0.6943, + "step": 2909 + }, + { + "epoch": 0.2601931330472103, + "grad_norm": 0.14835095932714482, + "learning_rate": 0.0001734794127478202, + "loss": 0.7169, + "step": 2910 + }, + { + "epoch": 0.26028254649499283, + "grad_norm": 0.14625620535449999, + "learning_rate": 0.00017345976630858218, + "loss": 0.6889, + "step": 2911 + }, + { + "epoch": 0.2603719599427754, + "grad_norm": 0.1297851891805484, + "learning_rate": 0.00017344011370836227, + "loss": 0.7104, + "step": 2912 + }, + { + "epoch": 0.2604613733905579, + "grad_norm": 0.13991572243208578, + "learning_rate": 0.00017342045494880872, + "loss": 0.6617, + "step": 2913 + }, + { + "epoch": 0.2605507868383405, + "grad_norm": 0.14513939977291596, + "learning_rate": 0.0001734007900315703, + "loss": 0.6853, + "step": 2914 + }, + { + "epoch": 0.26064020028612306, + "grad_norm": 0.13324955777442757, + "learning_rate": 0.00017338111895829624, + "loss": 0.6567, + "step": 2915 + }, + { + "epoch": 0.2607296137339056, + "grad_norm": 0.14563556361989402, + "learning_rate": 0.00017336144173063636, + "loss": 0.7049, + "step": 2916 + }, + { + "epoch": 0.26081902718168815, + "grad_norm": 0.1449557623204389, + "learning_rate": 0.00017334175835024095, + "loss": 0.7002, + "step": 2917 + }, + { + "epoch": 0.26090844062947066, + "grad_norm": 0.14169449535942014, + "learning_rate": 0.00017332206881876086, + "loss": 0.6952, + "step": 2918 + }, + { + "epoch": 0.26099785407725323, + "grad_norm": 0.15097020604282413, + "learning_rate": 0.0001733023731378474, + "loss": 0.7378, + "step": 2919 + }, + { + "epoch": 0.26108726752503575, + "grad_norm": 0.13957322384116544, + "learning_rate": 0.00017328267130915244, + "loss": 0.6372, + "step": 2920 + }, + { + "epoch": 0.2611766809728183, + "grad_norm": 0.1284481165108627, + "learning_rate": 0.00017326296333432833, + "loss": 0.6746, + "step": 2921 + }, + { + "epoch": 0.26126609442060084, + "grad_norm": 0.13378145483334314, + "learning_rate": 0.000173243249215028, + "loss": 0.6729, + "step": 2922 + }, + { + "epoch": 0.2613555078683834, + "grad_norm": 0.126490315442072, + "learning_rate": 0.00017322352895290477, + "loss": 0.6586, + "step": 2923 + }, + { + "epoch": 0.261444921316166, + "grad_norm": 0.12766839453962422, + "learning_rate": 0.0001732038025496126, + "loss": 0.6791, + "step": 2924 + }, + { + "epoch": 0.2615343347639485, + "grad_norm": 0.1413825841736304, + "learning_rate": 0.0001731840700068059, + "loss": 0.6945, + "step": 2925 + }, + { + "epoch": 0.26162374821173107, + "grad_norm": 0.14753407711476174, + "learning_rate": 0.00017316433132613969, + "loss": 0.7028, + "step": 2926 + }, + { + "epoch": 0.2617131616595136, + "grad_norm": 0.12139983632159071, + "learning_rate": 0.00017314458650926934, + "loss": 0.6688, + "step": 2927 + }, + { + "epoch": 0.26180257510729615, + "grad_norm": 0.13315987727444425, + "learning_rate": 0.00017312483555785086, + "loss": 0.6842, + "step": 2928 + }, + { + "epoch": 0.26189198855507867, + "grad_norm": 0.12302831112309963, + "learning_rate": 0.00017310507847354077, + "loss": 0.6747, + "step": 2929 + }, + { + "epoch": 0.26198140200286124, + "grad_norm": 0.12287650864574676, + "learning_rate": 0.00017308531525799597, + "loss": 0.645, + "step": 2930 + }, + { + "epoch": 0.26207081545064376, + "grad_norm": 0.15127155483310536, + "learning_rate": 0.0001730655459128741, + "loss": 0.6611, + "step": 2931 + }, + { + "epoch": 0.26216022889842633, + "grad_norm": 0.1475165296745581, + "learning_rate": 0.0001730457704398331, + "loss": 0.6846, + "step": 2932 + }, + { + "epoch": 0.26224964234620884, + "grad_norm": 0.1433228599443013, + "learning_rate": 0.00017302598884053153, + "loss": 0.6799, + "step": 2933 + }, + { + "epoch": 0.2623390557939914, + "grad_norm": 0.12262415891082779, + "learning_rate": 0.00017300620111662852, + "loss": 0.6278, + "step": 2934 + }, + { + "epoch": 0.262428469241774, + "grad_norm": 0.13696056694390138, + "learning_rate": 0.00017298640726978357, + "loss": 0.7029, + "step": 2935 + }, + { + "epoch": 0.2625178826895565, + "grad_norm": 0.13136745854973692, + "learning_rate": 0.00017296660730165678, + "loss": 0.6743, + "step": 2936 + }, + { + "epoch": 0.2626072961373391, + "grad_norm": 0.13850479157218024, + "learning_rate": 0.00017294680121390877, + "loss": 0.6624, + "step": 2937 + }, + { + "epoch": 0.2626967095851216, + "grad_norm": 0.11638103579511704, + "learning_rate": 0.00017292698900820064, + "loss": 0.6739, + "step": 2938 + }, + { + "epoch": 0.26278612303290416, + "grad_norm": 0.1307480749000159, + "learning_rate": 0.00017290717068619402, + "loss": 0.6551, + "step": 2939 + }, + { + "epoch": 0.2628755364806867, + "grad_norm": 0.13826825681635743, + "learning_rate": 0.00017288734624955102, + "loss": 0.6959, + "step": 2940 + }, + { + "epoch": 0.26296494992846925, + "grad_norm": 0.15276870252326533, + "learning_rate": 0.00017286751569993433, + "loss": 0.7051, + "step": 2941 + }, + { + "epoch": 0.26305436337625177, + "grad_norm": 0.11431566385147478, + "learning_rate": 0.0001728476790390071, + "loss": 0.716, + "step": 2942 + }, + { + "epoch": 0.26314377682403434, + "grad_norm": 0.1312389222718282, + "learning_rate": 0.00017282783626843302, + "loss": 0.6837, + "step": 2943 + }, + { + "epoch": 0.2632331902718169, + "grad_norm": 0.15124069951134927, + "learning_rate": 0.00017280798738987624, + "loss": 0.671, + "step": 2944 + }, + { + "epoch": 0.2633226037195994, + "grad_norm": 0.11979049784066613, + "learning_rate": 0.00017278813240500154, + "loss": 0.6185, + "step": 2945 + }, + { + "epoch": 0.263412017167382, + "grad_norm": 0.1448451409437601, + "learning_rate": 0.000172768271315474, + "loss": 0.6879, + "step": 2946 + }, + { + "epoch": 0.2635014306151645, + "grad_norm": 0.1227729951783725, + "learning_rate": 0.00017274840412295948, + "loss": 0.6861, + "step": 2947 + }, + { + "epoch": 0.2635908440629471, + "grad_norm": 0.11428978588133518, + "learning_rate": 0.00017272853082912418, + "loss": 0.6486, + "step": 2948 + }, + { + "epoch": 0.2636802575107296, + "grad_norm": 0.15410762747758125, + "learning_rate": 0.00017270865143563478, + "loss": 0.6878, + "step": 2949 + }, + { + "epoch": 0.26376967095851217, + "grad_norm": 0.14416873960160323, + "learning_rate": 0.00017268876594415863, + "loss": 0.6786, + "step": 2950 + }, + { + "epoch": 0.2638590844062947, + "grad_norm": 0.12707156445679096, + "learning_rate": 0.00017266887435636344, + "loss": 0.6528, + "step": 2951 + }, + { + "epoch": 0.26394849785407726, + "grad_norm": 0.12582275934179996, + "learning_rate": 0.00017264897667391754, + "loss": 0.6486, + "step": 2952 + }, + { + "epoch": 0.2640379113018598, + "grad_norm": 0.1395029864710135, + "learning_rate": 0.0001726290728984897, + "loss": 0.6831, + "step": 2953 + }, + { + "epoch": 0.26412732474964234, + "grad_norm": 0.1563540001659561, + "learning_rate": 0.00017260916303174923, + "loss": 0.7419, + "step": 2954 + }, + { + "epoch": 0.2642167381974249, + "grad_norm": 0.12940133883439334, + "learning_rate": 0.00017258924707536596, + "loss": 0.6865, + "step": 2955 + }, + { + "epoch": 0.26430615164520743, + "grad_norm": 0.11780268617009172, + "learning_rate": 0.00017256932503101018, + "loss": 0.6895, + "step": 2956 + }, + { + "epoch": 0.26439556509299, + "grad_norm": 0.12932237866053253, + "learning_rate": 0.00017254939690035276, + "loss": 0.6831, + "step": 2957 + }, + { + "epoch": 0.2644849785407725, + "grad_norm": 0.15075506720087042, + "learning_rate": 0.00017252946268506505, + "loss": 0.7603, + "step": 2958 + }, + { + "epoch": 0.2645743919885551, + "grad_norm": 0.13772314117405804, + "learning_rate": 0.00017250952238681889, + "loss": 0.7062, + "step": 2959 + }, + { + "epoch": 0.2646638054363376, + "grad_norm": 0.12245720557582407, + "learning_rate": 0.00017248957600728664, + "loss": 0.6769, + "step": 2960 + }, + { + "epoch": 0.2647532188841202, + "grad_norm": 0.14929653518585326, + "learning_rate": 0.0001724696235481412, + "loss": 0.6829, + "step": 2961 + }, + { + "epoch": 0.2648426323319027, + "grad_norm": 0.11691163077160172, + "learning_rate": 0.00017244966501105596, + "loss": 0.6804, + "step": 2962 + }, + { + "epoch": 0.26493204577968527, + "grad_norm": 0.1514353762833311, + "learning_rate": 0.0001724297003977048, + "loss": 0.7109, + "step": 2963 + }, + { + "epoch": 0.26502145922746784, + "grad_norm": 0.13547667384008028, + "learning_rate": 0.0001724097297097622, + "loss": 0.7342, + "step": 2964 + }, + { + "epoch": 0.26511087267525035, + "grad_norm": 0.11356346170836748, + "learning_rate": 0.00017238975294890297, + "loss": 0.6739, + "step": 2965 + }, + { + "epoch": 0.2652002861230329, + "grad_norm": 0.12768482884447324, + "learning_rate": 0.00017236977011680257, + "loss": 0.6491, + "step": 2966 + }, + { + "epoch": 0.26528969957081544, + "grad_norm": 0.13471089722031354, + "learning_rate": 0.00017234978121513699, + "loss": 0.6955, + "step": 2967 + }, + { + "epoch": 0.265379113018598, + "grad_norm": 0.126699361363956, + "learning_rate": 0.0001723297862455826, + "loss": 0.6928, + "step": 2968 + }, + { + "epoch": 0.2654685264663805, + "grad_norm": 0.12002017314377668, + "learning_rate": 0.00017230978520981643, + "loss": 0.668, + "step": 2969 + }, + { + "epoch": 0.2655579399141631, + "grad_norm": 0.12923311913395252, + "learning_rate": 0.00017228977810951584, + "loss": 0.6689, + "step": 2970 + }, + { + "epoch": 0.2656473533619456, + "grad_norm": 0.12453602035504126, + "learning_rate": 0.00017226976494635893, + "loss": 0.6508, + "step": 2971 + }, + { + "epoch": 0.2657367668097282, + "grad_norm": 0.12896678546554308, + "learning_rate": 0.00017224974572202409, + "loss": 0.6834, + "step": 2972 + }, + { + "epoch": 0.2658261802575107, + "grad_norm": 0.1355244729056894, + "learning_rate": 0.0001722297204381903, + "loss": 0.6531, + "step": 2973 + }, + { + "epoch": 0.2659155937052933, + "grad_norm": 0.11608207525352529, + "learning_rate": 0.00017220968909653715, + "loss": 0.649, + "step": 2974 + }, + { + "epoch": 0.26600500715307585, + "grad_norm": 0.1203119185240794, + "learning_rate": 0.00017218965169874456, + "loss": 0.6931, + "step": 2975 + }, + { + "epoch": 0.26609442060085836, + "grad_norm": 0.1250098775041532, + "learning_rate": 0.00017216960824649303, + "loss": 0.6768, + "step": 2976 + }, + { + "epoch": 0.26618383404864093, + "grad_norm": 0.13313393579934163, + "learning_rate": 0.00017214955874146363, + "loss": 0.7017, + "step": 2977 + }, + { + "epoch": 0.26627324749642345, + "grad_norm": 0.12416391141360913, + "learning_rate": 0.00017212950318533788, + "loss": 0.6907, + "step": 2978 + }, + { + "epoch": 0.266362660944206, + "grad_norm": 0.14827016292577017, + "learning_rate": 0.00017210944157979783, + "loss": 0.6924, + "step": 2979 + }, + { + "epoch": 0.26645207439198854, + "grad_norm": 0.1390528808208819, + "learning_rate": 0.00017208937392652594, + "loss": 0.728, + "step": 2980 + }, + { + "epoch": 0.2665414878397711, + "grad_norm": 0.13470574459897655, + "learning_rate": 0.0001720693002272054, + "loss": 0.6989, + "step": 2981 + }, + { + "epoch": 0.2666309012875536, + "grad_norm": 0.13651948050724072, + "learning_rate": 0.00017204922048351964, + "loss": 0.6691, + "step": 2982 + }, + { + "epoch": 0.2667203147353362, + "grad_norm": 0.13180150762142404, + "learning_rate": 0.0001720291346971528, + "loss": 0.6893, + "step": 2983 + }, + { + "epoch": 0.26680972818311877, + "grad_norm": 0.12568846487506, + "learning_rate": 0.0001720090428697894, + "loss": 0.704, + "step": 2984 + }, + { + "epoch": 0.2668991416309013, + "grad_norm": 0.14949941952202397, + "learning_rate": 0.00017198894500311453, + "loss": 0.7029, + "step": 2985 + }, + { + "epoch": 0.26698855507868385, + "grad_norm": 0.11509551882418091, + "learning_rate": 0.0001719688410988138, + "loss": 0.6771, + "step": 2986 + }, + { + "epoch": 0.26707796852646637, + "grad_norm": 0.12017509453906877, + "learning_rate": 0.00017194873115857328, + "loss": 0.6869, + "step": 2987 + }, + { + "epoch": 0.26716738197424894, + "grad_norm": 0.12031409898734885, + "learning_rate": 0.00017192861518407958, + "loss": 0.6692, + "step": 2988 + }, + { + "epoch": 0.26725679542203146, + "grad_norm": 0.13330845976866956, + "learning_rate": 0.00017190849317701975, + "loss": 0.6778, + "step": 2989 + }, + { + "epoch": 0.26734620886981403, + "grad_norm": 0.12279796467674928, + "learning_rate": 0.00017188836513908152, + "loss": 0.6848, + "step": 2990 + }, + { + "epoch": 0.26743562231759654, + "grad_norm": 0.12472314877447199, + "learning_rate": 0.00017186823107195287, + "loss": 0.6903, + "step": 2991 + }, + { + "epoch": 0.2675250357653791, + "grad_norm": 0.14201230261194847, + "learning_rate": 0.00017184809097732246, + "loss": 0.7041, + "step": 2992 + }, + { + "epoch": 0.2676144492131617, + "grad_norm": 0.13212324956227922, + "learning_rate": 0.00017182794485687944, + "loss": 0.6829, + "step": 2993 + }, + { + "epoch": 0.2677038626609442, + "grad_norm": 0.11524097455219574, + "learning_rate": 0.00017180779271231344, + "loss": 0.6773, + "step": 2994 + }, + { + "epoch": 0.2677932761087268, + "grad_norm": 0.12297535840184355, + "learning_rate": 0.0001717876345453146, + "loss": 0.6817, + "step": 2995 + }, + { + "epoch": 0.2678826895565093, + "grad_norm": 0.10744605424565695, + "learning_rate": 0.00017176747035757355, + "loss": 0.6767, + "step": 2996 + }, + { + "epoch": 0.26797210300429186, + "grad_norm": 0.1414480502871035, + "learning_rate": 0.0001717473001507814, + "loss": 0.7022, + "step": 2997 + }, + { + "epoch": 0.2680615164520744, + "grad_norm": 0.13165453714172898, + "learning_rate": 0.00017172712392662988, + "loss": 0.7325, + "step": 2998 + }, + { + "epoch": 0.26815092989985695, + "grad_norm": 0.12550944125125402, + "learning_rate": 0.00017170694168681106, + "loss": 0.6927, + "step": 2999 + }, + { + "epoch": 0.26824034334763946, + "grad_norm": 0.1318499367568541, + "learning_rate": 0.00017168675343301769, + "loss": 0.6954, + "step": 3000 + }, + { + "epoch": 0.26832975679542204, + "grad_norm": 0.1401550188558601, + "learning_rate": 0.00017166655916694284, + "loss": 0.7067, + "step": 3001 + }, + { + "epoch": 0.26841917024320455, + "grad_norm": 0.14529722394161085, + "learning_rate": 0.00017164635889028025, + "loss": 0.6848, + "step": 3002 + }, + { + "epoch": 0.2685085836909871, + "grad_norm": 0.13644837347977767, + "learning_rate": 0.00017162615260472402, + "loss": 0.7122, + "step": 3003 + }, + { + "epoch": 0.2685979971387697, + "grad_norm": 0.1381248235186665, + "learning_rate": 0.00017160594031196894, + "loss": 0.6646, + "step": 3004 + }, + { + "epoch": 0.2686874105865522, + "grad_norm": 0.16168223944246873, + "learning_rate": 0.00017158572201371008, + "loss": 0.7174, + "step": 3005 + }, + { + "epoch": 0.2687768240343348, + "grad_norm": 0.10989679923345014, + "learning_rate": 0.00017156549771164318, + "loss": 0.6767, + "step": 3006 + }, + { + "epoch": 0.2688662374821173, + "grad_norm": 0.11499294083188892, + "learning_rate": 0.00017154526740746442, + "loss": 0.6832, + "step": 3007 + }, + { + "epoch": 0.26895565092989987, + "grad_norm": 0.12270110677867556, + "learning_rate": 0.00017152503110287048, + "loss": 0.6788, + "step": 3008 + }, + { + "epoch": 0.2690450643776824, + "grad_norm": 0.13597181590804433, + "learning_rate": 0.00017150478879955858, + "loss": 0.6782, + "step": 3009 + }, + { + "epoch": 0.26913447782546496, + "grad_norm": 0.1295458940643419, + "learning_rate": 0.00017148454049922636, + "loss": 0.6638, + "step": 3010 + }, + { + "epoch": 0.2692238912732475, + "grad_norm": 0.12049188538642809, + "learning_rate": 0.00017146428620357212, + "loss": 0.698, + "step": 3011 + }, + { + "epoch": 0.26931330472103004, + "grad_norm": 0.12896891912520106, + "learning_rate": 0.00017144402591429448, + "loss": 0.6612, + "step": 3012 + }, + { + "epoch": 0.2694027181688126, + "grad_norm": 0.11460892219715033, + "learning_rate": 0.00017142375963309262, + "loss": 0.6713, + "step": 3013 + }, + { + "epoch": 0.26949213161659513, + "grad_norm": 0.12518528607492624, + "learning_rate": 0.00017140348736166636, + "loss": 0.6807, + "step": 3014 + }, + { + "epoch": 0.2695815450643777, + "grad_norm": 0.12123349402427727, + "learning_rate": 0.00017138320910171584, + "loss": 0.6966, + "step": 3015 + }, + { + "epoch": 0.2696709585121602, + "grad_norm": 0.14474528334629408, + "learning_rate": 0.00017136292485494175, + "loss": 0.6908, + "step": 3016 + }, + { + "epoch": 0.2697603719599428, + "grad_norm": 0.11426853691945782, + "learning_rate": 0.00017134263462304533, + "loss": 0.6836, + "step": 3017 + }, + { + "epoch": 0.2698497854077253, + "grad_norm": 0.13509485950254496, + "learning_rate": 0.00017132233840772836, + "loss": 0.6757, + "step": 3018 + }, + { + "epoch": 0.2699391988555079, + "grad_norm": 0.11711871263620013, + "learning_rate": 0.00017130203621069297, + "loss": 0.6376, + "step": 3019 + }, + { + "epoch": 0.2700286123032904, + "grad_norm": 0.1363185280714207, + "learning_rate": 0.00017128172803364188, + "loss": 0.7054, + "step": 3020 + }, + { + "epoch": 0.27011802575107297, + "grad_norm": 0.12408093784158518, + "learning_rate": 0.0001712614138782784, + "loss": 0.6737, + "step": 3021 + }, + { + "epoch": 0.2702074391988555, + "grad_norm": 0.13254225091581512, + "learning_rate": 0.00017124109374630616, + "loss": 0.6834, + "step": 3022 + }, + { + "epoch": 0.27029685264663805, + "grad_norm": 0.13731221287062417, + "learning_rate": 0.00017122076763942946, + "loss": 0.6977, + "step": 3023 + }, + { + "epoch": 0.2703862660944206, + "grad_norm": 0.12452341052921993, + "learning_rate": 0.00017120043555935298, + "loss": 0.6608, + "step": 3024 + }, + { + "epoch": 0.27047567954220314, + "grad_norm": 0.12938251303746084, + "learning_rate": 0.00017118009750778196, + "loss": 0.6603, + "step": 3025 + }, + { + "epoch": 0.2705650929899857, + "grad_norm": 0.11858505810310292, + "learning_rate": 0.00017115975348642212, + "loss": 0.681, + "step": 3026 + }, + { + "epoch": 0.2706545064377682, + "grad_norm": 0.11923702576617531, + "learning_rate": 0.00017113940349697967, + "loss": 0.6524, + "step": 3027 + }, + { + "epoch": 0.2707439198855508, + "grad_norm": 0.13952167901966964, + "learning_rate": 0.00017111904754116142, + "loss": 0.7129, + "step": 3028 + }, + { + "epoch": 0.2708333333333333, + "grad_norm": 0.09839438392055597, + "learning_rate": 0.0001710986856206745, + "loss": 0.6613, + "step": 3029 + }, + { + "epoch": 0.2709227467811159, + "grad_norm": 0.12369627249216338, + "learning_rate": 0.00017107831773722668, + "loss": 0.6858, + "step": 3030 + }, + { + "epoch": 0.2710121602288984, + "grad_norm": 0.133962362111956, + "learning_rate": 0.00017105794389252622, + "loss": 0.6635, + "step": 3031 + }, + { + "epoch": 0.271101573676681, + "grad_norm": 0.14601217746411846, + "learning_rate": 0.00017103756408828183, + "loss": 0.7038, + "step": 3032 + }, + { + "epoch": 0.27119098712446355, + "grad_norm": 0.12474475968691143, + "learning_rate": 0.0001710171783262027, + "loss": 0.6879, + "step": 3033 + }, + { + "epoch": 0.27128040057224606, + "grad_norm": 0.11454849981321781, + "learning_rate": 0.00017099678660799857, + "loss": 0.6422, + "step": 3034 + }, + { + "epoch": 0.27136981402002863, + "grad_norm": 0.12359121838449139, + "learning_rate": 0.00017097638893537976, + "loss": 0.6787, + "step": 3035 + }, + { + "epoch": 0.27145922746781115, + "grad_norm": 0.1236910341197007, + "learning_rate": 0.00017095598531005688, + "loss": 0.6555, + "step": 3036 + }, + { + "epoch": 0.2715486409155937, + "grad_norm": 0.1262987099701944, + "learning_rate": 0.0001709355757337412, + "loss": 0.669, + "step": 3037 + }, + { + "epoch": 0.27163805436337624, + "grad_norm": 0.13202067309259283, + "learning_rate": 0.00017091516020814447, + "loss": 0.7107, + "step": 3038 + }, + { + "epoch": 0.2717274678111588, + "grad_norm": 0.13592826673546418, + "learning_rate": 0.0001708947387349789, + "loss": 0.6972, + "step": 3039 + }, + { + "epoch": 0.2718168812589413, + "grad_norm": 0.12161658118691426, + "learning_rate": 0.0001708743113159572, + "loss": 0.6853, + "step": 3040 + }, + { + "epoch": 0.2719062947067239, + "grad_norm": 0.12742040315357933, + "learning_rate": 0.0001708538779527926, + "loss": 0.6735, + "step": 3041 + }, + { + "epoch": 0.2719957081545064, + "grad_norm": 0.1359168553388471, + "learning_rate": 0.00017083343864719884, + "loss": 0.737, + "step": 3042 + }, + { + "epoch": 0.272085121602289, + "grad_norm": 0.13610332420321428, + "learning_rate": 0.00017081299340089012, + "loss": 0.6813, + "step": 3043 + }, + { + "epoch": 0.27217453505007155, + "grad_norm": 0.10735506935765922, + "learning_rate": 0.00017079254221558115, + "loss": 0.641, + "step": 3044 + }, + { + "epoch": 0.27226394849785407, + "grad_norm": 0.121941623916289, + "learning_rate": 0.00017077208509298718, + "loss": 0.6935, + "step": 3045 + }, + { + "epoch": 0.27235336194563664, + "grad_norm": 0.13557618823275516, + "learning_rate": 0.0001707516220348239, + "loss": 0.704, + "step": 3046 + }, + { + "epoch": 0.27244277539341916, + "grad_norm": 0.12582068619145692, + "learning_rate": 0.00017073115304280754, + "loss": 0.6281, + "step": 3047 + }, + { + "epoch": 0.27253218884120173, + "grad_norm": 0.1345732674307166, + "learning_rate": 0.00017071067811865476, + "loss": 0.6972, + "step": 3048 + }, + { + "epoch": 0.27262160228898424, + "grad_norm": 0.14624416831686987, + "learning_rate": 0.00017069019726408282, + "loss": 0.6844, + "step": 3049 + }, + { + "epoch": 0.2727110157367668, + "grad_norm": 0.13629958244899848, + "learning_rate": 0.0001706697104808094, + "loss": 0.7185, + "step": 3050 + }, + { + "epoch": 0.27280042918454933, + "grad_norm": 0.13287708639431892, + "learning_rate": 0.00017064921777055272, + "loss": 0.6485, + "step": 3051 + }, + { + "epoch": 0.2728898426323319, + "grad_norm": 0.13995605728424018, + "learning_rate": 0.00017062871913503148, + "loss": 0.6731, + "step": 3052 + }, + { + "epoch": 0.2729792560801145, + "grad_norm": 0.11153428241235072, + "learning_rate": 0.00017060821457596487, + "loss": 0.6467, + "step": 3053 + }, + { + "epoch": 0.273068669527897, + "grad_norm": 0.14204474085572505, + "learning_rate": 0.00017058770409507254, + "loss": 0.7133, + "step": 3054 + }, + { + "epoch": 0.27315808297567956, + "grad_norm": 0.1154314026022467, + "learning_rate": 0.00017056718769407474, + "loss": 0.6669, + "step": 3055 + }, + { + "epoch": 0.2732474964234621, + "grad_norm": 0.12793132874856786, + "learning_rate": 0.00017054666537469213, + "loss": 0.7229, + "step": 3056 + }, + { + "epoch": 0.27333690987124465, + "grad_norm": 0.13954426211485368, + "learning_rate": 0.00017052613713864587, + "loss": 0.7033, + "step": 3057 + }, + { + "epoch": 0.27342632331902716, + "grad_norm": 0.125480799498345, + "learning_rate": 0.0001705056029876577, + "loss": 0.6778, + "step": 3058 + }, + { + "epoch": 0.27351573676680974, + "grad_norm": 0.12107099625137051, + "learning_rate": 0.00017048506292344974, + "loss": 0.6356, + "step": 3059 + }, + { + "epoch": 0.27360515021459225, + "grad_norm": 0.12804815141327885, + "learning_rate": 0.00017046451694774467, + "loss": 0.671, + "step": 3060 + }, + { + "epoch": 0.2736945636623748, + "grad_norm": 0.11541624000775119, + "learning_rate": 0.00017044396506226566, + "loss": 0.6762, + "step": 3061 + }, + { + "epoch": 0.2737839771101574, + "grad_norm": 0.13732526212659477, + "learning_rate": 0.0001704234072687364, + "loss": 0.666, + "step": 3062 + }, + { + "epoch": 0.2738733905579399, + "grad_norm": 0.13689827309578842, + "learning_rate": 0.000170402843568881, + "loss": 0.6796, + "step": 3063 + }, + { + "epoch": 0.2739628040057225, + "grad_norm": 0.13280909160363835, + "learning_rate": 0.00017038227396442415, + "loss": 0.7198, + "step": 3064 + }, + { + "epoch": 0.274052217453505, + "grad_norm": 0.13943213894782924, + "learning_rate": 0.00017036169845709097, + "loss": 0.6763, + "step": 3065 + }, + { + "epoch": 0.27414163090128757, + "grad_norm": 0.14092732463711594, + "learning_rate": 0.00017034111704860712, + "loss": 0.7539, + "step": 3066 + }, + { + "epoch": 0.2742310443490701, + "grad_norm": 0.135172115867354, + "learning_rate": 0.00017032052974069874, + "loss": 0.6667, + "step": 3067 + }, + { + "epoch": 0.27432045779685266, + "grad_norm": 0.13137420604129071, + "learning_rate": 0.00017029993653509243, + "loss": 0.6719, + "step": 3068 + }, + { + "epoch": 0.2744098712446352, + "grad_norm": 0.1170299265171342, + "learning_rate": 0.0001702793374335154, + "loss": 0.6368, + "step": 3069 + }, + { + "epoch": 0.27449928469241774, + "grad_norm": 0.13016041364095818, + "learning_rate": 0.00017025873243769517, + "loss": 0.6557, + "step": 3070 + }, + { + "epoch": 0.27458869814020026, + "grad_norm": 0.16051162757724788, + "learning_rate": 0.0001702381215493599, + "loss": 0.7381, + "step": 3071 + }, + { + "epoch": 0.27467811158798283, + "grad_norm": 0.13270359892672498, + "learning_rate": 0.0001702175047702382, + "loss": 0.6883, + "step": 3072 + }, + { + "epoch": 0.2747675250357654, + "grad_norm": 0.146014502644136, + "learning_rate": 0.00017019688210205918, + "loss": 0.6761, + "step": 3073 + }, + { + "epoch": 0.2748569384835479, + "grad_norm": 0.15705843526866575, + "learning_rate": 0.00017017625354655245, + "loss": 0.6763, + "step": 3074 + }, + { + "epoch": 0.2749463519313305, + "grad_norm": 0.13246241284658203, + "learning_rate": 0.00017015561910544807, + "loss": 0.6985, + "step": 3075 + }, + { + "epoch": 0.275035765379113, + "grad_norm": 0.15324374104476543, + "learning_rate": 0.00017013497878047668, + "loss": 0.7494, + "step": 3076 + }, + { + "epoch": 0.2751251788268956, + "grad_norm": 0.12611131242831838, + "learning_rate": 0.0001701143325733693, + "loss": 0.7034, + "step": 3077 + }, + { + "epoch": 0.2752145922746781, + "grad_norm": 0.14246075775881226, + "learning_rate": 0.0001700936804858575, + "loss": 0.6996, + "step": 3078 + }, + { + "epoch": 0.27530400572246067, + "grad_norm": 0.13461518693971977, + "learning_rate": 0.00017007302251967338, + "loss": 0.7152, + "step": 3079 + }, + { + "epoch": 0.2753934191702432, + "grad_norm": 0.1355594790730842, + "learning_rate": 0.0001700523586765495, + "loss": 0.6505, + "step": 3080 + }, + { + "epoch": 0.27548283261802575, + "grad_norm": 0.1225259646505168, + "learning_rate": 0.00017003168895821888, + "loss": 0.6373, + "step": 3081 + }, + { + "epoch": 0.2755722460658083, + "grad_norm": 0.12924214381757548, + "learning_rate": 0.00017001101336641512, + "loss": 0.6655, + "step": 3082 + }, + { + "epoch": 0.27566165951359084, + "grad_norm": 0.14128040347761098, + "learning_rate": 0.0001699903319028722, + "loss": 0.7345, + "step": 3083 + }, + { + "epoch": 0.2757510729613734, + "grad_norm": 0.11390018141635316, + "learning_rate": 0.00016996964456932466, + "loss": 0.6851, + "step": 3084 + }, + { + "epoch": 0.2758404864091559, + "grad_norm": 0.11686933551553898, + "learning_rate": 0.0001699489513675075, + "loss": 0.663, + "step": 3085 + }, + { + "epoch": 0.2759298998569385, + "grad_norm": 0.11523584197890306, + "learning_rate": 0.00016992825229915636, + "loss": 0.6552, + "step": 3086 + }, + { + "epoch": 0.276019313304721, + "grad_norm": 0.1259629517128753, + "learning_rate": 0.0001699075473660071, + "loss": 0.6767, + "step": 3087 + }, + { + "epoch": 0.2761087267525036, + "grad_norm": 0.1191663251355355, + "learning_rate": 0.00016988683656979624, + "loss": 0.6819, + "step": 3088 + }, + { + "epoch": 0.2761981402002861, + "grad_norm": 0.12266943137663087, + "learning_rate": 0.00016986611991226086, + "loss": 0.6461, + "step": 3089 + }, + { + "epoch": 0.2762875536480687, + "grad_norm": 0.13395760254735284, + "learning_rate": 0.00016984539739513835, + "loss": 0.6733, + "step": 3090 + }, + { + "epoch": 0.2763769670958512, + "grad_norm": 0.12124669105118877, + "learning_rate": 0.0001698246690201667, + "loss": 0.6607, + "step": 3091 + }, + { + "epoch": 0.27646638054363376, + "grad_norm": 0.1272185295230469, + "learning_rate": 0.00016980393478908438, + "loss": 0.6663, + "step": 3092 + }, + { + "epoch": 0.27655579399141633, + "grad_norm": 0.14586012498718895, + "learning_rate": 0.00016978319470363035, + "loss": 0.7013, + "step": 3093 + }, + { + "epoch": 0.27664520743919885, + "grad_norm": 0.12751853752678977, + "learning_rate": 0.0001697624487655441, + "loss": 0.6474, + "step": 3094 + }, + { + "epoch": 0.2767346208869814, + "grad_norm": 0.11686275630008985, + "learning_rate": 0.0001697416969765655, + "loss": 0.7012, + "step": 3095 + }, + { + "epoch": 0.27682403433476394, + "grad_norm": 0.11297895515153165, + "learning_rate": 0.000169720939338435, + "loss": 0.6476, + "step": 3096 + }, + { + "epoch": 0.2769134477825465, + "grad_norm": 0.1265157500026224, + "learning_rate": 0.0001697001758528935, + "loss": 0.6636, + "step": 3097 + }, + { + "epoch": 0.277002861230329, + "grad_norm": 0.11997187118661885, + "learning_rate": 0.00016967940652168247, + "loss": 0.6699, + "step": 3098 + }, + { + "epoch": 0.2770922746781116, + "grad_norm": 0.11976045850080837, + "learning_rate": 0.00016965863134654372, + "loss": 0.641, + "step": 3099 + }, + { + "epoch": 0.2771816881258941, + "grad_norm": 0.1335674176634491, + "learning_rate": 0.0001696378503292197, + "loss": 0.6848, + "step": 3100 + }, + { + "epoch": 0.2772711015736767, + "grad_norm": 0.11601982336450223, + "learning_rate": 0.0001696170634714533, + "loss": 0.6434, + "step": 3101 + }, + { + "epoch": 0.27736051502145925, + "grad_norm": 0.1278268537432063, + "learning_rate": 0.00016959627077498782, + "loss": 0.7014, + "step": 3102 + }, + { + "epoch": 0.27744992846924177, + "grad_norm": 0.13103572678536712, + "learning_rate": 0.00016957547224156718, + "loss": 0.6702, + "step": 3103 + }, + { + "epoch": 0.27753934191702434, + "grad_norm": 0.11501112562856704, + "learning_rate": 0.00016955466787293576, + "loss": 0.6409, + "step": 3104 + }, + { + "epoch": 0.27762875536480686, + "grad_norm": 0.1170015034026321, + "learning_rate": 0.00016953385767083827, + "loss": 0.657, + "step": 3105 + }, + { + "epoch": 0.2777181688125894, + "grad_norm": 0.14266385752512797, + "learning_rate": 0.00016951304163702013, + "loss": 0.6814, + "step": 3106 + }, + { + "epoch": 0.27780758226037194, + "grad_norm": 0.1464431791535085, + "learning_rate": 0.00016949221977322716, + "loss": 0.7317, + "step": 3107 + }, + { + "epoch": 0.2778969957081545, + "grad_norm": 0.15096239863888505, + "learning_rate": 0.00016947139208120564, + "loss": 0.6915, + "step": 3108 + }, + { + "epoch": 0.27798640915593703, + "grad_norm": 0.1334844333895743, + "learning_rate": 0.00016945055856270236, + "loss": 0.6929, + "step": 3109 + }, + { + "epoch": 0.2780758226037196, + "grad_norm": 1.2056643150968096, + "learning_rate": 0.0001694297192194646, + "loss": 0.453, + "step": 3110 + }, + { + "epoch": 0.2781652360515021, + "grad_norm": 0.1932695407745898, + "learning_rate": 0.00016940887405324015, + "loss": 0.6976, + "step": 3111 + }, + { + "epoch": 0.2782546494992847, + "grad_norm": 0.15039501312739323, + "learning_rate": 0.00016938802306577726, + "loss": 0.6744, + "step": 3112 + }, + { + "epoch": 0.27834406294706726, + "grad_norm": 0.19951088392239427, + "learning_rate": 0.00016936716625882468, + "loss": 0.7248, + "step": 3113 + }, + { + "epoch": 0.2784334763948498, + "grad_norm": 0.17478099534895242, + "learning_rate": 0.00016934630363413163, + "loss": 0.6395, + "step": 3114 + }, + { + "epoch": 0.27852288984263235, + "grad_norm": 0.12830701130629693, + "learning_rate": 0.00016932543519344783, + "loss": 0.6663, + "step": 3115 + }, + { + "epoch": 0.27861230329041486, + "grad_norm": 0.12911056765825263, + "learning_rate": 0.00016930456093852353, + "loss": 0.7011, + "step": 3116 + }, + { + "epoch": 0.27870171673819744, + "grad_norm": 0.12133868143381774, + "learning_rate": 0.00016928368087110938, + "loss": 0.6352, + "step": 3117 + }, + { + "epoch": 0.27879113018597995, + "grad_norm": 0.16159516551695755, + "learning_rate": 0.0001692627949929566, + "loss": 0.6908, + "step": 3118 + }, + { + "epoch": 0.2788805436337625, + "grad_norm": 0.16833925272837116, + "learning_rate": 0.00016924190330581685, + "loss": 0.6948, + "step": 3119 + }, + { + "epoch": 0.27896995708154504, + "grad_norm": 0.12231826289054558, + "learning_rate": 0.00016922100581144228, + "loss": 0.6146, + "step": 3120 + }, + { + "epoch": 0.2790593705293276, + "grad_norm": 0.1269430879390714, + "learning_rate": 0.0001692001025115856, + "loss": 0.6387, + "step": 3121 + }, + { + "epoch": 0.2791487839771102, + "grad_norm": 0.13009808650614688, + "learning_rate": 0.00016917919340799986, + "loss": 0.6697, + "step": 3122 + }, + { + "epoch": 0.2792381974248927, + "grad_norm": 0.11840845564169107, + "learning_rate": 0.00016915827850243868, + "loss": 0.6412, + "step": 3123 + }, + { + "epoch": 0.27932761087267527, + "grad_norm": 0.14128861206380092, + "learning_rate": 0.00016913735779665627, + "loss": 0.6701, + "step": 3124 + }, + { + "epoch": 0.2794170243204578, + "grad_norm": 0.14806368907872258, + "learning_rate": 0.00016911643129240714, + "loss": 0.7237, + "step": 3125 + }, + { + "epoch": 0.27950643776824036, + "grad_norm": 0.12541146653012641, + "learning_rate": 0.00016909549899144635, + "loss": 0.6566, + "step": 3126 + }, + { + "epoch": 0.2795958512160229, + "grad_norm": 0.13111771435430405, + "learning_rate": 0.00016907456089552953, + "loss": 0.6617, + "step": 3127 + }, + { + "epoch": 0.27968526466380544, + "grad_norm": 0.11594695720889933, + "learning_rate": 0.00016905361700641271, + "loss": 0.6572, + "step": 3128 + }, + { + "epoch": 0.27977467811158796, + "grad_norm": 0.13240823897184945, + "learning_rate": 0.00016903266732585243, + "loss": 0.6696, + "step": 3129 + }, + { + "epoch": 0.27986409155937053, + "grad_norm": 0.12665643580297414, + "learning_rate": 0.00016901171185560574, + "loss": 0.6708, + "step": 3130 + }, + { + "epoch": 0.2799535050071531, + "grad_norm": 0.14172200690581188, + "learning_rate": 0.00016899075059743007, + "loss": 0.647, + "step": 3131 + }, + { + "epoch": 0.2800429184549356, + "grad_norm": 0.1289873214802673, + "learning_rate": 0.00016896978355308352, + "loss": 0.6539, + "step": 3132 + }, + { + "epoch": 0.2801323319027182, + "grad_norm": 0.1231471533413054, + "learning_rate": 0.00016894881072432443, + "loss": 0.6793, + "step": 3133 + }, + { + "epoch": 0.2802217453505007, + "grad_norm": 0.13355764927271027, + "learning_rate": 0.00016892783211291194, + "loss": 0.6458, + "step": 3134 + }, + { + "epoch": 0.2803111587982833, + "grad_norm": 0.12564867790949769, + "learning_rate": 0.00016890684772060538, + "loss": 0.6503, + "step": 3135 + }, + { + "epoch": 0.2804005722460658, + "grad_norm": 0.12693974571953137, + "learning_rate": 0.00016888585754916476, + "loss": 0.6762, + "step": 3136 + }, + { + "epoch": 0.28048998569384836, + "grad_norm": 0.12959527461677692, + "learning_rate": 0.0001688648616003504, + "loss": 0.6866, + "step": 3137 + }, + { + "epoch": 0.2805793991416309, + "grad_norm": 1.5331530511887288, + "learning_rate": 0.0001688438598759233, + "loss": 0.5539, + "step": 3138 + }, + { + "epoch": 0.28066881258941345, + "grad_norm": 0.16462585038234076, + "learning_rate": 0.00016882285237764482, + "loss": 0.7042, + "step": 3139 + }, + { + "epoch": 0.28075822603719597, + "grad_norm": 0.13546013008028612, + "learning_rate": 0.0001688018391072768, + "loss": 0.6718, + "step": 3140 + }, + { + "epoch": 0.28084763948497854, + "grad_norm": 0.1832433516431572, + "learning_rate": 0.00016878082006658164, + "loss": 0.7193, + "step": 3141 + }, + { + "epoch": 0.2809370529327611, + "grad_norm": 0.15443084230155343, + "learning_rate": 0.00016875979525732214, + "loss": 0.6781, + "step": 3142 + }, + { + "epoch": 0.2810264663805436, + "grad_norm": 0.15663824278361876, + "learning_rate": 0.0001687387646812617, + "loss": 0.7254, + "step": 3143 + }, + { + "epoch": 0.2811158798283262, + "grad_norm": 0.15935239182237668, + "learning_rate": 0.00016871772834016406, + "loss": 0.7137, + "step": 3144 + }, + { + "epoch": 0.2812052932761087, + "grad_norm": 0.14574332442885343, + "learning_rate": 0.00016869668623579353, + "loss": 0.6488, + "step": 3145 + }, + { + "epoch": 0.2812947067238913, + "grad_norm": 0.1534667293244337, + "learning_rate": 0.00016867563836991492, + "loss": 0.6911, + "step": 3146 + }, + { + "epoch": 0.2813841201716738, + "grad_norm": 0.13996666892507928, + "learning_rate": 0.00016865458474429342, + "loss": 0.7067, + "step": 3147 + }, + { + "epoch": 0.2814735336194564, + "grad_norm": 0.14239761173974985, + "learning_rate": 0.00016863352536069482, + "loss": 0.7281, + "step": 3148 + }, + { + "epoch": 0.2815629470672389, + "grad_norm": 0.1403017603082169, + "learning_rate": 0.00016861246022088536, + "loss": 0.6741, + "step": 3149 + }, + { + "epoch": 0.28165236051502146, + "grad_norm": 0.13314831272641883, + "learning_rate": 0.0001685913893266317, + "loss": 0.6671, + "step": 3150 + }, + { + "epoch": 0.28174177396280403, + "grad_norm": 0.13367496395861628, + "learning_rate": 0.00016857031267970105, + "loss": 0.6825, + "step": 3151 + }, + { + "epoch": 0.28183118741058655, + "grad_norm": 0.13686136975950272, + "learning_rate": 0.00016854923028186111, + "loss": 0.6773, + "step": 3152 + }, + { + "epoch": 0.2819206008583691, + "grad_norm": 0.1776241464195531, + "learning_rate": 0.00016852814213488, + "loss": 0.722, + "step": 3153 + }, + { + "epoch": 0.28201001430615164, + "grad_norm": 0.15402201127179968, + "learning_rate": 0.00016850704824052635, + "loss": 0.6618, + "step": 3154 + }, + { + "epoch": 0.2820994277539342, + "grad_norm": 0.13730255333580807, + "learning_rate": 0.00016848594860056933, + "loss": 0.6466, + "step": 3155 + }, + { + "epoch": 0.2821888412017167, + "grad_norm": 0.15679093308390465, + "learning_rate": 0.00016846484321677852, + "loss": 0.7275, + "step": 3156 + }, + { + "epoch": 0.2822782546494993, + "grad_norm": 0.14652141596297183, + "learning_rate": 0.00016844373209092396, + "loss": 0.715, + "step": 3157 + }, + { + "epoch": 0.2823676680972818, + "grad_norm": 0.19776511599504776, + "learning_rate": 0.00016842261522477628, + "loss": 0.6848, + "step": 3158 + }, + { + "epoch": 0.2824570815450644, + "grad_norm": 0.1363385680795593, + "learning_rate": 0.00016840149262010648, + "loss": 0.6415, + "step": 3159 + }, + { + "epoch": 0.2825464949928469, + "grad_norm": 0.36374992491467534, + "learning_rate": 0.00016838036427868608, + "loss": 0.6848, + "step": 3160 + }, + { + "epoch": 0.28263590844062947, + "grad_norm": 0.1772284746811081, + "learning_rate": 0.00016835923020228712, + "loss": 0.721, + "step": 3161 + }, + { + "epoch": 0.28272532188841204, + "grad_norm": 2.4351226599275524, + "learning_rate": 0.0001683380903926821, + "loss": 0.7254, + "step": 3162 + }, + { + "epoch": 0.28281473533619456, + "grad_norm": 2.614824596493362, + "learning_rate": 0.00016831694485164398, + "loss": 0.7132, + "step": 3163 + }, + { + "epoch": 0.2829041487839771, + "grad_norm": 109.68370785980126, + "learning_rate": 0.00016829579358094616, + "loss": 0.8672, + "step": 3164 + }, + { + "epoch": 0.28299356223175964, + "grad_norm": 29.35739093871733, + "learning_rate": 0.00016827463658236264, + "loss": 1.2877, + "step": 3165 + }, + { + "epoch": 0.2830829756795422, + "grad_norm": 87.2164228972079, + "learning_rate": 0.0001682534738576678, + "loss": 2.2864, + "step": 3166 + }, + { + "epoch": 0.28317238912732473, + "grad_norm": 207.76613529543525, + "learning_rate": 0.00016823230540863654, + "loss": 1.204, + "step": 3167 + }, + { + "epoch": 0.2832618025751073, + "grad_norm": 180.59756077572592, + "learning_rate": 0.00016821113123704424, + "loss": 4.5641, + "step": 3168 + }, + { + "epoch": 0.2833512160228898, + "grad_norm": 35.48318123449796, + "learning_rate": 0.0001681899513446667, + "loss": 6.6929, + "step": 3169 + }, + { + "epoch": 0.2834406294706724, + "grad_norm": 32.849454082553436, + "learning_rate": 0.00016816876573328037, + "loss": 6.1566, + "step": 3170 + }, + { + "epoch": 0.28353004291845496, + "grad_norm": 5.151046440010461, + "learning_rate": 0.00016814757440466188, + "loss": 5.1408, + "step": 3171 + }, + { + "epoch": 0.2836194563662375, + "grad_norm": 9.235347061888291, + "learning_rate": 0.0001681263773605887, + "loss": 5.1482, + "step": 3172 + }, + { + "epoch": 0.28370886981402005, + "grad_norm": 10.317902021482372, + "learning_rate": 0.00016810517460283853, + "loss": 5.3977, + "step": 3173 + }, + { + "epoch": 0.28379828326180256, + "grad_norm": 13.927307798007565, + "learning_rate": 0.0001680839661331896, + "loss": 8.5175, + "step": 3174 + }, + { + "epoch": 0.28388769670958514, + "grad_norm": 12.4541704976026, + "learning_rate": 0.00016806275195342064, + "loss": 6.7214, + "step": 3175 + }, + { + "epoch": 0.28397711015736765, + "grad_norm": 5.887704042341981, + "learning_rate": 0.00016804153206531088, + "loss": 6.0616, + "step": 3176 + }, + { + "epoch": 0.2840665236051502, + "grad_norm": 6.559935356842627, + "learning_rate": 0.00016802030647064, + "loss": 5.2376, + "step": 3177 + }, + { + "epoch": 0.28415593705293274, + "grad_norm": 46.45485504538953, + "learning_rate": 0.00016799907517118818, + "loss": 7.0084, + "step": 3178 + }, + { + "epoch": 0.2842453505007153, + "grad_norm": 2.415634830838911, + "learning_rate": 0.00016797783816873603, + "loss": 8.3, + "step": 3179 + }, + { + "epoch": 0.2843347639484979, + "grad_norm": 11.808700353837585, + "learning_rate": 0.00016795659546506468, + "loss": 7.1274, + "step": 3180 + }, + { + "epoch": 0.2844241773962804, + "grad_norm": 4.0730642780949085, + "learning_rate": 0.00016793534706195575, + "loss": 4.9784, + "step": 3181 + }, + { + "epoch": 0.28451359084406297, + "grad_norm": 6.462691250820877, + "learning_rate": 0.0001679140929611913, + "loss": 4.9897, + "step": 3182 + }, + { + "epoch": 0.2846030042918455, + "grad_norm": 3.7010351361980987, + "learning_rate": 0.00016789283316455392, + "loss": 5.2373, + "step": 3183 + }, + { + "epoch": 0.28469241773962806, + "grad_norm": 3.638868737263044, + "learning_rate": 0.00016787156767382659, + "loss": 4.9326, + "step": 3184 + }, + { + "epoch": 0.2847818311874106, + "grad_norm": 3.2518693161758416, + "learning_rate": 0.00016785029649079287, + "loss": 4.8424, + "step": 3185 + }, + { + "epoch": 0.28487124463519314, + "grad_norm": 3.9058836635015233, + "learning_rate": 0.0001678290196172367, + "loss": 4.216, + "step": 3186 + }, + { + "epoch": 0.28496065808297566, + "grad_norm": 7.097924924912819, + "learning_rate": 0.0001678077370549426, + "loss": 5.4077, + "step": 3187 + }, + { + "epoch": 0.28505007153075823, + "grad_norm": 1.8634778211981005, + "learning_rate": 0.00016778644880569544, + "loss": 4.4358, + "step": 3188 + }, + { + "epoch": 0.28513948497854075, + "grad_norm": 6.496229410476049, + "learning_rate": 0.00016776515487128073, + "loss": 4.195, + "step": 3189 + }, + { + "epoch": 0.2852288984263233, + "grad_norm": 10.985428036114804, + "learning_rate": 0.00016774385525348428, + "loss": 4.9139, + "step": 3190 + }, + { + "epoch": 0.2853183118741059, + "grad_norm": 4.317511126326712, + "learning_rate": 0.00016772254995409255, + "loss": 4.5143, + "step": 3191 + }, + { + "epoch": 0.2854077253218884, + "grad_norm": 6.378455001594766, + "learning_rate": 0.00016770123897489228, + "loss": 4.4212, + "step": 3192 + }, + { + "epoch": 0.285497138769671, + "grad_norm": 6.453803665858282, + "learning_rate": 0.00016767992231767092, + "loss": 4.0856, + "step": 3193 + }, + { + "epoch": 0.2855865522174535, + "grad_norm": 4.960704229800028, + "learning_rate": 0.0001676585999842162, + "loss": 4.215, + "step": 3194 + }, + { + "epoch": 0.28567596566523606, + "grad_norm": 6.018432404001403, + "learning_rate": 0.0001676372719763164, + "loss": 4.0665, + "step": 3195 + }, + { + "epoch": 0.2857653791130186, + "grad_norm": 4.470799224053419, + "learning_rate": 0.0001676159382957603, + "loss": 4.6952, + "step": 3196 + }, + { + "epoch": 0.28585479256080115, + "grad_norm": 7.124103715965428, + "learning_rate": 0.0001675945989443371, + "loss": 4.2989, + "step": 3197 + }, + { + "epoch": 0.28594420600858367, + "grad_norm": 4.421703037458057, + "learning_rate": 0.0001675732539238365, + "loss": 4.4658, + "step": 3198 + }, + { + "epoch": 0.28603361945636624, + "grad_norm": 14.180540552842688, + "learning_rate": 0.00016755190323604872, + "loss": 3.8118, + "step": 3199 + }, + { + "epoch": 0.2861230329041488, + "grad_norm": 3.8915157128105946, + "learning_rate": 0.0001675305468827644, + "loss": 4.6976, + "step": 3200 + }, + { + "epoch": 0.2862124463519313, + "grad_norm": 10.5426965530204, + "learning_rate": 0.00016750918486577466, + "loss": 4.2535, + "step": 3201 + }, + { + "epoch": 0.2863018597997139, + "grad_norm": 2.8725676618041183, + "learning_rate": 0.00016748781718687111, + "loss": 3.9635, + "step": 3202 + }, + { + "epoch": 0.2863912732474964, + "grad_norm": 10.679724056550565, + "learning_rate": 0.00016746644384784586, + "loss": 4.6017, + "step": 3203 + }, + { + "epoch": 0.286480686695279, + "grad_norm": 3.2575396550080744, + "learning_rate": 0.00016744506485049144, + "loss": 8.6004, + "step": 3204 + }, + { + "epoch": 0.2865701001430615, + "grad_norm": 5.0146942224117685, + "learning_rate": 0.00016742368019660088, + "loss": 4.0338, + "step": 3205 + }, + { + "epoch": 0.2866595135908441, + "grad_norm": 4.015886204927873, + "learning_rate": 0.0001674022898879677, + "loss": 4.3393, + "step": 3206 + }, + { + "epoch": 0.2867489270386266, + "grad_norm": 2.1108289208421738, + "learning_rate": 0.00016738089392638586, + "loss": 3.7905, + "step": 3207 + }, + { + "epoch": 0.28683834048640916, + "grad_norm": 1.350897285408448, + "learning_rate": 0.0001673594923136498, + "loss": 3.9644, + "step": 3208 + }, + { + "epoch": 0.2869277539341917, + "grad_norm": 5.274073662759056, + "learning_rate": 0.00016733808505155448, + "loss": 4.7034, + "step": 3209 + }, + { + "epoch": 0.28701716738197425, + "grad_norm": 2.45164482811205, + "learning_rate": 0.0001673166721418953, + "loss": 3.8922, + "step": 3210 + }, + { + "epoch": 0.2871065808297568, + "grad_norm": 4.2409980159267135, + "learning_rate": 0.00016729525358646813, + "loss": 4.3398, + "step": 3211 + }, + { + "epoch": 0.28719599427753933, + "grad_norm": 10.99844151093623, + "learning_rate": 0.00016727382938706931, + "loss": 4.2308, + "step": 3212 + }, + { + "epoch": 0.2872854077253219, + "grad_norm": 1.806302160886743, + "learning_rate": 0.00016725239954549565, + "loss": 3.7958, + "step": 3213 + }, + { + "epoch": 0.2873748211731044, + "grad_norm": 1.3446774380805993, + "learning_rate": 0.00016723096406354447, + "loss": 4.0321, + "step": 3214 + }, + { + "epoch": 0.287464234620887, + "grad_norm": 3.7747025306950266, + "learning_rate": 0.00016720952294301355, + "loss": 4.0829, + "step": 3215 + }, + { + "epoch": 0.2875536480686695, + "grad_norm": 4.050413975679484, + "learning_rate": 0.00016718807618570106, + "loss": 3.9456, + "step": 3216 + }, + { + "epoch": 0.2876430615164521, + "grad_norm": 5.601574069740354, + "learning_rate": 0.0001671666237934058, + "loss": 4.2919, + "step": 3217 + }, + { + "epoch": 0.2877324749642346, + "grad_norm": 3.1073481836517947, + "learning_rate": 0.00016714516576792692, + "loss": 3.5554, + "step": 3218 + }, + { + "epoch": 0.28782188841201717, + "grad_norm": 2.7548122974009015, + "learning_rate": 0.00016712370211106406, + "loss": 4.3728, + "step": 3219 + }, + { + "epoch": 0.28791130185979974, + "grad_norm": 1.382035648858085, + "learning_rate": 0.0001671022328246174, + "loss": 4.3227, + "step": 3220 + }, + { + "epoch": 0.28800071530758226, + "grad_norm": 2.635587021907402, + "learning_rate": 0.00016708075791038745, + "loss": 3.7153, + "step": 3221 + }, + { + "epoch": 0.2880901287553648, + "grad_norm": 5.26217432694525, + "learning_rate": 0.00016705927737017544, + "loss": 3.9522, + "step": 3222 + }, + { + "epoch": 0.28817954220314734, + "grad_norm": 2.0192229198442764, + "learning_rate": 0.00016703779120578273, + "loss": 3.8813, + "step": 3223 + }, + { + "epoch": 0.2882689556509299, + "grad_norm": 6.301807895901321, + "learning_rate": 0.00016701629941901148, + "loss": 3.5374, + "step": 3224 + }, + { + "epoch": 0.28835836909871243, + "grad_norm": 8.973586046364293, + "learning_rate": 0.00016699480201166415, + "loss": 4.248, + "step": 3225 + }, + { + "epoch": 0.288447782546495, + "grad_norm": 3.2076139480865082, + "learning_rate": 0.00016697329898554365, + "loss": 3.6978, + "step": 3226 + }, + { + "epoch": 0.2885371959942775, + "grad_norm": 9.327864216335328, + "learning_rate": 0.00016695179034245346, + "loss": 4.2565, + "step": 3227 + }, + { + "epoch": 0.2886266094420601, + "grad_norm": 5.246376067166037, + "learning_rate": 0.00016693027608419747, + "loss": 4.1789, + "step": 3228 + }, + { + "epoch": 0.2887160228898426, + "grad_norm": 7.564462484738335, + "learning_rate": 0.00016690875621258006, + "loss": 4.0468, + "step": 3229 + }, + { + "epoch": 0.2888054363376252, + "grad_norm": 2.6362485776709925, + "learning_rate": 0.00016688723072940607, + "loss": 3.9839, + "step": 3230 + }, + { + "epoch": 0.28889484978540775, + "grad_norm": 8.963586766698459, + "learning_rate": 0.0001668656996364808, + "loss": 3.856, + "step": 3231 + }, + { + "epoch": 0.28898426323319026, + "grad_norm": 5.743271594841896, + "learning_rate": 0.0001668441629356101, + "loss": 3.8613, + "step": 3232 + }, + { + "epoch": 0.28907367668097284, + "grad_norm": 5.819192331109701, + "learning_rate": 0.00016682262062860014, + "loss": 3.3976, + "step": 3233 + }, + { + "epoch": 0.28916309012875535, + "grad_norm": 2.2685486027911903, + "learning_rate": 0.0001668010727172577, + "loss": 3.8783, + "step": 3234 + }, + { + "epoch": 0.2892525035765379, + "grad_norm": 4.573280297309724, + "learning_rate": 0.00016677951920338995, + "loss": 4.2271, + "step": 3235 + }, + { + "epoch": 0.28934191702432044, + "grad_norm": 7.563417220617187, + "learning_rate": 0.00016675796008880462, + "loss": 3.6916, + "step": 3236 + }, + { + "epoch": 0.289431330472103, + "grad_norm": 4.465865909185956, + "learning_rate": 0.00016673639537530976, + "loss": 3.3656, + "step": 3237 + }, + { + "epoch": 0.2895207439198855, + "grad_norm": 5.223181267008394, + "learning_rate": 0.00016671482506471402, + "loss": 3.7847, + "step": 3238 + }, + { + "epoch": 0.2896101573676681, + "grad_norm": 2.41025262545117, + "learning_rate": 0.0001666932491588265, + "loss": 3.1398, + "step": 3239 + }, + { + "epoch": 0.28969957081545067, + "grad_norm": 4.419790316227284, + "learning_rate": 0.00016667166765945668, + "loss": 3.2258, + "step": 3240 + }, + { + "epoch": 0.2897889842632332, + "grad_norm": 7.509490427616291, + "learning_rate": 0.00016665008056841466, + "loss": 3.258, + "step": 3241 + }, + { + "epoch": 0.28987839771101576, + "grad_norm": 3.22072974992524, + "learning_rate": 0.00016662848788751085, + "loss": 3.4594, + "step": 3242 + }, + { + "epoch": 0.28996781115879827, + "grad_norm": 17.676169503329938, + "learning_rate": 0.00016660688961855623, + "loss": 3.0872, + "step": 3243 + }, + { + "epoch": 0.29005722460658084, + "grad_norm": 31.768354680665166, + "learning_rate": 0.0001665852857633622, + "loss": 3.6153, + "step": 3244 + }, + { + "epoch": 0.29014663805436336, + "grad_norm": 10.035079364505094, + "learning_rate": 0.0001665636763237407, + "loss": 3.4406, + "step": 3245 + }, + { + "epoch": 0.29023605150214593, + "grad_norm": 13.703201349918155, + "learning_rate": 0.00016654206130150404, + "loss": 3.7742, + "step": 3246 + }, + { + "epoch": 0.29032546494992845, + "grad_norm": 5.32606490113053, + "learning_rate": 0.00016652044069846505, + "loss": 3.0207, + "step": 3247 + }, + { + "epoch": 0.290414878397711, + "grad_norm": 9.22708590907137, + "learning_rate": 0.00016649881451643705, + "loss": 2.8273, + "step": 3248 + }, + { + "epoch": 0.2905042918454936, + "grad_norm": 22.617424785024088, + "learning_rate": 0.0001664771827572338, + "loss": 3.3399, + "step": 3249 + }, + { + "epoch": 0.2905937052932761, + "grad_norm": 5.985496167035877, + "learning_rate": 0.0001664555454226695, + "loss": 3.4459, + "step": 3250 + }, + { + "epoch": 0.2906831187410587, + "grad_norm": 20.838301564674932, + "learning_rate": 0.00016643390251455884, + "loss": 2.6787, + "step": 3251 + }, + { + "epoch": 0.2907725321888412, + "grad_norm": 58.40398614888195, + "learning_rate": 0.00016641225403471701, + "loss": 5.1037, + "step": 3252 + }, + { + "epoch": 0.29086194563662376, + "grad_norm": 32.73418934604156, + "learning_rate": 0.00016639059998495968, + "loss": 3.7201, + "step": 3253 + }, + { + "epoch": 0.2909513590844063, + "grad_norm": 7.918587831003649, + "learning_rate": 0.00016636894036710286, + "loss": 4.1403, + "step": 3254 + }, + { + "epoch": 0.29104077253218885, + "grad_norm": 7.190108031473884, + "learning_rate": 0.0001663472751829632, + "loss": 3.9778, + "step": 3255 + }, + { + "epoch": 0.29113018597997137, + "grad_norm": 6.276891915691289, + "learning_rate": 0.0001663256044343577, + "loss": 4.1515, + "step": 3256 + }, + { + "epoch": 0.29121959942775394, + "grad_norm": 5.62464690744072, + "learning_rate": 0.00016630392812310384, + "loss": 3.145, + "step": 3257 + }, + { + "epoch": 0.29130901287553645, + "grad_norm": 5.654504064065954, + "learning_rate": 0.00016628224625101962, + "loss": 3.1494, + "step": 3258 + }, + { + "epoch": 0.291398426323319, + "grad_norm": 4.775276195214736, + "learning_rate": 0.00016626055881992344, + "loss": 3.6355, + "step": 3259 + }, + { + "epoch": 0.2914878397711016, + "grad_norm": 4.1927550154940905, + "learning_rate": 0.00016623886583163423, + "loss": 3.1856, + "step": 3260 + }, + { + "epoch": 0.2915772532188841, + "grad_norm": 10.460710403348065, + "learning_rate": 0.00016621716728797132, + "loss": 3.1766, + "step": 3261 + }, + { + "epoch": 0.2916666666666667, + "grad_norm": 4.93418365121729, + "learning_rate": 0.00016619546319075455, + "loss": 3.5236, + "step": 3262 + }, + { + "epoch": 0.2917560801144492, + "grad_norm": 6.118584449806641, + "learning_rate": 0.00016617375354180424, + "loss": 3.2093, + "step": 3263 + }, + { + "epoch": 0.2918454935622318, + "grad_norm": 15.388490379561576, + "learning_rate": 0.00016615203834294119, + "loss": 3.2324, + "step": 3264 + }, + { + "epoch": 0.2919349070100143, + "grad_norm": 0.9945376598482305, + "learning_rate": 0.0001661303175959865, + "loss": 2.9528, + "step": 3265 + }, + { + "epoch": 0.29202432045779686, + "grad_norm": 3.5068832474090756, + "learning_rate": 0.00016610859130276198, + "loss": 2.3786, + "step": 3266 + }, + { + "epoch": 0.2921137339055794, + "grad_norm": 1.9972887066455196, + "learning_rate": 0.00016608685946508972, + "loss": 2.385, + "step": 3267 + }, + { + "epoch": 0.29220314735336195, + "grad_norm": 9.825383433790877, + "learning_rate": 0.00016606512208479238, + "loss": 8.5167, + "step": 3268 + }, + { + "epoch": 0.2922925608011445, + "grad_norm": 7.95812706123854, + "learning_rate": 0.00016604337916369306, + "loss": 2.9636, + "step": 3269 + }, + { + "epoch": 0.29238197424892703, + "grad_norm": 6.315744521003476, + "learning_rate": 0.00016602163070361526, + "loss": 2.6699, + "step": 3270 + }, + { + "epoch": 0.2924713876967096, + "grad_norm": 4.18588287170997, + "learning_rate": 0.00016599987670638304, + "loss": 2.7516, + "step": 3271 + }, + { + "epoch": 0.2925608011444921, + "grad_norm": 8.105978640891665, + "learning_rate": 0.00016597811717382083, + "loss": 3.3026, + "step": 3272 + }, + { + "epoch": 0.2926502145922747, + "grad_norm": 8.71382884347448, + "learning_rate": 0.00016595635210775366, + "loss": 8.7628, + "step": 3273 + }, + { + "epoch": 0.2927396280400572, + "grad_norm": 5.706434427372548, + "learning_rate": 0.00016593458151000688, + "loss": 3.0653, + "step": 3274 + }, + { + "epoch": 0.2928290414878398, + "grad_norm": 8.873054548614077, + "learning_rate": 0.0001659128053824064, + "loss": 3.2467, + "step": 3275 + }, + { + "epoch": 0.2929184549356223, + "grad_norm": 2.90482695495245, + "learning_rate": 0.0001658910237267785, + "loss": 2.5935, + "step": 3276 + }, + { + "epoch": 0.29300786838340487, + "grad_norm": 3.601572141477614, + "learning_rate": 0.00016586923654495004, + "loss": 2.5304, + "step": 3277 + }, + { + "epoch": 0.2930972818311874, + "grad_norm": 6.006325033287063, + "learning_rate": 0.00016584744383874825, + "loss": 2.9035, + "step": 3278 + }, + { + "epoch": 0.29318669527896996, + "grad_norm": 3.5705874923207475, + "learning_rate": 0.00016582564561000088, + "loss": 2.7931, + "step": 3279 + }, + { + "epoch": 0.2932761087267525, + "grad_norm": 2.5981284983196833, + "learning_rate": 0.0001658038418605361, + "loss": 2.8167, + "step": 3280 + }, + { + "epoch": 0.29336552217453504, + "grad_norm": 5.663096277880741, + "learning_rate": 0.00016578203259218257, + "loss": 2.6177, + "step": 3281 + }, + { + "epoch": 0.2934549356223176, + "grad_norm": 4.100924467434517, + "learning_rate": 0.00016576021780676943, + "loss": 2.7843, + "step": 3282 + }, + { + "epoch": 0.29354434907010013, + "grad_norm": 1.4887516698186272, + "learning_rate": 0.00016573839750612623, + "loss": 2.8861, + "step": 3283 + }, + { + "epoch": 0.2936337625178827, + "grad_norm": 1.5973686009018366, + "learning_rate": 0.00016571657169208302, + "loss": 2.5327, + "step": 3284 + }, + { + "epoch": 0.2937231759656652, + "grad_norm": 2.0296153486778907, + "learning_rate": 0.00016569474036647028, + "loss": 2.4028, + "step": 3285 + }, + { + "epoch": 0.2938125894134478, + "grad_norm": 2.1152928961671704, + "learning_rate": 0.00016567290353111905, + "loss": 2.8299, + "step": 3286 + }, + { + "epoch": 0.2939020028612303, + "grad_norm": 4.050338947464087, + "learning_rate": 0.0001656510611878607, + "loss": 8.1436, + "step": 3287 + }, + { + "epoch": 0.2939914163090129, + "grad_norm": 2.9173165588357026, + "learning_rate": 0.00016562921333852714, + "loss": 3.1023, + "step": 3288 + }, + { + "epoch": 0.29408082975679545, + "grad_norm": 1.8092424946571801, + "learning_rate": 0.00016560735998495066, + "loss": 1.8331, + "step": 3289 + }, + { + "epoch": 0.29417024320457796, + "grad_norm": 0.7069139130056243, + "learning_rate": 0.0001655855011289642, + "loss": 2.6552, + "step": 3290 + }, + { + "epoch": 0.29425965665236054, + "grad_norm": 0.9118281330648345, + "learning_rate": 0.00016556363677240098, + "loss": 2.9961, + "step": 3291 + }, + { + "epoch": 0.29434907010014305, + "grad_norm": 1.2248105586804225, + "learning_rate": 0.00016554176691709467, + "loss": 2.6656, + "step": 3292 + }, + { + "epoch": 0.2944384835479256, + "grad_norm": 1.0583447113035311, + "learning_rate": 0.00016551989156487955, + "loss": 2.7415, + "step": 3293 + }, + { + "epoch": 0.29452789699570814, + "grad_norm": 1.161297841499713, + "learning_rate": 0.00016549801071759026, + "loss": 2.2794, + "step": 3294 + }, + { + "epoch": 0.2946173104434907, + "grad_norm": 0.4513871748885197, + "learning_rate": 0.00016547612437706189, + "loss": 2.1242, + "step": 3295 + }, + { + "epoch": 0.2947067238912732, + "grad_norm": 0.7044649982042712, + "learning_rate": 0.00016545423254513004, + "loss": 2.2613, + "step": 3296 + }, + { + "epoch": 0.2947961373390558, + "grad_norm": 1.058346184442964, + "learning_rate": 0.00016543233522363078, + "loss": 2.465, + "step": 3297 + }, + { + "epoch": 0.2948855507868383, + "grad_norm": 0.989868473132693, + "learning_rate": 0.00016541043241440057, + "loss": 2.9383, + "step": 3298 + }, + { + "epoch": 0.2949749642346209, + "grad_norm": 1.0136061465006916, + "learning_rate": 0.0001653885241192764, + "loss": 2.4374, + "step": 3299 + }, + { + "epoch": 0.29506437768240346, + "grad_norm": 0.6459533859125532, + "learning_rate": 0.00016536661034009567, + "loss": 2.4376, + "step": 3300 + }, + { + "epoch": 0.29515379113018597, + "grad_norm": 0.7746542702634331, + "learning_rate": 0.00016534469107869627, + "loss": 1.9412, + "step": 3301 + }, + { + "epoch": 0.29524320457796854, + "grad_norm": 1.4691960301050855, + "learning_rate": 0.00016532276633691656, + "loss": 2.9388, + "step": 3302 + }, + { + "epoch": 0.29533261802575106, + "grad_norm": 0.7778231283883283, + "learning_rate": 0.00016530083611659532, + "loss": 3.2579, + "step": 3303 + }, + { + "epoch": 0.29542203147353363, + "grad_norm": 1.76710950722173, + "learning_rate": 0.00016527890041957184, + "loss": 2.925, + "step": 3304 + }, + { + "epoch": 0.29551144492131615, + "grad_norm": 0.37436408222820977, + "learning_rate": 0.0001652569592476858, + "loss": 2.8484, + "step": 3305 + }, + { + "epoch": 0.2956008583690987, + "grad_norm": 1.1361311247662347, + "learning_rate": 0.0001652350126027774, + "loss": 2.5122, + "step": 3306 + }, + { + "epoch": 0.29569027181688123, + "grad_norm": 3.8807579804202224, + "learning_rate": 0.00016521306048668727, + "loss": 8.2657, + "step": 3307 + }, + { + "epoch": 0.2957796852646638, + "grad_norm": 0.6789597138616542, + "learning_rate": 0.00016519110290125652, + "loss": 1.9013, + "step": 3308 + }, + { + "epoch": 0.2958690987124464, + "grad_norm": 1.7397294264398404, + "learning_rate": 0.0001651691398483267, + "loss": 2.4978, + "step": 3309 + }, + { + "epoch": 0.2959585121602289, + "grad_norm": 1.7287508606163668, + "learning_rate": 0.00016514717132973982, + "loss": 2.2487, + "step": 3310 + }, + { + "epoch": 0.29604792560801146, + "grad_norm": 0.684890120748853, + "learning_rate": 0.00016512519734733836, + "loss": 2.7394, + "step": 3311 + }, + { + "epoch": 0.296137339055794, + "grad_norm": 0.6018278390854357, + "learning_rate": 0.00016510321790296525, + "loss": 2.4226, + "step": 3312 + }, + { + "epoch": 0.29622675250357655, + "grad_norm": 4.544869188428613, + "learning_rate": 0.0001650812329984639, + "loss": 8.1234, + "step": 3313 + }, + { + "epoch": 0.29631616595135907, + "grad_norm": 2.3969699861942217, + "learning_rate": 0.0001650592426356781, + "loss": 2.6414, + "step": 3314 + }, + { + "epoch": 0.29640557939914164, + "grad_norm": 1.584073450763068, + "learning_rate": 0.00016503724681645222, + "loss": 2.5424, + "step": 3315 + }, + { + "epoch": 0.29649499284692415, + "grad_norm": 1.4759641441100761, + "learning_rate": 0.000165015245542631, + "loss": 2.4051, + "step": 3316 + }, + { + "epoch": 0.2965844062947067, + "grad_norm": 1.0028332784675207, + "learning_rate": 0.00016499323881605964, + "loss": 2.8699, + "step": 3317 + }, + { + "epoch": 0.2966738197424893, + "grad_norm": 1.5465750567224135, + "learning_rate": 0.00016497122663858385, + "loss": 2.5306, + "step": 3318 + }, + { + "epoch": 0.2967632331902718, + "grad_norm": 1.399305348456039, + "learning_rate": 0.0001649492090120497, + "loss": 2.705, + "step": 3319 + }, + { + "epoch": 0.2968526466380544, + "grad_norm": 0.8234437079814507, + "learning_rate": 0.00016492718593830389, + "loss": 3.4965, + "step": 3320 + }, + { + "epoch": 0.2969420600858369, + "grad_norm": 1.3700671406487597, + "learning_rate": 0.00016490515741919334, + "loss": 2.8917, + "step": 3321 + }, + { + "epoch": 0.2970314735336195, + "grad_norm": 0.8165046931183331, + "learning_rate": 0.00016488312345656566, + "loss": 2.5722, + "step": 3322 + }, + { + "epoch": 0.297120886981402, + "grad_norm": 1.216577267655659, + "learning_rate": 0.0001648610840522688, + "loss": 2.4569, + "step": 3323 + }, + { + "epoch": 0.29721030042918456, + "grad_norm": 0.2432724666497758, + "learning_rate": 0.00016483903920815111, + "loss": 2.431, + "step": 3324 + }, + { + "epoch": 0.2972997138769671, + "grad_norm": 0.8368748903325377, + "learning_rate": 0.0001648169889260615, + "loss": 2.7847, + "step": 3325 + }, + { + "epoch": 0.29738912732474965, + "grad_norm": 0.9594928110802587, + "learning_rate": 0.00016479493320784938, + "loss": 7.8701, + "step": 3326 + }, + { + "epoch": 0.29747854077253216, + "grad_norm": 0.4334188166795502, + "learning_rate": 0.0001647728720553644, + "loss": 2.1084, + "step": 3327 + }, + { + "epoch": 0.29756795422031473, + "grad_norm": 0.27578466769029963, + "learning_rate": 0.00016475080547045687, + "loss": 2.6651, + "step": 3328 + }, + { + "epoch": 0.2976573676680973, + "grad_norm": 1.1717515568176564, + "learning_rate": 0.0001647287334549775, + "loss": 2.9626, + "step": 3329 + }, + { + "epoch": 0.2977467811158798, + "grad_norm": 0.662095906593713, + "learning_rate": 0.00016470665601077742, + "loss": 2.2709, + "step": 3330 + }, + { + "epoch": 0.2978361945636624, + "grad_norm": 1.959328681086918, + "learning_rate": 0.00016468457313970826, + "loss": 7.8767, + "step": 3331 + }, + { + "epoch": 0.2979256080114449, + "grad_norm": 2.891330036271476, + "learning_rate": 0.00016466248484362208, + "loss": 2.1316, + "step": 3332 + }, + { + "epoch": 0.2980150214592275, + "grad_norm": 0.5391182983734274, + "learning_rate": 0.00016464039112437138, + "loss": 2.1015, + "step": 3333 + }, + { + "epoch": 0.29810443490701, + "grad_norm": 0.7709134581671627, + "learning_rate": 0.00016461829198380912, + "loss": 2.3315, + "step": 3334 + }, + { + "epoch": 0.29819384835479257, + "grad_norm": 1.4912417826283253, + "learning_rate": 0.00016459618742378876, + "loss": 2.9617, + "step": 3335 + }, + { + "epoch": 0.2982832618025751, + "grad_norm": 0.327913760500544, + "learning_rate": 0.0001645740774461642, + "loss": 2.1557, + "step": 3336 + }, + { + "epoch": 0.29837267525035766, + "grad_norm": 0.7743710683028279, + "learning_rate": 0.00016455196205278968, + "loss": 2.5525, + "step": 3337 + }, + { + "epoch": 0.2984620886981402, + "grad_norm": 0.9512828462947668, + "learning_rate": 0.0001645298412455201, + "loss": 2.6712, + "step": 3338 + }, + { + "epoch": 0.29855150214592274, + "grad_norm": 0.3571354361087924, + "learning_rate": 0.0001645077150262107, + "loss": 2.8859, + "step": 3339 + }, + { + "epoch": 0.2986409155937053, + "grad_norm": 0.6050537334892501, + "learning_rate": 0.00016448558339671713, + "loss": 2.2212, + "step": 3340 + }, + { + "epoch": 0.29873032904148783, + "grad_norm": 0.8287789849103586, + "learning_rate": 0.00016446344635889554, + "loss": 2.7476, + "step": 3341 + }, + { + "epoch": 0.2988197424892704, + "grad_norm": 0.8907503586053893, + "learning_rate": 0.00016444130391460258, + "loss": 2.7224, + "step": 3342 + }, + { + "epoch": 0.2989091559370529, + "grad_norm": 0.6046166628484061, + "learning_rate": 0.00016441915606569526, + "loss": 2.5178, + "step": 3343 + }, + { + "epoch": 0.2989985693848355, + "grad_norm": 1.1565541094903726, + "learning_rate": 0.00016439700281403114, + "loss": 2.2929, + "step": 3344 + }, + { + "epoch": 0.299087982832618, + "grad_norm": 1.0364891524318145, + "learning_rate": 0.00016437484416146817, + "loss": 2.0602, + "step": 3345 + }, + { + "epoch": 0.2991773962804006, + "grad_norm": 0.5629705452325882, + "learning_rate": 0.00016435268010986476, + "loss": 2.8691, + "step": 3346 + }, + { + "epoch": 0.2992668097281831, + "grad_norm": 0.9326036306715856, + "learning_rate": 0.0001643305106610798, + "loss": 2.8963, + "step": 3347 + }, + { + "epoch": 0.29935622317596566, + "grad_norm": 0.4054934501226124, + "learning_rate": 0.00016430833581697254, + "loss": 2.4643, + "step": 3348 + }, + { + "epoch": 0.29944563662374823, + "grad_norm": 0.5770871797817814, + "learning_rate": 0.00016428615557940288, + "loss": 2.9348, + "step": 3349 + }, + { + "epoch": 0.29953505007153075, + "grad_norm": 0.42737345309128794, + "learning_rate": 0.000164263969950231, + "loss": 2.9523, + "step": 3350 + }, + { + "epoch": 0.2996244635193133, + "grad_norm": 1.1318340323979967, + "learning_rate": 0.0001642417789313175, + "loss": 2.3074, + "step": 3351 + }, + { + "epoch": 0.29971387696709584, + "grad_norm": 0.5549930347572761, + "learning_rate": 0.00016421958252452363, + "loss": 2.2997, + "step": 3352 + }, + { + "epoch": 0.2998032904148784, + "grad_norm": 0.30580413585100524, + "learning_rate": 0.00016419738073171093, + "loss": 2.7531, + "step": 3353 + }, + { + "epoch": 0.2998927038626609, + "grad_norm": 0.4992680128840708, + "learning_rate": 0.00016417517355474145, + "loss": 2.5534, + "step": 3354 + }, + { + "epoch": 0.2999821173104435, + "grad_norm": 0.8555315506655801, + "learning_rate": 0.00016415296099547765, + "loss": 2.3889, + "step": 3355 + }, + { + "epoch": 0.300071530758226, + "grad_norm": 0.5374833307714577, + "learning_rate": 0.0001641307430557825, + "loss": 2.7745, + "step": 3356 + }, + { + "epoch": 0.3001609442060086, + "grad_norm": 0.7588462586492398, + "learning_rate": 0.0001641085197375194, + "loss": 2.6868, + "step": 3357 + }, + { + "epoch": 0.30025035765379116, + "grad_norm": 1.0320791910509899, + "learning_rate": 0.00016408629104255212, + "loss": 2.8147, + "step": 3358 + }, + { + "epoch": 0.30033977110157367, + "grad_norm": 0.36706957144832614, + "learning_rate": 0.00016406405697274505, + "loss": 2.3805, + "step": 3359 + }, + { + "epoch": 0.30042918454935624, + "grad_norm": 0.8942866793175855, + "learning_rate": 0.00016404181752996289, + "loss": 2.0291, + "step": 3360 + }, + { + "epoch": 0.30051859799713876, + "grad_norm": 1.0108470681285882, + "learning_rate": 0.00016401957271607083, + "loss": 2.5097, + "step": 3361 + }, + { + "epoch": 0.30060801144492133, + "grad_norm": 0.9297984421575333, + "learning_rate": 0.0001639973225329345, + "loss": 2.6644, + "step": 3362 + }, + { + "epoch": 0.30069742489270385, + "grad_norm": 2.2368210906099346, + "learning_rate": 0.00016397506698242003, + "loss": 2.6212, + "step": 3363 + }, + { + "epoch": 0.3007868383404864, + "grad_norm": 1.14752691146757, + "learning_rate": 0.00016395280606639395, + "loss": 2.2551, + "step": 3364 + }, + { + "epoch": 0.30087625178826893, + "grad_norm": 0.8153144130110394, + "learning_rate": 0.00016393053978672328, + "loss": 2.0083, + "step": 3365 + }, + { + "epoch": 0.3009656652360515, + "grad_norm": 2.117097394372742, + "learning_rate": 0.00016390826814527545, + "loss": 2.8788, + "step": 3366 + }, + { + "epoch": 0.301055078683834, + "grad_norm": 0.982926618061747, + "learning_rate": 0.00016388599114391833, + "loss": 2.198, + "step": 3367 + }, + { + "epoch": 0.3011444921316166, + "grad_norm": 0.9333679756671526, + "learning_rate": 0.0001638637087845203, + "loss": 2.0636, + "step": 3368 + }, + { + "epoch": 0.30123390557939916, + "grad_norm": 2.7870472323589466, + "learning_rate": 0.00016384142106895015, + "loss": 3.0071, + "step": 3369 + }, + { + "epoch": 0.3013233190271817, + "grad_norm": 1.4091452958643897, + "learning_rate": 0.0001638191279990771, + "loss": 2.2766, + "step": 3370 + }, + { + "epoch": 0.30141273247496425, + "grad_norm": 0.9473391349591229, + "learning_rate": 0.00016379682957677087, + "loss": 2.4719, + "step": 3371 + }, + { + "epoch": 0.30150214592274677, + "grad_norm": 2.61825479727625, + "learning_rate": 0.00016377452580390158, + "loss": 2.372, + "step": 3372 + }, + { + "epoch": 0.30159155937052934, + "grad_norm": 1.5440488761415607, + "learning_rate": 0.00016375221668233985, + "loss": 2.3961, + "step": 3373 + }, + { + "epoch": 0.30168097281831185, + "grad_norm": 0.9834157801341683, + "learning_rate": 0.00016372990221395666, + "loss": 2.4281, + "step": 3374 + }, + { + "epoch": 0.3017703862660944, + "grad_norm": 2.090597008149283, + "learning_rate": 0.00016370758240062357, + "loss": 2.9972, + "step": 3375 + }, + { + "epoch": 0.30185979971387694, + "grad_norm": 0.8949828790516107, + "learning_rate": 0.00016368525724421248, + "loss": 2.1685, + "step": 3376 + }, + { + "epoch": 0.3019492131616595, + "grad_norm": 0.5022906090671218, + "learning_rate": 0.00016366292674659577, + "loss": 2.4493, + "step": 3377 + }, + { + "epoch": 0.3020386266094421, + "grad_norm": 0.6598053279714018, + "learning_rate": 0.0001636405909096463, + "loss": 2.1962, + "step": 3378 + }, + { + "epoch": 0.3021280400572246, + "grad_norm": 0.3338555218672446, + "learning_rate": 0.0001636182497352373, + "loss": 2.161, + "step": 3379 + }, + { + "epoch": 0.30221745350500717, + "grad_norm": 0.5019086258175716, + "learning_rate": 0.00016359590322524253, + "loss": 2.9451, + "step": 3380 + }, + { + "epoch": 0.3023068669527897, + "grad_norm": 0.5215104496013695, + "learning_rate": 0.0001635735513815362, + "loss": 2.3428, + "step": 3381 + }, + { + "epoch": 0.30239628040057226, + "grad_norm": 1.0577730985173084, + "learning_rate": 0.00016355119420599282, + "loss": 2.5563, + "step": 3382 + }, + { + "epoch": 0.3024856938483548, + "grad_norm": 0.8209611355613659, + "learning_rate": 0.00016352883170048758, + "loss": 2.5922, + "step": 3383 + }, + { + "epoch": 0.30257510729613735, + "grad_norm": 0.41235981916491987, + "learning_rate": 0.00016350646386689593, + "loss": 2.0652, + "step": 3384 + }, + { + "epoch": 0.30266452074391986, + "grad_norm": 0.9054826884070715, + "learning_rate": 0.0001634840907070939, + "loss": 2.4863, + "step": 3385 + }, + { + "epoch": 0.30275393419170243, + "grad_norm": 0.8996770122075122, + "learning_rate": 0.0001634617122229578, + "loss": 2.4683, + "step": 3386 + }, + { + "epoch": 0.302843347639485, + "grad_norm": 1.2299379342122156, + "learning_rate": 0.00016343932841636456, + "loss": 2.6695, + "step": 3387 + }, + { + "epoch": 0.3029327610872675, + "grad_norm": 0.7564041491236903, + "learning_rate": 0.00016341693928919145, + "loss": 2.5231, + "step": 3388 + }, + { + "epoch": 0.3030221745350501, + "grad_norm": 0.36199510146915537, + "learning_rate": 0.00016339454484331624, + "loss": 2.4269, + "step": 3389 + }, + { + "epoch": 0.3031115879828326, + "grad_norm": 0.7860241115992583, + "learning_rate": 0.00016337214508061712, + "loss": 2.2133, + "step": 3390 + }, + { + "epoch": 0.3032010014306152, + "grad_norm": 1.3612645947631856, + "learning_rate": 0.00016334974000297271, + "loss": 2.5295, + "step": 3391 + }, + { + "epoch": 0.3032904148783977, + "grad_norm": 0.9488024425630462, + "learning_rate": 0.0001633273296122621, + "loss": 1.9509, + "step": 3392 + }, + { + "epoch": 0.30337982832618027, + "grad_norm": 0.6538936570267675, + "learning_rate": 0.0001633049139103649, + "loss": 2.4248, + "step": 3393 + }, + { + "epoch": 0.3034692417739628, + "grad_norm": 1.1864889230765816, + "learning_rate": 0.00016328249289916097, + "loss": 2.1436, + "step": 3394 + }, + { + "epoch": 0.30355865522174535, + "grad_norm": 0.8844711211165162, + "learning_rate": 0.00016326006658053078, + "loss": 2.5944, + "step": 3395 + }, + { + "epoch": 0.30364806866952787, + "grad_norm": 0.8805664692879601, + "learning_rate": 0.00016323763495635523, + "loss": 2.06, + "step": 3396 + }, + { + "epoch": 0.30373748211731044, + "grad_norm": 1.1951249520124272, + "learning_rate": 0.00016321519802851557, + "loss": 2.4701, + "step": 3397 + }, + { + "epoch": 0.303826895565093, + "grad_norm": 27.584954020811512, + "learning_rate": 0.00016319275579889365, + "loss": 8.3517, + "step": 3398 + }, + { + "epoch": 0.30391630901287553, + "grad_norm": 31.27313851746402, + "learning_rate": 0.0001631703082693716, + "loss": 3.8382, + "step": 3399 + }, + { + "epoch": 0.3040057224606581, + "grad_norm": 4.098150132675263, + "learning_rate": 0.00016314785544183208, + "loss": 3.4649, + "step": 3400 + }, + { + "epoch": 0.3040951359084406, + "grad_norm": 5.1343132803700575, + "learning_rate": 0.00016312539731815816, + "loss": 2.6729, + "step": 3401 + }, + { + "epoch": 0.3041845493562232, + "grad_norm": 16.342881532859817, + "learning_rate": 0.00016310293390023344, + "loss": 2.666, + "step": 3402 + }, + { + "epoch": 0.3042739628040057, + "grad_norm": 6.944337998067373, + "learning_rate": 0.00016308046518994184, + "loss": 2.4821, + "step": 3403 + }, + { + "epoch": 0.3043633762517883, + "grad_norm": 2.3457516809573806, + "learning_rate": 0.00016305799118916783, + "loss": 3.4097, + "step": 3404 + }, + { + "epoch": 0.3044527896995708, + "grad_norm": 3.381427092868798, + "learning_rate": 0.00016303551189979625, + "loss": 3.1638, + "step": 3405 + }, + { + "epoch": 0.30454220314735336, + "grad_norm": 2.054301586774374, + "learning_rate": 0.0001630130273237124, + "loss": 2.8825, + "step": 3406 + }, + { + "epoch": 0.30463161659513593, + "grad_norm": 1.746561381754013, + "learning_rate": 0.00016299053746280206, + "loss": 8.5724, + "step": 3407 + }, + { + "epoch": 0.30472103004291845, + "grad_norm": 4.125844972739123, + "learning_rate": 0.00016296804231895142, + "loss": 2.7886, + "step": 3408 + }, + { + "epoch": 0.304810443490701, + "grad_norm": 1.6276681939539932, + "learning_rate": 0.00016294554189404708, + "loss": 2.5211, + "step": 3409 + }, + { + "epoch": 0.30489985693848354, + "grad_norm": 1.4446274001654127, + "learning_rate": 0.00016292303618997619, + "loss": 2.4787, + "step": 3410 + }, + { + "epoch": 0.3049892703862661, + "grad_norm": 0.9666598958868712, + "learning_rate": 0.00016290052520862624, + "loss": 2.894, + "step": 3411 + }, + { + "epoch": 0.3050786838340486, + "grad_norm": 1.202839131696372, + "learning_rate": 0.00016287800895188522, + "loss": 2.4761, + "step": 3412 + }, + { + "epoch": 0.3051680972818312, + "grad_norm": 2.8538732534534, + "learning_rate": 0.0001628554874216415, + "loss": 3.1315, + "step": 3413 + }, + { + "epoch": 0.3052575107296137, + "grad_norm": 1.5779471932181373, + "learning_rate": 0.00016283296061978398, + "loss": 3.0471, + "step": 3414 + }, + { + "epoch": 0.3053469241773963, + "grad_norm": 0.9784679109881879, + "learning_rate": 0.00016281042854820194, + "loss": 2.9086, + "step": 3415 + }, + { + "epoch": 0.3054363376251788, + "grad_norm": 2.4988054235242685, + "learning_rate": 0.0001627878912087851, + "loss": 2.7667, + "step": 3416 + }, + { + "epoch": 0.30552575107296137, + "grad_norm": 1.6384326976238741, + "learning_rate": 0.00016276534860342368, + "loss": 2.785, + "step": 3417 + }, + { + "epoch": 0.30561516452074394, + "grad_norm": 1.2597358098546538, + "learning_rate": 0.00016274280073400824, + "loss": 2.4543, + "step": 3418 + }, + { + "epoch": 0.30570457796852646, + "grad_norm": 0.5582283150361071, + "learning_rate": 0.00016272024760242992, + "loss": 2.1315, + "step": 3419 + }, + { + "epoch": 0.30579399141630903, + "grad_norm": 1.313828606442226, + "learning_rate": 0.00016269768921058013, + "loss": 3.1282, + "step": 3420 + }, + { + "epoch": 0.30588340486409155, + "grad_norm": 0.5967022682841737, + "learning_rate": 0.0001626751255603509, + "loss": 2.5149, + "step": 3421 + }, + { + "epoch": 0.3059728183118741, + "grad_norm": 0.6631037905008523, + "learning_rate": 0.00016265255665363454, + "loss": 2.6287, + "step": 3422 + }, + { + "epoch": 0.30606223175965663, + "grad_norm": 0.9754842189787103, + "learning_rate": 0.00016262998249232398, + "loss": 2.9518, + "step": 3423 + }, + { + "epoch": 0.3061516452074392, + "grad_norm": 0.45831818341661473, + "learning_rate": 0.00016260740307831237, + "loss": 2.2488, + "step": 3424 + }, + { + "epoch": 0.3062410586552217, + "grad_norm": 0.43530406143853806, + "learning_rate": 0.00016258481841349348, + "loss": 2.6614, + "step": 3425 + }, + { + "epoch": 0.3063304721030043, + "grad_norm": 1.6196142737185473, + "learning_rate": 0.0001625622284997615, + "loss": 2.8112, + "step": 3426 + }, + { + "epoch": 0.30641988555078686, + "grad_norm": 0.5682674334791078, + "learning_rate": 0.0001625396333390109, + "loss": 2.672, + "step": 3427 + }, + { + "epoch": 0.3065092989985694, + "grad_norm": 0.818339196673269, + "learning_rate": 0.00016251703293313687, + "loss": 2.5929, + "step": 3428 + }, + { + "epoch": 0.30659871244635195, + "grad_norm": 4.9748778313244, + "learning_rate": 0.00016249442728403474, + "loss": 8.3797, + "step": 3429 + }, + { + "epoch": 0.30668812589413447, + "grad_norm": 5.234869799781756, + "learning_rate": 0.00016247181639360045, + "loss": 2.2425, + "step": 3430 + }, + { + "epoch": 0.30677753934191704, + "grad_norm": 4.842809382012184, + "learning_rate": 0.00016244920026373038, + "loss": 3.1017, + "step": 3431 + }, + { + "epoch": 0.30686695278969955, + "grad_norm": 1.7152658308458353, + "learning_rate": 0.00016242657889632133, + "loss": 2.7807, + "step": 3432 + }, + { + "epoch": 0.3069563662374821, + "grad_norm": 2.7047666048660948, + "learning_rate": 0.0001624039522932705, + "loss": 2.3876, + "step": 3433 + }, + { + "epoch": 0.30704577968526464, + "grad_norm": 1.5253544501386374, + "learning_rate": 0.00016238132045647553, + "loss": 2.9763, + "step": 3434 + }, + { + "epoch": 0.3071351931330472, + "grad_norm": 1.609436490616568, + "learning_rate": 0.00016235868338783455, + "loss": 2.2176, + "step": 3435 + }, + { + "epoch": 0.3072246065808298, + "grad_norm": 1.6693642703399367, + "learning_rate": 0.00016233604108924609, + "loss": 2.0415, + "step": 3436 + }, + { + "epoch": 0.3073140200286123, + "grad_norm": 1.8539296658043232, + "learning_rate": 0.0001623133935626092, + "loss": 2.5161, + "step": 3437 + }, + { + "epoch": 0.30740343347639487, + "grad_norm": 0.5276478795560124, + "learning_rate": 0.00016229074080982317, + "loss": 2.1876, + "step": 3438 + }, + { + "epoch": 0.3074928469241774, + "grad_norm": 1.240298751503645, + "learning_rate": 0.000162268082832788, + "loss": 2.226, + "step": 3439 + }, + { + "epoch": 0.30758226037195996, + "grad_norm": 2.0040417753653355, + "learning_rate": 0.00016224541963340391, + "loss": 2.9307, + "step": 3440 + }, + { + "epoch": 0.3076716738197425, + "grad_norm": 1.510458144613273, + "learning_rate": 0.00016222275121357163, + "loss": 2.1238, + "step": 3441 + }, + { + "epoch": 0.30776108726752505, + "grad_norm": 0.5362060735605538, + "learning_rate": 0.00016220007757519238, + "loss": 2.1497, + "step": 3442 + }, + { + "epoch": 0.30785050071530756, + "grad_norm": 1.8973280272114998, + "learning_rate": 0.00016217739872016772, + "loss": 2.5873, + "step": 3443 + }, + { + "epoch": 0.30793991416309013, + "grad_norm": 1.118276986379085, + "learning_rate": 0.00016215471465039975, + "loss": 2.1738, + "step": 3444 + }, + { + "epoch": 0.30802932761087265, + "grad_norm": 1.839722853084098, + "learning_rate": 0.00016213202536779087, + "loss": 2.7929, + "step": 3445 + }, + { + "epoch": 0.3081187410586552, + "grad_norm": 1.1318911315731361, + "learning_rate": 0.00016210933087424412, + "loss": 2.7556, + "step": 3446 + }, + { + "epoch": 0.3082081545064378, + "grad_norm": 0.6266998876836054, + "learning_rate": 0.00016208663117166277, + "loss": 2.7803, + "step": 3447 + }, + { + "epoch": 0.3082975679542203, + "grad_norm": 1.3648957590939943, + "learning_rate": 0.00016206392626195063, + "loss": 2.337, + "step": 3448 + }, + { + "epoch": 0.3083869814020029, + "grad_norm": 2.1406512520154912, + "learning_rate": 0.00016204121614701197, + "loss": 2.7204, + "step": 3449 + }, + { + "epoch": 0.3084763948497854, + "grad_norm": 0.8563734319890518, + "learning_rate": 0.00016201850082875146, + "loss": 2.2504, + "step": 3450 + }, + { + "epoch": 0.30856580829756797, + "grad_norm": 0.40326763871230714, + "learning_rate": 0.00016199578030907415, + "loss": 2.2367, + "step": 3451 + }, + { + "epoch": 0.3086552217453505, + "grad_norm": 1.7793351599240768, + "learning_rate": 0.0001619730545898856, + "loss": 2.3376, + "step": 3452 + }, + { + "epoch": 0.30874463519313305, + "grad_norm": 0.43411991555810586, + "learning_rate": 0.00016195032367309183, + "loss": 2.3738, + "step": 3453 + }, + { + "epoch": 0.30883404864091557, + "grad_norm": 0.5889460291432865, + "learning_rate": 0.00016192758756059926, + "loss": 2.9191, + "step": 3454 + }, + { + "epoch": 0.30892346208869814, + "grad_norm": 0.8480134286412762, + "learning_rate": 0.00016190484625431468, + "loss": 2.7781, + "step": 3455 + }, + { + "epoch": 0.3090128755364807, + "grad_norm": 0.5821816799567469, + "learning_rate": 0.00016188209975614542, + "loss": 2.5682, + "step": 3456 + }, + { + "epoch": 0.30910228898426323, + "grad_norm": 0.42555066424670634, + "learning_rate": 0.00016185934806799916, + "loss": 3.0601, + "step": 3457 + }, + { + "epoch": 0.3091917024320458, + "grad_norm": 0.346479459983528, + "learning_rate": 0.0001618365911917841, + "loss": 2.7078, + "step": 3458 + }, + { + "epoch": 0.3092811158798283, + "grad_norm": 0.24681644333815786, + "learning_rate": 0.00016181382912940884, + "loss": 2.2413, + "step": 3459 + }, + { + "epoch": 0.3093705293276109, + "grad_norm": 0.36242379653271073, + "learning_rate": 0.00016179106188278234, + "loss": 2.3272, + "step": 3460 + }, + { + "epoch": 0.3094599427753934, + "grad_norm": 0.6474152248314234, + "learning_rate": 0.00016176828945381415, + "loss": 1.9221, + "step": 3461 + }, + { + "epoch": 0.309549356223176, + "grad_norm": 0.49838212902353135, + "learning_rate": 0.00016174551184441408, + "loss": 2.5846, + "step": 3462 + }, + { + "epoch": 0.3096387696709585, + "grad_norm": 0.4504349098803583, + "learning_rate": 0.00016172272905649253, + "loss": 2.2995, + "step": 3463 + }, + { + "epoch": 0.30972818311874106, + "grad_norm": 0.7274887650525198, + "learning_rate": 0.00016169994109196023, + "loss": 2.3256, + "step": 3464 + }, + { + "epoch": 0.3098175965665236, + "grad_norm": 0.43474623418630637, + "learning_rate": 0.00016167714795272837, + "loss": 2.9941, + "step": 3465 + }, + { + "epoch": 0.30990701001430615, + "grad_norm": 1.124044935520869, + "learning_rate": 0.00016165434964070862, + "loss": 2.5207, + "step": 3466 + }, + { + "epoch": 0.3099964234620887, + "grad_norm": 0.5954712026768604, + "learning_rate": 0.000161631546157813, + "loss": 2.4976, + "step": 3467 + }, + { + "epoch": 0.31008583690987124, + "grad_norm": 0.9346430005684103, + "learning_rate": 0.00016160873750595405, + "loss": 2.2887, + "step": 3468 + }, + { + "epoch": 0.3101752503576538, + "grad_norm": 0.7907224305917581, + "learning_rate": 0.00016158592368704472, + "loss": 2.5922, + "step": 3469 + }, + { + "epoch": 0.3102646638054363, + "grad_norm": 0.756876880845213, + "learning_rate": 0.00016156310470299832, + "loss": 2.0676, + "step": 3470 + }, + { + "epoch": 0.3103540772532189, + "grad_norm": 0.41700682931715594, + "learning_rate": 0.00016154028055572866, + "loss": 2.7136, + "step": 3471 + }, + { + "epoch": 0.3104434907010014, + "grad_norm": 0.5648243857469046, + "learning_rate": 0.00016151745124715002, + "loss": 3.256, + "step": 3472 + }, + { + "epoch": 0.310532904148784, + "grad_norm": 0.739518888714449, + "learning_rate": 0.000161494616779177, + "loss": 2.0131, + "step": 3473 + }, + { + "epoch": 0.3106223175965665, + "grad_norm": 0.7266814044098656, + "learning_rate": 0.00016147177715372476, + "loss": 2.1566, + "step": 3474 + }, + { + "epoch": 0.31071173104434907, + "grad_norm": 1.4028847937704134, + "learning_rate": 0.00016144893237270887, + "loss": 2.5351, + "step": 3475 + }, + { + "epoch": 0.31080114449213164, + "grad_norm": 1.13085520685719, + "learning_rate": 0.00016142608243804513, + "loss": 2.0319, + "step": 3476 + }, + { + "epoch": 0.31089055793991416, + "grad_norm": 1.5490329337547748, + "learning_rate": 0.0001614032273516501, + "loss": 3.0284, + "step": 3477 + }, + { + "epoch": 0.31097997138769673, + "grad_norm": 2.3040419037274225, + "learning_rate": 0.00016138036711544054, + "loss": 2.226, + "step": 3478 + }, + { + "epoch": 0.31106938483547925, + "grad_norm": 0.736904527833318, + "learning_rate": 0.0001613575017313337, + "loss": 2.3593, + "step": 3479 + }, + { + "epoch": 0.3111587982832618, + "grad_norm": 0.4990642198025826, + "learning_rate": 0.00016133463120124731, + "loss": 2.8791, + "step": 3480 + }, + { + "epoch": 0.31124821173104433, + "grad_norm": 1.429484776756485, + "learning_rate": 0.00016131175552709946, + "loss": 2.5542, + "step": 3481 + }, + { + "epoch": 0.3113376251788269, + "grad_norm": 1.1863807820672299, + "learning_rate": 0.00016128887471080874, + "loss": 2.4665, + "step": 3482 + }, + { + "epoch": 0.3114270386266094, + "grad_norm": 0.4716397381263984, + "learning_rate": 0.00016126598875429408, + "loss": 2.2972, + "step": 3483 + }, + { + "epoch": 0.311516452074392, + "grad_norm": 0.5288421219971439, + "learning_rate": 0.00016124309765947498, + "loss": 1.9555, + "step": 3484 + }, + { + "epoch": 0.3116058655221745, + "grad_norm": 1.0028260954374928, + "learning_rate": 0.00016122020142827123, + "loss": 2.0891, + "step": 3485 + }, + { + "epoch": 0.3116952789699571, + "grad_norm": 1.8094472630247496, + "learning_rate": 0.0001611973000626031, + "loss": 2.6879, + "step": 3486 + }, + { + "epoch": 0.31178469241773965, + "grad_norm": 1.1040729238561107, + "learning_rate": 0.00016117439356439132, + "loss": 2.5302, + "step": 3487 + }, + { + "epoch": 0.31187410586552217, + "grad_norm": 1.7510212673414312, + "learning_rate": 0.00016115148193555706, + "loss": 2.0736, + "step": 3488 + }, + { + "epoch": 0.31196351931330474, + "grad_norm": 0.3255401945227201, + "learning_rate": 0.00016112856517802183, + "loss": 2.7907, + "step": 3489 + }, + { + "epoch": 0.31205293276108725, + "grad_norm": 0.4545050752660696, + "learning_rate": 0.0001611056432937077, + "loss": 2.1911, + "step": 3490 + }, + { + "epoch": 0.3121423462088698, + "grad_norm": 0.31077530152862193, + "learning_rate": 0.00016108271628453703, + "loss": 2.5567, + "step": 3491 + }, + { + "epoch": 0.31223175965665234, + "grad_norm": 0.8854159468522601, + "learning_rate": 0.00016105978415243276, + "loss": 1.9712, + "step": 3492 + }, + { + "epoch": 0.3123211731044349, + "grad_norm": 0.30871474050833986, + "learning_rate": 0.00016103684689931807, + "loss": 2.0724, + "step": 3493 + }, + { + "epoch": 0.31241058655221743, + "grad_norm": 9.638145466377056, + "learning_rate": 0.0001610139045271168, + "loss": 8.7948, + "step": 3494 + }, + { + "epoch": 0.3125, + "grad_norm": 7.277624957834278, + "learning_rate": 0.00016099095703775302, + "loss": 2.344, + "step": 3495 + }, + { + "epoch": 0.31258941344778257, + "grad_norm": 7.896886203399324, + "learning_rate": 0.00016096800443315132, + "loss": 2.1172, + "step": 3496 + }, + { + "epoch": 0.3126788268955651, + "grad_norm": 2.573940204294962, + "learning_rate": 0.00016094504671523673, + "loss": 2.5126, + "step": 3497 + }, + { + "epoch": 0.31276824034334766, + "grad_norm": 11.643692666253742, + "learning_rate": 0.00016092208388593469, + "loss": 2.5756, + "step": 3498 + }, + { + "epoch": 0.3128576537911302, + "grad_norm": 2.992754101249407, + "learning_rate": 0.00016089911594717102, + "loss": 2.8507, + "step": 3499 + }, + { + "epoch": 0.31294706723891275, + "grad_norm": 0.6655351675258421, + "learning_rate": 0.00016087614290087208, + "loss": 2.2766, + "step": 3500 + }, + { + "epoch": 0.31303648068669526, + "grad_norm": 1.979673982703026, + "learning_rate": 0.00016085316474896452, + "loss": 2.3945, + "step": 3501 + }, + { + "epoch": 0.31312589413447783, + "grad_norm": 2.324981948768301, + "learning_rate": 0.00016083018149337558, + "loss": 2.528, + "step": 3502 + }, + { + "epoch": 0.31321530758226035, + "grad_norm": 1.6273817731769489, + "learning_rate": 0.0001608071931360327, + "loss": 2.4136, + "step": 3503 + }, + { + "epoch": 0.3133047210300429, + "grad_norm": 0.6952970273839586, + "learning_rate": 0.00016078419967886402, + "loss": 2.2806, + "step": 3504 + }, + { + "epoch": 0.3133941344778255, + "grad_norm": 0.9605937615051375, + "learning_rate": 0.00016076120112379792, + "loss": 2.4323, + "step": 3505 + }, + { + "epoch": 0.313483547925608, + "grad_norm": 0.5167971098730092, + "learning_rate": 0.00016073819747276327, + "loss": 2.5136, + "step": 3506 + }, + { + "epoch": 0.3135729613733906, + "grad_norm": 1.3599771029397414, + "learning_rate": 0.0001607151887276893, + "loss": 2.1255, + "step": 3507 + }, + { + "epoch": 0.3136623748211731, + "grad_norm": 0.8292140226550087, + "learning_rate": 0.00016069217489050584, + "loss": 2.6124, + "step": 3508 + }, + { + "epoch": 0.31375178826895567, + "grad_norm": 0.9655235111619024, + "learning_rate": 0.00016066915596314293, + "loss": 2.6025, + "step": 3509 + }, + { + "epoch": 0.3138412017167382, + "grad_norm": 1.151267478537801, + "learning_rate": 0.00016064613194753118, + "loss": 2.503, + "step": 3510 + }, + { + "epoch": 0.31393061516452075, + "grad_norm": 0.9294639348107754, + "learning_rate": 0.0001606231028456016, + "loss": 2.7267, + "step": 3511 + }, + { + "epoch": 0.31402002861230327, + "grad_norm": 1.9408238322363547, + "learning_rate": 0.0001606000686592856, + "loss": 2.1589, + "step": 3512 + }, + { + "epoch": 0.31410944206008584, + "grad_norm": 0.9087893206007729, + "learning_rate": 0.00016057702939051502, + "loss": 3.0305, + "step": 3513 + }, + { + "epoch": 0.31419885550786836, + "grad_norm": 12.45817834727231, + "learning_rate": 0.00016055398504122214, + "loss": 8.7324, + "step": 3514 + }, + { + "epoch": 0.31428826895565093, + "grad_norm": 6.4450722419909185, + "learning_rate": 0.00016053093561333966, + "loss": 2.7798, + "step": 3515 + }, + { + "epoch": 0.3143776824034335, + "grad_norm": 8.415064620283145, + "learning_rate": 0.00016050788110880072, + "loss": 2.6141, + "step": 3516 + }, + { + "epoch": 0.314467095851216, + "grad_norm": 4.041849417380432, + "learning_rate": 0.00016048482152953889, + "loss": 3.3189, + "step": 3517 + }, + { + "epoch": 0.3145565092989986, + "grad_norm": 3.0753959128574317, + "learning_rate": 0.0001604617568774881, + "loss": 2.4194, + "step": 3518 + }, + { + "epoch": 0.3146459227467811, + "grad_norm": 4.634228626599785, + "learning_rate": 0.0001604386871545828, + "loss": 2.3469, + "step": 3519 + }, + { + "epoch": 0.3147353361945637, + "grad_norm": 4.055487331499101, + "learning_rate": 0.00016041561236275777, + "loss": 2.3316, + "step": 3520 + }, + { + "epoch": 0.3148247496423462, + "grad_norm": 1.9295299875055132, + "learning_rate": 0.00016039253250394833, + "loss": 2.6072, + "step": 3521 + }, + { + "epoch": 0.31491416309012876, + "grad_norm": 0.8665218443804601, + "learning_rate": 0.0001603694475800901, + "loss": 2.6693, + "step": 3522 + }, + { + "epoch": 0.3150035765379113, + "grad_norm": 2.066173577281979, + "learning_rate": 0.00016034635759311922, + "loss": 2.4683, + "step": 3523 + }, + { + "epoch": 0.31509298998569385, + "grad_norm": 2.9572449152463918, + "learning_rate": 0.00016032326254497218, + "loss": 2.1165, + "step": 3524 + }, + { + "epoch": 0.3151824034334764, + "grad_norm": 1.5869771940713535, + "learning_rate": 0.000160300162437586, + "loss": 2.4227, + "step": 3525 + }, + { + "epoch": 0.31527181688125894, + "grad_norm": 0.593194143384899, + "learning_rate": 0.00016027705727289802, + "loss": 2.7482, + "step": 3526 + }, + { + "epoch": 0.3153612303290415, + "grad_norm": 1.9281164465663807, + "learning_rate": 0.00016025394705284602, + "loss": 8.2952, + "step": 3527 + }, + { + "epoch": 0.315450643776824, + "grad_norm": 0.7949887725394954, + "learning_rate": 0.00016023083177936823, + "loss": 2.1347, + "step": 3528 + }, + { + "epoch": 0.3155400572246066, + "grad_norm": 0.6837324178024037, + "learning_rate": 0.00016020771145440336, + "loss": 2.4217, + "step": 3529 + }, + { + "epoch": 0.3156294706723891, + "grad_norm": 0.7035721348951747, + "learning_rate": 0.00016018458607989044, + "loss": 2.1434, + "step": 3530 + }, + { + "epoch": 0.3157188841201717, + "grad_norm": 0.6661742687453105, + "learning_rate": 0.00016016145565776895, + "loss": 2.2723, + "step": 3531 + }, + { + "epoch": 0.3158082975679542, + "grad_norm": 0.5540062737334664, + "learning_rate": 0.00016013832018997882, + "loss": 2.4695, + "step": 3532 + }, + { + "epoch": 0.31589771101573677, + "grad_norm": 1.1727194262846325, + "learning_rate": 0.00016011517967846043, + "loss": 2.2628, + "step": 3533 + }, + { + "epoch": 0.3159871244635193, + "grad_norm": 0.9230167871146924, + "learning_rate": 0.00016009203412515455, + "loss": 2.1844, + "step": 3534 + }, + { + "epoch": 0.31607653791130186, + "grad_norm": 0.7894704461018311, + "learning_rate": 0.00016006888353200228, + "loss": 2.5973, + "step": 3535 + }, + { + "epoch": 0.31616595135908443, + "grad_norm": 0.6108613889957865, + "learning_rate": 0.00016004572790094535, + "loss": 2.6461, + "step": 3536 + }, + { + "epoch": 0.31625536480686695, + "grad_norm": 0.4932163744451215, + "learning_rate": 0.0001600225672339257, + "loss": 2.6153, + "step": 3537 + }, + { + "epoch": 0.3163447782546495, + "grad_norm": 1.5389335738702372, + "learning_rate": 0.00015999940153288582, + "loss": 3.151, + "step": 3538 + }, + { + "epoch": 0.31643419170243203, + "grad_norm": 1.1452338972315055, + "learning_rate": 0.00015997623079976863, + "loss": 1.8036, + "step": 3539 + }, + { + "epoch": 0.3165236051502146, + "grad_norm": 0.7657383211735571, + "learning_rate": 0.00015995305503651737, + "loss": 2.5541, + "step": 3540 + }, + { + "epoch": 0.3166130185979971, + "grad_norm": 1.08431801776986, + "learning_rate": 0.00015992987424507578, + "loss": 2.1592, + "step": 3541 + }, + { + "epoch": 0.3167024320457797, + "grad_norm": 0.4517433027881385, + "learning_rate": 0.000159906688427388, + "loss": 2.084, + "step": 3542 + }, + { + "epoch": 0.3167918454935622, + "grad_norm": 0.5671391541755353, + "learning_rate": 0.00015988349758539868, + "loss": 2.3235, + "step": 3543 + }, + { + "epoch": 0.3168812589413448, + "grad_norm": 0.30879834540595863, + "learning_rate": 0.00015986030172105266, + "loss": 2.2531, + "step": 3544 + }, + { + "epoch": 0.31697067238912735, + "grad_norm": 0.40269150949237525, + "learning_rate": 0.00015983710083629547, + "loss": 2.2256, + "step": 3545 + }, + { + "epoch": 0.31706008583690987, + "grad_norm": 0.9605908695416755, + "learning_rate": 0.00015981389493307288, + "loss": 2.1956, + "step": 3546 + }, + { + "epoch": 0.31714949928469244, + "grad_norm": 0.7459518178458757, + "learning_rate": 0.0001597906840133312, + "loss": 2.6742, + "step": 3547 + }, + { + "epoch": 0.31723891273247495, + "grad_norm": 1.0772543885463992, + "learning_rate": 0.000159767468079017, + "loss": 2.5476, + "step": 3548 + }, + { + "epoch": 0.3173283261802575, + "grad_norm": 0.2381672832347035, + "learning_rate": 0.00015974424713207746, + "loss": 2.7087, + "step": 3549 + }, + { + "epoch": 0.31741773962804004, + "grad_norm": 0.7309663185927715, + "learning_rate": 0.0001597210211744601, + "loss": 2.2279, + "step": 3550 + }, + { + "epoch": 0.3175071530758226, + "grad_norm": 1.2065741370444985, + "learning_rate": 0.0001596977902081128, + "loss": 2.3307, + "step": 3551 + }, + { + "epoch": 0.31759656652360513, + "grad_norm": 0.24710528116695413, + "learning_rate": 0.00015967455423498387, + "loss": 2.1626, + "step": 3552 + }, + { + "epoch": 0.3176859799713877, + "grad_norm": 0.7270158685218845, + "learning_rate": 0.00015965131325702223, + "loss": 2.0445, + "step": 3553 + }, + { + "epoch": 0.3177753934191702, + "grad_norm": 0.9464654881940598, + "learning_rate": 0.00015962806727617694, + "loss": 2.4852, + "step": 3554 + }, + { + "epoch": 0.3178648068669528, + "grad_norm": 0.9675221328429314, + "learning_rate": 0.00015960481629439768, + "loss": 2.6625, + "step": 3555 + }, + { + "epoch": 0.31795422031473536, + "grad_norm": 0.8455794527746799, + "learning_rate": 0.00015958156031363444, + "loss": 2.9036, + "step": 3556 + }, + { + "epoch": 0.3180436337625179, + "grad_norm": 0.9862107322629452, + "learning_rate": 0.0001595582993358377, + "loss": 2.0883, + "step": 3557 + }, + { + "epoch": 0.31813304721030045, + "grad_norm": 0.5606727714635409, + "learning_rate": 0.00015953503336295835, + "loss": 2.4992, + "step": 3558 + }, + { + "epoch": 0.31822246065808296, + "grad_norm": 1.2454342659449409, + "learning_rate": 0.00015951176239694764, + "loss": 2.2372, + "step": 3559 + }, + { + "epoch": 0.31831187410586553, + "grad_norm": 0.74761537629513, + "learning_rate": 0.00015948848643975726, + "loss": 2.3737, + "step": 3560 + }, + { + "epoch": 0.31840128755364805, + "grad_norm": 0.6589794918300171, + "learning_rate": 0.00015946520549333938, + "loss": 1.8333, + "step": 3561 + }, + { + "epoch": 0.3184907010014306, + "grad_norm": 0.6178943345876869, + "learning_rate": 0.00015944191955964655, + "loss": 2.4178, + "step": 3562 + }, + { + "epoch": 0.31858011444921314, + "grad_norm": 2.3097529103273144, + "learning_rate": 0.0001594186286406317, + "loss": 2.3551, + "step": 3563 + }, + { + "epoch": 0.3186695278969957, + "grad_norm": 0.8052227502782046, + "learning_rate": 0.00015939533273824822, + "loss": 1.9873, + "step": 3564 + }, + { + "epoch": 0.3187589413447783, + "grad_norm": 0.5904930647018856, + "learning_rate": 0.00015937203185444992, + "loss": 2.5773, + "step": 3565 + }, + { + "epoch": 0.3188483547925608, + "grad_norm": 2.345357973122811, + "learning_rate": 0.000159348725991191, + "loss": 8.229, + "step": 3566 + }, + { + "epoch": 0.31893776824034337, + "grad_norm": 0.8179813923108095, + "learning_rate": 0.00015932541515042615, + "loss": 2.1761, + "step": 3567 + }, + { + "epoch": 0.3190271816881259, + "grad_norm": 1.1820199560289373, + "learning_rate": 0.00015930209933411036, + "loss": 2.1043, + "step": 3568 + }, + { + "epoch": 0.31911659513590845, + "grad_norm": 0.5075677161130729, + "learning_rate": 0.00015927877854419908, + "loss": 2.2053, + "step": 3569 + }, + { + "epoch": 0.31920600858369097, + "grad_norm": 0.8712300455831171, + "learning_rate": 0.00015925545278264828, + "loss": 2.2305, + "step": 3570 + }, + { + "epoch": 0.31929542203147354, + "grad_norm": 0.46233134488519273, + "learning_rate": 0.00015923212205141418, + "loss": 2.4784, + "step": 3571 + }, + { + "epoch": 0.31938483547925606, + "grad_norm": 0.3561402684343024, + "learning_rate": 0.00015920878635245357, + "loss": 2.235, + "step": 3572 + }, + { + "epoch": 0.31947424892703863, + "grad_norm": 0.7900042960032385, + "learning_rate": 0.00015918544568772354, + "loss": 2.5824, + "step": 3573 + }, + { + "epoch": 0.3195636623748212, + "grad_norm": 1.257826778145137, + "learning_rate": 0.00015916210005918164, + "loss": 2.3211, + "step": 3574 + }, + { + "epoch": 0.3196530758226037, + "grad_norm": 0.41523826408848286, + "learning_rate": 0.00015913874946878588, + "loss": 2.0891, + "step": 3575 + }, + { + "epoch": 0.3197424892703863, + "grad_norm": 0.8198196427068926, + "learning_rate": 0.00015911539391849462, + "loss": 2.2291, + "step": 3576 + }, + { + "epoch": 0.3198319027181688, + "grad_norm": 1.7199179680758716, + "learning_rate": 0.00015909203341026666, + "loss": 2.5255, + "step": 3577 + }, + { + "epoch": 0.3199213161659514, + "grad_norm": 1.055033985823719, + "learning_rate": 0.00015906866794606126, + "loss": 2.6634, + "step": 3578 + }, + { + "epoch": 0.3200107296137339, + "grad_norm": 0.3253720286262901, + "learning_rate": 0.00015904529752783794, + "loss": 2.1798, + "step": 3579 + }, + { + "epoch": 0.32010014306151646, + "grad_norm": 0.3426953631674512, + "learning_rate": 0.00015902192215755688, + "loss": 2.4123, + "step": 3580 + }, + { + "epoch": 0.320189556509299, + "grad_norm": 0.5200698875010394, + "learning_rate": 0.00015899854183717852, + "loss": 2.0261, + "step": 3581 + }, + { + "epoch": 0.32027896995708155, + "grad_norm": 0.3915915445900581, + "learning_rate": 0.00015897515656866363, + "loss": 2.0023, + "step": 3582 + }, + { + "epoch": 0.32036838340486407, + "grad_norm": 0.6608288081374034, + "learning_rate": 0.00015895176635397364, + "loss": 3.074, + "step": 3583 + }, + { + "epoch": 0.32045779685264664, + "grad_norm": 0.4312861859072764, + "learning_rate": 0.00015892837119507014, + "loss": 2.115, + "step": 3584 + }, + { + "epoch": 0.3205472103004292, + "grad_norm": 0.3124726361369686, + "learning_rate": 0.0001589049710939154, + "loss": 2.3556, + "step": 3585 + }, + { + "epoch": 0.3206366237482117, + "grad_norm": 0.44880501487171715, + "learning_rate": 0.0001588815660524718, + "loss": 2.5186, + "step": 3586 + }, + { + "epoch": 0.3207260371959943, + "grad_norm": 1.6182632441891318, + "learning_rate": 0.0001588581560727024, + "loss": 2.8106, + "step": 3587 + }, + { + "epoch": 0.3208154506437768, + "grad_norm": 2.3597260667025566, + "learning_rate": 0.00015883474115657056, + "loss": 8.1006, + "step": 3588 + }, + { + "epoch": 0.3209048640915594, + "grad_norm": 0.4302149230632218, + "learning_rate": 0.00015881132130603998, + "loss": 2.5256, + "step": 3589 + }, + { + "epoch": 0.3209942775393419, + "grad_norm": 0.44603654858229325, + "learning_rate": 0.00015878789652307496, + "loss": 2.623, + "step": 3590 + }, + { + "epoch": 0.32108369098712447, + "grad_norm": 0.5033752356362895, + "learning_rate": 0.00015876446680964, + "loss": 2.6768, + "step": 3591 + }, + { + "epoch": 0.321173104434907, + "grad_norm": 0.7688891920051377, + "learning_rate": 0.00015874103216770023, + "loss": 2.3009, + "step": 3592 + }, + { + "epoch": 0.32126251788268956, + "grad_norm": 0.9756081271047176, + "learning_rate": 0.00015871759259922097, + "loss": 2.3345, + "step": 3593 + }, + { + "epoch": 0.32135193133047213, + "grad_norm": 0.4025439861423465, + "learning_rate": 0.0001586941481061682, + "loss": 2.578, + "step": 3594 + }, + { + "epoch": 0.32144134477825465, + "grad_norm": 1.4068489862355018, + "learning_rate": 0.0001586706986905081, + "loss": 2.1082, + "step": 3595 + }, + { + "epoch": 0.3215307582260372, + "grad_norm": 0.30117832354504126, + "learning_rate": 0.00015864724435420732, + "loss": 2.5268, + "step": 3596 + }, + { + "epoch": 0.32162017167381973, + "grad_norm": 1.1656082881699554, + "learning_rate": 0.000158623785099233, + "loss": 2.2297, + "step": 3597 + }, + { + "epoch": 0.3217095851216023, + "grad_norm": 1.371643804920025, + "learning_rate": 0.0001586003209275526, + "loss": 2.4313, + "step": 3598 + }, + { + "epoch": 0.3217989985693848, + "grad_norm": 0.3632264868960814, + "learning_rate": 0.00015857685184113412, + "loss": 2.7427, + "step": 3599 + }, + { + "epoch": 0.3218884120171674, + "grad_norm": 0.9637467885622107, + "learning_rate": 0.00015855337784194577, + "loss": 2.6294, + "step": 3600 + }, + { + "epoch": 0.3219778254649499, + "grad_norm": 0.2539502873615926, + "learning_rate": 0.00015852989893195635, + "loss": 2.5402, + "step": 3601 + }, + { + "epoch": 0.3220672389127325, + "grad_norm": 0.4702333380515738, + "learning_rate": 0.00015850641511313496, + "loss": 2.7656, + "step": 3602 + }, + { + "epoch": 0.322156652360515, + "grad_norm": 0.4296661709008919, + "learning_rate": 0.00015848292638745125, + "loss": 2.5669, + "step": 3603 + }, + { + "epoch": 0.32224606580829757, + "grad_norm": 1.1397831122111153, + "learning_rate": 0.0001584594327568751, + "loss": 2.2541, + "step": 3604 + }, + { + "epoch": 0.32233547925608014, + "grad_norm": 0.6287072571109867, + "learning_rate": 0.00015843593422337695, + "loss": 3.138, + "step": 3605 + }, + { + "epoch": 0.32242489270386265, + "grad_norm": 0.810160458556817, + "learning_rate": 0.00015841243078892756, + "loss": 2.1158, + "step": 3606 + }, + { + "epoch": 0.3225143061516452, + "grad_norm": 0.44432061344997315, + "learning_rate": 0.0001583889224554981, + "loss": 2.6436, + "step": 3607 + }, + { + "epoch": 0.32260371959942774, + "grad_norm": 0.5655216348841309, + "learning_rate": 0.0001583654092250603, + "loss": 2.1943, + "step": 3608 + }, + { + "epoch": 0.3226931330472103, + "grad_norm": 0.5955641299965896, + "learning_rate": 0.00015834189109958607, + "loss": 2.4792, + "step": 3609 + }, + { + "epoch": 0.32278254649499283, + "grad_norm": 1.2884123906904696, + "learning_rate": 0.00015831836808104788, + "loss": 2.184, + "step": 3610 + }, + { + "epoch": 0.3228719599427754, + "grad_norm": 0.4734064617175246, + "learning_rate": 0.0001582948401714186, + "loss": 2.0994, + "step": 3611 + }, + { + "epoch": 0.3229613733905579, + "grad_norm": 1.2311489004116778, + "learning_rate": 0.0001582713073726715, + "loss": 2.1207, + "step": 3612 + }, + { + "epoch": 0.3230507868383405, + "grad_norm": 1.055287569438127, + "learning_rate": 0.00015824776968678024, + "loss": 2.3369, + "step": 3613 + }, + { + "epoch": 0.32314020028612306, + "grad_norm": 0.58393176883534, + "learning_rate": 0.00015822422711571883, + "loss": 2.5826, + "step": 3614 + }, + { + "epoch": 0.3232296137339056, + "grad_norm": 0.4919609341132873, + "learning_rate": 0.00015820067966146185, + "loss": 1.8359, + "step": 3615 + }, + { + "epoch": 0.32331902718168815, + "grad_norm": 1.0994319838547353, + "learning_rate": 0.00015817712732598413, + "loss": 2.202, + "step": 3616 + }, + { + "epoch": 0.32340844062947066, + "grad_norm": 0.6223540616436825, + "learning_rate": 0.00015815357011126103, + "loss": 2.4003, + "step": 3617 + }, + { + "epoch": 0.32349785407725323, + "grad_norm": 0.8909324657280331, + "learning_rate": 0.0001581300080192682, + "loss": 2.4766, + "step": 3618 + }, + { + "epoch": 0.32358726752503575, + "grad_norm": 0.47771903874734734, + "learning_rate": 0.00015810644105198184, + "loss": 2.2447, + "step": 3619 + }, + { + "epoch": 0.3236766809728183, + "grad_norm": 0.7253474849278306, + "learning_rate": 0.0001580828692113784, + "loss": 2.1608, + "step": 3620 + }, + { + "epoch": 0.32376609442060084, + "grad_norm": 1.0992292062095517, + "learning_rate": 0.0001580592924994349, + "loss": 2.7029, + "step": 3621 + }, + { + "epoch": 0.3238555078683834, + "grad_norm": 1.4996368921128354, + "learning_rate": 0.00015803571091812865, + "loss": 2.6202, + "step": 3622 + }, + { + "epoch": 0.323944921316166, + "grad_norm": 1.0493103719774242, + "learning_rate": 0.0001580121244694374, + "loss": 2.5611, + "step": 3623 + }, + { + "epoch": 0.3240343347639485, + "grad_norm": 0.7115598161414122, + "learning_rate": 0.00015798853315533931, + "loss": 2.3671, + "step": 3624 + }, + { + "epoch": 0.32412374821173107, + "grad_norm": 0.6497449503120611, + "learning_rate": 0.00015796493697781304, + "loss": 1.8441, + "step": 3625 + }, + { + "epoch": 0.3242131616595136, + "grad_norm": 0.7569390534352259, + "learning_rate": 0.0001579413359388375, + "loss": 2.1068, + "step": 3626 + }, + { + "epoch": 0.32430257510729615, + "grad_norm": 1.0455184741939456, + "learning_rate": 0.00015791773004039206, + "loss": 2.4692, + "step": 3627 + }, + { + "epoch": 0.32439198855507867, + "grad_norm": 1.9954103668501635, + "learning_rate": 0.00015789411928445653, + "loss": 2.5405, + "step": 3628 + }, + { + "epoch": 0.32448140200286124, + "grad_norm": 1.6154289423665038, + "learning_rate": 0.00015787050367301118, + "loss": 2.2915, + "step": 3629 + }, + { + "epoch": 0.32457081545064376, + "grad_norm": 1.9359852823370598, + "learning_rate": 0.00015784688320803655, + "loss": 2.2778, + "step": 3630 + }, + { + "epoch": 0.32466022889842633, + "grad_norm": 1.0326431144934216, + "learning_rate": 0.00015782325789151367, + "loss": 2.4323, + "step": 3631 + }, + { + "epoch": 0.32474964234620884, + "grad_norm": 0.467532102038626, + "learning_rate": 0.00015779962772542402, + "loss": 2.2925, + "step": 3632 + }, + { + "epoch": 0.3248390557939914, + "grad_norm": 0.9831915262788554, + "learning_rate": 0.0001577759927117494, + "loss": 2.346, + "step": 3633 + }, + { + "epoch": 0.324928469241774, + "grad_norm": 0.39009522705523203, + "learning_rate": 0.00015775235285247203, + "loss": 2.0889, + "step": 3634 + }, + { + "epoch": 0.3250178826895565, + "grad_norm": 1.1803524571302442, + "learning_rate": 0.00015772870814957453, + "loss": 2.189, + "step": 3635 + }, + { + "epoch": 0.3251072961373391, + "grad_norm": 0.43255296289833955, + "learning_rate": 0.00015770505860504005, + "loss": 2.3729, + "step": 3636 + }, + { + "epoch": 0.3251967095851216, + "grad_norm": 0.3454337130012161, + "learning_rate": 0.000157681404220852, + "loss": 2.2496, + "step": 3637 + }, + { + "epoch": 0.32528612303290416, + "grad_norm": 0.32907106589694124, + "learning_rate": 0.00015765774499899423, + "loss": 1.8168, + "step": 3638 + }, + { + "epoch": 0.3253755364806867, + "grad_norm": 1.3165047641391505, + "learning_rate": 0.00015763408094145103, + "loss": 2.5489, + "step": 3639 + }, + { + "epoch": 0.32546494992846925, + "grad_norm": 0.5345599742958853, + "learning_rate": 0.00015761041205020703, + "loss": 2.6876, + "step": 3640 + }, + { + "epoch": 0.32555436337625177, + "grad_norm": 1.2948082905349088, + "learning_rate": 0.00015758673832724738, + "loss": 2.0438, + "step": 3641 + }, + { + "epoch": 0.32564377682403434, + "grad_norm": 0.5002170525584608, + "learning_rate": 0.00015756305977455753, + "loss": 2.3711, + "step": 3642 + }, + { + "epoch": 0.3257331902718169, + "grad_norm": 1.262417546084448, + "learning_rate": 0.00015753937639412336, + "loss": 2.6973, + "step": 3643 + }, + { + "epoch": 0.3258226037195994, + "grad_norm": 2.7205335962431465, + "learning_rate": 0.00015751568818793117, + "loss": 2.7549, + "step": 3644 + }, + { + "epoch": 0.325912017167382, + "grad_norm": 1.0541739694868772, + "learning_rate": 0.0001574919951579677, + "loss": 2.5959, + "step": 3645 + }, + { + "epoch": 0.3260014306151645, + "grad_norm": 1.8262210310596132, + "learning_rate": 0.00015746829730622, + "loss": 2.4469, + "step": 3646 + }, + { + "epoch": 0.3260908440629471, + "grad_norm": 0.8894097135319621, + "learning_rate": 0.00015744459463467564, + "loss": 2.3804, + "step": 3647 + }, + { + "epoch": 0.3261802575107296, + "grad_norm": 0.6388224237687361, + "learning_rate": 0.00015742088714532247, + "loss": 2.3378, + "step": 3648 + }, + { + "epoch": 0.32626967095851217, + "grad_norm": 1.7694880487536122, + "learning_rate": 0.00015739717484014888, + "loss": 2.5506, + "step": 3649 + }, + { + "epoch": 0.3263590844062947, + "grad_norm": 1.0683555903561808, + "learning_rate": 0.00015737345772114355, + "loss": 2.2892, + "step": 3650 + }, + { + "epoch": 0.32644849785407726, + "grad_norm": 1.0024346293505282, + "learning_rate": 0.0001573497357902956, + "loss": 2.193, + "step": 3651 + }, + { + "epoch": 0.3265379113018598, + "grad_norm": 2.3275012514543616, + "learning_rate": 0.00015732600904959455, + "loss": 2.1532, + "step": 3652 + }, + { + "epoch": 0.32662732474964234, + "grad_norm": 0.35475389833719334, + "learning_rate": 0.00015730227750103038, + "loss": 2.4453, + "step": 3653 + }, + { + "epoch": 0.3267167381974249, + "grad_norm": 1.1569038750247693, + "learning_rate": 0.0001572785411465934, + "loss": 3.009, + "step": 3654 + }, + { + "epoch": 0.32680615164520743, + "grad_norm": 0.9231109881097309, + "learning_rate": 0.00015725479998827434, + "loss": 2.3121, + "step": 3655 + }, + { + "epoch": 0.32689556509299, + "grad_norm": 0.27342837663310793, + "learning_rate": 0.00015723105402806436, + "loss": 2.2044, + "step": 3656 + }, + { + "epoch": 0.3269849785407725, + "grad_norm": 0.6941008839903752, + "learning_rate": 0.000157207303267955, + "loss": 2.5139, + "step": 3657 + }, + { + "epoch": 0.3270743919885551, + "grad_norm": 0.68817501331846, + "learning_rate": 0.00015718354770993817, + "loss": 2.9116, + "step": 3658 + }, + { + "epoch": 0.3271638054363376, + "grad_norm": 0.4481926118917088, + "learning_rate": 0.00015715978735600627, + "loss": 2.573, + "step": 3659 + }, + { + "epoch": 0.3272532188841202, + "grad_norm": 1.3176327426456467, + "learning_rate": 0.00015713602220815203, + "loss": 1.9576, + "step": 3660 + }, + { + "epoch": 0.3273426323319027, + "grad_norm": 0.9485492853839, + "learning_rate": 0.00015711225226836865, + "loss": 2.1047, + "step": 3661 + }, + { + "epoch": 0.32743204577968527, + "grad_norm": 0.616183484047012, + "learning_rate": 0.00015708847753864963, + "loss": 2.4908, + "step": 3662 + }, + { + "epoch": 0.32752145922746784, + "grad_norm": 0.7085327126148812, + "learning_rate": 0.0001570646980209889, + "loss": 2.0655, + "step": 3663 + }, + { + "epoch": 0.32761087267525035, + "grad_norm": 0.8199114343006428, + "learning_rate": 0.0001570409137173809, + "loss": 2.2703, + "step": 3664 + }, + { + "epoch": 0.3277002861230329, + "grad_norm": 0.7075941944850204, + "learning_rate": 0.00015701712462982037, + "loss": 2.1132, + "step": 3665 + }, + { + "epoch": 0.32778969957081544, + "grad_norm": 1.3005025611685408, + "learning_rate": 0.0001569933307603024, + "loss": 2.2207, + "step": 3666 + }, + { + "epoch": 0.327879113018598, + "grad_norm": 0.7979260146644583, + "learning_rate": 0.00015696953211082268, + "loss": 2.4425, + "step": 3667 + }, + { + "epoch": 0.3279685264663805, + "grad_norm": 0.5888440459046184, + "learning_rate": 0.00015694572868337706, + "loss": 2.3027, + "step": 3668 + }, + { + "epoch": 0.3280579399141631, + "grad_norm": 1.5096956086953086, + "learning_rate": 0.00015692192047996194, + "loss": 1.8574, + "step": 3669 + }, + { + "epoch": 0.3281473533619456, + "grad_norm": 1.4004578186733103, + "learning_rate": 0.00015689810750257413, + "loss": 2.2491, + "step": 3670 + }, + { + "epoch": 0.3282367668097282, + "grad_norm": 2.101329720072674, + "learning_rate": 0.00015687428975321078, + "loss": 2.5828, + "step": 3671 + }, + { + "epoch": 0.3283261802575107, + "grad_norm": 1.286412308427985, + "learning_rate": 0.00015685046723386937, + "loss": 2.2532, + "step": 3672 + }, + { + "epoch": 0.3284155937052933, + "grad_norm": 0.513748036068587, + "learning_rate": 0.00015682663994654795, + "loss": 2.4657, + "step": 3673 + }, + { + "epoch": 0.32850500715307585, + "grad_norm": 0.4360084738135119, + "learning_rate": 0.0001568028078932449, + "loss": 1.6765, + "step": 3674 + }, + { + "epoch": 0.32859442060085836, + "grad_norm": 0.7664080351247774, + "learning_rate": 0.00015677897107595892, + "loss": 2.722, + "step": 3675 + }, + { + "epoch": 0.32868383404864093, + "grad_norm": 1.0550730359750606, + "learning_rate": 0.0001567551294966892, + "loss": 2.1501, + "step": 3676 + }, + { + "epoch": 0.32877324749642345, + "grad_norm": 1.2637935869718455, + "learning_rate": 0.00015673128315743534, + "loss": 2.5862, + "step": 3677 + }, + { + "epoch": 0.328862660944206, + "grad_norm": 0.9894882152697456, + "learning_rate": 0.00015670743206019723, + "loss": 1.8552, + "step": 3678 + }, + { + "epoch": 0.32895207439198854, + "grad_norm": 1.6012208591533172, + "learning_rate": 0.00015668357620697533, + "loss": 2.3179, + "step": 3679 + }, + { + "epoch": 0.3290414878397711, + "grad_norm": 0.4713924170432133, + "learning_rate": 0.00015665971559977035, + "loss": 2.2319, + "step": 3680 + }, + { + "epoch": 0.3291309012875536, + "grad_norm": 0.6573586506559189, + "learning_rate": 0.00015663585024058342, + "loss": 2.3556, + "step": 3681 + }, + { + "epoch": 0.3292203147353362, + "grad_norm": 1.0416647770143292, + "learning_rate": 0.00015661198013141613, + "loss": 2.6515, + "step": 3682 + }, + { + "epoch": 0.32930972818311877, + "grad_norm": 0.3660498535859192, + "learning_rate": 0.00015658810527427046, + "loss": 2.6238, + "step": 3683 + }, + { + "epoch": 0.3293991416309013, + "grad_norm": 0.5763249698127224, + "learning_rate": 0.00015656422567114872, + "loss": 2.8534, + "step": 3684 + }, + { + "epoch": 0.32948855507868385, + "grad_norm": 0.7627050847405625, + "learning_rate": 0.0001565403413240537, + "loss": 2.0558, + "step": 3685 + }, + { + "epoch": 0.32957796852646637, + "grad_norm": 4.415839744865498, + "learning_rate": 0.00015651645223498854, + "loss": 7.9919, + "step": 3686 + }, + { + "epoch": 0.32966738197424894, + "grad_norm": 3.553126453620357, + "learning_rate": 0.00015649255840595675, + "loss": 2.8567, + "step": 3687 + }, + { + "epoch": 0.32975679542203146, + "grad_norm": 5.0259319607114294, + "learning_rate": 0.00015646865983896238, + "loss": 8.0414, + "step": 3688 + }, + { + "epoch": 0.32984620886981403, + "grad_norm": 0.6784153346735604, + "learning_rate": 0.00015644475653600964, + "loss": 2.3908, + "step": 3689 + }, + { + "epoch": 0.32993562231759654, + "grad_norm": 2.544521005666286, + "learning_rate": 0.00015642084849910336, + "loss": 2.3704, + "step": 3690 + }, + { + "epoch": 0.3300250357653791, + "grad_norm": 0.9558962426093142, + "learning_rate": 0.00015639693573024865, + "loss": 2.1125, + "step": 3691 + }, + { + "epoch": 0.3301144492131617, + "grad_norm": 2.3155557947843013, + "learning_rate": 0.0001563730182314511, + "loss": 2.5812, + "step": 3692 + }, + { + "epoch": 0.3302038626609442, + "grad_norm": 1.7753512561253793, + "learning_rate": 0.0001563490960047165, + "loss": 2.1774, + "step": 3693 + }, + { + "epoch": 0.3302932761087268, + "grad_norm": 2.8791272224611815, + "learning_rate": 0.00015632516905205135, + "loss": 2.4384, + "step": 3694 + }, + { + "epoch": 0.3303826895565093, + "grad_norm": 0.5717326974127993, + "learning_rate": 0.00015630123737546224, + "loss": 2.1665, + "step": 3695 + }, + { + "epoch": 0.33047210300429186, + "grad_norm": 0.5384240043817082, + "learning_rate": 0.00015627730097695638, + "loss": 2.4292, + "step": 3696 + }, + { + "epoch": 0.3305615164520744, + "grad_norm": 0.729819608526611, + "learning_rate": 0.00015625335985854126, + "loss": 2.1395, + "step": 3697 + }, + { + "epoch": 0.33065092989985695, + "grad_norm": 1.3959669469630205, + "learning_rate": 0.00015622941402222479, + "loss": 2.2458, + "step": 3698 + }, + { + "epoch": 0.33074034334763946, + "grad_norm": 3.196662968560822, + "learning_rate": 0.00015620546347001524, + "loss": 2.3499, + "step": 3699 + }, + { + "epoch": 0.33082975679542204, + "grad_norm": 3.7534390442910035, + "learning_rate": 0.00015618150820392136, + "loss": 2.2503, + "step": 3700 + }, + { + "epoch": 0.33091917024320455, + "grad_norm": 1.66363214222579, + "learning_rate": 0.00015615754822595224, + "loss": 1.8476, + "step": 3701 + }, + { + "epoch": 0.3310085836909871, + "grad_norm": 1.2263362137958662, + "learning_rate": 0.00015613358353811738, + "loss": 2.1695, + "step": 3702 + }, + { + "epoch": 0.3310979971387697, + "grad_norm": 1.2615462023748587, + "learning_rate": 0.00015610961414242664, + "loss": 2.2119, + "step": 3703 + }, + { + "epoch": 0.3311874105865522, + "grad_norm": 0.3885307534071092, + "learning_rate": 0.00015608564004089033, + "loss": 2.1746, + "step": 3704 + }, + { + "epoch": 0.3312768240343348, + "grad_norm": 0.9586159119422106, + "learning_rate": 0.00015606166123551912, + "loss": 2.3701, + "step": 3705 + }, + { + "epoch": 0.3313662374821173, + "grad_norm": 0.3245221467893126, + "learning_rate": 0.00015603767772832413, + "loss": 1.7564, + "step": 3706 + }, + { + "epoch": 0.33145565092989987, + "grad_norm": 0.7687609143820383, + "learning_rate": 0.0001560136895213167, + "loss": 2.0341, + "step": 3707 + }, + { + "epoch": 0.3315450643776824, + "grad_norm": 1.0896233127026353, + "learning_rate": 0.00015598969661650888, + "loss": 2.7008, + "step": 3708 + }, + { + "epoch": 0.33163447782546496, + "grad_norm": 0.4730508521875635, + "learning_rate": 0.00015596569901591277, + "loss": 1.679, + "step": 3709 + }, + { + "epoch": 0.3317238912732475, + "grad_norm": 1.15875259715131, + "learning_rate": 0.00015594169672154107, + "loss": 1.5838, + "step": 3710 + }, + { + "epoch": 0.33181330472103004, + "grad_norm": 1.6955932912461882, + "learning_rate": 0.00015591768973540683, + "loss": 1.9436, + "step": 3711 + }, + { + "epoch": 0.3319027181688126, + "grad_norm": 0.36101258497944694, + "learning_rate": 0.00015589367805952348, + "loss": 1.8788, + "step": 3712 + }, + { + "epoch": 0.33199213161659513, + "grad_norm": 0.8047759538735507, + "learning_rate": 0.00015586966169590488, + "loss": 2.3694, + "step": 3713 + }, + { + "epoch": 0.3320815450643777, + "grad_norm": 1.6389089460681203, + "learning_rate": 0.0001558456406465652, + "loss": 1.8665, + "step": 3714 + }, + { + "epoch": 0.3321709585121602, + "grad_norm": 5.493684353631802, + "learning_rate": 0.00015582161491351908, + "loss": 7.7531, + "step": 3715 + }, + { + "epoch": 0.3322603719599428, + "grad_norm": 4.889257160849119, + "learning_rate": 0.00015579758449878157, + "loss": 2.6607, + "step": 3716 + }, + { + "epoch": 0.3323497854077253, + "grad_norm": 1.1327906647458672, + "learning_rate": 0.000155773549404368, + "loss": 2.2341, + "step": 3717 + }, + { + "epoch": 0.3324391988555079, + "grad_norm": 2.0252355212566187, + "learning_rate": 0.00015574950963229419, + "loss": 2.0437, + "step": 3718 + }, + { + "epoch": 0.3325286123032904, + "grad_norm": 2.266988474073484, + "learning_rate": 0.00015572546518457636, + "loss": 2.6115, + "step": 3719 + }, + { + "epoch": 0.33261802575107297, + "grad_norm": 1.2931584559247706, + "learning_rate": 0.00015570141606323105, + "loss": 2.4582, + "step": 3720 + }, + { + "epoch": 0.3327074391988555, + "grad_norm": 1.7198714062123905, + "learning_rate": 0.00015567736227027525, + "loss": 2.8416, + "step": 3721 + }, + { + "epoch": 0.33279685264663805, + "grad_norm": 1.510614892317902, + "learning_rate": 0.00015565330380772633, + "loss": 1.8256, + "step": 3722 + }, + { + "epoch": 0.3328862660944206, + "grad_norm": 3.0209799670430466, + "learning_rate": 0.00015562924067760202, + "loss": 2.8134, + "step": 3723 + }, + { + "epoch": 0.33297567954220314, + "grad_norm": 1.1364380104901977, + "learning_rate": 0.00015560517288192046, + "loss": 2.3074, + "step": 3724 + }, + { + "epoch": 0.3330650929899857, + "grad_norm": 2.06958371221048, + "learning_rate": 0.00015558110042270023, + "loss": 1.7963, + "step": 3725 + }, + { + "epoch": 0.3331545064377682, + "grad_norm": 1.0154534042338057, + "learning_rate": 0.00015555702330196023, + "loss": 1.9351, + "step": 3726 + }, + { + "epoch": 0.3332439198855508, + "grad_norm": 0.5584619455940021, + "learning_rate": 0.00015553294152171977, + "loss": 2.1063, + "step": 3727 + }, + { + "epoch": 0.3333333333333333, + "grad_norm": 1.991136329418894, + "learning_rate": 0.00015550885508399856, + "loss": 2.3076, + "step": 3728 + }, + { + "epoch": 0.3334227467811159, + "grad_norm": 0.3179754808955083, + "learning_rate": 0.00015548476399081674, + "loss": 1.9675, + "step": 3729 + }, + { + "epoch": 0.3335121602288984, + "grad_norm": 1.795719793132848, + "learning_rate": 0.0001554606682441948, + "loss": 2.1505, + "step": 3730 + }, + { + "epoch": 0.333601573676681, + "grad_norm": 0.9220508839579649, + "learning_rate": 0.00015543656784615354, + "loss": 2.1396, + "step": 3731 + }, + { + "epoch": 0.33369098712446355, + "grad_norm": 7.605070520683048, + "learning_rate": 0.00015541246279871432, + "loss": 7.7617, + "step": 3732 + }, + { + "epoch": 0.33378040057224606, + "grad_norm": 17.85424185204044, + "learning_rate": 0.00015538835310389875, + "loss": 2.7183, + "step": 3733 + }, + { + "epoch": 0.33386981402002863, + "grad_norm": 3.97595675223657, + "learning_rate": 0.00015536423876372888, + "loss": 7.7361, + "step": 3734 + }, + { + "epoch": 0.33395922746781115, + "grad_norm": 28.606767557892855, + "learning_rate": 0.00015534011978022717, + "loss": 2.7069, + "step": 3735 + }, + { + "epoch": 0.3340486409155937, + "grad_norm": 19.015193883702644, + "learning_rate": 0.00015531599615541648, + "loss": 2.0131, + "step": 3736 + }, + { + "epoch": 0.33413805436337624, + "grad_norm": 9.200463236242701, + "learning_rate": 0.00015529186789131996, + "loss": 2.3141, + "step": 3737 + }, + { + "epoch": 0.3342274678111588, + "grad_norm": 7.691332725977047, + "learning_rate": 0.0001552677349899613, + "loss": 2.7319, + "step": 3738 + }, + { + "epoch": 0.3343168812589413, + "grad_norm": 5.902765618918872, + "learning_rate": 0.0001552435974533644, + "loss": 2.6711, + "step": 3739 + }, + { + "epoch": 0.3344062947067239, + "grad_norm": 5.170952623106583, + "learning_rate": 0.00015521945528355376, + "loss": 2.3891, + "step": 3740 + }, + { + "epoch": 0.3344957081545064, + "grad_norm": 6.417167997086952, + "learning_rate": 0.00015519530848255407, + "loss": 7.5417, + "step": 3741 + }, + { + "epoch": 0.334585121602289, + "grad_norm": 4.366001633740606, + "learning_rate": 0.00015517115705239047, + "loss": 2.1336, + "step": 3742 + }, + { + "epoch": 0.33467453505007155, + "grad_norm": 2.211335449774863, + "learning_rate": 0.0001551470009950886, + "loss": 2.0036, + "step": 3743 + }, + { + "epoch": 0.33476394849785407, + "grad_norm": 6.330714820436823, + "learning_rate": 0.00015512284031267437, + "loss": 2.818, + "step": 3744 + }, + { + "epoch": 0.33485336194563664, + "grad_norm": 3.045352380623487, + "learning_rate": 0.00015509867500717407, + "loss": 2.0833, + "step": 3745 + }, + { + "epoch": 0.33494277539341916, + "grad_norm": 0.6946788801813353, + "learning_rate": 0.00015507450508061443, + "loss": 1.8871, + "step": 3746 + }, + { + "epoch": 0.33503218884120173, + "grad_norm": 1.1677339197981582, + "learning_rate": 0.0001550503305350226, + "loss": 2.3609, + "step": 3747 + }, + { + "epoch": 0.33512160228898424, + "grad_norm": 3.0665623429478175, + "learning_rate": 0.000155026151372426, + "loss": 2.367, + "step": 3748 + }, + { + "epoch": 0.3352110157367668, + "grad_norm": 3.130805235085462, + "learning_rate": 0.00015500196759485254, + "loss": 2.7839, + "step": 3749 + }, + { + "epoch": 0.33530042918454933, + "grad_norm": 1.9429031339451122, + "learning_rate": 0.0001549777792043305, + "loss": 2.399, + "step": 3750 + }, + { + "epoch": 0.3353898426323319, + "grad_norm": 1.9726579601713359, + "learning_rate": 0.0001549535862028885, + "loss": 2.5027, + "step": 3751 + }, + { + "epoch": 0.3354792560801145, + "grad_norm": 1.8579002082156464, + "learning_rate": 0.0001549293885925556, + "loss": 2.8044, + "step": 3752 + }, + { + "epoch": 0.335568669527897, + "grad_norm": 3.4490277360947177, + "learning_rate": 0.0001549051863753612, + "loss": 2.0532, + "step": 3753 + }, + { + "epoch": 0.33565808297567956, + "grad_norm": 0.696036140651303, + "learning_rate": 0.00015488097955333515, + "loss": 1.5526, + "step": 3754 + }, + { + "epoch": 0.3357474964234621, + "grad_norm": 1.470315684690488, + "learning_rate": 0.00015485676812850761, + "loss": 2.0469, + "step": 3755 + }, + { + "epoch": 0.33583690987124465, + "grad_norm": 0.6915216220129974, + "learning_rate": 0.0001548325521029092, + "loss": 2.2727, + "step": 3756 + }, + { + "epoch": 0.33592632331902716, + "grad_norm": 2.362570998492373, + "learning_rate": 0.00015480833147857087, + "loss": 2.2011, + "step": 3757 + }, + { + "epoch": 0.33601573676680974, + "grad_norm": 0.9180934919941607, + "learning_rate": 0.00015478410625752393, + "loss": 2.3098, + "step": 3758 + }, + { + "epoch": 0.33610515021459225, + "grad_norm": 1.504416656687682, + "learning_rate": 0.00015475987644180016, + "loss": 2.1026, + "step": 3759 + }, + { + "epoch": 0.3361945636623748, + "grad_norm": 2.12792347227035, + "learning_rate": 0.00015473564203343174, + "loss": 2.4949, + "step": 3760 + }, + { + "epoch": 0.3362839771101574, + "grad_norm": 1.0967315326708338, + "learning_rate": 0.00015471140303445109, + "loss": 2.5717, + "step": 3761 + }, + { + "epoch": 0.3363733905579399, + "grad_norm": 2.478731148650882, + "learning_rate": 0.00015468715944689113, + "loss": 2.7363, + "step": 3762 + }, + { + "epoch": 0.3364628040057225, + "grad_norm": 1.3871734160824356, + "learning_rate": 0.0001546629112727852, + "loss": 1.7607, + "step": 3763 + }, + { + "epoch": 0.336552217453505, + "grad_norm": 1.8179984470197719, + "learning_rate": 0.00015463865851416685, + "loss": 1.9313, + "step": 3764 + }, + { + "epoch": 0.33664163090128757, + "grad_norm": 1.8585646109914322, + "learning_rate": 0.00015461440117307026, + "loss": 2.332, + "step": 3765 + }, + { + "epoch": 0.3367310443490701, + "grad_norm": 1.4658472490722814, + "learning_rate": 0.00015459013925152976, + "loss": 2.3842, + "step": 3766 + }, + { + "epoch": 0.33682045779685266, + "grad_norm": 1.4466134280775365, + "learning_rate": 0.00015456587275158024, + "loss": 2.1676, + "step": 3767 + }, + { + "epoch": 0.3369098712446352, + "grad_norm": 1.5025853368283466, + "learning_rate": 0.00015454160167525685, + "loss": 2.6169, + "step": 3768 + }, + { + "epoch": 0.33699928469241774, + "grad_norm": 1.1247203545013054, + "learning_rate": 0.00015451732602459522, + "loss": 2.7915, + "step": 3769 + }, + { + "epoch": 0.33708869814020026, + "grad_norm": 1.2989934514253052, + "learning_rate": 0.00015449304580163125, + "loss": 2.2158, + "step": 3770 + }, + { + "epoch": 0.33717811158798283, + "grad_norm": 1.2165425256673525, + "learning_rate": 0.00015446876100840137, + "loss": 2.1183, + "step": 3771 + }, + { + "epoch": 0.3372675250357654, + "grad_norm": 1.032890177263354, + "learning_rate": 0.0001544444716469423, + "loss": 2.3233, + "step": 3772 + }, + { + "epoch": 0.3373569384835479, + "grad_norm": 1.2881225805555951, + "learning_rate": 0.0001544201777192911, + "loss": 2.3132, + "step": 3773 + }, + { + "epoch": 0.3374463519313305, + "grad_norm": 0.5360801010396243, + "learning_rate": 0.00015439587922748537, + "loss": 1.8165, + "step": 3774 + }, + { + "epoch": 0.337535765379113, + "grad_norm": 0.39802279420517156, + "learning_rate": 0.00015437157617356292, + "loss": 2.1894, + "step": 3775 + }, + { + "epoch": 0.3376251788268956, + "grad_norm": 0.801371827393199, + "learning_rate": 0.00015434726855956206, + "loss": 1.8329, + "step": 3776 + }, + { + "epoch": 0.3377145922746781, + "grad_norm": 1.9799631477470057, + "learning_rate": 0.0001543229563875214, + "loss": 2.3453, + "step": 3777 + }, + { + "epoch": 0.33780400572246067, + "grad_norm": 0.5745400508551144, + "learning_rate": 0.00015429863965947996, + "loss": 1.9651, + "step": 3778 + }, + { + "epoch": 0.3378934191702432, + "grad_norm": 0.9104791001783941, + "learning_rate": 0.00015427431837747725, + "loss": 2.183, + "step": 3779 + }, + { + "epoch": 0.33798283261802575, + "grad_norm": 3.8244095829966014, + "learning_rate": 0.00015424999254355296, + "loss": 7.8938, + "step": 3780 + }, + { + "epoch": 0.3380722460658083, + "grad_norm": 2.210957483041519, + "learning_rate": 0.00015422566215974733, + "loss": 2.4276, + "step": 3781 + }, + { + "epoch": 0.33816165951359084, + "grad_norm": 2.8388190420785286, + "learning_rate": 0.00015420132722810092, + "loss": 2.7449, + "step": 3782 + }, + { + "epoch": 0.3382510729613734, + "grad_norm": 3.2397833429992144, + "learning_rate": 0.00015417698775065466, + "loss": 2.4961, + "step": 3783 + }, + { + "epoch": 0.3383404864091559, + "grad_norm": 1.2026616278112936, + "learning_rate": 0.00015415264372944983, + "loss": 2.4528, + "step": 3784 + }, + { + "epoch": 0.3384298998569385, + "grad_norm": 2.6642505967269385, + "learning_rate": 0.00015412829516652817, + "loss": 1.9841, + "step": 3785 + }, + { + "epoch": 0.338519313304721, + "grad_norm": 0.9478607163693157, + "learning_rate": 0.00015410394206393177, + "loss": 2.3247, + "step": 3786 + }, + { + "epoch": 0.3386087267525036, + "grad_norm": 0.7618146880965596, + "learning_rate": 0.00015407958442370312, + "loss": 1.9158, + "step": 3787 + }, + { + "epoch": 0.3386981402002861, + "grad_norm": 1.6910979168986284, + "learning_rate": 0.000154055222247885, + "loss": 2.0035, + "step": 3788 + }, + { + "epoch": 0.3387875536480687, + "grad_norm": 3.820461115035451, + "learning_rate": 0.00015403085553852068, + "loss": 2.1151, + "step": 3789 + }, + { + "epoch": 0.3388769670958512, + "grad_norm": 1.8953359587751504, + "learning_rate": 0.00015400648429765375, + "loss": 1.8867, + "step": 3790 + }, + { + "epoch": 0.33896638054363376, + "grad_norm": 2.4051260239214756, + "learning_rate": 0.00015398210852732825, + "loss": 2.1708, + "step": 3791 + }, + { + "epoch": 0.33905579399141633, + "grad_norm": 2.4719354784241157, + "learning_rate": 0.00015395772822958845, + "loss": 2.1843, + "step": 3792 + }, + { + "epoch": 0.33914520743919885, + "grad_norm": 2.5288338399879398, + "learning_rate": 0.00015393334340647917, + "loss": 7.6447, + "step": 3793 + }, + { + "epoch": 0.3392346208869814, + "grad_norm": 3.8331842626604247, + "learning_rate": 0.00015390895406004553, + "loss": 2.3502, + "step": 3794 + }, + { + "epoch": 0.33932403433476394, + "grad_norm": 2.7372096413970013, + "learning_rate": 0.00015388456019233302, + "loss": 2.2435, + "step": 3795 + }, + { + "epoch": 0.3394134477825465, + "grad_norm": 0.832186866212809, + "learning_rate": 0.0001538601618053875, + "loss": 2.2245, + "step": 3796 + }, + { + "epoch": 0.339502861230329, + "grad_norm": 4.173073170548899, + "learning_rate": 0.00015383575890125527, + "loss": 2.1601, + "step": 3797 + }, + { + "epoch": 0.3395922746781116, + "grad_norm": 2.1851474845128585, + "learning_rate": 0.00015381135148198293, + "loss": 2.2906, + "step": 3798 + }, + { + "epoch": 0.3396816881258941, + "grad_norm": 1.1247750774427425, + "learning_rate": 0.00015378693954961754, + "loss": 1.8776, + "step": 3799 + }, + { + "epoch": 0.3397711015736767, + "grad_norm": 2.4027806479617775, + "learning_rate": 0.0001537625231062065, + "loss": 2.1361, + "step": 3800 + }, + { + "epoch": 0.33986051502145925, + "grad_norm": 2.383349431253343, + "learning_rate": 0.00015373810215379757, + "loss": 2.5396, + "step": 3801 + }, + { + "epoch": 0.33994992846924177, + "grad_norm": 1.072390528064676, + "learning_rate": 0.0001537136766944389, + "loss": 1.9267, + "step": 3802 + }, + { + "epoch": 0.34003934191702434, + "grad_norm": 1.8453376755509217, + "learning_rate": 0.00015368924673017905, + "loss": 1.9822, + "step": 3803 + }, + { + "epoch": 0.34012875536480686, + "grad_norm": 0.9138412546874749, + "learning_rate": 0.00015366481226306692, + "loss": 2.2341, + "step": 3804 + }, + { + "epoch": 0.3402181688125894, + "grad_norm": 0.8725485423013115, + "learning_rate": 0.00015364037329515182, + "loss": 2.4044, + "step": 3805 + }, + { + "epoch": 0.34030758226037194, + "grad_norm": 2.5986511762272384, + "learning_rate": 0.00015361592982848335, + "loss": 2.3909, + "step": 3806 + }, + { + "epoch": 0.3403969957081545, + "grad_norm": 0.36433686209470023, + "learning_rate": 0.00015359148186511163, + "loss": 2.4072, + "step": 3807 + }, + { + "epoch": 0.34048640915593703, + "grad_norm": 0.8431666087835222, + "learning_rate": 0.000153567029407087, + "loss": 2.7754, + "step": 3808 + }, + { + "epoch": 0.3405758226037196, + "grad_norm": 16.892694766750697, + "learning_rate": 0.00015354257245646036, + "loss": 7.8659, + "step": 3809 + }, + { + "epoch": 0.3406652360515021, + "grad_norm": 18.574645227116598, + "learning_rate": 0.0001535181110152828, + "loss": 2.7337, + "step": 3810 + }, + { + "epoch": 0.3407546494992847, + "grad_norm": 5.583536214921084, + "learning_rate": 0.00015349364508560588, + "loss": 2.2734, + "step": 3811 + }, + { + "epoch": 0.34084406294706726, + "grad_norm": 4.262828816340249, + "learning_rate": 0.00015346917466948161, + "loss": 2.3229, + "step": 3812 + }, + { + "epoch": 0.3409334763948498, + "grad_norm": 18.01015388331853, + "learning_rate": 0.0001534446997689622, + "loss": 2.6809, + "step": 3813 + }, + { + "epoch": 0.34102288984263235, + "grad_norm": 3.451793070603724, + "learning_rate": 0.00015342022038610038, + "loss": 2.3368, + "step": 3814 + }, + { + "epoch": 0.34111230329041486, + "grad_norm": 3.99580469952151, + "learning_rate": 0.00015339573652294917, + "loss": 2.7984, + "step": 3815 + }, + { + "epoch": 0.34120171673819744, + "grad_norm": 5.75669172259856, + "learning_rate": 0.00015337124818156205, + "loss": 1.9591, + "step": 3816 + }, + { + "epoch": 0.34129113018597995, + "grad_norm": 6.419004038209799, + "learning_rate": 0.00015334675536399277, + "loss": 2.3906, + "step": 3817 + }, + { + "epoch": 0.3413805436337625, + "grad_norm": 3.0818259274371904, + "learning_rate": 0.00015332225807229556, + "loss": 2.7934, + "step": 3818 + }, + { + "epoch": 0.34146995708154504, + "grad_norm": 1.9748601065612592, + "learning_rate": 0.00015329775630852497, + "loss": 2.3931, + "step": 3819 + }, + { + "epoch": 0.3415593705293276, + "grad_norm": 7.200843562670934, + "learning_rate": 0.00015327325007473592, + "loss": 2.3697, + "step": 3820 + }, + { + "epoch": 0.3416487839771102, + "grad_norm": 2.6970514541009685, + "learning_rate": 0.00015324873937298374, + "loss": 1.9073, + "step": 3821 + }, + { + "epoch": 0.3417381974248927, + "grad_norm": 1.6629531641694826, + "learning_rate": 0.00015322422420532407, + "loss": 2.3078, + "step": 3822 + }, + { + "epoch": 0.34182761087267527, + "grad_norm": 1.256204389587846, + "learning_rate": 0.000153199704573813, + "loss": 2.0024, + "step": 3823 + }, + { + "epoch": 0.3419170243204578, + "grad_norm": 1.8904532709939266, + "learning_rate": 0.00015317518048050697, + "loss": 2.4904, + "step": 3824 + }, + { + "epoch": 0.34200643776824036, + "grad_norm": 1.598844172594998, + "learning_rate": 0.00015315065192746276, + "loss": 2.2309, + "step": 3825 + }, + { + "epoch": 0.3420958512160229, + "grad_norm": 1.2199601491039018, + "learning_rate": 0.00015312611891673752, + "loss": 1.9917, + "step": 3826 + }, + { + "epoch": 0.34218526466380544, + "grad_norm": 0.5116472780362376, + "learning_rate": 0.00015310158145038892, + "loss": 1.9522, + "step": 3827 + }, + { + "epoch": 0.34227467811158796, + "grad_norm": 0.7538940205200667, + "learning_rate": 0.0001530770395304748, + "loss": 2.0122, + "step": 3828 + }, + { + "epoch": 0.34236409155937053, + "grad_norm": 0.8496245457367648, + "learning_rate": 0.00015305249315905348, + "loss": 2.7631, + "step": 3829 + }, + { + "epoch": 0.3424535050071531, + "grad_norm": 10.51754036385407, + "learning_rate": 0.0001530279423381836, + "loss": 7.9412, + "step": 3830 + }, + { + "epoch": 0.3425429184549356, + "grad_norm": 3.113740080662465, + "learning_rate": 0.00015300338706992426, + "loss": 2.1327, + "step": 3831 + }, + { + "epoch": 0.3426323319027182, + "grad_norm": 8.637984918377132, + "learning_rate": 0.00015297882735633485, + "loss": 2.5732, + "step": 3832 + }, + { + "epoch": 0.3427217453505007, + "grad_norm": 2.9266680860737586, + "learning_rate": 0.00015295426319947514, + "loss": 2.2587, + "step": 3833 + }, + { + "epoch": 0.3428111587982833, + "grad_norm": 1.0273290605658554, + "learning_rate": 0.0001529296946014054, + "loss": 2.0641, + "step": 3834 + }, + { + "epoch": 0.3429005722460658, + "grad_norm": 2.5136663790041207, + "learning_rate": 0.00015290512156418602, + "loss": 2.0666, + "step": 3835 + }, + { + "epoch": 0.34298998569384836, + "grad_norm": 2.4243775107500696, + "learning_rate": 0.000152880544089878, + "loss": 2.1947, + "step": 3836 + }, + { + "epoch": 0.3430793991416309, + "grad_norm": 2.285105625381663, + "learning_rate": 0.00015285596218054265, + "loss": 2.5472, + "step": 3837 + }, + { + "epoch": 0.34316881258941345, + "grad_norm": 1.996419308082073, + "learning_rate": 0.00015283137583824158, + "loss": 2.4434, + "step": 3838 + }, + { + "epoch": 0.34325822603719597, + "grad_norm": 2.809090955807346, + "learning_rate": 0.0001528067850650368, + "loss": 2.0679, + "step": 3839 + }, + { + "epoch": 0.34334763948497854, + "grad_norm": 0.5833888052139382, + "learning_rate": 0.00015278218986299074, + "loss": 1.9161, + "step": 3840 + }, + { + "epoch": 0.3434370529327611, + "grad_norm": 1.9060655624819203, + "learning_rate": 0.00015275759023416618, + "loss": 2.1729, + "step": 3841 + }, + { + "epoch": 0.3435264663805436, + "grad_norm": 1.7293737443923198, + "learning_rate": 0.00015273298618062624, + "loss": 2.4737, + "step": 3842 + }, + { + "epoch": 0.3436158798283262, + "grad_norm": 2.011988390221331, + "learning_rate": 0.00015270837770443437, + "loss": 2.1631, + "step": 3843 + }, + { + "epoch": 0.3437052932761087, + "grad_norm": 0.8093158241314069, + "learning_rate": 0.0001526837648076546, + "loss": 2.3136, + "step": 3844 + }, + { + "epoch": 0.3437947067238913, + "grad_norm": 1.6565569401419837, + "learning_rate": 0.00015265914749235107, + "loss": 2.4175, + "step": 3845 + }, + { + "epoch": 0.3438841201716738, + "grad_norm": 1.193145489744912, + "learning_rate": 0.00015263452576058843, + "loss": 2.3176, + "step": 3846 + }, + { + "epoch": 0.3439735336194564, + "grad_norm": 2.0054298261925463, + "learning_rate": 0.0001526098996144317, + "loss": 1.846, + "step": 3847 + }, + { + "epoch": 0.3440629470672389, + "grad_norm": 4.808640928613871, + "learning_rate": 0.0001525852690559462, + "loss": 7.4903, + "step": 3848 + }, + { + "epoch": 0.34415236051502146, + "grad_norm": 2.6418165314757154, + "learning_rate": 0.00015256063408719772, + "loss": 2.5286, + "step": 3849 + }, + { + "epoch": 0.34424177396280403, + "grad_norm": 2.0584276524745118, + "learning_rate": 0.0001525359947102523, + "loss": 2.5905, + "step": 3850 + }, + { + "epoch": 0.34433118741058655, + "grad_norm": 1.6248713811629485, + "learning_rate": 0.00015251135092717648, + "loss": 2.1968, + "step": 3851 + }, + { + "epoch": 0.3444206008583691, + "grad_norm": 0.795001436588086, + "learning_rate": 0.00015248670274003708, + "loss": 2.6396, + "step": 3852 + }, + { + "epoch": 0.34451001430615164, + "grad_norm": 0.882506218259963, + "learning_rate": 0.00015246205015090127, + "loss": 2.0839, + "step": 3853 + }, + { + "epoch": 0.3445994277539342, + "grad_norm": 3.5004053224591085, + "learning_rate": 0.0001524373931618367, + "loss": 2.2929, + "step": 3854 + }, + { + "epoch": 0.3446888412017167, + "grad_norm": 1.173683120287808, + "learning_rate": 0.0001524127317749113, + "loss": 2.4684, + "step": 3855 + }, + { + "epoch": 0.3447782546494993, + "grad_norm": 1.404174751172951, + "learning_rate": 0.00015238806599219336, + "loss": 2.1556, + "step": 3856 + }, + { + "epoch": 0.3448676680972818, + "grad_norm": 1.4454163265937303, + "learning_rate": 0.0001523633958157516, + "loss": 2.8325, + "step": 3857 + }, + { + "epoch": 0.3449570815450644, + "grad_norm": 2.598568969354096, + "learning_rate": 0.00015233872124765512, + "loss": 2.4192, + "step": 3858 + }, + { + "epoch": 0.3450464949928469, + "grad_norm": 0.7662411215650726, + "learning_rate": 0.00015231404228997325, + "loss": 2.6109, + "step": 3859 + }, + { + "epoch": 0.34513590844062947, + "grad_norm": 0.9033247642477307, + "learning_rate": 0.00015228935894477582, + "loss": 2.1227, + "step": 3860 + }, + { + "epoch": 0.34522532188841204, + "grad_norm": 2.0472988209324505, + "learning_rate": 0.00015226467121413304, + "loss": 2.0553, + "step": 3861 + }, + { + "epoch": 0.34531473533619456, + "grad_norm": 1.0370072646745174, + "learning_rate": 0.0001522399791001154, + "loss": 2.3024, + "step": 3862 + }, + { + "epoch": 0.3454041487839771, + "grad_norm": 0.6685075072127765, + "learning_rate": 0.00015221528260479377, + "loss": 1.6086, + "step": 3863 + }, + { + "epoch": 0.34549356223175964, + "grad_norm": 1.1860402515102322, + "learning_rate": 0.0001521905817302395, + "loss": 2.2649, + "step": 3864 + }, + { + "epoch": 0.3455829756795422, + "grad_norm": 0.9659131723758074, + "learning_rate": 0.00015216587647852415, + "loss": 2.0411, + "step": 3865 + }, + { + "epoch": 0.34567238912732473, + "grad_norm": 1.8011530180625246, + "learning_rate": 0.0001521411668517197, + "loss": 3.0569, + "step": 3866 + }, + { + "epoch": 0.3457618025751073, + "grad_norm": 0.6010375410338117, + "learning_rate": 0.00015211645285189858, + "loss": 2.1438, + "step": 3867 + }, + { + "epoch": 0.3458512160228898, + "grad_norm": 1.134448813452294, + "learning_rate": 0.00015209173448113355, + "loss": 2.5341, + "step": 3868 + }, + { + "epoch": 0.3459406294706724, + "grad_norm": 0.6675563938815202, + "learning_rate": 0.0001520670117414976, + "loss": 2.1715, + "step": 3869 + }, + { + "epoch": 0.34603004291845496, + "grad_norm": 1.131694089087341, + "learning_rate": 0.00015204228463506424, + "loss": 2.2505, + "step": 3870 + }, + { + "epoch": 0.3461194563662375, + "grad_norm": 0.9696285485652893, + "learning_rate": 0.00015201755316390737, + "loss": 2.8544, + "step": 3871 + }, + { + "epoch": 0.34620886981402005, + "grad_norm": 1.0201360317002277, + "learning_rate": 0.00015199281733010116, + "loss": 2.6802, + "step": 3872 + }, + { + "epoch": 0.34629828326180256, + "grad_norm": 4.402077538400389, + "learning_rate": 0.0001519680771357201, + "loss": 7.592, + "step": 3873 + }, + { + "epoch": 0.34638769670958514, + "grad_norm": 1.6856962788727035, + "learning_rate": 0.00015194333258283918, + "loss": 2.4737, + "step": 3874 + }, + { + "epoch": 0.34647711015736765, + "grad_norm": 2.4013755555766583, + "learning_rate": 0.00015191858367353368, + "loss": 2.3174, + "step": 3875 + }, + { + "epoch": 0.3465665236051502, + "grad_norm": 2.778303702445335, + "learning_rate": 0.0001518938304098793, + "loss": 2.1188, + "step": 3876 + }, + { + "epoch": 0.34665593705293274, + "grad_norm": 1.7223650011036984, + "learning_rate": 0.00015186907279395202, + "loss": 2.3321, + "step": 3877 + }, + { + "epoch": 0.3467453505007153, + "grad_norm": 1.2631845711679532, + "learning_rate": 0.00015184431082782823, + "loss": 2.4847, + "step": 3878 + }, + { + "epoch": 0.3468347639484979, + "grad_norm": 0.9362424123037452, + "learning_rate": 0.00015181954451358473, + "loss": 2.412, + "step": 3879 + }, + { + "epoch": 0.3469241773962804, + "grad_norm": 2.4911940427351276, + "learning_rate": 0.0001517947738532986, + "loss": 1.8392, + "step": 3880 + }, + { + "epoch": 0.34701359084406297, + "grad_norm": 1.6508570767327289, + "learning_rate": 0.00015176999884904734, + "loss": 2.4387, + "step": 3881 + }, + { + "epoch": 0.3471030042918455, + "grad_norm": 1.1877885689439616, + "learning_rate": 0.0001517452195029088, + "loss": 2.2563, + "step": 3882 + }, + { + "epoch": 0.34719241773962806, + "grad_norm": 2.823174669810273, + "learning_rate": 0.00015172043581696118, + "loss": 2.2507, + "step": 3883 + }, + { + "epoch": 0.3472818311874106, + "grad_norm": 4.5533817582257425, + "learning_rate": 0.0001516956477932831, + "loss": 2.245, + "step": 3884 + }, + { + "epoch": 0.34737124463519314, + "grad_norm": 1.5023277231385639, + "learning_rate": 0.00015167085543395348, + "loss": 1.8471, + "step": 3885 + }, + { + "epoch": 0.34746065808297566, + "grad_norm": 3.022442929554159, + "learning_rate": 0.00015164605874105156, + "loss": 1.8028, + "step": 3886 + }, + { + "epoch": 0.34755007153075823, + "grad_norm": 3.520291744850785, + "learning_rate": 0.0001516212577166571, + "loss": 2.2171, + "step": 3887 + }, + { + "epoch": 0.34763948497854075, + "grad_norm": 1.179333289533881, + "learning_rate": 0.0001515964523628501, + "loss": 1.8084, + "step": 3888 + }, + { + "epoch": 0.3477288984263233, + "grad_norm": 2.4502300662849907, + "learning_rate": 0.00015157164268171097, + "loss": 2.27, + "step": 3889 + }, + { + "epoch": 0.3478183118741059, + "grad_norm": 4.273221313059897, + "learning_rate": 0.0001515468286753204, + "loss": 2.831, + "step": 3890 + }, + { + "epoch": 0.3479077253218884, + "grad_norm": 2.0032857628864438, + "learning_rate": 0.0001515220103457596, + "loss": 2.0938, + "step": 3891 + }, + { + "epoch": 0.347997138769671, + "grad_norm": 2.1410173107062342, + "learning_rate": 0.00015149718769511003, + "loss": 2.0458, + "step": 3892 + }, + { + "epoch": 0.3480865522174535, + "grad_norm": 2.1534074021797625, + "learning_rate": 0.00015147236072545348, + "loss": 2.1984, + "step": 3893 + }, + { + "epoch": 0.34817596566523606, + "grad_norm": 0.8029027559287655, + "learning_rate": 0.00015144752943887222, + "loss": 2.1415, + "step": 3894 + }, + { + "epoch": 0.3482653791130186, + "grad_norm": 2.325759295031453, + "learning_rate": 0.0001514226938374488, + "loss": 2.0716, + "step": 3895 + }, + { + "epoch": 0.34835479256080115, + "grad_norm": 0.8224884492572359, + "learning_rate": 0.00015139785392326616, + "loss": 2.1745, + "step": 3896 + }, + { + "epoch": 0.34844420600858367, + "grad_norm": 1.7919689559291374, + "learning_rate": 0.00015137300969840758, + "loss": 2.4173, + "step": 3897 + }, + { + "epoch": 0.34853361945636624, + "grad_norm": 2.3824984314948714, + "learning_rate": 0.0001513481611649567, + "loss": 2.2683, + "step": 3898 + }, + { + "epoch": 0.3486230329041488, + "grad_norm": 0.6500663178096967, + "learning_rate": 0.00015132330832499756, + "loss": 2.5733, + "step": 3899 + }, + { + "epoch": 0.3487124463519313, + "grad_norm": 1.4211885209881605, + "learning_rate": 0.00015129845118061453, + "loss": 2.0222, + "step": 3900 + }, + { + "epoch": 0.3488018597997139, + "grad_norm": 1.1676915150086549, + "learning_rate": 0.00015127358973389236, + "loss": 2.4683, + "step": 3901 + }, + { + "epoch": 0.3488912732474964, + "grad_norm": 1.0490746355965705, + "learning_rate": 0.00015124872398691617, + "loss": 2.1803, + "step": 3902 + }, + { + "epoch": 0.348980686695279, + "grad_norm": 0.9258108370736176, + "learning_rate": 0.00015122385394177135, + "loss": 2.9737, + "step": 3903 + }, + { + "epoch": 0.3490701001430615, + "grad_norm": 0.6512581149443394, + "learning_rate": 0.0001511989796005438, + "loss": 2.0047, + "step": 3904 + }, + { + "epoch": 0.3491595135908441, + "grad_norm": 1.669194756462639, + "learning_rate": 0.00015117410096531964, + "loss": 2.0877, + "step": 3905 + }, + { + "epoch": 0.3492489270386266, + "grad_norm": 1.1634796702564434, + "learning_rate": 0.00015114921803818546, + "loss": 1.6245, + "step": 3906 + }, + { + "epoch": 0.34933834048640916, + "grad_norm": 0.8432903475963925, + "learning_rate": 0.0001511243308212281, + "loss": 2.3876, + "step": 3907 + }, + { + "epoch": 0.3494277539341917, + "grad_norm": 0.5925096853605077, + "learning_rate": 0.00015109943931653486, + "loss": 2.055, + "step": 3908 + }, + { + "epoch": 0.34951716738197425, + "grad_norm": 1.6971793140122, + "learning_rate": 0.00015107454352619336, + "loss": 2.2159, + "step": 3909 + }, + { + "epoch": 0.3496065808297568, + "grad_norm": 0.9765305834524515, + "learning_rate": 0.00015104964345229158, + "loss": 2.1522, + "step": 3910 + }, + { + "epoch": 0.34969599427753933, + "grad_norm": 1.7906060736692029, + "learning_rate": 0.00015102473909691785, + "loss": 1.9775, + "step": 3911 + }, + { + "epoch": 0.3497854077253219, + "grad_norm": 0.6421828633748518, + "learning_rate": 0.0001509998304621609, + "loss": 1.9498, + "step": 3912 + }, + { + "epoch": 0.3498748211731044, + "grad_norm": 0.717353114802791, + "learning_rate": 0.00015097491755010974, + "loss": 1.9788, + "step": 3913 + }, + { + "epoch": 0.349964234620887, + "grad_norm": 1.4067620321794916, + "learning_rate": 0.0001509500003628538, + "loss": 2.1732, + "step": 3914 + }, + { + "epoch": 0.3500536480686695, + "grad_norm": 1.1866788266724775, + "learning_rate": 0.00015092507890248288, + "loss": 2.0171, + "step": 3915 + }, + { + "epoch": 0.3501430615164521, + "grad_norm": 0.7035344876115415, + "learning_rate": 0.0001509001531710871, + "loss": 2.1776, + "step": 3916 + }, + { + "epoch": 0.3502324749642346, + "grad_norm": 0.6042919541538117, + "learning_rate": 0.00015087522317075693, + "loss": 2.2072, + "step": 3917 + }, + { + "epoch": 0.35032188841201717, + "grad_norm": 1.903248326467396, + "learning_rate": 0.00015085028890358325, + "loss": 2.113, + "step": 3918 + }, + { + "epoch": 0.35041130185979974, + "grad_norm": 0.9342160401390807, + "learning_rate": 0.00015082535037165724, + "loss": 1.9908, + "step": 3919 + }, + { + "epoch": 0.35050071530758226, + "grad_norm": 1.4415694746884615, + "learning_rate": 0.00015080040757707046, + "loss": 2.6572, + "step": 3920 + }, + { + "epoch": 0.3505901287553648, + "grad_norm": 1.172290818664918, + "learning_rate": 0.0001507754605219149, + "loss": 1.9492, + "step": 3921 + }, + { + "epoch": 0.35067954220314734, + "grad_norm": 0.7619013990109681, + "learning_rate": 0.00015075050920828272, + "loss": 2.0565, + "step": 3922 + }, + { + "epoch": 0.3507689556509299, + "grad_norm": 1.3206720355129375, + "learning_rate": 0.00015072555363826665, + "loss": 2.4778, + "step": 3923 + }, + { + "epoch": 0.35085836909871243, + "grad_norm": 1.1923513619368553, + "learning_rate": 0.00015070059381395968, + "loss": 1.9559, + "step": 3924 + }, + { + "epoch": 0.350947782546495, + "grad_norm": 1.0778200642366689, + "learning_rate": 0.0001506756297374551, + "loss": 1.9999, + "step": 3925 + }, + { + "epoch": 0.3510371959942775, + "grad_norm": 0.8220761641029823, + "learning_rate": 0.00015065066141084667, + "loss": 2.2189, + "step": 3926 + }, + { + "epoch": 0.3511266094420601, + "grad_norm": 0.8717478083025805, + "learning_rate": 0.00015062568883622844, + "loss": 2.2855, + "step": 3927 + }, + { + "epoch": 0.3512160228898426, + "grad_norm": 1.2847178659330925, + "learning_rate": 0.00015060071201569486, + "loss": 2.4992, + "step": 3928 + }, + { + "epoch": 0.3513054363376252, + "grad_norm": 1.668020849711783, + "learning_rate": 0.00015057573095134062, + "loss": 2.8482, + "step": 3929 + }, + { + "epoch": 0.35139484978540775, + "grad_norm": 2.4322046230940213, + "learning_rate": 0.00015055074564526095, + "loss": 2.3884, + "step": 3930 + }, + { + "epoch": 0.35148426323319026, + "grad_norm": 0.6866083254973401, + "learning_rate": 0.00015052575609955125, + "loss": 2.1276, + "step": 3931 + }, + { + "epoch": 0.35157367668097284, + "grad_norm": 5.608845841393769, + "learning_rate": 0.00015050076231630744, + "loss": 3.2868, + "step": 3932 + }, + { + "epoch": 0.35166309012875535, + "grad_norm": 4.718532657969048, + "learning_rate": 0.00015047576429762566, + "loss": 7.5924, + "step": 3933 + }, + { + "epoch": 0.3517525035765379, + "grad_norm": 1.1734076300125083, + "learning_rate": 0.0001504507620456025, + "loss": 1.5721, + "step": 3934 + }, + { + "epoch": 0.35184191702432044, + "grad_norm": 1.1346500290452681, + "learning_rate": 0.00015042575556233488, + "loss": 1.9, + "step": 3935 + }, + { + "epoch": 0.351931330472103, + "grad_norm": 1.074308488864458, + "learning_rate": 0.00015040074484992, + "loss": 2.2424, + "step": 3936 + }, + { + "epoch": 0.3520207439198855, + "grad_norm": 0.9576217172535404, + "learning_rate": 0.00015037572991045552, + "loss": 2.1175, + "step": 3937 + }, + { + "epoch": 0.3521101573676681, + "grad_norm": 3.0218068495105226, + "learning_rate": 0.00015035071074603944, + "loss": 2.069, + "step": 3938 + }, + { + "epoch": 0.35219957081545067, + "grad_norm": 0.7886702372922914, + "learning_rate": 0.00015032568735877003, + "loss": 2.5159, + "step": 3939 + }, + { + "epoch": 0.3522889842632332, + "grad_norm": 0.8979688005936185, + "learning_rate": 0.000150300659750746, + "loss": 2.0227, + "step": 3940 + }, + { + "epoch": 0.35237839771101576, + "grad_norm": 2.9550607567755254, + "learning_rate": 0.00015027562792406643, + "loss": 2.1095, + "step": 3941 + }, + { + "epoch": 0.35246781115879827, + "grad_norm": 0.948160694921976, + "learning_rate": 0.0001502505918808306, + "loss": 1.7656, + "step": 3942 + }, + { + "epoch": 0.35255722460658084, + "grad_norm": 0.7964421175346575, + "learning_rate": 0.00015022555162313834, + "loss": 2.2552, + "step": 3943 + }, + { + "epoch": 0.35264663805436336, + "grad_norm": 1.0414668570389434, + "learning_rate": 0.00015020050715308972, + "loss": 1.9824, + "step": 3944 + }, + { + "epoch": 0.35273605150214593, + "grad_norm": 1.3470123456334835, + "learning_rate": 0.0001501754584727852, + "loss": 1.9206, + "step": 3945 + }, + { + "epoch": 0.35282546494992845, + "grad_norm": 3.730578706351914, + "learning_rate": 0.0001501504055843256, + "loss": 2.7458, + "step": 3946 + }, + { + "epoch": 0.352914878397711, + "grad_norm": 1.1874300469438324, + "learning_rate": 0.00015012534848981202, + "loss": 2.2993, + "step": 3947 + }, + { + "epoch": 0.3530042918454936, + "grad_norm": 1.6846867098638276, + "learning_rate": 0.000150100287191346, + "loss": 2.7248, + "step": 3948 + }, + { + "epoch": 0.3530937052932761, + "grad_norm": 4.94355413528653, + "learning_rate": 0.00015007522169102941, + "loss": 7.4309, + "step": 3949 + }, + { + "epoch": 0.3531831187410587, + "grad_norm": 2.681380541965447, + "learning_rate": 0.00015005015199096443, + "loss": 2.3239, + "step": 3950 + }, + { + "epoch": 0.3532725321888412, + "grad_norm": 2.4737338736303376, + "learning_rate": 0.00015002507809325365, + "loss": 2.6767, + "step": 3951 + }, + { + "epoch": 0.35336194563662376, + "grad_norm": 0.9565939331640486, + "learning_rate": 0.00015000000000000001, + "loss": 1.7688, + "step": 3952 + }, + { + "epoch": 0.3534513590844063, + "grad_norm": 3.866933005103597, + "learning_rate": 0.00014997491771330675, + "loss": 2.5526, + "step": 3953 + }, + { + "epoch": 0.35354077253218885, + "grad_norm": 2.716295772176822, + "learning_rate": 0.0001499498312352775, + "loss": 2.2945, + "step": 3954 + }, + { + "epoch": 0.35363018597997137, + "grad_norm": 0.9845779253889771, + "learning_rate": 0.0001499247405680162, + "loss": 2.4059, + "step": 3955 + }, + { + "epoch": 0.35371959942775394, + "grad_norm": 2.803201965514544, + "learning_rate": 0.00014989964571362723, + "loss": 2.6503, + "step": 3956 + }, + { + "epoch": 0.35380901287553645, + "grad_norm": 1.6288918803017987, + "learning_rate": 0.0001498745466742152, + "loss": 1.9242, + "step": 3957 + }, + { + "epoch": 0.353898426323319, + "grad_norm": 1.1587713090140495, + "learning_rate": 0.0001498494434518852, + "loss": 2.1011, + "step": 3958 + }, + { + "epoch": 0.3539878397711016, + "grad_norm": 1.177812299461689, + "learning_rate": 0.0001498243360487426, + "loss": 2.1907, + "step": 3959 + }, + { + "epoch": 0.3540772532188841, + "grad_norm": 1.6004834875344038, + "learning_rate": 0.00014979922446689306, + "loss": 2.2909, + "step": 3960 + }, + { + "epoch": 0.3541666666666667, + "grad_norm": 1.2109977894913848, + "learning_rate": 0.00014977410870844273, + "loss": 2.0891, + "step": 3961 + }, + { + "epoch": 0.3542560801144492, + "grad_norm": 1.2303119200279065, + "learning_rate": 0.00014974898877549806, + "loss": 1.6621, + "step": 3962 + }, + { + "epoch": 0.3543454935622318, + "grad_norm": 8.68555863824782, + "learning_rate": 0.0001497238646701657, + "loss": 7.5393, + "step": 3963 + }, + { + "epoch": 0.3544349070100143, + "grad_norm": 2.7953748025693734, + "learning_rate": 0.00014969873639455292, + "loss": 2.1399, + "step": 3964 + }, + { + "epoch": 0.35452432045779686, + "grad_norm": 5.970593512691094, + "learning_rate": 0.00014967360395076713, + "loss": 2.5218, + "step": 3965 + }, + { + "epoch": 0.3546137339055794, + "grad_norm": 1.837104577984149, + "learning_rate": 0.00014964846734091616, + "loss": 2.1769, + "step": 3966 + }, + { + "epoch": 0.35470314735336195, + "grad_norm": 1.265471280964002, + "learning_rate": 0.00014962332656710817, + "loss": 2.0183, + "step": 3967 + }, + { + "epoch": 0.3547925608011445, + "grad_norm": 2.437654203898232, + "learning_rate": 0.00014959818163145174, + "loss": 1.9232, + "step": 3968 + }, + { + "epoch": 0.35488197424892703, + "grad_norm": 1.5662578147483426, + "learning_rate": 0.00014957303253605573, + "loss": 2.0555, + "step": 3969 + }, + { + "epoch": 0.3549713876967096, + "grad_norm": 2.1829725113088387, + "learning_rate": 0.00014954787928302935, + "loss": 2.1342, + "step": 3970 + }, + { + "epoch": 0.3550608011444921, + "grad_norm": 1.023762625206508, + "learning_rate": 0.00014952272187448214, + "loss": 2.5497, + "step": 3971 + }, + { + "epoch": 0.3551502145922747, + "grad_norm": 4.59237214313226, + "learning_rate": 0.00014949756031252406, + "loss": 2.6326, + "step": 3972 + }, + { + "epoch": 0.3552396280400572, + "grad_norm": 1.1837174003618713, + "learning_rate": 0.0001494723945992654, + "loss": 2.4534, + "step": 3973 + }, + { + "epoch": 0.3553290414878398, + "grad_norm": 0.6155719245434195, + "learning_rate": 0.00014944722473681673, + "loss": 2.3841, + "step": 3974 + }, + { + "epoch": 0.3554184549356223, + "grad_norm": 0.8520520127418193, + "learning_rate": 0.00014942205072728903, + "loss": 2.0079, + "step": 3975 + }, + { + "epoch": 0.35550786838340487, + "grad_norm": 2.397267625789138, + "learning_rate": 0.00014939687257279363, + "loss": 2.7448, + "step": 3976 + }, + { + "epoch": 0.3555972818311874, + "grad_norm": 1.1390646887478857, + "learning_rate": 0.0001493716902754422, + "loss": 2.1295, + "step": 3977 + }, + { + "epoch": 0.35568669527896996, + "grad_norm": 0.8356952618379498, + "learning_rate": 0.00014934650383734672, + "loss": 2.0403, + "step": 3978 + }, + { + "epoch": 0.3557761087267525, + "grad_norm": 1.7211168236852317, + "learning_rate": 0.00014932131326061957, + "loss": 1.9763, + "step": 3979 + }, + { + "epoch": 0.35586552217453504, + "grad_norm": 1.2566745333435643, + "learning_rate": 0.00014929611854737343, + "loss": 2.1547, + "step": 3980 + }, + { + "epoch": 0.3559549356223176, + "grad_norm": 0.6382766620871363, + "learning_rate": 0.00014927091969972134, + "loss": 2.105, + "step": 3981 + }, + { + "epoch": 0.35604434907010013, + "grad_norm": 0.4905160881036997, + "learning_rate": 0.00014924571671977676, + "loss": 1.8617, + "step": 3982 + }, + { + "epoch": 0.3561337625178827, + "grad_norm": 1.245401510532915, + "learning_rate": 0.0001492205096096534, + "loss": 1.6587, + "step": 3983 + }, + { + "epoch": 0.3562231759656652, + "grad_norm": 0.7253379065138459, + "learning_rate": 0.00014919529837146528, + "loss": 2.1473, + "step": 3984 + }, + { + "epoch": 0.3563125894134478, + "grad_norm": 1.019978941192343, + "learning_rate": 0.00014917008300732696, + "loss": 2.6679, + "step": 3985 + }, + { + "epoch": 0.3564020028612303, + "grad_norm": 2.7126420800865465, + "learning_rate": 0.00014914486351935312, + "loss": 7.2155, + "step": 3986 + }, + { + "epoch": 0.3564914163090129, + "grad_norm": 0.475069319374286, + "learning_rate": 0.00014911963990965897, + "loss": 1.628, + "step": 3987 + }, + { + "epoch": 0.35658082975679545, + "grad_norm": 1.9126983115979432, + "learning_rate": 0.00014909441218035992, + "loss": 2.4649, + "step": 3988 + }, + { + "epoch": 0.35667024320457796, + "grad_norm": 1.674680887545925, + "learning_rate": 0.0001490691803335718, + "loss": 2.5042, + "step": 3989 + }, + { + "epoch": 0.35675965665236054, + "grad_norm": 1.761733808645462, + "learning_rate": 0.0001490439443714108, + "loss": 2.0952, + "step": 3990 + }, + { + "epoch": 0.35684907010014305, + "grad_norm": 0.8649049209309426, + "learning_rate": 0.0001490187042959934, + "loss": 1.7779, + "step": 3991 + }, + { + "epoch": 0.3569384835479256, + "grad_norm": 2.6612825429546114, + "learning_rate": 0.0001489934601094365, + "loss": 3.0149, + "step": 3992 + }, + { + "epoch": 0.35702789699570814, + "grad_norm": 0.8507104974180465, + "learning_rate": 0.00014896821181385725, + "loss": 2.0839, + "step": 3993 + }, + { + "epoch": 0.3571173104434907, + "grad_norm": 4.238795067194387, + "learning_rate": 0.0001489429594113732, + "loss": 2.7305, + "step": 3994 + }, + { + "epoch": 0.3572067238912732, + "grad_norm": 1.6482326678040948, + "learning_rate": 0.00014891770290410228, + "loss": 2.2305, + "step": 3995 + }, + { + "epoch": 0.3572961373390558, + "grad_norm": 1.8211132262933325, + "learning_rate": 0.0001488924422941627, + "loss": 2.5588, + "step": 3996 + }, + { + "epoch": 0.3573855507868383, + "grad_norm": 3.2789905866730216, + "learning_rate": 0.000148867177583673, + "loss": 2.3054, + "step": 3997 + }, + { + "epoch": 0.3574749642346209, + "grad_norm": 1.7438841595604275, + "learning_rate": 0.00014884190877475213, + "loss": 2.3948, + "step": 3998 + }, + { + "epoch": 0.35756437768240346, + "grad_norm": 1.5953282257706245, + "learning_rate": 0.00014881663586951938, + "loss": 2.4593, + "step": 3999 + }, + { + "epoch": 0.35765379113018597, + "grad_norm": 2.445937374641172, + "learning_rate": 0.00014879135887009435, + "loss": 1.8397, + "step": 4000 + }, + { + "epoch": 0.35774320457796854, + "grad_norm": 3.2501543281240255, + "learning_rate": 0.00014876607777859695, + "loss": 2.3064, + "step": 4001 + }, + { + "epoch": 0.35783261802575106, + "grad_norm": 1.1430772565968486, + "learning_rate": 0.0001487407925971475, + "loss": 2.2778, + "step": 4002 + }, + { + "epoch": 0.35792203147353363, + "grad_norm": 3.4212548934842286, + "learning_rate": 0.00014871550332786666, + "loss": 1.7766, + "step": 4003 + }, + { + "epoch": 0.35801144492131615, + "grad_norm": 2.5807285616200217, + "learning_rate": 0.00014869020997287536, + "loss": 2.2192, + "step": 4004 + }, + { + "epoch": 0.3581008583690987, + "grad_norm": 2.14736668648304, + "learning_rate": 0.00014866491253429497, + "loss": 1.9477, + "step": 4005 + }, + { + "epoch": 0.35819027181688123, + "grad_norm": 3.512478505916858, + "learning_rate": 0.00014863961101424712, + "loss": 2.1847, + "step": 4006 + }, + { + "epoch": 0.3582796852646638, + "grad_norm": 3.338313173327663, + "learning_rate": 0.00014861430541485387, + "loss": 2.1544, + "step": 4007 + }, + { + "epoch": 0.3583690987124464, + "grad_norm": 0.8886087799991312, + "learning_rate": 0.00014858899573823753, + "loss": 2.1062, + "step": 4008 + }, + { + "epoch": 0.3584585121602289, + "grad_norm": 2.043685076011066, + "learning_rate": 0.00014856368198652077, + "loss": 2.0668, + "step": 4009 + }, + { + "epoch": 0.35854792560801146, + "grad_norm": 3.4912101940693225, + "learning_rate": 0.00014853836416182668, + "loss": 1.6153, + "step": 4010 + }, + { + "epoch": 0.358637339055794, + "grad_norm": 0.6953966911168712, + "learning_rate": 0.0001485130422662786, + "loss": 2.2352, + "step": 4011 + }, + { + "epoch": 0.35872675250357655, + "grad_norm": 1.142691455563591, + "learning_rate": 0.00014848771630200023, + "loss": 1.985, + "step": 4012 + }, + { + "epoch": 0.35881616595135907, + "grad_norm": 14.907143255314415, + "learning_rate": 0.00014846238627111568, + "loss": 2.2423, + "step": 4013 + }, + { + "epoch": 0.35890557939914164, + "grad_norm": 4.667449723347233, + "learning_rate": 0.00014843705217574933, + "loss": 2.3726, + "step": 4014 + }, + { + "epoch": 0.35899499284692415, + "grad_norm": 18.62503525342935, + "learning_rate": 0.00014841171401802587, + "loss": 1.878, + "step": 4015 + }, + { + "epoch": 0.3590844062947067, + "grad_norm": 2.260009537409484, + "learning_rate": 0.00014838637180007047, + "loss": 2.1068, + "step": 4016 + }, + { + "epoch": 0.3591738197424893, + "grad_norm": 6.179027040182919, + "learning_rate": 0.00014836102552400848, + "loss": 2.6575, + "step": 4017 + }, + { + "epoch": 0.3592632331902718, + "grad_norm": 10.351668234291177, + "learning_rate": 0.00014833567519196566, + "loss": 2.2545, + "step": 4018 + }, + { + "epoch": 0.3593526466380544, + "grad_norm": 1.2425733203653764, + "learning_rate": 0.00014831032080606817, + "loss": 2.0841, + "step": 4019 + }, + { + "epoch": 0.3594420600858369, + "grad_norm": 1.1066456023426199, + "learning_rate": 0.00014828496236844242, + "loss": 2.0243, + "step": 4020 + }, + { + "epoch": 0.3595314735336195, + "grad_norm": 1.2312709946152978, + "learning_rate": 0.00014825959988121515, + "loss": 2.3213, + "step": 4021 + }, + { + "epoch": 0.359620886981402, + "grad_norm": 1.5650284760235917, + "learning_rate": 0.00014823423334651357, + "loss": 2.2126, + "step": 4022 + }, + { + "epoch": 0.35971030042918456, + "grad_norm": 1.0775957292926386, + "learning_rate": 0.00014820886276646506, + "loss": 2.1053, + "step": 4023 + }, + { + "epoch": 0.3597997138769671, + "grad_norm": 2.204762706831355, + "learning_rate": 0.00014818348814319747, + "loss": 2.4132, + "step": 4024 + }, + { + "epoch": 0.35988912732474965, + "grad_norm": 2.326836166193982, + "learning_rate": 0.0001481581094788389, + "loss": 2.6033, + "step": 4025 + }, + { + "epoch": 0.35997854077253216, + "grad_norm": 1.811605167466905, + "learning_rate": 0.00014813272677551787, + "loss": 2.3642, + "step": 4026 + }, + { + "epoch": 0.36006795422031473, + "grad_norm": 2.3805825398337537, + "learning_rate": 0.00014810734003536317, + "loss": 2.556, + "step": 4027 + }, + { + "epoch": 0.3601573676680973, + "grad_norm": 3.2035503254637256, + "learning_rate": 0.00014808194926050394, + "loss": 2.3746, + "step": 4028 + }, + { + "epoch": 0.3602467811158798, + "grad_norm": 1.199341733193606, + "learning_rate": 0.0001480565544530697, + "loss": 1.7276, + "step": 4029 + }, + { + "epoch": 0.3603361945636624, + "grad_norm": 2.454956357168009, + "learning_rate": 0.0001480311556151903, + "loss": 7.5132, + "step": 4030 + }, + { + "epoch": 0.3604256080114449, + "grad_norm": 2.8188891573462307, + "learning_rate": 0.00014800575274899588, + "loss": 2.4534, + "step": 4031 + }, + { + "epoch": 0.3605150214592275, + "grad_norm": 1.6913086931834935, + "learning_rate": 0.00014798034585661695, + "loss": 2.2494, + "step": 4032 + }, + { + "epoch": 0.36060443490701, + "grad_norm": 2.5031289266435466, + "learning_rate": 0.00014795493494018433, + "loss": 2.3807, + "step": 4033 + }, + { + "epoch": 0.36069384835479257, + "grad_norm": 2.2325640484891207, + "learning_rate": 0.00014792952000182926, + "loss": 2.6387, + "step": 4034 + }, + { + "epoch": 0.3607832618025751, + "grad_norm": 2.554558575731203, + "learning_rate": 0.00014790410104368324, + "loss": 1.9084, + "step": 4035 + }, + { + "epoch": 0.36087267525035766, + "grad_norm": 3.9737971399978123, + "learning_rate": 0.00014787867806787807, + "loss": 2.5821, + "step": 4036 + }, + { + "epoch": 0.3609620886981402, + "grad_norm": 2.869569058767425, + "learning_rate": 0.00014785325107654606, + "loss": 2.2201, + "step": 4037 + }, + { + "epoch": 0.36105150214592274, + "grad_norm": 4.024194895704352, + "learning_rate": 0.00014782782007181962, + "loss": 2.1626, + "step": 4038 + }, + { + "epoch": 0.3611409155937053, + "grad_norm": 2.4607161004963083, + "learning_rate": 0.0001478023850558317, + "loss": 2.4346, + "step": 4039 + }, + { + "epoch": 0.36123032904148783, + "grad_norm": 5.378617507754137, + "learning_rate": 0.00014777694603071548, + "loss": 2.2794, + "step": 4040 + }, + { + "epoch": 0.3613197424892704, + "grad_norm": 7.268371645653662, + "learning_rate": 0.0001477515029986045, + "loss": 2.4152, + "step": 4041 + }, + { + "epoch": 0.3614091559370529, + "grad_norm": 4.163281128148078, + "learning_rate": 0.00014772605596163261, + "loss": 1.8589, + "step": 4042 + }, + { + "epoch": 0.3614985693848355, + "grad_norm": 8.003166076535315, + "learning_rate": 0.00014770060492193406, + "loss": 2.0688, + "step": 4043 + }, + { + "epoch": 0.361587982832618, + "grad_norm": 5.999617069802661, + "learning_rate": 0.00014767514988164336, + "loss": 2.2368, + "step": 4044 + }, + { + "epoch": 0.3616773962804006, + "grad_norm": 20.416870305263878, + "learning_rate": 0.00014764969084289544, + "loss": 1.9553, + "step": 4045 + }, + { + "epoch": 0.3617668097281831, + "grad_norm": 6.992912246303223, + "learning_rate": 0.00014762422780782548, + "loss": 2.0656, + "step": 4046 + }, + { + "epoch": 0.36185622317596566, + "grad_norm": 12.760673955411614, + "learning_rate": 0.00014759876077856905, + "loss": 2.2123, + "step": 4047 + }, + { + "epoch": 0.36194563662374823, + "grad_norm": 4.146655854310035, + "learning_rate": 0.00014757328975726207, + "loss": 2.4009, + "step": 4048 + }, + { + "epoch": 0.36203505007153075, + "grad_norm": 5.134221573770991, + "learning_rate": 0.0001475478147460407, + "loss": 2.3723, + "step": 4049 + }, + { + "epoch": 0.3621244635193133, + "grad_norm": 4.439038581186141, + "learning_rate": 0.00014752233574704153, + "loss": 2.3685, + "step": 4050 + }, + { + "epoch": 0.36221387696709584, + "grad_norm": 1.7001571078054503, + "learning_rate": 0.0001474968527624015, + "loss": 2.1557, + "step": 4051 + }, + { + "epoch": 0.3623032904148784, + "grad_norm": 4.642767208434392, + "learning_rate": 0.00014747136579425772, + "loss": 2.6351, + "step": 4052 + }, + { + "epoch": 0.3623927038626609, + "grad_norm": 2.678451537668853, + "learning_rate": 0.00014744587484474784, + "loss": 2.3193, + "step": 4053 + }, + { + "epoch": 0.3624821173104435, + "grad_norm": 1.8023187139701624, + "learning_rate": 0.00014742037991600975, + "loss": 2.0115, + "step": 4054 + }, + { + "epoch": 0.362571530758226, + "grad_norm": 3.125499123194534, + "learning_rate": 0.00014739488101018168, + "loss": 1.9441, + "step": 4055 + }, + { + "epoch": 0.3626609442060086, + "grad_norm": 4.623958977619172, + "learning_rate": 0.00014736937812940217, + "loss": 2.7989, + "step": 4056 + }, + { + "epoch": 0.36275035765379116, + "grad_norm": 0.6088714805529146, + "learning_rate": 0.0001473438712758101, + "loss": 2.3835, + "step": 4057 + }, + { + "epoch": 0.36283977110157367, + "grad_norm": 4.622155433311755, + "learning_rate": 0.00014731836045154477, + "loss": 2.2079, + "step": 4058 + }, + { + "epoch": 0.36292918454935624, + "grad_norm": 1.608231668002686, + "learning_rate": 0.00014729284565874562, + "loss": 2.0989, + "step": 4059 + }, + { + "epoch": 0.36301859799713876, + "grad_norm": 1.6148509264527984, + "learning_rate": 0.0001472673268995527, + "loss": 2.4222, + "step": 4060 + }, + { + "epoch": 0.36310801144492133, + "grad_norm": 2.076964745595366, + "learning_rate": 0.0001472418041761061, + "loss": 1.8423, + "step": 4061 + }, + { + "epoch": 0.36319742489270385, + "grad_norm": 1.7559979713828169, + "learning_rate": 0.00014721627749054647, + "loss": 2.3565, + "step": 4062 + }, + { + "epoch": 0.3632868383404864, + "grad_norm": 1.0641205575071022, + "learning_rate": 0.00014719074684501468, + "loss": 2.6759, + "step": 4063 + }, + { + "epoch": 0.36337625178826893, + "grad_norm": 1.7874901861922627, + "learning_rate": 0.00014716521224165192, + "loss": 2.3817, + "step": 4064 + }, + { + "epoch": 0.3634656652360515, + "grad_norm": 2.5334712454117, + "learning_rate": 0.0001471396736825998, + "loss": 2.3054, + "step": 4065 + }, + { + "epoch": 0.363555078683834, + "grad_norm": 1.0390865953933863, + "learning_rate": 0.00014711413117000013, + "loss": 2.1164, + "step": 4066 + }, + { + "epoch": 0.3636444921316166, + "grad_norm": 1.2583352094754752, + "learning_rate": 0.0001470885847059952, + "loss": 2.2004, + "step": 4067 + }, + { + "epoch": 0.36373390557939916, + "grad_norm": 1.6950605943194876, + "learning_rate": 0.00014706303429272755, + "loss": 2.2768, + "step": 4068 + }, + { + "epoch": 0.3638233190271817, + "grad_norm": 0.9713431691263387, + "learning_rate": 0.00014703747993234003, + "loss": 2.062, + "step": 4069 + }, + { + "epoch": 0.36391273247496425, + "grad_norm": 0.5449214833609148, + "learning_rate": 0.00014701192162697591, + "loss": 1.6764, + "step": 4070 + }, + { + "epoch": 0.36400214592274677, + "grad_norm": 1.0781339266768457, + "learning_rate": 0.00014698635937877868, + "loss": 2.2212, + "step": 4071 + }, + { + "epoch": 0.36409155937052934, + "grad_norm": 0.6392164503005537, + "learning_rate": 0.0001469607931898922, + "loss": 1.976, + "step": 4072 + }, + { + "epoch": 0.36418097281831185, + "grad_norm": 1.3184714842392908, + "learning_rate": 0.00014693522306246076, + "loss": 2.3198, + "step": 4073 + }, + { + "epoch": 0.3642703862660944, + "grad_norm": 2.5188404707787586, + "learning_rate": 0.00014690964899862882, + "loss": 2.4794, + "step": 4074 + }, + { + "epoch": 0.36435979971387694, + "grad_norm": 1.3111616849743846, + "learning_rate": 0.0001468840710005413, + "loss": 2.1749, + "step": 4075 + }, + { + "epoch": 0.3644492131616595, + "grad_norm": 1.8213833775596577, + "learning_rate": 0.00014685848907034331, + "loss": 2.0658, + "step": 4076 + }, + { + "epoch": 0.3645386266094421, + "grad_norm": 2.182069232474671, + "learning_rate": 0.00014683290321018048, + "loss": 2.2854, + "step": 4077 + }, + { + "epoch": 0.3646280400572246, + "grad_norm": 1.4205832163223617, + "learning_rate": 0.0001468073134221986, + "loss": 2.0557, + "step": 4078 + }, + { + "epoch": 0.36471745350500717, + "grad_norm": 0.6294527004551101, + "learning_rate": 0.0001467817197085439, + "loss": 2.3003, + "step": 4079 + }, + { + "epoch": 0.3648068669527897, + "grad_norm": 0.9505517143504502, + "learning_rate": 0.0001467561220713628, + "loss": 2.2052, + "step": 4080 + }, + { + "epoch": 0.36489628040057226, + "grad_norm": 0.576559008464793, + "learning_rate": 0.00014673052051280227, + "loss": 1.9529, + "step": 4081 + }, + { + "epoch": 0.3649856938483548, + "grad_norm": 1.8935503533554572, + "learning_rate": 0.0001467049150350094, + "loss": 2.0474, + "step": 4082 + }, + { + "epoch": 0.36507510729613735, + "grad_norm": 2.8913539289087207, + "learning_rate": 0.00014667930564013173, + "loss": 1.9117, + "step": 4083 + }, + { + "epoch": 0.36516452074391986, + "grad_norm": 1.716022102853993, + "learning_rate": 0.00014665369233031705, + "loss": 1.7629, + "step": 4084 + }, + { + "epoch": 0.36525393419170243, + "grad_norm": 2.6257724327652237, + "learning_rate": 0.00014662807510771355, + "loss": 2.4412, + "step": 4085 + }, + { + "epoch": 0.365343347639485, + "grad_norm": 1.3725888280560732, + "learning_rate": 0.0001466024539744697, + "loss": 2.4704, + "step": 4086 + }, + { + "epoch": 0.3654327610872675, + "grad_norm": 2.9050312523591106, + "learning_rate": 0.0001465768289327343, + "loss": 2.1595, + "step": 4087 + }, + { + "epoch": 0.3655221745350501, + "grad_norm": 2.4637598368769216, + "learning_rate": 0.00014655119998465652, + "loss": 1.9748, + "step": 4088 + }, + { + "epoch": 0.3656115879828326, + "grad_norm": 2.6166553717739403, + "learning_rate": 0.00014652556713238578, + "loss": 2.2348, + "step": 4089 + }, + { + "epoch": 0.3657010014306152, + "grad_norm": 1.3373786609769336, + "learning_rate": 0.000146499930378072, + "loss": 2.6335, + "step": 4090 + }, + { + "epoch": 0.3657904148783977, + "grad_norm": 3.2859613639812704, + "learning_rate": 0.00014647428972386513, + "loss": 2.0118, + "step": 4091 + }, + { + "epoch": 0.36587982832618027, + "grad_norm": 1.481546088882758, + "learning_rate": 0.00014644864517191576, + "loss": 2.5704, + "step": 4092 + }, + { + "epoch": 0.3659692417739628, + "grad_norm": 2.6194284395728604, + "learning_rate": 0.00014642299672437461, + "loss": 2.5923, + "step": 4093 + }, + { + "epoch": 0.36605865522174535, + "grad_norm": 2.9205825541332278, + "learning_rate": 0.00014639734438339278, + "loss": 7.3108, + "step": 4094 + }, + { + "epoch": 0.36614806866952787, + "grad_norm": 2.70994547171034, + "learning_rate": 0.0001463716881511217, + "loss": 2.4581, + "step": 4095 + }, + { + "epoch": 0.36623748211731044, + "grad_norm": 1.1619718399864711, + "learning_rate": 0.00014634602802971312, + "loss": 2.3214, + "step": 4096 + }, + { + "epoch": 0.366326895565093, + "grad_norm": 1.6751027473577285, + "learning_rate": 0.0001463203640213192, + "loss": 1.8393, + "step": 4097 + }, + { + "epoch": 0.36641630901287553, + "grad_norm": 2.146345715395954, + "learning_rate": 0.00014629469612809223, + "loss": 2.3563, + "step": 4098 + }, + { + "epoch": 0.3665057224606581, + "grad_norm": 1.0045362343258821, + "learning_rate": 0.00014626902435218504, + "loss": 1.8204, + "step": 4099 + }, + { + "epoch": 0.3665951359084406, + "grad_norm": 3.325848579487246, + "learning_rate": 0.00014624334869575066, + "loss": 2.6007, + "step": 4100 + }, + { + "epoch": 0.3666845493562232, + "grad_norm": 6.45876484599626, + "learning_rate": 0.00014621766916094248, + "loss": 2.3025, + "step": 4101 + }, + { + "epoch": 0.3667739628040057, + "grad_norm": 5.314464590589516, + "learning_rate": 0.00014619198574991417, + "loss": 2.34, + "step": 4102 + }, + { + "epoch": 0.3668633762517883, + "grad_norm": 2.3798691998969383, + "learning_rate": 0.00014616629846481982, + "loss": 1.7911, + "step": 4103 + }, + { + "epoch": 0.3669527896995708, + "grad_norm": 1.9720460312088801, + "learning_rate": 0.00014614060730781377, + "loss": 2.2982, + "step": 4104 + }, + { + "epoch": 0.36704220314735336, + "grad_norm": 1.1230442895653643, + "learning_rate": 0.0001461149122810507, + "loss": 1.7597, + "step": 4105 + }, + { + "epoch": 0.36713161659513593, + "grad_norm": 2.5228967498139827, + "learning_rate": 0.00014608921338668562, + "loss": 2.2737, + "step": 4106 + }, + { + "epoch": 0.36722103004291845, + "grad_norm": 1.0240532533688635, + "learning_rate": 0.00014606351062687391, + "loss": 2.4027, + "step": 4107 + }, + { + "epoch": 0.367310443490701, + "grad_norm": 0.6887154822119846, + "learning_rate": 0.00014603780400377118, + "loss": 2.1137, + "step": 4108 + }, + { + "epoch": 0.36739985693848354, + "grad_norm": 2.826515203881585, + "learning_rate": 0.00014601209351953345, + "loss": 2.2256, + "step": 4109 + }, + { + "epoch": 0.3674892703862661, + "grad_norm": 0.8688734227577966, + "learning_rate": 0.00014598637917631697, + "loss": 1.8075, + "step": 4110 + }, + { + "epoch": 0.3675786838340486, + "grad_norm": 1.6122534544532445, + "learning_rate": 0.00014596066097627842, + "loss": 2.148, + "step": 4111 + }, + { + "epoch": 0.3676680972818312, + "grad_norm": 0.9545534507140369, + "learning_rate": 0.00014593493892157473, + "loss": 1.7738, + "step": 4112 + }, + { + "epoch": 0.3677575107296137, + "grad_norm": 2.100591600434631, + "learning_rate": 0.00014590921301436318, + "loss": 2.5124, + "step": 4113 + }, + { + "epoch": 0.3678469241773963, + "grad_norm": 1.2514048553285073, + "learning_rate": 0.0001458834832568014, + "loss": 2.1511, + "step": 4114 + }, + { + "epoch": 0.3679363376251788, + "grad_norm": 1.137289765698423, + "learning_rate": 0.00014585774965104732, + "loss": 1.9047, + "step": 4115 + }, + { + "epoch": 0.36802575107296137, + "grad_norm": 2.3865491083320016, + "learning_rate": 0.00014583201219925908, + "loss": 2.5745, + "step": 4116 + }, + { + "epoch": 0.36811516452074394, + "grad_norm": 1.823845381891305, + "learning_rate": 0.0001458062709035954, + "loss": 2.4382, + "step": 4117 + }, + { + "epoch": 0.36820457796852646, + "grad_norm": 1.8078538587895359, + "learning_rate": 0.00014578052576621507, + "loss": 2.4717, + "step": 4118 + }, + { + "epoch": 0.36829399141630903, + "grad_norm": 1.6373602729854522, + "learning_rate": 0.00014575477678927732, + "loss": 2.1897, + "step": 4119 + }, + { + "epoch": 0.36838340486409155, + "grad_norm": 0.719134916535176, + "learning_rate": 0.00014572902397494173, + "loss": 2.1915, + "step": 4120 + }, + { + "epoch": 0.3684728183118741, + "grad_norm": 2.771084878866889, + "learning_rate": 0.0001457032673253681, + "loss": 2.3597, + "step": 4121 + }, + { + "epoch": 0.36856223175965663, + "grad_norm": 0.6819665562513433, + "learning_rate": 0.00014567750684271665, + "loss": 2.2278, + "step": 4122 + }, + { + "epoch": 0.3686516452074392, + "grad_norm": 1.0981185279966674, + "learning_rate": 0.00014565174252914785, + "loss": 1.8082, + "step": 4123 + }, + { + "epoch": 0.3687410586552217, + "grad_norm": 2.576102223221518, + "learning_rate": 0.00014562597438682256, + "loss": 2.1404, + "step": 4124 + }, + { + "epoch": 0.3688304721030043, + "grad_norm": 0.9145091859320809, + "learning_rate": 0.0001456002024179019, + "loss": 2.0166, + "step": 4125 + }, + { + "epoch": 0.36891988555078686, + "grad_norm": 0.9813621860864639, + "learning_rate": 0.0001455744266245473, + "loss": 2.1995, + "step": 4126 + }, + { + "epoch": 0.3690092989985694, + "grad_norm": 2.372154066361026, + "learning_rate": 0.0001455486470089206, + "loss": 2.218, + "step": 4127 + }, + { + "epoch": 0.36909871244635195, + "grad_norm": 0.9503706147282198, + "learning_rate": 0.0001455228635731839, + "loss": 2.1505, + "step": 4128 + }, + { + "epoch": 0.36918812589413447, + "grad_norm": 0.5817741967909984, + "learning_rate": 0.00014549707631949957, + "loss": 1.877, + "step": 4129 + }, + { + "epoch": 0.36927753934191704, + "grad_norm": 0.9797816268354421, + "learning_rate": 0.00014547128525003045, + "loss": 1.972, + "step": 4130 + }, + { + "epoch": 0.36936695278969955, + "grad_norm": 1.0949236630996937, + "learning_rate": 0.0001454454903669395, + "loss": 2.7347, + "step": 4131 + }, + { + "epoch": 0.3694563662374821, + "grad_norm": 1.4970444491177397, + "learning_rate": 0.0001454196916723902, + "loss": 2.2069, + "step": 4132 + }, + { + "epoch": 0.36954577968526464, + "grad_norm": 0.9555716488945956, + "learning_rate": 0.00014539388916854617, + "loss": 1.7137, + "step": 4133 + }, + { + "epoch": 0.3696351931330472, + "grad_norm": 1.6637834886329828, + "learning_rate": 0.00014536808285757152, + "loss": 1.8353, + "step": 4134 + }, + { + "epoch": 0.3697246065808298, + "grad_norm": 2.214014472367327, + "learning_rate": 0.00014534227274163051, + "loss": 2.173, + "step": 4135 + }, + { + "epoch": 0.3698140200286123, + "grad_norm": 1.6819704345616047, + "learning_rate": 0.00014531645882288788, + "loss": 2.4778, + "step": 4136 + }, + { + "epoch": 0.36990343347639487, + "grad_norm": 0.5931895727354265, + "learning_rate": 0.00014529064110350856, + "loss": 1.884, + "step": 4137 + }, + { + "epoch": 0.3699928469241774, + "grad_norm": 1.142515405589147, + "learning_rate": 0.00014526481958565787, + "loss": 2.4604, + "step": 4138 + }, + { + "epoch": 0.37008226037195996, + "grad_norm": 0.687425206317996, + "learning_rate": 0.00014523899427150143, + "loss": 1.73, + "step": 4139 + }, + { + "epoch": 0.3701716738197425, + "grad_norm": 1.337753003669797, + "learning_rate": 0.00014521316516320515, + "loss": 2.216, + "step": 4140 + }, + { + "epoch": 0.37026108726752505, + "grad_norm": 1.1209664515022355, + "learning_rate": 0.00014518733226293534, + "loss": 2.4693, + "step": 4141 + }, + { + "epoch": 0.37035050071530756, + "grad_norm": 0.8741085633332456, + "learning_rate": 0.00014516149557285856, + "loss": 1.9087, + "step": 4142 + }, + { + "epoch": 0.37043991416309013, + "grad_norm": 2.527356627683493, + "learning_rate": 0.00014513565509514167, + "loss": 1.9047, + "step": 4143 + }, + { + "epoch": 0.37052932761087265, + "grad_norm": 0.7959461040824393, + "learning_rate": 0.00014510981083195188, + "loss": 1.9359, + "step": 4144 + }, + { + "epoch": 0.3706187410586552, + "grad_norm": 2.7846412208964466, + "learning_rate": 0.00014508396278545678, + "loss": 7.0666, + "step": 4145 + }, + { + "epoch": 0.3707081545064378, + "grad_norm": 2.2364785094100625, + "learning_rate": 0.0001450581109578241, + "loss": 2.7653, + "step": 4146 + }, + { + "epoch": 0.3707975679542203, + "grad_norm": 1.4853966251215458, + "learning_rate": 0.00014503225535122212, + "loss": 2.2131, + "step": 4147 + }, + { + "epoch": 0.3708869814020029, + "grad_norm": 2.3674028723493263, + "learning_rate": 0.00014500639596781926, + "loss": 1.8205, + "step": 4148 + }, + { + "epoch": 0.3709763948497854, + "grad_norm": 1.9952066659155439, + "learning_rate": 0.00014498053280978434, + "loss": 2.517, + "step": 4149 + }, + { + "epoch": 0.37106580829756797, + "grad_norm": 1.3050320286747397, + "learning_rate": 0.00014495466587928642, + "loss": 2.1597, + "step": 4150 + }, + { + "epoch": 0.3711552217453505, + "grad_norm": 2.7448764355886612, + "learning_rate": 0.00014492879517849497, + "loss": 1.9953, + "step": 4151 + }, + { + "epoch": 0.37124463519313305, + "grad_norm": 1.4568506796737268, + "learning_rate": 0.0001449029207095798, + "loss": 2.4988, + "step": 4152 + }, + { + "epoch": 0.37133404864091557, + "grad_norm": 2.5216998560046546, + "learning_rate": 0.00014487704247471078, + "loss": 2.1503, + "step": 4153 + }, + { + "epoch": 0.37142346208869814, + "grad_norm": 3.717805752490543, + "learning_rate": 0.00014485116047605848, + "loss": 2.4863, + "step": 4154 + }, + { + "epoch": 0.3715128755364807, + "grad_norm": 2.121271882060806, + "learning_rate": 0.00014482527471579353, + "loss": 1.9575, + "step": 4155 + }, + { + "epoch": 0.37160228898426323, + "grad_norm": 4.451591724153684, + "learning_rate": 0.00014479938519608687, + "loss": 2.3239, + "step": 4156 + }, + { + "epoch": 0.3716917024320458, + "grad_norm": 1.3264284698066404, + "learning_rate": 0.0001447734919191099, + "loss": 2.3127, + "step": 4157 + }, + { + "epoch": 0.3717811158798283, + "grad_norm": 2.5179642818209156, + "learning_rate": 0.00014474759488703425, + "loss": 2.0634, + "step": 4158 + }, + { + "epoch": 0.3718705293276109, + "grad_norm": 3.4779348643971355, + "learning_rate": 0.00014472169410203187, + "loss": 2.0433, + "step": 4159 + }, + { + "epoch": 0.3719599427753934, + "grad_norm": 1.3522419142099518, + "learning_rate": 0.00014469578956627496, + "loss": 1.9529, + "step": 4160 + }, + { + "epoch": 0.372049356223176, + "grad_norm": 2.1861559191353237, + "learning_rate": 0.0001446698812819362, + "loss": 1.6952, + "step": 4161 + }, + { + "epoch": 0.3721387696709585, + "grad_norm": 3.396076425014109, + "learning_rate": 0.00014464396925118847, + "loss": 1.8461, + "step": 4162 + }, + { + "epoch": 0.37222818311874106, + "grad_norm": 0.9574771252539457, + "learning_rate": 0.00014461805347620489, + "loss": 1.6814, + "step": 4163 + }, + { + "epoch": 0.3723175965665236, + "grad_norm": 1.4867338459650805, + "learning_rate": 0.00014459213395915906, + "loss": 2.2351, + "step": 4164 + }, + { + "epoch": 0.37240701001430615, + "grad_norm": 5.793445936136257, + "learning_rate": 0.00014456621070222484, + "loss": 1.9912, + "step": 4165 + }, + { + "epoch": 0.3724964234620887, + "grad_norm": 5.316790729654446, + "learning_rate": 0.00014454028370757636, + "loss": 2.2796, + "step": 4166 + }, + { + "epoch": 0.37258583690987124, + "grad_norm": 1.3978744997697536, + "learning_rate": 0.00014451435297738806, + "loss": 2.0674, + "step": 4167 + }, + { + "epoch": 0.3726752503576538, + "grad_norm": 4.60099214221072, + "learning_rate": 0.00014448841851383472, + "loss": 2.1308, + "step": 4168 + }, + { + "epoch": 0.3727646638054363, + "grad_norm": 4.566811211420432, + "learning_rate": 0.00014446248031909148, + "loss": 1.9741, + "step": 4169 + }, + { + "epoch": 0.3728540772532189, + "grad_norm": 1.538826893053723, + "learning_rate": 0.0001444365383953337, + "loss": 2.3477, + "step": 4170 + }, + { + "epoch": 0.3729434907010014, + "grad_norm": 5.239188985708282, + "learning_rate": 0.00014441059274473706, + "loss": 2.3536, + "step": 4171 + }, + { + "epoch": 0.373032904148784, + "grad_norm": 3.8020362091831066, + "learning_rate": 0.00014438464336947773, + "loss": 2.6075, + "step": 4172 + }, + { + "epoch": 0.3731223175965665, + "grad_norm": 3.406929453804694, + "learning_rate": 0.0001443586902717319, + "loss": 2.5018, + "step": 4173 + }, + { + "epoch": 0.37321173104434907, + "grad_norm": 4.404970470812928, + "learning_rate": 0.0001443327334536763, + "loss": 2.1792, + "step": 4174 + }, + { + "epoch": 0.37330114449213164, + "grad_norm": 1.3117579061860096, + "learning_rate": 0.00014430677291748788, + "loss": 2.151, + "step": 4175 + }, + { + "epoch": 0.37339055793991416, + "grad_norm": 1.1731452765758792, + "learning_rate": 0.00014428080866534396, + "loss": 2.0134, + "step": 4176 + }, + { + "epoch": 0.37347997138769673, + "grad_norm": 2.4298197927749943, + "learning_rate": 0.00014425484069942207, + "loss": 1.4599, + "step": 4177 + }, + { + "epoch": 0.37356938483547925, + "grad_norm": 3.0175020357860656, + "learning_rate": 0.00014422886902190014, + "loss": 1.867, + "step": 4178 + }, + { + "epoch": 0.3736587982832618, + "grad_norm": 0.9060060594920737, + "learning_rate": 0.00014420289363495638, + "loss": 2.0982, + "step": 4179 + }, + { + "epoch": 0.37374821173104433, + "grad_norm": 1.009420821973031, + "learning_rate": 0.00014417691454076932, + "loss": 2.218, + "step": 4180 + }, + { + "epoch": 0.3738376251788269, + "grad_norm": 1.5771229413668058, + "learning_rate": 0.00014415093174151777, + "loss": 1.6699, + "step": 4181 + }, + { + "epoch": 0.3739270386266094, + "grad_norm": 2.6116235342134964, + "learning_rate": 0.0001441249452393809, + "loss": 2.1, + "step": 4182 + }, + { + "epoch": 0.374016452074392, + "grad_norm": 0.41830223132993233, + "learning_rate": 0.0001440989550365382, + "loss": 1.5149, + "step": 4183 + }, + { + "epoch": 0.3741058655221745, + "grad_norm": 1.0871160391074577, + "learning_rate": 0.00014407296113516934, + "loss": 1.7321, + "step": 4184 + }, + { + "epoch": 0.3741952789699571, + "grad_norm": 2.5979168092600933, + "learning_rate": 0.00014404696353745452, + "loss": 7.2796, + "step": 4185 + }, + { + "epoch": 0.37428469241773965, + "grad_norm": 2.2041556264576405, + "learning_rate": 0.000144020962245574, + "loss": 1.7805, + "step": 4186 + }, + { + "epoch": 0.37437410586552217, + "grad_norm": 1.5147139612500706, + "learning_rate": 0.00014399495726170858, + "loss": 1.9671, + "step": 4187 + }, + { + "epoch": 0.37446351931330474, + "grad_norm": 0.4902232318064865, + "learning_rate": 0.0001439689485880392, + "loss": 1.8147, + "step": 4188 + }, + { + "epoch": 0.37455293276108725, + "grad_norm": 1.234648488696309, + "learning_rate": 0.00014394293622674724, + "loss": 1.9449, + "step": 4189 + }, + { + "epoch": 0.3746423462088698, + "grad_norm": 0.4870725696910934, + "learning_rate": 0.00014391692018001425, + "loss": 2.0023, + "step": 4190 + }, + { + "epoch": 0.37473175965665234, + "grad_norm": 1.2667576033555783, + "learning_rate": 0.00014389090045002225, + "loss": 2.2483, + "step": 4191 + }, + { + "epoch": 0.3748211731044349, + "grad_norm": 0.7650384371735294, + "learning_rate": 0.0001438648770389534, + "loss": 1.9814, + "step": 4192 + }, + { + "epoch": 0.37491058655221743, + "grad_norm": 0.9209025970831233, + "learning_rate": 0.0001438388499489903, + "loss": 2.0005, + "step": 4193 + }, + { + "epoch": 0.375, + "grad_norm": 0.6798695198938086, + "learning_rate": 0.00014381281918231578, + "loss": 1.9134, + "step": 4194 + }, + { + "epoch": 0.37508941344778257, + "grad_norm": 1.2558320161955026, + "learning_rate": 0.00014378678474111304, + "loss": 2.1453, + "step": 4195 + }, + { + "epoch": 0.3751788268955651, + "grad_norm": 1.1066497796806685, + "learning_rate": 0.00014376074662756557, + "loss": 1.9301, + "step": 4196 + }, + { + "epoch": 0.37526824034334766, + "grad_norm": 0.4414389397160692, + "learning_rate": 0.0001437347048438571, + "loss": 2.0148, + "step": 4197 + }, + { + "epoch": 0.3753576537911302, + "grad_norm": 1.4002063668189346, + "learning_rate": 0.00014370865939217176, + "loss": 2.1247, + "step": 4198 + }, + { + "epoch": 0.37544706723891275, + "grad_norm": 1.4291883096423128, + "learning_rate": 0.00014368261027469394, + "loss": 2.3088, + "step": 4199 + }, + { + "epoch": 0.37553648068669526, + "grad_norm": 1.1064422391268482, + "learning_rate": 0.00014365655749360833, + "loss": 2.1007, + "step": 4200 + }, + { + "epoch": 0.37562589413447783, + "grad_norm": 2.25944637814276, + "learning_rate": 0.0001436305010511, + "loss": 1.6501, + "step": 4201 + }, + { + "epoch": 0.37571530758226035, + "grad_norm": 0.5133641637138936, + "learning_rate": 0.00014360444094935424, + "loss": 2.1765, + "step": 4202 + }, + { + "epoch": 0.3758047210300429, + "grad_norm": 0.5730214812806929, + "learning_rate": 0.00014357837719055667, + "loss": 2.4544, + "step": 4203 + }, + { + "epoch": 0.3758941344778255, + "grad_norm": 1.3165993266093294, + "learning_rate": 0.00014355230977689323, + "loss": 1.9069, + "step": 4204 + }, + { + "epoch": 0.375983547925608, + "grad_norm": 2.20757545107527, + "learning_rate": 0.00014352623871055018, + "loss": 2.111, + "step": 4205 + }, + { + "epoch": 0.3760729613733906, + "grad_norm": 2.106189062491677, + "learning_rate": 0.00014350016399371405, + "loss": 2.3265, + "step": 4206 + }, + { + "epoch": 0.3761623748211731, + "grad_norm": 1.4562674602828425, + "learning_rate": 0.00014347408562857169, + "loss": 1.532, + "step": 4207 + }, + { + "epoch": 0.37625178826895567, + "grad_norm": 1.880912578970401, + "learning_rate": 0.00014344800361731027, + "loss": 2.4261, + "step": 4208 + }, + { + "epoch": 0.3763412017167382, + "grad_norm": 0.9083015621437197, + "learning_rate": 0.00014342191796211726, + "loss": 2.0809, + "step": 4209 + }, + { + "epoch": 0.37643061516452075, + "grad_norm": 1.5939124237144076, + "learning_rate": 0.00014339582866518044, + "loss": 2.4575, + "step": 4210 + }, + { + "epoch": 0.37652002861230327, + "grad_norm": 1.3810028979595235, + "learning_rate": 0.00014336973572868787, + "loss": 2.6556, + "step": 4211 + }, + { + "epoch": 0.37660944206008584, + "grad_norm": 1.8276114535829868, + "learning_rate": 0.00014334363915482795, + "loss": 2.6927, + "step": 4212 + }, + { + "epoch": 0.37669885550786836, + "grad_norm": 0.8721591910436706, + "learning_rate": 0.00014331753894578937, + "loss": 1.9793, + "step": 4213 + }, + { + "epoch": 0.37678826895565093, + "grad_norm": 1.7370279017978367, + "learning_rate": 0.00014329143510376108, + "loss": 1.9011, + "step": 4214 + }, + { + "epoch": 0.3768776824034335, + "grad_norm": 0.4466364789630725, + "learning_rate": 0.00014326532763093245, + "loss": 2.0311, + "step": 4215 + }, + { + "epoch": 0.376967095851216, + "grad_norm": 0.6978620517716846, + "learning_rate": 0.00014323921652949301, + "loss": 2.0333, + "step": 4216 + }, + { + "epoch": 0.3770565092989986, + "grad_norm": 1.0623019715676636, + "learning_rate": 0.00014321310180163272, + "loss": 1.6552, + "step": 4217 + }, + { + "epoch": 0.3771459227467811, + "grad_norm": 1.3591395522191412, + "learning_rate": 0.00014318698344954175, + "loss": 2.2237, + "step": 4218 + }, + { + "epoch": 0.3772353361945637, + "grad_norm": 0.6944122684119909, + "learning_rate": 0.00014316086147541065, + "loss": 2.2004, + "step": 4219 + }, + { + "epoch": 0.3773247496423462, + "grad_norm": 1.1162252625887872, + "learning_rate": 0.00014313473588143026, + "loss": 2.2415, + "step": 4220 + }, + { + "epoch": 0.37741416309012876, + "grad_norm": 1.1711095931958189, + "learning_rate": 0.0001431086066697916, + "loss": 2.123, + "step": 4221 + }, + { + "epoch": 0.3775035765379113, + "grad_norm": 0.6274611862990122, + "learning_rate": 0.0001430824738426862, + "loss": 1.9014, + "step": 4222 + }, + { + "epoch": 0.37759298998569385, + "grad_norm": 0.6724173466550072, + "learning_rate": 0.00014305633740230574, + "loss": 1.8346, + "step": 4223 + }, + { + "epoch": 0.3776824034334764, + "grad_norm": 0.8117595620330004, + "learning_rate": 0.00014303019735084226, + "loss": 2.0277, + "step": 4224 + }, + { + "epoch": 0.37777181688125894, + "grad_norm": 0.9276319587442488, + "learning_rate": 0.00014300405369048808, + "loss": 2.1766, + "step": 4225 + }, + { + "epoch": 0.3778612303290415, + "grad_norm": 1.764172922008727, + "learning_rate": 0.00014297790642343587, + "loss": 1.6412, + "step": 4226 + }, + { + "epoch": 0.377950643776824, + "grad_norm": 0.8710893986680289, + "learning_rate": 0.00014295175555187854, + "loss": 2.0301, + "step": 4227 + }, + { + "epoch": 0.3780400572246066, + "grad_norm": 0.698311892207525, + "learning_rate": 0.00014292560107800935, + "loss": 2.2432, + "step": 4228 + }, + { + "epoch": 0.3781294706723891, + "grad_norm": 1.5951780359669536, + "learning_rate": 0.00014289944300402186, + "loss": 1.761, + "step": 4229 + }, + { + "epoch": 0.3782188841201717, + "grad_norm": 1.505467174813502, + "learning_rate": 0.00014287328133210986, + "loss": 2.1759, + "step": 4230 + }, + { + "epoch": 0.3783082975679542, + "grad_norm": 1.1324396408315207, + "learning_rate": 0.00014284711606446754, + "loss": 2.1233, + "step": 4231 + }, + { + "epoch": 0.37839771101573677, + "grad_norm": 0.7436767986497527, + "learning_rate": 0.00014282094720328937, + "loss": 2.0194, + "step": 4232 + }, + { + "epoch": 0.3784871244635193, + "grad_norm": 1.7233150025917545, + "learning_rate": 0.00014279477475077006, + "loss": 2.1603, + "step": 4233 + }, + { + "epoch": 0.37857653791130186, + "grad_norm": 0.3971210238674272, + "learning_rate": 0.00014276859870910463, + "loss": 1.6991, + "step": 4234 + }, + { + "epoch": 0.37866595135908443, + "grad_norm": 0.7744128621568771, + "learning_rate": 0.00014274241908048856, + "loss": 1.6318, + "step": 4235 + }, + { + "epoch": 0.37875536480686695, + "grad_norm": 0.716900130327278, + "learning_rate": 0.00014271623586711738, + "loss": 1.8749, + "step": 4236 + }, + { + "epoch": 0.3788447782546495, + "grad_norm": 0.9267438299802273, + "learning_rate": 0.00014269004907118706, + "loss": 2.371, + "step": 4237 + }, + { + "epoch": 0.37893419170243203, + "grad_norm": 0.6361781513919065, + "learning_rate": 0.0001426638586948939, + "loss": 2.0468, + "step": 4238 + }, + { + "epoch": 0.3790236051502146, + "grad_norm": 1.138217559055771, + "learning_rate": 0.00014263766474043445, + "loss": 1.8878, + "step": 4239 + }, + { + "epoch": 0.3791130185979971, + "grad_norm": 1.984992491617034, + "learning_rate": 0.00014261146721000553, + "loss": 2.5026, + "step": 4240 + }, + { + "epoch": 0.3792024320457797, + "grad_norm": 0.6260858018425344, + "learning_rate": 0.00014258526610580433, + "loss": 1.5327, + "step": 4241 + }, + { + "epoch": 0.3792918454935622, + "grad_norm": 2.292297706828008, + "learning_rate": 0.0001425590614300283, + "loss": 2.3706, + "step": 4242 + }, + { + "epoch": 0.3793812589413448, + "grad_norm": 0.7598304670116762, + "learning_rate": 0.0001425328531848752, + "loss": 2.2074, + "step": 4243 + }, + { + "epoch": 0.37947067238912735, + "grad_norm": 1.0635773547395497, + "learning_rate": 0.00014250664137254303, + "loss": 1.7237, + "step": 4244 + }, + { + "epoch": 0.37956008583690987, + "grad_norm": 3.0621106855715206, + "learning_rate": 0.0001424804259952302, + "loss": 1.9544, + "step": 4245 + }, + { + "epoch": 0.37964949928469244, + "grad_norm": 1.4176692075960164, + "learning_rate": 0.00014245420705513535, + "loss": 2.6882, + "step": 4246 + }, + { + "epoch": 0.37973891273247495, + "grad_norm": 1.5822817828072446, + "learning_rate": 0.0001424279845544574, + "loss": 2.8586, + "step": 4247 + }, + { + "epoch": 0.3798283261802575, + "grad_norm": 7.844084933792233, + "learning_rate": 0.00014240175849539565, + "loss": 7.2128, + "step": 4248 + }, + { + "epoch": 0.37991773962804004, + "grad_norm": 7.546874786675972, + "learning_rate": 0.00014237552888014961, + "loss": 2.104, + "step": 4249 + }, + { + "epoch": 0.3800071530758226, + "grad_norm": 4.774539321031871, + "learning_rate": 0.00014234929571091916, + "loss": 2.0185, + "step": 4250 + }, + { + "epoch": 0.38009656652360513, + "grad_norm": 1.7760341215922428, + "learning_rate": 0.0001423230589899044, + "loss": 2.3299, + "step": 4251 + }, + { + "epoch": 0.3801859799713877, + "grad_norm": 6.074471389715662, + "learning_rate": 0.00014229681871930582, + "loss": 2.4017, + "step": 4252 + }, + { + "epoch": 0.3802753934191702, + "grad_norm": 1.0988154754346309, + "learning_rate": 0.00014227057490132414, + "loss": 1.6414, + "step": 4253 + }, + { + "epoch": 0.3803648068669528, + "grad_norm": 1.2860740541941758, + "learning_rate": 0.00014224432753816036, + "loss": 2.3526, + "step": 4254 + }, + { + "epoch": 0.38045422031473536, + "grad_norm": 1.6634323667770994, + "learning_rate": 0.00014221807663201586, + "loss": 2.1642, + "step": 4255 + }, + { + "epoch": 0.3805436337625179, + "grad_norm": 2.891287610552461, + "learning_rate": 0.0001421918221850923, + "loss": 2.632, + "step": 4256 + }, + { + "epoch": 0.38063304721030045, + "grad_norm": 1.8074314243965202, + "learning_rate": 0.0001421655641995915, + "loss": 2.1656, + "step": 4257 + }, + { + "epoch": 0.38072246065808296, + "grad_norm": 1.5670139805046033, + "learning_rate": 0.0001421393026777158, + "loss": 2.3615, + "step": 4258 + }, + { + "epoch": 0.38081187410586553, + "grad_norm": 1.8808358146837423, + "learning_rate": 0.00014211303762166766, + "loss": 2.0382, + "step": 4259 + }, + { + "epoch": 0.38090128755364805, + "grad_norm": 19.432152200250137, + "learning_rate": 0.00014208676903364992, + "loss": 7.3224, + "step": 4260 + }, + { + "epoch": 0.3809907010014306, + "grad_norm": 10.23603882330831, + "learning_rate": 0.00014206049691586564, + "loss": 2.092, + "step": 4261 + }, + { + "epoch": 0.38108011444921314, + "grad_norm": 9.670198184031825, + "learning_rate": 0.00014203422127051835, + "loss": 2.8108, + "step": 4262 + }, + { + "epoch": 0.3811695278969957, + "grad_norm": 5.224798103302557, + "learning_rate": 0.00014200794209981167, + "loss": 2.3314, + "step": 4263 + }, + { + "epoch": 0.3812589413447783, + "grad_norm": 2.263550338059581, + "learning_rate": 0.0001419816594059496, + "loss": 2.2185, + "step": 4264 + }, + { + "epoch": 0.3813483547925608, + "grad_norm": 4.804489625244433, + "learning_rate": 0.00014195537319113647, + "loss": 2.0529, + "step": 4265 + }, + { + "epoch": 0.38143776824034337, + "grad_norm": 3.1777871676622733, + "learning_rate": 0.00014192908345757687, + "loss": 2.2769, + "step": 4266 + }, + { + "epoch": 0.3815271816881259, + "grad_norm": 5.015661030982061, + "learning_rate": 0.0001419027902074757, + "loss": 2.609, + "step": 4267 + }, + { + "epoch": 0.38161659513590845, + "grad_norm": 1.412549613761954, + "learning_rate": 0.000141876493443038, + "loss": 2.1254, + "step": 4268 + }, + { + "epoch": 0.38170600858369097, + "grad_norm": 1.5931777065541461, + "learning_rate": 0.0001418501931664695, + "loss": 2.2115, + "step": 4269 + }, + { + "epoch": 0.38179542203147354, + "grad_norm": 1.7445377803291082, + "learning_rate": 0.0001418238893799758, + "loss": 2.0446, + "step": 4270 + }, + { + "epoch": 0.38188483547925606, + "grad_norm": 3.4273539790270044, + "learning_rate": 0.00014179758208576298, + "loss": 2.486, + "step": 4271 + }, + { + "epoch": 0.38197424892703863, + "grad_norm": 3.4665169159118, + "learning_rate": 0.00014177127128603745, + "loss": 2.7002, + "step": 4272 + }, + { + "epoch": 0.3820636623748212, + "grad_norm": 3.3342986489803246, + "learning_rate": 0.00014174495698300588, + "loss": 1.8434, + "step": 4273 + }, + { + "epoch": 0.3821530758226037, + "grad_norm": 4.441791314416712, + "learning_rate": 0.00014171863917887513, + "loss": 2.6445, + "step": 4274 + }, + { + "epoch": 0.3822424892703863, + "grad_norm": 11.846782654439107, + "learning_rate": 0.0001416923178758525, + "loss": 7.9005, + "step": 4275 + }, + { + "epoch": 0.3823319027181688, + "grad_norm": 5.381734641370538, + "learning_rate": 0.00014166599307614556, + "loss": 2.7587, + "step": 4276 + }, + { + "epoch": 0.3824213161659514, + "grad_norm": 2.6166737414237646, + "learning_rate": 0.00014163966478196208, + "loss": 2.3198, + "step": 4277 + }, + { + "epoch": 0.3825107296137339, + "grad_norm": 3.366564479302385, + "learning_rate": 0.0001416133329955102, + "loss": 1.8579, + "step": 4278 + }, + { + "epoch": 0.38260014306151646, + "grad_norm": 3.2800888588676296, + "learning_rate": 0.00014158699771899832, + "loss": 2.2027, + "step": 4279 + }, + { + "epoch": 0.382689556509299, + "grad_norm": 5.248644420066978, + "learning_rate": 0.0001415606589546352, + "loss": 7.3856, + "step": 4280 + }, + { + "epoch": 0.38277896995708155, + "grad_norm": 2.365365794450299, + "learning_rate": 0.0001415343167046298, + "loss": 2.3782, + "step": 4281 + }, + { + "epoch": 0.38286838340486407, + "grad_norm": 5.271478995606345, + "learning_rate": 0.0001415079709711914, + "loss": 2.1921, + "step": 4282 + }, + { + "epoch": 0.38295779685264664, + "grad_norm": 3.521642872681917, + "learning_rate": 0.0001414816217565296, + "loss": 2.2854, + "step": 4283 + }, + { + "epoch": 0.3830472103004292, + "grad_norm": 2.055517010890399, + "learning_rate": 0.00014145526906285432, + "loss": 2.5206, + "step": 4284 + }, + { + "epoch": 0.3831366237482117, + "grad_norm": 2.640052193144996, + "learning_rate": 0.00014142891289237563, + "loss": 2.379, + "step": 4285 + }, + { + "epoch": 0.3832260371959943, + "grad_norm": 3.245502878703753, + "learning_rate": 0.0001414025532473041, + "loss": 2.0303, + "step": 4286 + }, + { + "epoch": 0.3833154506437768, + "grad_norm": 2.036800917947464, + "learning_rate": 0.00014137619012985042, + "loss": 1.6065, + "step": 4287 + }, + { + "epoch": 0.3834048640915594, + "grad_norm": 3.9968211370915667, + "learning_rate": 0.00014134982354222563, + "loss": 2.0471, + "step": 4288 + }, + { + "epoch": 0.3834942775393419, + "grad_norm": 4.914714360489716, + "learning_rate": 0.00014132345348664106, + "loss": 1.9636, + "step": 4289 + }, + { + "epoch": 0.38358369098712447, + "grad_norm": 2.3654139908272214, + "learning_rate": 0.00014129707996530838, + "loss": 2.334, + "step": 4290 + }, + { + "epoch": 0.383673104434907, + "grad_norm": 1.8344183260965268, + "learning_rate": 0.00014127070298043947, + "loss": 2.5224, + "step": 4291 + }, + { + "epoch": 0.38376251788268956, + "grad_norm": 1.1121856962849375, + "learning_rate": 0.00014124432253424655, + "loss": 2.4576, + "step": 4292 + }, + { + "epoch": 0.38385193133047213, + "grad_norm": 2.4268917213066783, + "learning_rate": 0.0001412179386289421, + "loss": 1.589, + "step": 4293 + }, + { + "epoch": 0.38394134477825465, + "grad_norm": 2.3785439061325295, + "learning_rate": 0.00014119155126673895, + "loss": 2.0836, + "step": 4294 + }, + { + "epoch": 0.3840307582260372, + "grad_norm": 2.1268783183829973, + "learning_rate": 0.0001411651604498501, + "loss": 2.126, + "step": 4295 + }, + { + "epoch": 0.38412017167381973, + "grad_norm": 1.5660014384001042, + "learning_rate": 0.00014113876618048897, + "loss": 2.2709, + "step": 4296 + }, + { + "epoch": 0.3842095851216023, + "grad_norm": 2.0523634138135205, + "learning_rate": 0.00014111236846086922, + "loss": 2.0925, + "step": 4297 + }, + { + "epoch": 0.3842989985693848, + "grad_norm": 2.2367769089358314, + "learning_rate": 0.00014108596729320473, + "loss": 2.468, + "step": 4298 + }, + { + "epoch": 0.3843884120171674, + "grad_norm": 1.8121292174078785, + "learning_rate": 0.0001410595626797098, + "loss": 1.9253, + "step": 4299 + }, + { + "epoch": 0.3844778254649499, + "grad_norm": 2.557594361625371, + "learning_rate": 0.00014103315462259898, + "loss": 2.2209, + "step": 4300 + }, + { + "epoch": 0.3845672389127325, + "grad_norm": 1.911781362114929, + "learning_rate": 0.000141006743124087, + "loss": 2.2259, + "step": 4301 + }, + { + "epoch": 0.384656652360515, + "grad_norm": 1.7488520278936648, + "learning_rate": 0.000140980328186389, + "loss": 1.7905, + "step": 4302 + }, + { + "epoch": 0.38474606580829757, + "grad_norm": 1.5380351548915607, + "learning_rate": 0.00014095390981172038, + "loss": 2.2882, + "step": 4303 + }, + { + "epoch": 0.38483547925608014, + "grad_norm": 7.116195153504849, + "learning_rate": 0.00014092748800229683, + "loss": 1.9815, + "step": 4304 + }, + { + "epoch": 0.38492489270386265, + "grad_norm": 2.9753813586876596, + "learning_rate": 0.00014090106276033423, + "loss": 2.4918, + "step": 4305 + }, + { + "epoch": 0.3850143061516452, + "grad_norm": 4.074189540160927, + "learning_rate": 0.00014087463408804892, + "loss": 2.5851, + "step": 4306 + }, + { + "epoch": 0.38510371959942774, + "grad_norm": 2.46347411166739, + "learning_rate": 0.00014084820198765743, + "loss": 2.0523, + "step": 4307 + }, + { + "epoch": 0.3851931330472103, + "grad_norm": 2.8581759273145653, + "learning_rate": 0.00014082176646137653, + "loss": 7.1489, + "step": 4308 + }, + { + "epoch": 0.38528254649499283, + "grad_norm": 2.6048007964137487, + "learning_rate": 0.0001407953275114234, + "loss": 7.4178, + "step": 4309 + }, + { + "epoch": 0.3853719599427754, + "grad_norm": 1.3769827507703698, + "learning_rate": 0.00014076888514001542, + "loss": 1.8395, + "step": 4310 + }, + { + "epoch": 0.3854613733905579, + "grad_norm": 1.8764631523230346, + "learning_rate": 0.0001407424393493703, + "loss": 2.1116, + "step": 4311 + }, + { + "epoch": 0.3855507868383405, + "grad_norm": 1.4515989452640208, + "learning_rate": 0.00014071599014170598, + "loss": 1.775, + "step": 4312 + }, + { + "epoch": 0.38564020028612306, + "grad_norm": 1.507455433082206, + "learning_rate": 0.0001406895375192407, + "loss": 2.194, + "step": 4313 + }, + { + "epoch": 0.3857296137339056, + "grad_norm": 0.7697617552459602, + "learning_rate": 0.0001406630814841931, + "loss": 1.6957, + "step": 4314 + }, + { + "epoch": 0.38581902718168815, + "grad_norm": 2.3909205204430926, + "learning_rate": 0.00014063662203878195, + "loss": 2.3466, + "step": 4315 + }, + { + "epoch": 0.38590844062947066, + "grad_norm": 2.0313417992739016, + "learning_rate": 0.00014061015918522639, + "loss": 1.9463, + "step": 4316 + }, + { + "epoch": 0.38599785407725323, + "grad_norm": 0.7551888680038763, + "learning_rate": 0.0001405836929257458, + "loss": 2.0402, + "step": 4317 + }, + { + "epoch": 0.38608726752503575, + "grad_norm": 1.5265734675015576, + "learning_rate": 0.00014055722326255992, + "loss": 1.9046, + "step": 4318 + }, + { + "epoch": 0.3861766809728183, + "grad_norm": 2.3487761285215814, + "learning_rate": 0.0001405307501978887, + "loss": 1.9022, + "step": 4319 + }, + { + "epoch": 0.38626609442060084, + "grad_norm": 1.5155138936761856, + "learning_rate": 0.0001405042737339524, + "loss": 2.365, + "step": 4320 + }, + { + "epoch": 0.3863555078683834, + "grad_norm": 0.92645152063704, + "learning_rate": 0.0001404777938729716, + "loss": 1.8943, + "step": 4321 + }, + { + "epoch": 0.386444921316166, + "grad_norm": 1.6995329456694004, + "learning_rate": 0.00014045131061716712, + "loss": 1.6753, + "step": 4322 + }, + { + "epoch": 0.3865343347639485, + "grad_norm": 1.4167015812444952, + "learning_rate": 0.00014042482396876005, + "loss": 2.0031, + "step": 4323 + }, + { + "epoch": 0.38662374821173107, + "grad_norm": 2.5683962716096373, + "learning_rate": 0.0001403983339299718, + "loss": 2.3884, + "step": 4324 + }, + { + "epoch": 0.3867131616595136, + "grad_norm": 0.9356343617576858, + "learning_rate": 0.0001403718405030241, + "loss": 2.2455, + "step": 4325 + }, + { + "epoch": 0.38680257510729615, + "grad_norm": 1.333318609512809, + "learning_rate": 0.00014034534369013887, + "loss": 1.7794, + "step": 4326 + }, + { + "epoch": 0.38689198855507867, + "grad_norm": 2.9783371068709332, + "learning_rate": 0.0001403188434935384, + "loss": 2.1291, + "step": 4327 + }, + { + "epoch": 0.38698140200286124, + "grad_norm": 0.38099325227337355, + "learning_rate": 0.00014029233991544527, + "loss": 1.7677, + "step": 4328 + }, + { + "epoch": 0.38707081545064376, + "grad_norm": 2.412507107578532, + "learning_rate": 0.0001402658329580822, + "loss": 2.1444, + "step": 4329 + }, + { + "epoch": 0.38716022889842633, + "grad_norm": 1.205014866465113, + "learning_rate": 0.0001402393226236723, + "loss": 2.274, + "step": 4330 + }, + { + "epoch": 0.38724964234620884, + "grad_norm": 1.3873796857893113, + "learning_rate": 0.00014021280891443909, + "loss": 2.3013, + "step": 4331 + }, + { + "epoch": 0.3873390557939914, + "grad_norm": 0.6689969868660209, + "learning_rate": 0.0001401862918326061, + "loss": 1.5961, + "step": 4332 + }, + { + "epoch": 0.387428469241774, + "grad_norm": 2.5118234831898008, + "learning_rate": 0.0001401597713803974, + "loss": 1.8563, + "step": 4333 + }, + { + "epoch": 0.3875178826895565, + "grad_norm": 1.1947148685345343, + "learning_rate": 0.00014013324756003716, + "loss": 2.1828, + "step": 4334 + }, + { + "epoch": 0.3876072961373391, + "grad_norm": 1.102719941848399, + "learning_rate": 0.0001401067203737499, + "loss": 2.6174, + "step": 4335 + }, + { + "epoch": 0.3876967095851216, + "grad_norm": 1.1046925345223257, + "learning_rate": 0.00014008018982376044, + "loss": 2.1307, + "step": 4336 + }, + { + "epoch": 0.38778612303290416, + "grad_norm": 4.053995756883853, + "learning_rate": 0.0001400536559122939, + "loss": 1.7834, + "step": 4337 + }, + { + "epoch": 0.3878755364806867, + "grad_norm": 1.932589292104215, + "learning_rate": 0.00014002711864157557, + "loss": 1.9266, + "step": 4338 + }, + { + "epoch": 0.38796494992846925, + "grad_norm": 2.3128074764018156, + "learning_rate": 0.00014000057801383115, + "loss": 2.3244, + "step": 4339 + }, + { + "epoch": 0.38805436337625177, + "grad_norm": 2.606117642991561, + "learning_rate": 0.0001399740340312866, + "loss": 2.505, + "step": 4340 + }, + { + "epoch": 0.38814377682403434, + "grad_norm": 2.0342416313832583, + "learning_rate": 0.00013994748669616803, + "loss": 2.2108, + "step": 4341 + }, + { + "epoch": 0.3882331902718169, + "grad_norm": 1.6876210396148554, + "learning_rate": 0.00013992093601070203, + "loss": 2.1404, + "step": 4342 + }, + { + "epoch": 0.3883226037195994, + "grad_norm": 1.8571044250669106, + "learning_rate": 0.00013989438197711533, + "loss": 1.5638, + "step": 4343 + }, + { + "epoch": 0.388412017167382, + "grad_norm": 2.6455956043046838, + "learning_rate": 0.000139867824597635, + "loss": 1.9479, + "step": 4344 + }, + { + "epoch": 0.3885014306151645, + "grad_norm": 1.6232716135562335, + "learning_rate": 0.00013984126387448837, + "loss": 1.8907, + "step": 4345 + }, + { + "epoch": 0.3885908440629471, + "grad_norm": 0.913524314172393, + "learning_rate": 0.00013981469980990302, + "loss": 2.0083, + "step": 4346 + }, + { + "epoch": 0.3886802575107296, + "grad_norm": 5.906270586698847, + "learning_rate": 0.0001397881324061069, + "loss": 1.8654, + "step": 4347 + }, + { + "epoch": 0.38876967095851217, + "grad_norm": 0.902863788394115, + "learning_rate": 0.0001397615616653282, + "loss": 1.813, + "step": 4348 + }, + { + "epoch": 0.3888590844062947, + "grad_norm": 2.202451832432575, + "learning_rate": 0.00013973498758979532, + "loss": 2.7036, + "step": 4349 + }, + { + "epoch": 0.38894849785407726, + "grad_norm": 1.9894051949877751, + "learning_rate": 0.00013970841018173702, + "loss": 2.316, + "step": 4350 + }, + { + "epoch": 0.3890379113018598, + "grad_norm": 1.4203800784999616, + "learning_rate": 0.0001396818294433823, + "loss": 2.664, + "step": 4351 + }, + { + "epoch": 0.38912732474964234, + "grad_norm": 0.5372768599518499, + "learning_rate": 0.00013965524537696048, + "loss": 2.2421, + "step": 4352 + }, + { + "epoch": 0.3892167381974249, + "grad_norm": 1.1009356729549586, + "learning_rate": 0.00013962865798470113, + "loss": 2.5979, + "step": 4353 + }, + { + "epoch": 0.38930615164520743, + "grad_norm": 1.2071106150842779, + "learning_rate": 0.00013960206726883407, + "loss": 1.7951, + "step": 4354 + }, + { + "epoch": 0.38939556509299, + "grad_norm": 0.7914207493109967, + "learning_rate": 0.00013957547323158949, + "loss": 2.4262, + "step": 4355 + }, + { + "epoch": 0.3894849785407725, + "grad_norm": 0.7197289302759241, + "learning_rate": 0.00013954887587519773, + "loss": 1.848, + "step": 4356 + }, + { + "epoch": 0.3895743919885551, + "grad_norm": 1.0023926555615776, + "learning_rate": 0.00013952227520188957, + "loss": 2.0378, + "step": 4357 + }, + { + "epoch": 0.3896638054363376, + "grad_norm": 2.802312085212501, + "learning_rate": 0.00013949567121389586, + "loss": 2.1235, + "step": 4358 + }, + { + "epoch": 0.3897532188841202, + "grad_norm": 1.0575047048109514, + "learning_rate": 0.00013946906391344791, + "loss": 2.0733, + "step": 4359 + }, + { + "epoch": 0.3898426323319027, + "grad_norm": 1.0158810039000183, + "learning_rate": 0.00013944245330277724, + "loss": 1.8391, + "step": 4360 + }, + { + "epoch": 0.38993204577968527, + "grad_norm": 2.6913175444985296, + "learning_rate": 0.00013941583938411567, + "loss": 2.2654, + "step": 4361 + }, + { + "epoch": 0.39002145922746784, + "grad_norm": 2.738972146048006, + "learning_rate": 0.00013938922215969523, + "loss": 2.5313, + "step": 4362 + }, + { + "epoch": 0.39011087267525035, + "grad_norm": 1.434831657631616, + "learning_rate": 0.00013936260163174832, + "loss": 2.0387, + "step": 4363 + }, + { + "epoch": 0.3902002861230329, + "grad_norm": 2.5973964721638927, + "learning_rate": 0.00013933597780250753, + "loss": 2.0041, + "step": 4364 + }, + { + "epoch": 0.39028969957081544, + "grad_norm": 2.483134343489981, + "learning_rate": 0.0001393093506742058, + "loss": 1.948, + "step": 4365 + }, + { + "epoch": 0.390379113018598, + "grad_norm": 2.402349544462921, + "learning_rate": 0.0001392827202490763, + "loss": 1.9404, + "step": 4366 + }, + { + "epoch": 0.3904685264663805, + "grad_norm": 1.6247186151270603, + "learning_rate": 0.00013925608652935249, + "loss": 1.9492, + "step": 4367 + }, + { + "epoch": 0.3905579399141631, + "grad_norm": 0.7689296264955398, + "learning_rate": 0.0001392294495172681, + "loss": 2.5609, + "step": 4368 + }, + { + "epoch": 0.3906473533619456, + "grad_norm": 3.9003099400275723, + "learning_rate": 0.00013920280921505716, + "loss": 2.6525, + "step": 4369 + }, + { + "epoch": 0.3907367668097282, + "grad_norm": 2.1724398870482347, + "learning_rate": 0.00013917616562495396, + "loss": 2.1084, + "step": 4370 + }, + { + "epoch": 0.3908261802575107, + "grad_norm": 1.3090924247336473, + "learning_rate": 0.00013914951874919308, + "loss": 1.9285, + "step": 4371 + }, + { + "epoch": 0.3909155937052933, + "grad_norm": 2.423079899906142, + "learning_rate": 0.00013912286859000934, + "loss": 1.9438, + "step": 4372 + }, + { + "epoch": 0.39100500715307585, + "grad_norm": 1.6136043005671712, + "learning_rate": 0.00013909621514963784, + "loss": 2.1557, + "step": 4373 + }, + { + "epoch": 0.39109442060085836, + "grad_norm": 10.390240518523301, + "learning_rate": 0.00013906955843031403, + "loss": 7.2814, + "step": 4374 + }, + { + "epoch": 0.39118383404864093, + "grad_norm": 12.504177285225941, + "learning_rate": 0.00013904289843427348, + "loss": 2.2854, + "step": 4375 + }, + { + "epoch": 0.39127324749642345, + "grad_norm": 2.7517373470338518, + "learning_rate": 0.00013901623516375219, + "loss": 2.05, + "step": 4376 + }, + { + "epoch": 0.391362660944206, + "grad_norm": 2.395096330503738, + "learning_rate": 0.00013898956862098643, + "loss": 1.9066, + "step": 4377 + }, + { + "epoch": 0.39145207439198854, + "grad_norm": 5.389542804999165, + "learning_rate": 0.00013896289880821263, + "loss": 2.0701, + "step": 4378 + }, + { + "epoch": 0.3915414878397711, + "grad_norm": 5.318022621964242, + "learning_rate": 0.0001389362257276675, + "loss": 2.1461, + "step": 4379 + }, + { + "epoch": 0.3916309012875536, + "grad_norm": 2.626975668467672, + "learning_rate": 0.00013890954938158823, + "loss": 7.0166, + "step": 4380 + }, + { + "epoch": 0.3917203147353362, + "grad_norm": 3.9053307280534213, + "learning_rate": 0.000138882869772212, + "loss": 2.1998, + "step": 4381 + }, + { + "epoch": 0.39180972818311877, + "grad_norm": 8.158189493311923, + "learning_rate": 0.00013885618690177642, + "loss": 7.2929, + "step": 4382 + }, + { + "epoch": 0.3918991416309013, + "grad_norm": 1.72295300142801, + "learning_rate": 0.0001388295007725194, + "loss": 2.0219, + "step": 4383 + }, + { + "epoch": 0.39198855507868385, + "grad_norm": 2.816643773502382, + "learning_rate": 0.00013880281138667905, + "loss": 2.4938, + "step": 4384 + }, + { + "epoch": 0.39207796852646637, + "grad_norm": 3.6430706112817557, + "learning_rate": 0.00013877611874649375, + "loss": 2.4973, + "step": 4385 + }, + { + "epoch": 0.39216738197424894, + "grad_norm": 0.62209786771443, + "learning_rate": 0.0001387494228542022, + "loss": 1.8858, + "step": 4386 + }, + { + "epoch": 0.39225679542203146, + "grad_norm": 1.3714470313366902, + "learning_rate": 0.00013872272371204337, + "loss": 1.7964, + "step": 4387 + }, + { + "epoch": 0.39234620886981403, + "grad_norm": 2.3174808833501555, + "learning_rate": 0.00013869602132225646, + "loss": 2.2003, + "step": 4388 + }, + { + "epoch": 0.39243562231759654, + "grad_norm": 1.522746703847165, + "learning_rate": 0.00013866931568708098, + "loss": 1.6941, + "step": 4389 + }, + { + "epoch": 0.3925250357653791, + "grad_norm": 1.4733775910202045, + "learning_rate": 0.00013864260680875666, + "loss": 1.9662, + "step": 4390 + }, + { + "epoch": 0.3926144492131617, + "grad_norm": 2.072126445682258, + "learning_rate": 0.00013861589468952364, + "loss": 2.0239, + "step": 4391 + }, + { + "epoch": 0.3927038626609442, + "grad_norm": 2.1437289794834196, + "learning_rate": 0.0001385891793316221, + "loss": 2.1089, + "step": 4392 + }, + { + "epoch": 0.3927932761087268, + "grad_norm": 1.4914684184343714, + "learning_rate": 0.0001385624607372927, + "loss": 2.1089, + "step": 4393 + }, + { + "epoch": 0.3928826895565093, + "grad_norm": 3.8069004703385474, + "learning_rate": 0.00013853573890877633, + "loss": 1.7903, + "step": 4394 + }, + { + "epoch": 0.39297210300429186, + "grad_norm": 1.4427960162947047, + "learning_rate": 0.0001385090138483141, + "loss": 1.8723, + "step": 4395 + }, + { + "epoch": 0.3930615164520744, + "grad_norm": 2.9129947395126603, + "learning_rate": 0.0001384822855581473, + "loss": 2.0802, + "step": 4396 + }, + { + "epoch": 0.39315092989985695, + "grad_norm": 0.9787486395717667, + "learning_rate": 0.00013845555404051776, + "loss": 2.0015, + "step": 4397 + }, + { + "epoch": 0.39324034334763946, + "grad_norm": 1.6316414399292085, + "learning_rate": 0.00013842881929766732, + "loss": 1.9964, + "step": 4398 + }, + { + "epoch": 0.39332975679542204, + "grad_norm": 0.8690722704539119, + "learning_rate": 0.00013840208133183822, + "loss": 1.8807, + "step": 4399 + }, + { + "epoch": 0.39341917024320455, + "grad_norm": 3.1524708513735527, + "learning_rate": 0.0001383753401452729, + "loss": 1.8246, + "step": 4400 + }, + { + "epoch": 0.3935085836909871, + "grad_norm": 2.0132929960624146, + "learning_rate": 0.00013834859574021418, + "loss": 2.0345, + "step": 4401 + }, + { + "epoch": 0.3935979971387697, + "grad_norm": 1.7518899140108428, + "learning_rate": 0.00013832184811890508, + "loss": 2.4853, + "step": 4402 + }, + { + "epoch": 0.3936874105865522, + "grad_norm": 1.4423389302293823, + "learning_rate": 0.0001382950972835888, + "loss": 2.374, + "step": 4403 + }, + { + "epoch": 0.3937768240343348, + "grad_norm": 3.460784186112702, + "learning_rate": 0.000138268343236509, + "loss": 1.7292, + "step": 4404 + }, + { + "epoch": 0.3938662374821173, + "grad_norm": 1.9706520735227957, + "learning_rate": 0.00013824158597990947, + "loss": 2.5272, + "step": 4405 + }, + { + "epoch": 0.39395565092989987, + "grad_norm": 0.7206146668825345, + "learning_rate": 0.00013821482551603425, + "loss": 1.6416, + "step": 4406 + }, + { + "epoch": 0.3940450643776824, + "grad_norm": 2.4841631467945557, + "learning_rate": 0.00013818806184712781, + "loss": 2.3946, + "step": 4407 + }, + { + "epoch": 0.39413447782546496, + "grad_norm": 1.9061810770091003, + "learning_rate": 0.00013816129497543476, + "loss": 1.915, + "step": 4408 + }, + { + "epoch": 0.3942238912732475, + "grad_norm": 1.040538704782396, + "learning_rate": 0.00013813452490319997, + "loss": 2.3579, + "step": 4409 + }, + { + "epoch": 0.39431330472103004, + "grad_norm": 0.48470079084596474, + "learning_rate": 0.0001381077516326686, + "loss": 2.1635, + "step": 4410 + }, + { + "epoch": 0.3944027181688126, + "grad_norm": 2.9533464816070496, + "learning_rate": 0.00013808097516608618, + "loss": 1.6458, + "step": 4411 + }, + { + "epoch": 0.39449213161659513, + "grad_norm": 2.4378080985820705, + "learning_rate": 0.00013805419550569833, + "loss": 2.0284, + "step": 4412 + }, + { + "epoch": 0.3945815450643777, + "grad_norm": 3.39706718779822, + "learning_rate": 0.00013802741265375105, + "loss": 1.9337, + "step": 4413 + }, + { + "epoch": 0.3946709585121602, + "grad_norm": 2.127255900310173, + "learning_rate": 0.00013800062661249062, + "loss": 2.6313, + "step": 4414 + }, + { + "epoch": 0.3947603719599428, + "grad_norm": 0.9899917065980499, + "learning_rate": 0.00013797383738416353, + "loss": 2.1893, + "step": 4415 + }, + { + "epoch": 0.3948497854077253, + "grad_norm": 0.5035552614481876, + "learning_rate": 0.00013794704497101655, + "loss": 1.6415, + "step": 4416 + }, + { + "epoch": 0.3949391988555079, + "grad_norm": 0.6646112137052471, + "learning_rate": 0.00013792024937529673, + "loss": 1.7496, + "step": 4417 + }, + { + "epoch": 0.3950286123032904, + "grad_norm": 2.5917633437773477, + "learning_rate": 0.0001378934505992514, + "loss": 2.3754, + "step": 4418 + }, + { + "epoch": 0.39511802575107297, + "grad_norm": 1.8532064154908836, + "learning_rate": 0.00013786664864512814, + "loss": 2.3267, + "step": 4419 + }, + { + "epoch": 0.3952074391988555, + "grad_norm": 2.001098245606875, + "learning_rate": 0.00013783984351517476, + "loss": 2.8499, + "step": 4420 + }, + { + "epoch": 0.39529685264663805, + "grad_norm": 2.605027531018096, + "learning_rate": 0.00013781303521163943, + "loss": 2.0668, + "step": 4421 + }, + { + "epoch": 0.3953862660944206, + "grad_norm": 0.76775829066696, + "learning_rate": 0.0001377862237367705, + "loss": 2.076, + "step": 4422 + }, + { + "epoch": 0.39547567954220314, + "grad_norm": 1.3826796381292241, + "learning_rate": 0.0001377594090928166, + "loss": 2.0843, + "step": 4423 + }, + { + "epoch": 0.3955650929899857, + "grad_norm": 2.4799892854019103, + "learning_rate": 0.00013773259128202668, + "loss": 2.0933, + "step": 4424 + }, + { + "epoch": 0.3956545064377682, + "grad_norm": 0.7543342979772498, + "learning_rate": 0.0001377057703066499, + "loss": 1.8666, + "step": 4425 + }, + { + "epoch": 0.3957439198855508, + "grad_norm": 6.018222423196742, + "learning_rate": 0.0001376789461689357, + "loss": 1.9048, + "step": 4426 + }, + { + "epoch": 0.3958333333333333, + "grad_norm": 2.73049367911516, + "learning_rate": 0.0001376521188711338, + "loss": 2.3457, + "step": 4427 + }, + { + "epoch": 0.3959227467811159, + "grad_norm": 2.305178227401996, + "learning_rate": 0.00013762528841549415, + "loss": 1.8596, + "step": 4428 + }, + { + "epoch": 0.3960121602288984, + "grad_norm": 1.6052827099856202, + "learning_rate": 0.000137598454804267, + "loss": 1.5107, + "step": 4429 + }, + { + "epoch": 0.396101573676681, + "grad_norm": 1.3551009933743756, + "learning_rate": 0.00013757161803970282, + "loss": 2.1758, + "step": 4430 + }, + { + "epoch": 0.39619098712446355, + "grad_norm": 1.1339763712305575, + "learning_rate": 0.00013754477812405247, + "loss": 2.0148, + "step": 4431 + }, + { + "epoch": 0.39628040057224606, + "grad_norm": 3.955521652452196, + "learning_rate": 0.0001375179350595669, + "loss": 2.2293, + "step": 4432 + }, + { + "epoch": 0.39636981402002863, + "grad_norm": 1.162530342292299, + "learning_rate": 0.0001374910888484974, + "loss": 2.2495, + "step": 4433 + }, + { + "epoch": 0.39645922746781115, + "grad_norm": 2.983244067561607, + "learning_rate": 0.00013746423949309556, + "loss": 1.8577, + "step": 4434 + }, + { + "epoch": 0.3965486409155937, + "grad_norm": 1.8706043028941468, + "learning_rate": 0.00013743738699561323, + "loss": 2.6743, + "step": 4435 + }, + { + "epoch": 0.39663805436337624, + "grad_norm": 1.0627452157197617, + "learning_rate": 0.00013741053135830246, + "loss": 2.1913, + "step": 4436 + }, + { + "epoch": 0.3967274678111588, + "grad_norm": 1.8868685068240174, + "learning_rate": 0.00013738367258341557, + "loss": 2.0167, + "step": 4437 + }, + { + "epoch": 0.3968168812589413, + "grad_norm": 1.5671494611783703, + "learning_rate": 0.00013735681067320526, + "loss": 1.7898, + "step": 4438 + }, + { + "epoch": 0.3969062947067239, + "grad_norm": 1.4415459296015871, + "learning_rate": 0.00013732994562992433, + "loss": 2.1284, + "step": 4439 + }, + { + "epoch": 0.3969957081545064, + "grad_norm": 0.4250331722492445, + "learning_rate": 0.00013730307745582593, + "loss": 1.3892, + "step": 4440 + }, + { + "epoch": 0.397085121602289, + "grad_norm": 2.2395599641269692, + "learning_rate": 0.0001372762061531635, + "loss": 2.1984, + "step": 4441 + }, + { + "epoch": 0.39717453505007155, + "grad_norm": 1.9611869684344403, + "learning_rate": 0.00013724933172419066, + "loss": 2.2541, + "step": 4442 + }, + { + "epoch": 0.39726394849785407, + "grad_norm": 1.2911243795282346, + "learning_rate": 0.00013722245417116134, + "loss": 1.9991, + "step": 4443 + }, + { + "epoch": 0.39735336194563664, + "grad_norm": 2.1348732344313723, + "learning_rate": 0.00013719557349632977, + "loss": 2.0004, + "step": 4444 + }, + { + "epoch": 0.39744277539341916, + "grad_norm": 1.8160508387700383, + "learning_rate": 0.00013716868970195038, + "loss": 2.0539, + "step": 4445 + }, + { + "epoch": 0.39753218884120173, + "grad_norm": 0.6572956961738332, + "learning_rate": 0.00013714180279027785, + "loss": 1.9722, + "step": 4446 + }, + { + "epoch": 0.39762160228898424, + "grad_norm": 2.2541566582360253, + "learning_rate": 0.00013711491276356718, + "loss": 2.6222, + "step": 4447 + }, + { + "epoch": 0.3977110157367668, + "grad_norm": 1.38643992639287, + "learning_rate": 0.0001370880196240736, + "loss": 1.7865, + "step": 4448 + }, + { + "epoch": 0.39780042918454933, + "grad_norm": 1.2699323347698321, + "learning_rate": 0.00013706112337405263, + "loss": 2.0524, + "step": 4449 + }, + { + "epoch": 0.3978898426323319, + "grad_norm": 1.146128130517853, + "learning_rate": 0.00013703422401575995, + "loss": 1.9711, + "step": 4450 + }, + { + "epoch": 0.3979792560801145, + "grad_norm": 1.8418096541197124, + "learning_rate": 0.00013700732155145167, + "loss": 2.2407, + "step": 4451 + }, + { + "epoch": 0.398068669527897, + "grad_norm": 1.8280035127006158, + "learning_rate": 0.00013698041598338403, + "loss": 2.3731, + "step": 4452 + }, + { + "epoch": 0.39815808297567956, + "grad_norm": 1.0238113576350028, + "learning_rate": 0.0001369535073138135, + "loss": 1.8133, + "step": 4453 + }, + { + "epoch": 0.3982474964234621, + "grad_norm": 2.0969900657474603, + "learning_rate": 0.00013692659554499702, + "loss": 2.2165, + "step": 4454 + }, + { + "epoch": 0.39833690987124465, + "grad_norm": 0.7414445930206922, + "learning_rate": 0.0001368996806791915, + "loss": 1.9246, + "step": 4455 + }, + { + "epoch": 0.39842632331902716, + "grad_norm": 5.526168544016936, + "learning_rate": 0.0001368727627186544, + "loss": 7.1077, + "step": 4456 + }, + { + "epoch": 0.39851573676680974, + "grad_norm": 4.711907923690241, + "learning_rate": 0.00013684584166564315, + "loss": 2.1161, + "step": 4457 + }, + { + "epoch": 0.39860515021459225, + "grad_norm": 2.513996259126121, + "learning_rate": 0.0001368189175224157, + "loss": 1.9976, + "step": 4458 + }, + { + "epoch": 0.3986945636623748, + "grad_norm": 1.8667947696208478, + "learning_rate": 0.00013679199029123008, + "loss": 2.328, + "step": 4459 + }, + { + "epoch": 0.3987839771101574, + "grad_norm": 3.4005071349737097, + "learning_rate": 0.00013676505997434467, + "loss": 2.2369, + "step": 4460 + }, + { + "epoch": 0.3988733905579399, + "grad_norm": 1.7127740543463672, + "learning_rate": 0.0001367381265740181, + "loss": 2.0911, + "step": 4461 + }, + { + "epoch": 0.3989628040057225, + "grad_norm": 2.2545150821620616, + "learning_rate": 0.00013671119009250922, + "loss": 1.8412, + "step": 4462 + }, + { + "epoch": 0.399052217453505, + "grad_norm": 3.045894564223903, + "learning_rate": 0.00013668425053207713, + "loss": 1.995, + "step": 4463 + }, + { + "epoch": 0.39914163090128757, + "grad_norm": 2.5791521781931093, + "learning_rate": 0.0001366573078949813, + "loss": 2.2589, + "step": 4464 + }, + { + "epoch": 0.3992310443490701, + "grad_norm": 4.572074044397254, + "learning_rate": 0.00013663036218348128, + "loss": 1.877, + "step": 4465 + }, + { + "epoch": 0.39932045779685266, + "grad_norm": 2.5392315291187124, + "learning_rate": 0.00013660341339983707, + "loss": 2.1177, + "step": 4466 + }, + { + "epoch": 0.3994098712446352, + "grad_norm": 4.805068481273641, + "learning_rate": 0.00013657646154630876, + "loss": 1.8563, + "step": 4467 + }, + { + "epoch": 0.39949928469241774, + "grad_norm": 2.0722820201738483, + "learning_rate": 0.00013654950662515678, + "loss": 1.9146, + "step": 4468 + }, + { + "epoch": 0.39958869814020026, + "grad_norm": 2.394259533015507, + "learning_rate": 0.00013652254863864185, + "loss": 1.825, + "step": 4469 + }, + { + "epoch": 0.39967811158798283, + "grad_norm": 2.330477789208813, + "learning_rate": 0.00013649558758902484, + "loss": 2.2785, + "step": 4470 + }, + { + "epoch": 0.3997675250357654, + "grad_norm": 0.6220616919331468, + "learning_rate": 0.000136468623478567, + "loss": 2.0463, + "step": 4471 + }, + { + "epoch": 0.3998569384835479, + "grad_norm": 1.3143972341079908, + "learning_rate": 0.00013644165630952973, + "loss": 1.6723, + "step": 4472 + }, + { + "epoch": 0.3999463519313305, + "grad_norm": 0.7041230089046507, + "learning_rate": 0.00013641468608417478, + "loss": 2.2926, + "step": 4473 + }, + { + "epoch": 0.400035765379113, + "grad_norm": 6.028460597105277, + "learning_rate": 0.00013638771280476405, + "loss": 7.3805, + "step": 4474 + }, + { + "epoch": 0.4001251788268956, + "grad_norm": 3.6061833971737673, + "learning_rate": 0.00013636073647355982, + "loss": 1.7758, + "step": 4475 + }, + { + "epoch": 0.4002145922746781, + "grad_norm": 3.9385078282255104, + "learning_rate": 0.00013633375709282453, + "loss": 2.1581, + "step": 4476 + }, + { + "epoch": 0.40030400572246067, + "grad_norm": 1.11147564756328, + "learning_rate": 0.00013630677466482092, + "loss": 1.9149, + "step": 4477 + }, + { + "epoch": 0.4003934191702432, + "grad_norm": 2.4435288062967206, + "learning_rate": 0.00013627978919181197, + "loss": 2.3313, + "step": 4478 + }, + { + "epoch": 0.40048283261802575, + "grad_norm": 3.0857934610870594, + "learning_rate": 0.0001362528006760609, + "loss": 1.9499, + "step": 4479 + }, + { + "epoch": 0.4005722460658083, + "grad_norm": 2.4193265559707737, + "learning_rate": 0.0001362258091198312, + "loss": 2.2362, + "step": 4480 + }, + { + "epoch": 0.40066165951359084, + "grad_norm": 1.2410259011092108, + "learning_rate": 0.0001361988145253867, + "loss": 2.1483, + "step": 4481 + }, + { + "epoch": 0.4007510729613734, + "grad_norm": 2.8798687068120477, + "learning_rate": 0.00013617181689499128, + "loss": 1.991, + "step": 4482 + }, + { + "epoch": 0.4008404864091559, + "grad_norm": 2.144634607032039, + "learning_rate": 0.00013614481623090932, + "loss": 2.436, + "step": 4483 + }, + { + "epoch": 0.4009298998569385, + "grad_norm": 1.9395546093102014, + "learning_rate": 0.00013611781253540522, + "loss": 1.614, + "step": 4484 + }, + { + "epoch": 0.401019313304721, + "grad_norm": 2.1230183111607857, + "learning_rate": 0.00013609080581074382, + "loss": 2.0885, + "step": 4485 + }, + { + "epoch": 0.4011087267525036, + "grad_norm": 1.9912648963013342, + "learning_rate": 0.00013606379605919013, + "loss": 1.9886, + "step": 4486 + }, + { + "epoch": 0.4011981402002861, + "grad_norm": 2.182812292575346, + "learning_rate": 0.00013603678328300939, + "loss": 1.963, + "step": 4487 + }, + { + "epoch": 0.4012875536480687, + "grad_norm": 0.8184976682572865, + "learning_rate": 0.0001360097674844672, + "loss": 1.9495, + "step": 4488 + }, + { + "epoch": 0.4013769670958512, + "grad_norm": 0.8724001747501305, + "learning_rate": 0.0001359827486658293, + "loss": 2.1601, + "step": 4489 + }, + { + "epoch": 0.40146638054363376, + "grad_norm": 1.544113730716254, + "learning_rate": 0.00013595572682936172, + "loss": 2.5324, + "step": 4490 + }, + { + "epoch": 0.40155579399141633, + "grad_norm": 1.132393636118192, + "learning_rate": 0.00013592870197733073, + "loss": 1.943, + "step": 4491 + }, + { + "epoch": 0.40164520743919885, + "grad_norm": 16.128373198044535, + "learning_rate": 0.0001359016741120029, + "loss": 7.4345, + "step": 4492 + }, + { + "epoch": 0.4017346208869814, + "grad_norm": 6.244969524034744, + "learning_rate": 0.00013587464323564503, + "loss": 2.0857, + "step": 4493 + }, + { + "epoch": 0.40182403433476394, + "grad_norm": 8.783190581341193, + "learning_rate": 0.00013584760935052417, + "loss": 2.3573, + "step": 4494 + }, + { + "epoch": 0.4019134477825465, + "grad_norm": 2.214995978573128, + "learning_rate": 0.00013582057245890757, + "loss": 1.6944, + "step": 4495 + }, + { + "epoch": 0.402002861230329, + "grad_norm": 4.750269039441815, + "learning_rate": 0.00013579353256306287, + "loss": 2.2905, + "step": 4496 + }, + { + "epoch": 0.4020922746781116, + "grad_norm": 2.856549086112824, + "learning_rate": 0.00013576648966525778, + "loss": 1.9003, + "step": 4497 + }, + { + "epoch": 0.4021816881258941, + "grad_norm": 0.8310046684450315, + "learning_rate": 0.00013573944376776042, + "loss": 1.9201, + "step": 4498 + }, + { + "epoch": 0.4022711015736767, + "grad_norm": 2.617548327149148, + "learning_rate": 0.00013571239487283906, + "loss": 2.0715, + "step": 4499 + }, + { + "epoch": 0.40236051502145925, + "grad_norm": 3.866436212455224, + "learning_rate": 0.00013568534298276228, + "loss": 2.0217, + "step": 4500 + }, + { + "epoch": 0.40244992846924177, + "grad_norm": 0.5385577440434184, + "learning_rate": 0.00013565828809979885, + "loss": 1.9178, + "step": 4501 + }, + { + "epoch": 0.40253934191702434, + "grad_norm": 1.8970649879749797, + "learning_rate": 0.0001356312302262179, + "loss": 2.1838, + "step": 4502 + }, + { + "epoch": 0.40262875536480686, + "grad_norm": 2.0420157419942893, + "learning_rate": 0.0001356041693642887, + "loss": 1.9575, + "step": 4503 + }, + { + "epoch": 0.4027181688125894, + "grad_norm": 3.2736586441422837, + "learning_rate": 0.0001355771055162808, + "loss": 2.3754, + "step": 4504 + }, + { + "epoch": 0.40280758226037194, + "grad_norm": 2.2707521609981494, + "learning_rate": 0.00013555003868446404, + "loss": 1.9049, + "step": 4505 + }, + { + "epoch": 0.4028969957081545, + "grad_norm": 2.1597066587147045, + "learning_rate": 0.00013552296887110846, + "loss": 1.8414, + "step": 4506 + }, + { + "epoch": 0.40298640915593703, + "grad_norm": 0.8141519792361622, + "learning_rate": 0.00013549589607848438, + "loss": 2.1776, + "step": 4507 + }, + { + "epoch": 0.4030758226037196, + "grad_norm": 1.8228367093312736, + "learning_rate": 0.00013546882030886237, + "loss": 1.8606, + "step": 4508 + }, + { + "epoch": 0.4031652360515021, + "grad_norm": 1.0199438584771374, + "learning_rate": 0.00013544174156451323, + "loss": 2.5999, + "step": 4509 + }, + { + "epoch": 0.4032546494992847, + "grad_norm": 0.6478935589625727, + "learning_rate": 0.00013541465984770804, + "loss": 2.34, + "step": 4510 + }, + { + "epoch": 0.40334406294706726, + "grad_norm": 1.2846085377452083, + "learning_rate": 0.00013538757516071807, + "loss": 2.263, + "step": 4511 + }, + { + "epoch": 0.4034334763948498, + "grad_norm": 2.1012019285750996, + "learning_rate": 0.00013536048750581494, + "loss": 2.1943, + "step": 4512 + }, + { + "epoch": 0.40352288984263235, + "grad_norm": 1.1910562169386485, + "learning_rate": 0.0001353333968852704, + "loss": 2.5022, + "step": 4513 + }, + { + "epoch": 0.40361230329041486, + "grad_norm": 1.298934260304746, + "learning_rate": 0.00013530630330135655, + "loss": 1.8017, + "step": 4514 + }, + { + "epoch": 0.40370171673819744, + "grad_norm": 1.5807525364122101, + "learning_rate": 0.0001352792067563457, + "loss": 2.167, + "step": 4515 + }, + { + "epoch": 0.40379113018597995, + "grad_norm": 0.7658724853988201, + "learning_rate": 0.00013525210725251035, + "loss": 2.1787, + "step": 4516 + }, + { + "epoch": 0.4038805436337625, + "grad_norm": 1.2986418946986087, + "learning_rate": 0.00013522500479212337, + "loss": 2.018, + "step": 4517 + }, + { + "epoch": 0.40396995708154504, + "grad_norm": 1.1474527687062064, + "learning_rate": 0.00013519789937745775, + "loss": 1.5411, + "step": 4518 + }, + { + "epoch": 0.4040593705293276, + "grad_norm": 3.1195398482591883, + "learning_rate": 0.00013517079101078684, + "loss": 2.8715, + "step": 4519 + }, + { + "epoch": 0.4041487839771102, + "grad_norm": 1.3367131377381725, + "learning_rate": 0.00013514367969438414, + "loss": 1.8158, + "step": 4520 + }, + { + "epoch": 0.4042381974248927, + "grad_norm": 1.1075478877848663, + "learning_rate": 0.0001351165654305235, + "loss": 2.1662, + "step": 4521 + }, + { + "epoch": 0.40432761087267527, + "grad_norm": 0.9938297982751446, + "learning_rate": 0.00013508944822147891, + "loss": 2.2842, + "step": 4522 + }, + { + "epoch": 0.4044170243204578, + "grad_norm": 1.3171737236694394, + "learning_rate": 0.00013506232806952467, + "loss": 2.1868, + "step": 4523 + }, + { + "epoch": 0.40450643776824036, + "grad_norm": 1.2254485847863972, + "learning_rate": 0.0001350352049769353, + "loss": 2.2183, + "step": 4524 + }, + { + "epoch": 0.4045958512160229, + "grad_norm": 0.6898309614444863, + "learning_rate": 0.00013500807894598565, + "loss": 2.3857, + "step": 4525 + }, + { + "epoch": 0.40468526466380544, + "grad_norm": 1.5447714045943384, + "learning_rate": 0.00013498094997895069, + "loss": 1.5404, + "step": 4526 + }, + { + "epoch": 0.40477467811158796, + "grad_norm": 0.9809713871313013, + "learning_rate": 0.00013495381807810569, + "loss": 2.2427, + "step": 4527 + }, + { + "epoch": 0.40486409155937053, + "grad_norm": 2.016592415938993, + "learning_rate": 0.00013492668324572614, + "loss": 2.3837, + "step": 4528 + }, + { + "epoch": 0.4049535050071531, + "grad_norm": 2.0825660853917536, + "learning_rate": 0.0001348995454840879, + "loss": 1.8506, + "step": 4529 + }, + { + "epoch": 0.4050429184549356, + "grad_norm": 2.1111978859714817, + "learning_rate": 0.00013487240479546691, + "loss": 2.3831, + "step": 4530 + }, + { + "epoch": 0.4051323319027182, + "grad_norm": 1.6946017704398606, + "learning_rate": 0.00013484526118213942, + "loss": 2.0134, + "step": 4531 + }, + { + "epoch": 0.4052217453505007, + "grad_norm": 2.1110249372123095, + "learning_rate": 0.000134818114646382, + "loss": 2.0195, + "step": 4532 + }, + { + "epoch": 0.4053111587982833, + "grad_norm": 1.8758808769203899, + "learning_rate": 0.00013479096519047136, + "loss": 2.255, + "step": 4533 + }, + { + "epoch": 0.4054005722460658, + "grad_norm": 1.984879331150686, + "learning_rate": 0.00013476381281668447, + "loss": 2.2311, + "step": 4534 + }, + { + "epoch": 0.40548998569384836, + "grad_norm": 1.7622914779955787, + "learning_rate": 0.00013473665752729855, + "loss": 2.1429, + "step": 4535 + }, + { + "epoch": 0.4055793991416309, + "grad_norm": 1.3111456689538334, + "learning_rate": 0.00013470949932459117, + "loss": 2.1413, + "step": 4536 + }, + { + "epoch": 0.40566881258941345, + "grad_norm": 2.3148663485284144, + "learning_rate": 0.00013468233821083996, + "loss": 2.2644, + "step": 4537 + }, + { + "epoch": 0.40575822603719597, + "grad_norm": 1.2640039541156114, + "learning_rate": 0.0001346551741883229, + "loss": 2.1919, + "step": 4538 + }, + { + "epoch": 0.40584763948497854, + "grad_norm": 1.0799705471351437, + "learning_rate": 0.0001346280072593183, + "loss": 2.3563, + "step": 4539 + }, + { + "epoch": 0.4059370529327611, + "grad_norm": 2.278763844506131, + "learning_rate": 0.00013460083742610455, + "loss": 1.7312, + "step": 4540 + }, + { + "epoch": 0.4060264663805436, + "grad_norm": 1.1090254357896674, + "learning_rate": 0.00013457366469096029, + "loss": 2.3839, + "step": 4541 + }, + { + "epoch": 0.4061158798283262, + "grad_norm": 3.0494683603248083, + "learning_rate": 0.00013454648905616458, + "loss": 2.5298, + "step": 4542 + }, + { + "epoch": 0.4062052932761087, + "grad_norm": 0.712733857490477, + "learning_rate": 0.00013451931052399656, + "loss": 2.0295, + "step": 4543 + }, + { + "epoch": 0.4062947067238913, + "grad_norm": 2.1940327973630964, + "learning_rate": 0.00013449212909673563, + "loss": 1.6992, + "step": 4544 + }, + { + "epoch": 0.4063841201716738, + "grad_norm": 1.567326847258456, + "learning_rate": 0.00013446494477666146, + "loss": 1.8098, + "step": 4545 + }, + { + "epoch": 0.4064735336194564, + "grad_norm": 1.0284814148799715, + "learning_rate": 0.00013443775756605405, + "loss": 1.7329, + "step": 4546 + }, + { + "epoch": 0.4065629470672389, + "grad_norm": 0.8592650113632299, + "learning_rate": 0.0001344105674671935, + "loss": 1.6064, + "step": 4547 + }, + { + "epoch": 0.40665236051502146, + "grad_norm": 1.2437627688106443, + "learning_rate": 0.00013438337448236015, + "loss": 1.9701, + "step": 4548 + }, + { + "epoch": 0.40674177396280403, + "grad_norm": 5.185860418836919, + "learning_rate": 0.0001343561786138348, + "loss": 1.6873, + "step": 4549 + }, + { + "epoch": 0.40683118741058655, + "grad_norm": 6.222296756968849, + "learning_rate": 0.00013432897986389818, + "loss": 1.9104, + "step": 4550 + }, + { + "epoch": 0.4069206008583691, + "grad_norm": 3.3850802622970355, + "learning_rate": 0.00013430177823483148, + "loss": 2.3825, + "step": 4551 + }, + { + "epoch": 0.40701001430615164, + "grad_norm": 2.564080988632183, + "learning_rate": 0.00013427457372891608, + "loss": 2.2842, + "step": 4552 + }, + { + "epoch": 0.4070994277539342, + "grad_norm": 5.351182587890252, + "learning_rate": 0.00013424736634843357, + "loss": 1.9957, + "step": 4553 + }, + { + "epoch": 0.4071888412017167, + "grad_norm": 1.5017971204502443, + "learning_rate": 0.0001342201560956658, + "loss": 1.9846, + "step": 4554 + }, + { + "epoch": 0.4072782546494993, + "grad_norm": 3.006363478038251, + "learning_rate": 0.00013419294297289486, + "loss": 2.0315, + "step": 4555 + }, + { + "epoch": 0.4073676680972818, + "grad_norm": 5.65119100190454, + "learning_rate": 0.00013416572698240312, + "loss": 2.1055, + "step": 4556 + }, + { + "epoch": 0.4074570815450644, + "grad_norm": 2.6161186325556725, + "learning_rate": 0.00013413850812647312, + "loss": 1.8068, + "step": 4557 + }, + { + "epoch": 0.4075464949928469, + "grad_norm": 1.6915869976810645, + "learning_rate": 0.00013411128640738762, + "loss": 2.1674, + "step": 4558 + }, + { + "epoch": 0.40763590844062947, + "grad_norm": 4.836315731547431, + "learning_rate": 0.00013408406182742976, + "loss": 2.2584, + "step": 4559 + }, + { + "epoch": 0.40772532188841204, + "grad_norm": 2.2818053259696858, + "learning_rate": 0.00013405683438888282, + "loss": 1.9935, + "step": 4560 + }, + { + "epoch": 0.40781473533619456, + "grad_norm": 1.948764992321564, + "learning_rate": 0.00013402960409403028, + "loss": 2.027, + "step": 4561 + }, + { + "epoch": 0.4079041487839771, + "grad_norm": 2.6575532917020235, + "learning_rate": 0.00013400237094515592, + "loss": 1.5306, + "step": 4562 + }, + { + "epoch": 0.40799356223175964, + "grad_norm": 2.789363818550645, + "learning_rate": 0.0001339751349445438, + "loss": 2.2109, + "step": 4563 + }, + { + "epoch": 0.4080829756795422, + "grad_norm": 0.9873451247015351, + "learning_rate": 0.00013394789609447817, + "loss": 2.2833, + "step": 4564 + }, + { + "epoch": 0.40817238912732473, + "grad_norm": 1.7905509049406958, + "learning_rate": 0.00013392065439724344, + "loss": 2.4413, + "step": 4565 + }, + { + "epoch": 0.4082618025751073, + "grad_norm": 2.6587580987140877, + "learning_rate": 0.00013389340985512442, + "loss": 2.3325, + "step": 4566 + }, + { + "epoch": 0.4083512160228898, + "grad_norm": 1.7103889078913788, + "learning_rate": 0.00013386616247040606, + "loss": 2.2613, + "step": 4567 + }, + { + "epoch": 0.4084406294706724, + "grad_norm": 1.3342605221208212, + "learning_rate": 0.00013383891224537354, + "loss": 1.8271, + "step": 4568 + }, + { + "epoch": 0.40853004291845496, + "grad_norm": 2.9928401865320162, + "learning_rate": 0.0001338116591823123, + "loss": 2.4999, + "step": 4569 + }, + { + "epoch": 0.4086194563662375, + "grad_norm": 2.035847594479642, + "learning_rate": 0.0001337844032835081, + "loss": 1.9651, + "step": 4570 + }, + { + "epoch": 0.40870886981402005, + "grad_norm": 2.0777820227735897, + "learning_rate": 0.0001337571445512467, + "loss": 2.1556, + "step": 4571 + }, + { + "epoch": 0.40879828326180256, + "grad_norm": 2.5162752622951197, + "learning_rate": 0.00013372988298781442, + "loss": 2.0057, + "step": 4572 + }, + { + "epoch": 0.40888769670958514, + "grad_norm": 1.5645012418369655, + "learning_rate": 0.00013370261859549758, + "loss": 2.297, + "step": 4573 + }, + { + "epoch": 0.40897711015736765, + "grad_norm": 1.5091263834199415, + "learning_rate": 0.00013367535137658282, + "loss": 2.3485, + "step": 4574 + }, + { + "epoch": 0.4090665236051502, + "grad_norm": 1.6565491597049582, + "learning_rate": 0.00013364808133335703, + "loss": 2.0671, + "step": 4575 + }, + { + "epoch": 0.40915593705293274, + "grad_norm": 2.0611870747652383, + "learning_rate": 0.00013362080846810725, + "loss": 1.8717, + "step": 4576 + }, + { + "epoch": 0.4092453505007153, + "grad_norm": 2.138579996616167, + "learning_rate": 0.0001335935327831209, + "loss": 1.99, + "step": 4577 + }, + { + "epoch": 0.4093347639484979, + "grad_norm": 1.4431326773512712, + "learning_rate": 0.0001335662542806855, + "loss": 2.1294, + "step": 4578 + }, + { + "epoch": 0.4094241773962804, + "grad_norm": 2.098353667713893, + "learning_rate": 0.00013353897296308892, + "loss": 1.8905, + "step": 4579 + }, + { + "epoch": 0.40951359084406297, + "grad_norm": 2.4041337046054183, + "learning_rate": 0.00013351168883261915, + "loss": 2.0369, + "step": 4580 + }, + { + "epoch": 0.4096030042918455, + "grad_norm": 0.941739151936225, + "learning_rate": 0.00013348440189156455, + "loss": 1.6495, + "step": 4581 + }, + { + "epoch": 0.40969241773962806, + "grad_norm": 0.9870262694702304, + "learning_rate": 0.00013345711214221359, + "loss": 1.9854, + "step": 4582 + }, + { + "epoch": 0.4097818311874106, + "grad_norm": 1.1781683655706088, + "learning_rate": 0.00013342981958685502, + "loss": 1.7087, + "step": 4583 + }, + { + "epoch": 0.40987124463519314, + "grad_norm": 2.352956165393678, + "learning_rate": 0.00013340252422777788, + "loss": 2.054, + "step": 4584 + }, + { + "epoch": 0.40996065808297566, + "grad_norm": 3.758141286393368, + "learning_rate": 0.00013337522606727132, + "loss": 2.3883, + "step": 4585 + }, + { + "epoch": 0.41005007153075823, + "grad_norm": 2.4197180486095187, + "learning_rate": 0.00013334792510762491, + "loss": 2.2525, + "step": 4586 + }, + { + "epoch": 0.41013948497854075, + "grad_norm": 2.967293454462834, + "learning_rate": 0.0001333206213511283, + "loss": 2.0688, + "step": 4587 + }, + { + "epoch": 0.4102288984263233, + "grad_norm": 2.7706907861372314, + "learning_rate": 0.00013329331480007139, + "loss": 1.7605, + "step": 4588 + }, + { + "epoch": 0.4103183118741059, + "grad_norm": 0.8490068539607566, + "learning_rate": 0.0001332660054567444, + "loss": 1.8909, + "step": 4589 + }, + { + "epoch": 0.4104077253218884, + "grad_norm": 2.331361893262168, + "learning_rate": 0.00013323869332343768, + "loss": 1.7458, + "step": 4590 + }, + { + "epoch": 0.410497138769671, + "grad_norm": 2.6297235913895194, + "learning_rate": 0.00013321137840244192, + "loss": 1.8204, + "step": 4591 + }, + { + "epoch": 0.4105865522174535, + "grad_norm": 0.49753116091709765, + "learning_rate": 0.00013318406069604794, + "loss": 1.6781, + "step": 4592 + }, + { + "epoch": 0.41067596566523606, + "grad_norm": 0.8765670316392377, + "learning_rate": 0.00013315674020654688, + "loss": 1.6714, + "step": 4593 + }, + { + "epoch": 0.4107653791130186, + "grad_norm": 3.664349184206248, + "learning_rate": 0.00013312941693623004, + "loss": 2.1893, + "step": 4594 + }, + { + "epoch": 0.41085479256080115, + "grad_norm": 2.5445810078824045, + "learning_rate": 0.00013310209088738902, + "loss": 2.1467, + "step": 4595 + }, + { + "epoch": 0.41094420600858367, + "grad_norm": 2.7013818558300677, + "learning_rate": 0.00013307476206231563, + "loss": 2.3741, + "step": 4596 + }, + { + "epoch": 0.41103361945636624, + "grad_norm": 1.6440111313308339, + "learning_rate": 0.0001330474304633019, + "loss": 1.9206, + "step": 4597 + }, + { + "epoch": 0.4111230329041488, + "grad_norm": 3.7094410361768353, + "learning_rate": 0.00013302009609264005, + "loss": 2.2781, + "step": 4598 + }, + { + "epoch": 0.4112124463519313, + "grad_norm": 1.7789340194181895, + "learning_rate": 0.00013299275895262266, + "loss": 2.4518, + "step": 4599 + }, + { + "epoch": 0.4113018597997139, + "grad_norm": 2.8052476010800906, + "learning_rate": 0.00013296541904554238, + "loss": 1.981, + "step": 4600 + }, + { + "epoch": 0.4113912732474964, + "grad_norm": 3.510328713893564, + "learning_rate": 0.00013293807637369226, + "loss": 2.3093, + "step": 4601 + }, + { + "epoch": 0.411480686695279, + "grad_norm": 0.8574051219623832, + "learning_rate": 0.00013291073093936543, + "loss": 2.3675, + "step": 4602 + }, + { + "epoch": 0.4115701001430615, + "grad_norm": 2.7362568995168894, + "learning_rate": 0.00013288338274485532, + "loss": 2.4766, + "step": 4603 + }, + { + "epoch": 0.4116595135908441, + "grad_norm": 3.719992335801898, + "learning_rate": 0.00013285603179245565, + "loss": 2.0886, + "step": 4604 + }, + { + "epoch": 0.4117489270386266, + "grad_norm": 4.997104864878925, + "learning_rate": 0.00013282867808446025, + "loss": 2.3034, + "step": 4605 + }, + { + "epoch": 0.41183834048640916, + "grad_norm": 3.425643910440218, + "learning_rate": 0.0001328013216231633, + "loss": 1.992, + "step": 4606 + }, + { + "epoch": 0.4119277539341917, + "grad_norm": 4.0720744317529105, + "learning_rate": 0.00013277396241085908, + "loss": 1.828, + "step": 4607 + }, + { + "epoch": 0.41201716738197425, + "grad_norm": 4.173845332857577, + "learning_rate": 0.00013274660044984224, + "loss": 2.094, + "step": 4608 + }, + { + "epoch": 0.4121065808297568, + "grad_norm": 0.7644228055086941, + "learning_rate": 0.00013271923574240756, + "loss": 1.6571, + "step": 4609 + }, + { + "epoch": 0.41219599427753933, + "grad_norm": 6.878203410148376, + "learning_rate": 0.00013269186829085008, + "loss": 1.6541, + "step": 4610 + }, + { + "epoch": 0.4122854077253219, + "grad_norm": 2.908998411886879, + "learning_rate": 0.0001326644980974651, + "loss": 2.1233, + "step": 4611 + }, + { + "epoch": 0.4123748211731044, + "grad_norm": 3.030829485915841, + "learning_rate": 0.00013263712516454812, + "loss": 2.669, + "step": 4612 + }, + { + "epoch": 0.412464234620887, + "grad_norm": 1.6834008513648597, + "learning_rate": 0.00013260974949439485, + "loss": 1.8557, + "step": 4613 + }, + { + "epoch": 0.4125536480686695, + "grad_norm": 5.879303875378204, + "learning_rate": 0.00013258237108930128, + "loss": 1.7543, + "step": 4614 + }, + { + "epoch": 0.4126430615164521, + "grad_norm": 3.3435530056123293, + "learning_rate": 0.0001325549899515636, + "loss": 2.1055, + "step": 4615 + }, + { + "epoch": 0.4127324749642346, + "grad_norm": 1.9000090092396356, + "learning_rate": 0.00013252760608347826, + "loss": 1.8811, + "step": 4616 + }, + { + "epoch": 0.41282188841201717, + "grad_norm": 5.992555864123145, + "learning_rate": 0.00013250021948734184, + "loss": 2.3096, + "step": 4617 + }, + { + "epoch": 0.41291130185979974, + "grad_norm": 2.942173325952129, + "learning_rate": 0.00013247283016545126, + "loss": 2.2775, + "step": 4618 + }, + { + "epoch": 0.41300071530758226, + "grad_norm": 0.9931449444135138, + "learning_rate": 0.00013244543812010364, + "loss": 2.1882, + "step": 4619 + }, + { + "epoch": 0.4130901287553648, + "grad_norm": 1.762137794417201, + "learning_rate": 0.00013241804335359633, + "loss": 1.8747, + "step": 4620 + }, + { + "epoch": 0.41317954220314734, + "grad_norm": 15.478405131460256, + "learning_rate": 0.00013239064586822685, + "loss": 7.5989, + "step": 4621 + }, + { + "epoch": 0.4132689556509299, + "grad_norm": 7.465580456436995, + "learning_rate": 0.000132363245666293, + "loss": 1.3794, + "step": 4622 + }, + { + "epoch": 0.41335836909871243, + "grad_norm": 29.680210913892815, + "learning_rate": 0.00013233584275009288, + "loss": 2.1875, + "step": 4623 + }, + { + "epoch": 0.413447782546495, + "grad_norm": 8.363402124253893, + "learning_rate": 0.00013230843712192463, + "loss": 2.2377, + "step": 4624 + }, + { + "epoch": 0.4135371959942775, + "grad_norm": 15.185677110259935, + "learning_rate": 0.0001322810287840868, + "loss": 2.6611, + "step": 4625 + }, + { + "epoch": 0.4136266094420601, + "grad_norm": 3.2482676084670365, + "learning_rate": 0.00013225361773887804, + "loss": 1.8598, + "step": 4626 + }, + { + "epoch": 0.4137160228898426, + "grad_norm": 18.06732496964694, + "learning_rate": 0.00013222620398859738, + "loss": 2.0717, + "step": 4627 + }, + { + "epoch": 0.4138054363376252, + "grad_norm": 13.607511096823245, + "learning_rate": 0.00013219878753554384, + "loss": 2.014, + "step": 4628 + }, + { + "epoch": 0.41389484978540775, + "grad_norm": 22.996258537921747, + "learning_rate": 0.0001321713683820169, + "loss": 2.0915, + "step": 4629 + }, + { + "epoch": 0.41398426323319026, + "grad_norm": 8.776819566370774, + "learning_rate": 0.00013214394653031616, + "loss": 2.2525, + "step": 4630 + }, + { + "epoch": 0.41407367668097284, + "grad_norm": 14.568854502946488, + "learning_rate": 0.00013211652198274145, + "loss": 2.5645, + "step": 4631 + }, + { + "epoch": 0.41416309012875535, + "grad_norm": 5.20571241053806, + "learning_rate": 0.0001320890947415928, + "loss": 2.3481, + "step": 4632 + }, + { + "epoch": 0.4142525035765379, + "grad_norm": 2.1179640239617257, + "learning_rate": 0.00013206166480917055, + "loss": 2.0807, + "step": 4633 + }, + { + "epoch": 0.41434191702432044, + "grad_norm": 12.765823603088466, + "learning_rate": 0.0001320342321877752, + "loss": 1.9104, + "step": 4634 + }, + { + "epoch": 0.414431330472103, + "grad_norm": 6.560718095492326, + "learning_rate": 0.00013200679687970748, + "loss": 2.1364, + "step": 4635 + }, + { + "epoch": 0.4145207439198855, + "grad_norm": 4.129222531870963, + "learning_rate": 0.00013197935888726832, + "loss": 2.4313, + "step": 4636 + }, + { + "epoch": 0.4146101573676681, + "grad_norm": 48.214166523332615, + "learning_rate": 0.000131951918212759, + "loss": 2.0654, + "step": 4637 + }, + { + "epoch": 0.41469957081545067, + "grad_norm": 20.418767410644833, + "learning_rate": 0.00013192447485848088, + "loss": 2.372, + "step": 4638 + }, + { + "epoch": 0.4147889842632332, + "grad_norm": 14.502240109442067, + "learning_rate": 0.00013189702882673556, + "loss": 2.2536, + "step": 4639 + }, + { + "epoch": 0.41487839771101576, + "grad_norm": 11.553899399910858, + "learning_rate": 0.00013186958011982502, + "loss": 2.2431, + "step": 4640 + }, + { + "epoch": 0.41496781115879827, + "grad_norm": 2.709798350613302, + "learning_rate": 0.00013184212874005124, + "loss": 7.3757, + "step": 4641 + }, + { + "epoch": 0.41505722460658084, + "grad_norm": 10.253667724715832, + "learning_rate": 0.0001318146746897166, + "loss": 2.5473, + "step": 4642 + }, + { + "epoch": 0.41514663805436336, + "grad_norm": 8.770093842117694, + "learning_rate": 0.00013178721797112362, + "loss": 2.2485, + "step": 4643 + }, + { + "epoch": 0.41523605150214593, + "grad_norm": 2.461832448309122, + "learning_rate": 0.00013175975858657505, + "loss": 1.6458, + "step": 4644 + }, + { + "epoch": 0.41532546494992845, + "grad_norm": 3.937059287153402, + "learning_rate": 0.00013173229653837387, + "loss": 2.532, + "step": 4645 + }, + { + "epoch": 0.415414878397711, + "grad_norm": 3.877244415310884, + "learning_rate": 0.0001317048318288233, + "loss": 2.5531, + "step": 4646 + }, + { + "epoch": 0.4155042918454936, + "grad_norm": 2.8995823303913375, + "learning_rate": 0.0001316773644602268, + "loss": 2.2075, + "step": 4647 + }, + { + "epoch": 0.4155937052932761, + "grad_norm": 1.9890107991512667, + "learning_rate": 0.00013164989443488798, + "loss": 1.6079, + "step": 4648 + }, + { + "epoch": 0.4156831187410587, + "grad_norm": 2.513771690669936, + "learning_rate": 0.00013162242175511076, + "loss": 1.8266, + "step": 4649 + }, + { + "epoch": 0.4157725321888412, + "grad_norm": 1.5572255346221764, + "learning_rate": 0.0001315949464231992, + "loss": 2.1144, + "step": 4650 + }, + { + "epoch": 0.41586194563662376, + "grad_norm": 2.8559824803257774, + "learning_rate": 0.00013156746844145766, + "loss": 2.5545, + "step": 4651 + }, + { + "epoch": 0.4159513590844063, + "grad_norm": 2.160841280558734, + "learning_rate": 0.00013153998781219062, + "loss": 2.1208, + "step": 4652 + }, + { + "epoch": 0.41604077253218885, + "grad_norm": 5.564418233869081, + "learning_rate": 0.00013151250453770293, + "loss": 2.2088, + "step": 4653 + }, + { + "epoch": 0.41613018597997137, + "grad_norm": 2.6012410525384504, + "learning_rate": 0.00013148501862029954, + "loss": 2.1003, + "step": 4654 + }, + { + "epoch": 0.41621959942775394, + "grad_norm": 2.881361744463645, + "learning_rate": 0.00013145753006228565, + "loss": 2.379, + "step": 4655 + }, + { + "epoch": 0.41630901287553645, + "grad_norm": 6.6368861930036305, + "learning_rate": 0.00013143003886596669, + "loss": 2.8031, + "step": 4656 + }, + { + "epoch": 0.416398426323319, + "grad_norm": 1.6428741561463798, + "learning_rate": 0.00013140254503364837, + "loss": 2.0048, + "step": 4657 + }, + { + "epoch": 0.4164878397711016, + "grad_norm": 2.288253053122755, + "learning_rate": 0.00013137504856763652, + "loss": 2.2477, + "step": 4658 + }, + { + "epoch": 0.4165772532188841, + "grad_norm": 2.017834192462023, + "learning_rate": 0.0001313475494702372, + "loss": 1.9035, + "step": 4659 + }, + { + "epoch": 0.4166666666666667, + "grad_norm": 2.673767449988665, + "learning_rate": 0.0001313200477437568, + "loss": 2.2457, + "step": 4660 + }, + { + "epoch": 0.4167560801144492, + "grad_norm": 4.239623239496352, + "learning_rate": 0.00013129254339050181, + "loss": 2.2625, + "step": 4661 + }, + { + "epoch": 0.4168454935622318, + "grad_norm": 3.1379041576402273, + "learning_rate": 0.00013126503641277897, + "loss": 2.2641, + "step": 4662 + }, + { + "epoch": 0.4169349070100143, + "grad_norm": 3.635224255684741, + "learning_rate": 0.00013123752681289529, + "loss": 2.2083, + "step": 4663 + }, + { + "epoch": 0.41702432045779686, + "grad_norm": 2.109207087198758, + "learning_rate": 0.000131210014593158, + "loss": 1.9455, + "step": 4664 + }, + { + "epoch": 0.4171137339055794, + "grad_norm": 2.1485518374215444, + "learning_rate": 0.00013118249975587447, + "loss": 2.0093, + "step": 4665 + }, + { + "epoch": 0.41720314735336195, + "grad_norm": 1.6981874938937354, + "learning_rate": 0.0001311549823033523, + "loss": 2.0945, + "step": 4666 + }, + { + "epoch": 0.4172925608011445, + "grad_norm": 3.067242548076855, + "learning_rate": 0.0001311274622378994, + "loss": 2.301, + "step": 4667 + }, + { + "epoch": 0.41738197424892703, + "grad_norm": 1.4180868499808228, + "learning_rate": 0.0001310999395618239, + "loss": 2.3743, + "step": 4668 + }, + { + "epoch": 0.4174713876967096, + "grad_norm": 2.625106699370849, + "learning_rate": 0.00013107241427743398, + "loss": 2.1318, + "step": 4669 + }, + { + "epoch": 0.4175608011444921, + "grad_norm": 1.898944135642483, + "learning_rate": 0.0001310448863870382, + "loss": 2.7866, + "step": 4670 + }, + { + "epoch": 0.4176502145922747, + "grad_norm": 1.4800530045194846, + "learning_rate": 0.0001310173558929453, + "loss": 2.1573, + "step": 4671 + }, + { + "epoch": 0.4177396280400572, + "grad_norm": 2.5181499121659594, + "learning_rate": 0.00013098982279746422, + "loss": 2.0482, + "step": 4672 + }, + { + "epoch": 0.4178290414878398, + "grad_norm": 2.129673750091285, + "learning_rate": 0.0001309622871029041, + "loss": 2.0511, + "step": 4673 + }, + { + "epoch": 0.4179184549356223, + "grad_norm": 1.825569337631238, + "learning_rate": 0.00013093474881157438, + "loss": 1.9771, + "step": 4674 + }, + { + "epoch": 0.41800786838340487, + "grad_norm": 1.4314526230126825, + "learning_rate": 0.00013090720792578465, + "loss": 2.2614, + "step": 4675 + }, + { + "epoch": 0.4180972818311874, + "grad_norm": 2.186378293586268, + "learning_rate": 0.00013087966444784468, + "loss": 1.8492, + "step": 4676 + }, + { + "epoch": 0.41818669527896996, + "grad_norm": 2.184696776424839, + "learning_rate": 0.00013085211838006458, + "loss": 1.9767, + "step": 4677 + }, + { + "epoch": 0.4182761087267525, + "grad_norm": 2.1823127850380737, + "learning_rate": 0.00013082456972475458, + "loss": 1.9796, + "step": 4678 + }, + { + "epoch": 0.41836552217453504, + "grad_norm": 1.762806614836065, + "learning_rate": 0.0001307970184842251, + "loss": 2.0032, + "step": 4679 + }, + { + "epoch": 0.4184549356223176, + "grad_norm": 1.1034584656080768, + "learning_rate": 0.0001307694646607869, + "loss": 2.1005, + "step": 4680 + }, + { + "epoch": 0.41854434907010013, + "grad_norm": 3.3356871795862806, + "learning_rate": 0.00013074190825675087, + "loss": 2.0049, + "step": 4681 + }, + { + "epoch": 0.4186337625178827, + "grad_norm": 2.0418559381908734, + "learning_rate": 0.00013071434927442813, + "loss": 2.4784, + "step": 4682 + }, + { + "epoch": 0.4187231759656652, + "grad_norm": 1.0772559143663174, + "learning_rate": 0.00013068678771612996, + "loss": 1.9721, + "step": 4683 + }, + { + "epoch": 0.4188125894134478, + "grad_norm": 1.3079975254196814, + "learning_rate": 0.00013065922358416798, + "loss": 2.6361, + "step": 4684 + }, + { + "epoch": 0.4189020028612303, + "grad_norm": 1.984862939097251, + "learning_rate": 0.00013063165688085397, + "loss": 2.2618, + "step": 4685 + }, + { + "epoch": 0.4189914163090129, + "grad_norm": 1.2157098491189875, + "learning_rate": 0.00013060408760849987, + "loss": 1.9959, + "step": 4686 + }, + { + "epoch": 0.41908082975679545, + "grad_norm": 0.9657261193582612, + "learning_rate": 0.00013057651576941793, + "loss": 1.9791, + "step": 4687 + }, + { + "epoch": 0.41917024320457796, + "grad_norm": 1.8428013548752484, + "learning_rate": 0.00013054894136592052, + "loss": 2.018, + "step": 4688 + }, + { + "epoch": 0.41925965665236054, + "grad_norm": 2.4142112217715357, + "learning_rate": 0.00013052136440032028, + "loss": 1.7713, + "step": 4689 + }, + { + "epoch": 0.41934907010014305, + "grad_norm": 1.0558703173645596, + "learning_rate": 0.00013049378487493008, + "loss": 2.1359, + "step": 4690 + }, + { + "epoch": 0.4194384835479256, + "grad_norm": 4.643038670415777, + "learning_rate": 0.00013046620279206296, + "loss": 1.9923, + "step": 4691 + }, + { + "epoch": 0.41952789699570814, + "grad_norm": 0.9374432060284165, + "learning_rate": 0.00013043861815403225, + "loss": 1.9128, + "step": 4692 + }, + { + "epoch": 0.4196173104434907, + "grad_norm": 1.5177286912824117, + "learning_rate": 0.0001304110309631513, + "loss": 2.1328, + "step": 4693 + }, + { + "epoch": 0.4197067238912732, + "grad_norm": 0.8509962360455514, + "learning_rate": 0.000130383441221734, + "loss": 2.369, + "step": 4694 + }, + { + "epoch": 0.4197961373390558, + "grad_norm": 3.7865696746327107, + "learning_rate": 0.00013035584893209416, + "loss": 7.3655, + "step": 4695 + }, + { + "epoch": 0.4198855507868383, + "grad_norm": 4.003118702449356, + "learning_rate": 0.00013032825409654592, + "loss": 1.9902, + "step": 4696 + }, + { + "epoch": 0.4199749642346209, + "grad_norm": 3.1655909946050023, + "learning_rate": 0.00013030065671740363, + "loss": 2.2481, + "step": 4697 + }, + { + "epoch": 0.42006437768240346, + "grad_norm": 1.6052828031950352, + "learning_rate": 0.00013027305679698186, + "loss": 2.0371, + "step": 4698 + }, + { + "epoch": 0.42015379113018597, + "grad_norm": 4.706727091257522, + "learning_rate": 0.00013024545433759538, + "loss": 2.4673, + "step": 4699 + }, + { + "epoch": 0.42024320457796854, + "grad_norm": 2.0733395478139256, + "learning_rate": 0.00013021784934155915, + "loss": 2.6236, + "step": 4700 + }, + { + "epoch": 0.42033261802575106, + "grad_norm": 2.3015766714952317, + "learning_rate": 0.00013019024181118845, + "loss": 2.0365, + "step": 4701 + }, + { + "epoch": 0.42042203147353363, + "grad_norm": 2.6041343697735857, + "learning_rate": 0.00013016263174879858, + "loss": 2.3565, + "step": 4702 + }, + { + "epoch": 0.42051144492131615, + "grad_norm": 0.8013548176302863, + "learning_rate": 0.00013013501915670522, + "loss": 2.4092, + "step": 4703 + }, + { + "epoch": 0.4206008583690987, + "grad_norm": 2.0523393336973004, + "learning_rate": 0.0001301074040372242, + "loss": 1.7443, + "step": 4704 + }, + { + "epoch": 0.42069027181688123, + "grad_norm": 1.8674334752913622, + "learning_rate": 0.0001300797863926716, + "loss": 2.081, + "step": 4705 + }, + { + "epoch": 0.4207796852646638, + "grad_norm": 0.8203403297907276, + "learning_rate": 0.00013005216622536355, + "loss": 1.5068, + "step": 4706 + }, + { + "epoch": 0.4208690987124464, + "grad_norm": 1.859534385234513, + "learning_rate": 0.00013002454353761665, + "loss": 2.338, + "step": 4707 + }, + { + "epoch": 0.4209585121602289, + "grad_norm": 1.5941537653046847, + "learning_rate": 0.0001299969183317476, + "loss": 2.1213, + "step": 4708 + }, + { + "epoch": 0.42104792560801146, + "grad_norm": 1.0910708036284125, + "learning_rate": 0.0001299692906100732, + "loss": 1.8278, + "step": 4709 + }, + { + "epoch": 0.421137339055794, + "grad_norm": 1.6348552325933605, + "learning_rate": 0.00012994166037491058, + "loss": 2.0549, + "step": 4710 + }, + { + "epoch": 0.42122675250357655, + "grad_norm": 5.531767503280408, + "learning_rate": 0.00012991402762857707, + "loss": 7.1812, + "step": 4711 + }, + { + "epoch": 0.42131616595135907, + "grad_norm": 5.154140460240618, + "learning_rate": 0.00012988639237339022, + "loss": 2.0739, + "step": 4712 + }, + { + "epoch": 0.42140557939914164, + "grad_norm": 2.7366108088111427, + "learning_rate": 0.0001298587546116677, + "loss": 1.9517, + "step": 4713 + }, + { + "epoch": 0.42149499284692415, + "grad_norm": 1.8194274038411582, + "learning_rate": 0.00012983111434572748, + "loss": 1.8637, + "step": 4714 + }, + { + "epoch": 0.4215844062947067, + "grad_norm": 3.3043578374177125, + "learning_rate": 0.00012980347157788777, + "loss": 6.9976, + "step": 4715 + }, + { + "epoch": 0.4216738197424893, + "grad_norm": 4.351540796460803, + "learning_rate": 0.00012977582631046685, + "loss": 2.5019, + "step": 4716 + }, + { + "epoch": 0.4217632331902718, + "grad_norm": 2.332112229344236, + "learning_rate": 0.0001297481785457834, + "loss": 1.8759, + "step": 4717 + }, + { + "epoch": 0.4218526466380544, + "grad_norm": 2.1612230622423687, + "learning_rate": 0.00012972052828615606, + "loss": 1.7075, + "step": 4718 + }, + { + "epoch": 0.4219420600858369, + "grad_norm": 3.355997431332186, + "learning_rate": 0.00012969287553390397, + "loss": 2.1792, + "step": 4719 + }, + { + "epoch": 0.4220314735336195, + "grad_norm": 1.5697426077975283, + "learning_rate": 0.00012966522029134623, + "loss": 2.1121, + "step": 4720 + }, + { + "epoch": 0.422120886981402, + "grad_norm": 1.1313100494195303, + "learning_rate": 0.0001296375625608023, + "loss": 2.0369, + "step": 4721 + }, + { + "epoch": 0.42221030042918456, + "grad_norm": 1.2516569788295673, + "learning_rate": 0.0001296099023445918, + "loss": 1.8049, + "step": 4722 + }, + { + "epoch": 0.4222997138769671, + "grad_norm": 3.325649068193561, + "learning_rate": 0.00012958223964503452, + "loss": 2.3323, + "step": 4723 + }, + { + "epoch": 0.42238912732474965, + "grad_norm": 0.8046016058985275, + "learning_rate": 0.00012955457446445055, + "loss": 1.8095, + "step": 4724 + }, + { + "epoch": 0.42247854077253216, + "grad_norm": 1.4293614089341349, + "learning_rate": 0.00012952690680516016, + "loss": 1.7247, + "step": 4725 + }, + { + "epoch": 0.42256795422031473, + "grad_norm": 3.431282296094077, + "learning_rate": 0.0001294992366694837, + "loss": 2.4906, + "step": 4726 + }, + { + "epoch": 0.4226573676680973, + "grad_norm": 1.9692513716260585, + "learning_rate": 0.00012947156405974187, + "loss": 1.9484, + "step": 4727 + }, + { + "epoch": 0.4227467811158798, + "grad_norm": 2.2287033561840497, + "learning_rate": 0.0001294438889782556, + "loss": 2.2101, + "step": 4728 + }, + { + "epoch": 0.4228361945636624, + "grad_norm": 2.419255033724992, + "learning_rate": 0.00012941621142734594, + "loss": 1.8244, + "step": 4729 + }, + { + "epoch": 0.4229256080114449, + "grad_norm": 1.8247719169217187, + "learning_rate": 0.00012938853140933407, + "loss": 1.7356, + "step": 4730 + }, + { + "epoch": 0.4230150214592275, + "grad_norm": 0.8300697496749112, + "learning_rate": 0.0001293608489265416, + "loss": 1.8756, + "step": 4731 + }, + { + "epoch": 0.42310443490701, + "grad_norm": 0.879929494858071, + "learning_rate": 0.00012933316398129022, + "loss": 1.5089, + "step": 4732 + }, + { + "epoch": 0.42319384835479257, + "grad_norm": 1.5522507093338174, + "learning_rate": 0.00012930547657590179, + "loss": 2.2873, + "step": 4733 + }, + { + "epoch": 0.4232832618025751, + "grad_norm": 1.0261077753845267, + "learning_rate": 0.00012927778671269842, + "loss": 1.8963, + "step": 4734 + }, + { + "epoch": 0.42337267525035766, + "grad_norm": 8.98379460879351, + "learning_rate": 0.00012925009439400243, + "loss": 7.2226, + "step": 4735 + }, + { + "epoch": 0.4234620886981402, + "grad_norm": 5.025359805103211, + "learning_rate": 0.00012922239962213637, + "loss": 2.2378, + "step": 4736 + }, + { + "epoch": 0.42355150214592274, + "grad_norm": 3.1756895864484105, + "learning_rate": 0.00012919470239942292, + "loss": 1.7376, + "step": 4737 + }, + { + "epoch": 0.4236409155937053, + "grad_norm": 2.292127692075983, + "learning_rate": 0.00012916700272818505, + "loss": 1.8625, + "step": 4738 + }, + { + "epoch": 0.42373032904148783, + "grad_norm": 1.4312666751281564, + "learning_rate": 0.00012913930061074592, + "loss": 1.5796, + "step": 4739 + }, + { + "epoch": 0.4238197424892704, + "grad_norm": 0.9145225940222956, + "learning_rate": 0.00012911159604942879, + "loss": 1.9997, + "step": 4740 + }, + { + "epoch": 0.4239091559370529, + "grad_norm": 3.2362756893322846, + "learning_rate": 0.0001290838890465573, + "loss": 2.3562, + "step": 4741 + }, + { + "epoch": 0.4239985693848355, + "grad_norm": 4.866602884617317, + "learning_rate": 0.00012905617960445512, + "loss": 2.2967, + "step": 4742 + }, + { + "epoch": 0.424087982832618, + "grad_norm": 1.1918588035136592, + "learning_rate": 0.00012902846772544624, + "loss": 2.2065, + "step": 4743 + }, + { + "epoch": 0.4241773962804006, + "grad_norm": 3.6972268744343175, + "learning_rate": 0.00012900075341185487, + "loss": 1.8464, + "step": 4744 + }, + { + "epoch": 0.4242668097281831, + "grad_norm": 3.728262379597152, + "learning_rate": 0.0001289730366660053, + "loss": 2.1095, + "step": 4745 + }, + { + "epoch": 0.42435622317596566, + "grad_norm": 1.1796357669945075, + "learning_rate": 0.00012894531749022217, + "loss": 1.7061, + "step": 4746 + }, + { + "epoch": 0.42444563662374823, + "grad_norm": 1.865904256063892, + "learning_rate": 0.00012891759588683018, + "loss": 1.6306, + "step": 4747 + }, + { + "epoch": 0.42453505007153075, + "grad_norm": 2.4400208045297114, + "learning_rate": 0.0001288898718581544, + "loss": 1.6972, + "step": 4748 + }, + { + "epoch": 0.4246244635193133, + "grad_norm": 1.6364255899177351, + "learning_rate": 0.0001288621454065199, + "loss": 1.7571, + "step": 4749 + }, + { + "epoch": 0.42471387696709584, + "grad_norm": 0.8371995149344681, + "learning_rate": 0.00012883441653425214, + "loss": 2.1086, + "step": 4750 + }, + { + "epoch": 0.4248032904148784, + "grad_norm": 1.7495789899867296, + "learning_rate": 0.00012880668524367672, + "loss": 2.0479, + "step": 4751 + }, + { + "epoch": 0.4248927038626609, + "grad_norm": 1.9924281407167017, + "learning_rate": 0.00012877895153711935, + "loss": 2.625, + "step": 4752 + }, + { + "epoch": 0.4249821173104435, + "grad_norm": 2.6357411109300917, + "learning_rate": 0.0001287512154169061, + "loss": 2.2795, + "step": 4753 + }, + { + "epoch": 0.425071530758226, + "grad_norm": 2.09087930909215, + "learning_rate": 0.00012872347688536312, + "loss": 2.47, + "step": 4754 + }, + { + "epoch": 0.4251609442060086, + "grad_norm": 11.493547079687447, + "learning_rate": 0.00012869573594481685, + "loss": 7.2571, + "step": 4755 + }, + { + "epoch": 0.42525035765379116, + "grad_norm": 10.940621634037988, + "learning_rate": 0.00012866799259759386, + "loss": 1.7848, + "step": 4756 + }, + { + "epoch": 0.42533977110157367, + "grad_norm": 4.6139375972131145, + "learning_rate": 0.0001286402468460209, + "loss": 2.0454, + "step": 4757 + }, + { + "epoch": 0.42542918454935624, + "grad_norm": 1.968161281389542, + "learning_rate": 0.0001286124986924251, + "loss": 1.9102, + "step": 4758 + }, + { + "epoch": 0.42551859799713876, + "grad_norm": 7.8348935803949376, + "learning_rate": 0.00012858474813913352, + "loss": 2.3844, + "step": 4759 + }, + { + "epoch": 0.42560801144492133, + "grad_norm": 6.434664612188854, + "learning_rate": 0.00012855699518847367, + "loss": 2.2458, + "step": 4760 + }, + { + "epoch": 0.42569742489270385, + "grad_norm": 1.3605276918443618, + "learning_rate": 0.00012852923984277314, + "loss": 1.9124, + "step": 4761 + }, + { + "epoch": 0.4257868383404864, + "grad_norm": 3.6974906447735334, + "learning_rate": 0.0001285014821043597, + "loss": 2.2323, + "step": 4762 + }, + { + "epoch": 0.42587625178826893, + "grad_norm": 4.396317797645202, + "learning_rate": 0.00012847372197556138, + "loss": 2.1357, + "step": 4763 + }, + { + "epoch": 0.4259656652360515, + "grad_norm": 2.210053306091973, + "learning_rate": 0.00012844595945870637, + "loss": 2.1177, + "step": 4764 + }, + { + "epoch": 0.426055078683834, + "grad_norm": 0.917325359829935, + "learning_rate": 0.00012841819455612313, + "loss": 2.5547, + "step": 4765 + }, + { + "epoch": 0.4261444921316166, + "grad_norm": 1.8944228319692467, + "learning_rate": 0.0001283904272701402, + "loss": 1.8599, + "step": 4766 + }, + { + "epoch": 0.42623390557939916, + "grad_norm": 2.689014065056753, + "learning_rate": 0.0001283626576030864, + "loss": 2.0239, + "step": 4767 + }, + { + "epoch": 0.4263233190271817, + "grad_norm": 1.2479934382102265, + "learning_rate": 0.0001283348855572908, + "loss": 2.2986, + "step": 4768 + }, + { + "epoch": 0.42641273247496425, + "grad_norm": 1.053144010621932, + "learning_rate": 0.00012830711113508256, + "loss": 1.8103, + "step": 4769 + }, + { + "epoch": 0.42650214592274677, + "grad_norm": 1.2411180991153443, + "learning_rate": 0.0001282793343387911, + "loss": 2.0338, + "step": 4770 + }, + { + "epoch": 0.42659155937052934, + "grad_norm": 14.49238241705603, + "learning_rate": 0.000128251555170746, + "loss": 7.3239, + "step": 4771 + }, + { + "epoch": 0.42668097281831185, + "grad_norm": 3.3129327104075674, + "learning_rate": 0.00012822377363327713, + "loss": 2.1572, + "step": 4772 + }, + { + "epoch": 0.4267703862660944, + "grad_norm": 4.681731810677094, + "learning_rate": 0.00012819598972871443, + "loss": 2.2215, + "step": 4773 + }, + { + "epoch": 0.42685979971387694, + "grad_norm": 5.201042803840223, + "learning_rate": 0.0001281682034593881, + "loss": 2.3926, + "step": 4774 + }, + { + "epoch": 0.4269492131616595, + "grad_norm": 1.3381395607621396, + "learning_rate": 0.0001281404148276286, + "loss": 1.9382, + "step": 4775 + }, + { + "epoch": 0.4270386266094421, + "grad_norm": 0.8537457365611046, + "learning_rate": 0.00012811262383576646, + "loss": 1.9643, + "step": 4776 + }, + { + "epoch": 0.4271280400572246, + "grad_norm": 2.1342499899277696, + "learning_rate": 0.00012808483048613252, + "loss": 2.0108, + "step": 4777 + }, + { + "epoch": 0.42721745350500717, + "grad_norm": 2.9813344078958295, + "learning_rate": 0.00012805703478105778, + "loss": 2.7315, + "step": 4778 + }, + { + "epoch": 0.4273068669527897, + "grad_norm": 1.2045503160742308, + "learning_rate": 0.00012802923672287342, + "loss": 2.0076, + "step": 4779 + }, + { + "epoch": 0.42739628040057226, + "grad_norm": 1.639360673109062, + "learning_rate": 0.00012800143631391082, + "loss": 2.1041, + "step": 4780 + }, + { + "epoch": 0.4274856938483548, + "grad_norm": 1.6895738480136306, + "learning_rate": 0.00012797363355650154, + "loss": 1.8797, + "step": 4781 + }, + { + "epoch": 0.42757510729613735, + "grad_norm": 1.084893267220161, + "learning_rate": 0.00012794582845297744, + "loss": 2.1934, + "step": 4782 + }, + { + "epoch": 0.42766452074391986, + "grad_norm": 1.2546246248182986, + "learning_rate": 0.00012791802100567043, + "loss": 2.1743, + "step": 4783 + }, + { + "epoch": 0.42775393419170243, + "grad_norm": 0.7665613895352623, + "learning_rate": 0.00012789021121691274, + "loss": 2.0815, + "step": 4784 + }, + { + "epoch": 0.427843347639485, + "grad_norm": 0.8731128471738467, + "learning_rate": 0.0001278623990890367, + "loss": 2.046, + "step": 4785 + }, + { + "epoch": 0.4279327610872675, + "grad_norm": 1.4821059072339253, + "learning_rate": 0.0001278345846243749, + "loss": 1.9381, + "step": 4786 + }, + { + "epoch": 0.4280221745350501, + "grad_norm": 1.1336478155023875, + "learning_rate": 0.00012780676782526014, + "loss": 1.7638, + "step": 4787 + }, + { + "epoch": 0.4281115879828326, + "grad_norm": 0.9287757314031643, + "learning_rate": 0.0001277789486940253, + "loss": 2.0956, + "step": 4788 + }, + { + "epoch": 0.4282010014306152, + "grad_norm": 1.1533756255529022, + "learning_rate": 0.0001277511272330036, + "loss": 2.1378, + "step": 4789 + }, + { + "epoch": 0.4282904148783977, + "grad_norm": 9.958122837139136, + "learning_rate": 0.00012772330344452834, + "loss": 7.0709, + "step": 4790 + }, + { + "epoch": 0.42837982832618027, + "grad_norm": 1.2239241709772395, + "learning_rate": 0.00012769547733093312, + "loss": 1.9893, + "step": 4791 + }, + { + "epoch": 0.4284692417739628, + "grad_norm": 2.429961739882048, + "learning_rate": 0.0001276676488945517, + "loss": 1.9006, + "step": 4792 + }, + { + "epoch": 0.42855865522174535, + "grad_norm": 2.657122253125133, + "learning_rate": 0.00012763981813771795, + "loss": 1.7803, + "step": 4793 + }, + { + "epoch": 0.42864806866952787, + "grad_norm": 2.585298068856333, + "learning_rate": 0.00012761198506276603, + "loss": 1.836, + "step": 4794 + }, + { + "epoch": 0.42873748211731044, + "grad_norm": 1.0071376570598618, + "learning_rate": 0.00012758414967203028, + "loss": 2.2231, + "step": 4795 + }, + { + "epoch": 0.428826895565093, + "grad_norm": 0.9583650342908568, + "learning_rate": 0.00012755631196784522, + "loss": 1.9081, + "step": 4796 + }, + { + "epoch": 0.42891630901287553, + "grad_norm": 1.3351349749122678, + "learning_rate": 0.00012752847195254553, + "loss": 1.887, + "step": 4797 + }, + { + "epoch": 0.4290057224606581, + "grad_norm": 0.8283082532049664, + "learning_rate": 0.00012750062962846613, + "loss": 1.8204, + "step": 4798 + }, + { + "epoch": 0.4290951359084406, + "grad_norm": 1.31248949502499, + "learning_rate": 0.0001274727849979422, + "loss": 2.1034, + "step": 4799 + }, + { + "epoch": 0.4291845493562232, + "grad_norm": 1.6747270887405308, + "learning_rate": 0.0001274449380633089, + "loss": 1.7292, + "step": 4800 + }, + { + "epoch": 0.4292739628040057, + "grad_norm": 0.952881395035901, + "learning_rate": 0.0001274170888269018, + "loss": 1.9499, + "step": 4801 + }, + { + "epoch": 0.4293633762517883, + "grad_norm": 0.8062554490643253, + "learning_rate": 0.00012738923729105662, + "loss": 2.1676, + "step": 4802 + }, + { + "epoch": 0.4294527896995708, + "grad_norm": 2.230475999286249, + "learning_rate": 0.00012736138345810917, + "loss": 2.4008, + "step": 4803 + }, + { + "epoch": 0.42954220314735336, + "grad_norm": 1.171940623392539, + "learning_rate": 0.0001273335273303955, + "loss": 1.8865, + "step": 4804 + }, + { + "epoch": 0.42963161659513593, + "grad_norm": 0.8069499286285624, + "learning_rate": 0.00012730566891025195, + "loss": 1.8658, + "step": 4805 + }, + { + "epoch": 0.42972103004291845, + "grad_norm": 1.4025328297428141, + "learning_rate": 0.0001272778082000149, + "loss": 1.9784, + "step": 4806 + }, + { + "epoch": 0.429810443490701, + "grad_norm": 1.6465329980697847, + "learning_rate": 0.000127249945202021, + "loss": 1.8838, + "step": 4807 + }, + { + "epoch": 0.42989985693848354, + "grad_norm": 0.7934486360287419, + "learning_rate": 0.00012722207991860713, + "loss": 2.2896, + "step": 4808 + }, + { + "epoch": 0.4299892703862661, + "grad_norm": 1.0530336539505925, + "learning_rate": 0.0001271942123521103, + "loss": 1.994, + "step": 4809 + }, + { + "epoch": 0.4300786838340486, + "grad_norm": 1.1646407385448398, + "learning_rate": 0.0001271663425048677, + "loss": 2.1182, + "step": 4810 + }, + { + "epoch": 0.4301680972818312, + "grad_norm": 5.640899173212758, + "learning_rate": 0.00012713847037921678, + "loss": 7.0446, + "step": 4811 + }, + { + "epoch": 0.4302575107296137, + "grad_norm": 1.0586584046607823, + "learning_rate": 0.00012711059597749513, + "loss": 2.6108, + "step": 4812 + }, + { + "epoch": 0.4303469241773963, + "grad_norm": 1.0817092402294515, + "learning_rate": 0.00012708271930204052, + "loss": 2.1781, + "step": 4813 + }, + { + "epoch": 0.4304363376251788, + "grad_norm": 1.128263669415293, + "learning_rate": 0.00012705484035519096, + "loss": 2.0568, + "step": 4814 + }, + { + "epoch": 0.43052575107296137, + "grad_norm": 2.0127000532058674, + "learning_rate": 0.0001270269591392846, + "loss": 2.0287, + "step": 4815 + }, + { + "epoch": 0.43061516452074394, + "grad_norm": 1.0297040408902165, + "learning_rate": 0.00012699907565665982, + "loss": 1.7132, + "step": 4816 + }, + { + "epoch": 0.43070457796852646, + "grad_norm": 0.744476569948313, + "learning_rate": 0.0001269711899096552, + "loss": 1.3689, + "step": 4817 + }, + { + "epoch": 0.43079399141630903, + "grad_norm": 1.288127847464427, + "learning_rate": 0.0001269433019006094, + "loss": 1.902, + "step": 4818 + }, + { + "epoch": 0.43088340486409155, + "grad_norm": 1.6924113477206917, + "learning_rate": 0.00012691541163186148, + "loss": 1.4537, + "step": 4819 + }, + { + "epoch": 0.4309728183118741, + "grad_norm": 0.6574090593725923, + "learning_rate": 0.00012688751910575044, + "loss": 1.7666, + "step": 4820 + }, + { + "epoch": 0.43106223175965663, + "grad_norm": 0.5547073982373882, + "learning_rate": 0.00012685962432461563, + "loss": 1.8707, + "step": 4821 + }, + { + "epoch": 0.4311516452074392, + "grad_norm": 1.2713260934556947, + "learning_rate": 0.00012683172729079662, + "loss": 1.9418, + "step": 4822 + }, + { + "epoch": 0.4312410586552217, + "grad_norm": 1.6306246594397, + "learning_rate": 0.00012680382800663302, + "loss": 2.3456, + "step": 4823 + }, + { + "epoch": 0.4313304721030043, + "grad_norm": 1.1743142936264646, + "learning_rate": 0.00012677592647446472, + "loss": 2.1472, + "step": 4824 + }, + { + "epoch": 0.43141988555078686, + "grad_norm": 1.8872635916295988, + "learning_rate": 0.0001267480226966318, + "loss": 1.751, + "step": 4825 + }, + { + "epoch": 0.4315092989985694, + "grad_norm": 1.2454738881444654, + "learning_rate": 0.00012672011667547457, + "loss": 2.1174, + "step": 4826 + }, + { + "epoch": 0.43159871244635195, + "grad_norm": 1.3139596142492367, + "learning_rate": 0.0001266922084133334, + "loss": 2.2674, + "step": 4827 + }, + { + "epoch": 0.43168812589413447, + "grad_norm": 1.0190099299966349, + "learning_rate": 0.00012666429791254892, + "loss": 2.223, + "step": 4828 + }, + { + "epoch": 0.43177753934191704, + "grad_norm": 0.5410438279805829, + "learning_rate": 0.000126636385175462, + "loss": 1.5918, + "step": 4829 + }, + { + "epoch": 0.43186695278969955, + "grad_norm": 1.8396331013082585, + "learning_rate": 0.00012660847020441363, + "loss": 1.8403, + "step": 4830 + }, + { + "epoch": 0.4319563662374821, + "grad_norm": 1.5411032633968862, + "learning_rate": 0.00012658055300174498, + "loss": 2.0208, + "step": 4831 + }, + { + "epoch": 0.43204577968526464, + "grad_norm": 0.7428056297654608, + "learning_rate": 0.00012655263356979747, + "loss": 2.1787, + "step": 4832 + }, + { + "epoch": 0.4321351931330472, + "grad_norm": 0.5812889638776249, + "learning_rate": 0.00012652471191091266, + "loss": 2.3235, + "step": 4833 + }, + { + "epoch": 0.4322246065808298, + "grad_norm": 1.1015542919999752, + "learning_rate": 0.00012649678802743227, + "loss": 1.5498, + "step": 4834 + }, + { + "epoch": 0.4323140200286123, + "grad_norm": 1.7534172464647733, + "learning_rate": 0.00012646886192169826, + "loss": 1.8132, + "step": 4835 + }, + { + "epoch": 0.43240343347639487, + "grad_norm": 3.9764979361339727, + "learning_rate": 0.00012644093359605278, + "loss": 1.9756, + "step": 4836 + }, + { + "epoch": 0.4324928469241774, + "grad_norm": 1.3516440672046925, + "learning_rate": 0.00012641300305283814, + "loss": 1.6514, + "step": 4837 + }, + { + "epoch": 0.43258226037195996, + "grad_norm": 3.5374309176287886, + "learning_rate": 0.00012638507029439684, + "loss": 2.0595, + "step": 4838 + }, + { + "epoch": 0.4326716738197425, + "grad_norm": 1.9117415058761225, + "learning_rate": 0.00012635713532307152, + "loss": 1.7814, + "step": 4839 + }, + { + "epoch": 0.43276108726752505, + "grad_norm": 0.9277658876387895, + "learning_rate": 0.00012632919814120513, + "loss": 2.3244, + "step": 4840 + }, + { + "epoch": 0.43285050071530756, + "grad_norm": 1.674459453152977, + "learning_rate": 0.00012630125875114068, + "loss": 2.2144, + "step": 4841 + }, + { + "epoch": 0.43293991416309013, + "grad_norm": 1.6388102355888976, + "learning_rate": 0.00012627331715522143, + "loss": 1.9675, + "step": 4842 + }, + { + "epoch": 0.43302932761087265, + "grad_norm": 2.197266450642653, + "learning_rate": 0.0001262453733557908, + "loss": 1.677, + "step": 4843 + }, + { + "epoch": 0.4331187410586552, + "grad_norm": 1.1049764332688934, + "learning_rate": 0.00012621742735519239, + "loss": 2.0394, + "step": 4844 + }, + { + "epoch": 0.4332081545064378, + "grad_norm": 0.9423073166337989, + "learning_rate": 0.00012618947915577, + "loss": 2.3508, + "step": 4845 + }, + { + "epoch": 0.4332975679542203, + "grad_norm": 1.9705478492997885, + "learning_rate": 0.0001261615287598676, + "loss": 2.2634, + "step": 4846 + }, + { + "epoch": 0.4333869814020029, + "grad_norm": 1.0528139119424096, + "learning_rate": 0.00012613357616982946, + "loss": 2.2332, + "step": 4847 + }, + { + "epoch": 0.4334763948497854, + "grad_norm": 0.9172535592275372, + "learning_rate": 0.00012610562138799978, + "loss": 2.286, + "step": 4848 + }, + { + "epoch": 0.43356580829756797, + "grad_norm": 1.0637805022045874, + "learning_rate": 0.00012607766441672318, + "loss": 1.6309, + "step": 4849 + }, + { + "epoch": 0.4336552217453505, + "grad_norm": 1.7707331484606896, + "learning_rate": 0.00012604970525834436, + "loss": 2.3112, + "step": 4850 + }, + { + "epoch": 0.43374463519313305, + "grad_norm": 1.5156806926556088, + "learning_rate": 0.00012602174391520822, + "loss": 1.9647, + "step": 4851 + }, + { + "epoch": 0.43383404864091557, + "grad_norm": 2.310014260079505, + "learning_rate": 0.00012599378038965985, + "loss": 1.7504, + "step": 4852 + }, + { + "epoch": 0.43392346208869814, + "grad_norm": 1.344470732301466, + "learning_rate": 0.00012596581468404453, + "loss": 2.2382, + "step": 4853 + }, + { + "epoch": 0.4340128755364807, + "grad_norm": 5.56105041369347, + "learning_rate": 0.00012593784680070766, + "loss": 7.1391, + "step": 4854 + }, + { + "epoch": 0.43410228898426323, + "grad_norm": 2.4434625892298802, + "learning_rate": 0.0001259098767419949, + "loss": 2.2687, + "step": 4855 + }, + { + "epoch": 0.4341917024320458, + "grad_norm": 1.5884884250261413, + "learning_rate": 0.00012588190451025207, + "loss": 2.2435, + "step": 4856 + }, + { + "epoch": 0.4342811158798283, + "grad_norm": 1.0489257819072375, + "learning_rate": 0.0001258539301078252, + "loss": 1.9517, + "step": 4857 + }, + { + "epoch": 0.4343705293276109, + "grad_norm": 0.7157576221934521, + "learning_rate": 0.0001258259535370604, + "loss": 2.0681, + "step": 4858 + }, + { + "epoch": 0.4344599427753934, + "grad_norm": 0.5438431413944756, + "learning_rate": 0.00012579797480030406, + "loss": 1.652, + "step": 4859 + }, + { + "epoch": 0.434549356223176, + "grad_norm": 0.7691136146947207, + "learning_rate": 0.00012576999389990278, + "loss": 2.2894, + "step": 4860 + }, + { + "epoch": 0.4346387696709585, + "grad_norm": 1.483851332766039, + "learning_rate": 0.0001257420108382032, + "loss": 1.9435, + "step": 4861 + }, + { + "epoch": 0.43472818311874106, + "grad_norm": 2.9485525513452755, + "learning_rate": 0.00012571402561755227, + "loss": 1.7493, + "step": 4862 + }, + { + "epoch": 0.4348175965665236, + "grad_norm": 3.9090942499637316, + "learning_rate": 0.00012568603824029707, + "loss": 1.9643, + "step": 4863 + }, + { + "epoch": 0.43490701001430615, + "grad_norm": 0.5357450513144895, + "learning_rate": 0.00012565804870878484, + "loss": 1.8492, + "step": 4864 + }, + { + "epoch": 0.4349964234620887, + "grad_norm": 2.8426424146335427, + "learning_rate": 0.00012563005702536306, + "loss": 2.3232, + "step": 4865 + }, + { + "epoch": 0.43508583690987124, + "grad_norm": 1.4469934129863162, + "learning_rate": 0.00012560206319237936, + "loss": 2.0393, + "step": 4866 + }, + { + "epoch": 0.4351752503576538, + "grad_norm": 0.6767198482457067, + "learning_rate": 0.00012557406721218155, + "loss": 2.2096, + "step": 4867 + }, + { + "epoch": 0.4352646638054363, + "grad_norm": 0.8677735308138783, + "learning_rate": 0.00012554606908711757, + "loss": 2.2333, + "step": 4868 + }, + { + "epoch": 0.4353540772532189, + "grad_norm": 1.6007475524061139, + "learning_rate": 0.00012551806881953566, + "loss": 2.0291, + "step": 4869 + }, + { + "epoch": 0.4354434907010014, + "grad_norm": 2.4820199224054926, + "learning_rate": 0.0001254900664117841, + "loss": 1.8544, + "step": 4870 + }, + { + "epoch": 0.435532904148784, + "grad_norm": 1.4126876403602204, + "learning_rate": 0.0001254620618662115, + "loss": 2.4185, + "step": 4871 + }, + { + "epoch": 0.4356223175965665, + "grad_norm": 2.6182209256563254, + "learning_rate": 0.0001254340551851665, + "loss": 1.9303, + "step": 4872 + }, + { + "epoch": 0.43571173104434907, + "grad_norm": 4.767459036501769, + "learning_rate": 0.000125406046370998, + "loss": 1.9546, + "step": 4873 + }, + { + "epoch": 0.43580114449213164, + "grad_norm": 2.8524344292913297, + "learning_rate": 0.00012537803542605512, + "loss": 2.4803, + "step": 4874 + }, + { + "epoch": 0.43589055793991416, + "grad_norm": 3.140863565376866, + "learning_rate": 0.00012535002235268701, + "loss": 2.1764, + "step": 4875 + }, + { + "epoch": 0.43597997138769673, + "grad_norm": 1.322575944237815, + "learning_rate": 0.00012532200715324317, + "loss": 2.0459, + "step": 4876 + }, + { + "epoch": 0.43606938483547925, + "grad_norm": 1.5915675393050916, + "learning_rate": 0.0001252939898300731, + "loss": 1.7056, + "step": 4877 + }, + { + "epoch": 0.4361587982832618, + "grad_norm": 2.518828744526082, + "learning_rate": 0.0001252659703855267, + "loss": 2.1226, + "step": 4878 + }, + { + "epoch": 0.43624821173104433, + "grad_norm": 4.182255682965619, + "learning_rate": 0.00012523794882195391, + "loss": 2.1947, + "step": 4879 + }, + { + "epoch": 0.4363376251788269, + "grad_norm": 2.8771980036649376, + "learning_rate": 0.0001252099251417048, + "loss": 2.2892, + "step": 4880 + }, + { + "epoch": 0.4364270386266094, + "grad_norm": 2.68033527991553, + "learning_rate": 0.00012518189934712973, + "loss": 1.8305, + "step": 4881 + }, + { + "epoch": 0.436516452074392, + "grad_norm": 4.683072429525462, + "learning_rate": 0.00012515387144057915, + "loss": 1.8063, + "step": 4882 + }, + { + "epoch": 0.4366058655221745, + "grad_norm": 1.85831970930091, + "learning_rate": 0.00012512584142440378, + "loss": 2.0864, + "step": 4883 + }, + { + "epoch": 0.4366952789699571, + "grad_norm": 4.853059452317493, + "learning_rate": 0.00012509780930095442, + "loss": 1.9999, + "step": 4884 + }, + { + "epoch": 0.43678469241773965, + "grad_norm": 1.0257278700388175, + "learning_rate": 0.00012506977507258208, + "loss": 1.9479, + "step": 4885 + }, + { + "epoch": 0.43687410586552217, + "grad_norm": 1.6636982597383523, + "learning_rate": 0.000125041738741638, + "loss": 1.9825, + "step": 4886 + }, + { + "epoch": 0.43696351931330474, + "grad_norm": 2.8018372389286985, + "learning_rate": 0.00012501370031047356, + "loss": 2.1819, + "step": 4887 + }, + { + "epoch": 0.43705293276108725, + "grad_norm": 1.5130352717050533, + "learning_rate": 0.00012498565978144027, + "loss": 2.0273, + "step": 4888 + }, + { + "epoch": 0.4371423462088698, + "grad_norm": 0.9842068176219134, + "learning_rate": 0.00012495761715688983, + "loss": 1.8225, + "step": 4889 + }, + { + "epoch": 0.43723175965665234, + "grad_norm": 1.773202665793875, + "learning_rate": 0.00012492957243917424, + "loss": 2.0819, + "step": 4890 + }, + { + "epoch": 0.4373211731044349, + "grad_norm": 1.6211941811693247, + "learning_rate": 0.00012490152563064544, + "loss": 1.98, + "step": 4891 + }, + { + "epoch": 0.43741058655221743, + "grad_norm": 0.8273216984426565, + "learning_rate": 0.00012487347673365582, + "loss": 2.0298, + "step": 4892 + }, + { + "epoch": 0.4375, + "grad_norm": 1.2096756229373296, + "learning_rate": 0.00012484542575055775, + "loss": 1.9956, + "step": 4893 + }, + { + "epoch": 0.43758941344778257, + "grad_norm": 1.1109707282392138, + "learning_rate": 0.0001248173726837038, + "loss": 1.7764, + "step": 4894 + }, + { + "epoch": 0.4376788268955651, + "grad_norm": 1.5440213126109315, + "learning_rate": 0.00012478931753544676, + "loss": 1.9361, + "step": 4895 + }, + { + "epoch": 0.43776824034334766, + "grad_norm": 1.1568692289909641, + "learning_rate": 0.00012476126030813963, + "loss": 2.1411, + "step": 4896 + }, + { + "epoch": 0.4378576537911302, + "grad_norm": 1.659615304500026, + "learning_rate": 0.0001247332010041355, + "loss": 2.1599, + "step": 4897 + }, + { + "epoch": 0.43794706723891275, + "grad_norm": 1.799177422522495, + "learning_rate": 0.00012470513962578768, + "loss": 2.46, + "step": 4898 + }, + { + "epoch": 0.43803648068669526, + "grad_norm": 1.0779175512579942, + "learning_rate": 0.0001246770761754496, + "loss": 1.9685, + "step": 4899 + }, + { + "epoch": 0.43812589413447783, + "grad_norm": 2.1948784097133625, + "learning_rate": 0.000124649010655475, + "loss": 2.0431, + "step": 4900 + }, + { + "epoch": 0.43821530758226035, + "grad_norm": 1.6427358203298656, + "learning_rate": 0.00012462094306821758, + "loss": 1.7685, + "step": 4901 + }, + { + "epoch": 0.4383047210300429, + "grad_norm": 0.9167046117804443, + "learning_rate": 0.0001245928734160314, + "loss": 2.6297, + "step": 4902 + }, + { + "epoch": 0.4383941344778255, + "grad_norm": 0.9486255085262724, + "learning_rate": 0.00012456480170127069, + "loss": 1.6485, + "step": 4903 + }, + { + "epoch": 0.438483547925608, + "grad_norm": 0.7708879390456894, + "learning_rate": 0.00012453672792628968, + "loss": 2.1157, + "step": 4904 + }, + { + "epoch": 0.4385729613733906, + "grad_norm": 0.9523898925002873, + "learning_rate": 0.00012450865209344294, + "loss": 1.8753, + "step": 4905 + }, + { + "epoch": 0.4386623748211731, + "grad_norm": 0.8029906353564543, + "learning_rate": 0.00012448057420508517, + "loss": 2.2935, + "step": 4906 + }, + { + "epoch": 0.43875178826895567, + "grad_norm": 0.4173261600805615, + "learning_rate": 0.0001244524942635712, + "loss": 1.9303, + "step": 4907 + }, + { + "epoch": 0.4388412017167382, + "grad_norm": 1.0348248922295955, + "learning_rate": 0.00012442441227125602, + "loss": 1.9901, + "step": 4908 + }, + { + "epoch": 0.43893061516452075, + "grad_norm": 1.4188843451617281, + "learning_rate": 0.00012439632823049493, + "loss": 2.357, + "step": 4909 + }, + { + "epoch": 0.43902002861230327, + "grad_norm": 0.5939377042539815, + "learning_rate": 0.00012436824214364324, + "loss": 1.8124, + "step": 4910 + }, + { + "epoch": 0.43910944206008584, + "grad_norm": 0.5900811813901293, + "learning_rate": 0.00012434015401305653, + "loss": 2.4591, + "step": 4911 + }, + { + "epoch": 0.43919885550786836, + "grad_norm": 2.9806548895949168, + "learning_rate": 0.00012431206384109044, + "loss": 1.927, + "step": 4912 + }, + { + "epoch": 0.43928826895565093, + "grad_norm": 1.2262019201159515, + "learning_rate": 0.00012428397163010096, + "loss": 2.5327, + "step": 4913 + }, + { + "epoch": 0.4393776824034335, + "grad_norm": 1.1064141245046384, + "learning_rate": 0.00012425587738244413, + "loss": 2.2555, + "step": 4914 + }, + { + "epoch": 0.439467095851216, + "grad_norm": 0.9842269428016573, + "learning_rate": 0.00012422778110047613, + "loss": 1.9241, + "step": 4915 + }, + { + "epoch": 0.4395565092989986, + "grad_norm": 1.906585088795211, + "learning_rate": 0.0001241996827865534, + "loss": 2.3014, + "step": 4916 + }, + { + "epoch": 0.4396459227467811, + "grad_norm": 1.142578246686776, + "learning_rate": 0.00012417158244303249, + "loss": 1.8648, + "step": 4917 + }, + { + "epoch": 0.4397353361945637, + "grad_norm": 0.758201645473539, + "learning_rate": 0.00012414348007227014, + "loss": 1.9198, + "step": 4918 + }, + { + "epoch": 0.4398247496423462, + "grad_norm": 2.4890969585992666, + "learning_rate": 0.00012411537567662327, + "loss": 2.4102, + "step": 4919 + }, + { + "epoch": 0.43991416309012876, + "grad_norm": 1.8111057543526057, + "learning_rate": 0.000124087269258449, + "loss": 1.9461, + "step": 4920 + }, + { + "epoch": 0.4400035765379113, + "grad_norm": 1.8130717893779058, + "learning_rate": 0.00012405916082010456, + "loss": 1.6827, + "step": 4921 + }, + { + "epoch": 0.44009298998569385, + "grad_norm": 2.178815787244888, + "learning_rate": 0.00012403105036394728, + "loss": 2.0723, + "step": 4922 + }, + { + "epoch": 0.4401824034334764, + "grad_norm": 1.5934710474116431, + "learning_rate": 0.0001240029378923349, + "loss": 6.7394, + "step": 4923 + }, + { + "epoch": 0.44027181688125894, + "grad_norm": 2.1340172270026905, + "learning_rate": 0.0001239748234076251, + "loss": 1.8849, + "step": 4924 + }, + { + "epoch": 0.4403612303290415, + "grad_norm": 3.1263859055874588, + "learning_rate": 0.0001239467069121758, + "loss": 2.5504, + "step": 4925 + }, + { + "epoch": 0.440450643776824, + "grad_norm": 1.1895756072965185, + "learning_rate": 0.0001239185884083451, + "loss": 1.9735, + "step": 4926 + }, + { + "epoch": 0.4405400572246066, + "grad_norm": 1.9624040719196767, + "learning_rate": 0.00012389046789849128, + "loss": 1.9852, + "step": 4927 + }, + { + "epoch": 0.4406294706723891, + "grad_norm": 2.5381383114437566, + "learning_rate": 0.00012386234538497282, + "loss": 1.8037, + "step": 4928 + }, + { + "epoch": 0.4407188841201717, + "grad_norm": 0.791720222337283, + "learning_rate": 0.00012383422087014817, + "loss": 2.0506, + "step": 4929 + }, + { + "epoch": 0.4408082975679542, + "grad_norm": 0.7169578009852894, + "learning_rate": 0.00012380609435637627, + "loss": 1.917, + "step": 4930 + }, + { + "epoch": 0.44089771101573677, + "grad_norm": 1.257416253211743, + "learning_rate": 0.000123777965846016, + "loss": 1.8021, + "step": 4931 + }, + { + "epoch": 0.4409871244635193, + "grad_norm": 1.0451633910787868, + "learning_rate": 0.0001237498353414264, + "loss": 2.1624, + "step": 4932 + }, + { + "epoch": 0.44107653791130186, + "grad_norm": 0.6579354746649044, + "learning_rate": 0.00012372170284496683, + "loss": 2.4785, + "step": 4933 + }, + { + "epoch": 0.44116595135908443, + "grad_norm": 0.96695304785, + "learning_rate": 0.0001236935683589967, + "loss": 1.8858, + "step": 4934 + }, + { + "epoch": 0.44125536480686695, + "grad_norm": 0.7880624673574234, + "learning_rate": 0.00012366543188587555, + "loss": 1.9166, + "step": 4935 + }, + { + "epoch": 0.4413447782546495, + "grad_norm": 0.8125503951746849, + "learning_rate": 0.00012363729342796325, + "loss": 2.3474, + "step": 4936 + }, + { + "epoch": 0.44143419170243203, + "grad_norm": 0.6904612000341105, + "learning_rate": 0.0001236091529876197, + "loss": 1.978, + "step": 4937 + }, + { + "epoch": 0.4415236051502146, + "grad_norm": 0.6767978660349242, + "learning_rate": 0.000123581010567205, + "loss": 2.1546, + "step": 4938 + }, + { + "epoch": 0.4416130185979971, + "grad_norm": 0.627231118342395, + "learning_rate": 0.00012355286616907939, + "loss": 1.8651, + "step": 4939 + }, + { + "epoch": 0.4417024320457797, + "grad_norm": 0.9500916246665315, + "learning_rate": 0.00012352471979560338, + "loss": 1.8552, + "step": 4940 + }, + { + "epoch": 0.4417918454935622, + "grad_norm": 1.8705548076202465, + "learning_rate": 0.00012349657144913753, + "loss": 2.0327, + "step": 4941 + }, + { + "epoch": 0.4418812589413448, + "grad_norm": 1.9244942590062655, + "learning_rate": 0.00012346842113204257, + "loss": 2.2145, + "step": 4942 + }, + { + "epoch": 0.44197067238912735, + "grad_norm": 0.5263368105738983, + "learning_rate": 0.0001234402688466795, + "loss": 2.0153, + "step": 4943 + }, + { + "epoch": 0.44206008583690987, + "grad_norm": 0.7087480434684652, + "learning_rate": 0.0001234121145954094, + "loss": 2.0401, + "step": 4944 + }, + { + "epoch": 0.44214949928469244, + "grad_norm": 0.8354247426601242, + "learning_rate": 0.00012338395838059352, + "loss": 1.9049, + "step": 4945 + }, + { + "epoch": 0.44223891273247495, + "grad_norm": 1.1369018575688399, + "learning_rate": 0.00012335580020459325, + "loss": 1.6489, + "step": 4946 + }, + { + "epoch": 0.4423283261802575, + "grad_norm": 0.710069833359457, + "learning_rate": 0.00012332764006977028, + "loss": 2.1688, + "step": 4947 + }, + { + "epoch": 0.44241773962804004, + "grad_norm": 0.5331427282777466, + "learning_rate": 0.0001232994779784863, + "loss": 2.2275, + "step": 4948 + }, + { + "epoch": 0.4425071530758226, + "grad_norm": 0.4673353284525824, + "learning_rate": 0.0001232713139331032, + "loss": 1.9601, + "step": 4949 + }, + { + "epoch": 0.44259656652360513, + "grad_norm": 0.9971921384069002, + "learning_rate": 0.00012324314793598314, + "loss": 1.7637, + "step": 4950 + }, + { + "epoch": 0.4426859799713877, + "grad_norm": 0.4409661982052047, + "learning_rate": 0.00012321497998948834, + "loss": 1.9885, + "step": 4951 + }, + { + "epoch": 0.4427753934191702, + "grad_norm": 0.8182551309910349, + "learning_rate": 0.00012318681009598116, + "loss": 1.9464, + "step": 4952 + }, + { + "epoch": 0.4428648068669528, + "grad_norm": 1.31492730660061, + "learning_rate": 0.00012315863825782425, + "loss": 1.3927, + "step": 4953 + }, + { + "epoch": 0.44295422031473536, + "grad_norm": 1.3151279512966771, + "learning_rate": 0.00012313046447738035, + "loss": 1.8856, + "step": 4954 + }, + { + "epoch": 0.4430436337625179, + "grad_norm": 1.0966052317978674, + "learning_rate": 0.0001231022887570123, + "loss": 1.9317, + "step": 4955 + }, + { + "epoch": 0.44313304721030045, + "grad_norm": 1.2779243152551694, + "learning_rate": 0.00012307411109908315, + "loss": 2.0273, + "step": 4956 + }, + { + "epoch": 0.44322246065808296, + "grad_norm": 1.1856914319636962, + "learning_rate": 0.00012304593150595623, + "loss": 1.9323, + "step": 4957 + }, + { + "epoch": 0.44331187410586553, + "grad_norm": 0.9351578991061923, + "learning_rate": 0.00012301774997999483, + "loss": 1.926, + "step": 4958 + }, + { + "epoch": 0.44340128755364805, + "grad_norm": 1.919630908735787, + "learning_rate": 0.00012298956652356257, + "loss": 2.2001, + "step": 4959 + }, + { + "epoch": 0.4434907010014306, + "grad_norm": 0.9431210371277843, + "learning_rate": 0.00012296138113902308, + "loss": 2.282, + "step": 4960 + }, + { + "epoch": 0.44358011444921314, + "grad_norm": 0.9783136180407972, + "learning_rate": 0.00012293319382874037, + "loss": 2.0644, + "step": 4961 + }, + { + "epoch": 0.4436695278969957, + "grad_norm": 0.6134595707626666, + "learning_rate": 0.0001229050045950783, + "loss": 2.2096, + "step": 4962 + }, + { + "epoch": 0.4437589413447783, + "grad_norm": 2.575235327270852, + "learning_rate": 0.00012287681344040117, + "loss": 2.138, + "step": 4963 + }, + { + "epoch": 0.4438483547925608, + "grad_norm": 1.6107571709902146, + "learning_rate": 0.00012284862036707339, + "loss": 1.8936, + "step": 4964 + }, + { + "epoch": 0.44393776824034337, + "grad_norm": 1.43096626908315, + "learning_rate": 0.00012282042537745938, + "loss": 1.823, + "step": 4965 + }, + { + "epoch": 0.4440271816881259, + "grad_norm": 0.7444622349430405, + "learning_rate": 0.00012279222847392385, + "loss": 2.2387, + "step": 4966 + }, + { + "epoch": 0.44411659513590845, + "grad_norm": 0.7882772295108519, + "learning_rate": 0.0001227640296588316, + "loss": 2.0782, + "step": 4967 + }, + { + "epoch": 0.44420600858369097, + "grad_norm": 2.077423507311666, + "learning_rate": 0.00012273582893454775, + "loss": 2.1345, + "step": 4968 + }, + { + "epoch": 0.44429542203147354, + "grad_norm": 1.0235365665225566, + "learning_rate": 0.00012270762630343734, + "loss": 2.1993, + "step": 4969 + }, + { + "epoch": 0.44438483547925606, + "grad_norm": 2.227854994447671, + "learning_rate": 0.00012267942176786575, + "loss": 2.6677, + "step": 4970 + }, + { + "epoch": 0.44447424892703863, + "grad_norm": 0.8613677799353636, + "learning_rate": 0.00012265121533019843, + "loss": 1.662, + "step": 4971 + }, + { + "epoch": 0.4445636623748212, + "grad_norm": 2.6165710491881815, + "learning_rate": 0.00012262300699280104, + "loss": 1.6059, + "step": 4972 + }, + { + "epoch": 0.4446530758226037, + "grad_norm": 3.285663445799079, + "learning_rate": 0.00012259479675803939, + "loss": 2.2135, + "step": 4973 + }, + { + "epoch": 0.4447424892703863, + "grad_norm": 1.0941433856320137, + "learning_rate": 0.00012256658462827941, + "loss": 2.1169, + "step": 4974 + }, + { + "epoch": 0.4448319027181688, + "grad_norm": 1.5558552620517612, + "learning_rate": 0.00012253837060588723, + "loss": 2.094, + "step": 4975 + }, + { + "epoch": 0.4449213161659514, + "grad_norm": 0.7620618179905317, + "learning_rate": 0.00012251015469322916, + "loss": 2.0375, + "step": 4976 + }, + { + "epoch": 0.4450107296137339, + "grad_norm": 1.0949069513972256, + "learning_rate": 0.00012248193689267157, + "loss": 2.0075, + "step": 4977 + }, + { + "epoch": 0.44510014306151646, + "grad_norm": 0.5135148606426675, + "learning_rate": 0.00012245371720658109, + "loss": 2.2086, + "step": 4978 + }, + { + "epoch": 0.445189556509299, + "grad_norm": 1.3209447790631912, + "learning_rate": 0.00012242549563732443, + "loss": 1.7451, + "step": 4979 + }, + { + "epoch": 0.44527896995708155, + "grad_norm": 1.616546124376746, + "learning_rate": 0.00012239727218726857, + "loss": 7.1043, + "step": 4980 + }, + { + "epoch": 0.44536838340486407, + "grad_norm": 0.7955195709856671, + "learning_rate": 0.00012236904685878055, + "loss": 2.1616, + "step": 4981 + }, + { + "epoch": 0.44545779685264664, + "grad_norm": 1.0623581222317586, + "learning_rate": 0.0001223408196542276, + "loss": 1.9333, + "step": 4982 + }, + { + "epoch": 0.4455472103004292, + "grad_norm": 1.719530622802301, + "learning_rate": 0.00012231259057597703, + "loss": 2.145, + "step": 4983 + }, + { + "epoch": 0.4456366237482117, + "grad_norm": 1.2055932419498059, + "learning_rate": 0.00012228435962639646, + "loss": 2.0763, + "step": 4984 + }, + { + "epoch": 0.4457260371959943, + "grad_norm": 1.668006463074235, + "learning_rate": 0.00012225612680785358, + "loss": 2.1328, + "step": 4985 + }, + { + "epoch": 0.4458154506437768, + "grad_norm": 1.5201919589002142, + "learning_rate": 0.0001222278921227162, + "loss": 1.7576, + "step": 4986 + }, + { + "epoch": 0.4459048640915594, + "grad_norm": 0.6478254434560275, + "learning_rate": 0.00012219965557335236, + "loss": 1.9319, + "step": 4987 + }, + { + "epoch": 0.4459942775393419, + "grad_norm": 1.3041433163314622, + "learning_rate": 0.00012217141716213022, + "loss": 2.0535, + "step": 4988 + }, + { + "epoch": 0.44608369098712447, + "grad_norm": 3.960055512225314, + "learning_rate": 0.0001221431768914181, + "loss": 6.9383, + "step": 4989 + }, + { + "epoch": 0.446173104434907, + "grad_norm": 3.3074076896266673, + "learning_rate": 0.00012211493476358448, + "loss": 2.0617, + "step": 4990 + }, + { + "epoch": 0.44626251788268956, + "grad_norm": 3.6381806581497367, + "learning_rate": 0.00012208669078099798, + "loss": 2.4294, + "step": 4991 + }, + { + "epoch": 0.44635193133047213, + "grad_norm": 1.811469181880436, + "learning_rate": 0.0001220584449460274, + "loss": 1.6869, + "step": 4992 + }, + { + "epoch": 0.44644134477825465, + "grad_norm": 2.7903563693983764, + "learning_rate": 0.00012203019726104168, + "loss": 2.4285, + "step": 4993 + }, + { + "epoch": 0.4465307582260372, + "grad_norm": 2.459324733626582, + "learning_rate": 0.00012200194772840991, + "loss": 1.6675, + "step": 4994 + }, + { + "epoch": 0.44662017167381973, + "grad_norm": 1.2362077311505002, + "learning_rate": 0.0001219736963505014, + "loss": 2.435, + "step": 4995 + }, + { + "epoch": 0.4467095851216023, + "grad_norm": 1.7171293039965954, + "learning_rate": 0.00012194544312968548, + "loss": 1.9565, + "step": 4996 + }, + { + "epoch": 0.4467989985693848, + "grad_norm": 2.0641267193287125, + "learning_rate": 0.00012191718806833177, + "loss": 1.9544, + "step": 4997 + }, + { + "epoch": 0.4468884120171674, + "grad_norm": 1.5485957543723086, + "learning_rate": 0.00012188893116880993, + "loss": 1.9468, + "step": 4998 + }, + { + "epoch": 0.4469778254649499, + "grad_norm": 1.9730195037637361, + "learning_rate": 0.00012186067243348991, + "loss": 2.2058, + "step": 4999 + }, + { + "epoch": 0.4470672389127325, + "grad_norm": 0.8632646693491478, + "learning_rate": 0.00012183241186474166, + "loss": 2.0255, + "step": 5000 + }, + { + "epoch": 0.447156652360515, + "grad_norm": 1.648923313190767, + "learning_rate": 0.00012180414946493538, + "loss": 2.4449, + "step": 5001 + }, + { + "epoch": 0.44724606580829757, + "grad_norm": 0.8500571113707015, + "learning_rate": 0.00012177588523644146, + "loss": 1.6339, + "step": 5002 + }, + { + "epoch": 0.44733547925608014, + "grad_norm": 2.5522958396820927, + "learning_rate": 0.00012174761918163034, + "loss": 1.6359, + "step": 5003 + }, + { + "epoch": 0.44742489270386265, + "grad_norm": 1.3328486722347728, + "learning_rate": 0.00012171935130287265, + "loss": 1.8037, + "step": 5004 + }, + { + "epoch": 0.4475143061516452, + "grad_norm": 2.5691955705131, + "learning_rate": 0.00012169108160253919, + "loss": 2.0098, + "step": 5005 + }, + { + "epoch": 0.44760371959942774, + "grad_norm": 1.0138312954327033, + "learning_rate": 0.00012166281008300094, + "loss": 2.1802, + "step": 5006 + }, + { + "epoch": 0.4476931330472103, + "grad_norm": 3.147232506090543, + "learning_rate": 0.00012163453674662892, + "loss": 1.9341, + "step": 5007 + }, + { + "epoch": 0.44778254649499283, + "grad_norm": 3.534725610290357, + "learning_rate": 0.00012160626159579447, + "loss": 1.8588, + "step": 5008 + }, + { + "epoch": 0.4478719599427754, + "grad_norm": 2.261757719113133, + "learning_rate": 0.00012157798463286894, + "loss": 1.9702, + "step": 5009 + }, + { + "epoch": 0.4479613733905579, + "grad_norm": 1.9777149108971952, + "learning_rate": 0.00012154970586022389, + "loss": 2.0538, + "step": 5010 + }, + { + "epoch": 0.4480507868383405, + "grad_norm": 2.9559185983788248, + "learning_rate": 0.00012152142528023107, + "loss": 1.788, + "step": 5011 + }, + { + "epoch": 0.44814020028612306, + "grad_norm": 0.8809460428196623, + "learning_rate": 0.00012149314289526228, + "loss": 2.0202, + "step": 5012 + }, + { + "epoch": 0.4482296137339056, + "grad_norm": 0.7203544677465135, + "learning_rate": 0.00012146485870768954, + "loss": 1.6177, + "step": 5013 + }, + { + "epoch": 0.44831902718168815, + "grad_norm": 1.5143290826185984, + "learning_rate": 0.00012143657271988505, + "loss": 2.1698, + "step": 5014 + }, + { + "epoch": 0.44840844062947066, + "grad_norm": 0.8963865103666719, + "learning_rate": 0.00012140828493422107, + "loss": 2.0852, + "step": 5015 + }, + { + "epoch": 0.44849785407725323, + "grad_norm": 1.2196987161150636, + "learning_rate": 0.0001213799953530701, + "loss": 1.9084, + "step": 5016 + }, + { + "epoch": 0.44858726752503575, + "grad_norm": 0.9735860359430524, + "learning_rate": 0.00012135170397880473, + "loss": 2.1988, + "step": 5017 + }, + { + "epoch": 0.4486766809728183, + "grad_norm": 4.68822287678228, + "learning_rate": 0.00012132341081379776, + "loss": 6.7962, + "step": 5018 + }, + { + "epoch": 0.44876609442060084, + "grad_norm": 4.7719405768741145, + "learning_rate": 0.00012129511586042206, + "loss": 2.1874, + "step": 5019 + }, + { + "epoch": 0.4488555078683834, + "grad_norm": 9.592607400278098, + "learning_rate": 0.00012126681912105069, + "loss": 6.9644, + "step": 5020 + }, + { + "epoch": 0.448944921316166, + "grad_norm": 7.455641630216957, + "learning_rate": 0.00012123852059805691, + "loss": 2.2911, + "step": 5021 + }, + { + "epoch": 0.4490343347639485, + "grad_norm": 4.882262704156782, + "learning_rate": 0.00012121022029381406, + "loss": 1.7443, + "step": 5022 + }, + { + "epoch": 0.44912374821173107, + "grad_norm": 3.450838414570058, + "learning_rate": 0.00012118191821069565, + "loss": 1.8133, + "step": 5023 + }, + { + "epoch": 0.4492131616595136, + "grad_norm": 5.225516479596623, + "learning_rate": 0.00012115361435107531, + "loss": 1.7034, + "step": 5024 + }, + { + "epoch": 0.44930257510729615, + "grad_norm": 0.9062961327302584, + "learning_rate": 0.0001211253087173269, + "loss": 2.1475, + "step": 5025 + }, + { + "epoch": 0.44939198855507867, + "grad_norm": 5.0101822341198625, + "learning_rate": 0.00012109700131182437, + "loss": 2.0718, + "step": 5026 + }, + { + "epoch": 0.44948140200286124, + "grad_norm": 1.7782827531776335, + "learning_rate": 0.00012106869213694181, + "loss": 2.0375, + "step": 5027 + }, + { + "epoch": 0.44957081545064376, + "grad_norm": 1.24611883783407, + "learning_rate": 0.0001210403811950535, + "loss": 1.5191, + "step": 5028 + }, + { + "epoch": 0.44966022889842633, + "grad_norm": 3.8360754468875298, + "learning_rate": 0.00012101206848853381, + "loss": 2.11, + "step": 5029 + }, + { + "epoch": 0.44974964234620884, + "grad_norm": 0.7630082357629481, + "learning_rate": 0.00012098375401975731, + "loss": 2.0419, + "step": 5030 + }, + { + "epoch": 0.4498390557939914, + "grad_norm": 1.5815110901480705, + "learning_rate": 0.00012095543779109873, + "loss": 1.9728, + "step": 5031 + }, + { + "epoch": 0.449928469241774, + "grad_norm": 2.232989782487896, + "learning_rate": 0.00012092711980493285, + "loss": 1.6187, + "step": 5032 + }, + { + "epoch": 0.4500178826895565, + "grad_norm": 1.4547330915745995, + "learning_rate": 0.00012089880006363475, + "loss": 1.8438, + "step": 5033 + }, + { + "epoch": 0.4501072961373391, + "grad_norm": 2.953290786454451, + "learning_rate": 0.00012087047856957949, + "loss": 2.2311, + "step": 5034 + }, + { + "epoch": 0.4501967095851216, + "grad_norm": 2.1037001875995593, + "learning_rate": 0.00012084215532514243, + "loss": 2.458, + "step": 5035 + }, + { + "epoch": 0.45028612303290416, + "grad_norm": 1.5130674421204993, + "learning_rate": 0.00012081383033269896, + "loss": 2.3324, + "step": 5036 + }, + { + "epoch": 0.4503755364806867, + "grad_norm": 2.0193658161491177, + "learning_rate": 0.00012078550359462464, + "loss": 1.8441, + "step": 5037 + }, + { + "epoch": 0.45046494992846925, + "grad_norm": 2.3397615127842615, + "learning_rate": 0.00012075717511329529, + "loss": 1.828, + "step": 5038 + }, + { + "epoch": 0.45055436337625177, + "grad_norm": 1.2645980138952568, + "learning_rate": 0.00012072884489108669, + "loss": 2.0142, + "step": 5039 + }, + { + "epoch": 0.45064377682403434, + "grad_norm": 1.3938562754330341, + "learning_rate": 0.00012070051293037492, + "loss": 1.7167, + "step": 5040 + }, + { + "epoch": 0.4507331902718169, + "grad_norm": 2.9239947641211463, + "learning_rate": 0.00012067217923353615, + "loss": 2.3672, + "step": 5041 + }, + { + "epoch": 0.4508226037195994, + "grad_norm": 0.9191995111825173, + "learning_rate": 0.00012064384380294667, + "loss": 1.9784, + "step": 5042 + }, + { + "epoch": 0.450912017167382, + "grad_norm": 1.7137420642471204, + "learning_rate": 0.00012061550664098293, + "loss": 1.8081, + "step": 5043 + }, + { + "epoch": 0.4510014306151645, + "grad_norm": 1.8316644469629328, + "learning_rate": 0.00012058716775002152, + "loss": 1.7437, + "step": 5044 + }, + { + "epoch": 0.4510908440629471, + "grad_norm": 0.8737680023159871, + "learning_rate": 0.00012055882713243926, + "loss": 1.8061, + "step": 5045 + }, + { + "epoch": 0.4511802575107296, + "grad_norm": 3.3942088284596905, + "learning_rate": 0.00012053048479061298, + "loss": 1.5091, + "step": 5046 + }, + { + "epoch": 0.45126967095851217, + "grad_norm": 1.4122641244088283, + "learning_rate": 0.00012050214072691969, + "loss": 1.886, + "step": 5047 + }, + { + "epoch": 0.4513590844062947, + "grad_norm": 3.7543306453824816, + "learning_rate": 0.0001204737949437367, + "loss": 2.1042, + "step": 5048 + }, + { + "epoch": 0.45144849785407726, + "grad_norm": 2.451420938079215, + "learning_rate": 0.00012044544744344122, + "loss": 1.5584, + "step": 5049 + }, + { + "epoch": 0.4515379113018598, + "grad_norm": 3.826482323864907, + "learning_rate": 0.00012041709822841074, + "loss": 1.7465, + "step": 5050 + }, + { + "epoch": 0.45162732474964234, + "grad_norm": 3.3369593842262777, + "learning_rate": 0.00012038874730102288, + "loss": 1.6926, + "step": 5051 + }, + { + "epoch": 0.4517167381974249, + "grad_norm": 1.7490178422043083, + "learning_rate": 0.00012036039466365543, + "loss": 2.1086, + "step": 5052 + }, + { + "epoch": 0.45180615164520743, + "grad_norm": 1.8826946005626215, + "learning_rate": 0.00012033204031868626, + "loss": 2.0682, + "step": 5053 + }, + { + "epoch": 0.45189556509299, + "grad_norm": 2.646902512278915, + "learning_rate": 0.0001203036842684934, + "loss": 2.0765, + "step": 5054 + }, + { + "epoch": 0.4519849785407725, + "grad_norm": 1.579411520192089, + "learning_rate": 0.00012027532651545512, + "loss": 1.6057, + "step": 5055 + }, + { + "epoch": 0.4520743919885551, + "grad_norm": 3.933324911360376, + "learning_rate": 0.00012024696706194967, + "loss": 1.998, + "step": 5056 + }, + { + "epoch": 0.4521638054363376, + "grad_norm": 1.1802093303603964, + "learning_rate": 0.00012021860591035549, + "loss": 1.9176, + "step": 5057 + }, + { + "epoch": 0.4522532188841202, + "grad_norm": 3.0218661346181475, + "learning_rate": 0.00012019024306305131, + "loss": 2.0843, + "step": 5058 + }, + { + "epoch": 0.4523426323319027, + "grad_norm": 1.7119297170965682, + "learning_rate": 0.00012016187852241583, + "loss": 1.7297, + "step": 5059 + }, + { + "epoch": 0.45243204577968527, + "grad_norm": 3.0120808900803007, + "learning_rate": 0.00012013351229082792, + "loss": 2.1719, + "step": 5060 + }, + { + "epoch": 0.45252145922746784, + "grad_norm": 0.9602518486791851, + "learning_rate": 0.00012010514437066664, + "loss": 2.4203, + "step": 5061 + }, + { + "epoch": 0.45261087267525035, + "grad_norm": 0.4839832898935171, + "learning_rate": 0.00012007677476431123, + "loss": 1.936, + "step": 5062 + }, + { + "epoch": 0.4527002861230329, + "grad_norm": 1.2512229886695412, + "learning_rate": 0.00012004840347414092, + "loss": 1.828, + "step": 5063 + }, + { + "epoch": 0.45278969957081544, + "grad_norm": 0.5003848291560069, + "learning_rate": 0.00012002003050253522, + "loss": 1.8629, + "step": 5064 + }, + { + "epoch": 0.452879113018598, + "grad_norm": 0.9328154710243161, + "learning_rate": 0.0001199916558518738, + "loss": 2.0113, + "step": 5065 + }, + { + "epoch": 0.4529685264663805, + "grad_norm": 0.49460360998771136, + "learning_rate": 0.00011996327952453629, + "loss": 1.8382, + "step": 5066 + }, + { + "epoch": 0.4530579399141631, + "grad_norm": 0.5325520528817288, + "learning_rate": 0.00011993490152290266, + "loss": 1.7876, + "step": 5067 + }, + { + "epoch": 0.4531473533619456, + "grad_norm": 0.6216750261760914, + "learning_rate": 0.00011990652184935289, + "loss": 1.6235, + "step": 5068 + }, + { + "epoch": 0.4532367668097282, + "grad_norm": 1.2667811887933669, + "learning_rate": 0.00011987814050626722, + "loss": 2.1935, + "step": 5069 + }, + { + "epoch": 0.4533261802575107, + "grad_norm": 1.355616712932552, + "learning_rate": 0.00011984975749602588, + "loss": 1.9869, + "step": 5070 + }, + { + "epoch": 0.4534155937052933, + "grad_norm": 1.5358422542601398, + "learning_rate": 0.00011982137282100934, + "loss": 1.6285, + "step": 5071 + }, + { + "epoch": 0.45350500715307585, + "grad_norm": 0.9570486426193999, + "learning_rate": 0.00011979298648359823, + "loss": 1.9948, + "step": 5072 + }, + { + "epoch": 0.45359442060085836, + "grad_norm": 0.7367184191662982, + "learning_rate": 0.00011976459848617323, + "loss": 2.4785, + "step": 5073 + }, + { + "epoch": 0.45368383404864093, + "grad_norm": 2.1935181022819195, + "learning_rate": 0.00011973620883111521, + "loss": 1.818, + "step": 5074 + }, + { + "epoch": 0.45377324749642345, + "grad_norm": 0.5645900827441173, + "learning_rate": 0.00011970781752080523, + "loss": 1.7516, + "step": 5075 + }, + { + "epoch": 0.453862660944206, + "grad_norm": 2.135419930224523, + "learning_rate": 0.00011967942455762437, + "loss": 1.9115, + "step": 5076 + }, + { + "epoch": 0.45395207439198854, + "grad_norm": 0.9359731114734325, + "learning_rate": 0.00011965102994395394, + "loss": 2.2031, + "step": 5077 + }, + { + "epoch": 0.4540414878397711, + "grad_norm": 1.0070688588147891, + "learning_rate": 0.00011962263368217535, + "loss": 1.9311, + "step": 5078 + }, + { + "epoch": 0.4541309012875536, + "grad_norm": 0.8935837818255644, + "learning_rate": 0.0001195942357746702, + "loss": 2.1122, + "step": 5079 + }, + { + "epoch": 0.4542203147353362, + "grad_norm": 0.9773162769537682, + "learning_rate": 0.00011956583622382015, + "loss": 2.0686, + "step": 5080 + }, + { + "epoch": 0.45430972818311877, + "grad_norm": 1.9542110321570842, + "learning_rate": 0.000119537435032007, + "loss": 1.7428, + "step": 5081 + }, + { + "epoch": 0.4543991416309013, + "grad_norm": 0.9850348038905047, + "learning_rate": 0.00011950903220161285, + "loss": 1.8171, + "step": 5082 + }, + { + "epoch": 0.45448855507868385, + "grad_norm": 1.5532898265029103, + "learning_rate": 0.00011948062773501969, + "loss": 1.8494, + "step": 5083 + }, + { + "epoch": 0.45457796852646637, + "grad_norm": 0.9796993591001816, + "learning_rate": 0.00011945222163460979, + "loss": 1.9319, + "step": 5084 + }, + { + "epoch": 0.45466738197424894, + "grad_norm": 2.658029586026255, + "learning_rate": 0.00011942381390276556, + "loss": 2.0404, + "step": 5085 + }, + { + "epoch": 0.45475679542203146, + "grad_norm": 1.6026841548140716, + "learning_rate": 0.00011939540454186954, + "loss": 1.8487, + "step": 5086 + }, + { + "epoch": 0.45484620886981403, + "grad_norm": 0.8877943978178183, + "learning_rate": 0.00011936699355430436, + "loss": 2.0397, + "step": 5087 + }, + { + "epoch": 0.45493562231759654, + "grad_norm": 1.7455719711068969, + "learning_rate": 0.00011933858094245281, + "loss": 1.9461, + "step": 5088 + }, + { + "epoch": 0.4550250357653791, + "grad_norm": 1.1925982453204196, + "learning_rate": 0.00011931016670869784, + "loss": 2.4017, + "step": 5089 + }, + { + "epoch": 0.4551144492131617, + "grad_norm": 0.6443558523384139, + "learning_rate": 0.0001192817508554225, + "loss": 2.068, + "step": 5090 + }, + { + "epoch": 0.4552038626609442, + "grad_norm": 1.623166057909814, + "learning_rate": 0.00011925333338500999, + "loss": 1.6244, + "step": 5091 + }, + { + "epoch": 0.4552932761087268, + "grad_norm": 0.8931389183830775, + "learning_rate": 0.00011922491429984368, + "loss": 1.6614, + "step": 5092 + }, + { + "epoch": 0.4553826895565093, + "grad_norm": 1.4216684699324762, + "learning_rate": 0.00011919649360230702, + "loss": 1.7966, + "step": 5093 + }, + { + "epoch": 0.45547210300429186, + "grad_norm": 0.8791851256558251, + "learning_rate": 0.00011916807129478363, + "loss": 1.9666, + "step": 5094 + }, + { + "epoch": 0.4555615164520744, + "grad_norm": 2.803239150618587, + "learning_rate": 0.00011913964737965723, + "loss": 1.5414, + "step": 5095 + }, + { + "epoch": 0.45565092989985695, + "grad_norm": 2.5618145055080364, + "learning_rate": 0.00011911122185931176, + "loss": 2.1442, + "step": 5096 + }, + { + "epoch": 0.45574034334763946, + "grad_norm": 1.3407530824660754, + "learning_rate": 0.00011908279473613115, + "loss": 2.0512, + "step": 5097 + }, + { + "epoch": 0.45582975679542204, + "grad_norm": 1.3084395190725169, + "learning_rate": 0.0001190543660124996, + "loss": 1.6509, + "step": 5098 + }, + { + "epoch": 0.45591917024320455, + "grad_norm": 1.6440404918351417, + "learning_rate": 0.0001190259356908014, + "loss": 2.019, + "step": 5099 + }, + { + "epoch": 0.4560085836909871, + "grad_norm": 1.0931935128359103, + "learning_rate": 0.00011899750377342098, + "loss": 1.5962, + "step": 5100 + }, + { + "epoch": 0.4560979971387697, + "grad_norm": 2.2878652814202707, + "learning_rate": 0.0001189690702627428, + "loss": 2.1514, + "step": 5101 + }, + { + "epoch": 0.4561874105865522, + "grad_norm": 1.485176539998129, + "learning_rate": 0.00011894063516115163, + "loss": 2.0532, + "step": 5102 + }, + { + "epoch": 0.4562768240343348, + "grad_norm": 2.8664195126364755, + "learning_rate": 0.00011891219847103228, + "loss": 1.6062, + "step": 5103 + }, + { + "epoch": 0.4563662374821173, + "grad_norm": 1.7544781466450494, + "learning_rate": 0.00011888376019476966, + "loss": 1.6882, + "step": 5104 + }, + { + "epoch": 0.45645565092989987, + "grad_norm": 1.119267747781596, + "learning_rate": 0.00011885532033474889, + "loss": 2.4508, + "step": 5105 + }, + { + "epoch": 0.4565450643776824, + "grad_norm": 1.0394747466257694, + "learning_rate": 0.00011882687889335518, + "loss": 1.9945, + "step": 5106 + }, + { + "epoch": 0.45663447782546496, + "grad_norm": 2.0325564874153477, + "learning_rate": 0.00011879843587297387, + "loss": 1.84, + "step": 5107 + }, + { + "epoch": 0.4567238912732475, + "grad_norm": 0.745198945988941, + "learning_rate": 0.00011876999127599042, + "loss": 1.4755, + "step": 5108 + }, + { + "epoch": 0.45681330472103004, + "grad_norm": 0.7858120924472495, + "learning_rate": 0.00011874154510479052, + "loss": 1.6049, + "step": 5109 + }, + { + "epoch": 0.4569027181688126, + "grad_norm": 3.582160952510693, + "learning_rate": 0.00011871309736175984, + "loss": 1.3721, + "step": 5110 + }, + { + "epoch": 0.45699213161659513, + "grad_norm": 4.0075575257361615, + "learning_rate": 0.0001186846480492843, + "loss": 1.8882, + "step": 5111 + }, + { + "epoch": 0.4570815450643777, + "grad_norm": 1.6250494232225305, + "learning_rate": 0.00011865619716974984, + "loss": 1.8039, + "step": 5112 + }, + { + "epoch": 0.4571709585121602, + "grad_norm": 1.809191418360033, + "learning_rate": 0.00011862774472554272, + "loss": 1.9286, + "step": 5113 + }, + { + "epoch": 0.4572603719599428, + "grad_norm": 1.1693727192107568, + "learning_rate": 0.00011859929071904912, + "loss": 1.8839, + "step": 5114 + }, + { + "epoch": 0.4573497854077253, + "grad_norm": 1.5214839107839322, + "learning_rate": 0.00011857083515265546, + "loss": 2.0928, + "step": 5115 + }, + { + "epoch": 0.4574391988555079, + "grad_norm": 1.0334979901075034, + "learning_rate": 0.0001185423780287483, + "loss": 2.0525, + "step": 5116 + }, + { + "epoch": 0.4575286123032904, + "grad_norm": 0.5095553872912031, + "learning_rate": 0.0001185139193497143, + "loss": 2.5626, + "step": 5117 + }, + { + "epoch": 0.45761802575107297, + "grad_norm": 1.6590399369297635, + "learning_rate": 0.0001184854591179402, + "loss": 1.8001, + "step": 5118 + }, + { + "epoch": 0.4577074391988555, + "grad_norm": 1.8192075641624144, + "learning_rate": 0.000118456997335813, + "loss": 1.5532, + "step": 5119 + }, + { + "epoch": 0.45779685264663805, + "grad_norm": 0.6375112535656413, + "learning_rate": 0.00011842853400571971, + "loss": 2.3446, + "step": 5120 + }, + { + "epoch": 0.4578862660944206, + "grad_norm": 1.3484255410803907, + "learning_rate": 0.00011840006913004753, + "loss": 1.7557, + "step": 5121 + }, + { + "epoch": 0.45797567954220314, + "grad_norm": 3.312942560920212, + "learning_rate": 0.00011837160271118377, + "loss": 1.8929, + "step": 5122 + }, + { + "epoch": 0.4580650929899857, + "grad_norm": 1.6934146304208681, + "learning_rate": 0.00011834313475151591, + "loss": 1.9592, + "step": 5123 + }, + { + "epoch": 0.4581545064377682, + "grad_norm": 3.0791085079700475, + "learning_rate": 0.00011831466525343146, + "loss": 2.1531, + "step": 5124 + }, + { + "epoch": 0.4582439198855508, + "grad_norm": 2.121953885000069, + "learning_rate": 0.00011828619421931817, + "loss": 1.6336, + "step": 5125 + }, + { + "epoch": 0.4583333333333333, + "grad_norm": 2.359296716912432, + "learning_rate": 0.00011825772165156384, + "loss": 2.1852, + "step": 5126 + }, + { + "epoch": 0.4584227467811159, + "grad_norm": 4.076333700052645, + "learning_rate": 0.00011822924755255647, + "loss": 1.9898, + "step": 5127 + }, + { + "epoch": 0.4585121602288984, + "grad_norm": 0.8090966802260404, + "learning_rate": 0.00011820077192468409, + "loss": 1.839, + "step": 5128 + }, + { + "epoch": 0.458601573676681, + "grad_norm": 1.627440748434925, + "learning_rate": 0.00011817229477033495, + "loss": 1.7498, + "step": 5129 + }, + { + "epoch": 0.45869098712446355, + "grad_norm": 1.2612824437668995, + "learning_rate": 0.00011814381609189741, + "loss": 1.6992, + "step": 5130 + }, + { + "epoch": 0.45878040057224606, + "grad_norm": 10.250307519933552, + "learning_rate": 0.00011811533589175994, + "loss": 7.0867, + "step": 5131 + }, + { + "epoch": 0.45886981402002863, + "grad_norm": 21.808215251408473, + "learning_rate": 0.00011808685417231111, + "loss": 2.2289, + "step": 5132 + }, + { + "epoch": 0.45895922746781115, + "grad_norm": 10.09412720260158, + "learning_rate": 0.00011805837093593965, + "loss": 2.2849, + "step": 5133 + }, + { + "epoch": 0.4590486409155937, + "grad_norm": 5.404371145601023, + "learning_rate": 0.00011802988618503447, + "loss": 2.3961, + "step": 5134 + }, + { + "epoch": 0.45913805436337624, + "grad_norm": 5.912462915798831, + "learning_rate": 0.00011800139992198447, + "loss": 1.802, + "step": 5135 + }, + { + "epoch": 0.4592274678111588, + "grad_norm": 2.0386467637576726, + "learning_rate": 0.00011797291214917881, + "loss": 1.7053, + "step": 5136 + }, + { + "epoch": 0.4593168812589413, + "grad_norm": 19.91974348576734, + "learning_rate": 0.00011794442286900673, + "loss": 7.1902, + "step": 5137 + }, + { + "epoch": 0.4594062947067239, + "grad_norm": 2.9157819025634915, + "learning_rate": 0.00011791593208385756, + "loss": 2.0775, + "step": 5138 + }, + { + "epoch": 0.4594957081545064, + "grad_norm": 3.603607520908811, + "learning_rate": 0.0001178874397961208, + "loss": 2.2174, + "step": 5139 + }, + { + "epoch": 0.459585121602289, + "grad_norm": 2.6451864391079214, + "learning_rate": 0.00011785894600818608, + "loss": 1.9238, + "step": 5140 + }, + { + "epoch": 0.45967453505007155, + "grad_norm": 1.0509462140305104, + "learning_rate": 0.0001178304507224431, + "loss": 1.4169, + "step": 5141 + }, + { + "epoch": 0.45976394849785407, + "grad_norm": 2.0487154363295503, + "learning_rate": 0.0001178019539412818, + "loss": 1.7141, + "step": 5142 + }, + { + "epoch": 0.45985336194563664, + "grad_norm": 2.161908140709416, + "learning_rate": 0.00011777345566709206, + "loss": 1.8768, + "step": 5143 + }, + { + "epoch": 0.45994277539341916, + "grad_norm": 2.31281851464355, + "learning_rate": 0.00011774495590226411, + "loss": 1.7359, + "step": 5144 + }, + { + "epoch": 0.46003218884120173, + "grad_norm": 1.5929303949503364, + "learning_rate": 0.00011771645464918813, + "loss": 2.0973, + "step": 5145 + }, + { + "epoch": 0.46012160228898424, + "grad_norm": 2.806576272918829, + "learning_rate": 0.00011768795191025445, + "loss": 2.1307, + "step": 5146 + }, + { + "epoch": 0.4602110157367668, + "grad_norm": 2.2352438453435948, + "learning_rate": 0.00011765944768785366, + "loss": 2.1144, + "step": 5147 + }, + { + "epoch": 0.46030042918454933, + "grad_norm": 1.730099006377854, + "learning_rate": 0.0001176309419843763, + "loss": 1.8365, + "step": 5148 + }, + { + "epoch": 0.4603898426323319, + "grad_norm": 1.979803649494865, + "learning_rate": 0.00011760243480221313, + "loss": 2.5384, + "step": 5149 + }, + { + "epoch": 0.4604792560801145, + "grad_norm": 1.720493064878782, + "learning_rate": 0.000117573926143755, + "loss": 2.2353, + "step": 5150 + }, + { + "epoch": 0.460568669527897, + "grad_norm": 1.241873893388436, + "learning_rate": 0.00011754541601139292, + "loss": 1.6652, + "step": 5151 + }, + { + "epoch": 0.46065808297567956, + "grad_norm": 0.8378242033059946, + "learning_rate": 0.000117516904407518, + "loss": 2.0547, + "step": 5152 + }, + { + "epoch": 0.4607474964234621, + "grad_norm": 4.591172069256219, + "learning_rate": 0.00011748839133452143, + "loss": 1.7537, + "step": 5153 + }, + { + "epoch": 0.46083690987124465, + "grad_norm": 1.9248761808985564, + "learning_rate": 0.00011745987679479462, + "loss": 1.8959, + "step": 5154 + }, + { + "epoch": 0.46092632331902716, + "grad_norm": 1.748038949220072, + "learning_rate": 0.00011743136079072903, + "loss": 1.9434, + "step": 5155 + }, + { + "epoch": 0.46101573676680974, + "grad_norm": 2.1400904466074255, + "learning_rate": 0.00011740284332471628, + "loss": 2.1014, + "step": 5156 + }, + { + "epoch": 0.46110515021459225, + "grad_norm": 1.9928884888691796, + "learning_rate": 0.00011737432439914804, + "loss": 1.9108, + "step": 5157 + }, + { + "epoch": 0.4611945636623748, + "grad_norm": 1.545166034327489, + "learning_rate": 0.0001173458040164162, + "loss": 2.2972, + "step": 5158 + }, + { + "epoch": 0.4612839771101574, + "grad_norm": 1.5415691905999753, + "learning_rate": 0.00011731728217891275, + "loss": 2.2791, + "step": 5159 + }, + { + "epoch": 0.4613733905579399, + "grad_norm": 3.16272359547965, + "learning_rate": 0.00011728875888902975, + "loss": 2.6927, + "step": 5160 + }, + { + "epoch": 0.4614628040057225, + "grad_norm": 1.482422631034482, + "learning_rate": 0.00011726023414915941, + "loss": 2.0169, + "step": 5161 + }, + { + "epoch": 0.461552217453505, + "grad_norm": 0.9255584130761444, + "learning_rate": 0.00011723170796169409, + "loss": 1.715, + "step": 5162 + }, + { + "epoch": 0.46164163090128757, + "grad_norm": 4.324178556544227, + "learning_rate": 0.00011720318032902624, + "loss": 1.7605, + "step": 5163 + }, + { + "epoch": 0.4617310443490701, + "grad_norm": 10.155128237237022, + "learning_rate": 0.0001171746512535484, + "loss": 7.4549, + "step": 5164 + }, + { + "epoch": 0.46182045779685266, + "grad_norm": 3.1111400001199265, + "learning_rate": 0.00011714612073765332, + "loss": 1.8094, + "step": 5165 + }, + { + "epoch": 0.4619098712446352, + "grad_norm": 2.7921062424174585, + "learning_rate": 0.00011711758878373383, + "loss": 2.098, + "step": 5166 + }, + { + "epoch": 0.46199928469241774, + "grad_norm": 7.875068543286191, + "learning_rate": 0.0001170890553941828, + "loss": 1.9224, + "step": 5167 + }, + { + "epoch": 0.46208869814020026, + "grad_norm": 1.0204182255152168, + "learning_rate": 0.00011706052057139335, + "loss": 2.1257, + "step": 5168 + }, + { + "epoch": 0.46217811158798283, + "grad_norm": 1.0324101565666806, + "learning_rate": 0.00011703198431775863, + "loss": 2.0586, + "step": 5169 + }, + { + "epoch": 0.4622675250357654, + "grad_norm": 1.8995010760021658, + "learning_rate": 0.00011700344663567197, + "loss": 1.7765, + "step": 5170 + }, + { + "epoch": 0.4623569384835479, + "grad_norm": 1.8459540511277719, + "learning_rate": 0.00011697490752752678, + "loss": 2.2507, + "step": 5171 + }, + { + "epoch": 0.4624463519313305, + "grad_norm": 1.11542144772558, + "learning_rate": 0.00011694636699571657, + "loss": 2.3241, + "step": 5172 + }, + { + "epoch": 0.462535765379113, + "grad_norm": 0.962508016616869, + "learning_rate": 0.00011691782504263505, + "loss": 2.2643, + "step": 5173 + }, + { + "epoch": 0.4626251788268956, + "grad_norm": 1.7160680201354324, + "learning_rate": 0.00011688928167067597, + "loss": 1.7606, + "step": 5174 + }, + { + "epoch": 0.4627145922746781, + "grad_norm": 1.6766268236987742, + "learning_rate": 0.0001168607368822332, + "loss": 2.0993, + "step": 5175 + }, + { + "epoch": 0.46280400572246067, + "grad_norm": 0.6250931251948254, + "learning_rate": 0.00011683219067970084, + "loss": 1.6331, + "step": 5176 + }, + { + "epoch": 0.4628934191702432, + "grad_norm": 3.0052462287661608, + "learning_rate": 0.00011680364306547298, + "loss": 2.1737, + "step": 5177 + }, + { + "epoch": 0.46298283261802575, + "grad_norm": 0.7277192748314749, + "learning_rate": 0.00011677509404194382, + "loss": 1.752, + "step": 5178 + }, + { + "epoch": 0.4630722460658083, + "grad_norm": 1.4560635679013358, + "learning_rate": 0.0001167465436115078, + "loss": 2.3738, + "step": 5179 + }, + { + "epoch": 0.46316165951359084, + "grad_norm": 2.9017365050130115, + "learning_rate": 0.00011671799177655942, + "loss": 1.8046, + "step": 5180 + }, + { + "epoch": 0.4632510729613734, + "grad_norm": 2.021250695681336, + "learning_rate": 0.00011668943853949323, + "loss": 1.7829, + "step": 5181 + }, + { + "epoch": 0.4633404864091559, + "grad_norm": 2.0021858936452586, + "learning_rate": 0.000116660883902704, + "loss": 1.9584, + "step": 5182 + }, + { + "epoch": 0.4634298998569385, + "grad_norm": 3.340554104171567, + "learning_rate": 0.00011663232786858656, + "loss": 1.5241, + "step": 5183 + }, + { + "epoch": 0.463519313304721, + "grad_norm": 0.6378225966674198, + "learning_rate": 0.00011660377043953588, + "loss": 2.3058, + "step": 5184 + }, + { + "epoch": 0.4636087267525036, + "grad_norm": 0.6583283122913737, + "learning_rate": 0.000116575211617947, + "loss": 2.5586, + "step": 5185 + }, + { + "epoch": 0.4636981402002861, + "grad_norm": 3.1148314567399966, + "learning_rate": 0.00011654665140621515, + "loss": 1.98, + "step": 5186 + }, + { + "epoch": 0.4637875536480687, + "grad_norm": 0.9603961581722997, + "learning_rate": 0.00011651808980673561, + "loss": 1.8826, + "step": 5187 + }, + { + "epoch": 0.4638769670958512, + "grad_norm": 2.1475140918516504, + "learning_rate": 0.00011648952682190387, + "loss": 2.294, + "step": 5188 + }, + { + "epoch": 0.46396638054363376, + "grad_norm": 0.8896491239162271, + "learning_rate": 0.00011646096245411538, + "loss": 1.9642, + "step": 5189 + }, + { + "epoch": 0.46405579399141633, + "grad_norm": 2.907284158730975, + "learning_rate": 0.00011643239670576589, + "loss": 1.7367, + "step": 5190 + }, + { + "epoch": 0.46414520743919885, + "grad_norm": 1.3778174506341718, + "learning_rate": 0.00011640382957925111, + "loss": 1.928, + "step": 5191 + }, + { + "epoch": 0.4642346208869814, + "grad_norm": 1.869500052094811, + "learning_rate": 0.00011637526107696694, + "loss": 1.9981, + "step": 5192 + }, + { + "epoch": 0.46432403433476394, + "grad_norm": 0.9309329633891053, + "learning_rate": 0.00011634669120130943, + "loss": 2.0116, + "step": 5193 + }, + { + "epoch": 0.4644134477825465, + "grad_norm": 1.0871737140082123, + "learning_rate": 0.00011631811995467467, + "loss": 1.8697, + "step": 5194 + }, + { + "epoch": 0.464502861230329, + "grad_norm": 1.006983366309099, + "learning_rate": 0.0001162895473394589, + "loss": 1.9936, + "step": 5195 + }, + { + "epoch": 0.4645922746781116, + "grad_norm": 0.850068659814209, + "learning_rate": 0.00011626097335805843, + "loss": 1.6525, + "step": 5196 + }, + { + "epoch": 0.4646816881258941, + "grad_norm": 1.4343017378038807, + "learning_rate": 0.00011623239801286981, + "loss": 1.7482, + "step": 5197 + }, + { + "epoch": 0.4647711015736767, + "grad_norm": 2.1023611777660562, + "learning_rate": 0.00011620382130628952, + "loss": 2.3403, + "step": 5198 + }, + { + "epoch": 0.46486051502145925, + "grad_norm": 1.0226008262833364, + "learning_rate": 0.00011617524324071433, + "loss": 1.7926, + "step": 5199 + }, + { + "epoch": 0.46494992846924177, + "grad_norm": 0.8391558133377911, + "learning_rate": 0.00011614666381854107, + "loss": 1.724, + "step": 5200 + }, + { + "epoch": 0.46503934191702434, + "grad_norm": 2.3589071578464025, + "learning_rate": 0.00011611808304216658, + "loss": 1.7673, + "step": 5201 + }, + { + "epoch": 0.46512875536480686, + "grad_norm": 1.8037337440832382, + "learning_rate": 0.0001160895009139879, + "loss": 1.7411, + "step": 5202 + }, + { + "epoch": 0.4652181688125894, + "grad_norm": 1.2425409349785843, + "learning_rate": 0.00011606091743640224, + "loss": 1.5929, + "step": 5203 + }, + { + "epoch": 0.46530758226037194, + "grad_norm": 1.250402418348105, + "learning_rate": 0.00011603233261180683, + "loss": 2.1221, + "step": 5204 + }, + { + "epoch": 0.4653969957081545, + "grad_norm": 2.0816834100753376, + "learning_rate": 0.00011600374644259906, + "loss": 1.7004, + "step": 5205 + }, + { + "epoch": 0.46548640915593703, + "grad_norm": 1.360820794501669, + "learning_rate": 0.00011597515893117637, + "loss": 1.8506, + "step": 5206 + }, + { + "epoch": 0.4655758226037196, + "grad_norm": 0.9668172381728477, + "learning_rate": 0.00011594657007993644, + "loss": 2.3041, + "step": 5207 + }, + { + "epoch": 0.4656652360515021, + "grad_norm": 1.2451705513719185, + "learning_rate": 0.0001159179798912769, + "loss": 1.8041, + "step": 5208 + }, + { + "epoch": 0.4657546494992847, + "grad_norm": 2.4457014919872475, + "learning_rate": 0.0001158893883675956, + "loss": 2.1576, + "step": 5209 + }, + { + "epoch": 0.46584406294706726, + "grad_norm": 1.6001912452072238, + "learning_rate": 0.00011586079551129053, + "loss": 2.2604, + "step": 5210 + }, + { + "epoch": 0.4659334763948498, + "grad_norm": 2.472458911768164, + "learning_rate": 0.00011583220132475966, + "loss": 1.914, + "step": 5211 + }, + { + "epoch": 0.46602288984263235, + "grad_norm": 3.1194763547587256, + "learning_rate": 0.0001158036058104012, + "loss": 1.6734, + "step": 5212 + }, + { + "epoch": 0.46611230329041486, + "grad_norm": 2.559204475658275, + "learning_rate": 0.00011577500897061338, + "loss": 1.538, + "step": 5213 + }, + { + "epoch": 0.46620171673819744, + "grad_norm": 0.813592967088317, + "learning_rate": 0.00011574641080779464, + "loss": 1.7421, + "step": 5214 + }, + { + "epoch": 0.46629113018597995, + "grad_norm": 2.6278182680927773, + "learning_rate": 0.00011571781132434343, + "loss": 2.13, + "step": 5215 + }, + { + "epoch": 0.4663805436337625, + "grad_norm": 0.8946014267691955, + "learning_rate": 0.00011568921052265836, + "loss": 2.1184, + "step": 5216 + }, + { + "epoch": 0.46646995708154504, + "grad_norm": 1.3852123131665033, + "learning_rate": 0.00011566060840513817, + "loss": 1.9422, + "step": 5217 + }, + { + "epoch": 0.4665593705293276, + "grad_norm": 2.0874184761456838, + "learning_rate": 0.00011563200497418168, + "loss": 1.7833, + "step": 5218 + }, + { + "epoch": 0.4666487839771102, + "grad_norm": 1.412357082375577, + "learning_rate": 0.00011560340023218776, + "loss": 2.0143, + "step": 5219 + }, + { + "epoch": 0.4667381974248927, + "grad_norm": 2.560672734246172, + "learning_rate": 0.00011557479418155555, + "loss": 1.7887, + "step": 5220 + }, + { + "epoch": 0.46682761087267527, + "grad_norm": 2.2544374439090755, + "learning_rate": 0.00011554618682468416, + "loss": 1.8884, + "step": 5221 + }, + { + "epoch": 0.4669170243204578, + "grad_norm": 1.507470190708272, + "learning_rate": 0.00011551757816397285, + "loss": 1.803, + "step": 5222 + }, + { + "epoch": 0.46700643776824036, + "grad_norm": 2.436889507453688, + "learning_rate": 0.00011548896820182095, + "loss": 1.5293, + "step": 5223 + }, + { + "epoch": 0.4670958512160229, + "grad_norm": 0.6613946067508567, + "learning_rate": 0.00011546035694062806, + "loss": 1.9933, + "step": 5224 + }, + { + "epoch": 0.46718526466380544, + "grad_norm": 2.965570988023064, + "learning_rate": 0.0001154317443827937, + "loss": 1.7733, + "step": 5225 + }, + { + "epoch": 0.46727467811158796, + "grad_norm": 0.7786707613500337, + "learning_rate": 0.00011540313053071752, + "loss": 2.644, + "step": 5226 + }, + { + "epoch": 0.46736409155937053, + "grad_norm": 1.4523070951014014, + "learning_rate": 0.00011537451538679944, + "loss": 2.2282, + "step": 5227 + }, + { + "epoch": 0.4674535050071531, + "grad_norm": 2.786917906745827, + "learning_rate": 0.00011534589895343933, + "loss": 1.6734, + "step": 5228 + }, + { + "epoch": 0.4675429184549356, + "grad_norm": 1.417717951017341, + "learning_rate": 0.00011531728123303715, + "loss": 1.9383, + "step": 5229 + }, + { + "epoch": 0.4676323319027182, + "grad_norm": 1.4618918412161381, + "learning_rate": 0.00011528866222799313, + "loss": 1.9391, + "step": 5230 + }, + { + "epoch": 0.4677217453505007, + "grad_norm": 2.144007757317337, + "learning_rate": 0.00011526004194070748, + "loss": 2.2306, + "step": 5231 + }, + { + "epoch": 0.4678111587982833, + "grad_norm": 0.42918112868863734, + "learning_rate": 0.0001152314203735805, + "loss": 2.3095, + "step": 5232 + }, + { + "epoch": 0.4679005722460658, + "grad_norm": 2.18226374979862, + "learning_rate": 0.00011520279752901273, + "loss": 2.1306, + "step": 5233 + }, + { + "epoch": 0.46798998569384836, + "grad_norm": 1.2902595259723988, + "learning_rate": 0.00011517417340940468, + "loss": 1.9571, + "step": 5234 + }, + { + "epoch": 0.4680793991416309, + "grad_norm": 0.5039898205387388, + "learning_rate": 0.00011514554801715704, + "loss": 1.9271, + "step": 5235 + }, + { + "epoch": 0.46816881258941345, + "grad_norm": 1.4890583201697094, + "learning_rate": 0.00011511692135467054, + "loss": 1.8844, + "step": 5236 + }, + { + "epoch": 0.46825822603719597, + "grad_norm": 0.804176769946178, + "learning_rate": 0.00011508829342434615, + "loss": 2.0809, + "step": 5237 + }, + { + "epoch": 0.46834763948497854, + "grad_norm": 0.9144301981911862, + "learning_rate": 0.00011505966422858481, + "loss": 1.5454, + "step": 5238 + }, + { + "epoch": 0.4684370529327611, + "grad_norm": 0.7412405354600966, + "learning_rate": 0.00011503103376978759, + "loss": 1.9017, + "step": 5239 + }, + { + "epoch": 0.4685264663805436, + "grad_norm": 1.3459561153269983, + "learning_rate": 0.00011500240205035573, + "loss": 2.1282, + "step": 5240 + }, + { + "epoch": 0.4686158798283262, + "grad_norm": 1.1260184042059063, + "learning_rate": 0.00011497376907269053, + "loss": 2.2266, + "step": 5241 + }, + { + "epoch": 0.4687052932761087, + "grad_norm": 0.648366138394463, + "learning_rate": 0.00011494513483919342, + "loss": 2.244, + "step": 5242 + }, + { + "epoch": 0.4687947067238913, + "grad_norm": 1.0295849041702114, + "learning_rate": 0.00011491649935226584, + "loss": 1.5323, + "step": 5243 + }, + { + "epoch": 0.4688841201716738, + "grad_norm": 0.6567866814059299, + "learning_rate": 0.00011488786261430954, + "loss": 1.9015, + "step": 5244 + }, + { + "epoch": 0.4689735336194564, + "grad_norm": 1.5714960100841402, + "learning_rate": 0.00011485922462772616, + "loss": 1.9368, + "step": 5245 + }, + { + "epoch": 0.4690629470672389, + "grad_norm": 0.8270503860582787, + "learning_rate": 0.00011483058539491756, + "loss": 1.9684, + "step": 5246 + }, + { + "epoch": 0.46915236051502146, + "grad_norm": 0.8030540928308264, + "learning_rate": 0.00011480194491828567, + "loss": 2.1195, + "step": 5247 + }, + { + "epoch": 0.46924177396280403, + "grad_norm": 2.213522225374091, + "learning_rate": 0.00011477330320023255, + "loss": 1.9634, + "step": 5248 + }, + { + "epoch": 0.46933118741058655, + "grad_norm": 1.1178622617967078, + "learning_rate": 0.00011474466024316029, + "loss": 2.0963, + "step": 5249 + }, + { + "epoch": 0.4694206008583691, + "grad_norm": 1.0631127846630233, + "learning_rate": 0.0001147160160494712, + "loss": 2.0402, + "step": 5250 + }, + { + "epoch": 0.46951001430615164, + "grad_norm": 1.1679454915451606, + "learning_rate": 0.00011468737062156765, + "loss": 1.6835, + "step": 5251 + }, + { + "epoch": 0.4695994277539342, + "grad_norm": 1.115217322071618, + "learning_rate": 0.00011465872396185204, + "loss": 2.3755, + "step": 5252 + }, + { + "epoch": 0.4696888412017167, + "grad_norm": 1.2037970746668054, + "learning_rate": 0.00011463007607272695, + "loss": 2.2486, + "step": 5253 + }, + { + "epoch": 0.4697782546494993, + "grad_norm": 1.3819867888269766, + "learning_rate": 0.00011460142695659503, + "loss": 2.0816, + "step": 5254 + }, + { + "epoch": 0.4698676680972818, + "grad_norm": 1.4929976925707482, + "learning_rate": 0.00011457277661585912, + "loss": 2.4415, + "step": 5255 + }, + { + "epoch": 0.4699570815450644, + "grad_norm": 1.5754950917151505, + "learning_rate": 0.000114544125052922, + "loss": 1.491, + "step": 5256 + }, + { + "epoch": 0.4700464949928469, + "grad_norm": 1.5189019544000082, + "learning_rate": 0.00011451547227018666, + "loss": 1.8354, + "step": 5257 + }, + { + "epoch": 0.47013590844062947, + "grad_norm": 1.1988468270171841, + "learning_rate": 0.00011448681827005623, + "loss": 1.9915, + "step": 5258 + }, + { + "epoch": 0.47022532188841204, + "grad_norm": 2.92506325847224, + "learning_rate": 0.00011445816305493382, + "loss": 1.9008, + "step": 5259 + }, + { + "epoch": 0.47031473533619456, + "grad_norm": 3.6060543734914523, + "learning_rate": 0.00011442950662722274, + "loss": 2.203, + "step": 5260 + }, + { + "epoch": 0.4704041487839771, + "grad_norm": 4.737488377814296, + "learning_rate": 0.00011440084898932637, + "loss": 6.9491, + "step": 5261 + }, + { + "epoch": 0.47049356223175964, + "grad_norm": 3.1693719043337576, + "learning_rate": 0.00011437219014364819, + "loss": 2.2286, + "step": 5262 + }, + { + "epoch": 0.4705829756795422, + "grad_norm": 0.7642265199246102, + "learning_rate": 0.00011434353009259178, + "loss": 1.7593, + "step": 5263 + }, + { + "epoch": 0.47067238912732473, + "grad_norm": 1.2695468826235414, + "learning_rate": 0.00011431486883856082, + "loss": 2.2366, + "step": 5264 + }, + { + "epoch": 0.4707618025751073, + "grad_norm": 1.1532244381696712, + "learning_rate": 0.0001142862063839591, + "loss": 1.7925, + "step": 5265 + }, + { + "epoch": 0.4708512160228898, + "grad_norm": 1.7995428010864147, + "learning_rate": 0.00011425754273119049, + "loss": 2.2085, + "step": 5266 + }, + { + "epoch": 0.4709406294706724, + "grad_norm": 1.6368867814864048, + "learning_rate": 0.00011422887788265901, + "loss": 1.771, + "step": 5267 + }, + { + "epoch": 0.47103004291845496, + "grad_norm": 0.5309329586316915, + "learning_rate": 0.00011420021184076872, + "loss": 1.6681, + "step": 5268 + }, + { + "epoch": 0.4711194563662375, + "grad_norm": 1.8010459038609874, + "learning_rate": 0.00011417154460792381, + "loss": 1.9105, + "step": 5269 + }, + { + "epoch": 0.47120886981402005, + "grad_norm": 2.0379974450638314, + "learning_rate": 0.00011414287618652857, + "loss": 1.6297, + "step": 5270 + }, + { + "epoch": 0.47129828326180256, + "grad_norm": 1.7049615204140978, + "learning_rate": 0.00011411420657898737, + "loss": 1.8351, + "step": 5271 + }, + { + "epoch": 0.47138769670958514, + "grad_norm": 1.7944868702419268, + "learning_rate": 0.00011408553578770473, + "loss": 2.1464, + "step": 5272 + }, + { + "epoch": 0.47147711015736765, + "grad_norm": 1.1079834962668873, + "learning_rate": 0.0001140568638150852, + "loss": 2.1249, + "step": 5273 + }, + { + "epoch": 0.4715665236051502, + "grad_norm": 1.2574797808208065, + "learning_rate": 0.00011402819066353348, + "loss": 2.2329, + "step": 5274 + }, + { + "epoch": 0.47165593705293274, + "grad_norm": 1.6153801727199426, + "learning_rate": 0.00011399951633545438, + "loss": 1.7022, + "step": 5275 + }, + { + "epoch": 0.4717453505007153, + "grad_norm": 2.4857610082821147, + "learning_rate": 0.00011397084083325271, + "loss": 2.0815, + "step": 5276 + }, + { + "epoch": 0.4718347639484979, + "grad_norm": 2.1120098864019137, + "learning_rate": 0.00011394216415933355, + "loss": 1.5623, + "step": 5277 + }, + { + "epoch": 0.4719241773962804, + "grad_norm": 1.1670660348672577, + "learning_rate": 0.00011391348631610186, + "loss": 1.7946, + "step": 5278 + }, + { + "epoch": 0.47201359084406297, + "grad_norm": 0.7364866917003046, + "learning_rate": 0.0001138848073059629, + "loss": 1.9754, + "step": 5279 + }, + { + "epoch": 0.4721030042918455, + "grad_norm": 3.781543995517819, + "learning_rate": 0.0001138561271313219, + "loss": 1.6548, + "step": 5280 + }, + { + "epoch": 0.47219241773962806, + "grad_norm": 1.4030838146062932, + "learning_rate": 0.00011382744579458426, + "loss": 1.9652, + "step": 5281 + }, + { + "epoch": 0.4722818311874106, + "grad_norm": 1.51637977648709, + "learning_rate": 0.00011379876329815546, + "loss": 1.7012, + "step": 5282 + }, + { + "epoch": 0.47237124463519314, + "grad_norm": 1.8284114154512474, + "learning_rate": 0.00011377007964444104, + "loss": 1.8569, + "step": 5283 + }, + { + "epoch": 0.47246065808297566, + "grad_norm": 3.035147183422895, + "learning_rate": 0.00011374139483584667, + "loss": 2.6596, + "step": 5284 + }, + { + "epoch": 0.47255007153075823, + "grad_norm": 1.8311513927247316, + "learning_rate": 0.00011371270887477809, + "loss": 1.7063, + "step": 5285 + }, + { + "epoch": 0.47263948497854075, + "grad_norm": 3.1069432695397854, + "learning_rate": 0.00011368402176364121, + "loss": 1.7478, + "step": 5286 + }, + { + "epoch": 0.4727288984263233, + "grad_norm": 4.257967539644738, + "learning_rate": 0.00011365533350484194, + "loss": 2.2098, + "step": 5287 + }, + { + "epoch": 0.4728183118741059, + "grad_norm": 2.722882993418148, + "learning_rate": 0.00011362664410078632, + "loss": 1.9795, + "step": 5288 + }, + { + "epoch": 0.4729077253218884, + "grad_norm": 2.8466078146660845, + "learning_rate": 0.00011359795355388054, + "loss": 1.9867, + "step": 5289 + }, + { + "epoch": 0.472997138769671, + "grad_norm": 1.041962589365019, + "learning_rate": 0.0001135692618665308, + "loss": 1.6864, + "step": 5290 + }, + { + "epoch": 0.4730865522174535, + "grad_norm": 0.6216839053921887, + "learning_rate": 0.00011354056904114347, + "loss": 1.8167, + "step": 5291 + }, + { + "epoch": 0.47317596566523606, + "grad_norm": 1.5206463289226406, + "learning_rate": 0.00011351187508012496, + "loss": 1.8239, + "step": 5292 + }, + { + "epoch": 0.4732653791130186, + "grad_norm": 1.2515020200470706, + "learning_rate": 0.0001134831799858818, + "loss": 1.712, + "step": 5293 + }, + { + "epoch": 0.47335479256080115, + "grad_norm": 1.5463106746267383, + "learning_rate": 0.00011345448376082064, + "loss": 1.8733, + "step": 5294 + }, + { + "epoch": 0.47344420600858367, + "grad_norm": 2.430395563187288, + "learning_rate": 0.00011342578640734816, + "loss": 6.7887, + "step": 5295 + }, + { + "epoch": 0.47353361945636624, + "grad_norm": 2.3790997586711655, + "learning_rate": 0.00011339708792787119, + "loss": 2.0162, + "step": 5296 + }, + { + "epoch": 0.4736230329041488, + "grad_norm": 1.5169979892824061, + "learning_rate": 0.00011336838832479661, + "loss": 1.9911, + "step": 5297 + }, + { + "epoch": 0.4737124463519313, + "grad_norm": 1.2384370804709437, + "learning_rate": 0.00011333968760053149, + "loss": 1.8669, + "step": 5298 + }, + { + "epoch": 0.4738018597997139, + "grad_norm": 0.7698227050197761, + "learning_rate": 0.00011331098575748284, + "loss": 1.604, + "step": 5299 + }, + { + "epoch": 0.4738912732474964, + "grad_norm": 3.6570901071752124, + "learning_rate": 0.00011328228279805792, + "loss": 2.1694, + "step": 5300 + }, + { + "epoch": 0.473980686695279, + "grad_norm": 1.3384902418148874, + "learning_rate": 0.00011325357872466398, + "loss": 2.0842, + "step": 5301 + }, + { + "epoch": 0.4740701001430615, + "grad_norm": 4.750584623677142, + "learning_rate": 0.00011322487353970838, + "loss": 1.6741, + "step": 5302 + }, + { + "epoch": 0.4741595135908441, + "grad_norm": 4.073770170326213, + "learning_rate": 0.00011319616724559866, + "loss": 2.0146, + "step": 5303 + }, + { + "epoch": 0.4742489270386266, + "grad_norm": 2.4656719321837515, + "learning_rate": 0.00011316745984474226, + "loss": 1.8066, + "step": 5304 + }, + { + "epoch": 0.47433834048640916, + "grad_norm": 0.994265019630157, + "learning_rate": 0.00011313875133954695, + "loss": 1.7088, + "step": 5305 + }, + { + "epoch": 0.4744277539341917, + "grad_norm": 1.955936838414413, + "learning_rate": 0.00011311004173242041, + "loss": 1.825, + "step": 5306 + }, + { + "epoch": 0.47451716738197425, + "grad_norm": 2.286498074099502, + "learning_rate": 0.0001130813310257705, + "loss": 1.4298, + "step": 5307 + }, + { + "epoch": 0.4746065808297568, + "grad_norm": 1.1176825781699093, + "learning_rate": 0.00011305261922200519, + "loss": 1.9401, + "step": 5308 + }, + { + "epoch": 0.47469599427753933, + "grad_norm": 7.0629726785739795, + "learning_rate": 0.00011302390632353241, + "loss": 7.0366, + "step": 5309 + }, + { + "epoch": 0.4747854077253219, + "grad_norm": 6.785046937199972, + "learning_rate": 0.00011299519233276037, + "loss": 2.2328, + "step": 5310 + }, + { + "epoch": 0.4748748211731044, + "grad_norm": 2.0580397113935502, + "learning_rate": 0.00011296647725209726, + "loss": 2.1668, + "step": 5311 + }, + { + "epoch": 0.474964234620887, + "grad_norm": 4.72672632911913, + "learning_rate": 0.00011293776108395135, + "loss": 2.0972, + "step": 5312 + }, + { + "epoch": 0.4750536480686695, + "grad_norm": 1.0776478014548667, + "learning_rate": 0.00011290904383073104, + "loss": 1.9618, + "step": 5313 + }, + { + "epoch": 0.4751430615164521, + "grad_norm": 3.6554178907151402, + "learning_rate": 0.0001128803254948448, + "loss": 1.7554, + "step": 5314 + }, + { + "epoch": 0.4752324749642346, + "grad_norm": 0.6511893138340222, + "learning_rate": 0.00011285160607870124, + "loss": 1.8998, + "step": 5315 + }, + { + "epoch": 0.47532188841201717, + "grad_norm": 2.8379420862803353, + "learning_rate": 0.000112822885584709, + "loss": 2.0082, + "step": 5316 + }, + { + "epoch": 0.47541130185979974, + "grad_norm": 1.071449178158558, + "learning_rate": 0.0001127941640152768, + "loss": 2.2031, + "step": 5317 + }, + { + "epoch": 0.47550071530758226, + "grad_norm": 1.8189958039865617, + "learning_rate": 0.00011276544137281355, + "loss": 2.2838, + "step": 5318 + }, + { + "epoch": 0.4755901287553648, + "grad_norm": 0.8605758966569419, + "learning_rate": 0.00011273671765972813, + "loss": 2.0715, + "step": 5319 + }, + { + "epoch": 0.47567954220314734, + "grad_norm": 0.7080388248282724, + "learning_rate": 0.00011270799287842957, + "loss": 2.1672, + "step": 5320 + }, + { + "epoch": 0.4757689556509299, + "grad_norm": 0.6264741758362624, + "learning_rate": 0.00011267926703132703, + "loss": 1.8893, + "step": 5321 + }, + { + "epoch": 0.47585836909871243, + "grad_norm": 0.8462629093453838, + "learning_rate": 0.00011265054012082967, + "loss": 1.5128, + "step": 5322 + }, + { + "epoch": 0.475947782546495, + "grad_norm": 1.1411178525549206, + "learning_rate": 0.00011262181214934677, + "loss": 1.7196, + "step": 5323 + }, + { + "epoch": 0.4760371959942775, + "grad_norm": 1.067114194314278, + "learning_rate": 0.00011259308311928771, + "loss": 1.6693, + "step": 5324 + }, + { + "epoch": 0.4761266094420601, + "grad_norm": 1.6112013947087205, + "learning_rate": 0.00011256435303306203, + "loss": 1.938, + "step": 5325 + }, + { + "epoch": 0.4762160228898426, + "grad_norm": 1.0057982238011818, + "learning_rate": 0.00011253562189307921, + "loss": 2.2956, + "step": 5326 + }, + { + "epoch": 0.4763054363376252, + "grad_norm": 1.3781847955428232, + "learning_rate": 0.0001125068897017489, + "loss": 2.1961, + "step": 5327 + }, + { + "epoch": 0.47639484978540775, + "grad_norm": 1.1559122090194232, + "learning_rate": 0.00011247815646148087, + "loss": 2.3025, + "step": 5328 + }, + { + "epoch": 0.47648426323319026, + "grad_norm": 0.6666680661661155, + "learning_rate": 0.00011244942217468495, + "loss": 2.4636, + "step": 5329 + }, + { + "epoch": 0.47657367668097284, + "grad_norm": 1.7131483952105744, + "learning_rate": 0.00011242068684377101, + "loss": 1.7629, + "step": 5330 + }, + { + "epoch": 0.47666309012875535, + "grad_norm": 0.800319144864144, + "learning_rate": 0.00011239195047114903, + "loss": 2.0806, + "step": 5331 + }, + { + "epoch": 0.4767525035765379, + "grad_norm": 2.9600917115962324, + "learning_rate": 0.00011236321305922919, + "loss": 1.6294, + "step": 5332 + }, + { + "epoch": 0.47684191702432044, + "grad_norm": 0.49729005770943285, + "learning_rate": 0.00011233447461042157, + "loss": 2.09, + "step": 5333 + }, + { + "epoch": 0.476931330472103, + "grad_norm": 2.285399297654077, + "learning_rate": 0.00011230573512713644, + "loss": 2.0732, + "step": 5334 + }, + { + "epoch": 0.4770207439198855, + "grad_norm": 0.9760881811073979, + "learning_rate": 0.00011227699461178423, + "loss": 2.203, + "step": 5335 + }, + { + "epoch": 0.4771101573676681, + "grad_norm": 1.2384716324686533, + "learning_rate": 0.00011224825306677527, + "loss": 1.8679, + "step": 5336 + }, + { + "epoch": 0.47719957081545067, + "grad_norm": 2.506221128989676, + "learning_rate": 0.00011221951049452009, + "loss": 1.7523, + "step": 5337 + }, + { + "epoch": 0.4772889842632332, + "grad_norm": 2.3860692518013824, + "learning_rate": 0.00011219076689742936, + "loss": 1.7757, + "step": 5338 + }, + { + "epoch": 0.47737839771101576, + "grad_norm": 2.8150884456405727, + "learning_rate": 0.00011216202227791373, + "loss": 2.0177, + "step": 5339 + }, + { + "epoch": 0.47746781115879827, + "grad_norm": 0.5393982615065513, + "learning_rate": 0.00011213327663838396, + "loss": 1.9539, + "step": 5340 + }, + { + "epoch": 0.47755722460658084, + "grad_norm": 3.2266059381001675, + "learning_rate": 0.00011210452998125094, + "loss": 2.0328, + "step": 5341 + }, + { + "epoch": 0.47764663805436336, + "grad_norm": 0.526815933178825, + "learning_rate": 0.00011207578230892562, + "loss": 1.498, + "step": 5342 + }, + { + "epoch": 0.47773605150214593, + "grad_norm": 1.5499238243431055, + "learning_rate": 0.00011204703362381903, + "loss": 2.327, + "step": 5343 + }, + { + "epoch": 0.47782546494992845, + "grad_norm": 1.0675614263883633, + "learning_rate": 0.00011201828392834223, + "loss": 1.7954, + "step": 5344 + }, + { + "epoch": 0.477914878397711, + "grad_norm": 2.086888726754073, + "learning_rate": 0.00011198953322490653, + "loss": 2.482, + "step": 5345 + }, + { + "epoch": 0.4780042918454936, + "grad_norm": 0.6242960516758245, + "learning_rate": 0.00011196078151592314, + "loss": 1.7097, + "step": 5346 + }, + { + "epoch": 0.4780937052932761, + "grad_norm": 3.0819302839196254, + "learning_rate": 0.00011193202880380343, + "loss": 1.5007, + "step": 5347 + }, + { + "epoch": 0.4781831187410587, + "grad_norm": 0.6088166771618446, + "learning_rate": 0.00011190327509095889, + "loss": 1.9371, + "step": 5348 + }, + { + "epoch": 0.4782725321888412, + "grad_norm": 0.8287816649731925, + "learning_rate": 0.00011187452037980104, + "loss": 2.017, + "step": 5349 + }, + { + "epoch": 0.47836194563662376, + "grad_norm": 2.0327701522289208, + "learning_rate": 0.0001118457646727415, + "loss": 2.0196, + "step": 5350 + }, + { + "epoch": 0.4784513590844063, + "grad_norm": 0.8011861832408559, + "learning_rate": 0.00011181700797219199, + "loss": 2.3032, + "step": 5351 + }, + { + "epoch": 0.47854077253218885, + "grad_norm": 2.207366741541213, + "learning_rate": 0.0001117882502805643, + "loss": 1.5005, + "step": 5352 + }, + { + "epoch": 0.47863018597997137, + "grad_norm": 0.750772044321214, + "learning_rate": 0.00011175949160027031, + "loss": 2.028, + "step": 5353 + }, + { + "epoch": 0.47871959942775394, + "grad_norm": 2.075129579406014, + "learning_rate": 0.0001117307319337219, + "loss": 1.7537, + "step": 5354 + }, + { + "epoch": 0.47880901287553645, + "grad_norm": 1.7435210752334716, + "learning_rate": 0.00011170197128333122, + "loss": 1.9099, + "step": 5355 + }, + { + "epoch": 0.478898426323319, + "grad_norm": 0.6606665226732948, + "learning_rate": 0.00011167320965151033, + "loss": 1.3755, + "step": 5356 + }, + { + "epoch": 0.4789878397711016, + "grad_norm": 3.692036681837508, + "learning_rate": 0.00011164444704067145, + "loss": 2.0618, + "step": 5357 + }, + { + "epoch": 0.4790772532188841, + "grad_norm": 1.1083359101228452, + "learning_rate": 0.00011161568345322684, + "loss": 1.7963, + "step": 5358 + }, + { + "epoch": 0.4791666666666667, + "grad_norm": 2.1964451050083373, + "learning_rate": 0.00011158691889158892, + "loss": 1.7706, + "step": 5359 + }, + { + "epoch": 0.4792560801144492, + "grad_norm": 1.8348140489395142, + "learning_rate": 0.00011155815335817011, + "loss": 1.7366, + "step": 5360 + }, + { + "epoch": 0.4793454935622318, + "grad_norm": 1.122873083890961, + "learning_rate": 0.00011152938685538287, + "loss": 1.8426, + "step": 5361 + }, + { + "epoch": 0.4794349070100143, + "grad_norm": 1.1375851579925476, + "learning_rate": 0.00011150061938563993, + "loss": 1.6945, + "step": 5362 + }, + { + "epoch": 0.47952432045779686, + "grad_norm": 1.6347042733014636, + "learning_rate": 0.00011147185095135395, + "loss": 1.8542, + "step": 5363 + }, + { + "epoch": 0.4796137339055794, + "grad_norm": 1.000319844423841, + "learning_rate": 0.00011144308155493763, + "loss": 1.8449, + "step": 5364 + }, + { + "epoch": 0.47970314735336195, + "grad_norm": 1.6790570785680694, + "learning_rate": 0.00011141431119880392, + "loss": 2.2111, + "step": 5365 + }, + { + "epoch": 0.4797925608011445, + "grad_norm": 1.321635924879938, + "learning_rate": 0.00011138553988536571, + "loss": 1.5203, + "step": 5366 + }, + { + "epoch": 0.47988197424892703, + "grad_norm": 0.8957871875907993, + "learning_rate": 0.000111356767617036, + "loss": 1.8261, + "step": 5367 + }, + { + "epoch": 0.4799713876967096, + "grad_norm": 0.5350501660609056, + "learning_rate": 0.00011132799439622792, + "loss": 1.6439, + "step": 5368 + }, + { + "epoch": 0.4800608011444921, + "grad_norm": 10.118307370871749, + "learning_rate": 0.00011129922022535464, + "loss": 7.0337, + "step": 5369 + }, + { + "epoch": 0.4801502145922747, + "grad_norm": 4.529642485369165, + "learning_rate": 0.0001112704451068294, + "loss": 1.6962, + "step": 5370 + }, + { + "epoch": 0.4802396280400572, + "grad_norm": 5.579064999502023, + "learning_rate": 0.0001112416690430655, + "loss": 2.1506, + "step": 5371 + }, + { + "epoch": 0.4803290414878398, + "grad_norm": 1.2674250388608908, + "learning_rate": 0.00011121289203647644, + "loss": 1.8376, + "step": 5372 + }, + { + "epoch": 0.4804184549356223, + "grad_norm": 5.738039909518425, + "learning_rate": 0.00011118411408947567, + "loss": 2.0077, + "step": 5373 + }, + { + "epoch": 0.48050786838340487, + "grad_norm": 3.509646587479051, + "learning_rate": 0.00011115533520447674, + "loss": 2.1903, + "step": 5374 + }, + { + "epoch": 0.4805972818311874, + "grad_norm": 2.501473550388217, + "learning_rate": 0.00011112655538389331, + "loss": 2.0463, + "step": 5375 + }, + { + "epoch": 0.48068669527896996, + "grad_norm": 2.9806409285554607, + "learning_rate": 0.00011109777463013915, + "loss": 1.4411, + "step": 5376 + }, + { + "epoch": 0.4807761087267525, + "grad_norm": 4.176781627348656, + "learning_rate": 0.000111068992945628, + "loss": 2.038, + "step": 5377 + }, + { + "epoch": 0.48086552217453504, + "grad_norm": 2.216309449904493, + "learning_rate": 0.00011104021033277379, + "loss": 2.2544, + "step": 5378 + }, + { + "epoch": 0.4809549356223176, + "grad_norm": 2.962375962172933, + "learning_rate": 0.00011101142679399049, + "loss": 1.9726, + "step": 5379 + }, + { + "epoch": 0.48104434907010013, + "grad_norm": 3.908231605456549, + "learning_rate": 0.00011098264233169211, + "loss": 1.7884, + "step": 5380 + }, + { + "epoch": 0.4811337625178827, + "grad_norm": 2.4900751794226186, + "learning_rate": 0.00011095385694829278, + "loss": 2.3013, + "step": 5381 + }, + { + "epoch": 0.4812231759656652, + "grad_norm": 2.4828164644779975, + "learning_rate": 0.0001109250706462067, + "loss": 1.9283, + "step": 5382 + }, + { + "epoch": 0.4813125894134478, + "grad_norm": 4.176635135536188, + "learning_rate": 0.00011089628342784814, + "loss": 1.9335, + "step": 5383 + }, + { + "epoch": 0.4814020028612303, + "grad_norm": 14.894400489461791, + "learning_rate": 0.00011086749529563143, + "loss": 6.9854, + "step": 5384 + }, + { + "epoch": 0.4814914163090129, + "grad_norm": 8.252943530052553, + "learning_rate": 0.00011083870625197103, + "loss": 1.9147, + "step": 5385 + }, + { + "epoch": 0.48158082975679545, + "grad_norm": 7.634559327513815, + "learning_rate": 0.00011080991629928143, + "loss": 1.7741, + "step": 5386 + }, + { + "epoch": 0.48167024320457796, + "grad_norm": 8.048282117983325, + "learning_rate": 0.00011078112543997723, + "loss": 2.0122, + "step": 5387 + }, + { + "epoch": 0.48175965665236054, + "grad_norm": 0.7354672058150301, + "learning_rate": 0.00011075233367647302, + "loss": 2.4422, + "step": 5388 + }, + { + "epoch": 0.48184907010014305, + "grad_norm": 6.549314427658946, + "learning_rate": 0.00011072354101118357, + "loss": 1.9761, + "step": 5389 + }, + { + "epoch": 0.4819384835479256, + "grad_norm": 4.656107108263266, + "learning_rate": 0.00011069474744652371, + "loss": 1.5903, + "step": 5390 + }, + { + "epoch": 0.48202789699570814, + "grad_norm": 5.121107566056689, + "learning_rate": 0.00011066595298490827, + "loss": 1.6577, + "step": 5391 + }, + { + "epoch": 0.4821173104434907, + "grad_norm": 3.2837206918851427, + "learning_rate": 0.00011063715762875225, + "loss": 1.7251, + "step": 5392 + }, + { + "epoch": 0.4822067238912732, + "grad_norm": 1.8383719503087332, + "learning_rate": 0.00011060836138047066, + "loss": 2.583, + "step": 5393 + }, + { + "epoch": 0.4822961373390558, + "grad_norm": 6.947240907206247, + "learning_rate": 0.00011057956424247861, + "loss": 1.7442, + "step": 5394 + }, + { + "epoch": 0.4823855507868383, + "grad_norm": 1.5721942754874463, + "learning_rate": 0.00011055076621719132, + "loss": 2.0261, + "step": 5395 + }, + { + "epoch": 0.4824749642346209, + "grad_norm": 4.312782801521387, + "learning_rate": 0.00011052196730702396, + "loss": 2.457, + "step": 5396 + }, + { + "epoch": 0.48256437768240346, + "grad_norm": 4.0076375541668545, + "learning_rate": 0.00011049316751439194, + "loss": 1.89, + "step": 5397 + }, + { + "epoch": 0.48265379113018597, + "grad_norm": 1.5089902207749728, + "learning_rate": 0.00011046436684171062, + "loss": 1.9684, + "step": 5398 + }, + { + "epoch": 0.48274320457796854, + "grad_norm": 4.316457682980723, + "learning_rate": 0.00011043556529139549, + "loss": 1.82, + "step": 5399 + }, + { + "epoch": 0.48283261802575106, + "grad_norm": 2.083168786407278, + "learning_rate": 0.00011040676286586211, + "loss": 2.1037, + "step": 5400 + }, + { + "epoch": 0.48292203147353363, + "grad_norm": 1.3190570600731246, + "learning_rate": 0.00011037795956752608, + "loss": 1.8255, + "step": 5401 + }, + { + "epoch": 0.48301144492131615, + "grad_norm": 5.104398829970234, + "learning_rate": 0.00011034915539880313, + "loss": 2.3214, + "step": 5402 + }, + { + "epoch": 0.4831008583690987, + "grad_norm": 1.3486850394675725, + "learning_rate": 0.00011032035036210901, + "loss": 1.7801, + "step": 5403 + }, + { + "epoch": 0.48319027181688123, + "grad_norm": 3.3774600929583185, + "learning_rate": 0.00011029154445985961, + "loss": 1.7744, + "step": 5404 + }, + { + "epoch": 0.4832796852646638, + "grad_norm": 4.497055380770766, + "learning_rate": 0.00011026273769447076, + "loss": 2.0631, + "step": 5405 + }, + { + "epoch": 0.4833690987124464, + "grad_norm": 1.7227386660595723, + "learning_rate": 0.00011023393006835847, + "loss": 1.7874, + "step": 5406 + }, + { + "epoch": 0.4834585121602289, + "grad_norm": 7.531496647387823, + "learning_rate": 0.00011020512158393887, + "loss": 1.6242, + "step": 5407 + }, + { + "epoch": 0.48354792560801146, + "grad_norm": 0.8138025867029777, + "learning_rate": 0.00011017631224362803, + "loss": 2.012, + "step": 5408 + }, + { + "epoch": 0.483637339055794, + "grad_norm": 3.977901017202713, + "learning_rate": 0.00011014750204984217, + "loss": 2.0047, + "step": 5409 + }, + { + "epoch": 0.48372675250357655, + "grad_norm": 3.3703507611475354, + "learning_rate": 0.00011011869100499758, + "loss": 1.8262, + "step": 5410 + }, + { + "epoch": 0.48381616595135907, + "grad_norm": 5.415735796569874, + "learning_rate": 0.00011008987911151058, + "loss": 1.5858, + "step": 5411 + }, + { + "epoch": 0.48390557939914164, + "grad_norm": 2.4843574492574483, + "learning_rate": 0.00011006106637179763, + "loss": 2.1602, + "step": 5412 + }, + { + "epoch": 0.48399499284692415, + "grad_norm": 1.6541111015259815, + "learning_rate": 0.00011003225278827515, + "loss": 1.8795, + "step": 5413 + }, + { + "epoch": 0.4840844062947067, + "grad_norm": 3.7241166526852867, + "learning_rate": 0.0001100034383633598, + "loss": 2.1434, + "step": 5414 + }, + { + "epoch": 0.4841738197424893, + "grad_norm": 3.240984174921102, + "learning_rate": 0.00010997462309946811, + "loss": 1.7142, + "step": 5415 + }, + { + "epoch": 0.4842632331902718, + "grad_norm": 1.8223092414795763, + "learning_rate": 0.00010994580699901684, + "loss": 2.1221, + "step": 5416 + }, + { + "epoch": 0.4843526466380544, + "grad_norm": 2.3436077223186977, + "learning_rate": 0.00010991699006442275, + "loss": 2.1018, + "step": 5417 + }, + { + "epoch": 0.4844420600858369, + "grad_norm": 6.605434359551668, + "learning_rate": 0.00010988817229810268, + "loss": 1.754, + "step": 5418 + }, + { + "epoch": 0.4845314735336195, + "grad_norm": 4.66333380288137, + "learning_rate": 0.00010985935370247355, + "loss": 1.9982, + "step": 5419 + }, + { + "epoch": 0.484620886981402, + "grad_norm": 4.95048338736035, + "learning_rate": 0.00010983053427995234, + "loss": 1.8023, + "step": 5420 + }, + { + "epoch": 0.48471030042918456, + "grad_norm": 1.0700072673042549, + "learning_rate": 0.0001098017140329561, + "loss": 1.6521, + "step": 5421 + }, + { + "epoch": 0.4847997138769671, + "grad_norm": 4.332377307514259, + "learning_rate": 0.0001097728929639019, + "loss": 2.0803, + "step": 5422 + }, + { + "epoch": 0.48488912732474965, + "grad_norm": 3.3396930074561197, + "learning_rate": 0.00010974407107520697, + "loss": 1.7945, + "step": 5423 + }, + { + "epoch": 0.48497854077253216, + "grad_norm": 1.3095789242709717, + "learning_rate": 0.0001097152483692886, + "loss": 2.197, + "step": 5424 + }, + { + "epoch": 0.48506795422031473, + "grad_norm": 3.8180137030652137, + "learning_rate": 0.00010968642484856406, + "loss": 2.1793, + "step": 5425 + }, + { + "epoch": 0.4851573676680973, + "grad_norm": 2.397941458335952, + "learning_rate": 0.0001096576005154508, + "loss": 1.8072, + "step": 5426 + }, + { + "epoch": 0.4852467811158798, + "grad_norm": 12.390410571401349, + "learning_rate": 0.0001096287753723662, + "loss": 7.198, + "step": 5427 + }, + { + "epoch": 0.4853361945636624, + "grad_norm": 15.354626692335254, + "learning_rate": 0.00010959994942172786, + "loss": 1.7738, + "step": 5428 + }, + { + "epoch": 0.4854256080114449, + "grad_norm": 6.714417520074206, + "learning_rate": 0.00010957112266595338, + "loss": 2.2162, + "step": 5429 + }, + { + "epoch": 0.4855150214592275, + "grad_norm": 4.010179152449721, + "learning_rate": 0.00010954229510746035, + "loss": 1.9802, + "step": 5430 + }, + { + "epoch": 0.48560443490701, + "grad_norm": 7.359086633644148, + "learning_rate": 0.0001095134667486666, + "loss": 1.9978, + "step": 5431 + }, + { + "epoch": 0.48569384835479257, + "grad_norm": 4.022662535562891, + "learning_rate": 0.00010948463759198986, + "loss": 1.6613, + "step": 5432 + }, + { + "epoch": 0.4857832618025751, + "grad_norm": 5.4820304894179825, + "learning_rate": 0.00010945580763984801, + "loss": 2.1603, + "step": 5433 + }, + { + "epoch": 0.48587267525035766, + "grad_norm": 4.815839889061017, + "learning_rate": 0.00010942697689465902, + "loss": 2.3517, + "step": 5434 + }, + { + "epoch": 0.4859620886981402, + "grad_norm": 3.169068721254345, + "learning_rate": 0.00010939814535884083, + "loss": 1.8932, + "step": 5435 + }, + { + "epoch": 0.48605150214592274, + "grad_norm": 2.4415572635227623, + "learning_rate": 0.00010936931303481158, + "loss": 2.1721, + "step": 5436 + }, + { + "epoch": 0.4861409155937053, + "grad_norm": 3.3511159100711065, + "learning_rate": 0.00010934047992498932, + "loss": 2.2735, + "step": 5437 + }, + { + "epoch": 0.48623032904148783, + "grad_norm": 3.7483206016484782, + "learning_rate": 0.00010931164603179231, + "loss": 2.581, + "step": 5438 + }, + { + "epoch": 0.4863197424892704, + "grad_norm": 2.494934508096795, + "learning_rate": 0.0001092828113576388, + "loss": 2.0966, + "step": 5439 + }, + { + "epoch": 0.4864091559370529, + "grad_norm": 0.977324978772993, + "learning_rate": 0.00010925397590494712, + "loss": 1.6601, + "step": 5440 + }, + { + "epoch": 0.4864985693848355, + "grad_norm": 3.574014677518061, + "learning_rate": 0.00010922513967613563, + "loss": 2.1341, + "step": 5441 + }, + { + "epoch": 0.486587982832618, + "grad_norm": 2.5489700950286296, + "learning_rate": 0.00010919630267362282, + "loss": 1.5046, + "step": 5442 + }, + { + "epoch": 0.4866773962804006, + "grad_norm": 1.563900270662488, + "learning_rate": 0.00010916746489982723, + "loss": 1.6933, + "step": 5443 + }, + { + "epoch": 0.4867668097281831, + "grad_norm": 0.8572285744739823, + "learning_rate": 0.00010913862635716741, + "loss": 1.9167, + "step": 5444 + }, + { + "epoch": 0.48685622317596566, + "grad_norm": 0.7373595929291378, + "learning_rate": 0.00010910978704806203, + "loss": 1.78, + "step": 5445 + }, + { + "epoch": 0.48694563662374823, + "grad_norm": 2.218578817596414, + "learning_rate": 0.00010908094697492983, + "loss": 1.9094, + "step": 5446 + }, + { + "epoch": 0.48703505007153075, + "grad_norm": 2.695431529613323, + "learning_rate": 0.00010905210614018957, + "loss": 2.0147, + "step": 5447 + }, + { + "epoch": 0.4871244635193133, + "grad_norm": 0.9394868622824452, + "learning_rate": 0.0001090232645462601, + "loss": 2.1105, + "step": 5448 + }, + { + "epoch": 0.48721387696709584, + "grad_norm": 0.7932909517977512, + "learning_rate": 0.00010899442219556033, + "loss": 1.7316, + "step": 5449 + }, + { + "epoch": 0.4873032904148784, + "grad_norm": 1.450333932883308, + "learning_rate": 0.00010896557909050927, + "loss": 1.8224, + "step": 5450 + }, + { + "epoch": 0.4873927038626609, + "grad_norm": 2.0371923563290126, + "learning_rate": 0.00010893673523352585, + "loss": 1.799, + "step": 5451 + }, + { + "epoch": 0.4874821173104435, + "grad_norm": 1.9038338000348751, + "learning_rate": 0.00010890789062702926, + "loss": 2.2809, + "step": 5452 + }, + { + "epoch": 0.487571530758226, + "grad_norm": 1.1538990737695147, + "learning_rate": 0.00010887904527343866, + "loss": 2.0104, + "step": 5453 + }, + { + "epoch": 0.4876609442060086, + "grad_norm": 0.8163226550760462, + "learning_rate": 0.00010885019917517325, + "loss": 1.8626, + "step": 5454 + }, + { + "epoch": 0.48775035765379116, + "grad_norm": 1.252663236773344, + "learning_rate": 0.00010882135233465232, + "loss": 1.9376, + "step": 5455 + }, + { + "epoch": 0.48783977110157367, + "grad_norm": 1.5146558132056778, + "learning_rate": 0.00010879250475429523, + "loss": 1.7669, + "step": 5456 + }, + { + "epoch": 0.48792918454935624, + "grad_norm": 2.209320742940823, + "learning_rate": 0.0001087636564365214, + "loss": 1.7679, + "step": 5457 + }, + { + "epoch": 0.48801859799713876, + "grad_norm": 2.0561116246548847, + "learning_rate": 0.00010873480738375024, + "loss": 2.0346, + "step": 5458 + }, + { + "epoch": 0.48810801144492133, + "grad_norm": 2.5601405191000506, + "learning_rate": 0.00010870595759840137, + "loss": 1.9266, + "step": 5459 + }, + { + "epoch": 0.48819742489270385, + "grad_norm": 9.497603051973044, + "learning_rate": 0.00010867710708289434, + "loss": 6.7391, + "step": 5460 + }, + { + "epoch": 0.4882868383404864, + "grad_norm": 2.9696316388937376, + "learning_rate": 0.00010864825583964882, + "loss": 1.8793, + "step": 5461 + }, + { + "epoch": 0.48837625178826893, + "grad_norm": 2.163168982004279, + "learning_rate": 0.00010861940387108451, + "loss": 1.4759, + "step": 5462 + }, + { + "epoch": 0.4884656652360515, + "grad_norm": 3.165185020449424, + "learning_rate": 0.00010859055117962125, + "loss": 1.7591, + "step": 5463 + }, + { + "epoch": 0.488555078683834, + "grad_norm": 1.848498137547094, + "learning_rate": 0.00010856169776767882, + "loss": 1.8792, + "step": 5464 + }, + { + "epoch": 0.4886444921316166, + "grad_norm": 3.368117995158969, + "learning_rate": 0.0001085328436376771, + "loss": 1.7046, + "step": 5465 + }, + { + "epoch": 0.48873390557939916, + "grad_norm": 3.6714910150527365, + "learning_rate": 0.00010850398879203611, + "loss": 1.8953, + "step": 5466 + }, + { + "epoch": 0.4888233190271817, + "grad_norm": 2.366545137335965, + "learning_rate": 0.00010847513323317588, + "loss": 2.1318, + "step": 5467 + }, + { + "epoch": 0.48891273247496425, + "grad_norm": 2.0047682123462316, + "learning_rate": 0.00010844627696351644, + "loss": 1.7975, + "step": 5468 + }, + { + "epoch": 0.48900214592274677, + "grad_norm": 1.1959567516178535, + "learning_rate": 0.00010841741998547794, + "loss": 1.64, + "step": 5469 + }, + { + "epoch": 0.48909155937052934, + "grad_norm": 2.9585460032857878, + "learning_rate": 0.00010838856230148063, + "loss": 2.1631, + "step": 5470 + }, + { + "epoch": 0.48918097281831185, + "grad_norm": 3.0644379347920188, + "learning_rate": 0.0001083597039139447, + "loss": 1.8279, + "step": 5471 + }, + { + "epoch": 0.4892703862660944, + "grad_norm": 0.8479856593072697, + "learning_rate": 0.00010833084482529048, + "loss": 1.9414, + "step": 5472 + }, + { + "epoch": 0.48935979971387694, + "grad_norm": 1.6558596920095492, + "learning_rate": 0.0001083019850379384, + "loss": 2.2562, + "step": 5473 + }, + { + "epoch": 0.4894492131616595, + "grad_norm": 3.670519319505249, + "learning_rate": 0.00010827312455430884, + "loss": 2.25, + "step": 5474 + }, + { + "epoch": 0.4895386266094421, + "grad_norm": 1.5647936826696667, + "learning_rate": 0.00010824426337682235, + "loss": 1.7419, + "step": 5475 + }, + { + "epoch": 0.4896280400572246, + "grad_norm": 1.9853328686872649, + "learning_rate": 0.00010821540150789939, + "loss": 1.6531, + "step": 5476 + }, + { + "epoch": 0.48971745350500717, + "grad_norm": 0.9801826351104084, + "learning_rate": 0.00010818653894996067, + "loss": 1.9211, + "step": 5477 + }, + { + "epoch": 0.4898068669527897, + "grad_norm": 1.010335118616705, + "learning_rate": 0.00010815767570542681, + "loss": 1.8122, + "step": 5478 + }, + { + "epoch": 0.48989628040057226, + "grad_norm": 1.5870053283289278, + "learning_rate": 0.00010812881177671852, + "loss": 1.5394, + "step": 5479 + }, + { + "epoch": 0.4899856938483548, + "grad_norm": 1.4278557470351556, + "learning_rate": 0.00010809994716625662, + "loss": 1.7565, + "step": 5480 + }, + { + "epoch": 0.49007510729613735, + "grad_norm": 1.1073579888161738, + "learning_rate": 0.00010807108187646195, + "loss": 2.1213, + "step": 5481 + }, + { + "epoch": 0.49016452074391986, + "grad_norm": 0.8610970655576256, + "learning_rate": 0.00010804221590975535, + "loss": 1.8622, + "step": 5482 + }, + { + "epoch": 0.49025393419170243, + "grad_norm": 1.3888835978040432, + "learning_rate": 0.00010801334926855784, + "loss": 1.9003, + "step": 5483 + }, + { + "epoch": 0.490343347639485, + "grad_norm": 1.096630218418679, + "learning_rate": 0.0001079844819552904, + "loss": 1.936, + "step": 5484 + }, + { + "epoch": 0.4904327610872675, + "grad_norm": 1.3693607455009114, + "learning_rate": 0.0001079556139723741, + "loss": 1.6554, + "step": 5485 + }, + { + "epoch": 0.4905221745350501, + "grad_norm": 1.426786419929387, + "learning_rate": 0.00010792674532223006, + "loss": 1.7985, + "step": 5486 + }, + { + "epoch": 0.4906115879828326, + "grad_norm": 0.8656876740131367, + "learning_rate": 0.00010789787600727948, + "loss": 2.2263, + "step": 5487 + }, + { + "epoch": 0.4907010014306152, + "grad_norm": 0.7884039748253975, + "learning_rate": 0.00010786900602994359, + "loss": 1.9117, + "step": 5488 + }, + { + "epoch": 0.4907904148783977, + "grad_norm": 2.5526072404725952, + "learning_rate": 0.00010784013539264359, + "loss": 1.8003, + "step": 5489 + }, + { + "epoch": 0.49087982832618027, + "grad_norm": 11.852432022086978, + "learning_rate": 0.00010781126409780098, + "loss": 7.3178, + "step": 5490 + }, + { + "epoch": 0.4909692417739628, + "grad_norm": 1.303577863931744, + "learning_rate": 0.00010778239214783708, + "loss": 1.8595, + "step": 5491 + }, + { + "epoch": 0.49105865522174535, + "grad_norm": 2.3888529402587917, + "learning_rate": 0.00010775351954517332, + "loss": 1.907, + "step": 5492 + }, + { + "epoch": 0.49114806866952787, + "grad_norm": 4.283701107219301, + "learning_rate": 0.00010772464629223124, + "loss": 1.803, + "step": 5493 + }, + { + "epoch": 0.49123748211731044, + "grad_norm": 2.330782653006837, + "learning_rate": 0.00010769577239143242, + "loss": 2.0126, + "step": 5494 + }, + { + "epoch": 0.491326895565093, + "grad_norm": 1.5281932426344564, + "learning_rate": 0.00010766689784519845, + "loss": 2.0909, + "step": 5495 + }, + { + "epoch": 0.49141630901287553, + "grad_norm": 5.758858969691826, + "learning_rate": 0.00010763802265595102, + "loss": 1.7872, + "step": 5496 + }, + { + "epoch": 0.4915057224606581, + "grad_norm": 1.623490940007833, + "learning_rate": 0.00010760914682611188, + "loss": 1.6459, + "step": 5497 + }, + { + "epoch": 0.4915951359084406, + "grad_norm": 9.121066986048872, + "learning_rate": 0.00010758027035810276, + "loss": 1.9867, + "step": 5498 + }, + { + "epoch": 0.4916845493562232, + "grad_norm": 0.9952753026702729, + "learning_rate": 0.00010755139325434548, + "loss": 1.8624, + "step": 5499 + }, + { + "epoch": 0.4917739628040057, + "grad_norm": 2.8001790170223217, + "learning_rate": 0.00010752251551726205, + "loss": 2.1869, + "step": 5500 + }, + { + "epoch": 0.4918633762517883, + "grad_norm": 6.3281232322725725, + "learning_rate": 0.0001074936371492743, + "loss": 1.6605, + "step": 5501 + }, + { + "epoch": 0.4919527896995708, + "grad_norm": 4.768742099914598, + "learning_rate": 0.00010746475815280424, + "loss": 2.1061, + "step": 5502 + }, + { + "epoch": 0.49204220314735336, + "grad_norm": 2.2136202354200285, + "learning_rate": 0.00010743587853027391, + "loss": 2.0286, + "step": 5503 + }, + { + "epoch": 0.49213161659513593, + "grad_norm": 1.223599302778388, + "learning_rate": 0.00010740699828410545, + "loss": 1.8334, + "step": 5504 + }, + { + "epoch": 0.49222103004291845, + "grad_norm": 2.310163976905821, + "learning_rate": 0.00010737811741672101, + "loss": 2.2532, + "step": 5505 + }, + { + "epoch": 0.492310443490701, + "grad_norm": 2.588041785089314, + "learning_rate": 0.00010734923593054271, + "loss": 1.698, + "step": 5506 + }, + { + "epoch": 0.49239985693848354, + "grad_norm": 2.2029319476975013, + "learning_rate": 0.00010732035382799293, + "loss": 1.7333, + "step": 5507 + }, + { + "epoch": 0.4924892703862661, + "grad_norm": 2.8641208186558016, + "learning_rate": 0.00010729147111149392, + "loss": 1.6413, + "step": 5508 + }, + { + "epoch": 0.4925786838340486, + "grad_norm": 1.5400171226910835, + "learning_rate": 0.00010726258778346798, + "loss": 1.8557, + "step": 5509 + }, + { + "epoch": 0.4926680972818312, + "grad_norm": 1.5858567378661326, + "learning_rate": 0.0001072337038463376, + "loss": 1.8916, + "step": 5510 + }, + { + "epoch": 0.4927575107296137, + "grad_norm": 1.3825070946835503, + "learning_rate": 0.00010720481930252524, + "loss": 2.0072, + "step": 5511 + }, + { + "epoch": 0.4928469241773963, + "grad_norm": 0.6207231284686485, + "learning_rate": 0.00010717593415445335, + "loss": 1.7636, + "step": 5512 + }, + { + "epoch": 0.4929363376251788, + "grad_norm": 0.4351794992596643, + "learning_rate": 0.00010714704840454453, + "loss": 1.7684, + "step": 5513 + }, + { + "epoch": 0.49302575107296137, + "grad_norm": 6.085327662376776, + "learning_rate": 0.0001071181620552214, + "loss": 6.9862, + "step": 5514 + }, + { + "epoch": 0.49311516452074394, + "grad_norm": 1.3259597338113334, + "learning_rate": 0.00010708927510890665, + "loss": 1.3875, + "step": 5515 + }, + { + "epoch": 0.49320457796852646, + "grad_norm": 0.7193270747266911, + "learning_rate": 0.0001070603875680229, + "loss": 6.9304, + "step": 5516 + }, + { + "epoch": 0.49329399141630903, + "grad_norm": 1.180255102098295, + "learning_rate": 0.000107031499434993, + "loss": 2.03, + "step": 5517 + }, + { + "epoch": 0.49338340486409155, + "grad_norm": 0.9285001684477564, + "learning_rate": 0.00010700261071223973, + "loss": 1.6498, + "step": 5518 + }, + { + "epoch": 0.4934728183118741, + "grad_norm": 0.9610406081470222, + "learning_rate": 0.00010697372140218596, + "loss": 1.8863, + "step": 5519 + }, + { + "epoch": 0.49356223175965663, + "grad_norm": 0.8111128125497086, + "learning_rate": 0.00010694483150725458, + "loss": 1.7422, + "step": 5520 + }, + { + "epoch": 0.4936516452074392, + "grad_norm": 1.0690584427074847, + "learning_rate": 0.00010691594102986861, + "loss": 2.042, + "step": 5521 + }, + { + "epoch": 0.4937410586552217, + "grad_norm": 0.997312759203772, + "learning_rate": 0.000106887049972451, + "loss": 1.8018, + "step": 5522 + }, + { + "epoch": 0.4938304721030043, + "grad_norm": 0.8358198920513285, + "learning_rate": 0.00010685815833742481, + "loss": 2.1162, + "step": 5523 + }, + { + "epoch": 0.49391988555078686, + "grad_norm": 1.0297057974813106, + "learning_rate": 0.00010682926612721315, + "loss": 2.1579, + "step": 5524 + }, + { + "epoch": 0.4940092989985694, + "grad_norm": 0.9900070568314004, + "learning_rate": 0.00010680037334423925, + "loss": 2.1282, + "step": 5525 + }, + { + "epoch": 0.49409871244635195, + "grad_norm": 0.7619876823569968, + "learning_rate": 0.00010677147999092618, + "loss": 2.1667, + "step": 5526 + }, + { + "epoch": 0.49418812589413447, + "grad_norm": 0.8207105291533787, + "learning_rate": 0.00010674258606969729, + "loss": 2.0453, + "step": 5527 + }, + { + "epoch": 0.49427753934191704, + "grad_norm": 0.9288865572821203, + "learning_rate": 0.00010671369158297586, + "loss": 1.7375, + "step": 5528 + }, + { + "epoch": 0.49436695278969955, + "grad_norm": 0.816656555246648, + "learning_rate": 0.00010668479653318522, + "loss": 1.5631, + "step": 5529 + }, + { + "epoch": 0.4944563662374821, + "grad_norm": 0.9037218506806238, + "learning_rate": 0.00010665590092274876, + "loss": 1.8204, + "step": 5530 + }, + { + "epoch": 0.49454577968526464, + "grad_norm": 1.3426075908719053, + "learning_rate": 0.00010662700475408994, + "loss": 2.1723, + "step": 5531 + }, + { + "epoch": 0.4946351931330472, + "grad_norm": 2.3918386113925334, + "learning_rate": 0.00010659810802963224, + "loss": 1.7442, + "step": 5532 + }, + { + "epoch": 0.4947246065808298, + "grad_norm": 0.7602622367585514, + "learning_rate": 0.00010656921075179915, + "loss": 2.0205, + "step": 5533 + }, + { + "epoch": 0.4948140200286123, + "grad_norm": 0.6868005275813969, + "learning_rate": 0.00010654031292301432, + "loss": 1.9969, + "step": 5534 + }, + { + "epoch": 0.49490343347639487, + "grad_norm": 1.525827471709873, + "learning_rate": 0.00010651141454570135, + "loss": 1.9299, + "step": 5535 + }, + { + "epoch": 0.4949928469241774, + "grad_norm": 0.6082864520747148, + "learning_rate": 0.00010648251562228386, + "loss": 1.319, + "step": 5536 + }, + { + "epoch": 0.49508226037195996, + "grad_norm": 0.8557253503189552, + "learning_rate": 0.00010645361615518565, + "loss": 2.0331, + "step": 5537 + }, + { + "epoch": 0.4951716738197425, + "grad_norm": 0.39335325366377877, + "learning_rate": 0.00010642471614683045, + "loss": 1.9445, + "step": 5538 + }, + { + "epoch": 0.49526108726752505, + "grad_norm": 1.1644250352631738, + "learning_rate": 0.00010639581559964205, + "loss": 2.0199, + "step": 5539 + }, + { + "epoch": 0.49535050071530756, + "grad_norm": 0.5286324355176738, + "learning_rate": 0.00010636691451604434, + "loss": 1.3596, + "step": 5540 + }, + { + "epoch": 0.49543991416309013, + "grad_norm": 0.5655803057473382, + "learning_rate": 0.00010633801289846119, + "loss": 1.9025, + "step": 5541 + }, + { + "epoch": 0.49552932761087265, + "grad_norm": 0.4922157474207836, + "learning_rate": 0.00010630911074931655, + "loss": 1.5222, + "step": 5542 + }, + { + "epoch": 0.4956187410586552, + "grad_norm": 3.599419313115816, + "learning_rate": 0.00010628020807103441, + "loss": 6.7577, + "step": 5543 + }, + { + "epoch": 0.4957081545064378, + "grad_norm": 1.3500398028961675, + "learning_rate": 0.00010625130486603878, + "loss": 2.3878, + "step": 5544 + }, + { + "epoch": 0.4957975679542203, + "grad_norm": 2.3491588338657934, + "learning_rate": 0.00010622240113675382, + "loss": 2.2, + "step": 5545 + }, + { + "epoch": 0.4958869814020029, + "grad_norm": 0.9789569411466288, + "learning_rate": 0.00010619349688560354, + "loss": 2.0011, + "step": 5546 + }, + { + "epoch": 0.4959763948497854, + "grad_norm": 1.1032121391500413, + "learning_rate": 0.00010616459211501217, + "loss": 2.0932, + "step": 5547 + }, + { + "epoch": 0.49606580829756797, + "grad_norm": 2.035897655057854, + "learning_rate": 0.00010613568682740391, + "loss": 1.8644, + "step": 5548 + }, + { + "epoch": 0.4961552217453505, + "grad_norm": 1.2474743448270793, + "learning_rate": 0.00010610678102520301, + "loss": 1.9536, + "step": 5549 + }, + { + "epoch": 0.49624463519313305, + "grad_norm": 1.3032879072967947, + "learning_rate": 0.00010607787471083375, + "loss": 1.7565, + "step": 5550 + }, + { + "epoch": 0.49633404864091557, + "grad_norm": 1.9115697719627653, + "learning_rate": 0.00010604896788672048, + "loss": 2.4799, + "step": 5551 + }, + { + "epoch": 0.49642346208869814, + "grad_norm": 0.8383931549691321, + "learning_rate": 0.0001060200605552876, + "loss": 2.0526, + "step": 5552 + }, + { + "epoch": 0.4965128755364807, + "grad_norm": 1.2761806908770954, + "learning_rate": 0.00010599115271895948, + "loss": 1.7818, + "step": 5553 + }, + { + "epoch": 0.49660228898426323, + "grad_norm": 1.5526175392265957, + "learning_rate": 0.00010596224438016063, + "loss": 2.0109, + "step": 5554 + }, + { + "epoch": 0.4966917024320458, + "grad_norm": 1.2820010759387865, + "learning_rate": 0.00010593333554131552, + "loss": 1.8815, + "step": 5555 + }, + { + "epoch": 0.4967811158798283, + "grad_norm": 0.7250058855203563, + "learning_rate": 0.00010590442620484875, + "loss": 1.5544, + "step": 5556 + }, + { + "epoch": 0.4968705293276109, + "grad_norm": 2.2784879920608563, + "learning_rate": 0.00010587551637318489, + "loss": 2.3132, + "step": 5557 + }, + { + "epoch": 0.4969599427753934, + "grad_norm": 1.1823086205886615, + "learning_rate": 0.00010584660604874857, + "loss": 1.5622, + "step": 5558 + }, + { + "epoch": 0.497049356223176, + "grad_norm": 1.8220597798200155, + "learning_rate": 0.00010581769523396445, + "loss": 2.0326, + "step": 5559 + }, + { + "epoch": 0.4971387696709585, + "grad_norm": 0.9119330908205191, + "learning_rate": 0.00010578878393125724, + "loss": 2.0855, + "step": 5560 + }, + { + "epoch": 0.49722818311874106, + "grad_norm": 1.786460004857075, + "learning_rate": 0.00010575987214305174, + "loss": 2.0789, + "step": 5561 + }, + { + "epoch": 0.4973175965665236, + "grad_norm": 1.2094656082630124, + "learning_rate": 0.0001057309598717727, + "loss": 1.7714, + "step": 5562 + }, + { + "epoch": 0.49740701001430615, + "grad_norm": 1.7100785329514046, + "learning_rate": 0.000105702047119845, + "loss": 2.0226, + "step": 5563 + }, + { + "epoch": 0.4974964234620887, + "grad_norm": 1.424760256999157, + "learning_rate": 0.00010567313388969348, + "loss": 1.9145, + "step": 5564 + }, + { + "epoch": 0.49758583690987124, + "grad_norm": 0.6663378239102447, + "learning_rate": 0.00010564422018374307, + "loss": 2.1035, + "step": 5565 + }, + { + "epoch": 0.4976752503576538, + "grad_norm": 1.7870017754016416, + "learning_rate": 0.00010561530600441873, + "loss": 2.2511, + "step": 5566 + }, + { + "epoch": 0.4977646638054363, + "grad_norm": 3.6590298732400313, + "learning_rate": 0.00010558639135414545, + "loss": 1.7289, + "step": 5567 + }, + { + "epoch": 0.4978540772532189, + "grad_norm": 1.5384580776076155, + "learning_rate": 0.00010555747623534831, + "loss": 1.6176, + "step": 5568 + }, + { + "epoch": 0.4979434907010014, + "grad_norm": 5.784396061800231, + "learning_rate": 0.00010552856065045232, + "loss": 6.9572, + "step": 5569 + }, + { + "epoch": 0.498032904148784, + "grad_norm": 1.7395573300473417, + "learning_rate": 0.00010549964460188261, + "loss": 1.8805, + "step": 5570 + }, + { + "epoch": 0.4981223175965665, + "grad_norm": 1.8315043435042697, + "learning_rate": 0.00010547072809206437, + "loss": 1.7564, + "step": 5571 + }, + { + "epoch": 0.49821173104434907, + "grad_norm": 2.2596594025406955, + "learning_rate": 0.00010544181112342278, + "loss": 1.9739, + "step": 5572 + }, + { + "epoch": 0.49830114449213164, + "grad_norm": 2.580925754789769, + "learning_rate": 0.00010541289369838302, + "loss": 1.9099, + "step": 5573 + }, + { + "epoch": 0.49839055793991416, + "grad_norm": 0.6696781837511987, + "learning_rate": 0.00010538397581937048, + "loss": 1.8123, + "step": 5574 + }, + { + "epoch": 0.49847997138769673, + "grad_norm": 1.270269511272512, + "learning_rate": 0.00010535505748881031, + "loss": 1.9359, + "step": 5575 + }, + { + "epoch": 0.49856938483547925, + "grad_norm": 2.2986421342785133, + "learning_rate": 0.00010532613870912799, + "loss": 2.2136, + "step": 5576 + }, + { + "epoch": 0.4986587982832618, + "grad_norm": 1.0383460097821517, + "learning_rate": 0.00010529721948274882, + "loss": 1.7452, + "step": 5577 + }, + { + "epoch": 0.49874821173104433, + "grad_norm": 1.6410146823531366, + "learning_rate": 0.00010526829981209827, + "loss": 2.053, + "step": 5578 + }, + { + "epoch": 0.4988376251788269, + "grad_norm": 2.0174270081338967, + "learning_rate": 0.00010523937969960176, + "loss": 1.6133, + "step": 5579 + }, + { + "epoch": 0.4989270386266094, + "grad_norm": 2.142042939266398, + "learning_rate": 0.00010521045914768482, + "loss": 2.107, + "step": 5580 + }, + { + "epoch": 0.499016452074392, + "grad_norm": 6.533593192754675, + "learning_rate": 0.00010518153815877294, + "loss": 6.9893, + "step": 5581 + }, + { + "epoch": 0.4991058655221745, + "grad_norm": 3.8779409875369555, + "learning_rate": 0.00010515261673529173, + "loss": 2.0367, + "step": 5582 + }, + { + "epoch": 0.4991952789699571, + "grad_norm": 3.794206333943223, + "learning_rate": 0.00010512369487966678, + "loss": 1.9548, + "step": 5583 + }, + { + "epoch": 0.49928469241773965, + "grad_norm": 1.649797917558215, + "learning_rate": 0.00010509477259432372, + "loss": 1.6511, + "step": 5584 + }, + { + "epoch": 0.49937410586552217, + "grad_norm": 0.7805296447465824, + "learning_rate": 0.00010506584988168824, + "loss": 2.1267, + "step": 5585 + }, + { + "epoch": 0.49946351931330474, + "grad_norm": 2.1765854696119287, + "learning_rate": 0.00010503692674418603, + "loss": 1.707, + "step": 5586 + }, + { + "epoch": 0.49955293276108725, + "grad_norm": 1.828488628264457, + "learning_rate": 0.00010500800318424286, + "loss": 1.9213, + "step": 5587 + }, + { + "epoch": 0.4996423462088698, + "grad_norm": 1.3428971911015095, + "learning_rate": 0.00010497907920428454, + "loss": 1.5059, + "step": 5588 + }, + { + "epoch": 0.49973175965665234, + "grad_norm": 1.7045842186434503, + "learning_rate": 0.00010495015480673685, + "loss": 2.1235, + "step": 5589 + }, + { + "epoch": 0.4998211731044349, + "grad_norm": 1.7678134758024975, + "learning_rate": 0.00010492122999402562, + "loss": 1.4889, + "step": 5590 + }, + { + "epoch": 0.49991058655221743, + "grad_norm": 1.178673762659057, + "learning_rate": 0.00010489230476857681, + "loss": 1.4649, + "step": 5591 + }, + { + "epoch": 0.5, + "grad_norm": 1.9815898196381165, + "learning_rate": 0.00010486337913281632, + "loss": 2.1201, + "step": 5592 + }, + { + "epoch": 0.5000894134477826, + "grad_norm": 1.6972501953150605, + "learning_rate": 0.00010483445308917006, + "loss": 2.221, + "step": 5593 + }, + { + "epoch": 0.5001788268955651, + "grad_norm": 0.8112761142908832, + "learning_rate": 0.00010480552664006406, + "loss": 1.7607, + "step": 5594 + }, + { + "epoch": 0.5002682403433476, + "grad_norm": 0.4126768093931635, + "learning_rate": 0.00010477659978792438, + "loss": 1.6292, + "step": 5595 + }, + { + "epoch": 0.5003576537911302, + "grad_norm": 0.661093400368448, + "learning_rate": 0.00010474767253517701, + "loss": 1.6898, + "step": 5596 + }, + { + "epoch": 0.5004470672389127, + "grad_norm": 0.835824021894167, + "learning_rate": 0.00010471874488424808, + "loss": 1.8566, + "step": 5597 + }, + { + "epoch": 0.5005364806866953, + "grad_norm": 1.0700428357157385, + "learning_rate": 0.00010468981683756373, + "loss": 2.0461, + "step": 5598 + }, + { + "epoch": 0.5006258941344778, + "grad_norm": 0.8967023160926806, + "learning_rate": 0.00010466088839755012, + "loss": 1.5678, + "step": 5599 + }, + { + "epoch": 0.5007153075822603, + "grad_norm": 1.2499867342006883, + "learning_rate": 0.00010463195956663338, + "loss": 1.6786, + "step": 5600 + }, + { + "epoch": 0.5008047210300429, + "grad_norm": 2.0094726074354616, + "learning_rate": 0.00010460303034723985, + "loss": 2.2179, + "step": 5601 + }, + { + "epoch": 0.5008941344778255, + "grad_norm": 1.8566149818431132, + "learning_rate": 0.00010457410074179568, + "loss": 1.7183, + "step": 5602 + }, + { + "epoch": 0.5009835479256081, + "grad_norm": 2.085632605101912, + "learning_rate": 0.00010454517075272721, + "loss": 6.7774, + "step": 5603 + }, + { + "epoch": 0.5010729613733905, + "grad_norm": 3.5008373699174196, + "learning_rate": 0.00010451624038246075, + "loss": 2.0406, + "step": 5604 + }, + { + "epoch": 0.5011623748211731, + "grad_norm": 2.2340543313425782, + "learning_rate": 0.00010448730963342268, + "loss": 2.0194, + "step": 5605 + }, + { + "epoch": 0.5012517882689557, + "grad_norm": 2.0076933708678406, + "learning_rate": 0.00010445837850803939, + "loss": 2.1104, + "step": 5606 + }, + { + "epoch": 0.5013412017167382, + "grad_norm": 1.981928335014811, + "learning_rate": 0.00010442944700873722, + "loss": 1.9878, + "step": 5607 + }, + { + "epoch": 0.5014306151645207, + "grad_norm": 1.3097928584641925, + "learning_rate": 0.00010440051513794271, + "loss": 1.8142, + "step": 5608 + }, + { + "epoch": 0.5015200286123033, + "grad_norm": 0.6412125395280187, + "learning_rate": 0.00010437158289808233, + "loss": 2.0117, + "step": 5609 + }, + { + "epoch": 0.5016094420600858, + "grad_norm": 1.3042859630440962, + "learning_rate": 0.00010434265029158254, + "loss": 2.1426, + "step": 5610 + }, + { + "epoch": 0.5016988555078684, + "grad_norm": 1.619679761412519, + "learning_rate": 0.00010431371732086994, + "loss": 2.1199, + "step": 5611 + }, + { + "epoch": 0.501788268955651, + "grad_norm": 0.9844184107674274, + "learning_rate": 0.00010428478398837107, + "loss": 2.048, + "step": 5612 + }, + { + "epoch": 0.5018776824034334, + "grad_norm": 1.4674435911952841, + "learning_rate": 0.00010425585029651252, + "loss": 2.1356, + "step": 5613 + }, + { + "epoch": 0.501967095851216, + "grad_norm": 1.1640688194316136, + "learning_rate": 0.00010422691624772097, + "loss": 2.2699, + "step": 5614 + }, + { + "epoch": 0.5020565092989986, + "grad_norm": 0.8691918598546265, + "learning_rate": 0.00010419798184442307, + "loss": 1.5912, + "step": 5615 + }, + { + "epoch": 0.5021459227467812, + "grad_norm": 0.6990437206986071, + "learning_rate": 0.00010416904708904548, + "loss": 1.5359, + "step": 5616 + }, + { + "epoch": 0.5022353361945636, + "grad_norm": 0.7000119950635895, + "learning_rate": 0.00010414011198401492, + "loss": 1.8335, + "step": 5617 + }, + { + "epoch": 0.5023247496423462, + "grad_norm": 0.7136339054360559, + "learning_rate": 0.00010411117653175821, + "loss": 1.6713, + "step": 5618 + }, + { + "epoch": 0.5024141630901288, + "grad_norm": 0.9140101550310901, + "learning_rate": 0.0001040822407347021, + "loss": 2.1284, + "step": 5619 + }, + { + "epoch": 0.5025035765379113, + "grad_norm": 0.9091173364859143, + "learning_rate": 0.00010405330459527336, + "loss": 1.9879, + "step": 5620 + }, + { + "epoch": 0.5025929899856938, + "grad_norm": 1.1007496755507358, + "learning_rate": 0.00010402436811589887, + "loss": 2.4353, + "step": 5621 + }, + { + "epoch": 0.5026824034334764, + "grad_norm": 0.6157968690176758, + "learning_rate": 0.00010399543129900549, + "loss": 2.3099, + "step": 5622 + }, + { + "epoch": 0.5027718168812589, + "grad_norm": 0.6766611622395701, + "learning_rate": 0.00010396649414702011, + "loss": 1.8639, + "step": 5623 + }, + { + "epoch": 0.5028612303290415, + "grad_norm": 0.6786821700690604, + "learning_rate": 0.00010393755666236962, + "loss": 1.8172, + "step": 5624 + }, + { + "epoch": 0.5029506437768241, + "grad_norm": 0.8793421660131727, + "learning_rate": 0.00010390861884748107, + "loss": 2.2509, + "step": 5625 + }, + { + "epoch": 0.5030400572246065, + "grad_norm": 0.9324999870836295, + "learning_rate": 0.00010387968070478136, + "loss": 1.8136, + "step": 5626 + }, + { + "epoch": 0.5031294706723891, + "grad_norm": 0.7658916268770035, + "learning_rate": 0.00010385074223669748, + "loss": 1.771, + "step": 5627 + }, + { + "epoch": 0.5032188841201717, + "grad_norm": 1.245020310202632, + "learning_rate": 0.0001038218034456565, + "loss": 2.0126, + "step": 5628 + }, + { + "epoch": 0.5033082975679543, + "grad_norm": 0.8749939641963846, + "learning_rate": 0.00010379286433408553, + "loss": 2.1839, + "step": 5629 + }, + { + "epoch": 0.5033977110157367, + "grad_norm": 1.0843870890962495, + "learning_rate": 0.00010376392490441158, + "loss": 2.2176, + "step": 5630 + }, + { + "epoch": 0.5034871244635193, + "grad_norm": 1.3240610665686332, + "learning_rate": 0.00010373498515906177, + "loss": 2.0368, + "step": 5631 + }, + { + "epoch": 0.5035765379113019, + "grad_norm": 1.3683131310584031, + "learning_rate": 0.00010370604510046331, + "loss": 1.746, + "step": 5632 + }, + { + "epoch": 0.5036659513590844, + "grad_norm": 1.4719124318799284, + "learning_rate": 0.00010367710473104331, + "loss": 2.1453, + "step": 5633 + }, + { + "epoch": 0.503755364806867, + "grad_norm": 0.5112501034234005, + "learning_rate": 0.00010364816405322895, + "loss": 1.8281, + "step": 5634 + }, + { + "epoch": 0.5038447782546495, + "grad_norm": 1.1165481965813076, + "learning_rate": 0.00010361922306944751, + "loss": 1.4002, + "step": 5635 + }, + { + "epoch": 0.503934191702432, + "grad_norm": 1.3002788353911976, + "learning_rate": 0.0001035902817821262, + "loss": 1.7409, + "step": 5636 + }, + { + "epoch": 0.5040236051502146, + "grad_norm": 0.571537634042895, + "learning_rate": 0.00010356134019369227, + "loss": 1.6388, + "step": 5637 + }, + { + "epoch": 0.5041130185979972, + "grad_norm": 0.7913025676325968, + "learning_rate": 0.00010353239830657304, + "loss": 1.4495, + "step": 5638 + }, + { + "epoch": 0.5042024320457796, + "grad_norm": 2.755223732408897, + "learning_rate": 0.00010350345612319586, + "loss": 2.3315, + "step": 5639 + }, + { + "epoch": 0.5042918454935622, + "grad_norm": 0.6994578756828366, + "learning_rate": 0.00010347451364598804, + "loss": 2.3189, + "step": 5640 + }, + { + "epoch": 0.5043812589413448, + "grad_norm": 0.7666179059188953, + "learning_rate": 0.00010344557087737692, + "loss": 1.3674, + "step": 5641 + }, + { + "epoch": 0.5044706723891274, + "grad_norm": 1.5843668893880882, + "learning_rate": 0.00010341662781978996, + "loss": 2.1971, + "step": 5642 + }, + { + "epoch": 0.5045600858369099, + "grad_norm": 1.3894643602864298, + "learning_rate": 0.00010338768447565457, + "loss": 1.8427, + "step": 5643 + }, + { + "epoch": 0.5046494992846924, + "grad_norm": 1.2284957299892967, + "learning_rate": 0.00010335874084739814, + "loss": 1.5002, + "step": 5644 + }, + { + "epoch": 0.504738912732475, + "grad_norm": 6.077694031627705, + "learning_rate": 0.00010332979693744815, + "loss": 6.7154, + "step": 5645 + }, + { + "epoch": 0.5048283261802575, + "grad_norm": 0.9947127166091219, + "learning_rate": 0.00010330085274823218, + "loss": 1.913, + "step": 5646 + }, + { + "epoch": 0.5049177396280401, + "grad_norm": 1.9863852063217653, + "learning_rate": 0.00010327190828217763, + "loss": 2.1612, + "step": 5647 + }, + { + "epoch": 0.5050071530758226, + "grad_norm": 2.2261644832226892, + "learning_rate": 0.00010324296354171207, + "loss": 2.0225, + "step": 5648 + }, + { + "epoch": 0.5050965665236051, + "grad_norm": 1.2072260601287288, + "learning_rate": 0.00010321401852926312, + "loss": 2.0011, + "step": 5649 + }, + { + "epoch": 0.5051859799713877, + "grad_norm": 0.4655453444098918, + "learning_rate": 0.0001031850732472583, + "loss": 2.2987, + "step": 5650 + }, + { + "epoch": 0.5052753934191703, + "grad_norm": 1.9323513410267834, + "learning_rate": 0.00010315612769812524, + "loss": 1.6154, + "step": 5651 + }, + { + "epoch": 0.5053648068669528, + "grad_norm": 1.5990997118025092, + "learning_rate": 0.00010312718188429154, + "loss": 1.7992, + "step": 5652 + }, + { + "epoch": 0.5054542203147353, + "grad_norm": 1.2757435448488257, + "learning_rate": 0.00010309823580818489, + "loss": 1.8181, + "step": 5653 + }, + { + "epoch": 0.5055436337625179, + "grad_norm": 2.820622921176429, + "learning_rate": 0.00010306928947223294, + "loss": 1.6597, + "step": 5654 + }, + { + "epoch": 0.5056330472103004, + "grad_norm": 1.7106883495377019, + "learning_rate": 0.00010304034287886337, + "loss": 2.0411, + "step": 5655 + }, + { + "epoch": 0.505722460658083, + "grad_norm": 2.5605990856977003, + "learning_rate": 0.00010301139603050394, + "loss": 1.7929, + "step": 5656 + }, + { + "epoch": 0.5058118741058655, + "grad_norm": 1.2432303004622969, + "learning_rate": 0.00010298244892958235, + "loss": 1.9663, + "step": 5657 + }, + { + "epoch": 0.505901287553648, + "grad_norm": 0.6546946498666215, + "learning_rate": 0.00010295350157852637, + "loss": 1.779, + "step": 5658 + }, + { + "epoch": 0.5059907010014306, + "grad_norm": 3.1985783066658784, + "learning_rate": 0.00010292455397976379, + "loss": 1.6632, + "step": 5659 + }, + { + "epoch": 0.5060801144492132, + "grad_norm": 2.13552891410547, + "learning_rate": 0.0001028956061357224, + "loss": 1.8853, + "step": 5660 + }, + { + "epoch": 0.5061695278969958, + "grad_norm": 3.8068832153471677, + "learning_rate": 0.00010286665804883, + "loss": 2.0019, + "step": 5661 + }, + { + "epoch": 0.5062589413447782, + "grad_norm": 2.4521319274407882, + "learning_rate": 0.00010283770972151445, + "loss": 1.8893, + "step": 5662 + }, + { + "epoch": 0.5063483547925608, + "grad_norm": 2.532624949855948, + "learning_rate": 0.00010280876115620365, + "loss": 2.0299, + "step": 5663 + }, + { + "epoch": 0.5064377682403434, + "grad_norm": 3.1840595277854975, + "learning_rate": 0.00010277981235532541, + "loss": 1.7807, + "step": 5664 + }, + { + "epoch": 0.5065271816881259, + "grad_norm": 0.5393452001704778, + "learning_rate": 0.00010275086332130768, + "loss": 1.5087, + "step": 5665 + }, + { + "epoch": 0.5066165951359084, + "grad_norm": 2.6452040469237854, + "learning_rate": 0.00010272191405657836, + "loss": 1.83, + "step": 5666 + }, + { + "epoch": 0.506706008583691, + "grad_norm": 6.415509574476299, + "learning_rate": 0.00010269296456356541, + "loss": 6.715, + "step": 5667 + }, + { + "epoch": 0.5067954220314735, + "grad_norm": 1.3998745790795646, + "learning_rate": 0.00010266401484469674, + "loss": 6.6832, + "step": 5668 + }, + { + "epoch": 0.5068848354792561, + "grad_norm": 7.086190105138635, + "learning_rate": 0.00010263506490240038, + "loss": 2.0681, + "step": 5669 + }, + { + "epoch": 0.5069742489270386, + "grad_norm": 2.5666248168465784, + "learning_rate": 0.00010260611473910433, + "loss": 6.7059, + "step": 5670 + }, + { + "epoch": 0.5070636623748211, + "grad_norm": 4.232026419254703, + "learning_rate": 0.00010257716435723656, + "loss": 1.7024, + "step": 5671 + }, + { + "epoch": 0.5071530758226037, + "grad_norm": 2.341848113426602, + "learning_rate": 0.00010254821375922512, + "loss": 1.5324, + "step": 5672 + }, + { + "epoch": 0.5072424892703863, + "grad_norm": 2.9033110643831983, + "learning_rate": 0.0001025192629474981, + "loss": 2.0821, + "step": 5673 + }, + { + "epoch": 0.5073319027181689, + "grad_norm": 4.413850956976885, + "learning_rate": 0.0001024903119244835, + "loss": 2.1894, + "step": 5674 + }, + { + "epoch": 0.5074213161659513, + "grad_norm": 1.1729277215933198, + "learning_rate": 0.0001024613606926095, + "loss": 1.8379, + "step": 5675 + }, + { + "epoch": 0.5075107296137339, + "grad_norm": 7.17844915133744, + "learning_rate": 0.00010243240925430411, + "loss": 2.2223, + "step": 5676 + }, + { + "epoch": 0.5076001430615165, + "grad_norm": 2.084562933225809, + "learning_rate": 0.00010240345761199553, + "loss": 1.4871, + "step": 5677 + }, + { + "epoch": 0.507689556509299, + "grad_norm": 6.0541912765081785, + "learning_rate": 0.0001023745057681118, + "loss": 1.5763, + "step": 5678 + }, + { + "epoch": 0.5077789699570815, + "grad_norm": 0.9509500699312111, + "learning_rate": 0.00010234555372508119, + "loss": 1.7286, + "step": 5679 + }, + { + "epoch": 0.5078683834048641, + "grad_norm": 4.059625544658581, + "learning_rate": 0.00010231660148533183, + "loss": 2.1209, + "step": 5680 + }, + { + "epoch": 0.5079577968526466, + "grad_norm": 2.3094451507908196, + "learning_rate": 0.00010228764905129184, + "loss": 6.6529, + "step": 5681 + }, + { + "epoch": 0.5080472103004292, + "grad_norm": 2.0577082746999795, + "learning_rate": 0.00010225869642538955, + "loss": 2.0234, + "step": 5682 + }, + { + "epoch": 0.5081366237482118, + "grad_norm": 0.40758543253025725, + "learning_rate": 0.00010222974361005309, + "loss": 1.8695, + "step": 5683 + }, + { + "epoch": 0.5082260371959942, + "grad_norm": 2.666119328350672, + "learning_rate": 0.00010220079060771075, + "loss": 2.1684, + "step": 5684 + }, + { + "epoch": 0.5083154506437768, + "grad_norm": 1.3113158864780468, + "learning_rate": 0.00010217183742079073, + "loss": 2.0031, + "step": 5685 + }, + { + "epoch": 0.5084048640915594, + "grad_norm": 1.8461541961086925, + "learning_rate": 0.00010214288405172133, + "loss": 2.2259, + "step": 5686 + }, + { + "epoch": 0.508494277539342, + "grad_norm": 1.0949684586512374, + "learning_rate": 0.00010211393050293083, + "loss": 1.5631, + "step": 5687 + }, + { + "epoch": 0.5085836909871244, + "grad_norm": 0.5028129157997022, + "learning_rate": 0.00010208497677684754, + "loss": 1.439, + "step": 5688 + }, + { + "epoch": 0.508673104434907, + "grad_norm": 0.5617891876289036, + "learning_rate": 0.0001020560228758998, + "loss": 1.629, + "step": 5689 + }, + { + "epoch": 0.5087625178826896, + "grad_norm": 1.284118702574631, + "learning_rate": 0.00010202706880251584, + "loss": 1.6789, + "step": 5690 + }, + { + "epoch": 0.5088519313304721, + "grad_norm": 0.6256738505739928, + "learning_rate": 0.00010199811455912412, + "loss": 2.0334, + "step": 5691 + }, + { + "epoch": 0.5089413447782547, + "grad_norm": 1.6622029832765808, + "learning_rate": 0.00010196916014815292, + "loss": 1.432, + "step": 5692 + }, + { + "epoch": 0.5090307582260372, + "grad_norm": 1.3053308607787184, + "learning_rate": 0.00010194020557203063, + "loss": 1.4822, + "step": 5693 + }, + { + "epoch": 0.5091201716738197, + "grad_norm": 1.1885200660025779, + "learning_rate": 0.00010191125083318566, + "loss": 1.533, + "step": 5694 + }, + { + "epoch": 0.5092095851216023, + "grad_norm": 1.586411642155849, + "learning_rate": 0.00010188229593404639, + "loss": 1.711, + "step": 5695 + }, + { + "epoch": 0.5092989985693849, + "grad_norm": 1.387587678890204, + "learning_rate": 0.00010185334087704124, + "loss": 1.6675, + "step": 5696 + }, + { + "epoch": 0.5093884120171673, + "grad_norm": 1.279681854012981, + "learning_rate": 0.0001018243856645986, + "loss": 1.4863, + "step": 5697 + }, + { + "epoch": 0.5094778254649499, + "grad_norm": 2.5404111655082606, + "learning_rate": 0.00010179543029914695, + "loss": 2.4282, + "step": 5698 + }, + { + "epoch": 0.5095672389127325, + "grad_norm": 1.9803536249199771, + "learning_rate": 0.00010176647478311473, + "loss": 1.949, + "step": 5699 + }, + { + "epoch": 0.509656652360515, + "grad_norm": 0.6725011518378318, + "learning_rate": 0.00010173751911893041, + "loss": 1.7246, + "step": 5700 + }, + { + "epoch": 0.5097460658082976, + "grad_norm": 1.1166914930528464, + "learning_rate": 0.00010170856330902247, + "loss": 1.8685, + "step": 5701 + }, + { + "epoch": 0.5098354792560801, + "grad_norm": 2.4104366206438392, + "learning_rate": 0.00010167960735581936, + "loss": 2.2691, + "step": 5702 + }, + { + "epoch": 0.5099248927038627, + "grad_norm": 0.8054981237223712, + "learning_rate": 0.00010165065126174962, + "loss": 1.6247, + "step": 5703 + }, + { + "epoch": 0.5100143061516452, + "grad_norm": 2.4313900685818126, + "learning_rate": 0.00010162169502924177, + "loss": 1.873, + "step": 5704 + }, + { + "epoch": 0.5101037195994278, + "grad_norm": 2.958866663499573, + "learning_rate": 0.0001015927386607243, + "loss": 2.0033, + "step": 5705 + }, + { + "epoch": 0.5101931330472103, + "grad_norm": 0.6429426566333292, + "learning_rate": 0.00010156378215862578, + "loss": 1.8587, + "step": 5706 + }, + { + "epoch": 0.5102825464949928, + "grad_norm": 3.1622463799110556, + "learning_rate": 0.00010153482552537472, + "loss": 2.33, + "step": 5707 + }, + { + "epoch": 0.5103719599427754, + "grad_norm": 2.1480438782007085, + "learning_rate": 0.00010150586876339969, + "loss": 1.8182, + "step": 5708 + }, + { + "epoch": 0.510461373390558, + "grad_norm": 2.4074023497602055, + "learning_rate": 0.00010147691187512928, + "loss": 1.852, + "step": 5709 + }, + { + "epoch": 0.5105507868383404, + "grad_norm": 1.252286799724558, + "learning_rate": 0.00010144795486299205, + "loss": 1.8702, + "step": 5710 + }, + { + "epoch": 0.510640200286123, + "grad_norm": 1.8659090755783827, + "learning_rate": 0.0001014189977294166, + "loss": 1.7516, + "step": 5711 + }, + { + "epoch": 0.5107296137339056, + "grad_norm": 0.9070862889832503, + "learning_rate": 0.00010139004047683151, + "loss": 1.9051, + "step": 5712 + }, + { + "epoch": 0.5108190271816881, + "grad_norm": 1.0741367959967438, + "learning_rate": 0.00010136108310766544, + "loss": 1.6521, + "step": 5713 + }, + { + "epoch": 0.5109084406294707, + "grad_norm": 1.3710087675698306, + "learning_rate": 0.00010133212562434693, + "loss": 1.7372, + "step": 5714 + }, + { + "epoch": 0.5109978540772532, + "grad_norm": 1.2235059634782866, + "learning_rate": 0.00010130316802930467, + "loss": 1.6378, + "step": 5715 + }, + { + "epoch": 0.5110872675250357, + "grad_norm": 1.5275904195785854, + "learning_rate": 0.00010127421032496729, + "loss": 1.9928, + "step": 5716 + }, + { + "epoch": 0.5111766809728183, + "grad_norm": 2.038099029665658, + "learning_rate": 0.00010124525251376342, + "loss": 2.4768, + "step": 5717 + }, + { + "epoch": 0.5112660944206009, + "grad_norm": 0.7385983367211066, + "learning_rate": 0.00010121629459812172, + "loss": 1.9234, + "step": 5718 + }, + { + "epoch": 0.5113555078683834, + "grad_norm": 1.6355683462700332, + "learning_rate": 0.00010118733658047088, + "loss": 1.8406, + "step": 5719 + }, + { + "epoch": 0.5114449213161659, + "grad_norm": 1.8903213881964256, + "learning_rate": 0.00010115837846323954, + "loss": 1.7972, + "step": 5720 + }, + { + "epoch": 0.5115343347639485, + "grad_norm": 1.0716492276799576, + "learning_rate": 0.00010112942024885639, + "loss": 2.3358, + "step": 5721 + }, + { + "epoch": 0.5116237482117311, + "grad_norm": 3.250737393652117, + "learning_rate": 0.00010110046193975014, + "loss": 1.8633, + "step": 5722 + }, + { + "epoch": 0.5117131616595136, + "grad_norm": 1.4038013059217798, + "learning_rate": 0.0001010715035383495, + "loss": 1.7379, + "step": 5723 + }, + { + "epoch": 0.5118025751072961, + "grad_norm": 0.9376868526785175, + "learning_rate": 0.00010104254504708311, + "loss": 1.6036, + "step": 5724 + }, + { + "epoch": 0.5118919885550787, + "grad_norm": 1.0527501094924052, + "learning_rate": 0.00010101358646837971, + "loss": 1.8892, + "step": 5725 + }, + { + "epoch": 0.5119814020028612, + "grad_norm": 1.3375292345104726, + "learning_rate": 0.00010098462780466808, + "loss": 1.8296, + "step": 5726 + }, + { + "epoch": 0.5120708154506438, + "grad_norm": 2.583390918293357, + "learning_rate": 0.00010095566905837692, + "loss": 6.5901, + "step": 5727 + }, + { + "epoch": 0.5121602288984263, + "grad_norm": 1.424117426104808, + "learning_rate": 0.00010092671023193491, + "loss": 1.7355, + "step": 5728 + }, + { + "epoch": 0.5122496423462088, + "grad_norm": 0.695134521297633, + "learning_rate": 0.00010089775132777084, + "loss": 2.2041, + "step": 5729 + }, + { + "epoch": 0.5123390557939914, + "grad_norm": 1.4257348629269024, + "learning_rate": 0.00010086879234831345, + "loss": 1.7278, + "step": 5730 + }, + { + "epoch": 0.512428469241774, + "grad_norm": 0.6643312717452808, + "learning_rate": 0.00010083983329599151, + "loss": 1.6653, + "step": 5731 + }, + { + "epoch": 0.5125178826895566, + "grad_norm": 1.470888458842403, + "learning_rate": 0.00010081087417323374, + "loss": 1.7206, + "step": 5732 + }, + { + "epoch": 0.512607296137339, + "grad_norm": 0.7513176031784655, + "learning_rate": 0.00010078191498246897, + "loss": 1.6626, + "step": 5733 + }, + { + "epoch": 0.5126967095851216, + "grad_norm": 1.3779010386637287, + "learning_rate": 0.00010075295572612593, + "loss": 1.9222, + "step": 5734 + }, + { + "epoch": 0.5127861230329042, + "grad_norm": 0.8671511995805262, + "learning_rate": 0.00010072399640663334, + "loss": 1.7942, + "step": 5735 + }, + { + "epoch": 0.5128755364806867, + "grad_norm": 2.403417325021818, + "learning_rate": 0.00010069503702642011, + "loss": 1.6077, + "step": 5736 + }, + { + "epoch": 0.5129649499284692, + "grad_norm": 0.8076491297382253, + "learning_rate": 0.00010066607758791495, + "loss": 1.5033, + "step": 5737 + }, + { + "epoch": 0.5130543633762518, + "grad_norm": 0.5874171226751922, + "learning_rate": 0.00010063711809354665, + "loss": 2.1858, + "step": 5738 + }, + { + "epoch": 0.5131437768240343, + "grad_norm": 1.6036181800399294, + "learning_rate": 0.00010060815854574403, + "loss": 1.9752, + "step": 5739 + }, + { + "epoch": 0.5132331902718169, + "grad_norm": 1.7918945269485644, + "learning_rate": 0.00010057919894693593, + "loss": 1.869, + "step": 5740 + }, + { + "epoch": 0.5133226037195995, + "grad_norm": 0.8285378796446415, + "learning_rate": 0.00010055023929955106, + "loss": 2.4096, + "step": 5741 + }, + { + "epoch": 0.5134120171673819, + "grad_norm": 2.3414678979443666, + "learning_rate": 0.00010052127960601829, + "loss": 1.9076, + "step": 5742 + }, + { + "epoch": 0.5135014306151645, + "grad_norm": 3.629584744926598, + "learning_rate": 0.00010049231986876646, + "loss": 6.7041, + "step": 5743 + }, + { + "epoch": 0.5135908440629471, + "grad_norm": 1.5885042800042461, + "learning_rate": 0.00010046336009022435, + "loss": 1.9407, + "step": 5744 + }, + { + "epoch": 0.5136802575107297, + "grad_norm": 3.007369263042841, + "learning_rate": 0.00010043440027282078, + "loss": 2.1599, + "step": 5745 + }, + { + "epoch": 0.5137696709585121, + "grad_norm": 1.6015617991057183, + "learning_rate": 0.00010040544041898456, + "loss": 1.8989, + "step": 5746 + }, + { + "epoch": 0.5138590844062947, + "grad_norm": 2.5750676094798868, + "learning_rate": 0.0001003764805311446, + "loss": 1.9649, + "step": 5747 + }, + { + "epoch": 0.5139484978540773, + "grad_norm": 2.508405127655456, + "learning_rate": 0.00010034752061172961, + "loss": 2.3001, + "step": 5748 + }, + { + "epoch": 0.5140379113018598, + "grad_norm": 1.737713466662593, + "learning_rate": 0.00010031856066316852, + "loss": 1.8434, + "step": 5749 + }, + { + "epoch": 0.5141273247496424, + "grad_norm": 1.9021635362831413, + "learning_rate": 0.00010028960068789012, + "loss": 1.7392, + "step": 5750 + }, + { + "epoch": 0.5142167381974249, + "grad_norm": 3.7246902101323136, + "learning_rate": 0.00010026064068832328, + "loss": 2.2645, + "step": 5751 + }, + { + "epoch": 0.5143061516452074, + "grad_norm": 0.9929002928557505, + "learning_rate": 0.00010023168066689677, + "loss": 2.1644, + "step": 5752 + }, + { + "epoch": 0.51439556509299, + "grad_norm": 2.463034956445213, + "learning_rate": 0.00010020272062603953, + "loss": 1.6409, + "step": 5753 + }, + { + "epoch": 0.5144849785407726, + "grad_norm": 3.699073875890389, + "learning_rate": 0.00010017376056818035, + "loss": 1.9034, + "step": 5754 + }, + { + "epoch": 0.514574391988555, + "grad_norm": 1.0344794099515993, + "learning_rate": 0.00010014480049574808, + "loss": 1.8014, + "step": 5755 + }, + { + "epoch": 0.5146638054363376, + "grad_norm": 2.9169249683003944, + "learning_rate": 0.00010011584041117155, + "loss": 1.8288, + "step": 5756 + }, + { + "epoch": 0.5147532188841202, + "grad_norm": 3.3011949908153477, + "learning_rate": 0.00010008688031687964, + "loss": 2.2738, + "step": 5757 + }, + { + "epoch": 0.5148426323319027, + "grad_norm": 2.56839097710487, + "learning_rate": 0.00010005792021530121, + "loss": 1.9386, + "step": 5758 + }, + { + "epoch": 0.5149320457796852, + "grad_norm": 1.7606639273693023, + "learning_rate": 0.000100028960108865, + "loss": 1.6935, + "step": 5759 + }, + { + "epoch": 0.5150214592274678, + "grad_norm": 2.338263872897644, + "learning_rate": 0.0001, + "loss": 1.5335, + "step": 5760 + }, + { + "epoch": 0.5151108726752504, + "grad_norm": 1.8095080582796947, + "learning_rate": 9.997103989113501e-05, + "loss": 1.6742, + "step": 5761 + }, + { + "epoch": 0.5152002861230329, + "grad_norm": 1.2170126426037078, + "learning_rate": 9.994207978469885e-05, + "loss": 1.8034, + "step": 5762 + }, + { + "epoch": 0.5152896995708155, + "grad_norm": 2.0911538729582295, + "learning_rate": 9.991311968312039e-05, + "loss": 1.8328, + "step": 5763 + }, + { + "epoch": 0.515379113018598, + "grad_norm": 1.4429235524565494, + "learning_rate": 9.988415958882845e-05, + "loss": 1.6032, + "step": 5764 + }, + { + "epoch": 0.5154685264663805, + "grad_norm": 1.8184678280720201, + "learning_rate": 9.985519950425196e-05, + "loss": 1.7763, + "step": 5765 + }, + { + "epoch": 0.5155579399141631, + "grad_norm": 0.9041790474218557, + "learning_rate": 9.982623943181966e-05, + "loss": 1.8583, + "step": 5766 + }, + { + "epoch": 0.5156473533619457, + "grad_norm": 1.6983664433075698, + "learning_rate": 9.979727937396048e-05, + "loss": 2.2641, + "step": 5767 + }, + { + "epoch": 0.5157367668097281, + "grad_norm": 4.054417770410007, + "learning_rate": 9.976831933310324e-05, + "loss": 6.6776, + "step": 5768 + }, + { + "epoch": 0.5158261802575107, + "grad_norm": 1.6110623129980088, + "learning_rate": 9.973935931167677e-05, + "loss": 1.9435, + "step": 5769 + }, + { + "epoch": 0.5159155937052933, + "grad_norm": 1.0565505868188187, + "learning_rate": 9.971039931210993e-05, + "loss": 1.6638, + "step": 5770 + }, + { + "epoch": 0.5160050071530758, + "grad_norm": 1.3826582366258973, + "learning_rate": 9.968143933683149e-05, + "loss": 2.2306, + "step": 5771 + }, + { + "epoch": 0.5160944206008584, + "grad_norm": 0.4382411379521342, + "learning_rate": 9.965247938827041e-05, + "loss": 1.392, + "step": 5772 + }, + { + "epoch": 0.5161838340486409, + "grad_norm": 1.7947706121311258, + "learning_rate": 9.962351946885544e-05, + "loss": 1.687, + "step": 5773 + }, + { + "epoch": 0.5162732474964234, + "grad_norm": 1.6892716047400214, + "learning_rate": 9.959455958101546e-05, + "loss": 2.4222, + "step": 5774 + }, + { + "epoch": 0.516362660944206, + "grad_norm": 1.0822186478833236, + "learning_rate": 9.956559972717925e-05, + "loss": 1.9696, + "step": 5775 + }, + { + "epoch": 0.5164520743919886, + "grad_norm": 1.5130466696032046, + "learning_rate": 9.953663990977568e-05, + "loss": 1.9725, + "step": 5776 + }, + { + "epoch": 0.516541487839771, + "grad_norm": 1.7337963497663784, + "learning_rate": 9.950768013123358e-05, + "loss": 2.0847, + "step": 5777 + }, + { + "epoch": 0.5166309012875536, + "grad_norm": 1.295809875873938, + "learning_rate": 9.94787203939817e-05, + "loss": 2.0089, + "step": 5778 + }, + { + "epoch": 0.5167203147353362, + "grad_norm": 2.0219904977437766, + "learning_rate": 9.944976070044894e-05, + "loss": 1.7265, + "step": 5779 + }, + { + "epoch": 0.5168097281831188, + "grad_norm": 1.5415079761893946, + "learning_rate": 9.94208010530641e-05, + "loss": 1.646, + "step": 5780 + }, + { + "epoch": 0.5168991416309013, + "grad_norm": 2.975665638630423, + "learning_rate": 9.939184145425598e-05, + "loss": 1.9376, + "step": 5781 + }, + { + "epoch": 0.5169885550786838, + "grad_norm": 0.985174801275643, + "learning_rate": 9.936288190645336e-05, + "loss": 1.7129, + "step": 5782 + }, + { + "epoch": 0.5170779685264664, + "grad_norm": 1.4907156052965413, + "learning_rate": 9.933392241208507e-05, + "loss": 1.7863, + "step": 5783 + }, + { + "epoch": 0.5171673819742489, + "grad_norm": 1.9222346635811955, + "learning_rate": 9.930496297357993e-05, + "loss": 2.0573, + "step": 5784 + }, + { + "epoch": 0.5172567954220315, + "grad_norm": 0.5633465838608702, + "learning_rate": 9.927600359336666e-05, + "loss": 2.0911, + "step": 5785 + }, + { + "epoch": 0.517346208869814, + "grad_norm": 1.3434514726568854, + "learning_rate": 9.92470442738741e-05, + "loss": 1.9661, + "step": 5786 + }, + { + "epoch": 0.5174356223175965, + "grad_norm": 0.5471009868666326, + "learning_rate": 9.921808501753106e-05, + "loss": 1.7281, + "step": 5787 + }, + { + "epoch": 0.5175250357653791, + "grad_norm": 1.052650378068922, + "learning_rate": 9.918912582676629e-05, + "loss": 1.9106, + "step": 5788 + }, + { + "epoch": 0.5176144492131617, + "grad_norm": 0.6247300497464271, + "learning_rate": 9.916016670400851e-05, + "loss": 1.8496, + "step": 5789 + }, + { + "epoch": 0.5177038626609443, + "grad_norm": 0.46179241621098177, + "learning_rate": 9.913120765168657e-05, + "loss": 1.7136, + "step": 5790 + }, + { + "epoch": 0.5177932761087267, + "grad_norm": 1.127221110298839, + "learning_rate": 9.910224867222921e-05, + "loss": 1.9212, + "step": 5791 + }, + { + "epoch": 0.5178826895565093, + "grad_norm": 1.719408552872576, + "learning_rate": 9.907328976806511e-05, + "loss": 1.8213, + "step": 5792 + }, + { + "epoch": 0.5179721030042919, + "grad_norm": 0.5131278337736475, + "learning_rate": 9.904433094162311e-05, + "loss": 2.1266, + "step": 5793 + }, + { + "epoch": 0.5180615164520744, + "grad_norm": 0.7452592222126541, + "learning_rate": 9.901537219533194e-05, + "loss": 1.7504, + "step": 5794 + }, + { + "epoch": 0.5181509298998569, + "grad_norm": 1.5551709294907452, + "learning_rate": 9.89864135316203e-05, + "loss": 1.7212, + "step": 5795 + }, + { + "epoch": 0.5182403433476395, + "grad_norm": 0.8402684168505201, + "learning_rate": 9.895745495291693e-05, + "loss": 1.9105, + "step": 5796 + }, + { + "epoch": 0.518329756795422, + "grad_norm": 1.7548370064112508, + "learning_rate": 9.892849646165057e-05, + "loss": 2.0592, + "step": 5797 + }, + { + "epoch": 0.5184191702432046, + "grad_norm": 0.9073696456785215, + "learning_rate": 9.889953806024991e-05, + "loss": 1.9347, + "step": 5798 + }, + { + "epoch": 0.5185085836909872, + "grad_norm": 1.0091777045131736, + "learning_rate": 9.887057975114362e-05, + "loss": 1.8469, + "step": 5799 + }, + { + "epoch": 0.5185979971387696, + "grad_norm": 2.4004200994657845, + "learning_rate": 9.884162153676048e-05, + "loss": 6.6477, + "step": 5800 + }, + { + "epoch": 0.5186874105865522, + "grad_norm": 4.3066390356643485, + "learning_rate": 9.881266341952915e-05, + "loss": 1.9606, + "step": 5801 + }, + { + "epoch": 0.5187768240343348, + "grad_norm": 1.6094299862902501, + "learning_rate": 9.878370540187831e-05, + "loss": 1.8327, + "step": 5802 + }, + { + "epoch": 0.5188662374821174, + "grad_norm": 1.8683153037370896, + "learning_rate": 9.875474748623661e-05, + "loss": 1.8962, + "step": 5803 + }, + { + "epoch": 0.5189556509298998, + "grad_norm": 3.1431923189589193, + "learning_rate": 9.872578967503275e-05, + "loss": 2.1597, + "step": 5804 + }, + { + "epoch": 0.5190450643776824, + "grad_norm": 1.4047619094547905, + "learning_rate": 9.869683197069533e-05, + "loss": 1.7956, + "step": 5805 + }, + { + "epoch": 0.519134477825465, + "grad_norm": 2.1426842703205926, + "learning_rate": 9.866787437565308e-05, + "loss": 2.2121, + "step": 5806 + }, + { + "epoch": 0.5192238912732475, + "grad_norm": 3.018783817159486, + "learning_rate": 9.863891689233459e-05, + "loss": 2.0176, + "step": 5807 + }, + { + "epoch": 0.51931330472103, + "grad_norm": 1.5443480903596445, + "learning_rate": 9.860995952316851e-05, + "loss": 1.8221, + "step": 5808 + }, + { + "epoch": 0.5194027181688126, + "grad_norm": 2.344864467061415, + "learning_rate": 9.858100227058342e-05, + "loss": 2.4241, + "step": 5809 + }, + { + "epoch": 0.5194921316165951, + "grad_norm": 1.7350018450512592, + "learning_rate": 9.855204513700797e-05, + "loss": 2.1705, + "step": 5810 + }, + { + "epoch": 0.5195815450643777, + "grad_norm": 2.139244413669776, + "learning_rate": 9.852308812487075e-05, + "loss": 2.1784, + "step": 5811 + }, + { + "epoch": 0.5196709585121603, + "grad_norm": 3.6232323949208354, + "learning_rate": 9.84941312366003e-05, + "loss": 1.6313, + "step": 5812 + }, + { + "epoch": 0.5197603719599427, + "grad_norm": 0.5295567416729141, + "learning_rate": 9.846517447462527e-05, + "loss": 2.0221, + "step": 5813 + }, + { + "epoch": 0.5198497854077253, + "grad_norm": 1.2211508955158037, + "learning_rate": 9.843621784137424e-05, + "loss": 1.6324, + "step": 5814 + }, + { + "epoch": 0.5199391988555079, + "grad_norm": 2.9177763829992225, + "learning_rate": 9.840726133927571e-05, + "loss": 1.9876, + "step": 5815 + }, + { + "epoch": 0.5200286123032904, + "grad_norm": 1.1461197547667292, + "learning_rate": 9.837830497075824e-05, + "loss": 1.8161, + "step": 5816 + }, + { + "epoch": 0.5201180257510729, + "grad_norm": 2.3891496038241353, + "learning_rate": 9.834934873825038e-05, + "loss": 1.841, + "step": 5817 + }, + { + "epoch": 0.5202074391988555, + "grad_norm": 3.220284078660168, + "learning_rate": 9.832039264418067e-05, + "loss": 1.7917, + "step": 5818 + }, + { + "epoch": 0.520296852646638, + "grad_norm": 1.1996414566866587, + "learning_rate": 9.829143669097754e-05, + "loss": 1.8893, + "step": 5819 + }, + { + "epoch": 0.5203862660944206, + "grad_norm": 3.9501769882135327, + "learning_rate": 9.826248088106959e-05, + "loss": 1.7491, + "step": 5820 + }, + { + "epoch": 0.5204756795422032, + "grad_norm": 0.6380578211378596, + "learning_rate": 9.823352521688528e-05, + "loss": 2.0981, + "step": 5821 + }, + { + "epoch": 0.5205650929899857, + "grad_norm": 4.232536715354211, + "learning_rate": 9.820456970085307e-05, + "loss": 1.5964, + "step": 5822 + }, + { + "epoch": 0.5206545064377682, + "grad_norm": 1.1455586778389721, + "learning_rate": 9.817561433540141e-05, + "loss": 1.9554, + "step": 5823 + }, + { + "epoch": 0.5207439198855508, + "grad_norm": 1.2191360133394913, + "learning_rate": 9.81466591229588e-05, + "loss": 2.025, + "step": 5824 + }, + { + "epoch": 0.5208333333333334, + "grad_norm": 1.1007597502073883, + "learning_rate": 9.811770406595365e-05, + "loss": 2.2645, + "step": 5825 + }, + { + "epoch": 0.5209227467811158, + "grad_norm": 1.7212947402833874, + "learning_rate": 9.808874916681436e-05, + "loss": 1.5748, + "step": 5826 + }, + { + "epoch": 0.5210121602288984, + "grad_norm": 2.1466902637746386, + "learning_rate": 9.805979442796936e-05, + "loss": 2.3309, + "step": 5827 + }, + { + "epoch": 0.521101573676681, + "grad_norm": 2.3780465750604978, + "learning_rate": 9.80308398518471e-05, + "loss": 2.2108, + "step": 5828 + }, + { + "epoch": 0.5211909871244635, + "grad_norm": 0.7052769256308651, + "learning_rate": 9.800188544087592e-05, + "loss": 1.8869, + "step": 5829 + }, + { + "epoch": 0.5212804005722461, + "grad_norm": 2.075102377983756, + "learning_rate": 9.797293119748417e-05, + "loss": 1.8811, + "step": 5830 + }, + { + "epoch": 0.5213698140200286, + "grad_norm": 1.5928366300262484, + "learning_rate": 9.794397712410025e-05, + "loss": 1.5674, + "step": 5831 + }, + { + "epoch": 0.5214592274678111, + "grad_norm": 1.0642384480625384, + "learning_rate": 9.791502322315249e-05, + "loss": 1.7942, + "step": 5832 + }, + { + "epoch": 0.5215486409155937, + "grad_norm": 0.7765961550667112, + "learning_rate": 9.788606949706918e-05, + "loss": 1.8112, + "step": 5833 + }, + { + "epoch": 0.5216380543633763, + "grad_norm": 1.7643290088808858, + "learning_rate": 9.785711594827868e-05, + "loss": 1.9353, + "step": 5834 + }, + { + "epoch": 0.5217274678111588, + "grad_norm": 0.5806910127414951, + "learning_rate": 9.78281625792093e-05, + "loss": 1.6496, + "step": 5835 + }, + { + "epoch": 0.5218168812589413, + "grad_norm": 0.8046392792220508, + "learning_rate": 9.779920939228928e-05, + "loss": 1.6069, + "step": 5836 + }, + { + "epoch": 0.5219062947067239, + "grad_norm": 1.3495317460186305, + "learning_rate": 9.777025638994693e-05, + "loss": 1.7263, + "step": 5837 + }, + { + "epoch": 0.5219957081545065, + "grad_norm": 0.7304404999644081, + "learning_rate": 9.774130357461049e-05, + "loss": 1.7081, + "step": 5838 + }, + { + "epoch": 0.522085121602289, + "grad_norm": 1.6282627303123, + "learning_rate": 9.771235094870817e-05, + "loss": 2.5263, + "step": 5839 + }, + { + "epoch": 0.5221745350500715, + "grad_norm": 0.5420356029311696, + "learning_rate": 9.768339851466818e-05, + "loss": 1.8203, + "step": 5840 + }, + { + "epoch": 0.5222639484978541, + "grad_norm": 0.6788331922797379, + "learning_rate": 9.765444627491882e-05, + "loss": 1.7447, + "step": 5841 + }, + { + "epoch": 0.5223533619456366, + "grad_norm": 1.0764820157124517, + "learning_rate": 9.76254942318882e-05, + "loss": 1.8283, + "step": 5842 + }, + { + "epoch": 0.5224427753934192, + "grad_norm": 1.9313992616438058, + "learning_rate": 9.759654238800451e-05, + "loss": 2.2416, + "step": 5843 + }, + { + "epoch": 0.5225321888412017, + "grad_norm": 0.7073695861756115, + "learning_rate": 9.756759074569591e-05, + "loss": 1.4386, + "step": 5844 + }, + { + "epoch": 0.5226216022889842, + "grad_norm": 1.8402571927288651, + "learning_rate": 9.753863930739054e-05, + "loss": 2.0749, + "step": 5845 + }, + { + "epoch": 0.5227110157367668, + "grad_norm": 1.2807120644768357, + "learning_rate": 9.75096880755165e-05, + "loss": 1.7357, + "step": 5846 + }, + { + "epoch": 0.5228004291845494, + "grad_norm": 0.9582681229656309, + "learning_rate": 9.748073705250188e-05, + "loss": 1.7741, + "step": 5847 + }, + { + "epoch": 0.522889842632332, + "grad_norm": 1.0174231933003604, + "learning_rate": 9.745178624077488e-05, + "loss": 1.5174, + "step": 5848 + }, + { + "epoch": 0.5229792560801144, + "grad_norm": 0.8965718841103638, + "learning_rate": 9.742283564276347e-05, + "loss": 1.6626, + "step": 5849 + }, + { + "epoch": 0.523068669527897, + "grad_norm": 0.6009025413635768, + "learning_rate": 9.739388526089568e-05, + "loss": 1.8205, + "step": 5850 + }, + { + "epoch": 0.5231580829756796, + "grad_norm": 0.6870644157192652, + "learning_rate": 9.736493509759962e-05, + "loss": 1.6964, + "step": 5851 + }, + { + "epoch": 0.5232474964234621, + "grad_norm": 1.8415289309426848, + "learning_rate": 9.733598515530328e-05, + "loss": 2.0151, + "step": 5852 + }, + { + "epoch": 0.5233369098712446, + "grad_norm": 1.0397838682824618, + "learning_rate": 9.730703543643464e-05, + "loss": 1.8538, + "step": 5853 + }, + { + "epoch": 0.5234263233190272, + "grad_norm": 2.0832606760068635, + "learning_rate": 9.727808594342164e-05, + "loss": 1.799, + "step": 5854 + }, + { + "epoch": 0.5235157367668097, + "grad_norm": 1.2251231452954645, + "learning_rate": 9.724913667869233e-05, + "loss": 1.9269, + "step": 5855 + }, + { + "epoch": 0.5236051502145923, + "grad_norm": 0.4637965315933595, + "learning_rate": 9.722018764467461e-05, + "loss": 1.6779, + "step": 5856 + }, + { + "epoch": 0.5236945636623748, + "grad_norm": 1.0861155484169227, + "learning_rate": 9.719123884379637e-05, + "loss": 2.0932, + "step": 5857 + }, + { + "epoch": 0.5237839771101573, + "grad_norm": 1.1038001947119436, + "learning_rate": 9.716229027848556e-05, + "loss": 1.7785, + "step": 5858 + }, + { + "epoch": 0.5238733905579399, + "grad_norm": 0.6791595596400544, + "learning_rate": 9.713334195117004e-05, + "loss": 2.1431, + "step": 5859 + }, + { + "epoch": 0.5239628040057225, + "grad_norm": 1.5816669559491434, + "learning_rate": 9.710439386427764e-05, + "loss": 1.9288, + "step": 5860 + }, + { + "epoch": 0.524052217453505, + "grad_norm": 0.9709233624330212, + "learning_rate": 9.707544602023622e-05, + "loss": 1.8193, + "step": 5861 + }, + { + "epoch": 0.5241416309012875, + "grad_norm": 1.0481605938911145, + "learning_rate": 9.704649842147364e-05, + "loss": 2.2682, + "step": 5862 + }, + { + "epoch": 0.5242310443490701, + "grad_norm": 2.1286905852590006, + "learning_rate": 9.701755107041767e-05, + "loss": 2.0686, + "step": 5863 + }, + { + "epoch": 0.5243204577968527, + "grad_norm": 1.228497047720931, + "learning_rate": 9.698860396949608e-05, + "loss": 2.0897, + "step": 5864 + }, + { + "epoch": 0.5244098712446352, + "grad_norm": 1.7490044144626216, + "learning_rate": 9.695965712113666e-05, + "loss": 1.6473, + "step": 5865 + }, + { + "epoch": 0.5244992846924177, + "grad_norm": 1.965952749074139, + "learning_rate": 9.69307105277671e-05, + "loss": 2.4282, + "step": 5866 + }, + { + "epoch": 0.5245886981402003, + "grad_norm": 1.0754335540232782, + "learning_rate": 9.690176419181516e-05, + "loss": 2.1745, + "step": 5867 + }, + { + "epoch": 0.5246781115879828, + "grad_norm": 1.498197590122057, + "learning_rate": 9.687281811570847e-05, + "loss": 2.1224, + "step": 5868 + }, + { + "epoch": 0.5247675250357654, + "grad_norm": 2.5139192104944144, + "learning_rate": 9.68438723018748e-05, + "loss": 1.5856, + "step": 5869 + }, + { + "epoch": 0.524856938483548, + "grad_norm": 0.6397418221010601, + "learning_rate": 9.681492675274171e-05, + "loss": 1.7951, + "step": 5870 + }, + { + "epoch": 0.5249463519313304, + "grad_norm": 2.2716896158090214, + "learning_rate": 9.678598147073689e-05, + "loss": 2.028, + "step": 5871 + }, + { + "epoch": 0.525035765379113, + "grad_norm": 1.3666113123906403, + "learning_rate": 9.675703645828794e-05, + "loss": 1.8879, + "step": 5872 + }, + { + "epoch": 0.5251251788268956, + "grad_norm": 1.8043001531676957, + "learning_rate": 9.67280917178224e-05, + "loss": 1.5242, + "step": 5873 + }, + { + "epoch": 0.5252145922746781, + "grad_norm": 1.9730996505155598, + "learning_rate": 9.669914725176787e-05, + "loss": 1.7916, + "step": 5874 + }, + { + "epoch": 0.5253040057224606, + "grad_norm": 0.8106935417208088, + "learning_rate": 9.667020306255183e-05, + "loss": 1.7597, + "step": 5875 + }, + { + "epoch": 0.5253934191702432, + "grad_norm": 1.9671361271786945, + "learning_rate": 9.66412591526019e-05, + "loss": 1.8452, + "step": 5876 + }, + { + "epoch": 0.5254828326180258, + "grad_norm": 4.247677805161104, + "learning_rate": 9.661231552434546e-05, + "loss": 1.4222, + "step": 5877 + }, + { + "epoch": 0.5255722460658083, + "grad_norm": 0.9564861898801691, + "learning_rate": 9.658337218021007e-05, + "loss": 1.7592, + "step": 5878 + }, + { + "epoch": 0.5256616595135909, + "grad_norm": 1.9752192366499792, + "learning_rate": 9.655442912262311e-05, + "loss": 1.7555, + "step": 5879 + }, + { + "epoch": 0.5257510729613734, + "grad_norm": 2.3587782743223014, + "learning_rate": 9.652548635401201e-05, + "loss": 1.9165, + "step": 5880 + }, + { + "epoch": 0.5258404864091559, + "grad_norm": 0.9036223659216817, + "learning_rate": 9.64965438768042e-05, + "loss": 2.1958, + "step": 5881 + }, + { + "epoch": 0.5259298998569385, + "grad_norm": 1.410185681612369, + "learning_rate": 9.646760169342696e-05, + "loss": 1.8492, + "step": 5882 + }, + { + "epoch": 0.5260193133047211, + "grad_norm": 0.8524219379225796, + "learning_rate": 9.643865980630775e-05, + "loss": 1.7848, + "step": 5883 + }, + { + "epoch": 0.5261087267525035, + "grad_norm": 1.0826446364283835, + "learning_rate": 9.640971821787382e-05, + "loss": 1.5321, + "step": 5884 + }, + { + "epoch": 0.5261981402002861, + "grad_norm": 0.7081058754990627, + "learning_rate": 9.638077693055252e-05, + "loss": 1.7003, + "step": 5885 + }, + { + "epoch": 0.5262875536480687, + "grad_norm": 0.9470798417500508, + "learning_rate": 9.635183594677107e-05, + "loss": 2.0082, + "step": 5886 + }, + { + "epoch": 0.5263769670958512, + "grad_norm": 0.4619266118009254, + "learning_rate": 9.632289526895672e-05, + "loss": 1.8592, + "step": 5887 + }, + { + "epoch": 0.5264663805436338, + "grad_norm": 1.0181357569698306, + "learning_rate": 9.629395489953669e-05, + "loss": 1.6905, + "step": 5888 + }, + { + "epoch": 0.5265557939914163, + "grad_norm": 0.6847602290161321, + "learning_rate": 9.626501484093823e-05, + "loss": 1.77, + "step": 5889 + }, + { + "epoch": 0.5266452074391988, + "grad_norm": 1.1483246424760711, + "learning_rate": 9.623607509558846e-05, + "loss": 1.7921, + "step": 5890 + }, + { + "epoch": 0.5267346208869814, + "grad_norm": 1.8929356970603375, + "learning_rate": 9.620713566591449e-05, + "loss": 2.1981, + "step": 5891 + }, + { + "epoch": 0.526824034334764, + "grad_norm": 1.440417435984481, + "learning_rate": 9.61781965543435e-05, + "loss": 2.0841, + "step": 5892 + }, + { + "epoch": 0.5269134477825465, + "grad_norm": 0.8938639401909471, + "learning_rate": 9.614925776330254e-05, + "loss": 1.7282, + "step": 5893 + }, + { + "epoch": 0.527002861230329, + "grad_norm": 0.604248733121952, + "learning_rate": 9.612031929521869e-05, + "loss": 1.9859, + "step": 5894 + }, + { + "epoch": 0.5270922746781116, + "grad_norm": 0.4736416382052757, + "learning_rate": 9.609138115251894e-05, + "loss": 1.3792, + "step": 5895 + }, + { + "epoch": 0.5271816881258942, + "grad_norm": 0.5606718242070454, + "learning_rate": 9.606244333763038e-05, + "loss": 1.8395, + "step": 5896 + }, + { + "epoch": 0.5272711015736766, + "grad_norm": 0.6529294248474393, + "learning_rate": 9.603350585297991e-05, + "loss": 2.0746, + "step": 5897 + }, + { + "epoch": 0.5273605150214592, + "grad_norm": 0.48714757591445385, + "learning_rate": 9.600456870099454e-05, + "loss": 1.8168, + "step": 5898 + }, + { + "epoch": 0.5274499284692418, + "grad_norm": 1.5110313548491996, + "learning_rate": 9.597563188410116e-05, + "loss": 1.7801, + "step": 5899 + }, + { + "epoch": 0.5275393419170243, + "grad_norm": 0.9705882058209758, + "learning_rate": 9.594669540472666e-05, + "loss": 1.7672, + "step": 5900 + }, + { + "epoch": 0.5276287553648069, + "grad_norm": 0.7962532016494072, + "learning_rate": 9.591775926529793e-05, + "loss": 1.9588, + "step": 5901 + }, + { + "epoch": 0.5277181688125894, + "grad_norm": 1.3471229250503953, + "learning_rate": 9.588882346824177e-05, + "loss": 1.5904, + "step": 5902 + }, + { + "epoch": 0.5278075822603719, + "grad_norm": 0.35393214902365966, + "learning_rate": 9.585988801598506e-05, + "loss": 1.47, + "step": 5903 + }, + { + "epoch": 0.5278969957081545, + "grad_norm": 1.2885588129237646, + "learning_rate": 9.583095291095453e-05, + "loss": 1.7286, + "step": 5904 + }, + { + "epoch": 0.5279864091559371, + "grad_norm": 0.5837006157405129, + "learning_rate": 9.580201815557695e-05, + "loss": 2.0427, + "step": 5905 + }, + { + "epoch": 0.5280758226037195, + "grad_norm": 0.8559241393358288, + "learning_rate": 9.577308375227906e-05, + "loss": 1.8778, + "step": 5906 + }, + { + "epoch": 0.5281652360515021, + "grad_norm": 0.7013566237140234, + "learning_rate": 9.574414970348749e-05, + "loss": 1.9606, + "step": 5907 + }, + { + "epoch": 0.5282546494992847, + "grad_norm": 0.8398616520332585, + "learning_rate": 9.571521601162897e-05, + "loss": 2.2763, + "step": 5908 + }, + { + "epoch": 0.5283440629470673, + "grad_norm": 1.1313047973372392, + "learning_rate": 9.568628267913007e-05, + "loss": 1.3882, + "step": 5909 + }, + { + "epoch": 0.5284334763948498, + "grad_norm": 1.6510983356571005, + "learning_rate": 9.565734970841747e-05, + "loss": 1.6113, + "step": 5910 + }, + { + "epoch": 0.5285228898426323, + "grad_norm": 0.9286426399536342, + "learning_rate": 9.562841710191769e-05, + "loss": 1.5327, + "step": 5911 + }, + { + "epoch": 0.5286123032904149, + "grad_norm": 0.4626187663290597, + "learning_rate": 9.55994848620573e-05, + "loss": 1.9687, + "step": 5912 + }, + { + "epoch": 0.5287017167381974, + "grad_norm": 1.5228224109783948, + "learning_rate": 9.55705529912628e-05, + "loss": 2.0603, + "step": 5913 + }, + { + "epoch": 0.52879113018598, + "grad_norm": 0.5923677592640406, + "learning_rate": 9.554162149196066e-05, + "loss": 2.1727, + "step": 5914 + }, + { + "epoch": 0.5288805436337625, + "grad_norm": 0.7841255968745273, + "learning_rate": 9.551269036657736e-05, + "loss": 1.404, + "step": 5915 + }, + { + "epoch": 0.528969957081545, + "grad_norm": 3.0767082734859224, + "learning_rate": 9.548375961753926e-05, + "loss": 6.5497, + "step": 5916 + }, + { + "epoch": 0.5290593705293276, + "grad_norm": 1.6613203290668843, + "learning_rate": 9.545482924727282e-05, + "loss": 2.2644, + "step": 5917 + }, + { + "epoch": 0.5291487839771102, + "grad_norm": 0.9113436249536143, + "learning_rate": 9.542589925820435e-05, + "loss": 2.0318, + "step": 5918 + }, + { + "epoch": 0.5292381974248928, + "grad_norm": 1.6050819241089902, + "learning_rate": 9.539696965276019e-05, + "loss": 2.1077, + "step": 5919 + }, + { + "epoch": 0.5293276108726752, + "grad_norm": 1.2252587266031796, + "learning_rate": 9.536804043336664e-05, + "loss": 1.6522, + "step": 5920 + }, + { + "epoch": 0.5294170243204578, + "grad_norm": 1.0902874030845293, + "learning_rate": 9.533911160244993e-05, + "loss": 1.9967, + "step": 5921 + }, + { + "epoch": 0.5295064377682404, + "grad_norm": 2.113287323903709, + "learning_rate": 9.53101831624363e-05, + "loss": 2.3135, + "step": 5922 + }, + { + "epoch": 0.5295958512160229, + "grad_norm": 0.9196805762254102, + "learning_rate": 9.528125511575193e-05, + "loss": 1.6762, + "step": 5923 + }, + { + "epoch": 0.5296852646638054, + "grad_norm": 1.5305795915498015, + "learning_rate": 9.525232746482301e-05, + "loss": 1.5992, + "step": 5924 + }, + { + "epoch": 0.529774678111588, + "grad_norm": 1.3422408293045223, + "learning_rate": 9.522340021207564e-05, + "loss": 1.6351, + "step": 5925 + }, + { + "epoch": 0.5298640915593705, + "grad_norm": 1.1897686494299873, + "learning_rate": 9.519447335993595e-05, + "loss": 1.5804, + "step": 5926 + }, + { + "epoch": 0.5299535050071531, + "grad_norm": 1.1497242892070838, + "learning_rate": 9.516554691082995e-05, + "loss": 1.5425, + "step": 5927 + }, + { + "epoch": 0.5300429184549357, + "grad_norm": 2.597968629079806, + "learning_rate": 9.513662086718372e-05, + "loss": 1.6558, + "step": 5928 + }, + { + "epoch": 0.5301323319027181, + "grad_norm": 0.9416168340389391, + "learning_rate": 9.510769523142322e-05, + "loss": 1.6495, + "step": 5929 + }, + { + "epoch": 0.5302217453505007, + "grad_norm": 0.8934851507215246, + "learning_rate": 9.507877000597437e-05, + "loss": 2.0818, + "step": 5930 + }, + { + "epoch": 0.5303111587982833, + "grad_norm": 0.8937837612485133, + "learning_rate": 9.504984519326316e-05, + "loss": 2.073, + "step": 5931 + }, + { + "epoch": 0.5304005722460658, + "grad_norm": 1.0876265409985673, + "learning_rate": 9.502092079571547e-05, + "loss": 2.0051, + "step": 5932 + }, + { + "epoch": 0.5304899856938483, + "grad_norm": 1.4937715404825282, + "learning_rate": 9.499199681575716e-05, + "loss": 2.0827, + "step": 5933 + }, + { + "epoch": 0.5305793991416309, + "grad_norm": 0.5847472756949492, + "learning_rate": 9.496307325581398e-05, + "loss": 1.6962, + "step": 5934 + }, + { + "epoch": 0.5306688125894135, + "grad_norm": 1.9647278064780735, + "learning_rate": 9.49341501183118e-05, + "loss": 1.7167, + "step": 5935 + }, + { + "epoch": 0.530758226037196, + "grad_norm": 1.547793898186035, + "learning_rate": 9.490522740567633e-05, + "loss": 1.7197, + "step": 5936 + }, + { + "epoch": 0.5308476394849786, + "grad_norm": 0.9842310540988336, + "learning_rate": 9.487630512033325e-05, + "loss": 2.0072, + "step": 5937 + }, + { + "epoch": 0.530937052932761, + "grad_norm": 1.721577896519471, + "learning_rate": 9.484738326470828e-05, + "loss": 1.81, + "step": 5938 + }, + { + "epoch": 0.5310264663805436, + "grad_norm": 0.44903561741476145, + "learning_rate": 9.481846184122707e-05, + "loss": 1.7429, + "step": 5939 + }, + { + "epoch": 0.5311158798283262, + "grad_norm": 2.232962841390918, + "learning_rate": 9.478954085231522e-05, + "loss": 1.5707, + "step": 5940 + }, + { + "epoch": 0.5312052932761088, + "grad_norm": 1.3303162704385938, + "learning_rate": 9.476062030039825e-05, + "loss": 1.6229, + "step": 5941 + }, + { + "epoch": 0.5312947067238912, + "grad_norm": 1.96081851798222, + "learning_rate": 9.473170018790176e-05, + "loss": 1.9234, + "step": 5942 + }, + { + "epoch": 0.5313841201716738, + "grad_norm": 1.8560520341123437, + "learning_rate": 9.470278051725122e-05, + "loss": 1.97, + "step": 5943 + }, + { + "epoch": 0.5314735336194564, + "grad_norm": 1.1640981110556825, + "learning_rate": 9.467386129087202e-05, + "loss": 2.1103, + "step": 5944 + }, + { + "epoch": 0.531562947067239, + "grad_norm": 1.545770267412217, + "learning_rate": 9.464494251118968e-05, + "loss": 1.2842, + "step": 5945 + }, + { + "epoch": 0.5316523605150214, + "grad_norm": 0.954368542993865, + "learning_rate": 9.461602418062956e-05, + "loss": 1.6537, + "step": 5946 + }, + { + "epoch": 0.531741773962804, + "grad_norm": 0.9489475074723258, + "learning_rate": 9.458710630161698e-05, + "loss": 1.7062, + "step": 5947 + }, + { + "epoch": 0.5318311874105865, + "grad_norm": 0.8249177987465606, + "learning_rate": 9.455818887657725e-05, + "loss": 2.0342, + "step": 5948 + }, + { + "epoch": 0.5319206008583691, + "grad_norm": 0.5426908516127975, + "learning_rate": 9.452927190793566e-05, + "loss": 2.1226, + "step": 5949 + }, + { + "epoch": 0.5320100143061517, + "grad_norm": 3.904837689965429, + "learning_rate": 9.450035539811741e-05, + "loss": 2.1286, + "step": 5950 + }, + { + "epoch": 0.5320994277539342, + "grad_norm": 1.6859698329406378, + "learning_rate": 9.447143934954771e-05, + "loss": 1.5052, + "step": 5951 + }, + { + "epoch": 0.5321888412017167, + "grad_norm": 4.047196530798027, + "learning_rate": 9.444252376465171e-05, + "loss": 1.7366, + "step": 5952 + }, + { + "epoch": 0.5322782546494993, + "grad_norm": 2.5328333918709824, + "learning_rate": 9.441360864585456e-05, + "loss": 2.1145, + "step": 5953 + }, + { + "epoch": 0.5323676680972819, + "grad_norm": 2.077585160651717, + "learning_rate": 9.438469399558128e-05, + "loss": 1.7825, + "step": 5954 + }, + { + "epoch": 0.5324570815450643, + "grad_norm": 1.071758066379605, + "learning_rate": 9.435577981625697e-05, + "loss": 2.1446, + "step": 5955 + }, + { + "epoch": 0.5325464949928469, + "grad_norm": 2.441175719578654, + "learning_rate": 9.432686611030657e-05, + "loss": 1.5757, + "step": 5956 + }, + { + "epoch": 0.5326359084406295, + "grad_norm": 1.9068820692031272, + "learning_rate": 9.429795288015504e-05, + "loss": 1.6033, + "step": 5957 + }, + { + "epoch": 0.532725321888412, + "grad_norm": 1.678987412698081, + "learning_rate": 9.42690401282273e-05, + "loss": 1.5246, + "step": 5958 + }, + { + "epoch": 0.5328147353361946, + "grad_norm": 1.540862179792111, + "learning_rate": 9.424012785694827e-05, + "loss": 1.9524, + "step": 5959 + }, + { + "epoch": 0.5329041487839771, + "grad_norm": 0.5470422689686006, + "learning_rate": 9.421121606874278e-05, + "loss": 1.5055, + "step": 5960 + }, + { + "epoch": 0.5329935622317596, + "grad_norm": 1.1087082258239707, + "learning_rate": 9.418230476603558e-05, + "loss": 1.6096, + "step": 5961 + }, + { + "epoch": 0.5330829756795422, + "grad_norm": 1.1877984615194455, + "learning_rate": 9.415339395125147e-05, + "loss": 1.7104, + "step": 5962 + }, + { + "epoch": 0.5331723891273248, + "grad_norm": 1.0932295390502749, + "learning_rate": 9.412448362681516e-05, + "loss": 1.5219, + "step": 5963 + }, + { + "epoch": 0.5332618025751072, + "grad_norm": 0.9265580041853293, + "learning_rate": 9.409557379515127e-05, + "loss": 1.7226, + "step": 5964 + }, + { + "epoch": 0.5333512160228898, + "grad_norm": 1.0720891292688208, + "learning_rate": 9.406666445868448e-05, + "loss": 2.0075, + "step": 5965 + }, + { + "epoch": 0.5334406294706724, + "grad_norm": 1.4903006004872301, + "learning_rate": 9.40377556198394e-05, + "loss": 1.8343, + "step": 5966 + }, + { + "epoch": 0.533530042918455, + "grad_norm": 0.8727582405820333, + "learning_rate": 9.400884728104056e-05, + "loss": 1.9546, + "step": 5967 + }, + { + "epoch": 0.5336194563662375, + "grad_norm": 1.0525281484141504, + "learning_rate": 9.397993944471244e-05, + "loss": 1.8116, + "step": 5968 + }, + { + "epoch": 0.53370886981402, + "grad_norm": 0.3752490694669083, + "learning_rate": 9.395103211327955e-05, + "loss": 1.7323, + "step": 5969 + }, + { + "epoch": 0.5337982832618026, + "grad_norm": 0.5963980822715453, + "learning_rate": 9.39221252891663e-05, + "loss": 1.5959, + "step": 5970 + }, + { + "epoch": 0.5338876967095851, + "grad_norm": 1.0182367218875081, + "learning_rate": 9.389321897479703e-05, + "loss": 1.7056, + "step": 5971 + }, + { + "epoch": 0.5339771101573677, + "grad_norm": 1.4847814145463234, + "learning_rate": 9.386431317259609e-05, + "loss": 1.8142, + "step": 5972 + }, + { + "epoch": 0.5340665236051502, + "grad_norm": 0.6496879741482922, + "learning_rate": 9.383540788498784e-05, + "loss": 2.1509, + "step": 5973 + }, + { + "epoch": 0.5341559370529327, + "grad_norm": 2.433564382629384, + "learning_rate": 9.380650311439649e-05, + "loss": 1.848, + "step": 5974 + }, + { + "epoch": 0.5342453505007153, + "grad_norm": 0.9814394395670941, + "learning_rate": 9.37775988632462e-05, + "loss": 1.9239, + "step": 5975 + }, + { + "epoch": 0.5343347639484979, + "grad_norm": 1.4902775606157148, + "learning_rate": 9.374869513396123e-05, + "loss": 2.0256, + "step": 5976 + }, + { + "epoch": 0.5344241773962805, + "grad_norm": 1.554384740888087, + "learning_rate": 9.371979192896564e-05, + "loss": 1.9635, + "step": 5977 + }, + { + "epoch": 0.5345135908440629, + "grad_norm": 0.8412653501450313, + "learning_rate": 9.369088925068347e-05, + "loss": 1.7745, + "step": 5978 + }, + { + "epoch": 0.5346030042918455, + "grad_norm": 0.8108852394646998, + "learning_rate": 9.366198710153882e-05, + "loss": 1.5738, + "step": 5979 + }, + { + "epoch": 0.5346924177396281, + "grad_norm": 1.1526976559343467, + "learning_rate": 9.363308548395568e-05, + "loss": 1.8863, + "step": 5980 + }, + { + "epoch": 0.5347818311874106, + "grad_norm": 0.6975079026434479, + "learning_rate": 9.360418440035796e-05, + "loss": 1.7921, + "step": 5981 + }, + { + "epoch": 0.5348712446351931, + "grad_norm": 0.6704273208845039, + "learning_rate": 9.357528385316958e-05, + "loss": 1.8147, + "step": 5982 + }, + { + "epoch": 0.5349606580829757, + "grad_norm": 1.4357827476035379, + "learning_rate": 9.354638384481437e-05, + "loss": 1.5556, + "step": 5983 + }, + { + "epoch": 0.5350500715307582, + "grad_norm": 0.6520713161520594, + "learning_rate": 9.351748437771615e-05, + "loss": 1.6422, + "step": 5984 + }, + { + "epoch": 0.5351394849785408, + "grad_norm": 1.1887482870718624, + "learning_rate": 9.348858545429868e-05, + "loss": 1.6506, + "step": 5985 + }, + { + "epoch": 0.5352288984263234, + "grad_norm": 1.0147032371334745, + "learning_rate": 9.345968707698569e-05, + "loss": 1.9175, + "step": 5986 + }, + { + "epoch": 0.5353183118741058, + "grad_norm": 1.741489294401105, + "learning_rate": 9.343078924820087e-05, + "loss": 2.2435, + "step": 5987 + }, + { + "epoch": 0.5354077253218884, + "grad_norm": 1.4570400851126901, + "learning_rate": 9.340189197036779e-05, + "loss": 1.8322, + "step": 5988 + }, + { + "epoch": 0.535497138769671, + "grad_norm": 0.9842640542882807, + "learning_rate": 9.337299524591009e-05, + "loss": 1.3892, + "step": 5989 + }, + { + "epoch": 0.5355865522174535, + "grad_norm": 1.1652776050161555, + "learning_rate": 9.334409907725128e-05, + "loss": 1.8711, + "step": 5990 + }, + { + "epoch": 0.535675965665236, + "grad_norm": 2.2329732480784927, + "learning_rate": 9.33152034668148e-05, + "loss": 1.7355, + "step": 5991 + }, + { + "epoch": 0.5357653791130186, + "grad_norm": 1.844425510370239, + "learning_rate": 9.328630841702414e-05, + "loss": 2.1002, + "step": 5992 + }, + { + "epoch": 0.5358547925608012, + "grad_norm": 2.9849584555260056, + "learning_rate": 9.32574139303027e-05, + "loss": 2.0001, + "step": 5993 + }, + { + "epoch": 0.5359442060085837, + "grad_norm": 1.504360077153875, + "learning_rate": 9.322852000907383e-05, + "loss": 1.8782, + "step": 5994 + }, + { + "epoch": 0.5360336194563662, + "grad_norm": 3.528769484335482, + "learning_rate": 9.319962665576078e-05, + "loss": 2.104, + "step": 5995 + }, + { + "epoch": 0.5361230329041488, + "grad_norm": 1.914456335197087, + "learning_rate": 9.317073387278686e-05, + "loss": 2.0723, + "step": 5996 + }, + { + "epoch": 0.5362124463519313, + "grad_norm": 1.0760373766131655, + "learning_rate": 9.314184166257524e-05, + "loss": 1.3649, + "step": 5997 + }, + { + "epoch": 0.5363018597997139, + "grad_norm": 3.4034216738562746, + "learning_rate": 9.311295002754905e-05, + "loss": 1.9983, + "step": 5998 + }, + { + "epoch": 0.5363912732474965, + "grad_norm": 1.5270757355912699, + "learning_rate": 9.30840589701314e-05, + "loss": 2.098, + "step": 5999 + }, + { + "epoch": 0.5364806866952789, + "grad_norm": 2.918919099569353, + "learning_rate": 9.305516849274541e-05, + "loss": 1.8459, + "step": 6000 + }, + { + "epoch": 0.5365701001430615, + "grad_norm": 2.0014912036995507, + "learning_rate": 9.302627859781406e-05, + "loss": 1.7661, + "step": 6001 + }, + { + "epoch": 0.5366595135908441, + "grad_norm": 1.2114234402575048, + "learning_rate": 9.299738928776029e-05, + "loss": 1.6068, + "step": 6002 + }, + { + "epoch": 0.5367489270386266, + "grad_norm": 1.4760096257274735, + "learning_rate": 9.296850056500703e-05, + "loss": 2.0143, + "step": 6003 + }, + { + "epoch": 0.5368383404864091, + "grad_norm": 2.3448505630383307, + "learning_rate": 9.293961243197715e-05, + "loss": 2.0849, + "step": 6004 + }, + { + "epoch": 0.5369277539341917, + "grad_norm": 1.8530197542974687, + "learning_rate": 9.29107248910934e-05, + "loss": 2.0223, + "step": 6005 + }, + { + "epoch": 0.5370171673819742, + "grad_norm": 2.340267129897915, + "learning_rate": 9.28818379447786e-05, + "loss": 1.6022, + "step": 6006 + }, + { + "epoch": 0.5371065808297568, + "grad_norm": 2.059434298667918, + "learning_rate": 9.285295159545547e-05, + "loss": 1.7028, + "step": 6007 + }, + { + "epoch": 0.5371959942775394, + "grad_norm": 1.8369263313960544, + "learning_rate": 9.282406584554668e-05, + "loss": 1.714, + "step": 6008 + }, + { + "epoch": 0.5372854077253219, + "grad_norm": 0.8968551933283151, + "learning_rate": 9.279518069747479e-05, + "loss": 1.9279, + "step": 6009 + }, + { + "epoch": 0.5373748211731044, + "grad_norm": 0.9125819676443838, + "learning_rate": 9.276629615366242e-05, + "loss": 1.5272, + "step": 6010 + }, + { + "epoch": 0.537464234620887, + "grad_norm": 2.258250579036234, + "learning_rate": 9.273741221653204e-05, + "loss": 1.7629, + "step": 6011 + }, + { + "epoch": 0.5375536480686696, + "grad_norm": 1.5032578576802955, + "learning_rate": 9.270852888850615e-05, + "loss": 1.7287, + "step": 6012 + }, + { + "epoch": 0.537643061516452, + "grad_norm": 1.5240222799463943, + "learning_rate": 9.267964617200707e-05, + "loss": 1.7501, + "step": 6013 + }, + { + "epoch": 0.5377324749642346, + "grad_norm": 1.5511452196966429, + "learning_rate": 9.265076406945727e-05, + "loss": 1.8936, + "step": 6014 + }, + { + "epoch": 0.5378218884120172, + "grad_norm": 0.8794056563634832, + "learning_rate": 9.262188258327901e-05, + "loss": 2.2615, + "step": 6015 + }, + { + "epoch": 0.5379113018597997, + "grad_norm": 1.213184307777675, + "learning_rate": 9.259300171589456e-05, + "loss": 1.5129, + "step": 6016 + }, + { + "epoch": 0.5380007153075823, + "grad_norm": 1.555089834196486, + "learning_rate": 9.256412146972611e-05, + "loss": 1.6676, + "step": 6017 + }, + { + "epoch": 0.5380901287553648, + "grad_norm": 1.3032193892747639, + "learning_rate": 9.25352418471958e-05, + "loss": 1.8485, + "step": 6018 + }, + { + "epoch": 0.5381795422031473, + "grad_norm": 1.2419312270706044, + "learning_rate": 9.250636285072574e-05, + "loss": 1.579, + "step": 6019 + }, + { + "epoch": 0.5382689556509299, + "grad_norm": 2.065323803556897, + "learning_rate": 9.247748448273796e-05, + "loss": 2.134, + "step": 6020 + }, + { + "epoch": 0.5383583690987125, + "grad_norm": 1.913219866203495, + "learning_rate": 9.24486067456545e-05, + "loss": 1.9699, + "step": 6021 + }, + { + "epoch": 0.538447782546495, + "grad_norm": 1.8075236278546811, + "learning_rate": 9.241972964189726e-05, + "loss": 1.5424, + "step": 6022 + }, + { + "epoch": 0.5385371959942775, + "grad_norm": 1.746975993335768, + "learning_rate": 9.239085317388816e-05, + "loss": 1.8632, + "step": 6023 + }, + { + "epoch": 0.5386266094420601, + "grad_norm": 1.3493071295898704, + "learning_rate": 9.236197734404901e-05, + "loss": 1.8378, + "step": 6024 + }, + { + "epoch": 0.5387160228898427, + "grad_norm": 2.613417929655143, + "learning_rate": 9.233310215480157e-05, + "loss": 1.8377, + "step": 6025 + }, + { + "epoch": 0.5388054363376252, + "grad_norm": 1.031174991550182, + "learning_rate": 9.230422760856762e-05, + "loss": 1.6787, + "step": 6026 + }, + { + "epoch": 0.5388948497854077, + "grad_norm": 1.858333855822533, + "learning_rate": 9.227535370776877e-05, + "loss": 1.9491, + "step": 6027 + }, + { + "epoch": 0.5389842632331903, + "grad_norm": 1.8374611407993318, + "learning_rate": 9.22464804548267e-05, + "loss": 1.5961, + "step": 6028 + }, + { + "epoch": 0.5390736766809728, + "grad_norm": 0.7901810722728144, + "learning_rate": 9.221760785216295e-05, + "loss": 1.8753, + "step": 6029 + }, + { + "epoch": 0.5391630901287554, + "grad_norm": 3.065699838798716, + "learning_rate": 9.218873590219905e-05, + "loss": 1.9589, + "step": 6030 + }, + { + "epoch": 0.5392525035765379, + "grad_norm": 0.6897570371108304, + "learning_rate": 9.215986460735642e-05, + "loss": 1.914, + "step": 6031 + }, + { + "epoch": 0.5393419170243204, + "grad_norm": 2.3036114544350657, + "learning_rate": 9.213099397005646e-05, + "loss": 1.7983, + "step": 6032 + }, + { + "epoch": 0.539431330472103, + "grad_norm": 1.1890367160305357, + "learning_rate": 9.210212399272056e-05, + "loss": 1.5074, + "step": 6033 + }, + { + "epoch": 0.5395207439198856, + "grad_norm": 0.9540695388254871, + "learning_rate": 9.207325467776993e-05, + "loss": 1.8411, + "step": 6034 + }, + { + "epoch": 0.539610157367668, + "grad_norm": 1.591704125972033, + "learning_rate": 9.204438602762592e-05, + "loss": 2.0987, + "step": 6035 + }, + { + "epoch": 0.5396995708154506, + "grad_norm": 1.200385512163004, + "learning_rate": 9.201551804470962e-05, + "loss": 1.8148, + "step": 6036 + }, + { + "epoch": 0.5397889842632332, + "grad_norm": 1.0796535359293198, + "learning_rate": 9.198665073144218e-05, + "loss": 1.7089, + "step": 6037 + }, + { + "epoch": 0.5398783977110158, + "grad_norm": 1.7903561712280587, + "learning_rate": 9.195778409024468e-05, + "loss": 1.7659, + "step": 6038 + }, + { + "epoch": 0.5399678111587983, + "grad_norm": 0.41614885005760194, + "learning_rate": 9.19289181235381e-05, + "loss": 1.3788, + "step": 6039 + }, + { + "epoch": 0.5400572246065808, + "grad_norm": 0.9478980654101963, + "learning_rate": 9.190005283374343e-05, + "loss": 1.8348, + "step": 6040 + }, + { + "epoch": 0.5401466380543634, + "grad_norm": 0.7567571247585867, + "learning_rate": 9.187118822328149e-05, + "loss": 2.0604, + "step": 6041 + }, + { + "epoch": 0.5402360515021459, + "grad_norm": 0.8467766868141854, + "learning_rate": 9.184232429457323e-05, + "loss": 1.9364, + "step": 6042 + }, + { + "epoch": 0.5403254649499285, + "grad_norm": 0.4779308288602207, + "learning_rate": 9.181346105003936e-05, + "loss": 1.4503, + "step": 6043 + }, + { + "epoch": 0.540414878397711, + "grad_norm": 1.250155774814123, + "learning_rate": 9.178459849210063e-05, + "loss": 1.8708, + "step": 6044 + }, + { + "epoch": 0.5405042918454935, + "grad_norm": 1.0860345539922536, + "learning_rate": 9.175573662317769e-05, + "loss": 1.6598, + "step": 6045 + }, + { + "epoch": 0.5405937052932761, + "grad_norm": 1.1469950426046112, + "learning_rate": 9.172687544569118e-05, + "loss": 1.7389, + "step": 6046 + }, + { + "epoch": 0.5406831187410587, + "grad_norm": 1.6686937658866665, + "learning_rate": 9.169801496206165e-05, + "loss": 1.8755, + "step": 6047 + }, + { + "epoch": 0.5407725321888412, + "grad_norm": 0.9801445406598503, + "learning_rate": 9.166915517470953e-05, + "loss": 2.0258, + "step": 6048 + }, + { + "epoch": 0.5408619456366237, + "grad_norm": 2.2718019334193684, + "learning_rate": 9.164029608605531e-05, + "loss": 1.6175, + "step": 6049 + }, + { + "epoch": 0.5409513590844063, + "grad_norm": 2.1831117220280527, + "learning_rate": 9.161143769851941e-05, + "loss": 1.7165, + "step": 6050 + }, + { + "epoch": 0.5410407725321889, + "grad_norm": 1.865450828716928, + "learning_rate": 9.158258001452208e-05, + "loss": 1.974, + "step": 6051 + }, + { + "epoch": 0.5411301859799714, + "grad_norm": 2.6767599253099035, + "learning_rate": 9.155372303648359e-05, + "loss": 1.5207, + "step": 6052 + }, + { + "epoch": 0.5412195994277539, + "grad_norm": 1.2357587735957474, + "learning_rate": 9.152486676682415e-05, + "loss": 2.0754, + "step": 6053 + }, + { + "epoch": 0.5413090128755365, + "grad_norm": 2.6882400429104862, + "learning_rate": 9.149601120796391e-05, + "loss": 2.043, + "step": 6054 + }, + { + "epoch": 0.541398426323319, + "grad_norm": 2.2154618161067243, + "learning_rate": 9.146715636232291e-05, + "loss": 2.0314, + "step": 6055 + }, + { + "epoch": 0.5414878397711016, + "grad_norm": 1.5633869286877362, + "learning_rate": 9.14383022323212e-05, + "loss": 2.0054, + "step": 6056 + }, + { + "epoch": 0.5415772532188842, + "grad_norm": 2.9757796632754085, + "learning_rate": 9.140944882037879e-05, + "loss": 1.3996, + "step": 6057 + }, + { + "epoch": 0.5416666666666666, + "grad_norm": 9.164394610192474, + "learning_rate": 9.138059612891551e-05, + "loss": 1.6522, + "step": 6058 + }, + { + "epoch": 0.5417560801144492, + "grad_norm": 16.68519881727399, + "learning_rate": 9.13517441603512e-05, + "loss": 2.0322, + "step": 6059 + }, + { + "epoch": 0.5418454935622318, + "grad_norm": 1.903978079190418, + "learning_rate": 9.13228929171057e-05, + "loss": 2.0506, + "step": 6060 + }, + { + "epoch": 0.5419349070100143, + "grad_norm": 21.5201540050972, + "learning_rate": 9.129404240159864e-05, + "loss": 1.8921, + "step": 6061 + }, + { + "epoch": 0.5420243204577968, + "grad_norm": 7.925047000144746, + "learning_rate": 9.126519261624977e-05, + "loss": 2.0681, + "step": 6062 + }, + { + "epoch": 0.5421137339055794, + "grad_norm": 7.721922590398002, + "learning_rate": 9.123634356347863e-05, + "loss": 1.7028, + "step": 6063 + }, + { + "epoch": 0.542203147353362, + "grad_norm": 3.63972736344291, + "learning_rate": 9.12074952457048e-05, + "loss": 1.9784, + "step": 6064 + }, + { + "epoch": 0.5422925608011445, + "grad_norm": 2.232520947032164, + "learning_rate": 9.117864766534772e-05, + "loss": 1.5338, + "step": 6065 + }, + { + "epoch": 0.5423819742489271, + "grad_norm": 0.9854186364681472, + "learning_rate": 9.114980082482677e-05, + "loss": 1.6123, + "step": 6066 + }, + { + "epoch": 0.5424713876967096, + "grad_norm": 3.9643611234687164, + "learning_rate": 9.112095472656137e-05, + "loss": 2.1018, + "step": 6067 + }, + { + "epoch": 0.5425608011444921, + "grad_norm": 2.4781541460946954, + "learning_rate": 9.109210937297074e-05, + "loss": 1.8953, + "step": 6068 + }, + { + "epoch": 0.5426502145922747, + "grad_norm": 0.8414124400430669, + "learning_rate": 9.106326476647417e-05, + "loss": 2.0489, + "step": 6069 + }, + { + "epoch": 0.5427396280400573, + "grad_norm": 1.4014419972831644, + "learning_rate": 9.103442090949077e-05, + "loss": 1.5131, + "step": 6070 + }, + { + "epoch": 0.5428290414878397, + "grad_norm": 4.7683877977560805, + "learning_rate": 9.100557780443968e-05, + "loss": 2.1594, + "step": 6071 + }, + { + "epoch": 0.5429184549356223, + "grad_norm": 1.1648001406879207, + "learning_rate": 9.09767354537399e-05, + "loss": 1.5979, + "step": 6072 + }, + { + "epoch": 0.5430078683834049, + "grad_norm": 3.2364097133908096, + "learning_rate": 9.094789385981045e-05, + "loss": 2.055, + "step": 6073 + }, + { + "epoch": 0.5430972818311874, + "grad_norm": 1.5725029107825488, + "learning_rate": 9.09190530250702e-05, + "loss": 1.5517, + "step": 6074 + }, + { + "epoch": 0.54318669527897, + "grad_norm": 1.0958999852532023, + "learning_rate": 9.089021295193796e-05, + "loss": 1.8413, + "step": 6075 + }, + { + "epoch": 0.5432761087267525, + "grad_norm": 2.1655230556114313, + "learning_rate": 9.08613736428326e-05, + "loss": 1.7079, + "step": 6076 + }, + { + "epoch": 0.543365522174535, + "grad_norm": 0.9514864002087559, + "learning_rate": 9.083253510017279e-05, + "loss": 1.9632, + "step": 6077 + }, + { + "epoch": 0.5434549356223176, + "grad_norm": 1.5103763133543793, + "learning_rate": 9.08036973263772e-05, + "loss": 1.5329, + "step": 6078 + }, + { + "epoch": 0.5435443490701002, + "grad_norm": 2.940262683692675, + "learning_rate": 9.077486032386439e-05, + "loss": 1.7722, + "step": 6079 + }, + { + "epoch": 0.5436337625178826, + "grad_norm": 0.7571900512619761, + "learning_rate": 9.074602409505293e-05, + "loss": 2.0193, + "step": 6080 + }, + { + "epoch": 0.5437231759656652, + "grad_norm": 2.2048462781199127, + "learning_rate": 9.071718864236125e-05, + "loss": 1.9527, + "step": 6081 + }, + { + "epoch": 0.5438125894134478, + "grad_norm": 3.223997178393359, + "learning_rate": 9.06883539682077e-05, + "loss": 1.8847, + "step": 6082 + }, + { + "epoch": 0.5439020028612304, + "grad_norm": 0.7169730145304885, + "learning_rate": 9.065952007501067e-05, + "loss": 1.8483, + "step": 6083 + }, + { + "epoch": 0.5439914163090128, + "grad_norm": 2.6721130750026623, + "learning_rate": 9.063068696518843e-05, + "loss": 2.2479, + "step": 6084 + }, + { + "epoch": 0.5440808297567954, + "grad_norm": 3.7776305528965755, + "learning_rate": 9.060185464115918e-05, + "loss": 2.1718, + "step": 6085 + }, + { + "epoch": 0.544170243204578, + "grad_norm": 1.7705678486485663, + "learning_rate": 9.0573023105341e-05, + "loss": 1.7994, + "step": 6086 + }, + { + "epoch": 0.5442596566523605, + "grad_norm": 2.2013611800908217, + "learning_rate": 9.054419236015201e-05, + "loss": 1.8681, + "step": 6087 + }, + { + "epoch": 0.5443490701001431, + "grad_norm": 2.523537942178017, + "learning_rate": 9.05153624080102e-05, + "loss": 2.2578, + "step": 6088 + }, + { + "epoch": 0.5444384835479256, + "grad_norm": 1.6066390158587074, + "learning_rate": 9.048653325133343e-05, + "loss": 2.0925, + "step": 6089 + }, + { + "epoch": 0.5445278969957081, + "grad_norm": 4.81203824375336, + "learning_rate": 9.045770489253965e-05, + "loss": 1.6838, + "step": 6090 + }, + { + "epoch": 0.5446173104434907, + "grad_norm": 1.1726697770084742, + "learning_rate": 9.042887733404666e-05, + "loss": 1.6604, + "step": 6091 + }, + { + "epoch": 0.5447067238912733, + "grad_norm": 3.1702920394395973, + "learning_rate": 9.040005057827216e-05, + "loss": 1.454, + "step": 6092 + }, + { + "epoch": 0.5447961373390557, + "grad_norm": 3.441281955811958, + "learning_rate": 9.037122462763383e-05, + "loss": 1.836, + "step": 6093 + }, + { + "epoch": 0.5448855507868383, + "grad_norm": 0.6909442535406433, + "learning_rate": 9.034239948454925e-05, + "loss": 1.7566, + "step": 6094 + }, + { + "epoch": 0.5449749642346209, + "grad_norm": 3.50836842247887, + "learning_rate": 9.031357515143599e-05, + "loss": 1.8194, + "step": 6095 + }, + { + "epoch": 0.5450643776824035, + "grad_norm": 2.215544018961142, + "learning_rate": 9.028475163071141e-05, + "loss": 1.7536, + "step": 6096 + }, + { + "epoch": 0.545153791130186, + "grad_norm": 1.336245451254834, + "learning_rate": 9.025592892479303e-05, + "loss": 1.7519, + "step": 6097 + }, + { + "epoch": 0.5452432045779685, + "grad_norm": 2.698875306034751, + "learning_rate": 9.022710703609814e-05, + "loss": 1.8412, + "step": 6098 + }, + { + "epoch": 0.5453326180257511, + "grad_norm": 3.4283414014149187, + "learning_rate": 9.019828596704394e-05, + "loss": 1.7635, + "step": 6099 + }, + { + "epoch": 0.5454220314735336, + "grad_norm": 2.639055162403924, + "learning_rate": 9.01694657200477e-05, + "loss": 2.132, + "step": 6100 + }, + { + "epoch": 0.5455114449213162, + "grad_norm": 7.586768769786016, + "learning_rate": 9.014064629752647e-05, + "loss": 1.658, + "step": 6101 + }, + { + "epoch": 0.5456008583690987, + "grad_norm": 1.6310007723424813, + "learning_rate": 9.011182770189733e-05, + "loss": 2.1127, + "step": 6102 + }, + { + "epoch": 0.5456902718168812, + "grad_norm": 3.252068075286143, + "learning_rate": 9.008300993557723e-05, + "loss": 1.878, + "step": 6103 + }, + { + "epoch": 0.5457796852646638, + "grad_norm": 2.60443653814799, + "learning_rate": 9.005419300098316e-05, + "loss": 2.0211, + "step": 6104 + }, + { + "epoch": 0.5458690987124464, + "grad_norm": 2.4894974121549924, + "learning_rate": 9.002537690053191e-05, + "loss": 1.6125, + "step": 6105 + }, + { + "epoch": 0.545958512160229, + "grad_norm": 3.800848921400684, + "learning_rate": 8.999656163664023e-05, + "loss": 1.7943, + "step": 6106 + }, + { + "epoch": 0.5460479256080114, + "grad_norm": 1.200311578063266, + "learning_rate": 8.996774721172487e-05, + "loss": 1.6061, + "step": 6107 + }, + { + "epoch": 0.546137339055794, + "grad_norm": 1.533305588845268, + "learning_rate": 8.993893362820241e-05, + "loss": 2.0044, + "step": 6108 + }, + { + "epoch": 0.5462267525035766, + "grad_norm": 2.6232044386290463, + "learning_rate": 8.991012088848944e-05, + "loss": 1.7263, + "step": 6109 + }, + { + "epoch": 0.5463161659513591, + "grad_norm": 1.5642298759091657, + "learning_rate": 8.988130899500243e-05, + "loss": 1.9984, + "step": 6110 + }, + { + "epoch": 0.5464055793991416, + "grad_norm": 0.6183870922975446, + "learning_rate": 8.985249795015784e-05, + "loss": 1.8761, + "step": 6111 + }, + { + "epoch": 0.5464949928469242, + "grad_norm": 2.229223672931685, + "learning_rate": 8.9823687756372e-05, + "loss": 1.9821, + "step": 6112 + }, + { + "epoch": 0.5465844062947067, + "grad_norm": 1.3643207906105543, + "learning_rate": 8.979487841606115e-05, + "loss": 1.803, + "step": 6113 + }, + { + "epoch": 0.5466738197424893, + "grad_norm": 1.1999392301688105, + "learning_rate": 8.976606993164155e-05, + "loss": 1.8989, + "step": 6114 + }, + { + "epoch": 0.5467632331902719, + "grad_norm": 1.060175741091908, + "learning_rate": 8.97372623055293e-05, + "loss": 1.8368, + "step": 6115 + }, + { + "epoch": 0.5468526466380543, + "grad_norm": 1.2190915025522668, + "learning_rate": 8.970845554014044e-05, + "loss": 1.653, + "step": 6116 + }, + { + "epoch": 0.5469420600858369, + "grad_norm": 2.3655109550777826, + "learning_rate": 8.967964963789097e-05, + "loss": 1.6754, + "step": 6117 + }, + { + "epoch": 0.5470314735336195, + "grad_norm": 2.1174364808387636, + "learning_rate": 8.965084460119687e-05, + "loss": 2.165, + "step": 6118 + }, + { + "epoch": 0.547120886981402, + "grad_norm": 1.5392432348186842, + "learning_rate": 8.962204043247393e-05, + "loss": 2.0388, + "step": 6119 + }, + { + "epoch": 0.5472103004291845, + "grad_norm": 0.8534591097719173, + "learning_rate": 8.959323713413791e-05, + "loss": 1.6364, + "step": 6120 + }, + { + "epoch": 0.5472997138769671, + "grad_norm": 1.3825064409404697, + "learning_rate": 8.956443470860453e-05, + "loss": 1.5594, + "step": 6121 + }, + { + "epoch": 0.5473891273247496, + "grad_norm": 1.7028041354760581, + "learning_rate": 8.953563315828942e-05, + "loss": 1.8762, + "step": 6122 + }, + { + "epoch": 0.5474785407725322, + "grad_norm": 0.9444443411999103, + "learning_rate": 8.95068324856081e-05, + "loss": 1.7363, + "step": 6123 + }, + { + "epoch": 0.5475679542203148, + "grad_norm": 1.5534633287132464, + "learning_rate": 8.947803269297604e-05, + "loss": 1.8794, + "step": 6124 + }, + { + "epoch": 0.5476573676680973, + "grad_norm": 2.258688405267129, + "learning_rate": 8.944923378280871e-05, + "loss": 2.2193, + "step": 6125 + }, + { + "epoch": 0.5477467811158798, + "grad_norm": 1.0918280271259453, + "learning_rate": 8.942043575752141e-05, + "loss": 1.73, + "step": 6126 + }, + { + "epoch": 0.5478361945636624, + "grad_norm": 2.563554515462673, + "learning_rate": 8.939163861952935e-05, + "loss": 1.4457, + "step": 6127 + }, + { + "epoch": 0.547925608011445, + "grad_norm": 1.7672059466891086, + "learning_rate": 8.936284237124778e-05, + "loss": 2.1406, + "step": 6128 + }, + { + "epoch": 0.5480150214592274, + "grad_norm": 1.1926129634862475, + "learning_rate": 8.933404701509175e-05, + "loss": 1.6621, + "step": 6129 + }, + { + "epoch": 0.54810443490701, + "grad_norm": 1.8069814129159545, + "learning_rate": 8.930525255347634e-05, + "loss": 2.0019, + "step": 6130 + }, + { + "epoch": 0.5481938483547926, + "grad_norm": 1.861110561755898, + "learning_rate": 8.927645898881644e-05, + "loss": 1.4164, + "step": 6131 + }, + { + "epoch": 0.5482832618025751, + "grad_norm": 0.4824479172328434, + "learning_rate": 8.924766632352702e-05, + "loss": 1.8891, + "step": 6132 + }, + { + "epoch": 0.5483726752503576, + "grad_norm": 1.701735904670417, + "learning_rate": 8.92188745600228e-05, + "loss": 1.8898, + "step": 6133 + }, + { + "epoch": 0.5484620886981402, + "grad_norm": 1.0069372314181053, + "learning_rate": 8.919008370071859e-05, + "loss": 1.6093, + "step": 6134 + }, + { + "epoch": 0.5485515021459227, + "grad_norm": 0.7339577606488823, + "learning_rate": 8.916129374802899e-05, + "loss": 2.1981, + "step": 6135 + }, + { + "epoch": 0.5486409155937053, + "grad_norm": 0.8766759998633137, + "learning_rate": 8.913250470436858e-05, + "loss": 1.9059, + "step": 6136 + }, + { + "epoch": 0.5487303290414879, + "grad_norm": 0.5997054803792977, + "learning_rate": 8.910371657215191e-05, + "loss": 1.6481, + "step": 6137 + }, + { + "epoch": 0.5488197424892703, + "grad_norm": 0.6137186303963557, + "learning_rate": 8.907492935379331e-05, + "loss": 1.754, + "step": 6138 + }, + { + "epoch": 0.5489091559370529, + "grad_norm": 0.7410942483517099, + "learning_rate": 8.904614305170724e-05, + "loss": 1.9383, + "step": 6139 + }, + { + "epoch": 0.5489985693848355, + "grad_norm": 1.0776058321756765, + "learning_rate": 8.90173576683079e-05, + "loss": 1.8716, + "step": 6140 + }, + { + "epoch": 0.5490879828326181, + "grad_norm": 0.8030570068380808, + "learning_rate": 8.898857320600952e-05, + "loss": 1.8216, + "step": 6141 + }, + { + "epoch": 0.5491773962804005, + "grad_norm": 0.7160891829983902, + "learning_rate": 8.895978966722623e-05, + "loss": 1.5695, + "step": 6142 + }, + { + "epoch": 0.5492668097281831, + "grad_norm": 2.1122165565609596, + "learning_rate": 8.893100705437201e-05, + "loss": 2.1813, + "step": 6143 + }, + { + "epoch": 0.5493562231759657, + "grad_norm": 1.6843562367824167, + "learning_rate": 8.890222536986085e-05, + "loss": 1.737, + "step": 6144 + }, + { + "epoch": 0.5494456366237482, + "grad_norm": 0.5924058345972537, + "learning_rate": 8.887344461610668e-05, + "loss": 1.9308, + "step": 6145 + }, + { + "epoch": 0.5495350500715308, + "grad_norm": 2.2295415453597327, + "learning_rate": 8.884466479552328e-05, + "loss": 1.8025, + "step": 6146 + }, + { + "epoch": 0.5496244635193133, + "grad_norm": 0.7577624033755483, + "learning_rate": 8.881588591052434e-05, + "loss": 1.8416, + "step": 6147 + }, + { + "epoch": 0.5497138769670958, + "grad_norm": 5.8989003415203145, + "learning_rate": 8.878710796352358e-05, + "loss": 6.6113, + "step": 6148 + }, + { + "epoch": 0.5498032904148784, + "grad_norm": 1.6947057432086605, + "learning_rate": 8.875833095693451e-05, + "loss": 1.4845, + "step": 6149 + }, + { + "epoch": 0.549892703862661, + "grad_norm": 1.5730951938296067, + "learning_rate": 8.872955489317063e-05, + "loss": 1.7372, + "step": 6150 + }, + { + "epoch": 0.5499821173104434, + "grad_norm": 1.461493317903028, + "learning_rate": 8.870077977464537e-05, + "loss": 2.2222, + "step": 6151 + }, + { + "epoch": 0.550071530758226, + "grad_norm": 1.6674630750172068, + "learning_rate": 8.867200560377209e-05, + "loss": 1.784, + "step": 6152 + }, + { + "epoch": 0.5501609442060086, + "grad_norm": 1.63013778585236, + "learning_rate": 8.864323238296401e-05, + "loss": 1.9797, + "step": 6153 + }, + { + "epoch": 0.5502503576537912, + "grad_norm": 0.8158727396920051, + "learning_rate": 8.861446011463432e-05, + "loss": 1.8055, + "step": 6154 + }, + { + "epoch": 0.5503397711015737, + "grad_norm": 1.6078336041926595, + "learning_rate": 8.858568880119611e-05, + "loss": 1.8395, + "step": 6155 + }, + { + "epoch": 0.5504291845493562, + "grad_norm": 0.5905145229254708, + "learning_rate": 8.855691844506238e-05, + "loss": 1.7927, + "step": 6156 + }, + { + "epoch": 0.5505185979971388, + "grad_norm": 0.6091729873785626, + "learning_rate": 8.852814904864611e-05, + "loss": 1.9117, + "step": 6157 + }, + { + "epoch": 0.5506080114449213, + "grad_norm": 1.1282432674051885, + "learning_rate": 8.849938061436006e-05, + "loss": 2.2033, + "step": 6158 + }, + { + "epoch": 0.5506974248927039, + "grad_norm": 1.7024150288250135, + "learning_rate": 8.847061314461714e-05, + "loss": 1.6238, + "step": 6159 + }, + { + "epoch": 0.5507868383404864, + "grad_norm": 0.6268249334706336, + "learning_rate": 8.844184664182993e-05, + "loss": 1.6482, + "step": 6160 + }, + { + "epoch": 0.5508762517882689, + "grad_norm": 0.5541279379052562, + "learning_rate": 8.84130811084111e-05, + "loss": 1.6155, + "step": 6161 + }, + { + "epoch": 0.5509656652360515, + "grad_norm": 0.7592481520219128, + "learning_rate": 8.838431654677317e-05, + "loss": 1.7065, + "step": 6162 + }, + { + "epoch": 0.5510550786838341, + "grad_norm": 1.1384450430307835, + "learning_rate": 8.835555295932857e-05, + "loss": 2.0001, + "step": 6163 + }, + { + "epoch": 0.5511444921316166, + "grad_norm": 0.7351853767317774, + "learning_rate": 8.832679034848969e-05, + "loss": 1.9743, + "step": 6164 + }, + { + "epoch": 0.5512339055793991, + "grad_norm": 1.3558091376666475, + "learning_rate": 8.829802871666877e-05, + "loss": 2.0058, + "step": 6165 + }, + { + "epoch": 0.5513233190271817, + "grad_norm": 0.7721670998218971, + "learning_rate": 8.82692680662781e-05, + "loss": 1.7856, + "step": 6166 + }, + { + "epoch": 0.5514127324749643, + "grad_norm": 0.9733253259423372, + "learning_rate": 8.824050839972973e-05, + "loss": 1.7764, + "step": 6167 + }, + { + "epoch": 0.5515021459227468, + "grad_norm": 0.501723607592881, + "learning_rate": 8.821174971943572e-05, + "loss": 1.8299, + "step": 6168 + }, + { + "epoch": 0.5515915593705293, + "grad_norm": 0.5927110847510152, + "learning_rate": 8.818299202780805e-05, + "loss": 1.941, + "step": 6169 + }, + { + "epoch": 0.5516809728183119, + "grad_norm": 0.4204452856184363, + "learning_rate": 8.815423532725852e-05, + "loss": 1.635, + "step": 6170 + }, + { + "epoch": 0.5517703862660944, + "grad_norm": 1.173958314280935, + "learning_rate": 8.8125479620199e-05, + "loss": 1.6684, + "step": 6171 + }, + { + "epoch": 0.551859799713877, + "grad_norm": 1.0395692831228236, + "learning_rate": 8.809672490904111e-05, + "loss": 1.838, + "step": 6172 + }, + { + "epoch": 0.5519492131616596, + "grad_norm": 1.5012365631683193, + "learning_rate": 8.806797119619658e-05, + "loss": 1.4572, + "step": 6173 + }, + { + "epoch": 0.552038626609442, + "grad_norm": 1.1896793931184246, + "learning_rate": 8.803921848407687e-05, + "loss": 1.6994, + "step": 6174 + }, + { + "epoch": 0.5521280400572246, + "grad_norm": 1.2311302251919165, + "learning_rate": 8.80104667750935e-05, + "loss": 1.8845, + "step": 6175 + }, + { + "epoch": 0.5522174535050072, + "grad_norm": 1.4432106458374674, + "learning_rate": 8.798171607165778e-05, + "loss": 2.1344, + "step": 6176 + }, + { + "epoch": 0.5523068669527897, + "grad_norm": 0.9729310324402471, + "learning_rate": 8.795296637618101e-05, + "loss": 2.0108, + "step": 6177 + }, + { + "epoch": 0.5523962804005722, + "grad_norm": 0.7524358636030458, + "learning_rate": 8.792421769107442e-05, + "loss": 1.979, + "step": 6178 + }, + { + "epoch": 0.5524856938483548, + "grad_norm": 1.9113864770295579, + "learning_rate": 8.789547001874906e-05, + "loss": 1.3406, + "step": 6179 + }, + { + "epoch": 0.5525751072961373, + "grad_norm": 0.8440172650099048, + "learning_rate": 8.786672336161605e-05, + "loss": 1.7928, + "step": 6180 + }, + { + "epoch": 0.5526645207439199, + "grad_norm": 0.9396851774963517, + "learning_rate": 8.783797772208628e-05, + "loss": 1.9657, + "step": 6181 + }, + { + "epoch": 0.5527539341917024, + "grad_norm": 1.2679487785322854, + "learning_rate": 8.780923310257067e-05, + "loss": 2.2429, + "step": 6182 + }, + { + "epoch": 0.552843347639485, + "grad_norm": 1.084427483530512, + "learning_rate": 8.778048950547994e-05, + "loss": 2.0654, + "step": 6183 + }, + { + "epoch": 0.5529327610872675, + "grad_norm": 1.2153284021319197, + "learning_rate": 8.775174693322478e-05, + "loss": 2.0237, + "step": 6184 + }, + { + "epoch": 0.5530221745350501, + "grad_norm": 1.2718103637994234, + "learning_rate": 8.772300538821583e-05, + "loss": 1.5977, + "step": 6185 + }, + { + "epoch": 0.5531115879828327, + "grad_norm": 1.1556203015426387, + "learning_rate": 8.769426487286356e-05, + "loss": 1.9843, + "step": 6186 + }, + { + "epoch": 0.5532010014306151, + "grad_norm": 1.2960948690803678, + "learning_rate": 8.766552538957846e-05, + "loss": 2.0611, + "step": 6187 + }, + { + "epoch": 0.5532904148783977, + "grad_norm": 1.6047894675165595, + "learning_rate": 8.763678694077083e-05, + "loss": 2.2893, + "step": 6188 + }, + { + "epoch": 0.5533798283261803, + "grad_norm": 10.531514269716405, + "learning_rate": 8.760804952885098e-05, + "loss": 6.583, + "step": 6189 + }, + { + "epoch": 0.5534692417739628, + "grad_norm": 4.867058305127961, + "learning_rate": 8.757931315622903e-05, + "loss": 1.7151, + "step": 6190 + }, + { + "epoch": 0.5535586552217453, + "grad_norm": 3.4378467688461725, + "learning_rate": 8.755057782531509e-05, + "loss": 1.858, + "step": 6191 + }, + { + "epoch": 0.5536480686695279, + "grad_norm": 2.1575426322106392, + "learning_rate": 8.752184353851916e-05, + "loss": 2.1863, + "step": 6192 + }, + { + "epoch": 0.5537374821173104, + "grad_norm": 3.586230304266038, + "learning_rate": 8.749311029825111e-05, + "loss": 1.3837, + "step": 6193 + }, + { + "epoch": 0.553826895565093, + "grad_norm": 2.019407971648632, + "learning_rate": 8.74643781069208e-05, + "loss": 1.8994, + "step": 6194 + }, + { + "epoch": 0.5539163090128756, + "grad_norm": 2.3225017119260563, + "learning_rate": 8.7435646966938e-05, + "loss": 2.0414, + "step": 6195 + }, + { + "epoch": 0.554005722460658, + "grad_norm": 2.040475198105533, + "learning_rate": 8.74069168807123e-05, + "loss": 1.6202, + "step": 6196 + }, + { + "epoch": 0.5540951359084406, + "grad_norm": 1.199647619727308, + "learning_rate": 8.737818785065326e-05, + "loss": 1.888, + "step": 6197 + }, + { + "epoch": 0.5541845493562232, + "grad_norm": 1.01088859045883, + "learning_rate": 8.734945987917038e-05, + "loss": 1.8032, + "step": 6198 + }, + { + "epoch": 0.5542739628040058, + "grad_norm": 2.3420077048313326, + "learning_rate": 8.732073296867303e-05, + "loss": 1.9945, + "step": 6199 + }, + { + "epoch": 0.5543633762517882, + "grad_norm": 1.0628297588004836, + "learning_rate": 8.729200712157043e-05, + "loss": 2.1111, + "step": 6200 + }, + { + "epoch": 0.5544527896995708, + "grad_norm": 1.5544136287914176, + "learning_rate": 8.726328234027188e-05, + "loss": 1.4406, + "step": 6201 + }, + { + "epoch": 0.5545422031473534, + "grad_norm": 4.096266731104351, + "learning_rate": 8.723455862718649e-05, + "loss": 2.1217, + "step": 6202 + }, + { + "epoch": 0.5546316165951359, + "grad_norm": 1.2192016336039513, + "learning_rate": 8.720583598472322e-05, + "loss": 1.7542, + "step": 6203 + }, + { + "epoch": 0.5547210300429185, + "grad_norm": 2.9361605468735275, + "learning_rate": 8.717711441529104e-05, + "loss": 1.9051, + "step": 6204 + }, + { + "epoch": 0.554810443490701, + "grad_norm": 2.538091099871715, + "learning_rate": 8.71483939212988e-05, + "loss": 2.0632, + "step": 6205 + }, + { + "epoch": 0.5548998569384835, + "grad_norm": 1.3242370004925483, + "learning_rate": 8.711967450515524e-05, + "loss": 1.7857, + "step": 6206 + }, + { + "epoch": 0.5549892703862661, + "grad_norm": 3.2032598772874215, + "learning_rate": 8.709095616926897e-05, + "loss": 1.6785, + "step": 6207 + }, + { + "epoch": 0.5550786838340487, + "grad_norm": 1.4805275863320069, + "learning_rate": 8.706223891604866e-05, + "loss": 1.5007, + "step": 6208 + }, + { + "epoch": 0.5551680972818311, + "grad_norm": 3.100829537932172, + "learning_rate": 8.703352274790276e-05, + "loss": 1.9269, + "step": 6209 + }, + { + "epoch": 0.5552575107296137, + "grad_norm": 1.95505124257562, + "learning_rate": 8.700480766723964e-05, + "loss": 1.8912, + "step": 6210 + }, + { + "epoch": 0.5553469241773963, + "grad_norm": 1.9552670880782514, + "learning_rate": 8.69760936764676e-05, + "loss": 1.7644, + "step": 6211 + }, + { + "epoch": 0.5554363376251789, + "grad_norm": 2.2951888491650334, + "learning_rate": 8.694738077799488e-05, + "loss": 1.5953, + "step": 6212 + }, + { + "epoch": 0.5555257510729614, + "grad_norm": 1.541066776363617, + "learning_rate": 8.691866897422952e-05, + "loss": 1.5032, + "step": 6213 + }, + { + "epoch": 0.5556151645207439, + "grad_norm": 2.458687975809932, + "learning_rate": 8.688995826757961e-05, + "loss": 1.7558, + "step": 6214 + }, + { + "epoch": 0.5557045779685265, + "grad_norm": 1.7128699919006645, + "learning_rate": 8.686124866045308e-05, + "loss": 1.6954, + "step": 6215 + }, + { + "epoch": 0.555793991416309, + "grad_norm": 3.4047127146245613, + "learning_rate": 8.683254015525776e-05, + "loss": 1.6617, + "step": 6216 + }, + { + "epoch": 0.5558834048640916, + "grad_norm": 1.3149511269255794, + "learning_rate": 8.680383275440138e-05, + "loss": 1.8076, + "step": 6217 + }, + { + "epoch": 0.5559728183118741, + "grad_norm": 1.7971283065736756, + "learning_rate": 8.677512646029163e-05, + "loss": 1.9129, + "step": 6218 + }, + { + "epoch": 0.5560622317596566, + "grad_norm": 2.278282784223133, + "learning_rate": 8.674642127533605e-05, + "loss": 1.6232, + "step": 6219 + }, + { + "epoch": 0.5561516452074392, + "grad_norm": 0.7349907565230496, + "learning_rate": 8.671771720194211e-05, + "loss": 1.632, + "step": 6220 + }, + { + "epoch": 0.5562410586552218, + "grad_norm": 1.8078326610039273, + "learning_rate": 8.668901424251714e-05, + "loss": 1.9393, + "step": 6221 + }, + { + "epoch": 0.5563304721030042, + "grad_norm": 2.1306658782368486, + "learning_rate": 8.666031239946852e-05, + "loss": 1.7263, + "step": 6222 + }, + { + "epoch": 0.5564198855507868, + "grad_norm": 2.315086630700959, + "learning_rate": 8.66316116752034e-05, + "loss": 2.2455, + "step": 6223 + }, + { + "epoch": 0.5565092989985694, + "grad_norm": 0.8555931842946557, + "learning_rate": 8.660291207212882e-05, + "loss": 1.9097, + "step": 6224 + }, + { + "epoch": 0.556598712446352, + "grad_norm": 2.5510515975784513, + "learning_rate": 8.657421359265188e-05, + "loss": 1.9049, + "step": 6225 + }, + { + "epoch": 0.5566881258941345, + "grad_norm": 1.1061577104612468, + "learning_rate": 8.654551623917941e-05, + "loss": 1.5669, + "step": 6226 + }, + { + "epoch": 0.556777539341917, + "grad_norm": 1.3937285122308614, + "learning_rate": 8.651682001411821e-05, + "loss": 1.505, + "step": 6227 + }, + { + "epoch": 0.5568669527896996, + "grad_norm": 1.5439358697590302, + "learning_rate": 8.648812491987504e-05, + "loss": 2.2118, + "step": 6228 + }, + { + "epoch": 0.5569563662374821, + "grad_norm": 1.7374481933736132, + "learning_rate": 8.645943095885655e-05, + "loss": 1.5881, + "step": 6229 + }, + { + "epoch": 0.5570457796852647, + "grad_norm": 1.4227819039680891, + "learning_rate": 8.643073813346922e-05, + "loss": 1.7643, + "step": 6230 + }, + { + "epoch": 0.5571351931330472, + "grad_norm": 1.8306997136628014, + "learning_rate": 8.640204644611948e-05, + "loss": 1.9453, + "step": 6231 + }, + { + "epoch": 0.5572246065808297, + "grad_norm": 1.8524751495353022, + "learning_rate": 8.63733558992137e-05, + "loss": 1.8509, + "step": 6232 + }, + { + "epoch": 0.5573140200286123, + "grad_norm": 1.0787993000800082, + "learning_rate": 8.634466649515811e-05, + "loss": 2.1001, + "step": 6233 + }, + { + "epoch": 0.5574034334763949, + "grad_norm": 16.340712639858452, + "learning_rate": 8.63159782363588e-05, + "loss": 7.2217, + "step": 6234 + }, + { + "epoch": 0.5574928469241774, + "grad_norm": 13.371362253143694, + "learning_rate": 8.62872911252219e-05, + "loss": 1.8641, + "step": 6235 + }, + { + "epoch": 0.5575822603719599, + "grad_norm": 6.04948549886162, + "learning_rate": 8.625860516415335e-05, + "loss": 1.8921, + "step": 6236 + }, + { + "epoch": 0.5576716738197425, + "grad_norm": 3.6174219660201543, + "learning_rate": 8.6229920355559e-05, + "loss": 2.031, + "step": 6237 + }, + { + "epoch": 0.557761087267525, + "grad_norm": 8.531212242706124, + "learning_rate": 8.620123670184455e-05, + "loss": 2.4213, + "step": 6238 + }, + { + "epoch": 0.5578505007153076, + "grad_norm": 5.670496027811528, + "learning_rate": 8.617255420541576e-05, + "loss": 1.7906, + "step": 6239 + }, + { + "epoch": 0.5579399141630901, + "grad_norm": 1.5224356342561147, + "learning_rate": 8.614387286867814e-05, + "loss": 1.8759, + "step": 6240 + }, + { + "epoch": 0.5580293276108726, + "grad_norm": 4.846895340651091, + "learning_rate": 8.611519269403712e-05, + "loss": 1.5483, + "step": 6241 + }, + { + "epoch": 0.5581187410586552, + "grad_norm": 6.449519830310931, + "learning_rate": 8.608651368389815e-05, + "loss": 2.1561, + "step": 6242 + }, + { + "epoch": 0.5582081545064378, + "grad_norm": 1.4102272159545421, + "learning_rate": 8.605783584066649e-05, + "loss": 2.0322, + "step": 6243 + }, + { + "epoch": 0.5582975679542204, + "grad_norm": 4.2200481174107045, + "learning_rate": 8.602915916674731e-05, + "loss": 1.697, + "step": 6244 + }, + { + "epoch": 0.5583869814020028, + "grad_norm": 3.8178561839814695, + "learning_rate": 8.600048366454565e-05, + "loss": 1.637, + "step": 6245 + }, + { + "epoch": 0.5584763948497854, + "grad_norm": 1.9537215578141618, + "learning_rate": 8.597180933646653e-05, + "loss": 1.8285, + "step": 6246 + }, + { + "epoch": 0.558565808297568, + "grad_norm": 1.891101922892752, + "learning_rate": 8.594313618491481e-05, + "loss": 2.0494, + "step": 6247 + }, + { + "epoch": 0.5586552217453505, + "grad_norm": 3.944517777555975, + "learning_rate": 8.591446421229528e-05, + "loss": 2.0778, + "step": 6248 + }, + { + "epoch": 0.558744635193133, + "grad_norm": 2.3940287905460913, + "learning_rate": 8.588579342101263e-05, + "loss": 1.7889, + "step": 6249 + }, + { + "epoch": 0.5588340486409156, + "grad_norm": 19.176329780071477, + "learning_rate": 8.585712381347145e-05, + "loss": 6.8801, + "step": 6250 + }, + { + "epoch": 0.5589234620886981, + "grad_norm": 10.578810099905603, + "learning_rate": 8.58284553920762e-05, + "loss": 1.8781, + "step": 6251 + }, + { + "epoch": 0.5590128755364807, + "grad_norm": 8.627401183064894, + "learning_rate": 8.57997881592313e-05, + "loss": 1.8045, + "step": 6252 + }, + { + "epoch": 0.5591022889842633, + "grad_norm": 2.0281695684817183, + "learning_rate": 8.577112211734104e-05, + "loss": 1.6816, + "step": 6253 + }, + { + "epoch": 0.5591917024320457, + "grad_norm": 6.7743190390600265, + "learning_rate": 8.574245726880953e-05, + "loss": 2.1894, + "step": 6254 + }, + { + "epoch": 0.5592811158798283, + "grad_norm": 9.515449410439357, + "learning_rate": 8.571379361604091e-05, + "loss": 1.7713, + "step": 6255 + }, + { + "epoch": 0.5593705293276109, + "grad_norm": 0.7528787524064231, + "learning_rate": 8.568513116143919e-05, + "loss": 1.5754, + "step": 6256 + }, + { + "epoch": 0.5594599427753935, + "grad_norm": 4.702794982222805, + "learning_rate": 8.565646990740824e-05, + "loss": 1.4137, + "step": 6257 + }, + { + "epoch": 0.5595493562231759, + "grad_norm": 6.485510344655877, + "learning_rate": 8.562780985635183e-05, + "loss": 1.5663, + "step": 6258 + }, + { + "epoch": 0.5596387696709585, + "grad_norm": 3.642733659030622, + "learning_rate": 8.559915101067366e-05, + "loss": 2.0113, + "step": 6259 + }, + { + "epoch": 0.5597281831187411, + "grad_norm": 2.531717236820563, + "learning_rate": 8.55704933727773e-05, + "loss": 1.8908, + "step": 6260 + }, + { + "epoch": 0.5598175965665236, + "grad_norm": 5.462864198783562, + "learning_rate": 8.554183694506622e-05, + "loss": 1.9629, + "step": 6261 + }, + { + "epoch": 0.5599070100143062, + "grad_norm": 2.376443240359283, + "learning_rate": 8.551318172994378e-05, + "loss": 1.7695, + "step": 6262 + }, + { + "epoch": 0.5599964234620887, + "grad_norm": 1.4250058632128337, + "learning_rate": 8.548452772981334e-05, + "loss": 1.9275, + "step": 6263 + }, + { + "epoch": 0.5600858369098712, + "grad_norm": 3.6295967487247616, + "learning_rate": 8.545587494707803e-05, + "loss": 1.5013, + "step": 6264 + }, + { + "epoch": 0.5601752503576538, + "grad_norm": 3.3741082130872213, + "learning_rate": 8.54272233841409e-05, + "loss": 1.9876, + "step": 6265 + }, + { + "epoch": 0.5602646638054364, + "grad_norm": 0.6968004567254262, + "learning_rate": 8.539857304340498e-05, + "loss": 1.6692, + "step": 6266 + }, + { + "epoch": 0.5603540772532188, + "grad_norm": 1.5223004045532786, + "learning_rate": 8.53699239272731e-05, + "loss": 2.1886, + "step": 6267 + }, + { + "epoch": 0.5604434907010014, + "grad_norm": 2.258916842746854, + "learning_rate": 8.5341276038148e-05, + "loss": 2.0468, + "step": 6268 + }, + { + "epoch": 0.560532904148784, + "grad_norm": 2.0879180184115795, + "learning_rate": 8.531262937843236e-05, + "loss": 1.5873, + "step": 6269 + }, + { + "epoch": 0.5606223175965666, + "grad_norm": 0.7889611767529627, + "learning_rate": 8.528398395052879e-05, + "loss": 2.3994, + "step": 6270 + }, + { + "epoch": 0.560711731044349, + "grad_norm": 1.7562821612611539, + "learning_rate": 8.525533975683972e-05, + "loss": 1.6697, + "step": 6271 + }, + { + "epoch": 0.5608011444921316, + "grad_norm": 1.9641253656252637, + "learning_rate": 8.522669679976749e-05, + "loss": 2.2762, + "step": 6272 + }, + { + "epoch": 0.5608905579399142, + "grad_norm": 2.129787596247542, + "learning_rate": 8.519805508171437e-05, + "loss": 1.9034, + "step": 6273 + }, + { + "epoch": 0.5609799713876967, + "grad_norm": 1.136767421241018, + "learning_rate": 8.516941460508247e-05, + "loss": 1.7951, + "step": 6274 + }, + { + "epoch": 0.5610693848354793, + "grad_norm": 2.9712067620715965, + "learning_rate": 8.514077537227388e-05, + "loss": 1.6124, + "step": 6275 + }, + { + "epoch": 0.5611587982832618, + "grad_norm": 0.5787650114921702, + "learning_rate": 8.511213738569046e-05, + "loss": 1.6924, + "step": 6276 + }, + { + "epoch": 0.5612482117310443, + "grad_norm": 1.4306416844909022, + "learning_rate": 8.508350064773415e-05, + "loss": 2.1274, + "step": 6277 + }, + { + "epoch": 0.5613376251788269, + "grad_norm": 1.302533530905165, + "learning_rate": 8.50548651608066e-05, + "loss": 1.9014, + "step": 6278 + }, + { + "epoch": 0.5614270386266095, + "grad_norm": 2.1182867835729273, + "learning_rate": 8.50262309273095e-05, + "loss": 1.5472, + "step": 6279 + }, + { + "epoch": 0.5615164520743919, + "grad_norm": 1.2237863877721442, + "learning_rate": 8.49975979496443e-05, + "loss": 1.7482, + "step": 6280 + }, + { + "epoch": 0.5616058655221745, + "grad_norm": 0.5584536134048779, + "learning_rate": 8.496896623021245e-05, + "loss": 2.0483, + "step": 6281 + }, + { + "epoch": 0.5616952789699571, + "grad_norm": 1.212485025548346, + "learning_rate": 8.494033577141525e-05, + "loss": 1.862, + "step": 6282 + }, + { + "epoch": 0.5617846924177397, + "grad_norm": 1.2840565383648543, + "learning_rate": 8.491170657565386e-05, + "loss": 1.6216, + "step": 6283 + }, + { + "epoch": 0.5618741058655222, + "grad_norm": 1.031509346683812, + "learning_rate": 8.488307864532946e-05, + "loss": 2.1036, + "step": 6284 + }, + { + "epoch": 0.5619635193133047, + "grad_norm": 1.2608296089202589, + "learning_rate": 8.485445198284298e-05, + "loss": 1.8782, + "step": 6285 + }, + { + "epoch": 0.5620529327610873, + "grad_norm": 6.551657462669413, + "learning_rate": 8.482582659059534e-05, + "loss": 7.0099, + "step": 6286 + }, + { + "epoch": 0.5621423462088698, + "grad_norm": 1.6368601334530444, + "learning_rate": 8.47972024709873e-05, + "loss": 1.8991, + "step": 6287 + }, + { + "epoch": 0.5622317596566524, + "grad_norm": 2.094452616639558, + "learning_rate": 8.47685796264195e-05, + "loss": 2.1162, + "step": 6288 + }, + { + "epoch": 0.5623211731044349, + "grad_norm": 1.571351819858423, + "learning_rate": 8.473995805929257e-05, + "loss": 1.8405, + "step": 6289 + }, + { + "epoch": 0.5624105865522174, + "grad_norm": 0.9998142346397425, + "learning_rate": 8.471133777200688e-05, + "loss": 1.9978, + "step": 6290 + }, + { + "epoch": 0.5625, + "grad_norm": 1.7643283281643924, + "learning_rate": 8.468271876696286e-05, + "loss": 1.6866, + "step": 6291 + }, + { + "epoch": 0.5625894134477826, + "grad_norm": 0.671080060975345, + "learning_rate": 8.46541010465607e-05, + "loss": 1.3938, + "step": 6292 + }, + { + "epoch": 0.5626788268955651, + "grad_norm": 1.280422769118469, + "learning_rate": 8.462548461320057e-05, + "loss": 1.566, + "step": 6293 + }, + { + "epoch": 0.5627682403433476, + "grad_norm": 0.558688141699237, + "learning_rate": 8.459686946928249e-05, + "loss": 1.9534, + "step": 6294 + }, + { + "epoch": 0.5628576537911302, + "grad_norm": 0.46794345962823836, + "learning_rate": 8.456825561720634e-05, + "loss": 1.5007, + "step": 6295 + }, + { + "epoch": 0.5629470672389127, + "grad_norm": 1.2936538294517193, + "learning_rate": 8.453964305937197e-05, + "loss": 1.6653, + "step": 6296 + }, + { + "epoch": 0.5630364806866953, + "grad_norm": 1.421994816097851, + "learning_rate": 8.451103179817903e-05, + "loss": 1.9625, + "step": 6297 + }, + { + "epoch": 0.5631258941344778, + "grad_norm": 0.5268636906911727, + "learning_rate": 8.448242183602719e-05, + "loss": 1.6252, + "step": 6298 + }, + { + "epoch": 0.5632153075822603, + "grad_norm": 1.0624190736083443, + "learning_rate": 8.445381317531586e-05, + "loss": 2.0559, + "step": 6299 + }, + { + "epoch": 0.5633047210300429, + "grad_norm": 1.5586913516621335, + "learning_rate": 8.442520581844447e-05, + "loss": 2.0735, + "step": 6300 + }, + { + "epoch": 0.5633941344778255, + "grad_norm": 0.8298060143590774, + "learning_rate": 8.439659976781226e-05, + "loss": 1.973, + "step": 6301 + }, + { + "epoch": 0.5634835479256081, + "grad_norm": 0.6535458716188628, + "learning_rate": 8.436799502581836e-05, + "loss": 1.9558, + "step": 6302 + }, + { + "epoch": 0.5635729613733905, + "grad_norm": 0.8651037201509639, + "learning_rate": 8.433939159486186e-05, + "loss": 1.6453, + "step": 6303 + }, + { + "epoch": 0.5636623748211731, + "grad_norm": 0.6200731529909367, + "learning_rate": 8.431078947734164e-05, + "loss": 1.7897, + "step": 6304 + }, + { + "epoch": 0.5637517882689557, + "grad_norm": 0.4748885247728283, + "learning_rate": 8.428218867565659e-05, + "loss": 1.8969, + "step": 6305 + }, + { + "epoch": 0.5638412017167382, + "grad_norm": 0.938945905323199, + "learning_rate": 8.425358919220537e-05, + "loss": 1.7593, + "step": 6306 + }, + { + "epoch": 0.5639306151645207, + "grad_norm": 0.5951571074787472, + "learning_rate": 8.422499102938663e-05, + "loss": 1.618, + "step": 6307 + }, + { + "epoch": 0.5640200286123033, + "grad_norm": 1.5979695141716117, + "learning_rate": 8.419639418959884e-05, + "loss": 2.0827, + "step": 6308 + }, + { + "epoch": 0.5641094420600858, + "grad_norm": 0.5970516771274151, + "learning_rate": 8.416779867524039e-05, + "loss": 2.1148, + "step": 6309 + }, + { + "epoch": 0.5641988555078684, + "grad_norm": 0.9636277531242635, + "learning_rate": 8.413920448870954e-05, + "loss": 2.0682, + "step": 6310 + }, + { + "epoch": 0.564288268955651, + "grad_norm": 0.9095728590249235, + "learning_rate": 8.411061163240441e-05, + "loss": 1.8643, + "step": 6311 + }, + { + "epoch": 0.5643776824034334, + "grad_norm": 0.4277684168450317, + "learning_rate": 8.408202010872312e-05, + "loss": 1.9186, + "step": 6312 + }, + { + "epoch": 0.564467095851216, + "grad_norm": 1.2485499019637853, + "learning_rate": 8.40534299200636e-05, + "loss": 1.6707, + "step": 6313 + }, + { + "epoch": 0.5645565092989986, + "grad_norm": 1.2717903580156176, + "learning_rate": 8.402484106882364e-05, + "loss": 1.9986, + "step": 6314 + }, + { + "epoch": 0.5646459227467812, + "grad_norm": 0.484907825899921, + "learning_rate": 8.399625355740097e-05, + "loss": 1.852, + "step": 6315 + }, + { + "epoch": 0.5647353361945636, + "grad_norm": 1.4973456355606545, + "learning_rate": 8.396766738819319e-05, + "loss": 1.8261, + "step": 6316 + }, + { + "epoch": 0.5648247496423462, + "grad_norm": 0.7963287861919685, + "learning_rate": 8.393908256359776e-05, + "loss": 1.647, + "step": 6317 + }, + { + "epoch": 0.5649141630901288, + "grad_norm": 0.7579018602535971, + "learning_rate": 8.39104990860121e-05, + "loss": 1.6191, + "step": 6318 + }, + { + "epoch": 0.5650035765379113, + "grad_norm": 1.0382943842832566, + "learning_rate": 8.388191695783345e-05, + "loss": 1.995, + "step": 6319 + }, + { + "epoch": 0.5650929899856938, + "grad_norm": 1.11219193506216, + "learning_rate": 8.385333618145896e-05, + "loss": 2.246, + "step": 6320 + }, + { + "epoch": 0.5651824034334764, + "grad_norm": 0.6275675888344614, + "learning_rate": 8.382475675928568e-05, + "loss": 1.8398, + "step": 6321 + }, + { + "epoch": 0.5652718168812589, + "grad_norm": 1.12622124064658, + "learning_rate": 8.379617869371049e-05, + "loss": 1.9687, + "step": 6322 + }, + { + "epoch": 0.5653612303290415, + "grad_norm": 0.8849212754926369, + "learning_rate": 8.376760198713024e-05, + "loss": 1.5936, + "step": 6323 + }, + { + "epoch": 0.5654506437768241, + "grad_norm": 0.9787488616113698, + "learning_rate": 8.373902664194156e-05, + "loss": 2.0028, + "step": 6324 + }, + { + "epoch": 0.5655400572246065, + "grad_norm": 0.9179833312409595, + "learning_rate": 8.371045266054114e-05, + "loss": 1.8577, + "step": 6325 + }, + { + "epoch": 0.5656294706723891, + "grad_norm": 0.7439969579975612, + "learning_rate": 8.368188004532535e-05, + "loss": 2.0268, + "step": 6326 + }, + { + "epoch": 0.5657188841201717, + "grad_norm": 1.6405204613036075, + "learning_rate": 8.365330879869059e-05, + "loss": 1.935, + "step": 6327 + }, + { + "epoch": 0.5658082975679543, + "grad_norm": 1.065907136587979, + "learning_rate": 8.362473892303308e-05, + "loss": 1.9302, + "step": 6328 + }, + { + "epoch": 0.5658977110157367, + "grad_norm": 1.6511488493978004, + "learning_rate": 8.359617042074891e-05, + "loss": 1.8646, + "step": 6329 + }, + { + "epoch": 0.5659871244635193, + "grad_norm": 0.8425330771274369, + "learning_rate": 8.356760329423417e-05, + "loss": 1.744, + "step": 6330 + }, + { + "epoch": 0.5660765379113019, + "grad_norm": 0.43710517751393846, + "learning_rate": 8.353903754588463e-05, + "loss": 1.7154, + "step": 6331 + }, + { + "epoch": 0.5661659513590844, + "grad_norm": 0.713722706127901, + "learning_rate": 8.351047317809617e-05, + "loss": 1.7196, + "step": 6332 + }, + { + "epoch": 0.566255364806867, + "grad_norm": 0.784748433463082, + "learning_rate": 8.34819101932644e-05, + "loss": 1.7676, + "step": 6333 + }, + { + "epoch": 0.5663447782546495, + "grad_norm": 1.4249509174915822, + "learning_rate": 8.345334859378489e-05, + "loss": 1.9019, + "step": 6334 + }, + { + "epoch": 0.566434191702432, + "grad_norm": 1.1900199575793609, + "learning_rate": 8.342478838205302e-05, + "loss": 1.4283, + "step": 6335 + }, + { + "epoch": 0.5665236051502146, + "grad_norm": 1.593550142364258, + "learning_rate": 8.339622956046417e-05, + "loss": 2.007, + "step": 6336 + }, + { + "epoch": 0.5666130185979972, + "grad_norm": 0.6035015844367481, + "learning_rate": 8.336767213141348e-05, + "loss": 1.4835, + "step": 6337 + }, + { + "epoch": 0.5667024320457796, + "grad_norm": 1.5489177242321939, + "learning_rate": 8.333911609729601e-05, + "loss": 1.5825, + "step": 6338 + }, + { + "epoch": 0.5667918454935622, + "grad_norm": 0.9910379549000436, + "learning_rate": 8.331056146050676e-05, + "loss": 1.5579, + "step": 6339 + }, + { + "epoch": 0.5668812589413448, + "grad_norm": 0.6355416216063682, + "learning_rate": 8.328200822344058e-05, + "loss": 1.6649, + "step": 6340 + }, + { + "epoch": 0.5669706723891274, + "grad_norm": 1.338682847897039, + "learning_rate": 8.325345638849221e-05, + "loss": 2.1205, + "step": 6341 + }, + { + "epoch": 0.5670600858369099, + "grad_norm": 1.7721685331480912, + "learning_rate": 8.322490595805619e-05, + "loss": 1.7494, + "step": 6342 + }, + { + "epoch": 0.5671494992846924, + "grad_norm": 1.0168228189197623, + "learning_rate": 8.319635693452707e-05, + "loss": 1.2556, + "step": 6343 + }, + { + "epoch": 0.567238912732475, + "grad_norm": 0.9077859205951364, + "learning_rate": 8.31678093202992e-05, + "loss": 1.582, + "step": 6344 + }, + { + "epoch": 0.5673283261802575, + "grad_norm": 1.3488908678387415, + "learning_rate": 8.313926311776678e-05, + "loss": 1.941, + "step": 6345 + }, + { + "epoch": 0.5674177396280401, + "grad_norm": 1.5957408827205937, + "learning_rate": 8.311071832932404e-05, + "loss": 1.3715, + "step": 6346 + }, + { + "epoch": 0.5675071530758226, + "grad_norm": 1.4447694726721267, + "learning_rate": 8.308217495736496e-05, + "loss": 2.0194, + "step": 6347 + }, + { + "epoch": 0.5675965665236051, + "grad_norm": 0.47470518833531916, + "learning_rate": 8.305363300428346e-05, + "loss": 1.9045, + "step": 6348 + }, + { + "epoch": 0.5676859799713877, + "grad_norm": 1.1923940273082188, + "learning_rate": 8.302509247247325e-05, + "loss": 1.8121, + "step": 6349 + }, + { + "epoch": 0.5677753934191703, + "grad_norm": 1.2280818347704117, + "learning_rate": 8.299655336432806e-05, + "loss": 1.7404, + "step": 6350 + }, + { + "epoch": 0.5678648068669528, + "grad_norm": 1.1159511451303716, + "learning_rate": 8.296801568224142e-05, + "loss": 1.844, + "step": 6351 + }, + { + "epoch": 0.5679542203147353, + "grad_norm": 0.8896462703518693, + "learning_rate": 8.293947942860666e-05, + "loss": 2.2776, + "step": 6352 + }, + { + "epoch": 0.5680436337625179, + "grad_norm": 0.8777617267208834, + "learning_rate": 8.291094460581721e-05, + "loss": 1.819, + "step": 6353 + }, + { + "epoch": 0.5681330472103004, + "grad_norm": 0.6673795471040249, + "learning_rate": 8.288241121626621e-05, + "loss": 1.6143, + "step": 6354 + }, + { + "epoch": 0.568222460658083, + "grad_norm": 0.8165640964252998, + "learning_rate": 8.28538792623467e-05, + "loss": 1.4853, + "step": 6355 + }, + { + "epoch": 0.5683118741058655, + "grad_norm": 0.8884558674991849, + "learning_rate": 8.282534874645162e-05, + "loss": 2.1032, + "step": 6356 + }, + { + "epoch": 0.568401287553648, + "grad_norm": 0.6881475159365408, + "learning_rate": 8.279681967097381e-05, + "loss": 1.6406, + "step": 6357 + }, + { + "epoch": 0.5684907010014306, + "grad_norm": 0.6369642054853116, + "learning_rate": 8.276829203830596e-05, + "loss": 1.7117, + "step": 6358 + }, + { + "epoch": 0.5685801144492132, + "grad_norm": 1.8260075783080836, + "learning_rate": 8.27397658508406e-05, + "loss": 1.5042, + "step": 6359 + }, + { + "epoch": 0.5686695278969958, + "grad_norm": 0.9526438412084951, + "learning_rate": 8.271124111097026e-05, + "loss": 1.7514, + "step": 6360 + }, + { + "epoch": 0.5687589413447782, + "grad_norm": 1.3592721888291497, + "learning_rate": 8.268271782108727e-05, + "loss": 2.17, + "step": 6361 + }, + { + "epoch": 0.5688483547925608, + "grad_norm": 1.0164887305553896, + "learning_rate": 8.265419598358381e-05, + "loss": 1.9848, + "step": 6362 + }, + { + "epoch": 0.5689377682403434, + "grad_norm": 0.4753981318946522, + "learning_rate": 8.262567560085199e-05, + "loss": 1.6323, + "step": 6363 + }, + { + "epoch": 0.5690271816881259, + "grad_norm": 0.64932040164583, + "learning_rate": 8.259715667528377e-05, + "loss": 1.8394, + "step": 6364 + }, + { + "epoch": 0.5691165951359084, + "grad_norm": 1.1779721138123802, + "learning_rate": 8.256863920927099e-05, + "loss": 2.0722, + "step": 6365 + }, + { + "epoch": 0.569206008583691, + "grad_norm": 1.1155556732857899, + "learning_rate": 8.254012320520539e-05, + "loss": 2.1005, + "step": 6366 + }, + { + "epoch": 0.5692954220314735, + "grad_norm": 0.9704557712049606, + "learning_rate": 8.251160866547857e-05, + "loss": 1.7945, + "step": 6367 + }, + { + "epoch": 0.5693848354792561, + "grad_norm": 1.66811484856012, + "learning_rate": 8.248309559248203e-05, + "loss": 1.5716, + "step": 6368 + }, + { + "epoch": 0.5694742489270386, + "grad_norm": 1.8389096905758038, + "learning_rate": 8.245458398860709e-05, + "loss": 1.8477, + "step": 6369 + }, + { + "epoch": 0.5695636623748211, + "grad_norm": 0.5991564821075225, + "learning_rate": 8.242607385624501e-05, + "loss": 1.9021, + "step": 6370 + }, + { + "epoch": 0.5696530758226037, + "grad_norm": 1.4760726613346005, + "learning_rate": 8.23975651977869e-05, + "loss": 1.8931, + "step": 6371 + }, + { + "epoch": 0.5697424892703863, + "grad_norm": 1.0504062011204633, + "learning_rate": 8.236905801562373e-05, + "loss": 1.5367, + "step": 6372 + }, + { + "epoch": 0.5698319027181689, + "grad_norm": 0.9496247319193364, + "learning_rate": 8.234055231214634e-05, + "loss": 1.602, + "step": 6373 + }, + { + "epoch": 0.5699213161659513, + "grad_norm": 1.1556346022126138, + "learning_rate": 8.231204808974554e-05, + "loss": 1.5865, + "step": 6374 + }, + { + "epoch": 0.5700107296137339, + "grad_norm": 1.2915680704353785, + "learning_rate": 8.228354535081191e-05, + "loss": 1.8542, + "step": 6375 + }, + { + "epoch": 0.5701001430615165, + "grad_norm": 0.7920076643918855, + "learning_rate": 8.225504409773591e-05, + "loss": 1.8246, + "step": 6376 + }, + { + "epoch": 0.570189556509299, + "grad_norm": 0.8921948526838723, + "learning_rate": 8.222654433290795e-05, + "loss": 1.7301, + "step": 6377 + }, + { + "epoch": 0.5702789699570815, + "grad_norm": 1.4246078060273637, + "learning_rate": 8.219804605871826e-05, + "loss": 1.6462, + "step": 6378 + }, + { + "epoch": 0.5703683834048641, + "grad_norm": 1.1065566689293893, + "learning_rate": 8.216954927755692e-05, + "loss": 1.9859, + "step": 6379 + }, + { + "epoch": 0.5704577968526466, + "grad_norm": 1.0961609097123213, + "learning_rate": 8.214105399181393e-05, + "loss": 2.3854, + "step": 6380 + }, + { + "epoch": 0.5705472103004292, + "grad_norm": 1.853675245788332, + "learning_rate": 8.21125602038792e-05, + "loss": 1.6731, + "step": 6381 + }, + { + "epoch": 0.5706366237482118, + "grad_norm": 0.9470772519342872, + "learning_rate": 8.208406791614247e-05, + "loss": 2.0801, + "step": 6382 + }, + { + "epoch": 0.5707260371959942, + "grad_norm": 2.031455824468731, + "learning_rate": 8.20555771309933e-05, + "loss": 2.1032, + "step": 6383 + }, + { + "epoch": 0.5708154506437768, + "grad_norm": 1.6283373327547532, + "learning_rate": 8.202708785082121e-05, + "loss": 2.0395, + "step": 6384 + }, + { + "epoch": 0.5709048640915594, + "grad_norm": 0.7938863664607699, + "learning_rate": 8.199860007801557e-05, + "loss": 1.7461, + "step": 6385 + }, + { + "epoch": 0.570994277539342, + "grad_norm": 2.706068262262351, + "learning_rate": 8.197011381496558e-05, + "loss": 1.8839, + "step": 6386 + }, + { + "epoch": 0.5710836909871244, + "grad_norm": 1.5001390576272975, + "learning_rate": 8.194162906406033e-05, + "loss": 1.8825, + "step": 6387 + }, + { + "epoch": 0.571173104434907, + "grad_norm": 1.0755622835240426, + "learning_rate": 8.191314582768891e-05, + "loss": 1.6408, + "step": 6388 + }, + { + "epoch": 0.5712625178826896, + "grad_norm": 2.213413110539304, + "learning_rate": 8.18846641082401e-05, + "loss": 1.9716, + "step": 6389 + }, + { + "epoch": 0.5713519313304721, + "grad_norm": 0.6887612834454343, + "learning_rate": 8.18561839081026e-05, + "loss": 1.381, + "step": 6390 + }, + { + "epoch": 0.5714413447782547, + "grad_norm": 1.4854173181381165, + "learning_rate": 8.182770522966507e-05, + "loss": 1.8094, + "step": 6391 + }, + { + "epoch": 0.5715307582260372, + "grad_norm": 1.3876263931168056, + "learning_rate": 8.179922807531594e-05, + "loss": 1.9223, + "step": 6392 + }, + { + "epoch": 0.5716201716738197, + "grad_norm": 0.4417951579718184, + "learning_rate": 8.177075244744358e-05, + "loss": 1.8828, + "step": 6393 + }, + { + "epoch": 0.5717095851216023, + "grad_norm": 1.1686316944923658, + "learning_rate": 8.174227834843617e-05, + "loss": 1.7474, + "step": 6394 + }, + { + "epoch": 0.5717989985693849, + "grad_norm": 1.620950147276358, + "learning_rate": 8.171380578068185e-05, + "loss": 1.7097, + "step": 6395 + }, + { + "epoch": 0.5718884120171673, + "grad_norm": 1.4517127826378524, + "learning_rate": 8.168533474656855e-05, + "loss": 2.0011, + "step": 6396 + }, + { + "epoch": 0.5719778254649499, + "grad_norm": 1.393897569946749, + "learning_rate": 8.165686524848411e-05, + "loss": 2.1058, + "step": 6397 + }, + { + "epoch": 0.5720672389127325, + "grad_norm": 1.0284078944310486, + "learning_rate": 8.162839728881625e-05, + "loss": 1.582, + "step": 6398 + }, + { + "epoch": 0.572156652360515, + "grad_norm": 1.9555129917618015, + "learning_rate": 8.159993086995249e-05, + "loss": 1.6743, + "step": 6399 + }, + { + "epoch": 0.5722460658082976, + "grad_norm": 2.3460545098653554, + "learning_rate": 8.157146599428028e-05, + "loss": 2.0238, + "step": 6400 + }, + { + "epoch": 0.5723354792560801, + "grad_norm": 0.8173927314403898, + "learning_rate": 8.154300266418702e-05, + "loss": 2.0344, + "step": 6401 + }, + { + "epoch": 0.5724248927038627, + "grad_norm": 0.7822812284657884, + "learning_rate": 8.151454088205982e-05, + "loss": 1.6642, + "step": 6402 + }, + { + "epoch": 0.5725143061516452, + "grad_norm": 1.4332297004046537, + "learning_rate": 8.148608065028574e-05, + "loss": 1.6384, + "step": 6403 + }, + { + "epoch": 0.5726037195994278, + "grad_norm": 1.402952178287237, + "learning_rate": 8.145762197125173e-05, + "loss": 1.8355, + "step": 6404 + }, + { + "epoch": 0.5726931330472103, + "grad_norm": 0.7154241040966602, + "learning_rate": 8.142916484734458e-05, + "loss": 1.7014, + "step": 6405 + }, + { + "epoch": 0.5727825464949928, + "grad_norm": 1.309500845118811, + "learning_rate": 8.140070928095092e-05, + "loss": 1.659, + "step": 6406 + }, + { + "epoch": 0.5728719599427754, + "grad_norm": 0.7831861898400589, + "learning_rate": 8.137225527445727e-05, + "loss": 1.9293, + "step": 6407 + }, + { + "epoch": 0.572961373390558, + "grad_norm": 0.9563326272590194, + "learning_rate": 8.134380283025014e-05, + "loss": 2.1605, + "step": 6408 + }, + { + "epoch": 0.5730507868383404, + "grad_norm": 0.9770300794525341, + "learning_rate": 8.131535195071574e-05, + "loss": 1.4727, + "step": 6409 + }, + { + "epoch": 0.573140200286123, + "grad_norm": 1.2971479856438313, + "learning_rate": 8.128690263824017e-05, + "loss": 1.5281, + "step": 6410 + }, + { + "epoch": 0.5732296137339056, + "grad_norm": 0.8560934541174887, + "learning_rate": 8.12584548952095e-05, + "loss": 1.5059, + "step": 6411 + }, + { + "epoch": 0.5733190271816881, + "grad_norm": 1.6431090530342733, + "learning_rate": 8.123000872400959e-05, + "loss": 2.4228, + "step": 6412 + }, + { + "epoch": 0.5734084406294707, + "grad_norm": 0.8774360204960419, + "learning_rate": 8.120156412702615e-05, + "loss": 1.8037, + "step": 6413 + }, + { + "epoch": 0.5734978540772532, + "grad_norm": 0.8365520979517573, + "learning_rate": 8.117312110664482e-05, + "loss": 1.408, + "step": 6414 + }, + { + "epoch": 0.5735872675250357, + "grad_norm": 1.4335968982209975, + "learning_rate": 8.114467966525112e-05, + "loss": 1.4777, + "step": 6415 + }, + { + "epoch": 0.5736766809728183, + "grad_norm": 2.067889696925775, + "learning_rate": 8.111623980523035e-05, + "loss": 1.7294, + "step": 6416 + }, + { + "epoch": 0.5737660944206009, + "grad_norm": 1.175465413897075, + "learning_rate": 8.108780152896773e-05, + "loss": 1.9359, + "step": 6417 + }, + { + "epoch": 0.5738555078683834, + "grad_norm": 1.1939001232762458, + "learning_rate": 8.105936483884838e-05, + "loss": 1.9358, + "step": 6418 + }, + { + "epoch": 0.5739449213161659, + "grad_norm": 1.6877813159802173, + "learning_rate": 8.103092973725724e-05, + "loss": 1.529, + "step": 6419 + }, + { + "epoch": 0.5740343347639485, + "grad_norm": 0.758361633431353, + "learning_rate": 8.100249622657907e-05, + "loss": 1.6102, + "step": 6420 + }, + { + "epoch": 0.5741237482117311, + "grad_norm": 1.6942917131898705, + "learning_rate": 8.097406430919858e-05, + "loss": 1.887, + "step": 6421 + }, + { + "epoch": 0.5742131616595136, + "grad_norm": 0.9425204156825014, + "learning_rate": 8.094563398750039e-05, + "loss": 1.8184, + "step": 6422 + }, + { + "epoch": 0.5743025751072961, + "grad_norm": 1.479735804550438, + "learning_rate": 8.091720526386886e-05, + "loss": 1.6927, + "step": 6423 + }, + { + "epoch": 0.5743919885550787, + "grad_norm": 2.7595009780934987, + "learning_rate": 8.088877814068827e-05, + "loss": 2.502, + "step": 6424 + }, + { + "epoch": 0.5744814020028612, + "grad_norm": 0.679085906808698, + "learning_rate": 8.086035262034278e-05, + "loss": 2.2571, + "step": 6425 + }, + { + "epoch": 0.5745708154506438, + "grad_norm": 1.251323073031776, + "learning_rate": 8.083192870521638e-05, + "loss": 1.7858, + "step": 6426 + }, + { + "epoch": 0.5746602288984263, + "grad_norm": 0.6160331698963857, + "learning_rate": 8.0803506397693e-05, + "loss": 1.6965, + "step": 6427 + }, + { + "epoch": 0.5747496423462088, + "grad_norm": 1.5652471932127854, + "learning_rate": 8.077508570015632e-05, + "loss": 1.76, + "step": 6428 + }, + { + "epoch": 0.5748390557939914, + "grad_norm": 0.5752829063591766, + "learning_rate": 8.074666661499002e-05, + "loss": 1.58, + "step": 6429 + }, + { + "epoch": 0.574928469241774, + "grad_norm": 0.878292239667154, + "learning_rate": 8.071824914457751e-05, + "loss": 1.7916, + "step": 6430 + }, + { + "epoch": 0.5750178826895566, + "grad_norm": 0.6507670921116676, + "learning_rate": 8.068983329130218e-05, + "loss": 1.6059, + "step": 6431 + }, + { + "epoch": 0.575107296137339, + "grad_norm": 1.6159288875611668, + "learning_rate": 8.066141905754723e-05, + "loss": 1.9029, + "step": 6432 + }, + { + "epoch": 0.5751967095851216, + "grad_norm": 1.1633082229485907, + "learning_rate": 8.063300644569567e-05, + "loss": 1.7112, + "step": 6433 + }, + { + "epoch": 0.5752861230329042, + "grad_norm": 3.99264044494413, + "learning_rate": 8.060459545813049e-05, + "loss": 6.6759, + "step": 6434 + }, + { + "epoch": 0.5753755364806867, + "grad_norm": 2.358221091504515, + "learning_rate": 8.057618609723443e-05, + "loss": 1.8757, + "step": 6435 + }, + { + "epoch": 0.5754649499284692, + "grad_norm": 1.8295226809960372, + "learning_rate": 8.054777836539022e-05, + "loss": 1.9111, + "step": 6436 + }, + { + "epoch": 0.5755543633762518, + "grad_norm": 0.7080053058468473, + "learning_rate": 8.051937226498034e-05, + "loss": 1.7594, + "step": 6437 + }, + { + "epoch": 0.5756437768240343, + "grad_norm": 1.3239740411767997, + "learning_rate": 8.049096779838719e-05, + "loss": 1.5564, + "step": 6438 + }, + { + "epoch": 0.5757331902718169, + "grad_norm": 3.502831814851747, + "learning_rate": 8.0462564967993e-05, + "loss": 6.8762, + "step": 6439 + }, + { + "epoch": 0.5758226037195995, + "grad_norm": 0.7283485334653405, + "learning_rate": 8.043416377617988e-05, + "loss": 1.9057, + "step": 6440 + }, + { + "epoch": 0.5759120171673819, + "grad_norm": 1.5649081680739332, + "learning_rate": 8.040576422532984e-05, + "loss": 1.5439, + "step": 6441 + }, + { + "epoch": 0.5760014306151645, + "grad_norm": 1.6385026188947223, + "learning_rate": 8.037736631782465e-05, + "loss": 1.7976, + "step": 6442 + }, + { + "epoch": 0.5760908440629471, + "grad_norm": 0.8639204721444184, + "learning_rate": 8.034897005604608e-05, + "loss": 1.4158, + "step": 6443 + }, + { + "epoch": 0.5761802575107297, + "grad_norm": 1.572771321177526, + "learning_rate": 8.032057544237565e-05, + "loss": 1.9427, + "step": 6444 + }, + { + "epoch": 0.5762696709585121, + "grad_norm": 1.6760065868942697, + "learning_rate": 8.02921824791948e-05, + "loss": 2.0753, + "step": 6445 + }, + { + "epoch": 0.5763590844062947, + "grad_norm": 1.290101974533281, + "learning_rate": 8.026379116888481e-05, + "loss": 1.9271, + "step": 6446 + }, + { + "epoch": 0.5764484978540773, + "grad_norm": 1.3076439401614945, + "learning_rate": 8.02354015138268e-05, + "loss": 1.8304, + "step": 6447 + }, + { + "epoch": 0.5765379113018598, + "grad_norm": 1.2596039520250981, + "learning_rate": 8.020701351640182e-05, + "loss": 1.6213, + "step": 6448 + }, + { + "epoch": 0.5766273247496424, + "grad_norm": 0.5021934119090404, + "learning_rate": 8.017862717899066e-05, + "loss": 1.9429, + "step": 6449 + }, + { + "epoch": 0.5767167381974249, + "grad_norm": 0.5567998980781425, + "learning_rate": 8.015024250397415e-05, + "loss": 1.5682, + "step": 6450 + }, + { + "epoch": 0.5768061516452074, + "grad_norm": 0.801933586866022, + "learning_rate": 8.01218594937328e-05, + "loss": 1.3127, + "step": 6451 + }, + { + "epoch": 0.57689556509299, + "grad_norm": 1.1345643865106247, + "learning_rate": 8.009347815064712e-05, + "loss": 1.858, + "step": 6452 + }, + { + "epoch": 0.5769849785407726, + "grad_norm": 1.0698419765244216, + "learning_rate": 8.006509847709735e-05, + "loss": 2.1115, + "step": 6453 + }, + { + "epoch": 0.577074391988555, + "grad_norm": 0.6601484562470465, + "learning_rate": 8.003672047546373e-05, + "loss": 2.1191, + "step": 6454 + }, + { + "epoch": 0.5771638054363376, + "grad_norm": 0.9029469319540541, + "learning_rate": 8.000834414812625e-05, + "loss": 1.7514, + "step": 6455 + }, + { + "epoch": 0.5772532188841202, + "grad_norm": 1.2167670813186886, + "learning_rate": 7.997996949746477e-05, + "loss": 1.5651, + "step": 6456 + }, + { + "epoch": 0.5773426323319027, + "grad_norm": 1.3619357066642985, + "learning_rate": 7.995159652585908e-05, + "loss": 1.7165, + "step": 6457 + }, + { + "epoch": 0.5774320457796852, + "grad_norm": 0.6207235864721994, + "learning_rate": 7.99232252356888e-05, + "loss": 1.7656, + "step": 6458 + }, + { + "epoch": 0.5775214592274678, + "grad_norm": 1.3674382185406997, + "learning_rate": 7.989485562933338e-05, + "loss": 1.8246, + "step": 6459 + }, + { + "epoch": 0.5776108726752504, + "grad_norm": 0.6994635236042639, + "learning_rate": 7.98664877091721e-05, + "loss": 1.4034, + "step": 6460 + }, + { + "epoch": 0.5777002861230329, + "grad_norm": 0.9771387988394465, + "learning_rate": 7.983812147758422e-05, + "loss": 1.6335, + "step": 6461 + }, + { + "epoch": 0.5777896995708155, + "grad_norm": 1.1033881873234859, + "learning_rate": 7.980975693694872e-05, + "loss": 1.8385, + "step": 6462 + }, + { + "epoch": 0.577879113018598, + "grad_norm": 1.205939740889653, + "learning_rate": 7.97813940896445e-05, + "loss": 1.7325, + "step": 6463 + }, + { + "epoch": 0.5779685264663805, + "grad_norm": 0.8703205708744, + "learning_rate": 7.975303293805035e-05, + "loss": 2.2183, + "step": 6464 + }, + { + "epoch": 0.5780579399141631, + "grad_norm": 1.0362855345174764, + "learning_rate": 7.97246734845449e-05, + "loss": 1.7032, + "step": 6465 + }, + { + "epoch": 0.5781473533619457, + "grad_norm": 1.3131371913954468, + "learning_rate": 7.96963157315066e-05, + "loss": 2.1344, + "step": 6466 + }, + { + "epoch": 0.5782367668097281, + "grad_norm": 0.44506317495341224, + "learning_rate": 7.966795968131377e-05, + "loss": 1.7345, + "step": 6467 + }, + { + "epoch": 0.5783261802575107, + "grad_norm": 0.5247572935001975, + "learning_rate": 7.963960533634461e-05, + "loss": 2.0923, + "step": 6468 + }, + { + "epoch": 0.5784155937052933, + "grad_norm": 0.42045026055377366, + "learning_rate": 7.961125269897716e-05, + "loss": 1.5243, + "step": 6469 + }, + { + "epoch": 0.5785050071530758, + "grad_norm": 0.6461819513034858, + "learning_rate": 7.95829017715893e-05, + "loss": 1.7056, + "step": 6470 + }, + { + "epoch": 0.5785944206008584, + "grad_norm": 1.3985899467549048, + "learning_rate": 7.955455255655881e-05, + "loss": 1.6943, + "step": 6471 + }, + { + "epoch": 0.5786838340486409, + "grad_norm": 1.185335753169801, + "learning_rate": 7.952620505626333e-05, + "loss": 1.6048, + "step": 6472 + }, + { + "epoch": 0.5787732474964234, + "grad_norm": 0.8853669689564075, + "learning_rate": 7.949785927308032e-05, + "loss": 1.8139, + "step": 6473 + }, + { + "epoch": 0.578862660944206, + "grad_norm": 0.9172710862888815, + "learning_rate": 7.946951520938706e-05, + "loss": 1.7656, + "step": 6474 + }, + { + "epoch": 0.5789520743919886, + "grad_norm": 1.1079613842464446, + "learning_rate": 7.944117286756079e-05, + "loss": 1.3649, + "step": 6475 + }, + { + "epoch": 0.579041487839771, + "grad_norm": 2.7037745924855883, + "learning_rate": 7.94128322499785e-05, + "loss": 1.7435, + "step": 6476 + }, + { + "epoch": 0.5791309012875536, + "grad_norm": 1.016614209814845, + "learning_rate": 7.93844933590171e-05, + "loss": 1.5353, + "step": 6477 + }, + { + "epoch": 0.5792203147353362, + "grad_norm": 0.5793053168858945, + "learning_rate": 7.935615619705334e-05, + "loss": 1.8924, + "step": 6478 + }, + { + "epoch": 0.5793097281831188, + "grad_norm": 1.4072638304387317, + "learning_rate": 7.932782076646386e-05, + "loss": 1.7535, + "step": 6479 + }, + { + "epoch": 0.5793991416309013, + "grad_norm": 1.9941861334948165, + "learning_rate": 7.929948706962508e-05, + "loss": 1.7789, + "step": 6480 + }, + { + "epoch": 0.5794885550786838, + "grad_norm": 0.6253527465710091, + "learning_rate": 7.927115510891332e-05, + "loss": 1.6878, + "step": 6481 + }, + { + "epoch": 0.5795779685264664, + "grad_norm": 1.1104643732803752, + "learning_rate": 7.924282488670476e-05, + "loss": 1.87, + "step": 6482 + }, + { + "epoch": 0.5796673819742489, + "grad_norm": 8.316606386692515, + "learning_rate": 7.921449640537535e-05, + "loss": 6.6081, + "step": 6483 + }, + { + "epoch": 0.5797567954220315, + "grad_norm": 4.94539993080349, + "learning_rate": 7.918616966730108e-05, + "loss": 1.71, + "step": 6484 + }, + { + "epoch": 0.579846208869814, + "grad_norm": 4.581965578027085, + "learning_rate": 7.91578446748576e-05, + "loss": 1.9446, + "step": 6485 + }, + { + "epoch": 0.5799356223175965, + "grad_norm": 1.4277472354433256, + "learning_rate": 7.912952143042052e-05, + "loss": 1.9817, + "step": 6486 + }, + { + "epoch": 0.5800250357653791, + "grad_norm": 2.681033602156615, + "learning_rate": 7.910119993636528e-05, + "loss": 1.7716, + "step": 6487 + }, + { + "epoch": 0.5801144492131617, + "grad_norm": 4.294390244508989, + "learning_rate": 7.907288019506717e-05, + "loss": 2.0865, + "step": 6488 + }, + { + "epoch": 0.5802038626609443, + "grad_norm": 1.255141136255979, + "learning_rate": 7.904456220890132e-05, + "loss": 2.1567, + "step": 6489 + }, + { + "epoch": 0.5802932761087267, + "grad_norm": 3.008038634216335, + "learning_rate": 7.901624598024269e-05, + "loss": 1.8226, + "step": 6490 + }, + { + "epoch": 0.5803826895565093, + "grad_norm": 3.2279297436168286, + "learning_rate": 7.89879315114662e-05, + "loss": 1.7124, + "step": 6491 + }, + { + "epoch": 0.5804721030042919, + "grad_norm": 1.7682839404877362, + "learning_rate": 7.895961880494652e-05, + "loss": 2.2807, + "step": 6492 + }, + { + "epoch": 0.5805615164520744, + "grad_norm": 2.465906374837297, + "learning_rate": 7.893130786305821e-05, + "loss": 1.8294, + "step": 6493 + }, + { + "epoch": 0.5806509298998569, + "grad_norm": 3.3645403605828874, + "learning_rate": 7.890299868817564e-05, + "loss": 2.1587, + "step": 6494 + }, + { + "epoch": 0.5807403433476395, + "grad_norm": 1.9020153632853525, + "learning_rate": 7.887469128267312e-05, + "loss": 1.7175, + "step": 6495 + }, + { + "epoch": 0.580829756795422, + "grad_norm": 1.7437569565322852, + "learning_rate": 7.884638564892472e-05, + "loss": 1.2437, + "step": 6496 + }, + { + "epoch": 0.5809191702432046, + "grad_norm": 2.7447873916789476, + "learning_rate": 7.881808178930438e-05, + "loss": 1.4899, + "step": 6497 + }, + { + "epoch": 0.5810085836909872, + "grad_norm": 2.772413175914574, + "learning_rate": 7.878977970618595e-05, + "loss": 2.1886, + "step": 6498 + }, + { + "epoch": 0.5810979971387696, + "grad_norm": 1.6840809625423456, + "learning_rate": 7.876147940194311e-05, + "loss": 1.7193, + "step": 6499 + }, + { + "epoch": 0.5811874105865522, + "grad_norm": 4.099732394366122, + "learning_rate": 7.873318087894933e-05, + "loss": 2.0399, + "step": 6500 + }, + { + "epoch": 0.5812768240343348, + "grad_norm": 2.225550703678738, + "learning_rate": 7.870488413957797e-05, + "loss": 1.5527, + "step": 6501 + }, + { + "epoch": 0.5813662374821174, + "grad_norm": 1.846998940244715, + "learning_rate": 7.867658918620229e-05, + "loss": 1.5291, + "step": 6502 + }, + { + "epoch": 0.5814556509298998, + "grad_norm": 3.0457128625477, + "learning_rate": 7.86482960211953e-05, + "loss": 1.9308, + "step": 6503 + }, + { + "epoch": 0.5815450643776824, + "grad_norm": 3.4190684290808133, + "learning_rate": 7.862000464692991e-05, + "loss": 1.6622, + "step": 6504 + }, + { + "epoch": 0.581634477825465, + "grad_norm": 2.9796515151522196, + "learning_rate": 7.859171506577893e-05, + "loss": 1.4395, + "step": 6505 + }, + { + "epoch": 0.5817238912732475, + "grad_norm": 3.108667501620754, + "learning_rate": 7.856342728011498e-05, + "loss": 1.9266, + "step": 6506 + }, + { + "epoch": 0.58181330472103, + "grad_norm": 9.138250192925021, + "learning_rate": 7.853514129231049e-05, + "loss": 6.59, + "step": 6507 + }, + { + "epoch": 0.5819027181688126, + "grad_norm": 6.551829552365603, + "learning_rate": 7.850685710473775e-05, + "loss": 2.0979, + "step": 6508 + }, + { + "epoch": 0.5819921316165951, + "grad_norm": 4.881037240555335, + "learning_rate": 7.847857471976897e-05, + "loss": 1.6707, + "step": 6509 + }, + { + "epoch": 0.5820815450643777, + "grad_norm": 1.8250767312354241, + "learning_rate": 7.845029413977613e-05, + "loss": 1.7302, + "step": 6510 + }, + { + "epoch": 0.5821709585121603, + "grad_norm": 2.9449619486088214, + "learning_rate": 7.842201536713107e-05, + "loss": 1.8369, + "step": 6511 + }, + { + "epoch": 0.5822603719599427, + "grad_norm": 6.229067406636033, + "learning_rate": 7.839373840420554e-05, + "loss": 1.8996, + "step": 6512 + }, + { + "epoch": 0.5823497854077253, + "grad_norm": 2.2460873450964534, + "learning_rate": 7.83654632533711e-05, + "loss": 1.5205, + "step": 6513 + }, + { + "epoch": 0.5824391988555079, + "grad_norm": 1.359600712695774, + "learning_rate": 7.83371899169991e-05, + "loss": 1.7536, + "step": 6514 + }, + { + "epoch": 0.5825286123032904, + "grad_norm": 3.938293232197056, + "learning_rate": 7.830891839746083e-05, + "loss": 1.8538, + "step": 6515 + }, + { + "epoch": 0.5826180257510729, + "grad_norm": 3.4659102075874917, + "learning_rate": 7.828064869712739e-05, + "loss": 1.8777, + "step": 6516 + }, + { + "epoch": 0.5827074391988555, + "grad_norm": 1.6501899568139229, + "learning_rate": 7.82523808183697e-05, + "loss": 1.9901, + "step": 6517 + }, + { + "epoch": 0.582796852646638, + "grad_norm": 2.257138278734233, + "learning_rate": 7.822411476355854e-05, + "loss": 1.7906, + "step": 6518 + }, + { + "epoch": 0.5828862660944206, + "grad_norm": 2.7908496250151877, + "learning_rate": 7.819585053506461e-05, + "loss": 1.6473, + "step": 6519 + }, + { + "epoch": 0.5829756795422032, + "grad_norm": 2.6060617540248963, + "learning_rate": 7.816758813525836e-05, + "loss": 1.5998, + "step": 6520 + }, + { + "epoch": 0.5830650929899857, + "grad_norm": 0.898195688590259, + "learning_rate": 7.813932756651012e-05, + "loss": 2.1505, + "step": 6521 + }, + { + "epoch": 0.5831545064377682, + "grad_norm": 2.526598969771819, + "learning_rate": 7.811106883119008e-05, + "loss": 2.207, + "step": 6522 + }, + { + "epoch": 0.5832439198855508, + "grad_norm": 2.061883396284492, + "learning_rate": 7.808281193166829e-05, + "loss": 1.753, + "step": 6523 + }, + { + "epoch": 0.5833333333333334, + "grad_norm": 1.2820657862457656, + "learning_rate": 7.805455687031455e-05, + "loss": 1.6438, + "step": 6524 + }, + { + "epoch": 0.5834227467811158, + "grad_norm": 0.8759305344633427, + "learning_rate": 7.80263036494986e-05, + "loss": 1.631, + "step": 6525 + }, + { + "epoch": 0.5835121602288984, + "grad_norm": 8.156185656408095, + "learning_rate": 7.799805227159007e-05, + "loss": 6.7951, + "step": 6526 + }, + { + "epoch": 0.583601573676681, + "grad_norm": 3.0875385680866883, + "learning_rate": 7.796980273895833e-05, + "loss": 1.5622, + "step": 6527 + }, + { + "epoch": 0.5836909871244635, + "grad_norm": 4.003373584221746, + "learning_rate": 7.794155505397261e-05, + "loss": 1.6212, + "step": 6528 + }, + { + "epoch": 0.5837804005722461, + "grad_norm": 2.286189095808415, + "learning_rate": 7.791330921900205e-05, + "loss": 1.8399, + "step": 6529 + }, + { + "epoch": 0.5838698140200286, + "grad_norm": 1.4501711977092249, + "learning_rate": 7.788506523641556e-05, + "loss": 1.3328, + "step": 6530 + }, + { + "epoch": 0.5839592274678111, + "grad_norm": 4.05689042351577, + "learning_rate": 7.785682310858193e-05, + "loss": 2.0127, + "step": 6531 + }, + { + "epoch": 0.5840486409155937, + "grad_norm": 4.217605314808252, + "learning_rate": 7.782858283786976e-05, + "loss": 1.8364, + "step": 6532 + }, + { + "epoch": 0.5841380543633763, + "grad_norm": 1.6626588990293947, + "learning_rate": 7.780034442664764e-05, + "loss": 1.8737, + "step": 6533 + }, + { + "epoch": 0.5842274678111588, + "grad_norm": 2.2018790672984547, + "learning_rate": 7.777210787728382e-05, + "loss": 1.678, + "step": 6534 + }, + { + "epoch": 0.5843168812589413, + "grad_norm": 3.2262374515676675, + "learning_rate": 7.774387319214643e-05, + "loss": 1.7351, + "step": 6535 + }, + { + "epoch": 0.5844062947067239, + "grad_norm": 2.046864081307369, + "learning_rate": 7.771564037360355e-05, + "loss": 1.7173, + "step": 6536 + }, + { + "epoch": 0.5844957081545065, + "grad_norm": 0.9081951159658568, + "learning_rate": 7.768740942402301e-05, + "loss": 1.8326, + "step": 6537 + }, + { + "epoch": 0.584585121602289, + "grad_norm": 1.3304330823141117, + "learning_rate": 7.765918034577245e-05, + "loss": 2.0961, + "step": 6538 + }, + { + "epoch": 0.5846745350500715, + "grad_norm": 2.173648433270032, + "learning_rate": 7.763095314121945e-05, + "loss": 1.8421, + "step": 6539 + }, + { + "epoch": 0.5847639484978541, + "grad_norm": 1.3720466648968457, + "learning_rate": 7.760272781273142e-05, + "loss": 1.9127, + "step": 6540 + }, + { + "epoch": 0.5848533619456366, + "grad_norm": 1.5902167261287155, + "learning_rate": 7.757450436267558e-05, + "loss": 1.953, + "step": 6541 + }, + { + "epoch": 0.5849427753934192, + "grad_norm": 0.6090015808515532, + "learning_rate": 7.754628279341895e-05, + "loss": 1.7281, + "step": 6542 + }, + { + "epoch": 0.5850321888412017, + "grad_norm": 2.36444266959111, + "learning_rate": 7.751806310732847e-05, + "loss": 2.0532, + "step": 6543 + }, + { + "epoch": 0.5851216022889842, + "grad_norm": 1.0985033307426995, + "learning_rate": 7.748984530677089e-05, + "loss": 1.817, + "step": 6544 + }, + { + "epoch": 0.5852110157367668, + "grad_norm": 0.7921196027662811, + "learning_rate": 7.746162939411279e-05, + "loss": 1.4483, + "step": 6545 + }, + { + "epoch": 0.5853004291845494, + "grad_norm": 1.3110574782743512, + "learning_rate": 7.74334153717206e-05, + "loss": 1.6112, + "step": 6546 + }, + { + "epoch": 0.585389842632332, + "grad_norm": 2.7930679508926426, + "learning_rate": 7.740520324196064e-05, + "loss": 1.893, + "step": 6547 + }, + { + "epoch": 0.5854792560801144, + "grad_norm": 0.7789448815913588, + "learning_rate": 7.737699300719896e-05, + "loss": 1.7855, + "step": 6548 + }, + { + "epoch": 0.585568669527897, + "grad_norm": 1.7308406860264742, + "learning_rate": 7.734878466980159e-05, + "loss": 1.9075, + "step": 6549 + }, + { + "epoch": 0.5856580829756796, + "grad_norm": 2.175452906188264, + "learning_rate": 7.73205782321343e-05, + "loss": 1.7129, + "step": 6550 + }, + { + "epoch": 0.5857474964234621, + "grad_norm": 1.6678883507632085, + "learning_rate": 7.729237369656269e-05, + "loss": 2.2074, + "step": 6551 + }, + { + "epoch": 0.5858369098712446, + "grad_norm": 3.1331250464221845, + "learning_rate": 7.72641710654523e-05, + "loss": 1.8897, + "step": 6552 + }, + { + "epoch": 0.5859263233190272, + "grad_norm": 1.33188358532176, + "learning_rate": 7.723597034116838e-05, + "loss": 1.8107, + "step": 6553 + }, + { + "epoch": 0.5860157367668097, + "grad_norm": 1.4652829031511039, + "learning_rate": 7.720777152607619e-05, + "loss": 2.1316, + "step": 6554 + }, + { + "epoch": 0.5861051502145923, + "grad_norm": 1.933454585832967, + "learning_rate": 7.717957462254065e-05, + "loss": 1.8951, + "step": 6555 + }, + { + "epoch": 0.5861945636623748, + "grad_norm": 2.166695891755191, + "learning_rate": 7.715137963292665e-05, + "loss": 1.9184, + "step": 6556 + }, + { + "epoch": 0.5862839771101573, + "grad_norm": 1.2082824158952716, + "learning_rate": 7.712318655959884e-05, + "loss": 2.1926, + "step": 6557 + }, + { + "epoch": 0.5863733905579399, + "grad_norm": 1.2545559164251032, + "learning_rate": 7.709499540492171e-05, + "loss": 1.5413, + "step": 6558 + }, + { + "epoch": 0.5864628040057225, + "grad_norm": 1.5474335835236712, + "learning_rate": 7.70668061712597e-05, + "loss": 1.4172, + "step": 6559 + }, + { + "epoch": 0.586552217453505, + "grad_norm": 0.7678152902064488, + "learning_rate": 7.70386188609769e-05, + "loss": 2.4096, + "step": 6560 + }, + { + "epoch": 0.5866416309012875, + "grad_norm": 0.7159242874134073, + "learning_rate": 7.701043347643747e-05, + "loss": 1.3771, + "step": 6561 + }, + { + "epoch": 0.5867310443490701, + "grad_norm": 0.9162891666623887, + "learning_rate": 7.698225002000516e-05, + "loss": 1.9822, + "step": 6562 + }, + { + "epoch": 0.5868204577968527, + "grad_norm": 0.635067789560074, + "learning_rate": 7.695406849404379e-05, + "loss": 1.7528, + "step": 6563 + }, + { + "epoch": 0.5869098712446352, + "grad_norm": 0.8020618225298484, + "learning_rate": 7.692588890091686e-05, + "loss": 1.7959, + "step": 6564 + }, + { + "epoch": 0.5869992846924177, + "grad_norm": 0.4662975276239208, + "learning_rate": 7.689771124298774e-05, + "loss": 1.2761, + "step": 6565 + }, + { + "epoch": 0.5870886981402003, + "grad_norm": 1.2461633117791553, + "learning_rate": 7.686953552261966e-05, + "loss": 1.687, + "step": 6566 + }, + { + "epoch": 0.5871781115879828, + "grad_norm": 1.1452320017675843, + "learning_rate": 7.684136174217574e-05, + "loss": 1.7112, + "step": 6567 + }, + { + "epoch": 0.5872675250357654, + "grad_norm": 1.3048384320560658, + "learning_rate": 7.681318990401885e-05, + "loss": 1.8517, + "step": 6568 + }, + { + "epoch": 0.587356938483548, + "grad_norm": 1.5523731102233846, + "learning_rate": 7.678502001051168e-05, + "loss": 2.0436, + "step": 6569 + }, + { + "epoch": 0.5874463519313304, + "grad_norm": 1.4728108345981323, + "learning_rate": 7.675685206401689e-05, + "loss": 1.5393, + "step": 6570 + }, + { + "epoch": 0.587535765379113, + "grad_norm": 1.8063519834468125, + "learning_rate": 7.67286860668968e-05, + "loss": 1.7391, + "step": 6571 + }, + { + "epoch": 0.5876251788268956, + "grad_norm": 1.2265323197892086, + "learning_rate": 7.670052202151374e-05, + "loss": 1.8321, + "step": 6572 + }, + { + "epoch": 0.5877145922746781, + "grad_norm": 0.7012695275193325, + "learning_rate": 7.667235993022972e-05, + "loss": 1.9889, + "step": 6573 + }, + { + "epoch": 0.5878040057224606, + "grad_norm": 0.9237658570862796, + "learning_rate": 7.664419979540673e-05, + "loss": 1.5334, + "step": 6574 + }, + { + "epoch": 0.5878934191702432, + "grad_norm": 0.9644794357243063, + "learning_rate": 7.66160416194065e-05, + "loss": 2.0221, + "step": 6575 + }, + { + "epoch": 0.5879828326180258, + "grad_norm": 0.6441751298639011, + "learning_rate": 7.658788540459062e-05, + "loss": 1.8693, + "step": 6576 + }, + { + "epoch": 0.5880722460658083, + "grad_norm": 0.9288959905062785, + "learning_rate": 7.655973115332052e-05, + "loss": 1.5736, + "step": 6577 + }, + { + "epoch": 0.5881616595135909, + "grad_norm": 0.7183512138183218, + "learning_rate": 7.653157886795744e-05, + "loss": 1.7875, + "step": 6578 + }, + { + "epoch": 0.5882510729613734, + "grad_norm": 0.5429548313703877, + "learning_rate": 7.65034285508625e-05, + "loss": 1.5866, + "step": 6579 + }, + { + "epoch": 0.5883404864091559, + "grad_norm": 0.9951398868113132, + "learning_rate": 7.647528020439662e-05, + "loss": 1.5816, + "step": 6580 + }, + { + "epoch": 0.5884298998569385, + "grad_norm": 1.0879945992080173, + "learning_rate": 7.64471338309206e-05, + "loss": 1.6992, + "step": 6581 + }, + { + "epoch": 0.5885193133047211, + "grad_norm": 0.6920412722620991, + "learning_rate": 7.641898943279501e-05, + "loss": 2.0329, + "step": 6582 + }, + { + "epoch": 0.5886087267525035, + "grad_norm": 0.5417005860777215, + "learning_rate": 7.639084701238032e-05, + "loss": 1.6598, + "step": 6583 + }, + { + "epoch": 0.5886981402002861, + "grad_norm": 0.6564394165319358, + "learning_rate": 7.636270657203677e-05, + "loss": 1.7696, + "step": 6584 + }, + { + "epoch": 0.5887875536480687, + "grad_norm": 0.7967759151782017, + "learning_rate": 7.633456811412446e-05, + "loss": 1.7846, + "step": 6585 + }, + { + "epoch": 0.5888769670958512, + "grad_norm": 0.42156354336806, + "learning_rate": 7.630643164100335e-05, + "loss": 1.543, + "step": 6586 + }, + { + "epoch": 0.5889663805436338, + "grad_norm": 0.9302223307855698, + "learning_rate": 7.627829715503317e-05, + "loss": 2.1049, + "step": 6587 + }, + { + "epoch": 0.5890557939914163, + "grad_norm": 0.6032616193105631, + "learning_rate": 7.625016465857361e-05, + "loss": 2.0093, + "step": 6588 + }, + { + "epoch": 0.5891452074391988, + "grad_norm": 0.675860542544487, + "learning_rate": 7.622203415398402e-05, + "loss": 1.5779, + "step": 6589 + }, + { + "epoch": 0.5892346208869814, + "grad_norm": 0.9382193239908577, + "learning_rate": 7.619390564362374e-05, + "loss": 2.2399, + "step": 6590 + }, + { + "epoch": 0.589324034334764, + "grad_norm": 3.27502524243689, + "learning_rate": 7.616577912985185e-05, + "loss": 6.5542, + "step": 6591 + }, + { + "epoch": 0.5894134477825465, + "grad_norm": 1.0849302377595962, + "learning_rate": 7.613765461502724e-05, + "loss": 1.7158, + "step": 6592 + }, + { + "epoch": 0.589502861230329, + "grad_norm": 1.4060621051376885, + "learning_rate": 7.610953210150875e-05, + "loss": 1.6305, + "step": 6593 + }, + { + "epoch": 0.5895922746781116, + "grad_norm": 0.7865438680486359, + "learning_rate": 7.608141159165492e-05, + "loss": 1.8585, + "step": 6594 + }, + { + "epoch": 0.5896816881258942, + "grad_norm": 0.9738320001929073, + "learning_rate": 7.605329308782423e-05, + "loss": 1.9117, + "step": 6595 + }, + { + "epoch": 0.5897711015736766, + "grad_norm": 6.625095892725471, + "learning_rate": 7.602517659237492e-05, + "loss": 6.5427, + "step": 6596 + }, + { + "epoch": 0.5898605150214592, + "grad_norm": 0.730665332752895, + "learning_rate": 7.599706210766513e-05, + "loss": 1.6465, + "step": 6597 + }, + { + "epoch": 0.5899499284692418, + "grad_norm": 1.6764284056740109, + "learning_rate": 7.596894963605274e-05, + "loss": 1.7656, + "step": 6598 + }, + { + "epoch": 0.5900393419170243, + "grad_norm": 3.0557216860974057, + "learning_rate": 7.594083917989549e-05, + "loss": 1.7598, + "step": 6599 + }, + { + "epoch": 0.5901287553648069, + "grad_norm": 0.6710393235168893, + "learning_rate": 7.591273074155104e-05, + "loss": 1.7324, + "step": 6600 + }, + { + "epoch": 0.5902181688125894, + "grad_norm": 1.8262050753008268, + "learning_rate": 7.588462432337672e-05, + "loss": 1.6101, + "step": 6601 + }, + { + "epoch": 0.5903075822603719, + "grad_norm": 2.7199442786148813, + "learning_rate": 7.585651992772988e-05, + "loss": 1.7462, + "step": 6602 + }, + { + "epoch": 0.5903969957081545, + "grad_norm": 6.208466321620735, + "learning_rate": 7.582841755696754e-05, + "loss": 6.5083, + "step": 6603 + }, + { + "epoch": 0.5904864091559371, + "grad_norm": 1.233834960363346, + "learning_rate": 7.580031721344663e-05, + "loss": 1.7599, + "step": 6604 + }, + { + "epoch": 0.5905758226037195, + "grad_norm": 1.3952548127147948, + "learning_rate": 7.577221889952389e-05, + "loss": 1.5862, + "step": 6605 + }, + { + "epoch": 0.5906652360515021, + "grad_norm": 0.520960575501723, + "learning_rate": 7.57441226175559e-05, + "loss": 1.7493, + "step": 6606 + }, + { + "epoch": 0.5907546494992847, + "grad_norm": 1.6718233321500535, + "learning_rate": 7.571602836989906e-05, + "loss": 1.9489, + "step": 6607 + }, + { + "epoch": 0.5908440629470673, + "grad_norm": 0.6549980544882779, + "learning_rate": 7.568793615890954e-05, + "loss": 1.363, + "step": 6608 + }, + { + "epoch": 0.5909334763948498, + "grad_norm": 0.595121157722021, + "learning_rate": 7.565984598694349e-05, + "loss": 1.7978, + "step": 6609 + }, + { + "epoch": 0.5910228898426323, + "grad_norm": 1.1086515871475382, + "learning_rate": 7.563175785635678e-05, + "loss": 1.7508, + "step": 6610 + }, + { + "epoch": 0.5911123032904149, + "grad_norm": 0.9120744030676823, + "learning_rate": 7.56036717695051e-05, + "loss": 1.9238, + "step": 6611 + }, + { + "epoch": 0.5912017167381974, + "grad_norm": 0.7118599953633724, + "learning_rate": 7.557558772874398e-05, + "loss": 1.7062, + "step": 6612 + }, + { + "epoch": 0.59129113018598, + "grad_norm": 0.8085215781053217, + "learning_rate": 7.554750573642886e-05, + "loss": 1.8595, + "step": 6613 + }, + { + "epoch": 0.5913805436337625, + "grad_norm": 1.135179635755045, + "learning_rate": 7.551942579491489e-05, + "loss": 1.5549, + "step": 6614 + }, + { + "epoch": 0.591469957081545, + "grad_norm": 1.0426747552571365, + "learning_rate": 7.549134790655708e-05, + "loss": 1.8292, + "step": 6615 + }, + { + "epoch": 0.5915593705293276, + "grad_norm": 0.7219210121002089, + "learning_rate": 7.546327207371033e-05, + "loss": 1.4801, + "step": 6616 + }, + { + "epoch": 0.5916487839771102, + "grad_norm": 1.3062695904113089, + "learning_rate": 7.543519829872934e-05, + "loss": 1.8658, + "step": 6617 + }, + { + "epoch": 0.5917381974248928, + "grad_norm": 0.9523807050733787, + "learning_rate": 7.54071265839686e-05, + "loss": 1.4148, + "step": 6618 + }, + { + "epoch": 0.5918276108726752, + "grad_norm": 1.2558139750766868, + "learning_rate": 7.537905693178245e-05, + "loss": 2.0232, + "step": 6619 + }, + { + "epoch": 0.5919170243204578, + "grad_norm": 1.2748410303037234, + "learning_rate": 7.535098934452508e-05, + "loss": 2.1652, + "step": 6620 + }, + { + "epoch": 0.5920064377682404, + "grad_norm": 0.5760236783257447, + "learning_rate": 7.532292382455044e-05, + "loss": 1.5868, + "step": 6621 + }, + { + "epoch": 0.5920958512160229, + "grad_norm": 1.773141362259978, + "learning_rate": 7.529486037421235e-05, + "loss": 1.6146, + "step": 6622 + }, + { + "epoch": 0.5921852646638054, + "grad_norm": 1.9566014468470865, + "learning_rate": 7.52667989958645e-05, + "loss": 1.6023, + "step": 6623 + }, + { + "epoch": 0.592274678111588, + "grad_norm": 0.45862862485997397, + "learning_rate": 7.523873969186039e-05, + "loss": 1.6991, + "step": 6624 + }, + { + "epoch": 0.5923640915593705, + "grad_norm": 1.8582507185389883, + "learning_rate": 7.521068246455325e-05, + "loss": 1.7523, + "step": 6625 + }, + { + "epoch": 0.5924535050071531, + "grad_norm": 2.1247449231535156, + "learning_rate": 7.518262731629623e-05, + "loss": 1.9651, + "step": 6626 + }, + { + "epoch": 0.5925429184549357, + "grad_norm": 2.4869384631897207, + "learning_rate": 7.51545742494423e-05, + "loss": 6.5277, + "step": 6627 + }, + { + "epoch": 0.5926323319027181, + "grad_norm": 1.4371427602265519, + "learning_rate": 7.512652326634421e-05, + "loss": 1.965, + "step": 6628 + }, + { + "epoch": 0.5927217453505007, + "grad_norm": 1.3599784419770222, + "learning_rate": 7.509847436935455e-05, + "loss": 1.5639, + "step": 6629 + }, + { + "epoch": 0.5928111587982833, + "grad_norm": 0.7656804394669169, + "learning_rate": 7.50704275608258e-05, + "loss": 1.5577, + "step": 6630 + }, + { + "epoch": 0.5929005722460658, + "grad_norm": 1.4385420115983436, + "learning_rate": 7.504238284311019e-05, + "loss": 1.9399, + "step": 6631 + }, + { + "epoch": 0.5929899856938483, + "grad_norm": 1.4044727037858427, + "learning_rate": 7.501434021855977e-05, + "loss": 2.0644, + "step": 6632 + }, + { + "epoch": 0.5930793991416309, + "grad_norm": 1.3233901152737197, + "learning_rate": 7.498629968952648e-05, + "loss": 1.7471, + "step": 6633 + }, + { + "epoch": 0.5931688125894135, + "grad_norm": 1.0698780318125394, + "learning_rate": 7.495826125836203e-05, + "loss": 1.8266, + "step": 6634 + }, + { + "epoch": 0.593258226037196, + "grad_norm": 1.6498619486713424, + "learning_rate": 7.493022492741795e-05, + "loss": 1.5733, + "step": 6635 + }, + { + "epoch": 0.5933476394849786, + "grad_norm": 1.148982043288378, + "learning_rate": 7.49021906990456e-05, + "loss": 1.9808, + "step": 6636 + }, + { + "epoch": 0.593437052932761, + "grad_norm": 0.6391130360547062, + "learning_rate": 7.487415857559625e-05, + "loss": 1.4374, + "step": 6637 + }, + { + "epoch": 0.5935264663805436, + "grad_norm": 0.9053901442607939, + "learning_rate": 7.484612855942088e-05, + "loss": 1.8619, + "step": 6638 + }, + { + "epoch": 0.5936158798283262, + "grad_norm": 0.35313633098843317, + "learning_rate": 7.481810065287029e-05, + "loss": 1.4817, + "step": 6639 + }, + { + "epoch": 0.5937052932761088, + "grad_norm": 3.4848575744647197, + "learning_rate": 7.479007485829523e-05, + "loss": 6.6943, + "step": 6640 + }, + { + "epoch": 0.5937947067238912, + "grad_norm": 0.5124946256217238, + "learning_rate": 7.476205117804614e-05, + "loss": 1.6335, + "step": 6641 + }, + { + "epoch": 0.5938841201716738, + "grad_norm": 1.2579717173228107, + "learning_rate": 7.47340296144733e-05, + "loss": 1.8409, + "step": 6642 + }, + { + "epoch": 0.5939735336194564, + "grad_norm": 1.418461200684819, + "learning_rate": 7.470601016992687e-05, + "loss": 1.6817, + "step": 6643 + }, + { + "epoch": 0.594062947067239, + "grad_norm": 1.4990770177017763, + "learning_rate": 7.467799284675687e-05, + "loss": 1.6129, + "step": 6644 + }, + { + "epoch": 0.5941523605150214, + "grad_norm": 0.684103268304194, + "learning_rate": 7.464997764731304e-05, + "loss": 2.1642, + "step": 6645 + }, + { + "epoch": 0.594241773962804, + "grad_norm": 1.8003788372075271, + "learning_rate": 7.462196457394493e-05, + "loss": 1.7908, + "step": 6646 + }, + { + "epoch": 0.5943311874105865, + "grad_norm": 1.222144791189828, + "learning_rate": 7.459395362900201e-05, + "loss": 1.4937, + "step": 6647 + }, + { + "epoch": 0.5944206008583691, + "grad_norm": 1.1465543287544886, + "learning_rate": 7.456594481483355e-05, + "loss": 1.9794, + "step": 6648 + }, + { + "epoch": 0.5945100143061517, + "grad_norm": 1.4285849977625542, + "learning_rate": 7.453793813378853e-05, + "loss": 1.8427, + "step": 6649 + }, + { + "epoch": 0.5945994277539342, + "grad_norm": 1.9196582381918186, + "learning_rate": 7.450993358821589e-05, + "loss": 1.7463, + "step": 6650 + }, + { + "epoch": 0.5946888412017167, + "grad_norm": 1.0243490623666034, + "learning_rate": 7.448193118046435e-05, + "loss": 1.7225, + "step": 6651 + }, + { + "epoch": 0.5947782546494993, + "grad_norm": 0.6658162554362902, + "learning_rate": 7.445393091288247e-05, + "loss": 1.9641, + "step": 6652 + }, + { + "epoch": 0.5948676680972819, + "grad_norm": 1.4244371239450826, + "learning_rate": 7.442593278781848e-05, + "loss": 1.966, + "step": 6653 + }, + { + "epoch": 0.5949570815450643, + "grad_norm": 1.0361850816456533, + "learning_rate": 7.439793680762068e-05, + "loss": 1.6938, + "step": 6654 + }, + { + "epoch": 0.5950464949928469, + "grad_norm": 0.5266102632272847, + "learning_rate": 7.436994297463698e-05, + "loss": 1.7445, + "step": 6655 + }, + { + "epoch": 0.5951359084406295, + "grad_norm": 0.7727797685592711, + "learning_rate": 7.434195129121518e-05, + "loss": 1.9353, + "step": 6656 + }, + { + "epoch": 0.595225321888412, + "grad_norm": 1.357933399460367, + "learning_rate": 7.431396175970296e-05, + "loss": 1.8051, + "step": 6657 + }, + { + "epoch": 0.5953147353361946, + "grad_norm": 1.1073956282701238, + "learning_rate": 7.428597438244776e-05, + "loss": 1.8785, + "step": 6658 + }, + { + "epoch": 0.5954041487839771, + "grad_norm": 0.42837232735959596, + "learning_rate": 7.425798916179683e-05, + "loss": 1.8891, + "step": 6659 + }, + { + "epoch": 0.5954935622317596, + "grad_norm": 0.5888179662044306, + "learning_rate": 7.423000610009725e-05, + "loss": 1.6406, + "step": 6660 + }, + { + "epoch": 0.5955829756795422, + "grad_norm": 0.6576123654029926, + "learning_rate": 7.420202519969595e-05, + "loss": 1.6679, + "step": 6661 + }, + { + "epoch": 0.5956723891273248, + "grad_norm": 0.9333478525765216, + "learning_rate": 7.417404646293961e-05, + "loss": 1.7138, + "step": 6662 + }, + { + "epoch": 0.5957618025751072, + "grad_norm": 3.7618680871205874, + "learning_rate": 7.414606989217482e-05, + "loss": 6.5272, + "step": 6663 + }, + { + "epoch": 0.5958512160228898, + "grad_norm": 1.6243871063432358, + "learning_rate": 7.411809548974792e-05, + "loss": 1.5898, + "step": 6664 + }, + { + "epoch": 0.5959406294706724, + "grad_norm": 1.4932495030775959, + "learning_rate": 7.409012325800511e-05, + "loss": 1.5189, + "step": 6665 + }, + { + "epoch": 0.596030042918455, + "grad_norm": 0.760247594357995, + "learning_rate": 7.406215319929235e-05, + "loss": 1.6346, + "step": 6666 + }, + { + "epoch": 0.5961194563662375, + "grad_norm": 0.5104783316206147, + "learning_rate": 7.403418531595551e-05, + "loss": 1.7163, + "step": 6667 + }, + { + "epoch": 0.59620886981402, + "grad_norm": 1.5499513644523701, + "learning_rate": 7.400621961034018e-05, + "loss": 2.6763, + "step": 6668 + }, + { + "epoch": 0.5962982832618026, + "grad_norm": 1.270491245671205, + "learning_rate": 7.39782560847918e-05, + "loss": 1.6654, + "step": 6669 + }, + { + "epoch": 0.5963876967095851, + "grad_norm": 1.3646867325202618, + "learning_rate": 7.395029474165562e-05, + "loss": 2.1871, + "step": 6670 + }, + { + "epoch": 0.5964771101573677, + "grad_norm": 0.8890359246144929, + "learning_rate": 7.392233558327683e-05, + "loss": 1.6304, + "step": 6671 + }, + { + "epoch": 0.5965665236051502, + "grad_norm": 0.4193492767261564, + "learning_rate": 7.389437861200024e-05, + "loss": 1.7794, + "step": 6672 + }, + { + "epoch": 0.5966559370529327, + "grad_norm": 1.4989146189553673, + "learning_rate": 7.386642383017057e-05, + "loss": 1.6696, + "step": 6673 + }, + { + "epoch": 0.5967453505007153, + "grad_norm": 0.7220936249804157, + "learning_rate": 7.383847124013239e-05, + "loss": 1.4933, + "step": 6674 + }, + { + "epoch": 0.5968347639484979, + "grad_norm": 1.0815164213293724, + "learning_rate": 7.381052084423005e-05, + "loss": 1.5316, + "step": 6675 + }, + { + "epoch": 0.5969241773962805, + "grad_norm": 0.9065928309454752, + "learning_rate": 7.378257264480766e-05, + "loss": 1.9015, + "step": 6676 + }, + { + "epoch": 0.5970135908440629, + "grad_norm": 0.48452995297478424, + "learning_rate": 7.375462664420922e-05, + "loss": 1.8305, + "step": 6677 + }, + { + "epoch": 0.5971030042918455, + "grad_norm": 1.1632506616813687, + "learning_rate": 7.37266828447786e-05, + "loss": 1.8763, + "step": 6678 + }, + { + "epoch": 0.5971924177396281, + "grad_norm": 0.7502632560807753, + "learning_rate": 7.369874124885934e-05, + "loss": 1.565, + "step": 6679 + }, + { + "epoch": 0.5972818311874106, + "grad_norm": 1.0549410973698186, + "learning_rate": 7.367080185879489e-05, + "loss": 1.6327, + "step": 6680 + }, + { + "epoch": 0.5973712446351931, + "grad_norm": 0.6836828054197506, + "learning_rate": 7.364286467692848e-05, + "loss": 2.0759, + "step": 6681 + }, + { + "epoch": 0.5974606580829757, + "grad_norm": 0.8407469581527797, + "learning_rate": 7.361492970560322e-05, + "loss": 1.9773, + "step": 6682 + }, + { + "epoch": 0.5975500715307582, + "grad_norm": 0.8074611805321463, + "learning_rate": 7.358699694716189e-05, + "loss": 1.6475, + "step": 6683 + }, + { + "epoch": 0.5976394849785408, + "grad_norm": 1.5503563657233501, + "learning_rate": 7.35590664039472e-05, + "loss": 1.786, + "step": 6684 + }, + { + "epoch": 0.5977288984263234, + "grad_norm": 0.519698837399605, + "learning_rate": 7.353113807830175e-05, + "loss": 1.8407, + "step": 6685 + }, + { + "epoch": 0.5978183118741058, + "grad_norm": 0.473760311156471, + "learning_rate": 7.350321197256777e-05, + "loss": 1.6293, + "step": 6686 + }, + { + "epoch": 0.5979077253218884, + "grad_norm": 0.8301196392510366, + "learning_rate": 7.347528808908737e-05, + "loss": 2.067, + "step": 6687 + }, + { + "epoch": 0.597997138769671, + "grad_norm": 1.447219625155929, + "learning_rate": 7.344736643020256e-05, + "loss": 1.7948, + "step": 6688 + }, + { + "epoch": 0.5980865522174535, + "grad_norm": 0.5924538198677981, + "learning_rate": 7.341944699825503e-05, + "loss": 1.6701, + "step": 6689 + }, + { + "epoch": 0.598175965665236, + "grad_norm": 1.106350102260914, + "learning_rate": 7.33915297955864e-05, + "loss": 1.501, + "step": 6690 + }, + { + "epoch": 0.5982653791130186, + "grad_norm": 0.771350500398248, + "learning_rate": 7.3363614824538e-05, + "loss": 1.6829, + "step": 6691 + }, + { + "epoch": 0.5983547925608012, + "grad_norm": 0.46394106570907623, + "learning_rate": 7.333570208745109e-05, + "loss": 1.6367, + "step": 6692 + }, + { + "epoch": 0.5984442060085837, + "grad_norm": 0.6163909771266056, + "learning_rate": 7.330779158666661e-05, + "loss": 2.0232, + "step": 6693 + }, + { + "epoch": 0.5985336194563662, + "grad_norm": 0.6324788171850027, + "learning_rate": 7.327988332452545e-05, + "loss": 1.6278, + "step": 6694 + }, + { + "epoch": 0.5986230329041488, + "grad_norm": 0.5988368011205563, + "learning_rate": 7.325197730336819e-05, + "loss": 1.9627, + "step": 6695 + }, + { + "epoch": 0.5987124463519313, + "grad_norm": 1.578788568769494, + "learning_rate": 7.322407352553529e-05, + "loss": 2.1211, + "step": 6696 + }, + { + "epoch": 0.5988018597997139, + "grad_norm": 0.631616613632536, + "learning_rate": 7.319617199336701e-05, + "loss": 2.1271, + "step": 6697 + }, + { + "epoch": 0.5988912732474965, + "grad_norm": 1.7225336705962553, + "learning_rate": 7.316827270920339e-05, + "loss": 1.3283, + "step": 6698 + }, + { + "epoch": 0.5989806866952789, + "grad_norm": 0.9681136995080148, + "learning_rate": 7.314037567538436e-05, + "loss": 1.7813, + "step": 6699 + }, + { + "epoch": 0.5990701001430615, + "grad_norm": 0.8356567810964248, + "learning_rate": 7.311248089424958e-05, + "loss": 1.6085, + "step": 6700 + }, + { + "epoch": 0.5991595135908441, + "grad_norm": 0.583426114212829, + "learning_rate": 7.308458836813856e-05, + "loss": 1.5171, + "step": 6701 + }, + { + "epoch": 0.5992489270386266, + "grad_norm": 2.025504562592817, + "learning_rate": 7.305669809939062e-05, + "loss": 1.876, + "step": 6702 + }, + { + "epoch": 0.5993383404864091, + "grad_norm": 6.216550103678121, + "learning_rate": 7.302881009034484e-05, + "loss": 6.5724, + "step": 6703 + }, + { + "epoch": 0.5994277539341917, + "grad_norm": 2.3157921998556783, + "learning_rate": 7.30009243433402e-05, + "loss": 1.7137, + "step": 6704 + }, + { + "epoch": 0.5995171673819742, + "grad_norm": 2.034638858185497, + "learning_rate": 7.29730408607154e-05, + "loss": 1.7089, + "step": 6705 + }, + { + "epoch": 0.5996065808297568, + "grad_norm": 1.0088109031410548, + "learning_rate": 7.294515964480906e-05, + "loss": 1.7486, + "step": 6706 + }, + { + "epoch": 0.5996959942775394, + "grad_norm": 1.3748853945971, + "learning_rate": 7.291728069795948e-05, + "loss": 1.4789, + "step": 6707 + }, + { + "epoch": 0.5997854077253219, + "grad_norm": 1.6828505324256218, + "learning_rate": 7.28894040225049e-05, + "loss": 1.8605, + "step": 6708 + }, + { + "epoch": 0.5998748211731044, + "grad_norm": 0.9981876675786396, + "learning_rate": 7.286152962078326e-05, + "loss": 1.7152, + "step": 6709 + }, + { + "epoch": 0.599964234620887, + "grad_norm": 0.7696785986021666, + "learning_rate": 7.283365749513231e-05, + "loss": 2.0534, + "step": 6710 + }, + { + "epoch": 0.6000536480686696, + "grad_norm": 6.061641138281632, + "learning_rate": 7.280578764788975e-05, + "loss": 6.5517, + "step": 6711 + }, + { + "epoch": 0.600143061516452, + "grad_norm": 1.3098047074882688, + "learning_rate": 7.277792008139287e-05, + "loss": 1.6272, + "step": 6712 + }, + { + "epoch": 0.6002324749642346, + "grad_norm": 1.6247706016038834, + "learning_rate": 7.2750054797979e-05, + "loss": 1.9888, + "step": 6713 + }, + { + "epoch": 0.6003218884120172, + "grad_norm": 0.995755312437303, + "learning_rate": 7.272219179998511e-05, + "loss": 1.711, + "step": 6714 + }, + { + "epoch": 0.6004113018597997, + "grad_norm": 1.0304936914150113, + "learning_rate": 7.269433108974809e-05, + "loss": 1.8203, + "step": 6715 + }, + { + "epoch": 0.6005007153075823, + "grad_norm": 1.0179546511869586, + "learning_rate": 7.266647266960452e-05, + "loss": 2.5672, + "step": 6716 + }, + { + "epoch": 0.6005901287553648, + "grad_norm": 1.2627957017756537, + "learning_rate": 7.263861654189086e-05, + "loss": 2.1291, + "step": 6717 + }, + { + "epoch": 0.6006795422031473, + "grad_norm": 0.8666870207570224, + "learning_rate": 7.261076270894342e-05, + "loss": 1.6047, + "step": 6718 + }, + { + "epoch": 0.6007689556509299, + "grad_norm": 1.5446501873450773, + "learning_rate": 7.258291117309817e-05, + "loss": 2.0916, + "step": 6719 + }, + { + "epoch": 0.6008583690987125, + "grad_norm": 1.7389232590290884, + "learning_rate": 7.25550619366911e-05, + "loss": 1.6515, + "step": 6720 + }, + { + "epoch": 0.600947782546495, + "grad_norm": 1.5648283513475503, + "learning_rate": 7.252721500205783e-05, + "loss": 1.8461, + "step": 6721 + }, + { + "epoch": 0.6010371959942775, + "grad_norm": 2.4183548698042725, + "learning_rate": 7.249937037153387e-05, + "loss": 1.8568, + "step": 6722 + }, + { + "epoch": 0.6011266094420601, + "grad_norm": 0.8396586245800302, + "learning_rate": 7.24715280474545e-05, + "loss": 1.3752, + "step": 6723 + }, + { + "epoch": 0.6012160228898427, + "grad_norm": 1.448175148508641, + "learning_rate": 7.244368803215482e-05, + "loss": 1.7687, + "step": 6724 + }, + { + "epoch": 0.6013054363376252, + "grad_norm": 3.1083326118676915, + "learning_rate": 7.241585032796977e-05, + "loss": 6.7396, + "step": 6725 + }, + { + "epoch": 0.6013948497854077, + "grad_norm": 1.653646471758331, + "learning_rate": 7.238801493723398e-05, + "loss": 1.5945, + "step": 6726 + }, + { + "epoch": 0.6014842632331903, + "grad_norm": 1.9000630223096093, + "learning_rate": 7.236018186228206e-05, + "loss": 1.9091, + "step": 6727 + }, + { + "epoch": 0.6015736766809728, + "grad_norm": 1.2870796362000674, + "learning_rate": 7.233235110544833e-05, + "loss": 2.0302, + "step": 6728 + }, + { + "epoch": 0.6016630901287554, + "grad_norm": 0.8666322718594368, + "learning_rate": 7.230452266906689e-05, + "loss": 1.6249, + "step": 6729 + }, + { + "epoch": 0.6017525035765379, + "grad_norm": 2.4885892782951564, + "learning_rate": 7.227669655547167e-05, + "loss": 2.0058, + "step": 6730 + }, + { + "epoch": 0.6018419170243204, + "grad_norm": 1.2159469456143162, + "learning_rate": 7.224887276699645e-05, + "loss": 1.3612, + "step": 6731 + }, + { + "epoch": 0.601931330472103, + "grad_norm": 1.4005033964030507, + "learning_rate": 7.222105130597477e-05, + "loss": 1.8877, + "step": 6732 + }, + { + "epoch": 0.6020207439198856, + "grad_norm": 1.6573817939857673, + "learning_rate": 7.21932321747399e-05, + "loss": 1.6803, + "step": 6733 + }, + { + "epoch": 0.602110157367668, + "grad_norm": 1.032992722980614, + "learning_rate": 7.21654153756251e-05, + "loss": 1.7187, + "step": 6734 + }, + { + "epoch": 0.6021995708154506, + "grad_norm": 2.1042645418117187, + "learning_rate": 7.213760091096331e-05, + "loss": 1.6346, + "step": 6735 + }, + { + "epoch": 0.6022889842632332, + "grad_norm": 2.0534516791773547, + "learning_rate": 7.210978878308729e-05, + "loss": 1.875, + "step": 6736 + }, + { + "epoch": 0.6023783977110158, + "grad_norm": 0.4542229077915978, + "learning_rate": 7.208197899432958e-05, + "loss": 1.6504, + "step": 6737 + }, + { + "epoch": 0.6024678111587983, + "grad_norm": 1.0093234762542092, + "learning_rate": 7.20541715470226e-05, + "loss": 1.65, + "step": 6738 + }, + { + "epoch": 0.6025572246065808, + "grad_norm": 1.2361389983910924, + "learning_rate": 7.202636644349845e-05, + "loss": 1.6842, + "step": 6739 + }, + { + "epoch": 0.6026466380543634, + "grad_norm": 1.3818642818799798, + "learning_rate": 7.199856368608922e-05, + "loss": 2.1735, + "step": 6740 + }, + { + "epoch": 0.6027360515021459, + "grad_norm": 0.9845078221363892, + "learning_rate": 7.197076327712659e-05, + "loss": 1.9218, + "step": 6741 + }, + { + "epoch": 0.6028254649499285, + "grad_norm": 1.2019701134614982, + "learning_rate": 7.194296521894223e-05, + "loss": 1.6777, + "step": 6742 + }, + { + "epoch": 0.602914878397711, + "grad_norm": 1.117520958374545, + "learning_rate": 7.191516951386751e-05, + "loss": 1.8087, + "step": 6743 + }, + { + "epoch": 0.6030042918454935, + "grad_norm": 1.1201386363480126, + "learning_rate": 7.188737616423356e-05, + "loss": 1.9136, + "step": 6744 + }, + { + "epoch": 0.6030937052932761, + "grad_norm": 1.276795496281168, + "learning_rate": 7.185958517237146e-05, + "loss": 1.7046, + "step": 6745 + }, + { + "epoch": 0.6031831187410587, + "grad_norm": 0.7480162110655876, + "learning_rate": 7.183179654061191e-05, + "loss": 1.572, + "step": 6746 + }, + { + "epoch": 0.6032725321888412, + "grad_norm": 0.8728714079457692, + "learning_rate": 7.18040102712856e-05, + "loss": 1.552, + "step": 6747 + }, + { + "epoch": 0.6033619456366237, + "grad_norm": 0.42342483563682165, + "learning_rate": 7.17762263667229e-05, + "loss": 1.7964, + "step": 6748 + }, + { + "epoch": 0.6034513590844063, + "grad_norm": 0.7844641809579409, + "learning_rate": 7.1748444829254e-05, + "loss": 1.7182, + "step": 6749 + }, + { + "epoch": 0.6035407725321889, + "grad_norm": 0.5279340783143412, + "learning_rate": 7.172066566120892e-05, + "loss": 1.5142, + "step": 6750 + }, + { + "epoch": 0.6036301859799714, + "grad_norm": 0.9667479991645693, + "learning_rate": 7.169288886491746e-05, + "loss": 1.9368, + "step": 6751 + }, + { + "epoch": 0.6037195994277539, + "grad_norm": 0.7237334674215038, + "learning_rate": 7.166511444270924e-05, + "loss": 1.5578, + "step": 6752 + }, + { + "epoch": 0.6038090128755365, + "grad_norm": 0.5806746317457926, + "learning_rate": 7.16373423969136e-05, + "loss": 1.5287, + "step": 6753 + }, + { + "epoch": 0.603898426323319, + "grad_norm": 0.48912852684343616, + "learning_rate": 7.160957272985982e-05, + "loss": 1.9183, + "step": 6754 + }, + { + "epoch": 0.6039878397711016, + "grad_norm": 1.0055363102611654, + "learning_rate": 7.158180544387691e-05, + "loss": 1.8891, + "step": 6755 + }, + { + "epoch": 0.6040772532188842, + "grad_norm": 0.6437594267047039, + "learning_rate": 7.155404054129366e-05, + "loss": 1.9355, + "step": 6756 + }, + { + "epoch": 0.6041666666666666, + "grad_norm": 0.5943519010110869, + "learning_rate": 7.152627802443866e-05, + "loss": 1.7724, + "step": 6757 + }, + { + "epoch": 0.6042560801144492, + "grad_norm": 0.47721775966132585, + "learning_rate": 7.149851789564034e-05, + "loss": 1.4897, + "step": 6758 + }, + { + "epoch": 0.6043454935622318, + "grad_norm": 1.1540064642176349, + "learning_rate": 7.147076015722691e-05, + "loss": 1.8343, + "step": 6759 + }, + { + "epoch": 0.6044349070100143, + "grad_norm": 0.7474934726662269, + "learning_rate": 7.144300481152633e-05, + "loss": 1.566, + "step": 6760 + }, + { + "epoch": 0.6045243204577968, + "grad_norm": 0.7218410079720603, + "learning_rate": 7.141525186086647e-05, + "loss": 1.8919, + "step": 6761 + }, + { + "epoch": 0.6046137339055794, + "grad_norm": 0.4045172473474334, + "learning_rate": 7.138750130757493e-05, + "loss": 1.5728, + "step": 6762 + }, + { + "epoch": 0.604703147353362, + "grad_norm": 0.5712675006689861, + "learning_rate": 7.135975315397912e-05, + "loss": 1.8417, + "step": 6763 + }, + { + "epoch": 0.6047925608011445, + "grad_norm": 0.6291290961791799, + "learning_rate": 7.133200740240618e-05, + "loss": 1.5006, + "step": 6764 + }, + { + "epoch": 0.6048819742489271, + "grad_norm": 0.6516937320310678, + "learning_rate": 7.130426405518318e-05, + "loss": 2.1653, + "step": 6765 + }, + { + "epoch": 0.6049713876967096, + "grad_norm": 0.8174061252168736, + "learning_rate": 7.127652311463691e-05, + "loss": 1.9579, + "step": 6766 + }, + { + "epoch": 0.6050608011444921, + "grad_norm": 1.1985008774631303, + "learning_rate": 7.124878458309391e-05, + "loss": 1.7254, + "step": 6767 + }, + { + "epoch": 0.6051502145922747, + "grad_norm": 0.5517575673420785, + "learning_rate": 7.122104846288064e-05, + "loss": 2.1074, + "step": 6768 + }, + { + "epoch": 0.6052396280400573, + "grad_norm": 0.915572590449816, + "learning_rate": 7.119331475632332e-05, + "loss": 2.0519, + "step": 6769 + }, + { + "epoch": 0.6053290414878397, + "grad_norm": 1.8770695521057126, + "learning_rate": 7.116558346574788e-05, + "loss": 1.7834, + "step": 6770 + }, + { + "epoch": 0.6054184549356223, + "grad_norm": 1.181391329270799, + "learning_rate": 7.113785459348012e-05, + "loss": 1.7627, + "step": 6771 + }, + { + "epoch": 0.6055078683834049, + "grad_norm": 0.39898090733958197, + "learning_rate": 7.111012814184566e-05, + "loss": 1.6985, + "step": 6772 + }, + { + "epoch": 0.6055972818311874, + "grad_norm": 0.8903802220397776, + "learning_rate": 7.108240411316986e-05, + "loss": 1.4022, + "step": 6773 + }, + { + "epoch": 0.60568669527897, + "grad_norm": 0.9110746551383752, + "learning_rate": 7.105468250977786e-05, + "loss": 1.7664, + "step": 6774 + }, + { + "epoch": 0.6057761087267525, + "grad_norm": 0.49700721820222643, + "learning_rate": 7.10269633339947e-05, + "loss": 1.8871, + "step": 6775 + }, + { + "epoch": 0.605865522174535, + "grad_norm": 0.6381852080714389, + "learning_rate": 7.099924658814517e-05, + "loss": 1.6125, + "step": 6776 + }, + { + "epoch": 0.6059549356223176, + "grad_norm": 0.4476302662617976, + "learning_rate": 7.097153227455379e-05, + "loss": 1.7935, + "step": 6777 + }, + { + "epoch": 0.6060443490701002, + "grad_norm": 1.0382701252916613, + "learning_rate": 7.094382039554493e-05, + "loss": 2.1164, + "step": 6778 + }, + { + "epoch": 0.6061337625178826, + "grad_norm": 0.7532353933056665, + "learning_rate": 7.091611095344277e-05, + "loss": 1.6447, + "step": 6779 + }, + { + "epoch": 0.6062231759656652, + "grad_norm": 0.8320519682163141, + "learning_rate": 7.088840395057124e-05, + "loss": 1.5202, + "step": 6780 + }, + { + "epoch": 0.6063125894134478, + "grad_norm": 0.7524766385946466, + "learning_rate": 7.086069938925411e-05, + "loss": 1.7002, + "step": 6781 + }, + { + "epoch": 0.6064020028612304, + "grad_norm": 1.0789441213801751, + "learning_rate": 7.083299727181495e-05, + "loss": 1.6496, + "step": 6782 + }, + { + "epoch": 0.6064914163090128, + "grad_norm": 0.7389829526482183, + "learning_rate": 7.080529760057709e-05, + "loss": 1.8883, + "step": 6783 + }, + { + "epoch": 0.6065808297567954, + "grad_norm": 0.7593801486996256, + "learning_rate": 7.077760037786365e-05, + "loss": 1.9253, + "step": 6784 + }, + { + "epoch": 0.606670243204578, + "grad_norm": 3.4499365681119274, + "learning_rate": 7.074990560599759e-05, + "loss": 2.0066, + "step": 6785 + }, + { + "epoch": 0.6067596566523605, + "grad_norm": 1.1680515552818052, + "learning_rate": 7.072221328730162e-05, + "loss": 2.1482, + "step": 6786 + }, + { + "epoch": 0.6068490701001431, + "grad_norm": 1.6102660386425554, + "learning_rate": 7.069452342409825e-05, + "loss": 1.8035, + "step": 6787 + }, + { + "epoch": 0.6069384835479256, + "grad_norm": 1.6614089997920671, + "learning_rate": 7.066683601870978e-05, + "loss": 1.816, + "step": 6788 + }, + { + "epoch": 0.6070278969957081, + "grad_norm": 0.5042294981980366, + "learning_rate": 7.063915107345839e-05, + "loss": 1.8495, + "step": 6789 + }, + { + "epoch": 0.6071173104434907, + "grad_norm": 1.3913889797884489, + "learning_rate": 7.061146859066594e-05, + "loss": 1.8165, + "step": 6790 + }, + { + "epoch": 0.6072067238912733, + "grad_norm": 1.4407239025481677, + "learning_rate": 7.058378857265411e-05, + "loss": 1.7784, + "step": 6791 + }, + { + "epoch": 0.6072961373390557, + "grad_norm": 0.6480393888575475, + "learning_rate": 7.055611102174442e-05, + "loss": 1.5477, + "step": 6792 + }, + { + "epoch": 0.6073855507868383, + "grad_norm": 1.2524544273390774, + "learning_rate": 7.052843594025815e-05, + "loss": 1.6013, + "step": 6793 + }, + { + "epoch": 0.6074749642346209, + "grad_norm": 1.5067163510284458, + "learning_rate": 7.050076333051634e-05, + "loss": 1.9723, + "step": 6794 + }, + { + "epoch": 0.6075643776824035, + "grad_norm": 0.9156240008803632, + "learning_rate": 7.047309319483985e-05, + "loss": 2.0021, + "step": 6795 + }, + { + "epoch": 0.607653791130186, + "grad_norm": 1.4354050666670526, + "learning_rate": 7.044542553554943e-05, + "loss": 1.8458, + "step": 6796 + }, + { + "epoch": 0.6077432045779685, + "grad_norm": 4.352597271669645, + "learning_rate": 7.041776035496547e-05, + "loss": 6.4942, + "step": 6797 + }, + { + "epoch": 0.6078326180257511, + "grad_norm": 3.512652682328465, + "learning_rate": 7.039009765540822e-05, + "loss": 1.8643, + "step": 6798 + }, + { + "epoch": 0.6079220314735336, + "grad_norm": 2.8527906842807678, + "learning_rate": 7.036243743919773e-05, + "loss": 1.72, + "step": 6799 + }, + { + "epoch": 0.6080114449213162, + "grad_norm": 0.7007370407427025, + "learning_rate": 7.033477970865381e-05, + "loss": 1.4908, + "step": 6800 + }, + { + "epoch": 0.6081008583690987, + "grad_norm": 1.4686207399718125, + "learning_rate": 7.030712446609608e-05, + "loss": 1.6003, + "step": 6801 + }, + { + "epoch": 0.6081902718168812, + "grad_norm": 1.9534762681166336, + "learning_rate": 7.027947171384394e-05, + "loss": 1.7673, + "step": 6802 + }, + { + "epoch": 0.6082796852646638, + "grad_norm": 1.4572967011059563, + "learning_rate": 7.025182145421665e-05, + "loss": 1.6852, + "step": 6803 + }, + { + "epoch": 0.6083690987124464, + "grad_norm": 1.0922595486252442, + "learning_rate": 7.022417368953317e-05, + "loss": 1.6573, + "step": 6804 + }, + { + "epoch": 0.608458512160229, + "grad_norm": 2.2233656081965134, + "learning_rate": 7.019652842211226e-05, + "loss": 2.1995, + "step": 6805 + }, + { + "epoch": 0.6085479256080114, + "grad_norm": 0.5530125886045059, + "learning_rate": 7.016888565427253e-05, + "loss": 1.6618, + "step": 6806 + }, + { + "epoch": 0.608637339055794, + "grad_norm": 1.3798868915849414, + "learning_rate": 7.014124538833234e-05, + "loss": 1.4917, + "step": 6807 + }, + { + "epoch": 0.6087267525035766, + "grad_norm": 1.4208656674692781, + "learning_rate": 7.011360762660983e-05, + "loss": 1.5389, + "step": 6808 + }, + { + "epoch": 0.6088161659513591, + "grad_norm": 1.800444588574211, + "learning_rate": 7.008597237142293e-05, + "loss": 1.8183, + "step": 6809 + }, + { + "epoch": 0.6089055793991416, + "grad_norm": 0.7589355815548615, + "learning_rate": 7.005833962508943e-05, + "loss": 1.6323, + "step": 6810 + }, + { + "epoch": 0.6089949928469242, + "grad_norm": 2.532256910460711, + "learning_rate": 7.003070938992682e-05, + "loss": 1.8866, + "step": 6811 + }, + { + "epoch": 0.6090844062947067, + "grad_norm": 1.449059418879243, + "learning_rate": 7.000308166825243e-05, + "loss": 2.0168, + "step": 6812 + }, + { + "epoch": 0.6091738197424893, + "grad_norm": 1.8198130230214844, + "learning_rate": 6.997545646238335e-05, + "loss": 1.4801, + "step": 6813 + }, + { + "epoch": 0.6092632331902719, + "grad_norm": 1.2947769881193953, + "learning_rate": 6.994783377463645e-05, + "loss": 1.8982, + "step": 6814 + }, + { + "epoch": 0.6093526466380543, + "grad_norm": 1.3556170631945708, + "learning_rate": 6.992021360732848e-05, + "loss": 1.5383, + "step": 6815 + }, + { + "epoch": 0.6094420600858369, + "grad_norm": 1.1356619801849286, + "learning_rate": 6.989259596277582e-05, + "loss": 1.3068, + "step": 6816 + }, + { + "epoch": 0.6095314735336195, + "grad_norm": 2.1407006898293255, + "learning_rate": 6.98649808432948e-05, + "loss": 2.0346, + "step": 6817 + }, + { + "epoch": 0.609620886981402, + "grad_norm": 9.018447277684636, + "learning_rate": 6.983736825120144e-05, + "loss": 6.6303, + "step": 6818 + }, + { + "epoch": 0.6097103004291845, + "grad_norm": 5.56758000274621, + "learning_rate": 6.980975818881159e-05, + "loss": 1.9524, + "step": 6819 + }, + { + "epoch": 0.6097997138769671, + "grad_norm": 4.473308793349436, + "learning_rate": 6.978215065844087e-05, + "loss": 1.5057, + "step": 6820 + }, + { + "epoch": 0.6098891273247496, + "grad_norm": 0.5170316738283827, + "learning_rate": 6.975454566240465e-05, + "loss": 1.4825, + "step": 6821 + }, + { + "epoch": 0.6099785407725322, + "grad_norm": 3.0983056674428657, + "learning_rate": 6.972694320301813e-05, + "loss": 2.004, + "step": 6822 + }, + { + "epoch": 0.6100679542203148, + "grad_norm": 2.0173323982020888, + "learning_rate": 6.969934328259637e-05, + "loss": 1.6225, + "step": 6823 + }, + { + "epoch": 0.6101573676680973, + "grad_norm": 1.1560508134032257, + "learning_rate": 6.96717459034541e-05, + "loss": 1.3867, + "step": 6824 + }, + { + "epoch": 0.6102467811158798, + "grad_norm": 2.066775021518527, + "learning_rate": 6.964415106790586e-05, + "loss": 1.756, + "step": 6825 + }, + { + "epoch": 0.6103361945636624, + "grad_norm": 1.9025074166610219, + "learning_rate": 6.961655877826603e-05, + "loss": 1.6738, + "step": 6826 + }, + { + "epoch": 0.610425608011445, + "grad_norm": 2.20622019785495, + "learning_rate": 6.95889690368487e-05, + "loss": 1.8891, + "step": 6827 + }, + { + "epoch": 0.6105150214592274, + "grad_norm": 1.7468880341414412, + "learning_rate": 6.956138184596782e-05, + "loss": 1.9868, + "step": 6828 + }, + { + "epoch": 0.61060443490701, + "grad_norm": 1.5879100816499039, + "learning_rate": 6.953379720793703e-05, + "loss": 1.8819, + "step": 6829 + }, + { + "epoch": 0.6106938483547926, + "grad_norm": 1.4256516735703701, + "learning_rate": 6.950621512506993e-05, + "loss": 1.6138, + "step": 6830 + }, + { + "epoch": 0.6107832618025751, + "grad_norm": 1.2746964058688313, + "learning_rate": 6.947863559967976e-05, + "loss": 2.0613, + "step": 6831 + }, + { + "epoch": 0.6108726752503576, + "grad_norm": 1.4367521860595585, + "learning_rate": 6.945105863407951e-05, + "loss": 1.8926, + "step": 6832 + }, + { + "epoch": 0.6109620886981402, + "grad_norm": 2.735439699135291, + "learning_rate": 6.942348423058212e-05, + "loss": 2.0116, + "step": 6833 + }, + { + "epoch": 0.6110515021459227, + "grad_norm": 0.5983057494093351, + "learning_rate": 6.939591239150014e-05, + "loss": 1.6019, + "step": 6834 + }, + { + "epoch": 0.6111409155937053, + "grad_norm": 2.686443536195329, + "learning_rate": 6.936834311914606e-05, + "loss": 1.9149, + "step": 6835 + }, + { + "epoch": 0.6112303290414879, + "grad_norm": 1.7845331966249742, + "learning_rate": 6.934077641583201e-05, + "loss": 1.8179, + "step": 6836 + }, + { + "epoch": 0.6113197424892703, + "grad_norm": 1.1657120911459302, + "learning_rate": 6.931321228387005e-05, + "loss": 1.7601, + "step": 6837 + }, + { + "epoch": 0.6114091559370529, + "grad_norm": 1.8093423675635336, + "learning_rate": 6.928565072557191e-05, + "loss": 1.8422, + "step": 6838 + }, + { + "epoch": 0.6114985693848355, + "grad_norm": 3.664220041584136, + "learning_rate": 6.925809174324915e-05, + "loss": 1.6635, + "step": 6839 + }, + { + "epoch": 0.6115879828326181, + "grad_norm": 0.9784974398807905, + "learning_rate": 6.923053533921312e-05, + "loss": 2.1366, + "step": 6840 + }, + { + "epoch": 0.6116773962804005, + "grad_norm": 2.553160231989245, + "learning_rate": 6.920298151577491e-05, + "loss": 1.9295, + "step": 6841 + }, + { + "epoch": 0.6117668097281831, + "grad_norm": 3.273938388485559, + "learning_rate": 6.917543027524546e-05, + "loss": 1.8774, + "step": 6842 + }, + { + "epoch": 0.6118562231759657, + "grad_norm": 0.9964779646167456, + "learning_rate": 6.914788161993542e-05, + "loss": 1.6177, + "step": 6843 + }, + { + "epoch": 0.6119456366237482, + "grad_norm": 1.934632638719112, + "learning_rate": 6.912033555215532e-05, + "loss": 1.5139, + "step": 6844 + }, + { + "epoch": 0.6120350500715308, + "grad_norm": 2.7571117747769867, + "learning_rate": 6.909279207421536e-05, + "loss": 1.7548, + "step": 6845 + }, + { + "epoch": 0.6121244635193133, + "grad_norm": 0.9321536776724305, + "learning_rate": 6.906525118842563e-05, + "loss": 1.7342, + "step": 6846 + }, + { + "epoch": 0.6122138769670958, + "grad_norm": 1.2562898163981706, + "learning_rate": 6.903771289709591e-05, + "loss": 1.6578, + "step": 6847 + }, + { + "epoch": 0.6123032904148784, + "grad_norm": 1.5627094472843726, + "learning_rate": 6.901017720253583e-05, + "loss": 2.0685, + "step": 6848 + }, + { + "epoch": 0.612392703862661, + "grad_norm": 1.5606208762038958, + "learning_rate": 6.898264410705475e-05, + "loss": 1.5162, + "step": 6849 + }, + { + "epoch": 0.6124821173104434, + "grad_norm": 0.7201349390508338, + "learning_rate": 6.89551136129618e-05, + "loss": 2.0849, + "step": 6850 + }, + { + "epoch": 0.612571530758226, + "grad_norm": 1.326019768203148, + "learning_rate": 6.892758572256604e-05, + "loss": 1.8277, + "step": 6851 + }, + { + "epoch": 0.6126609442060086, + "grad_norm": 1.0930987251884827, + "learning_rate": 6.890006043817612e-05, + "loss": 1.8318, + "step": 6852 + }, + { + "epoch": 0.6127503576537912, + "grad_norm": 1.8265591145498132, + "learning_rate": 6.887253776210058e-05, + "loss": 2.2117, + "step": 6853 + }, + { + "epoch": 0.6128397711015737, + "grad_norm": 1.0600066739199014, + "learning_rate": 6.884501769664773e-05, + "loss": 2.1777, + "step": 6854 + }, + { + "epoch": 0.6129291845493562, + "grad_norm": 0.8782378336250124, + "learning_rate": 6.881750024412557e-05, + "loss": 1.6366, + "step": 6855 + }, + { + "epoch": 0.6130185979971388, + "grad_norm": 1.9413131563543031, + "learning_rate": 6.878998540684206e-05, + "loss": 1.5774, + "step": 6856 + }, + { + "epoch": 0.6131080114449213, + "grad_norm": 1.7734505241740615, + "learning_rate": 6.876247318710471e-05, + "loss": 2.0639, + "step": 6857 + }, + { + "epoch": 0.6131974248927039, + "grad_norm": 0.5961718705540063, + "learning_rate": 6.873496358722105e-05, + "loss": 2.0015, + "step": 6858 + }, + { + "epoch": 0.6132868383404864, + "grad_norm": 1.9991460198790092, + "learning_rate": 6.870745660949822e-05, + "loss": 1.8146, + "step": 6859 + }, + { + "epoch": 0.6133762517882689, + "grad_norm": 1.5370962232592356, + "learning_rate": 6.867995225624324e-05, + "loss": 1.7582, + "step": 6860 + }, + { + "epoch": 0.6134656652360515, + "grad_norm": 1.0627004609577697, + "learning_rate": 6.865245052976284e-05, + "loss": 1.7241, + "step": 6861 + }, + { + "epoch": 0.6135550786838341, + "grad_norm": 0.545118805522884, + "learning_rate": 6.862495143236353e-05, + "loss": 1.5151, + "step": 6862 + }, + { + "epoch": 0.6136444921316166, + "grad_norm": 2.2621647475942246, + "learning_rate": 6.859745496635167e-05, + "loss": 1.7374, + "step": 6863 + }, + { + "epoch": 0.6137339055793991, + "grad_norm": 1.0275974672321, + "learning_rate": 6.85699611340333e-05, + "loss": 1.9544, + "step": 6864 + }, + { + "epoch": 0.6138233190271817, + "grad_norm": 1.9389269976436703, + "learning_rate": 6.854246993771438e-05, + "loss": 2.013, + "step": 6865 + }, + { + "epoch": 0.6139127324749643, + "grad_norm": 1.0820135128153783, + "learning_rate": 6.851498137970049e-05, + "loss": 1.9271, + "step": 6866 + }, + { + "epoch": 0.6140021459227468, + "grad_norm": 1.9222237826106556, + "learning_rate": 6.84874954622971e-05, + "loss": 1.5935, + "step": 6867 + }, + { + "epoch": 0.6140915593705293, + "grad_norm": 3.001730393429673, + "learning_rate": 6.84600121878094e-05, + "loss": 2.2178, + "step": 6868 + }, + { + "epoch": 0.6141809728183119, + "grad_norm": 1.0049865873150619, + "learning_rate": 6.843253155854239e-05, + "loss": 1.6417, + "step": 6869 + }, + { + "epoch": 0.6142703862660944, + "grad_norm": 1.5191528803114656, + "learning_rate": 6.840505357680085e-05, + "loss": 1.7942, + "step": 6870 + }, + { + "epoch": 0.614359799713877, + "grad_norm": 14.079832251573796, + "learning_rate": 6.837757824488927e-05, + "loss": 6.6385, + "step": 6871 + }, + { + "epoch": 0.6144492131616596, + "grad_norm": 5.973851817022312, + "learning_rate": 6.835010556511201e-05, + "loss": 1.8041, + "step": 6872 + }, + { + "epoch": 0.614538626609442, + "grad_norm": 8.02736853669745, + "learning_rate": 6.832263553977321e-05, + "loss": 1.7687, + "step": 6873 + }, + { + "epoch": 0.6146280400572246, + "grad_norm": 3.124331964404777, + "learning_rate": 6.829516817117671e-05, + "loss": 1.8994, + "step": 6874 + }, + { + "epoch": 0.6147174535050072, + "grad_norm": 4.066432436183494, + "learning_rate": 6.826770346162614e-05, + "loss": 1.9904, + "step": 6875 + }, + { + "epoch": 0.6148068669527897, + "grad_norm": 4.098500567653356, + "learning_rate": 6.8240241413425e-05, + "loss": 1.5334, + "step": 6876 + }, + { + "epoch": 0.6148962804005722, + "grad_norm": 6.4023620462176805, + "learning_rate": 6.821278202887643e-05, + "loss": 2.119, + "step": 6877 + }, + { + "epoch": 0.6149856938483548, + "grad_norm": 2.5569844331423375, + "learning_rate": 6.818532531028342e-05, + "loss": 1.5432, + "step": 6878 + }, + { + "epoch": 0.6150751072961373, + "grad_norm": 15.545223141809979, + "learning_rate": 6.815787125994875e-05, + "loss": 6.689, + "step": 6879 + }, + { + "epoch": 0.6151645207439199, + "grad_norm": 4.583147733941015, + "learning_rate": 6.813041988017501e-05, + "loss": 1.8771, + "step": 6880 + }, + { + "epoch": 0.6152539341917024, + "grad_norm": 1.222693443610963, + "learning_rate": 6.810297117326445e-05, + "loss": 1.6209, + "step": 6881 + }, + { + "epoch": 0.615343347639485, + "grad_norm": 5.015431671766717, + "learning_rate": 6.807552514151915e-05, + "loss": 1.889, + "step": 6882 + }, + { + "epoch": 0.6154327610872675, + "grad_norm": 6.147768931753927, + "learning_rate": 6.804808178724105e-05, + "loss": 1.9866, + "step": 6883 + }, + { + "epoch": 0.6155221745350501, + "grad_norm": 1.5720321396810844, + "learning_rate": 6.802064111273173e-05, + "loss": 1.7892, + "step": 6884 + }, + { + "epoch": 0.6156115879828327, + "grad_norm": 3.1605824760694663, + "learning_rate": 6.799320312029256e-05, + "loss": 1.759, + "step": 6885 + }, + { + "epoch": 0.6157010014306151, + "grad_norm": 3.2521780521054007, + "learning_rate": 6.796576781222481e-05, + "loss": 1.3038, + "step": 6886 + }, + { + "epoch": 0.6157904148783977, + "grad_norm": 4.69365682201812, + "learning_rate": 6.793833519082946e-05, + "loss": 2.292, + "step": 6887 + }, + { + "epoch": 0.6158798283261803, + "grad_norm": 1.7710438242494444, + "learning_rate": 6.791090525840722e-05, + "loss": 2.0016, + "step": 6888 + }, + { + "epoch": 0.6159692417739628, + "grad_norm": 2.346861237274326, + "learning_rate": 6.788347801725859e-05, + "loss": 1.6252, + "step": 6889 + }, + { + "epoch": 0.6160586552217453, + "grad_norm": 4.065538205513124, + "learning_rate": 6.785605346968386e-05, + "loss": 2.0262, + "step": 6890 + }, + { + "epoch": 0.6161480686695279, + "grad_norm": 2.141406512969733, + "learning_rate": 6.782863161798311e-05, + "loss": 1.5381, + "step": 6891 + }, + { + "epoch": 0.6162374821173104, + "grad_norm": 2.1030302871891844, + "learning_rate": 6.780121246445617e-05, + "loss": 1.4345, + "step": 6892 + }, + { + "epoch": 0.616326895565093, + "grad_norm": 3.903400052468564, + "learning_rate": 6.777379601140264e-05, + "loss": 1.7252, + "step": 6893 + }, + { + "epoch": 0.6164163090128756, + "grad_norm": 8.499935382071527, + "learning_rate": 6.774638226112195e-05, + "loss": 6.7621, + "step": 6894 + }, + { + "epoch": 0.616505722460658, + "grad_norm": 5.370043368004912, + "learning_rate": 6.771897121591321e-05, + "loss": 2.0915, + "step": 6895 + }, + { + "epoch": 0.6165951359084406, + "grad_norm": 9.30219623793083, + "learning_rate": 6.76915628780754e-05, + "loss": 2.1342, + "step": 6896 + }, + { + "epoch": 0.6166845493562232, + "grad_norm": 2.1588575686361233, + "learning_rate": 6.766415724990718e-05, + "loss": 1.5221, + "step": 6897 + }, + { + "epoch": 0.6167739628040058, + "grad_norm": 3.634453490894981, + "learning_rate": 6.7636754333707e-05, + "loss": 1.5889, + "step": 6898 + }, + { + "epoch": 0.6168633762517882, + "grad_norm": 3.3527752188456756, + "learning_rate": 6.760935413177316e-05, + "loss": 1.8794, + "step": 6899 + }, + { + "epoch": 0.6169527896995708, + "grad_norm": 3.021202450991168, + "learning_rate": 6.75819566464037e-05, + "loss": 2.0294, + "step": 6900 + }, + { + "epoch": 0.6170422031473534, + "grad_norm": 4.3374413358989, + "learning_rate": 6.755456187989637e-05, + "loss": 1.6408, + "step": 6901 + }, + { + "epoch": 0.6171316165951359, + "grad_norm": 2.9945078049130722, + "learning_rate": 6.752716983454875e-05, + "loss": 1.533, + "step": 6902 + }, + { + "epoch": 0.6172210300429185, + "grad_norm": 1.9651288510950147, + "learning_rate": 6.749978051265819e-05, + "loss": 1.8904, + "step": 6903 + }, + { + "epoch": 0.617310443490701, + "grad_norm": 0.8271896637224959, + "learning_rate": 6.74723939165218e-05, + "loss": 1.3125, + "step": 6904 + }, + { + "epoch": 0.6173998569384835, + "grad_norm": 3.4815487984248734, + "learning_rate": 6.74450100484364e-05, + "loss": 1.7353, + "step": 6905 + }, + { + "epoch": 0.6174892703862661, + "grad_norm": 1.9779814222360392, + "learning_rate": 6.741762891069871e-05, + "loss": 1.8743, + "step": 6906 + }, + { + "epoch": 0.6175786838340487, + "grad_norm": 1.2367837709016842, + "learning_rate": 6.739025050560514e-05, + "loss": 1.5846, + "step": 6907 + }, + { + "epoch": 0.6176680972818311, + "grad_norm": 2.8621009068424983, + "learning_rate": 6.736287483545191e-05, + "loss": 1.5487, + "step": 6908 + }, + { + "epoch": 0.6177575107296137, + "grad_norm": 0.9954330470716786, + "learning_rate": 6.73355019025349e-05, + "loss": 1.6377, + "step": 6909 + }, + { + "epoch": 0.6178469241773963, + "grad_norm": 1.0447685295920668, + "learning_rate": 6.730813170914993e-05, + "loss": 1.7973, + "step": 6910 + }, + { + "epoch": 0.6179363376251789, + "grad_norm": 1.5772805637592198, + "learning_rate": 6.72807642575925e-05, + "loss": 1.3883, + "step": 6911 + }, + { + "epoch": 0.6180257510729614, + "grad_norm": 1.0153304954239455, + "learning_rate": 6.725339955015777e-05, + "loss": 1.8937, + "step": 6912 + }, + { + "epoch": 0.6181151645207439, + "grad_norm": 0.7992258631788806, + "learning_rate": 6.722603758914092e-05, + "loss": 1.6673, + "step": 6913 + }, + { + "epoch": 0.6182045779685265, + "grad_norm": 0.8235242379354528, + "learning_rate": 6.719867837683672e-05, + "loss": 1.9055, + "step": 6914 + }, + { + "epoch": 0.618293991416309, + "grad_norm": 1.0680770137943192, + "learning_rate": 6.717132191553977e-05, + "loss": 1.9341, + "step": 6915 + }, + { + "epoch": 0.6183834048640916, + "grad_norm": 0.7809240561622602, + "learning_rate": 6.714396820754436e-05, + "loss": 2.0866, + "step": 6916 + }, + { + "epoch": 0.6184728183118741, + "grad_norm": 0.619598578572507, + "learning_rate": 6.711661725514469e-05, + "loss": 1.9437, + "step": 6917 + }, + { + "epoch": 0.6185622317596566, + "grad_norm": 0.8565487426319724, + "learning_rate": 6.708926906063462e-05, + "loss": 1.7685, + "step": 6918 + }, + { + "epoch": 0.6186516452074392, + "grad_norm": 1.5910317977567436, + "learning_rate": 6.706192362630776e-05, + "loss": 1.7488, + "step": 6919 + }, + { + "epoch": 0.6187410586552218, + "grad_norm": 0.923187517620561, + "learning_rate": 6.70345809544576e-05, + "loss": 2.2146, + "step": 6920 + }, + { + "epoch": 0.6188304721030042, + "grad_norm": 1.3077688472688718, + "learning_rate": 6.700724104737736e-05, + "loss": 1.5753, + "step": 6921 + }, + { + "epoch": 0.6189198855507868, + "grad_norm": 1.6687595577298464, + "learning_rate": 6.697990390735997e-05, + "loss": 1.9654, + "step": 6922 + }, + { + "epoch": 0.6190092989985694, + "grad_norm": 0.3771551974059099, + "learning_rate": 6.695256953669812e-05, + "loss": 1.6052, + "step": 6923 + }, + { + "epoch": 0.619098712446352, + "grad_norm": 0.5885414234592253, + "learning_rate": 6.69252379376844e-05, + "loss": 1.6661, + "step": 6924 + }, + { + "epoch": 0.6191881258941345, + "grad_norm": 1.344369361211118, + "learning_rate": 6.689790911261099e-05, + "loss": 1.9644, + "step": 6925 + }, + { + "epoch": 0.619277539341917, + "grad_norm": 1.014320001034831, + "learning_rate": 6.687058306376997e-05, + "loss": 1.2318, + "step": 6926 + }, + { + "epoch": 0.6193669527896996, + "grad_norm": 0.5583005292451163, + "learning_rate": 6.684325979345315e-05, + "loss": 1.9371, + "step": 6927 + }, + { + "epoch": 0.6194563662374821, + "grad_norm": 0.7097625062031159, + "learning_rate": 6.681593930395209e-05, + "loss": 1.8315, + "step": 6928 + }, + { + "epoch": 0.6195457796852647, + "grad_norm": 0.8345058923446047, + "learning_rate": 6.678862159755809e-05, + "loss": 1.6295, + "step": 6929 + }, + { + "epoch": 0.6196351931330472, + "grad_norm": 0.8817543313849754, + "learning_rate": 6.676130667656235e-05, + "loss": 1.5544, + "step": 6930 + }, + { + "epoch": 0.6197246065808297, + "grad_norm": 1.1279498926934466, + "learning_rate": 6.673399454325565e-05, + "loss": 1.9179, + "step": 6931 + }, + { + "epoch": 0.6198140200286123, + "grad_norm": 0.8642981630705668, + "learning_rate": 6.670668519992864e-05, + "loss": 1.8067, + "step": 6932 + }, + { + "epoch": 0.6199034334763949, + "grad_norm": 0.59757806985686, + "learning_rate": 6.66793786488717e-05, + "loss": 1.6013, + "step": 6933 + }, + { + "epoch": 0.6199928469241774, + "grad_norm": 0.8300375401222958, + "learning_rate": 6.66520748923751e-05, + "loss": 1.7959, + "step": 6934 + }, + { + "epoch": 0.6200822603719599, + "grad_norm": 1.3228934413790359, + "learning_rate": 6.662477393272869e-05, + "loss": 2.0993, + "step": 6935 + }, + { + "epoch": 0.6201716738197425, + "grad_norm": 0.6080869902396736, + "learning_rate": 6.659747577222216e-05, + "loss": 1.7459, + "step": 6936 + }, + { + "epoch": 0.620261087267525, + "grad_norm": 0.648645361251727, + "learning_rate": 6.657018041314502e-05, + "loss": 1.4758, + "step": 6937 + }, + { + "epoch": 0.6203505007153076, + "grad_norm": 0.8016577771255294, + "learning_rate": 6.654288785778646e-05, + "loss": 1.5792, + "step": 6938 + }, + { + "epoch": 0.6204399141630901, + "grad_norm": 0.667566761461345, + "learning_rate": 6.651559810843548e-05, + "loss": 1.8968, + "step": 6939 + }, + { + "epoch": 0.6205293276108726, + "grad_norm": 1.3532814638516277, + "learning_rate": 6.648831116738083e-05, + "loss": 1.8768, + "step": 6940 + }, + { + "epoch": 0.6206187410586552, + "grad_norm": 1.0208435636629587, + "learning_rate": 6.646102703691111e-05, + "loss": 1.8815, + "step": 6941 + }, + { + "epoch": 0.6207081545064378, + "grad_norm": 0.6148358888308568, + "learning_rate": 6.643374571931451e-05, + "loss": 1.8557, + "step": 6942 + }, + { + "epoch": 0.6207975679542204, + "grad_norm": 1.2852796209180795, + "learning_rate": 6.640646721687913e-05, + "loss": 2.1711, + "step": 6943 + }, + { + "epoch": 0.6208869814020028, + "grad_norm": 1.0887909881570874, + "learning_rate": 6.637919153189279e-05, + "loss": 2.0924, + "step": 6944 + }, + { + "epoch": 0.6209763948497854, + "grad_norm": 0.7164200827672786, + "learning_rate": 6.635191866664303e-05, + "loss": 2.3057, + "step": 6945 + }, + { + "epoch": 0.621065808297568, + "grad_norm": 0.648845474315388, + "learning_rate": 6.632464862341721e-05, + "loss": 1.7859, + "step": 6946 + }, + { + "epoch": 0.6211552217453505, + "grad_norm": 0.47243868422112767, + "learning_rate": 6.629738140450241e-05, + "loss": 1.7982, + "step": 6947 + }, + { + "epoch": 0.621244635193133, + "grad_norm": 0.6815564052364791, + "learning_rate": 6.62701170121856e-05, + "loss": 1.4636, + "step": 6948 + }, + { + "epoch": 0.6213340486409156, + "grad_norm": 0.7339722039576867, + "learning_rate": 6.62428554487533e-05, + "loss": 1.6738, + "step": 6949 + }, + { + "epoch": 0.6214234620886981, + "grad_norm": 0.5061334645707547, + "learning_rate": 6.621559671649196e-05, + "loss": 1.5289, + "step": 6950 + }, + { + "epoch": 0.6215128755364807, + "grad_norm": 0.4520283062980822, + "learning_rate": 6.618834081768772e-05, + "loss": 1.5904, + "step": 6951 + }, + { + "epoch": 0.6216022889842633, + "grad_norm": 0.6672505721939149, + "learning_rate": 6.616108775462649e-05, + "loss": 1.7364, + "step": 6952 + }, + { + "epoch": 0.6216917024320457, + "grad_norm": 0.34642601055233835, + "learning_rate": 6.613383752959398e-05, + "loss": 1.5365, + "step": 6953 + }, + { + "epoch": 0.6217811158798283, + "grad_norm": 0.5539255704367634, + "learning_rate": 6.610659014487557e-05, + "loss": 1.7883, + "step": 6954 + }, + { + "epoch": 0.6218705293276109, + "grad_norm": 1.040323450076254, + "learning_rate": 6.607934560275657e-05, + "loss": 1.9853, + "step": 6955 + }, + { + "epoch": 0.6219599427753935, + "grad_norm": 0.5191251710612247, + "learning_rate": 6.605210390552185e-05, + "loss": 1.9465, + "step": 6956 + }, + { + "epoch": 0.6220493562231759, + "grad_norm": 0.7532362872228437, + "learning_rate": 6.602486505545621e-05, + "loss": 1.7882, + "step": 6957 + }, + { + "epoch": 0.6221387696709585, + "grad_norm": 0.9860323845938942, + "learning_rate": 6.59976290548441e-05, + "loss": 1.4646, + "step": 6958 + }, + { + "epoch": 0.6222281831187411, + "grad_norm": 0.7730810575538225, + "learning_rate": 6.597039590596976e-05, + "loss": 1.7837, + "step": 6959 + }, + { + "epoch": 0.6223175965665236, + "grad_norm": 0.6503952127218847, + "learning_rate": 6.594316561111724e-05, + "loss": 1.8231, + "step": 6960 + }, + { + "epoch": 0.6224070100143062, + "grad_norm": 2.55553718276518, + "learning_rate": 6.591593817257025e-05, + "loss": 6.5132, + "step": 6961 + }, + { + "epoch": 0.6224964234620887, + "grad_norm": 1.0625894728339684, + "learning_rate": 6.58887135926124e-05, + "loss": 1.5857, + "step": 6962 + }, + { + "epoch": 0.6225858369098712, + "grad_norm": 1.2187158803749032, + "learning_rate": 6.58614918735269e-05, + "loss": 1.7313, + "step": 6963 + }, + { + "epoch": 0.6226752503576538, + "grad_norm": 1.2202280375402708, + "learning_rate": 6.58342730175969e-05, + "loss": 2.0966, + "step": 6964 + }, + { + "epoch": 0.6227646638054364, + "grad_norm": 1.2662014224885363, + "learning_rate": 6.580705702710514e-05, + "loss": 1.7743, + "step": 6965 + }, + { + "epoch": 0.6228540772532188, + "grad_norm": 1.3880753921895976, + "learning_rate": 6.577984390433421e-05, + "loss": 1.8196, + "step": 6966 + }, + { + "epoch": 0.6229434907010014, + "grad_norm": 1.0737906796241972, + "learning_rate": 6.575263365156647e-05, + "loss": 1.6876, + "step": 6967 + }, + { + "epoch": 0.623032904148784, + "grad_norm": 1.0693302511746223, + "learning_rate": 6.572542627108393e-05, + "loss": 1.4032, + "step": 6968 + }, + { + "epoch": 0.6231223175965666, + "grad_norm": 1.2370353920745016, + "learning_rate": 6.569822176516853e-05, + "loss": 1.9314, + "step": 6969 + }, + { + "epoch": 0.623211731044349, + "grad_norm": 0.4919998848077668, + "learning_rate": 6.567102013610184e-05, + "loss": 1.3922, + "step": 6970 + }, + { + "epoch": 0.6233011444921316, + "grad_norm": 0.9376748603909927, + "learning_rate": 6.564382138616526e-05, + "loss": 1.7577, + "step": 6971 + }, + { + "epoch": 0.6233905579399142, + "grad_norm": 0.7720007946546734, + "learning_rate": 6.561662551763984e-05, + "loss": 1.8868, + "step": 6972 + }, + { + "epoch": 0.6234799713876967, + "grad_norm": 1.0593212473956062, + "learning_rate": 6.558943253280654e-05, + "loss": 1.2939, + "step": 6973 + }, + { + "epoch": 0.6235693848354793, + "grad_norm": 1.6696105010035105, + "learning_rate": 6.556224243394599e-05, + "loss": 1.7359, + "step": 6974 + }, + { + "epoch": 0.6236587982832618, + "grad_norm": 0.7631944460903385, + "learning_rate": 6.553505522333853e-05, + "loss": 1.5441, + "step": 6975 + }, + { + "epoch": 0.6237482117310443, + "grad_norm": 0.5811573367438886, + "learning_rate": 6.55078709032644e-05, + "loss": 1.9313, + "step": 6976 + }, + { + "epoch": 0.6238376251788269, + "grad_norm": 1.329753898881683, + "learning_rate": 6.548068947600346e-05, + "loss": 1.6914, + "step": 6977 + }, + { + "epoch": 0.6239270386266095, + "grad_norm": 0.9720610588292318, + "learning_rate": 6.545351094383544e-05, + "loss": 1.4258, + "step": 6978 + }, + { + "epoch": 0.6240164520743919, + "grad_norm": 0.503003937819595, + "learning_rate": 6.542633530903972e-05, + "loss": 1.6632, + "step": 6979 + }, + { + "epoch": 0.6241058655221745, + "grad_norm": 0.7054692770656014, + "learning_rate": 6.53991625738955e-05, + "loss": 2.6279, + "step": 6980 + }, + { + "epoch": 0.6241952789699571, + "grad_norm": 1.1471214011528865, + "learning_rate": 6.537199274068173e-05, + "loss": 1.7635, + "step": 6981 + }, + { + "epoch": 0.6242846924177397, + "grad_norm": 2.048186618193391, + "learning_rate": 6.534482581167707e-05, + "loss": 2.5067, + "step": 6982 + }, + { + "epoch": 0.6243741058655222, + "grad_norm": 1.6685379293630025, + "learning_rate": 6.531766178916008e-05, + "loss": 1.8696, + "step": 6983 + }, + { + "epoch": 0.6244635193133047, + "grad_norm": 1.3242662817814344, + "learning_rate": 6.529050067540887e-05, + "loss": 1.6593, + "step": 6984 + }, + { + "epoch": 0.6245529327610873, + "grad_norm": 2.194956552455627, + "learning_rate": 6.526334247270147e-05, + "loss": 1.5782, + "step": 6985 + }, + { + "epoch": 0.6246423462088698, + "grad_norm": 0.9695096544599456, + "learning_rate": 6.523618718331557e-05, + "loss": 1.6087, + "step": 6986 + }, + { + "epoch": 0.6247317596566524, + "grad_norm": 1.2769911236381697, + "learning_rate": 6.520903480952869e-05, + "loss": 2.0632, + "step": 6987 + }, + { + "epoch": 0.6248211731044349, + "grad_norm": 2.040997508603396, + "learning_rate": 6.518188535361803e-05, + "loss": 1.8901, + "step": 6988 + }, + { + "epoch": 0.6249105865522174, + "grad_norm": 2.021029233687889, + "learning_rate": 6.515473881786056e-05, + "loss": 1.3666, + "step": 6989 + }, + { + "epoch": 0.625, + "grad_norm": 1.3889716595646604, + "learning_rate": 6.512759520453308e-05, + "loss": 2.0554, + "step": 6990 + }, + { + "epoch": 0.6250894134477826, + "grad_norm": 2.191673596041293, + "learning_rate": 6.510045451591211e-05, + "loss": 6.6699, + "step": 6991 + }, + { + "epoch": 0.6251788268955651, + "grad_norm": 2.137392711062321, + "learning_rate": 6.507331675427387e-05, + "loss": 1.9074, + "step": 6992 + }, + { + "epoch": 0.6252682403433476, + "grad_norm": 1.6727691490990055, + "learning_rate": 6.504618192189435e-05, + "loss": 1.6565, + "step": 6993 + }, + { + "epoch": 0.6253576537911302, + "grad_norm": 0.7305910759654369, + "learning_rate": 6.501905002104935e-05, + "loss": 1.7843, + "step": 6994 + }, + { + "epoch": 0.6254470672389127, + "grad_norm": 2.864926321431669, + "learning_rate": 6.499192105401435e-05, + "loss": 1.6654, + "step": 6995 + }, + { + "epoch": 0.6255364806866953, + "grad_norm": 1.3348895822753388, + "learning_rate": 6.49647950230647e-05, + "loss": 1.5654, + "step": 6996 + }, + { + "epoch": 0.6256258941344778, + "grad_norm": 0.4723421861110578, + "learning_rate": 6.493767193047534e-05, + "loss": 1.5293, + "step": 6997 + }, + { + "epoch": 0.6257153075822603, + "grad_norm": 1.5327966664283648, + "learning_rate": 6.491055177852111e-05, + "loss": 1.8989, + "step": 6998 + }, + { + "epoch": 0.6258047210300429, + "grad_norm": 2.0329800770081103, + "learning_rate": 6.488343456947654e-05, + "loss": 1.8055, + "step": 6999 + }, + { + "epoch": 0.6258941344778255, + "grad_norm": 1.0730944273276317, + "learning_rate": 6.485632030561587e-05, + "loss": 1.8633, + "step": 7000 + }, + { + "epoch": 0.6259835479256081, + "grad_norm": 0.9062891783350528, + "learning_rate": 6.48292089892132e-05, + "loss": 1.9167, + "step": 7001 + }, + { + "epoch": 0.6260729613733905, + "grad_norm": 2.112308808336771, + "learning_rate": 6.480210062254225e-05, + "loss": 1.7257, + "step": 7002 + }, + { + "epoch": 0.6261623748211731, + "grad_norm": 1.2985080162205982, + "learning_rate": 6.477499520787665e-05, + "loss": 1.8316, + "step": 7003 + }, + { + "epoch": 0.6262517882689557, + "grad_norm": 0.7226668251410621, + "learning_rate": 6.474789274748964e-05, + "loss": 1.8424, + "step": 7004 + }, + { + "epoch": 0.6263412017167382, + "grad_norm": 1.1171140150962073, + "learning_rate": 6.472079324365433e-05, + "loss": 1.842, + "step": 7005 + }, + { + "epoch": 0.6264306151645207, + "grad_norm": 2.24360082097764, + "learning_rate": 6.469369669864346e-05, + "loss": 1.8147, + "step": 7006 + }, + { + "epoch": 0.6265200286123033, + "grad_norm": 2.142409359893413, + "learning_rate": 6.466660311472962e-05, + "loss": 6.6014, + "step": 7007 + }, + { + "epoch": 0.6266094420600858, + "grad_norm": 1.7242909634986578, + "learning_rate": 6.46395124941851e-05, + "loss": 2.0235, + "step": 7008 + }, + { + "epoch": 0.6266988555078684, + "grad_norm": 0.534341584016889, + "learning_rate": 6.461242483928194e-05, + "loss": 1.4876, + "step": 7009 + }, + { + "epoch": 0.626788268955651, + "grad_norm": 0.569504940340179, + "learning_rate": 6.4585340152292e-05, + "loss": 1.485, + "step": 7010 + }, + { + "epoch": 0.6268776824034334, + "grad_norm": 1.1551877598872664, + "learning_rate": 6.455825843548678e-05, + "loss": 2.1283, + "step": 7011 + }, + { + "epoch": 0.626967095851216, + "grad_norm": 1.1429296410373302, + "learning_rate": 6.453117969113767e-05, + "loss": 1.3973, + "step": 7012 + }, + { + "epoch": 0.6270565092989986, + "grad_norm": 1.2805484174395627, + "learning_rate": 6.450410392151564e-05, + "loss": 1.9945, + "step": 7013 + }, + { + "epoch": 0.6271459227467812, + "grad_norm": 0.5551836428403378, + "learning_rate": 6.447703112889158e-05, + "loss": 1.4711, + "step": 7014 + }, + { + "epoch": 0.6272353361945636, + "grad_norm": 1.3443036196219007, + "learning_rate": 6.4449961315536e-05, + "loss": 1.7378, + "step": 7015 + }, + { + "epoch": 0.6273247496423462, + "grad_norm": 0.997824557378801, + "learning_rate": 6.44228944837192e-05, + "loss": 1.4544, + "step": 7016 + }, + { + "epoch": 0.6274141630901288, + "grad_norm": 0.771965210468782, + "learning_rate": 6.43958306357113e-05, + "loss": 1.7864, + "step": 7017 + }, + { + "epoch": 0.6275035765379113, + "grad_norm": 0.606311843976514, + "learning_rate": 6.43687697737821e-05, + "loss": 2.3359, + "step": 7018 + }, + { + "epoch": 0.6275929899856938, + "grad_norm": 0.6842518454339627, + "learning_rate": 6.434171190020116e-05, + "loss": 1.3999, + "step": 7019 + }, + { + "epoch": 0.6276824034334764, + "grad_norm": 1.1379252236367101, + "learning_rate": 6.431465701723774e-05, + "loss": 1.4781, + "step": 7020 + }, + { + "epoch": 0.6277718168812589, + "grad_norm": 0.903915910128561, + "learning_rate": 6.428760512716096e-05, + "loss": 1.6953, + "step": 7021 + }, + { + "epoch": 0.6278612303290415, + "grad_norm": 0.8859012188931539, + "learning_rate": 6.426055623223963e-05, + "loss": 2.0297, + "step": 7022 + }, + { + "epoch": 0.6279506437768241, + "grad_norm": 1.4014370129062383, + "learning_rate": 6.423351033474223e-05, + "loss": 1.7433, + "step": 7023 + }, + { + "epoch": 0.6280400572246065, + "grad_norm": 0.7639816540605964, + "learning_rate": 6.420646743693714e-05, + "loss": 1.7955, + "step": 7024 + }, + { + "epoch": 0.6281294706723891, + "grad_norm": 0.4310737107128032, + "learning_rate": 6.417942754109242e-05, + "loss": 1.5301, + "step": 7025 + }, + { + "epoch": 0.6282188841201717, + "grad_norm": 1.5817028219010518, + "learning_rate": 6.415239064947587e-05, + "loss": 1.5888, + "step": 7026 + }, + { + "epoch": 0.6283082975679543, + "grad_norm": 3.947338654720035, + "learning_rate": 6.412535676435498e-05, + "loss": 6.5339, + "step": 7027 + }, + { + "epoch": 0.6283977110157367, + "grad_norm": 0.6263316837501627, + "learning_rate": 6.409832588799713e-05, + "loss": 1.6936, + "step": 7028 + }, + { + "epoch": 0.6284871244635193, + "grad_norm": 1.8553029151537475, + "learning_rate": 6.407129802266932e-05, + "loss": 1.694, + "step": 7029 + }, + { + "epoch": 0.6285765379113019, + "grad_norm": 1.4734057237472902, + "learning_rate": 6.404427317063832e-05, + "loss": 1.8737, + "step": 7030 + }, + { + "epoch": 0.6286659513590844, + "grad_norm": 1.4536310464785442, + "learning_rate": 6.401725133417071e-05, + "loss": 2.1561, + "step": 7031 + }, + { + "epoch": 0.628755364806867, + "grad_norm": 0.9993403194834606, + "learning_rate": 6.39902325155328e-05, + "loss": 1.7661, + "step": 7032 + }, + { + "epoch": 0.6288447782546495, + "grad_norm": 1.605557040374245, + "learning_rate": 6.396321671699061e-05, + "loss": 1.6169, + "step": 7033 + }, + { + "epoch": 0.628934191702432, + "grad_norm": 1.1863336837967888, + "learning_rate": 6.39362039408099e-05, + "loss": 1.6919, + "step": 7034 + }, + { + "epoch": 0.6290236051502146, + "grad_norm": 0.6653718496774691, + "learning_rate": 6.39091941892562e-05, + "loss": 1.6097, + "step": 7035 + }, + { + "epoch": 0.6291130185979972, + "grad_norm": 0.8477976742610615, + "learning_rate": 6.388218746459483e-05, + "loss": 2.0604, + "step": 7036 + }, + { + "epoch": 0.6292024320457796, + "grad_norm": 1.6578776642730086, + "learning_rate": 6.385518376909072e-05, + "loss": 1.907, + "step": 7037 + }, + { + "epoch": 0.6292918454935622, + "grad_norm": 1.0217698463159444, + "learning_rate": 6.382818310500871e-05, + "loss": 1.7506, + "step": 7038 + }, + { + "epoch": 0.6293812589413448, + "grad_norm": 1.355715921086487, + "learning_rate": 6.380118547461334e-05, + "loss": 2.0252, + "step": 7039 + }, + { + "epoch": 0.6294706723891274, + "grad_norm": 0.6740875794186745, + "learning_rate": 6.377419088016881e-05, + "loss": 1.7914, + "step": 7040 + }, + { + "epoch": 0.6295600858369099, + "grad_norm": 1.1948542646849831, + "learning_rate": 6.374719932393913e-05, + "loss": 1.7648, + "step": 7041 + }, + { + "epoch": 0.6296494992846924, + "grad_norm": 1.7094251729104672, + "learning_rate": 6.372021080818808e-05, + "loss": 1.8372, + "step": 7042 + }, + { + "epoch": 0.629738912732475, + "grad_norm": 1.1639713549828652, + "learning_rate": 6.36932253351791e-05, + "loss": 1.9033, + "step": 7043 + }, + { + "epoch": 0.6298283261802575, + "grad_norm": 0.6226038867386979, + "learning_rate": 6.366624290717548e-05, + "loss": 1.765, + "step": 7044 + }, + { + "epoch": 0.6299177396280401, + "grad_norm": 0.5939868553123321, + "learning_rate": 6.363926352644019e-05, + "loss": 1.7609, + "step": 7045 + }, + { + "epoch": 0.6300071530758226, + "grad_norm": 1.2419287049736292, + "learning_rate": 6.361228719523595e-05, + "loss": 1.5876, + "step": 7046 + }, + { + "epoch": 0.6300965665236051, + "grad_norm": 1.0352580607687025, + "learning_rate": 6.358531391582525e-05, + "loss": 1.6336, + "step": 7047 + }, + { + "epoch": 0.6301859799713877, + "grad_norm": 1.0654264774626245, + "learning_rate": 6.355834369047029e-05, + "loss": 2.1013, + "step": 7048 + }, + { + "epoch": 0.6302753934191703, + "grad_norm": 1.354064849624315, + "learning_rate": 6.353137652143305e-05, + "loss": 1.636, + "step": 7049 + }, + { + "epoch": 0.6303648068669528, + "grad_norm": 0.8340592333312573, + "learning_rate": 6.350441241097518e-05, + "loss": 1.631, + "step": 7050 + }, + { + "epoch": 0.6304542203147353, + "grad_norm": 2.1909908934298046, + "learning_rate": 6.347745136135816e-05, + "loss": 1.8894, + "step": 7051 + }, + { + "epoch": 0.6305436337625179, + "grad_norm": 1.0203607890477546, + "learning_rate": 6.345049337484323e-05, + "loss": 2.2347, + "step": 7052 + }, + { + "epoch": 0.6306330472103004, + "grad_norm": 1.1295565697896681, + "learning_rate": 6.342353845369127e-05, + "loss": 1.8433, + "step": 7053 + }, + { + "epoch": 0.630722460658083, + "grad_norm": 10.118405144712227, + "learning_rate": 6.339658660016295e-05, + "loss": 6.5954, + "step": 7054 + }, + { + "epoch": 0.6308118741058655, + "grad_norm": 2.001638836517026, + "learning_rate": 6.336963781651873e-05, + "loss": 6.4301, + "step": 7055 + }, + { + "epoch": 0.630901287553648, + "grad_norm": 3.2778450532184396, + "learning_rate": 6.334269210501875e-05, + "loss": 1.5582, + "step": 7056 + }, + { + "epoch": 0.6309907010014306, + "grad_norm": 4.078315984128975, + "learning_rate": 6.331574946792288e-05, + "loss": 1.7243, + "step": 7057 + }, + { + "epoch": 0.6310801144492132, + "grad_norm": 1.342219061882757, + "learning_rate": 6.328880990749079e-05, + "loss": 1.5478, + "step": 7058 + }, + { + "epoch": 0.6311695278969958, + "grad_norm": 1.6282160245038142, + "learning_rate": 6.32618734259819e-05, + "loss": 2.3588, + "step": 7059 + }, + { + "epoch": 0.6312589413447782, + "grad_norm": 3.8893762086379873, + "learning_rate": 6.323494002565534e-05, + "loss": 2.0663, + "step": 7060 + }, + { + "epoch": 0.6313483547925608, + "grad_norm": 1.5156296509579803, + "learning_rate": 6.320800970876992e-05, + "loss": 1.6158, + "step": 7061 + }, + { + "epoch": 0.6314377682403434, + "grad_norm": 1.9411783040122716, + "learning_rate": 6.318108247758434e-05, + "loss": 1.6862, + "step": 7062 + }, + { + "epoch": 0.6315271816881259, + "grad_norm": 3.3888215620658455, + "learning_rate": 6.315415833435687e-05, + "loss": 1.9209, + "step": 7063 + }, + { + "epoch": 0.6316165951359084, + "grad_norm": 13.129094154039926, + "learning_rate": 6.312723728134565e-05, + "loss": 6.8221, + "step": 7064 + }, + { + "epoch": 0.631706008583691, + "grad_norm": 4.309493802925427, + "learning_rate": 6.310031932080847e-05, + "loss": 1.6231, + "step": 7065 + }, + { + "epoch": 0.6317954220314735, + "grad_norm": 4.99253967642082, + "learning_rate": 6.3073404455003e-05, + "loss": 1.6921, + "step": 7066 + }, + { + "epoch": 0.6318848354792561, + "grad_norm": 3.891182467126472, + "learning_rate": 6.304649268618649e-05, + "loss": 2.1305, + "step": 7067 + }, + { + "epoch": 0.6319742489270386, + "grad_norm": 2.021782250084454, + "learning_rate": 6.3019584016616e-05, + "loss": 1.864, + "step": 7068 + }, + { + "epoch": 0.6320636623748211, + "grad_norm": 4.573662484725276, + "learning_rate": 6.299267844854835e-05, + "loss": 1.6174, + "step": 7069 + }, + { + "epoch": 0.6321530758226037, + "grad_norm": 3.11234243533563, + "learning_rate": 6.296577598424004e-05, + "loss": 2.1125, + "step": 7070 + }, + { + "epoch": 0.6322424892703863, + "grad_norm": 0.6898810991933599, + "learning_rate": 6.293887662594742e-05, + "loss": 1.8377, + "step": 7071 + }, + { + "epoch": 0.6323319027181689, + "grad_norm": 1.8037614938766242, + "learning_rate": 6.291198037592639e-05, + "loss": 1.4652, + "step": 7072 + }, + { + "epoch": 0.6324213161659513, + "grad_norm": 4.021297615645095, + "learning_rate": 6.288508723643283e-05, + "loss": 2.046, + "step": 7073 + }, + { + "epoch": 0.6325107296137339, + "grad_norm": 1.9327790510870462, + "learning_rate": 6.285819720972214e-05, + "loss": 1.8262, + "step": 7074 + }, + { + "epoch": 0.6326001430615165, + "grad_norm": 1.0446984944783497, + "learning_rate": 6.283131029804963e-05, + "loss": 1.25, + "step": 7075 + }, + { + "epoch": 0.632689556509299, + "grad_norm": 2.0699080117263304, + "learning_rate": 6.280442650367025e-05, + "loss": 1.4884, + "step": 7076 + }, + { + "epoch": 0.6327789699570815, + "grad_norm": 2.1949073907315557, + "learning_rate": 6.277754582883866e-05, + "loss": 1.3503, + "step": 7077 + }, + { + "epoch": 0.6328683834048641, + "grad_norm": 2.6821933209781266, + "learning_rate": 6.275066827580933e-05, + "loss": 2.0023, + "step": 7078 + }, + { + "epoch": 0.6329577968526466, + "grad_norm": 1.4700058941578364, + "learning_rate": 6.272379384683651e-05, + "loss": 1.7271, + "step": 7079 + }, + { + "epoch": 0.6330472103004292, + "grad_norm": 3.206348392523273, + "learning_rate": 6.269692254417408e-05, + "loss": 2.0536, + "step": 7080 + }, + { + "epoch": 0.6331366237482118, + "grad_norm": 2.0557999359622205, + "learning_rate": 6.267005437007567e-05, + "loss": 1.6688, + "step": 7081 + }, + { + "epoch": 0.6332260371959942, + "grad_norm": 0.51321650407555, + "learning_rate": 6.264318932679476e-05, + "loss": 1.6735, + "step": 7082 + }, + { + "epoch": 0.6333154506437768, + "grad_norm": 2.259459378851087, + "learning_rate": 6.261632741658443e-05, + "loss": 1.8966, + "step": 7083 + }, + { + "epoch": 0.6334048640915594, + "grad_norm": 1.0244202064870007, + "learning_rate": 6.258946864169757e-05, + "loss": 1.3118, + "step": 7084 + }, + { + "epoch": 0.633494277539342, + "grad_norm": 1.356178767720117, + "learning_rate": 6.256261300438676e-05, + "loss": 1.3957, + "step": 7085 + }, + { + "epoch": 0.6335836909871244, + "grad_norm": 1.0453318157633869, + "learning_rate": 6.253576050690442e-05, + "loss": 1.6112, + "step": 7086 + }, + { + "epoch": 0.633673104434907, + "grad_norm": 0.8002655966821579, + "learning_rate": 6.250891115150261e-05, + "loss": 1.634, + "step": 7087 + }, + { + "epoch": 0.6337625178826896, + "grad_norm": 1.7441050594979564, + "learning_rate": 6.248206494043313e-05, + "loss": 2.1627, + "step": 7088 + }, + { + "epoch": 0.6338519313304721, + "grad_norm": 0.8049676435306738, + "learning_rate": 6.245522187594757e-05, + "loss": 1.5979, + "step": 7089 + }, + { + "epoch": 0.6339413447782547, + "grad_norm": 0.4453384548140215, + "learning_rate": 6.242838196029719e-05, + "loss": 1.3837, + "step": 7090 + }, + { + "epoch": 0.6340307582260372, + "grad_norm": 1.6982659277113072, + "learning_rate": 6.240154519573304e-05, + "loss": 1.9154, + "step": 7091 + }, + { + "epoch": 0.6341201716738197, + "grad_norm": 1.0560996611678193, + "learning_rate": 6.237471158450585e-05, + "loss": 2.0521, + "step": 7092 + }, + { + "epoch": 0.6342095851216023, + "grad_norm": 0.5136774214540389, + "learning_rate": 6.234788112886623e-05, + "loss": 1.3656, + "step": 7093 + }, + { + "epoch": 0.6342989985693849, + "grad_norm": 0.9208002733592682, + "learning_rate": 6.232105383106432e-05, + "loss": 1.8659, + "step": 7094 + }, + { + "epoch": 0.6343884120171673, + "grad_norm": 1.7663168921467334, + "learning_rate": 6.22942296933501e-05, + "loss": 1.9637, + "step": 7095 + }, + { + "epoch": 0.6344778254649499, + "grad_norm": 0.8240278756112037, + "learning_rate": 6.226740871797334e-05, + "loss": 1.8369, + "step": 7096 + }, + { + "epoch": 0.6345672389127325, + "grad_norm": 0.9144177816680512, + "learning_rate": 6.224059090718341e-05, + "loss": 1.9168, + "step": 7097 + }, + { + "epoch": 0.634656652360515, + "grad_norm": 2.4194864919820835, + "learning_rate": 6.221377626322953e-05, + "loss": 1.7265, + "step": 7098 + }, + { + "epoch": 0.6347460658082976, + "grad_norm": 0.8004723683049134, + "learning_rate": 6.218696478836058e-05, + "loss": 1.7042, + "step": 7099 + }, + { + "epoch": 0.6348354792560801, + "grad_norm": 1.3456626305208883, + "learning_rate": 6.216015648482525e-05, + "loss": 1.7592, + "step": 7100 + }, + { + "epoch": 0.6349248927038627, + "grad_norm": 0.90713989381325, + "learning_rate": 6.21333513548719e-05, + "loss": 1.3461, + "step": 7101 + }, + { + "epoch": 0.6350143061516452, + "grad_norm": 0.9674834614845725, + "learning_rate": 6.210654940074861e-05, + "loss": 1.6585, + "step": 7102 + }, + { + "epoch": 0.6351037195994278, + "grad_norm": 0.5238568412433485, + "learning_rate": 6.20797506247033e-05, + "loss": 1.397, + "step": 7103 + }, + { + "epoch": 0.6351931330472103, + "grad_norm": 1.0891110283188585, + "learning_rate": 6.205295502898348e-05, + "loss": 2.1828, + "step": 7104 + }, + { + "epoch": 0.6352825464949928, + "grad_norm": 0.8779943642181782, + "learning_rate": 6.202616261583652e-05, + "loss": 1.8545, + "step": 7105 + }, + { + "epoch": 0.6353719599427754, + "grad_norm": 0.6024468635475211, + "learning_rate": 6.199937338750939e-05, + "loss": 1.6959, + "step": 7106 + }, + { + "epoch": 0.635461373390558, + "grad_norm": 1.0083788299192544, + "learning_rate": 6.197258734624896e-05, + "loss": 1.7238, + "step": 7107 + }, + { + "epoch": 0.6355507868383404, + "grad_norm": 1.2542735997039873, + "learning_rate": 6.194580449430168e-05, + "loss": 1.6049, + "step": 7108 + }, + { + "epoch": 0.635640200286123, + "grad_norm": 0.5845401094543574, + "learning_rate": 6.191902483391386e-05, + "loss": 1.9133, + "step": 7109 + }, + { + "epoch": 0.6357296137339056, + "grad_norm": 7.055907292794755, + "learning_rate": 6.18922483673314e-05, + "loss": 6.6375, + "step": 7110 + }, + { + "epoch": 0.6358190271816881, + "grad_norm": 2.440146527028173, + "learning_rate": 6.186547509680007e-05, + "loss": 1.5138, + "step": 7111 + }, + { + "epoch": 0.6359084406294707, + "grad_norm": 3.3309710834925443, + "learning_rate": 6.183870502456529e-05, + "loss": 2.005, + "step": 7112 + }, + { + "epoch": 0.6359978540772532, + "grad_norm": 0.7066915927038934, + "learning_rate": 6.181193815287218e-05, + "loss": 1.624, + "step": 7113 + }, + { + "epoch": 0.6360872675250357, + "grad_norm": 1.3600726531655407, + "learning_rate": 6.178517448396575e-05, + "loss": 2.0019, + "step": 7114 + }, + { + "epoch": 0.6361766809728183, + "grad_norm": 1.8235630227541555, + "learning_rate": 6.175841402009058e-05, + "loss": 1.3979, + "step": 7115 + }, + { + "epoch": 0.6362660944206009, + "grad_norm": 2.515381422458678, + "learning_rate": 6.173165676349103e-05, + "loss": 1.8244, + "step": 7116 + }, + { + "epoch": 0.6363555078683834, + "grad_norm": 0.8408293892183746, + "learning_rate": 6.170490271641123e-05, + "loss": 1.6299, + "step": 7117 + }, + { + "epoch": 0.6364449213161659, + "grad_norm": 1.7905749181204087, + "learning_rate": 6.167815188109496e-05, + "loss": 1.5618, + "step": 7118 + }, + { + "epoch": 0.6365343347639485, + "grad_norm": 1.2861734286481838, + "learning_rate": 6.165140425978584e-05, + "loss": 1.7534, + "step": 7119 + }, + { + "epoch": 0.6366237482117311, + "grad_norm": 1.1508215612797832, + "learning_rate": 6.16246598547271e-05, + "loss": 1.7628, + "step": 7120 + }, + { + "epoch": 0.6367131616595136, + "grad_norm": 0.7221986148783472, + "learning_rate": 6.159791866816182e-05, + "loss": 1.6236, + "step": 7121 + }, + { + "epoch": 0.6368025751072961, + "grad_norm": 1.3167210876544697, + "learning_rate": 6.157118070233269e-05, + "loss": 2.172, + "step": 7122 + }, + { + "epoch": 0.6368919885550787, + "grad_norm": 1.347204530047025, + "learning_rate": 6.154444595948227e-05, + "loss": 1.9351, + "step": 7123 + }, + { + "epoch": 0.6369814020028612, + "grad_norm": 0.8760710442906268, + "learning_rate": 6.15177144418527e-05, + "loss": 1.6689, + "step": 7124 + }, + { + "epoch": 0.6370708154506438, + "grad_norm": 1.029276900115003, + "learning_rate": 6.149098615168594e-05, + "loss": 1.7651, + "step": 7125 + }, + { + "epoch": 0.6371602288984263, + "grad_norm": 1.130601748092791, + "learning_rate": 6.14642610912237e-05, + "loss": 1.8587, + "step": 7126 + }, + { + "epoch": 0.6372496423462088, + "grad_norm": 0.6465961313259722, + "learning_rate": 6.143753926270727e-05, + "loss": 1.7544, + "step": 7127 + }, + { + "epoch": 0.6373390557939914, + "grad_norm": 1.0148120686866926, + "learning_rate": 6.141082066837791e-05, + "loss": 1.8773, + "step": 7128 + }, + { + "epoch": 0.637428469241774, + "grad_norm": 1.0801868137467996, + "learning_rate": 6.13841053104764e-05, + "loss": 2.0809, + "step": 7129 + }, + { + "epoch": 0.6375178826895566, + "grad_norm": 0.616801637815129, + "learning_rate": 6.135739319124335e-05, + "loss": 1.7189, + "step": 7130 + }, + { + "epoch": 0.637607296137339, + "grad_norm": 0.8474469616079118, + "learning_rate": 6.133068431291904e-05, + "loss": 1.8365, + "step": 7131 + }, + { + "epoch": 0.6376967095851216, + "grad_norm": 1.174038404809612, + "learning_rate": 6.130397867774357e-05, + "loss": 2.1505, + "step": 7132 + }, + { + "epoch": 0.6377861230329042, + "grad_norm": 0.9586607372032853, + "learning_rate": 6.127727628795668e-05, + "loss": 1.8922, + "step": 7133 + }, + { + "epoch": 0.6378755364806867, + "grad_norm": 0.9445450614628572, + "learning_rate": 6.12505771457978e-05, + "loss": 1.8205, + "step": 7134 + }, + { + "epoch": 0.6379649499284692, + "grad_norm": 0.687739682044517, + "learning_rate": 6.122388125350625e-05, + "loss": 1.6295, + "step": 7135 + }, + { + "epoch": 0.6380543633762518, + "grad_norm": 0.8429113951571161, + "learning_rate": 6.119718861332098e-05, + "loss": 1.9277, + "step": 7136 + }, + { + "epoch": 0.6381437768240343, + "grad_norm": 1.2278731199406383, + "learning_rate": 6.117049922748063e-05, + "loss": 1.497, + "step": 7137 + }, + { + "epoch": 0.6382331902718169, + "grad_norm": 1.1169699415036232, + "learning_rate": 6.114381309822359e-05, + "loss": 1.7218, + "step": 7138 + }, + { + "epoch": 0.6383226037195995, + "grad_norm": 0.7743579448943511, + "learning_rate": 6.111713022778804e-05, + "loss": 1.5425, + "step": 7139 + }, + { + "epoch": 0.6384120171673819, + "grad_norm": 1.3994326736076845, + "learning_rate": 6.109045061841183e-05, + "loss": 1.9595, + "step": 7140 + }, + { + "epoch": 0.6385014306151645, + "grad_norm": 1.0588690421227707, + "learning_rate": 6.106377427233247e-05, + "loss": 2.162, + "step": 7141 + }, + { + "epoch": 0.6385908440629471, + "grad_norm": 0.46184722563738495, + "learning_rate": 6.103710119178738e-05, + "loss": 1.4545, + "step": 7142 + }, + { + "epoch": 0.6386802575107297, + "grad_norm": 0.9371743520449225, + "learning_rate": 6.1010431379013585e-05, + "loss": 1.6991, + "step": 7143 + }, + { + "epoch": 0.6387696709585121, + "grad_norm": 0.6050770154495873, + "learning_rate": 6.098376483624781e-05, + "loss": 1.641, + "step": 7144 + }, + { + "epoch": 0.6388590844062947, + "grad_norm": 0.8691323900699176, + "learning_rate": 6.095710156572654e-05, + "loss": 1.5292, + "step": 7145 + }, + { + "epoch": 0.6389484978540773, + "grad_norm": 0.7307206413601711, + "learning_rate": 6.0930441569686036e-05, + "loss": 1.6591, + "step": 7146 + }, + { + "epoch": 0.6390379113018598, + "grad_norm": 0.64275674290883, + "learning_rate": 6.090378485036221e-05, + "loss": 1.6706, + "step": 7147 + }, + { + "epoch": 0.6391273247496424, + "grad_norm": 0.8601723806611827, + "learning_rate": 6.0877131409990684e-05, + "loss": 1.6693, + "step": 7148 + }, + { + "epoch": 0.6392167381974249, + "grad_norm": 0.5814795876727318, + "learning_rate": 6.085048125080692e-05, + "loss": 1.7947, + "step": 7149 + }, + { + "epoch": 0.6393061516452074, + "grad_norm": 0.9840512697457562, + "learning_rate": 6.082383437504604e-05, + "loss": 1.6455, + "step": 7150 + }, + { + "epoch": 0.63939556509299, + "grad_norm": 4.702406629814146, + "learning_rate": 6.079719078494286e-05, + "loss": 6.5938, + "step": 7151 + }, + { + "epoch": 0.6394849785407726, + "grad_norm": 0.9352007299049896, + "learning_rate": 6.0770550482731924e-05, + "loss": 1.486, + "step": 7152 + }, + { + "epoch": 0.639574391988555, + "grad_norm": 1.6996333104687187, + "learning_rate": 6.0743913470647564e-05, + "loss": 1.7313, + "step": 7153 + }, + { + "epoch": 0.6396638054363376, + "grad_norm": 2.1363632107435255, + "learning_rate": 6.071727975092376e-05, + "loss": 1.6251, + "step": 7154 + }, + { + "epoch": 0.6397532188841202, + "grad_norm": 0.7395086375157326, + "learning_rate": 6.069064932579423e-05, + "loss": 1.5935, + "step": 7155 + }, + { + "epoch": 0.6398426323319027, + "grad_norm": 1.5717659294885544, + "learning_rate": 6.0664022197492475e-05, + "loss": 1.8138, + "step": 7156 + }, + { + "epoch": 0.6399320457796852, + "grad_norm": 1.2338048208234789, + "learning_rate": 6.0637398368251705e-05, + "loss": 1.5083, + "step": 7157 + }, + { + "epoch": 0.6400214592274678, + "grad_norm": 3.0849399358186655, + "learning_rate": 6.06107778403048e-05, + "loss": 1.5347, + "step": 7158 + }, + { + "epoch": 0.6401108726752504, + "grad_norm": 1.1762692535556565, + "learning_rate": 6.058416061588434e-05, + "loss": 1.9297, + "step": 7159 + }, + { + "epoch": 0.6402002861230329, + "grad_norm": 0.9610376727669074, + "learning_rate": 6.055754669722278e-05, + "loss": 1.5558, + "step": 7160 + }, + { + "epoch": 0.6402896995708155, + "grad_norm": 1.4473545207352605, + "learning_rate": 6.0530936086552095e-05, + "loss": 1.7383, + "step": 7161 + }, + { + "epoch": 0.640379113018598, + "grad_norm": 2.0603696133669804, + "learning_rate": 6.050432878610417e-05, + "loss": 1.7732, + "step": 7162 + }, + { + "epoch": 0.6404685264663805, + "grad_norm": 2.518004301388161, + "learning_rate": 6.047772479811047e-05, + "loss": 2.3714, + "step": 7163 + }, + { + "epoch": 0.6405579399141631, + "grad_norm": 0.3753104109482285, + "learning_rate": 6.0451124124802275e-05, + "loss": 1.688, + "step": 7164 + }, + { + "epoch": 0.6406473533619457, + "grad_norm": 1.592489707001158, + "learning_rate": 6.042452676841053e-05, + "loss": 1.5469, + "step": 7165 + }, + { + "epoch": 0.6407367668097281, + "grad_norm": 2.8955603326453865, + "learning_rate": 6.039793273116594e-05, + "loss": 1.6548, + "step": 7166 + }, + { + "epoch": 0.6408261802575107, + "grad_norm": 0.6076652684219572, + "learning_rate": 6.03713420152989e-05, + "loss": 1.4388, + "step": 7167 + }, + { + "epoch": 0.6409155937052933, + "grad_norm": 0.4485419273666167, + "learning_rate": 6.034475462303952e-05, + "loss": 1.6787, + "step": 7168 + }, + { + "epoch": 0.6410050071530758, + "grad_norm": 0.6549241614449619, + "learning_rate": 6.031817055661769e-05, + "loss": 1.6019, + "step": 7169 + }, + { + "epoch": 0.6410944206008584, + "grad_norm": 1.2336023565750378, + "learning_rate": 6.029158981826299e-05, + "loss": 1.7814, + "step": 7170 + }, + { + "epoch": 0.6411838340486409, + "grad_norm": 1.3712098869208391, + "learning_rate": 6.02650124102047e-05, + "loss": 1.6486, + "step": 7171 + }, + { + "epoch": 0.6412732474964234, + "grad_norm": 0.7598068682157859, + "learning_rate": 6.023843833467182e-05, + "loss": 1.9924, + "step": 7172 + }, + { + "epoch": 0.641362660944206, + "grad_norm": 0.8547044135398768, + "learning_rate": 6.02118675938931e-05, + "loss": 1.8308, + "step": 7173 + }, + { + "epoch": 0.6414520743919886, + "grad_norm": 1.5048664393434057, + "learning_rate": 6.0185300190097004e-05, + "loss": 1.846, + "step": 7174 + }, + { + "epoch": 0.641541487839771, + "grad_norm": 1.233005301247911, + "learning_rate": 6.0158736125511664e-05, + "loss": 1.9205, + "step": 7175 + }, + { + "epoch": 0.6416309012875536, + "grad_norm": 0.6271785475391026, + "learning_rate": 6.013217540236502e-05, + "loss": 1.7727, + "step": 7176 + }, + { + "epoch": 0.6417203147353362, + "grad_norm": 0.9515036630130581, + "learning_rate": 6.0105618022884694e-05, + "loss": 1.8614, + "step": 7177 + }, + { + "epoch": 0.6418097281831188, + "grad_norm": 0.627814854923539, + "learning_rate": 6.0079063989298e-05, + "loss": 1.5739, + "step": 7178 + }, + { + "epoch": 0.6418991416309013, + "grad_norm": 1.008099846290149, + "learning_rate": 6.005251330383199e-05, + "loss": 1.8346, + "step": 7179 + }, + { + "epoch": 0.6419885550786838, + "grad_norm": 0.6657663307646858, + "learning_rate": 6.002596596871346e-05, + "loss": 1.78, + "step": 7180 + }, + { + "epoch": 0.6420779685264664, + "grad_norm": 0.5692365693230448, + "learning_rate": 5.999942198616888e-05, + "loss": 1.6442, + "step": 7181 + }, + { + "epoch": 0.6421673819742489, + "grad_norm": 0.8344231096849963, + "learning_rate": 5.9972881358424436e-05, + "loss": 1.6645, + "step": 7182 + }, + { + "epoch": 0.6422567954220315, + "grad_norm": 0.9117442247466014, + "learning_rate": 5.994634408770612e-05, + "loss": 2.0635, + "step": 7183 + }, + { + "epoch": 0.642346208869814, + "grad_norm": 0.4985255120525201, + "learning_rate": 5.991981017623955e-05, + "loss": 1.733, + "step": 7184 + }, + { + "epoch": 0.6424356223175965, + "grad_norm": 1.6114749543096825, + "learning_rate": 5.9893279626250124e-05, + "loss": 1.9658, + "step": 7185 + }, + { + "epoch": 0.6425250357653791, + "grad_norm": 0.4608646564625952, + "learning_rate": 5.986675243996286e-05, + "loss": 1.7106, + "step": 7186 + }, + { + "epoch": 0.6426144492131617, + "grad_norm": 0.600482130873652, + "learning_rate": 5.9840228619602636e-05, + "loss": 1.5527, + "step": 7187 + }, + { + "epoch": 0.6427038626609443, + "grad_norm": 1.6223965907333198, + "learning_rate": 5.981370816739389e-05, + "loss": 2.1747, + "step": 7188 + }, + { + "epoch": 0.6427932761087267, + "grad_norm": 0.8229443174394104, + "learning_rate": 5.978719108556094e-05, + "loss": 1.5248, + "step": 7189 + }, + { + "epoch": 0.6428826895565093, + "grad_norm": 1.1187439517825029, + "learning_rate": 5.976067737632769e-05, + "loss": 1.4626, + "step": 7190 + }, + { + "epoch": 0.6429721030042919, + "grad_norm": 1.0863080336984243, + "learning_rate": 5.9734167041917856e-05, + "loss": 1.6124, + "step": 7191 + }, + { + "epoch": 0.6430615164520744, + "grad_norm": 0.7926253709458997, + "learning_rate": 5.9707660084554774e-05, + "loss": 1.5218, + "step": 7192 + }, + { + "epoch": 0.6431509298998569, + "grad_norm": 0.4781540533822633, + "learning_rate": 5.968115650646161e-05, + "loss": 1.5155, + "step": 7193 + }, + { + "epoch": 0.6432403433476395, + "grad_norm": 1.4583684026951502, + "learning_rate": 5.9654656309861155e-05, + "loss": 1.688, + "step": 7194 + }, + { + "epoch": 0.643329756795422, + "grad_norm": 1.017288845195456, + "learning_rate": 5.9628159496975935e-05, + "loss": 1.6697, + "step": 7195 + }, + { + "epoch": 0.6434191702432046, + "grad_norm": 1.0042585300915172, + "learning_rate": 5.9601666070028194e-05, + "loss": 1.9474, + "step": 7196 + }, + { + "epoch": 0.6435085836909872, + "grad_norm": 1.0359719009616455, + "learning_rate": 5.9575176031239964e-05, + "loss": 1.7514, + "step": 7197 + }, + { + "epoch": 0.6435979971387696, + "grad_norm": 0.8763450416475485, + "learning_rate": 5.954868938283291e-05, + "loss": 1.779, + "step": 7198 + }, + { + "epoch": 0.6436874105865522, + "grad_norm": 0.6219579462320839, + "learning_rate": 5.9522206127028414e-05, + "loss": 1.6534, + "step": 7199 + }, + { + "epoch": 0.6437768240343348, + "grad_norm": 0.5358950196362295, + "learning_rate": 5.9495726266047605e-05, + "loss": 1.4975, + "step": 7200 + }, + { + "epoch": 0.6438662374821174, + "grad_norm": 0.6466690024483406, + "learning_rate": 5.9469249802111324e-05, + "loss": 1.45, + "step": 7201 + }, + { + "epoch": 0.6439556509298998, + "grad_norm": 1.3527736760952949, + "learning_rate": 5.94427767374401e-05, + "loss": 1.8453, + "step": 7202 + }, + { + "epoch": 0.6440450643776824, + "grad_norm": 1.0867310139083124, + "learning_rate": 5.941630707425418e-05, + "loss": 2.3196, + "step": 7203 + }, + { + "epoch": 0.644134477825465, + "grad_norm": 0.6111653174391125, + "learning_rate": 5.938984081477363e-05, + "loss": 1.7015, + "step": 7204 + }, + { + "epoch": 0.6442238912732475, + "grad_norm": 0.6001338498529587, + "learning_rate": 5.936337796121807e-05, + "loss": 2.0151, + "step": 7205 + }, + { + "epoch": 0.64431330472103, + "grad_norm": 0.8561662582795484, + "learning_rate": 5.9336918515806914e-05, + "loss": 1.6037, + "step": 7206 + }, + { + "epoch": 0.6444027181688126, + "grad_norm": 1.849369244850278, + "learning_rate": 5.931046248075931e-05, + "loss": 2.1061, + "step": 7207 + }, + { + "epoch": 0.6444921316165951, + "grad_norm": 0.9792761006375045, + "learning_rate": 5.9284009858294076e-05, + "loss": 1.5122, + "step": 7208 + }, + { + "epoch": 0.6445815450643777, + "grad_norm": 0.4707810625486213, + "learning_rate": 5.925756065062975e-05, + "loss": 1.6466, + "step": 7209 + }, + { + "epoch": 0.6446709585121603, + "grad_norm": 1.2991935580601652, + "learning_rate": 5.9231114859984584e-05, + "loss": 2.3526, + "step": 7210 + }, + { + "epoch": 0.6447603719599427, + "grad_norm": 1.0805372960993218, + "learning_rate": 5.920467248857661e-05, + "loss": 1.4722, + "step": 7211 + }, + { + "epoch": 0.6448497854077253, + "grad_norm": 0.7059724546158623, + "learning_rate": 5.9178233538623486e-05, + "loss": 1.8771, + "step": 7212 + }, + { + "epoch": 0.6449391988555079, + "grad_norm": 1.3435750278478085, + "learning_rate": 5.9151798012342605e-05, + "loss": 1.7806, + "step": 7213 + }, + { + "epoch": 0.6450286123032904, + "grad_norm": 0.6438334731130547, + "learning_rate": 5.91253659119511e-05, + "loss": 1.668, + "step": 7214 + }, + { + "epoch": 0.6451180257510729, + "grad_norm": 0.9616718861469259, + "learning_rate": 5.9098937239665796e-05, + "loss": 1.7493, + "step": 7215 + }, + { + "epoch": 0.6452074391988555, + "grad_norm": 0.6254327603963616, + "learning_rate": 5.9072511997703226e-05, + "loss": 1.7633, + "step": 7216 + }, + { + "epoch": 0.645296852646638, + "grad_norm": 0.7908776359382075, + "learning_rate": 5.904609018827961e-05, + "loss": 1.6031, + "step": 7217 + }, + { + "epoch": 0.6453862660944206, + "grad_norm": 0.7849040268044011, + "learning_rate": 5.9019671813610986e-05, + "loss": 1.5871, + "step": 7218 + }, + { + "epoch": 0.6454756795422032, + "grad_norm": 0.5634886074184908, + "learning_rate": 5.899325687591302e-05, + "loss": 1.625, + "step": 7219 + }, + { + "epoch": 0.6455650929899857, + "grad_norm": 0.653929295269001, + "learning_rate": 5.896684537740103e-05, + "loss": 1.5254, + "step": 7220 + }, + { + "epoch": 0.6456545064377682, + "grad_norm": 0.9394577116681647, + "learning_rate": 5.89404373202902e-05, + "loss": 1.8735, + "step": 7221 + }, + { + "epoch": 0.6457439198855508, + "grad_norm": 0.7091895448392102, + "learning_rate": 5.891403270679527e-05, + "loss": 2.0336, + "step": 7222 + }, + { + "epoch": 0.6458333333333334, + "grad_norm": 1.2170217946849418, + "learning_rate": 5.8887631539130826e-05, + "loss": 1.8815, + "step": 7223 + }, + { + "epoch": 0.6459227467811158, + "grad_norm": 0.6455688980369243, + "learning_rate": 5.886123381951103e-05, + "loss": 1.7062, + "step": 7224 + }, + { + "epoch": 0.6460121602288984, + "grad_norm": 0.8887052441885465, + "learning_rate": 5.883483955014992e-05, + "loss": 2.0413, + "step": 7225 + }, + { + "epoch": 0.646101573676681, + "grad_norm": 0.7804824082036448, + "learning_rate": 5.8808448733261076e-05, + "loss": 1.653, + "step": 7226 + }, + { + "epoch": 0.6461909871244635, + "grad_norm": 0.8109279043554136, + "learning_rate": 5.878206137105791e-05, + "loss": 1.8069, + "step": 7227 + }, + { + "epoch": 0.6462804005722461, + "grad_norm": 0.5404584392062705, + "learning_rate": 5.875567746575348e-05, + "loss": 1.636, + "step": 7228 + }, + { + "epoch": 0.6463698140200286, + "grad_norm": 0.5405283557150092, + "learning_rate": 5.872929701956054e-05, + "loss": 1.7039, + "step": 7229 + }, + { + "epoch": 0.6464592274678111, + "grad_norm": 1.1583524694338727, + "learning_rate": 5.870292003469164e-05, + "loss": 1.7922, + "step": 7230 + }, + { + "epoch": 0.6465486409155937, + "grad_norm": 0.47629468074411124, + "learning_rate": 5.867654651335893e-05, + "loss": 1.7404, + "step": 7231 + }, + { + "epoch": 0.6466380543633763, + "grad_norm": 0.7662847292166188, + "learning_rate": 5.86501764577744e-05, + "loss": 1.6308, + "step": 7232 + }, + { + "epoch": 0.6467274678111588, + "grad_norm": 1.180665928067402, + "learning_rate": 5.862380987014959e-05, + "loss": 1.9097, + "step": 7233 + }, + { + "epoch": 0.6468168812589413, + "grad_norm": 0.6404262752755193, + "learning_rate": 5.8597446752695915e-05, + "loss": 1.696, + "step": 7234 + }, + { + "epoch": 0.6469062947067239, + "grad_norm": 0.9225195641107715, + "learning_rate": 5.857108710762439e-05, + "loss": 1.7765, + "step": 7235 + }, + { + "epoch": 0.6469957081545065, + "grad_norm": 1.0493180537954396, + "learning_rate": 5.854473093714572e-05, + "loss": 1.8778, + "step": 7236 + }, + { + "epoch": 0.647085121602289, + "grad_norm": 0.912256604378396, + "learning_rate": 5.851837824347042e-05, + "loss": 1.5259, + "step": 7237 + }, + { + "epoch": 0.6471745350500715, + "grad_norm": 0.5762174240684242, + "learning_rate": 5.8492029028808615e-05, + "loss": 2.1298, + "step": 7238 + }, + { + "epoch": 0.6472639484978541, + "grad_norm": 0.7776138066768693, + "learning_rate": 5.846568329537023e-05, + "loss": 1.9553, + "step": 7239 + }, + { + "epoch": 0.6473533619456366, + "grad_norm": 0.8424971858946336, + "learning_rate": 5.8439341045364815e-05, + "loss": 1.7382, + "step": 7240 + }, + { + "epoch": 0.6474427753934192, + "grad_norm": 1.244997101417072, + "learning_rate": 5.8413002281001686e-05, + "loss": 1.7259, + "step": 7241 + }, + { + "epoch": 0.6475321888412017, + "grad_norm": 0.8566181373826273, + "learning_rate": 5.8386667004489835e-05, + "loss": 2.1587, + "step": 7242 + }, + { + "epoch": 0.6476216022889842, + "grad_norm": 0.6807395113637748, + "learning_rate": 5.836033521803796e-05, + "loss": 1.862, + "step": 7243 + }, + { + "epoch": 0.6477110157367668, + "grad_norm": 0.559408453256638, + "learning_rate": 5.833400692385444e-05, + "loss": 1.8286, + "step": 7244 + }, + { + "epoch": 0.6478004291845494, + "grad_norm": 0.7686306533801435, + "learning_rate": 5.8307682124147466e-05, + "loss": 1.3827, + "step": 7245 + }, + { + "epoch": 0.647889842632332, + "grad_norm": 1.1325723179117981, + "learning_rate": 5.8281360821124884e-05, + "loss": 1.7864, + "step": 7246 + }, + { + "epoch": 0.6479792560801144, + "grad_norm": 0.8094184309392946, + "learning_rate": 5.8255043016994145e-05, + "loss": 1.7738, + "step": 7247 + }, + { + "epoch": 0.648068669527897, + "grad_norm": 0.7600426990588561, + "learning_rate": 5.8228728713962543e-05, + "loss": 1.7569, + "step": 7248 + }, + { + "epoch": 0.6481580829756796, + "grad_norm": 2.4708474910514098, + "learning_rate": 5.820241791423704e-05, + "loss": 1.9861, + "step": 7249 + }, + { + "epoch": 0.6482474964234621, + "grad_norm": 1.3712232156433948, + "learning_rate": 5.8176110620024236e-05, + "loss": 1.4969, + "step": 7250 + }, + { + "epoch": 0.6483369098712446, + "grad_norm": 1.703883764617488, + "learning_rate": 5.814980683353053e-05, + "loss": 1.5131, + "step": 7251 + }, + { + "epoch": 0.6484263233190272, + "grad_norm": 1.1528955814185606, + "learning_rate": 5.812350655696197e-05, + "loss": 1.5939, + "step": 7252 + }, + { + "epoch": 0.6485157367668097, + "grad_norm": 0.9390170550100451, + "learning_rate": 5.809720979252435e-05, + "loss": 2.3714, + "step": 7253 + }, + { + "epoch": 0.6486051502145923, + "grad_norm": 1.6941638344873478, + "learning_rate": 5.807091654242318e-05, + "loss": 2.033, + "step": 7254 + }, + { + "epoch": 0.6486945636623748, + "grad_norm": 3.138139436963737, + "learning_rate": 5.8044626808863557e-05, + "loss": 6.5579, + "step": 7255 + }, + { + "epoch": 0.6487839771101573, + "grad_norm": 1.4544790041063438, + "learning_rate": 5.801834059405041e-05, + "loss": 1.7235, + "step": 7256 + }, + { + "epoch": 0.6488733905579399, + "grad_norm": 0.5591825529248924, + "learning_rate": 5.799205790018838e-05, + "loss": 1.7023, + "step": 7257 + }, + { + "epoch": 0.6489628040057225, + "grad_norm": 1.161526844536513, + "learning_rate": 5.796577872948165e-05, + "loss": 1.8888, + "step": 7258 + }, + { + "epoch": 0.649052217453505, + "grad_norm": 1.9601565998141885, + "learning_rate": 5.793950308413432e-05, + "loss": 1.7188, + "step": 7259 + }, + { + "epoch": 0.6491416309012875, + "grad_norm": 1.3839300451996583, + "learning_rate": 5.7913230966350116e-05, + "loss": 1.8285, + "step": 7260 + }, + { + "epoch": 0.6492310443490701, + "grad_norm": 0.642956601593466, + "learning_rate": 5.788696237833237e-05, + "loss": 1.6603, + "step": 7261 + }, + { + "epoch": 0.6493204577968527, + "grad_norm": 1.7299934630933804, + "learning_rate": 5.786069732228423e-05, + "loss": 1.786, + "step": 7262 + }, + { + "epoch": 0.6494098712446352, + "grad_norm": 1.7992932237241401, + "learning_rate": 5.783443580040854e-05, + "loss": 1.898, + "step": 7263 + }, + { + "epoch": 0.6494992846924177, + "grad_norm": 1.1383377430015955, + "learning_rate": 5.780817781490777e-05, + "loss": 2.1442, + "step": 7264 + }, + { + "epoch": 0.6495886981402003, + "grad_norm": 2.7242765753116265, + "learning_rate": 5.778192336798416e-05, + "loss": 2.2834, + "step": 7265 + }, + { + "epoch": 0.6496781115879828, + "grad_norm": 0.7092785856131725, + "learning_rate": 5.775567246183966e-05, + "loss": 1.7543, + "step": 7266 + }, + { + "epoch": 0.6497675250357654, + "grad_norm": 1.114512522291279, + "learning_rate": 5.772942509867588e-05, + "loss": 1.6201, + "step": 7267 + }, + { + "epoch": 0.649856938483548, + "grad_norm": 1.5935653941215822, + "learning_rate": 5.7703181280694184e-05, + "loss": 1.8027, + "step": 7268 + }, + { + "epoch": 0.6499463519313304, + "grad_norm": 0.7296896854012844, + "learning_rate": 5.767694101009562e-05, + "loss": 1.7751, + "step": 7269 + }, + { + "epoch": 0.650035765379113, + "grad_norm": 1.2222267754845106, + "learning_rate": 5.765070428908086e-05, + "loss": 1.7859, + "step": 7270 + }, + { + "epoch": 0.6501251788268956, + "grad_norm": 0.7801421379271709, + "learning_rate": 5.762447111985039e-05, + "loss": 1.3914, + "step": 7271 + }, + { + "epoch": 0.6502145922746781, + "grad_norm": 1.124682339078583, + "learning_rate": 5.759824150460435e-05, + "loss": 1.6408, + "step": 7272 + }, + { + "epoch": 0.6503040057224606, + "grad_norm": 0.43418369696472564, + "learning_rate": 5.7572015445542594e-05, + "loss": 1.6559, + "step": 7273 + }, + { + "epoch": 0.6503934191702432, + "grad_norm": 0.801385833839956, + "learning_rate": 5.7545792944864696e-05, + "loss": 1.6718, + "step": 7274 + }, + { + "epoch": 0.6504828326180258, + "grad_norm": 0.9949275338944862, + "learning_rate": 5.751957400476984e-05, + "loss": 1.558, + "step": 7275 + }, + { + "epoch": 0.6505722460658083, + "grad_norm": 0.5720006381051559, + "learning_rate": 5.7493358627456995e-05, + "loss": 1.6169, + "step": 7276 + }, + { + "epoch": 0.6506616595135909, + "grad_norm": 4.398654132952512, + "learning_rate": 5.7467146815124874e-05, + "loss": 6.4474, + "step": 7277 + }, + { + "epoch": 0.6507510729613734, + "grad_norm": 1.3952085787863224, + "learning_rate": 5.744093856997175e-05, + "loss": 1.7103, + "step": 7278 + }, + { + "epoch": 0.6508404864091559, + "grad_norm": 1.0047318699517993, + "learning_rate": 5.741473389419565e-05, + "loss": 1.6491, + "step": 7279 + }, + { + "epoch": 0.6509298998569385, + "grad_norm": 1.0294316971181658, + "learning_rate": 5.7388532789994476e-05, + "loss": 2.049, + "step": 7280 + }, + { + "epoch": 0.6510193133047211, + "grad_norm": 0.9217606554217636, + "learning_rate": 5.7362335259565556e-05, + "loss": 1.8393, + "step": 7281 + }, + { + "epoch": 0.6511087267525035, + "grad_norm": 0.7532706094522316, + "learning_rate": 5.733614130510609e-05, + "loss": 1.7045, + "step": 7282 + }, + { + "epoch": 0.6511981402002861, + "grad_norm": 1.0894082098719906, + "learning_rate": 5.730995092881297e-05, + "loss": 1.7135, + "step": 7283 + }, + { + "epoch": 0.6512875536480687, + "grad_norm": 0.8618619310575364, + "learning_rate": 5.728376413288267e-05, + "loss": 1.8304, + "step": 7284 + }, + { + "epoch": 0.6513769670958512, + "grad_norm": 0.5802536034281083, + "learning_rate": 5.725758091951148e-05, + "loss": 1.5923, + "step": 7285 + }, + { + "epoch": 0.6514663805436338, + "grad_norm": 0.8820495781560629, + "learning_rate": 5.723140129089535e-05, + "loss": 1.8353, + "step": 7286 + }, + { + "epoch": 0.6515557939914163, + "grad_norm": 2.3176852765807543, + "learning_rate": 5.720522524922995e-05, + "loss": 6.4221, + "step": 7287 + }, + { + "epoch": 0.6516452074391988, + "grad_norm": 1.0584988584502497, + "learning_rate": 5.717905279671068e-05, + "loss": 1.618, + "step": 7288 + }, + { + "epoch": 0.6517346208869814, + "grad_norm": 1.329332427968024, + "learning_rate": 5.715288393553247e-05, + "loss": 1.9702, + "step": 7289 + }, + { + "epoch": 0.651824034334764, + "grad_norm": 0.5658557511238259, + "learning_rate": 5.712671866789015e-05, + "loss": 1.6733, + "step": 7290 + }, + { + "epoch": 0.6519134477825465, + "grad_norm": 1.0177943435014443, + "learning_rate": 5.710055699597816e-05, + "loss": 1.6843, + "step": 7291 + }, + { + "epoch": 0.652002861230329, + "grad_norm": 0.6545125129269027, + "learning_rate": 5.707439892199068e-05, + "loss": 2.019, + "step": 7292 + }, + { + "epoch": 0.6520922746781116, + "grad_norm": 0.7625954332379405, + "learning_rate": 5.7048244448121447e-05, + "loss": 2.1052, + "step": 7293 + }, + { + "epoch": 0.6521816881258942, + "grad_norm": 0.5147576965684216, + "learning_rate": 5.7022093576564165e-05, + "loss": 1.9374, + "step": 7294 + }, + { + "epoch": 0.6522711015736766, + "grad_norm": 0.523187831872023, + "learning_rate": 5.6995946309511924e-05, + "loss": 1.8964, + "step": 7295 + }, + { + "epoch": 0.6523605150214592, + "grad_norm": 0.557664894474007, + "learning_rate": 5.696980264915777e-05, + "loss": 1.7603, + "step": 7296 + }, + { + "epoch": 0.6524499284692418, + "grad_norm": 0.4726101692129084, + "learning_rate": 5.69436625976943e-05, + "loss": 1.9089, + "step": 7297 + }, + { + "epoch": 0.6525393419170243, + "grad_norm": 0.660703141216849, + "learning_rate": 5.691752615731384e-05, + "loss": 1.9639, + "step": 7298 + }, + { + "epoch": 0.6526287553648069, + "grad_norm": 0.39076133538974184, + "learning_rate": 5.689139333020842e-05, + "loss": 1.6824, + "step": 7299 + }, + { + "epoch": 0.6527181688125894, + "grad_norm": 0.5967535166526838, + "learning_rate": 5.686526411856978e-05, + "loss": 1.4594, + "step": 7300 + }, + { + "epoch": 0.6528075822603719, + "grad_norm": 0.5040088251978101, + "learning_rate": 5.6839138524589344e-05, + "loss": 1.4508, + "step": 7301 + }, + { + "epoch": 0.6528969957081545, + "grad_norm": 0.5877826650167339, + "learning_rate": 5.681301655045823e-05, + "loss": 1.4772, + "step": 7302 + }, + { + "epoch": 0.6529864091559371, + "grad_norm": 0.675192283646439, + "learning_rate": 5.678689819836731e-05, + "loss": 1.7197, + "step": 7303 + }, + { + "epoch": 0.6530758226037195, + "grad_norm": 0.5839202101952956, + "learning_rate": 5.6760783470506996e-05, + "loss": 1.5328, + "step": 7304 + }, + { + "epoch": 0.6531652360515021, + "grad_norm": 0.8536735529063643, + "learning_rate": 5.673467236906758e-05, + "loss": 1.7537, + "step": 7305 + }, + { + "epoch": 0.6532546494992847, + "grad_norm": 1.1383428970577543, + "learning_rate": 5.6708564896238944e-05, + "loss": 1.7064, + "step": 7306 + }, + { + "epoch": 0.6533440629470673, + "grad_norm": 0.5050727214410571, + "learning_rate": 5.6682461054210635e-05, + "loss": 1.6712, + "step": 7307 + }, + { + "epoch": 0.6534334763948498, + "grad_norm": 1.066486326757957, + "learning_rate": 5.6656360845172076e-05, + "loss": 1.5983, + "step": 7308 + }, + { + "epoch": 0.6535228898426323, + "grad_norm": 1.6380695772240528, + "learning_rate": 5.663026427131215e-05, + "loss": 1.7918, + "step": 7309 + }, + { + "epoch": 0.6536123032904149, + "grad_norm": 0.8372848656136969, + "learning_rate": 5.6604171334819564e-05, + "loss": 1.748, + "step": 7310 + }, + { + "epoch": 0.6537017167381974, + "grad_norm": 1.8258586946990272, + "learning_rate": 5.657808203788277e-05, + "loss": 2.0122, + "step": 7311 + }, + { + "epoch": 0.65379113018598, + "grad_norm": 1.0623722452274054, + "learning_rate": 5.6551996382689776e-05, + "loss": 1.3915, + "step": 7312 + }, + { + "epoch": 0.6538805436337625, + "grad_norm": 0.6872434820554219, + "learning_rate": 5.6525914371428344e-05, + "loss": 1.2285, + "step": 7313 + }, + { + "epoch": 0.653969957081545, + "grad_norm": 0.5583221290944317, + "learning_rate": 5.649983600628599e-05, + "loss": 1.4368, + "step": 7314 + }, + { + "epoch": 0.6540593705293276, + "grad_norm": 1.373300052499209, + "learning_rate": 5.647376128944984e-05, + "loss": 1.5039, + "step": 7315 + }, + { + "epoch": 0.6541487839771102, + "grad_norm": 1.4280043449613458, + "learning_rate": 5.6447690223106775e-05, + "loss": 1.5794, + "step": 7316 + }, + { + "epoch": 0.6542381974248928, + "grad_norm": 1.1617600587825572, + "learning_rate": 5.642162280944336e-05, + "loss": 1.5105, + "step": 7317 + }, + { + "epoch": 0.6543276108726752, + "grad_norm": 1.5908413227833413, + "learning_rate": 5.6395559050645794e-05, + "loss": 1.9577, + "step": 7318 + }, + { + "epoch": 0.6544170243204578, + "grad_norm": 1.2858114406285162, + "learning_rate": 5.6369498948900014e-05, + "loss": 1.5958, + "step": 7319 + }, + { + "epoch": 0.6545064377682404, + "grad_norm": 2.4021156265297763, + "learning_rate": 5.63434425063917e-05, + "loss": 1.8967, + "step": 7320 + }, + { + "epoch": 0.6545958512160229, + "grad_norm": 1.3572957820188196, + "learning_rate": 5.6317389725306066e-05, + "loss": 1.7819, + "step": 7321 + }, + { + "epoch": 0.6546852646638054, + "grad_norm": 1.2590012457817366, + "learning_rate": 5.629134060782828e-05, + "loss": 1.6063, + "step": 7322 + }, + { + "epoch": 0.654774678111588, + "grad_norm": 2.212916043212328, + "learning_rate": 5.626529515614294e-05, + "loss": 2.2607, + "step": 7323 + }, + { + "epoch": 0.6548640915593705, + "grad_norm": 2.0537464559972434, + "learning_rate": 5.6239253372434465e-05, + "loss": 1.8957, + "step": 7324 + }, + { + "epoch": 0.6549535050071531, + "grad_norm": 1.8386715293050941, + "learning_rate": 5.621321525888697e-05, + "loss": 1.7493, + "step": 7325 + }, + { + "epoch": 0.6550429184549357, + "grad_norm": 2.2667331957188335, + "learning_rate": 5.618718081768426e-05, + "loss": 1.885, + "step": 7326 + }, + { + "epoch": 0.6551323319027181, + "grad_norm": 1.6161392728359176, + "learning_rate": 5.616115005100975e-05, + "loss": 6.4379, + "step": 7327 + }, + { + "epoch": 0.6552217453505007, + "grad_norm": 2.273197542460971, + "learning_rate": 5.613512296104663e-05, + "loss": 1.8375, + "step": 7328 + }, + { + "epoch": 0.6553111587982833, + "grad_norm": 3.1068985457071103, + "learning_rate": 5.6109099549977786e-05, + "loss": 2.1023, + "step": 7329 + }, + { + "epoch": 0.6554005722460658, + "grad_norm": 0.8401062708394986, + "learning_rate": 5.608307981998574e-05, + "loss": 1.8922, + "step": 7330 + }, + { + "epoch": 0.6554899856938483, + "grad_norm": 1.2121489328260946, + "learning_rate": 5.6057063773252794e-05, + "loss": 1.9096, + "step": 7331 + }, + { + "epoch": 0.6555793991416309, + "grad_norm": 5.060985963494534, + "learning_rate": 5.603105141196081e-05, + "loss": 2.4698, + "step": 7332 + }, + { + "epoch": 0.6556688125894135, + "grad_norm": 1.02705010678193, + "learning_rate": 5.600504273829144e-05, + "loss": 2.0157, + "step": 7333 + }, + { + "epoch": 0.655758226037196, + "grad_norm": 2.3379857801411155, + "learning_rate": 5.5979037754426003e-05, + "loss": 1.6175, + "step": 7334 + }, + { + "epoch": 0.6558476394849786, + "grad_norm": 3.410707783055544, + "learning_rate": 5.5953036462545505e-05, + "loss": 1.5381, + "step": 7335 + }, + { + "epoch": 0.655937052932761, + "grad_norm": 1.2928595068140873, + "learning_rate": 5.592703886483064e-05, + "loss": 1.7312, + "step": 7336 + }, + { + "epoch": 0.6560264663805436, + "grad_norm": 1.7670561392670263, + "learning_rate": 5.590104496346185e-05, + "loss": 1.7593, + "step": 7337 + }, + { + "epoch": 0.6561158798283262, + "grad_norm": 2.1144318762601393, + "learning_rate": 5.5875054760619104e-05, + "loss": 1.5515, + "step": 7338 + }, + { + "epoch": 0.6562052932761088, + "grad_norm": 2.063988121405982, + "learning_rate": 5.584906825848224e-05, + "loss": 1.7368, + "step": 7339 + }, + { + "epoch": 0.6562947067238912, + "grad_norm": 0.5831053739405619, + "learning_rate": 5.582308545923074e-05, + "loss": 1.3428, + "step": 7340 + }, + { + "epoch": 0.6563841201716738, + "grad_norm": 2.1870496749436, + "learning_rate": 5.579710636504362e-05, + "loss": 1.9924, + "step": 7341 + }, + { + "epoch": 0.6564735336194564, + "grad_norm": 2.161273800318886, + "learning_rate": 5.577113097809989e-05, + "loss": 1.8455, + "step": 7342 + }, + { + "epoch": 0.656562947067239, + "grad_norm": 0.8494593625683502, + "learning_rate": 5.574515930057795e-05, + "loss": 1.6449, + "step": 7343 + }, + { + "epoch": 0.6566523605150214, + "grad_norm": 0.8521577683611035, + "learning_rate": 5.571919133465605e-05, + "loss": 1.5133, + "step": 7344 + }, + { + "epoch": 0.656741773962804, + "grad_norm": 2.128492243055363, + "learning_rate": 5.569322708251215e-05, + "loss": 1.8322, + "step": 7345 + }, + { + "epoch": 0.6568311874105865, + "grad_norm": 0.921538813540504, + "learning_rate": 5.5667266546323723e-05, + "loss": 1.5825, + "step": 7346 + }, + { + "epoch": 0.6569206008583691, + "grad_norm": 1.6223158583921387, + "learning_rate": 5.564130972826813e-05, + "loss": 2.2989, + "step": 7347 + }, + { + "epoch": 0.6570100143061517, + "grad_norm": 1.1055485357580477, + "learning_rate": 5.561535663052231e-05, + "loss": 1.7685, + "step": 7348 + }, + { + "epoch": 0.6570994277539342, + "grad_norm": 0.6008080015967439, + "learning_rate": 5.558940725526291e-05, + "loss": 1.8121, + "step": 7349 + }, + { + "epoch": 0.6571888412017167, + "grad_norm": 1.2816116467864882, + "learning_rate": 5.5563461604666325e-05, + "loss": 1.618, + "step": 7350 + }, + { + "epoch": 0.6572782546494993, + "grad_norm": 0.7809529260570702, + "learning_rate": 5.553751968090857e-05, + "loss": 1.6143, + "step": 7351 + }, + { + "epoch": 0.6573676680972819, + "grad_norm": 0.6809092609131682, + "learning_rate": 5.55115814861653e-05, + "loss": 1.6939, + "step": 7352 + }, + { + "epoch": 0.6574570815450643, + "grad_norm": 0.45131828347853575, + "learning_rate": 5.548564702261196e-05, + "loss": 1.4123, + "step": 7353 + }, + { + "epoch": 0.6575464949928469, + "grad_norm": 1.123642889141304, + "learning_rate": 5.545971629242369e-05, + "loss": 1.7066, + "step": 7354 + }, + { + "epoch": 0.6576359084406295, + "grad_norm": 1.0186381791318546, + "learning_rate": 5.543378929777514e-05, + "loss": 1.7614, + "step": 7355 + }, + { + "epoch": 0.657725321888412, + "grad_norm": 0.8533346031869824, + "learning_rate": 5.540786604084091e-05, + "loss": 1.6537, + "step": 7356 + }, + { + "epoch": 0.6578147353361946, + "grad_norm": 0.7546208731581144, + "learning_rate": 5.538194652379514e-05, + "loss": 2.0539, + "step": 7357 + }, + { + "epoch": 0.6579041487839771, + "grad_norm": 0.9502765993328983, + "learning_rate": 5.5356030748811575e-05, + "loss": 1.9371, + "step": 7358 + }, + { + "epoch": 0.6579935622317596, + "grad_norm": 1.1911602081340922, + "learning_rate": 5.5330118718063795e-05, + "loss": 1.7007, + "step": 7359 + }, + { + "epoch": 0.6580829756795422, + "grad_norm": 0.5626843357974801, + "learning_rate": 5.530421043372507e-05, + "loss": 1.7184, + "step": 7360 + }, + { + "epoch": 0.6581723891273248, + "grad_norm": 0.7960469111497911, + "learning_rate": 5.5278305897968185e-05, + "loss": 1.3874, + "step": 7361 + }, + { + "epoch": 0.6582618025751072, + "grad_norm": 0.6274017471489018, + "learning_rate": 5.525240511296577e-05, + "loss": 2.0988, + "step": 7362 + }, + { + "epoch": 0.6583512160228898, + "grad_norm": 1.1935019498927628, + "learning_rate": 5.522650808089011e-05, + "loss": 1.6058, + "step": 7363 + }, + { + "epoch": 0.6584406294706724, + "grad_norm": 1.1032681312839534, + "learning_rate": 5.520061480391313e-05, + "loss": 1.5404, + "step": 7364 + }, + { + "epoch": 0.658530042918455, + "grad_norm": 0.8248896536744735, + "learning_rate": 5.517472528420653e-05, + "loss": 2.0069, + "step": 7365 + }, + { + "epoch": 0.6586194563662375, + "grad_norm": 1.0583906857388707, + "learning_rate": 5.514883952394154e-05, + "loss": 1.8662, + "step": 7366 + }, + { + "epoch": 0.65870886981402, + "grad_norm": 0.7762504487457765, + "learning_rate": 5.512295752528922e-05, + "loss": 1.4262, + "step": 7367 + }, + { + "epoch": 0.6587982832618026, + "grad_norm": 0.38422114385156647, + "learning_rate": 5.50970792904203e-05, + "loss": 1.366, + "step": 7368 + }, + { + "epoch": 0.6588876967095851, + "grad_norm": 0.660750465321797, + "learning_rate": 5.507120482150501e-05, + "loss": 2.0432, + "step": 7369 + }, + { + "epoch": 0.6589771101573677, + "grad_norm": 0.391265424703362, + "learning_rate": 5.5045334120713565e-05, + "loss": 1.6353, + "step": 7370 + }, + { + "epoch": 0.6590665236051502, + "grad_norm": 0.3931580009411282, + "learning_rate": 5.501946719021569e-05, + "loss": 1.4911, + "step": 7371 + }, + { + "epoch": 0.6591559370529327, + "grad_norm": 0.7462345550059045, + "learning_rate": 5.4993604032180746e-05, + "loss": 1.387, + "step": 7372 + }, + { + "epoch": 0.6592453505007153, + "grad_norm": 0.5320292205679149, + "learning_rate": 5.496774464877787e-05, + "loss": 1.7107, + "step": 7373 + }, + { + "epoch": 0.6593347639484979, + "grad_norm": 1.2825131085232075, + "learning_rate": 5.494188904217592e-05, + "loss": 1.832, + "step": 7374 + }, + { + "epoch": 0.6594241773962805, + "grad_norm": 0.3556720744951579, + "learning_rate": 5.491603721454327e-05, + "loss": 1.5299, + "step": 7375 + }, + { + "epoch": 0.6595135908440629, + "grad_norm": 1.134804085012986, + "learning_rate": 5.489018916804813e-05, + "loss": 1.8313, + "step": 7376 + }, + { + "epoch": 0.6596030042918455, + "grad_norm": 0.43694957133385365, + "learning_rate": 5.4864344904858345e-05, + "loss": 1.5144, + "step": 7377 + }, + { + "epoch": 0.6596924177396281, + "grad_norm": 0.44043387749743357, + "learning_rate": 5.483850442714145e-05, + "loss": 1.6517, + "step": 7378 + }, + { + "epoch": 0.6597818311874106, + "grad_norm": 0.8793956512935084, + "learning_rate": 5.481266773706468e-05, + "loss": 1.6459, + "step": 7379 + }, + { + "epoch": 0.6598712446351931, + "grad_norm": 0.558842872096171, + "learning_rate": 5.4786834836794855e-05, + "loss": 1.7897, + "step": 7380 + }, + { + "epoch": 0.6599606580829757, + "grad_norm": 0.6871764697268133, + "learning_rate": 5.4761005728498594e-05, + "loss": 1.5349, + "step": 7381 + }, + { + "epoch": 0.6600500715307582, + "grad_norm": 0.8300005858192422, + "learning_rate": 5.4735180414342134e-05, + "loss": 1.65, + "step": 7382 + }, + { + "epoch": 0.6601394849785408, + "grad_norm": 0.7314600719785715, + "learning_rate": 5.4709358896491445e-05, + "loss": 2.1215, + "step": 7383 + }, + { + "epoch": 0.6602288984263234, + "grad_norm": 0.865259605318468, + "learning_rate": 5.468354117711212e-05, + "loss": 1.4641, + "step": 7384 + }, + { + "epoch": 0.6603183118741058, + "grad_norm": 0.9151445779220387, + "learning_rate": 5.465772725836951e-05, + "loss": 1.7566, + "step": 7385 + }, + { + "epoch": 0.6604077253218884, + "grad_norm": 0.7720667545741066, + "learning_rate": 5.463191714242851e-05, + "loss": 1.6384, + "step": 7386 + }, + { + "epoch": 0.660497138769671, + "grad_norm": 0.7025896003490117, + "learning_rate": 5.4606110831453836e-05, + "loss": 1.7041, + "step": 7387 + }, + { + "epoch": 0.6605865522174535, + "grad_norm": 0.5091602597092021, + "learning_rate": 5.458030832760985e-05, + "loss": 1.5445, + "step": 7388 + }, + { + "epoch": 0.660675965665236, + "grad_norm": 1.194532411311789, + "learning_rate": 5.4554509633060524e-05, + "loss": 1.8484, + "step": 7389 + }, + { + "epoch": 0.6607653791130186, + "grad_norm": 0.8372014973400453, + "learning_rate": 5.452871474996955e-05, + "loss": 1.812, + "step": 7390 + }, + { + "epoch": 0.6608547925608012, + "grad_norm": 0.4802385362996391, + "learning_rate": 5.450292368050043e-05, + "loss": 1.4113, + "step": 7391 + }, + { + "epoch": 0.6609442060085837, + "grad_norm": 0.5218758372935178, + "learning_rate": 5.447713642681612e-05, + "loss": 1.9681, + "step": 7392 + }, + { + "epoch": 0.6610336194563662, + "grad_norm": 0.9159079108671213, + "learning_rate": 5.44513529910794e-05, + "loss": 1.165, + "step": 7393 + }, + { + "epoch": 0.6611230329041488, + "grad_norm": 1.3498835896544656, + "learning_rate": 5.442557337545273e-05, + "loss": 1.9319, + "step": 7394 + }, + { + "epoch": 0.6612124463519313, + "grad_norm": 1.0376294974843363, + "learning_rate": 5.4399797582098144e-05, + "loss": 2.016, + "step": 7395 + }, + { + "epoch": 0.6613018597997139, + "grad_norm": 1.0749402412936857, + "learning_rate": 5.437402561317746e-05, + "loss": 1.4097, + "step": 7396 + }, + { + "epoch": 0.6613912732474965, + "grad_norm": 1.934974508292019, + "learning_rate": 5.434825747085215e-05, + "loss": 1.7532, + "step": 7397 + }, + { + "epoch": 0.6614806866952789, + "grad_norm": 1.5562174315174069, + "learning_rate": 5.432249315728336e-05, + "loss": 1.844, + "step": 7398 + }, + { + "epoch": 0.6615701001430615, + "grad_norm": 0.9316726706408134, + "learning_rate": 5.429673267463193e-05, + "loss": 1.2838, + "step": 7399 + }, + { + "epoch": 0.6616595135908441, + "grad_norm": 1.4717489668671482, + "learning_rate": 5.427097602505831e-05, + "loss": 2.0003, + "step": 7400 + }, + { + "epoch": 0.6617489270386266, + "grad_norm": 1.3323969688745956, + "learning_rate": 5.42452232107227e-05, + "loss": 1.3746, + "step": 7401 + }, + { + "epoch": 0.6618383404864091, + "grad_norm": 1.0592666252548566, + "learning_rate": 5.4219474233785e-05, + "loss": 1.7706, + "step": 7402 + }, + { + "epoch": 0.6619277539341917, + "grad_norm": 1.4832919686835815, + "learning_rate": 5.419372909640466e-05, + "loss": 1.9595, + "step": 7403 + }, + { + "epoch": 0.6620171673819742, + "grad_norm": 0.7749291688425538, + "learning_rate": 5.416798780074091e-05, + "loss": 1.8583, + "step": 7404 + }, + { + "epoch": 0.6621065808297568, + "grad_norm": 0.8039797979703428, + "learning_rate": 5.414225034895273e-05, + "loss": 1.882, + "step": 7405 + }, + { + "epoch": 0.6621959942775394, + "grad_norm": 1.4017212933168788, + "learning_rate": 5.411651674319862e-05, + "loss": 1.9961, + "step": 7406 + }, + { + "epoch": 0.6622854077253219, + "grad_norm": 0.6044903840475992, + "learning_rate": 5.409078698563682e-05, + "loss": 1.7836, + "step": 7407 + }, + { + "epoch": 0.6623748211731044, + "grad_norm": 0.7149051019489548, + "learning_rate": 5.4065061078425315e-05, + "loss": 1.7842, + "step": 7408 + }, + { + "epoch": 0.662464234620887, + "grad_norm": 0.8345594385232182, + "learning_rate": 5.403933902372162e-05, + "loss": 1.5594, + "step": 7409 + }, + { + "epoch": 0.6625536480686696, + "grad_norm": 0.7630476407182826, + "learning_rate": 5.401362082368306e-05, + "loss": 1.8516, + "step": 7410 + }, + { + "epoch": 0.662643061516452, + "grad_norm": 0.9414683775074755, + "learning_rate": 5.3987906480466586e-05, + "loss": 1.6806, + "step": 7411 + }, + { + "epoch": 0.6627324749642346, + "grad_norm": 0.760091905355426, + "learning_rate": 5.3962195996228825e-05, + "loss": 1.6393, + "step": 7412 + }, + { + "epoch": 0.6628218884120172, + "grad_norm": 1.090434803648923, + "learning_rate": 5.3936489373126075e-05, + "loss": 1.5367, + "step": 7413 + }, + { + "epoch": 0.6629113018597997, + "grad_norm": 0.7389625894396974, + "learning_rate": 5.391078661331439e-05, + "loss": 1.4172, + "step": 7414 + }, + { + "epoch": 0.6630007153075823, + "grad_norm": 0.40816899212806473, + "learning_rate": 5.388508771894931e-05, + "loss": 1.6315, + "step": 7415 + }, + { + "epoch": 0.6630901287553648, + "grad_norm": 1.49843757399479, + "learning_rate": 5.385939269218625e-05, + "loss": 2.2721, + "step": 7416 + }, + { + "epoch": 0.6631795422031473, + "grad_norm": 1.007411776209617, + "learning_rate": 5.383370153518019e-05, + "loss": 6.2774, + "step": 7417 + }, + { + "epoch": 0.6632689556509299, + "grad_norm": 1.1750590178764548, + "learning_rate": 5.3808014250085836e-05, + "loss": 1.4594, + "step": 7418 + }, + { + "epoch": 0.6633583690987125, + "grad_norm": 1.3758172189139828, + "learning_rate": 5.3782330839057573e-05, + "loss": 1.6638, + "step": 7419 + }, + { + "epoch": 0.663447782546495, + "grad_norm": 0.8544161039546262, + "learning_rate": 5.375665130424936e-05, + "loss": 1.6616, + "step": 7420 + }, + { + "epoch": 0.6635371959942775, + "grad_norm": 0.699866744154163, + "learning_rate": 5.373097564781496e-05, + "loss": 1.3092, + "step": 7421 + }, + { + "epoch": 0.6636266094420601, + "grad_norm": 2.64760290251266, + "learning_rate": 5.3705303871907795e-05, + "loss": 6.5051, + "step": 7422 + }, + { + "epoch": 0.6637160228898427, + "grad_norm": 1.3926171500972657, + "learning_rate": 5.3679635978680843e-05, + "loss": 1.895, + "step": 7423 + }, + { + "epoch": 0.6638054363376252, + "grad_norm": 2.109333406710138, + "learning_rate": 5.365397197028685e-05, + "loss": 6.5912, + "step": 7424 + }, + { + "epoch": 0.6638948497854077, + "grad_norm": 1.0380345371931596, + "learning_rate": 5.3628311848878333e-05, + "loss": 1.4687, + "step": 7425 + }, + { + "epoch": 0.6639842632331903, + "grad_norm": 1.5414515617537323, + "learning_rate": 5.360265561660725e-05, + "loss": 1.7775, + "step": 7426 + }, + { + "epoch": 0.6640736766809728, + "grad_norm": 0.678559527495712, + "learning_rate": 5.35770032756254e-05, + "loss": 1.6343, + "step": 7427 + }, + { + "epoch": 0.6641630901287554, + "grad_norm": 1.1820207750509053, + "learning_rate": 5.3551354828084276e-05, + "loss": 1.6954, + "step": 7428 + }, + { + "epoch": 0.6642525035765379, + "grad_norm": 0.4105759916457698, + "learning_rate": 5.352571027613489e-05, + "loss": 1.2561, + "step": 7429 + }, + { + "epoch": 0.6643419170243204, + "grad_norm": 0.7405789074826594, + "learning_rate": 5.350006962192804e-05, + "loss": 1.7782, + "step": 7430 + }, + { + "epoch": 0.664431330472103, + "grad_norm": 0.8481044933688392, + "learning_rate": 5.34744328676142e-05, + "loss": 1.5347, + "step": 7431 + }, + { + "epoch": 0.6645207439198856, + "grad_norm": 0.5840389779257366, + "learning_rate": 5.344880001534349e-05, + "loss": 1.4732, + "step": 7432 + }, + { + "epoch": 0.664610157367668, + "grad_norm": 0.613121002780577, + "learning_rate": 5.342317106726574e-05, + "loss": 1.418, + "step": 7433 + }, + { + "epoch": 0.6646995708154506, + "grad_norm": 0.5634072875661909, + "learning_rate": 5.339754602553034e-05, + "loss": 1.3724, + "step": 7434 + }, + { + "epoch": 0.6647889842632332, + "grad_norm": 0.7905569729510024, + "learning_rate": 5.3371924892286484e-05, + "loss": 1.7838, + "step": 7435 + }, + { + "epoch": 0.6648783977110158, + "grad_norm": 0.4717248524424835, + "learning_rate": 5.3346307669683005e-05, + "loss": 1.4296, + "step": 7436 + }, + { + "epoch": 0.6649678111587983, + "grad_norm": 1.0815288936553564, + "learning_rate": 5.332069435986832e-05, + "loss": 2.2383, + "step": 7437 + }, + { + "epoch": 0.6650572246065808, + "grad_norm": 0.7769737842948685, + "learning_rate": 5.329508496499058e-05, + "loss": 1.7059, + "step": 7438 + }, + { + "epoch": 0.6651466380543634, + "grad_norm": 0.9783964211994571, + "learning_rate": 5.326947948719775e-05, + "loss": 1.6819, + "step": 7439 + }, + { + "epoch": 0.6652360515021459, + "grad_norm": 0.8652743104913789, + "learning_rate": 5.324387792863719e-05, + "loss": 1.8377, + "step": 7440 + }, + { + "epoch": 0.6653254649499285, + "grad_norm": 0.5790388390897719, + "learning_rate": 5.3218280291456126e-05, + "loss": 1.489, + "step": 7441 + }, + { + "epoch": 0.665414878397711, + "grad_norm": 0.8908407271391877, + "learning_rate": 5.319268657780143e-05, + "loss": 6.3923, + "step": 7442 + }, + { + "epoch": 0.6655042918454935, + "grad_norm": 2.358919708263631, + "learning_rate": 5.316709678981955e-05, + "loss": 1.923, + "step": 7443 + }, + { + "epoch": 0.6655937052932761, + "grad_norm": 1.6097046315328314, + "learning_rate": 5.314151092965669e-05, + "loss": 1.6207, + "step": 7444 + }, + { + "epoch": 0.6656831187410587, + "grad_norm": 1.0530843208986043, + "learning_rate": 5.311592899945873e-05, + "loss": 2.0372, + "step": 7445 + }, + { + "epoch": 0.6657725321888412, + "grad_norm": 0.8467956308587321, + "learning_rate": 5.3090351001371185e-05, + "loss": 1.5176, + "step": 7446 + }, + { + "epoch": 0.6658619456366237, + "grad_norm": 1.8087923795519143, + "learning_rate": 5.306477693753924e-05, + "loss": 1.6185, + "step": 7447 + }, + { + "epoch": 0.6659513590844063, + "grad_norm": 3.2433381559797843, + "learning_rate": 5.303920681010781e-05, + "loss": 6.4624, + "step": 7448 + }, + { + "epoch": 0.6660407725321889, + "grad_norm": 0.6710117053363869, + "learning_rate": 5.301364062122136e-05, + "loss": 1.6668, + "step": 7449 + }, + { + "epoch": 0.6661301859799714, + "grad_norm": 0.8867960147449536, + "learning_rate": 5.298807837302411e-05, + "loss": 1.4798, + "step": 7450 + }, + { + "epoch": 0.6662195994277539, + "grad_norm": 2.308692724763356, + "learning_rate": 5.2962520067660004e-05, + "loss": 1.7242, + "step": 7451 + }, + { + "epoch": 0.6663090128755365, + "grad_norm": 1.2372107125090377, + "learning_rate": 5.2936965707272446e-05, + "loss": 1.5623, + "step": 7452 + }, + { + "epoch": 0.666398426323319, + "grad_norm": 0.9657073183711193, + "learning_rate": 5.291141529400483e-05, + "loss": 1.5634, + "step": 7453 + }, + { + "epoch": 0.6664878397711016, + "grad_norm": 2.3843825256659756, + "learning_rate": 5.288586882999989e-05, + "loss": 1.7618, + "step": 7454 + }, + { + "epoch": 0.6665772532188842, + "grad_norm": 1.8853120924571678, + "learning_rate": 5.286032631740023e-05, + "loss": 1.5321, + "step": 7455 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 1.9681742613610202, + "learning_rate": 5.283478775834811e-05, + "loss": 1.9105, + "step": 7456 + }, + { + "epoch": 0.6667560801144492, + "grad_norm": 1.4124205844824544, + "learning_rate": 5.280925315498536e-05, + "loss": 1.484, + "step": 7457 + }, + { + "epoch": 0.6668454935622318, + "grad_norm": 1.1976286359593955, + "learning_rate": 5.278372250945354e-05, + "loss": 1.625, + "step": 7458 + }, + { + "epoch": 0.6669349070100143, + "grad_norm": 1.846238617569942, + "learning_rate": 5.2758195823893896e-05, + "loss": 1.6337, + "step": 7459 + }, + { + "epoch": 0.6670243204577968, + "grad_norm": 2.0413999661194384, + "learning_rate": 5.273267310044732e-05, + "loss": 1.994, + "step": 7460 + }, + { + "epoch": 0.6671137339055794, + "grad_norm": 0.7535252527645097, + "learning_rate": 5.270715434125435e-05, + "loss": 1.651, + "step": 7461 + }, + { + "epoch": 0.667203147353362, + "grad_norm": 1.189411938645001, + "learning_rate": 5.2681639548455284e-05, + "loss": 1.9814, + "step": 7462 + }, + { + "epoch": 0.6672925608011445, + "grad_norm": 1.8039632940176495, + "learning_rate": 5.2656128724189916e-05, + "loss": 1.9469, + "step": 7463 + }, + { + "epoch": 0.6673819742489271, + "grad_norm": 0.511460667866729, + "learning_rate": 5.263062187059785e-05, + "loss": 1.4345, + "step": 7464 + }, + { + "epoch": 0.6674713876967096, + "grad_norm": 0.7220760947304127, + "learning_rate": 5.260511898981837e-05, + "loss": 1.8492, + "step": 7465 + }, + { + "epoch": 0.6675608011444921, + "grad_norm": 0.44124182280122126, + "learning_rate": 5.2579620083990244e-05, + "loss": 1.3123, + "step": 7466 + }, + { + "epoch": 0.6676502145922747, + "grad_norm": 0.9839877000124762, + "learning_rate": 5.2554125155252175e-05, + "loss": 1.5748, + "step": 7467 + }, + { + "epoch": 0.6677396280400573, + "grad_norm": 0.9487692185274359, + "learning_rate": 5.25286342057423e-05, + "loss": 1.846, + "step": 7468 + }, + { + "epoch": 0.6678290414878397, + "grad_norm": 1.0348642854798773, + "learning_rate": 5.2503147237598546e-05, + "loss": 1.6923, + "step": 7469 + }, + { + "epoch": 0.6679184549356223, + "grad_norm": 0.9316269082289234, + "learning_rate": 5.247766425295848e-05, + "loss": 1.3901, + "step": 7470 + }, + { + "epoch": 0.6680078683834049, + "grad_norm": 0.9572239027963486, + "learning_rate": 5.245218525395934e-05, + "loss": 1.5636, + "step": 7471 + }, + { + "epoch": 0.6680972818311874, + "grad_norm": 0.8026769277709601, + "learning_rate": 5.242671024273798e-05, + "loss": 2.2223, + "step": 7472 + }, + { + "epoch": 0.66818669527897, + "grad_norm": 1.344644761048616, + "learning_rate": 5.240123922143096e-05, + "loss": 1.6227, + "step": 7473 + }, + { + "epoch": 0.6682761087267525, + "grad_norm": 1.6274887905647712, + "learning_rate": 5.2375772192174534e-05, + "loss": 2.033, + "step": 7474 + }, + { + "epoch": 0.668365522174535, + "grad_norm": 0.47830182839090374, + "learning_rate": 5.235030915710457e-05, + "loss": 1.2915, + "step": 7475 + }, + { + "epoch": 0.6684549356223176, + "grad_norm": 0.43572555127402113, + "learning_rate": 5.2324850118356674e-05, + "loss": 1.3899, + "step": 7476 + }, + { + "epoch": 0.6685443490701002, + "grad_norm": 1.8788309975867927, + "learning_rate": 5.229939507806598e-05, + "loss": 1.5669, + "step": 7477 + }, + { + "epoch": 0.6686337625178826, + "grad_norm": 0.760272083532738, + "learning_rate": 5.2273944038367416e-05, + "loss": 1.4827, + "step": 7478 + }, + { + "epoch": 0.6687231759656652, + "grad_norm": 0.7981895992560002, + "learning_rate": 5.224849700139557e-05, + "loss": 1.5878, + "step": 7479 + }, + { + "epoch": 0.6688125894134478, + "grad_norm": 0.4183357523548222, + "learning_rate": 5.222305396928453e-05, + "loss": 1.5221, + "step": 7480 + }, + { + "epoch": 0.6689020028612304, + "grad_norm": 1.1333422425831945, + "learning_rate": 5.219761494416828e-05, + "loss": 1.9917, + "step": 7481 + }, + { + "epoch": 0.6689914163090128, + "grad_norm": 1.1905860399576944, + "learning_rate": 5.2172179928180395e-05, + "loss": 1.6731, + "step": 7482 + }, + { + "epoch": 0.6690808297567954, + "grad_norm": 0.5289708541713857, + "learning_rate": 5.214674892345397e-05, + "loss": 1.7595, + "step": 7483 + }, + { + "epoch": 0.669170243204578, + "grad_norm": 1.3110775216041926, + "learning_rate": 5.2121321932121916e-05, + "loss": 1.5734, + "step": 7484 + }, + { + "epoch": 0.6692596566523605, + "grad_norm": 1.012972071979961, + "learning_rate": 5.209589895631681e-05, + "loss": 1.5109, + "step": 7485 + }, + { + "epoch": 0.6693490701001431, + "grad_norm": 1.104938272921112, + "learning_rate": 5.207047999817076e-05, + "loss": 1.6637, + "step": 7486 + }, + { + "epoch": 0.6694384835479256, + "grad_norm": 0.6175596090409425, + "learning_rate": 5.2045065059815676e-05, + "loss": 1.3819, + "step": 7487 + }, + { + "epoch": 0.6695278969957081, + "grad_norm": 2.1094216265137025, + "learning_rate": 5.201965414338308e-05, + "loss": 1.9108, + "step": 7488 + }, + { + "epoch": 0.6696173104434907, + "grad_norm": 1.0729731470247537, + "learning_rate": 5.199424725100413e-05, + "loss": 1.7373, + "step": 7489 + }, + { + "epoch": 0.6697067238912733, + "grad_norm": 0.8808453841650143, + "learning_rate": 5.1968844384809734e-05, + "loss": 2.1123, + "step": 7490 + }, + { + "epoch": 0.6697961373390557, + "grad_norm": 1.6486161346852268, + "learning_rate": 5.194344554693032e-05, + "loss": 1.6494, + "step": 7491 + }, + { + "epoch": 0.6698855507868383, + "grad_norm": 2.0682689243228185, + "learning_rate": 5.1918050739496074e-05, + "loss": 1.7081, + "step": 7492 + }, + { + "epoch": 0.6699749642346209, + "grad_norm": 0.5416254041087772, + "learning_rate": 5.189265996463689e-05, + "loss": 1.7681, + "step": 7493 + }, + { + "epoch": 0.6700643776824035, + "grad_norm": 1.5416257142759042, + "learning_rate": 5.186727322448214e-05, + "loss": 1.7126, + "step": 7494 + }, + { + "epoch": 0.670153791130186, + "grad_norm": 1.3260974051012102, + "learning_rate": 5.1841890521161085e-05, + "loss": 1.8539, + "step": 7495 + }, + { + "epoch": 0.6702432045779685, + "grad_norm": 0.4419328125316366, + "learning_rate": 5.181651185680256e-05, + "loss": 1.617, + "step": 7496 + }, + { + "epoch": 0.6703326180257511, + "grad_norm": 0.5771980807424507, + "learning_rate": 5.1791137233534946e-05, + "loss": 1.3397, + "step": 7497 + }, + { + "epoch": 0.6704220314735336, + "grad_norm": 1.1941620014653402, + "learning_rate": 5.1765766653486446e-05, + "loss": 1.9451, + "step": 7498 + }, + { + "epoch": 0.6705114449213162, + "grad_norm": 1.0787677435776075, + "learning_rate": 5.174040011878487e-05, + "loss": 1.64, + "step": 7499 + }, + { + "epoch": 0.6706008583690987, + "grad_norm": 0.5290317145814452, + "learning_rate": 5.171503763155758e-05, + "loss": 1.6678, + "step": 7500 + }, + { + "epoch": 0.6706902718168812, + "grad_norm": 0.7576878767465244, + "learning_rate": 5.168967919393186e-05, + "loss": 1.6774, + "step": 7501 + }, + { + "epoch": 0.6707796852646638, + "grad_norm": 0.9361781623283102, + "learning_rate": 5.166432480803435e-05, + "loss": 1.6806, + "step": 7502 + }, + { + "epoch": 0.6708690987124464, + "grad_norm": 0.5216360647751797, + "learning_rate": 5.1638974475991554e-05, + "loss": 2.0733, + "step": 7503 + }, + { + "epoch": 0.670958512160229, + "grad_norm": 0.5905463101286709, + "learning_rate": 5.1613628199929544e-05, + "loss": 1.9613, + "step": 7504 + }, + { + "epoch": 0.6710479256080114, + "grad_norm": 1.2522726912319737, + "learning_rate": 5.158828598197416e-05, + "loss": 1.7001, + "step": 7505 + }, + { + "epoch": 0.671137339055794, + "grad_norm": 0.42966513306299964, + "learning_rate": 5.1562947824250704e-05, + "loss": 1.7219, + "step": 7506 + }, + { + "epoch": 0.6712267525035766, + "grad_norm": 0.7122965493601354, + "learning_rate": 5.1537613728884335e-05, + "loss": 1.5669, + "step": 7507 + }, + { + "epoch": 0.6713161659513591, + "grad_norm": 1.0066301609133552, + "learning_rate": 5.151228369799976e-05, + "loss": 2.0237, + "step": 7508 + }, + { + "epoch": 0.6714055793991416, + "grad_norm": 0.9268959992820968, + "learning_rate": 5.1486957733721405e-05, + "loss": 2.2367, + "step": 7509 + }, + { + "epoch": 0.6714949928469242, + "grad_norm": 1.0674294099762234, + "learning_rate": 5.146163583817336e-05, + "loss": 1.6184, + "step": 7510 + }, + { + "epoch": 0.6715844062947067, + "grad_norm": 1.5557062247676101, + "learning_rate": 5.143631801347926e-05, + "loss": 1.6094, + "step": 7511 + }, + { + "epoch": 0.6716738197424893, + "grad_norm": 1.4736227523962064, + "learning_rate": 5.14110042617625e-05, + "loss": 1.8794, + "step": 7512 + }, + { + "epoch": 0.6717632331902719, + "grad_norm": 1.3959715796031236, + "learning_rate": 5.138569458514617e-05, + "loss": 1.9509, + "step": 7513 + }, + { + "epoch": 0.6718526466380543, + "grad_norm": 1.2055000920387102, + "learning_rate": 5.136038898575286e-05, + "loss": 1.7342, + "step": 7514 + }, + { + "epoch": 0.6719420600858369, + "grad_norm": 1.3378256551462329, + "learning_rate": 5.133508746570502e-05, + "loss": 1.984, + "step": 7515 + }, + { + "epoch": 0.6720314735336195, + "grad_norm": 1.026954275441856, + "learning_rate": 5.130979002712466e-05, + "loss": 1.8682, + "step": 7516 + }, + { + "epoch": 0.672120886981402, + "grad_norm": 1.0242915846075353, + "learning_rate": 5.128449667213337e-05, + "loss": 1.6941, + "step": 7517 + }, + { + "epoch": 0.6722103004291845, + "grad_norm": 1.2666914320327005, + "learning_rate": 5.1259207402852506e-05, + "loss": 1.5761, + "step": 7518 + }, + { + "epoch": 0.6722997138769671, + "grad_norm": 1.5911211397008547, + "learning_rate": 5.1233922221403094e-05, + "loss": 1.5973, + "step": 7519 + }, + { + "epoch": 0.6723891273247496, + "grad_norm": 1.2563115520158836, + "learning_rate": 5.120864112990569e-05, + "loss": 1.5975, + "step": 7520 + }, + { + "epoch": 0.6724785407725322, + "grad_norm": 1.1421171346568342, + "learning_rate": 5.118336413048064e-05, + "loss": 1.8575, + "step": 7521 + }, + { + "epoch": 0.6725679542203148, + "grad_norm": 0.5566336772347854, + "learning_rate": 5.115809122524787e-05, + "loss": 1.7654, + "step": 7522 + }, + { + "epoch": 0.6726573676680973, + "grad_norm": 0.6062468142727844, + "learning_rate": 5.113282241632702e-05, + "loss": 1.7543, + "step": 7523 + }, + { + "epoch": 0.6727467811158798, + "grad_norm": 0.7086513885472548, + "learning_rate": 5.110755770583736e-05, + "loss": 1.6677, + "step": 7524 + }, + { + "epoch": 0.6728361945636624, + "grad_norm": 0.7429534219549173, + "learning_rate": 5.108229709589776e-05, + "loss": 1.5051, + "step": 7525 + }, + { + "epoch": 0.672925608011445, + "grad_norm": 0.7722249322153968, + "learning_rate": 5.1057040588626816e-05, + "loss": 1.9625, + "step": 7526 + }, + { + "epoch": 0.6730150214592274, + "grad_norm": 0.7098825733483547, + "learning_rate": 5.103178818614277e-05, + "loss": 1.4826, + "step": 7527 + }, + { + "epoch": 0.67310443490701, + "grad_norm": 0.9830186715429259, + "learning_rate": 5.100653989056352e-05, + "loss": 1.828, + "step": 7528 + }, + { + "epoch": 0.6731938483547926, + "grad_norm": 0.8422983194517605, + "learning_rate": 5.098129570400658e-05, + "loss": 1.4752, + "step": 7529 + }, + { + "epoch": 0.6732832618025751, + "grad_norm": 0.5359313928765542, + "learning_rate": 5.095605562858923e-05, + "loss": 1.9938, + "step": 7530 + }, + { + "epoch": 0.6733726752503576, + "grad_norm": 1.1535453710555055, + "learning_rate": 5.093081966642822e-05, + "loss": 1.4909, + "step": 7531 + }, + { + "epoch": 0.6734620886981402, + "grad_norm": 0.8191502524097676, + "learning_rate": 5.09055878196401e-05, + "loss": 1.8219, + "step": 7532 + }, + { + "epoch": 0.6735515021459227, + "grad_norm": 0.5873454305067464, + "learning_rate": 5.0880360090341084e-05, + "loss": 1.9444, + "step": 7533 + }, + { + "epoch": 0.6736409155937053, + "grad_norm": 1.2920427227058229, + "learning_rate": 5.08551364806469e-05, + "loss": 1.8354, + "step": 7534 + }, + { + "epoch": 0.6737303290414879, + "grad_norm": 0.5205585386947105, + "learning_rate": 5.0829916992673035e-05, + "loss": 1.7615, + "step": 7535 + }, + { + "epoch": 0.6738197424892703, + "grad_norm": 0.44835691593876464, + "learning_rate": 5.080470162853472e-05, + "loss": 1.5383, + "step": 7536 + }, + { + "epoch": 0.6739091559370529, + "grad_norm": 0.6146776372741721, + "learning_rate": 5.0779490390346626e-05, + "loss": 1.1867, + "step": 7537 + }, + { + "epoch": 0.6739985693848355, + "grad_norm": 0.8574745631200538, + "learning_rate": 5.075428328022325e-05, + "loss": 1.7061, + "step": 7538 + }, + { + "epoch": 0.6740879828326181, + "grad_norm": 0.3706609798512897, + "learning_rate": 5.0729080300278676e-05, + "loss": 1.6001, + "step": 7539 + }, + { + "epoch": 0.6741773962804005, + "grad_norm": 0.6250904743988861, + "learning_rate": 5.07038814526266e-05, + "loss": 1.4797, + "step": 7540 + }, + { + "epoch": 0.6742668097281831, + "grad_norm": 0.8900656279262298, + "learning_rate": 5.0678686739380455e-05, + "loss": 1.9115, + "step": 7541 + }, + { + "epoch": 0.6743562231759657, + "grad_norm": 0.5869762889069866, + "learning_rate": 5.065349616265329e-05, + "loss": 1.4671, + "step": 7542 + }, + { + "epoch": 0.6744456366237482, + "grad_norm": 0.9722915377432678, + "learning_rate": 5.062830972455781e-05, + "loss": 1.8133, + "step": 7543 + }, + { + "epoch": 0.6745350500715308, + "grad_norm": 8.845058065708235, + "learning_rate": 5.060312742720639e-05, + "loss": 6.3281, + "step": 7544 + }, + { + "epoch": 0.6746244635193133, + "grad_norm": 2.2200011960585764, + "learning_rate": 5.0577949272711e-05, + "loss": 1.821, + "step": 7545 + }, + { + "epoch": 0.6747138769670958, + "grad_norm": 3.2758232713542843, + "learning_rate": 5.0552775263183294e-05, + "loss": 1.7146, + "step": 7546 + }, + { + "epoch": 0.6748032904148784, + "grad_norm": 1.3871256067295792, + "learning_rate": 5.052760540073467e-05, + "loss": 1.3967, + "step": 7547 + }, + { + "epoch": 0.674892703862661, + "grad_norm": 0.7332153667160455, + "learning_rate": 5.050243968747599e-05, + "loss": 1.5016, + "step": 7548 + }, + { + "epoch": 0.6749821173104434, + "grad_norm": 1.3029388762553349, + "learning_rate": 5.047727812551786e-05, + "loss": 1.6224, + "step": 7549 + }, + { + "epoch": 0.675071530758226, + "grad_norm": 2.2200133587647053, + "learning_rate": 5.04521207169707e-05, + "loss": 1.7422, + "step": 7550 + }, + { + "epoch": 0.6751609442060086, + "grad_norm": 1.643530798822227, + "learning_rate": 5.0426967463944285e-05, + "loss": 1.9166, + "step": 7551 + }, + { + "epoch": 0.6752503576537912, + "grad_norm": 2.1084805824562594, + "learning_rate": 5.040181836854825e-05, + "loss": 1.8907, + "step": 7552 + }, + { + "epoch": 0.6753397711015737, + "grad_norm": 1.3334794747495802, + "learning_rate": 5.037667343289185e-05, + "loss": 1.4783, + "step": 7553 + }, + { + "epoch": 0.6754291845493562, + "grad_norm": 2.2437604508567612, + "learning_rate": 5.035153265908388e-05, + "loss": 1.7924, + "step": 7554 + }, + { + "epoch": 0.6755185979971388, + "grad_norm": 1.5835421875713773, + "learning_rate": 5.032639604923289e-05, + "loss": 1.491, + "step": 7555 + }, + { + "epoch": 0.6756080114449213, + "grad_norm": 2.5544334886827937, + "learning_rate": 5.0301263605447093e-05, + "loss": 1.8455, + "step": 7556 + }, + { + "epoch": 0.6756974248927039, + "grad_norm": 1.3852278236176783, + "learning_rate": 5.0276135329834284e-05, + "loss": 1.4987, + "step": 7557 + }, + { + "epoch": 0.6757868383404864, + "grad_norm": 1.1212061664385184, + "learning_rate": 5.0251011224502e-05, + "loss": 1.5108, + "step": 7558 + }, + { + "epoch": 0.6758762517882689, + "grad_norm": 2.225187015699295, + "learning_rate": 5.0225891291557284e-05, + "loss": 2.0985, + "step": 7559 + }, + { + "epoch": 0.6759656652360515, + "grad_norm": 1.3084346811606493, + "learning_rate": 5.020077553310694e-05, + "loss": 1.761, + "step": 7560 + }, + { + "epoch": 0.6760550786838341, + "grad_norm": 2.4746019117164155, + "learning_rate": 5.0175663951257424e-05, + "loss": 1.8769, + "step": 7561 + }, + { + "epoch": 0.6761444921316166, + "grad_norm": 0.7960136667231723, + "learning_rate": 5.015055654811484e-05, + "loss": 1.334, + "step": 7562 + }, + { + "epoch": 0.6762339055793991, + "grad_norm": 1.353561926852468, + "learning_rate": 5.012545332578479e-05, + "loss": 1.4703, + "step": 7563 + }, + { + "epoch": 0.6763233190271817, + "grad_norm": 2.0307682530012086, + "learning_rate": 5.0100354286372806e-05, + "loss": 1.6125, + "step": 7564 + }, + { + "epoch": 0.6764127324749643, + "grad_norm": 0.8869227865490169, + "learning_rate": 5.007525943198382e-05, + "loss": 1.801, + "step": 7565 + }, + { + "epoch": 0.6765021459227468, + "grad_norm": 2.872254635518897, + "learning_rate": 5.0050168764722524e-05, + "loss": 1.8733, + "step": 7566 + }, + { + "epoch": 0.6765915593705293, + "grad_norm": 1.0446477769739448, + "learning_rate": 5.002508228669329e-05, + "loss": 1.6707, + "step": 7567 + }, + { + "epoch": 0.6766809728183119, + "grad_norm": 1.0294327294567633, + "learning_rate": 5.000000000000002e-05, + "loss": 1.879, + "step": 7568 + }, + { + "epoch": 0.6767703862660944, + "grad_norm": 0.4362941781613886, + "learning_rate": 4.9974921906746363e-05, + "loss": 1.587, + "step": 7569 + }, + { + "epoch": 0.676859799713877, + "grad_norm": 0.7134138797525181, + "learning_rate": 4.9949848009035584e-05, + "loss": 1.8058, + "step": 7570 + }, + { + "epoch": 0.6769492131616596, + "grad_norm": 1.1788638633440949, + "learning_rate": 4.992477830897061e-05, + "loss": 1.8199, + "step": 7571 + }, + { + "epoch": 0.677038626609442, + "grad_norm": 1.2525966821776744, + "learning_rate": 4.989971280865401e-05, + "loss": 1.6036, + "step": 7572 + }, + { + "epoch": 0.6771280400572246, + "grad_norm": 0.8537739818278764, + "learning_rate": 4.987465151018802e-05, + "loss": 1.676, + "step": 7573 + }, + { + "epoch": 0.6772174535050072, + "grad_norm": 1.8850651905475855, + "learning_rate": 4.984959441567443e-05, + "loss": 1.7442, + "step": 7574 + }, + { + "epoch": 0.6773068669527897, + "grad_norm": 0.6689010182365052, + "learning_rate": 4.9824541527214797e-05, + "loss": 1.2668, + "step": 7575 + }, + { + "epoch": 0.6773962804005722, + "grad_norm": 1.4463392512868214, + "learning_rate": 4.979949284691031e-05, + "loss": 1.4622, + "step": 7576 + }, + { + "epoch": 0.6774856938483548, + "grad_norm": 1.204166264908556, + "learning_rate": 4.977444837686165e-05, + "loss": 1.5239, + "step": 7577 + }, + { + "epoch": 0.6775751072961373, + "grad_norm": 0.9268681842240435, + "learning_rate": 4.974940811916943e-05, + "loss": 1.8189, + "step": 7578 + }, + { + "epoch": 0.6776645207439199, + "grad_norm": 0.6087339532934813, + "learning_rate": 4.9724372075933615e-05, + "loss": 1.6658, + "step": 7579 + }, + { + "epoch": 0.6777539341917024, + "grad_norm": 1.7982001830954766, + "learning_rate": 4.9699340249254e-05, + "loss": 1.6604, + "step": 7580 + }, + { + "epoch": 0.677843347639485, + "grad_norm": 1.342384582230569, + "learning_rate": 4.9674312641230015e-05, + "loss": 1.6562, + "step": 7581 + }, + { + "epoch": 0.6779327610872675, + "grad_norm": 1.4444038306736018, + "learning_rate": 4.9649289253960606e-05, + "loss": 1.7661, + "step": 7582 + }, + { + "epoch": 0.6780221745350501, + "grad_norm": 1.2190695060237717, + "learning_rate": 4.9624270089544464e-05, + "loss": 2.1209, + "step": 7583 + }, + { + "epoch": 0.6781115879828327, + "grad_norm": 0.7446968072140729, + "learning_rate": 4.959925515008002e-05, + "loss": 1.4608, + "step": 7584 + }, + { + "epoch": 0.6782010014306151, + "grad_norm": 1.4228369632371778, + "learning_rate": 4.9574244437665154e-05, + "loss": 2.2541, + "step": 7585 + }, + { + "epoch": 0.6782904148783977, + "grad_norm": 0.5384692509370794, + "learning_rate": 4.9549237954397495e-05, + "loss": 1.6504, + "step": 7586 + }, + { + "epoch": 0.6783798283261803, + "grad_norm": 0.6735100626889152, + "learning_rate": 4.952423570237437e-05, + "loss": 1.604, + "step": 7587 + }, + { + "epoch": 0.6784692417739628, + "grad_norm": 1.2342711100462842, + "learning_rate": 4.949923768369259e-05, + "loss": 1.9595, + "step": 7588 + }, + { + "epoch": 0.6785586552217453, + "grad_norm": 0.7595151699345152, + "learning_rate": 4.9474243900448755e-05, + "loss": 1.6661, + "step": 7589 + }, + { + "epoch": 0.6786480686695279, + "grad_norm": 0.4886968070602839, + "learning_rate": 4.9449254354739074e-05, + "loss": 1.6717, + "step": 7590 + }, + { + "epoch": 0.6787374821173104, + "grad_norm": 0.7289407150168522, + "learning_rate": 4.9424269048659375e-05, + "loss": 1.1957, + "step": 7591 + }, + { + "epoch": 0.678826895565093, + "grad_norm": 0.869292473385238, + "learning_rate": 4.939928798430515e-05, + "loss": 1.8026, + "step": 7592 + }, + { + "epoch": 0.6789163090128756, + "grad_norm": 1.0908612531354551, + "learning_rate": 4.9374311163771567e-05, + "loss": 2.1074, + "step": 7593 + }, + { + "epoch": 0.679005722460658, + "grad_norm": 0.6056854215480206, + "learning_rate": 4.9349338589153335e-05, + "loss": 1.8088, + "step": 7594 + }, + { + "epoch": 0.6790951359084406, + "grad_norm": 1.0351324035987046, + "learning_rate": 4.9324370262544905e-05, + "loss": 1.5078, + "step": 7595 + }, + { + "epoch": 0.6791845493562232, + "grad_norm": 0.3703683542139931, + "learning_rate": 4.929940618604037e-05, + "loss": 1.3961, + "step": 7596 + }, + { + "epoch": 0.6792739628040058, + "grad_norm": 1.446663412360361, + "learning_rate": 4.927444636173334e-05, + "loss": 2.1144, + "step": 7597 + }, + { + "epoch": 0.6793633762517882, + "grad_norm": 0.585237479344024, + "learning_rate": 4.92494907917173e-05, + "loss": 1.403, + "step": 7598 + }, + { + "epoch": 0.6794527896995708, + "grad_norm": 6.028494877072345, + "learning_rate": 4.9224539478085144e-05, + "loss": 6.551, + "step": 7599 + }, + { + "epoch": 0.6795422031473534, + "grad_norm": 4.422016003401475, + "learning_rate": 4.919959242292954e-05, + "loss": 1.9258, + "step": 7600 + }, + { + "epoch": 0.6796316165951359, + "grad_norm": 1.350752027563287, + "learning_rate": 4.9174649628342805e-05, + "loss": 6.2966, + "step": 7601 + }, + { + "epoch": 0.6797210300429185, + "grad_norm": 3.6958382037226616, + "learning_rate": 4.914971109641678e-05, + "loss": 1.7924, + "step": 7602 + }, + { + "epoch": 0.679810443490701, + "grad_norm": 1.1008892575269005, + "learning_rate": 4.912477682924309e-05, + "loss": 1.4339, + "step": 7603 + }, + { + "epoch": 0.6798998569384835, + "grad_norm": 1.9434565172799365, + "learning_rate": 4.909984682891291e-05, + "loss": 1.8304, + "step": 7604 + }, + { + "epoch": 0.6799892703862661, + "grad_norm": 4.402411683082708, + "learning_rate": 4.907492109751711e-05, + "loss": 1.7492, + "step": 7605 + }, + { + "epoch": 0.6800786838340487, + "grad_norm": 1.6911968748824755, + "learning_rate": 4.904999963714618e-05, + "loss": 1.7049, + "step": 7606 + }, + { + "epoch": 0.6801680972818311, + "grad_norm": 0.7478835210494721, + "learning_rate": 4.902508244989028e-05, + "loss": 1.8495, + "step": 7607 + }, + { + "epoch": 0.6802575107296137, + "grad_norm": 2.878361968136322, + "learning_rate": 4.900016953783912e-05, + "loss": 1.7875, + "step": 7608 + }, + { + "epoch": 0.6803469241773963, + "grad_norm": 3.7943873704870064, + "learning_rate": 4.8975260903082157e-05, + "loss": 1.713, + "step": 7609 + }, + { + "epoch": 0.6804363376251789, + "grad_norm": 0.8152823601264912, + "learning_rate": 4.895035654770846e-05, + "loss": 1.9051, + "step": 7610 + }, + { + "epoch": 0.6805257510729614, + "grad_norm": 1.1494914966937453, + "learning_rate": 4.892545647380664e-05, + "loss": 1.6263, + "step": 7611 + }, + { + "epoch": 0.6806151645207439, + "grad_norm": 4.622888146739374, + "learning_rate": 4.890056068346518e-05, + "loss": 1.9265, + "step": 7612 + }, + { + "epoch": 0.6807045779685265, + "grad_norm": 0.7940229562725221, + "learning_rate": 4.887566917877194e-05, + "loss": 1.6164, + "step": 7613 + }, + { + "epoch": 0.680793991416309, + "grad_norm": 0.8217055826086009, + "learning_rate": 4.885078196181458e-05, + "loss": 1.8777, + "step": 7614 + }, + { + "epoch": 0.6808834048640916, + "grad_norm": 0.9683217108514913, + "learning_rate": 4.882589903468041e-05, + "loss": 1.9341, + "step": 7615 + }, + { + "epoch": 0.6809728183118741, + "grad_norm": 2.1296041984186833, + "learning_rate": 4.880102039945624e-05, + "loss": 2.02, + "step": 7616 + }, + { + "epoch": 0.6810622317596566, + "grad_norm": 0.703291801150973, + "learning_rate": 4.8776146058228665e-05, + "loss": 1.4011, + "step": 7617 + }, + { + "epoch": 0.6811516452074392, + "grad_norm": 0.6406882054464713, + "learning_rate": 4.875127601308386e-05, + "loss": 1.7826, + "step": 7618 + }, + { + "epoch": 0.6812410586552218, + "grad_norm": 0.6487570459943234, + "learning_rate": 4.8726410266107634e-05, + "loss": 1.6201, + "step": 7619 + }, + { + "epoch": 0.6813304721030042, + "grad_norm": 1.4265624492884645, + "learning_rate": 4.870154881938546e-05, + "loss": 1.5618, + "step": 7620 + }, + { + "epoch": 0.6814198855507868, + "grad_norm": 1.3256106364097302, + "learning_rate": 4.867669167500247e-05, + "loss": 1.8186, + "step": 7621 + }, + { + "epoch": 0.6815092989985694, + "grad_norm": 0.7841991468555048, + "learning_rate": 4.865183883504333e-05, + "loss": 1.6842, + "step": 7622 + }, + { + "epoch": 0.681598712446352, + "grad_norm": 0.6872275166254608, + "learning_rate": 4.862699030159246e-05, + "loss": 2.0538, + "step": 7623 + }, + { + "epoch": 0.6816881258941345, + "grad_norm": 1.4900002346396268, + "learning_rate": 4.86021460767339e-05, + "loss": 1.9073, + "step": 7624 + }, + { + "epoch": 0.681777539341917, + "grad_norm": 1.0267980569783253, + "learning_rate": 4.8577306162551196e-05, + "loss": 1.6664, + "step": 7625 + }, + { + "epoch": 0.6818669527896996, + "grad_norm": 0.5474582549755691, + "learning_rate": 4.8552470561127775e-05, + "loss": 1.5443, + "step": 7626 + }, + { + "epoch": 0.6819563662374821, + "grad_norm": 0.6966290130916118, + "learning_rate": 4.852763927454653e-05, + "loss": 1.6337, + "step": 7627 + }, + { + "epoch": 0.6820457796852647, + "grad_norm": 1.1921075115740525, + "learning_rate": 4.850281230489e-05, + "loss": 1.6547, + "step": 7628 + }, + { + "epoch": 0.6821351931330472, + "grad_norm": 0.582496920505101, + "learning_rate": 4.84779896542404e-05, + "loss": 1.4396, + "step": 7629 + }, + { + "epoch": 0.6822246065808297, + "grad_norm": 0.8204635606375176, + "learning_rate": 4.845317132467963e-05, + "loss": 1.7154, + "step": 7630 + }, + { + "epoch": 0.6823140200286123, + "grad_norm": 0.5192274642039514, + "learning_rate": 4.842835731828908e-05, + "loss": 1.575, + "step": 7631 + }, + { + "epoch": 0.6824034334763949, + "grad_norm": 0.5179586544420217, + "learning_rate": 4.840354763714991e-05, + "loss": 1.6398, + "step": 7632 + }, + { + "epoch": 0.6824928469241774, + "grad_norm": 0.6608691952202175, + "learning_rate": 4.83787422833429e-05, + "loss": 1.8742, + "step": 7633 + }, + { + "epoch": 0.6825822603719599, + "grad_norm": 1.3710404098367235, + "learning_rate": 4.835394125894843e-05, + "loss": 1.5135, + "step": 7634 + }, + { + "epoch": 0.6826716738197425, + "grad_norm": 0.7604899766643368, + "learning_rate": 4.832914456604658e-05, + "loss": 1.7924, + "step": 7635 + }, + { + "epoch": 0.682761087267525, + "grad_norm": 0.7568194491151492, + "learning_rate": 4.830435220671693e-05, + "loss": 1.5716, + "step": 7636 + }, + { + "epoch": 0.6828505007153076, + "grad_norm": 0.6872395264522388, + "learning_rate": 4.8279564183038825e-05, + "loss": 1.6386, + "step": 7637 + }, + { + "epoch": 0.6829399141630901, + "grad_norm": 0.6471046709788715, + "learning_rate": 4.825478049709124e-05, + "loss": 1.9854, + "step": 7638 + }, + { + "epoch": 0.6830293276108726, + "grad_norm": 0.9176671212181358, + "learning_rate": 4.823000115095266e-05, + "loss": 1.3908, + "step": 7639 + }, + { + "epoch": 0.6831187410586552, + "grad_norm": 1.2028991101520583, + "learning_rate": 4.82052261467014e-05, + "loss": 1.8247, + "step": 7640 + }, + { + "epoch": 0.6832081545064378, + "grad_norm": 0.8153695738883215, + "learning_rate": 4.81804554864153e-05, + "loss": 1.765, + "step": 7641 + }, + { + "epoch": 0.6832975679542204, + "grad_norm": 0.7622282089089027, + "learning_rate": 4.815568917217178e-05, + "loss": 1.3811, + "step": 7642 + }, + { + "epoch": 0.6833869814020028, + "grad_norm": 0.707104693013162, + "learning_rate": 4.813092720604799e-05, + "loss": 1.6656, + "step": 7643 + }, + { + "epoch": 0.6834763948497854, + "grad_norm": 4.86028975458027, + "learning_rate": 4.8106169590120745e-05, + "loss": 6.5024, + "step": 7644 + }, + { + "epoch": 0.683565808297568, + "grad_norm": 1.5146755671861603, + "learning_rate": 4.8081416326466346e-05, + "loss": 2.0717, + "step": 7645 + }, + { + "epoch": 0.6836552217453505, + "grad_norm": 1.9534335065407828, + "learning_rate": 4.805666741716085e-05, + "loss": 1.9331, + "step": 7646 + }, + { + "epoch": 0.683744635193133, + "grad_norm": 1.9069766115242435, + "learning_rate": 4.8031922864279924e-05, + "loss": 1.8931, + "step": 7647 + }, + { + "epoch": 0.6838340486409156, + "grad_norm": 0.9625193021305177, + "learning_rate": 4.800718266989888e-05, + "loss": 1.7157, + "step": 7648 + }, + { + "epoch": 0.6839234620886981, + "grad_norm": 0.7838248713455553, + "learning_rate": 4.798244683609262e-05, + "loss": 1.788, + "step": 7649 + }, + { + "epoch": 0.6840128755364807, + "grad_norm": 1.138462604088308, + "learning_rate": 4.795771536493576e-05, + "loss": 1.8976, + "step": 7650 + }, + { + "epoch": 0.6841022889842633, + "grad_norm": 1.1085901960994229, + "learning_rate": 4.793298825850243e-05, + "loss": 1.7941, + "step": 7651 + }, + { + "epoch": 0.6841917024320457, + "grad_norm": 1.1862755914757932, + "learning_rate": 4.790826551886649e-05, + "loss": 1.6806, + "step": 7652 + }, + { + "epoch": 0.6842811158798283, + "grad_norm": 0.8489026401985122, + "learning_rate": 4.788354714810141e-05, + "loss": 1.6112, + "step": 7653 + }, + { + "epoch": 0.6843705293276109, + "grad_norm": 1.2259689670406166, + "learning_rate": 4.7858833148280294e-05, + "loss": 1.8913, + "step": 7654 + }, + { + "epoch": 0.6844599427753935, + "grad_norm": 0.6270071889263759, + "learning_rate": 4.78341235214759e-05, + "loss": 1.6426, + "step": 7655 + }, + { + "epoch": 0.6845493562231759, + "grad_norm": 1.2635904037320655, + "learning_rate": 4.7809418269760545e-05, + "loss": 1.542, + "step": 7656 + }, + { + "epoch": 0.6846387696709585, + "grad_norm": 0.7495981737858456, + "learning_rate": 4.778471739520624e-05, + "loss": 1.7504, + "step": 7657 + }, + { + "epoch": 0.6847281831187411, + "grad_norm": 1.2901426416918576, + "learning_rate": 4.7760020899884664e-05, + "loss": 1.7781, + "step": 7658 + }, + { + "epoch": 0.6848175965665236, + "grad_norm": 0.778213027655443, + "learning_rate": 4.7735328785867004e-05, + "loss": 1.6351, + "step": 7659 + }, + { + "epoch": 0.6849070100143062, + "grad_norm": 1.8527184504069674, + "learning_rate": 4.771064105522417e-05, + "loss": 1.7887, + "step": 7660 + }, + { + "epoch": 0.6849964234620887, + "grad_norm": 1.9049411303232986, + "learning_rate": 4.7685957710026784e-05, + "loss": 1.6485, + "step": 7661 + }, + { + "epoch": 0.6850858369098712, + "grad_norm": 1.024535057850674, + "learning_rate": 4.766127875234492e-05, + "loss": 1.5981, + "step": 7662 + }, + { + "epoch": 0.6851752503576538, + "grad_norm": 0.8350317615681948, + "learning_rate": 4.763660418424839e-05, + "loss": 1.7625, + "step": 7663 + }, + { + "epoch": 0.6852646638054364, + "grad_norm": 1.8351107359343286, + "learning_rate": 4.7611934007806666e-05, + "loss": 2.1284, + "step": 7664 + }, + { + "epoch": 0.6853540772532188, + "grad_norm": 0.59883920808454, + "learning_rate": 4.758726822508874e-05, + "loss": 2.0813, + "step": 7665 + }, + { + "epoch": 0.6854434907010014, + "grad_norm": 0.7481060438082022, + "learning_rate": 4.756260683816333e-05, + "loss": 1.482, + "step": 7666 + }, + { + "epoch": 0.685532904148784, + "grad_norm": 1.1595465476648306, + "learning_rate": 4.753794984909874e-05, + "loss": 1.7473, + "step": 7667 + }, + { + "epoch": 0.6856223175965666, + "grad_norm": 0.7197106850146694, + "learning_rate": 4.751329725996295e-05, + "loss": 1.5158, + "step": 7668 + }, + { + "epoch": 0.685711731044349, + "grad_norm": 0.8441822669142743, + "learning_rate": 4.748864907282357e-05, + "loss": 1.4561, + "step": 7669 + }, + { + "epoch": 0.6858011444921316, + "grad_norm": 0.6433295183677517, + "learning_rate": 4.746400528974772e-05, + "loss": 1.905, + "step": 7670 + }, + { + "epoch": 0.6858905579399142, + "grad_norm": 0.6473532361608192, + "learning_rate": 4.7439365912802314e-05, + "loss": 1.6534, + "step": 7671 + }, + { + "epoch": 0.6859799713876967, + "grad_norm": 0.8088009182022915, + "learning_rate": 4.741473094405386e-05, + "loss": 1.6215, + "step": 7672 + }, + { + "epoch": 0.6860693848354793, + "grad_norm": 0.6282016565775067, + "learning_rate": 4.739010038556831e-05, + "loss": 1.5517, + "step": 7673 + }, + { + "epoch": 0.6861587982832618, + "grad_norm": 2.017400329779044, + "learning_rate": 4.736547423941157e-05, + "loss": 1.9095, + "step": 7674 + }, + { + "epoch": 0.6862482117310443, + "grad_norm": 0.4682607252548282, + "learning_rate": 4.734085250764896e-05, + "loss": 1.984, + "step": 7675 + }, + { + "epoch": 0.6863376251788269, + "grad_norm": 0.6937765664168475, + "learning_rate": 4.7316235192345416e-05, + "loss": 1.1502, + "step": 7676 + }, + { + "epoch": 0.6864270386266095, + "grad_norm": 1.585559263293998, + "learning_rate": 4.729162229556561e-05, + "loss": 1.7893, + "step": 7677 + }, + { + "epoch": 0.6865164520743919, + "grad_norm": 7.909410143344542, + "learning_rate": 4.726701381937382e-05, + "loss": 6.5626, + "step": 7678 + }, + { + "epoch": 0.6866058655221745, + "grad_norm": 1.6496569176330877, + "learning_rate": 4.724240976583386e-05, + "loss": 1.7291, + "step": 7679 + }, + { + "epoch": 0.6866952789699571, + "grad_norm": 3.0627675494432354, + "learning_rate": 4.7217810137009274e-05, + "loss": 1.7055, + "step": 7680 + }, + { + "epoch": 0.6867846924177397, + "grad_norm": 2.90916697543523, + "learning_rate": 4.7193214934963206e-05, + "loss": 1.9512, + "step": 7681 + }, + { + "epoch": 0.6868741058655222, + "grad_norm": 1.1815230113713486, + "learning_rate": 4.716862416175844e-05, + "loss": 1.7535, + "step": 7682 + }, + { + "epoch": 0.6869635193133047, + "grad_norm": 1.7797186605324264, + "learning_rate": 4.7144037819457345e-05, + "loss": 1.7766, + "step": 7683 + }, + { + "epoch": 0.6870529327610873, + "grad_norm": 1.8848175136582728, + "learning_rate": 4.7119455910122e-05, + "loss": 2.0113, + "step": 7684 + }, + { + "epoch": 0.6871423462088698, + "grad_norm": 1.4515708471099702, + "learning_rate": 4.709487843581399e-05, + "loss": 1.6666, + "step": 7685 + }, + { + "epoch": 0.6872317596566524, + "grad_norm": 0.7849280643051834, + "learning_rate": 4.707030539859465e-05, + "loss": 1.506, + "step": 7686 + }, + { + "epoch": 0.6873211731044349, + "grad_norm": 1.4174413409018947, + "learning_rate": 4.7045736800524856e-05, + "loss": 1.7161, + "step": 7687 + }, + { + "epoch": 0.6874105865522174, + "grad_norm": 2.20896054704636, + "learning_rate": 4.702117264366517e-05, + "loss": 1.3375, + "step": 7688 + }, + { + "epoch": 0.6875, + "grad_norm": 1.8540542700012872, + "learning_rate": 4.699661293007579e-05, + "loss": 1.7157, + "step": 7689 + }, + { + "epoch": 0.6875894134477826, + "grad_norm": 0.7829974122012613, + "learning_rate": 4.6972057661816426e-05, + "loss": 2.0249, + "step": 7690 + }, + { + "epoch": 0.6876788268955651, + "grad_norm": 1.0384025258145546, + "learning_rate": 4.6947506840946555e-05, + "loss": 1.3617, + "step": 7691 + }, + { + "epoch": 0.6877682403433476, + "grad_norm": 1.0469445095140204, + "learning_rate": 4.6922960469525245e-05, + "loss": 1.5127, + "step": 7692 + }, + { + "epoch": 0.6878576537911302, + "grad_norm": 1.5397072078220477, + "learning_rate": 4.68984185496111e-05, + "loss": 1.7517, + "step": 7693 + }, + { + "epoch": 0.6879470672389127, + "grad_norm": 1.4187541361230331, + "learning_rate": 4.687388108326243e-05, + "loss": 1.6814, + "step": 7694 + }, + { + "epoch": 0.6880364806866953, + "grad_norm": 0.528714553427588, + "learning_rate": 4.684934807253727e-05, + "loss": 1.5229, + "step": 7695 + }, + { + "epoch": 0.6881258941344778, + "grad_norm": 1.6587920159209344, + "learning_rate": 4.6824819519493057e-05, + "loss": 2.1454, + "step": 7696 + }, + { + "epoch": 0.6882153075822603, + "grad_norm": 0.9603284166843957, + "learning_rate": 4.6800295426187e-05, + "loss": 1.7163, + "step": 7697 + }, + { + "epoch": 0.6883047210300429, + "grad_norm": 0.9684317361444154, + "learning_rate": 4.677577579467597e-05, + "loss": 1.9056, + "step": 7698 + }, + { + "epoch": 0.6883941344778255, + "grad_norm": 0.8596438730535886, + "learning_rate": 4.67512606270163e-05, + "loss": 1.8544, + "step": 7699 + }, + { + "epoch": 0.6884835479256081, + "grad_norm": 0.5804324781087034, + "learning_rate": 4.67267499252641e-05, + "loss": 1.6298, + "step": 7700 + }, + { + "epoch": 0.6885729613733905, + "grad_norm": 1.6600239870899056, + "learning_rate": 4.670224369147505e-05, + "loss": 2.0741, + "step": 7701 + }, + { + "epoch": 0.6886623748211731, + "grad_norm": 0.4081385895438012, + "learning_rate": 4.6677741927704434e-05, + "loss": 1.4925, + "step": 7702 + }, + { + "epoch": 0.6887517882689557, + "grad_norm": 0.9533345949042878, + "learning_rate": 4.6653244636007255e-05, + "loss": 1.9368, + "step": 7703 + }, + { + "epoch": 0.6888412017167382, + "grad_norm": 0.6030787390708312, + "learning_rate": 4.6628751818437985e-05, + "loss": 1.6627, + "step": 7704 + }, + { + "epoch": 0.6889306151645207, + "grad_norm": 1.1635228951877004, + "learning_rate": 4.660426347705085e-05, + "loss": 1.6267, + "step": 7705 + }, + { + "epoch": 0.6890200286123033, + "grad_norm": 0.642152721012575, + "learning_rate": 4.6579779613899644e-05, + "loss": 2.0636, + "step": 7706 + }, + { + "epoch": 0.6891094420600858, + "grad_norm": 0.7131539083423083, + "learning_rate": 4.6555300231037836e-05, + "loss": 1.5891, + "step": 7707 + }, + { + "epoch": 0.6891988555078684, + "grad_norm": 0.9675801889936652, + "learning_rate": 4.653082533051839e-05, + "loss": 1.8589, + "step": 7708 + }, + { + "epoch": 0.689288268955651, + "grad_norm": 0.5530952436855328, + "learning_rate": 4.650635491439412e-05, + "loss": 1.5252, + "step": 7709 + }, + { + "epoch": 0.6893776824034334, + "grad_norm": 0.6082017191476382, + "learning_rate": 4.6481888984717225e-05, + "loss": 2.0143, + "step": 7710 + }, + { + "epoch": 0.689467095851216, + "grad_norm": 0.54690782867793, + "learning_rate": 4.6457427543539654e-05, + "loss": 1.7462, + "step": 7711 + }, + { + "epoch": 0.6895565092989986, + "grad_norm": 0.5508726001053647, + "learning_rate": 4.6432970592913026e-05, + "loss": 1.9425, + "step": 7712 + }, + { + "epoch": 0.6896459227467812, + "grad_norm": 0.7152863177327438, + "learning_rate": 4.640851813488842e-05, + "loss": 1.4318, + "step": 7713 + }, + { + "epoch": 0.6897353361945636, + "grad_norm": 0.4982010853102423, + "learning_rate": 4.638407017151667e-05, + "loss": 1.6901, + "step": 7714 + }, + { + "epoch": 0.6898247496423462, + "grad_norm": 0.5404872493196211, + "learning_rate": 4.6359626704848215e-05, + "loss": 1.7021, + "step": 7715 + }, + { + "epoch": 0.6899141630901288, + "grad_norm": 0.6675066217703668, + "learning_rate": 4.633518773693307e-05, + "loss": 1.4791, + "step": 7716 + }, + { + "epoch": 0.6900035765379113, + "grad_norm": 0.4815119971743937, + "learning_rate": 4.631075326982093e-05, + "loss": 2.1097, + "step": 7717 + }, + { + "epoch": 0.6900929899856938, + "grad_norm": 1.0565771514266133, + "learning_rate": 4.6286323305561105e-05, + "loss": 1.868, + "step": 7718 + }, + { + "epoch": 0.6901824034334764, + "grad_norm": 0.5846971781343848, + "learning_rate": 4.626189784620245e-05, + "loss": 1.8875, + "step": 7719 + }, + { + "epoch": 0.6902718168812589, + "grad_norm": 0.6593557353780936, + "learning_rate": 4.623747689379351e-05, + "loss": 1.6729, + "step": 7720 + }, + { + "epoch": 0.6903612303290415, + "grad_norm": 0.9641922243883083, + "learning_rate": 4.621306045038249e-05, + "loss": 1.9374, + "step": 7721 + }, + { + "epoch": 0.6904506437768241, + "grad_norm": 0.41136130653091496, + "learning_rate": 4.618864851801707e-05, + "loss": 1.7392, + "step": 7722 + }, + { + "epoch": 0.6905400572246065, + "grad_norm": 0.6413526338729212, + "learning_rate": 4.6164241098744776e-05, + "loss": 1.7914, + "step": 7723 + }, + { + "epoch": 0.6906294706723891, + "grad_norm": 0.5452942719955257, + "learning_rate": 4.613983819461253e-05, + "loss": 1.973, + "step": 7724 + }, + { + "epoch": 0.6907188841201717, + "grad_norm": 0.5752759344543092, + "learning_rate": 4.6115439807667005e-05, + "loss": 2.0727, + "step": 7725 + }, + { + "epoch": 0.6908082975679543, + "grad_norm": 0.4912270423247292, + "learning_rate": 4.6091045939954514e-05, + "loss": 1.71, + "step": 7726 + }, + { + "epoch": 0.6908977110157367, + "grad_norm": 0.4396495831112731, + "learning_rate": 4.606665659352085e-05, + "loss": 1.5234, + "step": 7727 + }, + { + "epoch": 0.6909871244635193, + "grad_norm": 0.49398251121952547, + "learning_rate": 4.604227177041156e-05, + "loss": 1.6951, + "step": 7728 + }, + { + "epoch": 0.6910765379113019, + "grad_norm": 0.5671939796776282, + "learning_rate": 4.601789147267177e-05, + "loss": 1.7238, + "step": 7729 + }, + { + "epoch": 0.6911659513590844, + "grad_norm": 0.5726188572207053, + "learning_rate": 4.5993515702346235e-05, + "loss": 1.9516, + "step": 7730 + }, + { + "epoch": 0.691255364806867, + "grad_norm": 0.48297584449403314, + "learning_rate": 4.596914446147932e-05, + "loss": 1.8264, + "step": 7731 + }, + { + "epoch": 0.6913447782546495, + "grad_norm": 10.314526216160617, + "learning_rate": 4.594477775211503e-05, + "loss": 6.5626, + "step": 7732 + }, + { + "epoch": 0.691434191702432, + "grad_norm": 3.3810348574134084, + "learning_rate": 4.5920415576296914e-05, + "loss": 2.1042, + "step": 7733 + }, + { + "epoch": 0.6915236051502146, + "grad_norm": 3.102397656249422, + "learning_rate": 4.589605793606824e-05, + "loss": 1.7492, + "step": 7734 + }, + { + "epoch": 0.6916130185979972, + "grad_norm": 2.227068016779591, + "learning_rate": 4.5871704833471876e-05, + "loss": 1.7311, + "step": 7735 + }, + { + "epoch": 0.6917024320457796, + "grad_norm": 0.8004048542363031, + "learning_rate": 4.584735627055019e-05, + "loss": 1.3486, + "step": 7736 + }, + { + "epoch": 0.6917918454935622, + "grad_norm": 2.2512015356416453, + "learning_rate": 4.5823012249345396e-05, + "loss": 1.8953, + "step": 7737 + }, + { + "epoch": 0.6918812589413448, + "grad_norm": 2.4944236043558314, + "learning_rate": 4.579867277189911e-05, + "loss": 1.8376, + "step": 7738 + }, + { + "epoch": 0.6919706723891274, + "grad_norm": 1.2392854752984765, + "learning_rate": 4.5774337840252666e-05, + "loss": 2.072, + "step": 7739 + }, + { + "epoch": 0.6920600858369099, + "grad_norm": 0.9068033960233782, + "learning_rate": 4.575000745644703e-05, + "loss": 1.7857, + "step": 7740 + }, + { + "epoch": 0.6921494992846924, + "grad_norm": 2.768758112865511, + "learning_rate": 4.5725681622522795e-05, + "loss": 1.8256, + "step": 7741 + }, + { + "epoch": 0.692238912732475, + "grad_norm": 2.237001675352032, + "learning_rate": 4.570136034052005e-05, + "loss": 1.4102, + "step": 7742 + }, + { + "epoch": 0.6923283261802575, + "grad_norm": 2.1729648581928296, + "learning_rate": 4.567704361247863e-05, + "loss": 2.1227, + "step": 7743 + }, + { + "epoch": 0.6924177396280401, + "grad_norm": 0.7095969335681578, + "learning_rate": 4.5652731440437965e-05, + "loss": 1.6003, + "step": 7744 + }, + { + "epoch": 0.6925071530758226, + "grad_norm": 1.7696178193281416, + "learning_rate": 4.5628423826437085e-05, + "loss": 1.6867, + "step": 7745 + }, + { + "epoch": 0.6925965665236051, + "grad_norm": 1.6618691783330954, + "learning_rate": 4.5604120772514655e-05, + "loss": 1.7427, + "step": 7746 + }, + { + "epoch": 0.6926859799713877, + "grad_norm": 0.9913182093916527, + "learning_rate": 4.557982228070891e-05, + "loss": 2.0961, + "step": 7747 + }, + { + "epoch": 0.6927753934191703, + "grad_norm": 0.7808456809483901, + "learning_rate": 4.5555528353057716e-05, + "loss": 2.2063, + "step": 7748 + }, + { + "epoch": 0.6928648068669528, + "grad_norm": 12.526846269977694, + "learning_rate": 4.553123899159867e-05, + "loss": 6.5472, + "step": 7749 + }, + { + "epoch": 0.6929542203147353, + "grad_norm": 2.55354035125348, + "learning_rate": 4.5506954198368744e-05, + "loss": 1.8333, + "step": 7750 + }, + { + "epoch": 0.6930436337625179, + "grad_norm": 2.825688696559299, + "learning_rate": 4.54826739754048e-05, + "loss": 1.6021, + "step": 7751 + }, + { + "epoch": 0.6931330472103004, + "grad_norm": 2.5091324585800807, + "learning_rate": 4.545839832474318e-05, + "loss": 1.7956, + "step": 7752 + }, + { + "epoch": 0.693222460658083, + "grad_norm": 2.089504912514312, + "learning_rate": 4.543412724841979e-05, + "loss": 2.0501, + "step": 7753 + }, + { + "epoch": 0.6933118741058655, + "grad_norm": 1.90688641514683, + "learning_rate": 4.5409860748470246e-05, + "loss": 1.4832, + "step": 7754 + }, + { + "epoch": 0.693401287553648, + "grad_norm": 4.666325956842294, + "learning_rate": 4.538559882692979e-05, + "loss": 1.8474, + "step": 7755 + }, + { + "epoch": 0.6934907010014306, + "grad_norm": 1.1603886858295809, + "learning_rate": 4.536134148583313e-05, + "loss": 1.7013, + "step": 7756 + }, + { + "epoch": 0.6935801144492132, + "grad_norm": 0.7498208957157405, + "learning_rate": 4.5337088727214835e-05, + "loss": 1.7251, + "step": 7757 + }, + { + "epoch": 0.6936695278969958, + "grad_norm": 3.5045951337131642, + "learning_rate": 4.531284055310887e-05, + "loss": 1.8118, + "step": 7758 + }, + { + "epoch": 0.6937589413447782, + "grad_norm": 2.1994212928641836, + "learning_rate": 4.5288596965548924e-05, + "loss": 1.7635, + "step": 7759 + }, + { + "epoch": 0.6938483547925608, + "grad_norm": 0.6043919231789977, + "learning_rate": 4.5264357966568306e-05, + "loss": 1.4932, + "step": 7760 + }, + { + "epoch": 0.6939377682403434, + "grad_norm": 1.4651430794357707, + "learning_rate": 4.5240123558199846e-05, + "loss": 1.9034, + "step": 7761 + }, + { + "epoch": 0.6940271816881259, + "grad_norm": 2.403513986180296, + "learning_rate": 4.521589374247609e-05, + "loss": 1.7678, + "step": 7762 + }, + { + "epoch": 0.6941165951359084, + "grad_norm": 0.8983651571343826, + "learning_rate": 4.519166852142917e-05, + "loss": 2.0456, + "step": 7763 + }, + { + "epoch": 0.694206008583691, + "grad_norm": 0.9934797732936325, + "learning_rate": 4.516744789709081e-05, + "loss": 1.4676, + "step": 7764 + }, + { + "epoch": 0.6942954220314735, + "grad_norm": 1.3259373601919762, + "learning_rate": 4.5143231871492375e-05, + "loss": 1.7625, + "step": 7765 + }, + { + "epoch": 0.6943848354792561, + "grad_norm": 0.8450808995221094, + "learning_rate": 4.5119020446664875e-05, + "loss": 1.7092, + "step": 7766 + }, + { + "epoch": 0.6944742489270386, + "grad_norm": 0.9392293635597222, + "learning_rate": 4.509481362463881e-05, + "loss": 1.751, + "step": 7767 + }, + { + "epoch": 0.6945636623748211, + "grad_norm": 0.8068049187421982, + "learning_rate": 4.507061140744442e-05, + "loss": 1.6074, + "step": 7768 + }, + { + "epoch": 0.6946530758226037, + "grad_norm": 0.8019741117900738, + "learning_rate": 4.504641379711154e-05, + "loss": 1.8103, + "step": 7769 + }, + { + "epoch": 0.6947424892703863, + "grad_norm": 0.592419471939661, + "learning_rate": 4.502222079566951e-05, + "loss": 1.4501, + "step": 7770 + }, + { + "epoch": 0.6948319027181689, + "grad_norm": 1.0035649857129172, + "learning_rate": 4.499803240514745e-05, + "loss": 1.7844, + "step": 7771 + }, + { + "epoch": 0.6949213161659513, + "grad_norm": 0.7952831169803376, + "learning_rate": 4.497384862757403e-05, + "loss": 1.6913, + "step": 7772 + }, + { + "epoch": 0.6950107296137339, + "grad_norm": 0.40353666898098717, + "learning_rate": 4.494966946497743e-05, + "loss": 1.6172, + "step": 7773 + }, + { + "epoch": 0.6951001430615165, + "grad_norm": 0.5649931398222562, + "learning_rate": 4.492549491938557e-05, + "loss": 1.6119, + "step": 7774 + }, + { + "epoch": 0.695189556509299, + "grad_norm": 0.9455181167902624, + "learning_rate": 4.4901324992825975e-05, + "loss": 2.0038, + "step": 7775 + }, + { + "epoch": 0.6952789699570815, + "grad_norm": 0.9285652167544687, + "learning_rate": 4.487715968732568e-05, + "loss": 1.8058, + "step": 7776 + }, + { + "epoch": 0.6953683834048641, + "grad_norm": 0.393979076166626, + "learning_rate": 4.4852999004911425e-05, + "loss": 1.5753, + "step": 7777 + }, + { + "epoch": 0.6954577968526466, + "grad_norm": 1.371017498134071, + "learning_rate": 4.482884294760954e-05, + "loss": 2.0422, + "step": 7778 + }, + { + "epoch": 0.6955472103004292, + "grad_norm": 0.6289740858877009, + "learning_rate": 4.480469151744596e-05, + "loss": 1.4573, + "step": 7779 + }, + { + "epoch": 0.6956366237482118, + "grad_norm": 0.5459833311293134, + "learning_rate": 4.4780544716446294e-05, + "loss": 1.5785, + "step": 7780 + }, + { + "epoch": 0.6957260371959942, + "grad_norm": 0.8960627105998531, + "learning_rate": 4.475640254663561e-05, + "loss": 1.5795, + "step": 7781 + }, + { + "epoch": 0.6958154506437768, + "grad_norm": 0.7477095233289742, + "learning_rate": 4.473226501003873e-05, + "loss": 1.3408, + "step": 7782 + }, + { + "epoch": 0.6959048640915594, + "grad_norm": 1.4967901430151747, + "learning_rate": 4.470813210868008e-05, + "loss": 1.5754, + "step": 7783 + }, + { + "epoch": 0.695994277539342, + "grad_norm": 6.167730196025352, + "learning_rate": 4.4684003844583534e-05, + "loss": 6.4558, + "step": 7784 + }, + { + "epoch": 0.6960836909871244, + "grad_norm": 1.0206762959951035, + "learning_rate": 4.465988021977282e-05, + "loss": 1.4716, + "step": 7785 + }, + { + "epoch": 0.696173104434907, + "grad_norm": 2.305892393230462, + "learning_rate": 4.4635761236271144e-05, + "loss": 1.9267, + "step": 7786 + }, + { + "epoch": 0.6962625178826896, + "grad_norm": 1.6121212645809453, + "learning_rate": 4.461164689610129e-05, + "loss": 1.7323, + "step": 7787 + }, + { + "epoch": 0.6963519313304721, + "grad_norm": 0.5913136547401758, + "learning_rate": 4.458753720128571e-05, + "loss": 1.4194, + "step": 7788 + }, + { + "epoch": 0.6964413447782547, + "grad_norm": 0.8171565458750942, + "learning_rate": 4.4563432153846494e-05, + "loss": 1.9141, + "step": 7789 + }, + { + "epoch": 0.6965307582260372, + "grad_norm": 1.5268408542037826, + "learning_rate": 4.453933175580525e-05, + "loss": 1.7832, + "step": 7790 + }, + { + "epoch": 0.6966201716738197, + "grad_norm": 1.353826552333259, + "learning_rate": 4.451523600918327e-05, + "loss": 2.1956, + "step": 7791 + }, + { + "epoch": 0.6967095851216023, + "grad_norm": 1.440811513185375, + "learning_rate": 4.4491144916001425e-05, + "loss": 2.0643, + "step": 7792 + }, + { + "epoch": 0.6967989985693849, + "grad_norm": 0.750446778946379, + "learning_rate": 4.4467058478280235e-05, + "loss": 1.8904, + "step": 7793 + }, + { + "epoch": 0.6968884120171673, + "grad_norm": 1.608365634765629, + "learning_rate": 4.444297669803981e-05, + "loss": 1.927, + "step": 7794 + }, + { + "epoch": 0.6969778254649499, + "grad_norm": 0.8866719867281987, + "learning_rate": 4.441889957729979e-05, + "loss": 1.659, + "step": 7795 + }, + { + "epoch": 0.6970672389127325, + "grad_norm": 0.8464083512611802, + "learning_rate": 4.439482711807955e-05, + "loss": 1.8964, + "step": 7796 + }, + { + "epoch": 0.697156652360515, + "grad_norm": 1.1240113127950708, + "learning_rate": 4.4370759322398006e-05, + "loss": 1.3996, + "step": 7797 + }, + { + "epoch": 0.6972460658082976, + "grad_norm": 8.038413620674028, + "learning_rate": 4.434669619227368e-05, + "loss": 6.4586, + "step": 7798 + }, + { + "epoch": 0.6973354792560801, + "grad_norm": 1.5747061729066865, + "learning_rate": 4.432263772972475e-05, + "loss": 1.6691, + "step": 7799 + }, + { + "epoch": 0.6974248927038627, + "grad_norm": 6.017258584454689, + "learning_rate": 4.4298583936768976e-05, + "loss": 6.3998, + "step": 7800 + }, + { + "epoch": 0.6975143061516452, + "grad_norm": 2.1607691216207443, + "learning_rate": 4.427453481542366e-05, + "loss": 1.783, + "step": 7801 + }, + { + "epoch": 0.6976037195994278, + "grad_norm": 2.171105438197301, + "learning_rate": 4.4250490367705824e-05, + "loss": 1.7109, + "step": 7802 + }, + { + "epoch": 0.6976931330472103, + "grad_norm": 2.328084884754204, + "learning_rate": 4.4226450595632055e-05, + "loss": 1.575, + "step": 7803 + }, + { + "epoch": 0.6977825464949928, + "grad_norm": 2.190280783280943, + "learning_rate": 4.420241550121849e-05, + "loss": 1.5844, + "step": 7804 + }, + { + "epoch": 0.6978719599427754, + "grad_norm": 1.1604919961522893, + "learning_rate": 4.41783850864809e-05, + "loss": 1.4599, + "step": 7805 + }, + { + "epoch": 0.697961373390558, + "grad_norm": 2.955296198284339, + "learning_rate": 4.4154359353434824e-05, + "loss": 2.0755, + "step": 7806 + }, + { + "epoch": 0.6980507868383404, + "grad_norm": 1.6841619502634806, + "learning_rate": 4.4130338304095146e-05, + "loss": 1.617, + "step": 7807 + }, + { + "epoch": 0.698140200286123, + "grad_norm": 1.4144286500470413, + "learning_rate": 4.4106321940476516e-05, + "loss": 1.7491, + "step": 7808 + }, + { + "epoch": 0.6982296137339056, + "grad_norm": 2.0959077578573972, + "learning_rate": 4.408231026459321e-05, + "loss": 1.3301, + "step": 7809 + }, + { + "epoch": 0.6983190271816881, + "grad_norm": 2.65042308987001, + "learning_rate": 4.405830327845896e-05, + "loss": 1.6065, + "step": 7810 + }, + { + "epoch": 0.6984084406294707, + "grad_norm": 1.9145044851038437, + "learning_rate": 4.403430098408726e-05, + "loss": 1.9085, + "step": 7811 + }, + { + "epoch": 0.6984978540772532, + "grad_norm": 1.5366651978098562, + "learning_rate": 4.401030338349115e-05, + "loss": 1.6108, + "step": 7812 + }, + { + "epoch": 0.6985872675250357, + "grad_norm": 2.8574189465487008, + "learning_rate": 4.3986310478683265e-05, + "loss": 1.7889, + "step": 7813 + }, + { + "epoch": 0.6986766809728183, + "grad_norm": 1.9983944277713397, + "learning_rate": 4.3962322271675915e-05, + "loss": 1.7859, + "step": 7814 + }, + { + "epoch": 0.6987660944206009, + "grad_norm": 0.9454167905511233, + "learning_rate": 4.393833876448089e-05, + "loss": 1.7931, + "step": 7815 + }, + { + "epoch": 0.6988555078683834, + "grad_norm": 1.59162465627431, + "learning_rate": 4.3914359959109686e-05, + "loss": 1.5708, + "step": 7816 + }, + { + "epoch": 0.6989449213161659, + "grad_norm": 1.7886341289476297, + "learning_rate": 4.389038585757341e-05, + "loss": 1.3276, + "step": 7817 + }, + { + "epoch": 0.6990343347639485, + "grad_norm": 1.4468157162845943, + "learning_rate": 4.3866416461882676e-05, + "loss": 1.5584, + "step": 7818 + }, + { + "epoch": 0.6991237482117311, + "grad_norm": 0.7290795445048431, + "learning_rate": 4.3842451774047755e-05, + "loss": 1.4342, + "step": 7819 + }, + { + "epoch": 0.6992131616595136, + "grad_norm": 0.8082121640877529, + "learning_rate": 4.381849179607867e-05, + "loss": 1.7445, + "step": 7820 + }, + { + "epoch": 0.6993025751072961, + "grad_norm": 1.6166147043354147, + "learning_rate": 4.379453652998479e-05, + "loss": 6.5689, + "step": 7821 + }, + { + "epoch": 0.6993919885550787, + "grad_norm": 1.3330878122189052, + "learning_rate": 4.377058597777524e-05, + "loss": 1.8856, + "step": 7822 + }, + { + "epoch": 0.6994814020028612, + "grad_norm": 1.2653395518117336, + "learning_rate": 4.3746640141458786e-05, + "loss": 1.4335, + "step": 7823 + }, + { + "epoch": 0.6995708154506438, + "grad_norm": 0.7254113064961636, + "learning_rate": 4.372269902304363e-05, + "loss": 1.5294, + "step": 7824 + }, + { + "epoch": 0.6996602288984263, + "grad_norm": 0.5047508346003584, + "learning_rate": 4.369876262453776e-05, + "loss": 1.4818, + "step": 7825 + }, + { + "epoch": 0.6997496423462088, + "grad_norm": 1.412984312742882, + "learning_rate": 4.367483094794866e-05, + "loss": 1.5509, + "step": 7826 + }, + { + "epoch": 0.6998390557939914, + "grad_norm": 2.794807696165727, + "learning_rate": 4.365090399528349e-05, + "loss": 2.1345, + "step": 7827 + }, + { + "epoch": 0.699928469241774, + "grad_norm": 1.62602305060419, + "learning_rate": 4.362698176854892e-05, + "loss": 1.789, + "step": 7828 + }, + { + "epoch": 0.7000178826895566, + "grad_norm": 1.3397482832899423, + "learning_rate": 4.360306426975136e-05, + "loss": 1.7908, + "step": 7829 + }, + { + "epoch": 0.700107296137339, + "grad_norm": 0.5994348183154702, + "learning_rate": 4.357915150089665e-05, + "loss": 1.4572, + "step": 7830 + }, + { + "epoch": 0.7001967095851216, + "grad_norm": 1.1234463359550846, + "learning_rate": 4.355524346399037e-05, + "loss": 1.7694, + "step": 7831 + }, + { + "epoch": 0.7002861230329042, + "grad_norm": 0.6989747658016571, + "learning_rate": 4.3531340161037684e-05, + "loss": 1.5106, + "step": 7832 + }, + { + "epoch": 0.7003755364806867, + "grad_norm": 0.5429867186160733, + "learning_rate": 4.350744159404323e-05, + "loss": 1.8984, + "step": 7833 + }, + { + "epoch": 0.7004649499284692, + "grad_norm": 0.720420254890666, + "learning_rate": 4.348354776501149e-05, + "loss": 1.508, + "step": 7834 + }, + { + "epoch": 0.7005543633762518, + "grad_norm": 0.9444865872459858, + "learning_rate": 4.345965867594631e-05, + "loss": 1.8075, + "step": 7835 + }, + { + "epoch": 0.7006437768240343, + "grad_norm": 0.8770017143317539, + "learning_rate": 4.3435774328851276e-05, + "loss": 1.8672, + "step": 7836 + }, + { + "epoch": 0.7007331902718169, + "grad_norm": 0.7567079625988181, + "learning_rate": 4.3411894725729576e-05, + "loss": 1.6093, + "step": 7837 + }, + { + "epoch": 0.7008226037195995, + "grad_norm": 1.317478819858559, + "learning_rate": 4.338801986858388e-05, + "loss": 6.4072, + "step": 7838 + }, + { + "epoch": 0.7009120171673819, + "grad_norm": 0.8410942852276143, + "learning_rate": 4.336414975941656e-05, + "loss": 1.2358, + "step": 7839 + }, + { + "epoch": 0.7010014306151645, + "grad_norm": 0.9232513965183456, + "learning_rate": 4.3340284400229666e-05, + "loss": 1.4441, + "step": 7840 + }, + { + "epoch": 0.7010908440629471, + "grad_norm": 0.6001433080084255, + "learning_rate": 4.331642379302466e-05, + "loss": 1.5775, + "step": 7841 + }, + { + "epoch": 0.7011802575107297, + "grad_norm": 0.7966298803278843, + "learning_rate": 4.329256793980274e-05, + "loss": 2.0395, + "step": 7842 + }, + { + "epoch": 0.7012696709585121, + "grad_norm": 1.102059113757878, + "learning_rate": 4.326871684256469e-05, + "loss": 2.0027, + "step": 7843 + }, + { + "epoch": 0.7013590844062947, + "grad_norm": 1.304862342721905, + "learning_rate": 4.324487050331082e-05, + "loss": 1.5526, + "step": 7844 + }, + { + "epoch": 0.7014484978540773, + "grad_norm": 1.6487797706478104, + "learning_rate": 4.3221028924041105e-05, + "loss": 1.6284, + "step": 7845 + }, + { + "epoch": 0.7015379113018598, + "grad_norm": 0.7624207447462205, + "learning_rate": 4.3197192106755125e-05, + "loss": 1.9142, + "step": 7846 + }, + { + "epoch": 0.7016273247496424, + "grad_norm": 2.823247788279633, + "learning_rate": 4.317336005345204e-05, + "loss": 6.1786, + "step": 7847 + }, + { + "epoch": 0.7017167381974249, + "grad_norm": 0.7664713909154597, + "learning_rate": 4.314953276613066e-05, + "loss": 1.366, + "step": 7848 + }, + { + "epoch": 0.7018061516452074, + "grad_norm": 0.8764762124496673, + "learning_rate": 4.312571024678926e-05, + "loss": 1.5067, + "step": 7849 + }, + { + "epoch": 0.70189556509299, + "grad_norm": 1.0651593977924512, + "learning_rate": 4.310189249742588e-05, + "loss": 1.7861, + "step": 7850 + }, + { + "epoch": 0.7019849785407726, + "grad_norm": 1.0990720884247354, + "learning_rate": 4.307807952003804e-05, + "loss": 1.6094, + "step": 7851 + }, + { + "epoch": 0.702074391988555, + "grad_norm": 1.1839479684985914, + "learning_rate": 4.305427131662296e-05, + "loss": 1.9655, + "step": 7852 + }, + { + "epoch": 0.7021638054363376, + "grad_norm": 1.2923463020913963, + "learning_rate": 4.303046788917732e-05, + "loss": 1.5695, + "step": 7853 + }, + { + "epoch": 0.7022532188841202, + "grad_norm": 0.806910400423658, + "learning_rate": 4.3006669239697596e-05, + "loss": 1.3301, + "step": 7854 + }, + { + "epoch": 0.7023426323319027, + "grad_norm": 1.0572350483683999, + "learning_rate": 4.298287537017965e-05, + "loss": 1.6312, + "step": 7855 + }, + { + "epoch": 0.7024320457796852, + "grad_norm": 0.8702220016665231, + "learning_rate": 4.29590862826191e-05, + "loss": 1.7143, + "step": 7856 + }, + { + "epoch": 0.7025214592274678, + "grad_norm": 1.3535829175866767, + "learning_rate": 4.293530197901112e-05, + "loss": 1.619, + "step": 7857 + }, + { + "epoch": 0.7026108726752504, + "grad_norm": 1.2578889653378182, + "learning_rate": 4.291152246135042e-05, + "loss": 1.8291, + "step": 7858 + }, + { + "epoch": 0.7027002861230329, + "grad_norm": 0.5536856688210992, + "learning_rate": 4.288774773163138e-05, + "loss": 1.7504, + "step": 7859 + }, + { + "epoch": 0.7027896995708155, + "grad_norm": 0.5121645009306038, + "learning_rate": 4.286397779184796e-05, + "loss": 1.5081, + "step": 7860 + }, + { + "epoch": 0.702879113018598, + "grad_norm": 0.9269871812341817, + "learning_rate": 4.2840212643993725e-05, + "loss": 1.7666, + "step": 7861 + }, + { + "epoch": 0.7029685264663805, + "grad_norm": 2.8769732910480474, + "learning_rate": 4.2816452290061826e-05, + "loss": 6.3401, + "step": 7862 + }, + { + "epoch": 0.7030579399141631, + "grad_norm": 2.8693105879960212, + "learning_rate": 4.279269673204504e-05, + "loss": 2.1833, + "step": 7863 + }, + { + "epoch": 0.7031473533619457, + "grad_norm": 1.3715701712453126, + "learning_rate": 4.276894597193567e-05, + "loss": 1.6778, + "step": 7864 + }, + { + "epoch": 0.7032367668097281, + "grad_norm": 1.4175818768265216, + "learning_rate": 4.274520001172567e-05, + "loss": 1.5857, + "step": 7865 + }, + { + "epoch": 0.7033261802575107, + "grad_norm": 2.249311671856957, + "learning_rate": 4.2721458853406646e-05, + "loss": 1.4831, + "step": 7866 + }, + { + "epoch": 0.7034155937052933, + "grad_norm": 2.7151378341831633, + "learning_rate": 4.2697722498969616e-05, + "loss": 1.819, + "step": 7867 + }, + { + "epoch": 0.7035050071530758, + "grad_norm": 3.6365498338729942, + "learning_rate": 4.267399095040546e-05, + "loss": 1.9794, + "step": 7868 + }, + { + "epoch": 0.7035944206008584, + "grad_norm": 1.1319333805671363, + "learning_rate": 4.265026420970443e-05, + "loss": 1.2536, + "step": 7869 + }, + { + "epoch": 0.7036838340486409, + "grad_norm": 2.274766608031757, + "learning_rate": 4.2626542278856464e-05, + "loss": 1.7253, + "step": 7870 + }, + { + "epoch": 0.7037732474964234, + "grad_norm": 4.553890195733891, + "learning_rate": 4.2602825159851156e-05, + "loss": 1.5332, + "step": 7871 + }, + { + "epoch": 0.703862660944206, + "grad_norm": 1.145276800289338, + "learning_rate": 4.257911285467754e-05, + "loss": 1.6133, + "step": 7872 + }, + { + "epoch": 0.7039520743919886, + "grad_norm": 1.7013885806228455, + "learning_rate": 4.2555405365324385e-05, + "loss": 1.8308, + "step": 7873 + }, + { + "epoch": 0.704041487839771, + "grad_norm": 2.200850534154294, + "learning_rate": 4.2531702693780005e-05, + "loss": 1.614, + "step": 7874 + }, + { + "epoch": 0.7041309012875536, + "grad_norm": 1.5556519442524137, + "learning_rate": 4.250800484203232e-05, + "loss": 1.4517, + "step": 7875 + }, + { + "epoch": 0.7042203147353362, + "grad_norm": 0.7600186009251112, + "learning_rate": 4.2484311812068836e-05, + "loss": 1.6776, + "step": 7876 + }, + { + "epoch": 0.7043097281831188, + "grad_norm": 2.233041130358626, + "learning_rate": 4.246062360587669e-05, + "loss": 1.941, + "step": 7877 + }, + { + "epoch": 0.7043991416309013, + "grad_norm": 1.6945022074590834, + "learning_rate": 4.243694022544251e-05, + "loss": 1.664, + "step": 7878 + }, + { + "epoch": 0.7044885550786838, + "grad_norm": 0.6611462715026684, + "learning_rate": 4.241326167275265e-05, + "loss": 1.4872, + "step": 7879 + }, + { + "epoch": 0.7045779685264664, + "grad_norm": 1.0505109803017458, + "learning_rate": 4.238958794979302e-05, + "loss": 1.6759, + "step": 7880 + }, + { + "epoch": 0.7046673819742489, + "grad_norm": 1.1532155387875107, + "learning_rate": 4.236591905854898e-05, + "loss": 1.9857, + "step": 7881 + }, + { + "epoch": 0.7047567954220315, + "grad_norm": 0.9486669587220286, + "learning_rate": 4.23422550010058e-05, + "loss": 1.8781, + "step": 7882 + }, + { + "epoch": 0.704846208869814, + "grad_norm": 0.9297722844973342, + "learning_rate": 4.231859577914802e-05, + "loss": 1.5052, + "step": 7883 + }, + { + "epoch": 0.7049356223175965, + "grad_norm": 1.2234520387142716, + "learning_rate": 4.229494139495995e-05, + "loss": 1.4606, + "step": 7884 + }, + { + "epoch": 0.7050250357653791, + "grad_norm": 1.1399575851885446, + "learning_rate": 4.2271291850425455e-05, + "loss": 1.9254, + "step": 7885 + }, + { + "epoch": 0.7051144492131617, + "grad_norm": 1.1042888674938651, + "learning_rate": 4.224764714752803e-05, + "loss": 1.6496, + "step": 7886 + }, + { + "epoch": 0.7052038626609443, + "grad_norm": 0.6692793866453187, + "learning_rate": 4.2224007288250645e-05, + "loss": 2.1678, + "step": 7887 + }, + { + "epoch": 0.7052932761087267, + "grad_norm": 1.553002900091995, + "learning_rate": 4.2200372274576e-05, + "loss": 1.8491, + "step": 7888 + }, + { + "epoch": 0.7053826895565093, + "grad_norm": 0.9144924835461979, + "learning_rate": 4.2176742108486334e-05, + "loss": 1.3604, + "step": 7889 + }, + { + "epoch": 0.7054721030042919, + "grad_norm": 0.44335190994918877, + "learning_rate": 4.2153116791963465e-05, + "loss": 1.4239, + "step": 7890 + }, + { + "epoch": 0.7055615164520744, + "grad_norm": 1.0749742794474604, + "learning_rate": 4.212949632698887e-05, + "loss": 1.6789, + "step": 7891 + }, + { + "epoch": 0.7056509298998569, + "grad_norm": 0.5456817060737791, + "learning_rate": 4.210588071554349e-05, + "loss": 1.4994, + "step": 7892 + }, + { + "epoch": 0.7057403433476395, + "grad_norm": 0.5815346136419263, + "learning_rate": 4.208226995960798e-05, + "loss": 1.7724, + "step": 7893 + }, + { + "epoch": 0.705829756795422, + "grad_norm": 0.7606429703167905, + "learning_rate": 4.205866406116258e-05, + "loss": 1.4819, + "step": 7894 + }, + { + "epoch": 0.7059191702432046, + "grad_norm": 1.2002764246234254, + "learning_rate": 4.203506302218697e-05, + "loss": 1.8421, + "step": 7895 + }, + { + "epoch": 0.7060085836909872, + "grad_norm": 1.0293090008035066, + "learning_rate": 4.2011466844660655e-05, + "loss": 1.5783, + "step": 7896 + }, + { + "epoch": 0.7060979971387696, + "grad_norm": 0.4588339471566865, + "learning_rate": 4.1987875530562624e-05, + "loss": 1.6474, + "step": 7897 + }, + { + "epoch": 0.7061874105865522, + "grad_norm": 0.7263202107955904, + "learning_rate": 4.1964289081871376e-05, + "loss": 1.8687, + "step": 7898 + }, + { + "epoch": 0.7062768240343348, + "grad_norm": 0.5652238726585623, + "learning_rate": 4.1940707500565114e-05, + "loss": 1.986, + "step": 7899 + }, + { + "epoch": 0.7063662374821174, + "grad_norm": 0.5407808683814628, + "learning_rate": 4.191713078862163e-05, + "loss": 1.9201, + "step": 7900 + }, + { + "epoch": 0.7064556509298998, + "grad_norm": 0.74102035793984, + "learning_rate": 4.189355894801821e-05, + "loss": 1.6585, + "step": 7901 + }, + { + "epoch": 0.7065450643776824, + "grad_norm": 0.7159868281388707, + "learning_rate": 4.186999198073182e-05, + "loss": 1.5553, + "step": 7902 + }, + { + "epoch": 0.706634477825465, + "grad_norm": 0.5157687397876111, + "learning_rate": 4.1846429888739005e-05, + "loss": 1.4358, + "step": 7903 + }, + { + "epoch": 0.7067238912732475, + "grad_norm": 1.2981823431501458, + "learning_rate": 4.182287267401587e-05, + "loss": 1.8463, + "step": 7904 + }, + { + "epoch": 0.70681330472103, + "grad_norm": 0.3550161443698066, + "learning_rate": 4.17993203385382e-05, + "loss": 1.5242, + "step": 7905 + }, + { + "epoch": 0.7069027181688126, + "grad_norm": 0.5715062030893584, + "learning_rate": 4.1775772884281185e-05, + "loss": 1.425, + "step": 7906 + }, + { + "epoch": 0.7069921316165951, + "grad_norm": 0.8620875971931147, + "learning_rate": 4.17522303132198e-05, + "loss": 1.8133, + "step": 7907 + }, + { + "epoch": 0.7070815450643777, + "grad_norm": 0.7151761261969811, + "learning_rate": 4.17286926273285e-05, + "loss": 1.8172, + "step": 7908 + }, + { + "epoch": 0.7071709585121603, + "grad_norm": 0.9629517952145589, + "learning_rate": 4.170515982858139e-05, + "loss": 1.6731, + "step": 7909 + }, + { + "epoch": 0.7072603719599427, + "grad_norm": 0.6190969731461266, + "learning_rate": 4.168163191895211e-05, + "loss": 1.7853, + "step": 7910 + }, + { + "epoch": 0.7073497854077253, + "grad_norm": 0.7621790632613029, + "learning_rate": 4.1658108900413975e-05, + "loss": 1.3372, + "step": 7911 + }, + { + "epoch": 0.7074391988555079, + "grad_norm": 0.47632492225216766, + "learning_rate": 4.163459077493974e-05, + "loss": 1.4745, + "step": 7912 + }, + { + "epoch": 0.7075286123032904, + "grad_norm": 0.6078806921447879, + "learning_rate": 4.16110775445019e-05, + "loss": 1.4974, + "step": 7913 + }, + { + "epoch": 0.7076180257510729, + "grad_norm": 1.8396867768672365, + "learning_rate": 4.158756921107251e-05, + "loss": 6.3136, + "step": 7914 + }, + { + "epoch": 0.7077074391988555, + "grad_norm": 1.0363636742258648, + "learning_rate": 4.15640657766231e-05, + "loss": 1.723, + "step": 7915 + }, + { + "epoch": 0.707796852646638, + "grad_norm": 0.5447616437290459, + "learning_rate": 4.1540567243124886e-05, + "loss": 1.4728, + "step": 7916 + }, + { + "epoch": 0.7078862660944206, + "grad_norm": 0.3543039073952818, + "learning_rate": 4.1517073612548764e-05, + "loss": 1.4327, + "step": 7917 + }, + { + "epoch": 0.7079756795422032, + "grad_norm": 0.8293350767260207, + "learning_rate": 4.1493584886865026e-05, + "loss": 1.8309, + "step": 7918 + }, + { + "epoch": 0.7080650929899857, + "grad_norm": 0.5923234614518237, + "learning_rate": 4.147010106804365e-05, + "loss": 1.6215, + "step": 7919 + }, + { + "epoch": 0.7081545064377682, + "grad_norm": 1.1012286238172246, + "learning_rate": 4.144662215805426e-05, + "loss": 1.8025, + "step": 7920 + }, + { + "epoch": 0.7082439198855508, + "grad_norm": 0.7960430312981652, + "learning_rate": 4.142314815886591e-05, + "loss": 1.4176, + "step": 7921 + }, + { + "epoch": 0.7083333333333334, + "grad_norm": 0.8479890473772491, + "learning_rate": 4.1399679072447384e-05, + "loss": 2.098, + "step": 7922 + }, + { + "epoch": 0.7084227467811158, + "grad_norm": 0.46890810211416056, + "learning_rate": 4.137621490076701e-05, + "loss": 1.5994, + "step": 7923 + }, + { + "epoch": 0.7085121602288984, + "grad_norm": 0.663513789021977, + "learning_rate": 4.135275564579268e-05, + "loss": 1.8518, + "step": 7924 + }, + { + "epoch": 0.708601573676681, + "grad_norm": 1.0142162342851948, + "learning_rate": 4.1329301309491955e-05, + "loss": 1.8835, + "step": 7925 + }, + { + "epoch": 0.7086909871244635, + "grad_norm": 0.5287609068150159, + "learning_rate": 4.130585189383183e-05, + "loss": 1.567, + "step": 7926 + }, + { + "epoch": 0.7087804005722461, + "grad_norm": 0.7769768853928058, + "learning_rate": 4.128240740077902e-05, + "loss": 1.5746, + "step": 7927 + }, + { + "epoch": 0.7088698140200286, + "grad_norm": 0.8814265721170375, + "learning_rate": 4.1258967832299835e-05, + "loss": 1.9003, + "step": 7928 + }, + { + "epoch": 0.7089592274678111, + "grad_norm": 0.5693838578993222, + "learning_rate": 4.123553319035999e-05, + "loss": 1.3826, + "step": 7929 + }, + { + "epoch": 0.7090486409155937, + "grad_norm": 1.2245712861567422, + "learning_rate": 4.121210347692506e-05, + "loss": 1.843, + "step": 7930 + }, + { + "epoch": 0.7091380543633763, + "grad_norm": 0.7755218854562544, + "learning_rate": 4.1188678693960034e-05, + "loss": 1.5187, + "step": 7931 + }, + { + "epoch": 0.7092274678111588, + "grad_norm": 0.6189429357606318, + "learning_rate": 4.116525884342947e-05, + "loss": 1.5765, + "step": 7932 + }, + { + "epoch": 0.7093168812589413, + "grad_norm": 1.1104663458628659, + "learning_rate": 4.114184392729758e-05, + "loss": 1.4131, + "step": 7933 + }, + { + "epoch": 0.7094062947067239, + "grad_norm": 0.6120803017598608, + "learning_rate": 4.1118433947528215e-05, + "loss": 1.5851, + "step": 7934 + }, + { + "epoch": 0.7094957081545065, + "grad_norm": 1.2576978503290714, + "learning_rate": 4.109502890608463e-05, + "loss": 1.5982, + "step": 7935 + }, + { + "epoch": 0.709585121602289, + "grad_norm": 0.5117180771879308, + "learning_rate": 4.107162880492984e-05, + "loss": 1.8484, + "step": 7936 + }, + { + "epoch": 0.7096745350500715, + "grad_norm": 0.9070212685605117, + "learning_rate": 4.104823364602638e-05, + "loss": 1.667, + "step": 7937 + }, + { + "epoch": 0.7097639484978541, + "grad_norm": 0.9609846877687505, + "learning_rate": 4.1024843431336355e-05, + "loss": 1.5654, + "step": 7938 + }, + { + "epoch": 0.7098533619456366, + "grad_norm": 0.5603314807940065, + "learning_rate": 4.100145816282154e-05, + "loss": 1.9076, + "step": 7939 + }, + { + "epoch": 0.7099427753934192, + "grad_norm": 0.5577929992036279, + "learning_rate": 4.097807784244313e-05, + "loss": 1.5324, + "step": 7940 + }, + { + "epoch": 0.7100321888412017, + "grad_norm": 0.6962541612233896, + "learning_rate": 4.095470247216205e-05, + "loss": 1.5511, + "step": 7941 + }, + { + "epoch": 0.7101216022889842, + "grad_norm": 0.7365868694001798, + "learning_rate": 4.0931332053938766e-05, + "loss": 1.8967, + "step": 7942 + }, + { + "epoch": 0.7102110157367668, + "grad_norm": 1.3748737089453038, + "learning_rate": 4.090796658973333e-05, + "loss": 1.8068, + "step": 7943 + }, + { + "epoch": 0.7103004291845494, + "grad_norm": 0.689025723594016, + "learning_rate": 4.0884606081505374e-05, + "loss": 2.1618, + "step": 7944 + }, + { + "epoch": 0.710389842632332, + "grad_norm": 0.37476819013924845, + "learning_rate": 4.0861250531214136e-05, + "loss": 1.5447, + "step": 7945 + }, + { + "epoch": 0.7104792560801144, + "grad_norm": 0.9328784848202587, + "learning_rate": 4.083789994081837e-05, + "loss": 1.5126, + "step": 7946 + }, + { + "epoch": 0.710568669527897, + "grad_norm": 0.83527069665024, + "learning_rate": 4.081455431227648e-05, + "loss": 1.8759, + "step": 7947 + }, + { + "epoch": 0.7106580829756796, + "grad_norm": 0.5488326678014402, + "learning_rate": 4.0791213647546475e-05, + "loss": 1.8603, + "step": 7948 + }, + { + "epoch": 0.7107474964234621, + "grad_norm": 0.5921760691874338, + "learning_rate": 4.0767877948585845e-05, + "loss": 1.7173, + "step": 7949 + }, + { + "epoch": 0.7108369098712446, + "grad_norm": 0.5513398547232992, + "learning_rate": 4.0744547217351715e-05, + "loss": 1.7245, + "step": 7950 + }, + { + "epoch": 0.7109263233190272, + "grad_norm": 0.43991787358677314, + "learning_rate": 4.072122145580093e-05, + "loss": 1.5109, + "step": 7951 + }, + { + "epoch": 0.7110157367668097, + "grad_norm": 1.8428228595249376, + "learning_rate": 4.069790066588967e-05, + "loss": 6.227, + "step": 7952 + }, + { + "epoch": 0.7111051502145923, + "grad_norm": 1.7263632252591046, + "learning_rate": 4.067458484957386e-05, + "loss": 1.6375, + "step": 7953 + }, + { + "epoch": 0.7111945636623748, + "grad_norm": 1.2449609063955926, + "learning_rate": 4.0651274008809004e-05, + "loss": 6.2408, + "step": 7954 + }, + { + "epoch": 0.7112839771101573, + "grad_norm": 0.6622786895917621, + "learning_rate": 4.06279681455501e-05, + "loss": 1.6422, + "step": 7955 + }, + { + "epoch": 0.7113733905579399, + "grad_norm": 0.6735896544017036, + "learning_rate": 4.060466726175179e-05, + "loss": 1.7559, + "step": 7956 + }, + { + "epoch": 0.7114628040057225, + "grad_norm": 0.3206004002822093, + "learning_rate": 4.0581371359368315e-05, + "loss": 1.2058, + "step": 7957 + }, + { + "epoch": 0.711552217453505, + "grad_norm": 0.45283036366612917, + "learning_rate": 4.0558080440353455e-05, + "loss": 1.4437, + "step": 7958 + }, + { + "epoch": 0.7116416309012875, + "grad_norm": 0.859526788908486, + "learning_rate": 4.0534794506660645e-05, + "loss": 1.3681, + "step": 7959 + }, + { + "epoch": 0.7117310443490701, + "grad_norm": 0.8812079871977047, + "learning_rate": 4.0511513560242766e-05, + "loss": 1.6958, + "step": 7960 + }, + { + "epoch": 0.7118204577968527, + "grad_norm": 0.8274759978906423, + "learning_rate": 4.0488237603052396e-05, + "loss": 1.8539, + "step": 7961 + }, + { + "epoch": 0.7119098712446352, + "grad_norm": 0.708986557101766, + "learning_rate": 4.04649666370417e-05, + "loss": 1.7665, + "step": 7962 + }, + { + "epoch": 0.7119992846924177, + "grad_norm": 1.4738932224150192, + "learning_rate": 4.044170066416233e-05, + "loss": 1.8159, + "step": 7963 + }, + { + "epoch": 0.7120886981402003, + "grad_norm": 1.1700192956689566, + "learning_rate": 4.041843968636555e-05, + "loss": 1.556, + "step": 7964 + }, + { + "epoch": 0.7121781115879828, + "grad_norm": 0.48702891254951713, + "learning_rate": 4.0395183705602354e-05, + "loss": 1.4376, + "step": 7965 + }, + { + "epoch": 0.7122675250357654, + "grad_norm": 0.9861583191235183, + "learning_rate": 4.037193272382308e-05, + "loss": 1.5615, + "step": 7966 + }, + { + "epoch": 0.712356938483548, + "grad_norm": 1.1868364548625323, + "learning_rate": 4.034868674297779e-05, + "loss": 1.6802, + "step": 7967 + }, + { + "epoch": 0.7124463519313304, + "grad_norm": 0.7607817219060158, + "learning_rate": 4.0325445765016145e-05, + "loss": 1.9709, + "step": 7968 + }, + { + "epoch": 0.712535765379113, + "grad_norm": 0.6989936198639423, + "learning_rate": 4.030220979188726e-05, + "loss": 1.5419, + "step": 7969 + }, + { + "epoch": 0.7126251788268956, + "grad_norm": 1.416305370649085, + "learning_rate": 4.027897882553994e-05, + "loss": 1.8755, + "step": 7970 + }, + { + "epoch": 0.7127145922746781, + "grad_norm": 0.7655155199296704, + "learning_rate": 4.025575286792254e-05, + "loss": 1.6085, + "step": 7971 + }, + { + "epoch": 0.7128040057224606, + "grad_norm": 0.3802813486657194, + "learning_rate": 4.0232531920983e-05, + "loss": 1.4896, + "step": 7972 + }, + { + "epoch": 0.7128934191702432, + "grad_norm": 1.0714536116215012, + "learning_rate": 4.020931598666882e-05, + "loss": 1.6727, + "step": 7973 + }, + { + "epoch": 0.7129828326180258, + "grad_norm": 1.059865555115892, + "learning_rate": 4.018610506692713e-05, + "loss": 1.5569, + "step": 7974 + }, + { + "epoch": 0.7130722460658083, + "grad_norm": 0.5280699580775315, + "learning_rate": 4.0162899163704545e-05, + "loss": 1.9667, + "step": 7975 + }, + { + "epoch": 0.7131616595135909, + "grad_norm": 0.4330903242456768, + "learning_rate": 4.0139698278947336e-05, + "loss": 1.6525, + "step": 7976 + }, + { + "epoch": 0.7132510729613734, + "grad_norm": 1.1844904059053007, + "learning_rate": 4.0116502414601384e-05, + "loss": 1.6823, + "step": 7977 + }, + { + "epoch": 0.7133404864091559, + "grad_norm": 0.6446904294754514, + "learning_rate": 4.009331157261198e-05, + "loss": 1.4769, + "step": 7978 + }, + { + "epoch": 0.7134298998569385, + "grad_norm": 0.9757207319437914, + "learning_rate": 4.007012575492425e-05, + "loss": 1.8744, + "step": 7979 + }, + { + "epoch": 0.7135193133047211, + "grad_norm": 1.5093231648439278, + "learning_rate": 4.004694496348267e-05, + "loss": 1.8502, + "step": 7980 + }, + { + "epoch": 0.7136087267525035, + "grad_norm": 1.4283910671052589, + "learning_rate": 4.0023769200231395e-05, + "loss": 2.1144, + "step": 7981 + }, + { + "epoch": 0.7136981402002861, + "grad_norm": 1.0638195155859382, + "learning_rate": 4.0000598467114214e-05, + "loss": 2.0248, + "step": 7982 + }, + { + "epoch": 0.7137875536480687, + "grad_norm": 0.960869121089069, + "learning_rate": 3.997743276607434e-05, + "loss": 1.707, + "step": 7983 + }, + { + "epoch": 0.7138769670958512, + "grad_norm": 1.2251528576982895, + "learning_rate": 3.995427209905469e-05, + "loss": 1.7024, + "step": 7984 + }, + { + "epoch": 0.7139663805436338, + "grad_norm": 0.8458872945440088, + "learning_rate": 3.993111646799772e-05, + "loss": 1.6583, + "step": 7985 + }, + { + "epoch": 0.7140557939914163, + "grad_norm": 0.5204696416040014, + "learning_rate": 3.990796587484548e-05, + "loss": 1.7389, + "step": 7986 + }, + { + "epoch": 0.7141452074391988, + "grad_norm": 0.5224979055187916, + "learning_rate": 3.988482032153955e-05, + "loss": 1.5953, + "step": 7987 + }, + { + "epoch": 0.7142346208869814, + "grad_norm": 0.9477768728951195, + "learning_rate": 3.986167981002118e-05, + "loss": 1.7795, + "step": 7988 + }, + { + "epoch": 0.714324034334764, + "grad_norm": 2.4727679734164645, + "learning_rate": 3.983854434223107e-05, + "loss": 6.5983, + "step": 7989 + }, + { + "epoch": 0.7144134477825465, + "grad_norm": 1.542719424514163, + "learning_rate": 3.981541392010958e-05, + "loss": 1.327, + "step": 7990 + }, + { + "epoch": 0.714502861230329, + "grad_norm": 1.6582305494062195, + "learning_rate": 3.979228854559668e-05, + "loss": 1.7662, + "step": 7991 + }, + { + "epoch": 0.7145922746781116, + "grad_norm": 0.9321612429890022, + "learning_rate": 3.9769168220631745e-05, + "loss": 1.5356, + "step": 7992 + }, + { + "epoch": 0.7146816881258942, + "grad_norm": 0.8485174844254856, + "learning_rate": 3.974605294715402e-05, + "loss": 1.9004, + "step": 7993 + }, + { + "epoch": 0.7147711015736766, + "grad_norm": 1.8162811767703404, + "learning_rate": 3.972294272710202e-05, + "loss": 1.8451, + "step": 7994 + }, + { + "epoch": 0.7148605150214592, + "grad_norm": 1.2890027435339115, + "learning_rate": 3.9699837562414024e-05, + "loss": 1.5831, + "step": 7995 + }, + { + "epoch": 0.7149499284692418, + "grad_norm": 1.14224184279114, + "learning_rate": 3.967673745502785e-05, + "loss": 1.9007, + "step": 7996 + }, + { + "epoch": 0.7150393419170243, + "grad_norm": 0.8352280798055138, + "learning_rate": 3.965364240688083e-05, + "loss": 1.365, + "step": 7997 + }, + { + "epoch": 0.7151287553648069, + "grad_norm": 3.387097459399782, + "learning_rate": 3.963055241990994e-05, + "loss": 2.0888, + "step": 7998 + }, + { + "epoch": 0.7152181688125894, + "grad_norm": 1.0120173939854022, + "learning_rate": 3.96074674960517e-05, + "loss": 1.475, + "step": 7999 + }, + { + "epoch": 0.7153075822603719, + "grad_norm": 1.665624553082025, + "learning_rate": 3.958438763724224e-05, + "loss": 1.3825, + "step": 8000 + }, + { + "epoch": 0.7153969957081545, + "grad_norm": 3.419014604387571, + "learning_rate": 3.956131284541722e-05, + "loss": 1.7488, + "step": 8001 + }, + { + "epoch": 0.7154864091559371, + "grad_norm": 1.838998735086009, + "learning_rate": 3.953824312251193e-05, + "loss": 1.8514, + "step": 8002 + }, + { + "epoch": 0.7155758226037195, + "grad_norm": 1.6186937614073191, + "learning_rate": 3.951517847046113e-05, + "loss": 1.8338, + "step": 8003 + }, + { + "epoch": 0.7156652360515021, + "grad_norm": 2.4967334139398103, + "learning_rate": 3.949211889119928e-05, + "loss": 1.8335, + "step": 8004 + }, + { + "epoch": 0.7157546494992847, + "grad_norm": 3.2396861226640086, + "learning_rate": 3.946906438666037e-05, + "loss": 1.9758, + "step": 8005 + }, + { + "epoch": 0.7158440629470673, + "grad_norm": 0.6740450510880649, + "learning_rate": 3.9446014958777863e-05, + "loss": 1.605, + "step": 8006 + }, + { + "epoch": 0.7159334763948498, + "grad_norm": 2.059064914613794, + "learning_rate": 3.942297060948498e-05, + "loss": 1.5846, + "step": 8007 + }, + { + "epoch": 0.7160228898426323, + "grad_norm": 3.3574486780763633, + "learning_rate": 3.9399931340714436e-05, + "loss": 1.5299, + "step": 8008 + }, + { + "epoch": 0.7161123032904149, + "grad_norm": 3.168961761885192, + "learning_rate": 3.937689715439842e-05, + "loss": 6.1846, + "step": 8009 + }, + { + "epoch": 0.7162017167381974, + "grad_norm": 1.5354294533225457, + "learning_rate": 3.935386805246882e-05, + "loss": 1.8857, + "step": 8010 + }, + { + "epoch": 0.71629113018598, + "grad_norm": 1.1929723777488757, + "learning_rate": 3.933084403685712e-05, + "loss": 1.6006, + "step": 8011 + }, + { + "epoch": 0.7163805436337625, + "grad_norm": 1.4999778596586357, + "learning_rate": 3.930782510949418e-05, + "loss": 1.5721, + "step": 8012 + }, + { + "epoch": 0.716469957081545, + "grad_norm": 1.5835526260219523, + "learning_rate": 3.9284811272310715e-05, + "loss": 1.5413, + "step": 8013 + }, + { + "epoch": 0.7165593705293276, + "grad_norm": 0.6594528744029362, + "learning_rate": 3.9261802527236765e-05, + "loss": 1.5277, + "step": 8014 + }, + { + "epoch": 0.7166487839771102, + "grad_norm": 1.962543325780012, + "learning_rate": 3.92387988762021e-05, + "loss": 1.8673, + "step": 8015 + }, + { + "epoch": 0.7167381974248928, + "grad_norm": 1.577717957523665, + "learning_rate": 3.921580032113602e-05, + "loss": 1.6552, + "step": 8016 + }, + { + "epoch": 0.7168276108726752, + "grad_norm": 1.4613682467792768, + "learning_rate": 3.919280686396732e-05, + "loss": 1.4063, + "step": 8017 + }, + { + "epoch": 0.7169170243204578, + "grad_norm": 0.49310071263240024, + "learning_rate": 3.916981850662448e-05, + "loss": 1.3112, + "step": 8018 + }, + { + "epoch": 0.7170064377682404, + "grad_norm": 0.7571815599829552, + "learning_rate": 3.9146835251035485e-05, + "loss": 1.5268, + "step": 8019 + }, + { + "epoch": 0.7170958512160229, + "grad_norm": 0.9361327341442635, + "learning_rate": 3.9123857099127936e-05, + "loss": 1.6733, + "step": 8020 + }, + { + "epoch": 0.7171852646638054, + "grad_norm": 1.5775772550194307, + "learning_rate": 3.910088405282897e-05, + "loss": 1.7627, + "step": 8021 + }, + { + "epoch": 0.717274678111588, + "grad_norm": 1.2803871611418292, + "learning_rate": 3.907791611406534e-05, + "loss": 1.7402, + "step": 8022 + }, + { + "epoch": 0.7173640915593705, + "grad_norm": 1.0028843632763382, + "learning_rate": 3.9054953284763284e-05, + "loss": 1.5548, + "step": 8023 + }, + { + "epoch": 0.7174535050071531, + "grad_norm": 1.104389924973163, + "learning_rate": 3.9031995566848687e-05, + "loss": 1.6965, + "step": 8024 + }, + { + "epoch": 0.7175429184549357, + "grad_norm": 1.259236933639634, + "learning_rate": 3.900904296224702e-05, + "loss": 1.7377, + "step": 8025 + }, + { + "epoch": 0.7176323319027181, + "grad_norm": 0.779606109821572, + "learning_rate": 3.89860954728832e-05, + "loss": 1.4978, + "step": 8026 + }, + { + "epoch": 0.7177217453505007, + "grad_norm": 0.6284642596641117, + "learning_rate": 3.896315310068194e-05, + "loss": 1.8958, + "step": 8027 + }, + { + "epoch": 0.7178111587982833, + "grad_norm": 0.5073562081731843, + "learning_rate": 3.8940215847567274e-05, + "loss": 1.2966, + "step": 8028 + }, + { + "epoch": 0.7179005722460658, + "grad_norm": 1.164569760786038, + "learning_rate": 3.891728371546297e-05, + "loss": 1.5356, + "step": 8029 + }, + { + "epoch": 0.7179899856938483, + "grad_norm": 0.5552145740149679, + "learning_rate": 3.88943567062923e-05, + "loss": 1.2828, + "step": 8030 + }, + { + "epoch": 0.7180793991416309, + "grad_norm": 0.8010315189237053, + "learning_rate": 3.887143482197818e-05, + "loss": 1.8588, + "step": 8031 + }, + { + "epoch": 0.7181688125894135, + "grad_norm": 0.6828621051312652, + "learning_rate": 3.884851806444296e-05, + "loss": 1.2552, + "step": 8032 + }, + { + "epoch": 0.718258226037196, + "grad_norm": 1.1224723947270245, + "learning_rate": 3.882560643560869e-05, + "loss": 1.6603, + "step": 8033 + }, + { + "epoch": 0.7183476394849786, + "grad_norm": 0.6218368515580914, + "learning_rate": 3.880269993739691e-05, + "loss": 1.6252, + "step": 8034 + }, + { + "epoch": 0.718437052932761, + "grad_norm": 0.5529291570053224, + "learning_rate": 3.8779798571728786e-05, + "loss": 1.363, + "step": 8035 + }, + { + "epoch": 0.7185264663805436, + "grad_norm": 9.714570124512436, + "learning_rate": 3.8756902340525046e-05, + "loss": 6.3249, + "step": 8036 + }, + { + "epoch": 0.7186158798283262, + "grad_norm": 4.439193807683781, + "learning_rate": 3.8734011245705924e-05, + "loss": 1.7049, + "step": 8037 + }, + { + "epoch": 0.7187052932761088, + "grad_norm": 4.977510140199275, + "learning_rate": 3.871112528919128e-05, + "loss": 1.8329, + "step": 8038 + }, + { + "epoch": 0.7187947067238912, + "grad_norm": 0.9580977594868848, + "learning_rate": 3.868824447290058e-05, + "loss": 1.3574, + "step": 8039 + }, + { + "epoch": 0.7188841201716738, + "grad_norm": 1.9853509751816667, + "learning_rate": 3.866536879875269e-05, + "loss": 6.4647, + "step": 8040 + }, + { + "epoch": 0.7189735336194564, + "grad_norm": 3.6949073907357923, + "learning_rate": 3.86424982686663e-05, + "loss": 6.2928, + "step": 8041 + }, + { + "epoch": 0.719062947067239, + "grad_norm": 6.097093045936809, + "learning_rate": 3.861963288455949e-05, + "loss": 1.642, + "step": 8042 + }, + { + "epoch": 0.7191523605150214, + "grad_norm": 2.537701535718649, + "learning_rate": 3.8596772648349924e-05, + "loss": 1.7702, + "step": 8043 + }, + { + "epoch": 0.719241773962804, + "grad_norm": 1.8914472919853593, + "learning_rate": 3.857391756195487e-05, + "loss": 1.8981, + "step": 8044 + }, + { + "epoch": 0.7193311874105865, + "grad_norm": 2.7631373800792063, + "learning_rate": 3.85510676272912e-05, + "loss": 1.6311, + "step": 8045 + }, + { + "epoch": 0.7194206008583691, + "grad_norm": 4.970018612629714, + "learning_rate": 3.852822284627524e-05, + "loss": 1.6785, + "step": 8046 + }, + { + "epoch": 0.7195100143061517, + "grad_norm": 2.300400681434123, + "learning_rate": 3.8505383220823e-05, + "loss": 1.6239, + "step": 8047 + }, + { + "epoch": 0.7195994277539342, + "grad_norm": 2.6037580123348847, + "learning_rate": 3.848254875285e-05, + "loss": 1.7493, + "step": 8048 + }, + { + "epoch": 0.7196888412017167, + "grad_norm": 3.0002246739686274, + "learning_rate": 3.845971944427135e-05, + "loss": 1.7882, + "step": 8049 + }, + { + "epoch": 0.7197782546494993, + "grad_norm": 3.439000203569714, + "learning_rate": 3.8436895297001726e-05, + "loss": 1.6834, + "step": 8050 + }, + { + "epoch": 0.7198676680972819, + "grad_norm": 1.4438317590672238, + "learning_rate": 3.841407631295532e-05, + "loss": 1.6564, + "step": 8051 + }, + { + "epoch": 0.7199570815450643, + "grad_norm": 2.7827878831934436, + "learning_rate": 3.8391262494045955e-05, + "loss": 1.6478, + "step": 8052 + }, + { + "epoch": 0.7200464949928469, + "grad_norm": 11.333811718528176, + "learning_rate": 3.8368453842187026e-05, + "loss": 6.3573, + "step": 8053 + }, + { + "epoch": 0.7201359084406295, + "grad_norm": 2.575453478869461, + "learning_rate": 3.8345650359291384e-05, + "loss": 1.6703, + "step": 8054 + }, + { + "epoch": 0.720225321888412, + "grad_norm": 3.959576045687741, + "learning_rate": 3.8322852047271615e-05, + "loss": 1.8548, + "step": 8055 + }, + { + "epoch": 0.7203147353361946, + "grad_norm": 3.6182294316972228, + "learning_rate": 3.830005890803979e-05, + "loss": 1.7041, + "step": 8056 + }, + { + "epoch": 0.7204041487839771, + "grad_norm": 1.904440991158417, + "learning_rate": 3.8277270943507484e-05, + "loss": 1.7773, + "step": 8057 + }, + { + "epoch": 0.7204935622317596, + "grad_norm": 1.4619818581152513, + "learning_rate": 3.8254488155585924e-05, + "loss": 1.8214, + "step": 8058 + }, + { + "epoch": 0.7205829756795422, + "grad_norm": 3.221491126968318, + "learning_rate": 3.8231710546185895e-05, + "loss": 2.0346, + "step": 8059 + }, + { + "epoch": 0.7206723891273248, + "grad_norm": 1.885002827998933, + "learning_rate": 3.8208938117217674e-05, + "loss": 1.6428, + "step": 8060 + }, + { + "epoch": 0.7207618025751072, + "grad_norm": 0.9622237042998945, + "learning_rate": 3.8186170870591185e-05, + "loss": 1.609, + "step": 8061 + }, + { + "epoch": 0.7208512160228898, + "grad_norm": 0.7030030982417357, + "learning_rate": 3.8163408808215904e-05, + "loss": 1.2647, + "step": 8062 + }, + { + "epoch": 0.7209406294706724, + "grad_norm": 1.5761812393136967, + "learning_rate": 3.814065193200084e-05, + "loss": 1.8355, + "step": 8063 + }, + { + "epoch": 0.721030042918455, + "grad_norm": 2.399362732877736, + "learning_rate": 3.8117900243854595e-05, + "loss": 1.8115, + "step": 8064 + }, + { + "epoch": 0.7211194563662375, + "grad_norm": 1.1517176859941203, + "learning_rate": 3.809515374568535e-05, + "loss": 1.6582, + "step": 8065 + }, + { + "epoch": 0.72120886981402, + "grad_norm": 0.6049283738589426, + "learning_rate": 3.807241243940077e-05, + "loss": 1.4198, + "step": 8066 + }, + { + "epoch": 0.7212982832618026, + "grad_norm": 1.113874022549067, + "learning_rate": 3.804967632690817e-05, + "loss": 1.7006, + "step": 8067 + }, + { + "epoch": 0.7213876967095851, + "grad_norm": 1.6829892226643213, + "learning_rate": 3.802694541011439e-05, + "loss": 1.8821, + "step": 8068 + }, + { + "epoch": 0.7214771101573677, + "grad_norm": 1.7283482765220923, + "learning_rate": 3.8004219690925856e-05, + "loss": 1.4884, + "step": 8069 + }, + { + "epoch": 0.7215665236051502, + "grad_norm": 0.8043744775055816, + "learning_rate": 3.7981499171248594e-05, + "loss": 1.8129, + "step": 8070 + }, + { + "epoch": 0.7216559370529327, + "grad_norm": 0.7705669355143098, + "learning_rate": 3.795878385298804e-05, + "loss": 1.7476, + "step": 8071 + }, + { + "epoch": 0.7217453505007153, + "grad_norm": 0.7468419327018794, + "learning_rate": 3.793607373804937e-05, + "loss": 1.5897, + "step": 8072 + }, + { + "epoch": 0.7218347639484979, + "grad_norm": 1.0161263817196275, + "learning_rate": 3.7913368828337285e-05, + "loss": 1.4792, + "step": 8073 + }, + { + "epoch": 0.7219241773962805, + "grad_norm": 0.875764611780845, + "learning_rate": 3.789066912575593e-05, + "loss": 1.3291, + "step": 8074 + }, + { + "epoch": 0.7220135908440629, + "grad_norm": 1.4760928757697083, + "learning_rate": 3.78679746322091e-05, + "loss": 1.47, + "step": 8075 + }, + { + "epoch": 0.7221030042918455, + "grad_norm": 0.6938777273894836, + "learning_rate": 3.784528534960029e-05, + "loss": 1.7528, + "step": 8076 + }, + { + "epoch": 0.7221924177396281, + "grad_norm": 0.6097225217057946, + "learning_rate": 3.782260127983229e-05, + "loss": 1.7055, + "step": 8077 + }, + { + "epoch": 0.7222818311874106, + "grad_norm": 1.468512589078429, + "learning_rate": 3.7799922424807634e-05, + "loss": 1.9393, + "step": 8078 + }, + { + "epoch": 0.7223712446351931, + "grad_norm": 1.4545394446510251, + "learning_rate": 3.777724878642839e-05, + "loss": 1.6762, + "step": 8079 + }, + { + "epoch": 0.7224606580829757, + "grad_norm": 0.6393307484029467, + "learning_rate": 3.7754580366596115e-05, + "loss": 1.3671, + "step": 8080 + }, + { + "epoch": 0.7225500715307582, + "grad_norm": 0.42993058089888975, + "learning_rate": 3.773191716721202e-05, + "loss": 1.8012, + "step": 8081 + }, + { + "epoch": 0.7226394849785408, + "grad_norm": 1.3398358712425524, + "learning_rate": 3.7709259190176816e-05, + "loss": 2.0591, + "step": 8082 + }, + { + "epoch": 0.7227288984263234, + "grad_norm": 0.7577747183707557, + "learning_rate": 3.768660643739083e-05, + "loss": 1.8062, + "step": 8083 + }, + { + "epoch": 0.7228183118741058, + "grad_norm": 0.3336002284422223, + "learning_rate": 3.766395891075394e-05, + "loss": 1.3345, + "step": 8084 + }, + { + "epoch": 0.7229077253218884, + "grad_norm": 0.4172206144331433, + "learning_rate": 3.764131661216549e-05, + "loss": 1.7097, + "step": 8085 + }, + { + "epoch": 0.722997138769671, + "grad_norm": 7.260118092598869, + "learning_rate": 3.7618679543524503e-05, + "loss": 6.2489, + "step": 8086 + }, + { + "epoch": 0.7230865522174535, + "grad_norm": 1.402946459965293, + "learning_rate": 3.759604770672953e-05, + "loss": 1.6441, + "step": 8087 + }, + { + "epoch": 0.723175965665236, + "grad_norm": 2.8021662645985135, + "learning_rate": 3.757342110367871e-05, + "loss": 2.0017, + "step": 8088 + }, + { + "epoch": 0.7232653791130186, + "grad_norm": 1.5472272629225823, + "learning_rate": 3.755079973626959e-05, + "loss": 1.4388, + "step": 8089 + }, + { + "epoch": 0.7233547925608012, + "grad_norm": 1.4058437053651434, + "learning_rate": 3.752818360639956e-05, + "loss": 1.5718, + "step": 8090 + }, + { + "epoch": 0.7234442060085837, + "grad_norm": 0.9688174596498683, + "learning_rate": 3.7505572715965284e-05, + "loss": 1.7483, + "step": 8091 + }, + { + "epoch": 0.7235336194563662, + "grad_norm": 1.2243637058882288, + "learning_rate": 3.748296706686315e-05, + "loss": 1.6182, + "step": 8092 + }, + { + "epoch": 0.7236230329041488, + "grad_norm": 2.4571386713277263, + "learning_rate": 3.74603666609891e-05, + "loss": 1.8105, + "step": 8093 + }, + { + "epoch": 0.7237124463519313, + "grad_norm": 2.0711162436992288, + "learning_rate": 3.7437771500238526e-05, + "loss": 1.9021, + "step": 8094 + }, + { + "epoch": 0.7238018597997139, + "grad_norm": 1.0071911959842543, + "learning_rate": 3.741518158650648e-05, + "loss": 1.6481, + "step": 8095 + }, + { + "epoch": 0.7238912732474965, + "grad_norm": 1.5913484641919007, + "learning_rate": 3.739259692168764e-05, + "loss": 1.6082, + "step": 8096 + }, + { + "epoch": 0.7239806866952789, + "grad_norm": 1.4180809468438345, + "learning_rate": 3.737001750767604e-05, + "loss": 1.421, + "step": 8097 + }, + { + "epoch": 0.7240701001430615, + "grad_norm": 2.6746484480813644, + "learning_rate": 3.734744334636544e-05, + "loss": 1.9602, + "step": 8098 + }, + { + "epoch": 0.7241595135908441, + "grad_norm": 1.3341968233701185, + "learning_rate": 3.732487443964914e-05, + "loss": 1.7026, + "step": 8099 + }, + { + "epoch": 0.7242489270386266, + "grad_norm": 0.6285667231701786, + "learning_rate": 3.730231078941988e-05, + "loss": 2.06, + "step": 8100 + }, + { + "epoch": 0.7243383404864091, + "grad_norm": 1.729477385859225, + "learning_rate": 3.727975239757011e-05, + "loss": 1.8103, + "step": 8101 + }, + { + "epoch": 0.7244277539341917, + "grad_norm": 0.9469283975196053, + "learning_rate": 3.725719926599175e-05, + "loss": 1.387, + "step": 8102 + }, + { + "epoch": 0.7245171673819742, + "grad_norm": 0.9080733854300378, + "learning_rate": 3.723465139657632e-05, + "loss": 1.4531, + "step": 8103 + }, + { + "epoch": 0.7246065808297568, + "grad_norm": 0.559333141369615, + "learning_rate": 3.72121087912149e-05, + "loss": 1.6205, + "step": 8104 + }, + { + "epoch": 0.7246959942775394, + "grad_norm": 0.4825261807573586, + "learning_rate": 3.7189571451798065e-05, + "loss": 1.4271, + "step": 8105 + }, + { + "epoch": 0.7247854077253219, + "grad_norm": 0.8352981773239138, + "learning_rate": 3.7167039380216005e-05, + "loss": 2.0637, + "step": 8106 + }, + { + "epoch": 0.7248748211731044, + "grad_norm": 1.0418506771131903, + "learning_rate": 3.714451257835852e-05, + "loss": 1.5807, + "step": 8107 + }, + { + "epoch": 0.724964234620887, + "grad_norm": 1.3301240261989977, + "learning_rate": 3.71219910481148e-05, + "loss": 1.7178, + "step": 8108 + }, + { + "epoch": 0.7250536480686696, + "grad_norm": 0.616250120158056, + "learning_rate": 3.7099474791373736e-05, + "loss": 2.1997, + "step": 8109 + }, + { + "epoch": 0.725143061516452, + "grad_norm": 0.6502228847053922, + "learning_rate": 3.707696381002381e-05, + "loss": 2.118, + "step": 8110 + }, + { + "epoch": 0.7252324749642346, + "grad_norm": 1.236179848006066, + "learning_rate": 3.705445810595291e-05, + "loss": 2.0946, + "step": 8111 + }, + { + "epoch": 0.7253218884120172, + "grad_norm": 0.743719027657325, + "learning_rate": 3.7031957681048604e-05, + "loss": 1.5829, + "step": 8112 + }, + { + "epoch": 0.7254113018597997, + "grad_norm": 1.3574173709201602, + "learning_rate": 3.700946253719798e-05, + "loss": 2.2097, + "step": 8113 + }, + { + "epoch": 0.7255007153075823, + "grad_norm": 0.49636295835234545, + "learning_rate": 3.6986972676287626e-05, + "loss": 1.6706, + "step": 8114 + }, + { + "epoch": 0.7255901287553648, + "grad_norm": 1.14715424680997, + "learning_rate": 3.6964488100203776e-05, + "loss": 1.9992, + "step": 8115 + }, + { + "epoch": 0.7256795422031473, + "grad_norm": 1.0510031958555621, + "learning_rate": 3.6942008810832184e-05, + "loss": 1.4209, + "step": 8116 + }, + { + "epoch": 0.7257689556509299, + "grad_norm": 1.0501907587200996, + "learning_rate": 3.6919534810058154e-05, + "loss": 1.4364, + "step": 8117 + }, + { + "epoch": 0.7258583690987125, + "grad_norm": 0.9016445033831222, + "learning_rate": 3.68970660997666e-05, + "loss": 1.5641, + "step": 8118 + }, + { + "epoch": 0.725947782546495, + "grad_norm": 0.8184123610191956, + "learning_rate": 3.687460268184185e-05, + "loss": 1.5268, + "step": 8119 + }, + { + "epoch": 0.7260371959942775, + "grad_norm": 0.4435328229846703, + "learning_rate": 3.685214455816796e-05, + "loss": 1.7512, + "step": 8120 + }, + { + "epoch": 0.7261266094420601, + "grad_norm": 0.6767480901263527, + "learning_rate": 3.682969173062842e-05, + "loss": 1.4948, + "step": 8121 + }, + { + "epoch": 0.7262160228898427, + "grad_norm": 0.9710735856846198, + "learning_rate": 3.6807244201106394e-05, + "loss": 1.6343, + "step": 8122 + }, + { + "epoch": 0.7263054363376252, + "grad_norm": 0.5715272410057851, + "learning_rate": 3.6784801971484405e-05, + "loss": 1.4915, + "step": 8123 + }, + { + "epoch": 0.7263948497854077, + "grad_norm": 0.6158381217649749, + "learning_rate": 3.6762365043644806e-05, + "loss": 1.7243, + "step": 8124 + }, + { + "epoch": 0.7264842632331903, + "grad_norm": 0.5480096356716522, + "learning_rate": 3.673993341946924e-05, + "loss": 1.4761, + "step": 8125 + }, + { + "epoch": 0.7265736766809728, + "grad_norm": 0.621968635042734, + "learning_rate": 3.671750710083906e-05, + "loss": 1.6029, + "step": 8126 + }, + { + "epoch": 0.7266630901287554, + "grad_norm": 0.5361476111915261, + "learning_rate": 3.6695086089635156e-05, + "loss": 1.9002, + "step": 8127 + }, + { + "epoch": 0.7267525035765379, + "grad_norm": 0.41757286672225197, + "learning_rate": 3.667267038773791e-05, + "loss": 1.4075, + "step": 8128 + }, + { + "epoch": 0.7268419170243204, + "grad_norm": 0.4652160535985314, + "learning_rate": 3.6650259997027315e-05, + "loss": 2.0586, + "step": 8129 + }, + { + "epoch": 0.726931330472103, + "grad_norm": 0.7035208416902358, + "learning_rate": 3.66278549193829e-05, + "loss": 2.1558, + "step": 8130 + }, + { + "epoch": 0.7270207439198856, + "grad_norm": 0.768439346499258, + "learning_rate": 3.6605455156683766e-05, + "loss": 1.8868, + "step": 8131 + }, + { + "epoch": 0.727110157367668, + "grad_norm": 0.6379312691845569, + "learning_rate": 3.658306071080855e-05, + "loss": 1.8023, + "step": 8132 + }, + { + "epoch": 0.7271995708154506, + "grad_norm": 0.9505722598647462, + "learning_rate": 3.6560671583635467e-05, + "loss": 1.7814, + "step": 8133 + }, + { + "epoch": 0.7272889842632332, + "grad_norm": 0.70291837713496, + "learning_rate": 3.6538287777042215e-05, + "loss": 1.5754, + "step": 8134 + }, + { + "epoch": 0.7273783977110158, + "grad_norm": 0.5963239671105878, + "learning_rate": 3.6515909292906126e-05, + "loss": 1.5744, + "step": 8135 + }, + { + "epoch": 0.7274678111587983, + "grad_norm": 0.7741200603787094, + "learning_rate": 3.649353613310409e-05, + "loss": 1.4684, + "step": 8136 + }, + { + "epoch": 0.7275572246065808, + "grad_norm": 0.5721798461293635, + "learning_rate": 3.6471168299512405e-05, + "loss": 1.5065, + "step": 8137 + }, + { + "epoch": 0.7276466380543634, + "grad_norm": 1.4545738022016004, + "learning_rate": 3.644880579400719e-05, + "loss": 1.9604, + "step": 8138 + }, + { + "epoch": 0.7277360515021459, + "grad_norm": 0.522770066630751, + "learning_rate": 3.6426448618463836e-05, + "loss": 1.8873, + "step": 8139 + }, + { + "epoch": 0.7278254649499285, + "grad_norm": 0.617224854709574, + "learning_rate": 3.640409677475748e-05, + "loss": 1.6157, + "step": 8140 + }, + { + "epoch": 0.727914878397711, + "grad_norm": 0.737046459277262, + "learning_rate": 3.6381750264762734e-05, + "loss": 1.4598, + "step": 8141 + }, + { + "epoch": 0.7280042918454935, + "grad_norm": 1.1367368061168732, + "learning_rate": 3.6359409090353744e-05, + "loss": 1.53, + "step": 8142 + }, + { + "epoch": 0.7280937052932761, + "grad_norm": 1.5324992898900054, + "learning_rate": 3.633707325340425e-05, + "loss": 1.5829, + "step": 8143 + }, + { + "epoch": 0.7281831187410587, + "grad_norm": 0.8621428531964925, + "learning_rate": 3.631474275578754e-05, + "loss": 1.549, + "step": 8144 + }, + { + "epoch": 0.7282725321888412, + "grad_norm": 0.4778748897494385, + "learning_rate": 3.6292417599376436e-05, + "loss": 1.3307, + "step": 8145 + }, + { + "epoch": 0.7283619456366237, + "grad_norm": 0.669321603004757, + "learning_rate": 3.627009778604333e-05, + "loss": 1.4025, + "step": 8146 + }, + { + "epoch": 0.7284513590844063, + "grad_norm": 1.1487463301109766, + "learning_rate": 3.624778331766019e-05, + "loss": 1.4886, + "step": 8147 + }, + { + "epoch": 0.7285407725321889, + "grad_norm": 1.255327824436943, + "learning_rate": 3.6225474196098444e-05, + "loss": 1.8793, + "step": 8148 + }, + { + "epoch": 0.7286301859799714, + "grad_norm": 5.13284805454595, + "learning_rate": 3.620317042322915e-05, + "loss": 6.2186, + "step": 8149 + }, + { + "epoch": 0.7287195994277539, + "grad_norm": 1.9792314015075065, + "learning_rate": 3.6180872000922935e-05, + "loss": 1.5742, + "step": 8150 + }, + { + "epoch": 0.7288090128755365, + "grad_norm": 3.041886961292758, + "learning_rate": 3.615857893104986e-05, + "loss": 1.6386, + "step": 8151 + }, + { + "epoch": 0.728898426323319, + "grad_norm": 2.273563312530059, + "learning_rate": 3.613629121547969e-05, + "loss": 2.002, + "step": 8152 + }, + { + "epoch": 0.7289878397711016, + "grad_norm": 0.5828145668476702, + "learning_rate": 3.611400885608168e-05, + "loss": 1.5578, + "step": 8153 + }, + { + "epoch": 0.7290772532188842, + "grad_norm": 2.0681783615533713, + "learning_rate": 3.6091731854724566e-05, + "loss": 1.7752, + "step": 8154 + }, + { + "epoch": 0.7291666666666666, + "grad_norm": 1.5519273647967877, + "learning_rate": 3.606946021327672e-05, + "loss": 1.3738, + "step": 8155 + }, + { + "epoch": 0.7292560801144492, + "grad_norm": 1.8259337490127592, + "learning_rate": 3.604719393360606e-05, + "loss": 1.6736, + "step": 8156 + }, + { + "epoch": 0.7293454935622318, + "grad_norm": 1.2441448722431487, + "learning_rate": 3.6024933017579984e-05, + "loss": 2.0079, + "step": 8157 + }, + { + "epoch": 0.7294349070100143, + "grad_norm": 0.7530811821398756, + "learning_rate": 3.600267746706552e-05, + "loss": 1.7768, + "step": 8158 + }, + { + "epoch": 0.7295243204577968, + "grad_norm": 1.3660124933800708, + "learning_rate": 3.59804272839292e-05, + "loss": 1.5328, + "step": 8159 + }, + { + "epoch": 0.7296137339055794, + "grad_norm": 1.6070715826451496, + "learning_rate": 3.595818247003713e-05, + "loss": 1.4846, + "step": 8160 + }, + { + "epoch": 0.729703147353362, + "grad_norm": 1.4221565768805213, + "learning_rate": 3.593594302725498e-05, + "loss": 1.8594, + "step": 8161 + }, + { + "epoch": 0.7297925608011445, + "grad_norm": 0.9275807652749838, + "learning_rate": 3.591370895744789e-05, + "loss": 1.6476, + "step": 8162 + }, + { + "epoch": 0.7298819742489271, + "grad_norm": 0.7792840738625793, + "learning_rate": 3.5891480262480635e-05, + "loss": 1.4326, + "step": 8163 + }, + { + "epoch": 0.7299713876967096, + "grad_norm": 0.8465564286814863, + "learning_rate": 3.5869256944217535e-05, + "loss": 1.2999, + "step": 8164 + }, + { + "epoch": 0.7300608011444921, + "grad_norm": 1.467380913538008, + "learning_rate": 3.584703900452234e-05, + "loss": 1.1996, + "step": 8165 + }, + { + "epoch": 0.7301502145922747, + "grad_norm": 0.7500015516007714, + "learning_rate": 3.582482644525854e-05, + "loss": 1.4524, + "step": 8166 + }, + { + "epoch": 0.7302396280400573, + "grad_norm": 5.880745219189027, + "learning_rate": 3.580261926828908e-05, + "loss": 6.3362, + "step": 8167 + }, + { + "epoch": 0.7303290414878397, + "grad_norm": 1.5656284351136693, + "learning_rate": 3.578041747547638e-05, + "loss": 1.7522, + "step": 8168 + }, + { + "epoch": 0.7304184549356223, + "grad_norm": 1.4670782622536447, + "learning_rate": 3.57582210686825e-05, + "loss": 1.6383, + "step": 8169 + }, + { + "epoch": 0.7305078683834049, + "grad_norm": 1.8193015962809282, + "learning_rate": 3.5736030049769074e-05, + "loss": 2.0589, + "step": 8170 + }, + { + "epoch": 0.7305972818311874, + "grad_norm": 0.7350382875408662, + "learning_rate": 3.5713844420597155e-05, + "loss": 1.6312, + "step": 8171 + }, + { + "epoch": 0.73068669527897, + "grad_norm": 0.7043875965517151, + "learning_rate": 3.569166418302747e-05, + "loss": 1.3609, + "step": 8172 + }, + { + "epoch": 0.7307761087267525, + "grad_norm": 1.8606889120296002, + "learning_rate": 3.566948933892025e-05, + "loss": 1.6435, + "step": 8173 + }, + { + "epoch": 0.730865522174535, + "grad_norm": 1.29720703574208, + "learning_rate": 3.564731989013527e-05, + "loss": 1.4075, + "step": 8174 + }, + { + "epoch": 0.7309549356223176, + "grad_norm": 0.425423660313272, + "learning_rate": 3.5625155838531877e-05, + "loss": 1.4412, + "step": 8175 + }, + { + "epoch": 0.7310443490701002, + "grad_norm": 0.7371839520237057, + "learning_rate": 3.560299718596889e-05, + "loss": 1.4741, + "step": 8176 + }, + { + "epoch": 0.7311337625178826, + "grad_norm": 0.8447753511729827, + "learning_rate": 3.558084393430475e-05, + "loss": 1.9081, + "step": 8177 + }, + { + "epoch": 0.7312231759656652, + "grad_norm": 1.9591742678168023, + "learning_rate": 3.555869608539743e-05, + "loss": 2.1489, + "step": 8178 + }, + { + "epoch": 0.7313125894134478, + "grad_norm": 0.7738820866303779, + "learning_rate": 3.5536553641104465e-05, + "loss": 1.3843, + "step": 8179 + }, + { + "epoch": 0.7314020028612304, + "grad_norm": 0.4231764283788071, + "learning_rate": 3.5514416603282876e-05, + "loss": 1.6765, + "step": 8180 + }, + { + "epoch": 0.7314914163090128, + "grad_norm": 1.1857124069823712, + "learning_rate": 3.549228497378932e-05, + "loss": 1.7406, + "step": 8181 + }, + { + "epoch": 0.7315808297567954, + "grad_norm": 1.0274179874167482, + "learning_rate": 3.547015875447989e-05, + "loss": 1.8753, + "step": 8182 + }, + { + "epoch": 0.731670243204578, + "grad_norm": 0.8415032294925476, + "learning_rate": 3.544803794721031e-05, + "loss": 1.3713, + "step": 8183 + }, + { + "epoch": 0.7317596566523605, + "grad_norm": 1.001082424777426, + "learning_rate": 3.542592255383586e-05, + "loss": 1.7747, + "step": 8184 + }, + { + "epoch": 0.7318490701001431, + "grad_norm": 0.743886387480144, + "learning_rate": 3.5403812576211246e-05, + "loss": 1.6524, + "step": 8185 + }, + { + "epoch": 0.7319384835479256, + "grad_norm": 0.9424234529813283, + "learning_rate": 3.538170801619088e-05, + "loss": 1.4694, + "step": 8186 + }, + { + "epoch": 0.7320278969957081, + "grad_norm": 0.7683603085297889, + "learning_rate": 3.535960887562866e-05, + "loss": 1.3565, + "step": 8187 + }, + { + "epoch": 0.7321173104434907, + "grad_norm": 0.3852143791305563, + "learning_rate": 3.533751515637794e-05, + "loss": 1.3012, + "step": 8188 + }, + { + "epoch": 0.7322067238912733, + "grad_norm": 1.0262639275438554, + "learning_rate": 3.531542686029173e-05, + "loss": 1.6479, + "step": 8189 + }, + { + "epoch": 0.7322961373390557, + "grad_norm": 1.0197357957397237, + "learning_rate": 3.5293343989222593e-05, + "loss": 1.9615, + "step": 8190 + }, + { + "epoch": 0.7323855507868383, + "grad_norm": 0.6253624729112404, + "learning_rate": 3.527126654502252e-05, + "loss": 1.488, + "step": 8191 + }, + { + "epoch": 0.7324749642346209, + "grad_norm": 0.8342269083514993, + "learning_rate": 3.5249194529543137e-05, + "loss": 1.5505, + "step": 8192 + }, + { + "epoch": 0.7325643776824035, + "grad_norm": 1.1070613521423678, + "learning_rate": 3.5227127944635606e-05, + "loss": 2.0296, + "step": 8193 + }, + { + "epoch": 0.732653791130186, + "grad_norm": 0.7305507094314498, + "learning_rate": 3.520506679215064e-05, + "loss": 1.7206, + "step": 8194 + }, + { + "epoch": 0.7327432045779685, + "grad_norm": 0.5762380349101212, + "learning_rate": 3.51830110739385e-05, + "loss": 1.8069, + "step": 8195 + }, + { + "epoch": 0.7328326180257511, + "grad_norm": 0.45360901583950486, + "learning_rate": 3.516096079184891e-05, + "loss": 1.408, + "step": 8196 + }, + { + "epoch": 0.7329220314735336, + "grad_norm": 0.8049356862098216, + "learning_rate": 3.513891594773123e-05, + "loss": 1.5777, + "step": 8197 + }, + { + "epoch": 0.7330114449213162, + "grad_norm": 0.5060872167140869, + "learning_rate": 3.5116876543434374e-05, + "loss": 1.2066, + "step": 8198 + }, + { + "epoch": 0.7331008583690987, + "grad_norm": 0.4561635619616763, + "learning_rate": 3.509484258080665e-05, + "loss": 1.5152, + "step": 8199 + }, + { + "epoch": 0.7331902718168812, + "grad_norm": 1.0767408629909665, + "learning_rate": 3.507281406169614e-05, + "loss": 1.9469, + "step": 8200 + }, + { + "epoch": 0.7332796852646638, + "grad_norm": 0.5717068861672462, + "learning_rate": 3.505079098795032e-05, + "loss": 1.5569, + "step": 8201 + }, + { + "epoch": 0.7333690987124464, + "grad_norm": 1.1885608932603682, + "learning_rate": 3.502877336141619e-05, + "loss": 1.8143, + "step": 8202 + }, + { + "epoch": 0.733458512160229, + "grad_norm": 0.9176370655659511, + "learning_rate": 3.5006761183940386e-05, + "loss": 1.6823, + "step": 8203 + }, + { + "epoch": 0.7335479256080114, + "grad_norm": 0.5788524783264506, + "learning_rate": 3.498475445736905e-05, + "loss": 1.7018, + "step": 8204 + }, + { + "epoch": 0.733637339055794, + "grad_norm": 0.5489061261956456, + "learning_rate": 3.4962753183547806e-05, + "loss": 1.3989, + "step": 8205 + }, + { + "epoch": 0.7337267525035766, + "grad_norm": 1.1276874451209211, + "learning_rate": 3.4940757364321906e-05, + "loss": 1.9059, + "step": 8206 + }, + { + "epoch": 0.7338161659513591, + "grad_norm": 0.9727038229531751, + "learning_rate": 3.491876700153612e-05, + "loss": 1.7631, + "step": 8207 + }, + { + "epoch": 0.7339055793991416, + "grad_norm": 0.5700708010370995, + "learning_rate": 3.489678209703475e-05, + "loss": 1.4335, + "step": 8208 + }, + { + "epoch": 0.7339949928469242, + "grad_norm": 0.6129671291070046, + "learning_rate": 3.487480265266164e-05, + "loss": 1.9462, + "step": 8209 + }, + { + "epoch": 0.7340844062947067, + "grad_norm": 0.6260016939660326, + "learning_rate": 3.485282867026021e-05, + "loss": 1.9328, + "step": 8210 + }, + { + "epoch": 0.7341738197424893, + "grad_norm": 0.8975897158106554, + "learning_rate": 3.483086015167333e-05, + "loss": 1.8184, + "step": 8211 + }, + { + "epoch": 0.7342632331902719, + "grad_norm": 0.43994245559210804, + "learning_rate": 3.48088970987435e-05, + "loss": 1.4338, + "step": 8212 + }, + { + "epoch": 0.7343526466380543, + "grad_norm": 0.5434416630170513, + "learning_rate": 3.4786939513312744e-05, + "loss": 1.5813, + "step": 8213 + }, + { + "epoch": 0.7344420600858369, + "grad_norm": 0.7568247912729444, + "learning_rate": 3.4764987397222614e-05, + "loss": 1.3476, + "step": 8214 + }, + { + "epoch": 0.7345314735336195, + "grad_norm": 1.57493885223088, + "learning_rate": 3.474304075231424e-05, + "loss": 2.183, + "step": 8215 + }, + { + "epoch": 0.734620886981402, + "grad_norm": 0.4327187936286653, + "learning_rate": 3.472109958042819e-05, + "loss": 1.9481, + "step": 8216 + }, + { + "epoch": 0.7347103004291845, + "grad_norm": 0.8483839324580463, + "learning_rate": 3.4699163883404685e-05, + "loss": 1.2826, + "step": 8217 + }, + { + "epoch": 0.7347997138769671, + "grad_norm": 0.8887520286285552, + "learning_rate": 3.467723366308348e-05, + "loss": 2.1425, + "step": 8218 + }, + { + "epoch": 0.7348891273247496, + "grad_norm": 1.0261401264915386, + "learning_rate": 3.465530892130375e-05, + "loss": 1.7708, + "step": 8219 + }, + { + "epoch": 0.7349785407725322, + "grad_norm": 0.648654344989215, + "learning_rate": 3.4633389659904324e-05, + "loss": 1.4982, + "step": 8220 + }, + { + "epoch": 0.7350679542203148, + "grad_norm": 0.6311990652808934, + "learning_rate": 3.461147588072362e-05, + "loss": 1.6578, + "step": 8221 + }, + { + "epoch": 0.7351573676680973, + "grad_norm": 0.5813175806595601, + "learning_rate": 3.458956758559945e-05, + "loss": 1.4479, + "step": 8222 + }, + { + "epoch": 0.7352467811158798, + "grad_norm": 0.7211696005594183, + "learning_rate": 3.4567664776369236e-05, + "loss": 1.6847, + "step": 8223 + }, + { + "epoch": 0.7353361945636624, + "grad_norm": 0.4639749781945933, + "learning_rate": 3.4545767454869995e-05, + "loss": 1.6125, + "step": 8224 + }, + { + "epoch": 0.735425608011445, + "grad_norm": 0.8881136229941742, + "learning_rate": 3.452387562293814e-05, + "loss": 1.9119, + "step": 8225 + }, + { + "epoch": 0.7355150214592274, + "grad_norm": 0.8992131309264473, + "learning_rate": 3.4501989282409776e-05, + "loss": 1.762, + "step": 8226 + }, + { + "epoch": 0.73560443490701, + "grad_norm": 0.4747562416573698, + "learning_rate": 3.448010843512046e-05, + "loss": 1.7725, + "step": 8227 + }, + { + "epoch": 0.7356938483547926, + "grad_norm": 0.5076830254163289, + "learning_rate": 3.4458233082905334e-05, + "loss": 1.4793, + "step": 8228 + }, + { + "epoch": 0.7357832618025751, + "grad_norm": 0.6962480012839924, + "learning_rate": 3.443636322759908e-05, + "loss": 1.6287, + "step": 8229 + }, + { + "epoch": 0.7358726752503576, + "grad_norm": 0.6460784605438914, + "learning_rate": 3.4414498871035816e-05, + "loss": 1.4075, + "step": 8230 + }, + { + "epoch": 0.7359620886981402, + "grad_norm": 1.7569431093976975, + "learning_rate": 3.4392640015049325e-05, + "loss": 1.9179, + "step": 8231 + }, + { + "epoch": 0.7360515021459227, + "grad_norm": 0.6574774746606109, + "learning_rate": 3.437078666147292e-05, + "loss": 1.687, + "step": 8232 + }, + { + "epoch": 0.7361409155937053, + "grad_norm": 1.1470687159663637, + "learning_rate": 3.434893881213934e-05, + "loss": 1.3291, + "step": 8233 + }, + { + "epoch": 0.7362303290414879, + "grad_norm": 1.1531286077500895, + "learning_rate": 3.432709646888095e-05, + "loss": 1.6199, + "step": 8234 + }, + { + "epoch": 0.7363197424892703, + "grad_norm": 1.416922851289972, + "learning_rate": 3.430525963352973e-05, + "loss": 1.5096, + "step": 8235 + }, + { + "epoch": 0.7364091559370529, + "grad_norm": 0.6170602125870583, + "learning_rate": 3.428342830791701e-05, + "loss": 1.5185, + "step": 8236 + }, + { + "epoch": 0.7364985693848355, + "grad_norm": 0.4481614101327138, + "learning_rate": 3.426160249387379e-05, + "loss": 1.2808, + "step": 8237 + }, + { + "epoch": 0.7365879828326181, + "grad_norm": 1.2615386710414882, + "learning_rate": 3.423978219323062e-05, + "loss": 1.9753, + "step": 8238 + }, + { + "epoch": 0.7366773962804005, + "grad_norm": 0.791972474417649, + "learning_rate": 3.421796740781745e-05, + "loss": 1.585, + "step": 8239 + }, + { + "epoch": 0.7367668097281831, + "grad_norm": 0.33525605986628826, + "learning_rate": 3.4196158139463915e-05, + "loss": 1.2352, + "step": 8240 + }, + { + "epoch": 0.7368562231759657, + "grad_norm": 1.0082158559055805, + "learning_rate": 3.417435438999914e-05, + "loss": 1.6474, + "step": 8241 + }, + { + "epoch": 0.7369456366237482, + "grad_norm": 0.8722368743411494, + "learning_rate": 3.4152556161251744e-05, + "loss": 1.7462, + "step": 8242 + }, + { + "epoch": 0.7370350500715308, + "grad_norm": 0.5587108612713523, + "learning_rate": 3.413076345504995e-05, + "loss": 1.4897, + "step": 8243 + }, + { + "epoch": 0.7371244635193133, + "grad_norm": 1.0041234492068631, + "learning_rate": 3.410897627322152e-05, + "loss": 1.6812, + "step": 8244 + }, + { + "epoch": 0.7372138769670958, + "grad_norm": 1.1837674157127216, + "learning_rate": 3.408719461759362e-05, + "loss": 1.7932, + "step": 8245 + }, + { + "epoch": 0.7373032904148784, + "grad_norm": 1.1348226672503303, + "learning_rate": 3.406541848999312e-05, + "loss": 1.7276, + "step": 8246 + }, + { + "epoch": 0.737392703862661, + "grad_norm": 0.75885030183011, + "learning_rate": 3.404364789224637e-05, + "loss": 1.3852, + "step": 8247 + }, + { + "epoch": 0.7374821173104434, + "grad_norm": 1.2066534809936647, + "learning_rate": 3.402188282617914e-05, + "loss": 2.2536, + "step": 8248 + }, + { + "epoch": 0.737571530758226, + "grad_norm": 1.475489969685909, + "learning_rate": 3.4000123293616995e-05, + "loss": 1.5188, + "step": 8249 + }, + { + "epoch": 0.7376609442060086, + "grad_norm": 1.3776728902256834, + "learning_rate": 3.397836929638476e-05, + "loss": 1.7672, + "step": 8250 + }, + { + "epoch": 0.7377503576537912, + "grad_norm": 0.43283143975186045, + "learning_rate": 3.395662083630696e-05, + "loss": 1.4361, + "step": 8251 + }, + { + "epoch": 0.7378397711015737, + "grad_norm": 2.0785365967226244, + "learning_rate": 3.393487791520765e-05, + "loss": 1.4254, + "step": 8252 + }, + { + "epoch": 0.7379291845493562, + "grad_norm": 1.0647701893035157, + "learning_rate": 3.391314053491031e-05, + "loss": 1.4491, + "step": 8253 + }, + { + "epoch": 0.7380185979971388, + "grad_norm": 0.42738795365316284, + "learning_rate": 3.389140869723806e-05, + "loss": 1.5234, + "step": 8254 + }, + { + "epoch": 0.7381080114449213, + "grad_norm": 0.808842621923257, + "learning_rate": 3.3869682404013516e-05, + "loss": 1.6328, + "step": 8255 + }, + { + "epoch": 0.7381974248927039, + "grad_norm": 0.5312376140301435, + "learning_rate": 3.3847961657058845e-05, + "loss": 1.5934, + "step": 8256 + }, + { + "epoch": 0.7382868383404864, + "grad_norm": 1.3156769670370263, + "learning_rate": 3.382624645819574e-05, + "loss": 1.8235, + "step": 8257 + }, + { + "epoch": 0.7383762517882689, + "grad_norm": 0.6393771246154036, + "learning_rate": 3.3804536809245455e-05, + "loss": 1.3989, + "step": 8258 + }, + { + "epoch": 0.7384656652360515, + "grad_norm": 0.6161330878600626, + "learning_rate": 3.37828327120287e-05, + "loss": 1.4316, + "step": 8259 + }, + { + "epoch": 0.7385550786838341, + "grad_norm": 1.0385225577710273, + "learning_rate": 3.376113416836579e-05, + "loss": 1.7834, + "step": 8260 + }, + { + "epoch": 0.7386444921316166, + "grad_norm": 1.3754666962765012, + "learning_rate": 3.373944118007657e-05, + "loss": 1.6703, + "step": 8261 + }, + { + "epoch": 0.7387339055793991, + "grad_norm": 1.076354734825202, + "learning_rate": 3.371775374898038e-05, + "loss": 1.5255, + "step": 8262 + }, + { + "epoch": 0.7388233190271817, + "grad_norm": 0.594827877983172, + "learning_rate": 3.369607187689618e-05, + "loss": 1.7044, + "step": 8263 + }, + { + "epoch": 0.7389127324749643, + "grad_norm": 0.6605289987660693, + "learning_rate": 3.3674395565642324e-05, + "loss": 1.6858, + "step": 8264 + }, + { + "epoch": 0.7390021459227468, + "grad_norm": 0.8835175023132555, + "learning_rate": 3.365272481703681e-05, + "loss": 1.4723, + "step": 8265 + }, + { + "epoch": 0.7390915593705293, + "grad_norm": 0.5169460037973818, + "learning_rate": 3.3631059632897135e-05, + "loss": 1.7662, + "step": 8266 + }, + { + "epoch": 0.7391809728183119, + "grad_norm": 0.46427736992990676, + "learning_rate": 3.360940001504037e-05, + "loss": 1.4278, + "step": 8267 + }, + { + "epoch": 0.7392703862660944, + "grad_norm": 0.577587876475361, + "learning_rate": 3.358774596528298e-05, + "loss": 1.5314, + "step": 8268 + }, + { + "epoch": 0.739359799713877, + "grad_norm": 4.181040361819913, + "learning_rate": 3.35660974854412e-05, + "loss": 6.2731, + "step": 8269 + }, + { + "epoch": 0.7394492131616596, + "grad_norm": 1.1693616390930803, + "learning_rate": 3.354445457733054e-05, + "loss": 1.6437, + "step": 8270 + }, + { + "epoch": 0.739538626609442, + "grad_norm": 1.8694869074926723, + "learning_rate": 3.352281724276623e-05, + "loss": 1.9999, + "step": 8271 + }, + { + "epoch": 0.7396280400572246, + "grad_norm": 0.9990254955687049, + "learning_rate": 3.3501185483562994e-05, + "loss": 2.2542, + "step": 8272 + }, + { + "epoch": 0.7397174535050072, + "grad_norm": 0.5223299443172369, + "learning_rate": 3.347955930153498e-05, + "loss": 1.5115, + "step": 8273 + }, + { + "epoch": 0.7398068669527897, + "grad_norm": 0.9196142086197244, + "learning_rate": 3.3457938698496e-05, + "loss": 1.619, + "step": 8274 + }, + { + "epoch": 0.7398962804005722, + "grad_norm": 1.3175579573342093, + "learning_rate": 3.343632367625932e-05, + "loss": 1.4271, + "step": 8275 + }, + { + "epoch": 0.7399856938483548, + "grad_norm": 1.0481040277028644, + "learning_rate": 3.34147142366378e-05, + "loss": 1.4559, + "step": 8276 + }, + { + "epoch": 0.7400751072961373, + "grad_norm": 0.5798011154628862, + "learning_rate": 3.339311038144378e-05, + "loss": 1.9604, + "step": 8277 + }, + { + "epoch": 0.7401645207439199, + "grad_norm": 1.1635463591074091, + "learning_rate": 3.337151211248918e-05, + "loss": 1.5309, + "step": 8278 + }, + { + "epoch": 0.7402539341917024, + "grad_norm": 0.9762622554578244, + "learning_rate": 3.3349919431585366e-05, + "loss": 1.4332, + "step": 8279 + }, + { + "epoch": 0.740343347639485, + "grad_norm": 1.6467385445516765, + "learning_rate": 3.332833234054331e-05, + "loss": 1.8593, + "step": 8280 + }, + { + "epoch": 0.7404327610872675, + "grad_norm": 1.607813805529186, + "learning_rate": 3.330675084117354e-05, + "loss": 1.5937, + "step": 8281 + }, + { + "epoch": 0.7405221745350501, + "grad_norm": 1.2834627001152181, + "learning_rate": 3.3285174935285954e-05, + "loss": 1.7143, + "step": 8282 + }, + { + "epoch": 0.7406115879828327, + "grad_norm": 1.9514133518791796, + "learning_rate": 3.3263604624690257e-05, + "loss": 1.7996, + "step": 8283 + }, + { + "epoch": 0.7407010014306151, + "grad_norm": 0.9707629658413547, + "learning_rate": 3.32420399111954e-05, + "loss": 1.8029, + "step": 8284 + }, + { + "epoch": 0.7407904148783977, + "grad_norm": 1.391419303798603, + "learning_rate": 3.322048079661004e-05, + "loss": 1.5961, + "step": 8285 + }, + { + "epoch": 0.7408798283261803, + "grad_norm": 1.6631100299205108, + "learning_rate": 3.3198927282742334e-05, + "loss": 1.5417, + "step": 8286 + }, + { + "epoch": 0.7409692417739628, + "grad_norm": 0.604583373296135, + "learning_rate": 3.3177379371399886e-05, + "loss": 1.5664, + "step": 8287 + }, + { + "epoch": 0.7410586552217453, + "grad_norm": 1.1006505266771112, + "learning_rate": 3.315583706438994e-05, + "loss": 1.5964, + "step": 8288 + }, + { + "epoch": 0.7411480686695279, + "grad_norm": 0.8319349520866935, + "learning_rate": 3.31343003635192e-05, + "loss": 1.8246, + "step": 8289 + }, + { + "epoch": 0.7412374821173104, + "grad_norm": 0.4267814500816897, + "learning_rate": 3.311276927059393e-05, + "loss": 1.7603, + "step": 8290 + }, + { + "epoch": 0.741326895565093, + "grad_norm": 1.2402055612111773, + "learning_rate": 3.3091243787419944e-05, + "loss": 1.7067, + "step": 8291 + }, + { + "epoch": 0.7414163090128756, + "grad_norm": 0.5146222801213908, + "learning_rate": 3.306972391580255e-05, + "loss": 1.4743, + "step": 8292 + }, + { + "epoch": 0.741505722460658, + "grad_norm": 0.7146803838268994, + "learning_rate": 3.304820965754656e-05, + "loss": 1.5383, + "step": 8293 + }, + { + "epoch": 0.7415951359084406, + "grad_norm": 0.8000619304278754, + "learning_rate": 3.302670101445636e-05, + "loss": 1.6365, + "step": 8294 + }, + { + "epoch": 0.7416845493562232, + "grad_norm": 1.1192138439438344, + "learning_rate": 3.3005197988335904e-05, + "loss": 1.9058, + "step": 8295 + }, + { + "epoch": 0.7417739628040058, + "grad_norm": 0.5337824892041874, + "learning_rate": 3.2983700580988505e-05, + "loss": 1.5775, + "step": 8296 + }, + { + "epoch": 0.7418633762517882, + "grad_norm": 0.7779717299609863, + "learning_rate": 3.2962208794217275e-05, + "loss": 1.628, + "step": 8297 + }, + { + "epoch": 0.7419527896995708, + "grad_norm": 0.6233848629875948, + "learning_rate": 3.2940722629824604e-05, + "loss": 1.7296, + "step": 8298 + }, + { + "epoch": 0.7420422031473534, + "grad_norm": 0.5758177711253027, + "learning_rate": 3.291924208961253e-05, + "loss": 1.5836, + "step": 8299 + }, + { + "epoch": 0.7421316165951359, + "grad_norm": 0.6539028962026187, + "learning_rate": 3.289776717538262e-05, + "loss": 1.9887, + "step": 8300 + }, + { + "epoch": 0.7422210300429185, + "grad_norm": 0.7940629112114691, + "learning_rate": 3.287629788893596e-05, + "loss": 1.61, + "step": 8301 + }, + { + "epoch": 0.742310443490701, + "grad_norm": 0.5865582885659341, + "learning_rate": 3.2854834232073105e-05, + "loss": 1.8093, + "step": 8302 + }, + { + "epoch": 0.7423998569384835, + "grad_norm": 0.8816863350987975, + "learning_rate": 3.283337620659421e-05, + "loss": 1.6316, + "step": 8303 + }, + { + "epoch": 0.7424892703862661, + "grad_norm": 0.6152024123084563, + "learning_rate": 3.281192381429894e-05, + "loss": 1.639, + "step": 8304 + }, + { + "epoch": 0.7425786838340487, + "grad_norm": 1.0749828787582196, + "learning_rate": 3.279047705698647e-05, + "loss": 1.8872, + "step": 8305 + }, + { + "epoch": 0.7426680972818311, + "grad_norm": 0.8445578471161911, + "learning_rate": 3.276903593645555e-05, + "loss": 1.702, + "step": 8306 + }, + { + "epoch": 0.7427575107296137, + "grad_norm": 0.44278121653784114, + "learning_rate": 3.2747600454504366e-05, + "loss": 1.6693, + "step": 8307 + }, + { + "epoch": 0.7428469241773963, + "grad_norm": 0.6277770121565233, + "learning_rate": 3.2726170612930716e-05, + "loss": 1.4598, + "step": 8308 + }, + { + "epoch": 0.7429363376251789, + "grad_norm": 1.2440425140382527, + "learning_rate": 3.270474641353192e-05, + "loss": 1.8637, + "step": 8309 + }, + { + "epoch": 0.7430257510729614, + "grad_norm": 0.6953806446974545, + "learning_rate": 3.26833278581047e-05, + "loss": 1.6261, + "step": 8310 + }, + { + "epoch": 0.7431151645207439, + "grad_norm": 1.2277913482306062, + "learning_rate": 3.266191494844552e-05, + "loss": 2.0339, + "step": 8311 + }, + { + "epoch": 0.7432045779685265, + "grad_norm": 0.608194970484472, + "learning_rate": 3.264050768635022e-05, + "loss": 1.6019, + "step": 8312 + }, + { + "epoch": 0.743293991416309, + "grad_norm": 1.0626540468733656, + "learning_rate": 3.261910607361417e-05, + "loss": 1.74, + "step": 8313 + }, + { + "epoch": 0.7433834048640916, + "grad_norm": 1.0846509194005767, + "learning_rate": 3.259771011203232e-05, + "loss": 1.5961, + "step": 8314 + }, + { + "epoch": 0.7434728183118741, + "grad_norm": 1.1328843629498515, + "learning_rate": 3.257631980339916e-05, + "loss": 1.8517, + "step": 8315 + }, + { + "epoch": 0.7435622317596566, + "grad_norm": 0.6597853522141697, + "learning_rate": 3.2554935149508584e-05, + "loss": 1.5276, + "step": 8316 + }, + { + "epoch": 0.7436516452074392, + "grad_norm": 1.324252674913824, + "learning_rate": 3.253355615215416e-05, + "loss": 1.7937, + "step": 8317 + }, + { + "epoch": 0.7437410586552218, + "grad_norm": 1.8219392872486493, + "learning_rate": 3.251218281312889e-05, + "loss": 1.9132, + "step": 8318 + }, + { + "epoch": 0.7438304721030042, + "grad_norm": 0.5358136765526518, + "learning_rate": 3.249081513422534e-05, + "loss": 1.7468, + "step": 8319 + }, + { + "epoch": 0.7439198855507868, + "grad_norm": 0.864975324924435, + "learning_rate": 3.246945311723564e-05, + "loss": 1.6954, + "step": 8320 + }, + { + "epoch": 0.7440092989985694, + "grad_norm": 1.7220896222538804, + "learning_rate": 3.244809676395131e-05, + "loss": 1.9405, + "step": 8321 + }, + { + "epoch": 0.744098712446352, + "grad_norm": 1.08114444829307, + "learning_rate": 3.2426746076163514e-05, + "loss": 1.4597, + "step": 8322 + }, + { + "epoch": 0.7441881258941345, + "grad_norm": 0.47826455319355327, + "learning_rate": 3.240540105566293e-05, + "loss": 1.5163, + "step": 8323 + }, + { + "epoch": 0.744277539341917, + "grad_norm": 0.852595961018296, + "learning_rate": 3.238406170423972e-05, + "loss": 1.4371, + "step": 8324 + }, + { + "epoch": 0.7443669527896996, + "grad_norm": 2.4234130736970076, + "learning_rate": 3.2362728023683594e-05, + "loss": 1.9193, + "step": 8325 + }, + { + "epoch": 0.7444563662374821, + "grad_norm": 0.9297191274761152, + "learning_rate": 3.234140001578383e-05, + "loss": 1.4111, + "step": 8326 + }, + { + "epoch": 0.7445457796852647, + "grad_norm": 0.746296634393825, + "learning_rate": 3.23200776823291e-05, + "loss": 1.6818, + "step": 8327 + }, + { + "epoch": 0.7446351931330472, + "grad_norm": 0.9207766222635512, + "learning_rate": 3.2298761025107706e-05, + "loss": 1.5127, + "step": 8328 + }, + { + "epoch": 0.7447246065808297, + "grad_norm": 1.7675642999177448, + "learning_rate": 3.22774500459075e-05, + "loss": 1.881, + "step": 8329 + }, + { + "epoch": 0.7448140200286123, + "grad_norm": 1.646021288440473, + "learning_rate": 3.2256144746515735e-05, + "loss": 1.8019, + "step": 8330 + }, + { + "epoch": 0.7449034334763949, + "grad_norm": 0.6706826618507257, + "learning_rate": 3.223484512871927e-05, + "loss": 1.5515, + "step": 8331 + }, + { + "epoch": 0.7449928469241774, + "grad_norm": 1.4350636167648627, + "learning_rate": 3.221355119430456e-05, + "loss": 1.6006, + "step": 8332 + }, + { + "epoch": 0.7450822603719599, + "grad_norm": 1.5457329967577718, + "learning_rate": 3.219226294505743e-05, + "loss": 1.5853, + "step": 8333 + }, + { + "epoch": 0.7451716738197425, + "grad_norm": 1.4397300103301196, + "learning_rate": 3.2170980382763306e-05, + "loss": 1.5195, + "step": 8334 + }, + { + "epoch": 0.745261087267525, + "grad_norm": 0.953061285276552, + "learning_rate": 3.214970350920716e-05, + "loss": 1.7953, + "step": 8335 + }, + { + "epoch": 0.7453505007153076, + "grad_norm": 1.376402660484465, + "learning_rate": 3.212843232617343e-05, + "loss": 1.6922, + "step": 8336 + }, + { + "epoch": 0.7454399141630901, + "grad_norm": 1.9879816594464053, + "learning_rate": 3.21071668354461e-05, + "loss": 1.818, + "step": 8337 + }, + { + "epoch": 0.7455293276108726, + "grad_norm": 1.504155409883328, + "learning_rate": 3.2085907038808695e-05, + "loss": 1.4898, + "step": 8338 + }, + { + "epoch": 0.7456187410586552, + "grad_norm": 0.9494394776037126, + "learning_rate": 3.2064652938044246e-05, + "loss": 1.9093, + "step": 8339 + }, + { + "epoch": 0.7457081545064378, + "grad_norm": 1.5144348190210204, + "learning_rate": 3.204340453493534e-05, + "loss": 1.7464, + "step": 8340 + }, + { + "epoch": 0.7457975679542204, + "grad_norm": 1.6382398086257108, + "learning_rate": 3.2022161831264e-05, + "loss": 1.728, + "step": 8341 + }, + { + "epoch": 0.7458869814020028, + "grad_norm": 1.4994052134999942, + "learning_rate": 3.200092482881184e-05, + "loss": 1.8021, + "step": 8342 + }, + { + "epoch": 0.7459763948497854, + "grad_norm": 0.6662888167705272, + "learning_rate": 3.197969352936003e-05, + "loss": 1.7446, + "step": 8343 + }, + { + "epoch": 0.746065808297568, + "grad_norm": 1.090975392742594, + "learning_rate": 3.1958467934689153e-05, + "loss": 1.5331, + "step": 8344 + }, + { + "epoch": 0.7461552217453505, + "grad_norm": 0.9527721592703211, + "learning_rate": 3.193724804657936e-05, + "loss": 1.549, + "step": 8345 + }, + { + "epoch": 0.746244635193133, + "grad_norm": 1.3441406461139944, + "learning_rate": 3.1916033866810436e-05, + "loss": 2.023, + "step": 8346 + }, + { + "epoch": 0.7463340486409156, + "grad_norm": 0.7718064088210185, + "learning_rate": 3.189482539716149e-05, + "loss": 1.5105, + "step": 8347 + }, + { + "epoch": 0.7464234620886981, + "grad_norm": 0.953411742539727, + "learning_rate": 3.1873622639411293e-05, + "loss": 1.5447, + "step": 8348 + }, + { + "epoch": 0.7465128755364807, + "grad_norm": 1.5220359400794548, + "learning_rate": 3.185242559533812e-05, + "loss": 1.6768, + "step": 8349 + }, + { + "epoch": 0.7466022889842633, + "grad_norm": 6.042758681505995, + "learning_rate": 3.183123426671968e-05, + "loss": 6.174, + "step": 8350 + }, + { + "epoch": 0.7466917024320457, + "grad_norm": 1.2017562601175962, + "learning_rate": 3.181004865533329e-05, + "loss": 1.6393, + "step": 8351 + }, + { + "epoch": 0.7467811158798283, + "grad_norm": 3.535521733601947, + "learning_rate": 3.178886876295578e-05, + "loss": 1.9392, + "step": 8352 + }, + { + "epoch": 0.7468705293276109, + "grad_norm": 1.5650107685573749, + "learning_rate": 3.176769459136346e-05, + "loss": 1.3764, + "step": 8353 + }, + { + "epoch": 0.7469599427753935, + "grad_norm": 1.794755853419576, + "learning_rate": 3.174652614233222e-05, + "loss": 2.0016, + "step": 8354 + }, + { + "epoch": 0.7470493562231759, + "grad_norm": 1.244743747885071, + "learning_rate": 3.172536341763738e-05, + "loss": 1.7513, + "step": 8355 + }, + { + "epoch": 0.7471387696709585, + "grad_norm": 1.7787764737627363, + "learning_rate": 3.170420641905384e-05, + "loss": 1.6515, + "step": 8356 + }, + { + "epoch": 0.7472281831187411, + "grad_norm": 2.67688576254741, + "learning_rate": 3.1683055148356044e-05, + "loss": 1.9171, + "step": 8357 + }, + { + "epoch": 0.7473175965665236, + "grad_norm": 1.4222203073568485, + "learning_rate": 3.1661909607317894e-05, + "loss": 1.5576, + "step": 8358 + }, + { + "epoch": 0.7474070100143062, + "grad_norm": 0.7634831571062767, + "learning_rate": 3.164076979771287e-05, + "loss": 1.4863, + "step": 8359 + }, + { + "epoch": 0.7474964234620887, + "grad_norm": 1.8911136535676165, + "learning_rate": 3.161963572131393e-05, + "loss": 1.7571, + "step": 8360 + }, + { + "epoch": 0.7475858369098712, + "grad_norm": 2.0398737896920007, + "learning_rate": 3.159850737989355e-05, + "loss": 1.6227, + "step": 8361 + }, + { + "epoch": 0.7476752503576538, + "grad_norm": 2.2928582284995502, + "learning_rate": 3.1577384775223754e-05, + "loss": 1.7751, + "step": 8362 + }, + { + "epoch": 0.7477646638054364, + "grad_norm": 1.396273827743179, + "learning_rate": 3.1556267909076076e-05, + "loss": 1.7862, + "step": 8363 + }, + { + "epoch": 0.7478540772532188, + "grad_norm": 1.8437809461638612, + "learning_rate": 3.153515678322152e-05, + "loss": 1.807, + "step": 8364 + }, + { + "epoch": 0.7479434907010014, + "grad_norm": 3.145192190613082, + "learning_rate": 3.1514051399430654e-05, + "loss": 1.8189, + "step": 8365 + }, + { + "epoch": 0.748032904148784, + "grad_norm": 1.5820686764704317, + "learning_rate": 3.149295175947365e-05, + "loss": 1.8614, + "step": 8366 + }, + { + "epoch": 0.7481223175965666, + "grad_norm": 0.4741151695296759, + "learning_rate": 3.1471857865120016e-05, + "loss": 1.3344, + "step": 8367 + }, + { + "epoch": 0.748211731044349, + "grad_norm": 0.6714580530021468, + "learning_rate": 3.145076971813891e-05, + "loss": 1.505, + "step": 8368 + }, + { + "epoch": 0.7483011444921316, + "grad_norm": 3.015829979627591, + "learning_rate": 3.1429687320298976e-05, + "loss": 1.8559, + "step": 8369 + }, + { + "epoch": 0.7483905579399142, + "grad_norm": 1.5264799993351628, + "learning_rate": 3.1408610673368333e-05, + "loss": 1.8359, + "step": 8370 + }, + { + "epoch": 0.7484799713876967, + "grad_norm": 0.46446207574553156, + "learning_rate": 3.138753977911467e-05, + "loss": 1.4948, + "step": 8371 + }, + { + "epoch": 0.7485693848354793, + "grad_norm": 0.8726127863944164, + "learning_rate": 3.1366474639305185e-05, + "loss": 1.3526, + "step": 8372 + }, + { + "epoch": 0.7486587982832618, + "grad_norm": 1.768494935111213, + "learning_rate": 3.134541525570659e-05, + "loss": 1.799, + "step": 8373 + }, + { + "epoch": 0.7487482117310443, + "grad_norm": 1.7852124153819926, + "learning_rate": 3.132436163008512e-05, + "loss": 1.7594, + "step": 8374 + }, + { + "epoch": 0.7488376251788269, + "grad_norm": 0.6909052955928573, + "learning_rate": 3.1303313764206486e-05, + "loss": 1.759, + "step": 8375 + }, + { + "epoch": 0.7489270386266095, + "grad_norm": 0.9143919924672703, + "learning_rate": 3.1282271659835946e-05, + "loss": 1.7465, + "step": 8376 + }, + { + "epoch": 0.7490164520743919, + "grad_norm": 0.9733201730329738, + "learning_rate": 3.1261235318738336e-05, + "loss": 1.7962, + "step": 8377 + }, + { + "epoch": 0.7491058655221745, + "grad_norm": 1.7841955628821011, + "learning_rate": 3.124020474267787e-05, + "loss": 1.7544, + "step": 8378 + }, + { + "epoch": 0.7491952789699571, + "grad_norm": 0.7927580077582607, + "learning_rate": 3.1219179933418365e-05, + "loss": 1.7689, + "step": 8379 + }, + { + "epoch": 0.7492846924177397, + "grad_norm": 0.4938833226781699, + "learning_rate": 3.1198160892723225e-05, + "loss": 1.35, + "step": 8380 + }, + { + "epoch": 0.7493741058655222, + "grad_norm": 0.9562638314563402, + "learning_rate": 3.117714762235522e-05, + "loss": 1.5989, + "step": 8381 + }, + { + "epoch": 0.7494635193133047, + "grad_norm": 1.149821143133765, + "learning_rate": 3.1156140124076714e-05, + "loss": 1.878, + "step": 8382 + }, + { + "epoch": 0.7495529327610873, + "grad_norm": 0.4486798967313245, + "learning_rate": 3.113513839964963e-05, + "loss": 1.4564, + "step": 8383 + }, + { + "epoch": 0.7496423462088698, + "grad_norm": 0.41928428914970295, + "learning_rate": 3.1114142450835294e-05, + "loss": 1.636, + "step": 8384 + }, + { + "epoch": 0.7497317596566524, + "grad_norm": 0.8590484271263603, + "learning_rate": 3.1093152279394635e-05, + "loss": 1.7776, + "step": 8385 + }, + { + "epoch": 0.7498211731044349, + "grad_norm": 0.8325501545532243, + "learning_rate": 3.1072167887088065e-05, + "loss": 1.7368, + "step": 8386 + }, + { + "epoch": 0.7499105865522174, + "grad_norm": 0.4684364987003822, + "learning_rate": 3.105118927567554e-05, + "loss": 1.5961, + "step": 8387 + }, + { + "epoch": 0.75, + "grad_norm": 0.5380723752855503, + "learning_rate": 3.103021644691651e-05, + "loss": 1.5381, + "step": 8388 + }, + { + "epoch": 0.7500894134477826, + "grad_norm": 0.611414127733802, + "learning_rate": 3.1009249402569954e-05, + "loss": 1.7503, + "step": 8389 + }, + { + "epoch": 0.7501788268955651, + "grad_norm": 0.6090953771419809, + "learning_rate": 3.098828814439429e-05, + "loss": 1.6941, + "step": 8390 + }, + { + "epoch": 0.7502682403433476, + "grad_norm": 0.6666783559311001, + "learning_rate": 3.096733267414758e-05, + "loss": 1.4594, + "step": 8391 + }, + { + "epoch": 0.7503576537911302, + "grad_norm": 0.5699831252632362, + "learning_rate": 3.094638299358732e-05, + "loss": 1.6534, + "step": 8392 + }, + { + "epoch": 0.7504470672389127, + "grad_norm": 0.6417446596191463, + "learning_rate": 3.0925439104470456e-05, + "loss": 1.8908, + "step": 8393 + }, + { + "epoch": 0.7505364806866953, + "grad_norm": 0.7226856772622392, + "learning_rate": 3.090450100855367e-05, + "loss": 1.7398, + "step": 8394 + }, + { + "epoch": 0.7506258941344778, + "grad_norm": 0.41807768671707934, + "learning_rate": 3.08835687075929e-05, + "loss": 1.4988, + "step": 8395 + }, + { + "epoch": 0.7507153075822603, + "grad_norm": 0.7657998944227135, + "learning_rate": 3.086264220334375e-05, + "loss": 1.5216, + "step": 8396 + }, + { + "epoch": 0.7508047210300429, + "grad_norm": 1.0445133314130282, + "learning_rate": 3.084172149756134e-05, + "loss": 1.5175, + "step": 8397 + }, + { + "epoch": 0.7508941344778255, + "grad_norm": 0.5567773044424026, + "learning_rate": 3.082080659200018e-05, + "loss": 1.412, + "step": 8398 + }, + { + "epoch": 0.7509835479256081, + "grad_norm": 0.5296418401081049, + "learning_rate": 3.079989748841444e-05, + "loss": 1.8655, + "step": 8399 + }, + { + "epoch": 0.7510729613733905, + "grad_norm": 0.41612594520350576, + "learning_rate": 3.077899418855772e-05, + "loss": 1.4994, + "step": 8400 + }, + { + "epoch": 0.7511623748211731, + "grad_norm": 0.7172724269091894, + "learning_rate": 3.075809669418316e-05, + "loss": 1.2401, + "step": 8401 + }, + { + "epoch": 0.7512517882689557, + "grad_norm": 1.3241802238969076, + "learning_rate": 3.07372050070434e-05, + "loss": 1.6902, + "step": 8402 + }, + { + "epoch": 0.7513412017167382, + "grad_norm": 0.7947084264042591, + "learning_rate": 3.071631912889065e-05, + "loss": 1.8494, + "step": 8403 + }, + { + "epoch": 0.7514306151645207, + "grad_norm": 0.516276845801112, + "learning_rate": 3.0695439061476504e-05, + "loss": 1.7079, + "step": 8404 + }, + { + "epoch": 0.7515200286123033, + "grad_norm": 1.126016665196676, + "learning_rate": 3.0674564806552187e-05, + "loss": 1.5199, + "step": 8405 + }, + { + "epoch": 0.7516094420600858, + "grad_norm": 0.9115947225900446, + "learning_rate": 3.0653696365868424e-05, + "loss": 1.523, + "step": 8406 + }, + { + "epoch": 0.7516988555078684, + "grad_norm": 0.9803139373383086, + "learning_rate": 3.0632833741175336e-05, + "loss": 1.4988, + "step": 8407 + }, + { + "epoch": 0.751788268955651, + "grad_norm": 0.9135954337562726, + "learning_rate": 3.061197693422278e-05, + "loss": 1.8973, + "step": 8408 + }, + { + "epoch": 0.7518776824034334, + "grad_norm": 0.4533144198694893, + "learning_rate": 3.059112594675987e-05, + "loss": 2.1964, + "step": 8409 + }, + { + "epoch": 0.751967095851216, + "grad_norm": 0.7683676284710058, + "learning_rate": 3.057028078053541e-05, + "loss": 1.9462, + "step": 8410 + }, + { + "epoch": 0.7520565092989986, + "grad_norm": 0.8959440029775062, + "learning_rate": 3.054944143729769e-05, + "loss": 1.7166, + "step": 8411 + }, + { + "epoch": 0.7521459227467812, + "grad_norm": 0.693826593222945, + "learning_rate": 3.0528607918794395e-05, + "loss": 1.822, + "step": 8412 + }, + { + "epoch": 0.7522353361945636, + "grad_norm": 7.706873781640672, + "learning_rate": 3.0507780226772863e-05, + "loss": 6.294, + "step": 8413 + }, + { + "epoch": 0.7523247496423462, + "grad_norm": 1.6208284370093906, + "learning_rate": 3.048695836297988e-05, + "loss": 1.3678, + "step": 8414 + }, + { + "epoch": 0.7524141630901288, + "grad_norm": 2.653382699183966, + "learning_rate": 3.046614232916174e-05, + "loss": 1.5837, + "step": 8415 + }, + { + "epoch": 0.7525035765379113, + "grad_norm": 3.075815248767311, + "learning_rate": 3.0445332127064275e-05, + "loss": 1.8887, + "step": 8416 + }, + { + "epoch": 0.7525929899856938, + "grad_norm": 0.9137196242072962, + "learning_rate": 3.042452775843284e-05, + "loss": 1.8371, + "step": 8417 + }, + { + "epoch": 0.7526824034334764, + "grad_norm": 0.9199122668831776, + "learning_rate": 3.040372922501219e-05, + "loss": 1.5097, + "step": 8418 + }, + { + "epoch": 0.7527718168812589, + "grad_norm": 2.3928652807739805, + "learning_rate": 3.0382936528546735e-05, + "loss": 1.6084, + "step": 8419 + }, + { + "epoch": 0.7528612303290415, + "grad_norm": 3.4956800930917473, + "learning_rate": 3.036214967078034e-05, + "loss": 1.691, + "step": 8420 + }, + { + "epoch": 0.7529506437768241, + "grad_norm": 0.7857967832838739, + "learning_rate": 3.0341368653456283e-05, + "loss": 1.4104, + "step": 8421 + }, + { + "epoch": 0.7530400572246065, + "grad_norm": 1.0354954900084266, + "learning_rate": 3.032059347831755e-05, + "loss": 1.9268, + "step": 8422 + }, + { + "epoch": 0.7531294706723891, + "grad_norm": 2.06984572011281, + "learning_rate": 3.0299824147106516e-05, + "loss": 1.8224, + "step": 8423 + }, + { + "epoch": 0.7532188841201717, + "grad_norm": 1.0688789630073015, + "learning_rate": 3.0279060661565028e-05, + "loss": 1.325, + "step": 8424 + }, + { + "epoch": 0.7533082975679543, + "grad_norm": 1.154645861634598, + "learning_rate": 3.025830302343452e-05, + "loss": 1.5075, + "step": 8425 + }, + { + "epoch": 0.7533977110157367, + "grad_norm": 0.8942893418761617, + "learning_rate": 3.023755123445594e-05, + "loss": 1.9366, + "step": 8426 + }, + { + "epoch": 0.7534871244635193, + "grad_norm": 6.575767257914733, + "learning_rate": 3.0216805296369654e-05, + "loss": 6.4422, + "step": 8427 + }, + { + "epoch": 0.7535765379113019, + "grad_norm": 4.95744251881035, + "learning_rate": 3.0196065210915637e-05, + "loss": 1.9862, + "step": 8428 + }, + { + "epoch": 0.7536659513590844, + "grad_norm": 5.216487112896184, + "learning_rate": 3.017533097983333e-05, + "loss": 1.9613, + "step": 8429 + }, + { + "epoch": 0.753755364806867, + "grad_norm": 1.778030016435559, + "learning_rate": 3.015460260486168e-05, + "loss": 2.1451, + "step": 8430 + }, + { + "epoch": 0.7538447782546495, + "grad_norm": 1.501292181620695, + "learning_rate": 3.0133880087739184e-05, + "loss": 1.6808, + "step": 8431 + }, + { + "epoch": 0.753934191702432, + "grad_norm": 3.425232442244291, + "learning_rate": 3.0113163430203772e-05, + "loss": 1.7107, + "step": 8432 + }, + { + "epoch": 0.7540236051502146, + "grad_norm": 2.3573033252003976, + "learning_rate": 3.009245263399293e-05, + "loss": 1.488, + "step": 8433 + }, + { + "epoch": 0.7541130185979972, + "grad_norm": 2.8636003891256125, + "learning_rate": 3.0071747700843667e-05, + "loss": 1.8516, + "step": 8434 + }, + { + "epoch": 0.7542024320457796, + "grad_norm": 0.5788226926555212, + "learning_rate": 3.0051048632492463e-05, + "loss": 1.4339, + "step": 8435 + }, + { + "epoch": 0.7542918454935622, + "grad_norm": 2.208883078546732, + "learning_rate": 3.003035543067534e-05, + "loss": 1.9769, + "step": 8436 + }, + { + "epoch": 0.7543812589413448, + "grad_norm": 1.954059639328183, + "learning_rate": 3.000966809712783e-05, + "loss": 1.5883, + "step": 8437 + }, + { + "epoch": 0.7544706723891274, + "grad_norm": 2.7702362362338566, + "learning_rate": 2.9988986633584902e-05, + "loss": 2.1042, + "step": 8438 + }, + { + "epoch": 0.7545600858369099, + "grad_norm": 1.0258947891347818, + "learning_rate": 2.9968311041781116e-05, + "loss": 1.4696, + "step": 8439 + }, + { + "epoch": 0.7546494992846924, + "grad_norm": 1.5159912564191131, + "learning_rate": 2.9947641323450535e-05, + "loss": 1.3971, + "step": 8440 + }, + { + "epoch": 0.754738912732475, + "grad_norm": 1.4701708691832096, + "learning_rate": 2.992697748032661e-05, + "loss": 1.5436, + "step": 8441 + }, + { + "epoch": 0.7548283261802575, + "grad_norm": 1.7279125257405124, + "learning_rate": 2.990631951414252e-05, + "loss": 1.4169, + "step": 8442 + }, + { + "epoch": 0.7549177396280401, + "grad_norm": 1.8488203212042278, + "learning_rate": 2.9885667426630737e-05, + "loss": 1.7835, + "step": 8443 + }, + { + "epoch": 0.7550071530758226, + "grad_norm": 1.2557261466338636, + "learning_rate": 2.9865021219523337e-05, + "loss": 1.5431, + "step": 8444 + }, + { + "epoch": 0.7550965665236051, + "grad_norm": 1.2767076230278882, + "learning_rate": 2.9844380894551916e-05, + "loss": 2.0671, + "step": 8445 + }, + { + "epoch": 0.7551859799713877, + "grad_norm": 1.3266541777472163, + "learning_rate": 2.9823746453447565e-05, + "loss": 1.5973, + "step": 8446 + }, + { + "epoch": 0.7552753934191703, + "grad_norm": 1.4179802630529286, + "learning_rate": 2.9803117897940826e-05, + "loss": 1.5859, + "step": 8447 + }, + { + "epoch": 0.7553648068669528, + "grad_norm": 0.8695212292123699, + "learning_rate": 2.9782495229761808e-05, + "loss": 1.4184, + "step": 8448 + }, + { + "epoch": 0.7554542203147353, + "grad_norm": 1.6987945679718413, + "learning_rate": 2.9761878450640112e-05, + "loss": 1.5109, + "step": 8449 + }, + { + "epoch": 0.7555436337625179, + "grad_norm": 1.0247163436192608, + "learning_rate": 2.9741267562304854e-05, + "loss": 1.9026, + "step": 8450 + }, + { + "epoch": 0.7556330472103004, + "grad_norm": 1.6393335506318591, + "learning_rate": 2.972066256648465e-05, + "loss": 1.4599, + "step": 8451 + }, + { + "epoch": 0.755722460658083, + "grad_norm": 1.416021990683106, + "learning_rate": 2.9700063464907578e-05, + "loss": 1.9734, + "step": 8452 + }, + { + "epoch": 0.7558118741058655, + "grad_norm": 0.5109959924865557, + "learning_rate": 2.967947025930128e-05, + "loss": 1.3917, + "step": 8453 + }, + { + "epoch": 0.755901287553648, + "grad_norm": 0.513468207402394, + "learning_rate": 2.9658882951392918e-05, + "loss": 1.5382, + "step": 8454 + }, + { + "epoch": 0.7559907010014306, + "grad_norm": 1.376529400296604, + "learning_rate": 2.963830154290903e-05, + "loss": 1.8525, + "step": 8455 + }, + { + "epoch": 0.7560801144492132, + "grad_norm": 1.3370143695005114, + "learning_rate": 2.9617726035575855e-05, + "loss": 1.6897, + "step": 8456 + }, + { + "epoch": 0.7561695278969958, + "grad_norm": 0.48364804654633986, + "learning_rate": 2.9597156431119023e-05, + "loss": 1.5242, + "step": 8457 + }, + { + "epoch": 0.7562589413447782, + "grad_norm": 0.6404852928336997, + "learning_rate": 2.957659273126362e-05, + "loss": 1.6768, + "step": 8458 + }, + { + "epoch": 0.7563483547925608, + "grad_norm": 0.856584175002122, + "learning_rate": 2.9556034937734332e-05, + "loss": 1.6964, + "step": 8459 + }, + { + "epoch": 0.7564377682403434, + "grad_norm": 0.5934540703462905, + "learning_rate": 2.9535483052255365e-05, + "loss": 1.6241, + "step": 8460 + }, + { + "epoch": 0.7565271816881259, + "grad_norm": 0.4373928213405828, + "learning_rate": 2.9514937076550286e-05, + "loss": 1.6514, + "step": 8461 + }, + { + "epoch": 0.7566165951359084, + "grad_norm": 0.5551720521085657, + "learning_rate": 2.9494397012342322e-05, + "loss": 1.6358, + "step": 8462 + }, + { + "epoch": 0.756706008583691, + "grad_norm": 0.9191226877041923, + "learning_rate": 2.9473862861354128e-05, + "loss": 1.7572, + "step": 8463 + }, + { + "epoch": 0.7567954220314735, + "grad_norm": 0.5869740151936041, + "learning_rate": 2.945333462530788e-05, + "loss": 1.5003, + "step": 8464 + }, + { + "epoch": 0.7568848354792561, + "grad_norm": 0.9596028958438706, + "learning_rate": 2.9432812305925295e-05, + "loss": 2.4031, + "step": 8465 + }, + { + "epoch": 0.7569742489270386, + "grad_norm": 0.5687949998058801, + "learning_rate": 2.941229590492748e-05, + "loss": 1.6312, + "step": 8466 + }, + { + "epoch": 0.7570636623748211, + "grad_norm": 0.6478024317916221, + "learning_rate": 2.9391785424035167e-05, + "loss": 1.8318, + "step": 8467 + }, + { + "epoch": 0.7571530758226037, + "grad_norm": 0.5042486651279203, + "learning_rate": 2.9371280864968565e-05, + "loss": 1.546, + "step": 8468 + }, + { + "epoch": 0.7572424892703863, + "grad_norm": 0.5009683505951745, + "learning_rate": 2.935078222944727e-05, + "loss": 1.5081, + "step": 8469 + }, + { + "epoch": 0.7573319027181689, + "grad_norm": 0.44767739123409783, + "learning_rate": 2.933028951919058e-05, + "loss": 1.3834, + "step": 8470 + }, + { + "epoch": 0.7574213161659513, + "grad_norm": 0.6963143691458261, + "learning_rate": 2.93098027359172e-05, + "loss": 1.8498, + "step": 8471 + }, + { + "epoch": 0.7575107296137339, + "grad_norm": 0.528909510961551, + "learning_rate": 2.9289321881345254e-05, + "loss": 1.5654, + "step": 8472 + }, + { + "epoch": 0.7576001430615165, + "grad_norm": 0.5438552699781166, + "learning_rate": 2.9268846957192485e-05, + "loss": 1.8334, + "step": 8473 + }, + { + "epoch": 0.757689556509299, + "grad_norm": 0.5330663379269381, + "learning_rate": 2.9248377965176134e-05, + "loss": 1.2179, + "step": 8474 + }, + { + "epoch": 0.7577789699570815, + "grad_norm": 0.6054341791207303, + "learning_rate": 2.9227914907012845e-05, + "loss": 1.7896, + "step": 8475 + }, + { + "epoch": 0.7578683834048641, + "grad_norm": 0.41324780907177794, + "learning_rate": 2.9207457784418835e-05, + "loss": 1.2122, + "step": 8476 + }, + { + "epoch": 0.7579577968526466, + "grad_norm": 0.42224225023085576, + "learning_rate": 2.91870065991099e-05, + "loss": 1.317, + "step": 8477 + }, + { + "epoch": 0.7580472103004292, + "grad_norm": 1.144182142607541, + "learning_rate": 2.9166561352801182e-05, + "loss": 1.7485, + "step": 8478 + }, + { + "epoch": 0.7581366237482118, + "grad_norm": 0.7544734978365456, + "learning_rate": 2.91461220472074e-05, + "loss": 1.4005, + "step": 8479 + }, + { + "epoch": 0.7582260371959942, + "grad_norm": 0.7051911286332851, + "learning_rate": 2.912568868404284e-05, + "loss": 1.8249, + "step": 8480 + }, + { + "epoch": 0.7583154506437768, + "grad_norm": 1.010466918316333, + "learning_rate": 2.9105261265021133e-05, + "loss": 1.7054, + "step": 8481 + }, + { + "epoch": 0.7584048640915594, + "grad_norm": 0.6521202265564325, + "learning_rate": 2.9084839791855544e-05, + "loss": 1.687, + "step": 8482 + }, + { + "epoch": 0.758494277539342, + "grad_norm": 0.6574809103840147, + "learning_rate": 2.9064424266258805e-05, + "loss": 1.8391, + "step": 8483 + }, + { + "epoch": 0.7585836909871244, + "grad_norm": 0.6026087549408757, + "learning_rate": 2.9044014689943132e-05, + "loss": 1.7405, + "step": 8484 + }, + { + "epoch": 0.758673104434907, + "grad_norm": 1.0664795315871909, + "learning_rate": 2.902361106462028e-05, + "loss": 1.8989, + "step": 8485 + }, + { + "epoch": 0.7587625178826896, + "grad_norm": 0.6711005549558695, + "learning_rate": 2.9003213392001426e-05, + "loss": 1.4745, + "step": 8486 + }, + { + "epoch": 0.7588519313304721, + "grad_norm": 0.5989808138218232, + "learning_rate": 2.8982821673797322e-05, + "loss": 1.8787, + "step": 8487 + }, + { + "epoch": 0.7589413447782547, + "grad_norm": 0.5279421936457401, + "learning_rate": 2.8962435911718222e-05, + "loss": 1.6438, + "step": 8488 + }, + { + "epoch": 0.7590307582260372, + "grad_norm": 0.8877464234486624, + "learning_rate": 2.8942056107473802e-05, + "loss": 1.8541, + "step": 8489 + }, + { + "epoch": 0.7591201716738197, + "grad_norm": 0.40018297122269536, + "learning_rate": 2.89216822627733e-05, + "loss": 1.451, + "step": 8490 + }, + { + "epoch": 0.7592095851216023, + "grad_norm": 0.8940256018535166, + "learning_rate": 2.8901314379325517e-05, + "loss": 1.7375, + "step": 8491 + }, + { + "epoch": 0.7592989985693849, + "grad_norm": 0.6016597389961027, + "learning_rate": 2.8880952458838593e-05, + "loss": 1.871, + "step": 8492 + }, + { + "epoch": 0.7593884120171673, + "grad_norm": 0.9395355959214781, + "learning_rate": 2.886059650302031e-05, + "loss": 1.7204, + "step": 8493 + }, + { + "epoch": 0.7594778254649499, + "grad_norm": 0.3580579904026937, + "learning_rate": 2.8840246513577907e-05, + "loss": 1.3548, + "step": 8494 + }, + { + "epoch": 0.7595672389127325, + "grad_norm": 1.2204866799209453, + "learning_rate": 2.8819902492218066e-05, + "loss": 1.9987, + "step": 8495 + }, + { + "epoch": 0.759656652360515, + "grad_norm": 0.44266653678751894, + "learning_rate": 2.879956444064703e-05, + "loss": 1.415, + "step": 8496 + }, + { + "epoch": 0.7597460658082976, + "grad_norm": 0.36781329583434874, + "learning_rate": 2.877923236057054e-05, + "loss": 1.443, + "step": 8497 + }, + { + "epoch": 0.7598354792560801, + "grad_norm": 0.7946742840754094, + "learning_rate": 2.8758906253693818e-05, + "loss": 1.6011, + "step": 8498 + }, + { + "epoch": 0.7599248927038627, + "grad_norm": 4.883496132835038, + "learning_rate": 2.8738586121721634e-05, + "loss": 6.3055, + "step": 8499 + }, + { + "epoch": 0.7600143061516452, + "grad_norm": 0.9504771075059435, + "learning_rate": 2.8718271966358124e-05, + "loss": 1.3896, + "step": 8500 + }, + { + "epoch": 0.7601037195994278, + "grad_norm": 1.873231258685163, + "learning_rate": 2.869796378930706e-05, + "loss": 1.6873, + "step": 8501 + }, + { + "epoch": 0.7601931330472103, + "grad_norm": 1.1662661475341753, + "learning_rate": 2.8677661592271666e-05, + "loss": 1.5391, + "step": 8502 + }, + { + "epoch": 0.7602825464949928, + "grad_norm": 1.8343429449992874, + "learning_rate": 2.8657365376954692e-05, + "loss": 2.0005, + "step": 8503 + }, + { + "epoch": 0.7603719599427754, + "grad_norm": 0.8907849344399054, + "learning_rate": 2.8637075145058257e-05, + "loss": 2.0684, + "step": 8504 + }, + { + "epoch": 0.760461373390558, + "grad_norm": 1.8917730144494271, + "learning_rate": 2.8616790898284207e-05, + "loss": 1.9676, + "step": 8505 + }, + { + "epoch": 0.7605507868383404, + "grad_norm": 1.8484657855865911, + "learning_rate": 2.859651263833366e-05, + "loss": 1.8488, + "step": 8506 + }, + { + "epoch": 0.760640200286123, + "grad_norm": 0.7836827031398331, + "learning_rate": 2.857624036690737e-05, + "loss": 1.652, + "step": 8507 + }, + { + "epoch": 0.7607296137339056, + "grad_norm": 0.7312744877132079, + "learning_rate": 2.8555974085705573e-05, + "loss": 1.6676, + "step": 8508 + }, + { + "epoch": 0.7608190271816881, + "grad_norm": 0.9566606167582373, + "learning_rate": 2.853571379642792e-05, + "loss": 1.7493, + "step": 8509 + }, + { + "epoch": 0.7609084406294707, + "grad_norm": 1.3970867950371835, + "learning_rate": 2.8515459500773633e-05, + "loss": 1.569, + "step": 8510 + }, + { + "epoch": 0.7609978540772532, + "grad_norm": 0.5886845560461152, + "learning_rate": 2.849521120044144e-05, + "loss": 1.3159, + "step": 8511 + }, + { + "epoch": 0.7610872675250357, + "grad_norm": 1.043016076777894, + "learning_rate": 2.847496889712952e-05, + "loss": 1.7236, + "step": 8512 + }, + { + "epoch": 0.7611766809728183, + "grad_norm": 0.4966861925110457, + "learning_rate": 2.845473259253557e-05, + "loss": 1.6604, + "step": 8513 + }, + { + "epoch": 0.7612660944206009, + "grad_norm": 0.8550610922731635, + "learning_rate": 2.8434502288356835e-05, + "loss": 2.0058, + "step": 8514 + }, + { + "epoch": 0.7613555078683834, + "grad_norm": 1.0379840766425446, + "learning_rate": 2.8414277986289928e-05, + "loss": 1.7292, + "step": 8515 + }, + { + "epoch": 0.7614449213161659, + "grad_norm": 0.9132764952614656, + "learning_rate": 2.839405968803108e-05, + "loss": 1.8459, + "step": 8516 + }, + { + "epoch": 0.7615343347639485, + "grad_norm": 0.49300085533012583, + "learning_rate": 2.8373847395275966e-05, + "loss": 1.5819, + "step": 8517 + }, + { + "epoch": 0.7616237482117311, + "grad_norm": 0.5368470379614031, + "learning_rate": 2.8353641109719764e-05, + "loss": 1.5898, + "step": 8518 + }, + { + "epoch": 0.7617131616595136, + "grad_norm": 0.9742300885551047, + "learning_rate": 2.833344083305719e-05, + "loss": 1.6843, + "step": 8519 + }, + { + "epoch": 0.7618025751072961, + "grad_norm": 1.0110160477582728, + "learning_rate": 2.8313246566982345e-05, + "loss": 1.944, + "step": 8520 + }, + { + "epoch": 0.7618919885550787, + "grad_norm": 0.46608444478335614, + "learning_rate": 2.8293058313188935e-05, + "loss": 1.4124, + "step": 8521 + }, + { + "epoch": 0.7619814020028612, + "grad_norm": 5.09795464490002, + "learning_rate": 2.827287607337016e-05, + "loss": 6.3072, + "step": 8522 + }, + { + "epoch": 0.7620708154506438, + "grad_norm": 1.3818094582323235, + "learning_rate": 2.8252699849218613e-05, + "loss": 1.4251, + "step": 8523 + }, + { + "epoch": 0.7621602288984263, + "grad_norm": 2.1105153724218733, + "learning_rate": 2.823252964242644e-05, + "loss": 1.4586, + "step": 8524 + }, + { + "epoch": 0.7622496423462088, + "grad_norm": 2.6404806952602957, + "learning_rate": 2.8212365454685408e-05, + "loss": 1.6867, + "step": 8525 + }, + { + "epoch": 0.7623390557939914, + "grad_norm": 1.0051426210408159, + "learning_rate": 2.8192207287686555e-05, + "loss": 1.7056, + "step": 8526 + }, + { + "epoch": 0.762428469241774, + "grad_norm": 0.9512454160801639, + "learning_rate": 2.8172055143120546e-05, + "loss": 1.5887, + "step": 8527 + }, + { + "epoch": 0.7625178826895566, + "grad_norm": 1.4449336927286358, + "learning_rate": 2.815190902267757e-05, + "loss": 1.4906, + "step": 8528 + }, + { + "epoch": 0.762607296137339, + "grad_norm": 1.5290720898474646, + "learning_rate": 2.8131768928047176e-05, + "loss": 1.4322, + "step": 8529 + }, + { + "epoch": 0.7626967095851216, + "grad_norm": 1.7133447252857035, + "learning_rate": 2.8111634860918524e-05, + "loss": 1.5561, + "step": 8530 + }, + { + "epoch": 0.7627861230329042, + "grad_norm": 0.6200044154305522, + "learning_rate": 2.809150682298024e-05, + "loss": 1.5725, + "step": 8531 + }, + { + "epoch": 0.7628755364806867, + "grad_norm": 0.555086329586969, + "learning_rate": 2.807138481592043e-05, + "loss": 1.291, + "step": 8532 + }, + { + "epoch": 0.7629649499284692, + "grad_norm": 2.4774447140464924, + "learning_rate": 2.8051268841426713e-05, + "loss": 1.7701, + "step": 8533 + }, + { + "epoch": 0.7630543633762518, + "grad_norm": 1.141359955866877, + "learning_rate": 2.803115890118623e-05, + "loss": 1.5668, + "step": 8534 + }, + { + "epoch": 0.7631437768240343, + "grad_norm": 0.9888822549422854, + "learning_rate": 2.8011054996885477e-05, + "loss": 1.3946, + "step": 8535 + }, + { + "epoch": 0.7632331902718169, + "grad_norm": 0.9948008643294065, + "learning_rate": 2.7990957130210617e-05, + "loss": 1.7384, + "step": 8536 + }, + { + "epoch": 0.7633226037195995, + "grad_norm": 1.151166180815625, + "learning_rate": 2.797086530284725e-05, + "loss": 1.6189, + "step": 8537 + }, + { + "epoch": 0.7634120171673819, + "grad_norm": 1.2461595911993737, + "learning_rate": 2.795077951648035e-05, + "loss": 1.7333, + "step": 8538 + }, + { + "epoch": 0.7635014306151645, + "grad_norm": 0.7090170884622458, + "learning_rate": 2.7930699772794623e-05, + "loss": 1.5916, + "step": 8539 + }, + { + "epoch": 0.7635908440629471, + "grad_norm": 1.0591460415960083, + "learning_rate": 2.7910626073474045e-05, + "loss": 1.8136, + "step": 8540 + }, + { + "epoch": 0.7636802575107297, + "grad_norm": 1.4230966535784983, + "learning_rate": 2.7890558420202185e-05, + "loss": 1.6216, + "step": 8541 + }, + { + "epoch": 0.7637696709585121, + "grad_norm": 0.9690213022495549, + "learning_rate": 2.787049681466214e-05, + "loss": 1.504, + "step": 8542 + }, + { + "epoch": 0.7638590844062947, + "grad_norm": 0.37202381924437006, + "learning_rate": 2.7850441258536386e-05, + "loss": 1.2724, + "step": 8543 + }, + { + "epoch": 0.7639484978540773, + "grad_norm": 0.7916545957131075, + "learning_rate": 2.783039175350699e-05, + "loss": 1.9451, + "step": 8544 + }, + { + "epoch": 0.7640379113018598, + "grad_norm": 0.9157446865893206, + "learning_rate": 2.7810348301255486e-05, + "loss": 1.404, + "step": 8545 + }, + { + "epoch": 0.7641273247496424, + "grad_norm": 0.39328907011016434, + "learning_rate": 2.779031090346287e-05, + "loss": 1.3786, + "step": 8546 + }, + { + "epoch": 0.7642167381974249, + "grad_norm": 0.5183271258673819, + "learning_rate": 2.7770279561809686e-05, + "loss": 1.7406, + "step": 8547 + }, + { + "epoch": 0.7643061516452074, + "grad_norm": 0.8429283804243837, + "learning_rate": 2.775025427797594e-05, + "loss": 1.797, + "step": 8548 + }, + { + "epoch": 0.76439556509299, + "grad_norm": 1.3168328431009497, + "learning_rate": 2.7730235053641096e-05, + "loss": 1.4343, + "step": 8549 + }, + { + "epoch": 0.7644849785407726, + "grad_norm": 0.8637410534219434, + "learning_rate": 2.7710221890484157e-05, + "loss": 1.8164, + "step": 8550 + }, + { + "epoch": 0.764574391988555, + "grad_norm": 0.32829331879379214, + "learning_rate": 2.7690214790183622e-05, + "loss": 1.3792, + "step": 8551 + }, + { + "epoch": 0.7646638054363376, + "grad_norm": 0.6830538496560611, + "learning_rate": 2.7670213754417396e-05, + "loss": 1.5867, + "step": 8552 + }, + { + "epoch": 0.7647532188841202, + "grad_norm": 0.4716268506182502, + "learning_rate": 2.7650218784863047e-05, + "loss": 1.7907, + "step": 8553 + }, + { + "epoch": 0.7648426323319027, + "grad_norm": 0.9519200022033076, + "learning_rate": 2.7630229883197433e-05, + "loss": 2.2145, + "step": 8554 + }, + { + "epoch": 0.7649320457796852, + "grad_norm": 0.6331476033650256, + "learning_rate": 2.761024705109705e-05, + "loss": 1.6571, + "step": 8555 + }, + { + "epoch": 0.7650214592274678, + "grad_norm": 0.7833371469691145, + "learning_rate": 2.7590270290237852e-05, + "loss": 1.6185, + "step": 8556 + }, + { + "epoch": 0.7651108726752504, + "grad_norm": 0.6501309955042138, + "learning_rate": 2.75702996022952e-05, + "loss": 1.9103, + "step": 8557 + }, + { + "epoch": 0.7652002861230329, + "grad_norm": 0.6163163137068747, + "learning_rate": 2.755033498894405e-05, + "loss": 1.6123, + "step": 8558 + }, + { + "epoch": 0.7652896995708155, + "grad_norm": 0.5170341623038027, + "learning_rate": 2.7530376451858807e-05, + "loss": 1.4509, + "step": 8559 + }, + { + "epoch": 0.765379113018598, + "grad_norm": 0.7780072362255823, + "learning_rate": 2.7510423992713374e-05, + "loss": 1.5837, + "step": 8560 + }, + { + "epoch": 0.7654685264663805, + "grad_norm": 0.47268152792794, + "learning_rate": 2.749047761318113e-05, + "loss": 1.421, + "step": 8561 + }, + { + "epoch": 0.7655579399141631, + "grad_norm": 0.5647400129810027, + "learning_rate": 2.7470537314934997e-05, + "loss": 1.3865, + "step": 8562 + }, + { + "epoch": 0.7656473533619457, + "grad_norm": 0.4606966913634567, + "learning_rate": 2.7450603099647266e-05, + "loss": 1.3319, + "step": 8563 + }, + { + "epoch": 0.7657367668097281, + "grad_norm": 0.9139451896380226, + "learning_rate": 2.7430674968989832e-05, + "loss": 1.8494, + "step": 8564 + }, + { + "epoch": 0.7658261802575107, + "grad_norm": 1.1702029724976653, + "learning_rate": 2.7410752924634088e-05, + "loss": 1.4246, + "step": 8565 + }, + { + "epoch": 0.7659155937052933, + "grad_norm": 0.7954862308755053, + "learning_rate": 2.7390836968250766e-05, + "loss": 2.0955, + "step": 8566 + }, + { + "epoch": 0.7660050071530758, + "grad_norm": 0.7774007185292194, + "learning_rate": 2.737092710151029e-05, + "loss": 1.7387, + "step": 8567 + }, + { + "epoch": 0.7660944206008584, + "grad_norm": 0.61457595598191, + "learning_rate": 2.735102332608247e-05, + "loss": 1.7209, + "step": 8568 + }, + { + "epoch": 0.7661838340486409, + "grad_norm": 0.5065990972637557, + "learning_rate": 2.7331125643636567e-05, + "loss": 1.6268, + "step": 8569 + }, + { + "epoch": 0.7662732474964234, + "grad_norm": 0.6950625336678604, + "learning_rate": 2.7311234055841382e-05, + "loss": 1.8173, + "step": 8570 + }, + { + "epoch": 0.766362660944206, + "grad_norm": 1.0384258735760663, + "learning_rate": 2.7291348564365248e-05, + "loss": 1.6747, + "step": 8571 + }, + { + "epoch": 0.7664520743919886, + "grad_norm": 0.46918031014335954, + "learning_rate": 2.7271469170875863e-05, + "loss": 1.5409, + "step": 8572 + }, + { + "epoch": 0.766541487839771, + "grad_norm": 0.5064682981025816, + "learning_rate": 2.7251595877040538e-05, + "loss": 1.4203, + "step": 8573 + }, + { + "epoch": 0.7666309012875536, + "grad_norm": 0.4595781347113193, + "learning_rate": 2.7231728684525992e-05, + "loss": 1.5834, + "step": 8574 + }, + { + "epoch": 0.7667203147353362, + "grad_norm": 0.5350278013801475, + "learning_rate": 2.7211867594998486e-05, + "loss": 1.7888, + "step": 8575 + }, + { + "epoch": 0.7668097281831188, + "grad_norm": 0.6011183884018443, + "learning_rate": 2.7192012610123774e-05, + "loss": 1.2265, + "step": 8576 + }, + { + "epoch": 0.7668991416309013, + "grad_norm": 0.833897230517596, + "learning_rate": 2.7172163731567e-05, + "loss": 1.6094, + "step": 8577 + }, + { + "epoch": 0.7669885550786838, + "grad_norm": 0.6551023302677387, + "learning_rate": 2.7152320960992905e-05, + "loss": 1.6966, + "step": 8578 + }, + { + "epoch": 0.7670779685264664, + "grad_norm": 0.40032341688813, + "learning_rate": 2.71324843000657e-05, + "loss": 1.7629, + "step": 8579 + }, + { + "epoch": 0.7671673819742489, + "grad_norm": 0.5176678470141177, + "learning_rate": 2.711265375044897e-05, + "loss": 1.6939, + "step": 8580 + }, + { + "epoch": 0.7672567954220315, + "grad_norm": 0.7839144757325142, + "learning_rate": 2.709282931380598e-05, + "loss": 1.8428, + "step": 8581 + }, + { + "epoch": 0.767346208869814, + "grad_norm": 0.6780705622955636, + "learning_rate": 2.7073010991799376e-05, + "loss": 1.6488, + "step": 8582 + }, + { + "epoch": 0.7674356223175965, + "grad_norm": 0.5987235128578955, + "learning_rate": 2.705319878609124e-05, + "loss": 1.3253, + "step": 8583 + }, + { + "epoch": 0.7675250357653791, + "grad_norm": 0.6565519351100396, + "learning_rate": 2.7033392698343218e-05, + "loss": 1.7647, + "step": 8584 + }, + { + "epoch": 0.7676144492131617, + "grad_norm": 0.9533014923865274, + "learning_rate": 2.7013592730216465e-05, + "loss": 1.5955, + "step": 8585 + }, + { + "epoch": 0.7677038626609443, + "grad_norm": 0.5461138547359775, + "learning_rate": 2.69937988833715e-05, + "loss": 1.9525, + "step": 8586 + }, + { + "epoch": 0.7677932761087267, + "grad_norm": 0.651513526800422, + "learning_rate": 2.697401115946847e-05, + "loss": 1.7594, + "step": 8587 + }, + { + "epoch": 0.7678826895565093, + "grad_norm": 0.5163520233754659, + "learning_rate": 2.6954229560166923e-05, + "loss": 1.7329, + "step": 8588 + }, + { + "epoch": 0.7679721030042919, + "grad_norm": 1.0168542716471598, + "learning_rate": 2.6934454087125926e-05, + "loss": 1.7314, + "step": 8589 + }, + { + "epoch": 0.7680615164520744, + "grad_norm": 1.1624711374225536, + "learning_rate": 2.6914684742004028e-05, + "loss": 1.9747, + "step": 8590 + }, + { + "epoch": 0.7681509298998569, + "grad_norm": 0.8492213171259236, + "learning_rate": 2.689492152645928e-05, + "loss": 1.6687, + "step": 8591 + }, + { + "epoch": 0.7682403433476395, + "grad_norm": 1.163602574872497, + "learning_rate": 2.6875164442149147e-05, + "loss": 2.0615, + "step": 8592 + }, + { + "epoch": 0.768329756795422, + "grad_norm": 1.2988306472181699, + "learning_rate": 2.685541349073066e-05, + "loss": 1.7459, + "step": 8593 + }, + { + "epoch": 0.7684191702432046, + "grad_norm": 0.5470315564581506, + "learning_rate": 2.6835668673860314e-05, + "loss": 1.4319, + "step": 8594 + }, + { + "epoch": 0.7685085836909872, + "grad_norm": 0.7531559055115571, + "learning_rate": 2.6815929993194067e-05, + "loss": 1.7791, + "step": 8595 + }, + { + "epoch": 0.7685979971387696, + "grad_norm": 0.5604950114103617, + "learning_rate": 2.679619745038743e-05, + "loss": 1.8492, + "step": 8596 + }, + { + "epoch": 0.7686874105865522, + "grad_norm": 0.8973565832344739, + "learning_rate": 2.6776471047095263e-05, + "loss": 1.6726, + "step": 8597 + }, + { + "epoch": 0.7687768240343348, + "grad_norm": 1.5527697443256476, + "learning_rate": 2.675675078497204e-05, + "loss": 1.8776, + "step": 8598 + }, + { + "epoch": 0.7688662374821174, + "grad_norm": 0.8157716325494735, + "learning_rate": 2.67370366656717e-05, + "loss": 1.4167, + "step": 8599 + }, + { + "epoch": 0.7689556509298998, + "grad_norm": 0.6199178327428765, + "learning_rate": 2.6717328690847565e-05, + "loss": 1.629, + "step": 8600 + }, + { + "epoch": 0.7690450643776824, + "grad_norm": 1.319258478574011, + "learning_rate": 2.669762686215259e-05, + "loss": 1.6607, + "step": 8601 + }, + { + "epoch": 0.769134477825465, + "grad_norm": 6.014399109576478, + "learning_rate": 2.6677931181239158e-05, + "loss": 6.2559, + "step": 8602 + }, + { + "epoch": 0.7692238912732475, + "grad_norm": 2.1660259193405347, + "learning_rate": 2.6658241649759062e-05, + "loss": 6.2151, + "step": 8603 + }, + { + "epoch": 0.76931330472103, + "grad_norm": 1.2510778635159052, + "learning_rate": 2.6638558269363654e-05, + "loss": 1.6286, + "step": 8604 + }, + { + "epoch": 0.7694027181688126, + "grad_norm": 3.79391221122697, + "learning_rate": 2.6618881041703804e-05, + "loss": 1.8911, + "step": 8605 + }, + { + "epoch": 0.7694921316165951, + "grad_norm": 6.16620316348517, + "learning_rate": 2.659920996842975e-05, + "loss": 6.4471, + "step": 8606 + }, + { + "epoch": 0.7695815450643777, + "grad_norm": 0.9999643374188395, + "learning_rate": 2.6579545051191302e-05, + "loss": 1.7421, + "step": 8607 + }, + { + "epoch": 0.7696709585121603, + "grad_norm": 1.987175056626487, + "learning_rate": 2.6559886291637748e-05, + "loss": 1.3845, + "step": 8608 + }, + { + "epoch": 0.7697603719599427, + "grad_norm": 3.5657763879948927, + "learning_rate": 2.6540233691417837e-05, + "loss": 1.6268, + "step": 8609 + }, + { + "epoch": 0.7698497854077253, + "grad_norm": 3.441751786918885, + "learning_rate": 2.652058725217983e-05, + "loss": 1.6998, + "step": 8610 + }, + { + "epoch": 0.7699391988555079, + "grad_norm": 0.9746029088140326, + "learning_rate": 2.6500946975571405e-05, + "loss": 1.7472, + "step": 8611 + }, + { + "epoch": 0.7700286123032904, + "grad_norm": 1.418861327123767, + "learning_rate": 2.6481312863239804e-05, + "loss": 1.6931, + "step": 8612 + }, + { + "epoch": 0.7701180257510729, + "grad_norm": 2.1584110754666503, + "learning_rate": 2.646168491683172e-05, + "loss": 1.5355, + "step": 8613 + }, + { + "epoch": 0.7702074391988555, + "grad_norm": 2.7233883152843377, + "learning_rate": 2.6442063137993255e-05, + "loss": 1.6103, + "step": 8614 + }, + { + "epoch": 0.770296852646638, + "grad_norm": 1.4980637665994616, + "learning_rate": 2.6422447528370152e-05, + "loss": 1.3239, + "step": 8615 + }, + { + "epoch": 0.7703862660944206, + "grad_norm": 0.666509192668997, + "learning_rate": 2.640283808960754e-05, + "loss": 1.6137, + "step": 8616 + }, + { + "epoch": 0.7704756795422032, + "grad_norm": 1.3652061632419623, + "learning_rate": 2.638323482334999e-05, + "loss": 1.827, + "step": 8617 + }, + { + "epoch": 0.7705650929899857, + "grad_norm": 1.986077929416074, + "learning_rate": 2.636363773124163e-05, + "loss": 2.0122, + "step": 8618 + }, + { + "epoch": 0.7706545064377682, + "grad_norm": 1.745314028753051, + "learning_rate": 2.634404681492607e-05, + "loss": 1.7059, + "step": 8619 + }, + { + "epoch": 0.7707439198855508, + "grad_norm": 0.6229578829565182, + "learning_rate": 2.6324462076046318e-05, + "loss": 1.584, + "step": 8620 + }, + { + "epoch": 0.7708333333333334, + "grad_norm": 0.6812875756071326, + "learning_rate": 2.630488351624496e-05, + "loss": 1.5373, + "step": 8621 + }, + { + "epoch": 0.7709227467811158, + "grad_norm": 1.3056618499810466, + "learning_rate": 2.6285311137164013e-05, + "loss": 1.8635, + "step": 8622 + }, + { + "epoch": 0.7710121602288984, + "grad_norm": 0.6377671510337007, + "learning_rate": 2.6265744940445003e-05, + "loss": 1.5377, + "step": 8623 + }, + { + "epoch": 0.771101573676681, + "grad_norm": 1.1723957942212668, + "learning_rate": 2.624618492772891e-05, + "loss": 1.7618, + "step": 8624 + }, + { + "epoch": 0.7711909871244635, + "grad_norm": 0.611609337066035, + "learning_rate": 2.622663110065625e-05, + "loss": 1.5428, + "step": 8625 + }, + { + "epoch": 0.7712804005722461, + "grad_norm": 0.5478729558113091, + "learning_rate": 2.6207083460866912e-05, + "loss": 1.4623, + "step": 8626 + }, + { + "epoch": 0.7713698140200286, + "grad_norm": 0.7649258339058862, + "learning_rate": 2.6187542010000367e-05, + "loss": 1.5043, + "step": 8627 + }, + { + "epoch": 0.7714592274678111, + "grad_norm": 0.8335777607681644, + "learning_rate": 2.616800674969553e-05, + "loss": 1.4701, + "step": 8628 + }, + { + "epoch": 0.7715486409155937, + "grad_norm": 0.6403539370193976, + "learning_rate": 2.61484776815908e-05, + "loss": 1.5678, + "step": 8629 + }, + { + "epoch": 0.7716380543633763, + "grad_norm": 0.6352331442850229, + "learning_rate": 2.612895480732408e-05, + "loss": 1.8649, + "step": 8630 + }, + { + "epoch": 0.7717274678111588, + "grad_norm": 0.5222529550114967, + "learning_rate": 2.610943812853268e-05, + "loss": 1.8733, + "step": 8631 + }, + { + "epoch": 0.7718168812589413, + "grad_norm": 0.543292327845605, + "learning_rate": 2.6089927646853474e-05, + "loss": 1.6888, + "step": 8632 + }, + { + "epoch": 0.7719062947067239, + "grad_norm": 1.0214365327419341, + "learning_rate": 2.6070423363922803e-05, + "loss": 1.6303, + "step": 8633 + }, + { + "epoch": 0.7719957081545065, + "grad_norm": 1.1085490091257089, + "learning_rate": 2.6050925281376403e-05, + "loss": 2.0115, + "step": 8634 + }, + { + "epoch": 0.772085121602289, + "grad_norm": 0.4483665616778065, + "learning_rate": 2.603143340084957e-05, + "loss": 1.5284, + "step": 8635 + }, + { + "epoch": 0.7721745350500715, + "grad_norm": 0.47137572084206464, + "learning_rate": 2.601194772397715e-05, + "loss": 1.9264, + "step": 8636 + }, + { + "epoch": 0.7722639484978541, + "grad_norm": 2.7315890450723663, + "learning_rate": 2.5992468252393275e-05, + "loss": 6.2722, + "step": 8637 + }, + { + "epoch": 0.7723533619456366, + "grad_norm": 0.7198597103432712, + "learning_rate": 2.5972994987731714e-05, + "loss": 1.7588, + "step": 8638 + }, + { + "epoch": 0.7724427753934192, + "grad_norm": 0.353148737390585, + "learning_rate": 2.59535279316257e-05, + "loss": 1.4629, + "step": 8639 + }, + { + "epoch": 0.7725321888412017, + "grad_norm": 0.5606767230235005, + "learning_rate": 2.5934067085707834e-05, + "loss": 1.4382, + "step": 8640 + }, + { + "epoch": 0.7726216022889842, + "grad_norm": 0.708547290501387, + "learning_rate": 2.591461245161032e-05, + "loss": 1.7544, + "step": 8641 + }, + { + "epoch": 0.7727110157367668, + "grad_norm": 0.5951559375927208, + "learning_rate": 2.589516403096478e-05, + "loss": 1.4112, + "step": 8642 + }, + { + "epoch": 0.7728004291845494, + "grad_norm": 0.749709748495195, + "learning_rate": 2.5875721825402342e-05, + "loss": 1.4338, + "step": 8643 + }, + { + "epoch": 0.772889842632332, + "grad_norm": 0.5410645475764947, + "learning_rate": 2.585628583655362e-05, + "loss": 1.7151, + "step": 8644 + }, + { + "epoch": 0.7729792560801144, + "grad_norm": 0.7728504517363993, + "learning_rate": 2.583685606604863e-05, + "loss": 1.8071, + "step": 8645 + }, + { + "epoch": 0.773068669527897, + "grad_norm": 0.5137222638537396, + "learning_rate": 2.581743251551697e-05, + "loss": 1.4139, + "step": 8646 + }, + { + "epoch": 0.7731580829756796, + "grad_norm": 0.5016783415424629, + "learning_rate": 2.5798015186587643e-05, + "loss": 1.6797, + "step": 8647 + }, + { + "epoch": 0.7732474964234621, + "grad_norm": 0.9647216153459417, + "learning_rate": 2.5778604080889202e-05, + "loss": 1.7182, + "step": 8648 + }, + { + "epoch": 0.7733369098712446, + "grad_norm": 1.3322401071566068, + "learning_rate": 2.5759199200049534e-05, + "loss": 6.2214, + "step": 8649 + }, + { + "epoch": 0.7734263233190272, + "grad_norm": 0.897964113310362, + "learning_rate": 2.5739800545696237e-05, + "loss": 1.5821, + "step": 8650 + }, + { + "epoch": 0.7735157367668097, + "grad_norm": 0.980764052488084, + "learning_rate": 2.5720408119456152e-05, + "loss": 1.5282, + "step": 8651 + }, + { + "epoch": 0.7736051502145923, + "grad_norm": 0.6362133718511629, + "learning_rate": 2.5701021922955727e-05, + "loss": 1.7047, + "step": 8652 + }, + { + "epoch": 0.7736945636623748, + "grad_norm": 0.4831391388345164, + "learning_rate": 2.56816419578209e-05, + "loss": 1.6369, + "step": 8653 + }, + { + "epoch": 0.7737839771101573, + "grad_norm": 1.1777971869156345, + "learning_rate": 2.5662268225676976e-05, + "loss": 1.4881, + "step": 8654 + }, + { + "epoch": 0.7738733905579399, + "grad_norm": 1.1224055964575665, + "learning_rate": 2.5642900728148832e-05, + "loss": 1.6694, + "step": 8655 + }, + { + "epoch": 0.7739628040057225, + "grad_norm": 1.5479363425783506, + "learning_rate": 2.5623539466860813e-05, + "loss": 1.6977, + "step": 8656 + }, + { + "epoch": 0.774052217453505, + "grad_norm": 0.4961128160243538, + "learning_rate": 2.5604184443436707e-05, + "loss": 1.722, + "step": 8657 + }, + { + "epoch": 0.7741416309012875, + "grad_norm": 0.9177794814074093, + "learning_rate": 2.5584835659499807e-05, + "loss": 1.5508, + "step": 8658 + }, + { + "epoch": 0.7742310443490701, + "grad_norm": 1.740003542489393, + "learning_rate": 2.5565493116672902e-05, + "loss": 1.5334, + "step": 8659 + }, + { + "epoch": 0.7743204577968527, + "grad_norm": 1.1560629925490502, + "learning_rate": 2.5546156816578158e-05, + "loss": 1.5069, + "step": 8660 + }, + { + "epoch": 0.7744098712446352, + "grad_norm": 0.8530337072620802, + "learning_rate": 2.552682676083733e-05, + "loss": 1.6077, + "step": 8661 + }, + { + "epoch": 0.7744992846924177, + "grad_norm": 0.9502990721732425, + "learning_rate": 2.5507502951071637e-05, + "loss": 1.8156, + "step": 8662 + }, + { + "epoch": 0.7745886981402003, + "grad_norm": 1.487233668596834, + "learning_rate": 2.5488185388901642e-05, + "loss": 1.8677, + "step": 8663 + }, + { + "epoch": 0.7746781115879828, + "grad_norm": 0.580352058808744, + "learning_rate": 2.54688740759476e-05, + "loss": 1.6153, + "step": 8664 + }, + { + "epoch": 0.7747675250357654, + "grad_norm": 0.705713950880571, + "learning_rate": 2.5449569013829066e-05, + "loss": 1.7248, + "step": 8665 + }, + { + "epoch": 0.774856938483548, + "grad_norm": 0.9251648156752172, + "learning_rate": 2.543027020416514e-05, + "loss": 1.845, + "step": 8666 + }, + { + "epoch": 0.7749463519313304, + "grad_norm": 0.7827228318647993, + "learning_rate": 2.541097764857442e-05, + "loss": 1.9987, + "step": 8667 + }, + { + "epoch": 0.775035765379113, + "grad_norm": 1.2186834356774292, + "learning_rate": 2.5391691348674894e-05, + "loss": 6.1004, + "step": 8668 + }, + { + "epoch": 0.7751251788268956, + "grad_norm": 0.9842953197080483, + "learning_rate": 2.537241130608411e-05, + "loss": 1.5083, + "step": 8669 + }, + { + "epoch": 0.7752145922746781, + "grad_norm": 1.3822961100872972, + "learning_rate": 2.5353137522419067e-05, + "loss": 1.6728, + "step": 8670 + }, + { + "epoch": 0.7753040057224606, + "grad_norm": 1.2303257192978199, + "learning_rate": 2.5333869999296223e-05, + "loss": 1.5752, + "step": 8671 + }, + { + "epoch": 0.7753934191702432, + "grad_norm": 0.6368270715789228, + "learning_rate": 2.5314608738331537e-05, + "loss": 1.5436, + "step": 8672 + }, + { + "epoch": 0.7754828326180258, + "grad_norm": 0.7602911276754141, + "learning_rate": 2.529535374114044e-05, + "loss": 1.9666, + "step": 8673 + }, + { + "epoch": 0.7755722460658083, + "grad_norm": 0.5643082932751391, + "learning_rate": 2.527610500933778e-05, + "loss": 1.4932, + "step": 8674 + }, + { + "epoch": 0.7756616595135909, + "grad_norm": 0.8879657893100611, + "learning_rate": 2.525686254453795e-05, + "loss": 1.8396, + "step": 8675 + }, + { + "epoch": 0.7757510729613734, + "grad_norm": 0.8355211435166774, + "learning_rate": 2.5237626348354813e-05, + "loss": 1.4501, + "step": 8676 + }, + { + "epoch": 0.7758404864091559, + "grad_norm": 0.6553941165462377, + "learning_rate": 2.5218396422401614e-05, + "loss": 1.4284, + "step": 8677 + }, + { + "epoch": 0.7759298998569385, + "grad_norm": 0.8007525230089657, + "learning_rate": 2.5199172768291248e-05, + "loss": 2.0887, + "step": 8678 + }, + { + "epoch": 0.7760193133047211, + "grad_norm": 0.8197737475478739, + "learning_rate": 2.51799553876359e-05, + "loss": 1.3564, + "step": 8679 + }, + { + "epoch": 0.7761087267525035, + "grad_norm": 1.4791998796426178, + "learning_rate": 2.5160744282047333e-05, + "loss": 1.9842, + "step": 8680 + }, + { + "epoch": 0.7761981402002861, + "grad_norm": 0.8523038522871342, + "learning_rate": 2.5141539453136755e-05, + "loss": 1.4807, + "step": 8681 + }, + { + "epoch": 0.7762875536480687, + "grad_norm": 0.6441201232662996, + "learning_rate": 2.5122340902514897e-05, + "loss": 1.5804, + "step": 8682 + }, + { + "epoch": 0.7763769670958512, + "grad_norm": 0.7388456098806273, + "learning_rate": 2.510314863179184e-05, + "loss": 1.4663, + "step": 8683 + }, + { + "epoch": 0.7764663805436338, + "grad_norm": 0.515050115018055, + "learning_rate": 2.508396264257725e-05, + "loss": 1.4965, + "step": 8684 + }, + { + "epoch": 0.7765557939914163, + "grad_norm": 0.7971661241967803, + "learning_rate": 2.5064782936480248e-05, + "loss": 1.7335, + "step": 8685 + }, + { + "epoch": 0.7766452074391988, + "grad_norm": 0.6581901570733721, + "learning_rate": 2.5045609515109403e-05, + "loss": 1.6187, + "step": 8686 + }, + { + "epoch": 0.7767346208869814, + "grad_norm": 0.8079531623187707, + "learning_rate": 2.502644238007279e-05, + "loss": 1.7037, + "step": 8687 + }, + { + "epoch": 0.776824034334764, + "grad_norm": 0.5024542386275053, + "learning_rate": 2.500728153297788e-05, + "loss": 1.3349, + "step": 8688 + }, + { + "epoch": 0.7769134477825465, + "grad_norm": 0.6212479428276629, + "learning_rate": 2.498812697543169e-05, + "loss": 1.6333, + "step": 8689 + }, + { + "epoch": 0.777002861230329, + "grad_norm": 0.7594567098912406, + "learning_rate": 2.4968978709040713e-05, + "loss": 1.8016, + "step": 8690 + }, + { + "epoch": 0.7770922746781116, + "grad_norm": 0.5077261450932946, + "learning_rate": 2.4949836735410882e-05, + "loss": 1.3747, + "step": 8691 + }, + { + "epoch": 0.7771816881258942, + "grad_norm": 0.4786434030731836, + "learning_rate": 2.4930701056147586e-05, + "loss": 1.4747, + "step": 8692 + }, + { + "epoch": 0.7772711015736766, + "grad_norm": 0.5994209707101068, + "learning_rate": 2.491157167285578e-05, + "loss": 1.4695, + "step": 8693 + }, + { + "epoch": 0.7773605150214592, + "grad_norm": 0.5882281012208332, + "learning_rate": 2.489244858713974e-05, + "loss": 1.3405, + "step": 8694 + }, + { + "epoch": 0.7774499284692418, + "grad_norm": 0.5318264700679345, + "learning_rate": 2.4873331800603327e-05, + "loss": 1.2912, + "step": 8695 + }, + { + "epoch": 0.7775393419170243, + "grad_norm": 1.0076639128158817, + "learning_rate": 2.485422131484987e-05, + "loss": 1.5707, + "step": 8696 + }, + { + "epoch": 0.7776287553648069, + "grad_norm": 0.5918605471907542, + "learning_rate": 2.4835117131482067e-05, + "loss": 1.7931, + "step": 8697 + }, + { + "epoch": 0.7777181688125894, + "grad_norm": 0.4686451901076295, + "learning_rate": 2.4816019252102273e-05, + "loss": 1.2525, + "step": 8698 + }, + { + "epoch": 0.7778075822603719, + "grad_norm": 0.7509369584725215, + "learning_rate": 2.479692767831211e-05, + "loss": 1.6706, + "step": 8699 + }, + { + "epoch": 0.7778969957081545, + "grad_norm": 0.657435926229035, + "learning_rate": 2.4777842411712805e-05, + "loss": 1.7633, + "step": 8700 + }, + { + "epoch": 0.7779864091559371, + "grad_norm": 0.9252312823311182, + "learning_rate": 2.4758763453905044e-05, + "loss": 1.6833, + "step": 8701 + }, + { + "epoch": 0.7780758226037195, + "grad_norm": 0.3768289923833078, + "learning_rate": 2.473969080648889e-05, + "loss": 1.7247, + "step": 8702 + }, + { + "epoch": 0.7781652360515021, + "grad_norm": 0.7620612297668116, + "learning_rate": 2.472062447106398e-05, + "loss": 1.779, + "step": 8703 + }, + { + "epoch": 0.7782546494992847, + "grad_norm": 0.5272086511114792, + "learning_rate": 2.4701564449229374e-05, + "loss": 1.4226, + "step": 8704 + }, + { + "epoch": 0.7783440629470673, + "grad_norm": 1.0371511431994922, + "learning_rate": 2.468251074258362e-05, + "loss": 1.4459, + "step": 8705 + }, + { + "epoch": 0.7784334763948498, + "grad_norm": 0.6201279976343707, + "learning_rate": 2.4663463352724737e-05, + "loss": 1.4108, + "step": 8706 + }, + { + "epoch": 0.7785228898426323, + "grad_norm": 0.8523636192513271, + "learning_rate": 2.4644422281250223e-05, + "loss": 1.4899, + "step": 8707 + }, + { + "epoch": 0.7786123032904149, + "grad_norm": 0.891135647124909, + "learning_rate": 2.462538752975698e-05, + "loss": 1.8653, + "step": 8708 + }, + { + "epoch": 0.7787017167381974, + "grad_norm": 0.9666600982673315, + "learning_rate": 2.4606359099841457e-05, + "loss": 1.6843, + "step": 8709 + }, + { + "epoch": 0.77879113018598, + "grad_norm": 1.2189825519918858, + "learning_rate": 2.4587336993099574e-05, + "loss": 1.775, + "step": 8710 + }, + { + "epoch": 0.7788805436337625, + "grad_norm": 0.5023539172573116, + "learning_rate": 2.4568321211126598e-05, + "loss": 1.6284, + "step": 8711 + }, + { + "epoch": 0.778969957081545, + "grad_norm": 1.0578024929089591, + "learning_rate": 2.4549311755517457e-05, + "loss": 1.6848, + "step": 8712 + }, + { + "epoch": 0.7790593705293276, + "grad_norm": 0.3528748951863757, + "learning_rate": 2.4530308627866438e-05, + "loss": 1.514, + "step": 8713 + }, + { + "epoch": 0.7791487839771102, + "grad_norm": 0.7638998528610563, + "learning_rate": 2.451131182976727e-05, + "loss": 1.4054, + "step": 8714 + }, + { + "epoch": 0.7792381974248928, + "grad_norm": 0.6804167294752693, + "learning_rate": 2.4492321362813207e-05, + "loss": 1.6844, + "step": 8715 + }, + { + "epoch": 0.7793276108726752, + "grad_norm": 0.4752874807478382, + "learning_rate": 2.4473337228596994e-05, + "loss": 1.5431, + "step": 8716 + }, + { + "epoch": 0.7794170243204578, + "grad_norm": 0.6331779948593012, + "learning_rate": 2.445435942871074e-05, + "loss": 1.7634, + "step": 8717 + }, + { + "epoch": 0.7795064377682404, + "grad_norm": 0.596035106711962, + "learning_rate": 2.4435387964746127e-05, + "loss": 1.8165, + "step": 8718 + }, + { + "epoch": 0.7795958512160229, + "grad_norm": 0.5639314486896575, + "learning_rate": 2.4416422838294273e-05, + "loss": 2.1381, + "step": 8719 + }, + { + "epoch": 0.7796852646638054, + "grad_norm": 0.43698615856862144, + "learning_rate": 2.439746405094575e-05, + "loss": 1.3768, + "step": 8720 + }, + { + "epoch": 0.779774678111588, + "grad_norm": 0.9325182792990236, + "learning_rate": 2.4378511604290632e-05, + "loss": 1.863, + "step": 8721 + }, + { + "epoch": 0.7798640915593705, + "grad_norm": 0.5415703033314861, + "learning_rate": 2.4359565499918402e-05, + "loss": 1.8192, + "step": 8722 + }, + { + "epoch": 0.7799535050071531, + "grad_norm": 0.7347297677422211, + "learning_rate": 2.4340625739418055e-05, + "loss": 2.0605, + "step": 8723 + }, + { + "epoch": 0.7800429184549357, + "grad_norm": 0.49167242446196874, + "learning_rate": 2.4321692324378087e-05, + "loss": 1.6042, + "step": 8724 + }, + { + "epoch": 0.7801323319027181, + "grad_norm": 0.47626499235236147, + "learning_rate": 2.4302765256386327e-05, + "loss": 1.6927, + "step": 8725 + }, + { + "epoch": 0.7802217453505007, + "grad_norm": 0.8623868764047017, + "learning_rate": 2.4283844537030252e-05, + "loss": 1.682, + "step": 8726 + }, + { + "epoch": 0.7803111587982833, + "grad_norm": 0.5091706378262036, + "learning_rate": 2.4264930167896727e-05, + "loss": 1.5964, + "step": 8727 + }, + { + "epoch": 0.7804005722460658, + "grad_norm": 0.8056756730854929, + "learning_rate": 2.4246022150572024e-05, + "loss": 1.824, + "step": 8728 + }, + { + "epoch": 0.7804899856938483, + "grad_norm": 0.4286508869454822, + "learning_rate": 2.422712048664194e-05, + "loss": 1.312, + "step": 8729 + }, + { + "epoch": 0.7805793991416309, + "grad_norm": 0.6868666619489626, + "learning_rate": 2.420822517769179e-05, + "loss": 1.5941, + "step": 8730 + }, + { + "epoch": 0.7806688125894135, + "grad_norm": 0.5296771145548043, + "learning_rate": 2.4189336225306225e-05, + "loss": 1.8961, + "step": 8731 + }, + { + "epoch": 0.780758226037196, + "grad_norm": 0.7159427052233202, + "learning_rate": 2.417045363106948e-05, + "loss": 1.6565, + "step": 8732 + }, + { + "epoch": 0.7808476394849786, + "grad_norm": 0.6395419774698005, + "learning_rate": 2.4151577396565205e-05, + "loss": 1.5852, + "step": 8733 + }, + { + "epoch": 0.780937052932761, + "grad_norm": 0.9926935825781762, + "learning_rate": 2.413270752337653e-05, + "loss": 1.5495, + "step": 8734 + }, + { + "epoch": 0.7810264663805436, + "grad_norm": 0.7113473901110179, + "learning_rate": 2.4113844013086083e-05, + "loss": 1.5327, + "step": 8735 + }, + { + "epoch": 0.7811158798283262, + "grad_norm": 0.32809314871996165, + "learning_rate": 2.409498686727587e-05, + "loss": 1.5644, + "step": 8736 + }, + { + "epoch": 0.7812052932761088, + "grad_norm": 0.5198178719542047, + "learning_rate": 2.4076136087527435e-05, + "loss": 1.3714, + "step": 8737 + }, + { + "epoch": 0.7812947067238912, + "grad_norm": 1.0299769116223798, + "learning_rate": 2.4057291675421768e-05, + "loss": 1.7994, + "step": 8738 + }, + { + "epoch": 0.7813841201716738, + "grad_norm": 0.656989065507624, + "learning_rate": 2.4038453632539338e-05, + "loss": 1.7366, + "step": 8739 + }, + { + "epoch": 0.7814735336194564, + "grad_norm": 0.3966157225756014, + "learning_rate": 2.4019621960460058e-05, + "loss": 1.7777, + "step": 8740 + }, + { + "epoch": 0.781562947067239, + "grad_norm": 0.43740403525372723, + "learning_rate": 2.4000796660763346e-05, + "loss": 1.8753, + "step": 8741 + }, + { + "epoch": 0.7816523605150214, + "grad_norm": 0.561079342919207, + "learning_rate": 2.3981977735028018e-05, + "loss": 1.7221, + "step": 8742 + }, + { + "epoch": 0.781741773962804, + "grad_norm": 0.4398036323946011, + "learning_rate": 2.3963165184832403e-05, + "loss": 1.6912, + "step": 8743 + }, + { + "epoch": 0.7818311874105865, + "grad_norm": 0.7211282822671423, + "learning_rate": 2.3944359011754336e-05, + "loss": 1.9241, + "step": 8744 + }, + { + "epoch": 0.7819206008583691, + "grad_norm": 0.6375144629368913, + "learning_rate": 2.3925559217370987e-05, + "loss": 1.5866, + "step": 8745 + }, + { + "epoch": 0.7820100143061517, + "grad_norm": 0.39864493170074466, + "learning_rate": 2.3906765803259078e-05, + "loss": 1.5131, + "step": 8746 + }, + { + "epoch": 0.7820994277539342, + "grad_norm": 0.5689089584783897, + "learning_rate": 2.388797877099489e-05, + "loss": 1.5918, + "step": 8747 + }, + { + "epoch": 0.7821888412017167, + "grad_norm": 1.0866504629846125, + "learning_rate": 2.386919812215398e-05, + "loss": 1.5725, + "step": 8748 + }, + { + "epoch": 0.7822782546494993, + "grad_norm": 0.6973010379214899, + "learning_rate": 2.3850423858311466e-05, + "loss": 1.5889, + "step": 8749 + }, + { + "epoch": 0.7823676680972819, + "grad_norm": 0.9229572849735583, + "learning_rate": 2.3831655981041977e-05, + "loss": 1.4719, + "step": 8750 + }, + { + "epoch": 0.7824570815450643, + "grad_norm": 0.7370341486130655, + "learning_rate": 2.381289449191948e-05, + "loss": 1.7465, + "step": 8751 + }, + { + "epoch": 0.7825464949928469, + "grad_norm": 0.6329937686460192, + "learning_rate": 2.379413939251751e-05, + "loss": 1.8035, + "step": 8752 + }, + { + "epoch": 0.7826359084406295, + "grad_norm": 0.8374461372302813, + "learning_rate": 2.3775390684409037e-05, + "loss": 1.7809, + "step": 8753 + }, + { + "epoch": 0.782725321888412, + "grad_norm": 0.38561597261243485, + "learning_rate": 2.375664836916649e-05, + "loss": 1.231, + "step": 8754 + }, + { + "epoch": 0.7828147353361946, + "grad_norm": 0.44800787490632205, + "learning_rate": 2.3737912448361798e-05, + "loss": 1.7815, + "step": 8755 + }, + { + "epoch": 0.7829041487839771, + "grad_norm": 0.5484486636047582, + "learning_rate": 2.3719182923566263e-05, + "loss": 1.7926, + "step": 8756 + }, + { + "epoch": 0.7829935622317596, + "grad_norm": 0.6850117985298543, + "learning_rate": 2.3700459796350726e-05, + "loss": 1.7767, + "step": 8757 + }, + { + "epoch": 0.7830829756795422, + "grad_norm": 0.3835447242381383, + "learning_rate": 2.36817430682855e-05, + "loss": 1.278, + "step": 8758 + }, + { + "epoch": 0.7831723891273248, + "grad_norm": 0.5079765479238779, + "learning_rate": 2.3663032740940293e-05, + "loss": 1.4788, + "step": 8759 + }, + { + "epoch": 0.7832618025751072, + "grad_norm": 0.6549283588487307, + "learning_rate": 2.364432881588431e-05, + "loss": 1.7834, + "step": 8760 + }, + { + "epoch": 0.7833512160228898, + "grad_norm": 0.45508752443323697, + "learning_rate": 2.362563129468631e-05, + "loss": 1.5573, + "step": 8761 + }, + { + "epoch": 0.7834406294706724, + "grad_norm": 1.1349505633540538, + "learning_rate": 2.360694017891436e-05, + "loss": 2.0262, + "step": 8762 + }, + { + "epoch": 0.783530042918455, + "grad_norm": 0.44923296762700604, + "learning_rate": 2.358825547013607e-05, + "loss": 2.0618, + "step": 8763 + }, + { + "epoch": 0.7836194563662375, + "grad_norm": 0.6911235442178492, + "learning_rate": 2.3569577169918532e-05, + "loss": 1.4147, + "step": 8764 + }, + { + "epoch": 0.78370886981402, + "grad_norm": 0.6593638176011106, + "learning_rate": 2.355090527982823e-05, + "loss": 1.8319, + "step": 8765 + }, + { + "epoch": 0.7837982832618026, + "grad_norm": 0.8291638283002056, + "learning_rate": 2.353223980143118e-05, + "loss": 1.4772, + "step": 8766 + }, + { + "epoch": 0.7838876967095851, + "grad_norm": 0.5648942062194696, + "learning_rate": 2.351358073629282e-05, + "loss": 1.8143, + "step": 8767 + }, + { + "epoch": 0.7839771101573677, + "grad_norm": 1.043780073673575, + "learning_rate": 2.3494928085978073e-05, + "loss": 1.8202, + "step": 8768 + }, + { + "epoch": 0.7840665236051502, + "grad_norm": 0.5208589654981648, + "learning_rate": 2.3476281852051308e-05, + "loss": 1.4096, + "step": 8769 + }, + { + "epoch": 0.7841559370529327, + "grad_norm": 0.5752132476578664, + "learning_rate": 2.345764203607641e-05, + "loss": 1.5984, + "step": 8770 + }, + { + "epoch": 0.7842453505007153, + "grad_norm": 0.6289556486071634, + "learning_rate": 2.343900863961659e-05, + "loss": 1.4279, + "step": 8771 + }, + { + "epoch": 0.7843347639484979, + "grad_norm": 0.6332861504321875, + "learning_rate": 2.342038166423466e-05, + "loss": 1.6048, + "step": 8772 + }, + { + "epoch": 0.7844241773962805, + "grad_norm": 0.5479712764247445, + "learning_rate": 2.3401761111492836e-05, + "loss": 1.39, + "step": 8773 + }, + { + "epoch": 0.7845135908440629, + "grad_norm": 0.5976674920757096, + "learning_rate": 2.338314698295281e-05, + "loss": 1.7368, + "step": 8774 + }, + { + "epoch": 0.7846030042918455, + "grad_norm": 1.0433336130641575, + "learning_rate": 2.3364539280175734e-05, + "loss": 1.4828, + "step": 8775 + }, + { + "epoch": 0.7846924177396281, + "grad_norm": 0.37834952436639835, + "learning_rate": 2.3345938004722168e-05, + "loss": 1.1929, + "step": 8776 + }, + { + "epoch": 0.7847818311874106, + "grad_norm": 0.3965608061424073, + "learning_rate": 2.3327343158152205e-05, + "loss": 1.3425, + "step": 8777 + }, + { + "epoch": 0.7848712446351931, + "grad_norm": 0.8344509655833618, + "learning_rate": 2.3308754742025406e-05, + "loss": 1.9095, + "step": 8778 + }, + { + "epoch": 0.7849606580829757, + "grad_norm": 0.5420903671433732, + "learning_rate": 2.3290172757900696e-05, + "loss": 1.5826, + "step": 8779 + }, + { + "epoch": 0.7850500715307582, + "grad_norm": 0.6479562628965775, + "learning_rate": 2.3271597207336526e-05, + "loss": 1.6512, + "step": 8780 + }, + { + "epoch": 0.7851394849785408, + "grad_norm": 0.3035211380678379, + "learning_rate": 2.3253028091890893e-05, + "loss": 1.2726, + "step": 8781 + }, + { + "epoch": 0.7852288984263234, + "grad_norm": 1.2434940147714397, + "learning_rate": 2.3234465413121086e-05, + "loss": 2.2792, + "step": 8782 + }, + { + "epoch": 0.7853183118741058, + "grad_norm": 0.368417859872348, + "learning_rate": 2.321590917258395e-05, + "loss": 1.3048, + "step": 8783 + }, + { + "epoch": 0.7854077253218884, + "grad_norm": 0.5542569782024608, + "learning_rate": 2.3197359371835802e-05, + "loss": 1.7513, + "step": 8784 + }, + { + "epoch": 0.785497138769671, + "grad_norm": 0.35647215163378815, + "learning_rate": 2.3178816012432346e-05, + "loss": 1.6727, + "step": 8785 + }, + { + "epoch": 0.7855865522174535, + "grad_norm": 0.38795984266189054, + "learning_rate": 2.3160279095928817e-05, + "loss": 1.3875, + "step": 8786 + }, + { + "epoch": 0.785675965665236, + "grad_norm": 0.6308966009773402, + "learning_rate": 2.3141748623879878e-05, + "loss": 1.3692, + "step": 8787 + }, + { + "epoch": 0.7857653791130186, + "grad_norm": 0.8008526075263188, + "learning_rate": 2.3123224597839664e-05, + "loss": 1.6645, + "step": 8788 + }, + { + "epoch": 0.7858547925608012, + "grad_norm": 0.8481230680033363, + "learning_rate": 2.3104707019361782e-05, + "loss": 2.1485, + "step": 8789 + }, + { + "epoch": 0.7859442060085837, + "grad_norm": 0.8383666301664398, + "learning_rate": 2.3086195889999228e-05, + "loss": 1.5973, + "step": 8790 + }, + { + "epoch": 0.7860336194563662, + "grad_norm": 0.8272917071826013, + "learning_rate": 2.3067691211304544e-05, + "loss": 1.6235, + "step": 8791 + }, + { + "epoch": 0.7861230329041488, + "grad_norm": 0.7566910906551472, + "learning_rate": 2.3049192984829715e-05, + "loss": 1.335, + "step": 8792 + }, + { + "epoch": 0.7862124463519313, + "grad_norm": 0.5574262231082718, + "learning_rate": 2.3030701212126106e-05, + "loss": 1.4785, + "step": 8793 + }, + { + "epoch": 0.7863018597997139, + "grad_norm": 0.6155339232500912, + "learning_rate": 2.3012215894744593e-05, + "loss": 1.7216, + "step": 8794 + }, + { + "epoch": 0.7863912732474965, + "grad_norm": 0.5804275061773462, + "learning_rate": 2.299373703423563e-05, + "loss": 1.7223, + "step": 8795 + }, + { + "epoch": 0.7864806866952789, + "grad_norm": 0.4113866277640799, + "learning_rate": 2.2975264632148896e-05, + "loss": 1.3467, + "step": 8796 + }, + { + "epoch": 0.7865701001430615, + "grad_norm": 0.5224164405781955, + "learning_rate": 2.2956798690033708e-05, + "loss": 1.3371, + "step": 8797 + }, + { + "epoch": 0.7866595135908441, + "grad_norm": 0.7244174178472477, + "learning_rate": 2.2938339209438797e-05, + "loss": 1.4307, + "step": 8798 + }, + { + "epoch": 0.7867489270386266, + "grad_norm": 0.6816911998873948, + "learning_rate": 2.2919886191912277e-05, + "loss": 1.4207, + "step": 8799 + }, + { + "epoch": 0.7868383404864091, + "grad_norm": 0.6111543274602306, + "learning_rate": 2.290143963900181e-05, + "loss": 1.8671, + "step": 8800 + }, + { + "epoch": 0.7869277539341917, + "grad_norm": 0.4195338414207445, + "learning_rate": 2.2882999552254492e-05, + "loss": 1.2824, + "step": 8801 + }, + { + "epoch": 0.7870171673819742, + "grad_norm": 0.8298440284497017, + "learning_rate": 2.2864565933216865e-05, + "loss": 1.9048, + "step": 8802 + }, + { + "epoch": 0.7871065808297568, + "grad_norm": 0.48795954567603345, + "learning_rate": 2.2846138783434944e-05, + "loss": 1.6975, + "step": 8803 + }, + { + "epoch": 0.7871959942775394, + "grad_norm": 0.6820523819610834, + "learning_rate": 2.282771810445421e-05, + "loss": 1.6859, + "step": 8804 + }, + { + "epoch": 0.7872854077253219, + "grad_norm": 0.6993028984805831, + "learning_rate": 2.280930389781952e-05, + "loss": 1.6966, + "step": 8805 + }, + { + "epoch": 0.7873748211731044, + "grad_norm": 0.4616522816362692, + "learning_rate": 2.2790896165075305e-05, + "loss": 1.5489, + "step": 8806 + }, + { + "epoch": 0.787464234620887, + "grad_norm": 0.7945394026257957, + "learning_rate": 2.2772494907765406e-05, + "loss": 1.6636, + "step": 8807 + }, + { + "epoch": 0.7875536480686696, + "grad_norm": 0.4667048358523697, + "learning_rate": 2.275410012743303e-05, + "loss": 1.5121, + "step": 8808 + }, + { + "epoch": 0.787643061516452, + "grad_norm": 0.4135690657722232, + "learning_rate": 2.2735711825621052e-05, + "loss": 1.709, + "step": 8809 + }, + { + "epoch": 0.7877324749642346, + "grad_norm": 0.5710340284079849, + "learning_rate": 2.2717330003871573e-05, + "loss": 1.6214, + "step": 8810 + }, + { + "epoch": 0.7878218884120172, + "grad_norm": 0.7491098097152826, + "learning_rate": 2.26989546637263e-05, + "loss": 1.5293, + "step": 8811 + }, + { + "epoch": 0.7879113018597997, + "grad_norm": 0.5293765724744698, + "learning_rate": 2.2680585806726373e-05, + "loss": 1.3551, + "step": 8812 + }, + { + "epoch": 0.7880007153075823, + "grad_norm": 0.42489975324773277, + "learning_rate": 2.266222343441231e-05, + "loss": 1.4222, + "step": 8813 + }, + { + "epoch": 0.7880901287553648, + "grad_norm": 0.5220538579025328, + "learning_rate": 2.264386754832416e-05, + "loss": 1.4895, + "step": 8814 + }, + { + "epoch": 0.7881795422031473, + "grad_norm": 0.4543476935891215, + "learning_rate": 2.2625518150001425e-05, + "loss": 1.6752, + "step": 8815 + }, + { + "epoch": 0.7882689556509299, + "grad_norm": 0.8552848720278878, + "learning_rate": 2.2607175240983026e-05, + "loss": 1.6869, + "step": 8816 + }, + { + "epoch": 0.7883583690987125, + "grad_norm": 0.4906711661846381, + "learning_rate": 2.2588838822807378e-05, + "loss": 1.6717, + "step": 8817 + }, + { + "epoch": 0.788447782546495, + "grad_norm": 0.4273386047403816, + "learning_rate": 2.2570508897012355e-05, + "loss": 1.4074, + "step": 8818 + }, + { + "epoch": 0.7885371959942775, + "grad_norm": 0.5480998643709266, + "learning_rate": 2.2552185465135224e-05, + "loss": 1.8726, + "step": 8819 + }, + { + "epoch": 0.7886266094420601, + "grad_norm": 0.48623677868539883, + "learning_rate": 2.2533868528712755e-05, + "loss": 1.6908, + "step": 8820 + }, + { + "epoch": 0.7887160228898427, + "grad_norm": 0.6949718411211344, + "learning_rate": 2.2515558089281196e-05, + "loss": 1.5881, + "step": 8821 + }, + { + "epoch": 0.7888054363376252, + "grad_norm": 0.8621646774361368, + "learning_rate": 2.2497254148376157e-05, + "loss": 1.8185, + "step": 8822 + }, + { + "epoch": 0.7888948497854077, + "grad_norm": 0.48238697361091293, + "learning_rate": 2.247895670753287e-05, + "loss": 1.5521, + "step": 8823 + }, + { + "epoch": 0.7889842632331903, + "grad_norm": 0.6143085187109676, + "learning_rate": 2.2460665768285826e-05, + "loss": 1.6269, + "step": 8824 + }, + { + "epoch": 0.7890736766809728, + "grad_norm": 1.3471596643309711, + "learning_rate": 2.2442381332169115e-05, + "loss": 1.9687, + "step": 8825 + }, + { + "epoch": 0.7891630901287554, + "grad_norm": 0.5297589225717557, + "learning_rate": 2.2424103400716203e-05, + "loss": 1.5769, + "step": 8826 + }, + { + "epoch": 0.7892525035765379, + "grad_norm": 0.4069680006209874, + "learning_rate": 2.240583197546008e-05, + "loss": 1.4315, + "step": 8827 + }, + { + "epoch": 0.7893419170243204, + "grad_norm": 0.9250350865329277, + "learning_rate": 2.23875670579331e-05, + "loss": 1.802, + "step": 8828 + }, + { + "epoch": 0.789431330472103, + "grad_norm": 0.6914722902863291, + "learning_rate": 2.236930864966713e-05, + "loss": 1.6928, + "step": 8829 + }, + { + "epoch": 0.7895207439198856, + "grad_norm": 0.8219772317133162, + "learning_rate": 2.235105675219349e-05, + "loss": 1.9791, + "step": 8830 + }, + { + "epoch": 0.789610157367668, + "grad_norm": 0.4398670541177345, + "learning_rate": 2.2332811367042948e-05, + "loss": 1.6815, + "step": 8831 + }, + { + "epoch": 0.7896995708154506, + "grad_norm": 0.6047288417724898, + "learning_rate": 2.2314572495745746e-05, + "loss": 1.5492, + "step": 8832 + }, + { + "epoch": 0.7897889842632332, + "grad_norm": 0.8650456239913795, + "learning_rate": 2.2296340139831494e-05, + "loss": 1.7203, + "step": 8833 + }, + { + "epoch": 0.7898783977110158, + "grad_norm": 0.7789731605060366, + "learning_rate": 2.2278114300829356e-05, + "loss": 1.7412, + "step": 8834 + }, + { + "epoch": 0.7899678111587983, + "grad_norm": 1.0781281712598174, + "learning_rate": 2.2259894980267937e-05, + "loss": 1.5221, + "step": 8835 + }, + { + "epoch": 0.7900572246065808, + "grad_norm": 0.5236325972598549, + "learning_rate": 2.224168217967518e-05, + "loss": 1.6204, + "step": 8836 + }, + { + "epoch": 0.7901466380543634, + "grad_norm": 0.9870630224009684, + "learning_rate": 2.2223475900578674e-05, + "loss": 2.0069, + "step": 8837 + }, + { + "epoch": 0.7902360515021459, + "grad_norm": 1.0572138755528142, + "learning_rate": 2.220527614450533e-05, + "loss": 1.3713, + "step": 8838 + }, + { + "epoch": 0.7903254649499285, + "grad_norm": 0.7056871630446621, + "learning_rate": 2.2187082912981493e-05, + "loss": 1.6009, + "step": 8839 + }, + { + "epoch": 0.790414878397711, + "grad_norm": 0.6065462024986674, + "learning_rate": 2.216889620753304e-05, + "loss": 1.4441, + "step": 8840 + }, + { + "epoch": 0.7905042918454935, + "grad_norm": 0.7290215687605089, + "learning_rate": 2.215071602968529e-05, + "loss": 1.5998, + "step": 8841 + }, + { + "epoch": 0.7905937052932761, + "grad_norm": 0.9466079855445855, + "learning_rate": 2.213254238096295e-05, + "loss": 1.5887, + "step": 8842 + }, + { + "epoch": 0.7906831187410587, + "grad_norm": 0.41615201673508184, + "learning_rate": 2.211437526289023e-05, + "loss": 1.2434, + "step": 8843 + }, + { + "epoch": 0.7907725321888412, + "grad_norm": 1.9442211278175994, + "learning_rate": 2.20962146769908e-05, + "loss": 6.2008, + "step": 8844 + }, + { + "epoch": 0.7908619456366237, + "grad_norm": 0.6835316228385628, + "learning_rate": 2.2078060624787757e-05, + "loss": 1.8287, + "step": 8845 + }, + { + "epoch": 0.7909513590844063, + "grad_norm": 0.6370156743899956, + "learning_rate": 2.2059913107803697e-05, + "loss": 1.52, + "step": 8846 + }, + { + "epoch": 0.7910407725321889, + "grad_norm": 0.45122019274107017, + "learning_rate": 2.2041772127560566e-05, + "loss": 1.4212, + "step": 8847 + }, + { + "epoch": 0.7911301859799714, + "grad_norm": 0.378270170414208, + "learning_rate": 2.2023637685579856e-05, + "loss": 1.7782, + "step": 8848 + }, + { + "epoch": 0.7912195994277539, + "grad_norm": 0.6778245626731694, + "learning_rate": 2.2005509783382517e-05, + "loss": 1.4765, + "step": 8849 + }, + { + "epoch": 0.7913090128755365, + "grad_norm": 1.07183313020938, + "learning_rate": 2.198738842248882e-05, + "loss": 1.8305, + "step": 8850 + }, + { + "epoch": 0.791398426323319, + "grad_norm": 0.747518530160169, + "learning_rate": 2.196927360441866e-05, + "loss": 1.5003, + "step": 8851 + }, + { + "epoch": 0.7914878397711016, + "grad_norm": 0.38790931566548553, + "learning_rate": 2.1951165330691324e-05, + "loss": 1.6369, + "step": 8852 + }, + { + "epoch": 0.7915772532188842, + "grad_norm": 0.5545705085890462, + "learning_rate": 2.1933063602825455e-05, + "loss": 1.8546, + "step": 8853 + }, + { + "epoch": 0.7916666666666666, + "grad_norm": 0.5162179651958814, + "learning_rate": 2.1914968422339266e-05, + "loss": 1.9908, + "step": 8854 + }, + { + "epoch": 0.7917560801144492, + "grad_norm": 0.5528803191130696, + "learning_rate": 2.1896879790750403e-05, + "loss": 1.6393, + "step": 8855 + }, + { + "epoch": 0.7918454935622318, + "grad_norm": 0.6783873833704922, + "learning_rate": 2.1878797709575847e-05, + "loss": 1.6029, + "step": 8856 + }, + { + "epoch": 0.7919349070100143, + "grad_norm": 0.4630174220232602, + "learning_rate": 2.186072218033224e-05, + "loss": 1.6871, + "step": 8857 + }, + { + "epoch": 0.7920243204577968, + "grad_norm": 0.4948891540063965, + "learning_rate": 2.1842653204535466e-05, + "loss": 1.5779, + "step": 8858 + }, + { + "epoch": 0.7921137339055794, + "grad_norm": 0.6251197150419202, + "learning_rate": 2.1824590783700982e-05, + "loss": 1.7501, + "step": 8859 + }, + { + "epoch": 0.792203147353362, + "grad_norm": 0.6110621875753455, + "learning_rate": 2.1806534919343647e-05, + "loss": 1.4893, + "step": 8860 + }, + { + "epoch": 0.7922925608011445, + "grad_norm": 0.5820951792032002, + "learning_rate": 2.1788485612977827e-05, + "loss": 1.7174, + "step": 8861 + }, + { + "epoch": 0.7923819742489271, + "grad_norm": 0.6120194915678993, + "learning_rate": 2.1770442866117236e-05, + "loss": 1.6852, + "step": 8862 + }, + { + "epoch": 0.7924713876967096, + "grad_norm": 0.5932759671316766, + "learning_rate": 2.1752406680275126e-05, + "loss": 1.3641, + "step": 8863 + }, + { + "epoch": 0.7925608011444921, + "grad_norm": 0.8656817257687736, + "learning_rate": 2.1734377056964172e-05, + "loss": 1.2827, + "step": 8864 + }, + { + "epoch": 0.7926502145922747, + "grad_norm": 0.5380151049053667, + "learning_rate": 2.1716353997696482e-05, + "loss": 1.419, + "step": 8865 + }, + { + "epoch": 0.7927396280400573, + "grad_norm": 0.9381668971110662, + "learning_rate": 2.169833750398368e-05, + "loss": 1.4773, + "step": 8866 + }, + { + "epoch": 0.7928290414878397, + "grad_norm": 4.417912498628269, + "learning_rate": 2.1680327577336712e-05, + "loss": 6.2169, + "step": 8867 + }, + { + "epoch": 0.7929184549356223, + "grad_norm": 1.1294475247553974, + "learning_rate": 2.1662324219266083e-05, + "loss": 1.7506, + "step": 8868 + }, + { + "epoch": 0.7930078683834049, + "grad_norm": 1.424012172866076, + "learning_rate": 2.1644327431281742e-05, + "loss": 1.5814, + "step": 8869 + }, + { + "epoch": 0.7930972818311874, + "grad_norm": 1.2592999108466194, + "learning_rate": 2.1626337214892978e-05, + "loss": 1.8034, + "step": 8870 + }, + { + "epoch": 0.79318669527897, + "grad_norm": 0.5450568101653908, + "learning_rate": 2.1608353571608685e-05, + "loss": 1.2919, + "step": 8871 + }, + { + "epoch": 0.7932761087267525, + "grad_norm": 0.5992475096195513, + "learning_rate": 2.1590376502937136e-05, + "loss": 1.6349, + "step": 8872 + }, + { + "epoch": 0.793365522174535, + "grad_norm": 1.244009976645464, + "learning_rate": 2.1572406010385983e-05, + "loss": 1.9945, + "step": 8873 + }, + { + "epoch": 0.7934549356223176, + "grad_norm": 0.8180873475735791, + "learning_rate": 2.1554442095462422e-05, + "loss": 1.5217, + "step": 8874 + }, + { + "epoch": 0.7935443490701002, + "grad_norm": 0.578441741284651, + "learning_rate": 2.1536484759673092e-05, + "loss": 1.7034, + "step": 8875 + }, + { + "epoch": 0.7936337625178826, + "grad_norm": 0.6701612795693784, + "learning_rate": 2.1518534004523993e-05, + "loss": 1.5907, + "step": 8876 + }, + { + "epoch": 0.7937231759656652, + "grad_norm": 1.0283442970820356, + "learning_rate": 2.150058983152068e-05, + "loss": 1.7303, + "step": 8877 + }, + { + "epoch": 0.7938125894134478, + "grad_norm": 1.0373434532462817, + "learning_rate": 2.1482652242168077e-05, + "loss": 1.6166, + "step": 8878 + }, + { + "epoch": 0.7939020028612304, + "grad_norm": 0.8086304498864421, + "learning_rate": 2.146472123797062e-05, + "loss": 1.7541, + "step": 8879 + }, + { + "epoch": 0.7939914163090128, + "grad_norm": 0.6011568603631318, + "learning_rate": 2.1446796820432167e-05, + "loss": 1.5362, + "step": 8880 + }, + { + "epoch": 0.7940808297567954, + "grad_norm": 0.6333202002751344, + "learning_rate": 2.1428878991055966e-05, + "loss": 1.9908, + "step": 8881 + }, + { + "epoch": 0.794170243204578, + "grad_norm": 0.9790905461894314, + "learning_rate": 2.1410967751344803e-05, + "loss": 1.5887, + "step": 8882 + }, + { + "epoch": 0.7942596566523605, + "grad_norm": 0.534482164121778, + "learning_rate": 2.1393063102800847e-05, + "loss": 1.3684, + "step": 8883 + }, + { + "epoch": 0.7943490701001431, + "grad_norm": 0.6236108408523684, + "learning_rate": 2.137516504692577e-05, + "loss": 1.5405, + "step": 8884 + }, + { + "epoch": 0.7944384835479256, + "grad_norm": 0.6333751014712323, + "learning_rate": 2.135727358522064e-05, + "loss": 1.3913, + "step": 8885 + }, + { + "epoch": 0.7945278969957081, + "grad_norm": 0.5351911801318727, + "learning_rate": 2.1339388719186028e-05, + "loss": 1.7182, + "step": 8886 + }, + { + "epoch": 0.7946173104434907, + "grad_norm": 0.7381415950505065, + "learning_rate": 2.1321510450321858e-05, + "loss": 1.9312, + "step": 8887 + }, + { + "epoch": 0.7947067238912733, + "grad_norm": 0.7064023867793168, + "learning_rate": 2.1303638780127588e-05, + "loss": 2.1794, + "step": 8888 + }, + { + "epoch": 0.7947961373390557, + "grad_norm": 0.47768334871307766, + "learning_rate": 2.128577371010212e-05, + "loss": 1.6769, + "step": 8889 + }, + { + "epoch": 0.7948855507868383, + "grad_norm": 9.935875993615426, + "learning_rate": 2.126791524174372e-05, + "loss": 6.3725, + "step": 8890 + }, + { + "epoch": 0.7949749642346209, + "grad_norm": 2.933107022309229, + "learning_rate": 2.1250063376550154e-05, + "loss": 1.8405, + "step": 8891 + }, + { + "epoch": 0.7950643776824035, + "grad_norm": 2.3379493372434625, + "learning_rate": 2.1232218116018722e-05, + "loss": 1.4845, + "step": 8892 + }, + { + "epoch": 0.795153791130186, + "grad_norm": 2.930607430021387, + "learning_rate": 2.1214379461646005e-05, + "loss": 1.9757, + "step": 8893 + }, + { + "epoch": 0.7952432045779685, + "grad_norm": 1.3779676414211866, + "learning_rate": 2.1196547414928137e-05, + "loss": 1.9089, + "step": 8894 + }, + { + "epoch": 0.7953326180257511, + "grad_norm": 0.7381413198656419, + "learning_rate": 2.1178721977360684e-05, + "loss": 1.6917, + "step": 8895 + }, + { + "epoch": 0.7954220314735336, + "grad_norm": 2.1916555376338764, + "learning_rate": 2.1160903150438605e-05, + "loss": 1.5465, + "step": 8896 + }, + { + "epoch": 0.7955114449213162, + "grad_norm": 2.350178872698639, + "learning_rate": 2.114309093565637e-05, + "loss": 2.0781, + "step": 8897 + }, + { + "epoch": 0.7956008583690987, + "grad_norm": 0.9594370179857831, + "learning_rate": 2.112528533450786e-05, + "loss": 1.5287, + "step": 8898 + }, + { + "epoch": 0.7956902718168812, + "grad_norm": 0.6192140404660275, + "learning_rate": 2.1107486348486406e-05, + "loss": 1.4969, + "step": 8899 + }, + { + "epoch": 0.7957796852646638, + "grad_norm": 8.067598471821865, + "learning_rate": 2.1089693979084825e-05, + "loss": 6.2314, + "step": 8900 + }, + { + "epoch": 0.7958690987124464, + "grad_norm": 0.7204733410364713, + "learning_rate": 2.107190822779529e-05, + "loss": 1.9684, + "step": 8901 + }, + { + "epoch": 0.795958512160229, + "grad_norm": 1.1388723391851703, + "learning_rate": 2.1054129096109486e-05, + "loss": 1.8002, + "step": 8902 + }, + { + "epoch": 0.7960479256080114, + "grad_norm": 1.730991805345988, + "learning_rate": 2.103635658551856e-05, + "loss": 1.5673, + "step": 8903 + }, + { + "epoch": 0.796137339055794, + "grad_norm": 1.7788664798803684, + "learning_rate": 2.101859069751301e-05, + "loss": 1.7745, + "step": 8904 + }, + { + "epoch": 0.7962267525035766, + "grad_norm": 1.0293431139239353, + "learning_rate": 2.1000831433582856e-05, + "loss": 1.4781, + "step": 8905 + }, + { + "epoch": 0.7963161659513591, + "grad_norm": 1.0590015561479236, + "learning_rate": 2.0983078795217603e-05, + "loss": 1.5447, + "step": 8906 + }, + { + "epoch": 0.7964055793991416, + "grad_norm": 0.865771436663074, + "learning_rate": 2.0965332783906087e-05, + "loss": 1.2906, + "step": 8907 + }, + { + "epoch": 0.7964949928469242, + "grad_norm": 0.9542582895970422, + "learning_rate": 2.0947593401136657e-05, + "loss": 1.6286, + "step": 8908 + }, + { + "epoch": 0.7965844062947067, + "grad_norm": 1.0346995085196573, + "learning_rate": 2.0929860648397126e-05, + "loss": 1.6271, + "step": 8909 + }, + { + "epoch": 0.7966738197424893, + "grad_norm": 0.7303544237465517, + "learning_rate": 2.0912134527174664e-05, + "loss": 1.338, + "step": 8910 + }, + { + "epoch": 0.7967632331902719, + "grad_norm": 0.5943152063184814, + "learning_rate": 2.0894415038955962e-05, + "loss": 1.7462, + "step": 8911 + }, + { + "epoch": 0.7968526466380543, + "grad_norm": 1.4743163647975845, + "learning_rate": 2.0876702185227137e-05, + "loss": 1.961, + "step": 8912 + }, + { + "epoch": 0.7969420600858369, + "grad_norm": 1.0896153489316678, + "learning_rate": 2.085899596747375e-05, + "loss": 1.7552, + "step": 8913 + }, + { + "epoch": 0.7970314735336195, + "grad_norm": 0.7877035390788136, + "learning_rate": 2.084129638718081e-05, + "loss": 1.3678, + "step": 8914 + }, + { + "epoch": 0.797120886981402, + "grad_norm": 0.47933508402582975, + "learning_rate": 2.082360344583272e-05, + "loss": 1.3666, + "step": 8915 + }, + { + "epoch": 0.7972103004291845, + "grad_norm": 0.654929366642452, + "learning_rate": 2.080591714491339e-05, + "loss": 1.4256, + "step": 8916 + }, + { + "epoch": 0.7972997138769671, + "grad_norm": 0.8113579578289901, + "learning_rate": 2.0788237485906135e-05, + "loss": 1.7591, + "step": 8917 + }, + { + "epoch": 0.7973891273247496, + "grad_norm": 0.8690178436925138, + "learning_rate": 2.0770564470293775e-05, + "loss": 1.2874, + "step": 8918 + }, + { + "epoch": 0.7974785407725322, + "grad_norm": 0.6458172953958171, + "learning_rate": 2.0752898099558437e-05, + "loss": 1.5088, + "step": 8919 + }, + { + "epoch": 0.7975679542203148, + "grad_norm": 0.6023478138609298, + "learning_rate": 2.0735238375181875e-05, + "loss": 1.5565, + "step": 8920 + }, + { + "epoch": 0.7976573676680973, + "grad_norm": 0.5296281401504093, + "learning_rate": 2.0717585298645127e-05, + "loss": 1.8302, + "step": 8921 + }, + { + "epoch": 0.7977467811158798, + "grad_norm": 0.4055585187448653, + "learning_rate": 2.069993887142874e-05, + "loss": 1.3692, + "step": 8922 + }, + { + "epoch": 0.7978361945636624, + "grad_norm": 0.9481939472428204, + "learning_rate": 2.0682299095012747e-05, + "loss": 1.5034, + "step": 8923 + }, + { + "epoch": 0.797925608011445, + "grad_norm": 1.374812499611513, + "learning_rate": 2.0664665970876496e-05, + "loss": 2.0124, + "step": 8924 + }, + { + "epoch": 0.7980150214592274, + "grad_norm": 0.7114550544645324, + "learning_rate": 2.064703950049891e-05, + "loss": 1.6418, + "step": 8925 + }, + { + "epoch": 0.79810443490701, + "grad_norm": 0.543574862590364, + "learning_rate": 2.0629419685358286e-05, + "loss": 1.5103, + "step": 8926 + }, + { + "epoch": 0.7981938483547926, + "grad_norm": 0.43675719101126925, + "learning_rate": 2.0611806526932364e-05, + "loss": 1.5093, + "step": 8927 + }, + { + "epoch": 0.7982832618025751, + "grad_norm": 1.3636820711364865, + "learning_rate": 2.0594200026698363e-05, + "loss": 1.6847, + "step": 8928 + }, + { + "epoch": 0.7983726752503576, + "grad_norm": 0.7006710929546508, + "learning_rate": 2.0576600186132934e-05, + "loss": 1.6538, + "step": 8929 + }, + { + "epoch": 0.7984620886981402, + "grad_norm": 0.9801347317088681, + "learning_rate": 2.0559007006712106e-05, + "loss": 1.761, + "step": 8930 + }, + { + "epoch": 0.7985515021459227, + "grad_norm": 0.5568872024608355, + "learning_rate": 2.0541420489911413e-05, + "loss": 1.6865, + "step": 8931 + }, + { + "epoch": 0.7986409155937053, + "grad_norm": 0.5161113811855642, + "learning_rate": 2.052384063720585e-05, + "loss": 1.3759, + "step": 8932 + }, + { + "epoch": 0.7987303290414879, + "grad_norm": 1.2876595622433105, + "learning_rate": 2.0506267450069737e-05, + "loss": 1.8068, + "step": 8933 + }, + { + "epoch": 0.7988197424892703, + "grad_norm": 0.6882931603444523, + "learning_rate": 2.048870092997702e-05, + "loss": 1.4708, + "step": 8934 + }, + { + "epoch": 0.7989091559370529, + "grad_norm": 0.6623480328754255, + "learning_rate": 2.0471141078400912e-05, + "loss": 2.0711, + "step": 8935 + }, + { + "epoch": 0.7989985693848355, + "grad_norm": 0.5395348859279747, + "learning_rate": 2.0453587896814142e-05, + "loss": 1.5795, + "step": 8936 + }, + { + "epoch": 0.7990879828326181, + "grad_norm": 0.6640883006544026, + "learning_rate": 2.0436041386688932e-05, + "loss": 1.5483, + "step": 8937 + }, + { + "epoch": 0.7991773962804005, + "grad_norm": 1.3663534469138114, + "learning_rate": 2.0418501549496792e-05, + "loss": 1.6052, + "step": 8938 + }, + { + "epoch": 0.7992668097281831, + "grad_norm": 0.7163270228362316, + "learning_rate": 2.040096838670881e-05, + "loss": 1.6373, + "step": 8939 + }, + { + "epoch": 0.7993562231759657, + "grad_norm": 0.42429520410270083, + "learning_rate": 2.0383441899795518e-05, + "loss": 1.8988, + "step": 8940 + }, + { + "epoch": 0.7994456366237482, + "grad_norm": 0.8486700204005653, + "learning_rate": 2.0365922090226784e-05, + "loss": 1.6967, + "step": 8941 + }, + { + "epoch": 0.7995350500715308, + "grad_norm": 0.966479725389996, + "learning_rate": 2.034840895947199e-05, + "loss": 1.8275, + "step": 8942 + }, + { + "epoch": 0.7996244635193133, + "grad_norm": 0.5824896481752494, + "learning_rate": 2.033090250899997e-05, + "loss": 1.6695, + "step": 8943 + }, + { + "epoch": 0.7997138769670958, + "grad_norm": 0.8144129392133312, + "learning_rate": 2.0313402740278908e-05, + "loss": 1.7149, + "step": 8944 + }, + { + "epoch": 0.7998032904148784, + "grad_norm": 0.6006439701736795, + "learning_rate": 2.0295909654776524e-05, + "loss": 1.721, + "step": 8945 + }, + { + "epoch": 0.799892703862661, + "grad_norm": 0.8770750555785828, + "learning_rate": 2.0278423253959934e-05, + "loss": 1.8364, + "step": 8946 + }, + { + "epoch": 0.7999821173104434, + "grad_norm": 0.39834590430830513, + "learning_rate": 2.026094353929572e-05, + "loss": 1.4696, + "step": 8947 + }, + { + "epoch": 0.800071530758226, + "grad_norm": 0.4602234998419379, + "learning_rate": 2.024347051224985e-05, + "loss": 1.3856, + "step": 8948 + }, + { + "epoch": 0.8001609442060086, + "grad_norm": 0.6539528165625438, + "learning_rate": 2.0226004174287827e-05, + "loss": 1.7023, + "step": 8949 + }, + { + "epoch": 0.8002503576537912, + "grad_norm": 0.7190204043411917, + "learning_rate": 2.0208544526874475e-05, + "loss": 1.8595, + "step": 8950 + }, + { + "epoch": 0.8003397711015737, + "grad_norm": 0.4037355704063401, + "learning_rate": 2.0191091571474108e-05, + "loss": 1.5786, + "step": 8951 + }, + { + "epoch": 0.8004291845493562, + "grad_norm": 0.4719499120078768, + "learning_rate": 2.0173645309550548e-05, + "loss": 1.4472, + "step": 8952 + }, + { + "epoch": 0.8005185979971388, + "grad_norm": 0.5022077886658639, + "learning_rate": 2.0156205742566892e-05, + "loss": 1.5003, + "step": 8953 + }, + { + "epoch": 0.8006080114449213, + "grad_norm": 0.48794540141963794, + "learning_rate": 2.013877287198588e-05, + "loss": 1.5529, + "step": 8954 + }, + { + "epoch": 0.8006974248927039, + "grad_norm": 0.4475174017962914, + "learning_rate": 2.0121346699269516e-05, + "loss": 1.72, + "step": 8955 + }, + { + "epoch": 0.8007868383404864, + "grad_norm": 0.8909540942142223, + "learning_rate": 2.0103927225879336e-05, + "loss": 1.6311, + "step": 8956 + }, + { + "epoch": 0.8008762517882689, + "grad_norm": 0.7919057879824242, + "learning_rate": 2.008651445327633e-05, + "loss": 1.7553, + "step": 8957 + }, + { + "epoch": 0.8009656652360515, + "grad_norm": 0.7465358432497401, + "learning_rate": 2.00691083829208e-05, + "loss": 1.4951, + "step": 8958 + }, + { + "epoch": 0.8010550786838341, + "grad_norm": 0.8456693446252592, + "learning_rate": 2.0051709016272625e-05, + "loss": 1.7401, + "step": 8959 + }, + { + "epoch": 0.8011444921316166, + "grad_norm": 0.3736946567074374, + "learning_rate": 2.0034316354791062e-05, + "loss": 1.5944, + "step": 8960 + }, + { + "epoch": 0.8012339055793991, + "grad_norm": 0.5535208367181146, + "learning_rate": 2.001693039993482e-05, + "loss": 1.6994, + "step": 8961 + }, + { + "epoch": 0.8013233190271817, + "grad_norm": 0.4481973428026893, + "learning_rate": 1.9999551153162022e-05, + "loss": 1.8371, + "step": 8962 + }, + { + "epoch": 0.8014127324749643, + "grad_norm": 1.4772773709765412, + "learning_rate": 1.998217861593028e-05, + "loss": 1.8185, + "step": 8963 + }, + { + "epoch": 0.8015021459227468, + "grad_norm": 0.5481434591889549, + "learning_rate": 1.996481278969655e-05, + "loss": 1.5012, + "step": 8964 + }, + { + "epoch": 0.8015915593705293, + "grad_norm": 0.37375594969074577, + "learning_rate": 1.9947453675917316e-05, + "loss": 1.3833, + "step": 8965 + }, + { + "epoch": 0.8016809728183119, + "grad_norm": 0.5882588145130937, + "learning_rate": 1.9930101276048485e-05, + "loss": 2.0389, + "step": 8966 + }, + { + "epoch": 0.8017703862660944, + "grad_norm": 0.6183282359098647, + "learning_rate": 1.9912755591545317e-05, + "loss": 1.4189, + "step": 8967 + }, + { + "epoch": 0.801859799713877, + "grad_norm": 0.6722745311851103, + "learning_rate": 1.9895416623862662e-05, + "loss": 1.5452, + "step": 8968 + }, + { + "epoch": 0.8019492131616596, + "grad_norm": 0.6318327838570038, + "learning_rate": 1.9878084374454653e-05, + "loss": 1.7806, + "step": 8969 + }, + { + "epoch": 0.802038626609442, + "grad_norm": 0.41065743139370703, + "learning_rate": 1.986075884477494e-05, + "loss": 1.8198, + "step": 8970 + }, + { + "epoch": 0.8021280400572246, + "grad_norm": 0.5906159629534734, + "learning_rate": 1.984344003627663e-05, + "loss": 1.7343, + "step": 8971 + }, + { + "epoch": 0.8022174535050072, + "grad_norm": 0.6954447195980875, + "learning_rate": 1.9826127950412167e-05, + "loss": 1.7967, + "step": 8972 + }, + { + "epoch": 0.8023068669527897, + "grad_norm": 0.43212188925409034, + "learning_rate": 1.9808822588633535e-05, + "loss": 1.6334, + "step": 8973 + }, + { + "epoch": 0.8023962804005722, + "grad_norm": 0.7664447625641502, + "learning_rate": 1.97915239523921e-05, + "loss": 1.6718, + "step": 8974 + }, + { + "epoch": 0.8024856938483548, + "grad_norm": 0.43505314615221324, + "learning_rate": 1.9774232043138685e-05, + "loss": 1.5592, + "step": 8975 + }, + { + "epoch": 0.8025751072961373, + "grad_norm": 1.219874102973011, + "learning_rate": 1.9756946862323535e-05, + "loss": 1.5787, + "step": 8976 + }, + { + "epoch": 0.8026645207439199, + "grad_norm": 0.6795824572219007, + "learning_rate": 1.9739668411396383e-05, + "loss": 1.6861, + "step": 8977 + }, + { + "epoch": 0.8027539341917024, + "grad_norm": 0.6931508748026021, + "learning_rate": 1.9722396691806267e-05, + "loss": 1.6986, + "step": 8978 + }, + { + "epoch": 0.802843347639485, + "grad_norm": 0.4571646070697766, + "learning_rate": 1.97051317050018e-05, + "loss": 1.812, + "step": 8979 + }, + { + "epoch": 0.8029327610872675, + "grad_norm": 0.528324510141174, + "learning_rate": 1.9687873452430995e-05, + "loss": 1.6195, + "step": 8980 + }, + { + "epoch": 0.8030221745350501, + "grad_norm": 0.6255073107645673, + "learning_rate": 1.967062193554119e-05, + "loss": 1.7033, + "step": 8981 + }, + { + "epoch": 0.8031115879828327, + "grad_norm": 0.7507615020647738, + "learning_rate": 1.965337715577934e-05, + "loss": 1.7246, + "step": 8982 + }, + { + "epoch": 0.8032010014306151, + "grad_norm": 0.517554556889417, + "learning_rate": 1.9636139114591747e-05, + "loss": 1.5281, + "step": 8983 + }, + { + "epoch": 0.8032904148783977, + "grad_norm": 0.7742563682204893, + "learning_rate": 1.961890781342408e-05, + "loss": 1.9544, + "step": 8984 + }, + { + "epoch": 0.8033798283261803, + "grad_norm": 0.44442988144687273, + "learning_rate": 1.9601683253721536e-05, + "loss": 1.5801, + "step": 8985 + }, + { + "epoch": 0.8034692417739628, + "grad_norm": 0.5545827854541852, + "learning_rate": 1.9584465436928745e-05, + "loss": 1.8488, + "step": 8986 + }, + { + "epoch": 0.8035586552217453, + "grad_norm": 0.4556786827592548, + "learning_rate": 1.9567254364489694e-05, + "loss": 1.5627, + "step": 8987 + }, + { + "epoch": 0.8036480686695279, + "grad_norm": 0.8991679640393474, + "learning_rate": 1.955005003784789e-05, + "loss": 1.5787, + "step": 8988 + }, + { + "epoch": 0.8037374821173104, + "grad_norm": 0.6756295705599039, + "learning_rate": 1.9532852458446228e-05, + "loss": 1.7118, + "step": 8989 + }, + { + "epoch": 0.803826895565093, + "grad_norm": 0.6024475890344169, + "learning_rate": 1.9515661627727044e-05, + "loss": 1.4322, + "step": 8990 + }, + { + "epoch": 0.8039163090128756, + "grad_norm": 0.42617947895660924, + "learning_rate": 1.9498477547132154e-05, + "loss": 1.9715, + "step": 8991 + }, + { + "epoch": 0.804005722460658, + "grad_norm": 0.943996526045535, + "learning_rate": 1.9481300218102692e-05, + "loss": 1.9677, + "step": 8992 + }, + { + "epoch": 0.8040951359084406, + "grad_norm": 0.8892753802956983, + "learning_rate": 1.9464129642079355e-05, + "loss": 1.4978, + "step": 8993 + }, + { + "epoch": 0.8041845493562232, + "grad_norm": 0.44179648347095263, + "learning_rate": 1.9446965820502218e-05, + "loss": 1.91, + "step": 8994 + }, + { + "epoch": 0.8042739628040058, + "grad_norm": 0.4119011327023683, + "learning_rate": 1.9429808754810717e-05, + "loss": 1.5603, + "step": 8995 + }, + { + "epoch": 0.8043633762517882, + "grad_norm": 0.4446204962809333, + "learning_rate": 1.9412658446443887e-05, + "loss": 1.5254, + "step": 8996 + }, + { + "epoch": 0.8044527896995708, + "grad_norm": 0.42832313275368405, + "learning_rate": 1.9395514896840093e-05, + "loss": 1.5183, + "step": 8997 + }, + { + "epoch": 0.8045422031473534, + "grad_norm": 0.6986187492323395, + "learning_rate": 1.93783781074371e-05, + "loss": 1.4354, + "step": 8998 + }, + { + "epoch": 0.8046316165951359, + "grad_norm": 0.7799490760167144, + "learning_rate": 1.9361248079672158e-05, + "loss": 1.6188, + "step": 8999 + }, + { + "epoch": 0.8047210300429185, + "grad_norm": 0.43234947473260266, + "learning_rate": 1.934412481498198e-05, + "loss": 1.3284, + "step": 9000 + }, + { + "epoch": 0.804810443490701, + "grad_norm": 0.6109915479852845, + "learning_rate": 1.932700831480262e-05, + "loss": 1.5975, + "step": 9001 + }, + { + "epoch": 0.8048998569384835, + "grad_norm": 0.5022437357738915, + "learning_rate": 1.930989858056965e-05, + "loss": 1.6038, + "step": 9002 + }, + { + "epoch": 0.8049892703862661, + "grad_norm": 0.4439469879079757, + "learning_rate": 1.929279561371803e-05, + "loss": 1.4524, + "step": 9003 + }, + { + "epoch": 0.8050786838340487, + "grad_norm": 0.8320496957853428, + "learning_rate": 1.927569941568218e-05, + "loss": 1.8791, + "step": 9004 + }, + { + "epoch": 0.8051680972818311, + "grad_norm": 0.6514182081787294, + "learning_rate": 1.9258609987895926e-05, + "loss": 1.5116, + "step": 9005 + }, + { + "epoch": 0.8052575107296137, + "grad_norm": 0.5630782152303478, + "learning_rate": 1.9241527331792562e-05, + "loss": 1.632, + "step": 9006 + }, + { + "epoch": 0.8053469241773963, + "grad_norm": 0.4499849178031312, + "learning_rate": 1.922445144880475e-05, + "loss": 1.8575, + "step": 9007 + }, + { + "epoch": 0.8054363376251789, + "grad_norm": 0.6828355093062608, + "learning_rate": 1.9207382340364634e-05, + "loss": 1.6957, + "step": 9008 + }, + { + "epoch": 0.8055257510729614, + "grad_norm": 0.6059291460667351, + "learning_rate": 1.9190320007903796e-05, + "loss": 1.905, + "step": 9009 + }, + { + "epoch": 0.8056151645207439, + "grad_norm": 0.3942073546749884, + "learning_rate": 1.9173264452853222e-05, + "loss": 1.3902, + "step": 9010 + }, + { + "epoch": 0.8057045779685265, + "grad_norm": 0.5173695035261816, + "learning_rate": 1.9156215676643375e-05, + "loss": 1.5321, + "step": 9011 + }, + { + "epoch": 0.805793991416309, + "grad_norm": 1.0274485830478166, + "learning_rate": 1.913917368070406e-05, + "loss": 1.7624, + "step": 9012 + }, + { + "epoch": 0.8058834048640916, + "grad_norm": 0.5246034967972613, + "learning_rate": 1.912213846646459e-05, + "loss": 1.4851, + "step": 9013 + }, + { + "epoch": 0.8059728183118741, + "grad_norm": 0.6539147692920829, + "learning_rate": 1.9105110035353714e-05, + "loss": 1.5424, + "step": 9014 + }, + { + "epoch": 0.8060622317596566, + "grad_norm": 0.3653372862854768, + "learning_rate": 1.9088088388799542e-05, + "loss": 1.4992, + "step": 9015 + }, + { + "epoch": 0.8061516452074392, + "grad_norm": 0.6298792037729242, + "learning_rate": 1.9071073528229655e-05, + "loss": 1.9364, + "step": 9016 + }, + { + "epoch": 0.8062410586552218, + "grad_norm": 0.4850023639745562, + "learning_rate": 1.9054065455071136e-05, + "loss": 1.9535, + "step": 9017 + }, + { + "epoch": 0.8063304721030042, + "grad_norm": 0.37370139833353816, + "learning_rate": 1.9037064170750373e-05, + "loss": 1.5633, + "step": 9018 + }, + { + "epoch": 0.8064198855507868, + "grad_norm": 0.5413066778452067, + "learning_rate": 1.9020069676693252e-05, + "loss": 1.733, + "step": 9019 + }, + { + "epoch": 0.8065092989985694, + "grad_norm": 0.5671554827756782, + "learning_rate": 1.9003081974325122e-05, + "loss": 1.7782, + "step": 9020 + }, + { + "epoch": 0.806598712446352, + "grad_norm": 0.5790008060954608, + "learning_rate": 1.898610106507066e-05, + "loss": 1.6089, + "step": 9021 + }, + { + "epoch": 0.8066881258941345, + "grad_norm": 0.6889360959091796, + "learning_rate": 1.8969126950354055e-05, + "loss": 1.433, + "step": 9022 + }, + { + "epoch": 0.806777539341917, + "grad_norm": 4.072088215683785, + "learning_rate": 1.8952159631598922e-05, + "loss": 6.1796, + "step": 9023 + }, + { + "epoch": 0.8068669527896996, + "grad_norm": 0.7010918462942427, + "learning_rate": 1.8935199110228275e-05, + "loss": 1.5615, + "step": 9024 + }, + { + "epoch": 0.8069563662374821, + "grad_norm": 1.3072785262502034, + "learning_rate": 1.8918245387664602e-05, + "loss": 1.5729, + "step": 9025 + }, + { + "epoch": 0.8070457796852647, + "grad_norm": 1.042310639961517, + "learning_rate": 1.8901298465329743e-05, + "loss": 1.569, + "step": 9026 + }, + { + "epoch": 0.8071351931330472, + "grad_norm": 1.2221150879339455, + "learning_rate": 1.8884358344645025e-05, + "loss": 1.5033, + "step": 9027 + }, + { + "epoch": 0.8072246065808297, + "grad_norm": 0.5061808692323017, + "learning_rate": 1.886742502703125e-05, + "loss": 1.4299, + "step": 9028 + }, + { + "epoch": 0.8073140200286123, + "grad_norm": 0.7362869460251154, + "learning_rate": 1.88504985139085e-05, + "loss": 1.6106, + "step": 9029 + }, + { + "epoch": 0.8074034334763949, + "grad_norm": 0.7240272684470658, + "learning_rate": 1.883357880669646e-05, + "loss": 1.5415, + "step": 9030 + }, + { + "epoch": 0.8074928469241774, + "grad_norm": 0.818866200460353, + "learning_rate": 1.8816665906814178e-05, + "loss": 1.46, + "step": 9031 + }, + { + "epoch": 0.8075822603719599, + "grad_norm": 0.823508570488937, + "learning_rate": 1.879975981568004e-05, + "loss": 1.4775, + "step": 9032 + }, + { + "epoch": 0.8076716738197425, + "grad_norm": 0.9773752230124568, + "learning_rate": 1.8782860534711998e-05, + "loss": 1.7232, + "step": 9033 + }, + { + "epoch": 0.807761087267525, + "grad_norm": 0.6041925715033533, + "learning_rate": 1.8765968065327367e-05, + "loss": 1.4857, + "step": 9034 + }, + { + "epoch": 0.8078505007153076, + "grad_norm": 1.1145595626067495, + "learning_rate": 1.8749082408942876e-05, + "loss": 1.6391, + "step": 9035 + }, + { + "epoch": 0.8079399141630901, + "grad_norm": 0.42169437963494455, + "learning_rate": 1.8732203566974705e-05, + "loss": 1.3697, + "step": 9036 + }, + { + "epoch": 0.8080293276108726, + "grad_norm": 0.7636073098589621, + "learning_rate": 1.8715331540838487e-05, + "loss": 1.536, + "step": 9037 + }, + { + "epoch": 0.8081187410586552, + "grad_norm": 0.42134795130210295, + "learning_rate": 1.8698466331949238e-05, + "loss": 1.497, + "step": 9038 + }, + { + "epoch": 0.8082081545064378, + "grad_norm": 1.057307245946494, + "learning_rate": 1.8681607941721425e-05, + "loss": 1.8065, + "step": 9039 + }, + { + "epoch": 0.8082975679542204, + "grad_norm": 0.7412233075926226, + "learning_rate": 1.866475637156898e-05, + "loss": 1.8157, + "step": 9040 + }, + { + "epoch": 0.8083869814020028, + "grad_norm": 1.01773295099773, + "learning_rate": 1.8647911622905168e-05, + "loss": 1.5605, + "step": 9041 + }, + { + "epoch": 0.8084763948497854, + "grad_norm": 0.98308775778128, + "learning_rate": 1.8631073697142754e-05, + "loss": 1.8527, + "step": 9042 + }, + { + "epoch": 0.808565808297568, + "grad_norm": 0.35966225544665914, + "learning_rate": 1.8614242595693908e-05, + "loss": 1.5651, + "step": 9043 + }, + { + "epoch": 0.8086552217453505, + "grad_norm": 0.46681069681271037, + "learning_rate": 1.8597418319970262e-05, + "loss": 1.6207, + "step": 9044 + }, + { + "epoch": 0.808744635193133, + "grad_norm": 0.5313292044018106, + "learning_rate": 1.8580600871382857e-05, + "loss": 1.3818, + "step": 9045 + }, + { + "epoch": 0.8088340486409156, + "grad_norm": 0.45563103967748286, + "learning_rate": 1.8563790251342095e-05, + "loss": 1.3093, + "step": 9046 + }, + { + "epoch": 0.8089234620886981, + "grad_norm": 0.492165875178899, + "learning_rate": 1.85469864612579e-05, + "loss": 1.4337, + "step": 9047 + }, + { + "epoch": 0.8090128755364807, + "grad_norm": 0.4020247202364185, + "learning_rate": 1.8530189502539607e-05, + "loss": 1.2537, + "step": 9048 + }, + { + "epoch": 0.8091022889842633, + "grad_norm": 0.4021196941014107, + "learning_rate": 1.8513399376595895e-05, + "loss": 1.6581, + "step": 9049 + }, + { + "epoch": 0.8091917024320457, + "grad_norm": 0.6114787191914478, + "learning_rate": 1.849661608483495e-05, + "loss": 1.5335, + "step": 9050 + }, + { + "epoch": 0.8092811158798283, + "grad_norm": 0.8478195004363677, + "learning_rate": 1.847983962866443e-05, + "loss": 1.9264, + "step": 9051 + }, + { + "epoch": 0.8093705293276109, + "grad_norm": 0.7505842851426588, + "learning_rate": 1.846307000949129e-05, + "loss": 1.3005, + "step": 9052 + }, + { + "epoch": 0.8094599427753935, + "grad_norm": 0.4008701283410487, + "learning_rate": 1.844630722872199e-05, + "loss": 1.5333, + "step": 9053 + }, + { + "epoch": 0.8095493562231759, + "grad_norm": 0.44411798867671815, + "learning_rate": 1.8429551287762435e-05, + "loss": 1.5712, + "step": 9054 + }, + { + "epoch": 0.8096387696709585, + "grad_norm": 0.8514476817357096, + "learning_rate": 1.8412802188017885e-05, + "loss": 1.5268, + "step": 9055 + }, + { + "epoch": 0.8097281831187411, + "grad_norm": 0.6387046339768798, + "learning_rate": 1.839605993089307e-05, + "loss": 1.5788, + "step": 9056 + }, + { + "epoch": 0.8098175965665236, + "grad_norm": 0.6884304327260468, + "learning_rate": 1.8379324517792163e-05, + "loss": 1.5486, + "step": 9057 + }, + { + "epoch": 0.8099070100143062, + "grad_norm": 0.5454858508453034, + "learning_rate": 1.8362595950118733e-05, + "loss": 1.4985, + "step": 9058 + }, + { + "epoch": 0.8099964234620887, + "grad_norm": 0.4072937106482229, + "learning_rate": 1.8345874229275816e-05, + "loss": 1.4803, + "step": 9059 + }, + { + "epoch": 0.8100858369098712, + "grad_norm": 0.39909703811367775, + "learning_rate": 1.8329159356665793e-05, + "loss": 1.5281, + "step": 9060 + }, + { + "epoch": 0.8101752503576538, + "grad_norm": 2.0446575027448355, + "learning_rate": 1.8312451333690538e-05, + "loss": 6.1844, + "step": 9061 + }, + { + "epoch": 0.8102646638054364, + "grad_norm": 0.44758694937791743, + "learning_rate": 1.8295750161751334e-05, + "loss": 1.7091, + "step": 9062 + }, + { + "epoch": 0.8103540772532188, + "grad_norm": 1.1763799664040426, + "learning_rate": 1.8279055842248915e-05, + "loss": 1.6012, + "step": 9063 + }, + { + "epoch": 0.8104434907010014, + "grad_norm": 1.1209365860187397, + "learning_rate": 1.826236837658334e-05, + "loss": 1.9578, + "step": 9064 + }, + { + "epoch": 0.810532904148784, + "grad_norm": 0.40926583791829674, + "learning_rate": 1.8245687766154262e-05, + "loss": 1.334, + "step": 9065 + }, + { + "epoch": 0.8106223175965666, + "grad_norm": 0.5778137212248341, + "learning_rate": 1.822901401236059e-05, + "loss": 1.6169, + "step": 9066 + }, + { + "epoch": 0.810711731044349, + "grad_norm": 0.8312894796881477, + "learning_rate": 1.821234711660077e-05, + "loss": 2.0421, + "step": 9067 + }, + { + "epoch": 0.8108011444921316, + "grad_norm": 0.9812948561448509, + "learning_rate": 1.819568708027264e-05, + "loss": 1.8079, + "step": 9068 + }, + { + "epoch": 0.8108905579399142, + "grad_norm": 0.8333399729507986, + "learning_rate": 1.817903390477341e-05, + "loss": 1.4078, + "step": 9069 + }, + { + "epoch": 0.8109799713876967, + "grad_norm": 0.9116299335409241, + "learning_rate": 1.8162387591499796e-05, + "loss": 1.5186, + "step": 9070 + }, + { + "epoch": 0.8110693848354793, + "grad_norm": 0.6225592979818272, + "learning_rate": 1.8145748141847908e-05, + "loss": 1.5053, + "step": 9071 + }, + { + "epoch": 0.8111587982832618, + "grad_norm": 0.7760084367273029, + "learning_rate": 1.8129115557213262e-05, + "loss": 1.7689, + "step": 9072 + }, + { + "epoch": 0.8112482117310443, + "grad_norm": 0.7438982153451691, + "learning_rate": 1.811248983899082e-05, + "loss": 1.6774, + "step": 9073 + }, + { + "epoch": 0.8113376251788269, + "grad_norm": 0.8336887837230368, + "learning_rate": 1.809587098857498e-05, + "loss": 1.6542, + "step": 9074 + }, + { + "epoch": 0.8114270386266095, + "grad_norm": 0.6367268708341894, + "learning_rate": 1.8079259007359506e-05, + "loss": 1.6057, + "step": 9075 + }, + { + "epoch": 0.8115164520743919, + "grad_norm": 0.7681920998416515, + "learning_rate": 1.8062653896737647e-05, + "loss": 1.8019, + "step": 9076 + }, + { + "epoch": 0.8116058655221745, + "grad_norm": 0.589755851501819, + "learning_rate": 1.804605565810207e-05, + "loss": 1.6442, + "step": 9077 + }, + { + "epoch": 0.8116952789699571, + "grad_norm": 3.7023622181737097, + "learning_rate": 1.8029464292844778e-05, + "loss": 6.267, + "step": 9078 + }, + { + "epoch": 0.8117846924177397, + "grad_norm": 1.3257408288109178, + "learning_rate": 1.8012879802357374e-05, + "loss": 2.1816, + "step": 9079 + }, + { + "epoch": 0.8118741058655222, + "grad_norm": 0.8863944738070563, + "learning_rate": 1.79963021880307e-05, + "loss": 1.5382, + "step": 9080 + }, + { + "epoch": 0.8119635193133047, + "grad_norm": 0.8049964543210352, + "learning_rate": 1.797973145125512e-05, + "loss": 1.3723, + "step": 9081 + }, + { + "epoch": 0.8120529327610873, + "grad_norm": 0.7019310528256006, + "learning_rate": 1.7963167593420438e-05, + "loss": 1.728, + "step": 9082 + }, + { + "epoch": 0.8121423462088698, + "grad_norm": 1.1900804654812192, + "learning_rate": 1.7946610615915792e-05, + "loss": 2.003, + "step": 9083 + }, + { + "epoch": 0.8122317596566524, + "grad_norm": 0.4281352022597375, + "learning_rate": 1.793006052012981e-05, + "loss": 1.5194, + "step": 9084 + }, + { + "epoch": 0.8123211731044349, + "grad_norm": 1.1219470329502623, + "learning_rate": 1.7913517307450544e-05, + "loss": 1.7636, + "step": 9085 + }, + { + "epoch": 0.8124105865522174, + "grad_norm": 0.9139856354750135, + "learning_rate": 1.7896980979265443e-05, + "loss": 1.9352, + "step": 9086 + }, + { + "epoch": 0.8125, + "grad_norm": 0.5814317774221361, + "learning_rate": 1.7880451536961394e-05, + "loss": 1.4249, + "step": 9087 + }, + { + "epoch": 0.8125894134477826, + "grad_norm": 0.8030205103508665, + "learning_rate": 1.7863928981924726e-05, + "loss": 1.392, + "step": 9088 + }, + { + "epoch": 0.8126788268955651, + "grad_norm": 0.9301399871793422, + "learning_rate": 1.7847413315541118e-05, + "loss": 1.6904, + "step": 9089 + }, + { + "epoch": 0.8127682403433476, + "grad_norm": 1.0253823886083906, + "learning_rate": 1.7830904539195726e-05, + "loss": 1.5713, + "step": 9090 + }, + { + "epoch": 0.8128576537911302, + "grad_norm": 0.643383822603743, + "learning_rate": 1.7814402654273167e-05, + "loss": 1.786, + "step": 9091 + }, + { + "epoch": 0.8129470672389127, + "grad_norm": 0.7214026355272533, + "learning_rate": 1.7797907662157355e-05, + "loss": 1.7327, + "step": 9092 + }, + { + "epoch": 0.8130364806866953, + "grad_norm": 1.3652531597467632, + "learning_rate": 1.7781419564231805e-05, + "loss": 1.6525, + "step": 9093 + }, + { + "epoch": 0.8131258941344778, + "grad_norm": 1.2591312147633527, + "learning_rate": 1.776493836187927e-05, + "loss": 1.8708, + "step": 9094 + }, + { + "epoch": 0.8132153075822603, + "grad_norm": 0.8215723669857175, + "learning_rate": 1.774846405648204e-05, + "loss": 1.636, + "step": 9095 + }, + { + "epoch": 0.8133047210300429, + "grad_norm": 0.5155007207687636, + "learning_rate": 1.7731996649421802e-05, + "loss": 1.3329, + "step": 9096 + }, + { + "epoch": 0.8133941344778255, + "grad_norm": 0.9095651065510195, + "learning_rate": 1.771553614207967e-05, + "loss": 1.541, + "step": 9097 + }, + { + "epoch": 0.8134835479256081, + "grad_norm": 1.1813335787823294, + "learning_rate": 1.769908253583612e-05, + "loss": 1.6906, + "step": 9098 + }, + { + "epoch": 0.8135729613733905, + "grad_norm": 0.6551072658037062, + "learning_rate": 1.7682635832071125e-05, + "loss": 1.5951, + "step": 9099 + }, + { + "epoch": 0.8136623748211731, + "grad_norm": 0.9254788783447953, + "learning_rate": 1.766619603216405e-05, + "loss": 1.6001, + "step": 9100 + }, + { + "epoch": 0.8137517882689557, + "grad_norm": 0.43113523873712056, + "learning_rate": 1.7649763137493682e-05, + "loss": 1.4133, + "step": 9101 + }, + { + "epoch": 0.8138412017167382, + "grad_norm": 0.8223129520580248, + "learning_rate": 1.7633337149438246e-05, + "loss": 1.2861, + "step": 9102 + }, + { + "epoch": 0.8139306151645207, + "grad_norm": 1.166110448569968, + "learning_rate": 1.7616918069375322e-05, + "loss": 1.9223, + "step": 9103 + }, + { + "epoch": 0.8140200286123033, + "grad_norm": 1.0790727601883674, + "learning_rate": 1.7600505898681997e-05, + "loss": 1.3782, + "step": 9104 + }, + { + "epoch": 0.8141094420600858, + "grad_norm": 0.4962538544173503, + "learning_rate": 1.7584100638734745e-05, + "loss": 1.4996, + "step": 9105 + }, + { + "epoch": 0.8141988555078684, + "grad_norm": 0.775606493186888, + "learning_rate": 1.7567702290909393e-05, + "loss": 1.9049, + "step": 9106 + }, + { + "epoch": 0.814288268955651, + "grad_norm": 0.49430786404255966, + "learning_rate": 1.7551310856581316e-05, + "loss": 1.7358, + "step": 9107 + }, + { + "epoch": 0.8143776824034334, + "grad_norm": 1.2558405317279777, + "learning_rate": 1.7534926337125257e-05, + "loss": 1.802, + "step": 9108 + }, + { + "epoch": 0.814467095851216, + "grad_norm": 0.8402306587050054, + "learning_rate": 1.751854873391531e-05, + "loss": 1.4417, + "step": 9109 + }, + { + "epoch": 0.8145565092989986, + "grad_norm": 0.9767531837289276, + "learning_rate": 1.750217804832506e-05, + "loss": 1.6101, + "step": 9110 + }, + { + "epoch": 0.8146459227467812, + "grad_norm": 0.8823350174935795, + "learning_rate": 1.7485814281727532e-05, + "loss": 1.2448, + "step": 9111 + }, + { + "epoch": 0.8147353361945636, + "grad_norm": 0.7438124494841041, + "learning_rate": 1.7469457435495063e-05, + "loss": 1.7578, + "step": 9112 + }, + { + "epoch": 0.8148247496423462, + "grad_norm": 0.9661581410226368, + "learning_rate": 1.7453107510999568e-05, + "loss": 2.0521, + "step": 9113 + }, + { + "epoch": 0.8149141630901288, + "grad_norm": 0.9576498488872817, + "learning_rate": 1.7436764509612237e-05, + "loss": 1.6828, + "step": 9114 + }, + { + "epoch": 0.8150035765379113, + "grad_norm": 1.0502667270604327, + "learning_rate": 1.742042843270375e-05, + "loss": 2.0475, + "step": 9115 + }, + { + "epoch": 0.8150929899856938, + "grad_norm": 0.6142560709701723, + "learning_rate": 1.7404099281644237e-05, + "loss": 1.7495, + "step": 9116 + }, + { + "epoch": 0.8151824034334764, + "grad_norm": 0.4610919060494467, + "learning_rate": 1.7387777057803134e-05, + "loss": 1.6368, + "step": 9117 + }, + { + "epoch": 0.8152718168812589, + "grad_norm": 0.6030079044358985, + "learning_rate": 1.737146176254939e-05, + "loss": 1.734, + "step": 9118 + }, + { + "epoch": 0.8153612303290415, + "grad_norm": 0.5339687822450468, + "learning_rate": 1.735515339725137e-05, + "loss": 1.54, + "step": 9119 + }, + { + "epoch": 0.8154506437768241, + "grad_norm": 0.3580305745521264, + "learning_rate": 1.7338851963276825e-05, + "loss": 1.5288, + "step": 9120 + }, + { + "epoch": 0.8155400572246065, + "grad_norm": 0.8662253507090908, + "learning_rate": 1.7322557461992926e-05, + "loss": 1.6216, + "step": 9121 + }, + { + "epoch": 0.8156294706723891, + "grad_norm": 0.6836491268892675, + "learning_rate": 1.7306269894766312e-05, + "loss": 1.8454, + "step": 9122 + }, + { + "epoch": 0.8157188841201717, + "grad_norm": 0.6100806657113349, + "learning_rate": 1.728998926296296e-05, + "loss": 1.3753, + "step": 9123 + }, + { + "epoch": 0.8158082975679543, + "grad_norm": 0.4588467592937981, + "learning_rate": 1.727371556794831e-05, + "loss": 1.467, + "step": 9124 + }, + { + "epoch": 0.8158977110157367, + "grad_norm": 0.5137765394936664, + "learning_rate": 1.725744881108725e-05, + "loss": 1.5973, + "step": 9125 + }, + { + "epoch": 0.8159871244635193, + "grad_norm": 0.6238417403047418, + "learning_rate": 1.7241188993743984e-05, + "loss": 1.8657, + "step": 9126 + }, + { + "epoch": 0.8160765379113019, + "grad_norm": 0.9030469852036006, + "learning_rate": 1.7224936117282276e-05, + "loss": 1.812, + "step": 9127 + }, + { + "epoch": 0.8161659513590844, + "grad_norm": 0.46094228038399143, + "learning_rate": 1.7208690183065236e-05, + "loss": 1.457, + "step": 9128 + }, + { + "epoch": 0.816255364806867, + "grad_norm": 0.6685664191966032, + "learning_rate": 1.719245119245534e-05, + "loss": 1.5118, + "step": 9129 + }, + { + "epoch": 0.8163447782546495, + "grad_norm": 0.3394962313582769, + "learning_rate": 1.7176219146814542e-05, + "loss": 1.4498, + "step": 9130 + }, + { + "epoch": 0.816434191702432, + "grad_norm": 0.5307070166138222, + "learning_rate": 1.715999404750426e-05, + "loss": 1.7698, + "step": 9131 + }, + { + "epoch": 0.8165236051502146, + "grad_norm": 0.5673271585991841, + "learning_rate": 1.7143775895885195e-05, + "loss": 1.3864, + "step": 9132 + }, + { + "epoch": 0.8166130185979972, + "grad_norm": 0.5479316449093935, + "learning_rate": 1.712756469331759e-05, + "loss": 1.7613, + "step": 9133 + }, + { + "epoch": 0.8167024320457796, + "grad_norm": 0.5789425903483498, + "learning_rate": 1.7111360441161038e-05, + "loss": 1.6106, + "step": 9134 + }, + { + "epoch": 0.8167918454935622, + "grad_norm": 0.4144141035872836, + "learning_rate": 1.7095163140774596e-05, + "loss": 1.532, + "step": 9135 + }, + { + "epoch": 0.8168812589413448, + "grad_norm": 0.6250261892350881, + "learning_rate": 1.707897279351671e-05, + "loss": 1.5942, + "step": 9136 + }, + { + "epoch": 0.8169706723891274, + "grad_norm": 3.7986369133927846, + "learning_rate": 1.7062789400745215e-05, + "loss": 6.1238, + "step": 9137 + }, + { + "epoch": 0.8170600858369099, + "grad_norm": 0.9527832519173114, + "learning_rate": 1.704661296381741e-05, + "loss": 1.8744, + "step": 9138 + }, + { + "epoch": 0.8171494992846924, + "grad_norm": 0.4691252822874966, + "learning_rate": 1.703044348409002e-05, + "loss": 1.4323, + "step": 9139 + }, + { + "epoch": 0.817238912732475, + "grad_norm": 0.9395926148745312, + "learning_rate": 1.701428096291908e-05, + "loss": 2.001, + "step": 9140 + }, + { + "epoch": 0.8173283261802575, + "grad_norm": 3.4884803335113452, + "learning_rate": 1.6998125401660202e-05, + "loss": 6.1982, + "step": 9141 + }, + { + "epoch": 0.8174177396280401, + "grad_norm": 0.6197977027978798, + "learning_rate": 1.698197680166832e-05, + "loss": 1.4156, + "step": 9142 + }, + { + "epoch": 0.8175071530758226, + "grad_norm": 0.6594180167649204, + "learning_rate": 1.6965835164297773e-05, + "loss": 1.5007, + "step": 9143 + }, + { + "epoch": 0.8175965665236051, + "grad_norm": 1.3489322601047753, + "learning_rate": 1.6949700490902344e-05, + "loss": 1.824, + "step": 9144 + }, + { + "epoch": 0.8176859799713877, + "grad_norm": 0.6629520927549148, + "learning_rate": 1.693357278283526e-05, + "loss": 1.7752, + "step": 9145 + }, + { + "epoch": 0.8177753934191703, + "grad_norm": 0.6789208854084103, + "learning_rate": 1.6917452041449077e-05, + "loss": 1.7573, + "step": 9146 + }, + { + "epoch": 0.8178648068669528, + "grad_norm": 0.6766646701565487, + "learning_rate": 1.6901338268095866e-05, + "loss": 1.4727, + "step": 9147 + }, + { + "epoch": 0.8179542203147353, + "grad_norm": 0.9343194002093479, + "learning_rate": 1.688523146412705e-05, + "loss": 1.5391, + "step": 9148 + }, + { + "epoch": 0.8180436337625179, + "grad_norm": 0.6377114256601653, + "learning_rate": 1.68691316308935e-05, + "loss": 1.2961, + "step": 9149 + }, + { + "epoch": 0.8181330472103004, + "grad_norm": 0.6461418378023706, + "learning_rate": 1.6853038769745467e-05, + "loss": 1.4934, + "step": 9150 + }, + { + "epoch": 0.818222460658083, + "grad_norm": 0.5210892630081975, + "learning_rate": 1.6836952882032698e-05, + "loss": 1.6342, + "step": 9151 + }, + { + "epoch": 0.8183118741058655, + "grad_norm": 0.7694662694394581, + "learning_rate": 1.682087396910422e-05, + "loss": 1.5857, + "step": 9152 + }, + { + "epoch": 0.818401287553648, + "grad_norm": 0.4991825021476439, + "learning_rate": 1.68048020323086e-05, + "loss": 1.7914, + "step": 9153 + }, + { + "epoch": 0.8184907010014306, + "grad_norm": 0.9689713165400681, + "learning_rate": 1.6788737072993744e-05, + "loss": 1.6762, + "step": 9154 + }, + { + "epoch": 0.8185801144492132, + "grad_norm": 0.7572483983496991, + "learning_rate": 1.6772679092507025e-05, + "loss": 1.3216, + "step": 9155 + }, + { + "epoch": 0.8186695278969958, + "grad_norm": 0.6770751825295317, + "learning_rate": 1.6756628092195214e-05, + "loss": 1.6904, + "step": 9156 + }, + { + "epoch": 0.8187589413447782, + "grad_norm": 0.6996575617819096, + "learning_rate": 1.6740584073404454e-05, + "loss": 1.6572, + "step": 9157 + }, + { + "epoch": 0.8188483547925608, + "grad_norm": 0.46914065732672217, + "learning_rate": 1.6724547037480355e-05, + "loss": 1.6068, + "step": 9158 + }, + { + "epoch": 0.8189377682403434, + "grad_norm": 0.8412468434039497, + "learning_rate": 1.6708516985767953e-05, + "loss": 1.5721, + "step": 9159 + }, + { + "epoch": 0.8190271816881259, + "grad_norm": 0.6139012045973221, + "learning_rate": 1.6692493919611606e-05, + "loss": 1.7273, + "step": 9160 + }, + { + "epoch": 0.8191165951359084, + "grad_norm": 0.5863939329268009, + "learning_rate": 1.6676477840355166e-05, + "loss": 1.739, + "step": 9161 + }, + { + "epoch": 0.819206008583691, + "grad_norm": 1.0580945090813407, + "learning_rate": 1.666046874934195e-05, + "loss": 1.665, + "step": 9162 + }, + { + "epoch": 0.8192954220314735, + "grad_norm": 0.5853212935373445, + "learning_rate": 1.6644466647914546e-05, + "loss": 1.7027, + "step": 9163 + }, + { + "epoch": 0.8193848354792561, + "grad_norm": 0.7398290766174012, + "learning_rate": 1.662847153741506e-05, + "loss": 1.7836, + "step": 9164 + }, + { + "epoch": 0.8194742489270386, + "grad_norm": 0.5391203920482417, + "learning_rate": 1.6612483419185e-05, + "loss": 1.1391, + "step": 9165 + }, + { + "epoch": 0.8195636623748211, + "grad_norm": 0.634603319167339, + "learning_rate": 1.659650229456522e-05, + "loss": 1.5316, + "step": 9166 + }, + { + "epoch": 0.8196530758226037, + "grad_norm": 0.7933064513042736, + "learning_rate": 1.658052816489607e-05, + "loss": 1.2884, + "step": 9167 + }, + { + "epoch": 0.8197424892703863, + "grad_norm": 0.6344152546619144, + "learning_rate": 1.656456103151728e-05, + "loss": 1.7148, + "step": 9168 + }, + { + "epoch": 0.8198319027181689, + "grad_norm": 0.514575049775389, + "learning_rate": 1.6548600895767997e-05, + "loss": 1.3685, + "step": 9169 + }, + { + "epoch": 0.8199213161659513, + "grad_norm": 0.5979785051820528, + "learning_rate": 1.6532647758986786e-05, + "loss": 1.6132, + "step": 9170 + }, + { + "epoch": 0.8200107296137339, + "grad_norm": 0.6393030697308959, + "learning_rate": 1.6516701622511588e-05, + "loss": 1.5123, + "step": 9171 + }, + { + "epoch": 0.8201001430615165, + "grad_norm": 0.8903673564652961, + "learning_rate": 1.65007624876798e-05, + "loss": 1.8458, + "step": 9172 + }, + { + "epoch": 0.820189556509299, + "grad_norm": 0.6756079060404864, + "learning_rate": 1.6484830355828242e-05, + "loss": 1.4179, + "step": 9173 + }, + { + "epoch": 0.8202789699570815, + "grad_norm": 0.6268545913938118, + "learning_rate": 1.6468905228293073e-05, + "loss": 1.876, + "step": 9174 + }, + { + "epoch": 0.8203683834048641, + "grad_norm": 0.3784212532698263, + "learning_rate": 1.6452987106409935e-05, + "loss": 1.4291, + "step": 9175 + }, + { + "epoch": 0.8204577968526466, + "grad_norm": 0.6146561018328373, + "learning_rate": 1.6437075991513905e-05, + "loss": 1.9067, + "step": 9176 + }, + { + "epoch": 0.8205472103004292, + "grad_norm": 0.6292982155816707, + "learning_rate": 1.6421171884939368e-05, + "loss": 1.7792, + "step": 9177 + }, + { + "epoch": 0.8206366237482118, + "grad_norm": 0.7725231547361482, + "learning_rate": 1.640527478802021e-05, + "loss": 1.7803, + "step": 9178 + }, + { + "epoch": 0.8207260371959942, + "grad_norm": 0.6374273009592472, + "learning_rate": 1.638938470208973e-05, + "loss": 1.627, + "step": 9179 + }, + { + "epoch": 0.8208154506437768, + "grad_norm": 0.7522628838984833, + "learning_rate": 1.6373501628480535e-05, + "loss": 1.8003, + "step": 9180 + }, + { + "epoch": 0.8209048640915594, + "grad_norm": 0.6153199362247048, + "learning_rate": 1.6357625568524783e-05, + "loss": 1.3687, + "step": 9181 + }, + { + "epoch": 0.820994277539342, + "grad_norm": 0.5767110301204266, + "learning_rate": 1.6341756523553954e-05, + "loss": 1.4201, + "step": 9182 + }, + { + "epoch": 0.8210836909871244, + "grad_norm": 0.8178482234455751, + "learning_rate": 1.6325894494898975e-05, + "loss": 1.513, + "step": 9183 + }, + { + "epoch": 0.821173104434907, + "grad_norm": 0.43243502247478693, + "learning_rate": 1.631003948389016e-05, + "loss": 1.6949, + "step": 9184 + }, + { + "epoch": 0.8212625178826896, + "grad_norm": 0.9171885888202121, + "learning_rate": 1.629419149185729e-05, + "loss": 2.1611, + "step": 9185 + }, + { + "epoch": 0.8213519313304721, + "grad_norm": 0.6400737298930187, + "learning_rate": 1.627835052012947e-05, + "loss": 1.4284, + "step": 9186 + }, + { + "epoch": 0.8214413447782547, + "grad_norm": 0.3744407079052654, + "learning_rate": 1.626251657003528e-05, + "loss": 1.3578, + "step": 9187 + }, + { + "epoch": 0.8215307582260372, + "grad_norm": 0.700833265446914, + "learning_rate": 1.6246689642902725e-05, + "loss": 1.9048, + "step": 9188 + }, + { + "epoch": 0.8216201716738197, + "grad_norm": 0.3685088341609464, + "learning_rate": 1.6230869740059106e-05, + "loss": 1.3633, + "step": 9189 + }, + { + "epoch": 0.8217095851216023, + "grad_norm": 0.41112891754640196, + "learning_rate": 1.6215056862831324e-05, + "loss": 1.4829, + "step": 9190 + }, + { + "epoch": 0.8217989985693849, + "grad_norm": 0.6828525440134243, + "learning_rate": 1.6199251012545512e-05, + "loss": 1.5885, + "step": 9191 + }, + { + "epoch": 0.8218884120171673, + "grad_norm": 1.1980510946991172, + "learning_rate": 1.6183452190527316e-05, + "loss": 1.9835, + "step": 9192 + }, + { + "epoch": 0.8219778254649499, + "grad_norm": 0.49056750551166955, + "learning_rate": 1.616766039810178e-05, + "loss": 1.7173, + "step": 9193 + }, + { + "epoch": 0.8220672389127325, + "grad_norm": 0.5143601044316046, + "learning_rate": 1.6151875636593306e-05, + "loss": 1.5123, + "step": 9194 + }, + { + "epoch": 0.822156652360515, + "grad_norm": 0.8817643772528395, + "learning_rate": 1.613609790732572e-05, + "loss": 1.7934, + "step": 9195 + }, + { + "epoch": 0.8222460658082976, + "grad_norm": 0.4640596226638034, + "learning_rate": 1.6120327211622375e-05, + "loss": 1.2408, + "step": 9196 + }, + { + "epoch": 0.8223354792560801, + "grad_norm": 0.7602895816723405, + "learning_rate": 1.6104563550805875e-05, + "loss": 1.5552, + "step": 9197 + }, + { + "epoch": 0.8224248927038627, + "grad_norm": 0.4914856676627757, + "learning_rate": 1.6088806926198297e-05, + "loss": 1.7641, + "step": 9198 + }, + { + "epoch": 0.8225143061516452, + "grad_norm": 0.5100070518371163, + "learning_rate": 1.6073057339121166e-05, + "loss": 1.582, + "step": 9199 + }, + { + "epoch": 0.8226037195994278, + "grad_norm": 0.6882284984367286, + "learning_rate": 1.605731479089534e-05, + "loss": 1.3892, + "step": 9200 + }, + { + "epoch": 0.8226931330472103, + "grad_norm": 0.802337819266485, + "learning_rate": 1.6041579282841145e-05, + "loss": 1.537, + "step": 9201 + }, + { + "epoch": 0.8227825464949928, + "grad_norm": 0.5668690062014227, + "learning_rate": 1.6025850816278297e-05, + "loss": 1.5714, + "step": 9202 + }, + { + "epoch": 0.8228719599427754, + "grad_norm": 0.8053524776183673, + "learning_rate": 1.601012939252592e-05, + "loss": 1.7329, + "step": 9203 + }, + { + "epoch": 0.822961373390558, + "grad_norm": 0.5019660318585969, + "learning_rate": 1.5994415012902587e-05, + "loss": 1.9606, + "step": 9204 + }, + { + "epoch": 0.8230507868383404, + "grad_norm": 0.37438577132927325, + "learning_rate": 1.597870767872619e-05, + "loss": 1.5068, + "step": 9205 + }, + { + "epoch": 0.823140200286123, + "grad_norm": 1.1241884823810129, + "learning_rate": 1.5963007391314113e-05, + "loss": 1.6591, + "step": 9206 + }, + { + "epoch": 0.8232296137339056, + "grad_norm": 0.81292207003051, + "learning_rate": 1.5947314151983105e-05, + "loss": 1.9548, + "step": 9207 + }, + { + "epoch": 0.8233190271816881, + "grad_norm": 0.5280082995663966, + "learning_rate": 1.5931627962049378e-05, + "loss": 1.7566, + "step": 9208 + }, + { + "epoch": 0.8234084406294707, + "grad_norm": 0.6193069342957817, + "learning_rate": 1.591594882282844e-05, + "loss": 1.733, + "step": 9209 + }, + { + "epoch": 0.8234978540772532, + "grad_norm": 0.6795021028798665, + "learning_rate": 1.5900276735635367e-05, + "loss": 1.4751, + "step": 9210 + }, + { + "epoch": 0.8235872675250357, + "grad_norm": 0.3728266033771286, + "learning_rate": 1.5884611701784504e-05, + "loss": 1.4201, + "step": 9211 + }, + { + "epoch": 0.8236766809728183, + "grad_norm": 0.8403163290255499, + "learning_rate": 1.5868953722589663e-05, + "loss": 1.7875, + "step": 9212 + }, + { + "epoch": 0.8237660944206009, + "grad_norm": 0.5458807502026644, + "learning_rate": 1.585330279936409e-05, + "loss": 1.5858, + "step": 9213 + }, + { + "epoch": 0.8238555078683834, + "grad_norm": 0.5601410857196565, + "learning_rate": 1.5837658933420375e-05, + "loss": 1.726, + "step": 9214 + }, + { + "epoch": 0.8239449213161659, + "grad_norm": 0.5889101865276896, + "learning_rate": 1.5822022126070556e-05, + "loss": 1.4009, + "step": 9215 + }, + { + "epoch": 0.8240343347639485, + "grad_norm": 0.5481758519859096, + "learning_rate": 1.580639237862608e-05, + "loss": 1.9114, + "step": 9216 + }, + { + "epoch": 0.8241237482117311, + "grad_norm": 0.42737132951307144, + "learning_rate": 1.57907696923978e-05, + "loss": 1.5843, + "step": 9217 + }, + { + "epoch": 0.8242131616595136, + "grad_norm": 0.5003661509370715, + "learning_rate": 1.5775154068695963e-05, + "loss": 1.2405, + "step": 9218 + }, + { + "epoch": 0.8243025751072961, + "grad_norm": 0.39663080130894046, + "learning_rate": 1.5759545508830252e-05, + "loss": 1.508, + "step": 9219 + }, + { + "epoch": 0.8243919885550787, + "grad_norm": 0.3974539276909978, + "learning_rate": 1.5743944014109713e-05, + "loss": 1.7082, + "step": 9220 + }, + { + "epoch": 0.8244814020028612, + "grad_norm": 0.4198078398088354, + "learning_rate": 1.5728349585842827e-05, + "loss": 1.4748, + "step": 9221 + }, + { + "epoch": 0.8245708154506438, + "grad_norm": 0.5698516108646444, + "learning_rate": 1.571276222533751e-05, + "loss": 1.3718, + "step": 9222 + }, + { + "epoch": 0.8246602288984263, + "grad_norm": 0.5684578622318325, + "learning_rate": 1.5697181933900985e-05, + "loss": 1.503, + "step": 9223 + }, + { + "epoch": 0.8247496423462088, + "grad_norm": 0.7022846867293439, + "learning_rate": 1.5681608712840046e-05, + "loss": 1.6727, + "step": 9224 + }, + { + "epoch": 0.8248390557939914, + "grad_norm": 0.3798543167581772, + "learning_rate": 1.5666042563460737e-05, + "loss": 1.5255, + "step": 9225 + }, + { + "epoch": 0.824928469241774, + "grad_norm": 0.3487479971515124, + "learning_rate": 1.565048348706858e-05, + "loss": 1.1226, + "step": 9226 + }, + { + "epoch": 0.8250178826895566, + "grad_norm": 0.35120984381292497, + "learning_rate": 1.563493148496853e-05, + "loss": 1.2885, + "step": 9227 + }, + { + "epoch": 0.825107296137339, + "grad_norm": 0.5032212422333977, + "learning_rate": 1.5619386558464865e-05, + "loss": 1.4671, + "step": 9228 + }, + { + "epoch": 0.8251967095851216, + "grad_norm": 0.43258250604003906, + "learning_rate": 1.5603848708861347e-05, + "loss": 1.1977, + "step": 9229 + }, + { + "epoch": 0.8252861230329042, + "grad_norm": 0.44838347446776716, + "learning_rate": 1.5588317937461105e-05, + "loss": 1.8471, + "step": 9230 + }, + { + "epoch": 0.8253755364806867, + "grad_norm": 0.4967048287918016, + "learning_rate": 1.55727942455667e-05, + "loss": 1.5427, + "step": 9231 + }, + { + "epoch": 0.8254649499284692, + "grad_norm": 0.5635532186663348, + "learning_rate": 1.5557277634480083e-05, + "loss": 1.4978, + "step": 9232 + }, + { + "epoch": 0.8255543633762518, + "grad_norm": 0.47461175285797835, + "learning_rate": 1.554176810550263e-05, + "loss": 1.6051, + "step": 9233 + }, + { + "epoch": 0.8256437768240343, + "grad_norm": 0.4861856718200829, + "learning_rate": 1.552626565993507e-05, + "loss": 1.6519, + "step": 9234 + }, + { + "epoch": 0.8257331902718169, + "grad_norm": 0.4083527420801717, + "learning_rate": 1.551077029907758e-05, + "loss": 1.3285, + "step": 9235 + }, + { + "epoch": 0.8258226037195995, + "grad_norm": 0.715154437768005, + "learning_rate": 1.5495282024229775e-05, + "loss": 1.9372, + "step": 9236 + }, + { + "epoch": 0.8259120171673819, + "grad_norm": 0.7646421021736579, + "learning_rate": 1.547980083669056e-05, + "loss": 1.6302, + "step": 9237 + }, + { + "epoch": 0.8260014306151645, + "grad_norm": 0.9169819844726512, + "learning_rate": 1.5464326737758428e-05, + "loss": 1.555, + "step": 9238 + }, + { + "epoch": 0.8260908440629471, + "grad_norm": 0.889749739110311, + "learning_rate": 1.544885972873109e-05, + "loss": 1.4491, + "step": 9239 + }, + { + "epoch": 0.8261802575107297, + "grad_norm": 0.3884172157991068, + "learning_rate": 1.543339981090578e-05, + "loss": 1.2079, + "step": 9240 + }, + { + "epoch": 0.8262696709585121, + "grad_norm": 0.6154345058714971, + "learning_rate": 1.541794698557909e-05, + "loss": 1.5501, + "step": 9241 + }, + { + "epoch": 0.8263590844062947, + "grad_norm": 0.6406121539422955, + "learning_rate": 1.5402501254047065e-05, + "loss": 1.7379, + "step": 9242 + }, + { + "epoch": 0.8264484978540773, + "grad_norm": 0.6796885878273781, + "learning_rate": 1.5387062617605064e-05, + "loss": 1.4687, + "step": 9243 + }, + { + "epoch": 0.8265379113018598, + "grad_norm": 0.5843544153991541, + "learning_rate": 1.5371631077547942e-05, + "loss": 1.4876, + "step": 9244 + }, + { + "epoch": 0.8266273247496424, + "grad_norm": 0.8648658740246746, + "learning_rate": 1.5356206635169912e-05, + "loss": 1.794, + "step": 9245 + }, + { + "epoch": 0.8267167381974249, + "grad_norm": 0.544840717947441, + "learning_rate": 1.5340789291764612e-05, + "loss": 1.5298, + "step": 9246 + }, + { + "epoch": 0.8268061516452074, + "grad_norm": 0.6464563528831899, + "learning_rate": 1.532537904862509e-05, + "loss": 1.7942, + "step": 9247 + }, + { + "epoch": 0.82689556509299, + "grad_norm": 0.6771436859577263, + "learning_rate": 1.530997590704375e-05, + "loss": 1.297, + "step": 9248 + }, + { + "epoch": 0.8269849785407726, + "grad_norm": 0.5672430191361489, + "learning_rate": 1.529457986831244e-05, + "loss": 1.7074, + "step": 9249 + }, + { + "epoch": 0.827074391988555, + "grad_norm": 0.9175484305105193, + "learning_rate": 1.5279190933722443e-05, + "loss": 1.8932, + "step": 9250 + }, + { + "epoch": 0.8271638054363376, + "grad_norm": 0.5459323161832623, + "learning_rate": 1.5263809104564353e-05, + "loss": 1.9137, + "step": 9251 + }, + { + "epoch": 0.8272532188841202, + "grad_norm": 0.48108561575637176, + "learning_rate": 1.5248434382128263e-05, + "loss": 1.5351, + "step": 9252 + }, + { + "epoch": 0.8273426323319027, + "grad_norm": 0.9365559462533519, + "learning_rate": 1.5233066767703663e-05, + "loss": 1.7292, + "step": 9253 + }, + { + "epoch": 0.8274320457796852, + "grad_norm": 0.9688865906014377, + "learning_rate": 1.5217706262579356e-05, + "loss": 1.5326, + "step": 9254 + }, + { + "epoch": 0.8275214592274678, + "grad_norm": 0.531846118350985, + "learning_rate": 1.5202352868043624e-05, + "loss": 1.5002, + "step": 9255 + }, + { + "epoch": 0.8276108726752504, + "grad_norm": 1.089320511859906, + "learning_rate": 1.5187006585384179e-05, + "loss": 1.6977, + "step": 9256 + }, + { + "epoch": 0.8277002861230329, + "grad_norm": 0.5082937520327719, + "learning_rate": 1.5171667415888046e-05, + "loss": 1.3078, + "step": 9257 + }, + { + "epoch": 0.8277896995708155, + "grad_norm": 1.1596135252557789, + "learning_rate": 1.515633536084171e-05, + "loss": 1.5892, + "step": 9258 + }, + { + "epoch": 0.827879113018598, + "grad_norm": 0.68998680454605, + "learning_rate": 1.5141010421531066e-05, + "loss": 1.5855, + "step": 9259 + }, + { + "epoch": 0.8279685264663805, + "grad_norm": 0.7174926147078723, + "learning_rate": 1.5125692599241391e-05, + "loss": 1.7334, + "step": 9260 + }, + { + "epoch": 0.8280579399141631, + "grad_norm": 0.6922174552408957, + "learning_rate": 1.5110381895257408e-05, + "loss": 1.5105, + "step": 9261 + }, + { + "epoch": 0.8281473533619457, + "grad_norm": 0.42309404063366934, + "learning_rate": 1.5095078310863142e-05, + "loss": 1.439, + "step": 9262 + }, + { + "epoch": 0.8282367668097281, + "grad_norm": 0.43088376777722837, + "learning_rate": 1.5079781847342123e-05, + "loss": 1.6778, + "step": 9263 + }, + { + "epoch": 0.8283261802575107, + "grad_norm": 0.8292299405209201, + "learning_rate": 1.5064492505977234e-05, + "loss": 1.7637, + "step": 9264 + }, + { + "epoch": 0.8284155937052933, + "grad_norm": 0.5251390960331617, + "learning_rate": 1.5049210288050796e-05, + "loss": 1.4916, + "step": 9265 + }, + { + "epoch": 0.8285050071530758, + "grad_norm": 0.5802438492218556, + "learning_rate": 1.5033935194844484e-05, + "loss": 1.3665, + "step": 9266 + }, + { + "epoch": 0.8285944206008584, + "grad_norm": 0.4607379980816203, + "learning_rate": 1.501866722763945e-05, + "loss": 1.5928, + "step": 9267 + }, + { + "epoch": 0.8286838340486409, + "grad_norm": 0.4376109298566681, + "learning_rate": 1.5003406387716134e-05, + "loss": 1.643, + "step": 9268 + }, + { + "epoch": 0.8287732474964234, + "grad_norm": 0.5808301145197525, + "learning_rate": 1.4988152676354472e-05, + "loss": 1.3331, + "step": 9269 + }, + { + "epoch": 0.828862660944206, + "grad_norm": 0.9030991816518821, + "learning_rate": 1.4972906094833805e-05, + "loss": 1.6059, + "step": 9270 + }, + { + "epoch": 0.8289520743919886, + "grad_norm": 1.1423049987614218, + "learning_rate": 1.4957666644432788e-05, + "loss": 1.8822, + "step": 9271 + }, + { + "epoch": 0.829041487839771, + "grad_norm": 0.5874544161770905, + "learning_rate": 1.4942434326429544e-05, + "loss": 1.4732, + "step": 9272 + }, + { + "epoch": 0.8291309012875536, + "grad_norm": 0.7308651007259142, + "learning_rate": 1.4927209142101662e-05, + "loss": 1.7609, + "step": 9273 + }, + { + "epoch": 0.8292203147353362, + "grad_norm": 0.6597772121001125, + "learning_rate": 1.4911991092725985e-05, + "loss": 1.6304, + "step": 9274 + }, + { + "epoch": 0.8293097281831188, + "grad_norm": 1.0955452197555615, + "learning_rate": 1.489678017957884e-05, + "loss": 1.493, + "step": 9275 + }, + { + "epoch": 0.8293991416309013, + "grad_norm": 0.7809800389867922, + "learning_rate": 1.4881576403936004e-05, + "loss": 1.4631, + "step": 9276 + }, + { + "epoch": 0.8294885550786838, + "grad_norm": 0.5875133285565531, + "learning_rate": 1.4866379767072525e-05, + "loss": 1.6974, + "step": 9277 + }, + { + "epoch": 0.8295779685264664, + "grad_norm": 0.5186461868222111, + "learning_rate": 1.485119027026296e-05, + "loss": 1.5996, + "step": 9278 + }, + { + "epoch": 0.8296673819742489, + "grad_norm": 1.0416325586270525, + "learning_rate": 1.4836007914781225e-05, + "loss": 1.6397, + "step": 9279 + }, + { + "epoch": 0.8297567954220315, + "grad_norm": 0.7827814618261079, + "learning_rate": 1.4820832701900667e-05, + "loss": 1.7971, + "step": 9280 + }, + { + "epoch": 0.829846208869814, + "grad_norm": 0.6045978062580437, + "learning_rate": 1.4805664632894024e-05, + "loss": 1.5776, + "step": 9281 + }, + { + "epoch": 0.8299356223175965, + "grad_norm": 0.3162232677741206, + "learning_rate": 1.4790503709033365e-05, + "loss": 1.3527, + "step": 9282 + }, + { + "epoch": 0.8300250357653791, + "grad_norm": 0.5447860269504505, + "learning_rate": 1.4775349931590266e-05, + "loss": 1.6379, + "step": 9283 + }, + { + "epoch": 0.8301144492131617, + "grad_norm": 0.5215392945684665, + "learning_rate": 1.4760203301835652e-05, + "loss": 1.5493, + "step": 9284 + }, + { + "epoch": 0.8302038626609443, + "grad_norm": 0.8506553574726453, + "learning_rate": 1.4745063821039806e-05, + "loss": 1.58, + "step": 9285 + }, + { + "epoch": 0.8302932761087267, + "grad_norm": 0.7539171568320208, + "learning_rate": 1.4729931490472515e-05, + "loss": 1.7839, + "step": 9286 + }, + { + "epoch": 0.8303826895565093, + "grad_norm": 0.5766608026249131, + "learning_rate": 1.4714806311402918e-05, + "loss": 1.762, + "step": 9287 + }, + { + "epoch": 0.8304721030042919, + "grad_norm": 0.49360613112628415, + "learning_rate": 1.4699688285099489e-05, + "loss": 1.4221, + "step": 9288 + }, + { + "epoch": 0.8305615164520744, + "grad_norm": 0.5428190662542691, + "learning_rate": 1.4684577412830191e-05, + "loss": 1.9271, + "step": 9289 + }, + { + "epoch": 0.8306509298998569, + "grad_norm": 0.47872843178394237, + "learning_rate": 1.4669473695862368e-05, + "loss": 1.8459, + "step": 9290 + }, + { + "epoch": 0.8307403433476395, + "grad_norm": 0.442446573160868, + "learning_rate": 1.4654377135462715e-05, + "loss": 1.2905, + "step": 9291 + }, + { + "epoch": 0.830829756795422, + "grad_norm": 0.5157673541057248, + "learning_rate": 1.4639287732897377e-05, + "loss": 1.4528, + "step": 9292 + }, + { + "epoch": 0.8309191702432046, + "grad_norm": 0.612809502379158, + "learning_rate": 1.4624205489431886e-05, + "loss": 1.58, + "step": 9293 + }, + { + "epoch": 0.8310085836909872, + "grad_norm": 0.5488178512965093, + "learning_rate": 1.4609130406331172e-05, + "loss": 1.5763, + "step": 9294 + }, + { + "epoch": 0.8310979971387696, + "grad_norm": 0.5112066122836684, + "learning_rate": 1.4594062484859595e-05, + "loss": 1.3494, + "step": 9295 + }, + { + "epoch": 0.8311874105865522, + "grad_norm": 0.7986257450244737, + "learning_rate": 1.4579001726280828e-05, + "loss": 1.9246, + "step": 9296 + }, + { + "epoch": 0.8312768240343348, + "grad_norm": 0.4561939570366489, + "learning_rate": 1.4563948131858018e-05, + "loss": 1.5363, + "step": 9297 + }, + { + "epoch": 0.8313662374821174, + "grad_norm": 1.1185988466971846, + "learning_rate": 1.4548901702853701e-05, + "loss": 1.5359, + "step": 9298 + }, + { + "epoch": 0.8314556509298998, + "grad_norm": 0.5272713436216321, + "learning_rate": 1.4533862440529799e-05, + "loss": 1.5572, + "step": 9299 + }, + { + "epoch": 0.8315450643776824, + "grad_norm": 0.5934432903685676, + "learning_rate": 1.4518830346147638e-05, + "loss": 1.7697, + "step": 9300 + }, + { + "epoch": 0.831634477825465, + "grad_norm": 0.43526948823223127, + "learning_rate": 1.4503805420967964e-05, + "loss": 1.4616, + "step": 9301 + }, + { + "epoch": 0.8317238912732475, + "grad_norm": 0.7123337023432665, + "learning_rate": 1.4488787666250858e-05, + "loss": 1.5181, + "step": 9302 + }, + { + "epoch": 0.83181330472103, + "grad_norm": 0.47880292495059573, + "learning_rate": 1.4473777083255857e-05, + "loss": 1.3763, + "step": 9303 + }, + { + "epoch": 0.8319027181688126, + "grad_norm": 0.9825050617387384, + "learning_rate": 1.4458773673241899e-05, + "loss": 1.8681, + "step": 9304 + }, + { + "epoch": 0.8319921316165951, + "grad_norm": 0.4321602746284928, + "learning_rate": 1.4443777437467265e-05, + "loss": 1.2533, + "step": 9305 + }, + { + "epoch": 0.8320815450643777, + "grad_norm": 0.6153658149632535, + "learning_rate": 1.4428788377189672e-05, + "loss": 1.6253, + "step": 9306 + }, + { + "epoch": 0.8321709585121603, + "grad_norm": 0.48673238144305536, + "learning_rate": 1.4413806493666293e-05, + "loss": 1.6302, + "step": 9307 + }, + { + "epoch": 0.8322603719599427, + "grad_norm": 0.5276955723744292, + "learning_rate": 1.4398831788153588e-05, + "loss": 1.3682, + "step": 9308 + }, + { + "epoch": 0.8323497854077253, + "grad_norm": 0.9714988427651261, + "learning_rate": 1.4383864261907476e-05, + "loss": 1.7481, + "step": 9309 + }, + { + "epoch": 0.8324391988555079, + "grad_norm": 1.0522662019963396, + "learning_rate": 1.4368903916183296e-05, + "loss": 1.8502, + "step": 9310 + }, + { + "epoch": 0.8325286123032904, + "grad_norm": 0.7429701433711382, + "learning_rate": 1.4353950752235702e-05, + "loss": 1.3877, + "step": 9311 + }, + { + "epoch": 0.8326180257510729, + "grad_norm": 0.9781809227496259, + "learning_rate": 1.433900477131882e-05, + "loss": 1.8736, + "step": 9312 + }, + { + "epoch": 0.8327074391988555, + "grad_norm": 1.097957033146018, + "learning_rate": 1.4324065974686162e-05, + "loss": 1.8202, + "step": 9313 + }, + { + "epoch": 0.832796852646638, + "grad_norm": 0.49041187196933417, + "learning_rate": 1.4309134363590615e-05, + "loss": 1.4777, + "step": 9314 + }, + { + "epoch": 0.8328862660944206, + "grad_norm": 1.180976455819507, + "learning_rate": 1.4294209939284509e-05, + "loss": 1.8682, + "step": 9315 + }, + { + "epoch": 0.8329756795422032, + "grad_norm": 0.41337788348862525, + "learning_rate": 1.4279292703019486e-05, + "loss": 1.3331, + "step": 9316 + }, + { + "epoch": 0.8330650929899857, + "grad_norm": 1.0037174270915772, + "learning_rate": 1.426438265604666e-05, + "loss": 1.636, + "step": 9317 + }, + { + "epoch": 0.8331545064377682, + "grad_norm": 0.6660366068276529, + "learning_rate": 1.4249479799616538e-05, + "loss": 1.9755, + "step": 9318 + }, + { + "epoch": 0.8332439198855508, + "grad_norm": 0.5471653868859253, + "learning_rate": 1.4234584134978956e-05, + "loss": 1.6871, + "step": 9319 + }, + { + "epoch": 0.8333333333333334, + "grad_norm": 1.229698915455957, + "learning_rate": 1.4219695663383214e-05, + "loss": 6.2304, + "step": 9320 + }, + { + "epoch": 0.8334227467811158, + "grad_norm": 1.3590151658668326, + "learning_rate": 1.4204814386078036e-05, + "loss": 1.9342, + "step": 9321 + }, + { + "epoch": 0.8335121602288984, + "grad_norm": 0.7408254123667977, + "learning_rate": 1.4189940304311444e-05, + "loss": 1.3104, + "step": 9322 + }, + { + "epoch": 0.833601573676681, + "grad_norm": 0.7638867492365032, + "learning_rate": 1.417507341933092e-05, + "loss": 1.3716, + "step": 9323 + }, + { + "epoch": 0.8336909871244635, + "grad_norm": 0.5648561703801099, + "learning_rate": 1.4160213732383364e-05, + "loss": 1.448, + "step": 9324 + }, + { + "epoch": 0.8337804005722461, + "grad_norm": 0.6444135120741602, + "learning_rate": 1.4145361244714995e-05, + "loss": 1.3015, + "step": 9325 + }, + { + "epoch": 0.8338698140200286, + "grad_norm": 0.7230554532785515, + "learning_rate": 1.4130515957571488e-05, + "loss": 1.281, + "step": 9326 + }, + { + "epoch": 0.8339592274678111, + "grad_norm": 0.5978696024351565, + "learning_rate": 1.4115677872197908e-05, + "loss": 1.5014, + "step": 9327 + }, + { + "epoch": 0.8340486409155937, + "grad_norm": 0.847565862146144, + "learning_rate": 1.41008469898387e-05, + "loss": 1.4821, + "step": 9328 + }, + { + "epoch": 0.8341380543633763, + "grad_norm": 0.5930235167142371, + "learning_rate": 1.4086023311737716e-05, + "loss": 1.7822, + "step": 9329 + }, + { + "epoch": 0.8342274678111588, + "grad_norm": 0.6500321016355616, + "learning_rate": 1.4071206839138217e-05, + "loss": 1.4705, + "step": 9330 + }, + { + "epoch": 0.8343168812589413, + "grad_norm": 0.40854275705605075, + "learning_rate": 1.4056397573282808e-05, + "loss": 1.3663, + "step": 9331 + }, + { + "epoch": 0.8344062947067239, + "grad_norm": 0.669392576414901, + "learning_rate": 1.4041595515413542e-05, + "loss": 1.7345, + "step": 9332 + }, + { + "epoch": 0.8344957081545065, + "grad_norm": 0.5409494377739568, + "learning_rate": 1.4026800666771867e-05, + "loss": 1.2319, + "step": 9333 + }, + { + "epoch": 0.834585121602289, + "grad_norm": 2.5241343400248284, + "learning_rate": 1.4012013028598547e-05, + "loss": 6.2592, + "step": 9334 + }, + { + "epoch": 0.8346745350500715, + "grad_norm": 0.5960985917265262, + "learning_rate": 1.3997232602133892e-05, + "loss": 1.8224, + "step": 9335 + }, + { + "epoch": 0.8347639484978541, + "grad_norm": 0.6715358776947224, + "learning_rate": 1.3982459388617452e-05, + "loss": 1.4187, + "step": 9336 + }, + { + "epoch": 0.8348533619456366, + "grad_norm": 1.0037516901994035, + "learning_rate": 1.3967693389288261e-05, + "loss": 1.8288, + "step": 9337 + }, + { + "epoch": 0.8349427753934192, + "grad_norm": 0.6395460184958778, + "learning_rate": 1.3952934605384749e-05, + "loss": 1.369, + "step": 9338 + }, + { + "epoch": 0.8350321888412017, + "grad_norm": 0.6736140569897062, + "learning_rate": 1.393818303814467e-05, + "loss": 1.9466, + "step": 9339 + }, + { + "epoch": 0.8351216022889842, + "grad_norm": 0.5058581712546869, + "learning_rate": 1.3923438688805235e-05, + "loss": 1.5967, + "step": 9340 + }, + { + "epoch": 0.8352110157367668, + "grad_norm": 0.5531734808983999, + "learning_rate": 1.3908701558603054e-05, + "loss": 1.4205, + "step": 9341 + }, + { + "epoch": 0.8353004291845494, + "grad_norm": 1.0830978669666962, + "learning_rate": 1.3893971648774095e-05, + "loss": 1.89, + "step": 9342 + }, + { + "epoch": 0.835389842632332, + "grad_norm": 0.5818970862173779, + "learning_rate": 1.3879248960553737e-05, + "loss": 1.5032, + "step": 9343 + }, + { + "epoch": 0.8354792560801144, + "grad_norm": 1.004800912261973, + "learning_rate": 1.386453349517679e-05, + "loss": 1.7948, + "step": 9344 + }, + { + "epoch": 0.835568669527897, + "grad_norm": 0.9738395706994971, + "learning_rate": 1.384982525387738e-05, + "loss": 2.2813, + "step": 9345 + }, + { + "epoch": 0.8356580829756796, + "grad_norm": 0.5243623642827892, + "learning_rate": 1.3835124237889074e-05, + "loss": 1.4865, + "step": 9346 + }, + { + "epoch": 0.8357474964234621, + "grad_norm": 5.122690641117008, + "learning_rate": 1.3820430448444866e-05, + "loss": 6.1197, + "step": 9347 + }, + { + "epoch": 0.8358369098712446, + "grad_norm": 0.6515108457265675, + "learning_rate": 1.3805743886777022e-05, + "loss": 1.8673, + "step": 9348 + }, + { + "epoch": 0.8359263233190272, + "grad_norm": 0.6316740077008212, + "learning_rate": 1.3791064554117394e-05, + "loss": 1.557, + "step": 9349 + }, + { + "epoch": 0.8360157367668097, + "grad_norm": 1.1224006854649282, + "learning_rate": 1.377639245169704e-05, + "loss": 1.5864, + "step": 9350 + }, + { + "epoch": 0.8361051502145923, + "grad_norm": 1.0152838223933611, + "learning_rate": 1.376172758074653e-05, + "loss": 1.5224, + "step": 9351 + }, + { + "epoch": 0.8361945636623748, + "grad_norm": 0.692362524479008, + "learning_rate": 1.3747069942495794e-05, + "loss": 1.7789, + "step": 9352 + }, + { + "epoch": 0.8362839771101573, + "grad_norm": 0.4422941398583763, + "learning_rate": 1.3732419538174112e-05, + "loss": 1.6528, + "step": 9353 + }, + { + "epoch": 0.8363733905579399, + "grad_norm": 0.6752820842524214, + "learning_rate": 1.3717776369010216e-05, + "loss": 1.4093, + "step": 9354 + }, + { + "epoch": 0.8364628040057225, + "grad_norm": 1.04423049604426, + "learning_rate": 1.370314043623222e-05, + "loss": 1.371, + "step": 9355 + }, + { + "epoch": 0.836552217453505, + "grad_norm": 0.8904537720723777, + "learning_rate": 1.368851174106761e-05, + "loss": 1.5821, + "step": 9356 + }, + { + "epoch": 0.8366416309012875, + "grad_norm": 0.5935342793442057, + "learning_rate": 1.3673890284743285e-05, + "loss": 1.4425, + "step": 9357 + }, + { + "epoch": 0.8367310443490701, + "grad_norm": 0.5384389604661115, + "learning_rate": 1.3659276068485549e-05, + "loss": 1.8438, + "step": 9358 + }, + { + "epoch": 0.8368204577968527, + "grad_norm": 0.5724997191872834, + "learning_rate": 1.3644669093520035e-05, + "loss": 1.5752, + "step": 9359 + }, + { + "epoch": 0.8369098712446352, + "grad_norm": 0.6368626992506828, + "learning_rate": 1.363006936107183e-05, + "loss": 1.7495, + "step": 9360 + }, + { + "epoch": 0.8369992846924177, + "grad_norm": 0.5731756350697264, + "learning_rate": 1.3615476872365419e-05, + "loss": 1.4773, + "step": 9361 + }, + { + "epoch": 0.8370886981402003, + "grad_norm": 5.835948049616781, + "learning_rate": 1.3600891628624601e-05, + "loss": 6.1467, + "step": 9362 + }, + { + "epoch": 0.8371781115879828, + "grad_norm": 1.1793649657805279, + "learning_rate": 1.3586313631072668e-05, + "loss": 1.694, + "step": 9363 + }, + { + "epoch": 0.8372675250357654, + "grad_norm": 2.341499288130569, + "learning_rate": 1.357174288093228e-05, + "loss": 1.6046, + "step": 9364 + }, + { + "epoch": 0.837356938483548, + "grad_norm": 2.58590504202287, + "learning_rate": 1.355717937942541e-05, + "loss": 1.6025, + "step": 9365 + }, + { + "epoch": 0.8374463519313304, + "grad_norm": 1.914477161405429, + "learning_rate": 1.3542623127773523e-05, + "loss": 1.8834, + "step": 9366 + }, + { + "epoch": 0.837535765379113, + "grad_norm": 0.9015620107455018, + "learning_rate": 1.3528074127197432e-05, + "loss": 1.5273, + "step": 9367 + }, + { + "epoch": 0.8376251788268956, + "grad_norm": 0.6835645706135125, + "learning_rate": 1.3513532378917281e-05, + "loss": 1.4156, + "step": 9368 + }, + { + "epoch": 0.8377145922746781, + "grad_norm": 1.7838231180091275, + "learning_rate": 1.3498997884152776e-05, + "loss": 1.7705, + "step": 9369 + }, + { + "epoch": 0.8378040057224606, + "grad_norm": 1.9581133723091069, + "learning_rate": 1.3484470644122826e-05, + "loss": 1.654, + "step": 9370 + }, + { + "epoch": 0.8378934191702432, + "grad_norm": 1.0489516489395765, + "learning_rate": 1.3469950660045838e-05, + "loss": 1.6944, + "step": 9371 + }, + { + "epoch": 0.8379828326180258, + "grad_norm": 0.7044581627328353, + "learning_rate": 1.3455437933139614e-05, + "loss": 1.5854, + "step": 9372 + }, + { + "epoch": 0.8380722460658083, + "grad_norm": 0.7240390256106516, + "learning_rate": 1.344093246462126e-05, + "loss": 1.6845, + "step": 9373 + }, + { + "epoch": 0.8381616595135909, + "grad_norm": 0.9019441365208902, + "learning_rate": 1.3426434255707365e-05, + "loss": 1.6992, + "step": 9374 + }, + { + "epoch": 0.8382510729613734, + "grad_norm": 0.6548215355899277, + "learning_rate": 1.3411943307613883e-05, + "loss": 1.4722, + "step": 9375 + }, + { + "epoch": 0.8383404864091559, + "grad_norm": 1.016354091465214, + "learning_rate": 1.339745962155613e-05, + "loss": 1.5564, + "step": 9376 + }, + { + "epoch": 0.8384298998569385, + "grad_norm": 1.0418956235815464, + "learning_rate": 1.3382983198748855e-05, + "loss": 1.5627, + "step": 9377 + }, + { + "epoch": 0.8385193133047211, + "grad_norm": 1.0002370155344076, + "learning_rate": 1.33685140404062e-05, + "loss": 1.4671, + "step": 9378 + }, + { + "epoch": 0.8386087267525035, + "grad_norm": 0.595728560651765, + "learning_rate": 1.3354052147741625e-05, + "loss": 1.8787, + "step": 9379 + }, + { + "epoch": 0.8386981402002861, + "grad_norm": 0.5332315300433131, + "learning_rate": 1.333959752196805e-05, + "loss": 1.5225, + "step": 9380 + }, + { + "epoch": 0.8387875536480687, + "grad_norm": 0.6747304023757227, + "learning_rate": 1.3325150164297796e-05, + "loss": 1.5803, + "step": 9381 + }, + { + "epoch": 0.8388769670958512, + "grad_norm": 0.7171186012427989, + "learning_rate": 1.3310710075942479e-05, + "loss": 1.5025, + "step": 9382 + }, + { + "epoch": 0.8389663805436338, + "grad_norm": 1.4186082759134195, + "learning_rate": 1.3296277258113254e-05, + "loss": 1.7893, + "step": 9383 + }, + { + "epoch": 0.8390557939914163, + "grad_norm": 0.6973265055008929, + "learning_rate": 1.328185171202052e-05, + "loss": 1.9029, + "step": 9384 + }, + { + "epoch": 0.8391452074391988, + "grad_norm": 0.49265739782284773, + "learning_rate": 1.3267433438874155e-05, + "loss": 1.6325, + "step": 9385 + }, + { + "epoch": 0.8392346208869814, + "grad_norm": 0.7521500475235219, + "learning_rate": 1.3253022439883412e-05, + "loss": 1.9888, + "step": 9386 + }, + { + "epoch": 0.839324034334764, + "grad_norm": 0.6263669799244831, + "learning_rate": 1.3238618716256923e-05, + "loss": 1.5575, + "step": 9387 + }, + { + "epoch": 0.8394134477825465, + "grad_norm": 0.7206656937549523, + "learning_rate": 1.322422226920268e-05, + "loss": 1.8721, + "step": 9388 + }, + { + "epoch": 0.839502861230329, + "grad_norm": 1.223968920277615, + "learning_rate": 1.3209833099928114e-05, + "loss": 1.854, + "step": 9389 + }, + { + "epoch": 0.8395922746781116, + "grad_norm": 0.42221349099892447, + "learning_rate": 1.3195451209640041e-05, + "loss": 1.59, + "step": 9390 + }, + { + "epoch": 0.8396816881258942, + "grad_norm": 0.37191455081816804, + "learning_rate": 1.3181076599544629e-05, + "loss": 1.4861, + "step": 9391 + }, + { + "epoch": 0.8397711015736766, + "grad_norm": 0.5687298126307365, + "learning_rate": 1.3166709270847511e-05, + "loss": 1.7411, + "step": 9392 + }, + { + "epoch": 0.8398605150214592, + "grad_norm": 0.841142678810602, + "learning_rate": 1.3152349224753579e-05, + "loss": 1.8542, + "step": 9393 + }, + { + "epoch": 0.8399499284692418, + "grad_norm": 0.4959398047733695, + "learning_rate": 1.3137996462467239e-05, + "loss": 1.6609, + "step": 9394 + }, + { + "epoch": 0.8400393419170243, + "grad_norm": 0.47308009840157234, + "learning_rate": 1.312365098519226e-05, + "loss": 1.7329, + "step": 9395 + }, + { + "epoch": 0.8401287553648069, + "grad_norm": 0.613362813501066, + "learning_rate": 1.31093127941317e-05, + "loss": 1.6682, + "step": 9396 + }, + { + "epoch": 0.8402181688125894, + "grad_norm": 0.36536197434872486, + "learning_rate": 1.3094981890488167e-05, + "loss": 1.2355, + "step": 9397 + }, + { + "epoch": 0.8403075822603719, + "grad_norm": 0.3966869168883564, + "learning_rate": 1.3080658275463565e-05, + "loss": 1.6256, + "step": 9398 + }, + { + "epoch": 0.8403969957081545, + "grad_norm": 0.539774077099906, + "learning_rate": 1.3066341950259165e-05, + "loss": 1.4749, + "step": 9399 + }, + { + "epoch": 0.8404864091559371, + "grad_norm": 0.4849379203296311, + "learning_rate": 1.3052032916075674e-05, + "loss": 1.6786, + "step": 9400 + }, + { + "epoch": 0.8405758226037195, + "grad_norm": 0.48593189262225683, + "learning_rate": 1.3037731174113188e-05, + "loss": 1.9048, + "step": 9401 + }, + { + "epoch": 0.8406652360515021, + "grad_norm": 0.6167426296534121, + "learning_rate": 1.3023436725571158e-05, + "loss": 1.4309, + "step": 9402 + }, + { + "epoch": 0.8407546494992847, + "grad_norm": 0.4000704417133254, + "learning_rate": 1.3009149571648438e-05, + "loss": 1.5352, + "step": 9403 + }, + { + "epoch": 0.8408440629470673, + "grad_norm": 0.5550616962059658, + "learning_rate": 1.2994869713543289e-05, + "loss": 1.8228, + "step": 9404 + }, + { + "epoch": 0.8409334763948498, + "grad_norm": 0.595542773412791, + "learning_rate": 1.2980597152453344e-05, + "loss": 1.2195, + "step": 9405 + }, + { + "epoch": 0.8410228898426323, + "grad_norm": 0.3608133764245701, + "learning_rate": 1.2966331889575644e-05, + "loss": 1.4358, + "step": 9406 + }, + { + "epoch": 0.8411123032904149, + "grad_norm": 0.525801524471028, + "learning_rate": 1.2952073926106556e-05, + "loss": 1.3997, + "step": 9407 + }, + { + "epoch": 0.8412017167381974, + "grad_norm": 0.5053643076783064, + "learning_rate": 1.29378232632419e-05, + "loss": 1.5379, + "step": 9408 + }, + { + "epoch": 0.84129113018598, + "grad_norm": 0.7480655514146755, + "learning_rate": 1.2923579902176886e-05, + "loss": 1.8768, + "step": 9409 + }, + { + "epoch": 0.8413805436337625, + "grad_norm": 0.6779840888631165, + "learning_rate": 1.2909343844106014e-05, + "loss": 1.6019, + "step": 9410 + }, + { + "epoch": 0.841469957081545, + "grad_norm": 0.6474297859773384, + "learning_rate": 1.289511509022332e-05, + "loss": 1.6857, + "step": 9411 + }, + { + "epoch": 0.8415593705293276, + "grad_norm": 0.5251941030410111, + "learning_rate": 1.2880893641722147e-05, + "loss": 1.8238, + "step": 9412 + }, + { + "epoch": 0.8416487839771102, + "grad_norm": 0.5419872634904861, + "learning_rate": 1.2866679499795198e-05, + "loss": 1.601, + "step": 9413 + }, + { + "epoch": 0.8417381974248928, + "grad_norm": 0.631264848859024, + "learning_rate": 1.2852472665634607e-05, + "loss": 1.515, + "step": 9414 + }, + { + "epoch": 0.8418276108726752, + "grad_norm": 0.568495752927162, + "learning_rate": 1.28382731404319e-05, + "loss": 1.787, + "step": 9415 + }, + { + "epoch": 0.8419170243204578, + "grad_norm": 0.3758959428942526, + "learning_rate": 1.2824080925377945e-05, + "loss": 1.542, + "step": 9416 + }, + { + "epoch": 0.8420064377682404, + "grad_norm": 0.558669519125111, + "learning_rate": 1.2809896021663037e-05, + "loss": 1.6602, + "step": 9417 + }, + { + "epoch": 0.8420958512160229, + "grad_norm": 0.4172957007867305, + "learning_rate": 1.2795718430476854e-05, + "loss": 1.4578, + "step": 9418 + }, + { + "epoch": 0.8421852646638054, + "grad_norm": 0.5074155549816675, + "learning_rate": 1.278154815300845e-05, + "loss": 1.6408, + "step": 9419 + }, + { + "epoch": 0.842274678111588, + "grad_norm": 0.5079269117555505, + "learning_rate": 1.2767385190446257e-05, + "loss": 1.5809, + "step": 9420 + }, + { + "epoch": 0.8423640915593705, + "grad_norm": 0.6425737753732138, + "learning_rate": 1.2753229543978151e-05, + "loss": 1.6703, + "step": 9421 + }, + { + "epoch": 0.8424535050071531, + "grad_norm": 0.8955868880973271, + "learning_rate": 1.2739081214791293e-05, + "loss": 1.8447, + "step": 9422 + }, + { + "epoch": 0.8425429184549357, + "grad_norm": 0.3828520186240439, + "learning_rate": 1.2724940204072311e-05, + "loss": 1.4283, + "step": 9423 + }, + { + "epoch": 0.8426323319027181, + "grad_norm": 0.7473010183254398, + "learning_rate": 1.271080651300719e-05, + "loss": 1.3649, + "step": 9424 + }, + { + "epoch": 0.8427217453505007, + "grad_norm": 0.840600795346558, + "learning_rate": 1.2696680142781313e-05, + "loss": 1.9554, + "step": 9425 + }, + { + "epoch": 0.8428111587982833, + "grad_norm": 0.5065990496432575, + "learning_rate": 1.2682561094579448e-05, + "loss": 1.7548, + "step": 9426 + }, + { + "epoch": 0.8429005722460658, + "grad_norm": 0.4771773477470263, + "learning_rate": 1.2668449369585723e-05, + "loss": 1.7134, + "step": 9427 + }, + { + "epoch": 0.8429899856938483, + "grad_norm": 1.0721302966182862, + "learning_rate": 1.2654344968983668e-05, + "loss": 1.8146, + "step": 9428 + }, + { + "epoch": 0.8430793991416309, + "grad_norm": 0.39711337903146504, + "learning_rate": 1.2640247893956236e-05, + "loss": 1.2983, + "step": 9429 + }, + { + "epoch": 0.8431688125894135, + "grad_norm": 0.393727397252051, + "learning_rate": 1.2626158145685696e-05, + "loss": 1.285, + "step": 9430 + }, + { + "epoch": 0.843258226037196, + "grad_norm": 0.41816916884690564, + "learning_rate": 1.2612075725353722e-05, + "loss": 1.4729, + "step": 9431 + }, + { + "epoch": 0.8433476394849786, + "grad_norm": 0.5374115048155593, + "learning_rate": 1.259800063414146e-05, + "loss": 1.815, + "step": 9432 + }, + { + "epoch": 0.843437052932761, + "grad_norm": 0.5791340672613614, + "learning_rate": 1.258393287322932e-05, + "loss": 1.6273, + "step": 9433 + }, + { + "epoch": 0.8435264663805436, + "grad_norm": 0.4154607840802728, + "learning_rate": 1.2569872443797148e-05, + "loss": 1.4473, + "step": 9434 + }, + { + "epoch": 0.8436158798283262, + "grad_norm": 0.43792146195880655, + "learning_rate": 1.2555819347024211e-05, + "loss": 1.2016, + "step": 9435 + }, + { + "epoch": 0.8437052932761088, + "grad_norm": 0.40563045476124265, + "learning_rate": 1.2541773584089079e-05, + "loss": 1.7277, + "step": 9436 + }, + { + "epoch": 0.8437947067238912, + "grad_norm": 0.48598658752871815, + "learning_rate": 1.2527735156169773e-05, + "loss": 1.5856, + "step": 9437 + }, + { + "epoch": 0.8438841201716738, + "grad_norm": 0.43850781165630476, + "learning_rate": 1.2513704064443677e-05, + "loss": 1.6059, + "step": 9438 + }, + { + "epoch": 0.8439735336194564, + "grad_norm": 1.0147522980100794, + "learning_rate": 1.249968031008757e-05, + "loss": 1.5343, + "step": 9439 + }, + { + "epoch": 0.844062947067239, + "grad_norm": 0.5964233038407692, + "learning_rate": 1.2485663894277611e-05, + "loss": 1.7095, + "step": 9440 + }, + { + "epoch": 0.8441523605150214, + "grad_norm": 0.4542279948633442, + "learning_rate": 1.2471654818189316e-05, + "loss": 1.4904, + "step": 9441 + }, + { + "epoch": 0.844241773962804, + "grad_norm": 0.6489502352052284, + "learning_rate": 1.2457653082997634e-05, + "loss": 1.612, + "step": 9442 + }, + { + "epoch": 0.8443311874105865, + "grad_norm": 0.5456883972712274, + "learning_rate": 1.2443658689876847e-05, + "loss": 1.6099, + "step": 9443 + }, + { + "epoch": 0.8444206008583691, + "grad_norm": 0.47157962557005517, + "learning_rate": 1.2429671640000695e-05, + "loss": 1.8667, + "step": 9444 + }, + { + "epoch": 0.8445100143061517, + "grad_norm": 0.6814840620066227, + "learning_rate": 1.2415691934542183e-05, + "loss": 1.9535, + "step": 9445 + }, + { + "epoch": 0.8445994277539342, + "grad_norm": 1.064331884490301, + "learning_rate": 1.2401719574673854e-05, + "loss": 1.5919, + "step": 9446 + }, + { + "epoch": 0.8446888412017167, + "grad_norm": 0.41565224510176746, + "learning_rate": 1.2387754561567488e-05, + "loss": 1.4291, + "step": 9447 + }, + { + "epoch": 0.8447782546494993, + "grad_norm": 0.5499799123971907, + "learning_rate": 1.237379689639434e-05, + "loss": 1.6255, + "step": 9448 + }, + { + "epoch": 0.8448676680972819, + "grad_norm": 0.6806007334714859, + "learning_rate": 1.2359846580325041e-05, + "loss": 1.338, + "step": 9449 + }, + { + "epoch": 0.8449570815450643, + "grad_norm": 0.4886265526718232, + "learning_rate": 1.2345903614529552e-05, + "loss": 1.5903, + "step": 9450 + }, + { + "epoch": 0.8450464949928469, + "grad_norm": 0.5438075880894127, + "learning_rate": 1.233196800017724e-05, + "loss": 1.3769, + "step": 9451 + }, + { + "epoch": 0.8451359084406295, + "grad_norm": 0.6918636445922597, + "learning_rate": 1.2318039738436936e-05, + "loss": 1.8582, + "step": 9452 + }, + { + "epoch": 0.845225321888412, + "grad_norm": 0.5289827220730459, + "learning_rate": 1.230411883047673e-05, + "loss": 1.6711, + "step": 9453 + }, + { + "epoch": 0.8453147353361946, + "grad_norm": 0.5372312621945624, + "learning_rate": 1.2290205277464161e-05, + "loss": 1.6583, + "step": 9454 + }, + { + "epoch": 0.8454041487839771, + "grad_norm": 0.5497787542795848, + "learning_rate": 1.2276299080566178e-05, + "loss": 1.504, + "step": 9455 + }, + { + "epoch": 0.8454935622317596, + "grad_norm": 0.5433078488019147, + "learning_rate": 1.2262400240949023e-05, + "loss": 1.6385, + "step": 9456 + }, + { + "epoch": 0.8455829756795422, + "grad_norm": 0.3961952426408401, + "learning_rate": 1.22485087597784e-05, + "loss": 1.4318, + "step": 9457 + }, + { + "epoch": 0.8456723891273248, + "grad_norm": 4.951283805992648, + "learning_rate": 1.2234624638219372e-05, + "loss": 6.2042, + "step": 9458 + }, + { + "epoch": 0.8457618025751072, + "grad_norm": 1.0305036152135922, + "learning_rate": 1.2220747877436378e-05, + "loss": 2.0626, + "step": 9459 + }, + { + "epoch": 0.8458512160228898, + "grad_norm": 1.5475512991523654, + "learning_rate": 1.2206878478593276e-05, + "loss": 1.8046, + "step": 9460 + }, + { + "epoch": 0.8459406294706724, + "grad_norm": 0.9877545869393026, + "learning_rate": 1.2193016442853221e-05, + "loss": 1.4878, + "step": 9461 + }, + { + "epoch": 0.846030042918455, + "grad_norm": 0.7152752656534495, + "learning_rate": 1.2179161771378845e-05, + "loss": 1.529, + "step": 9462 + }, + { + "epoch": 0.8461194563662375, + "grad_norm": 0.4824230468271646, + "learning_rate": 1.2165314465332122e-05, + "loss": 1.6907, + "step": 9463 + }, + { + "epoch": 0.84620886981402, + "grad_norm": 0.6355081609177412, + "learning_rate": 1.2151474525874374e-05, + "loss": 1.8379, + "step": 9464 + }, + { + "epoch": 0.8462982832618026, + "grad_norm": 0.828202800818075, + "learning_rate": 1.2137641954166346e-05, + "loss": 1.9156, + "step": 9465 + }, + { + "epoch": 0.8463876967095851, + "grad_norm": 0.9078277308111952, + "learning_rate": 1.212381675136821e-05, + "loss": 1.5014, + "step": 9466 + }, + { + "epoch": 0.8464771101573677, + "grad_norm": 1.0445901191931968, + "learning_rate": 1.2109998918639431e-05, + "loss": 1.7908, + "step": 9467 + }, + { + "epoch": 0.8465665236051502, + "grad_norm": 0.5670207327257557, + "learning_rate": 1.209618845713889e-05, + "loss": 1.4273, + "step": 9468 + }, + { + "epoch": 0.8466559370529327, + "grad_norm": 0.546226941210924, + "learning_rate": 1.2082385368024884e-05, + "loss": 1.887, + "step": 9469 + }, + { + "epoch": 0.8467453505007153, + "grad_norm": 1.044901749823619, + "learning_rate": 1.2068589652455008e-05, + "loss": 1.6151, + "step": 9470 + }, + { + "epoch": 0.8468347639484979, + "grad_norm": 0.985099099331604, + "learning_rate": 1.205480131158634e-05, + "loss": 1.6155, + "step": 9471 + }, + { + "epoch": 0.8469241773962805, + "grad_norm": 0.650406631570501, + "learning_rate": 1.2041020346575272e-05, + "loss": 1.2889, + "step": 9472 + }, + { + "epoch": 0.8470135908440629, + "grad_norm": 1.086142221578396, + "learning_rate": 1.2027246758577593e-05, + "loss": 1.6582, + "step": 9473 + }, + { + "epoch": 0.8471030042918455, + "grad_norm": 1.1323122294390524, + "learning_rate": 1.2013480548748512e-05, + "loss": 1.9996, + "step": 9474 + }, + { + "epoch": 0.8471924177396281, + "grad_norm": 0.40954806047485515, + "learning_rate": 1.199972171824253e-05, + "loss": 1.5775, + "step": 9475 + }, + { + "epoch": 0.8472818311874106, + "grad_norm": 0.6429771972429497, + "learning_rate": 1.198597026821361e-05, + "loss": 1.3756, + "step": 9476 + }, + { + "epoch": 0.8473712446351931, + "grad_norm": 0.8928610227913917, + "learning_rate": 1.1972226199815074e-05, + "loss": 1.617, + "step": 9477 + }, + { + "epoch": 0.8474606580829757, + "grad_norm": 1.7176045319211148, + "learning_rate": 1.1958489514199634e-05, + "loss": 1.7819, + "step": 9478 + }, + { + "epoch": 0.8475500715307582, + "grad_norm": 1.1126702098821657, + "learning_rate": 1.1944760212519313e-05, + "loss": 1.7156, + "step": 9479 + }, + { + "epoch": 0.8476394849785408, + "grad_norm": 0.4478319712799585, + "learning_rate": 1.1931038295925645e-05, + "loss": 1.5869, + "step": 9480 + }, + { + "epoch": 0.8477288984263234, + "grad_norm": 0.4382018854473035, + "learning_rate": 1.1917323765569411e-05, + "loss": 1.2862, + "step": 9481 + }, + { + "epoch": 0.8478183118741058, + "grad_norm": 1.0411495469792884, + "learning_rate": 1.1903616622600866e-05, + "loss": 1.9579, + "step": 9482 + }, + { + "epoch": 0.8479077253218884, + "grad_norm": 4.8640253224509635, + "learning_rate": 1.1889916868169614e-05, + "loss": 5.9867, + "step": 9483 + }, + { + "epoch": 0.847997138769671, + "grad_norm": 0.4124037430873191, + "learning_rate": 1.1876224503424615e-05, + "loss": 1.544, + "step": 9484 + }, + { + "epoch": 0.8480865522174535, + "grad_norm": 0.7174690281165947, + "learning_rate": 1.1862539529514228e-05, + "loss": 1.4344, + "step": 9485 + }, + { + "epoch": 0.848175965665236, + "grad_norm": 1.2186913423804802, + "learning_rate": 1.184886194758621e-05, + "loss": 1.4918, + "step": 9486 + }, + { + "epoch": 0.8482653791130186, + "grad_norm": 1.0996688400119607, + "learning_rate": 1.183519175878769e-05, + "loss": 1.4129, + "step": 9487 + }, + { + "epoch": 0.8483547925608012, + "grad_norm": 1.3566789475437488, + "learning_rate": 1.182152896426515e-05, + "loss": 2.1154, + "step": 9488 + }, + { + "epoch": 0.8484442060085837, + "grad_norm": 1.0984583303797604, + "learning_rate": 1.1807873565164506e-05, + "loss": 1.7243, + "step": 9489 + }, + { + "epoch": 0.8485336194563662, + "grad_norm": 0.6149772600686447, + "learning_rate": 1.1794225562630978e-05, + "loss": 1.7758, + "step": 9490 + }, + { + "epoch": 0.8486230329041488, + "grad_norm": 0.5991803707148634, + "learning_rate": 1.1780584957809227e-05, + "loss": 1.4448, + "step": 9491 + }, + { + "epoch": 0.8487124463519313, + "grad_norm": 1.3441281724655938, + "learning_rate": 1.1766951751843292e-05, + "loss": 1.6624, + "step": 9492 + }, + { + "epoch": 0.8488018597997139, + "grad_norm": 0.7462918861356155, + "learning_rate": 1.1753325945876515e-05, + "loss": 1.4796, + "step": 9493 + }, + { + "epoch": 0.8488912732474965, + "grad_norm": 0.8137956761180493, + "learning_rate": 1.173970754105176e-05, + "loss": 1.5207, + "step": 9494 + }, + { + "epoch": 0.8489806866952789, + "grad_norm": 0.47999750488335574, + "learning_rate": 1.1726096538511122e-05, + "loss": 1.6477, + "step": 9495 + }, + { + "epoch": 0.8490701001430615, + "grad_norm": 0.5079489761564318, + "learning_rate": 1.1712492939396157e-05, + "loss": 1.7143, + "step": 9496 + }, + { + "epoch": 0.8491595135908441, + "grad_norm": 0.8420634358663268, + "learning_rate": 1.1698896744847809e-05, + "loss": 1.5326, + "step": 9497 + }, + { + "epoch": 0.8492489270386266, + "grad_norm": 0.8236576144318039, + "learning_rate": 1.168530795600632e-05, + "loss": 1.4586, + "step": 9498 + }, + { + "epoch": 0.8493383404864091, + "grad_norm": 0.5129881745063006, + "learning_rate": 1.1671726574011399e-05, + "loss": 1.372, + "step": 9499 + }, + { + "epoch": 0.8494277539341917, + "grad_norm": 0.6533412700858054, + "learning_rate": 1.1658152600002104e-05, + "loss": 1.4897, + "step": 9500 + }, + { + "epoch": 0.8495171673819742, + "grad_norm": 0.5188491240221667, + "learning_rate": 1.1644586035116856e-05, + "loss": 1.6581, + "step": 9501 + }, + { + "epoch": 0.8496065808297568, + "grad_norm": 0.5537836562537983, + "learning_rate": 1.1631026880493468e-05, + "loss": 1.6357, + "step": 9502 + }, + { + "epoch": 0.8496959942775394, + "grad_norm": 0.5324972006300944, + "learning_rate": 1.1617475137269152e-05, + "loss": 1.4934, + "step": 9503 + }, + { + "epoch": 0.8497854077253219, + "grad_norm": 0.6321087932066992, + "learning_rate": 1.1603930806580444e-05, + "loss": 1.5633, + "step": 9504 + }, + { + "epoch": 0.8498748211731044, + "grad_norm": 0.395213406527626, + "learning_rate": 1.15903938895633e-05, + "loss": 1.4808, + "step": 9505 + }, + { + "epoch": 0.849964234620887, + "grad_norm": 0.33399004026999685, + "learning_rate": 1.157686438735307e-05, + "loss": 1.3582, + "step": 9506 + }, + { + "epoch": 0.8500536480686696, + "grad_norm": 0.4407594087388458, + "learning_rate": 1.156334230108439e-05, + "loss": 1.4978, + "step": 9507 + }, + { + "epoch": 0.850143061516452, + "grad_norm": 0.8382578651422923, + "learning_rate": 1.1549827631891418e-05, + "loss": 1.6273, + "step": 9508 + }, + { + "epoch": 0.8502324749642346, + "grad_norm": 0.4423856333206398, + "learning_rate": 1.1536320380907596e-05, + "loss": 1.6971, + "step": 9509 + }, + { + "epoch": 0.8503218884120172, + "grad_norm": 0.5251841996254687, + "learning_rate": 1.1522820549265723e-05, + "loss": 1.6576, + "step": 9510 + }, + { + "epoch": 0.8504113018597997, + "grad_norm": 0.4152299543079439, + "learning_rate": 1.1509328138098041e-05, + "loss": 1.2402, + "step": 9511 + }, + { + "epoch": 0.8505007153075823, + "grad_norm": 0.43521275744630566, + "learning_rate": 1.1495843148536157e-05, + "loss": 1.3603, + "step": 9512 + }, + { + "epoch": 0.8505901287553648, + "grad_norm": 0.5339636095081783, + "learning_rate": 1.1482365581711008e-05, + "loss": 2.0219, + "step": 9513 + }, + { + "epoch": 0.8506795422031473, + "grad_norm": 0.42309507047947753, + "learning_rate": 1.1468895438752947e-05, + "loss": 1.8876, + "step": 9514 + }, + { + "epoch": 0.8507689556509299, + "grad_norm": 0.4100996296301473, + "learning_rate": 1.1455432720791714e-05, + "loss": 1.596, + "step": 9515 + }, + { + "epoch": 0.8508583690987125, + "grad_norm": 0.433795297487167, + "learning_rate": 1.1441977428956396e-05, + "loss": 1.6984, + "step": 9516 + }, + { + "epoch": 0.850947782546495, + "grad_norm": 0.4306964342394804, + "learning_rate": 1.1428529564375502e-05, + "loss": 1.1696, + "step": 9517 + }, + { + "epoch": 0.8510371959942775, + "grad_norm": 0.5246133220296171, + "learning_rate": 1.1415089128176847e-05, + "loss": 1.8172, + "step": 9518 + }, + { + "epoch": 0.8511266094420601, + "grad_norm": 0.49588270568503856, + "learning_rate": 1.1401656121487692e-05, + "loss": 1.7505, + "step": 9519 + }, + { + "epoch": 0.8512160228898427, + "grad_norm": 0.5225842899812342, + "learning_rate": 1.1388230545434653e-05, + "loss": 1.4441, + "step": 9520 + }, + { + "epoch": 0.8513054363376252, + "grad_norm": 0.4501659765371375, + "learning_rate": 1.1374812401143653e-05, + "loss": 1.7417, + "step": 9521 + }, + { + "epoch": 0.8513948497854077, + "grad_norm": 0.6646627406324452, + "learning_rate": 1.1361401689740137e-05, + "loss": 1.4707, + "step": 9522 + }, + { + "epoch": 0.8514842632331903, + "grad_norm": 0.6010774297505285, + "learning_rate": 1.1347998412348825e-05, + "loss": 2.0182, + "step": 9523 + }, + { + "epoch": 0.8515736766809728, + "grad_norm": 0.6430856745890434, + "learning_rate": 1.13346025700938e-05, + "loss": 1.8473, + "step": 9524 + }, + { + "epoch": 0.8516630901287554, + "grad_norm": 0.5922339603389115, + "learning_rate": 1.1321214164098582e-05, + "loss": 1.4826, + "step": 9525 + }, + { + "epoch": 0.8517525035765379, + "grad_norm": 0.5209019239567381, + "learning_rate": 1.1307833195486062e-05, + "loss": 1.8978, + "step": 9526 + }, + { + "epoch": 0.8518419170243204, + "grad_norm": 0.42609475756411747, + "learning_rate": 1.1294459665378432e-05, + "loss": 1.6835, + "step": 9527 + }, + { + "epoch": 0.851931330472103, + "grad_norm": 0.5223913278832606, + "learning_rate": 1.1281093574897338e-05, + "loss": 1.4354, + "step": 9528 + }, + { + "epoch": 0.8520207439198856, + "grad_norm": 0.8545221314405078, + "learning_rate": 1.1267734925163787e-05, + "loss": 1.9546, + "step": 9529 + }, + { + "epoch": 0.852110157367668, + "grad_norm": 0.9584005756439631, + "learning_rate": 1.1254383717298134e-05, + "loss": 1.9313, + "step": 9530 + }, + { + "epoch": 0.8521995708154506, + "grad_norm": 0.6524066707762853, + "learning_rate": 1.1241039952420173e-05, + "loss": 1.5929, + "step": 9531 + }, + { + "epoch": 0.8522889842632332, + "grad_norm": 0.3023167498292897, + "learning_rate": 1.1227703631648978e-05, + "loss": 1.3177, + "step": 9532 + }, + { + "epoch": 0.8523783977110158, + "grad_norm": 0.29774910292463497, + "learning_rate": 1.1214374756103064e-05, + "loss": 1.1951, + "step": 9533 + }, + { + "epoch": 0.8524678111587983, + "grad_norm": 0.3934742757444131, + "learning_rate": 1.1201053326900313e-05, + "loss": 1.452, + "step": 9534 + }, + { + "epoch": 0.8525572246065808, + "grad_norm": 0.6998759919748265, + "learning_rate": 1.1187739345157977e-05, + "loss": 1.4117, + "step": 9535 + }, + { + "epoch": 0.8526466380543634, + "grad_norm": 0.5456339874147318, + "learning_rate": 1.1174432811992685e-05, + "loss": 1.3654, + "step": 9536 + }, + { + "epoch": 0.8527360515021459, + "grad_norm": 0.488108138925094, + "learning_rate": 1.1161133728520467e-05, + "loss": 1.4123, + "step": 9537 + }, + { + "epoch": 0.8528254649499285, + "grad_norm": 0.36932784721804784, + "learning_rate": 1.1147842095856642e-05, + "loss": 1.4132, + "step": 9538 + }, + { + "epoch": 0.852914878397711, + "grad_norm": 0.6826310694522884, + "learning_rate": 1.1134557915115994e-05, + "loss": 1.7643, + "step": 9539 + }, + { + "epoch": 0.8530042918454935, + "grad_norm": 0.3445617307353773, + "learning_rate": 1.112128118741268e-05, + "loss": 1.1789, + "step": 9540 + }, + { + "epoch": 0.8530937052932761, + "grad_norm": 0.8682277058728206, + "learning_rate": 1.1108011913860128e-05, + "loss": 1.4601, + "step": 9541 + }, + { + "epoch": 0.8531831187410587, + "grad_norm": 0.6653885975219622, + "learning_rate": 1.1094750095571282e-05, + "loss": 2.1739, + "step": 9542 + }, + { + "epoch": 0.8532725321888412, + "grad_norm": 0.46313141010449743, + "learning_rate": 1.1081495733658409e-05, + "loss": 1.9677, + "step": 9543 + }, + { + "epoch": 0.8533619456366237, + "grad_norm": 0.6291117298952967, + "learning_rate": 1.1068248829233063e-05, + "loss": 1.4672, + "step": 9544 + }, + { + "epoch": 0.8534513590844063, + "grad_norm": 0.8922696372792952, + "learning_rate": 1.10550093834063e-05, + "loss": 1.8055, + "step": 9545 + }, + { + "epoch": 0.8535407725321889, + "grad_norm": 0.5341631061218521, + "learning_rate": 1.1041777397288488e-05, + "loss": 1.7973, + "step": 9546 + }, + { + "epoch": 0.8536301859799714, + "grad_norm": 0.6127008699472747, + "learning_rate": 1.1028552871989362e-05, + "loss": 1.2661, + "step": 9547 + }, + { + "epoch": 0.8537195994277539, + "grad_norm": 0.8150391170940798, + "learning_rate": 1.1015335808618055e-05, + "loss": 1.5567, + "step": 9548 + }, + { + "epoch": 0.8538090128755365, + "grad_norm": 0.3987029669180094, + "learning_rate": 1.100212620828307e-05, + "loss": 1.6258, + "step": 9549 + }, + { + "epoch": 0.853898426323319, + "grad_norm": 0.8316086147920618, + "learning_rate": 1.0988924072092266e-05, + "loss": 2.0412, + "step": 9550 + }, + { + "epoch": 0.8539878397711016, + "grad_norm": 0.4583684068550533, + "learning_rate": 1.0975729401152934e-05, + "loss": 1.4536, + "step": 9551 + }, + { + "epoch": 0.8540772532188842, + "grad_norm": 0.5758734046530547, + "learning_rate": 1.0962542196571634e-05, + "loss": 1.7418, + "step": 9552 + }, + { + "epoch": 0.8541666666666666, + "grad_norm": 0.6785406372126193, + "learning_rate": 1.0949362459454393e-05, + "loss": 1.5609, + "step": 9553 + }, + { + "epoch": 0.8542560801144492, + "grad_norm": 0.6396113370934444, + "learning_rate": 1.0936190190906603e-05, + "loss": 1.3559, + "step": 9554 + }, + { + "epoch": 0.8543454935622318, + "grad_norm": 0.4527790217686671, + "learning_rate": 1.0923025392032937e-05, + "loss": 1.5253, + "step": 9555 + }, + { + "epoch": 0.8544349070100143, + "grad_norm": 0.5062077178012462, + "learning_rate": 1.0909868063937567e-05, + "loss": 1.4998, + "step": 9556 + }, + { + "epoch": 0.8545243204577968, + "grad_norm": 0.47380676048520537, + "learning_rate": 1.0896718207723988e-05, + "loss": 1.5863, + "step": 9557 + }, + { + "epoch": 0.8546137339055794, + "grad_norm": 0.639279886488448, + "learning_rate": 1.0883575824495029e-05, + "loss": 1.4344, + "step": 9558 + }, + { + "epoch": 0.854703147353362, + "grad_norm": 0.7047869820276216, + "learning_rate": 1.0870440915352942e-05, + "loss": 1.6763, + "step": 9559 + }, + { + "epoch": 0.8547925608011445, + "grad_norm": 0.5464139781088589, + "learning_rate": 1.0857313481399355e-05, + "loss": 1.5302, + "step": 9560 + }, + { + "epoch": 0.8548819742489271, + "grad_norm": 0.6528238548930616, + "learning_rate": 1.0844193523735202e-05, + "loss": 2.0796, + "step": 9561 + }, + { + "epoch": 0.8549713876967096, + "grad_norm": 0.5140101845319972, + "learning_rate": 1.0831081043460868e-05, + "loss": 1.5765, + "step": 9562 + }, + { + "epoch": 0.8550608011444921, + "grad_norm": 0.5642562604700747, + "learning_rate": 1.081797604167608e-05, + "loss": 1.8326, + "step": 9563 + }, + { + "epoch": 0.8551502145922747, + "grad_norm": 0.8357127105897059, + "learning_rate": 1.0804878519479943e-05, + "loss": 1.7485, + "step": 9564 + }, + { + "epoch": 0.8552396280400573, + "grad_norm": 0.5349214222808499, + "learning_rate": 1.079178847797091e-05, + "loss": 1.2325, + "step": 9565 + }, + { + "epoch": 0.8553290414878397, + "grad_norm": 0.4565508692834867, + "learning_rate": 1.0778705918246867e-05, + "loss": 1.7107, + "step": 9566 + }, + { + "epoch": 0.8554184549356223, + "grad_norm": 0.751877329666437, + "learning_rate": 1.0765630841404994e-05, + "loss": 1.9397, + "step": 9567 + }, + { + "epoch": 0.8555078683834049, + "grad_norm": 0.7611309641497033, + "learning_rate": 1.0752563248541891e-05, + "loss": 1.729, + "step": 9568 + }, + { + "epoch": 0.8555972818311874, + "grad_norm": 0.4646755002398883, + "learning_rate": 1.0739503140753516e-05, + "loss": 1.8733, + "step": 9569 + }, + { + "epoch": 0.85568669527897, + "grad_norm": 0.3784304095820343, + "learning_rate": 1.0726450519135222e-05, + "loss": 1.3876, + "step": 9570 + }, + { + "epoch": 0.8557761087267525, + "grad_norm": 0.7246758274112417, + "learning_rate": 1.0713405384781727e-05, + "loss": 1.5519, + "step": 9571 + }, + { + "epoch": 0.855865522174535, + "grad_norm": 0.6044449920899148, + "learning_rate": 1.0700367738787064e-05, + "loss": 1.7277, + "step": 9572 + }, + { + "epoch": 0.8559549356223176, + "grad_norm": 0.4791764812243888, + "learning_rate": 1.0687337582244727e-05, + "loss": 1.8736, + "step": 9573 + }, + { + "epoch": 0.8560443490701002, + "grad_norm": 0.4523316481623316, + "learning_rate": 1.067431491624753e-05, + "loss": 1.3855, + "step": 9574 + }, + { + "epoch": 0.8561337625178826, + "grad_norm": 0.47913599124060724, + "learning_rate": 1.0661299741887654e-05, + "loss": 1.6403, + "step": 9575 + }, + { + "epoch": 0.8562231759656652, + "grad_norm": 0.47003852469490903, + "learning_rate": 1.0648292060256649e-05, + "loss": 1.3166, + "step": 9576 + }, + { + "epoch": 0.8563125894134478, + "grad_norm": 0.46437172127254045, + "learning_rate": 1.0635291872445518e-05, + "loss": 1.6523, + "step": 9577 + }, + { + "epoch": 0.8564020028612304, + "grad_norm": 0.6447483303164622, + "learning_rate": 1.0622299179544516e-05, + "loss": 1.8977, + "step": 9578 + }, + { + "epoch": 0.8564914163090128, + "grad_norm": 0.4588302171723099, + "learning_rate": 1.0609313982643331e-05, + "loss": 2.1287, + "step": 9579 + }, + { + "epoch": 0.8565808297567954, + "grad_norm": 4.340528268463359, + "learning_rate": 1.0596336282831054e-05, + "loss": 6.1048, + "step": 9580 + }, + { + "epoch": 0.856670243204578, + "grad_norm": 0.9292673022603314, + "learning_rate": 1.0583366081196066e-05, + "loss": 1.4739, + "step": 9581 + }, + { + "epoch": 0.8567596566523605, + "grad_norm": 2.056045222420623, + "learning_rate": 1.0570403378826166e-05, + "loss": 1.8529, + "step": 9582 + }, + { + "epoch": 0.8568490701001431, + "grad_norm": 1.5860088818885087, + "learning_rate": 1.0557448176808537e-05, + "loss": 1.5175, + "step": 9583 + }, + { + "epoch": 0.8569384835479256, + "grad_norm": 1.683778602835672, + "learning_rate": 1.0544500476229713e-05, + "loss": 1.7656, + "step": 9584 + }, + { + "epoch": 0.8570278969957081, + "grad_norm": 0.5700459327344, + "learning_rate": 1.0531560278175611e-05, + "loss": 1.5711, + "step": 9585 + }, + { + "epoch": 0.8571173104434907, + "grad_norm": 0.4165726884282964, + "learning_rate": 1.0518627583731477e-05, + "loss": 1.519, + "step": 9586 + }, + { + "epoch": 0.8572067238912733, + "grad_norm": 1.2222417861273733, + "learning_rate": 1.0505702393981987e-05, + "loss": 1.3743, + "step": 9587 + }, + { + "epoch": 0.8572961373390557, + "grad_norm": 1.3711858790939173, + "learning_rate": 1.0492784710011184e-05, + "loss": 1.5018, + "step": 9588 + }, + { + "epoch": 0.8573855507868383, + "grad_norm": 1.4935691949677183, + "learning_rate": 1.04798745329024e-05, + "loss": 1.6351, + "step": 9589 + }, + { + "epoch": 0.8574749642346209, + "grad_norm": 1.131437202236348, + "learning_rate": 1.0466971863738406e-05, + "loss": 1.7861, + "step": 9590 + }, + { + "epoch": 0.8575643776824035, + "grad_norm": 0.4635947693816503, + "learning_rate": 1.045407670360139e-05, + "loss": 1.5155, + "step": 9591 + }, + { + "epoch": 0.857653791130186, + "grad_norm": 0.6703868956611526, + "learning_rate": 1.0441189053572809e-05, + "loss": 1.561, + "step": 9592 + }, + { + "epoch": 0.8577432045779685, + "grad_norm": 0.8017982269582014, + "learning_rate": 1.0428308914733531e-05, + "loss": 1.6777, + "step": 9593 + }, + { + "epoch": 0.8578326180257511, + "grad_norm": 1.350406539834385, + "learning_rate": 1.0415436288163826e-05, + "loss": 2.003, + "step": 9594 + }, + { + "epoch": 0.8579220314735336, + "grad_norm": 0.7453421168590242, + "learning_rate": 1.0402571174943276e-05, + "loss": 1.5221, + "step": 9595 + }, + { + "epoch": 0.8580114449213162, + "grad_norm": 0.6110584377847877, + "learning_rate": 1.0389713576150883e-05, + "loss": 1.4549, + "step": 9596 + }, + { + "epoch": 0.8581008583690987, + "grad_norm": 0.5349723004276606, + "learning_rate": 1.0376863492864975e-05, + "loss": 1.9237, + "step": 9597 + }, + { + "epoch": 0.8581902718168812, + "grad_norm": 0.5413660874562505, + "learning_rate": 1.0364020926163298e-05, + "loss": 1.6067, + "step": 9598 + }, + { + "epoch": 0.8582796852646638, + "grad_norm": 0.9187053187005384, + "learning_rate": 1.0351185877122938e-05, + "loss": 1.6957, + "step": 9599 + }, + { + "epoch": 0.8583690987124464, + "grad_norm": 0.7198054795628895, + "learning_rate": 1.0338358346820353e-05, + "loss": 1.501, + "step": 9600 + }, + { + "epoch": 0.858458512160229, + "grad_norm": 0.644296059953282, + "learning_rate": 1.0325538336331364e-05, + "loss": 1.4234, + "step": 9601 + }, + { + "epoch": 0.8585479256080114, + "grad_norm": 0.6227013323793232, + "learning_rate": 1.0312725846731175e-05, + "loss": 1.5781, + "step": 9602 + }, + { + "epoch": 0.858637339055794, + "grad_norm": 0.47836686101517134, + "learning_rate": 1.0299920879094372e-05, + "loss": 1.6601, + "step": 9603 + }, + { + "epoch": 0.8587267525035766, + "grad_norm": 0.4776221938834239, + "learning_rate": 1.0287123434494827e-05, + "loss": 1.4094, + "step": 9604 + }, + { + "epoch": 0.8588161659513591, + "grad_norm": 0.35478565861696876, + "learning_rate": 1.027433351400594e-05, + "loss": 1.3907, + "step": 9605 + }, + { + "epoch": 0.8589055793991416, + "grad_norm": 0.7957785791402386, + "learning_rate": 1.0261551118700318e-05, + "loss": 1.6294, + "step": 9606 + }, + { + "epoch": 0.8589949928469242, + "grad_norm": 0.5464213601469712, + "learning_rate": 1.0248776249650027e-05, + "loss": 1.6078, + "step": 9607 + }, + { + "epoch": 0.8590844062947067, + "grad_norm": 0.6802755805841603, + "learning_rate": 1.0236008907926508e-05, + "loss": 1.6066, + "step": 9608 + }, + { + "epoch": 0.8591738197424893, + "grad_norm": 0.4808536257401718, + "learning_rate": 1.0223249094600485e-05, + "loss": 1.3961, + "step": 9609 + }, + { + "epoch": 0.8592632331902719, + "grad_norm": 0.48923973459572445, + "learning_rate": 1.0210496810742143e-05, + "loss": 1.3929, + "step": 9610 + }, + { + "epoch": 0.8593526466380543, + "grad_norm": 0.7522793361472829, + "learning_rate": 1.0197752057420995e-05, + "loss": 1.6214, + "step": 9611 + }, + { + "epoch": 0.8594420600858369, + "grad_norm": 0.9534868539886454, + "learning_rate": 1.018501483570592e-05, + "loss": 6.0913, + "step": 9612 + }, + { + "epoch": 0.8595314735336195, + "grad_norm": 0.42344225464274937, + "learning_rate": 1.0172285146665195e-05, + "loss": 1.5185, + "step": 9613 + }, + { + "epoch": 0.859620886981402, + "grad_norm": 0.4645132016972657, + "learning_rate": 1.0159562991366444e-05, + "loss": 1.5453, + "step": 9614 + }, + { + "epoch": 0.8597103004291845, + "grad_norm": 0.4741621730972156, + "learning_rate": 1.0146848370876627e-05, + "loss": 1.3714, + "step": 9615 + }, + { + "epoch": 0.8597997138769671, + "grad_norm": 0.4308214944366189, + "learning_rate": 1.013414128626211e-05, + "loss": 1.7748, + "step": 9616 + }, + { + "epoch": 0.8598891273247496, + "grad_norm": 0.5370640215356691, + "learning_rate": 1.0121441738588644e-05, + "loss": 1.6109, + "step": 9617 + }, + { + "epoch": 0.8599785407725322, + "grad_norm": 2.91774105745873, + "learning_rate": 1.0108749728921319e-05, + "loss": 6.1346, + "step": 9618 + }, + { + "epoch": 0.8600679542203148, + "grad_norm": 0.46846809960683156, + "learning_rate": 1.0096065258324606e-05, + "loss": 1.4091, + "step": 9619 + }, + { + "epoch": 0.8601573676680973, + "grad_norm": 0.4060883442953093, + "learning_rate": 1.0083388327862298e-05, + "loss": 1.6046, + "step": 9620 + }, + { + "epoch": 0.8602467811158798, + "grad_norm": 0.38087152741477753, + "learning_rate": 1.0070718938597623e-05, + "loss": 1.3431, + "step": 9621 + }, + { + "epoch": 0.8603361945636624, + "grad_norm": 0.6891611003424831, + "learning_rate": 1.0058057091593154e-05, + "loss": 1.2074, + "step": 9622 + }, + { + "epoch": 0.860425608011445, + "grad_norm": 0.9477028281981928, + "learning_rate": 1.0045402787910818e-05, + "loss": 1.8607, + "step": 9623 + }, + { + "epoch": 0.8605150214592274, + "grad_norm": 0.9677473554307602, + "learning_rate": 1.0032756028611878e-05, + "loss": 1.9194, + "step": 9624 + }, + { + "epoch": 0.86060443490701, + "grad_norm": 0.6495444871296352, + "learning_rate": 1.0020116814757085e-05, + "loss": 1.3129, + "step": 9625 + }, + { + "epoch": 0.8606938483547926, + "grad_norm": 0.5482360574871846, + "learning_rate": 1.0007485147406404e-05, + "loss": 1.739, + "step": 9626 + }, + { + "epoch": 0.8607832618025751, + "grad_norm": 0.5960575763043625, + "learning_rate": 9.99486102761925e-06, + "loss": 1.3774, + "step": 9627 + }, + { + "epoch": 0.8608726752503576, + "grad_norm": 0.8820214935289441, + "learning_rate": 9.982244456454427e-06, + "loss": 1.6612, + "step": 9628 + }, + { + "epoch": 0.8609620886981402, + "grad_norm": 1.2636908090007455, + "learning_rate": 9.969635434970037e-06, + "loss": 1.8938, + "step": 9629 + }, + { + "epoch": 0.8610515021459227, + "grad_norm": 0.9284715778036442, + "learning_rate": 9.957033964223582e-06, + "loss": 1.8396, + "step": 9630 + }, + { + "epoch": 0.8611409155937053, + "grad_norm": 0.4453715190743825, + "learning_rate": 9.944440045271953e-06, + "loss": 1.4562, + "step": 9631 + }, + { + "epoch": 0.8612303290414879, + "grad_norm": 1.530033857915378, + "learning_rate": 9.931853679171377e-06, + "loss": 1.598, + "step": 9632 + }, + { + "epoch": 0.8613197424892703, + "grad_norm": 1.0530740937359915, + "learning_rate": 9.919274866977457e-06, + "loss": 1.6975, + "step": 9633 + }, + { + "epoch": 0.8614091559370529, + "grad_norm": 1.0780580646000837, + "learning_rate": 9.90670360974517e-06, + "loss": 1.9291, + "step": 9634 + }, + { + "epoch": 0.8614985693848355, + "grad_norm": 0.6693935899880978, + "learning_rate": 9.894139908528843e-06, + "loss": 1.7962, + "step": 9635 + }, + { + "epoch": 0.8615879828326181, + "grad_norm": 0.6606888332090072, + "learning_rate": 9.881583764382175e-06, + "loss": 1.6451, + "step": 9636 + }, + { + "epoch": 0.8616773962804005, + "grad_norm": 0.46502871112163485, + "learning_rate": 9.869035178358266e-06, + "loss": 1.4799, + "step": 9637 + }, + { + "epoch": 0.8617668097281831, + "grad_norm": 0.5669703828794241, + "learning_rate": 9.856494151509488e-06, + "loss": 1.3141, + "step": 9638 + }, + { + "epoch": 0.8618562231759657, + "grad_norm": 0.7728999359198011, + "learning_rate": 9.84396068488771e-06, + "loss": 1.5567, + "step": 9639 + }, + { + "epoch": 0.8619456366237482, + "grad_norm": 0.8458128447245306, + "learning_rate": 9.831434779544057e-06, + "loss": 1.8199, + "step": 9640 + }, + { + "epoch": 0.8620350500715308, + "grad_norm": 0.7849589892714159, + "learning_rate": 9.818916436529069e-06, + "loss": 1.3833, + "step": 9641 + }, + { + "epoch": 0.8621244635193133, + "grad_norm": 0.7496197270947149, + "learning_rate": 9.80640565689267e-06, + "loss": 1.7791, + "step": 9642 + }, + { + "epoch": 0.8622138769670958, + "grad_norm": 0.6086616632880409, + "learning_rate": 9.793902441684077e-06, + "loss": 1.8461, + "step": 9643 + }, + { + "epoch": 0.8623032904148784, + "grad_norm": 0.41614900605861604, + "learning_rate": 9.781406791951952e-06, + "loss": 1.4886, + "step": 9644 + }, + { + "epoch": 0.862392703862661, + "grad_norm": 0.4942716974498269, + "learning_rate": 9.76891870874428e-06, + "loss": 1.7265, + "step": 9645 + }, + { + "epoch": 0.8624821173104434, + "grad_norm": 0.3955536519132566, + "learning_rate": 9.756438193108419e-06, + "loss": 1.5234, + "step": 9646 + }, + { + "epoch": 0.862571530758226, + "grad_norm": 0.8398038665764358, + "learning_rate": 9.743965246091102e-06, + "loss": 2.0731, + "step": 9647 + }, + { + "epoch": 0.8626609442060086, + "grad_norm": 0.4029192867878009, + "learning_rate": 9.731499868738447e-06, + "loss": 1.3053, + "step": 9648 + }, + { + "epoch": 0.8627503576537912, + "grad_norm": 0.5926439224891431, + "learning_rate": 9.719042062095851e-06, + "loss": 1.3822, + "step": 9649 + }, + { + "epoch": 0.8628397711015737, + "grad_norm": 0.60328983502736, + "learning_rate": 9.706591827208166e-06, + "loss": 1.7155, + "step": 9650 + }, + { + "epoch": 0.8629291845493562, + "grad_norm": 0.32141679351949687, + "learning_rate": 9.694149165119603e-06, + "loss": 1.2792, + "step": 9651 + }, + { + "epoch": 0.8630185979971388, + "grad_norm": 0.42609073748223014, + "learning_rate": 9.68171407687365e-06, + "loss": 1.1944, + "step": 9652 + }, + { + "epoch": 0.8631080114449213, + "grad_norm": 0.5951468442006166, + "learning_rate": 9.66928656351329e-06, + "loss": 1.3383, + "step": 9653 + }, + { + "epoch": 0.8631974248927039, + "grad_norm": 0.7046781332049675, + "learning_rate": 9.656866626080763e-06, + "loss": 1.7259, + "step": 9654 + }, + { + "epoch": 0.8632868383404864, + "grad_norm": 0.5059036269317316, + "learning_rate": 9.644454265617731e-06, + "loss": 1.5342, + "step": 9655 + }, + { + "epoch": 0.8633762517882689, + "grad_norm": 0.5075806148512485, + "learning_rate": 9.632049483165184e-06, + "loss": 1.672, + "step": 9656 + }, + { + "epoch": 0.8634656652360515, + "grad_norm": 0.3661252314830899, + "learning_rate": 9.619652279763536e-06, + "loss": 1.478, + "step": 9657 + }, + { + "epoch": 0.8635550786838341, + "grad_norm": 0.550237616535287, + "learning_rate": 9.607262656452475e-06, + "loss": 1.8122, + "step": 9658 + }, + { + "epoch": 0.8636444921316166, + "grad_norm": 0.7209423609922219, + "learning_rate": 9.59488061427114e-06, + "loss": 1.8671, + "step": 9659 + }, + { + "epoch": 0.8637339055793991, + "grad_norm": 0.4481801308882162, + "learning_rate": 9.582506154257976e-06, + "loss": 1.558, + "step": 9660 + }, + { + "epoch": 0.8638233190271817, + "grad_norm": 0.5415719240289604, + "learning_rate": 9.57013927745083e-06, + "loss": 1.5129, + "step": 9661 + }, + { + "epoch": 0.8639127324749643, + "grad_norm": 0.5703767921491335, + "learning_rate": 9.557779984886905e-06, + "loss": 1.832, + "step": 9662 + }, + { + "epoch": 0.8640021459227468, + "grad_norm": 0.5251848875235138, + "learning_rate": 9.545428277602731e-06, + "loss": 1.7077, + "step": 9663 + }, + { + "epoch": 0.8640915593705293, + "grad_norm": 0.7354726949383517, + "learning_rate": 9.533084156634242e-06, + "loss": 1.7763, + "step": 9664 + }, + { + "epoch": 0.8641809728183119, + "grad_norm": 0.3912230843313184, + "learning_rate": 9.520747623016747e-06, + "loss": 1.549, + "step": 9665 + }, + { + "epoch": 0.8642703862660944, + "grad_norm": 0.9356908765869837, + "learning_rate": 9.508418677784847e-06, + "loss": 1.8086, + "step": 9666 + }, + { + "epoch": 0.864359799713877, + "grad_norm": 0.6780054983087196, + "learning_rate": 9.496097321972597e-06, + "loss": 1.5652, + "step": 9667 + }, + { + "epoch": 0.8644492131616596, + "grad_norm": 1.0388130593917144, + "learning_rate": 9.48378355661339e-06, + "loss": 1.574, + "step": 9668 + }, + { + "epoch": 0.864538626609442, + "grad_norm": 0.7833516964267665, + "learning_rate": 9.471477382739912e-06, + "loss": 1.7046, + "step": 9669 + }, + { + "epoch": 0.8646280400572246, + "grad_norm": 0.9152471870359753, + "learning_rate": 9.459178801384304e-06, + "loss": 1.5745, + "step": 9670 + }, + { + "epoch": 0.8647174535050072, + "grad_norm": 0.4812551716736381, + "learning_rate": 9.446887813578031e-06, + "loss": 1.6483, + "step": 9671 + }, + { + "epoch": 0.8648068669527897, + "grad_norm": 0.5991425782034643, + "learning_rate": 9.434604420351911e-06, + "loss": 1.8876, + "step": 9672 + }, + { + "epoch": 0.8648962804005722, + "grad_norm": 0.7336249911980858, + "learning_rate": 9.422328622736142e-06, + "loss": 1.6391, + "step": 9673 + }, + { + "epoch": 0.8649856938483548, + "grad_norm": 0.51298685429687, + "learning_rate": 9.41006042176027e-06, + "loss": 1.2172, + "step": 9674 + }, + { + "epoch": 0.8650751072961373, + "grad_norm": 0.7805864890793808, + "learning_rate": 9.397799818453235e-06, + "loss": 1.8193, + "step": 9675 + }, + { + "epoch": 0.8651645207439199, + "grad_norm": 0.7365940313089312, + "learning_rate": 9.385546813843326e-06, + "loss": 1.6136, + "step": 9676 + }, + { + "epoch": 0.8652539341917024, + "grad_norm": 0.4090582937025843, + "learning_rate": 9.373301408958157e-06, + "loss": 1.7309, + "step": 9677 + }, + { + "epoch": 0.865343347639485, + "grad_norm": 0.5209309544157275, + "learning_rate": 9.361063604824738e-06, + "loss": 1.5268, + "step": 9678 + }, + { + "epoch": 0.8654327610872675, + "grad_norm": 0.549134815022632, + "learning_rate": 9.34883340246946e-06, + "loss": 1.5809, + "step": 9679 + }, + { + "epoch": 0.8655221745350501, + "grad_norm": 0.48025333358698447, + "learning_rate": 9.336610802918044e-06, + "loss": 1.7157, + "step": 9680 + }, + { + "epoch": 0.8656115879828327, + "grad_norm": 0.5936389717650448, + "learning_rate": 9.324395807195585e-06, + "loss": 1.7103, + "step": 9681 + }, + { + "epoch": 0.8657010014306151, + "grad_norm": 0.6223170535769162, + "learning_rate": 9.312188416326562e-06, + "loss": 1.7468, + "step": 9682 + }, + { + "epoch": 0.8657904148783977, + "grad_norm": 0.437269810459592, + "learning_rate": 9.299988631334755e-06, + "loss": 1.15, + "step": 9683 + }, + { + "epoch": 0.8658798283261803, + "grad_norm": 0.6832690961927617, + "learning_rate": 9.287796453243358e-06, + "loss": 1.661, + "step": 9684 + }, + { + "epoch": 0.8659692417739628, + "grad_norm": 0.44760714648550004, + "learning_rate": 9.275611883074941e-06, + "loss": 1.5403, + "step": 9685 + }, + { + "epoch": 0.8660586552217453, + "grad_norm": 0.654563709943089, + "learning_rate": 9.263434921851377e-06, + "loss": 1.6673, + "step": 9686 + }, + { + "epoch": 0.8661480686695279, + "grad_norm": 0.5992148623968355, + "learning_rate": 9.251265570593914e-06, + "loss": 1.3896, + "step": 9687 + }, + { + "epoch": 0.8662374821173104, + "grad_norm": 0.6514107434525406, + "learning_rate": 9.23910383032326e-06, + "loss": 1.8847, + "step": 9688 + }, + { + "epoch": 0.866326895565093, + "grad_norm": 0.808111941272789, + "learning_rate": 9.226949702059329e-06, + "loss": 2.0347, + "step": 9689 + }, + { + "epoch": 0.8664163090128756, + "grad_norm": 0.6979173184271455, + "learning_rate": 9.214803186821497e-06, + "loss": 1.6077, + "step": 9690 + }, + { + "epoch": 0.866505722460658, + "grad_norm": 0.8961247771833609, + "learning_rate": 9.202664285628504e-06, + "loss": 1.6022, + "step": 9691 + }, + { + "epoch": 0.8665951359084406, + "grad_norm": 0.4627239479271428, + "learning_rate": 9.190532999498392e-06, + "loss": 1.6698, + "step": 9692 + }, + { + "epoch": 0.8666845493562232, + "grad_norm": 0.6564529850780734, + "learning_rate": 9.178409329448601e-06, + "loss": 2.1444, + "step": 9693 + }, + { + "epoch": 0.8667739628040058, + "grad_norm": 0.4280387983318897, + "learning_rate": 9.16629327649593e-06, + "loss": 1.5565, + "step": 9694 + }, + { + "epoch": 0.8668633762517882, + "grad_norm": 0.5338599731157518, + "learning_rate": 9.154184841656544e-06, + "loss": 1.6881, + "step": 9695 + }, + { + "epoch": 0.8669527896995708, + "grad_norm": 0.49174321511175156, + "learning_rate": 9.142084025945984e-06, + "loss": 1.4757, + "step": 9696 + }, + { + "epoch": 0.8670422031473534, + "grad_norm": 0.5478275088968345, + "learning_rate": 9.129990830379087e-06, + "loss": 1.6589, + "step": 9697 + }, + { + "epoch": 0.8671316165951359, + "grad_norm": 0.38409022406374516, + "learning_rate": 9.117905255970116e-06, + "loss": 1.4603, + "step": 9698 + }, + { + "epoch": 0.8672210300429185, + "grad_norm": 0.4067406731422428, + "learning_rate": 9.105827303732695e-06, + "loss": 1.6126, + "step": 9699 + }, + { + "epoch": 0.867310443490701, + "grad_norm": 0.8099986685027852, + "learning_rate": 9.093756974679746e-06, + "loss": 1.6274, + "step": 9700 + }, + { + "epoch": 0.8673998569384835, + "grad_norm": 0.4575132081726312, + "learning_rate": 9.081694269823582e-06, + "loss": 1.6053, + "step": 9701 + }, + { + "epoch": 0.8674892703862661, + "grad_norm": 0.7481265704336569, + "learning_rate": 9.069639190175972e-06, + "loss": 1.7042, + "step": 9702 + }, + { + "epoch": 0.8675786838340487, + "grad_norm": 0.7769742548214825, + "learning_rate": 9.057591736747883e-06, + "loss": 1.9815, + "step": 9703 + }, + { + "epoch": 0.8676680972818311, + "grad_norm": 0.7114611281194081, + "learning_rate": 9.045551910549744e-06, + "loss": 1.7852, + "step": 9704 + }, + { + "epoch": 0.8677575107296137, + "grad_norm": 0.4488730487518198, + "learning_rate": 9.033519712591332e-06, + "loss": 1.6376, + "step": 9705 + }, + { + "epoch": 0.8678469241773963, + "grad_norm": 0.45792658791012286, + "learning_rate": 9.021495143881753e-06, + "loss": 1.6531, + "step": 9706 + }, + { + "epoch": 0.8679363376251789, + "grad_norm": 0.5195108174826393, + "learning_rate": 9.0094782054295e-06, + "loss": 1.636, + "step": 9707 + }, + { + "epoch": 0.8680257510729614, + "grad_norm": 0.45612338807089764, + "learning_rate": 8.997468898242422e-06, + "loss": 1.3605, + "step": 9708 + }, + { + "epoch": 0.8681151645207439, + "grad_norm": 0.5166265367977506, + "learning_rate": 8.985467223327726e-06, + "loss": 1.5174, + "step": 9709 + }, + { + "epoch": 0.8682045779685265, + "grad_norm": 0.38292012887313776, + "learning_rate": 8.973473181691993e-06, + "loss": 1.7317, + "step": 9710 + }, + { + "epoch": 0.868293991416309, + "grad_norm": 0.35120889713404724, + "learning_rate": 8.96148677434111e-06, + "loss": 1.2519, + "step": 9711 + }, + { + "epoch": 0.8683834048640916, + "grad_norm": 0.6971790988521369, + "learning_rate": 8.949508002280382e-06, + "loss": 2.0459, + "step": 9712 + }, + { + "epoch": 0.8684728183118741, + "grad_norm": 0.7831868795530676, + "learning_rate": 8.937536866514462e-06, + "loss": 1.6778, + "step": 9713 + }, + { + "epoch": 0.8685622317596566, + "grad_norm": 0.7669202908386522, + "learning_rate": 8.925573368047358e-06, + "loss": 1.3982, + "step": 9714 + }, + { + "epoch": 0.8686516452074392, + "grad_norm": 0.49314386132490995, + "learning_rate": 8.91361750788241e-06, + "loss": 1.6286, + "step": 9715 + }, + { + "epoch": 0.8687410586552218, + "grad_norm": 0.5941541408141751, + "learning_rate": 8.901669287022384e-06, + "loss": 1.3976, + "step": 9716 + }, + { + "epoch": 0.8688304721030042, + "grad_norm": 0.7301279369707556, + "learning_rate": 8.889728706469314e-06, + "loss": 1.7392, + "step": 9717 + }, + { + "epoch": 0.8689198855507868, + "grad_norm": 0.5903358528091389, + "learning_rate": 8.877795767224672e-06, + "loss": 1.6058, + "step": 9718 + }, + { + "epoch": 0.8690092989985694, + "grad_norm": 0.5249425028297839, + "learning_rate": 8.86587047028926e-06, + "loss": 1.6666, + "step": 9719 + }, + { + "epoch": 0.869098712446352, + "grad_norm": 0.3938518935400792, + "learning_rate": 8.853952816663213e-06, + "loss": 1.5355, + "step": 9720 + }, + { + "epoch": 0.8691881258941345, + "grad_norm": 0.8632345404662571, + "learning_rate": 8.842042807346051e-06, + "loss": 1.5754, + "step": 9721 + }, + { + "epoch": 0.869277539341917, + "grad_norm": 0.6518564189466506, + "learning_rate": 8.830140443336699e-06, + "loss": 1.9675, + "step": 9722 + }, + { + "epoch": 0.8693669527896996, + "grad_norm": 0.6938502776624257, + "learning_rate": 8.818245725633356e-06, + "loss": 1.2991, + "step": 9723 + }, + { + "epoch": 0.8694563662374821, + "grad_norm": 0.3921098653967205, + "learning_rate": 8.806358655233615e-06, + "loss": 1.7835, + "step": 9724 + }, + { + "epoch": 0.8695457796852647, + "grad_norm": 1.622947808481055, + "learning_rate": 8.794479233134456e-06, + "loss": 6.2009, + "step": 9725 + }, + { + "epoch": 0.8696351931330472, + "grad_norm": 0.8178708265848053, + "learning_rate": 8.78260746033217e-06, + "loss": 1.7593, + "step": 9726 + }, + { + "epoch": 0.8697246065808297, + "grad_norm": 0.8441928377016024, + "learning_rate": 8.770743337822418e-06, + "loss": 1.6138, + "step": 9727 + }, + { + "epoch": 0.8698140200286123, + "grad_norm": 0.4903159154777815, + "learning_rate": 8.758886866600257e-06, + "loss": 1.4313, + "step": 9728 + }, + { + "epoch": 0.8699034334763949, + "grad_norm": 0.4749029611084203, + "learning_rate": 8.74703804766005e-06, + "loss": 1.628, + "step": 9729 + }, + { + "epoch": 0.8699928469241774, + "grad_norm": 0.956931179887936, + "learning_rate": 8.735196881995589e-06, + "loss": 1.6255, + "step": 9730 + }, + { + "epoch": 0.8700822603719599, + "grad_norm": 0.3829031310052665, + "learning_rate": 8.723363370599924e-06, + "loss": 1.4075, + "step": 9731 + }, + { + "epoch": 0.8701716738197425, + "grad_norm": 0.6500835225662502, + "learning_rate": 8.71153751446553e-06, + "loss": 1.3005, + "step": 9732 + }, + { + "epoch": 0.870261087267525, + "grad_norm": 0.7599304414008264, + "learning_rate": 8.699719314584265e-06, + "loss": 1.7659, + "step": 9733 + }, + { + "epoch": 0.8703505007153076, + "grad_norm": 0.5204572837820541, + "learning_rate": 8.687908771947251e-06, + "loss": 1.4963, + "step": 9734 + }, + { + "epoch": 0.8704399141630901, + "grad_norm": 0.7003833776284063, + "learning_rate": 8.676105887545039e-06, + "loss": 2.0208, + "step": 9735 + }, + { + "epoch": 0.8705293276108726, + "grad_norm": 0.4246925484875429, + "learning_rate": 8.66431066236757e-06, + "loss": 1.8082, + "step": 9736 + }, + { + "epoch": 0.8706187410586552, + "grad_norm": 0.5329326386760447, + "learning_rate": 8.652523097404042e-06, + "loss": 1.5242, + "step": 9737 + }, + { + "epoch": 0.8707081545064378, + "grad_norm": 0.5050804334959852, + "learning_rate": 8.640743193643075e-06, + "loss": 1.5676, + "step": 9738 + }, + { + "epoch": 0.8707975679542204, + "grad_norm": 0.4205825711791025, + "learning_rate": 8.628970952072667e-06, + "loss": 1.3096, + "step": 9739 + }, + { + "epoch": 0.8708869814020028, + "grad_norm": 0.4143108067164644, + "learning_rate": 8.617206373680098e-06, + "loss": 1.4005, + "step": 9740 + }, + { + "epoch": 0.8709763948497854, + "grad_norm": 0.5946877703769281, + "learning_rate": 8.605449459452075e-06, + "loss": 1.5217, + "step": 9741 + }, + { + "epoch": 0.871065808297568, + "grad_norm": 0.5915271607449617, + "learning_rate": 8.593700210374622e-06, + "loss": 1.5648, + "step": 9742 + }, + { + "epoch": 0.8711552217453505, + "grad_norm": 0.5625885592150067, + "learning_rate": 8.58195862743314e-06, + "loss": 1.4742, + "step": 9743 + }, + { + "epoch": 0.871244635193133, + "grad_norm": 0.3471397274320741, + "learning_rate": 8.570224711612385e-06, + "loss": 1.2419, + "step": 9744 + }, + { + "epoch": 0.8713340486409156, + "grad_norm": 0.4848741349173154, + "learning_rate": 8.55849846389648e-06, + "loss": 1.527, + "step": 9745 + }, + { + "epoch": 0.8714234620886981, + "grad_norm": 0.616318201270939, + "learning_rate": 8.546779885268863e-06, + "loss": 1.4806, + "step": 9746 + }, + { + "epoch": 0.8715128755364807, + "grad_norm": 0.5362321494473367, + "learning_rate": 8.535068976712368e-06, + "loss": 1.733, + "step": 9747 + }, + { + "epoch": 0.8716022889842633, + "grad_norm": 0.4498418527263881, + "learning_rate": 8.523365739209188e-06, + "loss": 1.6449, + "step": 9748 + }, + { + "epoch": 0.8716917024320457, + "grad_norm": 0.6404824197696085, + "learning_rate": 8.511670173740816e-06, + "loss": 1.5862, + "step": 9749 + }, + { + "epoch": 0.8717811158798283, + "grad_norm": 0.7761813057401092, + "learning_rate": 8.499982281288221e-06, + "loss": 1.589, + "step": 9750 + }, + { + "epoch": 0.8718705293276109, + "grad_norm": 0.6395700763678995, + "learning_rate": 8.488302062831576e-06, + "loss": 1.596, + "step": 9751 + }, + { + "epoch": 0.8719599427753935, + "grad_norm": 0.43551460301855766, + "learning_rate": 8.476629519350532e-06, + "loss": 1.7196, + "step": 9752 + }, + { + "epoch": 0.8720493562231759, + "grad_norm": 0.4415889186667038, + "learning_rate": 8.464964651824048e-06, + "loss": 1.7511, + "step": 9753 + }, + { + "epoch": 0.8721387696709585, + "grad_norm": 0.49231809945801736, + "learning_rate": 8.453307461230409e-06, + "loss": 1.7247, + "step": 9754 + }, + { + "epoch": 0.8722281831187411, + "grad_norm": 0.47298823986336536, + "learning_rate": 8.441657948547322e-06, + "loss": 1.4752, + "step": 9755 + }, + { + "epoch": 0.8723175965665236, + "grad_norm": 0.3217054289909142, + "learning_rate": 8.430016114751805e-06, + "loss": 1.442, + "step": 9756 + }, + { + "epoch": 0.8724070100143062, + "grad_norm": 0.44788050584872835, + "learning_rate": 8.418381960820243e-06, + "loss": 1.348, + "step": 9757 + }, + { + "epoch": 0.8724964234620887, + "grad_norm": 0.4384147408082307, + "learning_rate": 8.40675548772839e-06, + "loss": 1.2709, + "step": 9758 + }, + { + "epoch": 0.8725858369098712, + "grad_norm": 0.5377299350439585, + "learning_rate": 8.395136696451355e-06, + "loss": 1.6004, + "step": 9759 + }, + { + "epoch": 0.8726752503576538, + "grad_norm": 0.49286155738115484, + "learning_rate": 8.383525587963558e-06, + "loss": 1.6624, + "step": 9760 + }, + { + "epoch": 0.8727646638054364, + "grad_norm": 0.4680437016377549, + "learning_rate": 8.371922163238821e-06, + "loss": 1.4657, + "step": 9761 + }, + { + "epoch": 0.8728540772532188, + "grad_norm": 0.5446828364129973, + "learning_rate": 8.36032642325033e-06, + "loss": 1.635, + "step": 9762 + }, + { + "epoch": 0.8729434907010014, + "grad_norm": 0.5813059736519085, + "learning_rate": 8.348738368970566e-06, + "loss": 1.4747, + "step": 9763 + }, + { + "epoch": 0.873032904148784, + "grad_norm": 0.4355339011911433, + "learning_rate": 8.337158001371449e-06, + "loss": 1.4563, + "step": 9764 + }, + { + "epoch": 0.8731223175965666, + "grad_norm": 0.5641082807712384, + "learning_rate": 8.325585321424178e-06, + "loss": 1.6743, + "step": 9765 + }, + { + "epoch": 0.873211731044349, + "grad_norm": 0.7653182835279374, + "learning_rate": 8.314020330099348e-06, + "loss": 1.7497, + "step": 9766 + }, + { + "epoch": 0.8733011444921316, + "grad_norm": 0.3782029100130729, + "learning_rate": 8.302463028366924e-06, + "loss": 1.5593, + "step": 9767 + }, + { + "epoch": 0.8733905579399142, + "grad_norm": 0.44875434607877784, + "learning_rate": 8.290913417196177e-06, + "loss": 1.4585, + "step": 9768 + }, + { + "epoch": 0.8734799713876967, + "grad_norm": 0.51286964538177, + "learning_rate": 8.279371497555755e-06, + "loss": 1.6053, + "step": 9769 + }, + { + "epoch": 0.8735693848354793, + "grad_norm": 0.6178128584530229, + "learning_rate": 8.26783727041367e-06, + "loss": 1.8819, + "step": 9770 + }, + { + "epoch": 0.8736587982832618, + "grad_norm": 0.5215297639300852, + "learning_rate": 8.256310736737294e-06, + "loss": 1.3403, + "step": 9771 + }, + { + "epoch": 0.8737482117310443, + "grad_norm": 0.5579653710806919, + "learning_rate": 8.244791897493342e-06, + "loss": 1.594, + "step": 9772 + }, + { + "epoch": 0.8738376251788269, + "grad_norm": 0.9976958868973235, + "learning_rate": 8.233280753647887e-06, + "loss": 1.8343, + "step": 9773 + }, + { + "epoch": 0.8739270386266095, + "grad_norm": 0.3808876990707221, + "learning_rate": 8.221777306166346e-06, + "loss": 1.3898, + "step": 9774 + }, + { + "epoch": 0.8740164520743919, + "grad_norm": 0.7498613983745707, + "learning_rate": 8.210281556013489e-06, + "loss": 1.6419, + "step": 9775 + }, + { + "epoch": 0.8741058655221745, + "grad_norm": 0.4846537614379837, + "learning_rate": 8.19879350415349e-06, + "loss": 1.7895, + "step": 9776 + }, + { + "epoch": 0.8741952789699571, + "grad_norm": 0.46848777415456305, + "learning_rate": 8.18731315154978e-06, + "loss": 1.2836, + "step": 9777 + }, + { + "epoch": 0.8742846924177397, + "grad_norm": 1.057453406860426, + "learning_rate": 8.175840499165244e-06, + "loss": 1.6096, + "step": 9778 + }, + { + "epoch": 0.8743741058655222, + "grad_norm": 0.4730845762075146, + "learning_rate": 8.16437554796209e-06, + "loss": 1.5556, + "step": 9779 + }, + { + "epoch": 0.8744635193133047, + "grad_norm": 1.2095905358664527, + "learning_rate": 8.152918298901836e-06, + "loss": 6.0401, + "step": 9780 + }, + { + "epoch": 0.8745529327610873, + "grad_norm": 0.5118363226993203, + "learning_rate": 8.141468752945392e-06, + "loss": 1.3468, + "step": 9781 + }, + { + "epoch": 0.8746423462088698, + "grad_norm": 0.4125943298252071, + "learning_rate": 8.130026911053045e-06, + "loss": 1.4282, + "step": 9782 + }, + { + "epoch": 0.8747317596566524, + "grad_norm": 0.6662414150315219, + "learning_rate": 8.118592774184385e-06, + "loss": 1.387, + "step": 9783 + }, + { + "epoch": 0.8748211731044349, + "grad_norm": 0.9826467959696426, + "learning_rate": 8.107166343298377e-06, + "loss": 1.9793, + "step": 9784 + }, + { + "epoch": 0.8749105865522174, + "grad_norm": 0.4701792811582305, + "learning_rate": 8.095747619353345e-06, + "loss": 1.5972, + "step": 9785 + }, + { + "epoch": 0.875, + "grad_norm": 0.5207836554259891, + "learning_rate": 8.084336603306974e-06, + "loss": 1.5234, + "step": 9786 + }, + { + "epoch": 0.8750894134477826, + "grad_norm": 0.4088178203891561, + "learning_rate": 8.072933296116303e-06, + "loss": 1.6338, + "step": 9787 + }, + { + "epoch": 0.8751788268955651, + "grad_norm": 0.5709277579095636, + "learning_rate": 8.061537698737675e-06, + "loss": 1.7434, + "step": 9788 + }, + { + "epoch": 0.8752682403433476, + "grad_norm": 0.728386053791918, + "learning_rate": 8.05014981212685e-06, + "loss": 1.7818, + "step": 9789 + }, + { + "epoch": 0.8753576537911302, + "grad_norm": 0.5496393827275402, + "learning_rate": 8.038769637238907e-06, + "loss": 1.6756, + "step": 9790 + }, + { + "epoch": 0.8754470672389127, + "grad_norm": 0.5480906192506764, + "learning_rate": 8.027397175028305e-06, + "loss": 1.8359, + "step": 9791 + }, + { + "epoch": 0.8755364806866953, + "grad_norm": 1.029523765339341, + "learning_rate": 8.016032426448817e-06, + "loss": 1.6297, + "step": 9792 + }, + { + "epoch": 0.8756258941344778, + "grad_norm": 0.5478511402448946, + "learning_rate": 8.00467539245362e-06, + "loss": 1.3883, + "step": 9793 + }, + { + "epoch": 0.8757153075822603, + "grad_norm": 0.7471635400704184, + "learning_rate": 7.993326073995189e-06, + "loss": 1.855, + "step": 9794 + }, + { + "epoch": 0.8758047210300429, + "grad_norm": 0.5308435929912083, + "learning_rate": 7.981984472025372e-06, + "loss": 1.5405, + "step": 9795 + }, + { + "epoch": 0.8758941344778255, + "grad_norm": 0.6767895763456201, + "learning_rate": 7.97065058749541e-06, + "loss": 1.9503, + "step": 9796 + }, + { + "epoch": 0.8759835479256081, + "grad_norm": 0.403768260704286, + "learning_rate": 7.959324421355797e-06, + "loss": 1.4095, + "step": 9797 + }, + { + "epoch": 0.8760729613733905, + "grad_norm": 0.47036879078378774, + "learning_rate": 7.948005974556539e-06, + "loss": 1.9163, + "step": 9798 + }, + { + "epoch": 0.8761623748211731, + "grad_norm": 0.5269608484233403, + "learning_rate": 7.936695248046822e-06, + "loss": 1.8918, + "step": 9799 + }, + { + "epoch": 0.8762517882689557, + "grad_norm": 0.43347455699696585, + "learning_rate": 7.925392242775288e-06, + "loss": 1.4974, + "step": 9800 + }, + { + "epoch": 0.8763412017167382, + "grad_norm": 0.8430269611165304, + "learning_rate": 7.91409695968991e-06, + "loss": 1.4897, + "step": 9801 + }, + { + "epoch": 0.8764306151645207, + "grad_norm": 0.7181096587583968, + "learning_rate": 7.90280939973802e-06, + "loss": 1.7518, + "step": 9802 + }, + { + "epoch": 0.8765200286123033, + "grad_norm": 0.36496982380003723, + "learning_rate": 7.891529563866274e-06, + "loss": 1.2493, + "step": 9803 + }, + { + "epoch": 0.8766094420600858, + "grad_norm": 0.5477126447118205, + "learning_rate": 7.8802574530207e-06, + "loss": 1.81, + "step": 9804 + }, + { + "epoch": 0.8766988555078684, + "grad_norm": 0.5826375926010359, + "learning_rate": 7.86899306814668e-06, + "loss": 1.5473, + "step": 9805 + }, + { + "epoch": 0.876788268955651, + "grad_norm": 0.5419661076684816, + "learning_rate": 7.857736410188953e-06, + "loss": 1.3992, + "step": 9806 + }, + { + "epoch": 0.8768776824034334, + "grad_norm": 0.5787471376900584, + "learning_rate": 7.846487480091603e-06, + "loss": 1.6672, + "step": 9807 + }, + { + "epoch": 0.876967095851216, + "grad_norm": 0.39348633968352037, + "learning_rate": 7.835246278798037e-06, + "loss": 1.5117, + "step": 9808 + }, + { + "epoch": 0.8770565092989986, + "grad_norm": 0.5994260203046093, + "learning_rate": 7.824012807251058e-06, + "loss": 1.7299, + "step": 9809 + }, + { + "epoch": 0.8771459227467812, + "grad_norm": 0.5829197030602922, + "learning_rate": 7.812787066392825e-06, + "loss": 1.4665, + "step": 9810 + }, + { + "epoch": 0.8772353361945636, + "grad_norm": 0.7202445781317052, + "learning_rate": 7.80156905716477e-06, + "loss": 1.9361, + "step": 9811 + }, + { + "epoch": 0.8773247496423462, + "grad_norm": 0.5936700192539756, + "learning_rate": 7.790358780507789e-06, + "loss": 1.5493, + "step": 9812 + }, + { + "epoch": 0.8774141630901288, + "grad_norm": 0.5345884837259567, + "learning_rate": 7.779156237362084e-06, + "loss": 1.8573, + "step": 9813 + }, + { + "epoch": 0.8775035765379113, + "grad_norm": 0.4435069430103818, + "learning_rate": 7.767961428667136e-06, + "loss": 1.662, + "step": 9814 + }, + { + "epoch": 0.8775929899856938, + "grad_norm": 0.6173627263986046, + "learning_rate": 7.756774355361884e-06, + "loss": 1.3882, + "step": 9815 + }, + { + "epoch": 0.8776824034334764, + "grad_norm": 0.5200297846974944, + "learning_rate": 7.745595018384578e-06, + "loss": 1.5246, + "step": 9816 + }, + { + "epoch": 0.8777718168812589, + "grad_norm": 0.8145345231770994, + "learning_rate": 7.734423418672786e-06, + "loss": 1.3571, + "step": 9817 + }, + { + "epoch": 0.8778612303290415, + "grad_norm": 0.43455347616595363, + "learning_rate": 7.723259557163487e-06, + "loss": 1.5946, + "step": 9818 + }, + { + "epoch": 0.8779506437768241, + "grad_norm": 0.664041157558289, + "learning_rate": 7.71210343479295e-06, + "loss": 1.7332, + "step": 9819 + }, + { + "epoch": 0.8780400572246065, + "grad_norm": 0.668877448622541, + "learning_rate": 7.70095505249685e-06, + "loss": 1.9062, + "step": 9820 + }, + { + "epoch": 0.8781294706723891, + "grad_norm": 0.3885477278977798, + "learning_rate": 7.689814411210195e-06, + "loss": 1.154, + "step": 9821 + }, + { + "epoch": 0.8782188841201717, + "grad_norm": 0.5776291024130501, + "learning_rate": 7.678681511867304e-06, + "loss": 1.5396, + "step": 9822 + }, + { + "epoch": 0.8783082975679543, + "grad_norm": 0.8751730419072263, + "learning_rate": 7.667556355401906e-06, + "loss": 1.6769, + "step": 9823 + }, + { + "epoch": 0.8783977110157367, + "grad_norm": 0.5010189818915564, + "learning_rate": 7.656438942747058e-06, + "loss": 2.0707, + "step": 9824 + }, + { + "epoch": 0.8784871244635193, + "grad_norm": 0.8909398670215658, + "learning_rate": 7.645329274835122e-06, + "loss": 1.7076, + "step": 9825 + }, + { + "epoch": 0.8785765379113019, + "grad_norm": 0.5986516539301178, + "learning_rate": 7.634227352597901e-06, + "loss": 1.8141, + "step": 9826 + }, + { + "epoch": 0.8786659513590844, + "grad_norm": 0.5506434935764659, + "learning_rate": 7.623133176966491e-06, + "loss": 1.5407, + "step": 9827 + }, + { + "epoch": 0.878755364806867, + "grad_norm": 0.6387163637690039, + "learning_rate": 7.612046748871327e-06, + "loss": 1.6303, + "step": 9828 + }, + { + "epoch": 0.8788447782546495, + "grad_norm": 0.5845429694143657, + "learning_rate": 7.600968069242232e-06, + "loss": 1.8296, + "step": 9829 + }, + { + "epoch": 0.878934191702432, + "grad_norm": 0.5734062364708509, + "learning_rate": 7.589897139008362e-06, + "loss": 1.696, + "step": 9830 + }, + { + "epoch": 0.8790236051502146, + "grad_norm": 0.9425099390572554, + "learning_rate": 7.578833959098209e-06, + "loss": 1.632, + "step": 9831 + }, + { + "epoch": 0.8791130185979972, + "grad_norm": 0.7327248826274723, + "learning_rate": 7.567778530439606e-06, + "loss": 1.5687, + "step": 9832 + }, + { + "epoch": 0.8792024320457796, + "grad_norm": 0.5319824796096894, + "learning_rate": 7.5567308539598256e-06, + "loss": 1.842, + "step": 9833 + }, + { + "epoch": 0.8792918454935622, + "grad_norm": 1.942842538593742, + "learning_rate": 7.545690930585381e-06, + "loss": 6.1118, + "step": 9834 + }, + { + "epoch": 0.8793812589413448, + "grad_norm": 0.4217254644847827, + "learning_rate": 7.534658761242164e-06, + "loss": 1.8416, + "step": 9835 + }, + { + "epoch": 0.8794706723891274, + "grad_norm": 0.33371647090042134, + "learning_rate": 7.52363434685548e-06, + "loss": 1.4488, + "step": 9836 + }, + { + "epoch": 0.8795600858369099, + "grad_norm": 0.4790375787391621, + "learning_rate": 7.512617688349866e-06, + "loss": 1.4282, + "step": 9837 + }, + { + "epoch": 0.8796494992846924, + "grad_norm": 0.49868489839598934, + "learning_rate": 7.501608786649328e-06, + "loss": 1.4202, + "step": 9838 + }, + { + "epoch": 0.879738912732475, + "grad_norm": 0.6316226312627164, + "learning_rate": 7.490607642677139e-06, + "loss": 1.7458, + "step": 9839 + }, + { + "epoch": 0.8798283261802575, + "grad_norm": 0.5605891436886701, + "learning_rate": 7.479614257355971e-06, + "loss": 1.6542, + "step": 9840 + }, + { + "epoch": 0.8799177396280401, + "grad_norm": 0.6390011277446342, + "learning_rate": 7.468628631607822e-06, + "loss": 1.697, + "step": 9841 + }, + { + "epoch": 0.8800071530758226, + "grad_norm": 0.5596079930897059, + "learning_rate": 7.45765076635404e-06, + "loss": 1.8712, + "step": 9842 + }, + { + "epoch": 0.8800965665236051, + "grad_norm": 0.46987832016724973, + "learning_rate": 7.446680662515315e-06, + "loss": 1.4235, + "step": 9843 + }, + { + "epoch": 0.8801859799713877, + "grad_norm": 0.7297501542198431, + "learning_rate": 7.435718321011731e-06, + "loss": 1.4203, + "step": 9844 + }, + { + "epoch": 0.8802753934191703, + "grad_norm": 0.8273975260480613, + "learning_rate": 7.424763742762642e-06, + "loss": 1.7606, + "step": 9845 + }, + { + "epoch": 0.8803648068669528, + "grad_norm": 0.6830058501673935, + "learning_rate": 7.41381692868679e-06, + "loss": 1.6688, + "step": 9846 + }, + { + "epoch": 0.8804542203147353, + "grad_norm": 0.46008285380675845, + "learning_rate": 7.402877879702341e-06, + "loss": 1.5975, + "step": 9847 + }, + { + "epoch": 0.8805436337625179, + "grad_norm": 0.6029593588615169, + "learning_rate": 7.391946596726673e-06, + "loss": 1.4404, + "step": 9848 + }, + { + "epoch": 0.8806330472103004, + "grad_norm": 0.48015635337949064, + "learning_rate": 7.381023080676608e-06, + "loss": 1.613, + "step": 9849 + }, + { + "epoch": 0.880722460658083, + "grad_norm": 0.6900886337486314, + "learning_rate": 7.3701073324682905e-06, + "loss": 1.4561, + "step": 9850 + }, + { + "epoch": 0.8808118741058655, + "grad_norm": 0.6434376814381219, + "learning_rate": 7.3591993530171984e-06, + "loss": 1.5514, + "step": 9851 + }, + { + "epoch": 0.880901287553648, + "grad_norm": 0.5307677617966673, + "learning_rate": 7.348299143238157e-06, + "loss": 1.5861, + "step": 9852 + }, + { + "epoch": 0.8809907010014306, + "grad_norm": 0.5753728738749739, + "learning_rate": 7.33740670404538e-06, + "loss": 1.5884, + "step": 9853 + }, + { + "epoch": 0.8810801144492132, + "grad_norm": 0.8725035302403845, + "learning_rate": 7.326522036352401e-06, + "loss": 1.4736, + "step": 9854 + }, + { + "epoch": 0.8811695278969958, + "grad_norm": 0.5431960986354878, + "learning_rate": 7.315645141072103e-06, + "loss": 1.7786, + "step": 9855 + }, + { + "epoch": 0.8812589413447782, + "grad_norm": 0.5699774577815483, + "learning_rate": 7.30477601911671e-06, + "loss": 1.7335, + "step": 9856 + }, + { + "epoch": 0.8813483547925608, + "grad_norm": 0.49639749160082264, + "learning_rate": 7.293914671397795e-06, + "loss": 1.2901, + "step": 9857 + }, + { + "epoch": 0.8814377682403434, + "grad_norm": 0.525605667614179, + "learning_rate": 7.283061098826294e-06, + "loss": 1.4979, + "step": 9858 + }, + { + "epoch": 0.8815271816881259, + "grad_norm": 0.7385198942248339, + "learning_rate": 7.272215302312502e-06, + "loss": 1.6311, + "step": 9859 + }, + { + "epoch": 0.8816165951359084, + "grad_norm": 0.5220377406246922, + "learning_rate": 7.261377282766002e-06, + "loss": 1.6523, + "step": 9860 + }, + { + "epoch": 0.881706008583691, + "grad_norm": 0.4115554730155303, + "learning_rate": 7.250547041095812e-06, + "loss": 1.4152, + "step": 9861 + }, + { + "epoch": 0.8817954220314735, + "grad_norm": 0.5951784550720544, + "learning_rate": 7.239724578210216e-06, + "loss": 1.4652, + "step": 9862 + }, + { + "epoch": 0.8818848354792561, + "grad_norm": 0.4971996343043114, + "learning_rate": 7.2289098950168995e-06, + "loss": 1.3917, + "step": 9863 + }, + { + "epoch": 0.8819742489270386, + "grad_norm": 0.623822234330484, + "learning_rate": 7.2181029924228814e-06, + "loss": 1.318, + "step": 9864 + }, + { + "epoch": 0.8820636623748211, + "grad_norm": 0.4903302723042842, + "learning_rate": 7.207303871334492e-06, + "loss": 1.656, + "step": 9865 + }, + { + "epoch": 0.8821530758226037, + "grad_norm": 0.719557819063778, + "learning_rate": 7.1965125326574735e-06, + "loss": 1.6263, + "step": 9866 + }, + { + "epoch": 0.8822424892703863, + "grad_norm": 0.5279569581157357, + "learning_rate": 7.185728977296857e-06, + "loss": 1.8308, + "step": 9867 + }, + { + "epoch": 0.8823319027181689, + "grad_norm": 0.7713294722058734, + "learning_rate": 7.174953206157064e-06, + "loss": 1.3551, + "step": 9868 + }, + { + "epoch": 0.8824213161659513, + "grad_norm": 0.6606469752071129, + "learning_rate": 7.16418522014185e-06, + "loss": 1.6737, + "step": 9869 + }, + { + "epoch": 0.8825107296137339, + "grad_norm": 0.6043848047064604, + "learning_rate": 7.153425020154314e-06, + "loss": 1.6974, + "step": 9870 + }, + { + "epoch": 0.8826001430615165, + "grad_norm": 0.4496281861754786, + "learning_rate": 7.142672607096878e-06, + "loss": 1.5188, + "step": 9871 + }, + { + "epoch": 0.882689556509299, + "grad_norm": 0.35421449186524817, + "learning_rate": 7.1319279818713445e-06, + "loss": 1.328, + "step": 9872 + }, + { + "epoch": 0.8827789699570815, + "grad_norm": 0.4916233350983188, + "learning_rate": 7.121191145378858e-06, + "loss": 1.9102, + "step": 9873 + }, + { + "epoch": 0.8828683834048641, + "grad_norm": 0.5860017829862226, + "learning_rate": 7.110462098519899e-06, + "loss": 1.5995, + "step": 9874 + }, + { + "epoch": 0.8829577968526466, + "grad_norm": 0.4081846975351659, + "learning_rate": 7.099740842194313e-06, + "loss": 1.6472, + "step": 9875 + }, + { + "epoch": 0.8830472103004292, + "grad_norm": 0.5139318919309686, + "learning_rate": 7.08902737730125e-06, + "loss": 1.7356, + "step": 9876 + }, + { + "epoch": 0.8831366237482118, + "grad_norm": 1.7974319909376808, + "learning_rate": 7.078321704739266e-06, + "loss": 5.9993, + "step": 9877 + }, + { + "epoch": 0.8832260371959942, + "grad_norm": 0.9632082277270282, + "learning_rate": 7.067623825406222e-06, + "loss": 1.6983, + "step": 9878 + }, + { + "epoch": 0.8833154506437768, + "grad_norm": 0.7280470838660438, + "learning_rate": 7.056933740199323e-06, + "loss": 1.5566, + "step": 9879 + }, + { + "epoch": 0.8834048640915594, + "grad_norm": 1.0346770632362876, + "learning_rate": 7.0462514500151285e-06, + "loss": 1.9191, + "step": 9880 + }, + { + "epoch": 0.883494277539342, + "grad_norm": 0.509027625489463, + "learning_rate": 7.035576955749601e-06, + "loss": 1.4204, + "step": 9881 + }, + { + "epoch": 0.8835836909871244, + "grad_norm": 0.5686475118278774, + "learning_rate": 7.0249102582979455e-06, + "loss": 1.5654, + "step": 9882 + }, + { + "epoch": 0.883673104434907, + "grad_norm": 0.642641287070679, + "learning_rate": 7.01425135855478e-06, + "loss": 1.7778, + "step": 9883 + }, + { + "epoch": 0.8837625178826896, + "grad_norm": 0.5185779648569937, + "learning_rate": 7.003600257414067e-06, + "loss": 1.5016, + "step": 9884 + }, + { + "epoch": 0.8838519313304721, + "grad_norm": 0.5746935761474471, + "learning_rate": 6.99295695576907e-06, + "loss": 1.6498, + "step": 9885 + }, + { + "epoch": 0.8839413447782547, + "grad_norm": 0.6117569685063047, + "learning_rate": 6.9823214545124525e-06, + "loss": 1.9618, + "step": 9886 + }, + { + "epoch": 0.8840307582260372, + "grad_norm": 0.7171372101756917, + "learning_rate": 6.971693754536201e-06, + "loss": 1.5593, + "step": 9887 + }, + { + "epoch": 0.8841201716738197, + "grad_norm": 1.2865199554566988, + "learning_rate": 6.961073856731648e-06, + "loss": 1.4458, + "step": 9888 + }, + { + "epoch": 0.8842095851216023, + "grad_norm": 0.6354208340066292, + "learning_rate": 6.950461761989458e-06, + "loss": 1.5433, + "step": 9889 + }, + { + "epoch": 0.8842989985693849, + "grad_norm": 0.5991992241158873, + "learning_rate": 6.9398574711996844e-06, + "loss": 1.5327, + "step": 9890 + }, + { + "epoch": 0.8843884120171673, + "grad_norm": 0.5499784088859698, + "learning_rate": 6.929260985251662e-06, + "loss": 1.429, + "step": 9891 + }, + { + "epoch": 0.8844778254649499, + "grad_norm": 0.9559758830310554, + "learning_rate": 6.918672305034124e-06, + "loss": 1.9671, + "step": 9892 + }, + { + "epoch": 0.8845672389127325, + "grad_norm": 0.8116396384183957, + "learning_rate": 6.908091431435138e-06, + "loss": 1.8442, + "step": 9893 + }, + { + "epoch": 0.884656652360515, + "grad_norm": 0.5035856041047627, + "learning_rate": 6.897518365342059e-06, + "loss": 1.7074, + "step": 9894 + }, + { + "epoch": 0.8847460658082976, + "grad_norm": 0.47826376115667757, + "learning_rate": 6.8869531076417136e-06, + "loss": 1.4631, + "step": 9895 + }, + { + "epoch": 0.8848354792560801, + "grad_norm": 0.7171848827663813, + "learning_rate": 6.876395659220148e-06, + "loss": 1.5465, + "step": 9896 + }, + { + "epoch": 0.8849248927038627, + "grad_norm": 0.6714594356599249, + "learning_rate": 6.865846020962807e-06, + "loss": 1.6353, + "step": 9897 + }, + { + "epoch": 0.8850143061516452, + "grad_norm": 0.4356046392399832, + "learning_rate": 6.855304193754497e-06, + "loss": 1.6581, + "step": 9898 + }, + { + "epoch": 0.8851037195994278, + "grad_norm": 0.7126100764149857, + "learning_rate": 6.844770178479321e-06, + "loss": 1.492, + "step": 9899 + }, + { + "epoch": 0.8851931330472103, + "grad_norm": 0.8435416827661643, + "learning_rate": 6.834243976020771e-06, + "loss": 1.62, + "step": 9900 + }, + { + "epoch": 0.8852825464949928, + "grad_norm": 0.572378897993924, + "learning_rate": 6.823725587261654e-06, + "loss": 1.6695, + "step": 9901 + }, + { + "epoch": 0.8853719599427754, + "grad_norm": 0.4846529110268892, + "learning_rate": 6.813215013084151e-06, + "loss": 1.4881, + "step": 9902 + }, + { + "epoch": 0.885461373390558, + "grad_norm": 0.3857348962866903, + "learning_rate": 6.8027122543697586e-06, + "loss": 1.5319, + "step": 9903 + }, + { + "epoch": 0.8855507868383404, + "grad_norm": 0.41794686949807675, + "learning_rate": 6.7922173119993606e-06, + "loss": 1.7152, + "step": 9904 + }, + { + "epoch": 0.885640200286123, + "grad_norm": 0.5081424658739572, + "learning_rate": 6.781730186853108e-06, + "loss": 1.6166, + "step": 9905 + }, + { + "epoch": 0.8857296137339056, + "grad_norm": 0.5874065709164795, + "learning_rate": 6.771250879810565e-06, + "loss": 1.5918, + "step": 9906 + }, + { + "epoch": 0.8858190271816881, + "grad_norm": 0.4569088061035808, + "learning_rate": 6.760779391750627e-06, + "loss": 1.6905, + "step": 9907 + }, + { + "epoch": 0.8859084406294707, + "grad_norm": 0.4415960212802462, + "learning_rate": 6.750315723551492e-06, + "loss": 1.5546, + "step": 9908 + }, + { + "epoch": 0.8859978540772532, + "grad_norm": 0.27484798921975206, + "learning_rate": 6.739859876090793e-06, + "loss": 1.1699, + "step": 9909 + }, + { + "epoch": 0.8860872675250357, + "grad_norm": 0.5153759934307718, + "learning_rate": 6.729411850245404e-06, + "loss": 1.4621, + "step": 9910 + }, + { + "epoch": 0.8861766809728183, + "grad_norm": 0.34330829787152595, + "learning_rate": 6.718971646891603e-06, + "loss": 1.2733, + "step": 9911 + }, + { + "epoch": 0.8862660944206009, + "grad_norm": 0.3760328482165097, + "learning_rate": 6.708539266905001e-06, + "loss": 1.3759, + "step": 9912 + }, + { + "epoch": 0.8863555078683834, + "grad_norm": 0.585695245044587, + "learning_rate": 6.6981147111605305e-06, + "loss": 1.4531, + "step": 9913 + }, + { + "epoch": 0.8864449213161659, + "grad_norm": 0.36834608992561363, + "learning_rate": 6.687697980532504e-06, + "loss": 1.3327, + "step": 9914 + }, + { + "epoch": 0.8865343347639485, + "grad_norm": 0.3093763676077789, + "learning_rate": 6.677289075894544e-06, + "loss": 1.286, + "step": 9915 + }, + { + "epoch": 0.8866237482117311, + "grad_norm": 0.6890480153479798, + "learning_rate": 6.666887998119653e-06, + "loss": 1.9618, + "step": 9916 + }, + { + "epoch": 0.8867131616595136, + "grad_norm": 0.4657143670385415, + "learning_rate": 6.656494748080144e-06, + "loss": 1.6041, + "step": 9917 + }, + { + "epoch": 0.8868025751072961, + "grad_norm": 0.5207724700566881, + "learning_rate": 6.646109326647709e-06, + "loss": 1.8536, + "step": 9918 + }, + { + "epoch": 0.8868919885550787, + "grad_norm": 0.48937856826287146, + "learning_rate": 6.635731734693329e-06, + "loss": 1.3396, + "step": 9919 + }, + { + "epoch": 0.8869814020028612, + "grad_norm": 0.3767218049787446, + "learning_rate": 6.625361973087363e-06, + "loss": 1.4003, + "step": 9920 + }, + { + "epoch": 0.8870708154506438, + "grad_norm": 0.5961211577758566, + "learning_rate": 6.6150000426995486e-06, + "loss": 1.34, + "step": 9921 + }, + { + "epoch": 0.8871602288984263, + "grad_norm": 0.6019475004605496, + "learning_rate": 6.604645944398858e-06, + "loss": 1.8494, + "step": 9922 + }, + { + "epoch": 0.8872496423462088, + "grad_norm": 0.5056402311698089, + "learning_rate": 6.594299679053739e-06, + "loss": 1.4132, + "step": 9923 + }, + { + "epoch": 0.8873390557939914, + "grad_norm": 0.6008868033626555, + "learning_rate": 6.583961247531911e-06, + "loss": 1.639, + "step": 9924 + }, + { + "epoch": 0.887428469241774, + "grad_norm": 0.5939851010096602, + "learning_rate": 6.573630650700424e-06, + "loss": 1.4643, + "step": 9925 + }, + { + "epoch": 0.8875178826895566, + "grad_norm": 0.5810345745004649, + "learning_rate": 6.563307889425707e-06, + "loss": 1.7124, + "step": 9926 + }, + { + "epoch": 0.887607296137339, + "grad_norm": 0.45483530383191567, + "learning_rate": 6.5529929645735235e-06, + "loss": 1.7866, + "step": 9927 + }, + { + "epoch": 0.8876967095851216, + "grad_norm": 0.6337643539029449, + "learning_rate": 6.542685877008959e-06, + "loss": 1.8213, + "step": 9928 + }, + { + "epoch": 0.8877861230329042, + "grad_norm": 0.6331272417701209, + "learning_rate": 6.532386627596454e-06, + "loss": 1.8027, + "step": 9929 + }, + { + "epoch": 0.8878755364806867, + "grad_norm": 0.40612232042970553, + "learning_rate": 6.522095217199797e-06, + "loss": 1.6068, + "step": 9930 + }, + { + "epoch": 0.8879649499284692, + "grad_norm": 0.47805352317896227, + "learning_rate": 6.511811646682131e-06, + "loss": 1.6903, + "step": 9931 + }, + { + "epoch": 0.8880543633762518, + "grad_norm": 0.707920465618905, + "learning_rate": 6.501535916905932e-06, + "loss": 1.6408, + "step": 9932 + }, + { + "epoch": 0.8881437768240343, + "grad_norm": 0.6318273669356677, + "learning_rate": 6.491268028732977e-06, + "loss": 1.6858, + "step": 9933 + }, + { + "epoch": 0.8882331902718169, + "grad_norm": 0.4747681491693669, + "learning_rate": 6.4810079830244455e-06, + "loss": 1.3854, + "step": 9934 + }, + { + "epoch": 0.8883226037195995, + "grad_norm": 0.5045213916217248, + "learning_rate": 6.470755780640847e-06, + "loss": 1.9052, + "step": 9935 + }, + { + "epoch": 0.8884120171673819, + "grad_norm": 0.4369306018719156, + "learning_rate": 6.460511422441984e-06, + "loss": 1.4835, + "step": 9936 + }, + { + "epoch": 0.8885014306151645, + "grad_norm": 0.4959547040843057, + "learning_rate": 6.450274909287068e-06, + "loss": 1.7015, + "step": 9937 + }, + { + "epoch": 0.8885908440629471, + "grad_norm": 0.7602932835948052, + "learning_rate": 6.440046242034625e-06, + "loss": 1.7118, + "step": 9938 + }, + { + "epoch": 0.8886802575107297, + "grad_norm": 0.5569236453195332, + "learning_rate": 6.429825421542512e-06, + "loss": 1.6005, + "step": 9939 + }, + { + "epoch": 0.8887696709585121, + "grad_norm": 0.5549570229055559, + "learning_rate": 6.4196124486679225e-06, + "loss": 1.8808, + "step": 9940 + }, + { + "epoch": 0.8888590844062947, + "grad_norm": 0.5235234532120144, + "learning_rate": 6.409407324267447e-06, + "loss": 1.6774, + "step": 9941 + }, + { + "epoch": 0.8889484978540773, + "grad_norm": 0.6471936163366855, + "learning_rate": 6.399210049196924e-06, + "loss": 1.3374, + "step": 9942 + }, + { + "epoch": 0.8890379113018598, + "grad_norm": 0.32831920687816146, + "learning_rate": 6.3890206243116255e-06, + "loss": 1.298, + "step": 9943 + }, + { + "epoch": 0.8891273247496424, + "grad_norm": 0.6446233481690445, + "learning_rate": 6.378839050466101e-06, + "loss": 1.5933, + "step": 9944 + }, + { + "epoch": 0.8892167381974249, + "grad_norm": 0.5135454409790168, + "learning_rate": 6.36866532851429e-06, + "loss": 1.3204, + "step": 9945 + }, + { + "epoch": 0.8893061516452074, + "grad_norm": 0.6480743035124573, + "learning_rate": 6.3584994593094305e-06, + "loss": 2.1032, + "step": 9946 + }, + { + "epoch": 0.88939556509299, + "grad_norm": 0.5608739321356325, + "learning_rate": 6.348341443704153e-06, + "loss": 1.7075, + "step": 9947 + }, + { + "epoch": 0.8894849785407726, + "grad_norm": 0.43915707183833724, + "learning_rate": 6.338191282550354e-06, + "loss": 1.7203, + "step": 9948 + }, + { + "epoch": 0.889574391988555, + "grad_norm": 0.5497653643545887, + "learning_rate": 6.328048976699352e-06, + "loss": 1.664, + "step": 9949 + }, + { + "epoch": 0.8896638054363376, + "grad_norm": 0.7557557922327903, + "learning_rate": 6.317914527001745e-06, + "loss": 1.4823, + "step": 9950 + }, + { + "epoch": 0.8897532188841202, + "grad_norm": 0.5615581835695145, + "learning_rate": 6.307787934307507e-06, + "loss": 1.9256, + "step": 9951 + }, + { + "epoch": 0.8898426323319027, + "grad_norm": 0.5291477725491129, + "learning_rate": 6.297669199465961e-06, + "loss": 1.8505, + "step": 9952 + }, + { + "epoch": 0.8899320457796852, + "grad_norm": 0.4069514276496185, + "learning_rate": 6.287558323325715e-06, + "loss": 1.5465, + "step": 9953 + }, + { + "epoch": 0.8900214592274678, + "grad_norm": 0.5512917447834261, + "learning_rate": 6.277455306734781e-06, + "loss": 1.6695, + "step": 9954 + }, + { + "epoch": 0.8901108726752504, + "grad_norm": 0.5457199016772145, + "learning_rate": 6.267360150540491e-06, + "loss": 1.8067, + "step": 9955 + }, + { + "epoch": 0.8902002861230329, + "grad_norm": 0.40654913177148605, + "learning_rate": 6.2572728555894796e-06, + "loss": 1.487, + "step": 9956 + }, + { + "epoch": 0.8902896995708155, + "grad_norm": 0.47629592801333753, + "learning_rate": 6.247193422727804e-06, + "loss": 1.3923, + "step": 9957 + }, + { + "epoch": 0.890379113018598, + "grad_norm": 0.35483387473614886, + "learning_rate": 6.237121852800798e-06, + "loss": 1.5526, + "step": 9958 + }, + { + "epoch": 0.8904685264663805, + "grad_norm": 0.49201451937379165, + "learning_rate": 6.227058146653131e-06, + "loss": 1.5031, + "step": 9959 + }, + { + "epoch": 0.8905579399141631, + "grad_norm": 0.8511005174657539, + "learning_rate": 6.217002305128849e-06, + "loss": 1.5484, + "step": 9960 + }, + { + "epoch": 0.8906473533619457, + "grad_norm": 0.6586763921081072, + "learning_rate": 6.206954329071335e-06, + "loss": 1.8553, + "step": 9961 + }, + { + "epoch": 0.8907367668097281, + "grad_norm": 0.590035729410713, + "learning_rate": 6.19691421932328e-06, + "loss": 1.2531, + "step": 9962 + }, + { + "epoch": 0.8908261802575107, + "grad_norm": 0.7549751898912404, + "learning_rate": 6.186881976726733e-06, + "loss": 1.6089, + "step": 9963 + }, + { + "epoch": 0.8909155937052933, + "grad_norm": 0.5530257476619276, + "learning_rate": 6.17685760212311e-06, + "loss": 1.8925, + "step": 9964 + }, + { + "epoch": 0.8910050071530758, + "grad_norm": 0.388307675606156, + "learning_rate": 6.166841096353126e-06, + "loss": 1.4322, + "step": 9965 + }, + { + "epoch": 0.8910944206008584, + "grad_norm": 0.42291029294564864, + "learning_rate": 6.1568324602568675e-06, + "loss": 1.4667, + "step": 9966 + }, + { + "epoch": 0.8911838340486409, + "grad_norm": 0.3623268102239769, + "learning_rate": 6.146831694673727e-06, + "loss": 1.3331, + "step": 9967 + }, + { + "epoch": 0.8912732474964234, + "grad_norm": 0.46064538980643344, + "learning_rate": 6.136838800442457e-06, + "loss": 1.6087, + "step": 9968 + }, + { + "epoch": 0.891362660944206, + "grad_norm": 0.5334389439053085, + "learning_rate": 6.126853778401187e-06, + "loss": 1.3792, + "step": 9969 + }, + { + "epoch": 0.8914520743919886, + "grad_norm": 0.49497928986453216, + "learning_rate": 6.11687662938728e-06, + "loss": 1.3969, + "step": 9970 + }, + { + "epoch": 0.891541487839771, + "grad_norm": 0.7458275631256034, + "learning_rate": 6.1069073542375675e-06, + "loss": 1.5786, + "step": 9971 + }, + { + "epoch": 0.8916309012875536, + "grad_norm": 0.5318559016643309, + "learning_rate": 6.0969459537881575e-06, + "loss": 1.2886, + "step": 9972 + }, + { + "epoch": 0.8917203147353362, + "grad_norm": 0.4837855988466403, + "learning_rate": 6.086992428874472e-06, + "loss": 1.6705, + "step": 9973 + }, + { + "epoch": 0.8918097281831188, + "grad_norm": 0.7927317200299222, + "learning_rate": 6.077046780331308e-06, + "loss": 1.7819, + "step": 9974 + }, + { + "epoch": 0.8918991416309013, + "grad_norm": 0.42673052228368236, + "learning_rate": 6.06710900899281e-06, + "loss": 1.6099, + "step": 9975 + }, + { + "epoch": 0.8919885550786838, + "grad_norm": 0.5773416666037321, + "learning_rate": 6.057179115692435e-06, + "loss": 1.8244, + "step": 9976 + }, + { + "epoch": 0.8920779685264664, + "grad_norm": 0.5326497396459211, + "learning_rate": 6.047257101262982e-06, + "loss": 1.7086, + "step": 9977 + }, + { + "epoch": 0.8921673819742489, + "grad_norm": 0.43175844418121223, + "learning_rate": 6.037342966536619e-06, + "loss": 1.6784, + "step": 9978 + }, + { + "epoch": 0.8922567954220315, + "grad_norm": 0.336981911347863, + "learning_rate": 6.027436712344814e-06, + "loss": 1.5124, + "step": 9979 + }, + { + "epoch": 0.892346208869814, + "grad_norm": 0.6348288068374224, + "learning_rate": 6.017538339518403e-06, + "loss": 1.7, + "step": 9980 + }, + { + "epoch": 0.8924356223175965, + "grad_norm": 0.4158234269843196, + "learning_rate": 6.007647848887565e-06, + "loss": 1.5022, + "step": 9981 + }, + { + "epoch": 0.8925250357653791, + "grad_norm": 0.5322585828515742, + "learning_rate": 5.997765241281783e-06, + "loss": 1.2976, + "step": 9982 + }, + { + "epoch": 0.8926144492131617, + "grad_norm": 1.01462304809976, + "learning_rate": 5.987890517529893e-06, + "loss": 1.9101, + "step": 9983 + }, + { + "epoch": 0.8927038626609443, + "grad_norm": 0.796493028362036, + "learning_rate": 5.978023678460099e-06, + "loss": 1.4899, + "step": 9984 + }, + { + "epoch": 0.8927932761087267, + "grad_norm": 0.45531358670419614, + "learning_rate": 5.968164724899894e-06, + "loss": 1.4601, + "step": 9985 + }, + { + "epoch": 0.8928826895565093, + "grad_norm": 0.5023421289354594, + "learning_rate": 5.958313657676173e-06, + "loss": 1.6147, + "step": 9986 + }, + { + "epoch": 0.8929721030042919, + "grad_norm": 0.8364414128821215, + "learning_rate": 5.948470477615098e-06, + "loss": 1.5245, + "step": 9987 + }, + { + "epoch": 0.8930615164520744, + "grad_norm": 0.6454814817978647, + "learning_rate": 5.938635185542218e-06, + "loss": 1.3498, + "step": 9988 + }, + { + "epoch": 0.8931509298998569, + "grad_norm": 1.1132576984980431, + "learning_rate": 5.928807782282431e-06, + "loss": 1.657, + "step": 9989 + }, + { + "epoch": 0.8932403433476395, + "grad_norm": 0.47987088777678044, + "learning_rate": 5.918988268659898e-06, + "loss": 1.844, + "step": 9990 + }, + { + "epoch": 0.893329756795422, + "grad_norm": 0.5165061225677293, + "learning_rate": 5.909176645498193e-06, + "loss": 1.3484, + "step": 9991 + }, + { + "epoch": 0.8934191702432046, + "grad_norm": 0.5795463322857448, + "learning_rate": 5.899372913620238e-06, + "loss": 1.8547, + "step": 9992 + }, + { + "epoch": 0.8935085836909872, + "grad_norm": 0.4684367017889553, + "learning_rate": 5.889577073848207e-06, + "loss": 1.8525, + "step": 9993 + }, + { + "epoch": 0.8935979971387696, + "grad_norm": 0.6313601710224458, + "learning_rate": 5.879789127003699e-06, + "loss": 1.6674, + "step": 9994 + }, + { + "epoch": 0.8936874105865522, + "grad_norm": 0.7115888225894847, + "learning_rate": 5.870009073907623e-06, + "loss": 1.4068, + "step": 9995 + }, + { + "epoch": 0.8937768240343348, + "grad_norm": 0.49657260886555965, + "learning_rate": 5.86023691538019e-06, + "loss": 1.6562, + "step": 9996 + }, + { + "epoch": 0.8938662374821174, + "grad_norm": 0.6650935616082988, + "learning_rate": 5.850472652240991e-06, + "loss": 1.7652, + "step": 9997 + }, + { + "epoch": 0.8939556509298998, + "grad_norm": 0.36762707975467535, + "learning_rate": 5.840716285308956e-06, + "loss": 1.4543, + "step": 9998 + }, + { + "epoch": 0.8940450643776824, + "grad_norm": 0.42805924975570137, + "learning_rate": 5.8309678154023216e-06, + "loss": 1.5146, + "step": 9999 + }, + { + "epoch": 0.894134477825465, + "grad_norm": 1.0371015403835335, + "learning_rate": 5.821227243338712e-06, + "loss": 1.74, + "step": 10000 + }, + { + "epoch": 0.8942238912732475, + "grad_norm": 0.6958980301161797, + "learning_rate": 5.811494569935016e-06, + "loss": 1.8585, + "step": 10001 + }, + { + "epoch": 0.89431330472103, + "grad_norm": 0.5333527653456183, + "learning_rate": 5.801769796007517e-06, + "loss": 1.7871, + "step": 10002 + }, + { + "epoch": 0.8944027181688126, + "grad_norm": 0.5578484468605981, + "learning_rate": 5.792052922371826e-06, + "loss": 1.8325, + "step": 10003 + }, + { + "epoch": 0.8944921316165951, + "grad_norm": 0.6175511716238139, + "learning_rate": 5.782343949842894e-06, + "loss": 1.691, + "step": 10004 + }, + { + "epoch": 0.8945815450643777, + "grad_norm": 0.5732634222573693, + "learning_rate": 5.7726428792349574e-06, + "loss": 1.6759, + "step": 10005 + }, + { + "epoch": 0.8946709585121603, + "grad_norm": 0.7831571300755391, + "learning_rate": 5.762949711361698e-06, + "loss": 1.5664, + "step": 10006 + }, + { + "epoch": 0.8947603719599427, + "grad_norm": 0.8447482117572988, + "learning_rate": 5.753264447036022e-06, + "loss": 1.7405, + "step": 10007 + }, + { + "epoch": 0.8948497854077253, + "grad_norm": 0.5386019337706867, + "learning_rate": 5.743587087070235e-06, + "loss": 1.6288, + "step": 10008 + }, + { + "epoch": 0.8949391988555079, + "grad_norm": 0.7382762630551234, + "learning_rate": 5.733917632275976e-06, + "loss": 1.6546, + "step": 10009 + }, + { + "epoch": 0.8950286123032904, + "grad_norm": 0.4791774156878048, + "learning_rate": 5.7242560834641855e-06, + "loss": 1.5696, + "step": 10010 + }, + { + "epoch": 0.8951180257510729, + "grad_norm": 0.3666892231367613, + "learning_rate": 5.714602441445194e-06, + "loss": 1.1941, + "step": 10011 + }, + { + "epoch": 0.8952074391988555, + "grad_norm": 0.5763912476875577, + "learning_rate": 5.704956707028619e-06, + "loss": 1.5424, + "step": 10012 + }, + { + "epoch": 0.895296852646638, + "grad_norm": 0.664526158254151, + "learning_rate": 5.695318881023437e-06, + "loss": 1.4694, + "step": 10013 + }, + { + "epoch": 0.8953862660944206, + "grad_norm": 0.7962618130571713, + "learning_rate": 5.685688964237979e-06, + "loss": 1.5965, + "step": 10014 + }, + { + "epoch": 0.8954756795422032, + "grad_norm": 0.5285713884992388, + "learning_rate": 5.676066957479898e-06, + "loss": 1.3815, + "step": 10015 + }, + { + "epoch": 0.8955650929899857, + "grad_norm": 0.4588182486410199, + "learning_rate": 5.66645286155616e-06, + "loss": 1.6875, + "step": 10016 + }, + { + "epoch": 0.8956545064377682, + "grad_norm": 0.5493136555063179, + "learning_rate": 5.656846677273086e-06, + "loss": 1.4855, + "step": 10017 + }, + { + "epoch": 0.8957439198855508, + "grad_norm": 1.7156822220917758, + "learning_rate": 5.647248405436356e-06, + "loss": 6.1952, + "step": 10018 + }, + { + "epoch": 0.8958333333333334, + "grad_norm": 0.42019039281232995, + "learning_rate": 5.637658046850924e-06, + "loss": 1.3989, + "step": 10019 + }, + { + "epoch": 0.8959227467811158, + "grad_norm": 0.4677979067437049, + "learning_rate": 5.628075602321181e-06, + "loss": 1.522, + "step": 10020 + }, + { + "epoch": 0.8960121602288984, + "grad_norm": 0.7579232199922867, + "learning_rate": 5.618501072650761e-06, + "loss": 1.5327, + "step": 10021 + }, + { + "epoch": 0.896101573676681, + "grad_norm": 0.5590126079550679, + "learning_rate": 5.608934458642656e-06, + "loss": 1.3041, + "step": 10022 + }, + { + "epoch": 0.8961909871244635, + "grad_norm": 1.0853343202031485, + "learning_rate": 5.599375761099246e-06, + "loss": 1.6826, + "step": 10023 + }, + { + "epoch": 0.8962804005722461, + "grad_norm": 0.5576865684387482, + "learning_rate": 5.589824980822167e-06, + "loss": 1.3421, + "step": 10024 + }, + { + "epoch": 0.8963698140200286, + "grad_norm": 0.3790991160496642, + "learning_rate": 5.580282118612446e-06, + "loss": 1.7164, + "step": 10025 + }, + { + "epoch": 0.8964592274678111, + "grad_norm": 0.6695736617046792, + "learning_rate": 5.570747175270441e-06, + "loss": 1.9767, + "step": 10026 + }, + { + "epoch": 0.8965486409155937, + "grad_norm": 0.732046524184887, + "learning_rate": 5.561220151595825e-06, + "loss": 1.5096, + "step": 10027 + }, + { + "epoch": 0.8966380543633763, + "grad_norm": 0.5711731271884196, + "learning_rate": 5.551701048387614e-06, + "loss": 1.5203, + "step": 10028 + }, + { + "epoch": 0.8967274678111588, + "grad_norm": 0.8177718780111697, + "learning_rate": 5.542189866444203e-06, + "loss": 1.6498, + "step": 10029 + }, + { + "epoch": 0.8968168812589413, + "grad_norm": 0.8611701056756847, + "learning_rate": 5.53268660656322e-06, + "loss": 6.2129, + "step": 10030 + }, + { + "epoch": 0.8969062947067239, + "grad_norm": 0.4342051147994749, + "learning_rate": 5.523191269541728e-06, + "loss": 1.5356, + "step": 10031 + }, + { + "epoch": 0.8969957081545065, + "grad_norm": 0.4899959944570955, + "learning_rate": 5.5137038561761115e-06, + "loss": 1.4024, + "step": 10032 + }, + { + "epoch": 0.897085121602289, + "grad_norm": 0.5104797718757593, + "learning_rate": 5.5042243672620006e-06, + "loss": 1.5378, + "step": 10033 + }, + { + "epoch": 0.8971745350500715, + "grad_norm": 0.5735638900032362, + "learning_rate": 5.494752803594505e-06, + "loss": 1.5643, + "step": 10034 + }, + { + "epoch": 0.8972639484978541, + "grad_norm": 1.0332985687031382, + "learning_rate": 5.485289165967933e-06, + "loss": 1.8692, + "step": 10035 + }, + { + "epoch": 0.8973533619456366, + "grad_norm": 0.40060170693717395, + "learning_rate": 5.475833455176027e-06, + "loss": 1.3032, + "step": 10036 + }, + { + "epoch": 0.8974427753934192, + "grad_norm": 0.7223483554009752, + "learning_rate": 5.466385672011809e-06, + "loss": 1.5517, + "step": 10037 + }, + { + "epoch": 0.8975321888412017, + "grad_norm": 0.6829596613517463, + "learning_rate": 5.4569458172676665e-06, + "loss": 1.9639, + "step": 10038 + }, + { + "epoch": 0.8976216022889842, + "grad_norm": 0.8048398862444384, + "learning_rate": 5.4475138917352894e-06, + "loss": 2.0008, + "step": 10039 + }, + { + "epoch": 0.8977110157367668, + "grad_norm": 0.6253962504891418, + "learning_rate": 5.4380898962057336e-06, + "loss": 1.4596, + "step": 10040 + }, + { + "epoch": 0.8978004291845494, + "grad_norm": 0.6314688589706414, + "learning_rate": 5.428673831469366e-06, + "loss": 1.486, + "step": 10041 + }, + { + "epoch": 0.897889842632332, + "grad_norm": 0.48352964341847043, + "learning_rate": 5.419265698315923e-06, + "loss": 1.7031, + "step": 10042 + }, + { + "epoch": 0.8979792560801144, + "grad_norm": 0.44151802522365646, + "learning_rate": 5.40986549753445e-06, + "loss": 1.8299, + "step": 10043 + }, + { + "epoch": 0.898068669527897, + "grad_norm": 0.4038842358426235, + "learning_rate": 5.400473229913305e-06, + "loss": 1.525, + "step": 10044 + }, + { + "epoch": 0.8981580829756796, + "grad_norm": 0.43702769482894177, + "learning_rate": 5.3910888962402265e-06, + "loss": 1.7164, + "step": 10045 + }, + { + "epoch": 0.8982474964234621, + "grad_norm": 0.4188580438701568, + "learning_rate": 5.381712497302261e-06, + "loss": 1.6679, + "step": 10046 + }, + { + "epoch": 0.8983369098712446, + "grad_norm": 0.43444345252961947, + "learning_rate": 5.372344033885801e-06, + "loss": 1.4912, + "step": 10047 + }, + { + "epoch": 0.8984263233190272, + "grad_norm": 0.628321466177299, + "learning_rate": 5.362983506776564e-06, + "loss": 1.3783, + "step": 10048 + }, + { + "epoch": 0.8985157367668097, + "grad_norm": 0.7042107264077557, + "learning_rate": 5.353630916759622e-06, + "loss": 2.0021, + "step": 10049 + }, + { + "epoch": 0.8986051502145923, + "grad_norm": 0.481746028738965, + "learning_rate": 5.344286264619347e-06, + "loss": 1.5718, + "step": 10050 + }, + { + "epoch": 0.8986945636623748, + "grad_norm": 0.5344726068432343, + "learning_rate": 5.334949551139457e-06, + "loss": 1.6011, + "step": 10051 + }, + { + "epoch": 0.8987839771101573, + "grad_norm": 1.3171491171691503, + "learning_rate": 5.325620777103035e-06, + "loss": 6.055, + "step": 10052 + }, + { + "epoch": 0.8988733905579399, + "grad_norm": 0.5723040928929711, + "learning_rate": 5.316299943292435e-06, + "loss": 1.8997, + "step": 10053 + }, + { + "epoch": 0.8989628040057225, + "grad_norm": 0.5171820234781292, + "learning_rate": 5.306987050489442e-06, + "loss": 1.3139, + "step": 10054 + }, + { + "epoch": 0.899052217453505, + "grad_norm": 0.69920057537903, + "learning_rate": 5.297682099475066e-06, + "loss": 1.6887, + "step": 10055 + }, + { + "epoch": 0.8991416309012875, + "grad_norm": 0.5313723760651959, + "learning_rate": 5.2883850910297235e-06, + "loss": 1.6466, + "step": 10056 + }, + { + "epoch": 0.8992310443490701, + "grad_norm": 0.5622511889981967, + "learning_rate": 5.27909602593315e-06, + "loss": 1.7583, + "step": 10057 + }, + { + "epoch": 0.8993204577968527, + "grad_norm": 0.5054945518010772, + "learning_rate": 5.2698149049643874e-06, + "loss": 1.5873, + "step": 10058 + }, + { + "epoch": 0.8994098712446352, + "grad_norm": 0.5040636236731952, + "learning_rate": 5.260541728901847e-06, + "loss": 1.5425, + "step": 10059 + }, + { + "epoch": 0.8994992846924177, + "grad_norm": 0.43330050730795994, + "learning_rate": 5.25127649852325e-06, + "loss": 1.5488, + "step": 10060 + }, + { + "epoch": 0.8995886981402003, + "grad_norm": 0.4561469045838878, + "learning_rate": 5.2420192146056645e-06, + "loss": 1.6452, + "step": 10061 + }, + { + "epoch": 0.8996781115879828, + "grad_norm": 0.5951688522697675, + "learning_rate": 5.232769877925503e-06, + "loss": 1.8011, + "step": 10062 + }, + { + "epoch": 0.8997675250357654, + "grad_norm": 0.6684486711668859, + "learning_rate": 5.2235284892584776e-06, + "loss": 1.6026, + "step": 10063 + }, + { + "epoch": 0.899856938483548, + "grad_norm": 0.38549861251847045, + "learning_rate": 5.214295049379658e-06, + "loss": 1.6144, + "step": 10064 + }, + { + "epoch": 0.8999463519313304, + "grad_norm": 0.6116717654072167, + "learning_rate": 5.205069559063425e-06, + "loss": 1.691, + "step": 10065 + }, + { + "epoch": 0.900035765379113, + "grad_norm": 0.6098197569055911, + "learning_rate": 5.195852019083558e-06, + "loss": 1.5373, + "step": 10066 + }, + { + "epoch": 0.9001251788268956, + "grad_norm": 0.7030994152067434, + "learning_rate": 5.18664243021304e-06, + "loss": 1.7567, + "step": 10067 + }, + { + "epoch": 0.9002145922746781, + "grad_norm": 0.5417079427423546, + "learning_rate": 5.177440793224342e-06, + "loss": 1.8931, + "step": 10068 + }, + { + "epoch": 0.9003040057224606, + "grad_norm": 0.5812620615008149, + "learning_rate": 5.168247108889179e-06, + "loss": 1.5165, + "step": 10069 + }, + { + "epoch": 0.9003934191702432, + "grad_norm": 0.7786043131247301, + "learning_rate": 5.159061377978591e-06, + "loss": 1.6172, + "step": 10070 + }, + { + "epoch": 0.9004828326180258, + "grad_norm": 0.4525209786241275, + "learning_rate": 5.149883601262984e-06, + "loss": 1.4217, + "step": 10071 + }, + { + "epoch": 0.9005722460658083, + "grad_norm": 0.7306276179671635, + "learning_rate": 5.1407137795121075e-06, + "loss": 1.5187, + "step": 10072 + }, + { + "epoch": 0.9006616595135909, + "grad_norm": 0.7433093675210966, + "learning_rate": 5.131551913494981e-06, + "loss": 1.8145, + "step": 10073 + }, + { + "epoch": 0.9007510729613734, + "grad_norm": 0.7615557670265923, + "learning_rate": 5.122398003980033e-06, + "loss": 1.4611, + "step": 10074 + }, + { + "epoch": 0.9008404864091559, + "grad_norm": 0.4997302070703065, + "learning_rate": 5.1132520517349735e-06, + "loss": 1.3163, + "step": 10075 + }, + { + "epoch": 0.9009298998569385, + "grad_norm": 0.6850862692627809, + "learning_rate": 5.104114057526876e-06, + "loss": 1.9332, + "step": 10076 + }, + { + "epoch": 0.9010193133047211, + "grad_norm": 0.4937721735528908, + "learning_rate": 5.09498402212214e-06, + "loss": 1.5293, + "step": 10077 + }, + { + "epoch": 0.9011087267525035, + "grad_norm": 0.6729733753857942, + "learning_rate": 5.085861946286463e-06, + "loss": 1.6966, + "step": 10078 + }, + { + "epoch": 0.9011981402002861, + "grad_norm": 0.43867433733213956, + "learning_rate": 5.076747830784923e-06, + "loss": 1.6018, + "step": 10079 + }, + { + "epoch": 0.9012875536480687, + "grad_norm": 0.6414401166610888, + "learning_rate": 5.067641676381918e-06, + "loss": 1.827, + "step": 10080 + }, + { + "epoch": 0.9013769670958512, + "grad_norm": 0.6502838056512924, + "learning_rate": 5.058543483841116e-06, + "loss": 1.6297, + "step": 10081 + }, + { + "epoch": 0.9014663805436338, + "grad_norm": 0.3877852404483874, + "learning_rate": 5.04945325392564e-06, + "loss": 1.287, + "step": 10082 + }, + { + "epoch": 0.9015557939914163, + "grad_norm": 0.5223276472491214, + "learning_rate": 5.040370987397858e-06, + "loss": 1.5609, + "step": 10083 + }, + { + "epoch": 0.9016452074391988, + "grad_norm": 0.7716277804108881, + "learning_rate": 5.03129668501946e-06, + "loss": 1.7323, + "step": 10084 + }, + { + "epoch": 0.9017346208869814, + "grad_norm": 0.7805814046201232, + "learning_rate": 5.022230347551515e-06, + "loss": 1.6464, + "step": 10085 + }, + { + "epoch": 0.901824034334764, + "grad_norm": 0.38377517525370086, + "learning_rate": 5.013171975754427e-06, + "loss": 1.3948, + "step": 10086 + }, + { + "epoch": 0.9019134477825465, + "grad_norm": 0.8994763369789552, + "learning_rate": 5.004121570387876e-06, + "loss": 1.4633, + "step": 10087 + }, + { + "epoch": 0.902002861230329, + "grad_norm": 0.41577681046228604, + "learning_rate": 4.995079132210922e-06, + "loss": 1.4915, + "step": 10088 + }, + { + "epoch": 0.9020922746781116, + "grad_norm": 0.5666987995741091, + "learning_rate": 4.986044661981948e-06, + "loss": 1.5287, + "step": 10089 + }, + { + "epoch": 0.9021816881258942, + "grad_norm": 0.5747048100877613, + "learning_rate": 4.977018160458646e-06, + "loss": 1.6968, + "step": 10090 + }, + { + "epoch": 0.9022711015736766, + "grad_norm": 0.5550345273715855, + "learning_rate": 4.967999628398101e-06, + "loss": 1.6365, + "step": 10091 + }, + { + "epoch": 0.9023605150214592, + "grad_norm": 0.784908243346529, + "learning_rate": 4.958989066556641e-06, + "loss": 1.8189, + "step": 10092 + }, + { + "epoch": 0.9024499284692418, + "grad_norm": 0.6837445745382368, + "learning_rate": 4.949986475689983e-06, + "loss": 1.5922, + "step": 10093 + }, + { + "epoch": 0.9025393419170243, + "grad_norm": 0.45353999587298066, + "learning_rate": 4.9409918565531675e-06, + "loss": 1.937, + "step": 10094 + }, + { + "epoch": 0.9026287553648069, + "grad_norm": 0.6875571792395843, + "learning_rate": 4.93200520990057e-06, + "loss": 1.6972, + "step": 10095 + }, + { + "epoch": 0.9027181688125894, + "grad_norm": 1.5488450229490878, + "learning_rate": 4.923026536485875e-06, + "loss": 6.137, + "step": 10096 + }, + { + "epoch": 0.9028075822603719, + "grad_norm": 0.5897877066100258, + "learning_rate": 4.914055837062137e-06, + "loss": 1.6312, + "step": 10097 + }, + { + "epoch": 0.9028969957081545, + "grad_norm": 0.4418606137247023, + "learning_rate": 4.905093112381687e-06, + "loss": 1.2875, + "step": 10098 + }, + { + "epoch": 0.9029864091559371, + "grad_norm": 0.7256520890590675, + "learning_rate": 4.896138363196235e-06, + "loss": 1.6677, + "step": 10099 + }, + { + "epoch": 0.9030758226037195, + "grad_norm": 0.9647595717574604, + "learning_rate": 4.8871915902568125e-06, + "loss": 1.5414, + "step": 10100 + }, + { + "epoch": 0.9031652360515021, + "grad_norm": 0.5449385562002343, + "learning_rate": 4.878252794313754e-06, + "loss": 1.6519, + "step": 10101 + }, + { + "epoch": 0.9032546494992847, + "grad_norm": 0.6332434455661162, + "learning_rate": 4.869321976116737e-06, + "loss": 1.5583, + "step": 10102 + }, + { + "epoch": 0.9033440629470673, + "grad_norm": 0.5375716002965282, + "learning_rate": 4.860399136414828e-06, + "loss": 1.5034, + "step": 10103 + }, + { + "epoch": 0.9034334763948498, + "grad_norm": 0.5851797791411785, + "learning_rate": 4.8514842759563306e-06, + "loss": 1.3582, + "step": 10104 + }, + { + "epoch": 0.9035228898426323, + "grad_norm": 0.535547298931117, + "learning_rate": 4.842577395488934e-06, + "loss": 1.4702, + "step": 10105 + }, + { + "epoch": 0.9036123032904149, + "grad_norm": 0.570712850557299, + "learning_rate": 4.833678495759664e-06, + "loss": 1.4477, + "step": 10106 + }, + { + "epoch": 0.9037017167381974, + "grad_norm": 0.8830429059013121, + "learning_rate": 4.8247875775148335e-06, + "loss": 1.763, + "step": 10107 + }, + { + "epoch": 0.90379113018598, + "grad_norm": 0.445591772072697, + "learning_rate": 4.815904641500124e-06, + "loss": 1.3247, + "step": 10108 + }, + { + "epoch": 0.9038805436337625, + "grad_norm": 0.49926072407156025, + "learning_rate": 4.80702968846054e-06, + "loss": 1.4363, + "step": 10109 + }, + { + "epoch": 0.903969957081545, + "grad_norm": 0.6479721859796923, + "learning_rate": 4.79816271914042e-06, + "loss": 1.5865, + "step": 10110 + }, + { + "epoch": 0.9040593705293276, + "grad_norm": 0.6392178956286044, + "learning_rate": 4.789303734283423e-06, + "loss": 1.4869, + "step": 10111 + }, + { + "epoch": 0.9041487839771102, + "grad_norm": 0.537078664090611, + "learning_rate": 4.780452734632524e-06, + "loss": 1.7265, + "step": 10112 + }, + { + "epoch": 0.9042381974248928, + "grad_norm": 0.6165838645989239, + "learning_rate": 4.771609720930059e-06, + "loss": 1.5191, + "step": 10113 + }, + { + "epoch": 0.9043276108726752, + "grad_norm": 0.5098690291805817, + "learning_rate": 4.762774693917693e-06, + "loss": 1.969, + "step": 10114 + }, + { + "epoch": 0.9044170243204578, + "grad_norm": 0.41454344231986584, + "learning_rate": 4.753947654336388e-06, + "loss": 1.3694, + "step": 10115 + }, + { + "epoch": 0.9045064377682404, + "grad_norm": 0.5320283818234204, + "learning_rate": 4.7451286029264405e-06, + "loss": 1.669, + "step": 10116 + }, + { + "epoch": 0.9045958512160229, + "grad_norm": 0.4360880086304415, + "learning_rate": 4.73631754042756e-06, + "loss": 1.5762, + "step": 10117 + }, + { + "epoch": 0.9046852646638054, + "grad_norm": 0.6755300476540885, + "learning_rate": 4.727514467578653e-06, + "loss": 1.2712, + "step": 10118 + }, + { + "epoch": 0.904774678111588, + "grad_norm": 0.8697825835837268, + "learning_rate": 4.718719385118053e-06, + "loss": 1.7487, + "step": 10119 + }, + { + "epoch": 0.9048640915593705, + "grad_norm": 0.6430536212782236, + "learning_rate": 4.7099322937833925e-06, + "loss": 1.4603, + "step": 10120 + }, + { + "epoch": 0.9049535050071531, + "grad_norm": 0.42167792906788865, + "learning_rate": 4.701153194311625e-06, + "loss": 1.5414, + "step": 10121 + }, + { + "epoch": 0.9050429184549357, + "grad_norm": 0.4461137681965416, + "learning_rate": 4.69238208743904e-06, + "loss": 1.6381, + "step": 10122 + }, + { + "epoch": 0.9051323319027181, + "grad_norm": 0.4540079835170972, + "learning_rate": 4.6836189739012715e-06, + "loss": 2.0891, + "step": 10123 + }, + { + "epoch": 0.9052217453505007, + "grad_norm": 0.38310036245180634, + "learning_rate": 4.6748638544332644e-06, + "loss": 1.583, + "step": 10124 + }, + { + "epoch": 0.9053111587982833, + "grad_norm": 0.5373647679482608, + "learning_rate": 4.66611672976931e-06, + "loss": 1.4927, + "step": 10125 + }, + { + "epoch": 0.9054005722460658, + "grad_norm": 0.8686000795307784, + "learning_rate": 4.6573776006430205e-06, + "loss": 1.5758, + "step": 10126 + }, + { + "epoch": 0.9054899856938483, + "grad_norm": 0.5728100866362607, + "learning_rate": 4.6486464677873094e-06, + "loss": 1.5091, + "step": 10127 + }, + { + "epoch": 0.9055793991416309, + "grad_norm": 0.3875334056639, + "learning_rate": 4.639923331934471e-06, + "loss": 1.3487, + "step": 10128 + }, + { + "epoch": 0.9056688125894135, + "grad_norm": 0.4113740154700533, + "learning_rate": 4.631208193816083e-06, + "loss": 1.5582, + "step": 10129 + }, + { + "epoch": 0.905758226037196, + "grad_norm": 0.8410796297959515, + "learning_rate": 4.622501054163098e-06, + "loss": 1.6663, + "step": 10130 + }, + { + "epoch": 0.9058476394849786, + "grad_norm": 0.6899469855358648, + "learning_rate": 4.613801913705773e-06, + "loss": 1.719, + "step": 10131 + }, + { + "epoch": 0.905937052932761, + "grad_norm": 0.5851375014356174, + "learning_rate": 4.605110773173682e-06, + "loss": 1.6181, + "step": 10132 + }, + { + "epoch": 0.9060264663805436, + "grad_norm": 0.6220168924122702, + "learning_rate": 4.59642763329573e-06, + "loss": 1.4849, + "step": 10133 + }, + { + "epoch": 0.9061158798283262, + "grad_norm": 0.40707161136252906, + "learning_rate": 4.5877524948001905e-06, + "loss": 1.3933, + "step": 10134 + }, + { + "epoch": 0.9062052932761088, + "grad_norm": 0.3755231190923393, + "learning_rate": 4.5790853584146035e-06, + "loss": 1.5753, + "step": 10135 + }, + { + "epoch": 0.9062947067238912, + "grad_norm": 0.379560057503046, + "learning_rate": 4.570426224865876e-06, + "loss": 1.3914, + "step": 10136 + }, + { + "epoch": 0.9063841201716738, + "grad_norm": 0.43741078432009406, + "learning_rate": 4.561775094880283e-06, + "loss": 1.5551, + "step": 10137 + }, + { + "epoch": 0.9064735336194564, + "grad_norm": 0.50604623503256, + "learning_rate": 4.5531319691833326e-06, + "loss": 1.749, + "step": 10138 + }, + { + "epoch": 0.906562947067239, + "grad_norm": 0.4656705244469675, + "learning_rate": 4.544496848499946e-06, + "loss": 1.7672, + "step": 10139 + }, + { + "epoch": 0.9066523605150214, + "grad_norm": 0.4514870306098492, + "learning_rate": 4.535869733554332e-06, + "loss": 1.5411, + "step": 10140 + }, + { + "epoch": 0.906741773962804, + "grad_norm": 0.40437400559553455, + "learning_rate": 4.527250625070012e-06, + "loss": 1.2527, + "step": 10141 + }, + { + "epoch": 0.9068311874105865, + "grad_norm": 0.47829620583511523, + "learning_rate": 4.518639523769897e-06, + "loss": 1.5752, + "step": 10142 + }, + { + "epoch": 0.9069206008583691, + "grad_norm": 0.7098207585713291, + "learning_rate": 4.510036430376152e-06, + "loss": 2.0565, + "step": 10143 + }, + { + "epoch": 0.9070100143061517, + "grad_norm": 0.7098662460457059, + "learning_rate": 4.501441345610347e-06, + "loss": 1.837, + "step": 10144 + }, + { + "epoch": 0.9070994277539342, + "grad_norm": 0.5204993145047319, + "learning_rate": 4.492854270193325e-06, + "loss": 1.7614, + "step": 10145 + }, + { + "epoch": 0.9071888412017167, + "grad_norm": 1.4223092594762565, + "learning_rate": 4.4842752048452676e-06, + "loss": 6.0672, + "step": 10146 + }, + { + "epoch": 0.9072782546494993, + "grad_norm": 0.5956525477105695, + "learning_rate": 4.475704150285687e-06, + "loss": 1.7776, + "step": 10147 + }, + { + "epoch": 0.9073676680972819, + "grad_norm": 0.5116131097883408, + "learning_rate": 4.4671411072334526e-06, + "loss": 1.2807, + "step": 10148 + }, + { + "epoch": 0.9074570815450643, + "grad_norm": 0.5721455742623011, + "learning_rate": 4.458586076406701e-06, + "loss": 1.7761, + "step": 10149 + }, + { + "epoch": 0.9075464949928469, + "grad_norm": 0.48820717126963203, + "learning_rate": 4.450039058522948e-06, + "loss": 1.5627, + "step": 10150 + }, + { + "epoch": 0.9076359084406295, + "grad_norm": 0.41990544606057667, + "learning_rate": 4.441500054299042e-06, + "loss": 1.4262, + "step": 10151 + }, + { + "epoch": 0.907725321888412, + "grad_norm": 0.6270299169463628, + "learning_rate": 4.432969064451109e-06, + "loss": 1.608, + "step": 10152 + }, + { + "epoch": 0.9078147353361946, + "grad_norm": 0.4532252120355065, + "learning_rate": 4.424446089694645e-06, + "loss": 1.7222, + "step": 10153 + }, + { + "epoch": 0.9079041487839771, + "grad_norm": 0.40349126348409614, + "learning_rate": 4.415931130744477e-06, + "loss": 1.5497, + "step": 10154 + }, + { + "epoch": 0.9079935622317596, + "grad_norm": 0.5206562183006449, + "learning_rate": 4.407424188314713e-06, + "loss": 1.3553, + "step": 10155 + }, + { + "epoch": 0.9080829756795422, + "grad_norm": 0.45257843755740346, + "learning_rate": 4.398925263118836e-06, + "loss": 1.3178, + "step": 10156 + }, + { + "epoch": 0.9081723891273248, + "grad_norm": 0.5343722739704524, + "learning_rate": 4.390434355869643e-06, + "loss": 1.475, + "step": 10157 + }, + { + "epoch": 0.9082618025751072, + "grad_norm": 0.6184449416200871, + "learning_rate": 4.381951467279244e-06, + "loss": 1.515, + "step": 10158 + }, + { + "epoch": 0.9083512160228898, + "grad_norm": 0.781805740163572, + "learning_rate": 4.373476598059112e-06, + "loss": 1.4304, + "step": 10159 + }, + { + "epoch": 0.9084406294706724, + "grad_norm": 0.3850771915040515, + "learning_rate": 4.365009748920012e-06, + "loss": 1.4054, + "step": 10160 + }, + { + "epoch": 0.908530042918455, + "grad_norm": 0.6131172927142468, + "learning_rate": 4.356550920572044e-06, + "loss": 1.7375, + "step": 10161 + }, + { + "epoch": 0.9086194563662375, + "grad_norm": 0.836715428647674, + "learning_rate": 4.348100113724629e-06, + "loss": 1.9678, + "step": 10162 + }, + { + "epoch": 0.90870886981402, + "grad_norm": 0.4365641539173706, + "learning_rate": 4.339657329086566e-06, + "loss": 1.4399, + "step": 10163 + }, + { + "epoch": 0.9087982832618026, + "grad_norm": 0.5328950539997092, + "learning_rate": 4.331222567365878e-06, + "loss": 1.4607, + "step": 10164 + }, + { + "epoch": 0.9088876967095851, + "grad_norm": 0.5173146924661403, + "learning_rate": 4.322795829270043e-06, + "loss": 1.4766, + "step": 10165 + }, + { + "epoch": 0.9089771101573677, + "grad_norm": 0.6342861999650498, + "learning_rate": 4.314377115505763e-06, + "loss": 1.5455, + "step": 10166 + }, + { + "epoch": 0.9090665236051502, + "grad_norm": 0.3485393174750181, + "learning_rate": 4.305966426779118e-06, + "loss": 1.4797, + "step": 10167 + }, + { + "epoch": 0.9091559370529327, + "grad_norm": 0.9006562127876908, + "learning_rate": 4.297563763795509e-06, + "loss": 1.6149, + "step": 10168 + }, + { + "epoch": 0.9092453505007153, + "grad_norm": 0.4873906069283725, + "learning_rate": 4.289169127259629e-06, + "loss": 1.4104, + "step": 10169 + }, + { + "epoch": 0.9093347639484979, + "grad_norm": 0.3545680887171886, + "learning_rate": 4.280782517875548e-06, + "loss": 1.4039, + "step": 10170 + }, + { + "epoch": 0.9094241773962805, + "grad_norm": 0.4587502660063021, + "learning_rate": 4.272403936346647e-06, + "loss": 1.5989, + "step": 10171 + }, + { + "epoch": 0.9095135908440629, + "grad_norm": 0.49796777815637283, + "learning_rate": 4.26403338337561e-06, + "loss": 1.5045, + "step": 10172 + }, + { + "epoch": 0.9096030042918455, + "grad_norm": 0.7489943446871727, + "learning_rate": 4.255670859664474e-06, + "loss": 6.0416, + "step": 10173 + }, + { + "epoch": 0.9096924177396281, + "grad_norm": 0.39531593603276766, + "learning_rate": 4.2473163659146e-06, + "loss": 1.5279, + "step": 10174 + }, + { + "epoch": 0.9097818311874106, + "grad_norm": 0.5896592721436175, + "learning_rate": 4.238969902826662e-06, + "loss": 1.3985, + "step": 10175 + }, + { + "epoch": 0.9098712446351931, + "grad_norm": 0.37392508411461867, + "learning_rate": 4.230631471100655e-06, + "loss": 1.3898, + "step": 10176 + }, + { + "epoch": 0.9099606580829757, + "grad_norm": 0.7361790607986982, + "learning_rate": 4.222301071435952e-06, + "loss": 1.736, + "step": 10177 + }, + { + "epoch": 0.9100500715307582, + "grad_norm": 0.6809592819502938, + "learning_rate": 4.213978704531152e-06, + "loss": 1.7379, + "step": 10178 + }, + { + "epoch": 0.9101394849785408, + "grad_norm": 0.5801748358629101, + "learning_rate": 4.205664371084306e-06, + "loss": 1.581, + "step": 10179 + }, + { + "epoch": 0.9102288984263234, + "grad_norm": 0.4424910506431924, + "learning_rate": 4.19735807179269e-06, + "loss": 1.4132, + "step": 10180 + }, + { + "epoch": 0.9103183118741058, + "grad_norm": 0.655498638676839, + "learning_rate": 4.189059807352958e-06, + "loss": 1.5205, + "step": 10181 + }, + { + "epoch": 0.9104077253218884, + "grad_norm": 0.5097649307444855, + "learning_rate": 4.180769578461063e-06, + "loss": 1.6852, + "step": 10182 + }, + { + "epoch": 0.910497138769671, + "grad_norm": 0.4756808963781449, + "learning_rate": 4.172487385812307e-06, + "loss": 1.4915, + "step": 10183 + }, + { + "epoch": 0.9105865522174535, + "grad_norm": 0.5538743694037898, + "learning_rate": 4.164213230101299e-06, + "loss": 1.6692, + "step": 10184 + }, + { + "epoch": 0.910675965665236, + "grad_norm": 0.8908780175088208, + "learning_rate": 4.155947112021985e-06, + "loss": 1.4098, + "step": 10185 + }, + { + "epoch": 0.9107653791130186, + "grad_norm": 0.6087487654258877, + "learning_rate": 4.147689032267643e-06, + "loss": 1.7791, + "step": 10186 + }, + { + "epoch": 0.9108547925608012, + "grad_norm": 0.7054662367196115, + "learning_rate": 4.139438991530853e-06, + "loss": 1.3127, + "step": 10187 + }, + { + "epoch": 0.9109442060085837, + "grad_norm": 0.5675338454675848, + "learning_rate": 4.131196990503561e-06, + "loss": 1.7413, + "step": 10188 + }, + { + "epoch": 0.9110336194563662, + "grad_norm": 0.6801684967618511, + "learning_rate": 4.1229630298769914e-06, + "loss": 1.7134, + "step": 10189 + }, + { + "epoch": 0.9111230329041488, + "grad_norm": 0.45212409772269235, + "learning_rate": 4.114737110341715e-06, + "loss": 1.728, + "step": 10190 + }, + { + "epoch": 0.9112124463519313, + "grad_norm": 0.5305595897724577, + "learning_rate": 4.106519232587647e-06, + "loss": 1.9857, + "step": 10191 + }, + { + "epoch": 0.9113018597997139, + "grad_norm": 0.4474652591242352, + "learning_rate": 4.098309397303978e-06, + "loss": 1.3517, + "step": 10192 + }, + { + "epoch": 0.9113912732474965, + "grad_norm": 0.569111646396901, + "learning_rate": 4.090107605179294e-06, + "loss": 1.632, + "step": 10193 + }, + { + "epoch": 0.9114806866952789, + "grad_norm": 0.4686648561070638, + "learning_rate": 4.081913856901476e-06, + "loss": 1.564, + "step": 10194 + }, + { + "epoch": 0.9115701001430615, + "grad_norm": 0.5373778160129681, + "learning_rate": 4.073728153157674e-06, + "loss": 1.6254, + "step": 10195 + }, + { + "epoch": 0.9116595135908441, + "grad_norm": 0.4759295496683307, + "learning_rate": 4.065550494634451e-06, + "loss": 1.5796, + "step": 10196 + }, + { + "epoch": 0.9117489270386266, + "grad_norm": 0.5387067547057248, + "learning_rate": 4.057380882017658e-06, + "loss": 1.5124, + "step": 10197 + }, + { + "epoch": 0.9118383404864091, + "grad_norm": 0.6013354368279105, + "learning_rate": 4.049219315992458e-06, + "loss": 1.676, + "step": 10198 + }, + { + "epoch": 0.9119277539341917, + "grad_norm": 0.5778907721512345, + "learning_rate": 4.041065797243349e-06, + "loss": 1.7129, + "step": 10199 + }, + { + "epoch": 0.9120171673819742, + "grad_norm": 0.4859471534994306, + "learning_rate": 4.032920326454159e-06, + "loss": 1.2781, + "step": 10200 + }, + { + "epoch": 0.9121065808297568, + "grad_norm": 0.8162482391157377, + "learning_rate": 4.0247829043080445e-06, + "loss": 1.6151, + "step": 10201 + }, + { + "epoch": 0.9121959942775394, + "grad_norm": 0.36615301423751295, + "learning_rate": 4.016653531487491e-06, + "loss": 1.3978, + "step": 10202 + }, + { + "epoch": 0.9122854077253219, + "grad_norm": 0.637782236022841, + "learning_rate": 4.008532208674276e-06, + "loss": 1.422, + "step": 10203 + }, + { + "epoch": 0.9123748211731044, + "grad_norm": 0.40549006819585703, + "learning_rate": 4.000418936549533e-06, + "loss": 1.7503, + "step": 10204 + }, + { + "epoch": 0.912464234620887, + "grad_norm": 0.8656671407960064, + "learning_rate": 3.992313715793727e-06, + "loss": 6.0841, + "step": 10205 + }, + { + "epoch": 0.9125536480686696, + "grad_norm": 0.5733932447130735, + "learning_rate": 3.984216547086606e-06, + "loss": 1.4035, + "step": 10206 + }, + { + "epoch": 0.912643061516452, + "grad_norm": 0.40656570411957765, + "learning_rate": 3.97612743110729e-06, + "loss": 1.6997, + "step": 10207 + }, + { + "epoch": 0.9127324749642346, + "grad_norm": 0.47109438147969146, + "learning_rate": 3.968046368534217e-06, + "loss": 1.2027, + "step": 10208 + }, + { + "epoch": 0.9128218884120172, + "grad_norm": 0.5058829948980255, + "learning_rate": 3.9599733600450995e-06, + "loss": 1.6861, + "step": 10209 + }, + { + "epoch": 0.9129113018597997, + "grad_norm": 0.5416202811371839, + "learning_rate": 3.95190840631704e-06, + "loss": 1.8439, + "step": 10210 + }, + { + "epoch": 0.9130007153075823, + "grad_norm": 0.6528696141190833, + "learning_rate": 3.94385150802643e-06, + "loss": 1.9204, + "step": 10211 + }, + { + "epoch": 0.9130901287553648, + "grad_norm": 0.5376196906582593, + "learning_rate": 3.9358026658489535e-06, + "loss": 1.5658, + "step": 10212 + }, + { + "epoch": 0.9131795422031473, + "grad_norm": 1.214874499297681, + "learning_rate": 3.927761880459735e-06, + "loss": 1.6248, + "step": 10213 + }, + { + "epoch": 0.9132689556509299, + "grad_norm": 0.423870200006217, + "learning_rate": 3.91972915253308e-06, + "loss": 1.585, + "step": 10214 + }, + { + "epoch": 0.9133583690987125, + "grad_norm": 0.4057363623257469, + "learning_rate": 3.9117044827427066e-06, + "loss": 1.7772, + "step": 10215 + }, + { + "epoch": 0.913447782546495, + "grad_norm": 0.4798288008050428, + "learning_rate": 3.90368787176163e-06, + "loss": 1.7239, + "step": 10216 + }, + { + "epoch": 0.9135371959942775, + "grad_norm": 0.5141315226591864, + "learning_rate": 3.895679320262202e-06, + "loss": 1.6974, + "step": 10217 + }, + { + "epoch": 0.9136266094420601, + "grad_norm": 0.696885880407775, + "learning_rate": 3.8876788289160855e-06, + "loss": 1.6153, + "step": 10218 + }, + { + "epoch": 0.9137160228898427, + "grad_norm": 0.6837167357858456, + "learning_rate": 3.879686398394267e-06, + "loss": 2.0634, + "step": 10219 + }, + { + "epoch": 0.9138054363376252, + "grad_norm": 0.45073676335401813, + "learning_rate": 3.871702029367064e-06, + "loss": 1.537, + "step": 10220 + }, + { + "epoch": 0.9138948497854077, + "grad_norm": 0.5331000430731719, + "learning_rate": 3.863725722504119e-06, + "loss": 1.8361, + "step": 10221 + }, + { + "epoch": 0.9139842632331903, + "grad_norm": 0.35409924851530516, + "learning_rate": 3.8557574784744085e-06, + "loss": 1.4578, + "step": 10222 + }, + { + "epoch": 0.9140736766809728, + "grad_norm": 0.5486358602502844, + "learning_rate": 3.847797297946198e-06, + "loss": 1.2423, + "step": 10223 + }, + { + "epoch": 0.9141630901287554, + "grad_norm": 1.0701230408969657, + "learning_rate": 3.839845181587098e-06, + "loss": 6.1684, + "step": 10224 + }, + { + "epoch": 0.9142525035765379, + "grad_norm": 0.713719427494771, + "learning_rate": 3.831901130064064e-06, + "loss": 1.5862, + "step": 10225 + }, + { + "epoch": 0.9143419170243204, + "grad_norm": 0.5658804993840124, + "learning_rate": 3.823965144043318e-06, + "loss": 1.4262, + "step": 10226 + }, + { + "epoch": 0.914431330472103, + "grad_norm": 0.753852233609641, + "learning_rate": 3.816037224190483e-06, + "loss": 1.8809, + "step": 10227 + }, + { + "epoch": 0.9145207439198856, + "grad_norm": 0.5191204416476255, + "learning_rate": 3.8081173711704497e-06, + "loss": 1.3797, + "step": 10228 + }, + { + "epoch": 0.914610157367668, + "grad_norm": 0.528651098018859, + "learning_rate": 3.8002055856474206e-06, + "loss": 1.5525, + "step": 10229 + }, + { + "epoch": 0.9146995708154506, + "grad_norm": 0.5460602205182352, + "learning_rate": 3.7923018682849864e-06, + "loss": 1.5515, + "step": 10230 + }, + { + "epoch": 0.9147889842632332, + "grad_norm": 0.7739514004223362, + "learning_rate": 3.784406219746006e-06, + "loss": 1.7257, + "step": 10231 + }, + { + "epoch": 0.9148783977110158, + "grad_norm": 0.45827721419754497, + "learning_rate": 3.7765186406926722e-06, + "loss": 1.3946, + "step": 10232 + }, + { + "epoch": 0.9149678111587983, + "grad_norm": 0.5774937044161993, + "learning_rate": 3.768639131786511e-06, + "loss": 1.8317, + "step": 10233 + }, + { + "epoch": 0.9150572246065808, + "grad_norm": 0.621442330371376, + "learning_rate": 3.760767693688361e-06, + "loss": 1.5951, + "step": 10234 + }, + { + "epoch": 0.9151466380543634, + "grad_norm": 0.5766627370012698, + "learning_rate": 3.752904327058404e-06, + "loss": 1.9473, + "step": 10235 + }, + { + "epoch": 0.9152360515021459, + "grad_norm": 0.4486301377883255, + "learning_rate": 3.745049032556125e-06, + "loss": 1.4989, + "step": 10236 + }, + { + "epoch": 0.9153254649499285, + "grad_norm": 1.0108886474403151, + "learning_rate": 3.7372018108403405e-06, + "loss": 1.7598, + "step": 10237 + }, + { + "epoch": 0.915414878397711, + "grad_norm": 0.4822557949792332, + "learning_rate": 3.729362662569169e-06, + "loss": 1.6579, + "step": 10238 + }, + { + "epoch": 0.9155042918454935, + "grad_norm": 0.613573573886884, + "learning_rate": 3.7215315884000957e-06, + "loss": 2.0786, + "step": 10239 + }, + { + "epoch": 0.9155937052932761, + "grad_norm": 1.0479247237679925, + "learning_rate": 3.7137085889898947e-06, + "loss": 5.9018, + "step": 10240 + }, + { + "epoch": 0.9156831187410587, + "grad_norm": 0.48717739811148614, + "learning_rate": 3.705893664994664e-06, + "loss": 1.5053, + "step": 10241 + }, + { + "epoch": 0.9157725321888412, + "grad_norm": 0.37844157469343476, + "learning_rate": 3.6980868170698456e-06, + "loss": 1.515, + "step": 10242 + }, + { + "epoch": 0.9158619456366237, + "grad_norm": 0.6782139497606116, + "learning_rate": 3.6902880458701826e-06, + "loss": 1.6608, + "step": 10243 + }, + { + "epoch": 0.9159513590844063, + "grad_norm": 0.45076504368481496, + "learning_rate": 3.6824973520497408e-06, + "loss": 1.5822, + "step": 10244 + }, + { + "epoch": 0.9160407725321889, + "grad_norm": 0.6327839384875144, + "learning_rate": 3.6747147362619304e-06, + "loss": 1.4475, + "step": 10245 + }, + { + "epoch": 0.9161301859799714, + "grad_norm": 0.38035075465388046, + "learning_rate": 3.666940199159463e-06, + "loss": 1.4174, + "step": 10246 + }, + { + "epoch": 0.9162195994277539, + "grad_norm": 0.7762483148723274, + "learning_rate": 3.6591737413943616e-06, + "loss": 1.4979, + "step": 10247 + }, + { + "epoch": 0.9163090128755365, + "grad_norm": 0.4404050556267454, + "learning_rate": 3.6514153636180383e-06, + "loss": 1.5279, + "step": 10248 + }, + { + "epoch": 0.916398426323319, + "grad_norm": 0.41598860026779544, + "learning_rate": 3.643665066481128e-06, + "loss": 1.3978, + "step": 10249 + }, + { + "epoch": 0.9164878397711016, + "grad_norm": 0.42527280178936716, + "learning_rate": 3.635922850633666e-06, + "loss": 1.5678, + "step": 10250 + }, + { + "epoch": 0.9165772532188842, + "grad_norm": 0.5227331995750172, + "learning_rate": 3.6281887167249895e-06, + "loss": 1.5216, + "step": 10251 + }, + { + "epoch": 0.9166666666666666, + "grad_norm": 0.473947731489007, + "learning_rate": 3.6204626654037233e-06, + "loss": 1.6069, + "step": 10252 + }, + { + "epoch": 0.9167560801144492, + "grad_norm": 0.8384136362644158, + "learning_rate": 3.612744697317849e-06, + "loss": 1.7414, + "step": 10253 + }, + { + "epoch": 0.9168454935622318, + "grad_norm": 0.7035299679589188, + "learning_rate": 3.6050348131146825e-06, + "loss": 1.6267, + "step": 10254 + }, + { + "epoch": 0.9169349070100143, + "grad_norm": 0.5003623237489346, + "learning_rate": 3.597333013440829e-06, + "loss": 1.6926, + "step": 10255 + }, + { + "epoch": 0.9170243204577968, + "grad_norm": 0.39837204625741823, + "learning_rate": 3.5896392989422377e-06, + "loss": 1.4798, + "step": 10256 + }, + { + "epoch": 0.9171137339055794, + "grad_norm": 0.4654277160568861, + "learning_rate": 3.5819536702641485e-06, + "loss": 1.6016, + "step": 10257 + }, + { + "epoch": 0.917203147353362, + "grad_norm": 0.36445793717992303, + "learning_rate": 3.5742761280511685e-06, + "loss": 1.4183, + "step": 10258 + }, + { + "epoch": 0.9172925608011445, + "grad_norm": 0.9131081403750184, + "learning_rate": 3.566606672947204e-06, + "loss": 1.6544, + "step": 10259 + }, + { + "epoch": 0.9173819742489271, + "grad_norm": 0.719993044982243, + "learning_rate": 3.5589453055954737e-06, + "loss": 1.8925, + "step": 10260 + }, + { + "epoch": 0.9174713876967096, + "grad_norm": 0.35004329593999994, + "learning_rate": 3.5512920266385085e-06, + "loss": 1.5508, + "step": 10261 + }, + { + "epoch": 0.9175608011444921, + "grad_norm": 0.385574613827311, + "learning_rate": 3.5436468367182284e-06, + "loss": 1.5196, + "step": 10262 + }, + { + "epoch": 0.9176502145922747, + "grad_norm": 0.5021231186879397, + "learning_rate": 3.536009736475787e-06, + "loss": 1.5892, + "step": 10263 + }, + { + "epoch": 0.9177396280400573, + "grad_norm": 0.4913771927241536, + "learning_rate": 3.5283807265517053e-06, + "loss": 1.876, + "step": 10264 + }, + { + "epoch": 0.9178290414878397, + "grad_norm": 0.4129803709880747, + "learning_rate": 3.5207598075858383e-06, + "loss": 1.5959, + "step": 10265 + }, + { + "epoch": 0.9179184549356223, + "grad_norm": 0.44523415750809076, + "learning_rate": 3.5131469802173076e-06, + "loss": 1.5663, + "step": 10266 + }, + { + "epoch": 0.9180078683834049, + "grad_norm": 0.4895029635175224, + "learning_rate": 3.5055422450846253e-06, + "loss": 1.9036, + "step": 10267 + }, + { + "epoch": 0.9180972818311874, + "grad_norm": 0.46606804548404485, + "learning_rate": 3.4979456028255806e-06, + "loss": 1.4159, + "step": 10268 + }, + { + "epoch": 0.91818669527897, + "grad_norm": 0.511245543815447, + "learning_rate": 3.4903570540772866e-06, + "loss": 1.7885, + "step": 10269 + }, + { + "epoch": 0.9182761087267525, + "grad_norm": 0.7532159040921754, + "learning_rate": 3.482776599476201e-06, + "loss": 1.7184, + "step": 10270 + }, + { + "epoch": 0.918365522174535, + "grad_norm": 0.40879469512162714, + "learning_rate": 3.4752042396580807e-06, + "loss": 1.7012, + "step": 10271 + }, + { + "epoch": 0.9184549356223176, + "grad_norm": 0.53563511628696, + "learning_rate": 3.467639975257997e-06, + "loss": 1.822, + "step": 10272 + }, + { + "epoch": 0.9185443490701002, + "grad_norm": 0.5632106410420329, + "learning_rate": 3.4600838069103635e-06, + "loss": 1.6354, + "step": 10273 + }, + { + "epoch": 0.9186337625178826, + "grad_norm": 0.41644277284185643, + "learning_rate": 3.4525357352489295e-06, + "loss": 1.7029, + "step": 10274 + }, + { + "epoch": 0.9187231759656652, + "grad_norm": 0.3592884920660952, + "learning_rate": 3.4449957609066996e-06, + "loss": 1.4836, + "step": 10275 + }, + { + "epoch": 0.9188125894134478, + "grad_norm": 1.2975297236791108, + "learning_rate": 3.43746388451609e-06, + "loss": 1.7122, + "step": 10276 + }, + { + "epoch": 0.9189020028612304, + "grad_norm": 0.7205859183283377, + "learning_rate": 3.429940106708751e-06, + "loss": 1.7423, + "step": 10277 + }, + { + "epoch": 0.9189914163090128, + "grad_norm": 0.46137304303172344, + "learning_rate": 3.422424428115711e-06, + "loss": 1.8218, + "step": 10278 + }, + { + "epoch": 0.9190808297567954, + "grad_norm": 0.6030011746044626, + "learning_rate": 3.4149168493673113e-06, + "loss": 1.6795, + "step": 10279 + }, + { + "epoch": 0.919170243204578, + "grad_norm": 0.569697659171585, + "learning_rate": 3.40741737109318e-06, + "loss": 1.1764, + "step": 10280 + }, + { + "epoch": 0.9192596566523605, + "grad_norm": 0.5591058983644438, + "learning_rate": 3.3999259939222927e-06, + "loss": 1.7788, + "step": 10281 + }, + { + "epoch": 0.9193490701001431, + "grad_norm": 1.5951718541883246, + "learning_rate": 3.3924427184829575e-06, + "loss": 6.1042, + "step": 10282 + }, + { + "epoch": 0.9194384835479256, + "grad_norm": 0.5791215042766659, + "learning_rate": 3.3849675454027727e-06, + "loss": 1.2465, + "step": 10283 + }, + { + "epoch": 0.9195278969957081, + "grad_norm": 0.42845329707413926, + "learning_rate": 3.3775004753086812e-06, + "loss": 1.4155, + "step": 10284 + }, + { + "epoch": 0.9196173104434907, + "grad_norm": 0.5931338490060607, + "learning_rate": 3.3700415088269377e-06, + "loss": 1.2842, + "step": 10285 + }, + { + "epoch": 0.9197067238912733, + "grad_norm": 0.4687894813427341, + "learning_rate": 3.362590646583108e-06, + "loss": 1.5325, + "step": 10286 + }, + { + "epoch": 0.9197961373390557, + "grad_norm": 0.46004212585761634, + "learning_rate": 3.3551478892020926e-06, + "loss": 1.588, + "step": 10287 + }, + { + "epoch": 0.9198855507868383, + "grad_norm": 0.7473489609887167, + "learning_rate": 3.3477132373081254e-06, + "loss": 1.7895, + "step": 10288 + }, + { + "epoch": 0.9199749642346209, + "grad_norm": 0.6382503435153809, + "learning_rate": 3.3402866915246854e-06, + "loss": 1.9492, + "step": 10289 + }, + { + "epoch": 0.9200643776824035, + "grad_norm": 0.5561788327077584, + "learning_rate": 3.3328682524746967e-06, + "loss": 1.6989, + "step": 10290 + }, + { + "epoch": 0.920153791130186, + "grad_norm": 0.6712494517204414, + "learning_rate": 3.325457920780295e-06, + "loss": 1.4563, + "step": 10291 + }, + { + "epoch": 0.9202432045779685, + "grad_norm": 0.5156985671544354, + "learning_rate": 3.318055697062983e-06, + "loss": 1.5461, + "step": 10292 + }, + { + "epoch": 0.9203326180257511, + "grad_norm": 0.4602492676283537, + "learning_rate": 3.310661581943586e-06, + "loss": 1.5896, + "step": 10293 + }, + { + "epoch": 0.9204220314735336, + "grad_norm": 0.5037153120664148, + "learning_rate": 3.3032755760422196e-06, + "loss": 1.5078, + "step": 10294 + }, + { + "epoch": 0.9205114449213162, + "grad_norm": 0.6230007411646122, + "learning_rate": 3.2958976799783326e-06, + "loss": 2.1039, + "step": 10295 + }, + { + "epoch": 0.9206008583690987, + "grad_norm": 0.4724944015697878, + "learning_rate": 3.288527894370752e-06, + "loss": 1.4006, + "step": 10296 + }, + { + "epoch": 0.9206902718168812, + "grad_norm": 0.5499391613974901, + "learning_rate": 3.281166219837517e-06, + "loss": 1.6982, + "step": 10297 + }, + { + "epoch": 0.9207796852646638, + "grad_norm": 0.6843348071907952, + "learning_rate": 3.273812656996067e-06, + "loss": 1.521, + "step": 10298 + }, + { + "epoch": 0.9208690987124464, + "grad_norm": 0.6712674367597704, + "learning_rate": 3.2664672064631528e-06, + "loss": 1.6893, + "step": 10299 + }, + { + "epoch": 0.920958512160229, + "grad_norm": 0.452607035794084, + "learning_rate": 3.2591298688547932e-06, + "loss": 1.4914, + "step": 10300 + }, + { + "epoch": 0.9210479256080114, + "grad_norm": 0.5179176132689459, + "learning_rate": 3.2518006447863847e-06, + "loss": 1.4364, + "step": 10301 + }, + { + "epoch": 0.921137339055794, + "grad_norm": 0.4690767111558384, + "learning_rate": 3.244479534872602e-06, + "loss": 1.8446, + "step": 10302 + }, + { + "epoch": 0.9212267525035766, + "grad_norm": 0.41385052725196325, + "learning_rate": 3.2371665397274763e-06, + "loss": 1.5582, + "step": 10303 + }, + { + "epoch": 0.9213161659513591, + "grad_norm": 0.6142936889720416, + "learning_rate": 3.2298616599643285e-06, + "loss": 1.9427, + "step": 10304 + }, + { + "epoch": 0.9214055793991416, + "grad_norm": 0.7957847992140766, + "learning_rate": 3.2225648961958344e-06, + "loss": 1.4649, + "step": 10305 + }, + { + "epoch": 0.9214949928469242, + "grad_norm": 0.44225371321211154, + "learning_rate": 3.215276249033927e-06, + "loss": 1.5137, + "step": 10306 + }, + { + "epoch": 0.9215844062947067, + "grad_norm": 0.38323101926943337, + "learning_rate": 3.207995719089918e-06, + "loss": 1.2157, + "step": 10307 + }, + { + "epoch": 0.9216738197424893, + "grad_norm": 0.3566527040188397, + "learning_rate": 3.200723306974418e-06, + "loss": 1.2244, + "step": 10308 + }, + { + "epoch": 0.9217632331902719, + "grad_norm": 0.7152358044911609, + "learning_rate": 3.1934590132973283e-06, + "loss": 1.8268, + "step": 10309 + }, + { + "epoch": 0.9218526466380543, + "grad_norm": 0.4307000460460902, + "learning_rate": 3.186202838667951e-06, + "loss": 1.4947, + "step": 10310 + }, + { + "epoch": 0.9219420600858369, + "grad_norm": 0.43316452577025444, + "learning_rate": 3.1789547836947986e-06, + "loss": 1.6439, + "step": 10311 + }, + { + "epoch": 0.9220314735336195, + "grad_norm": 0.38722057042737507, + "learning_rate": 3.171714848985785e-06, + "loss": 1.3726, + "step": 10312 + }, + { + "epoch": 0.922120886981402, + "grad_norm": 0.7059444865497068, + "learning_rate": 3.164483035148114e-06, + "loss": 1.6057, + "step": 10313 + }, + { + "epoch": 0.9222103004291845, + "grad_norm": 0.5673415504655204, + "learning_rate": 3.157259342788299e-06, + "loss": 1.7784, + "step": 10314 + }, + { + "epoch": 0.9222997138769671, + "grad_norm": 0.5717825196602958, + "learning_rate": 3.150043772512179e-06, + "loss": 1.6153, + "step": 10315 + }, + { + "epoch": 0.9223891273247496, + "grad_norm": 0.3079969582242839, + "learning_rate": 3.1428363249249247e-06, + "loss": 1.3157, + "step": 10316 + }, + { + "epoch": 0.9224785407725322, + "grad_norm": 0.45576551465520865, + "learning_rate": 3.1356370006310197e-06, + "loss": 1.6773, + "step": 10317 + }, + { + "epoch": 0.9225679542203148, + "grad_norm": 0.46211566521978636, + "learning_rate": 3.1284458002342475e-06, + "loss": 1.2797, + "step": 10318 + }, + { + "epoch": 0.9226573676680973, + "grad_norm": 0.5723651471996088, + "learning_rate": 3.121262724337748e-06, + "loss": 1.6188, + "step": 10319 + }, + { + "epoch": 0.9227467811158798, + "grad_norm": 0.5390453483621378, + "learning_rate": 3.1140877735439387e-06, + "loss": 1.3002, + "step": 10320 + }, + { + "epoch": 0.9228361945636624, + "grad_norm": 0.5196930304823886, + "learning_rate": 3.1069209484545725e-06, + "loss": 1.942, + "step": 10321 + }, + { + "epoch": 0.922925608011445, + "grad_norm": 0.5286414014214107, + "learning_rate": 3.0997622496707456e-06, + "loss": 1.6592, + "step": 10322 + }, + { + "epoch": 0.9230150214592274, + "grad_norm": 0.6618427406798264, + "learning_rate": 3.0926116777928116e-06, + "loss": 1.6587, + "step": 10323 + }, + { + "epoch": 0.92310443490701, + "grad_norm": 0.5948328345138243, + "learning_rate": 3.0854692334205125e-06, + "loss": 1.6903, + "step": 10324 + }, + { + "epoch": 0.9231938483547926, + "grad_norm": 0.6650197349573387, + "learning_rate": 3.0783349171528697e-06, + "loss": 1.4809, + "step": 10325 + }, + { + "epoch": 0.9232832618025751, + "grad_norm": 0.5014196385126195, + "learning_rate": 3.0712087295882154e-06, + "loss": 1.6064, + "step": 10326 + }, + { + "epoch": 0.9233726752503576, + "grad_norm": 0.9234995287517188, + "learning_rate": 3.064090671324238e-06, + "loss": 1.6869, + "step": 10327 + }, + { + "epoch": 0.9234620886981402, + "grad_norm": 0.4844437500200469, + "learning_rate": 3.0569807429579044e-06, + "loss": 1.9347, + "step": 10328 + }, + { + "epoch": 0.9235515021459227, + "grad_norm": 0.48078214163299426, + "learning_rate": 3.0498789450855046e-06, + "loss": 1.4106, + "step": 10329 + }, + { + "epoch": 0.9236409155937053, + "grad_norm": 0.8063086438626385, + "learning_rate": 3.0427852783026843e-06, + "loss": 5.989, + "step": 10330 + }, + { + "epoch": 0.9237303290414879, + "grad_norm": 0.48462777195158535, + "learning_rate": 3.0356997432043565e-06, + "loss": 1.8394, + "step": 10331 + }, + { + "epoch": 0.9238197424892703, + "grad_norm": 0.5332230587645532, + "learning_rate": 3.0286223403848014e-06, + "loss": 2.3868, + "step": 10332 + }, + { + "epoch": 0.9239091559370529, + "grad_norm": 0.5533136490462285, + "learning_rate": 3.021553070437577e-06, + "loss": 1.4837, + "step": 10333 + }, + { + "epoch": 0.9239985693848355, + "grad_norm": 0.6107484981755926, + "learning_rate": 3.0144919339555654e-06, + "loss": 1.2384, + "step": 10334 + }, + { + "epoch": 0.9240879828326181, + "grad_norm": 0.7792144540013869, + "learning_rate": 3.0074389315309928e-06, + "loss": 1.4759, + "step": 10335 + }, + { + "epoch": 0.9241773962804005, + "grad_norm": 0.3658258426605809, + "learning_rate": 3.000394063755396e-06, + "loss": 1.4268, + "step": 10336 + }, + { + "epoch": 0.9242668097281831, + "grad_norm": 0.6954027061095535, + "learning_rate": 2.9933573312195708e-06, + "loss": 1.7342, + "step": 10337 + }, + { + "epoch": 0.9243562231759657, + "grad_norm": 0.4741215307943145, + "learning_rate": 2.9863287345137216e-06, + "loss": 1.3631, + "step": 10338 + }, + { + "epoch": 0.9244456366237482, + "grad_norm": 0.776725710486551, + "learning_rate": 2.979308274227344e-06, + "loss": 1.5192, + "step": 10339 + }, + { + "epoch": 0.9245350500715308, + "grad_norm": 0.35777666138290837, + "learning_rate": 2.9722959509491888e-06, + "loss": 1.5311, + "step": 10340 + }, + { + "epoch": 0.9246244635193133, + "grad_norm": 0.419255183913999, + "learning_rate": 2.965291765267386e-06, + "loss": 1.7058, + "step": 10341 + }, + { + "epoch": 0.9247138769670958, + "grad_norm": 0.3978228072232694, + "learning_rate": 2.958295717769399e-06, + "loss": 1.5359, + "step": 10342 + }, + { + "epoch": 0.9248032904148784, + "grad_norm": 0.35020439420629795, + "learning_rate": 2.9513078090419365e-06, + "loss": 1.3389, + "step": 10343 + }, + { + "epoch": 0.924892703862661, + "grad_norm": 0.7147240726921387, + "learning_rate": 2.944328039671085e-06, + "loss": 1.9017, + "step": 10344 + }, + { + "epoch": 0.9249821173104434, + "grad_norm": 0.5285325760606838, + "learning_rate": 2.93735641024222e-06, + "loss": 1.4146, + "step": 10345 + }, + { + "epoch": 0.925071530758226, + "grad_norm": 0.48465541969749854, + "learning_rate": 2.930392921340053e-06, + "loss": 2.0503, + "step": 10346 + }, + { + "epoch": 0.9251609442060086, + "grad_norm": 0.40302467495347866, + "learning_rate": 2.9234375735486153e-06, + "loss": 1.5013, + "step": 10347 + }, + { + "epoch": 0.9252503576537912, + "grad_norm": 0.7505511186985279, + "learning_rate": 2.916490367451219e-06, + "loss": 1.799, + "step": 10348 + }, + { + "epoch": 0.9253397711015737, + "grad_norm": 0.3644271139733947, + "learning_rate": 2.909551303630531e-06, + "loss": 1.4423, + "step": 10349 + }, + { + "epoch": 0.9254291845493562, + "grad_norm": 0.5425002103144949, + "learning_rate": 2.9026203826685195e-06, + "loss": 1.4849, + "step": 10350 + }, + { + "epoch": 0.9255185979971388, + "grad_norm": 0.7087047948016014, + "learning_rate": 2.8956976051464636e-06, + "loss": 1.5288, + "step": 10351 + }, + { + "epoch": 0.9256080114449213, + "grad_norm": 0.7798259807780753, + "learning_rate": 2.8887829716449876e-06, + "loss": 1.6841, + "step": 10352 + }, + { + "epoch": 0.9256974248927039, + "grad_norm": 0.552120974384933, + "learning_rate": 2.8818764827440057e-06, + "loss": 1.7616, + "step": 10353 + }, + { + "epoch": 0.9257868383404864, + "grad_norm": 0.42349850133653666, + "learning_rate": 2.8749781390227437e-06, + "loss": 1.4909, + "step": 10354 + }, + { + "epoch": 0.9258762517882689, + "grad_norm": 0.4330419027094413, + "learning_rate": 2.8680879410597716e-06, + "loss": 1.7818, + "step": 10355 + }, + { + "epoch": 0.9259656652360515, + "grad_norm": 0.6015163425906425, + "learning_rate": 2.861205889432972e-06, + "loss": 1.6124, + "step": 10356 + }, + { + "epoch": 0.9260550786838341, + "grad_norm": 0.5574739553591188, + "learning_rate": 2.854331984719505e-06, + "loss": 1.6791, + "step": 10357 + }, + { + "epoch": 0.9261444921316166, + "grad_norm": 0.48279733735568375, + "learning_rate": 2.8474662274958987e-06, + "loss": 1.8125, + "step": 10358 + }, + { + "epoch": 0.9262339055793991, + "grad_norm": 0.43925344506009617, + "learning_rate": 2.8406086183379586e-06, + "loss": 1.864, + "step": 10359 + }, + { + "epoch": 0.9263233190271817, + "grad_norm": 0.32379964749726886, + "learning_rate": 2.8337591578208366e-06, + "loss": 1.4501, + "step": 10360 + }, + { + "epoch": 0.9264127324749643, + "grad_norm": 0.6914207163280396, + "learning_rate": 2.826917846518995e-06, + "loss": 1.6414, + "step": 10361 + }, + { + "epoch": 0.9265021459227468, + "grad_norm": 0.41960449089332025, + "learning_rate": 2.820084685006208e-06, + "loss": 1.2428, + "step": 10362 + }, + { + "epoch": 0.9265915593705293, + "grad_norm": 0.4331177422583481, + "learning_rate": 2.8132596738555397e-06, + "loss": 1.3937, + "step": 10363 + }, + { + "epoch": 0.9266809728183119, + "grad_norm": 0.541543965078598, + "learning_rate": 2.8064428136394096e-06, + "loss": 1.9584, + "step": 10364 + }, + { + "epoch": 0.9267703862660944, + "grad_norm": 1.136598523790612, + "learning_rate": 2.799634104929538e-06, + "loss": 6.0487, + "step": 10365 + }, + { + "epoch": 0.926859799713877, + "grad_norm": 0.46722658500867015, + "learning_rate": 2.7928335482969802e-06, + "loss": 1.8297, + "step": 10366 + }, + { + "epoch": 0.9269492131616596, + "grad_norm": 0.49032514034449926, + "learning_rate": 2.7860411443120684e-06, + "loss": 1.5511, + "step": 10367 + }, + { + "epoch": 0.927038626609442, + "grad_norm": 0.3957321450213774, + "learning_rate": 2.7792568935444796e-06, + "loss": 1.5493, + "step": 10368 + }, + { + "epoch": 0.9271280400572246, + "grad_norm": 0.45146269353830915, + "learning_rate": 2.772480796563204e-06, + "loss": 1.362, + "step": 10369 + }, + { + "epoch": 0.9272174535050072, + "grad_norm": 0.6941072226891816, + "learning_rate": 2.765712853936553e-06, + "loss": 1.6622, + "step": 10370 + }, + { + "epoch": 0.9273068669527897, + "grad_norm": 0.37513324156199307, + "learning_rate": 2.7589530662321285e-06, + "loss": 1.3884, + "step": 10371 + }, + { + "epoch": 0.9273962804005722, + "grad_norm": 0.43345984517323033, + "learning_rate": 2.7522014340168547e-06, + "loss": 1.2178, + "step": 10372 + }, + { + "epoch": 0.9274856938483548, + "grad_norm": 0.3789175909238038, + "learning_rate": 2.745457957857023e-06, + "loss": 1.7064, + "step": 10373 + }, + { + "epoch": 0.9275751072961373, + "grad_norm": 0.9023430793810272, + "learning_rate": 2.7387226383181696e-06, + "loss": 6.0324, + "step": 10374 + }, + { + "epoch": 0.9276645207439199, + "grad_norm": 0.5188949847889992, + "learning_rate": 2.7319954759651877e-06, + "loss": 1.6625, + "step": 10375 + }, + { + "epoch": 0.9277539341917024, + "grad_norm": 0.47571523599786886, + "learning_rate": 2.7252764713622814e-06, + "loss": 1.6658, + "step": 10376 + }, + { + "epoch": 0.927843347639485, + "grad_norm": 0.4227782074049056, + "learning_rate": 2.718565625072955e-06, + "loss": 1.4673, + "step": 10377 + }, + { + "epoch": 0.9279327610872675, + "grad_norm": 0.6152309443398528, + "learning_rate": 2.711862937660037e-06, + "loss": 1.6138, + "step": 10378 + }, + { + "epoch": 0.9280221745350501, + "grad_norm": 0.7201737845647407, + "learning_rate": 2.7051684096856876e-06, + "loss": 1.6716, + "step": 10379 + }, + { + "epoch": 0.9281115879828327, + "grad_norm": 0.4482874564402255, + "learning_rate": 2.6984820417113587e-06, + "loss": 1.5098, + "step": 10380 + }, + { + "epoch": 0.9282010014306151, + "grad_norm": 0.5189357893948207, + "learning_rate": 2.6918038342978345e-06, + "loss": 1.2933, + "step": 10381 + }, + { + "epoch": 0.9282904148783977, + "grad_norm": 0.3536213750685428, + "learning_rate": 2.685133788005201e-06, + "loss": 1.672, + "step": 10382 + }, + { + "epoch": 0.9283798283261803, + "grad_norm": 0.47082117641797067, + "learning_rate": 2.678471903392865e-06, + "loss": 1.779, + "step": 10383 + }, + { + "epoch": 0.9284692417739628, + "grad_norm": 0.4839796252339522, + "learning_rate": 2.6718181810195696e-06, + "loss": 1.5766, + "step": 10384 + }, + { + "epoch": 0.9285586552217453, + "grad_norm": 0.41331676231745446, + "learning_rate": 2.6651726214433235e-06, + "loss": 1.5513, + "step": 10385 + }, + { + "epoch": 0.9286480686695279, + "grad_norm": 0.5919754353359834, + "learning_rate": 2.6585352252215036e-06, + "loss": 1.5665, + "step": 10386 + }, + { + "epoch": 0.9287374821173104, + "grad_norm": 0.9005730078895169, + "learning_rate": 2.651905992910786e-06, + "loss": 1.7836, + "step": 10387 + }, + { + "epoch": 0.928826895565093, + "grad_norm": 0.47240678164298155, + "learning_rate": 2.6452849250671373e-06, + "loss": 1.7497, + "step": 10388 + }, + { + "epoch": 0.9289163090128756, + "grad_norm": 0.7873024813334345, + "learning_rate": 2.6386720222458693e-06, + "loss": 1.7043, + "step": 10389 + }, + { + "epoch": 0.929005722460658, + "grad_norm": 0.4520306112925746, + "learning_rate": 2.6320672850016047e-06, + "loss": 1.2717, + "step": 10390 + }, + { + "epoch": 0.9290951359084406, + "grad_norm": 0.7317031774583419, + "learning_rate": 2.625470713888256e-06, + "loss": 1.6975, + "step": 10391 + }, + { + "epoch": 0.9291845493562232, + "grad_norm": 0.5437916683835826, + "learning_rate": 2.618882309459081e-06, + "loss": 1.8723, + "step": 10392 + }, + { + "epoch": 0.9292739628040058, + "grad_norm": 0.5788296287811444, + "learning_rate": 2.612302072266637e-06, + "loss": 1.5298, + "step": 10393 + }, + { + "epoch": 0.9293633762517882, + "grad_norm": 0.47228793843714406, + "learning_rate": 2.605730002862805e-06, + "loss": 1.444, + "step": 10394 + }, + { + "epoch": 0.9294527896995708, + "grad_norm": 0.5777578689460007, + "learning_rate": 2.5991661017987777e-06, + "loss": 1.5965, + "step": 10395 + }, + { + "epoch": 0.9295422031473534, + "grad_norm": 0.5484339470645895, + "learning_rate": 2.5926103696250703e-06, + "loss": 1.5175, + "step": 10396 + }, + { + "epoch": 0.9296316165951359, + "grad_norm": 0.34843117944524477, + "learning_rate": 2.586062806891476e-06, + "loss": 1.3985, + "step": 10397 + }, + { + "epoch": 0.9297210300429185, + "grad_norm": 0.43939125589222516, + "learning_rate": 2.5795234141471445e-06, + "loss": 1.814, + "step": 10398 + }, + { + "epoch": 0.929810443490701, + "grad_norm": 0.5351378020586924, + "learning_rate": 2.5729921919405377e-06, + "loss": 1.7642, + "step": 10399 + }, + { + "epoch": 0.9298998569384835, + "grad_norm": 0.5335143254059883, + "learning_rate": 2.5664691408194165e-06, + "loss": 1.9967, + "step": 10400 + }, + { + "epoch": 0.9299892703862661, + "grad_norm": 0.6067258419626347, + "learning_rate": 2.559954261330866e-06, + "loss": 1.4802, + "step": 10401 + }, + { + "epoch": 0.9300786838340487, + "grad_norm": 0.5987291756631589, + "learning_rate": 2.55344755402126e-06, + "loss": 1.4718, + "step": 10402 + }, + { + "epoch": 0.9301680972818311, + "grad_norm": 0.6046977338641717, + "learning_rate": 2.546949019436329e-06, + "loss": 1.7852, + "step": 10403 + }, + { + "epoch": 0.9302575107296137, + "grad_norm": 0.35378906154840484, + "learning_rate": 2.540458658121092e-06, + "loss": 1.4221, + "step": 10404 + }, + { + "epoch": 0.9303469241773963, + "grad_norm": 0.6678726920833421, + "learning_rate": 2.533976470619881e-06, + "loss": 1.5402, + "step": 10405 + }, + { + "epoch": 0.9304363376251789, + "grad_norm": 0.41378510634390886, + "learning_rate": 2.5275024574763496e-06, + "loss": 1.6623, + "step": 10406 + }, + { + "epoch": 0.9305257510729614, + "grad_norm": 0.5671077648516126, + "learning_rate": 2.5210366192334745e-06, + "loss": 1.9775, + "step": 10407 + }, + { + "epoch": 0.9306151645207439, + "grad_norm": 0.6396147285347724, + "learning_rate": 2.514578956433533e-06, + "loss": 1.8313, + "step": 10408 + }, + { + "epoch": 0.9307045779685265, + "grad_norm": 0.5134778172651616, + "learning_rate": 2.5081294696181255e-06, + "loss": 1.7752, + "step": 10409 + }, + { + "epoch": 0.930793991416309, + "grad_norm": 0.45765154595471835, + "learning_rate": 2.501688159328164e-06, + "loss": 1.3779, + "step": 10410 + }, + { + "epoch": 0.9308834048640916, + "grad_norm": 0.5733804540379512, + "learning_rate": 2.49525502610386e-06, + "loss": 1.3787, + "step": 10411 + }, + { + "epoch": 0.9309728183118741, + "grad_norm": 0.5210710644329355, + "learning_rate": 2.48883007048476e-06, + "loss": 1.4871, + "step": 10412 + }, + { + "epoch": 0.9310622317596566, + "grad_norm": 0.44122322071611075, + "learning_rate": 2.4824132930097222e-06, + "loss": 1.5985, + "step": 10413 + }, + { + "epoch": 0.9311516452074392, + "grad_norm": 0.5633655080409624, + "learning_rate": 2.4760046942169048e-06, + "loss": 1.4208, + "step": 10414 + }, + { + "epoch": 0.9312410586552218, + "grad_norm": 0.45165522528187374, + "learning_rate": 2.4696042746438108e-06, + "loss": 1.7571, + "step": 10415 + }, + { + "epoch": 0.9313304721030042, + "grad_norm": 0.41550479676832247, + "learning_rate": 2.4632120348272003e-06, + "loss": 1.6712, + "step": 10416 + }, + { + "epoch": 0.9314198855507868, + "grad_norm": 0.6312685448159724, + "learning_rate": 2.456827975303211e-06, + "loss": 1.4676, + "step": 10417 + }, + { + "epoch": 0.9315092989985694, + "grad_norm": 0.3606053558626124, + "learning_rate": 2.4504520966072476e-06, + "loss": 1.5953, + "step": 10418 + }, + { + "epoch": 0.931598712446352, + "grad_norm": 0.4666713573908732, + "learning_rate": 2.4440843992740714e-06, + "loss": 1.4297, + "step": 10419 + }, + { + "epoch": 0.9316881258941345, + "grad_norm": 0.4104507223129851, + "learning_rate": 2.4377248838376996e-06, + "loss": 1.2537, + "step": 10420 + }, + { + "epoch": 0.931777539341917, + "grad_norm": 0.31980313258919657, + "learning_rate": 2.4313735508315396e-06, + "loss": 1.4707, + "step": 10421 + }, + { + "epoch": 0.9318669527896996, + "grad_norm": 0.6621393487770424, + "learning_rate": 2.425030400788231e-06, + "loss": 1.691, + "step": 10422 + }, + { + "epoch": 0.9319563662374821, + "grad_norm": 0.500652557610808, + "learning_rate": 2.4186954342397815e-06, + "loss": 1.6069, + "step": 10423 + }, + { + "epoch": 0.9320457796852647, + "grad_norm": 0.8105798481976111, + "learning_rate": 2.4123686517175113e-06, + "loss": 1.5164, + "step": 10424 + }, + { + "epoch": 0.9321351931330472, + "grad_norm": 0.5442688120411582, + "learning_rate": 2.406050053752018e-06, + "loss": 1.5472, + "step": 10425 + }, + { + "epoch": 0.9322246065808297, + "grad_norm": 0.43871643510480823, + "learning_rate": 2.3997396408732443e-06, + "loss": 1.3644, + "step": 10426 + }, + { + "epoch": 0.9323140200286123, + "grad_norm": 0.41296402422235323, + "learning_rate": 2.3934374136104222e-06, + "loss": 1.3906, + "step": 10427 + }, + { + "epoch": 0.9324034334763949, + "grad_norm": 0.6931952017210843, + "learning_rate": 2.38714337249214e-06, + "loss": 1.3111, + "step": 10428 + }, + { + "epoch": 0.9324928469241774, + "grad_norm": 0.5729123380831004, + "learning_rate": 2.3808575180462533e-06, + "loss": 1.8929, + "step": 10429 + }, + { + "epoch": 0.9325822603719599, + "grad_norm": 1.8579441576986275, + "learning_rate": 2.374579850799963e-06, + "loss": 6.0151, + "step": 10430 + }, + { + "epoch": 0.9326716738197425, + "grad_norm": 0.3540684345929835, + "learning_rate": 2.3683103712797473e-06, + "loss": 1.6073, + "step": 10431 + }, + { + "epoch": 0.932761087267525, + "grad_norm": 0.6399676587271617, + "learning_rate": 2.3620490800114304e-06, + "loss": 1.5905, + "step": 10432 + }, + { + "epoch": 0.9328505007153076, + "grad_norm": 0.5761565015108012, + "learning_rate": 2.3557959775201478e-06, + "loss": 1.657, + "step": 10433 + }, + { + "epoch": 0.9329399141630901, + "grad_norm": 0.5146400326485254, + "learning_rate": 2.349551064330313e-06, + "loss": 1.4907, + "step": 10434 + }, + { + "epoch": 0.9330293276108726, + "grad_norm": 0.3672708306059257, + "learning_rate": 2.3433143409657188e-06, + "loss": 1.3649, + "step": 10435 + }, + { + "epoch": 0.9331187410586552, + "grad_norm": 0.6965559790708113, + "learning_rate": 2.337085807949413e-06, + "loss": 6.1193, + "step": 10436 + }, + { + "epoch": 0.9332081545064378, + "grad_norm": 0.6209284216305195, + "learning_rate": 2.3308654658037555e-06, + "loss": 1.6555, + "step": 10437 + }, + { + "epoch": 0.9332975679542204, + "grad_norm": 0.4856959982373667, + "learning_rate": 2.3246533150504735e-06, + "loss": 1.1785, + "step": 10438 + }, + { + "epoch": 0.9333869814020028, + "grad_norm": 0.5898937427229808, + "learning_rate": 2.3184493562105504e-06, + "loss": 1.4853, + "step": 10439 + }, + { + "epoch": 0.9334763948497854, + "grad_norm": 0.6226211509239952, + "learning_rate": 2.312253589804314e-06, + "loss": 1.6486, + "step": 10440 + }, + { + "epoch": 0.933565808297568, + "grad_norm": 0.7007963376902433, + "learning_rate": 2.3060660163513825e-06, + "loss": 1.8064, + "step": 10441 + }, + { + "epoch": 0.9336552217453505, + "grad_norm": 0.6305806563261009, + "learning_rate": 2.2998866363707184e-06, + "loss": 1.6917, + "step": 10442 + }, + { + "epoch": 0.933744635193133, + "grad_norm": 0.5900306319148718, + "learning_rate": 2.2937154503805623e-06, + "loss": 1.6776, + "step": 10443 + }, + { + "epoch": 0.9338340486409156, + "grad_norm": 0.5770088674220688, + "learning_rate": 2.287552458898501e-06, + "loss": 1.4038, + "step": 10444 + }, + { + "epoch": 0.9339234620886981, + "grad_norm": 0.5931204916994611, + "learning_rate": 2.2813976624414093e-06, + "loss": 1.5575, + "step": 10445 + }, + { + "epoch": 0.9340128755364807, + "grad_norm": 0.7137863737842046, + "learning_rate": 2.275251061525474e-06, + "loss": 2.0455, + "step": 10446 + }, + { + "epoch": 0.9341022889842633, + "grad_norm": 0.535235588230517, + "learning_rate": 2.269112656666217e-06, + "loss": 1.5925, + "step": 10447 + }, + { + "epoch": 0.9341917024320457, + "grad_norm": 0.5148790831915604, + "learning_rate": 2.2629824483784366e-06, + "loss": 1.7325, + "step": 10448 + }, + { + "epoch": 0.9342811158798283, + "grad_norm": 0.7770685951837437, + "learning_rate": 2.2568604371763e-06, + "loss": 1.9021, + "step": 10449 + }, + { + "epoch": 0.9343705293276109, + "grad_norm": 0.47147431396101175, + "learning_rate": 2.25074662357323e-06, + "loss": 1.4598, + "step": 10450 + }, + { + "epoch": 0.9344599427753935, + "grad_norm": 0.41165337369877536, + "learning_rate": 2.2446410080819824e-06, + "loss": 1.3685, + "step": 10451 + }, + { + "epoch": 0.9345493562231759, + "grad_norm": 0.5358896496686902, + "learning_rate": 2.238543591214637e-06, + "loss": 1.5026, + "step": 10452 + }, + { + "epoch": 0.9346387696709585, + "grad_norm": 0.7183381400909677, + "learning_rate": 2.232454373482584e-06, + "loss": 1.6032, + "step": 10453 + }, + { + "epoch": 0.9347281831187411, + "grad_norm": 0.3070177543063629, + "learning_rate": 2.226373355396505e-06, + "loss": 1.267, + "step": 10454 + }, + { + "epoch": 0.9348175965665236, + "grad_norm": 1.7192645802258912, + "learning_rate": 2.220300537466413e-06, + "loss": 6.0721, + "step": 10455 + }, + { + "epoch": 0.9349070100143062, + "grad_norm": 0.43055869102036076, + "learning_rate": 2.2142359202016237e-06, + "loss": 1.2468, + "step": 10456 + }, + { + "epoch": 0.9349964234620887, + "grad_norm": 0.33106671197686255, + "learning_rate": 2.208179504110763e-06, + "loss": 1.2421, + "step": 10457 + }, + { + "epoch": 0.9350858369098712, + "grad_norm": 0.6148185583354202, + "learning_rate": 2.202131289701803e-06, + "loss": 1.8493, + "step": 10458 + }, + { + "epoch": 0.9351752503576538, + "grad_norm": 0.33576043766852887, + "learning_rate": 2.1960912774819707e-06, + "loss": 1.3989, + "step": 10459 + }, + { + "epoch": 0.9352646638054364, + "grad_norm": 0.44372005565536066, + "learning_rate": 2.1900594679578503e-06, + "loss": 1.7212, + "step": 10460 + }, + { + "epoch": 0.9353540772532188, + "grad_norm": 0.4657151188902587, + "learning_rate": 2.1840358616353252e-06, + "loss": 1.3092, + "step": 10461 + }, + { + "epoch": 0.9354434907010014, + "grad_norm": 0.45793795316322183, + "learning_rate": 2.1780204590195583e-06, + "loss": 1.547, + "step": 10462 + }, + { + "epoch": 0.935532904148784, + "grad_norm": 0.6631550200634936, + "learning_rate": 2.172013260615091e-06, + "loss": 1.7445, + "step": 10463 + }, + { + "epoch": 0.9356223175965666, + "grad_norm": 0.35020069006721816, + "learning_rate": 2.166014266925731e-06, + "loss": 1.3158, + "step": 10464 + }, + { + "epoch": 0.935711731044349, + "grad_norm": 0.4639139304571572, + "learning_rate": 2.160023478454587e-06, + "loss": 1.6182, + "step": 10465 + }, + { + "epoch": 0.9358011444921316, + "grad_norm": 0.7708292944107539, + "learning_rate": 2.1540408957041235e-06, + "loss": 1.7078, + "step": 10466 + }, + { + "epoch": 0.9358905579399142, + "grad_norm": 0.44611864213095725, + "learning_rate": 2.148066519176084e-06, + "loss": 1.6713, + "step": 10467 + }, + { + "epoch": 0.9359799713876967, + "grad_norm": 0.5357751150608958, + "learning_rate": 2.142100349371512e-06, + "loss": 1.6236, + "step": 10468 + }, + { + "epoch": 0.9360693848354793, + "grad_norm": 0.5336397561405495, + "learning_rate": 2.1361423867908293e-06, + "loss": 6.1176, + "step": 10469 + }, + { + "epoch": 0.9361587982832618, + "grad_norm": 0.6689785248356097, + "learning_rate": 2.1301926319336696e-06, + "loss": 2.0077, + "step": 10470 + }, + { + "epoch": 0.9362482117310443, + "grad_norm": 0.48244197594753396, + "learning_rate": 2.124251085299067e-06, + "loss": 1.5899, + "step": 10471 + }, + { + "epoch": 0.9363376251788269, + "grad_norm": 0.603196447952397, + "learning_rate": 2.1183177473853346e-06, + "loss": 1.6741, + "step": 10472 + }, + { + "epoch": 0.9364270386266095, + "grad_norm": 0.5338452603422855, + "learning_rate": 2.112392618690062e-06, + "loss": 1.5462, + "step": 10473 + }, + { + "epoch": 0.9365164520743919, + "grad_norm": 0.4384510171714179, + "learning_rate": 2.1064756997102084e-06, + "loss": 1.6257, + "step": 10474 + }, + { + "epoch": 0.9366058655221745, + "grad_norm": 0.6350895365781384, + "learning_rate": 2.10056699094201e-06, + "loss": 1.4774, + "step": 10475 + }, + { + "epoch": 0.9366952789699571, + "grad_norm": 0.5372155594397214, + "learning_rate": 2.0946664928810367e-06, + "loss": 1.8277, + "step": 10476 + }, + { + "epoch": 0.9367846924177397, + "grad_norm": 0.8900957695033374, + "learning_rate": 2.0887742060221262e-06, + "loss": 1.603, + "step": 10477 + }, + { + "epoch": 0.9368741058655222, + "grad_norm": 0.4480022210532913, + "learning_rate": 2.082890130859505e-06, + "loss": 1.6125, + "step": 10478 + }, + { + "epoch": 0.9369635193133047, + "grad_norm": 0.5718103913192931, + "learning_rate": 2.077014267886612e-06, + "loss": 1.5553, + "step": 10479 + }, + { + "epoch": 0.9370529327610873, + "grad_norm": 0.42072482808605627, + "learning_rate": 2.0711466175962756e-06, + "loss": 1.48, + "step": 10480 + }, + { + "epoch": 0.9371423462088698, + "grad_norm": 0.784687242875288, + "learning_rate": 2.065287180480613e-06, + "loss": 1.7616, + "step": 10481 + }, + { + "epoch": 0.9372317596566524, + "grad_norm": 0.5777151029748263, + "learning_rate": 2.0594359570310196e-06, + "loss": 1.8469, + "step": 10482 + }, + { + "epoch": 0.9373211731044349, + "grad_norm": 0.3821366499046855, + "learning_rate": 2.0535929477382587e-06, + "loss": 1.4876, + "step": 10483 + }, + { + "epoch": 0.9374105865522174, + "grad_norm": 0.47218893100270454, + "learning_rate": 2.0477581530923717e-06, + "loss": 1.7274, + "step": 10484 + }, + { + "epoch": 0.9375, + "grad_norm": 0.5208445565489963, + "learning_rate": 2.0419315735827116e-06, + "loss": 1.4292, + "step": 10485 + }, + { + "epoch": 0.9375894134477826, + "grad_norm": 1.0322147287834498, + "learning_rate": 2.036113209697943e-06, + "loss": 1.7973, + "step": 10486 + }, + { + "epoch": 0.9376788268955651, + "grad_norm": 0.34501004300632326, + "learning_rate": 2.0303030619260644e-06, + "loss": 1.332, + "step": 10487 + }, + { + "epoch": 0.9377682403433476, + "grad_norm": 0.43154141485400066, + "learning_rate": 2.0245011307543416e-06, + "loss": 1.4364, + "step": 10488 + }, + { + "epoch": 0.9378576537911302, + "grad_norm": 0.9001802010137723, + "learning_rate": 2.018707416669374e-06, + "loss": 1.7636, + "step": 10489 + }, + { + "epoch": 0.9379470672389127, + "grad_norm": 0.5789791600831619, + "learning_rate": 2.012921920157096e-06, + "loss": 1.7339, + "step": 10490 + }, + { + "epoch": 0.9380364806866953, + "grad_norm": 0.396411682783821, + "learning_rate": 2.0071446417027073e-06, + "loss": 1.5522, + "step": 10491 + }, + { + "epoch": 0.9381258941344778, + "grad_norm": 0.5576598306340489, + "learning_rate": 2.0013755817907652e-06, + "loss": 1.9302, + "step": 10492 + }, + { + "epoch": 0.9382153075822603, + "grad_norm": 0.4169384732212549, + "learning_rate": 1.995614740905094e-06, + "loss": 1.4387, + "step": 10493 + }, + { + "epoch": 0.9383047210300429, + "grad_norm": 0.4554498559439292, + "learning_rate": 1.9898621195288515e-06, + "loss": 1.4238, + "step": 10494 + }, + { + "epoch": 0.9383941344778255, + "grad_norm": 0.4012193745147053, + "learning_rate": 1.984117718144518e-06, + "loss": 1.4637, + "step": 10495 + }, + { + "epoch": 0.9384835479256081, + "grad_norm": 0.38613120518671284, + "learning_rate": 1.9783815372338423e-06, + "loss": 1.6853, + "step": 10496 + }, + { + "epoch": 0.9385729613733905, + "grad_norm": 0.6306115066333433, + "learning_rate": 1.972653577277939e-06, + "loss": 1.596, + "step": 10497 + }, + { + "epoch": 0.9386623748211731, + "grad_norm": 0.7055636274824139, + "learning_rate": 1.96693383875719e-06, + "loss": 1.4728, + "step": 10498 + }, + { + "epoch": 0.9387517882689557, + "grad_norm": 0.8304341651679696, + "learning_rate": 1.9612223221513125e-06, + "loss": 1.7407, + "step": 10499 + }, + { + "epoch": 0.9388412017167382, + "grad_norm": 0.3770320124276049, + "learning_rate": 1.955519027939301e-06, + "loss": 1.552, + "step": 10500 + }, + { + "epoch": 0.9389306151645207, + "grad_norm": 0.7488891765696445, + "learning_rate": 1.949823956599528e-06, + "loss": 1.9038, + "step": 10501 + }, + { + "epoch": 0.9390200286123033, + "grad_norm": 0.631042439870792, + "learning_rate": 1.9441371086095784e-06, + "loss": 1.6498, + "step": 10502 + }, + { + "epoch": 0.9391094420600858, + "grad_norm": 0.4407003292927486, + "learning_rate": 1.938458484446437e-06, + "loss": 1.8646, + "step": 10503 + }, + { + "epoch": 0.9391988555078684, + "grad_norm": 0.4013381937295077, + "learning_rate": 1.9327880845863568e-06, + "loss": 1.7366, + "step": 10504 + }, + { + "epoch": 0.939288268955651, + "grad_norm": 0.443058555280905, + "learning_rate": 1.927125909504901e-06, + "loss": 1.8528, + "step": 10505 + }, + { + "epoch": 0.9393776824034334, + "grad_norm": 0.6492690008233709, + "learning_rate": 1.921471959676957e-06, + "loss": 1.3763, + "step": 10506 + }, + { + "epoch": 0.939467095851216, + "grad_norm": 0.5539268263604907, + "learning_rate": 1.915826235576712e-06, + "loss": 1.2891, + "step": 10507 + }, + { + "epoch": 0.9395565092989986, + "grad_norm": 0.5993260324951438, + "learning_rate": 1.910188737677665e-06, + "loss": 1.6186, + "step": 10508 + }, + { + "epoch": 0.9396459227467812, + "grad_norm": 0.580784384363062, + "learning_rate": 1.9045594664526155e-06, + "loss": 1.5233, + "step": 10509 + }, + { + "epoch": 0.9397353361945636, + "grad_norm": 0.6938043054932878, + "learning_rate": 1.8989384223736971e-06, + "loss": 1.5226, + "step": 10510 + }, + { + "epoch": 0.9398247496423462, + "grad_norm": 0.5683300770769171, + "learning_rate": 1.8933256059123438e-06, + "loss": 1.7999, + "step": 10511 + }, + { + "epoch": 0.9399141630901288, + "grad_norm": 0.39437484388195365, + "learning_rate": 1.88772101753929e-06, + "loss": 1.7265, + "step": 10512 + }, + { + "epoch": 0.9400035765379113, + "grad_norm": 0.3424648211155291, + "learning_rate": 1.8821246577245822e-06, + "loss": 1.2683, + "step": 10513 + }, + { + "epoch": 0.9400929899856938, + "grad_norm": 0.5429791139498739, + "learning_rate": 1.876536526937589e-06, + "loss": 1.8841, + "step": 10514 + }, + { + "epoch": 0.9401824034334764, + "grad_norm": 0.448139613594531, + "learning_rate": 1.8709566256469691e-06, + "loss": 1.3998, + "step": 10515 + }, + { + "epoch": 0.9402718168812589, + "grad_norm": 0.4392245326999892, + "learning_rate": 1.8653849543207036e-06, + "loss": 1.5221, + "step": 10516 + }, + { + "epoch": 0.9403612303290415, + "grad_norm": 0.35860586253257637, + "learning_rate": 1.8598215134260743e-06, + "loss": 1.4037, + "step": 10517 + }, + { + "epoch": 0.9404506437768241, + "grad_norm": 0.5751279606917152, + "learning_rate": 1.8542663034297191e-06, + "loss": 1.5526, + "step": 10518 + }, + { + "epoch": 0.9405400572246065, + "grad_norm": 0.6594173433920926, + "learning_rate": 1.8487193247974989e-06, + "loss": 1.7559, + "step": 10519 + }, + { + "epoch": 0.9406294706723891, + "grad_norm": 0.610259845216015, + "learning_rate": 1.843180577994652e-06, + "loss": 1.4522, + "step": 10520 + }, + { + "epoch": 0.9407188841201717, + "grad_norm": 0.7292320985570464, + "learning_rate": 1.8376500634857296e-06, + "loss": 1.5256, + "step": 10521 + }, + { + "epoch": 0.9408082975679543, + "grad_norm": 0.6779601110782327, + "learning_rate": 1.8321277817345274e-06, + "loss": 1.7217, + "step": 10522 + }, + { + "epoch": 0.9408977110157367, + "grad_norm": 0.46153216516397555, + "learning_rate": 1.8266137332042077e-06, + "loss": 1.5907, + "step": 10523 + }, + { + "epoch": 0.9409871244635193, + "grad_norm": 0.43714685646516377, + "learning_rate": 1.8211079183572344e-06, + "loss": 1.8628, + "step": 10524 + }, + { + "epoch": 0.9410765379113019, + "grad_norm": 0.5765961216907435, + "learning_rate": 1.8156103376553714e-06, + "loss": 1.6367, + "step": 10525 + }, + { + "epoch": 0.9411659513590844, + "grad_norm": 0.6068335660989285, + "learning_rate": 1.810120991559694e-06, + "loss": 1.451, + "step": 10526 + }, + { + "epoch": 0.941255364806867, + "grad_norm": 0.4467633709676807, + "learning_rate": 1.8046398805305898e-06, + "loss": 1.4354, + "step": 10527 + }, + { + "epoch": 0.9413447782546495, + "grad_norm": 0.7012363009357095, + "learning_rate": 1.7991670050277354e-06, + "loss": 1.633, + "step": 10528 + }, + { + "epoch": 0.941434191702432, + "grad_norm": 0.6038103716602644, + "learning_rate": 1.7937023655101636e-06, + "loss": 1.6195, + "step": 10529 + }, + { + "epoch": 0.9415236051502146, + "grad_norm": 0.35770426670029354, + "learning_rate": 1.7882459624361637e-06, + "loss": 1.3078, + "step": 10530 + }, + { + "epoch": 0.9416130185979972, + "grad_norm": 0.5088946248947341, + "learning_rate": 1.782797796263358e-06, + "loss": 1.8277, + "step": 10531 + }, + { + "epoch": 0.9417024320457796, + "grad_norm": 0.4772096975900439, + "learning_rate": 1.7773578674486923e-06, + "loss": 1.4025, + "step": 10532 + }, + { + "epoch": 0.9417918454935622, + "grad_norm": 0.6112226692685511, + "learning_rate": 1.7719261764484019e-06, + "loss": 1.7505, + "step": 10533 + }, + { + "epoch": 0.9418812589413448, + "grad_norm": 0.558700804368384, + "learning_rate": 1.7665027237180332e-06, + "loss": 1.9726, + "step": 10534 + }, + { + "epoch": 0.9419706723891274, + "grad_norm": 0.6644918207571843, + "learning_rate": 1.7610875097124446e-06, + "loss": 1.3072, + "step": 10535 + }, + { + "epoch": 0.9420600858369099, + "grad_norm": 0.4974292606095015, + "learning_rate": 1.7556805348858064e-06, + "loss": 1.9117, + "step": 10536 + }, + { + "epoch": 0.9421494992846924, + "grad_norm": 0.48715128632058147, + "learning_rate": 1.7502817996915778e-06, + "loss": 1.8763, + "step": 10537 + }, + { + "epoch": 0.942238912732475, + "grad_norm": 0.5017917623506457, + "learning_rate": 1.7448913045825742e-06, + "loss": 1.7652, + "step": 10538 + }, + { + "epoch": 0.9423283261802575, + "grad_norm": 0.4783530567684918, + "learning_rate": 1.739509050010868e-06, + "loss": 1.292, + "step": 10539 + }, + { + "epoch": 0.9424177396280401, + "grad_norm": 0.35307495104962117, + "learning_rate": 1.7341350364278642e-06, + "loss": 1.3963, + "step": 10540 + }, + { + "epoch": 0.9425071530758226, + "grad_norm": 0.43519159441312916, + "learning_rate": 1.7287692642842911e-06, + "loss": 1.6387, + "step": 10541 + }, + { + "epoch": 0.9425965665236051, + "grad_norm": 0.3944367907393476, + "learning_rate": 1.723411734030156e-06, + "loss": 1.3925, + "step": 10542 + }, + { + "epoch": 0.9426859799713877, + "grad_norm": 0.6914178576151122, + "learning_rate": 1.7180624461147876e-06, + "loss": 1.7788, + "step": 10543 + }, + { + "epoch": 0.9427753934191703, + "grad_norm": 0.5220071699980922, + "learning_rate": 1.7127214009868385e-06, + "loss": 1.7358, + "step": 10544 + }, + { + "epoch": 0.9428648068669528, + "grad_norm": 0.41340423962241896, + "learning_rate": 1.7073885990942174e-06, + "loss": 1.4992, + "step": 10545 + }, + { + "epoch": 0.9429542203147353, + "grad_norm": 0.32303336137537, + "learning_rate": 1.7020640408842325e-06, + "loss": 1.4487, + "step": 10546 + }, + { + "epoch": 0.9430436337625179, + "grad_norm": 0.674637561930435, + "learning_rate": 1.696747726803416e-06, + "loss": 1.6066, + "step": 10547 + }, + { + "epoch": 0.9431330472103004, + "grad_norm": 0.6397221299897912, + "learning_rate": 1.6914396572976444e-06, + "loss": 1.8092, + "step": 10548 + }, + { + "epoch": 0.943222460658083, + "grad_norm": 0.9779925515775153, + "learning_rate": 1.6861398328121059e-06, + "loss": 1.6696, + "step": 10549 + }, + { + "epoch": 0.9433118741058655, + "grad_norm": 0.5710809044941777, + "learning_rate": 1.6808482537912896e-06, + "loss": 1.6916, + "step": 10550 + }, + { + "epoch": 0.943401287553648, + "grad_norm": 0.4028068812867952, + "learning_rate": 1.6755649206789737e-06, + "loss": 1.3949, + "step": 10551 + }, + { + "epoch": 0.9434907010014306, + "grad_norm": 0.6209542619972693, + "learning_rate": 1.6702898339182925e-06, + "loss": 1.5246, + "step": 10552 + }, + { + "epoch": 0.9435801144492132, + "grad_norm": 0.44004100462546464, + "learning_rate": 1.6650229939516593e-06, + "loss": 1.5649, + "step": 10553 + }, + { + "epoch": 0.9436695278969958, + "grad_norm": 0.34832379828101234, + "learning_rate": 1.6597644012207759e-06, + "loss": 1.1191, + "step": 10554 + }, + { + "epoch": 0.9437589413447782, + "grad_norm": 0.4069523753247756, + "learning_rate": 1.6545140561667005e-06, + "loss": 1.4194, + "step": 10555 + }, + { + "epoch": 0.9438483547925608, + "grad_norm": 0.6909491026393362, + "learning_rate": 1.6492719592297478e-06, + "loss": 1.5811, + "step": 10556 + }, + { + "epoch": 0.9439377682403434, + "grad_norm": 0.5359747219310252, + "learning_rate": 1.6440381108495772e-06, + "loss": 1.5446, + "step": 10557 + }, + { + "epoch": 0.9440271816881259, + "grad_norm": 0.5835039498174784, + "learning_rate": 1.6388125114651486e-06, + "loss": 1.4206, + "step": 10558 + }, + { + "epoch": 0.9441165951359084, + "grad_norm": 0.40508614034483287, + "learning_rate": 1.6335951615147337e-06, + "loss": 1.4502, + "step": 10559 + }, + { + "epoch": 0.944206008583691, + "grad_norm": 0.6005659331561861, + "learning_rate": 1.6283860614358936e-06, + "loss": 1.6289, + "step": 10560 + }, + { + "epoch": 0.9442954220314735, + "grad_norm": 0.4723605187974817, + "learning_rate": 1.623185211665501e-06, + "loss": 1.656, + "step": 10561 + }, + { + "epoch": 0.9443848354792561, + "grad_norm": 0.9265042092482882, + "learning_rate": 1.6179926126397626e-06, + "loss": 6.3385, + "step": 10562 + }, + { + "epoch": 0.9444742489270386, + "grad_norm": 0.6154718085063524, + "learning_rate": 1.6128082647941744e-06, + "loss": 1.673, + "step": 10563 + }, + { + "epoch": 0.9445636623748211, + "grad_norm": 0.5542492324275761, + "learning_rate": 1.6076321685635332e-06, + "loss": 1.4456, + "step": 10564 + }, + { + "epoch": 0.9446530758226037, + "grad_norm": 0.41642924138718146, + "learning_rate": 1.602464324381936e-06, + "loss": 1.389, + "step": 10565 + }, + { + "epoch": 0.9447424892703863, + "grad_norm": 0.5376954518261186, + "learning_rate": 1.5973047326828472e-06, + "loss": 1.4553, + "step": 10566 + }, + { + "epoch": 0.9448319027181689, + "grad_norm": 0.469200103675631, + "learning_rate": 1.5921533938989542e-06, + "loss": 1.5314, + "step": 10567 + }, + { + "epoch": 0.9449213161659513, + "grad_norm": 0.43894160762113477, + "learning_rate": 1.5870103084623111e-06, + "loss": 1.4885, + "step": 10568 + }, + { + "epoch": 0.9450107296137339, + "grad_norm": 0.43034460593959, + "learning_rate": 1.5818754768042733e-06, + "loss": 1.5939, + "step": 10569 + }, + { + "epoch": 0.9451001430615165, + "grad_norm": 0.5828875487749565, + "learning_rate": 1.5767488993554736e-06, + "loss": 1.69, + "step": 10570 + }, + { + "epoch": 0.945189556509299, + "grad_norm": 0.5987823150831016, + "learning_rate": 1.5716305765458683e-06, + "loss": 1.6183, + "step": 10571 + }, + { + "epoch": 0.9452789699570815, + "grad_norm": 0.4228237076956824, + "learning_rate": 1.5665205088047474e-06, + "loss": 1.4685, + "step": 10572 + }, + { + "epoch": 0.9453683834048641, + "grad_norm": 0.7784001254460653, + "learning_rate": 1.561418696560668e-06, + "loss": 1.7011, + "step": 10573 + }, + { + "epoch": 0.9454577968526466, + "grad_norm": 0.4449833960355372, + "learning_rate": 1.5563251402415102e-06, + "loss": 1.7082, + "step": 10574 + }, + { + "epoch": 0.9455472103004292, + "grad_norm": 0.5912203850481648, + "learning_rate": 1.5512398402744876e-06, + "loss": 1.6438, + "step": 10575 + }, + { + "epoch": 0.9456366237482118, + "grad_norm": 0.33525095055074766, + "learning_rate": 1.5461627970860814e-06, + "loss": 1.4475, + "step": 10576 + }, + { + "epoch": 0.9457260371959942, + "grad_norm": 0.5127851136055476, + "learning_rate": 1.5410940111020956e-06, + "loss": 1.6213, + "step": 10577 + }, + { + "epoch": 0.9458154506437768, + "grad_norm": 0.5143066980585967, + "learning_rate": 1.5360334827476564e-06, + "loss": 1.4439, + "step": 10578 + }, + { + "epoch": 0.9459048640915594, + "grad_norm": 0.4468420481984176, + "learning_rate": 1.5309812124471579e-06, + "loss": 1.7864, + "step": 10579 + }, + { + "epoch": 0.945994277539342, + "grad_norm": 0.4656079540687474, + "learning_rate": 1.52593720062435e-06, + "loss": 1.5835, + "step": 10580 + }, + { + "epoch": 0.9460836909871244, + "grad_norm": 0.5466865833702735, + "learning_rate": 1.520901447702272e-06, + "loss": 1.7172, + "step": 10581 + }, + { + "epoch": 0.946173104434907, + "grad_norm": 0.47808263192762623, + "learning_rate": 1.5158739541032418e-06, + "loss": 1.9061, + "step": 10582 + }, + { + "epoch": 0.9462625178826896, + "grad_norm": 0.6564100509015189, + "learning_rate": 1.5108547202489443e-06, + "loss": 1.8395, + "step": 10583 + }, + { + "epoch": 0.9463519313304721, + "grad_norm": 0.5195598698592662, + "learning_rate": 1.5058437465602982e-06, + "loss": 1.6229, + "step": 10584 + }, + { + "epoch": 0.9464413447782547, + "grad_norm": 0.41378606515310695, + "learning_rate": 1.5008410334576006e-06, + "loss": 1.4455, + "step": 10585 + }, + { + "epoch": 0.9465307582260372, + "grad_norm": 0.3765356636904052, + "learning_rate": 1.495846581360394e-06, + "loss": 1.6293, + "step": 10586 + }, + { + "epoch": 0.9466201716738197, + "grad_norm": 0.5431589297484549, + "learning_rate": 1.4908603906875761e-06, + "loss": 1.6884, + "step": 10587 + }, + { + "epoch": 0.9467095851216023, + "grad_norm": 0.370009883555017, + "learning_rate": 1.4858824618573352e-06, + "loss": 1.7592, + "step": 10588 + }, + { + "epoch": 0.9467989985693849, + "grad_norm": 0.7779333336763111, + "learning_rate": 1.4809127952871592e-06, + "loss": 6.147, + "step": 10589 + }, + { + "epoch": 0.9468884120171673, + "grad_norm": 0.3402019827515543, + "learning_rate": 1.4759513913938372e-06, + "loss": 1.3984, + "step": 10590 + }, + { + "epoch": 0.9469778254649499, + "grad_norm": 0.4424268486114009, + "learning_rate": 1.4709982505934806e-06, + "loss": 1.4584, + "step": 10591 + }, + { + "epoch": 0.9470672389127325, + "grad_norm": 0.5460577271245034, + "learning_rate": 1.4660533733015236e-06, + "loss": 1.8219, + "step": 10592 + }, + { + "epoch": 0.947156652360515, + "grad_norm": 0.43748654112310864, + "learning_rate": 1.461116759932657e-06, + "loss": 1.2914, + "step": 10593 + }, + { + "epoch": 0.9472460658082976, + "grad_norm": 0.45357032415279963, + "learning_rate": 1.4561884109009384e-06, + "loss": 1.5394, + "step": 10594 + }, + { + "epoch": 0.9473354792560801, + "grad_norm": 0.4708640411596751, + "learning_rate": 1.4512683266196703e-06, + "loss": 1.6835, + "step": 10595 + }, + { + "epoch": 0.9474248927038627, + "grad_norm": 0.6851387162463103, + "learning_rate": 1.4463565075015228e-06, + "loss": 6.2329, + "step": 10596 + }, + { + "epoch": 0.9475143061516452, + "grad_norm": 0.5259655640768461, + "learning_rate": 1.441452953958422e-06, + "loss": 1.4775, + "step": 10597 + }, + { + "epoch": 0.9476037195994278, + "grad_norm": 0.39309134005687785, + "learning_rate": 1.43655766640165e-06, + "loss": 1.4975, + "step": 10598 + }, + { + "epoch": 0.9476931330472103, + "grad_norm": 0.472237303707616, + "learning_rate": 1.4316706452417338e-06, + "loss": 1.7224, + "step": 10599 + }, + { + "epoch": 0.9477825464949928, + "grad_norm": 0.4962927882839053, + "learning_rate": 1.4267918908885681e-06, + "loss": 1.7018, + "step": 10600 + }, + { + "epoch": 0.9478719599427754, + "grad_norm": 0.7414385804678012, + "learning_rate": 1.421921403751314e-06, + "loss": 1.5337, + "step": 10601 + }, + { + "epoch": 0.947961373390558, + "grad_norm": 0.6086850896050737, + "learning_rate": 1.4170591842384672e-06, + "loss": 1.7873, + "step": 10602 + }, + { + "epoch": 0.9480507868383404, + "grad_norm": 0.42583973787885254, + "learning_rate": 1.4122052327578128e-06, + "loss": 1.628, + "step": 10603 + }, + { + "epoch": 0.948140200286123, + "grad_norm": 0.5343710605403009, + "learning_rate": 1.4073595497164361e-06, + "loss": 1.7018, + "step": 10604 + }, + { + "epoch": 0.9482296137339056, + "grad_norm": 0.46457876369638235, + "learning_rate": 1.402522135520734e-06, + "loss": 1.4174, + "step": 10605 + }, + { + "epoch": 0.9483190271816881, + "grad_norm": 0.5040418136950662, + "learning_rate": 1.397692990576449e-06, + "loss": 1.6224, + "step": 10606 + }, + { + "epoch": 0.9484084406294707, + "grad_norm": 0.35023138537703474, + "learning_rate": 1.392872115288546e-06, + "loss": 1.4763, + "step": 10607 + }, + { + "epoch": 0.9484978540772532, + "grad_norm": 0.7168591893273847, + "learning_rate": 1.3880595100613792e-06, + "loss": 1.4178, + "step": 10608 + }, + { + "epoch": 0.9485872675250357, + "grad_norm": 0.8308604756469649, + "learning_rate": 1.3832551752985811e-06, + "loss": 1.9135, + "step": 10609 + }, + { + "epoch": 0.9486766809728183, + "grad_norm": 0.32758136664606285, + "learning_rate": 1.378459111403052e-06, + "loss": 1.3073, + "step": 10610 + }, + { + "epoch": 0.9487660944206009, + "grad_norm": 0.5261720921704592, + "learning_rate": 1.37367131877707e-06, + "loss": 1.7328, + "step": 10611 + }, + { + "epoch": 0.9488555078683834, + "grad_norm": 0.3870149822246841, + "learning_rate": 1.3688917978221583e-06, + "loss": 1.3456, + "step": 10612 + }, + { + "epoch": 0.9489449213161659, + "grad_norm": 0.40249276781724797, + "learning_rate": 1.364120548939174e-06, + "loss": 1.3852, + "step": 10613 + }, + { + "epoch": 0.9490343347639485, + "grad_norm": 0.39197380831980494, + "learning_rate": 1.3593575725282749e-06, + "loss": 1.2849, + "step": 10614 + }, + { + "epoch": 0.9491237482117311, + "grad_norm": 0.5369103902075172, + "learning_rate": 1.3546028689889302e-06, + "loss": 1.6, + "step": 10615 + }, + { + "epoch": 0.9492131616595136, + "grad_norm": 0.5347931072707075, + "learning_rate": 1.3498564387199098e-06, + "loss": 1.3983, + "step": 10616 + }, + { + "epoch": 0.9493025751072961, + "grad_norm": 0.4054492722337682, + "learning_rate": 1.3451182821192954e-06, + "loss": 1.7317, + "step": 10617 + }, + { + "epoch": 0.9493919885550787, + "grad_norm": 0.5440081270580964, + "learning_rate": 1.3403883995844579e-06, + "loss": 1.5554, + "step": 10618 + }, + { + "epoch": 0.9494814020028612, + "grad_norm": 0.5821669063346059, + "learning_rate": 1.3356667915121025e-06, + "loss": 1.4859, + "step": 10619 + }, + { + "epoch": 0.9495708154506438, + "grad_norm": 0.47474675161220264, + "learning_rate": 1.330953458298212e-06, + "loss": 1.5463, + "step": 10620 + }, + { + "epoch": 0.9496602288984263, + "grad_norm": 0.6195835524299116, + "learning_rate": 1.3262484003380927e-06, + "loss": 1.611, + "step": 10621 + }, + { + "epoch": 0.9497496423462088, + "grad_norm": 0.346937045824777, + "learning_rate": 1.321551618026351e-06, + "loss": 1.4408, + "step": 10622 + }, + { + "epoch": 0.9498390557939914, + "grad_norm": 0.6318456893608693, + "learning_rate": 1.3168631117569052e-06, + "loss": 1.4914, + "step": 10623 + }, + { + "epoch": 0.949928469241774, + "grad_norm": 0.522362784000784, + "learning_rate": 1.3121828819229743e-06, + "loss": 1.4931, + "step": 10624 + }, + { + "epoch": 0.9500178826895566, + "grad_norm": 0.8586153854336755, + "learning_rate": 1.3075109289170773e-06, + "loss": 1.5371, + "step": 10625 + }, + { + "epoch": 0.950107296137339, + "grad_norm": 0.5642063983328645, + "learning_rate": 1.3028472531310454e-06, + "loss": 1.6014, + "step": 10626 + }, + { + "epoch": 0.9501967095851216, + "grad_norm": 0.5395879779076578, + "learning_rate": 1.2981918549560213e-06, + "loss": 1.5725, + "step": 10627 + }, + { + "epoch": 0.9502861230329042, + "grad_norm": 0.48144917457347, + "learning_rate": 1.293544734782437e-06, + "loss": 1.8114, + "step": 10628 + }, + { + "epoch": 0.9503755364806867, + "grad_norm": 0.5263184669086827, + "learning_rate": 1.2889058930000586e-06, + "loss": 1.5708, + "step": 10629 + }, + { + "epoch": 0.9504649499284692, + "grad_norm": 0.547465534083484, + "learning_rate": 1.2842753299979305e-06, + "loss": 1.7772, + "step": 10630 + }, + { + "epoch": 0.9505543633762518, + "grad_norm": 1.5197357694122238, + "learning_rate": 1.2796530461644086e-06, + "loss": 6.1149, + "step": 10631 + }, + { + "epoch": 0.9506437768240343, + "grad_norm": 0.4739935072294854, + "learning_rate": 1.2750390418871604e-06, + "loss": 1.499, + "step": 10632 + }, + { + "epoch": 0.9507331902718169, + "grad_norm": 0.415329072393003, + "learning_rate": 1.2704333175531546e-06, + "loss": 1.4512, + "step": 10633 + }, + { + "epoch": 0.9508226037195995, + "grad_norm": 0.6448336030155489, + "learning_rate": 1.265835873548682e-06, + "loss": 1.8163, + "step": 10634 + }, + { + "epoch": 0.9509120171673819, + "grad_norm": 0.4011479233976127, + "learning_rate": 1.2612467102593006e-06, + "loss": 1.6693, + "step": 10635 + }, + { + "epoch": 0.9510014306151645, + "grad_norm": 0.6050369705946463, + "learning_rate": 1.256665828069925e-06, + "loss": 1.8584, + "step": 10636 + }, + { + "epoch": 0.9510908440629471, + "grad_norm": 0.3870169988743417, + "learning_rate": 1.2520932273647258e-06, + "loss": 1.4515, + "step": 10637 + }, + { + "epoch": 0.9511802575107297, + "grad_norm": 0.39566089438408664, + "learning_rate": 1.2475289085272178e-06, + "loss": 1.5228, + "step": 10638 + }, + { + "epoch": 0.9512696709585121, + "grad_norm": 0.4719266521474626, + "learning_rate": 1.2429728719401845e-06, + "loss": 1.4332, + "step": 10639 + }, + { + "epoch": 0.9513590844062947, + "grad_norm": 0.5171930916784105, + "learning_rate": 1.2384251179857643e-06, + "loss": 1.6226, + "step": 10640 + }, + { + "epoch": 0.9514484978540773, + "grad_norm": 0.6467403998836948, + "learning_rate": 1.233885647045341e-06, + "loss": 1.4157, + "step": 10641 + }, + { + "epoch": 0.9515379113018598, + "grad_norm": 0.42618639839464806, + "learning_rate": 1.2293544594996543e-06, + "loss": 1.3767, + "step": 10642 + }, + { + "epoch": 0.9516273247496424, + "grad_norm": 0.5508394360755282, + "learning_rate": 1.2248315557287337e-06, + "loss": 1.4609, + "step": 10643 + }, + { + "epoch": 0.9517167381974249, + "grad_norm": 0.6370856562417868, + "learning_rate": 1.2203169361118871e-06, + "loss": 1.6464, + "step": 10644 + }, + { + "epoch": 0.9518061516452074, + "grad_norm": 0.39494129332588496, + "learning_rate": 1.215810601027767e-06, + "loss": 1.6077, + "step": 10645 + }, + { + "epoch": 0.95189556509299, + "grad_norm": 0.4448776498977148, + "learning_rate": 1.2113125508543267e-06, + "loss": 1.5928, + "step": 10646 + }, + { + "epoch": 0.9519849785407726, + "grad_norm": 0.49231261408785826, + "learning_rate": 1.2068227859687753e-06, + "loss": 1.635, + "step": 10647 + }, + { + "epoch": 0.952074391988555, + "grad_norm": 0.3502598033258094, + "learning_rate": 1.2023413067476896e-06, + "loss": 1.343, + "step": 10648 + }, + { + "epoch": 0.9521638054363376, + "grad_norm": 0.6267763322312078, + "learning_rate": 1.1978681135669245e-06, + "loss": 1.8785, + "step": 10649 + }, + { + "epoch": 0.9522532188841202, + "grad_norm": 0.426383460729787, + "learning_rate": 1.1934032068016354e-06, + "loss": 1.4997, + "step": 10650 + }, + { + "epoch": 0.9523426323319027, + "grad_norm": 0.4458858333827367, + "learning_rate": 1.1889465868263005e-06, + "loss": 1.6794, + "step": 10651 + }, + { + "epoch": 0.9524320457796852, + "grad_norm": 0.5583387835832361, + "learning_rate": 1.1844982540146654e-06, + "loss": 1.5107, + "step": 10652 + }, + { + "epoch": 0.9525214592274678, + "grad_norm": 0.4579328593537116, + "learning_rate": 1.1800582087398316e-06, + "loss": 1.5173, + "step": 10653 + }, + { + "epoch": 0.9526108726752504, + "grad_norm": 0.5006651983338621, + "learning_rate": 1.1756264513741676e-06, + "loss": 1.4586, + "step": 10654 + }, + { + "epoch": 0.9527002861230329, + "grad_norm": 0.6005223659360562, + "learning_rate": 1.1712029822893654e-06, + "loss": 1.489, + "step": 10655 + }, + { + "epoch": 0.9527896995708155, + "grad_norm": 0.5605361115127887, + "learning_rate": 1.1667878018564171e-06, + "loss": 1.2715, + "step": 10656 + }, + { + "epoch": 0.952879113018598, + "grad_norm": 0.3626391255076893, + "learning_rate": 1.1623809104456262e-06, + "loss": 1.4349, + "step": 10657 + }, + { + "epoch": 0.9529685264663805, + "grad_norm": 0.5793245736981217, + "learning_rate": 1.157982308426564e-06, + "loss": 2.0103, + "step": 10658 + }, + { + "epoch": 0.9530579399141631, + "grad_norm": 0.45777921117192727, + "learning_rate": 1.1535919961681575e-06, + "loss": 1.7622, + "step": 10659 + }, + { + "epoch": 0.9531473533619457, + "grad_norm": 0.47628299967903187, + "learning_rate": 1.1492099740386231e-06, + "loss": 1.4513, + "step": 10660 + }, + { + "epoch": 0.9532367668097281, + "grad_norm": 0.45013536676146604, + "learning_rate": 1.144836242405467e-06, + "loss": 1.6509, + "step": 10661 + }, + { + "epoch": 0.9533261802575107, + "grad_norm": 0.626697103076785, + "learning_rate": 1.140470801635496e-06, + "loss": 1.6202, + "step": 10662 + }, + { + "epoch": 0.9534155937052933, + "grad_norm": 0.42146110830781525, + "learning_rate": 1.13611365209485e-06, + "loss": 1.638, + "step": 10663 + }, + { + "epoch": 0.9535050071530758, + "grad_norm": 0.7694014156252256, + "learning_rate": 1.1317647941489595e-06, + "loss": 1.8108, + "step": 10664 + }, + { + "epoch": 0.9535944206008584, + "grad_norm": 0.46238355838097295, + "learning_rate": 1.1274242281625547e-06, + "loss": 1.7342, + "step": 10665 + }, + { + "epoch": 0.9536838340486409, + "grad_norm": 0.5007012979746059, + "learning_rate": 1.1230919544996776e-06, + "loss": 1.7257, + "step": 10666 + }, + { + "epoch": 0.9537732474964234, + "grad_norm": 0.4818386953023412, + "learning_rate": 1.1187679735236489e-06, + "loss": 1.5903, + "step": 10667 + }, + { + "epoch": 0.953862660944206, + "grad_norm": 0.46317288255538674, + "learning_rate": 1.114452285597145e-06, + "loss": 1.5083, + "step": 10668 + }, + { + "epoch": 0.9539520743919886, + "grad_norm": 0.508333329918875, + "learning_rate": 1.110144891082099e-06, + "loss": 1.322, + "step": 10669 + }, + { + "epoch": 0.954041487839771, + "grad_norm": 0.7120651438705685, + "learning_rate": 1.1058457903397656e-06, + "loss": 6.0617, + "step": 10670 + }, + { + "epoch": 0.9541309012875536, + "grad_norm": 0.4008950948371433, + "learning_rate": 1.1015549837307237e-06, + "loss": 1.4987, + "step": 10671 + }, + { + "epoch": 0.9542203147353362, + "grad_norm": 0.4513836705782417, + "learning_rate": 1.0972724716148187e-06, + "loss": 1.6847, + "step": 10672 + }, + { + "epoch": 0.9543097281831188, + "grad_norm": 0.46128409182400326, + "learning_rate": 1.0929982543512296e-06, + "loss": 1.7773, + "step": 10673 + }, + { + "epoch": 0.9543991416309013, + "grad_norm": 0.644333310012295, + "learning_rate": 1.0887323322984366e-06, + "loss": 1.8575, + "step": 10674 + }, + { + "epoch": 0.9544885550786838, + "grad_norm": 0.5469103474850303, + "learning_rate": 1.084474705814198e-06, + "loss": 1.7457, + "step": 10675 + }, + { + "epoch": 0.9545779685264664, + "grad_norm": 0.5183682955501941, + "learning_rate": 1.0802253752556058e-06, + "loss": 1.6922, + "step": 10676 + }, + { + "epoch": 0.9546673819742489, + "grad_norm": 0.40637864405609275, + "learning_rate": 1.0759843409790527e-06, + "loss": 1.4006, + "step": 10677 + }, + { + "epoch": 0.9547567954220315, + "grad_norm": 0.49044702238938825, + "learning_rate": 1.0717516033402097e-06, + "loss": 1.672, + "step": 10678 + }, + { + "epoch": 0.954846208869814, + "grad_norm": 0.39751742738188645, + "learning_rate": 1.0675271626940931e-06, + "loss": 1.5166, + "step": 10679 + }, + { + "epoch": 0.9549356223175965, + "grad_norm": 0.5359248070864363, + "learning_rate": 1.063311019395008e-06, + "loss": 1.467, + "step": 10680 + }, + { + "epoch": 0.9550250357653791, + "grad_norm": 0.4701897569951864, + "learning_rate": 1.0591031737965273e-06, + "loss": 1.582, + "step": 10681 + }, + { + "epoch": 0.9551144492131617, + "grad_norm": 0.638903207442665, + "learning_rate": 1.0549036262515689e-06, + "loss": 1.6878, + "step": 10682 + }, + { + "epoch": 0.9552038626609443, + "grad_norm": 0.3947528587414448, + "learning_rate": 1.0507123771123505e-06, + "loss": 1.5251, + "step": 10683 + }, + { + "epoch": 0.9552932761087267, + "grad_norm": 0.4886882201942753, + "learning_rate": 1.0465294267303915e-06, + "loss": 1.6594, + "step": 10684 + }, + { + "epoch": 0.9553826895565093, + "grad_norm": 0.44946572525601414, + "learning_rate": 1.0423547754564888e-06, + "loss": 1.5201, + "step": 10685 + }, + { + "epoch": 0.9554721030042919, + "grad_norm": 0.4544515738420074, + "learning_rate": 1.0381884236407958e-06, + "loss": 1.6358, + "step": 10686 + }, + { + "epoch": 0.9555615164520744, + "grad_norm": 0.39892352440652534, + "learning_rate": 1.0340303716327215e-06, + "loss": 1.4389, + "step": 10687 + }, + { + "epoch": 0.9556509298998569, + "grad_norm": 0.3742185231501777, + "learning_rate": 1.0298806197809984e-06, + "loss": 1.5374, + "step": 10688 + }, + { + "epoch": 0.9557403433476395, + "grad_norm": 0.2948565491825822, + "learning_rate": 1.0257391684336703e-06, + "loss": 1.2696, + "step": 10689 + }, + { + "epoch": 0.955829756795422, + "grad_norm": 0.5468694087830893, + "learning_rate": 1.0216060179380481e-06, + "loss": 1.7693, + "step": 10690 + }, + { + "epoch": 0.9559191702432046, + "grad_norm": 0.40341746074504176, + "learning_rate": 1.0174811686408104e-06, + "loss": 1.38, + "step": 10691 + }, + { + "epoch": 0.9560085836909872, + "grad_norm": 0.5680176795850592, + "learning_rate": 1.01336462088788e-06, + "loss": 2.0397, + "step": 10692 + }, + { + "epoch": 0.9560979971387696, + "grad_norm": 0.7188351980215585, + "learning_rate": 1.0092563750245032e-06, + "loss": 1.6347, + "step": 10693 + }, + { + "epoch": 0.9561874105865522, + "grad_norm": 0.4998383729522865, + "learning_rate": 1.00515643139526e-06, + "loss": 1.8689, + "step": 10694 + }, + { + "epoch": 0.9562768240343348, + "grad_norm": 0.4691816987292365, + "learning_rate": 1.0010647903439862e-06, + "loss": 1.6588, + "step": 10695 + }, + { + "epoch": 0.9563662374821174, + "grad_norm": 0.5283586492152658, + "learning_rate": 9.96981452213852e-07, + "loss": 1.5121, + "step": 10696 + }, + { + "epoch": 0.9564556509298998, + "grad_norm": 0.41887739373622745, + "learning_rate": 9.929064173473057e-07, + "loss": 1.5856, + "step": 10697 + }, + { + "epoch": 0.9565450643776824, + "grad_norm": 0.4713014655957833, + "learning_rate": 9.888396860861404e-07, + "loss": 1.8403, + "step": 10698 + }, + { + "epoch": 0.956634477825465, + "grad_norm": 0.3919183746134381, + "learning_rate": 9.847812587714057e-07, + "loss": 1.3953, + "step": 10699 + }, + { + "epoch": 0.9567238912732475, + "grad_norm": 0.40386404117820895, + "learning_rate": 9.807311357434956e-07, + "loss": 1.4481, + "step": 10700 + }, + { + "epoch": 0.95681330472103, + "grad_norm": 0.34070622714482224, + "learning_rate": 9.766893173420721e-07, + "loss": 1.5511, + "step": 10701 + }, + { + "epoch": 0.9569027181688126, + "grad_norm": 0.44996953936325956, + "learning_rate": 9.726558039061308e-07, + "loss": 1.3692, + "step": 10702 + }, + { + "epoch": 0.9569921316165951, + "grad_norm": 0.4930579284314233, + "learning_rate": 9.68630595773956e-07, + "loss": 1.763, + "step": 10703 + }, + { + "epoch": 0.9570815450643777, + "grad_norm": 0.5338930204139105, + "learning_rate": 9.64613693283123e-07, + "loss": 1.8896, + "step": 10704 + }, + { + "epoch": 0.9571709585121603, + "grad_norm": 0.4120134410154897, + "learning_rate": 9.606050967705393e-07, + "loss": 1.7462, + "step": 10705 + }, + { + "epoch": 0.9572603719599427, + "grad_norm": 0.37574785590146625, + "learning_rate": 9.566048065724032e-07, + "loss": 1.4772, + "step": 10706 + }, + { + "epoch": 0.9573497854077253, + "grad_norm": 0.4841312194487389, + "learning_rate": 9.526128230242016e-07, + "loss": 1.5357, + "step": 10707 + }, + { + "epoch": 0.9574391988555079, + "grad_norm": 0.5586484943066261, + "learning_rate": 9.486291464607444e-07, + "loss": 1.347, + "step": 10708 + }, + { + "epoch": 0.9575286123032904, + "grad_norm": 0.45750270472349713, + "learning_rate": 9.446537772161423e-07, + "loss": 1.3035, + "step": 10709 + }, + { + "epoch": 0.9576180257510729, + "grad_norm": 0.44456811804550345, + "learning_rate": 9.406867156237842e-07, + "loss": 1.7652, + "step": 10710 + }, + { + "epoch": 0.9577074391988555, + "grad_norm": 0.5409460830907618, + "learning_rate": 9.367279620164149e-07, + "loss": 1.758, + "step": 10711 + }, + { + "epoch": 0.957796852646638, + "grad_norm": 0.6622202848976282, + "learning_rate": 9.327775167260244e-07, + "loss": 1.4318, + "step": 10712 + }, + { + "epoch": 0.9578862660944206, + "grad_norm": 0.3995015886942954, + "learning_rate": 9.288353800839366e-07, + "loss": 1.7214, + "step": 10713 + }, + { + "epoch": 0.9579756795422032, + "grad_norm": 0.7413793761939103, + "learning_rate": 9.249015524207872e-07, + "loss": 1.6137, + "step": 10714 + }, + { + "epoch": 0.9580650929899857, + "grad_norm": 0.4634335190270478, + "learning_rate": 9.209760340664897e-07, + "loss": 1.6132, + "step": 10715 + }, + { + "epoch": 0.9581545064377682, + "grad_norm": 0.3957833145193772, + "learning_rate": 9.170588253502698e-07, + "loss": 1.8617, + "step": 10716 + }, + { + "epoch": 0.9582439198855508, + "grad_norm": 0.36385526708704624, + "learning_rate": 9.13149926600676e-07, + "loss": 1.494, + "step": 10717 + }, + { + "epoch": 0.9583333333333334, + "grad_norm": 0.38564216490495684, + "learning_rate": 9.092493381455236e-07, + "loss": 1.441, + "step": 10718 + }, + { + "epoch": 0.9584227467811158, + "grad_norm": 0.33236940583311353, + "learning_rate": 9.05357060311951e-07, + "loss": 1.3767, + "step": 10719 + }, + { + "epoch": 0.9585121602288984, + "grad_norm": 0.6277452018597264, + "learning_rate": 9.014730934264192e-07, + "loss": 1.6162, + "step": 10720 + }, + { + "epoch": 0.958601573676681, + "grad_norm": 0.4414618911694844, + "learning_rate": 8.975974378146457e-07, + "loss": 1.6254, + "step": 10721 + }, + { + "epoch": 0.9586909871244635, + "grad_norm": 0.5335816079800092, + "learning_rate": 8.937300938017035e-07, + "loss": 1.6092, + "step": 10722 + }, + { + "epoch": 0.9587804005722461, + "grad_norm": 0.40564063699540476, + "learning_rate": 8.898710617119222e-07, + "loss": 1.4149, + "step": 10723 + }, + { + "epoch": 0.9588698140200286, + "grad_norm": 0.48579129669682386, + "learning_rate": 8.860203418689539e-07, + "loss": 1.6049, + "step": 10724 + }, + { + "epoch": 0.9589592274678111, + "grad_norm": 0.49927984463063924, + "learning_rate": 8.821779345957626e-07, + "loss": 1.6306, + "step": 10725 + }, + { + "epoch": 0.9590486409155937, + "grad_norm": 0.36199549142016757, + "learning_rate": 8.783438402146127e-07, + "loss": 1.3316, + "step": 10726 + }, + { + "epoch": 0.9591380543633763, + "grad_norm": 0.6505339874544671, + "learning_rate": 8.74518059047047e-07, + "loss": 1.7747, + "step": 10727 + }, + { + "epoch": 0.9592274678111588, + "grad_norm": 0.47859899840841785, + "learning_rate": 8.707005914139422e-07, + "loss": 1.5952, + "step": 10728 + }, + { + "epoch": 0.9593168812589413, + "grad_norm": 0.4525297283649028, + "learning_rate": 8.668914376354642e-07, + "loss": 1.6977, + "step": 10729 + }, + { + "epoch": 0.9594062947067239, + "grad_norm": 0.4147961296303402, + "learning_rate": 8.630905980310689e-07, + "loss": 1.748, + "step": 10730 + }, + { + "epoch": 0.9594957081545065, + "grad_norm": 0.6137265363840545, + "learning_rate": 8.592980729195455e-07, + "loss": 2.5065, + "step": 10731 + }, + { + "epoch": 0.959585121602289, + "grad_norm": 0.54503084525024, + "learning_rate": 8.555138626189618e-07, + "loss": 1.4919, + "step": 10732 + }, + { + "epoch": 0.9596745350500715, + "grad_norm": 0.3179330320861252, + "learning_rate": 8.517379674466863e-07, + "loss": 1.2823, + "step": 10733 + }, + { + "epoch": 0.9597639484978541, + "grad_norm": 0.5389131879123937, + "learning_rate": 8.479703877194212e-07, + "loss": 1.6721, + "step": 10734 + }, + { + "epoch": 0.9598533619456366, + "grad_norm": 0.4932495552561669, + "learning_rate": 8.442111237531247e-07, + "loss": 1.4397, + "step": 10735 + }, + { + "epoch": 0.9599427753934192, + "grad_norm": 0.4208120196233225, + "learning_rate": 8.404601758630892e-07, + "loss": 1.7136, + "step": 10736 + }, + { + "epoch": 0.9600321888412017, + "grad_norm": 0.3931366591994999, + "learning_rate": 8.367175443639075e-07, + "loss": 1.9107, + "step": 10737 + }, + { + "epoch": 0.9601216022889842, + "grad_norm": 0.4193450683745568, + "learning_rate": 8.329832295694618e-07, + "loss": 1.3628, + "step": 10738 + }, + { + "epoch": 0.9602110157367668, + "grad_norm": 0.4967594753468597, + "learning_rate": 8.29257231792957e-07, + "loss": 1.5068, + "step": 10739 + }, + { + "epoch": 0.9603004291845494, + "grad_norm": 0.4841957163423867, + "learning_rate": 8.255395513468767e-07, + "loss": 1.6767, + "step": 10740 + }, + { + "epoch": 0.960389842632332, + "grad_norm": 0.85023596181414, + "learning_rate": 8.218301885430268e-07, + "loss": 1.5211, + "step": 10741 + }, + { + "epoch": 0.9604792560801144, + "grad_norm": 0.39312672619173206, + "learning_rate": 8.181291436924921e-07, + "loss": 1.3781, + "step": 10742 + }, + { + "epoch": 0.960568669527897, + "grad_norm": 0.4597680702709099, + "learning_rate": 8.144364171056906e-07, + "loss": 1.5234, + "step": 10743 + }, + { + "epoch": 0.9606580829756796, + "grad_norm": 0.5525629216339552, + "learning_rate": 8.107520090923193e-07, + "loss": 1.6178, + "step": 10744 + }, + { + "epoch": 0.9607474964234621, + "grad_norm": 0.5173026675420954, + "learning_rate": 8.070759199613864e-07, + "loss": 1.3384, + "step": 10745 + }, + { + "epoch": 0.9608369098712446, + "grad_norm": 0.6440305297093855, + "learning_rate": 8.03408150021201e-07, + "loss": 1.6699, + "step": 10746 + }, + { + "epoch": 0.9609263233190272, + "grad_norm": 1.1787704684353095, + "learning_rate": 7.997486995793834e-07, + "loss": 6.1694, + "step": 10747 + }, + { + "epoch": 0.9610157367668097, + "grad_norm": 0.4610554786770623, + "learning_rate": 7.96097568942833e-07, + "loss": 1.4006, + "step": 10748 + }, + { + "epoch": 0.9611051502145923, + "grad_norm": 0.4580461179601357, + "learning_rate": 7.924547584177711e-07, + "loss": 1.7911, + "step": 10749 + }, + { + "epoch": 0.9611945636623748, + "grad_norm": 0.8031318495776381, + "learning_rate": 7.88820268309709e-07, + "loss": 1.4343, + "step": 10750 + }, + { + "epoch": 0.9612839771101573, + "grad_norm": 0.48743103295778484, + "learning_rate": 7.851940989234919e-07, + "loss": 1.5337, + "step": 10751 + }, + { + "epoch": 0.9613733905579399, + "grad_norm": 0.6411055754574507, + "learning_rate": 7.815762505632096e-07, + "loss": 1.5408, + "step": 10752 + }, + { + "epoch": 0.9614628040057225, + "grad_norm": 0.528455460728322, + "learning_rate": 7.779667235322974e-07, + "loss": 1.6288, + "step": 10753 + }, + { + "epoch": 0.961552217453505, + "grad_norm": 0.45255068829644046, + "learning_rate": 7.743655181335019e-07, + "loss": 1.6753, + "step": 10754 + }, + { + "epoch": 0.9616416309012875, + "grad_norm": 0.38089939527074124, + "learning_rate": 7.707726346688259e-07, + "loss": 1.4181, + "step": 10755 + }, + { + "epoch": 0.9617310443490701, + "grad_norm": 0.5694892129790518, + "learning_rate": 7.671880734396175e-07, + "loss": 1.6525, + "step": 10756 + }, + { + "epoch": 0.9618204577968527, + "grad_norm": 0.5239710080121439, + "learning_rate": 7.636118347465027e-07, + "loss": 1.8038, + "step": 10757 + }, + { + "epoch": 0.9619098712446352, + "grad_norm": 0.9383988103112776, + "learning_rate": 7.600439188894082e-07, + "loss": 1.7691, + "step": 10758 + }, + { + "epoch": 0.9619992846924177, + "grad_norm": 0.40333178387431096, + "learning_rate": 7.564843261675835e-07, + "loss": 1.6994, + "step": 10759 + }, + { + "epoch": 0.9620886981402003, + "grad_norm": 0.6307457642173766, + "learning_rate": 7.529330568795568e-07, + "loss": 1.9159, + "step": 10760 + }, + { + "epoch": 0.9621781115879828, + "grad_norm": 0.8623320799111948, + "learning_rate": 7.493901113231782e-07, + "loss": 1.9876, + "step": 10761 + }, + { + "epoch": 0.9622675250357654, + "grad_norm": 0.8048487838942913, + "learning_rate": 7.458554897955883e-07, + "loss": 1.3505, + "step": 10762 + }, + { + "epoch": 0.962356938483548, + "grad_norm": 0.6935613295424239, + "learning_rate": 7.423291925932275e-07, + "loss": 1.5511, + "step": 10763 + }, + { + "epoch": 0.9624463519313304, + "grad_norm": 0.9119997299329741, + "learning_rate": 7.388112200118479e-07, + "loss": 1.8709, + "step": 10764 + }, + { + "epoch": 0.962535765379113, + "grad_norm": 0.4755839858133393, + "learning_rate": 7.353015723464918e-07, + "loss": 1.6703, + "step": 10765 + }, + { + "epoch": 0.9626251788268956, + "grad_norm": 0.48179184241667344, + "learning_rate": 7.318002498915122e-07, + "loss": 1.5022, + "step": 10766 + }, + { + "epoch": 0.9627145922746781, + "grad_norm": 0.32784327705851724, + "learning_rate": 7.283072529405521e-07, + "loss": 1.504, + "step": 10767 + }, + { + "epoch": 0.9628040057224606, + "grad_norm": 0.38868031115905544, + "learning_rate": 7.248225817865884e-07, + "loss": 1.4907, + "step": 10768 + }, + { + "epoch": 0.9628934191702432, + "grad_norm": 0.5978204905247071, + "learning_rate": 7.213462367218537e-07, + "loss": 1.2355, + "step": 10769 + }, + { + "epoch": 0.9629828326180258, + "grad_norm": 0.48687601850211426, + "learning_rate": 7.17878218037904e-07, + "loss": 1.4784, + "step": 10770 + }, + { + "epoch": 0.9630722460658083, + "grad_norm": 0.4192415177102242, + "learning_rate": 7.144185260256175e-07, + "loss": 1.4647, + "step": 10771 + }, + { + "epoch": 0.9631616595135909, + "grad_norm": 0.6177195965236566, + "learning_rate": 7.1096716097514e-07, + "loss": 1.7551, + "step": 10772 + }, + { + "epoch": 0.9632510729613734, + "grad_norm": 0.610014354479814, + "learning_rate": 7.075241231759289e-07, + "loss": 1.5451, + "step": 10773 + }, + { + "epoch": 0.9633404864091559, + "grad_norm": 0.6172702825400961, + "learning_rate": 7.040894129167641e-07, + "loss": 1.6096, + "step": 10774 + }, + { + "epoch": 0.9634298998569385, + "grad_norm": 0.2775144169444109, + "learning_rate": 7.006630304856932e-07, + "loss": 1.141, + "step": 10775 + }, + { + "epoch": 0.9635193133047211, + "grad_norm": 0.4721030423603108, + "learning_rate": 6.972449761700861e-07, + "loss": 1.3613, + "step": 10776 + }, + { + "epoch": 0.9636087267525035, + "grad_norm": 0.5944729904632428, + "learning_rate": 6.938352502566358e-07, + "loss": 1.9522, + "step": 10777 + }, + { + "epoch": 0.9636981402002861, + "grad_norm": 0.7228293261020664, + "learning_rate": 6.904338530312693e-07, + "loss": 1.8575, + "step": 10778 + }, + { + "epoch": 0.9637875536480687, + "grad_norm": 0.6295274468289875, + "learning_rate": 6.870407847792915e-07, + "loss": 1.4742, + "step": 10779 + }, + { + "epoch": 0.9638769670958512, + "grad_norm": 0.4105096718919977, + "learning_rate": 6.836560457852636e-07, + "loss": 1.3469, + "step": 10780 + }, + { + "epoch": 0.9639663805436338, + "grad_norm": 1.016668246239126, + "learning_rate": 6.802796363330588e-07, + "loss": 1.6762, + "step": 10781 + }, + { + "epoch": 0.9640557939914163, + "grad_norm": 0.5276102361848412, + "learning_rate": 6.769115567058504e-07, + "loss": 1.584, + "step": 10782 + }, + { + "epoch": 0.9641452074391988, + "grad_norm": 0.3858773469579858, + "learning_rate": 6.735518071861235e-07, + "loss": 1.2333, + "step": 10783 + }, + { + "epoch": 0.9642346208869814, + "grad_norm": 0.43676940403090975, + "learning_rate": 6.702003880556418e-07, + "loss": 1.455, + "step": 10784 + }, + { + "epoch": 0.964324034334764, + "grad_norm": 0.4271198949743528, + "learning_rate": 6.668572995955025e-07, + "loss": 1.3608, + "step": 10785 + }, + { + "epoch": 0.9644134477825465, + "grad_norm": 0.48281970731852764, + "learning_rate": 6.635225420860702e-07, + "loss": 1.4672, + "step": 10786 + }, + { + "epoch": 0.964502861230329, + "grad_norm": 0.3998716146429733, + "learning_rate": 6.601961158070325e-07, + "loss": 1.6686, + "step": 10787 + }, + { + "epoch": 0.9645922746781116, + "grad_norm": 0.48059236656123383, + "learning_rate": 6.56878021037377e-07, + "loss": 1.6865, + "step": 10788 + }, + { + "epoch": 0.9646816881258942, + "grad_norm": 0.4498104524409132, + "learning_rate": 6.535682580553926e-07, + "loss": 1.5417, + "step": 10789 + }, + { + "epoch": 0.9647711015736766, + "grad_norm": 0.4293595952344546, + "learning_rate": 6.502668271386458e-07, + "loss": 1.3607, + "step": 10790 + }, + { + "epoch": 0.9648605150214592, + "grad_norm": 0.4581560725463539, + "learning_rate": 6.469737285640487e-07, + "loss": 1.5182, + "step": 10791 + }, + { + "epoch": 0.9649499284692418, + "grad_norm": 0.5452348879558668, + "learning_rate": 6.436889626077691e-07, + "loss": 1.9729, + "step": 10792 + }, + { + "epoch": 0.9650393419170243, + "grad_norm": 0.5545185664939196, + "learning_rate": 6.40412529545309e-07, + "loss": 1.7211, + "step": 10793 + }, + { + "epoch": 0.9651287553648069, + "grad_norm": 0.45991652316858894, + "learning_rate": 6.371444296514484e-07, + "loss": 1.5475, + "step": 10794 + }, + { + "epoch": 0.9652181688125894, + "grad_norm": 0.38322719711848796, + "learning_rate": 6.338846632002904e-07, + "loss": 1.4744, + "step": 10795 + }, + { + "epoch": 0.9653075822603719, + "grad_norm": 0.4072647117058974, + "learning_rate": 6.306332304652273e-07, + "loss": 1.4701, + "step": 10796 + }, + { + "epoch": 0.9653969957081545, + "grad_norm": 0.44348532560148113, + "learning_rate": 6.273901317189301e-07, + "loss": 1.6396, + "step": 10797 + }, + { + "epoch": 0.9654864091559371, + "grad_norm": 0.6445821498160508, + "learning_rate": 6.241553672334255e-07, + "loss": 1.5032, + "step": 10798 + }, + { + "epoch": 0.9655758226037195, + "grad_norm": 0.5602248167870505, + "learning_rate": 6.209289372799854e-07, + "loss": 1.6591, + "step": 10799 + }, + { + "epoch": 0.9656652360515021, + "grad_norm": 0.42328613373609963, + "learning_rate": 6.177108421292266e-07, + "loss": 1.4837, + "step": 10800 + }, + { + "epoch": 0.9657546494992847, + "grad_norm": 0.3872843200009055, + "learning_rate": 6.145010820510222e-07, + "loss": 1.6678, + "step": 10801 + }, + { + "epoch": 0.9658440629470673, + "grad_norm": 0.33548743519097807, + "learning_rate": 6.112996573145902e-07, + "loss": 1.3783, + "step": 10802 + }, + { + "epoch": 0.9659334763948498, + "grad_norm": 0.44976633496186535, + "learning_rate": 6.081065681884268e-07, + "loss": 1.6441, + "step": 10803 + }, + { + "epoch": 0.9660228898426323, + "grad_norm": 0.5728215288164944, + "learning_rate": 6.04921814940329e-07, + "loss": 1.6541, + "step": 10804 + }, + { + "epoch": 0.9661123032904149, + "grad_norm": 0.5015105372050223, + "learning_rate": 6.017453978374055e-07, + "loss": 1.7299, + "step": 10805 + }, + { + "epoch": 0.9662017167381974, + "grad_norm": 0.6352027844439851, + "learning_rate": 5.985773171460429e-07, + "loss": 1.5521, + "step": 10806 + }, + { + "epoch": 0.96629113018598, + "grad_norm": 0.43376242180303154, + "learning_rate": 5.954175731319622e-07, + "loss": 1.3572, + "step": 10807 + }, + { + "epoch": 0.9663805436337625, + "grad_norm": 0.648009812707285, + "learning_rate": 5.922661660601514e-07, + "loss": 1.7979, + "step": 10808 + }, + { + "epoch": 0.966469957081545, + "grad_norm": 0.5828967704906894, + "learning_rate": 5.891230961949324e-07, + "loss": 1.6498, + "step": 10809 + }, + { + "epoch": 0.9665593705293276, + "grad_norm": 0.42628070805029306, + "learning_rate": 5.859883637998942e-07, + "loss": 1.7174, + "step": 10810 + }, + { + "epoch": 0.9666487839771102, + "grad_norm": 0.6864244996607645, + "learning_rate": 5.8286196913796e-07, + "loss": 1.4073, + "step": 10811 + }, + { + "epoch": 0.9667381974248928, + "grad_norm": 0.2987088244844177, + "learning_rate": 5.7974391247132e-07, + "loss": 1.2351, + "step": 10812 + }, + { + "epoch": 0.9668276108726752, + "grad_norm": 0.5933571549603227, + "learning_rate": 5.766341940614872e-07, + "loss": 1.6298, + "step": 10813 + }, + { + "epoch": 0.9669170243204578, + "grad_norm": 0.42881709781788785, + "learning_rate": 5.735328141692642e-07, + "loss": 1.2241, + "step": 10814 + }, + { + "epoch": 0.9670064377682404, + "grad_norm": 0.440333485144609, + "learning_rate": 5.704397730547762e-07, + "loss": 1.5741, + "step": 10815 + }, + { + "epoch": 0.9670958512160229, + "grad_norm": 0.8598529114130562, + "learning_rate": 5.673550709774267e-07, + "loss": 1.5218, + "step": 10816 + }, + { + "epoch": 0.9671852646638054, + "grad_norm": 0.4477616311169692, + "learning_rate": 5.6427870819592e-07, + "loss": 1.6869, + "step": 10817 + }, + { + "epoch": 0.967274678111588, + "grad_norm": 0.5424059534156889, + "learning_rate": 5.612106849682719e-07, + "loss": 1.9186, + "step": 10818 + }, + { + "epoch": 0.9673640915593705, + "grad_norm": 0.3858270312145225, + "learning_rate": 5.581510015517988e-07, + "loss": 1.2612, + "step": 10819 + }, + { + "epoch": 0.9674535050071531, + "grad_norm": 0.42758597981563995, + "learning_rate": 5.550996582030954e-07, + "loss": 1.3826, + "step": 10820 + }, + { + "epoch": 0.9675429184549357, + "grad_norm": 0.5309839052915337, + "learning_rate": 5.520566551780792e-07, + "loss": 1.4971, + "step": 10821 + }, + { + "epoch": 0.9676323319027181, + "grad_norm": 0.5613576340016884, + "learning_rate": 5.490219927319795e-07, + "loss": 1.5682, + "step": 10822 + }, + { + "epoch": 0.9677217453505007, + "grad_norm": 0.805053155646391, + "learning_rate": 5.459956711192926e-07, + "loss": 1.8, + "step": 10823 + }, + { + "epoch": 0.9678111587982833, + "grad_norm": 0.5806197214992196, + "learning_rate": 5.429776905938489e-07, + "loss": 1.4337, + "step": 10824 + }, + { + "epoch": 0.9679005722460658, + "grad_norm": 0.3972456020743506, + "learning_rate": 5.399680514087458e-07, + "loss": 1.5416, + "step": 10825 + }, + { + "epoch": 0.9679899856938483, + "grad_norm": 0.6099977894758091, + "learning_rate": 5.369667538164036e-07, + "loss": 1.6148, + "step": 10826 + }, + { + "epoch": 0.9680793991416309, + "grad_norm": 0.3872212075206633, + "learning_rate": 5.339737980685433e-07, + "loss": 1.4975, + "step": 10827 + }, + { + "epoch": 0.9681688125894135, + "grad_norm": 0.4792620100837379, + "learning_rate": 5.30989184416164e-07, + "loss": 1.758, + "step": 10828 + }, + { + "epoch": 0.968258226037196, + "grad_norm": 0.5244080022145402, + "learning_rate": 5.28012913109599e-07, + "loss": 1.3635, + "step": 10829 + }, + { + "epoch": 0.9683476394849786, + "grad_norm": 0.49083453069913874, + "learning_rate": 5.250449843984706e-07, + "loss": 1.6418, + "step": 10830 + }, + { + "epoch": 0.968437052932761, + "grad_norm": 0.5655700968586852, + "learning_rate": 5.220853985316798e-07, + "loss": 1.5706, + "step": 10831 + }, + { + "epoch": 0.9685264663805436, + "grad_norm": 0.44413342606129347, + "learning_rate": 5.191341557574392e-07, + "loss": 1.5355, + "step": 10832 + }, + { + "epoch": 0.9686158798283262, + "grad_norm": 0.3878284676338503, + "learning_rate": 5.16191256323273e-07, + "loss": 1.5417, + "step": 10833 + }, + { + "epoch": 0.9687052932761088, + "grad_norm": 0.6052424184825171, + "learning_rate": 5.132567004760169e-07, + "loss": 1.6028, + "step": 10834 + }, + { + "epoch": 0.9687947067238912, + "grad_norm": 0.3804146356240884, + "learning_rate": 5.103304884617521e-07, + "loss": 1.4165, + "step": 10835 + }, + { + "epoch": 0.9688841201716738, + "grad_norm": 0.45560581008916, + "learning_rate": 5.074126205259266e-07, + "loss": 1.7421, + "step": 10836 + }, + { + "epoch": 0.9689735336194564, + "grad_norm": 0.4209921923513746, + "learning_rate": 5.045030969132447e-07, + "loss": 1.4207, + "step": 10837 + }, + { + "epoch": 0.969062947067239, + "grad_norm": 0.2741686267364805, + "learning_rate": 5.016019178677333e-07, + "loss": 1.0816, + "step": 10838 + }, + { + "epoch": 0.9691523605150214, + "grad_norm": 0.32778997913728963, + "learning_rate": 4.987090836327091e-07, + "loss": 1.3829, + "step": 10839 + }, + { + "epoch": 0.969241773962804, + "grad_norm": 0.44866503758401527, + "learning_rate": 4.958245944507777e-07, + "loss": 1.3635, + "step": 10840 + }, + { + "epoch": 0.9693311874105865, + "grad_norm": 0.5090588120880932, + "learning_rate": 4.929484505638682e-07, + "loss": 1.4489, + "step": 10841 + }, + { + "epoch": 0.9694206008583691, + "grad_norm": 0.5824775231995495, + "learning_rate": 4.900806522131984e-07, + "loss": 1.7798, + "step": 10842 + }, + { + "epoch": 0.9695100143061517, + "grad_norm": 0.7587488109043936, + "learning_rate": 4.872211996392872e-07, + "loss": 1.8231, + "step": 10843 + }, + { + "epoch": 0.9695994277539342, + "grad_norm": 0.4352524089324384, + "learning_rate": 4.843700930819539e-07, + "loss": 1.5967, + "step": 10844 + }, + { + "epoch": 0.9696888412017167, + "grad_norm": 0.5040520868525135, + "learning_rate": 4.815273327803182e-07, + "loss": 1.7529, + "step": 10845 + }, + { + "epoch": 0.9697782546494993, + "grad_norm": 0.674137355575859, + "learning_rate": 4.786929189727896e-07, + "loss": 1.355, + "step": 10846 + }, + { + "epoch": 0.9698676680972819, + "grad_norm": 0.4669754172496015, + "learning_rate": 4.758668518970999e-07, + "loss": 1.7971, + "step": 10847 + }, + { + "epoch": 0.9699570815450643, + "grad_norm": 0.4625591362377614, + "learning_rate": 4.7304913179025965e-07, + "loss": 1.4508, + "step": 10848 + }, + { + "epoch": 0.9700464949928469, + "grad_norm": 0.5565265282139258, + "learning_rate": 4.7023975888859095e-07, + "loss": 1.4064, + "step": 10849 + }, + { + "epoch": 0.9701359084406295, + "grad_norm": 0.47179558511107944, + "learning_rate": 4.674387334277164e-07, + "loss": 1.5127, + "step": 10850 + }, + { + "epoch": 0.970225321888412, + "grad_norm": 0.6223900655330387, + "learning_rate": 4.6464605564254803e-07, + "loss": 1.7535, + "step": 10851 + }, + { + "epoch": 0.9703147353361946, + "grad_norm": 0.495570488009554, + "learning_rate": 4.6186172576730967e-07, + "loss": 1.668, + "step": 10852 + }, + { + "epoch": 0.9704041487839771, + "grad_norm": 0.40799090298303986, + "learning_rate": 4.5908574403551454e-07, + "loss": 1.756, + "step": 10853 + }, + { + "epoch": 0.9704935622317596, + "grad_norm": 0.36768754306949847, + "learning_rate": 4.5631811067998743e-07, + "loss": 1.6014, + "step": 10854 + }, + { + "epoch": 0.9705829756795422, + "grad_norm": 0.3933906216109812, + "learning_rate": 4.5355882593283163e-07, + "loss": 1.4333, + "step": 10855 + }, + { + "epoch": 0.9706723891273248, + "grad_norm": 0.6895429931864561, + "learning_rate": 4.5080789002548417e-07, + "loss": 1.6287, + "step": 10856 + }, + { + "epoch": 0.9707618025751072, + "grad_norm": 0.5556052449798287, + "learning_rate": 4.4806530318864945e-07, + "loss": 1.6703, + "step": 10857 + }, + { + "epoch": 0.9708512160228898, + "grad_norm": 0.5715658454137521, + "learning_rate": 4.453310656523435e-07, + "loss": 1.5765, + "step": 10858 + }, + { + "epoch": 0.9709406294706724, + "grad_norm": 0.6634193857652306, + "learning_rate": 4.42605177645905e-07, + "loss": 1.6725, + "step": 10859 + }, + { + "epoch": 0.971030042918455, + "grad_norm": 0.6047982541628615, + "learning_rate": 4.39887639397929e-07, + "loss": 1.5254, + "step": 10860 + }, + { + "epoch": 0.9711194563662375, + "grad_norm": 0.630452106509254, + "learning_rate": 4.3717845113633307e-07, + "loss": 1.7535, + "step": 10861 + }, + { + "epoch": 0.97120886981402, + "grad_norm": 0.42440386799012353, + "learning_rate": 4.344776130883466e-07, + "loss": 1.6288, + "step": 10862 + }, + { + "epoch": 0.9712982832618026, + "grad_norm": 0.5688954300439865, + "learning_rate": 4.3178512548046613e-07, + "loss": 1.6885, + "step": 10863 + }, + { + "epoch": 0.9713876967095851, + "grad_norm": 0.3966837188603024, + "learning_rate": 4.291009885385333e-07, + "loss": 1.7349, + "step": 10864 + }, + { + "epoch": 0.9714771101573677, + "grad_norm": 0.4278605335205266, + "learning_rate": 4.264252024876458e-07, + "loss": 1.6618, + "step": 10865 + }, + { + "epoch": 0.9715665236051502, + "grad_norm": 0.40343238269329496, + "learning_rate": 4.237577675522131e-07, + "loss": 1.4349, + "step": 10866 + }, + { + "epoch": 0.9716559370529327, + "grad_norm": 0.44351300620908335, + "learning_rate": 4.210986839559672e-07, + "loss": 1.4342, + "step": 10867 + }, + { + "epoch": 0.9717453505007153, + "grad_norm": 0.5355212086471812, + "learning_rate": 4.184479519219187e-07, + "loss": 1.8135, + "step": 10868 + }, + { + "epoch": 0.9718347639484979, + "grad_norm": 0.47335919350919287, + "learning_rate": 4.1580557167236744e-07, + "loss": 1.7212, + "step": 10869 + }, + { + "epoch": 0.9719241773962805, + "grad_norm": 0.5870807909781257, + "learning_rate": 4.131715434289363e-07, + "loss": 1.7615, + "step": 10870 + }, + { + "epoch": 0.9720135908440629, + "grad_norm": 0.5465893140816073, + "learning_rate": 4.105458674125373e-07, + "loss": 1.4772, + "step": 10871 + }, + { + "epoch": 0.9721030042918455, + "grad_norm": 0.8784575897497048, + "learning_rate": 4.0792854384338333e-07, + "loss": 6.0223, + "step": 10872 + }, + { + "epoch": 0.9721924177396281, + "grad_norm": 0.5710768345063663, + "learning_rate": 4.0531957294098755e-07, + "loss": 1.5596, + "step": 10873 + }, + { + "epoch": 0.9722818311874106, + "grad_norm": 0.6686480481131685, + "learning_rate": 4.027189549241639e-07, + "loss": 1.4735, + "step": 10874 + }, + { + "epoch": 0.9723712446351931, + "grad_norm": 0.3675725825435639, + "learning_rate": 4.001266900110046e-07, + "loss": 1.7481, + "step": 10875 + }, + { + "epoch": 0.9724606580829757, + "grad_norm": 0.9757116107388132, + "learning_rate": 3.975427784189467e-07, + "loss": 1.6735, + "step": 10876 + }, + { + "epoch": 0.9725500715307582, + "grad_norm": 0.32448679604050057, + "learning_rate": 3.949672203646837e-07, + "loss": 1.2454, + "step": 10877 + }, + { + "epoch": 0.9726394849785408, + "grad_norm": 0.5862705591103118, + "learning_rate": 3.924000160642205e-07, + "loss": 1.5761, + "step": 10878 + }, + { + "epoch": 0.9727288984263234, + "grad_norm": 0.4463593464317693, + "learning_rate": 3.898411657328849e-07, + "loss": 1.115, + "step": 10879 + }, + { + "epoch": 0.9728183118741058, + "grad_norm": 0.7085700953031314, + "learning_rate": 3.872906695852607e-07, + "loss": 1.5352, + "step": 10880 + }, + { + "epoch": 0.9729077253218884, + "grad_norm": 0.42014086794920225, + "learning_rate": 3.847485278352658e-07, + "loss": 1.6355, + "step": 10881 + }, + { + "epoch": 0.972997138769671, + "grad_norm": 0.43445535886698994, + "learning_rate": 3.8221474069611854e-07, + "loss": 1.6316, + "step": 10882 + }, + { + "epoch": 0.9730865522174535, + "grad_norm": 0.4973203034952357, + "learning_rate": 3.7968930838030436e-07, + "loss": 1.4808, + "step": 10883 + }, + { + "epoch": 0.973175965665236, + "grad_norm": 0.4978253960280166, + "learning_rate": 3.771722310996428e-07, + "loss": 1.618, + "step": 10884 + }, + { + "epoch": 0.9732653791130186, + "grad_norm": 0.6430055414064979, + "learning_rate": 3.7466350906522065e-07, + "loss": 1.7568, + "step": 10885 + }, + { + "epoch": 0.9733547925608012, + "grad_norm": 0.5865434825336844, + "learning_rate": 3.721631424874694e-07, + "loss": 1.4682, + "step": 10886 + }, + { + "epoch": 0.9734442060085837, + "grad_norm": 1.0971807325621945, + "learning_rate": 3.696711315760659e-07, + "loss": 6.0656, + "step": 10887 + }, + { + "epoch": 0.9735336194563662, + "grad_norm": 0.46284926313020697, + "learning_rate": 3.671874765400207e-07, + "loss": 1.6525, + "step": 10888 + }, + { + "epoch": 0.9736230329041488, + "grad_norm": 0.6751550635330466, + "learning_rate": 3.6471217758763387e-07, + "loss": 1.3591, + "step": 10889 + }, + { + "epoch": 0.9737124463519313, + "grad_norm": 0.3080557048100687, + "learning_rate": 3.6224523492651706e-07, + "loss": 1.3567, + "step": 10890 + }, + { + "epoch": 0.9738018597997139, + "grad_norm": 0.43223012019377627, + "learning_rate": 3.5978664876354926e-07, + "loss": 1.2333, + "step": 10891 + }, + { + "epoch": 0.9738912732474965, + "grad_norm": 0.5733050884707181, + "learning_rate": 3.573364193049433e-07, + "loss": 1.7363, + "step": 10892 + }, + { + "epoch": 0.9739806866952789, + "grad_norm": 0.45813288483519854, + "learning_rate": 3.5489454675620147e-07, + "loss": 1.4472, + "step": 10893 + }, + { + "epoch": 0.9740701001430615, + "grad_norm": 0.48209758466103997, + "learning_rate": 3.524610313221155e-07, + "loss": 1.9437, + "step": 10894 + }, + { + "epoch": 0.9741595135908441, + "grad_norm": 0.41846017335471647, + "learning_rate": 3.5003587320676655e-07, + "loss": 1.6079, + "step": 10895 + }, + { + "epoch": 0.9742489270386266, + "grad_norm": 0.6083119774385431, + "learning_rate": 3.4761907261356976e-07, + "loss": 1.3025, + "step": 10896 + }, + { + "epoch": 0.9743383404864091, + "grad_norm": 0.3994148800488398, + "learning_rate": 3.4521062974520737e-07, + "loss": 1.4494, + "step": 10897 + }, + { + "epoch": 0.9744277539341917, + "grad_norm": 0.474059807092929, + "learning_rate": 3.4281054480368445e-07, + "loss": 1.5262, + "step": 10898 + }, + { + "epoch": 0.9745171673819742, + "grad_norm": 0.3002277196724023, + "learning_rate": 3.404188179902845e-07, + "loss": 1.0944, + "step": 10899 + }, + { + "epoch": 0.9746065808297568, + "grad_norm": 0.532055206727455, + "learning_rate": 3.380354495055915e-07, + "loss": 1.5327, + "step": 10900 + }, + { + "epoch": 0.9746959942775394, + "grad_norm": 0.6214335565843687, + "learning_rate": 3.356604395495122e-07, + "loss": 1.7231, + "step": 10901 + }, + { + "epoch": 0.9747854077253219, + "grad_norm": 0.6498553510278732, + "learning_rate": 3.332937883212206e-07, + "loss": 1.4407, + "step": 10902 + }, + { + "epoch": 0.9748748211731044, + "grad_norm": 0.4705064604385302, + "learning_rate": 3.3093549601921345e-07, + "loss": 1.5222, + "step": 10903 + }, + { + "epoch": 0.974964234620887, + "grad_norm": 0.4747753851868555, + "learning_rate": 3.2858556284127704e-07, + "loss": 1.5791, + "step": 10904 + }, + { + "epoch": 0.9750536480686696, + "grad_norm": 0.4933254904339153, + "learning_rate": 3.2624398898449814e-07, + "loss": 1.3781, + "step": 10905 + }, + { + "epoch": 0.975143061516452, + "grad_norm": 0.46456552151473135, + "learning_rate": 3.239107746452641e-07, + "loss": 1.669, + "step": 10906 + }, + { + "epoch": 0.9752324749642346, + "grad_norm": 0.6814849316414058, + "learning_rate": 3.215859200192517e-07, + "loss": 1.8821, + "step": 10907 + }, + { + "epoch": 0.9753218884120172, + "grad_norm": 0.4510968049739439, + "learning_rate": 3.1926942530144945e-07, + "loss": 1.4357, + "step": 10908 + }, + { + "epoch": 0.9754113018597997, + "grad_norm": 0.45324463137733767, + "learning_rate": 3.1696129068613525e-07, + "loss": 1.3716, + "step": 10909 + }, + { + "epoch": 0.9755007153075823, + "grad_norm": 0.46134598558308715, + "learning_rate": 3.1466151636689865e-07, + "loss": 1.6537, + "step": 10910 + }, + { + "epoch": 0.9755901287553648, + "grad_norm": 0.49158162626870855, + "learning_rate": 3.1237010253659657e-07, + "loss": 1.765, + "step": 10911 + }, + { + "epoch": 0.9756795422031473, + "grad_norm": 0.47837710365428815, + "learning_rate": 3.1008704938743084e-07, + "loss": 1.5794, + "step": 10912 + }, + { + "epoch": 0.9757689556509299, + "grad_norm": 0.5682721596099379, + "learning_rate": 3.078123571108704e-07, + "loss": 1.4538, + "step": 10913 + }, + { + "epoch": 0.9758583690987125, + "grad_norm": 0.41274299321360336, + "learning_rate": 3.05546025897685e-07, + "loss": 1.604, + "step": 10914 + }, + { + "epoch": 0.975947782546495, + "grad_norm": 0.47160124320632674, + "learning_rate": 3.0328805593795584e-07, + "loss": 1.3951, + "step": 10915 + }, + { + "epoch": 0.9760371959942775, + "grad_norm": 0.5635944543476841, + "learning_rate": 3.010384474210537e-07, + "loss": 1.6554, + "step": 10916 + }, + { + "epoch": 0.9761266094420601, + "grad_norm": 0.5299504346848761, + "learning_rate": 2.987972005356499e-07, + "loss": 1.4847, + "step": 10917 + }, + { + "epoch": 0.9762160228898427, + "grad_norm": 0.7462397056280106, + "learning_rate": 2.965643154697162e-07, + "loss": 1.6451, + "step": 10918 + }, + { + "epoch": 0.9763054363376252, + "grad_norm": 0.9613630282686434, + "learning_rate": 2.943397924105251e-07, + "loss": 5.9641, + "step": 10919 + }, + { + "epoch": 0.9763948497854077, + "grad_norm": 0.429365338065213, + "learning_rate": 2.921236315446385e-07, + "loss": 1.3577, + "step": 10920 + }, + { + "epoch": 0.9764842632331903, + "grad_norm": 0.38493052108537634, + "learning_rate": 2.899158330579299e-07, + "loss": 1.2239, + "step": 10921 + }, + { + "epoch": 0.9765736766809728, + "grad_norm": 0.40858535226489906, + "learning_rate": 2.877163971355623e-07, + "loss": 1.5346, + "step": 10922 + }, + { + "epoch": 0.9766630901287554, + "grad_norm": 0.800423707123127, + "learning_rate": 2.8552532396198815e-07, + "loss": 1.6862, + "step": 10923 + }, + { + "epoch": 0.9767525035765379, + "grad_norm": 0.454895883033421, + "learning_rate": 2.833426137209938e-07, + "loss": 1.8931, + "step": 10924 + }, + { + "epoch": 0.9768419170243204, + "grad_norm": 0.4083903107994011, + "learning_rate": 2.811682665956217e-07, + "loss": 1.9159, + "step": 10925 + }, + { + "epoch": 0.976931330472103, + "grad_norm": 0.5664679918713468, + "learning_rate": 2.7900228276823704e-07, + "loss": 1.9456, + "step": 10926 + }, + { + "epoch": 0.9770207439198856, + "grad_norm": 0.5429026595931923, + "learning_rate": 2.768446624204946e-07, + "loss": 1.6304, + "step": 10927 + }, + { + "epoch": 0.977110157367668, + "grad_norm": 0.5746052271793509, + "learning_rate": 2.746954057333606e-07, + "loss": 1.8567, + "step": 10928 + }, + { + "epoch": 0.9771995708154506, + "grad_norm": 0.4249742446121491, + "learning_rate": 2.7255451288707987e-07, + "loss": 1.6393, + "step": 10929 + }, + { + "epoch": 0.9772889842632332, + "grad_norm": 0.5385004512929104, + "learning_rate": 2.704219840612199e-07, + "loss": 1.8177, + "step": 10930 + }, + { + "epoch": 0.9773783977110158, + "grad_norm": 0.6082630925679311, + "learning_rate": 2.682978194346264e-07, + "loss": 1.5963, + "step": 10931 + }, + { + "epoch": 0.9774678111587983, + "grad_norm": 0.3642941909901673, + "learning_rate": 2.661820191854347e-07, + "loss": 1.4452, + "step": 10932 + }, + { + "epoch": 0.9775572246065808, + "grad_norm": 0.44116681119661433, + "learning_rate": 2.640745834911251e-07, + "loss": 1.9271, + "step": 10933 + }, + { + "epoch": 0.9776466380543634, + "grad_norm": 0.45931869778850104, + "learning_rate": 2.6197551252842287e-07, + "loss": 1.5384, + "step": 10934 + }, + { + "epoch": 0.9777360515021459, + "grad_norm": 0.39769232134408616, + "learning_rate": 2.598848064733761e-07, + "loss": 1.6405, + "step": 10935 + }, + { + "epoch": 0.9778254649499285, + "grad_norm": 0.7208761004636377, + "learning_rate": 2.5780246550134444e-07, + "loss": 1.6812, + "step": 10936 + }, + { + "epoch": 0.977914878397711, + "grad_norm": 0.5015150891657784, + "learning_rate": 2.5572848978695496e-07, + "loss": 1.8846, + "step": 10937 + }, + { + "epoch": 0.9780042918454935, + "grad_norm": 0.5010575185812672, + "learning_rate": 2.5366287950415737e-07, + "loss": 1.5551, + "step": 10938 + }, + { + "epoch": 0.9780937052932761, + "grad_norm": 0.48136218001311304, + "learning_rate": 2.516056348261908e-07, + "loss": 1.8985, + "step": 10939 + }, + { + "epoch": 0.9781831187410587, + "grad_norm": 0.3817762443806319, + "learning_rate": 2.495567559256062e-07, + "loss": 1.5871, + "step": 10940 + }, + { + "epoch": 0.9782725321888412, + "grad_norm": 0.4686408779051565, + "learning_rate": 2.475162429742106e-07, + "loss": 1.9513, + "step": 10941 + }, + { + "epoch": 0.9783619456366237, + "grad_norm": 0.7353890778358371, + "learning_rate": 2.45484096143167e-07, + "loss": 1.6996, + "step": 10942 + }, + { + "epoch": 0.9784513590844063, + "grad_norm": 0.45434915494172207, + "learning_rate": 2.434603156028947e-07, + "loss": 1.7341, + "step": 10943 + }, + { + "epoch": 0.9785407725321889, + "grad_norm": 0.40106222820089765, + "learning_rate": 2.414449015231357e-07, + "loss": 1.4438, + "step": 10944 + }, + { + "epoch": 0.9786301859799714, + "grad_norm": 0.4218602630240065, + "learning_rate": 2.394378540729214e-07, + "loss": 1.731, + "step": 10945 + }, + { + "epoch": 0.9787195994277539, + "grad_norm": 0.5328500582088457, + "learning_rate": 2.3743917342056166e-07, + "loss": 1.5693, + "step": 10946 + }, + { + "epoch": 0.9788090128755365, + "grad_norm": 0.3762880682175113, + "learning_rate": 2.3544885973370012e-07, + "loss": 1.6033, + "step": 10947 + }, + { + "epoch": 0.978898426323319, + "grad_norm": 0.4031395635356563, + "learning_rate": 2.3346691317924775e-07, + "loss": 1.3916, + "step": 10948 + }, + { + "epoch": 0.9789878397711016, + "grad_norm": 0.5494624510378806, + "learning_rate": 2.314933339234493e-07, + "loss": 1.4508, + "step": 10949 + }, + { + "epoch": 0.9790772532188842, + "grad_norm": 0.49069470105250923, + "learning_rate": 2.2952812213181684e-07, + "loss": 1.856, + "step": 10950 + }, + { + "epoch": 0.9791666666666666, + "grad_norm": 0.3565428432037383, + "learning_rate": 2.27571277969163e-07, + "loss": 1.3921, + "step": 10951 + }, + { + "epoch": 0.9792560801144492, + "grad_norm": 0.31756319254499144, + "learning_rate": 2.2562280159961203e-07, + "loss": 1.1121, + "step": 10952 + }, + { + "epoch": 0.9793454935622318, + "grad_norm": 0.789682352153936, + "learning_rate": 2.2368269318657764e-07, + "loss": 5.9936, + "step": 10953 + }, + { + "epoch": 0.9794349070100143, + "grad_norm": 0.5710480427382039, + "learning_rate": 2.2175095289278524e-07, + "loss": 1.6745, + "step": 10954 + }, + { + "epoch": 0.9795243204577968, + "grad_norm": 0.7581879021265204, + "learning_rate": 2.1982758088022747e-07, + "loss": 6.0242, + "step": 10955 + }, + { + "epoch": 0.9796137339055794, + "grad_norm": 0.4856758640097872, + "learning_rate": 2.1791257731024194e-07, + "loss": 1.7567, + "step": 10956 + }, + { + "epoch": 0.979703147353362, + "grad_norm": 0.48119198630032894, + "learning_rate": 2.160059423434113e-07, + "loss": 1.5996, + "step": 10957 + }, + { + "epoch": 0.9797925608011445, + "grad_norm": 0.41847305280450997, + "learning_rate": 2.141076761396521e-07, + "loss": 1.5677, + "step": 10958 + }, + { + "epoch": 0.9798819742489271, + "grad_norm": 0.4438566761900103, + "learning_rate": 2.1221777885817028e-07, + "loss": 1.4094, + "step": 10959 + }, + { + "epoch": 0.9799713876967096, + "grad_norm": 0.46824585223649023, + "learning_rate": 2.1033625065747242e-07, + "loss": 1.4448, + "step": 10960 + }, + { + "epoch": 0.9800608011444921, + "grad_norm": 0.4403492055193765, + "learning_rate": 2.084630916953656e-07, + "loss": 1.9244, + "step": 10961 + }, + { + "epoch": 0.9801502145922747, + "grad_norm": 0.7212310151387569, + "learning_rate": 2.0659830212893527e-07, + "loss": 1.8138, + "step": 10962 + }, + { + "epoch": 0.9802396280400573, + "grad_norm": 0.5577225083406591, + "learning_rate": 2.0474188211457856e-07, + "loss": 1.8312, + "step": 10963 + }, + { + "epoch": 0.9803290414878397, + "grad_norm": 0.4192212106148331, + "learning_rate": 2.0289383180801537e-07, + "loss": 1.4093, + "step": 10964 + }, + { + "epoch": 0.9804184549356223, + "grad_norm": 0.5618246683705693, + "learning_rate": 2.0105415136421058e-07, + "loss": 1.6639, + "step": 10965 + }, + { + "epoch": 0.9805078683834049, + "grad_norm": 0.6704949084191312, + "learning_rate": 1.9922284093746302e-07, + "loss": 1.2351, + "step": 10966 + }, + { + "epoch": 0.9805972818311874, + "grad_norm": 0.38800850453120445, + "learning_rate": 1.9739990068137203e-07, + "loss": 1.4857, + "step": 10967 + }, + { + "epoch": 0.98068669527897, + "grad_norm": 0.5206043266688591, + "learning_rate": 1.9558533074882646e-07, + "loss": 1.3712, + "step": 10968 + }, + { + "epoch": 0.9807761087267525, + "grad_norm": 0.9129707808702131, + "learning_rate": 1.9377913129199344e-07, + "loss": 1.8038, + "step": 10969 + }, + { + "epoch": 0.980865522174535, + "grad_norm": 0.42188831049124653, + "learning_rate": 1.919813024623851e-07, + "loss": 1.7445, + "step": 10970 + }, + { + "epoch": 0.9809549356223176, + "grad_norm": 0.4214705249313667, + "learning_rate": 1.9019184441075865e-07, + "loss": 1.3107, + "step": 10971 + }, + { + "epoch": 0.9810443490701002, + "grad_norm": 0.3889377977578689, + "learning_rate": 1.8841075728719404e-07, + "loss": 1.5987, + "step": 10972 + }, + { + "epoch": 0.9811337625178826, + "grad_norm": 0.4555439532761677, + "learning_rate": 1.8663804124108286e-07, + "loss": 1.4828, + "step": 10973 + }, + { + "epoch": 0.9812231759656652, + "grad_norm": 0.5299836203584023, + "learning_rate": 1.848736964211062e-07, + "loss": 1.4676, + "step": 10974 + }, + { + "epoch": 0.9813125894134478, + "grad_norm": 0.45663631833457113, + "learning_rate": 1.8311772297521234e-07, + "loss": 1.6773, + "step": 10975 + }, + { + "epoch": 0.9814020028612304, + "grad_norm": 0.4223667713962302, + "learning_rate": 1.813701210506946e-07, + "loss": 1.7366, + "step": 10976 + }, + { + "epoch": 0.9814914163090128, + "grad_norm": 0.5629326493244139, + "learning_rate": 1.7963089079411356e-07, + "loss": 1.4174, + "step": 10977 + }, + { + "epoch": 0.9815808297567954, + "grad_norm": 0.42479612323546195, + "learning_rate": 1.7790003235134133e-07, + "loss": 1.574, + "step": 10978 + }, + { + "epoch": 0.981670243204578, + "grad_norm": 0.5347246063795184, + "learning_rate": 1.7617754586752855e-07, + "loss": 1.7482, + "step": 10979 + }, + { + "epoch": 0.9817596566523605, + "grad_norm": 0.46369516102462943, + "learning_rate": 1.744634314871485e-07, + "loss": 1.3012, + "step": 10980 + }, + { + "epoch": 0.9818490701001431, + "grad_norm": 0.8000244439919386, + "learning_rate": 1.7275768935397507e-07, + "loss": 6.088, + "step": 10981 + }, + { + "epoch": 0.9819384835479256, + "grad_norm": 0.7385137016296269, + "learning_rate": 1.710603196110383e-07, + "loss": 1.815, + "step": 10982 + }, + { + "epoch": 0.9820278969957081, + "grad_norm": 0.36101882827984827, + "learning_rate": 1.693713224007243e-07, + "loss": 1.691, + "step": 10983 + }, + { + "epoch": 0.9821173104434907, + "grad_norm": 0.4735259340354632, + "learning_rate": 1.6769069786466418e-07, + "loss": 1.7685, + "step": 10984 + }, + { + "epoch": 0.9822067238912733, + "grad_norm": 0.6084111621108012, + "learning_rate": 1.66018446143823e-07, + "loss": 1.9602, + "step": 10985 + }, + { + "epoch": 0.9822961373390557, + "grad_norm": 0.5421789491210134, + "learning_rate": 1.6435456737843302e-07, + "loss": 1.9371, + "step": 10986 + }, + { + "epoch": 0.9823855507868383, + "grad_norm": 0.7992204563642753, + "learning_rate": 1.6269906170807148e-07, + "loss": 1.8504, + "step": 10987 + }, + { + "epoch": 0.9824749642346209, + "grad_norm": 0.4115053507853648, + "learning_rate": 1.6105192927154956e-07, + "loss": 1.3233, + "step": 10988 + }, + { + "epoch": 0.9825643776824035, + "grad_norm": 0.4805145077388102, + "learning_rate": 1.594131702070345e-07, + "loss": 1.6307, + "step": 10989 + }, + { + "epoch": 0.982653791130186, + "grad_norm": 0.3777112572136031, + "learning_rate": 1.5778278465197194e-07, + "loss": 1.7146, + "step": 10990 + }, + { + "epoch": 0.9827432045779685, + "grad_norm": 0.3724506435541902, + "learning_rate": 1.5616077274307473e-07, + "loss": 1.2428, + "step": 10991 + }, + { + "epoch": 0.9828326180257511, + "grad_norm": 0.5345930720370106, + "learning_rate": 1.545471346164007e-07, + "loss": 1.5482, + "step": 10992 + }, + { + "epoch": 0.9829220314735336, + "grad_norm": 0.7096259518100138, + "learning_rate": 1.5294187040726382e-07, + "loss": 1.886, + "step": 10993 + }, + { + "epoch": 0.9830114449213162, + "grad_norm": 0.6162680992587619, + "learning_rate": 1.5134498025031196e-07, + "loss": 1.7701, + "step": 10994 + }, + { + "epoch": 0.9831008583690987, + "grad_norm": 0.4456397562694238, + "learning_rate": 1.4975646427948244e-07, + "loss": 1.4508, + "step": 10995 + }, + { + "epoch": 0.9831902718168812, + "grad_norm": 0.47465398351088206, + "learning_rate": 1.4817632262797976e-07, + "loss": 1.7475, + "step": 10996 + }, + { + "epoch": 0.9832796852646638, + "grad_norm": 0.3625485563105599, + "learning_rate": 1.4660455542833128e-07, + "loss": 1.298, + "step": 10997 + }, + { + "epoch": 0.9833690987124464, + "grad_norm": 0.559413450364436, + "learning_rate": 1.45041162812376e-07, + "loss": 1.3849, + "step": 10998 + }, + { + "epoch": 0.983458512160229, + "grad_norm": 0.3778602528767451, + "learning_rate": 1.4348614491123125e-07, + "loss": 1.7624, + "step": 10999 + }, + { + "epoch": 0.9835479256080114, + "grad_norm": 0.32556151933159383, + "learning_rate": 1.419395018552927e-07, + "loss": 1.1869, + "step": 11000 + }, + { + "epoch": 0.983637339055794, + "grad_norm": 0.6072505821710176, + "learning_rate": 1.4040123377428993e-07, + "loss": 6.0707, + "step": 11001 + }, + { + "epoch": 0.9837267525035766, + "grad_norm": 0.4334513953412744, + "learning_rate": 1.3887134079724196e-07, + "loss": 1.501, + "step": 11002 + }, + { + "epoch": 0.9838161659513591, + "grad_norm": 0.45719348103105245, + "learning_rate": 1.3734982305245724e-07, + "loss": 1.5189, + "step": 11003 + }, + { + "epoch": 0.9839055793991416, + "grad_norm": 1.1654471255141698, + "learning_rate": 1.3583668066753375e-07, + "loss": 6.0476, + "step": 11004 + }, + { + "epoch": 0.9839949928469242, + "grad_norm": 0.4729713052451713, + "learning_rate": 1.3433191376938103e-07, + "loss": 1.4731, + "step": 11005 + }, + { + "epoch": 0.9840844062947067, + "grad_norm": 0.44305612871930217, + "learning_rate": 1.3283552248420927e-07, + "loss": 1.4745, + "step": 11006 + }, + { + "epoch": 0.9841738197424893, + "grad_norm": 0.5499413548027206, + "learning_rate": 1.3134750693751806e-07, + "loss": 1.3924, + "step": 11007 + }, + { + "epoch": 0.9842632331902719, + "grad_norm": 0.41045991175600066, + "learning_rate": 1.298678672540854e-07, + "loss": 1.5518, + "step": 11008 + }, + { + "epoch": 0.9843526466380543, + "grad_norm": 0.564859446306711, + "learning_rate": 1.2839660355803417e-07, + "loss": 1.5731, + "step": 11009 + }, + { + "epoch": 0.9844420600858369, + "grad_norm": 0.4529000005345977, + "learning_rate": 1.2693371597273241e-07, + "loss": 1.7034, + "step": 11010 + }, + { + "epoch": 0.9845314735336195, + "grad_norm": 0.542532866622761, + "learning_rate": 1.2547920462089302e-07, + "loss": 1.5416, + "step": 11011 + }, + { + "epoch": 0.984620886981402, + "grad_norm": 0.4442893817786571, + "learning_rate": 1.2403306962449624e-07, + "loss": 1.5076, + "step": 11012 + }, + { + "epoch": 0.9847103004291845, + "grad_norm": 0.410652001212386, + "learning_rate": 1.225953111048228e-07, + "loss": 1.5367, + "step": 11013 + }, + { + "epoch": 0.9847997138769671, + "grad_norm": 0.5962995005171094, + "learning_rate": 1.2116592918246516e-07, + "loss": 1.3663, + "step": 11014 + }, + { + "epoch": 0.9848891273247496, + "grad_norm": 0.4208961769356665, + "learning_rate": 1.197449239772941e-07, + "loss": 1.5319, + "step": 11015 + }, + { + "epoch": 0.9849785407725322, + "grad_norm": 0.7044170934223689, + "learning_rate": 1.1833229560848092e-07, + "loss": 1.8103, + "step": 11016 + }, + { + "epoch": 0.9850679542203148, + "grad_norm": 0.5277312985217072, + "learning_rate": 1.1692804419451975e-07, + "loss": 1.7133, + "step": 11017 + }, + { + "epoch": 0.9851573676680973, + "grad_norm": 0.5188006130166614, + "learning_rate": 1.1553216985318305e-07, + "loss": 1.6549, + "step": 11018 + }, + { + "epoch": 0.9852467811158798, + "grad_norm": 0.4267787088836138, + "learning_rate": 1.1414467270152163e-07, + "loss": 1.6974, + "step": 11019 + }, + { + "epoch": 0.9853361945636624, + "grad_norm": 0.433568195572592, + "learning_rate": 1.1276555285592017e-07, + "loss": 1.3732, + "step": 11020 + }, + { + "epoch": 0.985425608011445, + "grad_norm": 0.3975639868965074, + "learning_rate": 1.113948104320417e-07, + "loss": 1.556, + "step": 11021 + }, + { + "epoch": 0.9855150214592274, + "grad_norm": 0.44321441984865234, + "learning_rate": 1.1003244554483871e-07, + "loss": 1.303, + "step": 11022 + }, + { + "epoch": 0.98560443490701, + "grad_norm": 0.4320349342668757, + "learning_rate": 1.0867845830858647e-07, + "loss": 1.7285, + "step": 11023 + }, + { + "epoch": 0.9856938483547926, + "grad_norm": 0.44792900981276784, + "learning_rate": 1.0733284883682749e-07, + "loss": 1.2835, + "step": 11024 + }, + { + "epoch": 0.9857832618025751, + "grad_norm": 0.5078955914476221, + "learning_rate": 1.0599561724242702e-07, + "loss": 1.9117, + "step": 11025 + }, + { + "epoch": 0.9858726752503576, + "grad_norm": 0.48540163044421863, + "learning_rate": 1.046667636375287e-07, + "loss": 1.8042, + "step": 11026 + }, + { + "epoch": 0.9859620886981402, + "grad_norm": 0.802422133041355, + "learning_rate": 1.0334628813358782e-07, + "loss": 6.0333, + "step": 11027 + }, + { + "epoch": 0.9860515021459227, + "grad_norm": 0.4279447838185764, + "learning_rate": 1.0203419084134913e-07, + "loss": 1.5597, + "step": 11028 + }, + { + "epoch": 0.9861409155937053, + "grad_norm": 0.44122542740838544, + "learning_rate": 1.0073047187085794e-07, + "loss": 1.7194, + "step": 11029 + }, + { + "epoch": 0.9862303290414879, + "grad_norm": 0.33194192440421844, + "learning_rate": 9.9435131331449e-08, + "loss": 1.2882, + "step": 11030 + }, + { + "epoch": 0.9863197424892703, + "grad_norm": 0.35793905357859934, + "learning_rate": 9.814816933176874e-08, + "loss": 1.3667, + "step": 11031 + }, + { + "epoch": 0.9864091559370529, + "grad_norm": 0.3855104699575225, + "learning_rate": 9.686958597975304e-08, + "loss": 1.4721, + "step": 11032 + }, + { + "epoch": 0.9864985693848355, + "grad_norm": 0.3177004713145934, + "learning_rate": 9.559938138263836e-08, + "loss": 1.2606, + "step": 11033 + }, + { + "epoch": 0.9865879828326181, + "grad_norm": 0.33344368355231435, + "learning_rate": 9.433755564693947e-08, + "loss": 1.3376, + "step": 11034 + }, + { + "epoch": 0.9866773962804005, + "grad_norm": 0.5625549793283228, + "learning_rate": 9.308410887849394e-08, + "loss": 1.6995, + "step": 11035 + }, + { + "epoch": 0.9867668097281831, + "grad_norm": 0.4471073070866215, + "learning_rate": 9.18390411824288e-08, + "loss": 1.1917, + "step": 11036 + }, + { + "epoch": 0.9868562231759657, + "grad_norm": 0.6618278136882334, + "learning_rate": 9.060235266317163e-08, + "loss": 1.677, + "step": 11037 + }, + { + "epoch": 0.9869456366237482, + "grad_norm": 0.403636405074615, + "learning_rate": 8.937404342442834e-08, + "loss": 1.4907, + "step": 11038 + }, + { + "epoch": 0.9870350500715308, + "grad_norm": 0.6077203915348206, + "learning_rate": 8.815411356922764e-08, + "loss": 1.5398, + "step": 11039 + }, + { + "epoch": 0.9871244635193133, + "grad_norm": 0.46657086142306053, + "learning_rate": 8.694256319987659e-08, + "loss": 1.5459, + "step": 11040 + }, + { + "epoch": 0.9872138769670958, + "grad_norm": 0.6055871965212241, + "learning_rate": 8.573939241798278e-08, + "loss": 1.6873, + "step": 11041 + }, + { + "epoch": 0.9873032904148784, + "grad_norm": 0.5410454910996103, + "learning_rate": 8.454460132446552e-08, + "loss": 1.6345, + "step": 11042 + }, + { + "epoch": 0.987392703862661, + "grad_norm": 0.5777118354299748, + "learning_rate": 8.335819001952239e-08, + "loss": 1.7928, + "step": 11043 + }, + { + "epoch": 0.9874821173104434, + "grad_norm": 0.3542461621491375, + "learning_rate": 8.21801586026627e-08, + "loss": 1.3254, + "step": 11044 + }, + { + "epoch": 0.987571530758226, + "grad_norm": 0.36737776246169346, + "learning_rate": 8.101050717267411e-08, + "loss": 1.8082, + "step": 11045 + }, + { + "epoch": 0.9876609442060086, + "grad_norm": 0.5052701716266444, + "learning_rate": 7.984923582767812e-08, + "loss": 1.5269, + "step": 11046 + }, + { + "epoch": 0.9877503576537912, + "grad_norm": 0.4201563561282056, + "learning_rate": 7.869634466504128e-08, + "loss": 1.3414, + "step": 11047 + }, + { + "epoch": 0.9878397711015737, + "grad_norm": 0.4961536489141638, + "learning_rate": 7.755183378147512e-08, + "loss": 1.6938, + "step": 11048 + }, + { + "epoch": 0.9879291845493562, + "grad_norm": 0.4359839798105903, + "learning_rate": 7.641570327295844e-08, + "loss": 1.4933, + "step": 11049 + }, + { + "epoch": 0.9880185979971388, + "grad_norm": 0.7883302275098387, + "learning_rate": 7.528795323477055e-08, + "loss": 1.6614, + "step": 11050 + }, + { + "epoch": 0.9881080114449213, + "grad_norm": 0.33770236921901026, + "learning_rate": 7.416858376151359e-08, + "loss": 1.4854, + "step": 11051 + }, + { + "epoch": 0.9881974248927039, + "grad_norm": 0.6806872223716662, + "learning_rate": 7.305759494705689e-08, + "loss": 1.6449, + "step": 11052 + }, + { + "epoch": 0.9882868383404864, + "grad_norm": 0.4140275684722014, + "learning_rate": 7.195498688458147e-08, + "loss": 1.7579, + "step": 11053 + }, + { + "epoch": 0.9883762517882689, + "grad_norm": 0.33452996613801095, + "learning_rate": 7.08607596665467e-08, + "loss": 1.3447, + "step": 11054 + }, + { + "epoch": 0.9884656652360515, + "grad_norm": 0.4757567962055668, + "learning_rate": 6.977491338474585e-08, + "loss": 1.7663, + "step": 11055 + }, + { + "epoch": 0.9885550786838341, + "grad_norm": 0.2954074822982603, + "learning_rate": 6.869744813023937e-08, + "loss": 1.2202, + "step": 11056 + }, + { + "epoch": 0.9886444921316166, + "grad_norm": 0.545077875185113, + "learning_rate": 6.762836399338834e-08, + "loss": 1.8326, + "step": 11057 + }, + { + "epoch": 0.9887339055793991, + "grad_norm": 0.36747977284720024, + "learning_rate": 6.656766106385436e-08, + "loss": 1.3385, + "step": 11058 + }, + { + "epoch": 0.9888233190271817, + "grad_norm": 0.382678094426197, + "learning_rate": 6.551533943061072e-08, + "loss": 1.5502, + "step": 11059 + }, + { + "epoch": 0.9889127324749643, + "grad_norm": 0.49862199192134665, + "learning_rate": 6.447139918189793e-08, + "loss": 1.6867, + "step": 11060 + }, + { + "epoch": 0.9890021459227468, + "grad_norm": 0.8332654023239959, + "learning_rate": 6.343584040527927e-08, + "loss": 1.4881, + "step": 11061 + }, + { + "epoch": 0.9890915593705293, + "grad_norm": 0.4676370716577371, + "learning_rate": 6.240866318760752e-08, + "loss": 1.4545, + "step": 11062 + }, + { + "epoch": 0.9891809728183119, + "grad_norm": 0.574210768129249, + "learning_rate": 6.138986761502486e-08, + "loss": 1.58, + "step": 11063 + }, + { + "epoch": 0.9892703862660944, + "grad_norm": 0.5600041295520103, + "learning_rate": 6.037945377297405e-08, + "loss": 1.8807, + "step": 11064 + }, + { + "epoch": 0.989359799713877, + "grad_norm": 0.34636246604821763, + "learning_rate": 5.9377421746209525e-08, + "loss": 1.3496, + "step": 11065 + }, + { + "epoch": 0.9894492131616596, + "grad_norm": 0.5381450268644941, + "learning_rate": 5.838377161875297e-08, + "loss": 1.5389, + "step": 11066 + }, + { + "epoch": 0.989538626609442, + "grad_norm": 0.4418779719325782, + "learning_rate": 5.739850347395992e-08, + "loss": 1.7245, + "step": 11067 + }, + { + "epoch": 0.9896280400572246, + "grad_norm": 0.41512938914420694, + "learning_rate": 5.642161739445317e-08, + "loss": 1.7585, + "step": 11068 + }, + { + "epoch": 0.9897174535050072, + "grad_norm": 0.49375276440589666, + "learning_rate": 5.545311346215609e-08, + "loss": 1.4259, + "step": 11069 + }, + { + "epoch": 0.9898068669527897, + "grad_norm": 0.5438782556232535, + "learning_rate": 5.449299175831479e-08, + "loss": 1.7318, + "step": 11070 + }, + { + "epoch": 0.9898962804005722, + "grad_norm": 0.3939876074812595, + "learning_rate": 5.354125236343155e-08, + "loss": 1.374, + "step": 11071 + }, + { + "epoch": 0.9899856938483548, + "grad_norm": 0.3947484131680159, + "learning_rate": 5.25978953573536e-08, + "loss": 1.4801, + "step": 11072 + }, + { + "epoch": 0.9900751072961373, + "grad_norm": 0.5012688439162934, + "learning_rate": 5.166292081917323e-08, + "loss": 2.0495, + "step": 11073 + }, + { + "epoch": 0.9901645207439199, + "grad_norm": 0.6067516203988621, + "learning_rate": 5.0736328827316605e-08, + "loss": 1.5814, + "step": 11074 + }, + { + "epoch": 0.9902539341917024, + "grad_norm": 0.38144145387590295, + "learning_rate": 4.9818119459499325e-08, + "loss": 1.4866, + "step": 11075 + }, + { + "epoch": 0.990343347639485, + "grad_norm": 0.4555480608511263, + "learning_rate": 4.890829279272646e-08, + "loss": 1.5603, + "step": 11076 + }, + { + "epoch": 0.9904327610872675, + "grad_norm": 1.9443277922999356, + "learning_rate": 4.800684890330365e-08, + "loss": 1.324, + "step": 11077 + }, + { + "epoch": 0.9905221745350501, + "grad_norm": 0.6796094454019825, + "learning_rate": 4.711378786683707e-08, + "loss": 1.5891, + "step": 11078 + }, + { + "epoch": 0.9906115879828327, + "grad_norm": 0.7299601079064056, + "learning_rate": 4.6229109758222365e-08, + "loss": 1.7826, + "step": 11079 + }, + { + "epoch": 0.9907010014306151, + "grad_norm": 0.42796852677028413, + "learning_rate": 4.535281465165575e-08, + "loss": 1.9841, + "step": 11080 + }, + { + "epoch": 0.9907904148783977, + "grad_norm": 0.529610099184995, + "learning_rate": 4.448490262064509e-08, + "loss": 1.4975, + "step": 11081 + }, + { + "epoch": 0.9908798283261803, + "grad_norm": 0.42866097809785864, + "learning_rate": 4.362537373795439e-08, + "loss": 1.3293, + "step": 11082 + }, + { + "epoch": 0.9909692417739628, + "grad_norm": 0.61672932251297, + "learning_rate": 4.277422807570375e-08, + "loss": 1.6167, + "step": 11083 + }, + { + "epoch": 0.9910586552217453, + "grad_norm": 0.6040211616464407, + "learning_rate": 4.1931465705247195e-08, + "loss": 1.6801, + "step": 11084 + }, + { + "epoch": 0.9911480686695279, + "grad_norm": 0.4408969804031976, + "learning_rate": 4.109708669728374e-08, + "loss": 1.9228, + "step": 11085 + }, + { + "epoch": 0.9912374821173104, + "grad_norm": 0.6302292764733076, + "learning_rate": 4.027109112179073e-08, + "loss": 1.5198, + "step": 11086 + }, + { + "epoch": 0.991326895565093, + "grad_norm": 0.571811781341457, + "learning_rate": 3.945347904803498e-08, + "loss": 1.887, + "step": 11087 + }, + { + "epoch": 0.9914163090128756, + "grad_norm": 0.6487735017532034, + "learning_rate": 3.8644250544594975e-08, + "loss": 1.8002, + "step": 11088 + }, + { + "epoch": 0.991505722460658, + "grad_norm": 0.4749332865464607, + "learning_rate": 3.784340567934974e-08, + "loss": 1.832, + "step": 11089 + }, + { + "epoch": 0.9915951359084406, + "grad_norm": 0.40259796391323865, + "learning_rate": 3.7050944519445576e-08, + "loss": 1.7314, + "step": 11090 + }, + { + "epoch": 0.9916845493562232, + "grad_norm": 0.4486901626836514, + "learning_rate": 3.626686713135152e-08, + "loss": 1.5466, + "step": 11091 + }, + { + "epoch": 0.9917739628040058, + "grad_norm": 0.582935525106114, + "learning_rate": 3.54911735808372e-08, + "loss": 1.5918, + "step": 11092 + }, + { + "epoch": 0.9918633762517882, + "grad_norm": 0.4212415195468258, + "learning_rate": 3.472386393293947e-08, + "loss": 1.6645, + "step": 11093 + }, + { + "epoch": 0.9919527896995708, + "grad_norm": 0.35024210171409814, + "learning_rate": 3.3964938252040166e-08, + "loss": 1.4647, + "step": 11094 + }, + { + "epoch": 0.9920422031473534, + "grad_norm": 0.7260287302145844, + "learning_rate": 3.3214396601766176e-08, + "loss": 1.6692, + "step": 11095 + }, + { + "epoch": 0.9921316165951359, + "grad_norm": 0.5207550511085793, + "learning_rate": 3.247223904506713e-08, + "loss": 1.7638, + "step": 11096 + }, + { + "epoch": 0.9922210300429185, + "grad_norm": 0.35242752150835177, + "learning_rate": 3.173846564419325e-08, + "loss": 1.4107, + "step": 11097 + }, + { + "epoch": 0.992310443490701, + "grad_norm": 0.4683771184209668, + "learning_rate": 3.1013076460684186e-08, + "loss": 1.754, + "step": 11098 + }, + { + "epoch": 0.9923998569384835, + "grad_norm": 0.4627062432264586, + "learning_rate": 3.0296071555369065e-08, + "loss": 1.7672, + "step": 11099 + }, + { + "epoch": 0.9924892703862661, + "grad_norm": 0.366205373757405, + "learning_rate": 2.9587450988399768e-08, + "loss": 1.2753, + "step": 11100 + }, + { + "epoch": 0.9925786838340487, + "grad_norm": 0.5267376092481757, + "learning_rate": 2.888721481919543e-08, + "loss": 1.427, + "step": 11101 + }, + { + "epoch": 0.9926680972818311, + "grad_norm": 0.6066584739958581, + "learning_rate": 2.819536310648685e-08, + "loss": 1.554, + "step": 11102 + }, + { + "epoch": 0.9927575107296137, + "grad_norm": 0.47192766893834737, + "learning_rate": 2.7511895908294282e-08, + "loss": 1.4025, + "step": 11103 + }, + { + "epoch": 0.9928469241773963, + "grad_norm": 0.4188146202269345, + "learning_rate": 2.6836813281938543e-08, + "loss": 1.4449, + "step": 11104 + }, + { + "epoch": 0.9929363376251789, + "grad_norm": 0.4276123282527594, + "learning_rate": 2.617011528405211e-08, + "loss": 1.4792, + "step": 11105 + }, + { + "epoch": 0.9930257510729614, + "grad_norm": 0.6705033737737529, + "learning_rate": 2.551180197053471e-08, + "loss": 1.4818, + "step": 11106 + }, + { + "epoch": 0.9931151645207439, + "grad_norm": 0.4178462943855645, + "learning_rate": 2.4861873396608838e-08, + "loss": 1.8116, + "step": 11107 + }, + { + "epoch": 0.9932045779685265, + "grad_norm": 0.3966814990593017, + "learning_rate": 2.422032961677534e-08, + "loss": 1.5664, + "step": 11108 + }, + { + "epoch": 0.993293991416309, + "grad_norm": 0.6032773945710472, + "learning_rate": 2.3587170684835623e-08, + "loss": 2.0942, + "step": 11109 + }, + { + "epoch": 0.9933834048640916, + "grad_norm": 0.47930880108529644, + "learning_rate": 2.2962396653913864e-08, + "loss": 1.9269, + "step": 11110 + }, + { + "epoch": 0.9934728183118741, + "grad_norm": 0.4879037957723491, + "learning_rate": 2.234600757637928e-08, + "loss": 1.7337, + "step": 11111 + }, + { + "epoch": 0.9935622317596566, + "grad_norm": 0.6824947492591514, + "learning_rate": 2.1738003503946057e-08, + "loss": 1.7719, + "step": 11112 + }, + { + "epoch": 0.9936516452074392, + "grad_norm": 0.4438698051014863, + "learning_rate": 2.1138384487606742e-08, + "loss": 1.573, + "step": 11113 + }, + { + "epoch": 0.9937410586552218, + "grad_norm": 0.4613224010859725, + "learning_rate": 2.054715057765444e-08, + "loss": 1.714, + "step": 11114 + }, + { + "epoch": 0.9938304721030042, + "grad_norm": 0.6937254263210441, + "learning_rate": 1.9964301823660604e-08, + "loss": 1.7866, + "step": 11115 + }, + { + "epoch": 0.9939198855507868, + "grad_norm": 0.5257345357516046, + "learning_rate": 1.9389838274508355e-08, + "loss": 1.5876, + "step": 11116 + }, + { + "epoch": 0.9940092989985694, + "grad_norm": 0.4287799153153913, + "learning_rate": 1.8823759978392474e-08, + "loss": 1.5926, + "step": 11117 + }, + { + "epoch": 0.994098712446352, + "grad_norm": 0.6750795685481804, + "learning_rate": 1.8266066982774997e-08, + "loss": 1.7605, + "step": 11118 + }, + { + "epoch": 0.9941881258941345, + "grad_norm": 0.31388283720274235, + "learning_rate": 1.7716759334440724e-08, + "loss": 1.0705, + "step": 11119 + }, + { + "epoch": 0.994277539341917, + "grad_norm": 0.6643128102403251, + "learning_rate": 1.7175837079452804e-08, + "loss": 1.8328, + "step": 11120 + }, + { + "epoch": 0.9943669527896996, + "grad_norm": 0.7666735230369605, + "learning_rate": 1.6643300263186056e-08, + "loss": 1.6788, + "step": 11121 + }, + { + "epoch": 0.9944563662374821, + "grad_norm": 0.44756158133343726, + "learning_rate": 1.6119148930282546e-08, + "loss": 1.599, + "step": 11122 + }, + { + "epoch": 0.9945457796852647, + "grad_norm": 0.6893973089575096, + "learning_rate": 1.560338312472931e-08, + "loss": 1.316, + "step": 11123 + }, + { + "epoch": 0.9946351931330472, + "grad_norm": 0.6470246542075337, + "learning_rate": 1.5096002889758433e-08, + "loss": 6.2134, + "step": 11124 + }, + { + "epoch": 0.9947246065808297, + "grad_norm": 0.5304435652299406, + "learning_rate": 1.4597008267935863e-08, + "loss": 1.4551, + "step": 11125 + }, + { + "epoch": 0.9948140200286123, + "grad_norm": 0.3494829819231586, + "learning_rate": 1.4106399301117012e-08, + "loss": 1.451, + "step": 11126 + }, + { + "epoch": 0.9949034334763949, + "grad_norm": 0.5585956183671338, + "learning_rate": 1.3624176030435642e-08, + "loss": 1.7261, + "step": 11127 + }, + { + "epoch": 0.9949928469241774, + "grad_norm": 0.4992992726212977, + "learning_rate": 1.315033849634828e-08, + "loss": 1.7059, + "step": 11128 + }, + { + "epoch": 0.9950822603719599, + "grad_norm": 0.5021184346383859, + "learning_rate": 1.2684886738589808e-08, + "loss": 1.4297, + "step": 11129 + }, + { + "epoch": 0.9951716738197425, + "grad_norm": 0.5912239011366804, + "learning_rate": 1.2227820796184564e-08, + "loss": 1.4619, + "step": 11130 + }, + { + "epoch": 0.995261087267525, + "grad_norm": 0.37249130467857505, + "learning_rate": 1.1779140707490755e-08, + "loss": 1.3881, + "step": 11131 + }, + { + "epoch": 0.9953505007153076, + "grad_norm": 0.43657929933841755, + "learning_rate": 1.1338846510111633e-08, + "loss": 2.0055, + "step": 11132 + }, + { + "epoch": 0.9954399141630901, + "grad_norm": 0.5238794277978589, + "learning_rate": 1.0906938240995423e-08, + "loss": 2.0715, + "step": 11133 + }, + { + "epoch": 0.9955293276108726, + "grad_norm": 0.6583883050483867, + "learning_rate": 1.04834159363576e-08, + "loss": 1.911, + "step": 11134 + }, + { + "epoch": 0.9956187410586552, + "grad_norm": 0.48802009686852366, + "learning_rate": 1.0068279631725297e-08, + "loss": 1.5427, + "step": 11135 + }, + { + "epoch": 0.9957081545064378, + "grad_norm": 0.4908963678952002, + "learning_rate": 9.661529361892907e-09, + "loss": 1.5949, + "step": 11136 + }, + { + "epoch": 0.9957975679542204, + "grad_norm": 0.4618138119120315, + "learning_rate": 9.263165160999787e-09, + "loss": 1.9276, + "step": 11137 + }, + { + "epoch": 0.9958869814020028, + "grad_norm": 0.567378151872757, + "learning_rate": 8.873187062452548e-09, + "loss": 1.61, + "step": 11138 + }, + { + "epoch": 0.9959763948497854, + "grad_norm": 0.6850880799381931, + "learning_rate": 8.491595098947258e-09, + "loss": 1.7931, + "step": 11139 + }, + { + "epoch": 0.996065808297568, + "grad_norm": 0.5767242461614954, + "learning_rate": 8.118389302491647e-09, + "loss": 1.9118, + "step": 11140 + }, + { + "epoch": 0.9961552217453505, + "grad_norm": 0.5150586078871549, + "learning_rate": 7.753569704382902e-09, + "loss": 1.4386, + "step": 11141 + }, + { + "epoch": 0.996244635193133, + "grad_norm": 0.3392106001258487, + "learning_rate": 7.397136335229871e-09, + "loss": 1.3329, + "step": 11142 + }, + { + "epoch": 0.9963340486409156, + "grad_norm": 0.6160677747243056, + "learning_rate": 7.049089224919758e-09, + "loss": 1.7213, + "step": 11143 + }, + { + "epoch": 0.9964234620886981, + "grad_norm": 0.4104010812359982, + "learning_rate": 6.709428402629225e-09, + "loss": 1.6294, + "step": 11144 + }, + { + "epoch": 0.9965128755364807, + "grad_norm": 0.42648606791053173, + "learning_rate": 6.378153896868799e-09, + "loss": 1.4817, + "step": 11145 + }, + { + "epoch": 0.9966022889842633, + "grad_norm": 0.4463706412677534, + "learning_rate": 6.055265735405158e-09, + "loss": 1.56, + "step": 11146 + }, + { + "epoch": 0.9966917024320457, + "grad_norm": 0.5011222545076123, + "learning_rate": 5.740763945327743e-09, + "loss": 1.7127, + "step": 11147 + }, + { + "epoch": 0.9967811158798283, + "grad_norm": 0.3580218747856647, + "learning_rate": 5.434648553015453e-09, + "loss": 1.4729, + "step": 11148 + }, + { + "epoch": 0.9968705293276109, + "grad_norm": 0.5147454869992326, + "learning_rate": 5.136919584125544e-09, + "loss": 1.7145, + "step": 11149 + }, + { + "epoch": 0.9969599427753935, + "grad_norm": 0.46512901215908475, + "learning_rate": 4.847577063649133e-09, + "loss": 1.4339, + "step": 11150 + }, + { + "epoch": 0.9970493562231759, + "grad_norm": 0.4443769705851475, + "learning_rate": 4.566621015833495e-09, + "loss": 1.5259, + "step": 11151 + }, + { + "epoch": 0.9971387696709585, + "grad_norm": 0.4885522148309994, + "learning_rate": 4.2940514642597626e-09, + "loss": 1.4391, + "step": 11152 + }, + { + "epoch": 0.9972281831187411, + "grad_norm": 0.6663350420306036, + "learning_rate": 4.029868431765227e-09, + "loss": 1.5376, + "step": 11153 + }, + { + "epoch": 0.9973175965665236, + "grad_norm": 0.6136855033767873, + "learning_rate": 3.774071940532142e-09, + "loss": 6.0808, + "step": 11154 + }, + { + "epoch": 0.9974070100143062, + "grad_norm": 0.6610859401677263, + "learning_rate": 3.526662012010018e-09, + "loss": 1.8119, + "step": 11155 + }, + { + "epoch": 0.9974964234620887, + "grad_norm": 0.477216954000684, + "learning_rate": 3.2876386669267177e-09, + "loss": 1.5314, + "step": 11156 + }, + { + "epoch": 0.9975858369098712, + "grad_norm": 0.3939648955346901, + "learning_rate": 3.057001925355074e-09, + "loss": 1.5772, + "step": 11157 + }, + { + "epoch": 0.9976752503576538, + "grad_norm": 0.5307438675996965, + "learning_rate": 2.8347518066129675e-09, + "loss": 1.3511, + "step": 11158 + }, + { + "epoch": 0.9977646638054364, + "grad_norm": 0.3805060221218771, + "learning_rate": 2.620888329363247e-09, + "loss": 1.413, + "step": 11159 + }, + { + "epoch": 0.9978540772532188, + "grad_norm": 0.3922843547692914, + "learning_rate": 2.4154115115360144e-09, + "loss": 1.6166, + "step": 11160 + }, + { + "epoch": 0.9979434907010014, + "grad_norm": 0.38220798546268675, + "learning_rate": 2.218321370361931e-09, + "loss": 1.5206, + "step": 11161 + }, + { + "epoch": 0.998032904148784, + "grad_norm": 0.5322411332848873, + "learning_rate": 2.0296179223722176e-09, + "loss": 1.8297, + "step": 11162 + }, + { + "epoch": 0.9981223175965666, + "grad_norm": 0.38755374989306435, + "learning_rate": 1.8493011833875529e-09, + "loss": 1.4333, + "step": 11163 + }, + { + "epoch": 0.998211731044349, + "grad_norm": 0.432291716107734, + "learning_rate": 1.6773711685291738e-09, + "loss": 1.7194, + "step": 11164 + }, + { + "epoch": 0.9983011444921316, + "grad_norm": 0.3741889791975254, + "learning_rate": 1.5138278922299797e-09, + "loss": 1.651, + "step": 11165 + }, + { + "epoch": 0.9983905579399142, + "grad_norm": 0.4892076678835746, + "learning_rate": 1.3586713681901232e-09, + "loss": 1.4401, + "step": 11166 + }, + { + "epoch": 0.9984799713876967, + "grad_norm": 0.6378014709947528, + "learning_rate": 1.211901609443622e-09, + "loss": 2.0832, + "step": 11167 + }, + { + "epoch": 0.9985693848354793, + "grad_norm": 0.39743473424515885, + "learning_rate": 1.073518628269543e-09, + "loss": 1.4136, + "step": 11168 + }, + { + "epoch": 0.9986587982832618, + "grad_norm": 0.64897037784731, + "learning_rate": 9.435224363030238e-10, + "loss": 1.5601, + "step": 11169 + }, + { + "epoch": 0.9987482117310443, + "grad_norm": 0.37189781843136566, + "learning_rate": 8.219130444353518e-10, + "loss": 1.4487, + "step": 11170 + }, + { + "epoch": 0.9988376251788269, + "grad_norm": 0.3829150556909626, + "learning_rate": 7.086904628694769e-10, + "loss": 1.5627, + "step": 11171 + }, + { + "epoch": 0.9989270386266095, + "grad_norm": 0.4546203612669538, + "learning_rate": 6.038547010867035e-10, + "loss": 1.5502, + "step": 11172 + }, + { + "epoch": 0.9990164520743919, + "grad_norm": 0.49116394588719064, + "learning_rate": 5.074057678911004e-10, + "loss": 1.7357, + "step": 11173 + }, + { + "epoch": 0.9991058655221745, + "grad_norm": 0.7010782853774514, + "learning_rate": 4.1934367137619334e-10, + "loss": 1.6786, + "step": 11174 + }, + { + "epoch": 0.9991952789699571, + "grad_norm": 0.3659710729169934, + "learning_rate": 3.396684189249655e-10, + "loss": 1.4808, + "step": 11175 + }, + { + "epoch": 0.9992846924177397, + "grad_norm": 0.4659025738108818, + "learning_rate": 2.683800172098572e-10, + "loss": 1.9209, + "step": 11176 + }, + { + "epoch": 0.9993741058655222, + "grad_norm": 0.3641305436191415, + "learning_rate": 2.054784722149705e-10, + "loss": 1.0408, + "step": 11177 + }, + { + "epoch": 0.9994635193133047, + "grad_norm": 0.41361301562504876, + "learning_rate": 1.5096378922496712e-10, + "loss": 1.5199, + "step": 11178 + }, + { + "epoch": 0.9995529327610873, + "grad_norm": 0.6473165489122282, + "learning_rate": 1.0483597280286361e-10, + "loss": 1.6982, + "step": 11179 + }, + { + "epoch": 0.9996423462088698, + "grad_norm": 1.063554408092281, + "learning_rate": 6.709502681223611e-11, + "loss": 6.1324, + "step": 11180 + }, + { + "epoch": 0.9997317596566524, + "grad_norm": 0.44089849672467596, + "learning_rate": 3.774095442832248e-11, + "loss": 1.3847, + "step": 11181 + }, + { + "epoch": 0.9998211731044349, + "grad_norm": 0.41706640159831915, + "learning_rate": 1.6773758104715597e-11, + "loss": 1.521, + "step": 11182 + }, + { + "epoch": 0.9999105865522174, + "grad_norm": 0.8159058759925939, + "learning_rate": 4.193439617772299e-12, + "loss": 2.7196, + "step": 11183 + }, + { + "epoch": 1.0, + "grad_norm": 0.6699957024689727, + "learning_rate": 0.0, + "loss": 4.7685, + "step": 11184 + }, + { + "epoch": 1.0, + "step": 11184, + "total_flos": 4556966994247680.0, + "train_loss": 0.05963147350604271, + "train_runtime": 5028.2144, + "train_samples_per_second": 284.698, + "train_steps_per_second": 2.224 + } + ], + "logging_steps": 1.0, + "max_steps": 11184, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 300, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 4556966994247680.0, + "train_batch_size": 16, + "trial_name": null, + "trial_params": null +}